A60 ==== Yes, but which setting is responsible? Because I get an Overflow error as well. Werner ==== werner-huysegoms@freegates.be says... After some experiments: It's flag -21: If set (Overflow->+/-9E499), the integral is numerically evaluated; if reset (Overflow->error), it errors out. ------------------------------------------------------------------------- Ralf Fritzsch Bundesanstalt fuer Wasserbau Federal Waterways Engineering and Research Dienststelle Kueste Institute - Department Hamburg ------------------------------------------------------------------------- Unix _IS_ user friendly - it's just selective about who its friends are. ------------------------------------------------------------------------- -0600) ==== The research results are in: Indeed; that's why I included -21 CF in my preceding post :) What about getting the wrong answer when DEG mode is set? As of VERSION 1.19-6 with [CAS] VER 4.20010912, as I said, ->NUM doesn't seem to bother changing to RAD mode, and the delivered wrong numeric answer needs a D->R command to rectify it! Has anyone repeated this experiment? [r->] [OFF] -0600) ==== Here again is what I pointed out earlier (Exact mode): -3 CF -21 CF STD DEG '.S(0,oo,INV(v/(X^6+1)),X)' ->NUM My 49G delivers a wrong answer: 80.3391167439 Now do D->R and you'll get 1.40218210533 (which everyone else says is the right answer :) Repeating my version identification: VERSION: 1.19-6 VER [CAS]: 4.20010912 Buggy? [r->] [OFF] ==== I'm completely agree! We shouldn't left JYA alone here. When we start? Luis. ==== CU Michael ==== is there a was to change how many numbers to the right of the decimal are displayed on the 49? 9 is the default, could i change it to 12, or even 15? -0600) ==== Aaron asked: Floating-point (a/k/a real or approximate) numbers are always internally saved with exactly 12 significant digits (mantissa), and a separate power of ten (exponent); remembering this may help with understanding the limits of how many digits can be shown in various display modes. The default display mode (STD) actually shows *all* the saved significant digits (up to 12), omitting only any *trailing* zeros in the fractional portion. You may, however, request the calc to show a fixed number of digits following the decimal point, via the FIX command, whose numeric argument must be in the range 0 through 11 ( < 0 will act like 0 and > 11 will act like 11) However, the calc will never show more than 12 digits in total, because there simply are no more significant digits available, so if you enter 123456.789012 11 FIX, the display shows 123,456.789012 (only 6 digits following .) because all 12 existing significant digits have already been used up! On the other hand, if you divide the above by 1E9, then the display shows 0.00012345679 -- it stops at 11 digits following the decimal point (rounding if necessary), as requested; thus 11 (or the argument of FIX) is the *maximum* number of digits ever displayed *after* the decimal point in FIX mode, while 12 is the maximum number of digits ever displayed in total. In STD display mode, however, the display now shows 1.23456789012E-4 because the default STD mode is committed to showing all *significant* digits all the time, and if this can't be done under the above constraints, then it resorts to scientific mode to show all the significant digits. If you multiply the above by 1000 and display in STD mode, you'll see .123456789012 (this is the only range and mode which may *sometimes* show 12 digits after the decimal point, *if* no digit at all need appear *before* the decimal point). Nowadays, another object type (long float), which is used for temporary intermediate calculation results, may be displayed on the 49G stack, possibly with up to 15 significant digits, but UserRPL commands never produce such final answers, because they are inherently not as accurate to the last digit. In fact, the basic arithmetic philosophy of HP calcs, unique in this field, has been to perform intermediate internal calculations to 15 digits and then always *round* every answer to 12 digits -- not rounding just in the display, but rounding all final answers, whether displayed or not. These fine points in numerical computation (including one extra detail called unbiased rounding), combined with extra internal measures to retain high intermediate precision in some very tricky cases where other calcs don't, is responsible for more reliable results in general, even though merely carrying more digits may make other calcs look more accurate in a cosmetic sense, while they are not as truly accurate in the sense of how completely all of the *retained* digits can be relied upon to remain accurate. But who cares anymore? No one wears a tuxedo to school, and no one will keep paying for a tuxedo quality instrument to knock around in school, rather than in scientific research and engineering labs and in upper echelon boardrooms (financial models), which are the rarefied original environments for which the prior HP pocket calcs were designed (and that era lasted quite a while -- just about 30 years to date). ----------------------------------------------------------- With best wishes from: John H Meyers http://www.mum.edu ==== it's the 'append Alpha' symbol. In PRGM mode, switch to ALPHA, then press SHIFT XEQ. Hope this helps. Raymond Wee-Meng Lee schrieb im Newsbeitrag ==== Ahh, I finally found the answer how to run it after mucking around! Just go to user mode, and press the A,B,C,D,E keys to enter N,I,PV,PMT,FV respectively. Enter the number of months, N, and press the A key. When key A is pressed, it would jump to the routine at LBL A which would store the number in 01. After entering N,I,PV, to find PMT, press the D key. Wow! What a great calculator! weemeng ==== Wee-Meng Lee meinte LBL A = N LBL B = I LBL C = PV LBL D = PMT LBL E = FV LBL a = clear Registers You invoke the Programm by XEQ FIN and then it's use is similar to the HP12C HTH G237nter ==== Hey all, There's an HP-01 up for auction on Ebay if anyone's interested... http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?ViewItem&item=705133502 ==== Yes, in deed, if I have clear all variables in HOME - then it works (DERIV and DESOLVE) ! :-) Enno guide, the guide says that the calculator needs to be set in exact mode flag -105 clear and Numeric mode must not be set flag -03 clear. Also I have had these flags set as well and it didn't work, so I just clear all the variables out of home then it works. It seems to get the same error you get after I display a plot. Hopefully this helps you. ==== Check at www.hpmuseum.org. -- James ==== Albert Graef a 216crit : xcas is relatively usable, you can try it on a linux PC or even on windows (but it has more possibilities on linux). It's not only a CAS, since unlike on the 48 there is nothing available on the ipaq. There is a 2-d