B349 ==== Subject: Re: Copying ROM Cards > The description at hpcalc.org says that your program copies from a high > port to port 1. But David wants to copy from port 1 to a high port. > As far as I know, only COPY by Rick Grevelle fits this . > Unfortunately, COPY is not at hpcalc.org yet, except inside Goodies Disk > #10. It's also in the EduCALC Hacker's ROM, if I remember correctly. > It's very small, so if anybody wants it, just drop me email. Werner, if your program also copies from port 1 to a high port, please > let us know *and* update the description at hpcalc.org. Thanx! > There are two 'copy port' entries (by me) in the Memory Management section of hpcalc.org: - Copy Port : copies port n to port 1 - ROM Card Copier : copies port 1 to port n The descriptions are correct in both cases, seems you took the first entry you found, Joe, i.o 'ROM Card Copier' as I had specified. BTW both are quite a lot faster than Rick Grevelle's, and both suffer from the (not so safe on a GX) 'DEEPSLEEP' at the end. So, be sure to switch the hp back ON right after running either of these programs. I once found a french utility doing the same thing, with graphics and all.. But I can't find it anymore, it was much more userfriendly than mine. -- Werner Huysegoms remove the x before replying http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: EXIT for UserRPL programs A very simple scheme to allow an exit anywhere in a user program, and then continue with any higher-level program: Store the following programs ( EXIT and CALL ) in HOME: << 7 DOERR >> 'EXIT' STO << IFERR EVAL THEN ERRN B->R 7 =/ { ERRN DOERR } IFT END >> 'CALL' STO @ NotEqual symbol Sample UserRPL programs, using the above functions: @ This program will be called by other programs << 123 EXIT 888 >> 'Prog1' STO @ If you press its key, it will exit anyway, @ but with the message Error: Warning @ This program will be invoked by pressing its menu key << 'Prog1' CALL @ Call above program, permitting EXIT << 456 EXIT 999 >> CALL @ Permit EXIT from part of this program 789 @ Execution of main program continues >> 'Prog2' STO @ Final output of Prog2 is 123 456 789 ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: Re: EXIT for UserRPL programs (#3) [canceled previous, reposting slight improvement] Very simple scheme to allow an exit anywhere in a user program, and then continue with any higher-level program: Store the following programs ( EXIT and CALL ) in HOME: << Exit DOERR >> 'EXIT' STO << IFERR EVAL THEN ERRM Exit =/ { ERRN DOERR } IFT END >> 'CALL' STO @ NotEqual symbol Sample UserRPL programs, using the above functions: @ This program will exit and display Exit << 123 EXIT 888 >> 'Prog1' STO @ This program continues after each EXIT, displaying no message: << 'Prog1' CALL @ Call above program, continuing after its EXIT << 456 EXIT 999 >> CALL @ Continue below, after EXIT 789 >> 'Prog2' STO @ Final output of Prog2 is 123 456 789 ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: EXIT for UserRPL (#4) [sorry for canceling/reposting again, want to make it *clear* !!!] This is the simplest scheme I could come up with for allowing a UserRPL program to exit at any point it wants to, while also allowing any higher-level programs to continue: Store the following programs ( EXIT and RUN ) in HOME: << Exit DOERR >> 'EXIT' STO << IFERR EVAL THEN ERRM Exit =/ { ERRN DOERR } IFT END >> 'RUN' STO Sample programs: @ This program will exit and display Exit @ (if called by another program, it will *not* return): << 123 EXIT 888 >> 'Prg1' STO @ Similarly, but no message will appear, and this program @ *can* be called from other programs, *without* aborting: << << 123 EXIT 888 >> RUN >> 'Prg2' STO @ This program also continues running (no messages): << 'Prg1' RUN << 456 EXIT 999 >> RUN 789 >> 'Prg3' STO @ Final output of Prg3 is 123 456 789 Note that the RUN command suppresses only the specific message used by the EXIT command; all other error messages act normally, canceling the running program unless trapped by your own IFERR ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: Re: -1^(1/3) does not = -1 You are experiencing the same confusion and irritation I felt when I first > got my 48 and was using the built in UNITS. I was horrified to find lb. as a > unit of MASS, and to make matters worse, under FORCE they have lbf wich I > thought was pound-foot until I looked in the manual and found out it is > pond-force. Now if you are going to do something soo stupid, to make it a > little more user freindly, they at least could have used lbm as pound-mass > and left lb as it should be ... as a FORCE. Just venting my frustration at HP's blatant insult on my inteligence. William. > wsmeyer@3dmail.com http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own It's interesting the effort that College Physics professors go through to convince us that a pound is a unit of force, while most of us had previously thought it was a unit of mass. They call the Pound a unit of force, because it makes their discussions flow more easily. But that is their construction, and not a historical or universally fundamental truth. The pound was established as a unit of measure long before Newton, et. al., developed the Calculus. Until Newton's day, the weight of a thing and the mass of a thing were different names for the same concept, and that concept was addressed by units like Pound and Troy Ounce. When the Bank of England began publishing One Pound Notes, it was well-understood that the note could be redemmed for a certain quantity of Sterling Silver. It was a quantity of matter that the note promised, not some gravitational force vector. Otherwise, the bank note would also have to state the gravity or specific location at which the note would be redemmed. The physicists had to have some unit for mass in the British system, so they concocted the unit of a Slug. To the best of my knowledge, the Slug has never been used as a unit of measure in any practical setting, except perhaps in some field artillery handbooks. I have never seen a scale calibrated in slugs. Stones, yes. Slugs, no. Here's a physics question for you, and perhaps a significant economic question for the British: How much would a pound of Sterling Silver weigh on the surface of the moon? Can you calculate it? Can you even approximate it? The surface gravity there is roughly one-sixth of the surface gravity here on earth. The HP48 places the unit Pound in the Mass area, and the unit Pound-Force in the Forces area. I don't know which of the HP Engineers did it, but I applaud the choice. Dave. ----- I don't speak for HP when I post here. And if you think a pound of Sterling Silver weighs ANYTHING other than a pound on the surface of the moon, then you must agree with me about what the word Pound really signifies in the English language. ---------------------------------------------------------------------- ==== Subject: Re: -1^(1/3) does not = -1 > You are experiencing the same confusion and irritation I felt when I first > got my 48 and was using the built in UNITS. I was horrified to find lb. as a > unit of MASS, and to make matters worse, under FORCE they have lbf wich I > thought was pound-foot until I looked in the manual and found out it is > pond-force. It's just your ignorance showing through, not Hewlett-Packard's. > Now if you are going to do something soo stupid, to make it a > little more user freindly, they at least could have used lbm as pound-mass > and left lb as it should be ... as a FORCE. I'm glad to see that HP follows the rules of NIST (National Institute of Standards and Technology, the United States national standards laboratory, and the ones who should be establishing exactly the kind of standard we are talking about here) which says that pounds are mass units with the symbol lb (and no period at the end as in your first use of it, and no s at the end in the plural), and pounds force have the symbol lbf. See, e.g., NIST Special Publication 811 (1995), p. 54. The mass definitions are and always have been the primary definitions of pounds. In addition, pounds have always been the most commonly used units for mass when it is measured in English units. Pounds force are secondary units, whose definition comes from the definition of the pound and the definition of a fixed number as the standard acceleration of gravity, a concept whose primary use is in defining the relationships between these units. One of the things that leads to misunderstanding of these concepts is a failure to realize that weight is an ambiguous word which refers to mass more often than it refers to one particular kind of force in the jargon of archery or a different kind of force in the jargon of physics. When the standards agencies of the United States, Canada, the United Kingdom, Australia, New Zed, and South Africa got together in 1959 to agree on a common definition, they defined the pound (avoirdupois) as a unit of mass exactly equal to 0.453 592 37 kg. In the U.S. this was implemented by the Federal Register notice of 1 July 1959, which makes no mention whatsoever of pounds force (which are not defined in any other place in United States law either). Don't you think these experts in the field knew what they were doing? All the metrologists involved in that decision knew full well the difference between mass and force. They defined the pound as a unit of mass, as all the various pounds throughout history have been. Before that, pounds had already been units of mass defined as a slightly different fraction of the kilogram for 66 years in the United States, and before that they were the mass of independently maintained standards. The official primary standard in the U.S. was the Troy Pound of the Mint from 1828 until 1893, and an interesting thing about the troy weights is that they don't suffer from the same mass/force ambiguity that avoirdupois weights do--troy weight always refers to mass, never to force. Just venting my frustration at HP's blatant insult on my inteligence. William. > wsmeyer@3dmail.com You ought to study up a little more before you go showing us your inteligence [sic]. Gene Nygaard http://ourworld.compuserve.com/homepages/Gene_Nygaard/weight.htm http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: apndvarlst in ML I was wondering if someone could help me with the apndvarlst. I want to write it in ML. ---------------------------------------------------------------------- ==== Subject: Re: apndvarlst in ML have you seen my posting ? Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: apndvarlst in ML I'll look at >TCOMP soon, but it seems there isn't much left for a speedup:-( Only some primitives which could be combined, and at the end, COMPN, which is a primitive code object. But let's see;-) Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: apndvarlst in ML I have compiled the program and it works great!! I will have to replace the >TCOMP command next. Any suggestions or ideas? are silent : ) have you seen my posting ? > Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: ML version of apndvarlst ! Who can beat this ? > ... > some code > ... > *ASSEMBLE > *EqualPosComp CODE > * CPU A B C D D0 D1 > A=DAT1 A ->ob ...showing the register's contents is really great ! David -- Belgium WWW : http://poseidon.ulb.ac.be/groupe/dh/ ---------------------------------------------------------------------- ==== Subject: Re: Please help... 3-4 minutes only... Sorry about the TOTALLY unrelated post, but for my > Honors Lab Biology class, we have to survey a diverse > group of people (a large amount,too) on the ethics of Well, this is not the most well-read group by any means. If you want real numbers, post to alt.atheism. ------------::o alt.tv.star-trek.voyager & alt.zen seem very busy as well... ----------- :: o .---..-..-..-..-..-..-. . .-. .-. .-..-..-.. .-. `| |'| || .` | > / / / / / | .` || | ) / `-' `-'`-'`-' `-' `. ^ .' `--^--'`-'`-'`-'' `--^--' Join The BabyNous Cult : The Friendly Neighborhood Cult ---------------------------------------------------------------------- ==== Subject: Re: How do I...? How do I write (type) something like theorems into my HP 48G and save it to > be recalled later? ~Shay WoW, it's a tricky one! ( :) ) Just put it into a string and save it! If there are graphs you can put'em into GROBs If you want something to write and browse easely and quickly, go http://www.hpcalc.org and i think you will find easely something that correspond to your needs.. A lot of libraries for exams cheaters exist, just search! -- login:burlett Contact: Jean-Yves.Burlett@scinfo.u-nancy.fr / jburlett@lemel.fr ---------------------------------------------------------------------- ==== Subject: Re: wrong eigenvalues These are the correct eigenvals: 2*sqrt(-25+5*I*sqrt(15)), -2*sqrt(-25+5*I*sqrt(15)), 2*sqrt(-25-5*I*sqrt(15)), -2*sqrt(-25-5*I*sqrt(15)) do you get the same? What is your basis for stating that the eigenvalues should be pure > imaginary? > Calculation of eigenvalues on my GX-R(and egn.vetors)for the following > matrix results in totally wrong answer: > [0 0 1 0] > [0 0 0 1] > [-80 -80 0 0] > [80 -120 0 0] > egnvalues should be pure imaginary, thought their arguments aren't even > close to -pi. Anyone care to explain? ---------------------------------------------------------------------- ==== Subject: Re: wrong eigenvalues I'm sorry for the confusion. It was my fault, I had a wrong matrix. It should have been: > [0 0 1 0] > [0 0 0 1] > [ 80 -80 0 0] > [80 -120 0 0] so one minus-sign less. This way everything is OK. Sorry. M.Krzewicki ---------------------------------------------------------------------- ==== Subject: On the topic of TI-89 calculators, my opinions and experiences I have put descriptions of my calculator, computer and educational background at the bottom of this message for those that desire that. Would someone post (or send me by electronic mail) the equivalent of this forum (comp.sys.hp48) for the TI-89. I purchased a TI-89 27 November 1998. I have spent most of the last two weekends and much of my evenings attempting to learn how to use it. It already is exhibiting the failure mode that has destroyed other TI calculators that I have owned, key bounce errors; that is, the keys sometimes generate either no response to a push or multiple responses to a single push. I had expected that; but, based on previous experience with TI calculators, I had expected to get at least 6 months use out of the machine, not 2 weeks. There are major problems with the owner's manual. The chart of the calculator inside the back cover of the manual does not have page numbers that take you to descriptions of the functions of the keys (such as the discussion which begins on page 13), e.g. I just today, while writing this message learned what the black-and-white key with an upward pointing arrow does. That finally explains how to generate upper-case alphabetic characters, e.g., ABC. The organization of the manual and its indexing make it very poor for use as a reference manual; one almost is required to read the manual line-by-line and work all the examples to built an appropriate index because much useful information is buried in the unindexed examples and margin hints. The first chapter of the owner's manual, ``Getting Started'' is way too short and the examples do not contain references to where in the manual the extended explanations are available. The complete description the functions of all keys should have been in that first chapter in the book; but that information has to be extracted by observation of the calculator's behavior while attempting the examples though the book. Appendix A should have been the second chapter in the book. All of the functions of the calculator should be in the index, both under the function name as programmed into the calculator and under the common English mathematics name of the function or procedure, if that name is not the function name as it is programmed into the calculator. For example, quite by accident, I learned how to invert a matrix using the ^-1 notation; there is no ``matrix invert'' or ``invert matrix'' index entry. Nonetheless, the calculator would not accept the mathematically equivalent 1/<> form, which should not be used as division is not defined for matrices. A necessary third chapter is simply not present: the ``command line'' entry format for the various data types. It took quite some time to learn that format for a 3 by 3 matrix is, e.g., [1,2,3;4,5,6;7,8,9] and not [[1,2][3,4]] format, which does work for a 2 by 2 matrix. The owner's manual spends a huge amount of its space on examples of how to graph---a function that I seldom use. {This reminds me of a problem I had with my calculus and differential equations instructors---they frequently took ``side trips'' into graphing instead of staying on the topic and were impossible to speed back onto the main topic.} The calculator keyboard has a color choice problem similar to that of the HP48GX as compared to the much better choices of the HP48SX, which I also own. Like the HP48GX, one must distinguish the modified key functions by their positions as the colors are so difficult to use. I do not mind that there are dedicated buttons for x, y, z and t. But, I find it disturbing that those letters are not also present in their normal English alphabetic order. The ESC(ape) key should be in the ``normal'' upper left hand corner of the keyboard; it is not. It always should abort the current operation and restore the machine to a ``reasonable'' previous state; in particular, it should make all dialog panels and error message panels go away. It does this some of the time; but, unfortunately, a reset or battery removal is necessary all too often. Until I found, by extensive menu browsing, the (MEM)ory F1(reset) ALL function I had considerable problems restoring the machine to a usable, clean condition after the machine entered a unusable condition. Before that I had to reset it frequently by removing its batteries, which is an operation that one should not have to do every few minutes of use. The NEWPROB cleanup function does not give one a clean HOME screen! One has to use the fore mentioned (MEM)ory F1(reset) ALL function to do that and that function does more than generally is desired. The custom menu should have an option of being an additional choice in the HOME menu. The APPS button should have been a choice in the HOME menu as well. The calculator uses the very poor notation: -1 -1 -1 -1 -1 -1 sin cos tan sinh cosh tanh instead of asin, acos, atan, asinh, acosh and atanh, which are unambiguous. The TI-89 notation can easily be misunderstood as alternates for cosec (csc), sec, cot, csch, sech and coth. The `editors' do not have a selection for ``save'' to the variable given on the entry panel nor to the expression being built on the HOME command line, only for ``save a copy''. This adds an unnecessary complexity to generating data. If there are methods for doing these very normal operations, then the manual is quite remiss in not obviously describing those methods. Because of the problems with the editors, many of the statistics, regressions and graphing operations are crippled. I must note that the TI-89 has more _built-in_ regression and graph types than the HP48 family has with more extensive control over those operations. With the TI-89 capacities that indicate to me that it is intended for use in an MBA class of use instead of an engineering class of use, I am quite surprised to see that it does not have a time-value-of- money application builtin and in particular one that would work closely with the spreadsheet capacities which it does have builtin to do, e.g., internal rate of return problems. The eigenvalue and eigenvector computations are separate with no explanation of how to associate the respective eigenvalue to their associated eigenvectors. The eigenvector computation presents its results as a matrix without the manual explaining whether the eigenvectors are the rows or columns of that matrix. The expression factoring procedure does not do a complete job of factoring a univariate expression unless that variate is specified as an additional argument to the procedure. There are separate procedures for real and complex factoring. I searched for a partial fraction factoring procedure in the TI-89. While using the expand() function, I discovered that partial fraction factoring was included within that procedure automatically. I feel that it should be a separate function. When one attempts to do the quite reasonable operation of extending the expression just evaluated to continue a computation, it replaces that expression with ``ans(1)'' which then requires one to clear the command line and reenter the expression again by hand or by using the cursor keys to high light the expression from the stack in either the evaluated or unevaluated form and press ENTER. It is not clear how to cause the calculator to just leave the expression alone. On the other hand, the TI-89 does permit the easy deletion of individual result groups (the TI-89 places both the input expression and the resultant expression on the result stack) from the result stack. I tried to use the machine to do a ``simple'' sales tax computation while shopping and received a very surprising result: 915687 42.99*1.065 ------ 20000 I switched to my HP48GX which I also had with me to do the computation as I could not spend the time then to determine what was causing the surprising result. I first learned that doing the fore mentioned memory full reset restored the machine to returning an EXACT decimal result. Since then, I have learned that right-hand modifier option of the ENTER key would have returned the desired result and the surprise was caused by the fact that the machine was in EXACT instead of AUTO mode somehow. The menu brought up by the MODE key should offer a function key for resetting the machine to the machine's default mode settings. I was unable to locate a programmatic method of storing and restoring a full set of MODE settings. that one use a PC and dedicated software instead of using a computer and generally available Kermit or X-modem protocols. that users are familiar with PC windows-oriented software. This is a poor assumption. I personally would not recommend that this machine be used in any educational environment. ------------------------------------------------------------ To explain my calculator background: I have owned and used HP calculators since the days of the HP41C and own and occasionally use a HP25. Among the smaller, less expensive calculators, I prefer the older, smaller Sharp 506 models. A HP48GX with Erable 3.024, ALG48 and EQSTK is my normal calculator environment. To explain my computer background, I have done systems level programming since the days of IBM 7044 computers running IBSYS, a batch operating system (1964). I do not own a PC of either IBM or Apple compatible type nor, until recently, use such at work. That machine rapidly had Linux installed to get a sane operating system, which I normally use with the X Window system not started. My normal computer environments are SGI Irix and AT&T 3B1 UNIX. To explain my educational background, I have a BA Mathematics from the University of Iowa, 1972 with a specialization in algebra and computer science. I completed most of a MBA from Loyola University in Chicago before I abandon that course of study due to philosophical disagreement with the program and the instructors. I recently completed a year of calculus and a quarter of differential equations, which I had not taken at the University of Iowa, using the HP48GX with Erable for support. Randolph J. Herber, herber@dcdrjh.fnal.gov, +1 630 840 2966, CD/CDFTF PK-149F, Mail Stop 318, Fermilab, Kirk & Pine Rds., PO Box 500, Batavia, IL 60510-0500, USA. (Speaking for myself and not for US, US DOE, FNAL nor URA.) (Product, trade, or service marks herein belong to their respective owners.) ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences I purchased a TI-89 27 November 1998. I have spent most of the last >two weekends and much of my evenings attempting to learn how to use >it. It already is exhibiting the failure mode that has destroyed other >TI calculators that I have owned, key bounce errors; that is, the keys >sometimes generate either no response to a push or multiple responses >to a single push. I had expected that; but, based on previous >experience with TI calculators, I had expected to get at least 6 months >use out of the machine, not 2 weeks. The only TI calculator I bought had an early keyboard failure. That was eons ago. HP has 'always' stressed mechanical durability. In fact, I heard that HP sales demos of the HP 35 included an intentional drop. Also, on a plant tour in Corvallis, I remember how they showed us their inlayed plastic keys compared to the cheaper painted style of TIs. ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences > HP has 'always' stressed mechanical durability. In fact, I heard that > HP sales demos of the HP 35 included an intentional drop. Easier with LED displays (all plastic) than with LCD (glass), but of course the HP35 was one of the extremely tough models, apparently made with surviving trips through a snowblower in mind :) It also cost US$400, at which price some current models might be able to afford a little bit heavier-duty design; today's HPs are apparently vastly better than TI, but not exactly like the good old days at HP, where rows of expensive gold-plated spring pins connected two circuit boards, rather than today's construction alternatives. > Also, on a plant tour in Corvallis, I remember how they showed us their > inlaid plastic keys, compared to the cheaper painted style of TIs. And they could have showed the gold-plated metal keyboards underneath, as well, which have also given way to alternatives But handheld calculators were once the top professional tools of the top engineers; now they are everyday commodities for a different audience. ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences >But handheld calculators were once the top professional tools of the >top engineers; now they are everyday commodities for a different audience. ... more like an indifferent audience. Stephen J Thomas SJThomasOD@aol.com ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences The only TI calculator I bought had an early keyboard failure. That was eons > ago. HP has 'always' stressed mechanical durability. In fact, I heard that HP > sales demos of the HP 35 included an intentional drop. Also, on a plant tour > in Corvallis, I remember how they showed us their inlayed plastic keys > compared to the cheaper painted style of TIs. John: I haven't had my TI-89 barely two months and already several keys either fail to register or double-register on pressing them! And I have I haven't even used it that much! I MUCH prefer the solid durability and feel of my HP over the cheap plastic that TI uses. I posted a complaint on TI's website and others joined in saying they had the same problem. This seems to be a real failing with TI hardware. - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences I personally would not recommend that this machine be used in any > educational environment. Randolph J. Herber Randolph: Your review of the TI-89 was excellent, and I heartily agree with your summation quoted above. I too, own the TI-89. Aside from the built-in Derive software, the TI-89 is a badly organized piece of cheap plastic. The only thing I use it for is checking indefinite integrals while doing homework away from my computer, where I have Mathematica installed. For all my other calculating chores, my trusty HP with ALG48, Erable, and EQSTK is my first weapon of choice! Hopefully the next HP calculator will have a computer algebra system that will be even better than Derive, and HP will win the calculator war that TI is winning now among college students. - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences I think HP might win that 'war' if they would just drop their ram card prices to justifiable levels. That would allow new users to 'upgrade' machines to levels that would allow the use of Erable, etc. $130 for 128K is rediculous! Brian >> I personally would not recommend that this machine be used in any >> educational environment. >> Randolph J. Herber >Randolph: Your review of the TI-89 was excellent, and I heartily agree with >your summation quoted above. I too, own the TI-89. Aside from the built-in >Derive software, the TI-89 is a badly organized piece of cheap plastic. The >only thing I use it for is checking indefinite integrals while doing homework >away from my computer, where I have Mathematica installed. For all my other >calculating chores, my trusty HP with ALG48, Erable, and EQSTK is my first >weapon of choice! Hopefully the next HP calculator will have a computer >algebra system that will be even better than Derive, and HP will win the >calculator war that TI is winning now among college students. - J.C. >Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: off topic TI-89, on topic HP On Thu, 10 Dec 1998 22:08:31 -0500, Brian Denley prices to justifiable levels. That would allow new users to 'upgrade' >machines to levels that would allow the use of Erable, etc. $130 for 128K >is rediculous! *my* words :-! I really can't think of a reason why HP doesn't lower the prices, so those home made cards would not be very interesting. Now HP sell's nearly 0 cards in germany. Some time I was so crazy (and would have paid much to much for a mem card) and tried to buy a original card, but the german distributor for HP calcs told me they are to expensive so they don't have them on stock and I should go out and buy one from (snip - some third party cards made in germany, costing a sixth of the original cards...)! I really don't understand why HP doesn't want to sell their cards. Even to find a 32SII for a reasonable price (not over DM 100) is really hard here.... Often the price for a 32SII is like the price for a 38G (which is sold cheap, but no_one_knows..., or more exactly there are very few places where you can buy HP calcs in germany) A friend in New Zed wanted to buy a cable (with the software kit) recently. And it was on stock!!! But he'd have to buy $149 for the set of the cable and the software... ?!?! A friend in the states will buy him a cable now.... So I guess HP has set themself a limit. They won't sell to many calcs and accessories. Maybe they want to be some sort of secret. I knew much longer from other graphic Casio's and Co. before I discovered HP at our university. (You have the chance twice in the year for some days to buy a calc there... :-) If HP would have a better distribution structure and a bit lower prices for the calcs, but much cheaper ram-card-prices their secret wouldn't be longer one and much more people would give HP a try and would see that RPN is much more intuitive then the so_called_algebraic use. Since HP is very successfull selling printers they want to have areas where they could expand, when everyone will have a printer. So till now the calc market is hold down. Just a few (many of them lurking around in this NG) get the chance to use their calcs nowadays... Sssssilenceee, please don't tell your neighbor of what the HP48 is capable of, he might tell Santa Claus that he wishes one, and you know: what Santa Claus hears won't be a secret any longer ;-) BTW, maybe some of us couldn't use the calc in exams if the professors would knew of the inner HP48?! Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: Re: off topic TI-89, on topic HP The price of mem cards manufactured by the OEM, Seiko-Epson, is not determined solely by HP, although there is probably some leeway to adjust. The fact that Seiko-Epson also holds the patents on such niceties as the protective shutter might also be slightly dampening to the prospects of switching suppliers to, say, one of the big three German home-made card makers. Big-name companies like HP might, however, prefer to source their own accessories from very well established similar companies, however, to assure everything from quality control to the certainty of product availability. There might even be some sort of contract still in effect, dating back to when HP decided to rely on the Epson typewriter cards as the storage medium for the HP48; certainly the reason that you can't find Epson cards in the same sizes as those carrying the HP logo is due to some contract, which, if expired, might have allowed Epson to sell them to places like EduCalc, which did sell authentic Epson cards of other sizes under their own brand, for a moderately lesser price. Besides this, does Epson even manufacture the cards any longer? These cards might be pretty obsolete, except for the HP48, and again, some contract might spell out whether Epson has any obligation to keep supplying them for HP; if there isn't enough ongoing demand, or profit, then why would Epson even want to bother to continue producing them? Very similar situations seem to prevail regarding other obsolete equipment. For example, I still have to help maintain a 12-year-old odd-brand computer system which uses big fat old-vintage 80MB disks, mounted in special cages with custom power supplies; there is almost no place which refurbishes and keeps any stock of these things, and if we find one, they cost at least $500, for which we could just as well get a modern small disk of 50 times the capacity, except that new disks are incompatible with the old system. You might as well ask someone to part with a mint condition HP35 for the price of a modern cheap calculator of equivalent processing capability as to ask for original Seiko-Epson cards to be sold for the price of less stringently manufactured cards, which may not have gone through the certifications, testings, and legal hassles which seem to be part of the process of putting out a commercial product, unless, for some reason, dumping these cards at low prices might be perceived as some way to sell many more calculators, which evidently has been done (in a package deal, with calculator + card for one lower price). So, for top-quality name-brand memory cards, the price seems to stay high (except for cards from the package deal, which somehow now get sold for low price, sans calculator); for home-made cards, which may well be pretty good, but still may or may not be shielded, and can't use the patented protective shutters, the price is less. You can buy directly from those German suppliers, but possibly HP can not, or wouldn't care to. But listen, I used to assemble electronic instruments with my for, say, about 1/2 of HP price, just send me the card layout and specs, and I'll make one for you; if HP wants maybe 10,000 of them, I'll consider doing some moonlighting, even setting up my own factory, how's that? ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: Why HP48 10 reasons for I think HP48 is best than TI calcs 1-HP48 has infrared transfer. 2-HP48 has more programs available for download in the www. 3-HP48 come in three ways: HP48G, HP48GX and HP48G Plus. 4-HP48 has his own newsgroup and mail lists. 5-HP48 has much more documents and tecnical informations than TI. 6-HP48 is known worldwide* 7-HP48 is USERRPL, SYSRPL and ML programable. 8-HP48 has emulators, protocols, text editors, Jazz and Metakernel available. 9-HP48 (HP48GX) is expandable till 4 MB. 10-If HP48 dont solve a problem with its own resourses, it is possible to make programs it to solve the problem. I dont have nothing against TI calcs, I just think HP48 is the best. This is my opinion only. CMarangon http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: Why HP48 Don't get me wrong. I love my HP.. but I won't stand for inacurracy. >10 reasons for I think HP48 is best than TI calcs 4-HP48 has his own newsgroup and mail lists. Ticalc.org has 10+ mail lists, and we DO have a newsgroup 5-HP48 has much more documents and tecnical informations than TI. Umm... When it comes to programming tutorials, TI has you beat.. the z80 is the worlds easiest processor to program, and there are a number of excellant manuals on how to program it, for free, for download on the web. >7-HP48 is USERRPL, SYSRPL and ML programable. Isn't it Easy SYSRPL, real SYSRPL and ML? wouldn't that be equal to TI-Basic and z80 assembler? >8-HP48 has emulators, protocols, text editors, Jazz and Metakernel available. TI calcs do have emulators The HP is still better, just a little harder to learn saturn asm than z80 asm, and that's the only thing I ever missed. ---------------------------------------------------------------------- ==== Subject: Re: off topic TI-89, on topic HP >The price of mem cards manufactured by the OEM, Seiko-Epson, >is not determined solely by HP, although there is probably >some leeway to adjust. A really good question, how much the price is dependent from Seiko-Epson...? >The fact that Seiko-Epson also holds the patents on such >niceties as the protective shutter might also be slightly >dampening to the prospects of switching suppliers to, >say, one of the big three German home-made card makers. Sure those cards are not a real alternative to HP. Maybe Epson holds the prices up. But although the original cards are better then the home-made cards that's not a real reason why an original 1MB card costs around 700 DM (street price), while you get a 1MB professional *home-made* card for around 100 - 140 DM... Some years ago the *home-made* cards where also much more expensive! An internal upgrade to 384K was offered for 159 DM. Today you get the same upgrade for about 40 DM to 50 DM. The card prices also dropped due to the falling chip prices. But here's the point: The prices for the original HP cards are stable for years! HP (or Epson) didn't adjust the prices. I guess Epson lowered the prices which HP has to pay. If they didn't lower the prices HP could manufacture their own cards... But I admit, that contracts with Epson might prevent that. Maybe John is right and Epson doesn't have a real interest to manufacture those *obsolete* cards. That could be the reason for the high (compared to other products) prices. >You might as well ask someone to part with a mint condition >HP35 for the price of a modern cheap calculator of equivalent >processing capability as to ask for original Seiko-Epson >cards to be sold for the price of less stringently >manufactured cards, which may not have gone through the >certifications, testings, I don't think that most of us are asking that. I just wonder why the ram prices were not adjusted to a suitable level in the last years. Some years ago home-madel cards where cheaper then the original cards. But nowadays the gap is immense! >So, for top-quality name-brand memory cards, the price >seems to stay high It seems so :-( My guess is that not technical details or contracts are responsible Singapure and to Australia now caused a *slow-down* to the whole calc division (including the marketing department!). The distribution for HP calcs is not really covering the *world*. I mean here that there are just a few places where you can buy HP calcs (at least in germany). In many other countries it even seems much harder to find HP calcs (especially if you can't/won't pay immense prices). In germany the prices went up a bit, after SESCO became the only official distributor for HP calcs. The AUR was sold for 40 DM *everywhere*. After SESCO become the distributor (where all the other german suppliers have to buy their stock) the AUR went up to 60 DM. The prices for the calc went up also... As I've mentioned it seems impossible to find the cable for the HP48 at a reasonable price in New Zed. This points to a marketing problem... I think the marketing is the real key to increase HP's sales of calcs. I have sometimes observed that the *loudest* catches the attention, not *best*... Sure quality is important to keep the attention of the people and to make them to customers, but first you have to tell_that_you_are_the_best. The price is also important, but you don't have to be the cheapest, if your quality is of a higher standard then the competitors. >But listen, I used to assemble electronic instruments with my >for, say, about 1/2 of HP price, just send me the >card layout and specs, and I'll make one for you; >if HP wants maybe 10,000 of them, I'll consider doing some >moonlighting, even setting up my own factory, how's that? Mmmhh, now Mike's 1 MB cards cost 135 DM. John's cards for a price of 350 DM won't sell very good I guess ;-) But fortunately the mem card prices are not important for me, as I was lucky to find two of the original *package* cards John mentioned... :-) Better go, my optics book is waiting :-) Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: Re: off topic TI-89, on topic HP >On Thu, 10 Dec 1998 22:08:31 -0500, Brian Denley >I think HP might win that 'war' if they would just drop their ram card >>prices to justifiable levels. That would allow new users to 'upgrade' >>machines to levels that would allow the use of Erable, etc. $130 for 128K >>is rediculous! >*my* words :-! I really can't think of a reason why HP doesn't lower the prices, so >those home made cards would not be very interesting. Now HP sell's >nearly 0 cards in germany. Some time I was so crazy (and would have >paid much to much for a mem card) and tried to buy a original card, large remainder snipped. I have used RPN calculators since my Cambridge Scientific in about 1974/75 and HP calculators continuously since 1977 with my first purchase of a HP67 (In Saudi Arabia/Al Khobar). They / or the genre of RPN and a stack, are superior (in my opinion) to ordinary calculators. (In fact ordinary/algebraic2 calculators irritate me). While I am prepared to pay a premium for the best calculator it frankly pisses me off to be exploited by my chosen calculator manufacturer for (frankly) incidental periferals. Here in the UK there have been incidences where high range French cosmetic items have been sourced by food supermarkets. First UK Translation: Scent - equivalent - perfume: roses orange blossom - smell nice, i.e. not sweat. OverPricedStandardItem: £20+ Food Biased Supermarket Equivalent 1/2 price. Statement from OverpricedI Item Main Distibuter. We feel that the people who buy our product are happy to pay this rediculous amout because it makes them feel good to smell thus, knowing that is has been costly to do so. Well HP when I can buy from Germany a 1-Meg plus a 128k Mem Card Plus a Serial interface cable plus relevant software for less than the price of a SINGLE 128k Mem Card from supplier in the UK then I shall smell as the supermarket. Love James. ---------------------------------------------------------------------- ==== Subject: Re: off topic TI-89, on topic HP Well HP when I can buy from Germany a 1-Meg plus a 128k Mem Card Plus > a Serial interface cable plus relevant software for less than the > price of a SINGLE 128k Mem Card from supplier in the UK then I shall > smell as the supermarket. James. approved HP's are about $125 U.S. and nonapproved are about $100 U.S. - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: off topic TI-89, on topic HP ...and both of those prices are absolutely outrageous. Ram chips are dirt cheap! Brian > Well HP when I can buy from Germany a 1-Meg plus a 128k Mem Card Plus >> a Serial interface cable plus relevant software for less than the >> price of a SINGLE 128k Mem Card from supplier in the UK then I shall >> smell as the supermarket. >> James. approved HP's are about $125 U.S. and nonapproved are about $100 U.S. > - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences Most of your problems are due to the fact that you have not read the manual. There is no doubt that the manual is **** but most of the stuff is there somewhere. Try a bit harder before you complain again. -- Bob Wheeler --- (Reply to: bwheeler@echip.com) ECHIP, Inc. I have put descriptions of my calculator, computer and educational > background at the bottom of this message for those that desire that. Would someone post (or send me by electronic mail) the equivalent of > this forum (comp.sys.hp48) for the TI-89. I purchased a TI-89 27 November 1998. I have spent most of the last > two weekends and much of my evenings attempting to learn how to use > it. It already is exhibiting the failure mode that has destroyed other > TI calculators that I have owned, key bounce errors; that is, the keys > sometimes generate either no response to a push or multiple responses > to a single push. I had expected that; but, based on previous > experience with TI calculators, I had expected to get at least 6 months > use out of the machine, not 2 weeks. There are major problems with the owner's manual. The chart of the > calculator inside the back cover of the manual does not have page > numbers that take you to descriptions of the functions of the keys > (such as the discussion which begins on page 13), e.g. I just today, > while writing this message learned what the black-and-white key with an > upward pointing arrow does. That finally explains how to generate > upper-case alphabetic characters, e.g., ABC. The organization of the > manual and its indexing make it very poor for use as a reference > manual; one almost is required to read the manual line-by-line and work > all the examples to built an appropriate index because much useful > information is buried in the unindexed examples and margin hints. The > first chapter of the owner's manual, ``Getting Started'' is way too > short and the examples do not contain references to where in the manual > the extended explanations are available. The complete description the > functions of all keys should have been in that first chapter in the > book; but that information has to be extracted by observation of the > calculator's behavior while attempting the examples though the book. > Appendix A should have been the second chapter in the book. All of the > functions of the calculator should be in the index, both under the > function name as programmed into the calculator and under the common > English mathematics name of the function or procedure, if that name is > not the function name as it is programmed into the calculator. For > example, quite by accident, I learned how to invert a matrix using the > ^-1 notation; there is no ``matrix invert'' or ``invert matrix'' index > entry. Nonetheless, the calculator would not accept the mathematically > equivalent 1/<> form, which should not be used as division is > not defined for matrices. A necessary third chapter is simply not > present: the ``command line'' entry format for the various data types. > It took quite some time to learn that format for a 3 by 3 matrix is, > e.g., [1,2,3;4,5,6;7,8,9] and not [[1,2][3,4]] format, which does work > for a 2 by 2 matrix. The owner's manual spends a huge amount of its > space on examples of how to graph---a function that I seldom use. > {This reminds me of a problem I had with my calculus and differential > equations instructors---they frequently took ``side trips'' into > graphing instead of staying on the topic and were impossible to speed > back onto the main topic.