plotter, interactive geometry, a programming language with an interactive debugger, a matrixwriter. You can link it with the Gnu Scientific Library and some numerical functions are available (currently some linear algebra numeric LU, SVD, eigenvalues, as well as numeric rootfinders) No equationwriter, no 3-d plots, some CAS features are still missing and it's a *development version*, I don't expect a stable version before the end of this year. But it's a GPL soft, anyone can contribute (either programming in C++ or helping to make documentation...) About your questions re: the ipaq, you can run everything that runs on a desktop linux as soon as it fits on the screen and you have the memory for it (and it is open-source so that you can compile it) ==== Dear Daryl: First, make sure the 41 didn't just lock up on you. With the calculator off, press and hold the backarrow key and press ON. You might also try leaving the batteries out of the calculator for 2 days. Good luck! amccornack@cox.net ==== pull the batteries out for a day. ==== Since HP has officially ceased all calculator development, would it be possible for volunteers to band together and rewrite all the features of the calculator themselves? This way, the entire community could participate in bug-fixing, enhancements, customizations, etc. I think we could do something similar to what OpenBeOS is doing to BeOS (www.openbeos.org). In fact, shouldn't it be easier, because the calculator, however complex it is, is much simpler than a full-fledged, multitasking OS? I've been an HP fan for years, and just bought my HP49 two days ago, so I wouldn't know enough to contribute to such a project (yet). However, this could keep the development alive. By the way, HP is still manufacturing HP49's, even if they're not currently developing anything else, right? I know I'm wrong, so shut me up and put me in my place. --Andrew Huey ==== At this point it would be stupid to rewrite a ROM for a piece of hardware with a 4mHz processor. What would be really kewl is to write the HP ROM for the TI calculator. You would just have to get over the keyboard issue. This way, since TI still has calculator development ongoing you would probably be able to port it to a new calc platform. How kewl would it be to have the 10mHz processor with RPN and all the features of HP? The HP could live on in its reincarnate version. Micah participate ==== [crap removed] you're trolling, but you are doing it so subtly that not even you realize it. Why? This piece of hardware with a 4mHz processor happens to be one of the best calculators ever produced and a creation unlike any other you will probably ever see. That is one. The rest find it yourself. Troll. ==== Yes, and either method does save a bit of typing. When I mentioned the limit on the length of the command line, I meant the command line that is being sent to the 48. I suppose to be more accurate I should say that the 48 will put whatever it receives in the type C packet on stack level 1 and then execute OBJ->. Anyway, whatever you send to the 48 with Kermit's remote host command has to fit within a single packet. Well, actually, I did that last night, so I think that no one is going a bit too far. I surmise that the Kermit on the PC sends a type G subtype D command. On the calculator one can use D G PKT to get a directory listing from a server, either a PC or another calculator. Good question. Note that the 49G's FILER displays the type and size of variables but not the checksums. It would indeed. I have the PDL on my old PC, but for some reason or other I stopped using it and went back to using a text editor and Kermit. Still, I was surprised that they didn't release a fully updated version for the 48G series and 49G. And make certain companies a heck of a lot of money. -- James ==== Today I should be able to try the experiment and I will let all you know about. different but the concept is can I send and receive data from/to hp48s The doubt come from how the manual has been traslated in Italian: ....you can send and receive files... that's way I asked about data rather than files. Many thanks all of you! :-)) Massimo. ==== [Sorry if this shows up twice] Yeah, some kind soul gave us this link: http://www.satlan.com/survey/nicklin/data/pdl.zip But that's version 1.0, I think that the last version was 1.1, right? Hmm, if only an even kinder soul ... :-) ftp://ftp.stud.fh-heilbronn.de/pub/system/hp48/hacker/compiler/other/pdl/pdl .tar.gz Or use the right mouse button. Tom Wellige's HPShell is also worth a try! http://www.wellige.com/hpshell Bye. HPCC member #1046 - -0600) ==== Okay, JMP, here's how to get a complete listing of variables residing in the current directory of your HP48, just the way it's presented upon REMOTE DIR ( D G PKT) from Kermit SERVER: CAUTION: *48* ONLY, uses *unsupported* entry point, back up memory first! << PATH ->STR 10 CHR + :!appendVARS$: #2DC39h SYSEVAL >> I wonder where's that entry on a 49g? A sample (printed string) from my 48GX(R): { HOME JUNK } P5 68 Program 10744 LIBSL 88.5 Program 22700 LIBSM 114.5 Program 23319 off 17.5 Command 45169 dates 18 Global Name 47018 te 12 Global Name 11940 This gives you the variable name, size, type, and decimal checksum (no hidden variables, wrong sizes for commands and functions). For a complete catalog of your HP48, just print this for every directory. [r->] [OFF] -0600) ==== I have a suspicion that MEM DROP should have been inserted in the previously posted VARS pretty print program for HP48: << PATH ->STR 10 CHR + MEM DROP :!appendVARS$: #2DC39h SYSEVAL >> CAUTION: *48* ONLY, uses *unsupported* entry point, back up memory first! The above lists the current directory's variables in a printable (string) format, as would otherwise be seen only by sending a remote dir command ( D G PKT) to the HP calc's Kermit SERVER from another computer (or calc) which implements Kermit command packets and displays the results. Sample printable output: [variable name, size, type, and decimal checksum] { HOME JUNK } P5 68 Program 10744 LIBSL 88.5 Program 22700 LIBSM 114.5 Program 23319 off 17.5 Command 45169 dates 18 Global Name 47018 te 12 Global Name 11940 Can anyone find the 49g equivalent (and is it stable?) [r->] [OFF] ==== Veli-Pekka Nousiainen escribi227 en el mensaje to Ok, Hermano Pedro. ==== What would you expect from a calculator that can be implemented today? (please be serious, I know that HP48 users have over-average imagination ;-) I am thinking of a small design based on MIPS series processor from IDT. It might be nice. But a mono screen, I think. Not too much processing power, but better than a '49 anyway. Keyboard like the '48, can't think of anything better. Most probably RPN. Please re: to sskowron@et.put.poznan.pl, Stanislaw Skowronek ----<>---- United States Law prohibits disposition of these (commodities) (technical data) to the Soviet Bloc, Poland, Hungary, Romania, Libya, Laos, The Peoples Republic of China, North Korea, Vietnam, Cambodia or Cuba unless otherwise authorized by the United States. (on an invoice from 2002) ----<>---- ==== I like my HP48GX as it is with the exception of the processing speed. Just take the 48GX and make it much faster. :D Mark R. (technical ==== I'd be perfectly happy with an HP49G in an HP48GX's body... ;-) And a faster processor. -Jerrod H. (technical ==== With the man hours needed to be invested to get to a calc like the 48/49 and TI's dumbing down of College and University expectations from a Calc. I'm resigned to either: a long wait for a 49 replacement, or waiting for a cheap PDA capable of running EMU49 with an external numeric keypad. HP49 with: - keyboard action that registers press at around 50% travel and softer touch for faster action. I prefer the keyboard scheme and schematics on the 49 over the 48. Don't mind the rubber keys either. - Same toughness and build strength. - Simple real arithmatic to be very fast. This should be bread aand butter for a calc, it's unfortunate the the new object types slow the 49 down over the 48. - Slightly Sloping display face so you can read the display at lower angle and have calc further away. - Longer rechargable battery life measure in months not weeks or hours as on PDA's. Other nice wishes: - faster processor - More user memory. With all the day to day rubish I'm putting in there I can see myself running short of room in 5 years (I intend to keep my 49 for a long time). And yes I am putting programs and other such things in port 1 and 2. - say 50% higher resolution display but same size for clearer fonts. - smaller body in plan and thinner. HP42/32 is perfect size but display and keyboard restrict design. But there is a lot of wasted space on 49G for logo's, trim and curves. - More day to day features for solving and operation of calc rather than more extensions of CAS. Best of luck (technical ==== imagination * Hp49 in an HP48GX case * Faster processor * USB port * Better documentation than the HP49 has Santos Lucero ==== hp48gx, rom rev R with a 128K mem card and a 1M card. ;) ==== I'm running 1.19-6 on the 49 and I consistently have problems using the CUT function in the editor. After selecting, if I hit CUT, it either hangs and needs an ON-C to recover, or completely locks up, requiring a paper clip reboot. I have re-flashed but it still occurs. Any ideas? Simon ==== I use 1.19-6 and often used CUT, and have never had any difficulties with it. Could it be that you have something assigned to that key? Maybe an added library is causing problems? -- James ==== I concur. Some library has probably assigned itself to that key during a setup config. Get Keyman and find that rascal.. Mike ==== One of my students accidentally found a bug in the Solve aplet of her 39G the other day. Just START the Solve aplet and then go to the MODES view and change the Numeric Format to 'Fraction'. Now go back to the SYMB view and try to enter a formula such as A=1.5B involving a decimal point. You get the error message Invalid syntax. If you replace the dot with a comma then you find that the equation is now accepted. Changing into any other view such as the APLET view you will find that all the sizes are now given using commas instead of dots. Yet the MODES view is still showing the Decimal mark setting as 'Dot (.)' This bug occurs on the 39 and 40G and also the emulator. As far as I can see it will not cause a crash and changing out of Fraction mode fixes it immediately. In the 38G the modes of the aplet and the HOME were independant but in the 39/40G they are linked. There was considerable discussion during the design as to how to handle the Fraction mode when in an aplet and, as I recall, the decision was that if Fraction mode was chosen then it would be enforced in HOME but would revert to Standard mode while in the aplet. Clearly this is not happening correctly here. ==== do you have correct setting as for the Decimal mark in Home Modes? If you have comma there, it should caused the problem. Johny ==== Decimal mark in MODES to be 'Dot (.)' despite the fact that all the displays are using commas. I suspect that the reason goes back to the 38G. On the 38G the only aplet which allowed the user to change the numeric setting was the Solve aplet via the NUM SETUP view. Remember, in the 38G the setting for the aplet was independent to that of HOME and, apart from in Solve, there was no way to change the aplet setting except via directly assigning values to the aplet vars 'Format' and 'Digits'. These vars were independent from the HOME equivalents of 'HFormat' and 'HDigits'. On the 39G these vars were aliased to each other so that a change in the MODES view affected the aplet (and HOME) but changing to another aplet changed the setting again. This was the reason why we decided to put the name of the aplet at the top of the HOME view instead of allowing the user to put their own name there: it was to remind the user what aplet they were using so that if they changed aplets they would remember that the Angle and Numeric settings would probably change too. Getting back to the Solve aplet on the 38G, that was the only one that had it's own access to the aplet settings of Numeric Format (in the NUM SETUP view). The difference was that the list presented when the CHOOSE button was pressed did not include the Fraction possibility. ie the list was one element shorter. What I suspect is happening is that the arrays storing the choices for values of Numeric Format and for Decimal Mark are stored contiguously in the BIOS but that the array for Numeric Format is shorter by one element than the one displayed in the MODES view. Thus when the user chooses Fraction (the last choice) the choice overflows into the next array and affects the setting of Decimal Mark even though this is not visible in MODES. I'm very much guessing here since I was not in on the programming side but I suspect that the answer will be something similar to this. ==== [...] Nice to see you posting here again, Cyrille. The funniest part about this piece of program is that I remember having used the same sequence of instructions recently, and it was probably on your advice. -- David Haguenauer ==== TI Calc Con '02: http://www-s.ti.com/cgi-bin/discuss/sdbmessageview.cgi?databasetoopen=calcul ;-) -- Bhuvanesh ==== Didn't the guys at ACO say they will release the 1.19-7 ROM before going? At least that's what I read on HPCalc.org. ==== it didn't happen. Since then, I'm under the impression that the problem is getting permission from HP. -- James ==== I do have a new ROM ready for the HP49. Unfortunately, HP never answered my questions regarding if I could publish it or not. So I won't Regarding the CAS manual, try using the Advanced User Guide, all