} The calculator keyboard has a color choice problem similar to that of > the HP48GX as compared to the much better choices of the HP48SX, which > I also own. Like the HP48GX, one must distinguish the modified key > functions by their positions as the colors are so difficult to use. I do not mind that there are dedicated buttons for x, y, z and t. But, > I find it disturbing that those letters are not also present in their > normal English alphabetic order. The ESC(ape) key should be in the ``normal'' upper left hand corner of > the keyboard; it is not. It always should abort the current operation > and restore the machine to a ``reasonable'' previous state; in > particular, it should make all dialog panels and error message panels > go away. It does this some of the time; but, unfortunately, a reset or > battery removal is necessary all too often. Until I found, by > extensive menu browsing, the (MEM)ory F1(reset) ALL function I had > considerable problems restoring the machine to a usable, clean > condition after the machine entered a unusable condition. Before that > I had to reset it frequently by removing its batteries, which is an > operation that one should not have to do every few minutes of use. The NEWPROB cleanup function does not give one a clean HOME screen! > One has to use the fore mentioned (MEM)ory F1(reset) ALL function to do > that and that function does more than generally is desired. The custom menu should have an option of being an additional choice in > the HOME menu. The APPS button should have been a choice in the HOME > menu as well. The calculator uses the very poor notation: -1 -1 -1 -1 -1 -1 sin cos > tan sinh cosh tanh instead of asin, acos, atan, asinh, acosh and atanh, which are > unambiguous. The TI-89 notation can easily be misunderstood as > alternates for cosec (csc), sec, cot, csch, sech and coth. The `editors' do not have a selection for ``save'' to the variable > given on the entry panel nor to the expression being built on the HOME > command line, only for ``save a copy''. This adds an unnecessary > complexity to generating data. If there are methods for doing these > very normal operations, then the manual is quite remiss in not > obviously describing those methods. Because of the problems with the editors, many of the statistics, > regressions and graphing operations are crippled. I must note that the > TI-89 has more _built-in_ regression and graph types than the HP48 > family has with more extensive control over those operations. With the TI-89 capacities that indicate to me that it is intended for > use in an MBA class of use instead of an engineering class of use, I am > quite surprised to see that it does not have a time-value-of- money > application builtin and in particular one that would work closely with > the spreadsheet capacities which it does have builtin to do, e.g., > internal rate of return problems. The eigenvalue and eigenvector computations are separate with no > explanation of how to associate the respective eigenvalue to their > associated eigenvectors. The eigenvector computation presents its > results as a matrix without the manual explaining whether the > eigenvectors are the rows or columns of that matrix. The expression factoring procedure does not do a complete job of > factoring a univariate expression unless that variate is specified as > an additional argument to the procedure. There are separate procedures > for real and complex factoring. I searched for a partial fraction factoring procedure in the TI-89. > While using the expand() function, I discovered that partial fraction > factoring was included within that procedure automatically. I feel > that it should be a separate function. When one attempts to do the quite reasonable operation of extending the > expression just evaluated to continue a computation, it replaces that > expression with ``ans(1)'' which then requires one to clear the command > line and reenter the expression again by hand or by using the cursor > keys to high light the expression from the stack in either the > evaluated or unevaluated form and press ENTER. It is not clear how to > cause the calculator to just leave the expression alone. On the other > hand, the TI-89 does permit the easy deletion of individual result > groups (the TI-89 places both the input expression and the resultant > expression on the result stack) from the result stack. I tried to use the machine to do a ``simple'' sales tax computation > while shopping and received a very surprising result: 915687 > 42.99*1.065 ------ > 20000 I switched to my HP48GX which I also had with me to do the computation > as I could not spend the time then to determine what was causing the > surprising result. I first learned that doing the fore mentioned memory > full reset restored the machine to returning an EXACT decimal result. > Since then, I have learned that right-hand modifier option of the ENTER > key would have returned the desired result and the surprise was caused > by the fact that the machine was in EXACT instead of AUTO mode somehow. The menu brought up by the MODE key should offer a function key for > resetting the machine to the machine's default mode settings. I was > unable to locate a programmatic method of storing and restoring a > full set of MODE settings. that one use a PC and dedicated software instead of using a computer and > generally available Kermit or X-modem protocols. that users are familiar with PC windows-oriented software. This is > a poor assumption. I personally would not recommend that this machine be used in any > educational environment. ------------------------------------------------------------ To explain my calculator background: I have owned and used HP > calculators since the days of the HP41C and own and occasionally use a > HP25. Among the smaller, less expensive calculators, I prefer the > older, smaller Sharp 506 models. A HP48GX with Erable 3.024, ALG48 and > EQSTK is my normal calculator environment. To explain my computer background, I have done systems level > programming since the days of IBM 7044 computers running IBSYS, a batch > operating system (1964). I do not own a PC of either IBM or Apple > compatible type nor, until recently, use such at work. That machine > rapidly had Linux installed to get a sane operating system, which I > normally use with the X Window system not started. My normal computer > environments are SGI Irix and AT&T 3B1 UNIX. To explain my educational background, I have a BA Mathematics from the > University of Iowa, 1972 with a specialization in algebra and computer > science. I completed most of a MBA from Loyola University in Chicago > before I abandon that course of study due to philosophical disagreement > with the program and the instructors. I recently completed a year of > calculus and a quarter of differential equations, which I had not taken > at the University of Iowa, using the HP48GX with Erable for support. Randolph J. Herber, herber@dcdrjh.fnal.gov, +1 630 840 2966, CD/CDFTF PK-149F, > Mail Stop 318, Fermilab, Kirk & Pine Rds., PO Box 500, Batavia, IL 60510-0500, > USA. (Speaking for myself and not for US, US DOE, FNAL nor URA.) (Product, > trade, or service marks herein belong to their respective owners.) ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences >Most of your problems are due to the fact that you >have not read the manual. There is no doubt that >the manual is **** but most of the stuff is there >somewhere. Try a bit harder before you complain >again. -- >Bob Wheeler --- (Reply to: bwheeler@echip.com) > ECHIP, Inc. > The useful content of your message is nil. Do you want to try again? I expect not. If you can not offer any useful information, then do not offer anything. I have already spent about 80 hours of effort (reading and doing examples). Much of that effort was spent reading the manual. As I mentioned, I had found solutions to some of the questions and problems. Sometimes, I learned answers by working problems whose solutions were already known and examined the answers that the machine gave. In this way, I learned some of the forms in which the calculator gave its answers. That information should have been in the manual, e.g., whether the eigenvectors within the solution matrix are row or column. I had not had similar problems with either HP calculators or their documentation. Randolph J. Herber, herber@dcdrjh.fnal.gov, +1 630 840 2966, CD/CDFTF PK-149F, Mail Stop 318, Fermilab, Kirk & Pine Rds., PO Box 500, Batavia, IL 60510-0500, USA. (Speaking for myself and not for US, US DOE, FNAL nor URA.) (Product, trade, or service marks herein belong to their respective owners.) ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences > Most of your problems are due to the fact that you > have not read the manual. There is no doubt that > the manual is **** but most of the stuff is there > somewhere. Try a bit harder before you complain > again. I agree with Bob - you gotta read the manual before writing the machine off as inferior. Intuitiveness only goes so far, and I think it's generally agreed that TI's have a reputation for being more intuitive than HP's. It strikes me that one who uses an RPN-based machine loaded up with third party software doesn't know better than to RTFM. ALG48, for example (referred to in the original post), is nearly impossible to use without spending some quality time with the documentation. The given example of being unable to convert an exact result into a real is classic: I've got the TI89 Guidebook open in front of me, and the conversion from exact results to real results is shown on page 4 of Chapter 1! This whole thread is practically ludicrous enough to make me wonder if it's a troll... -- Karl E. Yeanoplos kyeanopl@ast.lmco.com http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences >> Most of your problems are due to the fact that you >> have not read the manual. There is no doubt that >> the manual is **** but most of the stuff is there >> somewhere. Try a bit harder before you complain >> again. >I agree with Bob - you gotta read the manual before writing the machine off as >inferior. Intuitiveness only goes so far, and I think it's generally agreed >that TI's have a reputation for being more intuitive than HP's. >It strikes me that one who uses an RPN-based machine loaded up with third >party software doesn't know better than to RTFM. ALG48, for example >(referred to in the original post), is nearly impossible to use without >spending some quality time with the documentation. >The given example of being unable to convert an exact result into a real is >classic: I've got the TI89 Guidebook open in front of me, and the conversion >from exact results to real results is shown on page 4 of Chapter 1! >This whole thread is practically ludicrous enough to make me wonder if it's a >troll... >Karl E. Yeanoplos >kyeanopl@ast.lmco.com >http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own Then, you did not read my posting. RTP! I most certainly did read the manual and have worked many of the examples. I mentioned in my original posting that I did find that method of obtaining AUTO or approximate results (use the right-shift option of the ENTER key). My problem is that I still do not know how the machine came to be in that state (EXACT mode) at that time. If the calculator is in EXACT mode and all values used in the expression are expressed numerically, then it would have very reasonable for the TI-89 to have returned the exact decimal represent- ation. The problem of an absence of a general mode reset to default modes without clearing variables and programs is also a real problem. After more than 3 decades of operating system support and programming, I do understand RTFM. I can also recognize a poorly written manual. The problem mostly is the poorly written and poorly organized manual, which also is not a _reference_ manual. Most of the problems that I have had with the machine are or can be explained as, in fact, problems with the manual. Those problems with the manual are a sufficient reason, to me, for not recommending the calculator for use in an educational setting. I have owned other TI calculators in the past. Except for the keyboard problem I mentioned in my original posting, I did not have such problems with the calculator's manual. I feel that 2 weeks is more than enough trying! I strongly disagree with your thought that ``TI's have a reputation for being more intuitive than HP's.'' They are very much like Microsoft Windows and MacOS and very contraintuitive. They are very much like light switches to someone who has never seen one before and tries to experiences in the Republic of South Vietnam, I know that light switches are _not_ intuitive; but, they are easy to explain. The TIs' GUI-like interface needs to be explained and not assumed to be common knowledge. Common knowledge is the reason that you find light switches to be simple and interpret that simpleness as being intuitive. Simplicity and intuitiveness are quite different attributes. Randolph J. Herber, herber@dcdrjh.fnal.gov, +1 630 840 2966, CD/CDFTF PK-149F, Mail Stop 318, Fermilab, Kirk & Pine Rds., PO Box 500, Batavia, IL 60510-0500, USA. (Speaking for myself and not for US, US DOE, FNAL nor URA.) (Product, trade, or service marks herein belong to their respective owners.) ---------------------------------------------------------------------- ==== Subject: On the topic of TI-89 calculators, my opinions and experiences RJH> I strongly disagree with your thought that ``TI's have a RJH> reputation for being more intuitive than HP's.'' They are very RJH> much like Microsoft Windows and MacOS and very contraintuitive. RJH> They are very much like light switches to someone who has never RJH> seen one before and tries to apply oil lamp methodologies to a RJH> South Vietnam, I know that light switches are _not_ intuitive; RJH> but, they are easy to explain. Since the TI is for highschool or college students it's OK for TI to assume that the people that buy the calc know light switches. RJH> The TIs' GUI-like interface needs RJH> to be explained and not assumed to be common knowledge. If it's a standard GUI (I have no TI) I don't think they have to explain it hust as they don't have to put in reading course or show you that the buttons have to be pushed. RJH> Common knowledge is the reason that you find light switches to be RJH> simple and interpret that simpleness as being intuitive. RJH> Simplicity and intuitiveness are quite different attributes. No but you can assume that some standards are known by your target group. You are a very unique customer because you know computers but don't know GUIs very well. I guess the majority of todays computer users don't know what to do with command line tools (and yes I think that's bad). cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... Unzip? Explode? Expand? These guys are SICK! ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences I am requesting yet again a suggestion (pointer, if you will) to an appropriate, more relevant news group for the TI-89 calculator specifically, for TI high level calculators (which I do consider the TI-89 to be), or, at worse, for TI calculators in general. I have seen several respondents in this news group, comp.sys.hp48, and in private mail, misunderstand my objections to the TI-89. Yes, I do dislike GUIs in general. I see three proper uses for GUIs: graphic applications, e.g., drawing and CAD systems; situations in which the program input must be tightly controlled, e.g., database input programs and program for use by people for whom must be minimal, e.g., information kiosks in transportation nexi and fast food restaurant cash registers. One may argue properly that the TI-89 is intended for the third category of uses. If so, then that is unfortunate. Nonetheless, that is _not_ my major objection to the TI-89. My major objection is the poorly written manual that comes with the calculator. My minor objections are the poor keyboard and the fact that special software and computers are required to transfer data twixt the calculator and other machines. HP properly used Kermit, a registered trademark used by permission of Henson Associates, Inc., and RS-232c for such transfers. This facilitates using most any computer with a serial port for these transfers. I do consider PCs running Microsoft Windows programming support systems or Apple MacOS to be special types of computers. I happen not to have the use of one and I own or have the use of several types of computers running UNIX variants and CP/M. >RJH> I strongly disagree with your thought that ``TI's have a >RJH> reputation for being more intuitive than HP's.'' They are very >RJH> much like Microsoft Windows and MacOS and very contraintuitive. >RJH> They are very much like light switches to someone who has never >RJH> seen one before and tries to apply oil lamp methodologies to a >RJH> South Vietnam, I know that light switches are _not_ intuitive; >RJH> but, they are easy to explain. >Since the TI is for highschool or college students it's OK for TI to assume >that the people that buy the calc know light switches. Light switches were used as an example of an _equivalent_ situation. I was not suggesting that the _specific_ case of light switch functions and operation knowledge was needed to know how to program a TI-89. I realize that English may not be your native language even though you do appear to be quite fluent in the English language. But, you do _not_ appear to have understood that the English phrase ``very much like'' introduces a metaphor. In the Republic of South Vietnam, we had hired people to do janitor work. One young lady broke several light bulbs in an effort to blow them out (she had equated them to oil lamps with which she was familiar) and this technique was effective for her purposes---the light did go out. When we explained and demonstrated the light switch to her, she handled the explanation by considering that the switch somehow moved the wick remotely. This understanding was sufficient for both her and our purposes---she could turn the light on and off and we did not have to continue replacing the light bulbs. Fortunately, we did not have any ``three-way'' light switches. A paraphrase of Arthur C. Clarke, radio engineer, science fiction author, and the author of the artificial earth-orbiting communications satellite in 1945: ... sufficiently advanced technology is indistinguishable from magic. Quoted from Lost Worlds of 2001: A Space Odyssey, by Arthur C. Clarke, approximately page 185. I am suggesting that TI should include in the manual which comes with the calculator an explanation of the calculator's interface to the user. >RJH> The TIs' GUI-like interface needs >RJH> to be explained and not assumed to be common knowledge. >If it's a standard GUI (I have no TI) I don't think they have to explain it >hust as they don't have to put in reading course or show you that the buttons >have to be pushed. There several different styles of GUIs each with their separate styles of operation. The assumptions and patterns of operation of the GUI do need to be explained. If a _standard_ style of GUI is being used, then, at minimum, a citation of the standard needs to be stated prominently, clearly, completely and early in the document for the system. It would be better if the assumptions and patterns of operation of the GUI were stated in the manual even if the GUI is of a standard style. Otherwise, a complete specification of that GUI assumptions and patterns of operation ought to be included in the manual. TI did not do that for the TI-89. If TI puts that information in a teachers' manual for classroom education, then TI should designate the manual that comes with the calculator as a students' manual and, preferably, include a copy of that teachers' manual with each calculator or include the students' manual in a prominent manner how a copy of the teachers' manual can be obtained, preferably, free by request or at minimal cost. >RJH> Common knowledge is the reason that you find light switches to be >RJH> simple and interpret that simpleness as being intuitive. >RJH> Simplicity and intuitiveness are quite different attributes. >No but you can assume that some standards are known by your target group. You >are a very unique customer because you know computers but don't know GUIs >very well. I guess the majority of todays computer users don't know what to >do with command line tools (and yes I think that's bad). I do _not_ think that I am unique in my needs, desires or life experiences. Most of the world's population do not have access to computers. Most of the world's population does have experience with inexpensive calculators at least at the least of observing local small merchants using them. TI needs to include such users. >cu >Balazs Fischer >Balazs.Fischer@studbox.uni-stuttgart.de >PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 >Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 >Encrypted mail preferred >... Unzip? Explode? Expand? These guys are SICK! Randolph J. Herber, herber@dcdrjh.fnal.gov, +1 630 840 2966, CD/CDFTF PK-149F, Mail Stop 318, Fermilab, Kirk & Pine Rds., PO Box 500, Batavia, IL 60510-0500, USA. (Speaking for myself and not for US, US DOE, FNAL nor URA.) (Product, trade, or service marks herein belong to their respective owners.) ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences <74otnh$e61$1@info3.fnal.govI am requesting yet again a suggestion (pointer, if you will) to an >appropriate, more relevant news group for the TI-89 calculator specifically, >for TI high level calculators (which I do consider the TI-89 to be), or, >at worse, for TI calculators in general. > bit.listserv.calc-ti You may be disappointed, though - the posters are mostly high-schoolers, and the standard of debate isn't generally up to that of this ng. Better bet: the discussion groups at Texas Instruments web-site. For TI 89 / 92: http://www-s.ti.com/cgi-bin/discuss/sdbmessage.cgi?databasetoopen=calcul That's straight out of my bookmark file. Try www.ti.com/calcs if that doesn't work. Also: http://www.ticalc.org/ - unofficial TI site, some discussion. -- |>>>>>>>>>>>>>> Stuart Dawson Dawson Engineering sd@dawson-eng.demon.co.uk http://www.dawson-eng.demon.co.uk/nexus.htm Belfast, Northern Ireland +44 1232 640669 <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<| ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences ] bit.listserv.calc-ti ] ] You may be disappointed, though - the posters are mostly high-schoolers, ] and the standard of debate isn't generally up to that of this ng. You can say that again, the contrast is enormous, especially considering the similarity of the topics under discussion. Dick -- Dick Smith Acorn Risc PC dick@risctex.demon.co.uk ---------------------------------------------------------------------- ==== Subject: Are TI calcs known Wordwidely? hello! I have a question for HP48 users (or TI users) arroud the world: Is TI92 known in your countries? I live in Brazil, and here 95% of users are HP48 users. Is the TI-92 known in Australia, Europe, Africa, Latina America and Asia? What is more knows? Ti or HP? CMarangon http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: Are TI calcs known Wordwidely? I think there are more buyers of TI and Casio or Sharp or whatever calcs, but I prefer HP calcs. The TI 92 can be bought here in Germany, and it's powerful, but it's not of the quality and long lasting concept as the HP calcs are, especially the HP48, which is in trade for nearly 10 years. Most TI's are interesting for beginners of higher grade maths, they use it for one or two years, then a new model comes out, so the old one is obsolete. The HP48, even the older revisions, are still useable because the concept of expandability is far more developped. (OMHO:-) Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences you could look at http://www.ticalc.org/ for further links, maybe a newsgroup... Hope this helps Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: On the topic of TI-89 calculators, my opinions and experiences >you could look at http://www.ticalc.org/ >for further links, maybe a newsgroup... >Hope this helps >Raymond >Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net http://www.ticalc.org/top.html. Nonetheless, your URL put on the path to the URL. Unfortunately, the FAQ available from that URL does not seem have any useful content: please see: http://www.ticalc.org/faq/9.html 2. The name or names of such newsgroups is one of the things that I had requested in my posting to which you referred. Could you offer the names of any? Yes, it did help. Randolph J. Herber, herber@dcdrjh.fnal.gov, +1 630 840 2966, CD/CDFTF PK-149F, Mail Stop 318, Fermilab, Kirk & Pine Rds., PO Box 500, Batavia, IL 60510-0500, USA. (Speaking for myself and not for US, US DOE, FNAL nor URA.) (Product, trade, or service marks herein belong to their respective owners.) ---------------------------------------------------------------------- ==== Subject: Re: Configuring EMU48 > Ho appena comprato il cavo per collegare il PC alla HP48GX bravo/good > Mi serve un emulatore, dove posso trovarlo? vai su (try) http://www.hpcalc.org ce ne sono per windows (emu48) , mac e unix (x48) on this site you can find emulators for windows, mac and unix. > Quale e' il migliore? ce n'e' solo uno per ogni tipo di macchina there's one for each operating system you've got > Come faccio a copiare la ROM della HP sul PC? c'e' un programma che lo fa in ogni emulatore there's a program for downloading rom in every hp48 emulators saluti/hello luca_ -- Luca Sironi * X _ 'underscore' * X | mailto:luca_@geocities.com * XXX * | | _ _ __ __ _ * XXX * * telnet://grex.cyberspace.org XXXXX*** | |_| || / _/ _` | * XXXXX*** | http://www.geocities.com/Paris/2481 * |_____,_____,_|_____ XXXXXXX*** * ______________________________________________________________| X * _-*+ ---------------------------------------------------------------------- ==== Subject: variable baudrate ? Hello hardware hackers, how can I transfer files between my HP48GX and another machine at 7200 bps ? It can do it since StringWriter has an ASCII send function at 7500 bps, but this is useless. 7200 is more useful for me (printing through the serial and parallel interfaces of my Casio FX-850P, overclocked at 1.8432MHz instead of 1.2288 MHz). ---------------------------------------------------------------------- ==== Subject: Re: variable baudrate ? Hello The exact speed at which StringWriter can send data are: 1200 baud 1920 2400 3840 4800 7680 9600 15360 At the time I did StringWriter (8 years ago), I didn't have the technical documentation. therefore the speed I've put in the StringWriter's documentation is an approximation, not the real baud rate. 7200 is NOT a standard baud rate, while all the others above are. The formula is ( the HP48 use a 32768 Hz crystal): 32768 / coeff / 16 = baud rate You can change the speed of the serial port by changing the value in the BAU register (#0010D) and by writing the coefficient. Jean-Yves >Hello hardware hackers, how can I transfer files between my HP48GX and another machine >at 7200 bps ? It can do it since StringWriter has an ASCII send >function at 7500 bps, but this is useless. 7200 is more useful >for me (printing through the serial and parallel interfaces of my >Casio FX-850P, overclocked at 1.8432MHz instead of 1.2288 MHz). ---------------------------------------------------------------------- ==== Subject: Re: need a rom dump?....Necesitas un rom dump? > Just to clarify the issue. Is a ROM dump illegal ? > When I buy HP48 with software I buy ROM too so it is my own version. > I cannot copy it, use it in other applications etc, because it is > copyrighted but is there a law against reading it ? You can dump it, modify and use it, do whatever you want, if you KEEP IT for your own, strictly personnal use. You are the owner of the ROM for YOUR use, but the ROM itself is the property of Hewlett-Packard, when it comes to be distributed. I can quote you the french note of copyright, but it's probably different in USA; here it is: << Les programmes qui contr.99lent ce produit font l'objet de copyrights et tous les droits en sont r.8eserv.8es. La reproduction, l'adaptation et la traduction de ces programmes sont .8egalement interdites sans autorisation .8ecrite pr.8ealable de Hewlett-Packard. >> So, copyright covers: reproducing the ROM, adapting it (to another product or hardware) and translating it is forbidden, unless HP gives your a prior and written authorisation. -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: Help running X48 for mac. >X48 was designed for UNIX platforms. Did you ever used a UNIX windows >manager with a smaller resolution than 1024x768? That's [CENSORED] as >well. We are talking of the *Macintosh* x48 implementation, and a Mac doesn't have the UNIX windows manager, but QuickTime. So I really don't give a fuck here of the Unix windows manager. x48 sur Mac est une grosse merde -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: Help running X48 for mac. X48 was designed for UNIX platforms. Did you ever used a UNIX windows >manager with a smaller resolution than 1024x768? That's [CENSORED] as >well. We are talking of the *Macintosh* x48 implementation, and a Mac doesn't > have the UNIX windows manager, but QuickTime. Er... QuickDraw actually. So I really don't give a fuck here of the Unix windows manager. x48 sur Mac est une grosse merde Since this is a calculator group, let's try some calculations and see where we get? x48 is too tall for the screen running at 1024x768 resolution. So the on-screen image of the 48 requires more than 768 pixels to display. A typical dot pitch for a monitor is 0.26mm. This gives us 0.26 * 768 = 199.68. In other words the original x48 displays the calculator at more than 20cm in height, somewhat larger than life size. Quite why x48 is set up to give a larger than life size display of the calculator I don't know but since that is the case it's not surprising that it doesn't fit onto a typical 15 display, regardless of OS. Of course the real problem may be that the MacOS doesn't allow you to kid yourself that you've got higher resolution than you really have by allowing you to blindly choose resolutions and refresh rates that are not supported by the monitor and video card. -- Bruce Horrocks (Mailto:Bruce.Horrocks@gecm.com) Camberley, Surrey GU15 3PD. Tel: 01276 686777 Fax: 01276 686623 GNET: 832 3032 ---------------------------------------------------------------------- ==== Subject: Re: Help running X48 for mac. >But I don't use that, because I have a 19inch monitor (you can't see >me, but I wearing my smuggest grin right now) and the big X48 looks >fine -- I use it instead of the standard calculator all the time. I don't have a 19 inches monitor. A lot of people don't. Being left at the door because of such a thing makes me real sad :(( All the friends I know on Mac DON'T have a 19 inches of bigger. > Come on HP, make a software 48 -- I'd buy one, and I bet plenty of other > people would too. Macintosh is crap. I use a borrowed one because I've nothing else, and it need this shitty thing to link my HP to. Macs should be sold with KY-Jelly. -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: black out, HELP ME!!!! > I've tried everything > Help me please!!!!!! Remove the small upper-right pad below, and *keep it pressed* for several seconds. If you press one time the reset below the calc, it should do a reset. Good luck. -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: New: Raton 1.0 , use your hp48 as a mouse I finished some weeks ago this little program, which took me several hours of work. however here it is http://members.xoom.com/bnll4 just plug your hp48 to com1 and use it as a mouse moving with the keys, btw there is a disclaimer read it before do anything. German http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: Raton 1.0 , use your hp48 as a mouse Oops, I just made a mistaked and deleted raton10.zip file from my page sorry guys... tomorrow the program will be upload to my page correctly. Probably on friday. German http://members.xoom.com/bnll4 http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Try to Recover Memory? Ugg :-) I just replaced my HP48GX with the GX because it began to clear its memory and 5 years old, but the $100 exchange HP offers is a better deal than a new and inserting & removing the card several times. The card does seem to be the problem, my GX's memory has cleared twice in a week. I'll just be chugging along and blink, fun's over. At one point, somehow, the calculator shut off & wouldn't come back on until I hit the The HP phone tech rep suggested replacing the battery. I guess I never thought of that, I would think the indicator would be showing if the battery were dead. I did it anyway. That was yesterday. Has anyone had similar problems? Can I pull the card apart and inspect it? Could I even buy a new chip and ... well I guess removing the solder would be a real pain. Ed ---------------------------------------------------------------------- ==== Subject: Re: Hardcase for the HP48GX / If you are worried about the LCD, put something like thin plastic, or cardboard in the pocket and make the HP face this always. >Does anyone know where I can buy a hardcase for the HP48GX? >>I looked at www.hpcalc.org, under the buying guide and I didn't see any >>mention of a hardcase. impression and an address: Cusack >P.O. Box 222 >Golden, Colorado 80402 I talked with them on the phone years ago. You might try calling the >directory service in Golden, CO for their number. Good luck. Richard M. Smith >(509) 754-1126 >rmsmith@pobox.com > ---------------------------------------------------------------------- ==== Subject: Re: IR Printer Question I thought of that, but that would extend the light beam length to well over the transmission range (depending on where the IR diode is on the printer. I haven't seen it.) >How about a mirror? >Brian >I would like to be able to use my HP48 with the 82240B IR printer. >>Because of space limitations it will be necessary to mout the two >>units side by side instead of the printer above the calculator. I was >>thinking that perhaps there is a way to conduct the IR signal from >>calculator to printer via some sort of fiber optic thingy. I do not >>wish to modify either device . Has anyone done this sort of thing? I >>would appreciate hearing from anyone with ANY suggestion as to how I >>might accomplish this. >>TIA >>Vic Fraenckel >>victorf@wizvax.net > > ---------------------------------------------------------------------- ==== Subject: Next generation HP? j> Somehow I'd guess they can squeeze more than a 12Mhz chipset in j> there, but if not, well that'll do. I second your hopes for a j> built in periodic table. There are more than enough periodic tables to be downloaded. j> just run BZ compression and you're gtg. What I would like to find j> is a thermodynamics card, with extensive steam tables for example, j> automatic interpolation, etc. Anyone seen one? I haven't tried them, since I never needed them but there are several libraries out there. One is about 80K big. Loot at http://www.hpcalc.org j> I've searched and j> searched and haven't found anything for thermo, or anything with a j> standard atmosphere table... I don't know all that's out there but I have seen atmosphere tables. j> I also tried a shockwave calculation j> program that calculates downstream values behind a normal shock, j> Rayleigh flow, Prandtl flow, Isentropic, and Fanno flow, but I j> found at higher Mach numbers (>3.5), the normal shock calculations j> were inaccurate, at least comparing to my tables...Anyone use this j> program and have you found a better one? What's the name of the program? I once used on for an exam but most of the time I just program the equations into the MES. BTW, you could try http://www.cyberbox.de . Try steam.zip, tbles.zip, thermo.zip, thermopr.zip and hpshokv1.zip. I never tried them except the last lib but maybe one of them does what you need. cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... Each young doctor means a new graveyard. ---------------------------------------------------------------------- ==== Subject: Re: Next generation HP? j> just run BZ compression and you're gtg. What I would like to find j> is a thermodynamics card, with extensive steam tables for example, j> automatic interpolation, etc. Anyone seen one? ---------------------:: o i have a franklin dictionary... ( so we know it's possible ) ...wouldn't it be nice to have something 'like' The Penquin Dictionary of Mathematics/Physics/Chemistry with all the accompaning Equation as hyper-links built into the next generation HP Calc...??? speaking of EQ's... what i find really annoying about the built in EQ's in that in the Plane Geometry section, it DOESN'T have the simple SIN COS TAN equations for a right triangle...( i can never keep them straight...!!! ) NOR the ASA or AAA equations...!!! Ack! ----------- :: o .---..-..-..-..-..-..-. . .-. .-. .-..-..-.. .-. `| |'| || .` | > / / / / / | .` || | ) / `-' `-'`-'`-' `-' `. ^ .' `--^--'`-'`-'`-'' `--^--' Join The BabyNous Cult : The Friendly Neighborhood Cult ---------------------------------------------------------------------- ==== Subject: Re: Next generation HP? > j> just run BZ compression and you're gtg. What I would like to find > j> is a thermodynamics card, with extensive steam tables for example, > j> automatic interpolation, etc. Anyone seen one? > ---------------------:: o i have a franklin dictionary... > ( so we know it's possible ) > ...wouldn't it be nice to have something 'like' > The Penquin Dictionary of Mathematics/Physics/Chemistry with all the > accompaning Equation as hyper-links built into the next generation HP > Calc...??? if you're gonna go there, how about a compressed version of The Handbook of Chemistry and Physics by the CRC Press...mmmmmm...as for your probs with trig functions, all I can say is keep practicing...sin - opposite over hypoteneuse, opposite over hypoteneuse...etc. JDog Revenge is profitable, gratitude is expensive. - Edward Gibbon http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: Next generation HP? j> Somehow I'd guess they can squeeze more than a 12Mhz chipset in > j> there, but if not, well that'll do. I second your hopes for a > j> built in periodic table. There are more than enough periodic tables to be downloaded. But wouldn't it be nice to have it built in? > j> just run BZ compression and you're gtg. What I would like to find > j> is a thermodynamics card, with extensive steam tables for example, > j> automatic interpolation, etc. Anyone seen one? I haven't tried them, since I never needed them but there are several > libraries out there. One is about 80K big. Loot at http://www.hpcalc.org j> I've searched and > j> searched and haven't found anything for thermo, or anything with a > j> standard atmosphere table... I don't know all that's out there but I have seen atmosphere tables. I'm looking for a ROM card with this stuff. Programs are cool and all, but they soak my memory. > j> I also tried a shockwave calculation > j> program that calculates downstream values behind a normal shock, > j> Rayleigh flow, Prandtl flow, Isentropic, and Fanno flow, but I > j> found at higher Mach numbers (>3.5), the normal shock calculations > j> were inaccurate, at least comparing to my tables...Anyone use this > j> program and have you found a better one? What's the name of the program? I once used on for an exam but most of the > time I just program the equations into the MES. > BTW, you could try http://www.cyberbox.de . Try steam.zip, tbles.zip, > thermo.zip, thermopr.zip and hpshokv1.zip. I never tried them except the last > lib but maybe one of them does what you need. hpshokv1.zip is what I have and tried, and found it to be inaccurate. At least in reference to the tables in the back of Gas Dynamics by Zucrow. Hopefully I'll be programming my own shock table calc program soon, which will include an algorithm to calculate oblique shocks as well as normal ones. JD Revenge is profitable, gratitude is expensive. - Edward Gibbon http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Bode plots L> Anyone know of a good program to print bode along with phane plots. The best program for this is Bode from Thomas Reitmayr. It draws the bode plot together with the phase plot and gives you all important points. If you want a program that can also do nichols, root-locus, nyquist plots, design simple PD and PID controllers, give you lots of data about your transfer function, do Routh stability testing and can draw the response for any input you should try Bode-Routh from APO. cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... Whatever you delete today, you desperately need tomorrow. ---------------------------------------------------------------------- ==== Subject: Trigo I search for a program who allows me to resolve trigonometry equations ---------------------------------------------------------------------- ==== Subject: HP-01 Digital Wrist Instrument Does anyone know how much these might be worth? I've got a fine condition one with manuals, box, and all (the stainless steel one). Works fine. Not sure I want to part with it (I bought it new directly from HP in '79) but am looking for an appraisal as it were. Jim Horn (does wearing an HP-01 certify me as a Nerd?) ---------------------------------------------------------------------- ==== Subject: Re: HP-01 Digital Wrist Instrument > Does anyone know how much these might be worth? I've got a fine condition > one with manuals, box, and all (the stainless steel one). Works fine. > Not sure I want to part with it (I bought it new directly from HP in '79) but > am looking for an appraisal as it were. Not sure, but if you sell it to anybody other than me, I'll tell Dad! > Jim Horn (does wearing an HP-01 certify me as a Nerd?) Yes, but so does posting messages on comp.sys.hp48! >:-O -Joe Horn- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: HP-01 Digital Wrist Instrument [J. Horn & J. Horn were discussing how much (money) an HP-01 is worth] describing it. It had extremely ingenious functions which could display real-time results of, say, multiplying an elapsed time by a cost per minute (or second). So, why didn't any of these ingenious functions ever find their way into any other product, perhaps card-sized, or whatever? It wasn't even imitated by Casio! (what, was it all patented, or something?) Perhaps an LCD-based product would not be as good at displaying real-time (changing) results? [the HP-01 had LED display, which has instant response, unlike slightly sluggish LCDs] Hey, if you buy it, Joe, I might come to the next Handheld conference just to see it! :) ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: Re: HP-01 Digital Wrist Instrument JHM notes: So, why didn't any of these ingenious functions ever find their way into any other product, perhaps card-sized, or whatever? It wasn't even imitated by Casio! (what, was it all patented, or something?) ---------------------------------------------------------:: o i had a really nice...( was it Casio...?... )...'credit-card' watch several years ago that did alot of really beezar things...like play musical tunes and calculate weird ass sh*t ... why don't they make stuff like that anymore... i've been looking around for new watch, but they genuinely seem too be getting more & MORE BBBOOORRRIIINNNGGG...!!! what is what...??? i would think, ( you would think? ) that technology would allow products to be getting weirder and weirder...???...but they aren't...??? ( just more examples of the Time Travelers suppressing technology...Like ROBOTS*...!!! ) -------- * Where are all the Dang Robots...and Robot kits...!!!??? what happened to Logo, and Turtles...??? There used to be some kits, but now they're gone...and how can this be explained, except by; Time Travelers from the future mucking with our timeline...??? ----------- :: o .---..-..-..-..-..-..-. . .-. .-. .-..-..-.. .-. `| |'| || .` | > / / / / / | .` || | ) / `-' `-'`-'`-' `-' `. ^ .' `--^--'`-'`-'`-'' `--^--' Join The BabyNous Cult : The Friendly Neighborhood Cult ---------------------------------------------------------------------- ==== Subject: Re: HP-01 Digital Wrist Instrument >why don't they make stuff like that anymore... >i've been looking around for new watch, but they genuinely seem too be getting >more & MORE BBBOOORRRIIINNNGGG...!!! what is what...??? IMO Casio's lack of innovation is a classic illustration of monopolistic complacency. They are still selling their Data Watch, for example, which is hardly changed after .... 