the commands are described. Otherzise, on www.hpcalc.org you have a complete CAS manual in English You should find all the information you need there going? At ==== Uhm.. he could save the new rom file somewhere, just to be sure he doesnt loose it; and then he could press the new document button and take a note of the url, just to remember it; and then he could press the button labeled send to... ;)))) -- The set of solutions is never empty. Two solutions together form a new problem. -- Mycroft Holmes ==== Yes, I like the idea of saving it on an FTP site. I think we should put together a petition to send to HP for the ROM to be released. I have found a few bugs in the latest. I especially like the part where it automatically reboots! Micah At ==== Seriously!! This stinks! ;-) C'mon HP... let us have it!! -Palm Man note labeled ==== Hewlett Packard is gone as we know it - Carly dropped the ball. Don't keep your hopes up. ==== Oh no! There is a Rom update but HP hasn't bothered to respond to give permision! That's terrible. thoughts to. ==== start. I did, however, fill out that stupid web form that allegedly sends a message to her staff. I doubt it really goes anywhere. Micah note labeled ==== I'll sign any petition in hopes of getting the ROM. Anyone up for organizing it? found automatically going? ==== Could it be that the folks at HP might be a little confused regarding calculator issues? I can imagine the following: Hey boss, some French guy wants to know whether it's all right to publish a new ROM revision for the 49G. The 49G? Isn't that a calculator? Forward it to the Applied Calculations -- James ==== for the HP Calculator community. J. Yuan. ==== Al d216 Sun, 3 Mar 2002 13:12:17 +0100, Marco Polo al dgeve: If you own a HP49, maybe I have some programs I developed that may suit your needs: - An userRPL program who use Kern method for choose heat exchangers (no needs to use Kern tables because they are stored in the calc; undocumented prog :-(, and I don't remember well how it works, but I remember it's quite simple) - A sysRPL lib for Lee Kesler Tables (again, no needs for LK tables because they are stored in the calc. The library is in beta test phase and is documented in italian). I have the same for an HP48GX, but it is a collection of userRPL programs, it's very slow and it needs some extra libs to work) for exchanger program, I don't remember where I stored it). Bye, Marco /-----------------------/ Marco Tinarelli - Bologna - Italy tittiXXYYZZ_tt@tin.it (togliere XXYYZZ dall'indirizzo per rispondere) ==== A place I have made many purchases from (including plug-in enginerring/math cards for the HP48GX) is Calcpro website www.calcpro.com. I have had good dealings with their manager, Paul Nelson. ==== A place where you can purchase the HP48G series (i.e., for HP48G/GX/G+) Advanced User's Reference Manual is Calcpro website www.calcpro.com. I have purchased from them many times and have had good dealings with their manager, Paul Nelson. ==== In a sysRPL routine, I do some calculations and then want to display the results on the stack, wait for either a few seconds duration or wait for a keypress (preferable) and then continue. I've tried different things with DO LOOP but it doesn't seem to work and GETTOUCH doesn't display the numbers on the stack! Help!! Simon ==== SH> In a sysRPL routine, I do some calculations and then want to display the SH> results on the stack, wait for either a few seconds duration or wait for a SH> keypress (preferable) and then continue. There are many ways to do something like this, for example WaitForKey 2DROP or SysDisplay VERYVERYSLOW Hope this helps. - Carsten ==== Carsten, anything until I hit a key. What I need is a way to display the stack and then wait for the keypress. A code snippit follows: ID Lev1 @ put var on stack %8 @ put 8 on stack %- @ perform subtraction WaitForKey @ display result and wait for keypress 2DROP @ delete key info ... @ continue with routine How can I force the calculation results to show on the stack and then wait for the keypress? Simon <5906b20c.0203040825.18a790af@posting.google.com> ==== SH> Carsten, SH> anything until I hit a key. What I need is a way to display the stack SH> and then wait for the keypress. A code snippit follows: SH> ID Lev1 @ put var on stack SH> %8 @ put 8 on stack SH> %- @ perform subtraction SH> WaitForKey @ display result and wait for keypress SH> 2DROP @ delete key info SH> ... @ continue with routine SH> How can I force the calculation results to show on the stack and then SH> wait for the keypress? Did you read my entire reply? Have you tried SysDisplay? - Carsten SH> Simon ==== That did work for me. I was trying to get the keypress method, but this will do for now. Simon <5906b20c.0203040825.18a790af@posting.google.com> -0600) ==== Don't expect anything to display automatically; SysDisplay (in the System Outer Loop) is what normally re-draws the entire screen (only when the key buffer is empty), but that occurs only when your program has returned (ended)! Suggestion: Investigate DO>STR and Disp5x7, which are employed by the user DISP command. Are you sure that you can't do everything you want in UserRPL at this time? (But of course there's no adventure in that :) [r->] [OFF] ==== Does anyone know of either a site or hard copy source for a user's guide to -- Scott M. Krasner skrasner@carolina.rr.com ==== Suddenly, my calc has started to output numeric results for all inputs. 2 3 / becomes 0.666666666667 instead of 2/3 as it should be, 2 SQRT becomes 1.414235... etc. How can this be changed? Flag -2 and -3 are both clear. thanx, Artur ==== Using Erable? 15 SF (Exact mode) Artur Meinild escribi227 en el mensaje becomes ==== Yes, I have erable, but the problems I encounter are with the built-in commands, such as standard division and squareroot. I'm convinced that until today my calc wouldn't convert squareroot 2 to numeric until I did it manually, and that it wouldn't round fractions to decimal fractions... Can anyone shed some light on this? - Artur ==== If I'm mot wrong (too much time with MK and Erable), the original 48, always evaluates 2 3 / and 2 squareroot... but not ' 2/3 ' and ' V2 ' (where V is the squareroot symbol)... Artur Meinild escribi227 en el mensaje commands, my calc it light on this? ==== Which program are you using on the PC? Do you know whether you can establish communications with another calculator? Am I correct in thinking that recently means that at one time you could get your calculator communicating with the same computer using the same cable? In any case, make sure that the communication parameters on the calculator and PC match. On the calculator, press APPS, choose I/O functio..., Transfer..., and set the options there. On the PC, pay particular attention to the speed (or baud) and parity. -- directly from your menu, without bothering to wade into the MODES CAS application menu. If you use the MENU command, then your menu is permanently saved in the current directory, and can be invoked at any later time via left-shift CUSTOM (see the MODE key). Example: { STOVX DERVX INTVX SOLVEVX ISOL TAYLR } XCASM [T]MENU Do HOME 256 ATTACH before entering or downloading the following: << 1. << IF { 14. 