20? years. Before they won the 70s war of Watch Gadgetry watches were changing rapidly. ---------------------------------------------------------------------- ==== Subject: Re: HP-01 Digital Wrist Instrument > Hey, if you buy it, Joe, I might come to the > next Handheld conference just to see it! :) Attend it either way, since the HP-01 seems to contain some sort of homing device that attracts them to these conferences! Every year, a large percentage of attendees wear HP-01's! You can even tell who is wearing one at a distance: one arm is longer than the other. And, of course, they are surrounded by babes. See you in Zaandam, Holland, in '99! -Joe- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: HP-01 Digital Wrist Instrument Well, I can remember that I almost bought one back in 1980 but just couldn't afford it on an apprentice pay. I regret that I didn't, but looking on the bright side I don't have to worry about Y2K compliance... ;) ---------------------------------------------------------------------- ==== Subject: Re: HP-01 Digital Wrist Instrument OK Gang - All but John and Joe - sorry for the semi-off-topic posting! But in '79 HP made its HP-01 available to its employees *new* just to get rid of them (sales were terrible) for $195 (stainless) or $245 (gold). The results were so positive they had to reopen the line to meet the demand. So they're out there... Of course, a $195 watch on a $18k income still wasn't pocket change. But compared to the list price of $695 - wow! John: Howzabout I lend it to Joe so you can? Joe: I was startled breathless at the hpmuseum value listings - seems a complete '01 is worth $1100 - $2000. Yikes! That's *too high* to sell it so I don't intend to. Wouldya like to borrow it? If you do, Dad will know - the tidal effects of moving a 4+ ounce (~120g) 'watch' from northern California to southern CA may be directly observable in Wisconsin. Or any pictures of you that he may get which would show the list to port... Will you be getting that to go over the Christmas break??? Jim Older than Joe but younger than Henry Horn ---------------------------------------------------------------------- ==== Subject: Re: HP-01 Digital Wrist Instrument Saw one of these for sale on E-bay for lots of $$$. I think the price was about $800.00 and the auction was still going. I'm not sure how high it went. You sure you want to part with a babe-magnet like the HP-01?. > Does anyone know how much these might be worth? I've got a fine condition > one with manuals, box, and all (the stainless steel one). Works fine. > Not sure I want to part with it (I bought it new directly from HP in '79) but > am looking for an appraisal as it were. > Jim Horn (does wearing an HP-01 certify me as a Nerd?) ---------------------------------------------------------------------- ==== Subject: Re: Future of RPL & Assembly > One delicate point here: How many are ready to dash out > and buy ROM version A of this new product ? John: Your comments were highly informative and humorous! I really laughed at the above quote - for I have a weakness: I'm one of the guys who would rush out and buy the newest version. Hopefully the new HP may have some sort of electronically upgradable ROM like TI is now using, and spare us another series of ROMs A-R! - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) I recently purchased a 48GX so that I could put Java on, and other libraries like ALG48. I just recently got my calculator to connect to my computer, so I tried Java 3.6. I also put a few other libraries on, such as UFL and QPI. Well, while I was playing around with it, the calculator went into what looked like a warm start. Then it came up with the Try to recover memory? prompt. I pressed yes, and waited. Then the prompt came up again. Finally, I pressed no. Needless to say, that erased everything. Next I tried HPsauce, Ranch, 2.2. I had the same problem, but instead of the repeating the memory recovery prompt, it would just remove some or all of my libraries without recovering the memory, and leave the variables alone. I recently posted a question about that problem, and Andre Schoorl answer back saying that I should try using Java 3.6. So I gave Java another shot. Well, now I had a total of 4 libraries and a few variables on my calc. Then, all I did was press the VAR button, when I got an Invalid Card error (or something very close to that). Two of my libraries were erased, Java being one of them. I have a rom, revision R, which should work fine with Java 3.6. Does anyone have any suggestions or ideas as to what could be happening. Could it have anything to do with the fact I have quite a few uesr-defined keys, and use them quite often? (To do things like round to 4, 3, or 2 places, etc.) The following is a list of libraries that I had installed during the most recent crash (for lack of a better word). I have cut down a few from when I started with Java, thinking that they were causing conflicts. 1037 - FASTD1.3 257 - UFL 1.0, level 3 905 - QPI 4.3 908 - Java 3.6 you in advance. And he who forgets.. Will be destined to remember.. To reply by email, remove the asterisk from my address ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) > I recently purchased a 48GX so that I could put Java on, and other > libraries like ALG48. I just recently got my calculator to connect to > my computer, so I tried Java 3.6. I also put a few other libraries > on, such as UFL and QPI. Well, while I was playing around with it, > the calculator went into what looked like a warm start. Then it came > up with the Try to recover memory? prompt. I pressed yes, and > waited. Then the prompt came up again. Finally, I pressed no. > Needless to say, that erased everything. you in advance. Mark: I know very little about Java. I don't use it because it takes up about 30 KB in a covered port, (ports 0 or 1) I use EQSTK 9.1, which runs have all the features of Java, but it's never crashed on me, and it does a great job of showing large equations and matrices. In addition, it is snap to use, you can toggle in and out of EQSTK very fast. It even runs with the MetaKernal program, which is a ROM card stack replacement, without crashing. I actually mix the two programs! To use EQSTK you will need the UFL program for fonts and the AIO48 program for showing symbolic matrices. If you have ALG48 and EQSTK, you'll be the best armed person in any linear algebra class when it comes to crunching matrices! Bernard Parisse, the author of Erable, modified EQSTK to show the different modes of Erable when using Erable 3.2, which is really great. Give it a try. - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) > I have a rom, revision R, which should work fine with Java 3.6. > Does anyone have any suggestions or ideas as to what could be > happening. Could it have anything to do with the fact I have quite a > few uesr-defined keys, and use them quite often? I'm using Java 3.6 and I have USER mode always ON, with more than half of the keyboard redefined, with programs that check stack's context, and such. I don't get problems with it, and I also have a R ROM. Where did you store the Java library? > 1037 - FASTD1.3 What is that lib for? :) -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) [...] >I have a rom, revision R, which should work fine with Java 3.6. >Does anyone have any suggestions or ideas as to what could be >happening. Could it have anything to do with the fact I have quite a >few uesr-defined keys, and use them quite often? (To do things like >round to 4, 3, or 2 places, etc.) The following is a list of >libraries that I had installed during the most recent crash (for lack >of a better word). I have cut down a few from when I started with >Java, thinking that they were causing conflicts. 1037 - FASTD1.3 >257 - UFL 1.0, level 3 >905 - QPI 4.3 >908 - Java 3.6 Which flavour of Java (and HPSauce) did you try? I'm wondering if it's a possible ML Decomp bug, or if it's something else. Since some flavours contain it and others don't, trying ones that do / do not contain it might help to isolate the problem. Also unless you can come up with a specific way to reproduce a problem -- Andre Schoorl PGP key available on request/keyserver Comp Engineering, UVic, Canada Linux/HP48 http://www.engr.uvic.ca/~aschoorl/ ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) >Which flavour of Java (and HPSauce) did you try? I'm wondering if it's >a possible ML Decomp bug, or if it's something else. Since some flavours >contain it and others don't, trying ones that do / do not contain it might >help to isolate the problem. Also unless you can come up with a specific way to reproduce a problem I've tried HPsauce Ranch, and Java flavors Cappuccino and Frappuccino. It has done it with both Cappuccino and Frappuccino, and after double checking the instructions, I realized that Cappuccino did not have ML Decomp. I guess right now I'll try with just Java and my variables. And he who forgets.. Will be destined to remember.. To reply by email, remove the asterisk from my address ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) I gotta ask: have you tried loading java WITHOUT your user key programs? I understand that they are valuable to you but that might help you find out if they are incompatible with java. Brian >Which flavour of Java (and HPSauce) did you try? I'm wondering if it's >>a possible ML Decomp bug, or if it's something else. Since some flavours >>contain it and others don't, trying ones that do / do not contain it might >>help to isolate the problem. >>Also unless you can come up with a specific way to reproduce a problem I've tried HPsauce Ranch, and Java flavors Cappuccino and Frappuccino. >It has done it with both Cappuccino and Frappuccino, and after double >checking the instructions, I realized that Cappuccino did not have ML >Decomp. I guess right now I'll try with just Java and my variables. >And he who forgets.. >Will be destined to remember.. To reply by email, remove the asterisk from my address ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) I love Java but no matter what version I use, even on a clean calc with nothing else on it I alway run into trouble. Sooner or later something happens and I loose half or all of the memory. I have faith, thought, I keep trying the new relases as the come out but it eventually crashes. I currently use EQstack which I've never had crash but it's functions are very limited. By the way, the crashes usually occur when I either press the VAR button or R-SHIFT 2 (Library) (similar to original poster). I have only tried frapacino, however. If the bug gets fixed I'll be one of the first to install it. Just my thoughts, Andrew >I gotta ask: have you tried loading java WITHOUT your user key programs? I >understand that they are valuable to you but that might help you find out if >they are incompatible with java. >Brian ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) I have written a message to Mr. Schoorl, but I haven't got an answer yet. So here's my message (some parts cut) about a problem with Java 3.6 crashing: >I'm using Java 3.6 with Hp48 GX, version R (1993). No home-made memory cards or anything just the basic 128k. The problem is that >after I've installed the Java library (in this case Frappucino), the UFL3 library, warm booted my calculator and activated Java (oval >menus+algebraics in small+java active) everything seems to be working fine. If I now turn off the calc and back on then choose the MTH >menu, there comes some gibberish on the screen, HP makes a beepy noise and then displays the message Try to recover memory?. >Sometimes it can recover the whole memory, but most of time everything disappears! So practically to reproduce this just upload Java36f.lib (frappucino) and ufl3.lib to your HP48. Store both libraries into port 0. Purge them from memory and do a warm start ON-C. Then choose the MTH or VAR menu and it crashes. It has done this every time! Yours Antti Havulinna ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) >I have written a message to Mr. Schoorl, but I haven't got an answer yet. So >here's my message (some parts cut) about a problem with Java 3.6 crashing: >I'm using Java 3.6 with Hp48 GX, version R (1993). No home-made memory >cards or anything just the basic 128k. The problem is that >after I've >installed the Java library (in this case Frappucino), the UFL3 library, warm >booted my calculator and activated Java (oval >menus+algebraics in >small+java active) everything seems to be working fine. If I now turn off >the calc and back on then choose the MTH >menu, there comes some gibberish >on the screen, HP makes a beepy noise and then displays the message Try to >recover memory?. >Sometimes it can recover the whole memory, but most of >time everything disappears! >So practically to reproduce this just upload Java36f.lib (frappucino) and >ufl3.lib to your HP48. Store both libraries into port 0. Purge them from >memory and do a warm start ON-C. Then choose the MTH or VAR menu and it >crashes. It has done this every time! Yours Antti Havulinna Correction, after the warm start activate java with small algebraics, oval menus and make java active. Then restart the calc and choose MTH or VAR menu! ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) >I have written a message to Mr. Schoorl, but I haven't got an answer yet. > So >here's my message (some parts cut) about a problem with Java 3.6 crashing: >I'm using Java 3.6 with Hp48 GX, version R (1993). No home-made memory >cards or anything just the basic 128k. The problem is that >after I've >installed the Java library (in this case Frappucino), the UFL3 library, > warm >booted my calculator and activated Java (oval >menus+algebraics in >small+java active) everything seems to be working fine. If I now turn off >the calc and back on then choose the MTH >menu, there comes some gibberish >on the screen, HP makes a beepy noise and then displays the message Try to >recover memory?. >Sometimes it can recover the whole memory, but most of >time everything disappears! >So practically to reproduce this just upload Java36f.lib (frappucino) and >ufl3.lib to your HP48. Store both libraries into port 0. Purge them from >memory and do a warm start ON-C. Then choose the MTH or VAR menu and it >crashes. It has done this every time! Yours Antti Havulinna Correction, after the warm start activate java with small algebraics, oval > menus and make java active. Then restart the calc and choose MTH or VAR > menu! I did exactly as you have described and did not experience any problems. I don't know what to say.. where did you download JAVA from on the web? I got mine from the Eric Rechlin's archive. If you got it somewhere else, then you might try that. -- ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) I did exactly as you have described and did not experience any problems. I >don't know what to say.. where did you download JAVA from on the web? I got >mine from the Eric Rechlin's archive. If you got it somewhere else, then you >might try that. -- > > Did you install the Frappucino library? I got mine directly from Andre Schoorl's site so I don't think that's the problem. Which rom do you have in your hp? R? I've got a friend with a hp48GX with R rom (1993) and Java crashes on his calc too! Weird :) Antti ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) [...] >Did you install the Frappucino library? I got mine directly from Andre >Schoorl's site so I don't think that's the problem. Which rom do you have in >your hp? R? I've got a friend with a hp48GX with R rom (1993) and Java >crashes on his calc too! Weird :) Might be a ROM R only kind of bug. I'll check it out. There's a few other minor things I might like to fix as well, so maybe I'll make a 3.6a. -- Andre Schoorl PGP key available on request/keyserver Comp Engineering, UVic, Canada Linux/HP48 http://www.engr.uvic.ca/~aschoorl/ ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) >I love Java but no matter what version I use, even on a clean calc with >nothing else on it I alway run into trouble. Sooner or later something >happens and I loose half or all of the memory. I have faith, thought, I keep trying the new relases as the come out but it >eventually crashes. I currently use EQstack which I've never had crash but >it's functions are very limited. By the way, the crashes usually occur when I either press the VAR button or >R-SHIFT 2 (Library) (similar to original poster). I have only tried >frapacino, however. If the bug gets fixed I'll be one of the first to install it. Please try Cappuccino or Decaf and see if the problems still occur. And as I have said previously, we need specific ways to reproduce bugs to be sure they are Java's fault and not something else. I have gotten several reports of things that weren't related to Java at all, but once people install it they automatically blame everything bad that happens on it. I'm not saying the problems you mention aren't Java's fault - they probably are, but it may be a combination of other external elements. And I'd like to remind everyone, that the biggest is not necessarily the best (i.e. Frappuccino). The other ones are there for a reason, but I find most people end up installing the biggest one. The default flavour in the code is actually Espresso, not Frappuccino. -- Andre Schoorl PGP key available on request/keyserver Comp Engineering, UVic, Canada Linux/HP48 http://www.engr.uvic.ca/~aschoorl/ ---------------------------------------------------------------------- ==== Subject: Re: Stack Replacement Trouble (Java 3.6, HPsauce 2.2) It happened to me too. It fails when you push VAR, and even MATH. The first time it failed I intoduced 3digit numbers in to the stack and the started +. Suddenly the screen appeared with noise, and then it said: Try to recover memory? Pushed yes, and then I lose everything, but the memory was occupied. I tried frapuccino three times and the first one in the list (don«t remenber the name). -=UssU=- >I love Java but no matter what version I use, even on a clean calc with >nothing else on it I alway run into trouble. Sooner or later something >happens and I loose half or all of the memory. I have faith, thought, I keep trying the new relases as the come out but it >eventually crashes. I currently use EQstack which I've never had crash but >it's functions are very limited. By the way, the crashes usually occur when I either press the VAR button or >R-SHIFT 2 (Library) (similar to original poster). I have only tried >frapacino, however. If the bug gets fixed I'll be one of the first to install it. Just my thoughts, >Andrew >>I gotta ask: have you tried loading java WITHOUT your user key programs? >I >>understand that they are valuable to you but that might help you find out >if >>they are incompatible with java. >>Brian > ---------------------------------------------------------------------- ==== Subject: some EQStk Tips >I have faith, thought, I keep trying the new relases as the come out but it >eventually crashes. I currently use EQstack which I've never had crash but >it's functions are very limited. What is limited with EQStk? It offers a 7-level-stack, which is very usefull. I hope a 7-level-interactive stack can be released soon. I have to ask (name snipped :-) again, if he allows to release his hacked version. The only thing which I found missing is an on-sreen-scroller. To scroll big EQ's or other objects I use TED (including the viewer VV). I have assigned VV to my Down-Arrow-Key. That works perfect for me. In addition I use FC to view/change my flags fast. Then I don't need to see the status area. In addition I have assigned some usefull *programs* (what a big word for such tiny structures ;-) to some keys, which help me to see when I switch between different modes. Here are some tips for the use with EQStk: Status area ~~~~~~~~~~~ While running EQStk+ you can't see the status area. Here are some optional suggestions to circumvent this. To see the change of RAD/DEG you can assign a tiny program to a user key. This will give you a short message which mode you use, when changing between RAD/DEG. (Sure user keys have to be active then. I suggest to set flag -61 and flag -62 by typing -61 SF and pressing LeftShift USER [ENTER-key]. For further explanation of USER keys and CST (custom) menus see chapter 30 in the manual). Type that program: << IF -17 FC? THEN RAD RADIANT ELSE DEG DEGREE END 1 DISP >> Type: 21.2 ASN It's not neccessary to show the POLAR mode, because you see in the representation of the stack, if POLAR mode is on. For example: POLAR mode active: (5, <53.1301023542) POLAR mode not active: (3, 4) To show that you are turning off USER mode you can assign that program to the USER key: << -62 CF USER OFF 1 DISP >> 61.2 ASN If you see the message USER OFF when pressing LeftShift USER [alpha-key] you're just quitting USER mode. If you don't see a message you're just entering USER mode. Another usefull idea would be to assign ASTK (to start EQStk+ /to switch between small and large font) to a user key. Or to add ASTK to your CST (custom) menu. This way one could fast start EQStk+ /toggle the font size. And remember CONT will quit EQStk+ and return to the built in stack display. Scrolling large objects ~~~~~~~~~~~~~~~~~~~~~~~ If an object doesn't fit on the stack you will see ; instead of : after the number of the line, like 1; This is a large ob... instead of 1: 34545 If you want to take a look on large objects like equations/matrices and so on I want to suggest to use another splendid library from Mika Hiskanen - TED. TED is an editor and includes a viewer for all objects. This viewer is called VV. With the use of VV you can also scroll large objects very fast and easy. VV starts much faster then the built in viewer. To view a really large equation it took about 1 minute with the built in viewer, but only 2 seconds with VV! You may want to include VV in a CST menu or assign VV to a user key. For example you could assign VV to the Down-Arrow key. To do this you would have to follow these steps: Type VV in a list: { VV } extract VV by the following key sequence with the command OBJ-> (I just write down the characters of the keys): H C A then DROP 1 and finally type 35.1 ASN That's it. The trick with the list is necessary, because you can't enter a command (like VV) on the stack direct. This way you can assign every built in command or library command to a key. Notice that the built in VIEW command can still be reached via the LeftShifted Down-Arrow-key. All mentioned libraries can be found at the superb HP48 Software Archive http://www.hpcalc.org Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: It messes up loop structures I haven't checked out why, but the new ML EqualPosComp posted her a few days ago messes up some loop structures. I have a ZERO_DO loop that should be executed 25 times and it's done after 3 or 4 times. I haven't changed anything else, I only replaced one apndvarlst with the new routine. BTW, is there any replacement for >TCOMP or >HCOMP? cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... Tighten it until it cracks, then back off a half turn. ---------------------------------------------------------------------- ==== Subject: Re: It messes up loop structures boundary=------------44CED8A15FDBBEBC2836735B --------------44CED8A15FDBBEBC2836735B Hello! I took the program that was posted a few days ago and compiled it. It works o.k. for me. The only thing is though, I didn't get a speed gain. I lost a little speed. Is the >TCOMP instruction the blame for that? If so, please help with a new >TCOMP instruction!!!!! Jeremy E. Laughery--Author of Exhackt the h and k are silent : ) > I haven't checked out why, but the new ML EqualPosComp posted her a few days > ago messes up some loop structures. I have a ZERO_DO loop that should be > executed 25 times and it's done after 3 or 4 times. I haven't changed > anything else, I only replaced one apndvarlst with the new routine. BTW, is there any replacement for >TCOMP or >HCOMP? cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de > PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 > Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 > Encrypted mail preferred ... Tighten it until it cracks, then back off a half turn. --------------44CED8A15FDBBEBC2836735B Hello!
    I took the program that was posted a few days ago and compiled it. It works o.k. for me.
The only thing is though, I didn't get a speed gain. I lost a little speed. Is the >TCOMP instruction the blame for that? If so, please help with a new >TCOMP instruction!!!!!
Jeremy E. Laughery--Author of Exhackt           &nbs p;  the h and k are silent    : )
I haven't checked out why, but the new ML EqualPosComp posted her a few days
ago messes up some loop structures. I have a ZERO_DO loop that should be
executed 25 times and it's done after 3 or 4 times. I haven't changed
anything else, I only replaced one apndvarlst with the new routine.

BTW, is there any replacement for >TCOMP or >HCOMP?

cu

Balazs Fischer

Balazs.Fischer@studbox.uni-stuttgart.de
PGP Key: htt p://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5
Fingerprint: 8F EF C0 BA 22 B4 15 B3  56 27 07 04 37 7B D6 00
Encrypted mail preferred

... Tighten it until it cracks, then back off a half turn.

--------------44CED8A15FDBBEBC2836735B-- ---------------------------------------------------------------------- ==== Subject: Re: It messes up loop structures I tested the apndvarlst replacement with a loop which ran 200 times, and no problem arose. It may be a side effect of some code in your loop. Please tell me what your code does, maybe we'll find the place where that side effect happens. The test program is easy: :: xVARS DUP x+ DUP x+ ( *Many times, til it has a reasonable size* ) % 200 COERCE #1+_ONE_DO (DO) INDEX@ UNCOERCE DISPROW1 ( *Show index* ) ' ID APNDV2 ( *The item to search for* ) ID APNDV2 ( *The apndvarlst replacement* ) LOOP ; The speed gain is marginal if you take only a little list. The slowest part is >TCOMP, I'll work on this. That's all for today Gretings Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: It messes up loop structures I tested the apndvarlst replacement with a loop which ran 200 times, and no problem arose. It may be a side effect of some code in your loop. Please tell me what your code does, maybe we'll find the place where that side effect happens. The test program is easy: :: xVARS DUP x+ DUP x+ ( *Many times, til it has a reasonable size* ) % 200 COERCE #1+_ONE_DO (DO) INDEX@ UNCOERCE DISPROW1 ( *Show index* ) ' ID APNDV2 ( *The item to search for* ) ID APNDV2 ( *The apndvarlst replacement* ) LOOP ; The speed gain is marginal if you take only a little list. The slowest part is >TCOMP, I'll work on this. That's all for today Gretings Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: It messes up loop structures it seems we have found the leak: [..] > ID EQUALPOSCOMP > #0< caseDROP > >TCOMP > LOOP > ; [..] The caseDROP jumps to the SEMI (;) so you leave the loop before you should. Sandwich the sequence .. ID EQUALPOSCOMP #0<> caseDROP >TCOMP .. into DOCOL (::) and SEMI (;), then it should work;-) The sequence should look like the one below: :: ID EQUALPOSCOMP #0<> caseDROP >TCOMP ; Hope this helps Raymond BTW: I'm working on a faster version of FINDVAR (#353ABh). It makes heavy use of apndvarlst and matchob? and is recursive, too. Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: My error! (was: It messes up loop structures) RH> it seems we have found the leak: RH> [..] RH>> ID EQUALPOSCOMP RH>> #0<> RH>> caseDROP RH>> >TCOMP RH>> LOOP RH>> ; RH> [..] RH> The caseDROP RH> jumps to the SEMI (;) RH> so you leave the loop before you should. Ooops, my error. A hint to all coders, don't code after 18h of work :-). RH> BTW: I'm working on a faster version of FINDVAR (#353ABh). RH> It makes heavy use of apndvarlst and matchob? and is recursive, RH> too. Would be cool but it wouldn't help much. GETVAR needs about 5 seconds for a list with 25 equations with 43 different variables. If you really want to speed up something how about NUMSOLVE? I could use that in my MES replacement. cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... Error propagates faster than truth. ---------------------------------------------------------------------- ==== Subject: It messes up loop structures R> I tested the apndvarlst replacement with a loop which ran 200 R> times, and no problem arose. R> It may be a side effect of some code in your loop. R> Please tell me what your code does, maybe we'll find the R> place where that side effect happens. OK here it comes: :: CK&DISPATCH1 FIVE :: DUPNULL{}? ?SEMI INNERCOMP NULL{} SWAP ZERO_DO SWAP ID GETVAR DUPNULL{}? ITE_DROP :: INNERDUP #2+ROLL SWAP ZERO_DO SWAP2DUP ID EQUALPOSCOMP #0<> caseDROP >TCOMP LOOP ; LOOP ; ZERO PTR 353AB ; Give it a list of equations. BTW it's recursive so the name has to be GETVAR. I needed that because sometimes it gets symbolic matrices to crunch. It works perfectly if you replace SWAP2DUP ID EQUALPOSCOMP #0<> caseDROP >TCOMP with SWAP apndvarlst. cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... Trust me, I'm a lawyer. ---------------------------------------------------------------------- ==== Subject: It messes up loop structures JL> Hello! JL> I took the program that was posted a few days ago and compiled JL> it. It works o.k. for me. It works flawlessly if you execute it without a loop. But inside a loop it messes up the index. I have a program called getvar that takes a list of equations and returns all variable contained in them. It has to take every single equation, check for variables in it and then add them to the list of variables if they are not already contained in that list. I started the program with a list of 25 equations and it stopped after 3. The program works perfectly with apndvarlst. JL> The only thing is though, I didn't get a speed gain. I lost a JL> little speed. Is JL> the >TCOMP instruction the blame for that? If so, please help with JL> the >a new >TCOMP JL> instruction!!!!! Yep, I could use that too. If I have time maybe I will code one but RL has a higher priority for me in the moment. cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... I didn't do anything - unless I was supposed to. ---------------------------------------------------------------------- ==== Subject: le CRC ? Salut, j'ai cherch.8e des routines de CRC, il faut que je puisse calculer le CRC d'un objet, et j'ai regard.8e des trucs qui contiennent le mot CRC dans la table EC de Jazz (dans le bouquin que j'ai, rien dessus, sauf un truc qui parle de 4 quartets .88 lire) Le premier que j'ai trouv.8e c'est OCRC, et l.88 j'ai rien compris. D'abord, j'ai un GOSUB, ce dernier lit l'adresse de l'objet un, fais un DROP dessus, passe l'objet avec un SKIPOB, et renvoie la longueur de l'objet dans Aa et R0a Ensuite, il y a un saut .88 DoCRCc, et il semble r.8ecup.8erer le CRC dans R1a. C'est cette routine qui me pose probl.8fme. Elle fait pointer D0 sur Ca; je pense que C pointe sur l'objet, mais j'en sais rien. Ensuite, D1 pointe sur l'adresse o.9d commen.8dent les 4 quartet du CRC. Dans mon bouquin, il indique qu'il faut tous les mettre .88 z.8ero, pour ensuite lire les quartets de l'objet. Je pensait donc trouver un A=0 4, voire plut.99t sur champ A vu que c'est plus rapide (enfin, le bouquin le dit). Ensuite, une boucle de lecture; on divise A par 16, on garde le reste; on boucle autant de fois qu'on peut lire 16 quartet, et .88 la fin, on place dans P le reste-1. Je trouve un truc diff.8erent dans DoCRCc, et je comprends pas. J'ai ceci: D0=C D1=(5) =CRC ST=0 15 L.88, normal. D0 semble pointer sur l'objet, D1 sur les quatre quartet du CRC, et on coupe toutes les interruptions, comme le bouquin le dit. Ensuite, je pensait trouver la mise .88 z.8ero des quatre quartets du CRC. Pas du tout! Il .8ecrit 4 quartets de C dans le CRC, l.88 je suis perdu: DAT1=C 4 mais C pointe sur l'objet! :( Ensuite, c'est pire, je comprend rien: C=A B P=C 0 ASR A P=P-1 C contient ici la taille de l'objet (enfin, je crois) mais seulement sur 2 quartet, et P contient le quartet de poids faible de la taille. Mais .88 quoi .8da sert?!? Le ASR je crois qu'il divise la taille par 16, pour calculer le nombre de lecture de champ W (ce que j'attendait) mais ce qui est avant, je comprend rien. Ensuite, il d.8ecr.8emente P. S'il est nul, on a donc aucun reste de la division par 16, non? Sinon, il ex.8ecute cela: C=DAT0 WP CD0EX C+P+1 D0=C Bon, moi je pensait lire par paquets de W, et ensuite m'int.8eresser au reste. En fait, eux ils commen.8dent par .8eliminer le reste (mais je sais pas trop comment) et .88 lire autant de quartets qu'il faut pour l'.8eliminer, et donc obtenir une valeur de D0 qui pointe juste apr.8fs, sur un multiple de 16 de quartets qui restent .88 lire. Cela me semble mieux, surtout si on a besoin d'un CRC sur un nombre de quartets inf.8erieur ou .8egal .88 15. Mais je comprend pas vraiment pourquoi il .8ecrit C (qui pointe sur l'objet) dans les quatres quartets du CRC, et de coller la taille (Aa) dans C sur deux quartets, si on teste qu'un seul quartet apr.8fs, qu'on place dans P. La suite je m'y retrouve un peu: A=A-1 A GOC FINI LOO C=DAT0 W D0=D0+ 16 A=A-1 A GONC LOO A=0 A A=DAT1 4 GOVLNG =AllowIntr On retire 1 .88 cause de la boucle par carry. Ensuite, si on n'avait que 15 quartets ou moins, fini. Sinon, il boucle autant de fois qu'il faut sur W, et ensuite il va lire le CRC. Le d.8ebut de la routine, et certains trucs, je comprends pas. On peut m'expliquer? Mer.8di. -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: le CRC ? Salut: Tout d'abord te dire que j'ai le meme probleme que toi avec le group hp.programming... Je vais copier presque literelment du libre Voyage au centre... Je ne vais pas solutioner ton probleme, je vais le replanter: Calculateur de CRC #00105 #00106 #00107h Adreses calculateur de Codes de Redondance Cyclique=20 La 48 a un circuit electronique pour calculer le CRC.=20 Pour calculer un CRC ( comme realise la fonction BYTES ) il suffit donc de forcer a zero les 4 quartets de l'adresse #00104h, puis de lire les quartets de l'object dont on veut calculer le CRC ( je ne comprends pas tres bien ca, parce que si ce que l'on veut est l'adresse de cet object on aurais bessoin de 5 quartets, mais on dit lire et pas ecrire: comme est-ce que je vais lire dans ces quartets l'addresse de l'object dont je=20 veux calculer le CRC), puis de relire les 4 quartets en #00104h. Cette lecture ne devant pasa etre perturbee, il faut donc interdire les interruptions lorsqu'on l'effectue ( par la instrution INTOFF ) Attention il ne faut pas oublier de re-autoriser ces ints ( INTON ) De plus, ces 4 quartets ayant des valeurs tres variables on peut pas les utilisser comme generateur des nombres aleatoires en ASM. J'essaierais... mais il y a quelqu'un qui sache la reponse?=20 ---------------------------------------------------------------------- ==== Subject: Re: le CRC ? D0=C > D1=(5) =CRC > ST=0 15 L.88, normal. D0 semble pointer sur l'objet, D1 sur les quatre quartet du > CRC, et on coupe toutes les interruptions, comme le bouquin le dit. > Ensuite, je pensait trouver la mise .88 z.8ero des quatre quartets du CRC. > Pas du tout! Il .8ecrit 4 quartets de C dans le CRC, l.88 je suis perdu: DAT1=C 4 mais C pointe sur l'objet! :( > Have a look at it again. It's not all that difficult: =DoCRCc ( A.A contains #nibs to scan ) D0=C ( D0 contains start adress ) =DoCRC C=0 A D1=(5) =CRC ( D1 points to hardware CRC ) ST=0 15 ( disable interrupts ) DAT1=C 4 ( set CRC to zero !!!! ) C=A B P=C 0 ( P contains lower nib of A.A ) ASR A ( A div 16 ) P=P-1 ( since C=DAT0 WP reads P+1 nibs.. ) GOC + ( if P was zero, skip ) C=DAT0 WP CD0EX ( adjust D0 ) C+P+1 D0=C + P= 0 A=A-1 A ( test A.A = 0 & decrement ) GOC ++ - C=DAT0 W A=A-1 A GONC - ++ A=0 A ( read hardware CRC ) A=DAT1 4 GOVLNG =AllowIntr You are right about one thing: usually the trick with P= lowest nibble and A.A div 16 handles the leftover nibs at the end. (See MOVEUP/MOVEDOWN for instance). Here it's done the other way round... no big deal. Here's a small program I use for one special lib completely written in ML, for which I have to provide the CRC at the end (teh def of the lib simply includes 0000): * * adds the CRC to the end of a library * the lib must have the space for the CRC included; this * prog merely pokes the right CRC in the 4 last nibs. * :: CK1NoBlame CK&DISPATCH1 # 8F :: CKREF CODE GOSBVL =SAVPTR A=DAT1 A D0=A D0=D0+ 5 ( point at length field ) A=DAT0 A A=A-CON A,4 ( do not include the CRC field itself ) GOSBVL =DoCRC DAT0=A 4 GOVLNG =GETPTRLOOP ENDCODE ; ; -- Werner Huysegoms remove the x before replying http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: le CRC ? > Have a look at it again. It's not all that difficult: =DoCRCc ( A.A contains #nibs to scan ) > D0=C ( D0 contains start adress ) > =DoCRC > C=0 A > D1=(5) =CRC ( D1 points to hardware CRC ) > ST=0 15 ( disable interrupts ) > DAT1=C 4 ( set CRC to zero !!!! ) > C=A B > P=C 0 ( P contains lower nib of A.A ) > ASR A ( A div 16 ) > P=P-1 ( since C=DAT0 WP reads P+1 nibs.. ) > GOC + ( if P was zero, skip ) > C=DAT0 WP Them when Paul Courbis says: lire l'object dont on veut calculer l'adresse he wants to say that you have to pass it by the bus, just read the object! In this case read it from D0 to C! :) > CD0EX ( adjust D0 ) > C+P+1 > D0=C > + P= 0 > A=A-1 A ( test A.A = 0 & decrement ) > GOC ++ > - C=DAT0 W > A=A-1 A > GONC - > ++ A=0 A ( read hardware CRC ) > A=DAT1 4 > GOVLNG =AllowIntr Werner Huysegoms > remove the x before replying ---------------------------------------------------------------------- ==== Subject: drapeau 14, et interruptions Salut, La routine AllowIntr contient des trucs que je comprends pas. Elle commence par mettre ST(15) .88 un, ensuite j'ai un INTON; le GOC qui suit doit .90tre l.88 pour faire un saut plus rapide qu'un GOTO, mais la carry doit .90tre .88 z.8ero. En fait, je me demande si elle est pas mise .88 un par le mat.8eriel si une interruption est en attente (j'ai lu un truc quelque part, mais impossible de mettre la main dessus). Si la carry est .88 un, il teste le drapeau 14, qui sert je sais pas .88 quoi, sauf que je suis pas cens.8e y toucher. S'il est .8egal .88 z.8ero, .8da semble aller vite: on le met .88 z.8ero, et on fait un RTI. Sinon, on teste le drapeau 15, pour voir s'il est .88 un. Probl.8fme: en entr.8ee de routine, on le met .88 un.. (.88 moins qu'on puisse appeller .88 partir de cette instruction de test?) Je comprend pas, .88 quoi sert le flag 14? Et qui peut m'expliquer .8da: ?ST=0 14 GOYES (B) GONC (B) o.9d (B) est {ÊST=0 14 RTI } Si le flag 14 est .88 z.8ero, il fait un saut .88 (B), mais dans le cas contraire.. il fait pareil. Pourquoi ne pas avoir un GOTO l.88? Ou un simple GONC pour aller plus vite si on sait que la carry est .8egale .88 z.8ero? Le pire est apr.8fs; voici le bout complet: ?ST=0 14 GOYES (B) GONC (B) ?ST=0 14 <-- !!!! GOYES (B) <-- !!!! RSI <-- !!!! Si flag 14 .88 z.8ero, saut .88 B. Sinon, saut .88 B. Comment le reste est-t'il jamais ex.8ecut.8e? Et puis, .88 quoi .8da sert de tester si le flag 14 est a z.8ero? S'il l'est, on saute .88 (B), qui le met .88 z.8ero (mais il est d.8ej.88 .88 z.8ero!) et fait un RSI. Or, s'il est d.8ej.88 .88 z.8ero, .8da sert .88 rien de le mettre .88 z.8ero, et de faire un RSI, vu qu'on en a un juste apr.8fs ce test .8etrange. Je comprend vraiment rien .88 cette part l.88. Et quand il y a une interruption, il se passe quoi? C'est quelque chose de mat.8eriel, ou bien une sorte de photographie de tous les registres est prise, pour ex.8ecuter un truc, et ensuite revenir l.88 o.9d on .8etait? -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: drapeau 14, et interruptions The CPU flag 14 is used by the interrupt handler to know if an interrupt has occured or not. the flag 13 is used to know if the interrupt has been processed or not. The AllowIntr program will simply check if during the time where the interrupts were not authorised an interrupt has occured. If yes, you must reset the Interrupt Handler with the instruction RSI (Reset System Interrupt). All you need to know actually is that DisableIntr will disable the interrupt handler, and AllowIntr will allow the interruptions :) >Salut, La routine AllowIntr contient des trucs que je comprends pas. Elle >commence par mettre ST(15) .88 un, ensuite j'ai un INTON; le GOC qui suit >doit .90tre l.88 pour faire un saut plus rapide qu'un GOTO, mais la carry >doit .90tre .88 z.8ero. En fait, je me demande si elle est pas mise .88 un par >le mat.8eriel si une interruption est en attente (j'ai lu un truc quelque >part, mais impossible de mettre la main dessus). Si la carry est .88 un, il teste le drapeau 14, qui sert je sais pas .88 >quoi, sauf que je suis pas cens.8e y toucher. S'il est .8egal .88 z.8ero, .8da >semble aller vite: on le met .88 z.8ero, et on fait un RTI. Sinon, on teste le drapeau 15, pour voir s'il est .88 un. Probl.8fme: en >entr.8ee de routine, on le met .88 un.. (.88 moins qu'on puisse appeller .88 >partir de cette instruction de test?) Je comprend pas, .88 quoi sert le flag 14? Et qui peut m'expliquer .8da: ?ST=0 14 > GOYES (B) > GONC (B) o.9d (B) est { ST=0 14 RTI } Si le flag 14 est .88 z.8ero, il fait un saut .88 (B), mais dans le cas >contraire.. il fait pareil. Pourquoi ne pas avoir un GOTO l.88? Ou un >simple GONC pour aller plus vite si on sait que la carry est .8egale .88 >z.8ero? Le pire est apr.8fs; voici le bout complet: ?ST=0 14 > GOYES (B) > GONC (B) > ?ST=0 14 <-- !!!! > GOYES (B) <-- !!!! > RSI <-- !!!! Si flag 14 .88 z.8ero, saut .88 B. Sinon, saut .88 B. Comment le reste est-t'il jamais ex.8ecut.8e? Et puis, .88 quoi .8da sert de >tester si le flag 14 est a z.8ero? S'il l'est, on saute .88 (B), qui le met >.88 z.8ero (mais il est d.8ej.88 .88 z.8ero!) et fait un RSI. Or, s'il est d.8ej.88 .88 >z.8ero, .8da sert .88 rien de le mettre .88 z.8ero, et de faire un RSI, vu qu'on >en a un juste apr.8fs ce test .8etrange. Je comprend vraiment rien .88 cette part l.88. Et quand il y a une interruption, il se passe quoi? C'est quelque chose >de mat.8eriel, ou bien une sorte de photographie de tous les registres est >prise, pour ex.8ecuter un truc, et ensuite revenir l.88 o.9d on .8etait? -Gilb- >43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: drapeau 14, et interruptions > All you need to know actually is that DisableIntr will disable the >interrupt handler, and AllowIntr will allow the interruptions :) All I need to know is not all I want to know :) I want to know how it works. Has someone disassembled the routine at 0000C ? Where is the information by HPreg about the interrupts available? -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: drapeau 14, et interruptions All you need to know actually is that DisableIntr will disable the >interrupt handler, and AllowIntr will allow the interruptions :) All I need to know is not all I want to know :) I want to know how it works. Has someone disassembled the routine at > 0000C ? Where is the information by HPreg about the interrupts > available? Inside hpcalc.org, inside recompilations, inside one french compilation I dont remember exactely in which, they is a file called JYLOG.ZIP or something like this, and inside this zip file is the ROM disasembly part of the interruptions routine. It's for the SX but it seems to be the same on the G -Gilb- > 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ~+ . . * ^ * + Saturn: More than a simply CPU? ' | () .-.,=``=. - o - Of course, it's a planet! '=/_ | * | '=._ | `=./`, ' . '=.__.=' `=' * + + O * ' . ---------------------------------------------------------------------- ==== Subject: Re: drapeau 14, et interruptions > Inside hpcalc.org, inside recompilations, inside one french compilation > I dont remember exactely in which, they is a file called JYLOG.ZIP or > something like this, and inside this zip file is the ROM disasembly part > of the interruptions routine. It's for the SX but it seems to be the same > on the G Does someone know which? Phone is very expensive in France (between 8 F to 16 FF / minute, that's us $1.6 to $3.2 / minute) and I cannot spend time downloading compilations, since I only got a 28.8 modem. I will probably desassemble it myself, I've been looking after interrupt information for one week, and I'm tired searching, if I had worked on it, I would have finished now. But I'm too lazy... -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: drapeau 14, et interruptions I think that you mean the address #0000F I've better than that, I have the source code :) Jean-Yves > All you need to know actually is that DisableIntr will disable the >>interrupt handler, and AllowIntr will allow the interruptions :) All I need to know is not all I want to know :) I want to know how it works. Has someone disassembled the routine at >0000C ? Where is the information by HPreg about the interrupts >available? -Gilb- >43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: drapeau 14, et interruptions > I think that you mean the address #0000F I've better than that, I have the source code :) Er... may I have it? :) I've already started disassembling the whole thing, but some parts are still unknown of use to me (I have to disassembled other areas for that). -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: drapeau 14, et interruptions >I've already started disassembling the whole thing, but some parts are >still unknown of use to me (I have to disassembled other areas for >that). Hmmm, Joe Ervin disassembled and commented the interrupt system from a HP48 version E (SX). The code should be pretty much the same as I don't know if Joe still reads this newsgroup or not. But it would be worth it to see if you can find this file. dan ---------------------------------------------------------------------- ==== Subject: le SKIPOB Salut, je me suis int.8eress.8e .88 la fonction de CRC de la HP, pour la comparer .88 la mienne; je divise la longueur par 16 et ensuite je garde le reste, dans une boucle. Il y a encore des parts de leur fonction de CRC qui me sont obscures, mais je verrais cela plus tard. Juste au d.8ebut de OCRC, j'ai un appel LM .88 l'adresse #05912h qui contient en son sein un appel .88 SKIPOB; Aa re.8doit l'adresse de l'objet au niveau un de la pile, il y a un DROP, et R1a re.8doit l'adresse de l'objet. SKIPOB semble donc pr en D0 le d.8ebut de l'objet. Je pensait trouver une s.8erie de tests, qui aurait concern.8e tous les objets du type: et la fonction aurait calcul.8e l'adresse situ.8ee apr.8fs ces objets de mani.8fre rapide, et en fait je tombe sur un truc que je comprend pas trop. La routine utilise deux niveaux de RSTK pour sauver D1 et Ba; ensuite il y a le drapeau ST 1 qui est utilis.8e, je crois pour marquer le premier passage, mais je ne suis pas s.9er, dans la bouche de recherche. Ainsi, si le drapeau est .88 z.8ero (ce qui est le cas lors de l'entr.8ee dans SKIPOB) on lit le prologue de l'objet point.8e par D0, et l'avance de 5 quartet. Ensuite, D1 pointe sur Aa; donc on fait pointer D1 sur ce qui se trouve .88 l'adresse du prologue de l'objet. Si on a une cha.94ne, on a donc 02A2C en Aa, et D1 pointe donc dessus. Ensuite, j'ai ceci: LC(5) =PRLG C'est quoi? Si le contenu point.8e par le prologue est .8egal, je comprend vraiment pas .88 quoi sert ce qui suit: ?C#A A GOYES 03050 D1=D1- 5 AD1EX PC=(A) 03050 ... Si jamais j'ai C==A, alors que j'ai appell.8e SKIPOB avec un GOSBVL, je vais me retrouver avec un saut .88 (A) ? L.88 je comprends pas du tout. Ensuite, si j'ai pas A==PRLG, il semble tester d'autres valeurs, et affecter Ba en fonction. J'ai donc: LC(5) #3130 ?C#A A GOYES 03061 B=B-1 A GOC 03071 ?B#0 A GOYES 0302E C=RSTK B=C A C=RSTK D1=C RTNCC Apparament, tant que B est diff.8erent de 0, il boucle sur ce qui doit .90tre la routine de recherche de fin de l'objet. L.88, j'iai Ba et D1 qui sont restaur.8es. Mais je comprend vraiment pas comment on recherche l'objet suivant. La valeur #3130 est utilis.8ee lors des tests, elle a une particularit.8e li.8e .88 ce qui est point.8e par les prologues d'objets? Ba est mis .88 z.8ero au d.8ebut, avec P mis .88 z.8ero, juste apr.8fs avoir sauv.8e Ba et D1 dans RSTK. Donc, quand j'arrive sur: B=B-1 A La retenue passe donc .88 1 obligatoirement, non? Aucune instruction .88 part la d.8ecr.8ementation de Ba ne semble intervenir dans cette portion de code; Ca re.8doit une valeur, Aa re.8doit la lecture de D0 (qui pointe quelque part apr.8fs le prologue de l'objet) La routine que j'ai .8ecrite est beaucoup plus longue: je lis le prologue de l'objet, je teste s'il s'agit d'un objet et dans ce cas, je calcule l'adresse de fin, et je retourne juste apr.8fs l'appel (ma routine est appell.8ee par GOSBVL). Si je n'ai pas un objet du type calculable (pour ainsi dire) je fais un appel .88 SKIPOB, qui me renvoie je sais pas comment apr.8fs l'objet. Comme pas mal d'objets ont un type qui indique la taille, c'est plus rapide que le SKIPOB, mais je comprend pas comment il fonctionne, et .8da me g.90ne beaucoup. J'ai essay.8e d'utiliser le DB de Jazz, mais je n'arrive pas du tout .88 circuler dans le code. Les touches + et - ex.8ecutent les instructions, / et * le font sans arr.90t (enfin, sauf si j'appuie sur ON pour arr.90ter le flot continu) et moi ce que je voudrait, c'est pouvoir remonter ou desc dans le flot d'instructions, sans ex.8ecuter d'instructions: bref, voir sous forme d.8esassembl.8ee des routines, et ainsi de suite. J'ai plac.8e un appel .88 DB dans mon source, mais je suis oblig.8e de suivre le flot des instructions avec +, je n'arrive pas .88 me d.8eplacer dans le flot visible, les touches fl.8ech.8ees semblent modifier l'adresse .88 partir de laquelle DB d.8esassemble, .88 partir du quartet de poids faible. Quelqu'un peut me conseiller un outil qui permet de voir la m.8emoire sous forme d.8esassembl.8ee, et de s'y d.8eplacer sans r.8ealiser d'ex.8ecution de code? Ce serait le meilleur outil pour moi pour .8etudier des routines, parce qu'actuellement, je dois rechercher l'adresse de d.8epart, ensuite la poser sur la pile, dupliquer la valeur et ajouter un chiffre entre 100 ou 500, et faire un DISXY. Exemple: au sein de certaines routines, je tombe sur des adresses comme #05912h dont EC ne dispose pas, alors je dois ressortir, poser #05912h et dupliquer, ajouter #100h (ou un truc au petit bonheur la chance) et apr.8fs DISXY, essayer de voir ce qui se passe. C'est long, ennuyeux quand je n'ai pas indiqu.8e une seconde valeur assez grande, et qu'il me manque un bout. Note: je n'arrive pas .88 poster sur hp48.programming sur le serveur sam.via.ecp.fr, il me dit que je n'ai pas le droit, alors je poste ici, tant pis. -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: le SKIPOB I'll reply in English. The question was how SKIPOB works, and a few other things. First, some RPL basics (you probably know this, but apparently not all of it): . Every object starts with the address of a prologue that executes the object. Objects in RPL are linked sequentially with the 'RPL inner loop' at the end: A=DAT0 A D0=D0+ 5 PC=(A) which executes the prologue of the next object ion the runstream - IF that object was represented by a pointer. The runstream may indeed contain either a pointer to an object or the object itself. When it is represented by a pointer, the object is executed 'indirectly', and when it is included in situ, it is executed 'directly'. The RPL inner loop can be seen to work perfectly for indirect execution: runstream object prologue +-----+ +-----+ +-----+ D0 -> | | -------> | A.A | -----> | (A) | +-----+ |.....| | | | | | | | | +-----+ | | | | +-----+ +-----+ How then, is direct execution handled? When the runstream contains the object itself, a metaprologue is executed. The first five nibbles of any object prologue are always the same, and their hexadecimal representation is the address of the metaprologue: #028FC (Actually, the first five nibbles read CF820 from low to high; but when these are read into the A register using A=DAT0 A, they show up with 'C' in nibble 0, 'F' in nibble 1 etc.) runstream prologue metaprologue (#028FC) adjusting code +-----+ +-----+ | | ------------------------------> | | +-----+ | | | | +-----+ +-----+ +-----+ +-----+ D0 -> | | -----> | A.A | -----> | (A) | |.....| |.....| | | | | | | | | | | +-----+ +-----+ +-----+ +-----+ | | +-----+ After execution of the inner loop, D0 is not pointing to the next object, but simply five nibbles into the current one. Moreover, the object prologue is not executed, and A.A does not point to the object as in the indirect case. The metaprologue must therefore adjust D0 and execute the prologue, for any object. Ten nibbles before every object prologue is a pointer to code that does just that, for that object type. Thus, the metaprologue (=PRLG) simply reads: =PRLG LC(2) #A A=A-C B PC=(A) (making use of the fact that apparently the last 2 nibbles of any object prologue are larger than #0A) Similarly, the 5 nibbles immediately preceding an object's prologue contain a pointer to code that skips the object. That's why SKIPOB's inner code does not contain any 'object type test', it simply uses the object's prologue address-5 and does a PC=(A). That's the 'skip code' for that object. Take a look at some of the simpler ones (DOREAL, for instance). The rest of SKIPOB is simply: keep track of the number of levels we went down into composite objects (in B.A I believe? Haven't looked at the code). Hope I've been of help. -- Werner Huysegoms remove the x before replying http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: le SKIPOB > First, some RPL basics (you probably know this, but apparently not all > of it): The only book I found is about ML. It tells about object's, but only how about they're internally encoded, nothing else. >which executes the prologue of the next object ion the runstream - IF >that object was represented by a pointer. You mean runstream can contain code, that is primitive code objetcs, it's that? > [..] When it is represented > by a pointer, the object is executed 'indirectly', and when it is included > in situ, it is executed 'directly'. Huh? But, if an object is included in situ, it's always the prologue that is being read, no? I don't understand what you meaned there. What is direct, and what is not? > runstream object prologue +-----+ +-----+ +-----+ > D0 -> | | -------> | A.A | -----> | (A) | > +-----+ |.....| | | > | | | | | | > +-----+ | | | | > +-----+ +-----+ I don't understand this. >How then, is direct execution handled? What is direct execution? > When the runstream contains the object itself, a metaprologue is > executed. The first five nibbles of any object prologue are always the > same, and their hexadecimal representation is the address of the >metaprologue: #028FC So, if i got a program, the first << which is #02D9Dh, is jumped into, and there is a metaprologue? I have peeked 5 nibbles at #02A2Ch, the string prologue address. What I found there is: #028FC. There, I got that: LC(2) #A A=A-C B PC=(A) A does contain the current address put into PC, is that right? #02A2Ch - #Ah = 02A22h, there I got #02D04h or is it code I have to disassemble? Code there seems nonsense. > After execution of the inner loop, D0 is not pointing to the next > object, but simply five nibbles into the current one. Moreover, the > object prologue is not executed, and A.A does not point to the object > as in the indirect case. My book of ML says D0 points to the stream of execution; each address there points to an object's prologue OR primitive-code object. How coulc D0 point inside the next object? And why do you speak about D0? And D1? I don't understand a word of that: >Moreover, the object prologue is not executed, and A.A does not point >to the object as in the indirect case. > The metaprologue must therefore adjust D0 and execute the prologue, > for any object. ????????????????????????????????? > Hope I've been of help. None. I'm more lost now than previously. In my book, object's are described. All descriptions look alike, except the STR one, where it's written: Indirect execution: DOCSTR Of course, there is not even a word about what is direct or indirect execution. And, I still don't understand SKIPOB. (et une seconde couche, une...) Here it is: 03019 ST=0 1 ; used below, so we only point to Aa ; one time, using D1 CD1EX ; RSTK=C ; we save D1 in RSTK C=B A ; RSTK=C ; and Ba into RSTK P= 0 B=0 A ; B is used for something ; and what is the question :-) ?ST=1 1 GOYES 03034 ; if we already read A to D1 and made ; D0+5, we skip D0+5 and directly point ; D1 to A 0302E A=DAT0 A ; the prologue is put into Aa D0=D0+ 5 ; here we are past the prologue 03034 D1=A ; D1 to the prologue A=DAT1 A ; we read it LC(5) =PRLG ; = #028FCh ?C#A A GOYES 03050 ; 3050 = we have a valid prologue ; that is, there is #028FC at *(prologue) D1=D1- 5 ; here I don't understand: we saved D1 AD1EX ; and B onto RSTK, and we jump to the PC=(A) ; prologue, without recovering them!!? ; So, if we got a valid prologue (it contains #028FCh wheen seeked) ; we go the the search loop. Otherwise, we make PC jump to it ; but what happens of the two values saved onto RSTK? 03050 LC(5) #3130 ; this one is unknown to me, what does the ?C#A A ; #3130h value is here for? GOYES 03061 B=B-1 A GOC 03071 ; B is decremented when we seek after ; the object, but it's set to 0 before ; so here we get #FFFFFh, no? 03061 ?B#0 A ; while B != 0 we are still searching GOYES 0302E ; for the next object, and loop to 0302E ; (it's just above) C=RSTK ; if B is equal to zero, looks like we B=C A ; found the next object (when there's one) C=RSTK ; so we recover here D1 and B, and return D1=C ; and D0 should point after the object RTNCC ; (as D1 gets its previous value) If I understand some of what you said, a prologue does contain #028FC when it's a valid one, then I got 10 nibbles before it some kind of SKIP for the object. So, when D0 points to the object's prologue, I should do: C=DAT0 A D0=C C=DAT0 A and C should contain #028FCh if D0 really pointed to a prologue. Is it that? I don't understand how this routine works. What is the #3130h for? Why is B set to (what looks like) #FFFFFh and equals 0 when the next objet's found? (this looks strongly false, anyway) What is direct or indirect execution? Where should I disassemble: (prologue-10) or somewhere else? Is there a SKIP routine written for each object, that can be called 10 nibbles before every prologue's address? Is the metaprologue used to check we really have a prologue, or is it used for something else? How can D0 point inside the object, as my book tells it points to the next objet's prologue? How is the next object searched? What will happend if I point to the last object in ROM? Will D0 or D1 turn to #00000h and search an object? The HP48 is more than 8 years old, I can't believe no one hasn't looked at that routine!! So, help me! I got no info on how object's are worked on, and that would help me a lot. -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: le SKIPOB which executes the prologue of the next object ion the runstream - IF >that object was represented by a pointer. You mean runstream can contain code, that is primitive code objetcs, > it's that? > Well.. no. Take the following program: << 1 10 >> In SysRPL, it looks like: :: x<< %1 % 10 x>> ; The space before the 10 is intentional: it shows that % 10 is an embedded object, as opposed to %1, which is a pointer to an object in ROM. The size of the program is: nibs :: 5 x<< 5 %1 5 % 10 21 ( 5 prologue + 16 body ) x>> 5 ; 5 -------- 46 %1 is called 'referenced by a pointer', % 10 is 'embedded'. Upon execution, %1 is executed 'indirectly' (because its not embedded!) and % 10 is executed 'directly'. Now, the reason I said 'no' to your question is that ANY object may be embedded EXCEPT primitive code objects (PCO's). A primitive code object's prologue contains (selfaddress + 5). It is in effect, its own prologue, and can only be executed indirectly. Most stack operations are implemented as PCO's eg DUP. > [..] When it is represented > by a pointer, the object is executed 'indirectly', and when it is included > in situ, it is executed 'directly'. Huh? But, if an object is included in situ, it's always the prologue > that is being read, no? I don't understand what you meaned there. What is direct, and what is not? runstream object prologue +-----+ +-----+ +-----+ > D0 -> | | -------> | A.A | -----> | (A) | > +-----+ |.....| | | > | | | | | | > +-----+ | | | | > +-----+ +-----+ I don't understand this. Well, look at the inner loop again: A=DAT0 A D0=D0+ 5 PC=(A) D0 points at the next 'slot' in the runstream. That slot may contain either a POINTER to the object (as with %1) or the object itself (% 10). The above drawing illustrates the case of indirect execution, ie the slot contains a pointer to the object. A=DAT0 A (now A.A points at the first 5 nibbles of the object, which contain the address of the prologue) D0=D0+ 5 (D0 points to the next 'slot') PC=(A) (execution continues with the prologue code) Now, the case of direct execution: runstream prologue metaprologue (#028FC) adjusting code +-----+ +-----+ | | ------------------------------> | | +-----+ | | | | +-----+ +-----+ +-----+ +-----+ D0 -> | | -----> | A.A | -----> | (A) | |.....| |.....| | | | | | | | | | | +-----+ +-----+ +-----+ +-----+ | | +-----+ This time, DO already points at the object. So, if the inner loop is executed: A=DAT0 A (A.A points at the prologue itself) D0=D0+ 5 (advances D0 5 nibbles. Will have to be corrected later on, to point past the embedded object, because that's where the next runstream 'slot' resides) PC=(A) (continues at the address pointed to by the first 5 nibbles of the prologue) Execution continues at the address pointed to by the first 5 nibbles of the to high. As code, it is: D=D-1 A HS=0 0 Which decrements available memory, and a 3-nibble NOP. Very useful for data objects who, when executed, are simply pushed on the stack and thus consume 5 nibbles - or 1 unit of D.A. As an address, it represents #028FC which is the address of the metaprologue. This metaprologue is what gets executed in the direct case. As you can see, it has to: . adjust D0 to point beyond the object io 5 nibs into it . make A.A point to the object as in the indirect case . execute the object's prologue =PRLG LC(2) #A A=A-C B PC=(A) Instead of testing which object type is being executed, it simply uses the address 10 nibbles before the object's prologue. That address points to code that handles the current object type's direct execution. Let's take a look at an object's prologue and direct execution =DOREAL D=D-1 A * address #02933 HS=0 0 GOC L0298F D1=D1- 5 * write to next stack position DAT1=A A A=DAT0 A * inner loop D0=D0+ 5 PC=(A) ... L0298F GOTO L02D52 * handles garbage collection This piece of ML is what gets executed for %1 in the above program. It is quite straightforward. Now at address 02929 (10 nibbles before 02933), we find the address 02C49, and there we find: #02C49 D0=D0- 5 AD0EX * A=D0 D0=A D0=D0+ 5 * skip the real D0=D0+ 16 D=D-1 A * object prologue repeated GOC L02C91 D1=D1- 5 DAT1=A A A=DAT0 A D0=D0+ 5 PC=(A) Again, easy to follow.. A.A is changed to point to the object itself, which is address D0-5. Then the object is skipped, which is 21 nibbles long, and finally, the object prologue is repeated. > (et une seconde couche, une...) Here it is: 03019 ST=0 1 ; used below, so we only point to Aa > ; one time, using D1 > CD1EX ; > RSTK=C ; we save D1 in RSTK > C=B A ; > RSTK=C ; and Ba into RSTK P= 0 > B=0 A ; B is used for something > ; and what is the question :-) As I told you in my previous post, Ba is used to keep track of the 'level'. Skipping is recursive: if you need to skip a composite object like a list or a secondary, it is done by skipping all of its composing objects, which may themselves be composite etc. When a composite is encountered, Ba is augmented, and when SEMI is skipped, Ba is decremented. ?ST=1 1 > GOYES 03034 ; if we already read A to D1 and made > ; D0+5, we skip D0+5 and directly point > ; D1 to A 0302E A=DAT0 A ; the prologue is put into Aa > D0=D0+ 5 ; here we are past the prologue 03034 D1=A ; D1 to the prologue > A=DAT1 A ; we read it > LC(5) =PRLG ; = #028FCh > ?C#A A > GOYES 03050 ; 3050 = we have a valid prologue > ; that is, there is #028FC at *(prologue) D1=D1- 5 ; here I don't understand: we saved D1 > AD1EX ; and B onto RSTK, and we jump to the > PC=(A) ; prologue, without recovering them!!? > No, you jump to the address pointed to by the 5 nibbles preceding the prologue, which is what contains the 'skip' code for that object type. > ; So, if we got a valid prologue (it contains #028FCh wheen seeked) > ; we go the the search loop. Otherwise, we make PC jump to it > ; but what happens of the two values saved onto RSTK? 03050 LC(5) #3130 ; this one is unknown to me, what does the > ?C#A A ; #3130h value is here for? > GOYES 03061 > #03130 is the first five nibbles of =SEMI, which is a PCO. So this code tests whether you encountered SEMI, and decrements Ba: > B=B-1 A > GOC 03071 ; B is decremented when we seek after > ; the object, but it's set to 0 before > ; so here we get #FFFFFh, no? 03061 ?B#0 A ; while B != 0 we are still searching > GOYES 0302E ; for the next object, and loop to 0302E > ; (it's just above) C=RSTK ; if B is equal to zero, looks like we > B=C A ; found the next object (when there's one) > C=RSTK ; so we recover here D1 and B, and return > D1=C ; and D0 should point after the object > RTNCC ; (as D1 gets its previous value) If I understand some of what you said, a prologue does contain #028FC > when it's a valid one, then I got 10 nibbles before it some kind of SKIP > for the object. > 5 nibbles before the onject is the skipob code address, 10 nibbles before is the direct execution code's address. I may have confused the two in my previous post. > So, when D0 points to the object's prologue, I should do: C=DAT0 A > D0=C > C=DAT0 A and C should contain #028FCh if D0 really pointed to a prologue. Is it that? > Yes. > I don't understand how this routine works. What is the #3130h for? Why > is B set to (what looks like) #FFFFFh and equals 0 when the next objet's > found? (this looks strongly false, anyway) What is direct or indirect > execution? Where should I disassemble: (prologue-10) or somewhere else? > Is there a SKIP routine written for each object, that can be called 10 > nibbles before every prologue's address? Is the metaprologue used to > check we really have a prologue, or is it used for something else? How > can D0 point inside the object, as my book tells it points to the next > objet's prologue? How is the next object searched? What will happend if > I point to the last object in ROM? Will D0 or D1 turn to #00000h and > search an object? > I believe I have answered about all your questions. If you have any more, read RPLMAN first, as Jean-Yves suggested. BTW I agree fully with him on the sheer beauty of the concept. I hope it will be preserved in 'the next generation'. -- Werner Huysegoms remove the x before replying http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: le SKIPOB Have a look to the RPLMANUAL The concept is described there. the Prologue object is probably the more intellignet thing ever created on the HP calculators. The concept is extremely clever and creative. A prologue can be executed in ML, in sysRPL or referenced indirectly. All prologue starts with an address #028FC (which is used to recognize if it's a prologue or not), but this address is also a program ! To summarize: If you execute a prologue by doing PC=(A) on it it will: -Push the object on the stack -Run it if it's a program. The 5 nibbles before contains the code to execute the prologue when called directly in systemRPL. Example: When you insert an object in a program, you can put its address or the object itself. The result is the same due to the prologue structure. The 10 nibbles before the prologue contains the program to calculate the size of the program. (used by SKIPOB). SKIPOB in order to calculate the size of an object, simply run the object which will calculate its size itself. Everytime I look to this part of code, I'm impressed. Jean-Yves > First, some RPL basics (you probably know this, but apparently not all >> of it): The only book I found is about ML. It tells about object's, but only how >about they're internally encoded, nothing else. >which executes the prologue of the next object ion the runstream - IF >>that object was represented by a pointer. You mean runstream can contain code, that is primitive code objetcs, >it's that? > [..] When it is represented >> by a pointer, the object is executed 'indirectly', and when it is included >> in situ, it is executed 'directly'. Huh? But, if an object is included in situ, it's always the prologue >that is being read, no? I don't understand what you meaned there. What is direct, and what is not? > runstream object prologue >> +-----+ +-----+ +-----+ >> D0 -> | | -------> | A.A | -----> | (A) | >> +-----+ |.....| | | >> | | | | | | >> +-----+ | | | | >> +-----+ +-----+ I don't understand this. >How then, is direct execution handled? What is direct execution? > When the runstream contains the object itself, a metaprologue is >> executed. The first five nibbles of any object prologue are always the >> same, and their hexadecimal representation is the address of the >>metaprologue: #028FC So, if i got a program, the first << which is #02D9Dh, is jumped into, >and there is a metaprologue? I have peeked 5 nibbles at #02A2Ch, the string prologue address. What I >found there is: #028FC. There, I got that: LC(2) #A >A=A-C B >PC=(A) A does contain the current address put into PC, is that right? #02A2Ch - #Ah = 02A22h, there I got #02D04h >or is it code I have to disassemble? Code there seems nonsense. > After execution of the inner loop, D0 is not pointing to the next >> object, but simply five nibbles into the current one. Moreover, the >> object prologue is not executed, and A.A does not point to the object >> as in the indirect case. My book of ML says D0 points to the stream of execution; each address >there points to an object's prologue OR primitive-code object. How coulc D0 point inside the next object? And why do you speak about D0? And D1? I don't understand a word of that: >Moreover, the object prologue is not executed, and A.A does not point >>to the object as in the indirect case. > The metaprologue must therefore adjust D0 and execute the prologue, >> for any object. ????????????????????????????????? > Hope I've been of help. None. I'm more lost now than previously. In my book, object's are described. All descriptions look alike, except >the STR one, where it's written: Indirect execution: DOCSTR Of course, there is not even a word about what is direct or indirect >execution. And, I still don't understand SKIPOB. (et une seconde couche, une...) Here it is: 03019 ST=0 1 ; used below, so we only point to Aa > ; one time, using D1 > CD1EX ; > RSTK=C ; we save D1 in RSTK > C=B A ; > RSTK=C ; and Ba into RSTK P= 0 > B=0 A ; B is used for something > ; and what is the question :-) ?ST=1 1 > GOYES 03034 ; if we already read A to D1 and made > ; D0+5, we skip D0+5 and directly point > ; D1 to A 0302E A=DAT0 A ; the prologue is put into Aa > D0=D0+ 5 ; here we are past the prologue 03034 D1=A ; D1 to the prologue > A=DAT1 A ; we read it > LC(5) =PRLG ; = #028FCh > ?C#A A > GOYES 03050 ; 3050 = we have a valid prologue > ; that is, there is #028FC at *(prologue) D1=D1- 5 ; here I don't understand: we saved D1 > AD1EX ; and B onto RSTK, and we jump to the > PC=(A) ; prologue, without recovering them!!? ; So, if we got a valid prologue (it contains #028FCh wheen seeked) >; we go the the search loop. Otherwise, we make PC jump to it >; but what happens of the two values saved onto RSTK? 03050 LC(5) #3130 ; this one is unknown to me, what does the > ?C#A A ; #3130h value is here for? > GOYES 03061 B=B-1 A > GOC 03071 ; B is decremented when we seek after > ; the object, but it's set to 0 before > ; so here we get #FFFFFh, no? 03061 ?B#0 A ; while B != 0 we are still searching > GOYES 0302E ; for the next object, and loop to 0302E > ; (it's just above) C=RSTK ; if B is equal to zero, looks like we > B=C A ; found the next object (when there's one) > C=RSTK ; so we recover here D1 and B, and return > D1=C ; and D0 should point after the object > RTNCC ; (as D1 gets its previous value) If I understand some of what you said, a prologue does contain #028FC >when it's a valid one, then I got 10 nibbles before it some kind of SKIP >for the object. So, when D0 points to the object's prologue, I should do: C=DAT0 A > D0=C > C=DAT0 A and C should contain #028FCh if D0 really pointed to a prologue. Is it that? I don't understand how this routine works. What is the #3130h for? Why >is B set to (what looks like) #FFFFFh and equals 0 when the next objet's >found? (this looks strongly false, anyway) What is direct or indirect >execution? Where should I disassemble: (prologue-10) or somewhere else? >Is there a SKIP routine written for each object, that can be called 10 >nibbles before every prologue's address? Is the metaprologue used to >check we really have a prologue, or is it used for something else? How >can D0 point inside the object, as my book tells it points to the next >objet's prologue? How is the next object searched? What will happend if >I point to the last object in ROM? Will D0 or D1 turn to #00000h and >search an object? The HP48 is more than 8 years old, I can't believe no one hasn't looked >at that routine!! So, help me! I got no info on how object's are worked >on, and that would help me a lot. -Gilb- >43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: emulators Can someone recommend an emulator that will run in either DOS, Win 95 or Win 3.1 (and include its ftp address)? -- ---------------------------------------------------------------------- ==== Subject: Re: emulators Hi Mark try this: http://hp48.ml.org/pc/emulators/emu48bin.zip Eva > Can someone recommend an emulator that will run in either DOS, Win 95 or > Win 3.1 (and include its ftp address)? > -- ---------------------------------------------------------------------- ==== Subject: Questions about European schools Balazs & Gerald *** ARE MY ANSWERS E> I don't think the Swedish school tempo is anything to be proud of. E> Sweden being a half commie country has a very selective school E> system. E> At age 14 you need to apply into gymnasium (high school). If you E> mess up on you grads at age 7-14 you can not get into a engineering E> high school. Engineering is not for everyone. In Germany you also have to choose your type of school after the first four years of elementary school. And still about half the people who start a study as an engineer fail. On my first day in the university the instructor said Look at the person sitting next to you. He will probably be gone before the Vordiplom (which is 2 years after the start). E> There are other types of high schools, but good luck getting a job. The unemployment rate is not very high in Sweden (at least compared to other European countries). And only 27% of the GDP are from the industry compared to 71% from services. E> Until a few years ago a student was a full blow electrical E> engineer when she finished high school. And why did anyone have to study for a EE diploma? ***Actually you became a ãgymnasie engineerä More or less a BS degree. I am not really sure what you mean by an EE. If it means an Electrical engineer then that's what you were after technical gymnasium. The way things are now you get a masters degree (4 years) right after gymnasium. Then you become a doctor. E> Because of tax reasons all E> the high school programs were cut down to from 4 years to 3 years. E> Read in MSNBC (http://www.msnbc.com/news/220448.asp ***Sweden yes, tax no. You would need a 20 GB HD to explain Swedish tax laws in your PC. The Third International Mathematics and Science Study ranked American 12th-graders, aged 17 and 18, 18th out of 21 countries ÷ far behind Sweden and the Netherlands and ahead only of Lithuania, Cyprus and South Africa. E> Everything runs like clock work in Sweden. All high schools E> programs in the Sweden have the exact same books and tempo. This is actually good because you don't have a problem if you move to another city. *** I totally agree with you on that point. I know that anybody who has gone through the math program A.B.C.D.E. (The math books) has the exact same math skill· knowledge as I do. As a compliment to passing the math test E level you also have to do a research project in mathematics. I chose to expand on the Swedish mathematics Lars H.9armander (won the Fields prize 1962) theories in differential equations. I got on a Swedish NG asked who has read E math. A few people had. No wasted time having to find out who knew what. Within 5 minuets I had found a person who wanted (and could) to do joint research with me. E> This cut back from 4 years to 3 disrupted the clock work and lead the E> universities to complain. The result was that the gymnasiums had to E> make up for the lack of knowledge by increasing the tempo. E> I am studying the technical gymnasium line. Last week a guy commit E> suicide in my class because he could not handle the presser. Most of the times there are many problems together and not only one that lead to a suicide. ***Per was a dear friend to me. He was what some people would call beyond genius. Sometimes unfortunately individuals like him have problems cooping with daily life. He had broken his leg the 2 days before a math test. Our professor has a simple rule: If you don't take the test you fail. Per had never failed a thing in his life. The rest I don't Understand. Of course a bced person does not kill themselves. I saw a posting on this NG That I should be happy with the Swedish school system. I guess I was shocked and blamed my math teacher for his sternness and needed to grovel. E> I have so many exams to study for E> that I somtimes don't have time to go to school. ??? How does that work??? You don't go to school because you learn? ***Well it sounds ironic but its true. Lets say you have to learn 300 pages of stuff you have never heard of and a research (mine was the one I mentioned) project that must be turned in after 7 days. There are only 24 hours in a day and school hogs up a few to many. You sit on your butt until it black and blue at a library. When they E> At this time of year the sun goes down at 2:00 PM. Now this is really depressing. ***Hmmm I think you just solved why the Swedish students study like they do. Who would want to do anything with the beautiful beaches in California outside Your window. I can see myself now with a wine cooler on the beach there·. oooh. I can't say anything about this since I don't know how you compare to the average student. But the only time I learned that intensive was a few weeks before important exams. ***Its on and off like I described and only during the last year. ALL universities are free (tax money) so it is very important to ãweedä out the week. A student who studies 2 years at uni. and then drops out costs to much money. That's why they ãtortureä you in gymnasium. E> The only thing that means ANYTHING is passing the next exam with good E> grades. E> If you want to go to a technical E> university then you must complete the technical gymnasium first. And that's only natural. E> Of course with the best grades in the class !. Every one has to be the E> best in the class! If that were true then the technical universities would be really really small. ***There are many different kinds of technical lines 8 of them give you a masters degree after 4 years. There is nothing between gymnasium and a masters. Each line takes in about 80 students a year. That comes to about 2560 technical students per university. The universities are split into ãhousesä with lot of other majors then just engineers. E> Remember if you did not have the grades as a 14 year old you can not E> get into technical gymnasium. Is there no way at all to switch schools? In Germany you can change from a lower to a higher school (and vice versa) if your grades are good (bad). ***Welcome to ãcommieä land. You study were you live. All the programs are The same. There is no reason to change. E> Young people who would never touch cigarettes are using E> methamphimines to keep there minds alert in school. Something is E> wrong here !!! Math is vary important, as it is the crutch to all E> other subjects. In my physics class we are working heavily with E> quantum physics. Today it was detailed analyzes of Schr.9adings wave E> equation. Does a 17 year need to kill themselves to learn know this E> stuff ? A 17 year old doesn't have to kill himself period. ***So true! E> All lot of other countries think that Swedish people E> are dry boring people.... no wonder, there is nothing left of them E> when they get out of school ! Actually I have never heard that Swedish people are boring. The only prejudice I have heard about your people is that they are very open about sex (a running gag in MWC was the Swedish Bikini Team, in the 70s Swedish porn was famous). ***I have always seen myself as a self respecting girl. That was until a patrol man at June lake Calif. fined my mother and I for sun tanning topless high up in the Sierra mountains when we were on vacation there !? E> Swedish people are quick to ridicule to american school system. E> Hmmmmm when was the last time Sweden put a man on the moon. With the help of german engineers :-))). Besides it would be very hard for Sweden to start a rocket to the moon, because they don't have a good launching site and AFAIK Sweden doesn't have any colonies. And it was a prestige project in a war. Since Sweden is smart enough to keep out of wars there was no need for a prestige project like this. ***If no one fought back during W.W.II you would probably be living next door. I don't think it was right to sell Swedish steel to the Nazi war machine and Then letting them send 10âs of thousands of killers through my home town to kill Norwegians and cart Jews back to extermination camps by Swedish rail. All that to avoid war. I would have had a better conscience if we had fought back. We are all global citizens. Its the amount of lives at the end of the war that counts not how many of your own lands men you save. To think otherwise is quite primitive. E> Who kicks Sadams butt with ultra high tech toys when he is a bad boy. Sweden Population: 8,865,051 (July 1997 est.) GDP: purchasing power parity - $184.3 ion (1996 est.) Military expenditures - dollar figure: $5.8 ion (FY94/95) Military expenditures - percent of GDP: 2.5% (FY94/95) United States Population: 267,954,764 (July 1997 est.) GDP: purchasing power parity - $7.61 trillion (1996 est.) Military expenditures - dollar figure: $267.2 ion (1997 est.) Military expenditures - percent of GDP: 3.4% (1997 est.) Cost of a Aircraft carrier: $1,000,000,000 (in 70s dollars I think). ***How many schools would that help out ? Well somebody's got to do it. We have not progressed that far On the evolutionary ladder. But as long as the US is the world police The cost of its war machine should be split up. Any questions why Sweden doesn't kick Saddam's ass? Besides Sweden is smart enough to stay out of wars and doesn't risk the youth of the nation. Just you look in the CIA worldbook of facts for international disputes and see how many Sweden has. ***read the above E> Who's hospitals have the coolest toys etc. etc. etc. etc. The life expectancy in Sweden is more than 3 years higher than in the US. ***Everybody is given retirement money when they turn 60 even if they never lifted a finger (Katz is turning green) and a lot of people never do lift a finger. That tends to keeps people ticking a little longer. E> All that in only 200 years! 1776-1998 not bad! Industrialization has started later in all countries. ***If I remember right there was a civil war 1861-1865 in the US And it really set back the south were almost all industry was, for quite some time. Britain was 1st. Britain did not long remain the only country to experience an Industrial Revolution. Attempts to specify dates for the Industrial Revolution in other countries are controversial and not particularly rewarding. Nonetheless, scholars generally agree that the Industrial Revolution occurred in France, Belgium, Germany, and the United States about the middle of the 19th century; in Sweden and Japan toward the end of the century. E> Naaaa we don't do much here in Sweden except ANALYZE other peoples E> accomplishments and give THEM Nobel prizes. In the fields I know about Sweden is at the frontiers of technology. With their fighter planes J35 Draken, Ja37 Viggen and JAS39 Gripen they have surprised eastern and western analysts and have shown that even a small country can build top fighter planes. The Jas39 Gripen was the first 4th generation fighter to fly, long before the Rafale or the Eurofighter. ***Swedes tend to be humble. E> May the European school corridors rock with Rap music and the E> smell of spray paint and pot. E> Why not ? It worked for America. America catches up with other countries in the education because they have some of the best universities of the world. And the technological, political and social leaders of the US generally are not from schools that rock with Rap music and smell of spray paint and pot. But many of the people that have low paying jobs are. ***Now who was it that said ãI SMOKED POT BUT DIDN'T INHALE ä If I am not mistaken it was good old C. Don't get me wrong I am not a drug liberal. PS If John M. is reading this I be getting back to you soon. I need to scan in some more pics. for you but I got TIME now. Tests are over! Eva hellqvisteva@hotmail.com ----------- The only absolute is that noting is absolute. We perceive the world as we perceive ourselves because we have not been introduced to other states of randomness. Eva ---------------------------------------------------------------------- ==== Subject: Questions about European schools Balazs & Gerald E> I have so many exams to study for E> that I somtimes don't have time to go to school. BF>> ??? How does that work??? You don't go to school because you learn? E> ***Well it sounds ironic but its true. Lets say you have to learn 300 E> pages E> of stuff you have never heard of and a research (mine was E> the one I mentioned) E> project that must be turned in after 7 days. There are only E> 24 hours in a day and school hogs E> up a few to many. You sit on your butt until it black and E> blue at a library. When they E> close you go home and do your research project until about 4: So this is not the usual situation but the extreme. E> At this time of year the sun goes down at 2:00 PM. BF>> Now this is really depressing. E> ***Hmmm I think you just solved why the Swedish students study like they E> do. E> Who would want to do anything with the beautiful beaches in E> California outside E> Your window. I can see myself now with a wine cooler on the E> beach there·. oooh. Oh, yeah. May I join in? All this cold and snow here in Germany is depressing me. E> ***Its on and off like I described and only during the last year. ALL E> universities E> are free (tax money) so it is very important to ãweedä out E> the week. E> A student who studies 2 years at uni. and then drops out E> costs to much money. E> That's why they ãtortureä you in gymnasium. I Germany the university is free too, but they only weed out there. As I already said,, about half don't complete the first two years. BF>> Actually I have never heard that Swedish people are boring. The BF>> only prejudice I have heard about your people is that they are very BF>> open about sex (a running gag in MWC was the Swedish Bikini Team, BF>> in the 70s Swedish porn was famous). E> ***I have always seen myself as a self respecting girl. That was E> until a patrol man at June lake Calif. fined my mother and I E> for sun tanning topless high up in the Sierra mountains when E> we were on vacation there !? :-))) Naughty girl. But the law in the US is a little bit strange regarding public nudity. What is absolutely normal in large parts of Europe is forbidden in most Amerikan states. BF>> colonies. And it was a prestige project in a war. Since Sweden is BF>> smart enough to keep out of wars there was no need for a prestige BF>> project like this. E> ***If no one fought back during W.W.II you would probably be living next E> door. E> I don't think it was right to sell Swedish steel to the Nazi E> war machine and E> Then letting them send 10âs of thousands of killers E> through my home town to E> kill Norwegians and cart Jews back to extermination camps E> by Swedish rail. E> All that to avoid war. I would have had a better conscience E> if we had fought back. E> We are all global citizens. Its the amount of lives at the E> end of the war that counts not how many of your own lands E> men you save. E> To think otherwise is quite primitive. WWII was quite unique because it was one of the few wars with a distinct good and evil side. There are not many wars with a clear distinction like this. And the Swedish army wouldn't have made a difference (at least not more than a week). BF>> Cost of a Aircraft carrier: $1,000,000,000 (in 70s dollars I think). E> ***How many schools would that help out ? E> Well somebody's got to do it. We have not progressed that far E> On the evolutionary ladder. But as long as the US is the world E> police E> The cost of its war machine should be split up. Only if the benefits of the war machine are split up too. The Americans fight for their own reasons just like the French, the British and the Russians. E> Who's hospitals have the coolest toys etc. etc. etc. etc. BF>> The life expectancy in Sweden is more than 3 years higher than in BF>> the US. E> ***Everybody is given retirement money when they turn 60 even if they E> never lifted a finger (Katz is turning green) and a lot of E> people never E> do lift a finger. That tends to keeps people ticking a little E> longer. :-) That's a way to replace the cool toys in hospitals (but Sweden has them too). E> All that in only 200 years! 