19. } OVER TYPE POS THEN DUP 1. ->LIST 'XCASV' 2. ->PRG 2. ->LIST END >> DOLIST >> 'XCASM' STO @ The following program is used by the menu made by the above: << -55. CF RCLF RCLVX DUP IFERR RCL THEN END -> fl. vx. xx. << -120. SF IFERR vx. DUP #8C27h SYSEVAL VTYPE -1. > { 316. DOERR } IFT EVAL THEN 1. ELSE 0. END ERRN SWAP xx. vx. IFERR STO THEN DROP2 END fl. STOF { DOERR } { DROP } IFTE >> >> 'XCASV' STO Backup your memory first, in case you enter #8C27h incorrectly :) [r->] [OFF] exactly as i guessed. did you open it up already, see what is inside it? is it rpn? <3C8955C0.80713DB4@miu.edu> -0600) ==== No need; mine's transparent :) It's the standard Sharp logic 4-banger (clear and press 1 + = + = + = ..., get Fibonacci sequence :) This pure junker advertising piece has absolutely nothing to do with HP (but maybe you'll find C.F. on a street corner trying to sell them, with free stock certificates thrown in, if HP sinks while she's at the helm :) [r->] [OFF] ==== [snip-snip] use HALT for your debugging uses... just throw it in the middle of any program and run it. When it HALTs, just do SST and once you are over the part you wanted debugged, just hit CONT. handy. ==== This is not up to me unfortunately. GPL. So nothing will prevent the release of a new ROM from then the But what happen if you make your opinion based on a narrow view? You are upset which I can understand but there's no need to be so harsh. agree Because I have been away for quite a long time and now my work doesn't involve much the HP49 anymore. I have to make a leaving. But don't worry, I'm still focusing on education, who knows you may hear from me sooner than you think :) have I usually tend to become defensive if you criticize something I've worked hard on, natural reaction I guess. I still believe the HP49 is the best thing you could have get in this situation. I did mean that, I don't like what HP has became. I just hope it's not going to be the next Digital, I believe that the people who are now leading HP have no passion about what they are doing except money making. There used to be a time where what you were making and how you could make a change was more important than how much money you would get out of it. But that's the past unfortunately. ==== Excellent news! -- This message was written with 100% recycled electrons Pivo ==== NOW you tell us?! (note: you might have said this somewhere else, but i never saw it) Wait one year... hm... are we going to get the *rom* from you, or just the sources to MK&Erable? JY, can you please, for the love of everything good, why the hell didn't you/they just use the 48g build quality for the 49g? ==== Wooohoooo! Great! Thomas -- Thomas Rast If you cannot convince them, confuse them. -- Harry S. Truman ==== auggggh.... That's precisely what I feared to hear from you... I wish I was born 5-10 years earlier. Samuel Thibault ==== Doesn't HP still own the copyright on the GX source? IANAL but wouldn't this prevent those parts of the 49 source that are derived from the GX source from being released? Or is HP going to GPL the GX source as well? Could you be so kind as to clue us in to what's going on. :) ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. -0600) ==== Reuters news posted on Fri Mar 8, 9:15 PM ET: http://story.news.yahoo.com/news?tmpl=story&u=/nm/20020309/bs_nm/tech_compaq _calpers_dc_10&cid=580 Nail-biting suspense :) [r->] [OFF] ==== Listed these on ebay, no reserve, started at $9.99 Thermal printer 82240b http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=1338033659 Business consultant 19bII http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=1338028546 Printer paper http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=1338037035 -- ==== Unfortunately, none that I know of. What you can do if you have enough free user or port 1 memory is RCL the entire directory, temporarily STO it under some name within user memory, port 0, or port 1, RCL or EVAL the variable that you want using the path name, and then purge the temporary directory using PGDIR, or if you used port 0 or 1, then PURGE will also work. -- James ==== If Test is a directory into port 5 with a variable A inside, 5:{Test A} RCL, doesn't work... Try it, please recover ONLY the object I want, not the whole directory (as I am doing now) Raymond Hellstern escribi227 en el mensaje ==== As I said above, I think the solution must be a RCL raplacement, but XRCL of Hack Library doesn't work... James M. Prange escribi227 en el mensaje ==== you're right. I know he has developed some fast access routines for hidden ports, which are used in his Filer48. Raymond R Lion schrieb im Newsbeitrag ==== Raymond Hellstern escribi227 en el mensaje -0600) ==== That is, indeed, the purpose of a special form of RCL (using a path list), which was designed, on the 48SX, to locate a specific variable and put its direct memory address pointer on the stack (thus recalling just that one variable), in pretty much the same way for any directory, whether currently stored in a user variable or in a port. In the 48GX, some of the ports (2-33) are no longer directly addressable in memory, however, which is what challenges the old 48SX programming. Now, the 48GX does manage, all the same, to recall entire port objects, but to do so for these covered ports, it now has to temporarily map the desired port into the same address space as port 1 occupies, then it must copy the desired object into temporary memory, so that when the memory mapping of port 1 is restored, the copied object is not covered again by port 1. In principle, while the covered port is temporarily mapped into memory, it should be possible to find any desired individual variable within any stored directory object, and then copy just that object, but RCL for 48GX port objects seems to have its own independent programming, which does not both remap ports and handle a list at the same time. I took a quick peek into the system programming for RCL, but I was not able to untangle it and come up with any simple alternative (one of the challenges in so doing is to try to manage to use only supported or stable functions which have never moved); there are many more people who know much more about this sort of programming, however, and I leave it to them to either try it themselves or to give up, as I did (and perhaps even HP did, if this was complained about before the final ROM version of the HP48G series in 1993). The 49G can do this form of RCL in all