1776-1998 not bad! BF>> Industrialization has started later in all countries. E> ***If I remember right there was a civil war 1861-1865 in the US That's correct. E> And it really set back the south were almost all industry E> was, for quite some time. That's wrong. The industry was in the North. Pennsylva alone produced more than ten times the amount of iron the Confederates produced together. Because of that and because they had a higher manpower the North won. Although both sides payed a high price for the war (about 620,000 dead more than the combined American death rate of WWI, WWII and the Korean war together) it really helped the northern industry and it made America the biggest military power in the world. It also was the first really industrialized total war. E> Naaaa we don't do much here in Sweden except ANALYZE other E> peoples accomplishments and give THEM Nobel prizes. BF>> In the fields I know about Sweden is at the frontiers of BF>> technology. With their fighter planes J35 Draken, Ja37 Viggen and BF>> JAS39 Gripen they have surprised eastern and western analysts and BF>> have shown that even a small country can build top fighter planes. BF>> The Jas39 Gripen was the first 4th BF>> generation fighter to fly, long before the Rafale or the BF>> Eurofighter. E> ***Swedes tend to be humble. And do more than analyze other peoples accomplishments and give them Nobel prizes. E> May the European school corridors rock with Rap music and the E> smell of spray paint and pot. E> Why not ? It worked for America. BF>> America catches up with other countries in the education because BF>> they have some of the best universities of the world. And the BF>> technological, political and social leaders of the US generally are BF>> not from schools that rock with Rap music and smell of spray paint BF>> and pot. But many of the people that have low paying jobs are. E> ***Now who was it that said ãI SMOKED POT BUT DIDN'T INHALE ä E> If I am not mistaken it was good old C. He didn't do it in school but in college. E> Don't get me wrong I am not a drug liberal. It's proven that cigarettes are worse than pot for your health. And they are highly addictive. E> PS If John M. is reading this E> I be getting back to you soon. E> I need to scan in some more pics. for you E> but I got TIME now. Tests are over! So no more learning from dusk till dawn? cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... A sucking chest wound is just nature's way of telling you to slow down. ---------------------------------------------------------------------- ==== Subject: Re: Questions about European schools Balazs & Gerald > ***Everybody is given retirement money when they turn 60 even if they > never lifted a finger (Katz is turning green) and a lot of ^^^^^^^^^^^^^^^^^^^^^ I was... until I read this line & then I turned bright red laughing! How well you know me, Eva, how well! :) Ach, the American Social (in)Security system is much the same way, I won't bore everyone with the gory details... > ***If I remember right there was a civil war 1861-1865 in the US > And it really set back the south were almost all industry was, for > quite some time. Prior to & during the Civil War, the North (Union) was where nearly all American industry was located; the South (Confederate) was very rural, mostly cash crops like cotton, sugar, rice, etc. were the basis for the Southern economy & exports to Europe for war materiel. Good luck on your studies, katz -- +----------------------------------------------------------------+ | A pig will walk in the wilderness, a pig will walk on the sea. | | A pig will walk wherever he wants, but no pig walks on me. | +----------------------------------------------------------------+ ---------------------------------------------------------------------- ==== Subject: Re: Questions about European schools Balazs & Gerald > ***Everybody is given retirement money when they turn 60 even if they > never lifted a finger (Katz is turning green) and a lot of ^^^^^^^^^^^^^^^^^^^^^ I was... until I read this line & then I turned bright red laughing! How well you know me, Eva, how well! :) Ach, the American Social (in)Security system is much the same way, I won't bore everyone with the gory details... > ***If I remember right there was a civil war 1861-1865 in the US > And it really set back the south were almost all industry was, for > quite some time. Prior to & during the Civil War, the North (Union) was where nearly all American industry was located; the South (Confederate) was very rural, mostly cash crops like cotton, sugar, rice, etc. were the basis for the Southern economy & exports to Europe for war materiel. Good luck on your studies, katz -- +----------------------------------------------------------------+ | A pig will walk in the wilderness, a pig will walk on the sea. | | A pig will walk wherever he wants, but no pig walks on me. | +----------------------------------------------------------------+ ---------------------------------------------------------------------- ==== Subject: HP Calculator Archive Newsletter (December 9, 1998) HP Calculator Archive Newsletter - http://www.hpcalc.org This is my weekly newsletter for comp.sys.hp48 about the HP Calculator Archive. Included in the HP Calculator Archive are HP48, HP38G, and HP28 programs, the HP Calculator ICQ List, documentation, and much more. The HP Calculator Archive, which I think is the best place to get software for and information about the HP48, HP38G, and HP28, is at . I am trying to collect every recent HP calculator program ever made to put on the archive. Currently I have over 3000 files totaling over 100MB. Since the file archives are mostly complete, I would like to expand the documentation section. Submissions of columns and book reviews would be nice; see and for examples. The most recent addition to the HP Calculator Archive is the HPedia: the HP Calculator Encyclopedia. It contains dozens of commonly used terms and abbreviations as well as important people and software. The HPedia was originally Carlos Marangon's idea. He proposed it to me along with some pictures. Khalil Haddad translated the HPedia into French, and a Portuguese version is in progress by Carlos Marangon. You can view the first edition of the HPedia at but new terms and corrections are welcome. If you have ever created any HP48/38/28 software, please check to see if I have the latest version(s) on my site. If not, please email any updated programs to and I will update my site. The HP Calculator ICQ users list is designed to unite the HP48/38/28 Internet community. For those of you who don't know about ICQ, ICQ is the Internet's largest communication network. For more information on ICQ, visit . Also, I am using this message as a shameless plug for my HP Calculator Archive CD. If you would like to have the contents of the HP Calculator Archive on a CD-R disc, please visit . What's New on the HP Calculator Archive? December 9th, 1998: Updated HP Money 98 on Apps/Misc Added Memory Reservation to Docs/Programming Added Emul48 2.0 to PC/Emulators Added D->LIB to Programming/Libs Added apndvarlst Replacement to Programming/Misc Added MGRID to Programming/Misc Added BZ Series to Utils/Compress December 8th, 1998: Updated Tetri on Games/Arcade Updated HP48 to HPComm 2.0 on PC/Link Updated RPN on Programming/Misc Updated Operating Tools on Utils/Misc Added Uplifter to PC/Emulators Added Programming in Extended User RPL to Programming/Misc Added Copy to Utils/Memory December 7th, 1998: Added HP Money 98 to Apps/Misc Added Section to Science/Physics December 6th, 1998: Updated Emu48 Service Pack to 8 on PC/Emulators December 5th, 1998: Updated BAW to 2.1 on Math/Statistic Updated STAT48 to 1.3 on Math/Statistic Updated Exhackt to 1.9c on Math/Symbolic December 4th, 1998: Updated ZeldaHP to 0.32 on Games/Adventure December 3rd, 1998: Updated Java and Source to 3.6 on Utils/Stack What's New on the HP Calc+ICQ Users List? Added the following users: Brent Torrenga: 723926 Jerome Tan: 1850588 Carlos Alberto Bez Birolo: 20712649 Ernesto Guevara: 23256078 Wenderson Teixeira: 1597462 Rene Felder: 19721679 Bruno Quesnel: 3434464 The Complete HP Calc+ICQ Users List: 210 Members The complete, most up-to-date version of the HP Calc+ICQ Users List is at . If you would like to be added to the HP Calc+ICQ Users List, send your REAL first and last name, ICQ nickname, ICQ UIN, and authorization status to . You can also page me at ICQ # 783944. Eric Rechlin Bismarck, ND, USA eric@ hpcalc.org http://www.hpcalc.org/ (spam protection enabled) ---------------------------------------------------------------------- ==== Subject: Re: Text editors My thoughts on your query (post and email): In the simplest form, you could simply type a file in an ASCII text editor of the following form: %%HP: T(3); put whatever text you like here After transferring to the HP48 with HPExplorer or Hyperterminal etc, this would appear as a variable on the HP48, and could be viewed simply by recalling the contents to the stack and pushing the down arrow, or LS-[+/-]. No extra software would be required, but it would be very slow for a text file of any reasonable size, and the default font size is a bit restrictive. To easily view and edit a larger text file, purpose built software is required, and these are almost exclusively libraries. The program Scribe is intended as a datafile manager/viewer for the HP48, and is easy to use (documentation is comprehensive, which helps). It uses a bit less than 15kb, which may be restrictive if you have a G model. A benefit of using Scribe is the availability of a compatible editor for PC's (Win95), ScribePro, which is available from http://home.utah-inter.net/clalor/ Scribe itself is available from http://www.hpcalc.org/ Hope this helps, Jeff Roulston -----Original Message----- ==== Subject: Text editors >I know this question has been asked before but I haven't seen an answer >that it is simple yet: I want to be able to build a text file on my >computer and then upload it to my hp48 for viewing. I have my hp48 >connected to my computer and and have succesfully used Hpexplorer but I >can't find a program that doesn't have jargon filled documentation. Once >I get one of these files to my hp what do i do with it and what should >the associated text file on my computer look like? Are they all >libraries and thus all be installed like libraries or do some just run >as is? Am I completely missing something here? >-Confused in Oregon > ---------------------------------------------------------------------- ==== Subject: CompSqr 1.0 Hello All! CompSqr is a program that completes the square of a quadratic equation. It is written in 100% SYS RPL and is 361 bytes in size. If you want to give it a try, it will be @ http://hp48.ml.org/math/symbolic within the next day or so I hope. Jeremy E. Laughery ---------------------------------------------------------------------- ==== Subject: Java 3.6 I installed Java 3.6 four times, and the four times it failed. I start working on it normally, but after a while, it says if i want to recover the memory. I don«t know if I do something wrong, or there is a bug. Maybe this happened to somebody else... -=UssU=- ---------------------------------------------------------------------- ==== Subject: Re: Java 3.6 Give a try to the MetaKernel. I didn't have any bug report for the last year... http://www-miaif.lip6.fr/gerald/english/mk.html Jean-Yves >I installed Java 3.6 four times, and the four times it failed. I start >working on it normally, but after a while, it says if i want to recover the >memory. I don«t know if I do something wrong, or there is a bug. Maybe this >happened to somebody else... >-=UssU=- ---------------------------------------------------------------------- ==== Subject: Re: Java 3.6 hehe A shameless plug! Brian >Give a try to the MetaKernel. I didn't have any bug report for the last year... http://www-miaif.lip6.fr/gerald/english/mk.html Jean-Yves >I installed Java 3.6 four times, and the four times it failed. I start >>working on it normally, but after a while, it says if i want to recover the >>memory. I don«t know if I do something wrong, or there is a bug. Maybe this >>happened to somebody else... >>-=UssU=- > > ---------------------------------------------------------------------- ==== Subject: Symbolic capabilities of HP48 and TI89 Many postings here assume that the HP48+Erable+Alg48 is comparable to the TI89. I am afraid that this is simply not true. In fact: 1) A 48GX equipped with the memory expansion necessary to run these programs costs much more than a TI89; 2) Erable is certainly a good program, especially if compared to the built-in symbolic capabilities of the HP; however it is still VERY UNSTABLE (I have had many crashes in a few days), and, more important, it sometimes gives a wrong answer, even to simple problems, without any warning. For instance I have found many simple limits in which an incorrect result is given. The TI89 Advanced Mathematics Software, on the other hand, is very stable, and gives consistent and reliable results. I have used both of them in the last few weeks and I can tell you that there is really a big difference. This is my experience, however I see that many who write on this newsgroup have a different opinion, so I would like to know from people who have used both (TI89 and HP48+Erable) how their experience has been. Alfredo Todini mc0736@mclink.it ---------------------------------------------------------------------- ==== Subject: Re: Symbolic capabilities of HP48 and TI89 Many postings here assume that the HP48+Erable+Alg48 is comparable to > the TI89. I am afraid that this is simply not true. In fact: > 1) A 48GX equipped with the memory expansion necessary to run these > programs costs much more than a TI89; That's true for a 48GX, but you can buy 768K-upgraded 48G in Germany for about the same price. > 2) Erable is certainly a good program, especially if compared to the > built-in symbolic capabilities of the HP; however it is still VERY > UNSTABLE (I have had many crashes in a few days), and, more important, > it sometimes gives a wrong answer, even to simple problems, without any > warning. For instance I have found many simple limits in which an > incorrect result is given. The TI89 Advanced Mathematics Software, on > the other hand, is very stable, and gives consistent and reliable > results. I have used both of them in the last few weeks and I can tell > you that there is really a big difference. Maybe you did not install a stable release of Erable. Current stable version is 3.2. It remains certainly some bugs, but I don't think that it is VERY UNSTABLE. Do you use any other program? Don't forget that another program may corrupt the memory and cause a TTRM hours later. And could you give examples of simple limits that are not solved correctly > This is my experience, however I see that many who write on this > newsgroup have a different opinion, so I would like to know from people > who have used both (TI89 and HP48+Erable) how their experience has been. I use both but my opinion is certainly not unbiaised! I would say that the TI89 is nice for problems that are not too complex but I get often a better result with Erable for more complex problems (just test some examples of the documentation of Erable or ALG48). The TI is a nice blackbox tool as Erable is best used when you know and need more maths. Bernard Parisse ---------------------------------------------------------------------- ==== Subject: Re: Symbolic capabilities of HP48 and TI89 I have reinstalled Erable (I have now installed version 3.2) and it seems much more stable, so probably there was something wrong with my previous installation. In the next few days I will be using it more to test it, but if it remains stable as it is now, I must say that my previous judgement was wrong. I have used it for a while and it works very well. However, one limit it does not calculate is the following: 'SQRT(X)/SQRT(X+SQRT(X+SQRT(X)))' 'X=infinity' LIMIT I get: LIMIT Error: Bad Argument Type This is strange, as it computes much more difficult limits. Alfredo Todini mc0736@mclink.it Maybe you did not install a stable release of Erable. Current stable > version > is 3.2. It remains certainly some bugs, but I don't think that it is > VERY UNSTABLE. Do you use any other program? Don't forget that another > program may corrupt the memory and cause a TTRM hours later. > And could you give examples of simple limits that are not solved > correctly > ---------------------------------------------------------------------- ==== Subject: Re: Symbolic capabilities of HP48 and TI89 > I have reinstalled Erable (I have now installed version 3.2) and it > seems much more stable, so probably there was something wrong with my > previous installation. Alfredo Todini Alfred: Erable is touchy program as far as loading is concerned. One method that has worked for me is to store all other libraries (UFL,QPI,ALG48,AI048, don't get Out of Memory errors. Being stored in different ports helps keep programs from intefering with each other. Since I did that I have had On Limits: Send Mr. Parisse examples of where Erable fails. He is genuinely interested in improving Erable any way he can. I sent him examples of where working on a better algorithm to compute those. Where I go to school the TI-92 and now the TI-89 is banned from use. The HP48 sure as heck beats using a lousy TI-85 or TI-86! - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: Symbolic capabilities of HP48 and TI89 >> Many postings here assume that the HP48+Erable+Alg48 is comparable to >> the TI89. I am afraid that this is simply not true. In fact: >> 1) A 48GX equipped with the memory expansion necessary to run these >> programs costs much more than a TI89; That's true for a 48GX, but you can buy 768K-upgraded 48G in Germany for >about >the same price. Yes, but an upgraded calc gave me lots of trouble. Now with a GX and original mem cards (including Erabel 3.2 :-) I don't have any problems! I personally would not advice to buy an upgraded calc... Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: Re: Symbolic capabilities of HP48 and TI89 Well! Considering symbolic answers to problems given on a final exam in a Circuits 2 course, one may say that the TI89/92 is well equipped with a remote brain that lets people get out of doing the work and the thinking that is involved in certain processes like Fourier transforms and Laplace transforms. There were two guys that had TI92's in this class and they blew the average grade for the whole class! For me, I had Erable and Alg48 installed and done quite well for the most part. It turns out that I never used the HP much on the final though. I used it to find numeric roots to solve DE's and to write numbers in polar form for convolution of two signals because approximate answers were o.k.. For the two guy with TI's, I looked back every now and then and they were just peck'n away. I thought one of guys was a real genius until I saw him using a TI92. Ain't that a shame; he lost an idol because of the TI : ( . When it comes to being a judge of what something can do, one must think of the devistation it can cause also. There are good things and bad things when it comes to both calculators. Everyone with TI's thought that I was sick because I could use an HP and program it System RPL. They didn't like the RPL notation that comes with using it : ( I like it a lot! Things are entered a lot easier. Don't get me wrong but I would switch to TI in a second if I wasn't interested in understanding the math that is involved in Electrical Engineering. The HP48GX suits me just fine for what an Electrical engineer needs to accomplish. That's all for now! documentation on what the TI can do. Jeremy E. Laughery--Author of Exhackt and CompSqr > Many postings here assume that the HP48+Erable+Alg48 is comparable to > the TI89. I am afraid that this is simply not true. In fact: > 1) A 48GX equipped with the memory expansion necessary to run these > programs costs much more than a TI89; > 2) Erable is certainly a good program, especially if compared to the > built-in symbolic capabilities of the HP; however it is still VERY > UNSTABLE (I have had many crashes in a few days), and, more important, > it sometimes gives a wrong answer, even to simple problems, without any > warning. For instance I have found many simple limits in which an > incorrect result is given. The TI89 Advanced Mathematics Software, on > the other hand, is very stable, and gives consistent and reliable > results. I have used both of them in the last few weeks and I can tell > you that there is really a big difference. > This is my experience, however I see that many who write on this > newsgroup have a different opinion, so I would like to know from people > who have used both (TI89 and HP48+Erable) how their experience has been. Alfredo Todini > mc0736@mclink.it ---------------------------------------------------------------------- ==== Subject: Re: Symbolic capabilities of HP48 and TI89 Ecrivez moi si vous connaissez l assembleur sur hp ,merci ---------------------------------------------------------------------- ==== Subject: Re: Symbolic capabilities of HP48 and TI89 MOUEZA a .8ecrit: > Ecrivez moi si vous connaissez l assembleur sur hp ,merci Que veux-tu savoir? Des bouquins, des programmes? Ecris moi .88: andrell@iname.com J'ai pas d.8ebut.8e depuis super long, et je sais .88 quel point on gal.8fre au d.8ebut :-)))) L.89che pas! :) -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Clean solver: small bug This discussion referrs to the Clen Solver programm posted to this newsgroup by Peter Karp on 1998/11/10 with the subject: Clean solver; please search on www.dejanews.com if you lost that post. My observations regard only the version which cleans up ONLY the VARIABLES of the equation and keeps the variable 'EQ'. The bug consist in the fact that in the variable called 'EQ' there is stored not only the last equation solved by the built in solver but also other stuff all in one list. This is a little bit nasty because if I want to access again to the equation I've to recall the content of the variable 'EQ' and then do EVAL PURGE or OBJ-> DROP DROP; so I've slightly modified Peter's programm as follows: %%HP: T(3); << DUP # 353ABh SYSEVAL { { CLEA << RCEQ EVAL DROP DUP # 353ABh SYSEVAL PURGE 0 MENU 'EQ' STO >> } } + 2 ->LIST STEQ 30 MENU >> which is 126 bytes long and has checksum # 34B4h . This version, hitting the 'CLEA' botton , purges all the variables of the equation and keeps the variable 'EQ' in which is stored only the equation. Please note that the SYSEVALS are HEXADECIMAL numbers and that entering them wrong you risk to clear all your memory. That's all, I'm sorry for my bad english and if somebody wants to contact me, please send a email to: 3912@bigfoot.com Bye, alex ---------------------------------------------------------------------- ==== Subject: Re: Clean solver idea Hi Alex, helpfull for some. Although I personally will continue to use the unmodified version. I'll explain why: >This discussion referrs to the Clen Solver programm posted to this >newsgroup by Peter Karp on 1998/11/10 with the subject: Clean solver; >please search on www.dejanews.com if you lost that post. My observations regard only the version which cleans up ONLY the >VARIABLES of the equation and keeps the variable 'EQ'. The bug consist >in the fact that in the variable called 'EQ' there is stored not only >the last equation solved by the built in solver but also other stuff >all in one list. You know the saying it's not a bug it's a feature?! The clean solver's way IS to store a user defined list to the variable EQ. This is how this tiny program works. > This is a little bit nasty because if I want to >access again to the equation I've to recall the content of the >variable 'EQ' and then do EVAL PURGE or OBJ-> DROP DROP; That's right. For me personally I didn't had a problem with that, because I store *all* my formulas into vars and just recall them to the stack to view them and *clean-solving* them when neccessary. Sometimes I know I will have to calculate with the same equation (and with some of the same values), but have to do another calculation between. Then you can just press VAR, PRG, MTH or whatever you want and use the calc like normal. The *strange* EQ (containing the equation with the other stuff) is still there. So all you have to do to call the same EQ again is to press LeftShift, SOLVE (key 7), ROOT, SOLVR. Then you get back to the last EQ-cleaning-solve-routine containing still your safed vars (like when you use the SOLVR routine without the *clean* approach). But you're still able to do CLR to end the routine *clean*. Most of the time (when you hadn't entered more than one other menu) you can get back easier and faster with two keystrokes via the LAST MENU command (RightShift MENU (key L)). Also when you deceide to CLR (clear) the vars, the EQ is still permanent (with those *strange* contents) and you are also able to get back to the last used EQ-clean-solver by doing LeftShift, SOLVE, ROOT, SOLVR when you want to. (If you would like to add this to a custom menu you can use 30 MENU to do this) > so I've >slightly modified Peter's programm as follows: %%HP: T(3); ><< DUP # 353ABh >SYSEVAL { { CLEA > << RCEQ EVAL DROP >DUP # 353ABh >SYSEVAL PURGE 0 >MENU 'EQ' STO >> } } + 2 ->LIST >STEQ 30 MENU >> The original version has no real bug, but the feature to act like described :-) BTW, when someone likes to delete the reserved variable EQ you can just add 'EQ' PURGE at the end of the inner program. For your convenience here the unmodified version: %%HP: T(3); << DUP # 353ABh SYSEVAL { { CLR << RCEQ EVAL DROP # 353ABh SYSEVAL PURGE 0 MENU @ add 'EQ' PURGE here if you want to >> } } + 2 ->LIST STEQ 30 MENU >> One *bug* - missing feature - is here that you can only solve for one EQ. It's not possible to solve for more then one EQ with this version of the clean solver ( Levy's MathTools solver calls the multisolver in this situations). I personally have never used the solver with more then one EQ. The routine could be adapted to handle those cases too, but would be larger then... Your Mr. Saubermann (translation: dein Herr *washy* ;-) Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: Re: Clean solver: small bug Here is an after the fact general clean-up (purges all numeric and solver-related variables, even if accumulated over solving several different equations): << { 0 13 } TVARS { EQ Mpar } + PURGE >> Leave out anything above which you do not want to delete (type 0 includes all variables containing real numbers; type 13 includes variables which contain units, as may be created using the Equation Library). ----------------- A very different idea is to mark a directory ahead of time, before commencing equation solving; any new variables which subsequently appear in the VAR menu to the left of the mark are then candidates for deletion: Items to store in HOME: << VARS 1 OVER 'keep' POS 1 - SUB >> 'Keep' STO << VARS 'keep' 'Keep' OVER STO POS 6 / CEIL 1 % 2 + MENU >> 'KEEP' STO (This is all case-sensitive, so please carefully distinguish k from K etc.) What to do to mark existing variables as to be kept: Simply type KEEP; a variable named 'keep' appears in the VAR menu. What to do to clean up afterwards: Simply press the VAR menu key for 'keep' (if you don't see it, you can either press NXT until you do, or just type KEEP again, which will find the right menu page). Note: I omitted any PURGE command, so that you can decide whether you really want to purge all the variables in the list which appears; this could protect you from forgetting all about the matter, but inadvertently pressing 'keep' later, after having created some big new masterpiece of a program which you did *not* want to purge! ---------------------- Trained mice might be another alternative; there seem to be plenty of them scurrying about my office, looking for something to do (or chew) -- now, TI calcs are so simple that even a stupid mouse could probably be trained to work them, but I don't know whether a hard to use HP with RPL could be entrusted to any of these lemming-like mindless drones, even if they do tend to spout about how it insults their intelligence every time they come across something they don't comprehend :) ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: clean up >Here is an after the fact general clean-up And another idea how to clean up the_last annoying variables: << VARS DUP ROT POS 1 SWAP SUB PURGE >> name it DELTO or something like that - put the name of the var on the stack which you like to purge to. Run DELTO. All variables (inlcuding the one you've named) will be deleted. If you always have the same variable structure and your first variable should always be the same you could do: << VARS DUP 'MyFirst' POS 1 - 1 SWAP SUB PURGE >> Here all variables up to your variable 'MyFirst' are deleted, but without the var 'MyFirst' itself. The variable 'MyFirst' acts as a marker here, like in John's suggestion. The above suggestions began to make fun, when you assign the program to a User_key (see chapter 30 in the manual) Isn't it great that the HP48 can be customized so easy? :-) Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: BZ test question I just have tried the new compiled different versions of BZ. (Available at hpcalc.org named BZ Series). I have tried all (bigger) BZ's. My free mem was about 70 kilobyte. BZ packed a directory from 3493.5 bytes to 1626.5 bytes (-53.44%) The timings with different versions of BZ were: BZ [seconds] 1 12.0 (original BZ1) 2 9.9 3 9.9 4 8.7 (original BZ4) 5 9.8 6 8.5 7 8.4 8 7.5 9 9.7 10 6.9 (fastest!) 11 12.0 15 9.8 18 7.6 copy of the origial source code documentation: ________________________________________ Examples for time compression of a same file: Buffer Time Mem needed: #1000 14.54 26624 #800 15.77 16384 #400 17.56 11264 This is the BZ4 #200 21.26 8704 #100 28.59 7424 This is the BZ1 #80 42.78 6784 #40 69.97 6464 Do you know how one can calculate the desired memory for a specific buffer size? And why are some versions with a bigger buffer-size slower the others :-? Looking forward your answers! Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: Re: BZ test question : The timings with different versions of BZ were: : : BZ [seconds] : : 1 12.0 (original BZ1) : 2 9.9 : 3 9.9 : 4 8.7 (original BZ4) : 5 9.8 : 6 8.5 : 7 8.4 : 8 7.5 : 9 9.7 : 10 6.9 (fastest!) : 11 12.0 : 15 9.8 : 18 7.6 : And why are some versions with a bigger buffer-size slower the others : :-? Somehow I get the impression that for 11, 15 and 18, you get the same times you got for 1, 5 and 8. Is the buffer size of these versions really larger than that of BZ10 ? Just my .02 Pete Commercial email sent to this address will be proofread ---------------------------------------------------------------------- ==== Subject: Re: BZ test question >The timings with different versions of BZ were: BZ [seconds] 1 12.0 (original BZ1) >2 9.9 >3 9.9 >4 8.7 (original BZ4) >5 9.8 >6 8.5 >7 8.4 >8 7.5 >9 9.7 >10 6.9 (fastest!) >11 12.0 >15 9.8 >18 7.6 Facts: 1. The ordinal number indicates the size of the hash table in units of #100. This multiple is the constant HSHSIZE defined at the start of the compressor source. 2. Increasing hash size will increase the memory use 10-fold, (The equate is HASHMEM EQU 10*(HSHSIZE) ) since the hash table contains pointers to the first and last matches with the same hash code (2*5 nibbles). 3. The code assumes #1000 is the maximum HSHSIZE (oridinal 10), and accordingly uses the X field in the following places: a) start of AddNew subroutine b) Start of RemoveOld subroutine 4. Using a size larger than #1000 will not work since only the X-fields are used. Note the following 'strange' facts... BZ1 = 12.0s BZ11 = 12.0s BZ5 = 9.8s BZ15 = 9.8s BZ8 = 7.5s BZ18 = 7.6s 5. The hash calculation mechanism is NOT automatically optimized for each different HSHSIZE. The calculation is always the same, and may not work equally well for all unique HSHSIZE values. In this case, your timings pretty much decrease as expected, but with a few exceptions which can be attributed precisely to this fact. To save searching trouble, this is how the hash is computed based on the next 5 nibbles to be matched (any shorter matches are always output as literals). * Compute hash A=DAT0 A A[A] = value C=A A C[A] = value CSR A CSR A C[A] = value >> 8 CSRB.F X CSRB.F X C[A] = value >> 10 = val2 B=A X A=A!C X A[X] = val ! val2 C=C&B X C[X] = val & val2 A=A-C X A[X] = (val ! val2) - (val & val2) LC(3) (HSHSIZE)-1 C=C&A X C[A] = hash modulo HSHSIZE I think the actual logical operation is XOR (a!b - a&b), but I won't bother to dig it up. Also, since the modulo is calculated using the AND operation, HSHSIZE should be a multiple of #100. I have NOT performed any significant analysis on how effective the hash really is. I merely tested a few variants on text and binary files, and kept the one which seemed to be the most effective. If you wish to extend the HSHSIZE, it is sufficient to expand the hash calculating bits like the above one. The subsequent code will use the entire C[A] field to calculate the address in the hash table. Note that the improvements from expanding the hash table size diminish quite quickly. You might wish to experiment with different hash functions instead. Just make it a MACRO (apart from the leading A=DAT0 or A=DAT1) and use it in AddNew and RemoveOld to save the trouble of copying the code.. -- --- --> Mika Heiskanen mheiskan@gamma.hut.fi http://www.hut.fi/~mheiskan ---------------------------------------------------------------------- ==== Subject: Re: BZ test question I do not know the different versions of BZ, but I think the timing variations come from different RPL overheads in each version. Another point may be that the position of the program object in the var list may have some (minor) influence to the startup speed of a program. Gr.9f§e an den Benchmarker ;-) Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: BZ test question Hi Raymond, >I do not know the different versions of BZ, >but I think the timing variations come from >different RPL overheads in each version. As far as I know only the buffer size is different... >Another point may be that the position of the program object in >the var list may have some (minor) influence to the startup speed >of a program. All BZ versions where at position one in the vars when I ran the test. When I retry later when a specific version was not the first one the timings were simular (I only looked for the tenth of seconds). Gr.9f§e an MagicRaymond :-) Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: Is there any way to transfer files from HP48G(X) to computer via IR? I've read somewhere that someone managed to do that, not particularly quickly, but nonetheless, it worked. I was wondering if anyone knew how to do that. If anyone has any suggestions, please email me at man@email.unc.edu (I'll try and check the group, but I may or may not remember to... exams and all...) Michael Nicholson ---------------------------------------------------------------------- ==== Subject: Laplace transform; reply-to: copernico@tuttocitta.it Can somebody tell me where I can find a good program or a library to perform Laplace Transforms? AS ---------------------------------------------------------------------- ==== Subject: Re: Laplace transform; reply-to: copernico@tuttocitta.it Well, Neopolys v6.5 is pretty good but lacks some importent features. It doesn't support the shift rules, Heavisides step function, Dirac's delta function or general transform of f(t) --> F(s). Actually you'll get a memory clear if you try to transform an algebraic like f(t). Anyone else has such severe problems with Neopolys LaPlace transforms? I should mention that I also use Java v3.4a - that might have a say in this (I dont use v3.6 since the Cappucino version has a bug that doesn't allow you to use the built in letters as subscripts on names when the calc is using comma as fraction mark - that's not just me, right? Neopolys v6.5 is actually by far the best program I've seen that can do LaPlace transforms - I use it myself - and it's also pretty nifty with everything regarding polynomials! A bit big though (~11k), if you ONLY need the LaPlace funtions. Antonino Sabetta skrev i meddelelsen <74oqi5$cqo$1@nslave1.tin.it>... >Can somebody tell me where I can find a good program or a library to perform >Laplace Transforms? AS ---------------------------------------------------------------------- ==== Subject: SYSEVAL command je recherche toutes les commandes SYSEVAL ainsi que ce .88 quoi elles se rapportent. I find all the SYSEVAL command http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: SYSEVAL command a .8ecrit dans le message : 74ouch$jor$1@nnrp1.dejanews.com... >je recherche toutes les commandes SYSEVAL ainsi que ce .88 quoi elles se >rapportent. I find all the SYSEVAL command > You should read this books : Tu devrais lire ces livres : Les secrets de la HP48G/Gx de Jean-Michel Ferrard TOME I & II S.8ebastien ---------------------------------------------------------------------- ==== Subject: Re: SYSEVAL command S.8ebastien a .8ecrit: > You should read this books : > Tu devrais lire ces livres : Les secrets de la HP48G/Gx de Jean-Michel Ferrard TOME I & II Ce bouquins ne sont plus dispos. Ils ne le sont plus .88 Maubert, ils ne le sont plus dans les 8 Fnacs que j'ai fait .88 Paris, et m.90me .88 l'.8etranger ils gal.8frent pour les obtenir (autant dire: ils cherchent encore aux derni.8fres nouvelles) Ces bouquins sont s.9erement bien, mais l.88 on gal.8fre trop. Je pense qu'on devrait rassembler tout ce qu'on a, et faire un nouveau bouquin, ou quelque chose. A force de pas avoir l'information disponible, et en particulier les bouquins, trop de gens programment pas sur HP. Sur les trois personnes avec moi qui avons commenc.8e l'assembleur, il reste que moi, les autres ont laiss.8e tomb.8e. La semaine derni.8fre, j'ai d.8epens.8e plus de 200 F en photocopies que j'ai de deux heures pas jour .88 r.8epondre .88 mes m.8els :-)) Faut qu'on se bouge franchement, si on veut avoir plus de programmeurs LM et sysRPL sur HP. Ya des gens motiv.8es, plein de bonnes id.8ees, faut pas qu'ils se tournent ailleurs!!!!! -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: SYSEVAL command F.H. Gilbert a .8ecrit dans le message : Ce bouquins ne sont plus dispos. > Alors essayes les doc HP fournies dans TOOLS.ZIP, le fichier RPLMAN.DOC. Bon courage ! S.8ebastien ---------------------------------------------------------------------- ==== Subject: Re: SYSEVAL command S.8ebastien a .8ecrit: > Alors essayes les doc HP fournies dans TOOLS.ZIP, le fichier RPLMAN.DOC. J'ai lu ce dit fichier (bri.8fvement), et il y a peu de choses vraiment exploitables. AUCUNE description dans RPLMAN.DOC de la structure interne et d.8etaill.8ee de l'objet biblioth.8fque. Ces docs sont mal foutues, et incompl.8ftes. Plein de fois on m'a envoy.8e vers des documents, o.9d le r.8epertoire ProgrammingDocumentation de www.hpcalc.org Je l'ai fait des dizaines de fois, et comme je n'obtenait toujours pas l'information recherch.8ee, et maintenant je ne le fais plus. Le manque d'information conduit au manque de programmeurs, il m'a fallu parfois demander plein de fois de suite pour obtenir un brin d'info utile. Le plus rigolo c'est de poster une question *simple* et d'obtenir en r.8eponse autre chose, qui n'a rien .88 voir et me renvoie vers un document qui, une fois consult.8e ne r.8epond pas .88 la question. Je suis impatient, j'ai des sources qui bloquent sur des points dont la r.8eponse, une fois obtenue, aurait du provenir vite et facilement de la question pos.8ee .88 l'origine. Pardonnez moi mon impatience :-) -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: tempob area Can someone tell me how the info before each object in TEMPOB is coded? I know there are flags there, but in which order and for which meaning? -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: tempob area > Can someone tell me how the info before each object in TEMPOB is coded? > I know there are flags there, but in which order and for which meaning? > -Gilb- > 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 > There is but one flag, and it is used in garbage collection: When gc starts, it clears all flags of the tempob objects. It then sweeps all areas that may contain pointers to objects: the stack, return stack, temporary environments etc. For every object it finds referenced, it sets the corresponding flag in the tempob area. When every reference has been checked, all tempob objects with a clear flag may be deleted, for they are not referenced anywhere. -- Werner Huysegoms remove the x before replying http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: tempob area > There is but one flag, and it is used in garbage collection: Cool.. is it one bit? One nibble? Someone told me it's more than that: there is the size of the object in tempob too, and you said nothing of it. Still seeking the answer.. :P -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: list -> string suppose I have a list like: { x y z} > Is there a way to get an element of the list and put it in a string > without the tick marks. For example, if I do: > { x y z} HEAD ->STRING I get : 'x' > but what I want is just: x Replace ->STR with + ----------------------------------------------------------- > With best wishes from: John H Meyers I do believe this is your shortest posting ever, John. Slow day? ---------------------------------------------------------------------- ==== Subject: Re: list -> string ( see below for reactant posting ) when i saw this the other day... something about it stuck me as very disturbing, but unitl i was responded to by nospam...i wasn't quite sure what it was... it was that ( this occurred to me yesterday ) this posting occurs at the end of thread of posts...and that this technique JHM is promoting ONLY works ( very elegantly ) with single character expressions ( single tick delimiters )... it does not work with anything larger...such as real expressions or eq's... Now i see what so disturbs me...John DIDN'T point this out...!!! and then, the other day, he posted a WHOLE thread all by himself on EXIT routines...and i thought then...is he off his medications...has he been taking too much cough medicine...??? what is what? we're worried about you, John...!!! suppose I have a list like: { x y z} > Is there a way to get an element of the list and put it in a string > without the tick marks. For example, if I do: > { x y z} HEAD ->STRING I get : 'x' > but what I want is just: x Replace ->STR with + ----------------------------------------------------------- > With best wishes from: John H Meyers I do believe this is your shortest posting ever, John. Slow day? ----------- :: o .---..-..-..-..-..-..-. . .-. .-. .-..-..-.. .-. `| |'| || .` | > / / / / / | .` || | ) / `-' `-'`-'`-' `-' `. ^ .' `--^--'`-'`-'`-'' `--^--' Join The BabyNous Cult : The Friendly Neighborhood Cult ---------------------------------------------------------------------- ==== Subject: Connections (Re: list -> string) > this technique JHM is promoting... 'name' + ==> name > ONLY works ( very elegantly ) with single character expressions It works with any single name of any length, but with algebraic expressions (a different object type than names), the tic marks remain. Basically, it produces the same thing you get when you enclose the original items in a list, e.g. { 'ABC' '2+3' } ==> { ABC '2+3' } by which I mean that if you type into the command line what I indicated on the left side, it appears in the display as what I indicated on the right side. These varying forms of displaying things, according to where the display is to appear, are the results of the machinations of the internal decompiler, which is responsible for making text representations of internal binary objects, which must be done for any display, printing, or editing of any object, and of course for specific commands to turn objects into strings. > Now i see what so disturbs me...John DIDN'T point this out...!!! The specific question concerned a list of names only; did you want me to keep going until I'd written a whole new AUR? See? -- the one time I stop after answering just what was asked, and you complain! > and then, the other day, he posted a WHOLE thread all by himself > on EXIT routines...and i thought then...is he off his medications... It was in response to a specific request, and also to similar past inquiries for which no solution was ever posted, in my recollection. Of course, this time, the person who asked just settled on 0 DOERR, and my elegant solutions, six months in the making, were tossed in the circular file by the customer, which in fact happens many a time in the custom software business :) BTW, I'm on meditations, not medications :) > I do believe this is your shortest posting ever, John. Slow day? Little David Susskind -- please shut up! Please don't talk! - Allan Sherman (My Son, the Folk Singer) Don't worry, TW, I'll make up for whatever I've left out this time, by filling you in on some more things, which, given the huge and clever programming projects I see you undertaking, might even interest you: Apropos of your your observations, I might as well comment here about the two things that the internal compiler does differently between the compiling of programs and the compiling of lists. I will illustrate this by showing both the original input you type, and the result of disassembling what the HP48 produces from it, using Jazz: << ... >> (You enter this, and see the same displayed) :: x<< ... x>> ; (Jazz reveals its true contents) Now, you can see that a program object is represented, in SysRPL, using the delimiters :: at the beginning and ; at the end, right? So, what about the << and >>, which the user manual talks about? Well, those visible symbols, which are what you believe to be program delimiters in UserRPL, are not actually delimiters at all; in fact, they are *commands*, which the built-in compiler requires you to type when you want to enter a program, and which are actually retained, as extra commands, *inside* what is the *real* program object, whose real delimiters (:: and ;) are never actually displayed at all! This explains why the smallest possible user program, viz. << >> actually occupies 10 bytes: each of its four ROM pointers (to :: and x<< and x>> and ; ) is 2.5 bytes long. Oh, and of course you get the idea that SysRPL likes to represent all the UserRPL commands with an x in front, like xCOS, etc. << << >> 20 bytes? No! -- 22.5 bytes! Why? :: x<< xSILENT' :: x<< x>> ; x>> ; ~~~~~~~~ ~~~~~~~~~~~~ what? << >> My gosh, what is this silent object all about? Well, for one thing, it is not normally displayed on your screen at all, so, just like the real program delimiters, you do not even know it is there. Now, why is there a quote symbol ['] in the name of this weird beast? You guessed it -- it quotes the next following object, so that the next object gets put on the stack, no matter what it is -- even if it is a *program*, which would otherwise get immediately *executed* As parents tell their children about Santa Claus, but then have to do unseen things to make their promises come true, so does HP have to do unseen things to make come true what the user manual promises: On page 29-2 (G/GX User's Guide): Actions for Certain Objects in Programs: Program: Put on the stack. Were it not for the Secret Santa -- the xSILENT' command -- which the UserRPL compiler secretly inserts for you, that promise would not come true! << 'ABC' >> :: x<< x' ID ABC xENDTIC x>> ; ~~ ~~~~~~ ~~~~~~~ ' ABC ' As you can again guess, the x' thingy is another of those things which quotes (puts on the stack) whatever follows; this is what prevents the above program from *executing* the contents of variable ABC, rather than just putting the name on the stack. The xENDTIC thingy is for what, then? Well, what would your aesthetic sense say if the program were displayed as << 'ABC >> ? (as in the FXRECV program from HP, if you've ever seen it :) Of course, the lack of a symmetrical set of quote marks around the name ABC would offend you, so the xENDTIC thingy is there to display the trailing quote mark when you display the program. Actually, it does one more thing for you during program execution: it is one of those commands, like the user program delimiters, which secretly perform the internal function CK0ATTNABORT. If you are completely used to making the assumption that the CANCEL key may be used to abort a running program, then now you can be told that you owe this gift to another Secret Santa which HP sprinkles throughout user programs: only those user commands which perform CK0ATTNABORT can actually be interrupted. Otherwise, if you get into an endless loop in SysRPL, say, pressing CANCEL is pretty futile. The internal operating system can not afford to be stopped dead in its tracks in the middle of critical operations, so actually it polls the ATTN flag (set by pressing CANCEL) only when it's good and ready; most internal functions ignore this flag, but many user commands have the specific duty of doing so for you. { 'ABC' << ... >> } ==> { ABC << ... >> } { ID ABC :: x<< ... x>> ; } As I was saying, a couple of hours ago, I think, I was going to show you the difference between how the HP48 compiles program objects, vs. how it compiles lists (see? -- I remembered what I started to say, so I guess I forgot to inhale) Perusing the above list, do you see that it contains *neither* the x' nor the xENDTIC commands surrounding ABC, nor does it contain xSILENT' in front of << >>, all of which were inserted automatically by the compiler, when we typed the above text inside those pesky << ... >> thingys, instead of inside { } ? Why does HP do this differently? Right, because HP does not expect you to EVAL (execute) this list like a program; if those extra thingys (some of them invisible) were included in the list, we'd have heard endless complaints about why 1 GET did not recall the name ABC from the above list, nor would 2 GET recall the program (and OBJ-> would break that list up into some weird things which we wouldn't even recognize). In fact, this might even explain why HP refuses to pay attention to the tics you typed around 'ABC' and thus re-displays the list without them after you finish typing it in. So what, you say? Well, if you are a real memory cheapskate, like me (who bought the HP48G/32K, before realizing how many years could be wasted figuring out how it works, if only one had more memory), and if you had written, say << ... >> IFT in a program, well, pretty soon you might find out that it would save five big ones (bytes, that is, not just nibbles) to write { ... } IFT instead (eliminating the useless program-delimiter commands). The nasty reality, however, is that this does not always work; if the *program* which you want to replace by a *list* contains *either* an inner program << ... >> or a quoted name 'XXX' anywhere within it, then if you don't know (or forget) the facts above, you will end up with a program that saves five bytes but doesn't work, a fact which it originally took me a while to figure out :) Many of the things which don't seem to make sense in this world, which lead us to rail and complain about the stupidity of its creator(s), can similarly be recognized for their more profound content of intelligence, once we rise to the level of knowledge which enables us to comprehend them in their fullness. That's why I write with the motto of my original Alma Mater in mind: Knowledge and Thoroughness, intending to convey as deep an appreciation as possible for the fullness of things, ranging from the inspired design of RPL to the wider connections between the smaller and larger things of life, each part of which has something to do with the whole. The great scientists and philosophers have been those who had the deepest sense of this wholeness, I would think: I do not know what I may appear to the world, but to myself I seem to have been only like a boy playing on the seashore, and diverting myself in now and then finding a smoother pebble or a prettier shell than ordinary, whilst the great ocean of Truth lay all undiscovered before me - Isaac Newton Two things fill the mind with ever new and increasing admiration and awe the more often and steadily we reflect upon them: the starry heavens above and the moral law within me. I do not seek or conjecture either of them as if they were veiled obscurities or extravagances beyond the horizon of my vision; I see them before me and connect them immediately with the consciousness of my existence. - Immanuel Kant Critique of Practical Reason A human being is part of a whole, called by us the Universe, a part limited in time and space. He experiences himself, his thoughts and feelings, as something separated from the rest--a kind of optical delusion of his consciousness. This delusion is a kind of prison for us, restricting us to our personal desires and to affection for a few persons nearest us. Our task must be to free ourselves from this prison by widening our circles of compassion to embrace all living creatures and the whole of nature in its beauty. - Albert Einstein ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: Re: Connections (Re: list -> string) > Actually, it does one more thing for you during program execution: > it is one of those commands, like the user program delimiters, > which secretly perform the internal function CK0ATTNABORT. > Interesting, which prompts me to ask: According to sad, all of <<. >>, xIF, xCASE, xEND and others ar all the -same- object, namely :: CK0ATTNABORT ;. Are they all really the same, or is there some subtle difference beyond the different names to allow the compiler to enforce its syntax rules. -- Steve Dickey ---------------------------------------------------------------------- ==== Subject: Re: Connections (Re: list -> string) > According to sad, all of <<. >>, xIF, xCASE, xEND and others > are all the -same- object, namely :: CK0ATTNABORT ; > Are they all really the same, or is there some subtle difference > beyond the different names to allow the compiler > to enforce its syntax rules. Just as identical programs may be stored into several different user variables having different names, so can identical SysRPL programs be stored into several different library commands; each of these separate SysRPL programs, in turn, calls the common internal function CK0ATTNABORT. The distinct user words x<<, xIF, etc. are all address pointers to the different ROM addresses of the different ROM library functions (in lib 1792 or lib 2), which all have the same content, but different names, and possibly different internal functions associated with compile-time actions. ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: Re: Connections (Re: list -> string) > Interesting, which prompts me to ask: > According to sad, all of <<. >>, xIF, xCASE, xEND and others ar all the -same- > object, namely :: CK0ATTNABORT ;. > Are they all really the same, or is there some subtle difference beyond the > different names to allow the compiler to enforce its syntax rules. -- > Steve Dickey > Now, if I really understood the property flags of library commands I would be able to show you the difference. For now, let it be said that using those property flags (3 nibbles in front of the romptr body in front of a named library command), it is possible to define syntax rules for the UserRPL compiler. That's where the objects differ. -- Werner Huysegoms remove the x before replying http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: EMU48 <-> HP 48 Excuse moi de te derranger Laurent, mais j'ai une HP48 sans cable de liaison et j'aimerais qd m.90me utiliser EMU48. Peux tu m'envoyer la ROM que tu utilise. Merci d'avance, Ed. > Does EMU48 could talk to a HPvia the serial port. I would like to use it to > transfer files and to play ---------------------------------------------------------------------- ==== Subject: Re: EMU48 <-> HP 48 Ecoute normalement je n'ai pa le droit de te donner ma ROM. En cherchant sur le Net j'ai trouv.8e un vieux Emu48 avec la ROM fournie. Je vais le rerechercher (l'ai d.8etruit apr.8fs avoir vu que c'.8etait nul) car j'ai un HP 48 SX et je cherche une ROM GX (plus de m.8emoire, carte constante int.8egr.8ee..) Je te recontacte quand j'ai trouv.8e Bye ---------------------------------------------------------------------- ==== Subject: HI !!! my friend jonathan chum Why don't you report me to your mom n' then see what i do with her? Fuckk you puto baboso!!!!!!!. chum??? let's say i CUM on your big mouth ---------------------------------------------------------------------- ==== Subject: Re: HI !!! my friend jonathan chum Why do you do such strange things? I did not even report you, just a suggestion to others. Why would you even dare to curse and say that you will exchange an illegal rom to others for an exchange of more warez crap?? This newsgroup will not tolerate such activities! I frown upon such things =( ---------------------------------------------------------------------- ==== Subject: Re: HI !!! my friend jonathan chum Oh yeah. _THIS_ person is definately intelligent. PLease expend your energy on more constructive replies. You are only embarrassing yourself with this kind of language. Why don't you report me to your mom n' then see what i do with her? > ****k you puto baboso!!!!!!!. > chum??? let's say i -- Al Arduengo ---------------------------------------------------------------------- ==== Subject: TinyWanda Idea for Getting Rich... i was just thinking about this 'right now'* so i thought i would post it to this group which i know is filled with smart people, some of which, perhaps, that would like to become fabulously wealthy, AND help the world out... ok, this is the idea... i just heard on the radio that bogus computer scanned and printed checks are flooding the Seattle area... awhile ago, i heard a similar story...and it occurred to me... what banks and stores need is a small device** upon which you could stick a piece of printed material into it, and it would tell you how it was printed...that is, what method was used to create it...high pressure printing, ink jet, laser coping...or whatever...??? eh? send my share ( 2% ) to a homeless shelter in your area. ............ * as i was reading this group, and listening to the radio... ** it might be built into the other check checking thinger they use...??? ----------- :: o .---..-..-..-..-..-..-. . .-. .-. .-..-..-.. .-. `| |'| || .` | > / / / / / | .` || | ) / `-' `-'`-'`-' `-' `. ^ .' `--^--'`-'`-'`-'' `--^--' Join The BabyNous Cult : The Friendly Neighborhood Cult ---------------------------------------------------------------------- ==== Subject: Re: TinyWanda Idea for Getting Rich... >i was just thinking about this 'right now'* so i thought i would post it to >this group which i know is filled with smart people, some of which, perhaps, >that would like to become fabulously wealthy, AND help the world out... ok, this is the idea... >i just heard on the radio that bogus computer scanned and printed checks are >flooding the Seattle area... >awhile ago, i heard a similar story...and it occurred to me... >what banks and stores need is a small device** upon which you could stick a >piece of printed material into it, and it would tell you how it was >printed...that is, what method was used to create it...high pressure printing, >ink jet, laser coping...or whatever...??? eh? >send my share ( 2% ) to a homeless shelter in your area. Ah.....the only problem is, a lot of legitimate checks are printed by laser printer. That is why the crooks are able to accomplish this sort of crime. Businesses, and even some home users, buy special laser cartridges, and check safety paper, and then print their own checks. It saves over the relatively high cost of pre-printed computer checks. ---------------------------------------------------------------------- ==== Subject: JASM locks Hello ASM-Cracks and Gurus Im newby in assembler and Ive a problem while compiling code of >3k: the J-Asm Beta 0.92 locks in the initialising screen. Ive a Gx without ram card, so maybe my memory is unsuisifience (theres no error), or Im waiting not enough (how long takes it to compile code of about 5k or 20k?)... Thanx to Julien Meyer for his ads-docs, they are great. Noel Rohrbach Bern, Switzerland ---------------------------------------------------------------------- ==== Subject: Re: JASM locks sorry, sorry my fault (of course...) please be patient >Hello ASM-Cracks and Gurus Im newby in assembler and Ive a problem while compiling code of >3k: > the J-Asm Beta 0.92 locks in the initialising screen. >Ive a Gx without ram card, so maybe my memory is unsuisifience (theres no error), or >Im waiting not enough (how long takes it to compile code of about 5k or 20k?)... Thanx to Julien Meyer for his ads-docs, they are great. Noel Rohrbach > Bern, Switzerland ---------------------------------------------------------------------- ==== Subject: Re: JASM locks Sorry, sorry my fault, forget it (please be patiente) please...... >Hello ASM-Cracks and Gurus Im newby in assembler and Ive a problem while compiling code of >3k: > the J-Asm Beta 0.92 locks in the initialising screen. >Ive a Gx without ram card, so maybe my memory is unsuisifience (theres no error), or >Im waiting not enough (how long takes it to compile code of about 5k or 20k?)... Thanx to Julien Meyer for his ads-docs, they are great. Noel Rohrbach > Bern, Switzerland ---------------------------------------------------------------------- ==== Subject: Re: Problem with Extended BASE v0.95. I don't know this program. but because of the error you get (AND Error), IBFIX should contains an Integer with how many digits you want to work on. Example, if you want to work on 8 bits, then you save #FF in IBFIX. Give a try. Who knows ? Jean-Yves >I've uploaded in my HP48GX the software Extended BASE v0.95. I'm able to convert from a base to another (->IB), ... >but I 'm not able to do aritmetical operation (+,-,*,/). Example: I would add 1 and 2 in binary base ( 2 in BBASE) >1 + 10 = 11 I put in to stack: 4: >3: >2: 1 >1: 10 And then i go in VAR menu >I press [LOGIC] >then the down arrow >then NXT >then NXT >now I see the menu with >[ADD][SUB][MUL][DIV][INP][CONV] i press [ADD] and the calculator show this message: AND Error: >Undefined Name >4: >3: # 1d >2: 1 >1: 'IBFIX' >[ADD][SUB][MUL][DIV][INP][CONV] why ? >where I'm wrong ? Could anyone show me to right way or the solution >for this problem ! I thank you ! Fabio Pifferini Please, reply me directly to my email >mailto:101097@ticino.com P.s. ---------------------------------------------------------------------- ==== Subject: HP----> PC Cable I am interested in purchasing a hp cable for a 48g calculator. If anyone ---------------------------------------------------------------------- ==== Subject: Re: HP----> PC Cable Try this place, I think they have them for $10. http://www.palmtoppaper.com I bought one for my HP 95LX and it also fits the 48G. ---------------------------------------------------------------------- ==== Hi! I know that Speed Up Module alteres the clock of the calculator and kills the battery faster, but it still sounds great. I am thinking on buying a calculator with the module, and I was wondering if it would give me troubles Has any of you had any trouble using Speed Up Module in these or any other way? Roberto P.8erez-Franco Creator of HPCalc+ICQ users list http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Hello Roberto, the Speed UP Module work with our Ramcards , ourt internal expansions and with the original HP Ramcards, but not with cards from Klotz and Krolop and Cynox. I don't know wether it works with the gridcard or not. Dieter THE DEVELOPER OF THE SPEED UP FOR HP48 Digitalis, HP-Handheld-Zentrum-Reutlingen Dipl.-Ing.(FH) Dieter Goller (email: DigitalisR@aol.com) Burgstr. 66 D-72764 Reutlingen Tel / Fax : 07121-47 88 62 http://www.Digitalis.de/ ---------------------------------------------------------------------- ==== > Hello Roberto, the Speed UP Module work with our Ramcards , ourt internal expansions and with > the original HP Ramcards, but not with cards from Klotz and Krolop and Cynox. And work with the Cynox internal expansions? > I don't know wether it works with the gridcard or not. Dieter THE DEVELOPER OF THE SPEED UP FOR HP48 > Digitalis, > HP-Handheld-Zentrum-Reutlingen > Dipl.-Ing.(FH) Dieter Goller (email: DigitalisR@aol.com) > Burgstr. 66 > D-72764 Reutlingen > Tel / Fax : 07121-47 88 62 > http://www.Digitalis.de/ > ~+ . . * ^ * + Saturn: More than a simply CPU? ' | () .-.,=``=. - o - Of course, it's a planet! '=/_ | * | '=._ | `=./`, ' . '=.__.=' `=' * + + O * ' . ---------------------------------------------------------------------- ==== > Has any of you had any trouble using Speed Up Module in these or any other > way? Roberto P.8erez-Franco Roberto: What Speed Up Module? Is one being marketed? The closest thing to one that I'm aware of is the MetaKernal ROM card, which DOES speed the calc greatly. I really like my MetaKernal. The only down side to using the MetaKernal ROM card is that it eats up port 1. Maybe the next version of programs in port 0 and 1. - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== a .8ecrit dans le message : 74rfs0$pra$1@nnrp1.dejanews.com... >one that I'm aware of is the MetaKernal ROM card, which DOES speed the calc >greatly. I really like my MetaKernal. The only down side to using the >MetaKernal ROM card is that it eats up port 1. Maybe the next version of >programs in port 0 and 1. - J.C. Ron Correction : the Metakernel does not speed up the calc : it takes full advantage of the saturn capabilities ! So it shows what we can do with an HP 48... S.8ebastien ---------------------------------------------------------------------- ==== > a .8ecrit dans le message : Correction : the Metakernel does not speed up the calc : it takes full > advantage of the saturn capabilities ! So it shows what we can do with an HP > 48... S.8ebastien Sebastien: What you say may very well be technically correct - but the speed of calculator operations is MUCH faster with the MetaKernal than without. The best example is the Equation Writer. The MetaKernal Equation Writer is really great. I haven't measured it exactly, but it must be 3-5 times faster than the built-in equation writer. I think it is safe to say MetaKernal speeds up the calculator as a general rule. Having the Filer is worth the price of the card alone. Memory management is a snap with the MetaKernal. I have 80 KB of notes ,procedures, etc... that I can access in a flash with the Filer. The Filer puts the TI-89's memory to shame. In addition, the Filer lets you add, move, purge libraries, rename files and programs, etc... very fast. It will be interesting to see if the next version of MetaKernal arrives before this so- called speed up module from HP comes out. Won't that be fun? - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== >Sebastien: What you say may very well be technically correct - but the speed >of calculator operations is MUCH faster with the MetaKernal than without. The >best example is the Equation Writer. The MetaKernal Equation Writer is really >great. I haven't measured it exactly, but it must be 3-5 times faster than the >built-in equation writer. I think it is safe to say MetaKernal speeds up the >calculator as a general rule. Having the Filer is worth the price of the card >alone. Memory management is a snap with the MetaKernal. I have 80 KB of notes >,procedures, etc... that I can access in a flash with the Filer. The Filer >puts the TI-89's memory to shame. In addition, the Filer lets you add, move, >purge libraries, rename files and programs, etc... very fast. It will be >interesting to see if the next version of MetaKernal arrives before this so- >called speed up module from HP comes out. Won't that be fun? - J.C. Ron OK, ok. Let me explain what I wanted to say. First I use the MetakernEl (with an E !) for 6 years now, and I've met J-Y Avenard many times because I'm near Paris. When I read something like : ... the MK speed up the 48 ? I have to be correct and to answer the truth : the Metakernel does not speed up the calc : it takes full advantage of the saturn capabilities ! So it shows what we can do with an HP 48... For example, with a GOOD equation writer, you CAN WORK on equation onto the 48. So : the MK does not speed up the internal EQW, it offers a better one ! The same for the memory management, the graphic environment and so on. BUT I can tell you that the common operations on the stack are well accelerated by the MK, because of the graphic management provided by the MK. And if you want to know what I think about the MK, ... well, it's the most beautiful work ever done on the 48 with JAZZ. For the current avaible version of the MK, ask JY. S.8ebastien ---------------------------------------------------------------------- ==== > First I use the MetakernEl (with an E !) > S.8ebastien Sebastien: You're right, it is MetaKernEl. I never could spell worth a hoot. Your English is excellent for being French. Do you speak and write English yourself or do you use a translation service? I tried using AltaVista's English to French translation so I could write Bernard Parisse in French but when I asked it to translate back it was all garbled. There are so many French speakers in the world of HP I may start to learn French! - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: HP28S - did it come with a case? Does anybody know if the 28S came with a carrying case? I have two of these but I can't remember any case for them. Anybody? Brian ---------------------------------------------------------------------- ==== Subject: Re: HP28S - did it come with a case? Brian >Does anybody know if the 28S came with a carrying case? I have two of these >but I can't remember any case for them. Anybody? >Brian ---------------------------------------------------------------------- ==== Subject: Re: HP28S - did it come with a case? Carrying case? No, not as part of the package, but there is an optional wallet-style (ie, 1/4 flap that tucks into a strap) leather case for the 18C/19B/28S type of folding calculator that comes in burgandy, brown or black (HP part no. 92169). Incidentally, for the IR printer that was introduced the same time as the folding calcs were there is also a wallet-style leather case, also in burgandy, brown or black. Since the 19BII and the IR printer are still made, the above cases are still available from your local retailer, or HP directly. PCN > Does anybody know if the 28S came with a carrying case? I have two of these > but I can't remember any case for them. Anybody? > Brian ---------------------------------------------------------------------- ==== Subject: Re: HP28S - did it come with a case? ] Does anybody know if the 28S came with a carrying case? I have two of these ] but I can't remember any case for them. Anybody? Not as far as I recall. Mine didn't I'm fairly sure. Probably because the book-style hard case didn't really need anything else. Dick -- Dick Smith Acorn Risc PC dick@risctex.demon.co.uk ---------------------------------------------------------------------- ==== Subject: Re: memory management > Note: the ROMs L to R have a bug > you cannot use RCL or EVAL commands on ports 2 and above. > like: if we store ADDTWO in port 2, > and try :2:ADDTWO RCL (or EVAL) it won't work. False; RCL and EVAL work exactly the same for all ports. The FAQ correctly describes a somewhat different bug, which is that if a *directory* is stored in a port variable, and if you try to use a special syntax to recall just one variable from that directory, e.g. :2:{MYDIR MYOBJECT} then RCL and EVAL with such an argument will work only from port 0 or port 1; an Undefined Name error occurs for other ports. BTW, a memory backup (ARCHIVE command) stored in a port is actually stored as a directory; therefore you can recall any individual variable from the backup in this manner, but only if you have stored the backup in port 0 or port 1. ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: New: Now is raton 1.0 in my page, use your hp48 as a mouse Now the link is working and you can download it it's about 8kb btw it needs ufl font library. German http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: jazz question I'm using jazz frequently, for larger and smaller projects, and i'm Very satisfied with it. Yet, i never use the assembler capabilities, because I never write in ML. Isn't there a light version without the assembler components but with the system rpl debugger? The whole thing is consuming all my 128K-card memory as you all can imagine. Would be gratefull for any help. M.Krzewicki. ---------------------------------------------------------------------- ==== Subject: Re: jazz question >I'm using jazz frequently, for larger and smaller projects, and i'm >Very satisfied with it. Yet, i never use the assembler capabilities, >because I never write in ML. Isn't there a light version without the >assembler components but with the system rpl debugger? The whole thing >is consuming all my 128K-card memory as you all can imagine. Would be >gratefull for any help. M.Krzewicki. The source is readily available as well as the tools to build it (i.e. GNU-Tools). So there is nothing stopping you from compiling your own custom version. -- Andre Schoorl PGP key available on request/keyserver Comp Engineering, UVic, Canada Linux/HP48 http://www.engr.uvic.ca/~aschoorl/ ---------------------------------------------------------------------- ==== Subject: Installing ShellOS Who can tell me how to install ShellOS, once I load shell.dir in my HP48GX? ---------------------------------------------------------------------- ==== Subject: Re: Installing ShellOS in that directory should be a program like setup or inst to start installation. Be sure to read the README before installing. Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: Next generation HP? ne reve pas trop .MOoi aussi j ai 1 gx , mais j ai bien peur qu helas il ne faille compter plus que sur ti , surtout que hp a sorti la g+. Si tu connais l assembleur sur hp ,j aurai des question a te poser , alors fais moi signe . ---------------------------------------------------------------------- ==== Subject: Re: Next generation HP? MOUEZA a vomi: > ne reve pas trop .MOoi aussi j ai 1 gx , mais j ai bien peur qu helas il > ne faille compter plus que sur ti Je te signale qu'en ce moment, il y a des mecs qui pr.8eparent la prochaine HP, et qui bossent en Australie, et ce genre de truc est chiant .88 lire. Tu as envie d'une TI, prend un bon pat.8e de vaseline. :-) (avec humour bien s.9er!) -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: Next generation HP? Just port all the stuff on the HP to a StrongARM(200mhz), load it up with the CRC and learn SOH-CAH-TOA Sin-opposite/hypot... > I have an HP 48GX that I bought a year ago and I love it. But, has anyone > out there has heard any rumors of HP putting out a new graphing calculator in > the near future? Correct me if i'm wrong, but it's my understanding that > even the GX came out years ago, maybe as early as '93? The HP Calculator > webpage has no info of such an event on the horizon. Just curious, cuz 4 Mhz JDog the Rocket Scientist Knowledge is Power - Sir Francis Bacon http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: OFF is the topic -----BEGIN PGP SIGNED MESSAGE----- when issuing the OFF command via menu PRG->RUN->OFF, this handy machine is turned off, as expected. And when you re-awake it via ON, the same stack appears, just as it was before. When issuing O F F (the characters, via alpha) and then ENTER the same happens. But: When turning the HP ON again, you can notice that the stack is quickly redisplayed because something is dropped from the stack that seems to have been there only whilst the machine was off. Question: Is this the string 'OFF' itself? If yes, why is it so, as the command has been correctly executed and 'OFF' should therefore be deleted from stack level 1 >before< turning the calc off? Does somebody have the answer? Peter - -- - -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--=- Peter Czaja - student of computer science PGP: BB 9E 5B E0 8F 36 A9 4C E-Mail: pczaja@uni-paderborn.de 13 D2 88 BE 60 44 17 9B -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: noconv iQCVAwUBNnE2kRex1VcOKgL1AQFaNgP/QUJe1vDlnP3kIIkQOrBn0nyvrw8KpGOG oMzPP+sX7Cv81rM0BCkr7p6LpZc51HEtL1tGJyTavwINdFJAUm5MWxTggnx0M9uQ DuXx+YDP7dMRnF7eQanTMTu2lHCrJ5m85PlpgOtiqKQvko0YeE3HX6TpaBLD/Tyw 0NlpTaEOoyw= =rSJe -----END PGP SIGNATURE----- ---------------------------------------------------------------------- ==== Subject: Re: OFF is the topic > -----BEGIN PGP SIGNED MESSAGE----- It's an authentic question, then :) > when issuing the OFF command via menu PRG->RUN->OFF, > this handy machine is turned off, as expected. > And when you re-awake it via ON, > the same stack appears, just as it was before. > When issuing O F F (the characters, via alpha) and then ENTER > the same happens. But: When turning the HP ON again, > you can notice that the stack is quickly redisplayed > because something is dropped from the stack > that seems to have been there only whilst the machine was off. > Question: Is this the string 'OFF' itself? It is the same screen image which was last present at the moment you pressed ENTER, after editing, in which the OFF that you typed still appears; no, it is not a display of anything new on the stack. > If yes, why is it so, as the command has been correctly executed > and 'OFF' should therefore be deleted from stack level 1 > before turning the calc off? The command line is not actually stack level 1, as you can see from the fact that the label (and contents, if any) of level 1 move up to make room for text entry as soon as you begin to type. The command line may thus be considered as an independent text editor; if you abort this editing via CANCEL, you do not find level 1 cleared to zero, as you would have found in older HP calculators with 4-level stacks, because those calcs actually did lift the stack and begin entering digits right into level 1 (x) at once, whereas the HP48 has a completely independent scratchpad for typing new information. If you complete your typing and press ENTER, whatever you typed is compiled into a program and evaluated (executed), which then may influence the contents of the stack; if you abort the edit via CANCEL, the editor goes away and the unchanged stack remains. > Peter Czaja - student of computer science That's the big mistake of computer science -- I mean, what we should be doing is just getting things done, not figuring out how, right? :) As a very small kid, I scientifically investigated a light bulb socket (a live one) by sticking in my finger, and the resulting jolt of knowledge kept me away from any further interest in knowledge of Electrical Engineering for a long while :) It wasn't the same with apple trees, though... But back to the subject: Now, the HP48 has only one processor; this same processor is used both for system tasks (including creating graphics on the screen) and for running user programs; therefore it is not possible for the HP48 to continuously display the intermediate contents of the stack while it is busy using that same stack for everything else, including all the intermediate processing for every user function, and even for making the very graphics which display on the screen itself. Therefore, the HP48 does not redisplay the stack until all programs have completed executing. Some user commands (e.g. DISP, CLLCD, ->LCD, etc.) immediately write to the screen, but stack contents are formatted and displayed only in that part of the system outer loop which executes just before waiting for you to press more keys, and then only if there are no keypresses already accumulated in the type-ahead key buffer, in which case preference is given to more computing, rather than to displaying intermediate results; this saves time and makes the calc more efficient. As to the OFF command, this command turns off the calculator immediately, even if the OFF command appears right in the middle of a program. The next time the calculator is turned on, the program containing OFF wakes right up and continues, immediately following OFF. Thus OFF is like any other function or command, in that the system remembers exactly who called it, and returns back to the caller; sometimes the calling program is the operating system itself, but it can also be in your own program, which will simply pause, like Sleeping Beauty, awaiting some prince to turn her on again later (whatever that means). I hope this post wasn't too OFF-topic, but if it was, then it's too late to back off now. ------------------------------------------------- It's over, and can't be helped, and that's one consolation, as they always say in Turkey, when they cut the wrong man's head off - Charles Dickens, The Pickwick Papers (as quoted by famed spam-canceler Richard E. Depew) ----------------------------------------------------------- With best wishes from: John H Meyers ---------------------------------------------------------------------- ==== Subject: Re: OFF is the topic if you look closely at the display when turning on again, you'll see that the letters 'OFF' are still there for a moment. The calc has gotten the signal to redraw the display, but this will be done afterwards. As usual, the display update has a lower priority than command execution. Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: SEARCH DESESPARETELY ROMDUMP ?????? CAN I HAVE A ROM FROM A HP 48G ???? PLEASE THANKS ---------------------------------------------------------------------- ==== Subject: Re: SEARCH DESESPARETELY ROMDUMP ?????? > CAN I HAVE A ROM FROM A HP 48G ???? PLEASE No. No one will give you one. It's copyrighted material. If you want to use an emulator, you will have to upload YOUR Hp's ROM to your computer, and use it there. If you need help to upload YOUR Rom, ask. If you don't have an HP and simply want to use one freely on computer, answer is no. :-)) Have a nice day dear HP fan! -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: SEARCH DESESPARETELY ROMDUMP ?????? If you have the HP, then figure out how to do the dump yourself. I thought Kermit was a frog yet I was able to figure out how to do it manually in 5 minutes. ---------------------------------------------------------------------- ==== Subject: Re: SEARCH DESESPARETELY ROMDUMP ?????? If you send me $60 I will send you a copy of the ROM (installed in a used HP48G calculator). >CAN I HAVE A ROM FROM A HP 48G ???? PLEASE THANKS > ---------------------------------------------------------------------- ==== Subject: Re: SEARCH DESESPARETELY ROMDUMP ?????? Good one ... Ha! >If you send me $60 I will send you a copy of the ROM (installed in a used >HP48G calculator). >CAN I HAVE A ROM FROM A HP 48G ???? PLEASE >>THANKS > > ---------------------------------------------------------------------- ==== Subject: Translate for me! ==== Subject: Re: Symbolic capabilities of HP48 and TI89 Fri, 11 Dec 1998 16:02:20 +0100 (MET) Jeremy Laughery comp.sys.hp48 1 , 2 Can someone please translate for me? Ecrivez moi si vous connaissez l assembleur sur hp ,merci ---------------------------------------------------------------------- ==== Subject: Re: Translate for me! if you mean this sentence: > Ecrivez moi si vous connaissez l assembleur sur hp ,merci Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: This too ne reve pas trop .MOoi aussi j ai 1 gx , mais j ai bien peur qu helas il ne faille compter plus que sur ti , surtout que hp a sorti la g+. Si tu connais l assembleur sur hp ,j aurai des question a te poser , alors fais moi signe . ---------------------------------------------------------------------- ==== Subject: Re: This too this one says: > ne reve pas trop .MOoi aussi j ai 1 gx , mais j ai bien peur qu helas il > ne faille compter plus que sur ti , surtout que hp a sorti la g+. Si tu > connais l assembleur sur hp ,j aurai des question a te poser , alors > fais > moi signe . Don't dream so much. I've a GX, too, but I'm afraid that you should not count more on it than on the ti, even that hp has brought the G+ If you know an assembler for the hp, I will have some questions to you, please stay in touch.. This translation is NOT exact, but you should get the meaning;-) Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== speedbr. I installed Browse lib and the config lib for it in port 6. I done a reboot and went to check for the libraries and only the cofig lib was there with nothing in it. So, suspecting that the author didn't add the visual commands before compiling the library, I tried to create a directory initialize when I reboot my HP. Instead, the HP goes into an infinite loop until I take the card out and do though. It may be the HP that is nuked! Does anyone know how to fix this problem? I found one solution which should have already been a solution. Never try out software on the HP all! programs that I download from the internet. ---------------------------------------------------------------------- ==== JL> speedbr. I installed JL> Browse lib and the config lib for it in port 6. I done a reboot JL> and went to check for the libraries and only the cofig lib was JL> there with nothing in it. So, suspecting that the author didn't JL> add the visual commands before compiling the library, I tried to JL> create a directory from the lib with Mika's hack library. My 1 JL> the HP goes into an infinite loop until I take the card out and do JL> good though. It may be the HP that is nuked! Does anyone know JL> how to fix this problem? Get a config aborter, pull out the card, install the aborter, activate it, put the card back in, delete speedbrowser, deactivate aborter, warmstart. cu Balazs Fischer Balazs.Fischer@studbox.uni-stuttgart.de PGP Key: http://pgp5.ai.mit.edu:11371/pks/lookup?op=get&search=0xED7447E5 Fingerprint: 8F EF C0 BA 22 B4 15 B3 56 27 07 04 37 7B D6 00 Encrypted mail preferred ... Just adding features because 'it won't hurt' - does. ---------------------------------------------------------------------- ==== the SpeedBrowser itself should not be decompiled, not because of the copyright, but if you recompile it, it may not work since it uses fixed ROMPTR's to be compatible with the built-in browser. Another point: I'm going to speed up FINDVAR... Raymond BTW: SpeedBrowser works fine in EMU48 w/CF.LIB in port 0. Even the FormMaster package works, but EMU48 is so fast that you don't need FormMaster there;-) Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== the CF.LIB has no visual command in it, and it is not splittable with JAZZ. Be sure to the CF.LIB in port 0 and the SpeedBrowservin port 1. The SpeedBrowser is 100% bug-free, and the CF.LIB is not splittable. Sorry, I should have mentionned this;-) The only problem with CF.LIB is that it may not be able to configure SpeedBrowser in any port combination, but I'm working on this... Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Best softwares See the latest softwares (games such as Monopoli, tools such as PCT 5.7) on my site: http://perso.cybercable.fr/harold ---------------------------------------------------------------------- ==== Subject: hp48 screen I have several hp48g's and s's that my students have had broken when loaning their calculators out. The only busted part (before tearing into it) is the LCD. I think that the part is manufactured by Epson and I found similar replacement items on Epson's old web site, but the new one doesn't include any similar items. Does anybody know where I can get a replacement LCD? http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: hp48 screen We've got some dead HP48's with good screens. Just e-mail your request to dead48@smi.com with your address. Phil Trent I have several hp48g's and s's that my students have had broken when loaning > their calculators out. The only busted part (before tearing into it) is the > LCD. I think that the part is manufactured by Epson and I found similar > replacement items on Epson's old web site, but the new one doesn't include > any similar items. Does anybody know where I can get a replacement LCD? http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Where to Get the HP48 FAQ X-Url: http://www.engr.uvic.ca/~aschoorl/faq/ -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Where to Get the HP48 FAQ This document briefly describes where to get the latest version of the comp.sys.hp48 FAQ. The main site is: http://www.engr.uvic.ca/~aschoorl/faq/ All formats of the FAQ are available at the above link, including the text, PostScript, HTML, and SGML versions. Text Postings Text only versions of the FAQ will be posted to the newsgroup comp.sys.hp48 and comp.sources.hp48 as needed (usually every two weeks). Also, remember that all official FAQs (including this one) are mirrored at rtfm.mit.edu. Specifically, this one is at: ftp://rtfm.mit.edu/pub/usenet-by-group/comp.sys.hp48/ Furthermore, the text FAQ is posted to comp.answers and news.answers. I will sign all text versions with PGP (Pretty Good Privacy) using the following key for authenticity, and provide MD5 Checksums for the remaining files. You can get a copy of this public key through the URL above. Type Bits KeyID Created Expires Algorithm Use sec+ 1024 2CFAA0BB 1997-06-21 ---------- DSS Sign and Encrypt sub 2048 F940E148 1997-06-21 ---------- Diffie-Hellman uid Andre P. Schoorl Alternate Sites Alternatively, you can download the FAQ at one the following locations: North America: -------------- Latest: ftp://ftp.cis.com/pub/hp48g/uploads/ Older: ftp://hpcvbbs.cv.hp.com/dist/hp48g/docs/ ftp://ftp.cis.com/pub/hp48g/documentation/ Europe: ------- Latest: ftp://ftp.stud.fh-heilbronn.de/pub/systems/hp48/incoming/ Older: ftp://ftp.stud.fh-heilbronn.de/pub/systems/hp48/info/FAQ/ Translations: ------------- Espanol (Spanish): http://www.alumnos.utfsm.cl/~aarrieta/hp48.html Portugues (Portugese): http://members.tripod.com/~area48/ Contacting Me I rely on your input in order to keep this FAQ up to date. If you have any suggestions or updates, feel free to mail them to me. My current address is: aschoorl@engr.uvic.ca If you're using a web enabled news reader, you can click here: mailto:aschoorl@engr.uvic.ca -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 5.0i for non-commercial use Charset: noconv iQA/AwUBNgQVJ+hzXOws+qC7EQLPPgCgiNAYT2b921GdNg7pGqtgKFQ1RRIAoPyG ZJjwHZ6Tv9/FH+GzqmijC0Sw =bBXh -----END PGP SIGNATURE----- -- Andre Schoorl PGP key available on request/keyserver Comp Engineering, UVic, Canada Linux/HP48 http://www.engr.uvic.ca/~aschoorl/ ---------------------------------------------------------------------- ==== Subject: Minimum ROM version for MetaKernel? I downloaded MetaKernel (test version), just to see what it is, and I tried it on Emu48, with my ROM (my HP is a 48GX-M)... Then I looked at the docs (btw, WHO has written such a horrible doc !?! Even in french, there's a mistake every 3 or 4 words!!) and saw that the version required for it to work was the ROM version R... But it seems to work fine for me... What function won't work on my Rev-M ROM? -- Erwann ABALEA eabalea@certplus.com ---------------------------------------------------------------------- ==== Subject: Re: Minimum ROM version for MetaKernel? > I downloaded MetaKernel (test version), just to see what it is, and I > tried it on Emu48, with my ROM (my HP is a 48GX-M)... Then I looked at the docs (btw, WHO has written such a horrible doc !?! > Even in french, there's a mistake every 3 or 4 words!!) and saw that the > version required for it to work was the ROM version R... -- > Erwann ABALEA Erwann: Go ahead and buy the ROM card version! The MetaKernal on ROM card is a terrific program. It speeds up the calculator by a factor of at least 3-5 times. It has a pretty print program for displaying algebraics, and a really great file management program, and the Equation Writer works at lightning speed. The manual included with the card is good. I love my MetaKernal. You might want to dump your ROM version M for a ROM version R, though. - J.C. Ron http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: COLOR LCD Hey HP48 enthusiasts! See this page and imagine how an HP48 color LCD would looks like. http://www.geocities.com/~hpfb/hpcolor.html CMarangon http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: COLOR LCD Dream on! Brian >Hey HP48 enthusiasts! See this page and imagine how an HP48 color LCD would looks like. http://www.geocities.com/~hpfb/hpcolor.html CMarangon http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: COLOR LCD > Hey HP48 enthusiasts! See this page and imagine how an HP48 color LCD would looks like. http://www.geocities.com/~hpfb/hpcolor.html CMarangon Oh God! It looks like DOS! Your page has brought back some of my worst Seriously though, I'd like to see a color HP as much as the next guy. It's kinda cool to see what the interface might look like. -- Brian Stone stone@StoneEntertainment.com ---------------------------------------------------------------------- ==== Subject: Re: COLOR LCD (Looking at http://www.geocities.com/~hpfb/hpcolor.html ) Ouch!!!! My eyes!!! Luckily, however, the batteries would soon die and your eyes would return to normal. Last time I checked, my math book has equations in B&W and I can understand them well, and CRC standard mathematical reference was in B&W. One of my math teachers only wears gray, and he looks fine. Math is best in B&W. Save your money for a Color Gameboy. >Hey HP48 enthusiasts! See this page and imagine how an HP48 color LCD would looks like. http://www.geocities.com/~hpfb/hpcolor.html CMarangon http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: D->LIB does Invalid Card Data > You might get http://www.provi.de/~detlef/files/48D2LIB.ZIP -- It contains > the (sRPL) source of a short library builder. The purpose of that program > was exactly what you're asking for -- to outline the structure of the > LIB object. Tried it twice. First time, Invalid Card Data and Warmrestart. Port 0 contents lost (nothing there except HP Wizard's Debugger, but I have a copy of it on my 512 Kb card). Second time, HP seems to turn off. After 5 minutes, I press ON. Port 0 lost (D->LIB was there) but flags and HOME still there. Doesn't seem to work on ROM revision R, unless someone knows what's really happening... An idea, Detlef? -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: lib structure >You might get http://www.provi.de/~detlef/files/48D2LIB.ZIP -- It >contains the (sRPL) source of a short library builder. The purpose of >that program was exactly what you're asking for -- to outline the >structure of the LIB object. I would prefer a ML source, I got nothing about sysRPL, and I'm lost inside it. I keep reading commands I don't know, and it's annoying. The old good books here have vanished into oblivion, and I found myself without book reference. I'll have a look at it anway :) -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: on peut faire une comparaison sur 4 quartet? Si je veut faire .8da: ? 4 Je suis oblig.8e de passer par WP? Je sais que j'ai des trucs du type r=DATn 4 mais pour les comparaisons....? (j'ai fil.8e mon bouquin .88 un d.8ebutant .88 Maubert, et le seul qui me reste Mer.8di :)) -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? if the fifth nibble does not care, you could even use comparisons of the type ?r1_r2 A because comparisons on the A field are somewhat faster, as far as I know;-) This is because the CPU is optimized for 'A' field operations (This info is taken from SASM.DOC again) Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? you are right when you say that ?A=B WP (for example) takes one cycle less than ?A=B A (if P set to 3), but you forgot that you have to set P= 3 (2 cyc.) before and P= 0 (2 cyc.) after the test, so you end up with more cycles if you use the ...WP version than the ...A version, and the code will be larger ;-) BTW: I learned Saturn assembly language ~15 years ago when the HP71B came out, so I think the descriptions you gave were for the HP48 community . Don't get me wrong, I'm not calling myself an HP god or somewhat like that, I'd just say some way experienced;-) I'd like to encourage you to make your postings in english, it's not my native language, too, but it may be the least common divisor (hope that was the right expression:-) Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? > Larger code is sometimes (most often) faster. I don't think so. There are many occasions where you can optimize both code size and speed. You have to do a bit more planning, but since it's just for fun, it may be worth it. A good example is using hybrid code on the HP48, take for example my flag browser or my charset browser. The 'time critical' routines like display are in ML, of course, but I use the POL utilities for the user interface handling because it makes no sense to reinvent the wheel, it saves many bytes, is compatible to the standard HP48 user interface (very important) and is fast. Even FormMaster uses a version of the POL utilities, and if it were faster, people would haveproblems to use it;-) The things mentioned above are examples of optimized code for speed and size. Of course it's faster to do 100 multiplications one after the other w/o a loop around it, you could save the loop management overhead, but I think that's not the point... (POL stands for ParOuterLoop) Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germanynet.de ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? x-HP-Fanatic: yes > you are right when you say that ?A=B WP (for example) > takes one cycle less than ?A=B A (if P set to 3), > but you forgot that you have to set P= 3 (2 cyc.) before and > P= 0 (2 cyc.) after the test, Those two assignments are outside the loop, in which I use WP, so that's > so you end up with more cycles if you use the ...WP > version than the ...A version, and the code will be larger ;-) Larger code is sometimes (most often) faster. > Don't get me wrong, I'm not calling myself an HP god or somewhat like > that, I'd just say some way experienced;-) :)) -Gilb- (andrell@iname.com) ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? >Larger code is sometimes (most often) faster. No, larger code is sometimes faster. But when properly written, faster code is usually smaller. dan ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? x-HP-Fanatic: yes > because comparisons on the A field are somewhat faster, as > far as I know;-) ? My test is ?A#C over 4 nibbles If we consider even addresses (or just plain odd, I don't remember) we got: ?A#C f if A is used, we got 13+f cycles if the test is verified otherwise 6+f cycles A is 5 wide, so it's 18 cycles if test verified, otherwise 11 cycles Using WP with P=3 (thus WP being 4 nibbles wide) we get: 17 cycles if test verified 10 cycles otherwise So, unless my book is wrong (I've already seen books give wrong info over Hp's stuff) I will use WP. Since P can remain at 3 inside the look, there's no problem. I save 1 processor cycle per loop. > This is because the CPU is optimized for 'A' field operations > (This info is taken from SASM.DOC again) The instructions are not coded the same way for A field and other ones. You got for almost every instruction a code for the A field, and then a second one will a nibble to fill; like that: if we consider the instruction ASRB.F, it's coded: 819F0 when using field A. If we use another field, it becomes: 819a0 where you must put in nibble a a value, according to the field concerned, using this table: Register Value of a P 0 WP 1 XS 2 X 3 S 4 M 5 B 6 W 7 Here is an another example: instruction ?B#0 f; it's coded: 8AD for field A followed by a signed two-nibble value; but if you want to use an another field, it's 9aD (with still a two signed nibble value afterwards). The processor cycles remain the same, 6+n if test not verified to 13+n for verified tests (n being the number of nibbles) for most tests. The only example of a real difference I remember is loading into C or A instructions: LA 6+n cycles LC 4+n cycles n being number of nibbles -Gilb- (andrell@iname.com) ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? if I get your question right, you wanted to ask if you have to use field specifiers or WP in comparisons, whereas you can give a number of nibbles in memory access functions ? As you can see, yes! The doc (yes you said you don't like docs, but..) SASM.DOC says it all;-) Raymond BTW: Have you noticed that nearly all replies were in english, even those from french people ? Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? > if I get your question right, you wanted to ask if > you have to use field specifiers or WP in comparisons, > whereas you can give a number of nibbles in memory access > functions ? As you can see, yes! Doesn't work for 4 Jazz doesn't accept: ?C#A 4 I have to P=3, test using WP, and then P=0 twice (on each part after the test). > The doc (yes you said you don't like docs, but..) SASM.DOC > says it all;-) I don't have a PC, so I can't use SASM, and I don't like searching in english files information, it's a annoying tongue, especially when it comes to technical information. > BTW: Have you noticed that nearly all replies were in english, > even those from french people ? Answer in english if you prefer. I have learned english, and if I've learned something I disklike, people can get too disgustings moments learning french. I prefer french. Sometimes I feel writing english, most time I don't. -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: on peut faire une comparaison sur 4 quartet? The best way to compare two values stored on 4 nibbles is by doing something like: A=0 A C=0 A A=DAT1 4 C=DAT0 4 ?A#C A And that's it. Jean-Yves > if I get your question right, you wanted to ask if >> you have to use field specifiers or WP in comparisons, >> whereas you can give a number of nibbles in memory access >> functions ? >> As you can see, yes! Doesn't work for 4 Jazz doesn't accept: ?C#A 4 I have to P=3, test using WP, and then P=0 twice (on each part after the >test). > The doc (yes you said you don't like docs, but..) SASM.DOC >> says it all;-) I don't have a PC, so I can't use SASM, and I don't like searching in >english files information, it's a annoying tongue, especially when it >comes to technical information. > BTW: Have you noticed that nearly all replies were in english, >> even those from french people ? Answer in english if you prefer. I have learned english, and if I've >learned something I disklike, people can get too disgustings moments >learning french. I prefer french. Sometimes I feel writing english, most time I don't. -Gilb- >43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: HP48GX Rom Versions How do I determine the ROM version in my GX and what is the latest version? Ed Yee -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Ed Yee University of Calgary - Faculty of Electrical Engineering E-Mail: eyee@acs.ucalgary.ca yeee@enel.ucalgary.ca Mechanical engineers build weapons Civil engineers build targets. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ---------------------------------------------------------------------- ==== Subject: Re: HP48GX Rom Versions On Sat, 12 Dec 1998 02:13:32 -0800, Ed Yee version? type VERSION to see the Rom version. R is the latest. Further info is hided in the FAQ :-) Peter E-Mail: karpfenteich@gmx.de _______________________________ Do you know the great Frequently Asked Questions? http://www.engr.uvic.ca/~aschoorl/faq/ and the superb HP48 Software Archive? http://www.hpcalc.org to look for *old* HP48 postings see http://www.dejanews.com ---------------------------------------------------------------------- ==== Subject: Re: HP48GX Rom Versions > How do I determine the ROM version in my GX and what is the latest > version? Just enter the VERSION command, and you'll have 2 strings on your stack. I don't know what is the latest. -- Erwann ABALEA eabalea@certplus.com ---------------------------------------------------------------------- ==== Subject: Help ! Help ! Please answer fast !! i have a hp48g with 768k ram on board ... got the device some days ago No i need to enter some sentences + formula that i can use in a university test :-) where is the option to store text and how does it work ? Please answer by mail only thanx in advance, Markus ---------------------------------------------------------------------- ==== Subject: Re: Help ! Help ! Please answer fast !! RTFM Brian <36725F35.A171E0D6@stud-mailer.uni-marburg.de>... >i have a hp48g with 768k ram on board ... got the device some days ago >No i need to enter some sentences + formula that i can use in a >university test :-) where is the option to store text and how does it >work ? >Please answer by mail only thanx in advance, Markus > ---------------------------------------------------------------------- ==== Subject: re: Help I also have the xfer cable to store a text edited in the pc in the hp ... but i don't know how to read the text i stored ... ---------------------------------------------------------------------- ==== Subject: ROM How can I have the ROM ,can you give it? ---------------------------------------------------------------------- ==== Subject: Re: ROM one or more people have (or know how to get) the upload software for the ROM, so please try that software. Look at that site: http://www.hpcalc.org for the upload software. If you have an HP48, it should be no problem to do so;-) Raymond BTW: Uploading the ROM from your HP48 to your PC only takes a few minutes, so why use an illegal copy ? Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: ROM send me your rom please at sbaudry@caramail.com ---------------------------------------------------------------------- ==== Subject: Re: ROM Every few days the same questions and proposals . Are you too lazy to upload your own ROM ? Sorry for the hard words, but as you might know, this kind of postings ('Gimme a ROM', 'Where can I get the ROM', etc.) are boring;-) Many postings from different authors have been made in the last few months regarding this theme, and Jean-Yves Avenard and others provided solutions in the form of easy-to-use upload programs, so there's no need to act against any copyrights. Hope this helps Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: ROM > send me your rom please at sbaudry@caramail.com Please don't ask for such stupid things... It is absolutely forbidden to send the ROMs, since they are copyrighted. Sorry for the obfuscating text below, if you want a translation, don't hesitate to ask... [French On] Les francais se font vraiment mal remarquer dans ce ng... entre ceux qui n'ecrivent qu'en francais et ceux qui posent des questions idiotes, il ne reste pas beaucoup de place pour ceux qui peuvent remonter le niveau, comme JY Avenard, et d'autres dont je ne connais pas le nom... [French Off] -- Erwann ABALEA eabalea@certplus.com ---------------------------------------------------------------------- ==== Subject: Re: ROM Erwann a .8ecrit: > Les francais se font vraiment mal remarquer dans ce ng... entre ceux qui > n'ecrivent qu'en francais Si j'ai envie d'.8ecrire en fran.8dais, je le fais. Je suis libre de poster dans la langue que je veux. Et tu toucheras pas .88 ma libert.8e, ni personne d'autre. J'.8epelle le mot: L I B E R T E Si tu l'.8ecris lettre par lettre, tu devrais y arriver. Me dire que je vais mal me comporter parce que je posterait dans ma langue plut.99t qu'en anglais est innaceptable. Je sais que je ne touche pas tout le monde quand je poste en fran.8dais, et souvent je fais l'effort d'utiliser l'anglais, comme d'autres font l'effort d'utiliser un fran.8dais qu'ils disent mauvais, et que je trouve bon, et encourageant. L'anglais ne me pose pas de probl.8fme sur le plan de son utilisation, mais je ne supporte pas qu'on d.8epr.8ecie une langue pour une autre, et si Internet doit .90tre 100 % anglais, je dis merde, et j'utilise la langue que je veux, et j'invite tout le monde .88 faire pareil. .8da commence .88 me gonfler d'aller sur Internet pour me retrouver avec des r.8fgles et des usages qui sont l'antinomie de la pr.8etendue libert.8e du r.8eseau. Je dois sembler un peu dur, mais qu'on touche pas .88 ma *libert.8e* d'*expression*, ni sur le net, ni ailleurs, et cela comprend ma langue propre, comme pour *chacun* d'entre vous. Libert.8e, libert.8e ch.8erie! -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: ROM Sorry, but this time I will answer in French. Because actually I'm ashame of some of my compatriot. La liberte des uns finis la ou commence celle des autres. Ce Newsgroup est un newgroup international. La langue utilisee est l'anglais. Tu peux continuer a poster en Francais, mais a partir de maintenant, je n'y repondrai plus. Ce qui va grandement limiter l'etendue des reponses. C'est une question de politesse. Si tu veux ecrire en francais, alors utilise les newsgroup francais. Je peux compr un message ecris en francais pour des cas precis comme une demande d'achat ou de vente sur la France. Pour tous les sujets generaux, par respect pour ceux qui peuvent nous lire on se doit de faire l'effort d'ecrire et de repondre en anglais. Il n'y a rin de plus rude que d'avoir dans un groupe anglophone, des francais se sentant superieurs a tous les autres et parlant a la barbe de tout le monde en francais. Ce qui est etrange c'est que je n'ai vu cela qu'avec des Francais et jamais d'autre nationalite (allemand, suedois etc...) il faut croire qu'ils ont recu une meilleur education que nous. Jean-Yves Les francais se font vraiment mal remarquer dans ce ng... entre ceux qui >> n'ecrivent qu'en francais Si j'ai envie d'.8ecrire en fran.8dais, je le fais. Je suis libre de poster dans la langue que je veux. Et tu toucheras pas >.88 ma libert.8e, ni personne d'autre. J'.8epelle le mot: L I B E R T E Si tu l'.8ecris lettre par lettre, tu devrais y arriver. Me dire que je vais mal me comporter parce que je posterait dans ma >langue plut.99t qu'en anglais est innaceptable. Je sais que je ne touche pas tout le monde quand je poste en fran.8dais, >et souvent je fais l'effort d'utiliser l'anglais, comme d'autres font >l'effort d'utiliser un fran.8dais qu'ils disent mauvais, et que je trouve >bon, et encourageant. L'anglais ne me pose pas de probl.8fme sur le plan de son utilisation, >mais je ne supporte pas qu'on d.8epr.8ecie une langue pour une autre, et si >Internet doit .90tre 100 % anglais, je dis merde, et j'utilise la langue >que je veux, et j'invite tout le monde .88 faire pareil. .8da commence .88 me gonfler d'aller sur Internet pour me retrouver avec des >r.8fgles et des usages qui sont l'antinomie de la pr.8etendue libert.8e du >r.8eseau. Je dois sembler un peu dur, mais qu'on touche pas .88 ma *libert.8e* >d'*expression*, ni sur le net, ni ailleurs, et cela comprend ma langue >propre, comme pour *chacun* d'entre vous. Libert.8e, libert.8e ch.8erie! -Gilb- >43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: ROM I Totally agree. ---------------------------------------------------------------------- ==== Subject: Re: ROM Who won? Brian > I Totally agree. > ---------------------------------------------------------------------- ==== Subject: Re: ROM x-HP-Fanatic: yes > La liberte des uns finis la ou commence celle des autres. J'ai donc pour libert.8e de choisir ma langue d'expression. > Ce Newsgroup est un newgroup international. La langue utilisee est > l'anglais. Je ne consid.8fre pas l'anglais comme langue internationale, d'autant qu'il ne s'agit PAS de la langue la plus parl.8ee dans le monde (surprise? cherche un peu et tu d.8ecouvriras que l'anglais ne l'est pas). > C'est une question de politesse. Non, une question de droit d'expression dans la langue de son choix. Je veux bien faire un effort pour que tu n'aies plus honte de moi, mais je pr.8ef.8ere encore qu'on me lise moins en fran.8dais, qu'on se foute de mon anglais minable. > Pour tous les sujets generaux, par respect pour ceux qui peuvent nous lire > on se doit de faire l'effort d'ecrire et de repondre en anglais. Et pourquoi on ne leur demande pas, par respect d'utiliser une autre langue que l'anglais? On peut se renvoyer l'argument, cher Jean-Yves. Je pourrait demander par respect que tous ceux que je ne peux pas lire en anglais me r.8epondent dans une autre langue. > Il n'y a rin de plus rude que d'avoir dans un groupe anglophone, des > francais se sentant superieurs a tous les autres et parlant a la barbe de > tout le monde en francais. Ecoute, si le groupe est r.8eserv.8e .88 la langue anglais, alors il faut le mod.8erer, et interdire de fait tout envoie non-anglais. Quelle jolie image on aura alors.. n'est-ce pas? J'aime beaucoup les anglais. J'ai pas envie qu'on m'impose une langue, et l'anglais n'.8etant pas la langue la plus parl.8ee dans le monde, elle n'est pas plus internationale que l'espagnol, ou une autre. Je pense pas qu'on soit sup.8erieurs, mais on a cette envie de partager tout ce qu'on sait, et d'emmener les autres avec soi dans le m.90me voyage. Je m'en fous un peu de programmer des trucs super tip-top, je pr.8ef.8ere encore passer mon temps .88 appr .88 d'autres .88 le faire, je trouve cela bien plus gratifiant, de rester dans l'ombre, tranquille. > Ce qui est etrange c'est que je n'ai vu cela qu'avec des Francais et >jamais d'autre nationalite (allemand, suedois etc...) il faut croire >qu'ils ont recu une meilleur education que nous. Cela est .8evident, nous sommes idiots et nombrilistes. Enfin, on peut se trouver tous les d.8efauts de la terre. Anglais, je vous aime!!! :-) (mais laissez moi respirer en fran.8dais, voil.88 tout) -Gilb- (andrell@iname.com) ---------------------------------------------------------------------- ==== Subject: hp user's library I'm curious, what ever happened to the HP User's Library? There was a wealth of programs for the HP 41 and 71. I still use my 41 in addition to the 48, and I was wondering if the 41 programs are still available anywhere? V ---------------------------------------------------------------------- ==== Subject: Re: hp user's library >I'm curious, what ever happened to the HP User's Library? There was a >wealth of programs for the HP 41 and 71. I still use my 41 in addition to the 48, and I was wondering if the 41 >programs are still available anywhere? > V > It is no more and for a long time. -- ************************************************************************** | civan@eskimo.com ~ CHUCK DINSMORE P.E. ~ Seattle, WA.| | NG Jacobson & Assoc. www.shoshone.com/NGJ/ Seattle, WA.| ************************************************************************** ---------------------------------------------------------------------- ==== Subject: Re: hp user's library please take a look at the following page: http://www.hpcc.org/index.html They may have information about this topic. Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Help with Postscript How can I print a postscipt doc?? -=UssU=- ---------------------------------------------------------------------- ==== Subject: Re: Help with Postscript > How can I print a postscipt doc?? > -=UssU=- Get yourself Ghostscript from http://www.cs.wisc.edu/~ghost/, together with a suited previewer. This works under nearly all circumstands. CU Stefan ---------------------------------------------------------------------- ==== Subject: MetaKernel Spain Does anyone that lives in spain use the MetaKernel? If so, please contact with me. -=UssU=- ---------------------------------------------------------------------- ==== Subject: Converting programs from 41CV to 48GX? I have plenty good programs for the 41 CV and just bought a 48 GX. Is there any application to convert the 41 CV programs to the 48 GX or is it necesaary to write them all?? Fidel eta@uio.telconet.net ---------------------------------------------------------------------- ==== Subject: Re: Converting programs from 41CV to 48GX? At one time there was a card for converting 41 programs to 48. However, I posted a month or so ago inquiring as to a source for this card. No response. ---------------------------------------------------------------------- ==== Subject: For people with problems with Java 3.6 Keywords: Java Stack FASTOFF For those of you experiencing seemingly random crashes with Java 3.6, particularly when power cycling and then pressing menu keys, please try the following: << OFF >> 91.3 ASN I believe it may be the new FASTOFF that is causing problems. Please e-mail me directly with your successes / failures with the above technique. I may be removing it and making a 3.6a, and at the same time including some other small fixes. Even if it is not the FASTOFF, I want to remove it anyways as something like it doesn't really belong in Java, but should be in an external library for better modularity. I'm experimenting with a Java Helper library with common user keys right now which will be designed for 3.7 beta's and beyond. -- Andre Schoorl PGP key available on request/keyserver Comp Engineering, UVic, Canada Linux/HP48 http://www.engr.uvic.ca/~aschoorl/ ---------------------------------------------------------------------- ==== Subject: FOR SALE: SPICE48 CARD BY SPARCOM, BEST OFFER? ---------------------------------------------------------------------- ==== Subject: le drop qui =?ISO-8859-1?Q?dispara=EEt=2E=2E=2E?= :P Salut, un de mes programmes me sert .88 tester le contenu de la pile; en fait, je place un entier syst.8fme sur la pile, qui correspond au prologue de l'objet que je veux. Exemple: je veux tester une cha.94ne, j'ai: PRG <02A2Ch> Code END L'objet code commence par lire l'entier syst.8fme, et le retire de la pile: A=DAT1.A AD1EX D1+5 C=DAT1.A D1=A D1+5 D+1.A Remarquez, je vire l'objet avec un tit drop. Ensuite, un test tout simple: ais-je maintenant un objet au niveau 1 de la pile (pr.8ec.8edemment le niveau 2), si oui je v.8erifie que son prologue correspond .88 l'entier syst.8fme, etc. Bon, faisons simple. J'ai fait un drop sur l'entier syst.8fme. Quand je fais ceci: A=DAT1.A ?A=0.A { LA 00201 ErrorJump } Quand l'erreur se produit, .88 ma surprise (enfin, pas tellement) je me retrouve avec mon gentil entier syst.8fme. Je fais alors un tit test: D1+5 D+1.A LA 00201 ErrorJump Et bien entendu, m.90me avec une pile vide, tout roule :-) Question: comment puis-je lire une valeur sur la pile (1) .8eventuellement provoquer une erreur par ErrorJump (2) (j'utilise ErrorJump pour afficher .8eventuellement des messages d'erreur personnalis.8es, dans la table $MESSAGE de ma librairie, o.9d sera plac.8e ce code, ainsi que d'autres dont je vais taire la fonction ;-) ErrorJump EQU #04ED1 J'ai franchement pas l'envie d'aller d.8esassembler ErrorJump, je regarde notre ami Jarod, et ma copine .88 cot.8e .88 cach.8e ma HP je sais pas o.9d.. :o) -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: Re: le drop qui dispara.94t... :P Hello When an error occurs, the HP48 error handler restore the stack in his previous state. Since you didn't save the stack after you've droped you system binary. It still appears. Jean-Yves >Salut, un de mes programmes me sert .88 tester le contenu de la pile; en fait, je >place un entier syst.8fme sur la pile, qui correspond au prologue de >l'objet que je veux. Exemple: je veux tester une cha.94ne, j'ai: PRG > <02A2Ch Code >END L'objet code commence par lire l'entier syst.8fme, et le retire de la >pile: A=DAT1.A > AD1EX > D1+5 > C=DAT1.A > D1=A > D1+5 > D+1.A Remarquez, je vire l'objet avec un tit drop. Ensuite, un test tout simple: ais-je maintenant un objet au niveau 1 de >la pile (pr.8ec.8edemment le niveau 2), si oui je v.8erifie que son prologue >correspond .88 l'entier syst.8fme, etc. Bon, faisons simple. J'ai fait un drop sur l'entier syst.8fme. Quand je >fais ceci: A=DAT1.A > ?A=0.A { > LA 00201 > ErrorJump > } Quand l'erreur se produit, .88 ma surprise (enfin, pas tellement) je me >retrouve avec mon gentil entier syst.8fme. Je fais alors un tit test: D1+5 > D+1.A > LA 00201 > ErrorJump Et bien entendu, m.90me avec une pile vide, tout roule :-) Question: comment puis-je lire une valeur sur la pile (1) >.8eventuellement provoquer une erreur par ErrorJump (2) (j'utilise ErrorJump pour afficher .8eventuellement des messages d'erreur >personnalis.8es, dans la table $MESSAGE de ma librairie, o.9d sera plac.8e ce >code, ainsi que d'autres dont je vais taire la fonction ;-) ErrorJump EQU #04ED1 J'ai franchement pas l'envie d'aller d.8esassembler ErrorJump, je regarde >notre ami Jarod, et ma copine .88 cot.8e .88 cach.8e ma HP je sais pas o.9d.. :o) -Gilb- >43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: offer HP38 complete for used 48SX (even empty, but I want the whole keyboard-lower HP box) I want to move my GX to the enveloppe of a SX, so I'll give my HP38 (french doc, bought less than one month ago, frozen-mint condition) for a S plastic enveloppe. Don't ask why. I just have the GX overlay colors, and S/SX colors are nicer, so I'll move my GX inners inside :-) Email to: andrell@iname.com -Gilb- 43E2 61F7 4ECA 6A7C AAFE FE2F C1B1 FFE6 D7B0 48E3 ---------------------------------------------------------------------- ==== Subject: It would be awesome if... Does anybody know if there is a program that can give the k value of a single slope segment when graphing slope fields. The HP is unfortunately very limited when is comes to getting valuable information out of some of its graphing capabilities. It would be awesome if one could specify a particular point in a slope field. Once a particular point has been specified a particular plot graphed from that slop segment could be plotted. It would be advisable to have this line superimposed in the slope field with its particular equation visible in the corner of the screen for a quick idea how that line flows through the field. This would be extremly helpful in calculus math. Of course one should be able to use TRACE and be able to plot as many lines as needed with out leaving the slope field. If there are no such programs I plan on writing one. If any body as any ideas or wishes please get back to me. All the best and a big hello to all who know me. Eva Hellqvisteva@hotmail.com ---------------------------------------------------------------------- ==== Subject: String to algebraic for Java Hello all! I am needing to find a program to convert a string to an algebraic in advance! Jeremy E. Laughery ---------------------------------------------------------------------- ==== Subject: Re: String to algebraic for Java >Hello all! > I am needing to find a program to convert a string to an algebraic >in advance! Jeremy E. Laughery String to algebraic? How about STR-> no?!? Anyways this gives me the opportunity to announce 3.6a, which I just made available at my site. -- Andre Schoorl PGP key available on request/keyserver Comp Engineering, UVic, Canada Linux/HP48 http://www.engr.uvic.ca/~aschoorl/ ---------------------------------------------------------------------- ==== Subject: StrongARM I'm very angry at HP. They make a wonderful calc(HP48GX) and then foul it up with a 4mhz processor. Damn! Well, I've been looking at this StrongARM from Intel(formerly Digital) and was wondering if anyone would be interested in porting the HP's ROM from the Saturn to the StrongARM. I'm not sure about the legality but if you saw an HP run at 200mhz, you wouldn't care either. Respond with thoughts David Cummings ---------------------------------------------------------------------- ==== Subject: Re: StrongARM > I'm very angry at HP. They make a wonderful calc(HP48GX) and then foul > it up with a 4mhz processor. Damn! Well, I've been looking at this > StrongARM from Intel(formerly Digital) and was wondering if anyone would > be interested in porting the HP's ROM from the Saturn to the StrongARM. > I'm not sure about the legality but if you saw an HP run at 200mhz, you > wouldn't care either. Respond with thoughts > David Cummings I also thought about running an HP emulator on one of those processors. Source code of an emulator is already available in c (x48 for unix or maybe emu48) an thus it should be easily portable. The main problem would be to design a pcb with ram, lcd, keyboard and have some development tools (gnu c cross compiler for StrongARM already exists). The next step would be to write code to initialize and access the hardware (not much: drawing pixels, reading time on a rtc, keyboard, etc), and after that we're ready for the emulation. We can easily fit this pcb with 4 meg of ram or more, and thus we could emulate memory expansion cards of the hp (maybe 16 or more, depending on the size of our kernel/emulator). At 200MHz we should be able to emulate the hp48 at least at 10x it's speed. Great! I really think we could do something here.... Daniel -- EMail: daniel.roggen@epfl.ch ICQ: 2510427 ---------------------------------------------------------------------- ==== Subject: Re: StrongARM What about porting emu48 to wince? > I'm very angry at HP. They make a wonderful calc(HP48GX) and then foul >> it up with a 4mhz processor. Damn! Well, I've been looking at this >> StrongARM from Intel(formerly Digital) and was wondering if anyone would >> be interested in porting the HP's ROM from the Saturn to the StrongARM. >> I'm not sure about the legality but if you saw an HP run at 200mhz, you >> wouldn't care either. Respond with thoughts >> David Cummings I also thought about running an HP emulator on one of those processors. >Source code of an emulator is already available in c (x48 for unix or maybe >emu48) an thus it should be easily portable. The main problem would be to >design a pcb with ram, lcd, keyboard and have some development tools (gnu c >cross compiler for StrongARM already exists). The next step would be to >write code to initialize and access the hardware (not much: drawing pixels, >reading time on a rtc, keyboard, etc), and after that we're ready for the >emulation. We can easily fit this pcb with 4 meg of ram or more, and thus we could >emulate memory expansion cards of the hp (maybe 16 or more, depending on the >size of our kernel/emulator). At 200MHz we should be able to emulate the hp48 at least at 10x it's speed. >Great! I really think we could do something here.... >Daniel -- >EMail: daniel.roggen@epfl.ch ICQ: 2510427 ---------------------------------------------------------------------- ==== Subject: Re: StrongARM > What about porting emu48 to wince? What about running a smaller OS? -- Matthias Andree ---------------------------------------------------------------------- ==== Subject: Re: StrongARM I'm not sure about WinCE. It seems that using a Microsoft product for such a high-powered and reliability oriented task seemes self-destructive. > What about porting emu48 to wince? >> I'm very angry at HP. They make a wonderful calc(HP48GX) and then foul >> it up with a 4mhz processor. Damn! Well, I've been looking at this >> StrongARM from Intel(formerly Digital) and was wondering if anyone would >> be interested in porting the HP's ROM from the Saturn to the StrongARM. >> I'm not sure about the legality but if you saw an HP run at 200mhz, you >> wouldn't care either. Respond with thoughts >> David Cummings I also thought about running an HP emulator on one of those processors. >Source code of an emulator is already available in c (x48 for unix or maybe >emu48) an thus it should be easily portable. The main problem would be to >design a pcb with ram, lcd, keyboard and have some development tools (gnu c >cross compiler for StrongARM already exists). The next step would be to >write code to initialize and access the hardware (not much: drawing pixels, >reading time on a rtc, keyboard, etc), and after that we're ready for the >emulation. We can easily fit this pcb with 4 meg of ram or more, and thus we could >emulate memory expansion cards of the hp (maybe 16 or more, depending on > the >size of our kernel/emulator). At 200MHz we should be able to emulate the hp48 at least at 10x it's speed. >Great! I really think we could do something here.... >Daniel -- >EMail: daniel.roggen@epfl.ch ICQ: 2510427 ---------------------------------------------------------------------- ==== Subject: Re: StrongARM >I'm not sure about WinCE. It seems that using a Microsoft product for such a >high-powered and reliability oriented task seemes self-destructive. Yet Another Hater. Don't forget that EMU is written buy just a few (one?) people. It is a *very fine program* but it doesn't have much support. ---------------------------------------------------------------------- ==== Subject: Re: StrongARM I'm not sure about WinCE. It seems that using a Microsoft product for such > a >high-powered and reliability oriented task seemes self-destructive. Yet Another Hater. Don't forget that EMU is written buy just a few (one?) people. It is a *very > fine program* but it doesn't have much support. When you mean that fixing bugs and add improvements on Emu is no support, you're right. BTW why don't you make the support? Christoph Gie§elink, one of the hard working developers :-( http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: StrongARM You can download Emu48CE at: http://nt.marin.esc.edu.ar/jpla/emu48ce/emu48ce.html It's by Jim on. >I'm not sure about WinCE. It seems that using a Microsoft product for such >a >>high-powered and reliability oriented task seemes self-destructive. >Yet Another Hater. Don't forget that EMU is written buy just a few (one?) people. It is a *very >fine program* but it doesn't have much support. ---------------------------------------------------------------------- ==== Subject: Re: StrongARM What about porting emu48 to wince? As far as I know, somebody already ported the EMU towards the EPOC platform for the PSION 3a or 3c. And it works quit well . ---------------------------------------------------------------------- ==== Subject: Re: StrongARM You should see mathcad running on a 450 MHz pentium II !! Brian >I'm very angry at HP. They make a wonderful calc(HP48GX) and then foul >it up with a 4mhz processor. Damn! Well, I've been looking at this >StrongARM from Intel(formerly Digital) and was wondering if anyone would >be interested in porting the HP's ROM from the Saturn to the StrongARM. >I'm not sure about the legality but if you saw an HP run at 200mhz, you >wouldn't care either. Respond with thoughts >David Cummings ---------------------------------------------------------------------- ==== Subject: To the French people in the house! This is as Jean says en international NG French is a vary beautiful language but unfortunately not everybody can help you out with your questions in french. There is also as Jean points out the question of etiquette. Now and then I read something that is really interesting in french on the NG and feel that it is ashamed that not everybody can share it. Don't worry if your english happens not to be as clean as some other peoples. This is not a language school. You have nothing to prove, what the french might not have in english they definitely make up when it comes to HP stuff and ML. Keep up the high quality stuff. Al the best Eva PS Libert.8e, libert.8e ch.8erie! Its 1998 not 1789-1799 ----------------- La liberte des uns finis la ou commence celle des autres. Ce Newsgroup est un newgroup international. La langue utilisee est l'anglais. Tu peux continuer a poster en Francais, mais a partir de maintenant, je n'y repondrai plus. Ce qui va grandement limiter l'etendue des reponses. C'est une question de politesse. Si tu veux ecrire en francais, alors utilise les newsgroup francais. Je peux compr un message ecris en francais pour des cas precis comme une demande d'achat ou de vente sur la France. Pour tous les sujets generaux, par respect pour ceux qui peuvent nous lire on se doit de faire l'effort d'ecrire et de repondre en anglais. Il n'y a rin de plus rude que d'avoir dans un groupe anglophone, des francais se sentant superieurs a tous les autres et parlant a la barbe de tout le monde en francais. Ce qui est etrange c'est que je n'ai vu cela qu'avec des Francais et jamais d'autre nationalite (allemand, suedois etc...) il faut croire qu'ils ont recu une meilleur education que nous. Jean-Yves Les francais se font vraiment mal remarquer dans ce ng... entre ceux qui >> n'ecrivent qu'en francais Si j'ai envie d'.8ecrire en fran.8dais, je le fais. Je suis libre de poster dans la langue que je veux. Et tu toucheras pas >.88 ma libert.8e, ni personne d'autre. J'.8epelle le mot: L I B E R T E Si tu l'.8ecris lettre par lettre, tu devrais y arriver. Me dire que je vais mal me comporter parce que je posterait dans ma >langue plut.99t qu'en anglais est innaceptable. Je sais que je ne touche pas tout le monde quand je poste en fran.8dais, >et souvent je fais l'effort d'utiliser l'anglais, comme d'autres font >l'effort d'utiliser un fran.8dais qu'ils disent mauvais, et que je trouve >bon, et encourageant. L'anglais ne me pose pas de probl.8fme sur le plan de son utilisation, >mais je ne supporte pas qu'on d.8epr.8ecie une langue pour une autre, et si >Internet doit .90tre 100 % anglais, je dis merde, et j'utilise la langue >que je veux, et j'invite tout le monde .88 faire pareil. .8da commence .88 me gonfler d'aller sur Internet pour me retrouver avec des >r.8fgles et des usages qui sont l'antinomie de la pr.8etendue libert.8e du >r.8eseau. Je dois sembler un peu dur, mais qu'on touche pas .88 ma *libert.8e* >d'*expression*, ni sur le net, ni ailleurs, et cela comprend ma langue >propre, comme pour *chacun* d'entre vous. Libert.8e, libert.8e ch.8erie! > ---------------------------------------------------------------------- ==== Subject: Re: To the French people in the house! I'm trying to read the French. I will continue to try. Like many American's, I took French lessons in grammar and high school but never really had the opportunity to have to speak the language like one would in France (or Quebec). It would be so much easier to learn the language by living for awhile in a French speaking setting. Obviously the French contribution to this NG and to the software base for the 48 is superb. Brian >This is as Jean says en international NG >French is a vary beautiful language but unfortunately >not everybody can help you out with your questions in french. >There is also as Jean points out the question of >etiquette. Now and then I read something that is really interesting in >french >on the NG and feel that it is ashamed that not everybody can share it. Don't worry if your english happens not to be as clean as some >other peoples. This is not a language school. >You have nothing to prove, what the french might not have in >english they definitely make up when it comes to HP stuff and >ML. Keep up the high quality stuff. Al the best >Eva PS >Libert.8e, libert.8e ch.8erie! >Its 1998 not 1789-1799 >----------------- La liberte des uns finis la ou commence celle des autres. Ce Newsgroup est un newgroup international. La langue utilisee est >l'anglais. Tu peux continuer a poster en Francais, mais a partir de maintenant, je >n'y >repondrai plus. Ce qui va grandement limiter l'etendue des reponses. C'est une question de politesse. Si tu veux ecrire en francais, alors utilise les newsgroup francais. Je peux compr un message ecris en francais pour des cas precis >comme >une demande d'achat ou de vente sur la France. Pour tous les sujets generaux, par respect pour ceux qui peuvent nous >lire >on se doit de faire l'effort d'ecrire et de repondre en anglais. Il n'y a rin de plus rude que d'avoir dans un groupe anglophone, des >francais se sentant superieurs a tous les autres et parlant a la barbe >de >tout le monde en francais. Ce qui est etrange c'est que je n'ai vu cela >qu'avec des Francais et jamais d'autre nationalite (allemand, suedois >etc...) il faut croire qu'ils ont recu une meilleur education que nous. Jean-Yves >Erwann a .8ecrit: > Les francais se font vraiment mal remarquer dans ce ng... entre ceux >qui > n'ecrivent qu'en francais >>Si j'ai envie d'.8ecrire en fran.8dais, je le fais. >>Je suis libre de poster dans la langue que je veux. Et tu toucheras pas >.88 ma libert.8e, ni personne d'autre. >>J'.8epelle le mot: L I B E R T E >>Si tu l'.8ecris lettre par lettre, tu devrais y arriver. >>Me dire que je vais mal me comporter parce que je posterait dans ma >>langue plut.99t qu'en anglais est innaceptable. >>Je sais que je ne touche pas tout le monde quand je poste en fran.8dais, >>et souvent je fais l'effort d'utiliser l'anglais, comme d'autres font >>l'effort d'utiliser un fran.8dais qu'ils disent mauvais, et que je trouve >bon, et encourageant. >>L'anglais ne me pose pas de probl.8fme sur le plan de son utilisation, >>mais je ne supporte pas qu'on d.8epr.8ecie une langue pour une autre, et si >Internet doit .90tre 100 % anglais, je dis merde, et j'utilise la langue >>que je veux, et j'invite tout le monde .88 faire pareil. >>.8da commence .88 me gonfler d'aller sur Internet pour me retrouver avec >des >>r.8fgles et des usages qui sont l'antinomie de la pr.8etendue libert.8e du >>r.8eseau. >>Je dois sembler un peu dur, mais qu'on touche pas .88 ma *libert.8e* >>d'*expression*, ni sur le net, ni ailleurs, et cela comprend ma langue >>propre, comme pour *chacun* d'entre vous. >> Libert.8e, libert.8e ch.8erie! > > ---------------------------------------------------------------------- ==== Subject: Re: To the French people in the house! x-HP-Fanatic: yes > This is as Jean says en international NG > French is a vary beautiful language but unfortunately > not everybody can help you out with your questions in french. I know, but my english is really awful.. >There is also as Jean points out the question of >etiquette. Now and then I read something that is really interesting >in french on the NG and feel that it is ashamed that not everybody >can share it. And for people whom don't know english? Is that not a shame? I've not read SASM.DOC or Jazz docs to their end because they're in english, and I don't get it very well with it. First time I read Jazz doc, I didn't understand very well. I was unable to find how to do a R0=C.F A (for non ML dudes out there, this instruction is saving 5 nibbles from register C to the save register R0) and I was very annoying, frustrating, and it fed me up. > Don't worry if your english happens not to be as clean as some > other peoples. This is not a language school. Writing something *wrong* not because we did wrong, but because we misused english, is a real pain in the HP for me. It happenned to me several times, I even said once something, while I intended to say the inverse of it. A friend appointed the mistake to me, and I felt a kind of despair... and shame of my poor english. This is why I used french: I've been flamed, in english, by email. I felt sad, and I didn't felt like using english. What should I say of flamers? Give their name to all? I won't. If they can't say to all what they think, and send insult in private, that's their biz. > You have nothing to prove, what the french might not have in > english they definitely make up when it comes to HP stuff and > ML. I do think spanish coders are fucking good ones. I've been wandering on the HP web's ring, and I've found not only impressive pages, but good ideas, clever presentation, and programs so useful I can't work without them anymore. Japanese pages are real great too, but too few people get with japanese (which is not as worse as french, speaking it is easy, you spell ALL letters, it's not like in french where you write something, and speak it some another weird way, bad one usually) We french are lamers. I'm one, so I know about it and other ones around me. Like Jean-Yves said to me, too many thing themselves as HP gods. When you ask them an advice, they smile at you, and say a shit like: search by yourself like I did. I'm tired of people who know howto and pride themselves as good coders as long as they keep their somewhat knowdledge of. And people like that, I've only seen french ones. Pouah! Shame on us, we stupid bozo frogs. -Gilb- (andrell@iname.com) ---------------------------------------------------------------------- ==== Subject: Re: To the French people in the house! [off-topic] Your English is not so awful, you know.... Keep practising it, and it'll be even better. I hope the only reason you post in French is only the fact that you're not Shakespeare... ;-) Then this reason is no more valid, and you'll be able to answer to some questions, in English! Just put somewhere in your message that English is not your native language... Compare the difficulty for a non French to understand you messages in French, and you to understand and write in English... You'll just see that it's far easier for you... The others will do the same, and realize that you preferred not to be lazy. After all, English is easy for some people (French, German, Asian, ...) due to the very light grammar. Erwann. (I'm French, too!) F.H. Gilbert a .8ecrit: [ Lot of stuff deleted] > I know, but my english is really awful.. [More stuff deleted] > -Gilb- (andrell@iname.com) ---------------------------------------------------------------------- ==== Subject: Re: To the French people in the house! newsgroup so far. But after all this discussion about languages, I realized that, when I'm saying Hy everybody, I hail only to those english-speaking readers outta here... That's really a shame, and it would be really excellent if we could understand each other without language barriers. But that's not the way things are, and I'm afraid it will take a long time before it comes true. So, please, let's concentrate on the present, and specifically on the case of this (our!) newsgroup. Before I say anything else, please know that I'm Brazilian, so my native language is Portuguese. I realized that I would lose a lot of stuff if I didn't speak other languages. As a programmer, my first and natural choice was English (you know, the language of the mnemonics). Currently, I'm learning German, because I'm working for a German company, and they pay my course (very nice!). Anyway, I don't think I'm less patriot or less Brazilian only because of that. I mean, I'm pretty sure I'm not! And I realy enjoy rading Sartre's books, and I'm not less nacionalist only because of that. What I mean is that, I'm only able to read his books because someone has translated them to Portuguese. Probably, a French who realized that many people in Brazil and Portugal were unable to read his work only because of the language barrier. So, please, I beg to the French programmers here, that if they can write their postings in English, do it knowing that they will be helping a lot on non-american readers, too. Fortunatelly to all of lot of knowledge, please don't want to isolate it from the people who unfortunatelly can't speak French. I'd rather posting in Portuguese, it would be easier for me, but I think I couldn't help many people then... and I couldn't be helped, either! I hope I haven't offended anyone with this posting; I guess whoever comprehension! Ricardo Pereira Jasinski. http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: To the French people in the house! What about a new newsgroup: comp.sys.hp48.en ? Would end all this arguing :) -- Robert Tiismus http://www.physic.ut.ee/~robert ---------------------------------------------------------------------- ==== Subject: Re: To the French people in the house! > What about a new newsgroup: comp.sys.hp48.en ? Would end all this > arguing :) (1) Nothing will end the arguing. Stubbornness and stupidity know no limits. Nice thought, though. (2) There's already a French-only newsgroup; no need for new ones. (3) This NG is open to anyone, but standard rules of netiquette are expected, one of which is that the best reply to moronic or meaningless postings is no reply at all. -Joe- -citizen: USA; race: human- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ---------------------------------------------------------------------- ==== Subject: Re: To the French people in the house! [off-topic] >It's true only because french guys are so open to other country and culture >that we have to learn other languages early in our study cycle while you, >americans guys, do no event know that there is a world outside the USA >(except perhaps for IRAK). > Please, don't use the word America for the USA or Americans for their inhabitants. For people that live in the American Continent it is an offense that one country decided to use the name of a continent to refer to their own country. So make an analogy, think how would you feel if the English decided to call their country Europe. Yes, I agree. Not only they don't realize that there is a world outside the USA but that there are many other countries in America besides them. S'il te pla.94t, nâutilisez pas le mot Am.8erique pour te r.8ef.8erer aux .83tats-Unis ou le mot Am.8ericains pour te r.8ef.8erer .88 ses habitants. Pour les autres habitants du Continent Am.8ericain, s'est une offense que un pays .88 d.8ecid.8e utiliser le nom de un continent pour se r.8ef.8erer .88 leur pays. Pour faire une analogie, pense-toi que est-ce que tu penserais si lâAngleterre sâ appellerai Europe Vraiment, les .83tats-uniens ne comprennent pas quâil y a un monde que n'est pas aux .83tats-Unis ni quâil y a des autres payses en Am.8erique. Pardon pour mon mauvais Fran.8dais, mais si on dit quâon doit .8ecrire en mauvais Anglais pour que les Anglaises comprennent mieux, je pense quâaussi je peux .8ecrire en un tr.8fs mauvais Fran.8dais pour que les Fran.8daises, qui ont fait logiciels tr.8fs bonnes pour lâHP, comprennent. ---------------------------------------------------------------------- ==== Subject: Re: To the French people in the house! [off-topic] I absolutely agree with you on that. But the Americans are not alone to be like that... The Anglophones of Canada are having the same thoughts. Just have a look on the message of Iain Bennett in the same newsgroup and you will understand. Ë la prochaine, Martin Savard <754d9c$5i6$1@taiwan.informatik.unirostock.de>... >It's true only because french guys are so open to other country and culture >>that we have to learn other languages early in our study cycle while you, >>americans guys, do no event know that there is a world outside the USA >>(except perhaps for IRAK). >Please, don't use the word America for the USA or Americans for their >inhabitants. For people that live in the American Continent it is an >offense that one country decided to use the name of a continent to refer to >their own country. So make an analogy, think how would you feel if the >English decided to call their country Europe. Yes, I agree. Not only they don't realize that there is a world outside the >USA but that there are many other countries in America besides them. S'il te pla.94t, nâutilisez pas le mot Am.8erique pour te r.8ef.8erer aux >.83tats-Unis ou le mot Am.8ericains pour te r.8ef.8erer .88 ses habitants. Pour les >autres habitants du Continent Am.8ericain, s'est une offense que un pays .88 >d.8ecid.8e utiliser le nom de un continent pour se r.8ef.8erer .88 leur pays. Pour >faire une analogie, pense-toi que est-ce que tu penserais si lâAngleterre s â >appellerai Europe Vraiment, les .83tats-uniens ne comprennent pas quâil y a un monde que n'est >pas aux .83tats-Unis ni quâil y a des autres payses en Am.8erique. Pardon pour mon mauvais Fran.8dais, mais si on dit quâon doit .8ecrire en >mauvais Anglais pour que les Anglaises comprennent mieux, je pense quâaussi >je peux .8ecrire en un tr.8fs mauvais Fran.8dais pour que les Fran.8daises, qui ont >fait logiciels tr.8fs bonnes pour lâHP, comprennent. ---------------------------------------------------------------------- ==== Subject: learn assembly language (part 1) x-HP-Fanatic: yes I do think much more could learn assembly. We need coders, and main problem is that information is too hard to get about ML. So here I give you what (some of) I learned since the last three weeks, when I started learning. I still consider myself a newbie, so this is not a definite introduction to assembly language. I got three weeks of ML, where some like Mr Avenard, Mr De Brebinson have years of coding. I'm a peanut to them :o) Please send feedback to my email address basics (binary numbers) to 4-gray displays, and such. I will do my best, but I'll probably do some mistakes, so please don't flame me. To write what you read here too me several hours (more than 4 hours) and I hope people will learn to code from all I'm doing. It's very time consuming, and if you got problems, I'll try to help as I can. Tomorrow, I'l post the part 2 F.H. Gilbert (andrell@iname.com) ! THIS IS FREE: DISTRIBUTE WIDELY TO ALL HP FANS ! Richard Steventon for advices, or just sending me emails, that is some of their time they gave to me :-) In this part: 1. Introduction: the RPL, and system RPL 2. Assembly language and Machine language 3. The binary base, and the nibbles 4. Converting from binary to decimal 5. Most significant bit, and Least significant bit 6. Fractionnal numbers 7. Calculating using base 2 Adding Substracting Negative numbers The signed binary The complement* of 2 Multiplying Dividing (this one is missing, I don't know howto) 1. Introduction: the RPL, and system RPL ---------------------------------------- The HP48 can be used with RPL, and so does the basic user (no judgment there). It's easy to learn, and it's the first langage you can use to program your HP. Each time you do use a command of RPL, it checks the arguments, so it's a rather safe langage, thus you should not loose any data when using it. If we look in the inners of the HP, we'll discover that each RPL command is in fact divided in several simple commands, each one which does a single (usually) and simple thing. The RPL is divided internally in elementary commands, which we call the system RPL. Each one of this system RPL commands usually ends up calling machine code, which is being executed directly by the processor of your HP, the Saturn. Look at this: RPL --> system RPL --> Machine language So you get it. Using the HP48 gives you three languages of development. RPL is secure and easy, but the slowest of all. The system RPL is faster, because no checks are made, so you must take care of that (you do checks at start, and then you no longer need them as you feed data to hungry commands) and finally we have the assembly language, which produces machine language: the fastest code, directly executed by the processor, and of course you get total control over the HP. If you want to learn system RPL, knowing RPL will be enough, all you will need is a complete list of all system RPL commands inside the ROM. The assembly language is not linked to RPL or system RPL, you're going to learn something real different here. You don't have in ML all the commands which are available to the system RPL coder. It is not a complex language, its commands are simple, and do very elementary things; what *may* seem complicated is that you will need to write a lot of elementary commands to do more complicated things than just crashing your HP ;) What you'll win is the fastest code around, and lack of sleep, but that's being a coder. 2. Assembly language and Machine language ----------------------------------------- The Saturn processor uses series of 0 and 1 to work. We are not able, because of our inferior brains*, to understand those series. * Mr Avenard is good because he's got a Saturn processor surgically implanted inside his brain. An another option is to eat a lot of Saturn processors, like Mika did, but it's quite difficult to eat just one, trust me! :o) So, we will group those series of 0 and 1 together, to create small paquets of information. The Saturn processor has been mostly design to manage this binary information using 4-bits paquets, each one being called a nibble. I will explain to you the binary base, and the hexadecimal one, which is the one you will mostly use on computers. Don't worry :) So, the machine language is a series of binary digits (0 or 1) or bits (the word resulting from the contraction of binary and digit) which are kept together to form paquets of 4-bits, the nibbles. The Saturn processor is only able to do internally very simple things, like to read some nibbles in memory, write some, and do calculations over the nibbles read. Each elementary instruction is a kind of command to us. We, ML coders, use something called mnemonics. What is that? A mnemonic is the name we give to each elementary command, it's something very simple, so it reminds us what it's used for (mnemosis --> memory, so mnemonic). We will write our programs using an assembly language, which is composed of mnemonics, one per command the processor is able to use. A compiler will translate those mnemonics into binary information, and that bits will be used by the processor to obey our orders! :-) Remember that: o le machine language is the ONLY language the Saturn processor is able to understand. It's too complicated because it's a series a bits (0 and 1) so we don't use it directly. o We will then translate those to mnemonics, which are simple words, each one exists per command available on the Saturn processor. It is important to not mix assembly language and machine language. The second one is the resulting form of the first one when it has been translated by a compiler. That is why the assembly language is called a low-level language. In the assembly language, each mnemonic is the corresponding of one instruction. That is why an assembly language is *specific* to a processor. When we write assembly language programs, we write mnemonics in a source file (source because it's the source of our work) and then we'll use a program called an assembler or compiler that will translate mnemonics to the machine language, a series of 0 and 1 binary digits, which are kept together using 4-bit paquets on the HP. 3. The binary base, and the nibbles ----------------------------------- We are going to do some maths. Don't get afraid: I'm myself bad at maths, I've been in litterature studies in school, and I do ML anyway :)) As I told you, the machine language is composed of series of zeroes and ones. That is because the processor work, in their inners, with components that let electrons go or not around. In the computer world, 1 usually means there is voltage there, and 0 there is not (or less). So, the Saturn processor keeps the bits as 4-bits paquets, which are called nibbles. This is something you must REMEMBER. Let's start from something we know: the decimal base. We got ten digits (and not eleven.. like HPSun/Julien likes to claim!!) so we use the decimal base easily. The decimal base is not the only one. Sailers use the sexagesimal one, base 60, because they use hours, minutes, and seconds. When you type a number, a decimal number like 10, on your HP, it displays 10. In fact, the HP translates it to bits inside, and the Saturn processor works using those, thus under binary base (I will teach you later that the Saturn processor is able to work using our decimal base, or its binary base, a very good processor isn't it?!). Decimal base means we use 10 symbols to write numbers: 0 1 2 3 4 5 6 7 8 9 Under the binary base, we only got TWO symbols: 0 1 So, zero in binary base is zero, one is one, and two? Well, if we use the symbol 2 under the decimal base to represent the number two, we will use that: 10 under the binary base to represent the number two, that is: we are going to combine zeroes and ones to represent numbers using the binary base. The binary base is very close linked to the Boole algebra, but I would prefer that someone well versed in mathematics to speak about it. So, here is a table (a small one) of decimal numbers, and their corresponding form in binary base: Decimal base Binary base 0 0 1 1 2 10 3 11 4 100 5 101 6 110 7 111 8 1000 9 1001 10 1010 ... ... I want you to look at the way we combine the 0 and 1, and learn how to continue. It's not something as easy: myself I had trouble understanding (told you I was awfully b.89.89.89d at maths..) Now, we are going to kept bits as paquets of 4-bits, because that's the way the Saturn likes. 4 paquet bits are called nibbles. So a nibble is composed of 4-bits. The ML coders have some specials words when they speak about bits, and the different sizes we use to group bits. When 4 bits are packed together, we call it a word of 4-bits. If we pack 32 bits together, we'll get a word of 32-bits. How many values can a paquet of 4-bits be? It's easy: we use a power of two to calculate: 2 power 4 = 16 So 4-bits packed together do create a nibble, and: ***************************************** * a nibble can have 16 different values * ***************************************** Let's calculate an other one, how many values can a word of 64-bits have? 2 power 64 = 1.8 * 10 ^ 19 A lot :-)) 4. Converting from binary to decimal ------------------------------------ We are going to examinate two nibbles kept together, which is call a byte. 4-bits --> a nibble 8-bits --> a byte We are going to give each bit a number, from right to left, we call that the rank of a bit. Here is an example: I have two nibbles together (a byte) which are: 01100101, so we have: rank: 7 6 5 4 3 2 1 0 bits: 0 1 1 0 0 1 0 1 As you see, I have started the rank by zero, and then I count from right to left. Each bit has a rank, a specific rank. Each bit has also something special. The rank is used to know which weight the bit has. What does that mean? It means that each bit is, according to its rank, an exponant of base 2. Let me explain, look at that: rank: 7 6 5 4 3 2 1 0 bits: 0 1 1 0 0 1 0 1 weight: 2^7 2^6 2^5 2^4 2^3 2^2 2^1 2^0 What I call weight is the value of the bit in the decimal base. Here is an example: what is the 00100110 value in decimal mode? Let's write it: rank: 7 6 5 4 3 2 1 0 bits: 0 0 1 0 0 1 1 0 weight: 2^7 2^6 2^5 2^4 2^3 2^2 2^1 2^0 Now, using the weight of each bit can be calculated, and we'll find the decimal value! Let's do it: We are going to multiply each bit by its weight, and we add each together, that is: (0 * 2^7) + (0 * 2^6) + (1 * 2^5) + (0 * 2^4) + (0 * 2^3) + (1 * 2^2) + (1 * 2^1) + (0 * 2^0) that is: 0 + 0 + 32 + 0 + 0 + 4 + 2 + 0 = 38 Use your HP to check: type [MTH]Ê[BASE] [BIN] and then the number: #00100110b. Now, press [DEC] and you get: #38d It works!! Cool no? :-) Now, you are able to convert a binary number to a decimal one. You write them down, you give each one a rank, from right to left and from zero to the last one. Then, you write the weight of each bit, and you can calculate the decimal value. What is the maximal value of a nibble? Two nibbles? With two nibbles we have: rank: 7 6 5 4 3 2 1 0 bits: 1 1 1 1 1 1 1 1 weight: 2^7 2^6 2^5 2^4 2^3 2^2 2^1 2^0 That is: (1 * 2^7) + (1 * 2^6) + (1 * 2^5) + (1 * 2^4) + (1 * 2^3) + (1 * 2^2) + (1 * 2^1) + (1 * 2^0) That is: 128 + 64 + 32 + 16 + 8 + 4 + 2 + 1 = 255 So a byte (two nibbles) can go from 0 to 255, that is 256 different values Remember!! 0 (ZERO) is a value too! And with a nibble we have: rank: 3 2 1 0 bits: 1 1 1 1 weight: 2^3 2^2 2^1 2^0 That is: (1 * 2^3) + (1 * 2^2) + (1 * 2^1) + (1 * 2^0) That is: 8 + 4 + 2 + 1 = 15 So with a nibble we can go from 0 to 15, that is 16 different values. To get the decimal form of a binary integer, we multiply each bit's value by ( 2 power (the bit rank) ) and we add all values. 5. Most significant bit, and Least significant bit -------------------------------------------------- This is something important, don't skip it. As I have shown you, each bit has a weight. The bit which is the rightmost has a rank of zero, and the bit which is leftmost has the highest rank (rank n-1 when we have n bits). This means that each bit has a different weight in the final value. The bit which is rightmost is called the Least significant bit, or LSB The bit which is leftmost is called the Most significant bit, or MSB In the nibble, the LSB and MSB are there: rank: 3 2 1 0 MSB LSB In the byte, the LSB and MSB are there: rank: 7 6 5 4 3 2 1 0 MSB LSB 6. Fractionnal numbers ---------------------- This is trickier :-) A fractionnal number has two parts: the left one, before the comma, which is the integer part, and the right one, after the comma, which is the fractionnal one. But.. how can we have fractionnal numbers using bits? We are going to give each bit a negative exponant weight. Instead of: rank: 3 2 1 0 bits: 1 1 1 1 weight: 2^3 2^2 2^1 2^0 we have for fractionnal numbers: rank: 3 2 1 0 bits: 1 1 1 1 weight: 2^-3 2^-2 2^-1 2^-0 Please note: ( 2^-3 ) is in fact: 2 ^ ( 1 / 3 ) ( 2^-2 ) is in fact: 2 ^ ( 1 / 2 ) What is the value of 0,(0110) in decimal? rank: 3 2 1 0 bits: 0 1 1 0 weight: 2^-3 2^-2 2^-1 2^-0 That is: (0 * 2^-3) + (1 * 2^-2) + (1 * 2^-1) + (0 * 2^-0) That is: 0 + 0,25 + 0,5 + 0 = 0,75 7. Calculating using base 2 --------------------------- Now I am going to explain to you how to add two binary values, substract two binary values, multiplying and dividing, this is going to help you understand how the Saturn does :-) Adding ------ You will follow these rules: 0 + 0 = 0 --> no carry 0 + 1 = 1 --> no carry 1 + 0 = 0 --> no carry 1 + 1 = 1 --> carry Here is an example: 0 1 0 1 + 1 1 0 1 ---------- 1 0 0 1 0 When I have 1 + 1, I put a zero, and I move the carry 1 to the left. I work here from right to left, adding two bits each time. Substracting ------------ You will follow these rules: 0 - 0 = 0 0 - 1 = -1 --> negative result 1 - 0 = 0 1 - 1 = 0 So here I need to explain to you how a negative number is represented using the binary base. Negative numbers ---------------- There are various ways to do that. I am going to explain two to you, which are important to know (and you'll use either one or another in your own programs, that will vary if you let the processor do calculations for you, or if you want to make it work the way you want, and rule the beast!!) The signed binary: the first one -------------------------------- We are going to use the Most significant bit here (now you know why I told you about the MSB and the LSB!). When the MSB is zero, the value is POSITIVE. When the MSB is one, the value is NEGATIVE. And we use all the other bits available to code the value. This is called signed binary and it's my favorite one (don't ask why or I'll call Mr Avenard, and he will eat your brain to feed upon!!) If we have one nibble, a positive zero is: 0000 a negative zero is: 1000 (yes, there are a +0 and -0 when using signed binaries) If you remember what I told you before, a nibble could go from 0 to 255, thus 256 values; here we cannot, because we use the MSB to tell if it's positive or negative. So a nibble goes from: 0000 to 0111 when it is positive 1000 to 1111 when it is negative that is: from positive zero to 7 when it is positive from negative zero to -7 when it is negative If we use a byte (two nibbles): from: 00000000 to 01111111 when it is positive 10000000 to 11111111 when it is negative that is: from positive 0 to 127 when it is positive from negative 0 to -127 when it is negative Remember!! We have TWO zeroes here, the negative one, and the positive one. The complement* of 2: the second one ------------------------------------ * In french (yes, I'm another HP-freak frenchie) it is called << Compl.8ement .88 deux >> so I traduced it here like it looks alike in french. When all bits in a word (whatever its size) are inverted, we get.. an inverted word. We call that complement of 1. The zeroes become ones, and the ones becomes zeroes. That is not difficult at all. 0101 becomes: 1010 1100 becomes: 0011 0010 becomes: 1101 But the complement of 2 is the true one for a word, and it's the complement of 1 plus one; that is: ( complement of 2 ) = ( complement of 1 ) + 1 Example: what is the complement of 2 of 01101110 ? First, we invert all bits, and get: 10010001 I add one and I get: 10010001 + 1 ---------- 10010010 So the complement of 2 of 01101110 is 10010010 It is more complicated to handle complement of 2 binary numbers. Please pay attention, it is not complicated, but you must be awake for that :) I am going to tell you about the way the complement of 2 numbers are encoded. I will work on a byte (8 bits) to explain it to you. I have a start value of 0000 0000 If I add one, and again and again, it goes up.. But, what will happen if I have 0000 0000 and I remove 1? I get: 0000 0000 - 1 = 1111 1111 As you see, when using complement of 2, the negative values start with 1 in their MSB. If the MSB is 0, the value is positive. Interesting thing: positives values go from 0000 0000 to 0111 1111 that's: 0 to 127 negative values go from 1000 0000 to 1111 1111 that's: -128 to -1 You see? Note: here I'm using 8 bits.. but we could use less (like 4 in a nibble) or 64.. a lot of bits to play with ;-) We don't get a negative zero and a positive zero here!! Snif! Snif! :'( (Snif! = sound of someone close to crying, french noise) But there is something MUCH MORE IMPORTANT! We cannot do ( 1000 0000 - 1 ) !! Why? We would go from a NEGATIVE value to a POSITIVE value. Also, we cannot do ( 0111 1111 + 1 ) because we would go from a POSITIVE value to a NEGATIVE value, and adding +1 should not (unless we redefine mathematics) give a negative value from a positive value :-) I want you to understand and know the differences between the signed binaries, and the complement of 2. The first one helps remember and understanding the second one, that's how I do, and I hope it's a good way to remember it, as the first one is very easy to remember, with its +0 and -0 :-)) Using the complement of 2 is a bit more complicated. Here is an example, how do I calculate 8 - 11 (which are decimal values) using 8 bits and complement of 2? 8 is : 0000 1000 11 is : 0000 1011 To substract 11, I'm going to add its negative value, so I have: ( 8 - 11 ) becomes ( 8 + -11 ) The complement of 2 of 0000 1011 is inversing all bits, and adding one, that is: 0000 1011 --inversing bits--> 1111 0100 adding one and I get: 1111 0101 So I now add the two values: 0000 1000 + 1111 0101 ----------- 1111 1101 As you see, the MSB here equals 1, so the result is known to be negative. But, I cannot use that final value directly. I have to calculate its complement of 2 and remember it's negative, and add a - sign in front of it. complement of 2 of 1111 1101 is: 0000 0011 that is finally: - 3 REMEMBER! When we use complement of 2 to do calculus, like the one I showed up here above, the final result IS NOT EXPLOITABLE directly, we have to: 1. calculate the absolute value 2. put the sign according to the MSB value Er.. I'm a bit tired. Let me breathe! :o) Multiplying ----------- I don't like this one (because sometimes I loose myself!) but I have to. To do a multiplication, we do adding and shifting bits. When you want to multiply a value, the slowest of the slowest methods is to add the number to itself, as long as needed to get the multiplicated value. This will use and eat up the processor's time (a baaaaaaaaaad thing to do, would say my friend Mr Avenard) We can do a multiplication using bit-shifting, and adding values. Here is an example: 0010 x 0011 ------ 0010 (that is 0010 * 1) + 0010 (that is 0010 * 1) + 0000 (well..) + 0000 -------- 00000110 Well, it's not very beautiful, but it works :-) As you see, I have done 4 multiplications. As the two left bits are zero, I could simplify it like below: 10 * 11 ---- 10 10 ---- 110 To multiply, I do: 1. I read the multiplicator (the second number) from right to left 2. When I find a bit set to 1, I do copy the number being multiplicated (the first number) 3. Otherwise, if I have found a zero, I do put zeroes, well, that's easy no? :) 4. Each time I do look at a new bit on the second number, I shift to left the bits (look above, the bits are written one place left at each line). 5. Finally, I add all values like we know to do. REMEMBER: all you have to do it look at the second number that multiplies, bit per bit from right to left, and do shifts; then, we add all values considering carry when there is one. The Saturn processor is able to do shifts fastly, like adding values, so it's a good way to multiply values. To shift a value one bit to the left is like doing a multiplication per 2. To shift a value one bit to the right is like doing a divide per 2. There is a way to optimise multiplications, and divisions. As we have seen, each 0 found requires shifting one bit to the left. To multiply fastly a value, we find out how many zeroes are inside the number, and do as much left bit shifts. Dividing -------- Here I must ask someone else to explain, because I don't understand how to do very well. I have deep trouble with all that is linked to mathematics, and I can't do this one :P The next part ------------- I think I will explain the hexadecimal base there (base 16) and some logic operations, like OR, AND, XOR.. (useful on graphics, you'll see) and I will speak of the Saturn processor. ---------------------------------------------------------------------- ==== Subject: Re: learn assembly language (part 1) F.H. Gilbert a .8ecrit: I do think much more could learn assembly. We need coders, and main > problem is that information is too hard to get about ML. > I wanted to learn assembly for the Saturn, but even with several books, I didn'tmanage to code even the single hello world... And I already know x86 and 6805 assembly... (6805 is for smart cards) > So here I give you what (some of) I learned since the last three weeks, > when I started learning. I still consider myself a newbie, so this is not a definite introduction > to assembly language. > Let's start from the start... That's a good point... ;-) > basics (binary numbers) to 4-gray displays, and such. I will do my best, > but I'll probably do some mistakes, so please don't flame me. > You'll certainly receive criticism, it's normal. > To write what you read here too me several hours (more than 4 hours) and > I hope people will learn to code from all I'm doing. for the HP48 (in the crypto area) Anyway, that's a very good start, please continue...! Erwann. ---------------------------------------------------------------------- ==== Subject: Re: learn assembly language (part 1) Hi Jean-Yves, why do you always put the original posting into your reply ? I made that accidently one time with a larger document... and then never again. Maybe you should check your browsers switches;-) Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: learn assembly language (part 1) very nice. Is seems you can catch things fast and form it into other words. Maybe this can help others who were afraid of trying to learn something 'secret' like Saturn assembly language to give it a try. Raymond Raymond Hellstern -Magic48ges- Email: 101.165571@germany.net ---------------------------------------------------------------------- ==== Subject: Re: learn assembly language (part 1) Then only thing I can say here: CONGRATULATIONS ! Very well done, and very easy to understand. I hope you will continue to share your Saturn's discovery. I was also glad to learn that I have a Saturn implented into my brain... I will ask my parents about that, since I can't remember any surgery operations. Bravo encore Jean-Yves I do think much more could learn assembly. We need coders, and main >problem is that information is too hard to get about ML. So here I give you what (some of) I learned since the last three weeks, >when I started learning. I still consider myself a newbie, so this is not a definite introduction >to assembly language. I got three weeks of ML, where some like Mr Avenard, Mr De Brebinson >have years of coding. I'm a peanut to them :o) >Please send feedback to my email address basics (binary numbers) to 4-gray displays, and such. I will do my best, >but I'll probably do some mistakes, so please don't flame me. To write what you read here too me several hours (more than 4 hours) and >I hope people will learn to code from all I'm doing. It's very time consuming, and if you got problems, I'll try to help as I >can. Tomorrow, I'l post the part 2 F.H. Gilbert (andrell@iname.com) ! THIS IS FREE: DISTRIBUTE WIDELY TO ALL HP FANS ! Richard Steventon for advices, or just sending me emails, that is some >of their time they gave to me :-) >In this part: 1. Introduction: the RPL, and system RPL >2. Assembly language and Machine language >3. The binary base, and the nibbles >4. Converting from binary to decimal >5. Most significant bit, and Least significant bit >6. Fractionnal numbers >7. Calculating using base 2 > Adding > Substracting > Negative numbers > The signed binary > The complement* of 2 > Multiplying > Dividing (this one is missing, I don't know howto) 1. Introduction: the RPL, and system RPL >---------------------------------------- The HP48 can be used with RPL, and so does the basic user (no judgment >there). It's easy to learn, and it's the first langage you can use to >program your HP. Each time you do use a command of RPL, it checks the arguments, so it's >a rather safe langage, thus you should not loose any data when using it. If we look in the inners of the HP, we'll discover that each RPL command >is in fact divided in several simple commands, each one which does a >single (usually) and simple thing. The RPL is divided internally in >elementary commands, which we call the system RPL. Each one of this system RPL commands usually ends up calling machine >code, which is being executed directly by the processor of your HP, the >Saturn. Look at this: RPL --> system RPL --> Machine language So you get it. Using the HP48 gives you three languages of development. RPL is secure >and easy, but the slowest of all. The system RPL is faster, because no >checks are made, so you must take care of that (you do checks at start, >and then you no longer need them as you feed data to hungry commands) >and finally we have the assembly language, which produces machine >language: the fastest code, directly executed by the processor, and of >course you get total control over the HP. If you want to learn system RPL, knowing RPL will be enough, all you >will need is a complete list of all system RPL commands inside the ROM. The assembly language is not linked to RPL or system RPL, you're going >to learn something real different here. You don't have in ML all the >commands which are available to the system RPL coder. It is not a complex language, its commands are simple, and do very >elementary things; what *may* seem complicated is that you will need to >write a lot of elementary commands to do more complicated things than >just crashing your HP ;) What you'll win is the fastest code around, and lack of sleep, but >that's being a coder. >2. Assembly language and Machine language >----------------------------------------- The Saturn processor uses series of 0 and 1 to work. We are not able, >because of our inferior brains*, to understand those series. * Mr Avenard is good because he's got a Saturn processor surgically >implanted inside his brain. An another option is to eat a lot of Saturn >processors, like Mika did, but it's quite difficult to eat just one, >trust me! :o) So, we will group those series of 0 and 1 together, to create small >paquets of information. The Saturn processor has been mostly design to >manage this binary information using 4-bits paquets, each one being >called a nibble. I will explain to you the binary base, and the hexadecimal one, which is >the one you will mostly use on computers. Don't worry :) So, the machine language is a series of binary digits (0 or 1) or bits >(the word resulting from the contraction of binary and digit) which >are kept together to form paquets of 4-bits, the nibbles. The Saturn processor is only able to do internally very simple things, >like to read some nibbles in memory, write some, and do calculations >over the nibbles read. Each elementary instruction is a kind of command >to us. We, ML coders, use something called mnemonics. What is that? A mnemonic is the name we give to each elementary command, it's >something very simple, so it reminds us what it's used for (mnemosis --memory, so mnemonic). We will write our programs using an assembly language, which is composed >of mnemonics, one per command the processor is able to use. A compiler >will translate those mnemonics into binary information, and that bits >will be used by the processor to obey our orders! :-) Remember that: o le machine language is the ONLY language the Saturn > processor is able to understand. It's too complicated > because it's a series a bits (0 and 1) so we don't > use it directly. o We will then translate those to mnemonics, which are > simple words, each one exists per command available > on the Saturn processor. It is important to not mix assembly language and machine language. >The second one is the resulting form of the first one when it has been >translated by a compiler. That is why the assembly language is called a low-level language. In >the assembly language, each mnemonic is the corresponding of one >instruction. That is why an assembly language is *specific* to a >processor. When we write assembly language programs, we write mnemonics >in a source file (source because it's the source of our work) and then >we'll use a program called an assembler or compiler that will >translate mnemonics to the machine language, a series of 0 and 1 binary >digits, which are kept together using 4-bit paquets on the HP. >3. The binary base, and the nibbles >----------------------------------- We are going to do some maths. Don't get afraid: I'm myself bad at >maths, I've been in litterature studies in school, and I do ML anyway >:)) As I told you, the machine language is composed of series of zeroes and >ones. That is because the processor work, in their inners, with >components that let electrons go or not around. In the computer world, 1 >usually means there is voltage there, and 0 there is not (or less). So, the Saturn processor keeps the bits as 4-bits paquets, which are >called nibbles. This is something you must REMEMBER. Let's start from something we know: the decimal base. We got ten digits >(and not eleven.. like HPSun/Julien likes to claim!!) so we use the >decimal base easily. The decimal base is not the only one. Sailers use the sexagesimal one, >base 60, because they use hours, minutes, and seconds. When you type a number, a decimal number like 10, on your HP, it >displays 10. In fact, the HP translates it to bits inside, and the >Saturn processor works using those, thus under binary base (I will teach >you later that the Saturn processor is able to work using our decimal >base, or its binary base, a very good processor isn't it?!). Decimal base means we use 10 symbols to write numbers: 0 1 2 3 4 5 6 7 8 9 Under the binary base, we only got TWO symbols: 0 1 So, zero in binary base is zero, one is one, and two? Well, if we use the symbol 2 under the decimal base to represent the >number two, we will use that: 10 under the binary base to represent the number two, that is: we are going >to combine zeroes and ones to represent numbers using the binary base. The binary base is very close linked to the Boole algebra, but I would >prefer that someone well versed in mathematics to speak about it. So, here is a table (a small one) of decimal numbers, and their >corresponding form in binary base: Decimal base Binary base > 0 0 > 1 1 > 2 10 > 3 11 > 4 100 > 5 101 > 6 110 > 7 111 > 8 1000 > 9 1001 > 10 1010 > ... ... I want you to look at the way we combine the 0 and 1, and learn how to >continue. It's not something as easy: myself I had trouble understanding >(told you I was awfully b.89.89.89d at maths..) Now, we are going to kept bits as paquets of 4-bits, because that's the >way the Saturn likes. 4 paquet bits are called nibbles. So a nibble is composed of 4-bits. The >ML coders have some specials words when they speak about bits, and the >different sizes we use to group bits. When 4 bits are packed together, we call it a word of 4-bits. If we >pack 32 bits together, we'll get a word of 32-bits. How many values can a paquet of 4-bits be? It's easy: we use a power of >two to calculate: 2 power 4 = 16 So 4-bits packed together do create a nibble, and: ***************************************** > * a nibble can have 16 different values * > ***************************************** Let's calculate an other one, how many values can a word of 64-bits >have? 2 power 64 = 1.8 * 10 ^ 19 A lot :-)) >4. Converting from binary to decimal >------------------------------------ We are going to examinate two nibbles kept together, which is call a >byte. 4-bits --> a nibble >8-bits --> a byte We are going to give each bit a number, from right to left, we call that >the rank of a bit. Here is an example: I have two nibbles together (a >byte) which are: 01100101, so we have: rank: 7 6 5 4 3 2 1 0 >bits: 0 1 1 0 0 1 0 1 As you see, I have started the rank by zero, and then I count from >right to left. Each bit has a rank, a specific rank. Each bit has also >something special. The rank is used to know which weight the bit has. What does that >mean? It means that each bit is, according to its rank, an exponant of >base 2. Let me explain, look at that: rank: 7 6 5 4 3 2 1 0 >bits: 0 1 1 0 0 1 0 1 >weight: 2^7 2^6 2^5 2^4 2^3 2^2 2^1 2^0 What I call weight is the value of the bit in the decimal base. Here is >an example: what is the 00100110 value in decimal mode? Let's write it: rank: 7 6 5 4 3 2 1 0 >bits: 0 0 1 0 0 1 1 0 >weight: 2^7 2^6 2^5 2^4 2^3 2^2 2^1 2^0 Now, using the weight of each bit can be calculated, and we'll find >the decimal value! Let's do it: We are going to multiply each bit by its weight, and we add each >together, that is: (0 * 2^7) + (0 * 2^6) + (1 * 2^5) + (0 * 2^4) + >(0 * 2^3) + (1 * 2^2) + (1 * 2^1) + (0 * 2^0) that is: 0 + 0 + 32 + 0 + 0 + 4 + 2 + 0 = 38 Use your HP to check: type [MTH] [BASE] [BIN] and then the number: >#00100110b. Now, press [DEC] and you get: #38d It works!! Cool no? :-) Now, you are able to convert a binary number to a decimal one. You write >them down, you give each one a rank, from right to left and from zero to >the last one. Then, you write the weight of each bit, and you can >calculate the decimal value. What is the maximal value of a nibble? Two nibbles? With two nibbles we have: rank: 7 6 5 4 3 2 1 0 >bits: 1 1 1 1 1 1 1 1 >weight: 2^7 2^6 2^5 2^4 2^3 2^2 2^1 2^0 That is: (1 * 2^7) + (1 * 2^6) + (1 * 2^5) + (1 * 2^4) + >(1 * 2^3) + (1 * 2^2) + (1 * 2^1) + (1 * 2^0) That is: 128 + 64 + 32 + 16 + 8 + 4 + 2 + 1 = 255 So a byte (two nibbles) can go from 0 to 255, that is 256 different >values Remember!! 0 (ZERO) is a value too! And with a nibble we have: rank: 3 2 1 0 >bits: 1 1 1 1 >weight: 2^3 2^2 2^1 2^0 That is: (1 * 2^3) + (1 * 2^2) + (1 * 2^1) + (1 * 2^0) That is: 8 + 4 + 2 + 1 = 15 So with a nibble we can go from 0 to 15, that is 16 different values. To get the decimal form of a binary integer, we multiply each bit's >value by ( 2 power (the bit rank) ) and we add all values. >5. Most significant bit, and Least significant bit >-------------------------------------------------- This is something important, don't skip it. As I have shown you, each bit has a weight. The bit which is the >rightmost has a rank of zero, and the bit which is leftmost has the >highest rank (rank n-1 when we have n bits). This means that each bit has a different weight in the final value. The bit which is rightmost is called the Least significant bit, or LSB The bit which is leftmost is called the Most significant bit, or MSB In the nibble, the LSB and MSB are there: rank: 3 2 1 0 > MSB LSB In the byte, the LSB and MSB are there: rank: 7 6 5 4 3 2 1 0 > MSB LSB >6. Fractionnal numbers >---------------------- This is trickier :-) A fractionnal number has two parts: the left one, before the comma, >which is the integer part, and the right one, after the comma, which is >the fractionnal one. But.. how can we have fractionnal numbers using bits? We are going to give each bit a negative exponant weight. Instead of: rank: 3 2 1 0 >bits: 1 1 1 1 >weight: 2^3 2^2 2^1 2^0 we have for fractionnal numbers: rank: 3 2 1 0 >bits: 1 1 1 1 >weight: 2^-3 2^-2 2^-1 2^-0 Please note: ( 2^-3 ) is in fact: 2 ^ ( 1 / 3 ) > ( 2^-2 ) is in fact: 2 ^ ( 1 / 2 ) What is the value of 0,(0110) in decimal? rank: 3 2 1 0 >bits: 0 1 1 0 >weight: 2^-3 2^-2 2^-1 2^-0 That is: (0 * 2^-3) + (1 * 2^-2) + (1 * 2^-1) + (0 * 2^-0) That is: 0 + 0,25 + 0,5 + 0 = 0,75 >7. Calculating using base 2 >--------------------------- Now I am going to explain to you how to add two binary values, substract >two binary values, multiplying and dividing, this is going to help you >understand how the Saturn does :-) Adding >------ You will follow these rules: 0 + 0 = 0 --> no carry > 0 + 1 = 1 --> no carry > 1 + 0 = 0 --> no carry > 1 + 1 = 1 --> carry Here is an example: 0 1 0 1 > + 1 1 0 1 > ---------- > 1 0 0 1 0 When I have 1 + 1, I put a zero, and I move the carry 1 to the left. I >work here from right to left, adding two bits each time. Substracting >------------ You will follow these rules: 0 - 0 = 0 > 0 - 1 = -1 --> negative result > 1 - 0 = 0 > 1 - 1 = 0 So here I need to explain to you how a negative number is represented >using the binary base. Negative numbers >---------------- There are various ways to do that. I am going to explain two to you, >which are important to know (and you'll use either one or another in >your own programs, that will vary if you let the processor do >calculations for you, or if you want to make it work the way you want, >and rule the beast!!) The signed binary: the first one >-------------------------------- We are going to use the Most significant bit here (now you know why I >told you about the MSB and the LSB!). When the MSB is zero, the value is POSITIVE. >When the MSB is one, the value is NEGATIVE. And we use all the other bits available to code the value. This is called signed binary and it's my favorite one (don't ask why >or I'll call Mr Avenard, and he will eat your brain to feed upon!!) If we have one nibble, a positive zero is: 0000 > a negative zero is: 1000 (yes, there are a +0 and -0 when using signed binaries) If you remember what I told you before, a nibble could go from 0 to 255, >thus 256 values; here we cannot, because we use the MSB to tell if it's >positive or negative. So a nibble goes from: 0000 to 0111 when it is positive > 1000 to 1111 when it is negative that is: from positive zero to 7 when it is positive > from negative zero to -7 when it is negative If we use a byte (two nibbles): from: 00000000 to 01111111 when it is positive > 10000000 to 11111111 when it is negative that is: from positive 0 to 127 when it is positive > from negative 0 to -127 when it is negative Remember!! We have TWO zeroes here, the negative one, > and the positive one. The complement* of 2: the second one >------------------------------------ * In french (yes, I'm another HP-freak frenchie) it is called << >Compl.8ement .88 deux >> so I traduced it here like it looks alike in >french. When all bits in a word (whatever its size) are inverted, we get.. an >inverted word. We call that complement of 1. The zeroes become ones, >and the ones becomes zeroes. That is not difficult at all. 0101 becomes: 1010 > 1100 becomes: 0011 > 0010 becomes: 1101 But the complement of 2 is the true one for a word, and it's the >complement of 1 plus one; that is: ( complement of 2 ) = ( complement of 1 ) + 1 Example: what is the complement of 2 of 01101110 ? First, we invert all bits, and get: 10010001 I add one and I get: 10010001 > + 1 > ---------- > 10010010 So the complement of 2 of 01101110 is 10010010 It is more complicated to handle complement of 2 binary numbers. >Please pay attention, it is not complicated, but you must be awake for >that :) I am going to tell you about the way the complement of 2 numbers are >encoded. I will work on a byte (8 bits) to explain it to you. I have a start value of 0000 0000 >If I add one, and again and again, it goes up.. But, what will happen if I have 0000 0000 >and I remove 1? I get: 0000 0000 - 1 = 1111 1111 As you see, when using complement of 2, the negative values start with >1 in their MSB. If the MSB is 0, the value is positive. Interesting thing: positives values go from 0000 0000 to 0111 1111 > that's: 0 to 127 negative values go from 1000 0000 to 1111 1111 > that's: -128 to -1 You see? Note: here I'm using 8 bits.. but we could use less (like 4 in a nibble) >or 64.. a lot of bits to play with ;-) We don't get a negative zero and a positive zero here!! Snif! Snif! :'( (Snif! = sound of someone close to crying, french noise) But there is something MUCH MORE IMPORTANT! We cannot do ( 1000 0000 - 1 ) !! Why? We would go from a NEGATIVE value to a POSITIVE value. Also, we cannot do ( 0111 1111 + 1 ) because we would go from a POSITIVE >value to a NEGATIVE value, and adding +1 should not (unless we redefine >mathematics) give a negative value from a positive value :-) I want you to understand and know the differences between the signed >binaries, and the complement of 2. The first one helps remember and >understanding the second one, that's how I do, and I hope it's a good >way to remember it, as the first one is very easy to remember, with its >+0 and -0 :-)) Using the complement of 2 is a bit more complicated. Here is an >example, how do I calculate 8 - 11 (which are decimal values) using 8 >bits and complement of 2? 8 is : 0000 1000 > 11 is : 0000 1011 To substract 11, I'm going to add its negative value, so I have: ( 8 - 11 ) becomes ( 8 + -11 ) The complement of 2 of 0000 1011 is inversing all bits, and adding >one, that is: 0000 1011 --inversing bits--> 1111 0100 adding one and I get: 1111 0101 So I now add the two values: 0000 1000 > + 1111 0101 > ----------- > 1111 1101 As you see, the MSB here equals 1, so the result is known to be >negative. But, I cannot use that final value directly. I have to >calculate its complement of 2 and remember it's negative, and add a - >sign in front of it. complement of 2 of 1111 1101 is: 0000 0011 that is finally: - 3 REMEMBER! When we use complement of 2 to do calculus, like the one I showed up >here above, the final result IS NOT EXPLOITABLE directly, we have to: 1. calculate the absolute value > 2. put the sign according to the MSB value Er.. I'm a bit tired. Let me breathe! :o) Multiplying >----------- I don't like this one (because sometimes I loose myself!) but I have to. >To do a multiplication, we do adding and shifting bits. When you want to multiply a value, the slowest of the slowest methods is >to add the number to itself, as long as needed to get the multiplicated >value. This will use and eat up the processor's time (a baaaaaaaaaad >thing to do, would say my friend Mr Avenard) We can do a multiplication using bit-shifting, and adding values. Here >is an example: 0010 > x 0011 > ------ > 0010 (that is 0010 * 1) > + 0010 (that is 0010 * 1) > + 0000 (well..) > + 0000 > -------- > 00000110 Well, it's not very beautiful, but it works :-) As you see, I have done 4 multiplications. As the two left bits are >zero, I could simplify it like below: 10 > * 11 > ---- > 10 > 10 > ---- > 110 To multiply, I do: 1. I read the multiplicator (the second number) from > right to left > 2. When I find a bit set to 1, I do copy the number > being multiplicated (the first number) > 3. Otherwise, if I have found a zero, I do put zeroes, > well, that's easy no? :) > 4. Each time I do look at a new bit on the second > number, I shift to left the bits (look above, the > bits are written one place left at each line). > 5. Finally, I add all values like we know to do. REMEMBER: all you have to do it look at the second number > that multiplies, bit per bit from right to left, > and do shifts; then, we add all values considering > carry when there is one. The Saturn processor is able to do shifts fastly, like adding values, so >it's a good way to multiply values. To shift a value one bit to the left is like doing a multiplication per >2. To shift a value one bit to the right is like doing a divide per 2. There is a way to optimise multiplications, and divisions. As we have >seen, each 0 found requires shifting one bit to the left. To multiply >fastly a value, we find out how many zeroes are inside the number, and >do as much left bit shifts. Dividing >-------- Here I must ask someone else to explain, because I don't understand how >to do very well. I have deep trouble with all that is linked to >mathematics, and I can't do this one :P The next part >------------- I think I will explain the hexadecimal base there (base 16) and some >logic operations, like OR, AND, XOR.. (useful on graphics, you'll see) >and I will speak of the Saturn processor. > ---------------------------------------------------------------------- ==== Subject: Re: learn assembly language (part 1) x-HP-Fanatic: yes