ports, however, so perhaps you can just copy its program into your 48GX, or replace that old dull-looking thing with a flashy 49G instead, which doubles your user memory to begin with, and also eliminates replacing those pesky RAM card batteries :) [r->] [OFF] ==== when you say taht old dull-looking thing? You are... (I can't find the words) ;-) John H Meyers escribi227 en el mensaje ==== For some keyboard shortcuts that have been added since the last update of the FAQ list, look under What's New from Release 1.19-5 at www.hpcalc.org/viewzip.php?id=3240&file=49.html or www.epita.fr/~avenar_j/hp/49.html. -- James ==== Some of you have requested some improvment in Expresso (a spreadsheet for hp49), so I have improve horizontal scrolling and calcul are executed much faster. You will find the new version on www.hpcalc.org in some days or on www.cberger.net. For the user of the previous version : there are incompatibilities, if you want to use the new version you will have to execute e2->3 on your sheet and then change the $LyCx into LC(y,x). ------------- Cyrille Berger ------------- -0600) ==== The Casio Card Watch MQ-11 on my desk led me to the site below: http://users.hoops.ne.jp/s-osaki/ Got an MQ-11? Do you know that it also does Biorhythms? I believe it contains the exact same chip as the Casio Biolator, and is merely a smaller version of the latter, although Casio suppressed any info mentioning this extra feature (thus making it an Easter egg :) [r->] [OFF] ==== First the flip-over-lcd-cover promo calculationator 4-oper-with-square-root pos, then this. what next? http://www.collectmad.com/collectibles/stcal.htm <-- this? ==== How do you use ptr in system RPL? I would like to make a header file with words that aren't in the compiler's table. I guess ptr xxx means ptr with three bytes that represent the five nibble address. If so, how do you enter the three bytes? ==== Of course I can, I even bought a pair of sun glasses be But it is. You make incorrect statement regarding how HP was funding its department. ACO had a budget, and quite a big one. Otherwise how can you pay so many engineers ? I can't release any detailed information otherwise I will see HP lawyers at my door again. The HP49 has been correctly funded and so did Xpander. Xpander (and any new development on the HP49) has been cancelled because HP focused on a different global strategy which didn't involve education anymore. you can always argue this decision but that's a different story. is No. What has been said is that after the official release of the HP49, all the engineers working on the HP49 have been assigned on different projects and that's perfectly normal. Not that I liked it, it just what usually happen. So all the development made after ROM 1.18 has been made thanks to the free time of many engineers. play That is incorrect. The HP6S project has been started after the HP49G. It hasn't been a big success either. I don't think you can buy a 6S anymore. Last time I've checked it was not in the HP catalog. Corvallis-knowledge hope The only reason is that nobody from the old Corvallis team was working for HP at that time except Jim Donnelly, and Jim's involvment in the HP48 was minimum as he worked on the HP library card (periodic table, solver etc..). I've met Diana Byrne on this matter just 2 weeks before she left for TI. We had access to all the ressources available at the time. got The ROM was incomplete that's correct, but the main problem was the incomplete HP Tools which was not available on the new HP unix platform. But what can you expect from a 10 year old project that has been relocated 2 times over the world. I am not. I still believe that the HP49 was developped during the good time of HP. It started to deteriorate AFTER (about 2 years ago) ==== I don't see any difference because in both case the user has to upgrade his calc by himself, and if he want to upgrade then he probably knows that the beta 1.19.6 is much more stable than the so-called stable 1.18. I agree about the doc, but it seems that unfortunately people do not read the docs hence managers decided to spare money providing doc on the web (and if you look at TI it's the same as HP). step by step is something that depends on the country habits to learn maths. It is well adapted to the French way, I don't know other countries way. And what whould you expect? same! Yes a student should know the main methods for integration before using the s-b-s feature. The calc shows what method it uses. yes, he should. The calc does not replace the teacher. Same applies on the other examples. I don't think specialized OS will survive. Probably only pocket PC, linux, even palmOS is not sure. We'll see. One of the main advantage of e.g. linux is that you develop your app for the desktop and just cross-compile it for the PDA it's a 2 minutes work to port. BTW I don't use java, linux is not a synonym to java by far. ==== I'm pretty much a newbie. A respected electronics Professor recommended the RPN way, and I'm so glad I followed his advice. It enabled me to score significantly higher marks than my classmates. In studying precalculus I find the Equation Writer and the Step-by-Step mode very useful. I love the solid feel of the 49g keyboard. I printed and bound Renee De Graeve's Calcul formel et Mathematiques avec la HP49G en mode algebrique, translated into English by Ivan Bertolotti, from the website www-fourier.ujf-grenoble.fr/~degraeve/usflan.pdf, and have found it a terrific resource. Nick Karagiaouroglou's Trig Marathon is a tremendous resource, as will be the upcoming Complex Number Marathom and Calculus Olympics. For me, HP is the place to be. I took a 4-hour electronics exam recently with a public utility. Their wording was a calculator will be the only study aid allowed during the test. I hadn't had any exposure to digital circuits, so I had 5 weeks to learn enough digital to pass the exam. I downloaded Robin Getz's Digitician from hpcalc.org website, which takes Karnaugh map data in decimal string format and uses the Quine-McCluskey algorthm to convert to a simplified Boolean expression. This proved to be very useful. I was able to guide my shaky Boolean abilities and check my answers with it. My digital textbook, Digital Systems by Ronald Tocci says, The following discussion (of the Karnaugh map method) will be limited to problems with up to four inputs, since even five- and six-input problems are too involved and are best done by a computer program.(pg. 122) It is interesting to note that Getz's Digitian, even while not written in machine code and sacrificing program memory and speed for low run time memory, is able to solve a 6-input circuit with 15 highs and 10 don't cares in 8 minutes. On the 3- and 4-input circuits on the test I took, the solve time seemed plenty fast, maybe around a minute. My wife says I love my calc more than I love her, and sometimes I'm afraid she's right! to bring a superior calc to market any time soon at under $400? Are you going to be able to get it on the shelf at the Best Buys and Circuit Cities and Office Maxes and their international equivalents next to the TIs? Are you going to be able to break TI's stranglehold on the schools? I sincerely hope you are successful in doing so. But it seems to me the odds of your success are maybe one in a thousand. And if you are successful, are you going to give your HP predecessors due credit? ==== James--- Believe me, you want external fixed-polarity DC input. Get a 5V wall-wart to supply the power. If you use AC, you'll need a transformer, rectifier, and regulator. You'll still need a regulator if you're allowing for messy DC voltage, _AND_ you'll need some sort of electronic switching to handle polarity insensitivity. All of this circuitry would take up a space roughly equivalent to the size of the 41 series card reader, and would generate a _SIGNIFICANT_ amount of heat. If you want something with a form factor closer to that of the 32SII, then you _MUST_ have external voltage regulation, and simply use a polarized DC input. As for the rest: I find the architecture of the Saturn processor to be a wonderous thing for numerical manipulation. A 4-bit data bus, although amazingly small by today's standards, works well when dealing with BCD. I would, however, like to see this processor further optimized and sped up. Speed cannot be too greatly increased, however, due to the heat that would be generated by the processor. Dissipation through the casing into one's hand would become uncomfortable, and power consumption would increase as a result. Optimizing the instruction set and architecture would greatly benefit the processor and its implementation. I only used the IR port on my 48GX once, and that was just to play with it. Ditch the IR capability, and add USB. RS232-C could be retained for backward compatibility with older PC's. _OPTIMIZE_THE_ROM_, especially in the areas of display update. When people are able to write 4-level grayscale games that play like lightning, restricted only by the decay rate of the LCD itself, it shouldn't take 15 seconds to load up and display or scroll/refresh a directory listing. Improve memory management techniques. Make the UI easy, speedy, and consistant. Have two versions: One with full graphic capability (I agree with better resolution/same physical size), and one with only alphanumeric capability. Have both versions be able to drive a 800x600 monitor by means of a USB device (remember the disk drives and monitor for the 41 that used HPIL?) or PC interface program. The PC interface should be able to use the calc as if it were the CPU. Am I asking too much? I don't believe so. Yes, I want a faster, smoother calc. I also want one that is easier to use with a PC. I want to retain the wonderful feel of the 48G series. With today's technological capabilities, it should be doable. I think most of it can be done in software, however. Henry C. Gernhardt, III ==== from time to time. I know because I've written a paper about the HP38G and intend relevant to the HP38/39/40. as was your 'graphing tip' in V20N6. Is it worth becoming a member? The strong point of this club is the rich diversity of interests of its members: some people collect calcs, some use financial models, some like programming (both at user and system level), maths, science, education, ... By joining, you'll become a valuable resource for the rest of members. HPCC member #1046 - ==== Just to follow up, actually what I'd like is a general purpose delay routine similar to what we had in previous calcs such as the 41, where you could display a value on the screen for a predetermined time and then move on. Simon ==== The HP way is building a calculator that will last ones whole career. ( Like a Tank) ie th HP67, 41 etc. There are some older professors here at school that still use HP67's and 41's. Mind you, they have had to do a kludge and rebuild the battery packs, but everything else still works fine. Harold A. Climer Physics/Astronomy Lab Instructor U. Tennessee At Chattanooga ==== Our instructor requires us to show our work on paper and hasn't included graphic calculators as a requirement, which I'm glad she has done. However, I'm in calculus and people aren't sure how to simplify quantities with exponents....something fishy here in the states ! Please students... just don't forget what you learned in algebra, cause I want more time spent in class on the current subject and topic ! I like to use the HP49 with my assignments because I get to learn how to use the calculator efficiently, check my work, and see what limitations the calculator has, if any, with current problems. By the way... this discussion has become pretty comical at times. I think if we just ignored the very first comment (although hard to do) they would have felt very dumb ! solve If you're in trouble because your calculator can't do a LaPlace transform, then you should quit school. The idea is that you learn how to do it by hand, and then you don't need the calculator other than for checking maybe. Pretty sad when students rely so heavily on a calculator - I get the picture that that's often the case in the States. ==== Is it possible to write a function in RPN on the 49G? I spent a good part of yesterday looking in the User Guide, the FAQ, the hpcalc site, tried experimenting on my calculator -- still haven't a clue. Take a trivial example. To create a function which doubles the number on top of the stack, I can put this on the command line: 'DBL(X) = 2*X' then press DEF. But that's algebraic notation. How do I write a doubling function in RPN? The actual function I'd like to write will be ugly with nested parentheses if I have to use algebraic. In RPN it's easy. -- ==== Try using the editor to view the DBL function you just created. That should give you a hint how to write functions in RPN. Basically, you just use << >> (above +), fill in your function in RPN, put the function name on the stack and store it. If you don't want a ton of temporary variables lying around you can use nested << >>. for your DBL function, it would be << -> X << X 2 * >> ==== X 2 * {X what_you_want} | or : X 2 * what_you_want SUBST a60i55$gvq$1@news-int.gatech.edu... should ==== A simple solution: << 2 * >> Which isn't a function. Conforming to your request: << -> X << 2 X * >> >> Get the HP48 manual from www.hpcalc.org, it will certainly be a good source of information, regarding RPN-mode use of the HP49. -- David Haguenauer ==== Am I the only old-timer who, when asked to double a number, does an ENTER + ? No constant entry (slow and many bytes); addition, not multiplication (faster). Jim Horn (Joe is old as dirt; I'm older than Joe) ==== I did about a year or so ago, but I don't know if they'd still do that (Doubt it) -Palm Man patient, aprr. keys earlier a recent ==== I had the same problem - bought a 49G with serial number ID93.... in November 2001 (in the US). I sent it to HP support in Oregon, complaining about keyboard, screen cover, and the serial port error. They sent me a replacement within a couple of weeks, but it was an ID94...., so still Indonesian made, with the bad keyboard (this one is actually worse than the one I had in the first place) and the rainbow-effect screen cover. So not much improvement there, but at least the replacement was free and the serial port error is gone. So my experience was that it doesn't really help to send the calc in. Might be different in South Africa, who knows. Andreas ==== What did you say when you talked to them? How were you able to convince them? J. Yuan. (exept would The year and ==== Hahaha! That was a good one :D It actually is true... i hate verbosity and unneeded duplicates. That's why you'll never see me quote an entire post on newsgroups ;) cheers from Tsiros ==== Never knew I had such a flag. What ROM do you have, George? ;-) Greetings, Nick. ==== 170780 :D ==== You seriously need an upgrade! Adelfos-Petros OOPS! Mine is FI200758 ==== Look into using the calculator's SERVER mode and Kermit's REMOTE HOST command on the PC. The entire command line (including any translated characters and prefixed control characters) must not exceed 90 characters if CKSM option 3 is chosen, or up to 92 characters if CKSM option 1 is chosen , so you may well need to SEND data as named objects to the calculator first. The calculator's entire stack will be returned to the PC after the REMOTE HOST. It would probably be more convenient to use one of the emulators on the PC though. Visit www.columbia.edu/kermit for information on Kermit in general, and www.columbia.edu/kermit/hp48.html for additional notes on the calculator's implementation of Kermit. Not at all. -- James -0600) ==== With MS-DOS kermit, you can define a macro in mskermit.ini, e.g. define z remote host %1 %2 %3 %4 %5 %6 %7 %8 %9 This allows you to abbreviate a command line of up to 9 words from the Kermit window, e.g. z CLEAR MEM PATH VARS or, without a macro, you can just say (with no word count limit): rem ho CLEAR MEM PATH VARS ... If you want to see something which no one else even knows about, send a remote dir command from Kermit -- what I'd like to know right now is, what's the internal function which produces that directory listing? (which you can't see on the calc itself!), and wouldn't it be neat to be able to print that out? Of course, the HP48 Program Development Link (PDL) is much more convenient for this ancient (and perhaps obsolete) purpose; HP discontinued that product many years ago, and offered no equivalent replacement product, but by searching this newsgroup at Google, you can find a download site. The PDL ought to work with HP49G as well (command names can be added, and the optional library which can be transferred to the calc to speed up some operations is entirely in UserRPL). One handy use of the PDL is its built-in instant HELP for all the original HP48S[X] commands -- just type any command name (case-insensitive) then press F1! IMHO, the PDL is better than any GUI product that has since been produced; put it on your Libretto (if you have one) and laugh at how ten more years of technology have accomplished nothing really new, other than to slow down the interface :) [r->] [OFF] ==== statistics library originally written for the HP39/40 which was upgraded for the HP49. In the original application it was possible to plot PDF's, Probability Distribution Functions, according to the documentation but I have not been able to do so in the HP49. The documentation for the HP49 is the same one as for the HP39/40 and it doesn't say that plotting PDF's work for the HP49 but it doesn't say it doesn't. Any help from users of StatPack for the HP49 would be appreciated. Santos Lucero ==== Where can I get a sync cable for my 49G, serial or USB(preferred)? Sorry to here HP will NOT be making calculators any more - I prefer them over TI. - Jamie ==== I've been both sick & busy ==== Hope you are alright now. Now I see why you prefer degrees there in the cold North. ;-) Yeap! Absolutely. Aderfe-Petro is vocativ. Greetings, Nick. ==== Hope you are alright now. Now I see why you prefer degrees there in the cold North. ;-) Yeap! Absolutely. Aderfe-Petro is vocative. Greetings, Nick. ==== Well... I understand nothing at all about Adelfos-Petros but I'm glad to Nick Karagiaouroglou escribi227 en el mensaje ==== X Veli-Pekka = Brother-Peter = ... PS: I'm not a monk ;-) ==== had in mind is done. Worked perfectly first time! -- ==== What's SST? I haven't found a corresponding command. Also, DBUG seems to just execute a program and then halt at the first halt command. How can I use it to 'record' a program? Roman <3c83dca7$0$27027$7a628cd7@news.club-internet.fr> <3c841396@news.svn.net> <0kqb8u4i2iai4ps61ksr7dsegdjg636kef@4ax.com> ==== RK> What's SST? I haven't found a corresponding command. Also, DBUG seems to RK> just execute a program and then halt at the first halt command. How can I RK> use it to 'record' a program? execute the program step-by-step with the command SST. The SST command and other debugging commands are in the RUN menu: LeftShift CAT NXT NXT RUN - Carsten ==== Try this nice approach from Marc Moline: http://groups.google.com/groups?hl=en&selm=3soqov%24bp4%40diable.upc.es [Scroll down till Macro-Recorder] For a more complete solution, try porting the RECK command from Simone Rapisarda's Smartkeys library to your '49G (yourself or by persuading a bored hacker). Here's the latest manual: http://groups.google.com/groups?hl=en&selm=105368.2B687BE0%40wolf.sublink.or g S/SX version available in GD#9. G/GX version (1.59G thanks to Joe Horn): http://www.hpcalc.org/hp48/utils/interface/smartkey.zip Bye. HPCC member #1046 - ==== Roman ==== I have an ID95xxx calc, hoping to get a CN one instead. I'll have to call them and ask. the no result) off the easely. ID93704789 recent (hehe), ==== When trying to apply INTVX to `EXP(3*LN(X))/X^4`, I basically get the same thing back with some Xt parameters. I was expecting 1/X. Can someone tell me what this is and of any documentation to help me better understand this. thanks danny ==== That means that the HP49G can't integrate it. No, you were probably expecting 'LN(X)' or 'LN(ABS(X))'. Your original expression 'EXP(3*LN(X))/X^4' is equal to '1/X'. The '49 can integrate this if you get rid of the exponential. Use for example EXP2POW, TEXPAND, TSIMP or EXPAND to convert the expression to something that the '49 can integrate. Try it out: 'EXP(3*LN(X))/X^4' TSIMP INTVX -> 'LN(ABS(X))' (if Rigorous mode is on, else merely 'LN(X)') ==== Try previous manipulation LNCOL