A399 > Remember, there was a great deal of debate for weeks on usenet > about whether the 49G was a prank or real. > How should HP have intepreted that.. > People think our new flagship model is a joke? This was in part fostered by a sort of joker outside of HP who independently announced it on his own, one midnight, not even as part of the French calculator group to whom it was first shown on the *next* evening by HP. It was also around April :-) Avoid major announcements on April 1, especially; that's the date that inaugurated a newspaper in the USA, a newspaper called The Daily Column, which put the U.S. President/VP team of Lyndon Johnson (LBJ) and Hubert Humphrey (HHH) in color on its front and back pages, and headlined Don't count LBJ out on Viet and election. The big joke was, that even as the delivery trucks were unloading those newspapers in the pre-dawn, all over New York City, for first-day readers to buy later in morning, LBJ had already been on TV, announcing that he wouldn't run. HP's announcements were also very inauspiciously handled, directed only to France, and sort of like a Tupperware party, rather than like a fully professional formal product announcement. So it looked like a bunch of clowns having a party joke, and the look of the first product photos blended in well enough with that same idea, much as it still looks to me (but cool I'm not :) ******* !Flash! See the !NEW!HP68! (just announced at HPCC) at: http://209.197.117.170/item/product2.htm (already available, USD $399) First customer's review: . > You misunderstood me. I agree, that the poor hardware choice > of HP49 was a big mistake. What I was trying to say, is > that you claimed shortage of Saturn CPU as a cause of > bad HP49 sales. Then you misunderstood me. I was just saying that it is difficult to find 49 *this* year, and it might be related to shortage of saturn CPU (it's just an hypothesis, I don't have precise info). > I question, why HP was still producing and selling > older HP48 model in US that was using the same limited supply > of the Saturn CPUs badly needed by the new HP49 ? > Maybe it was the case, that HP48 despite higher manufacturing > costs and lower piece price was selling better than HP49 ? > I don't know, but here in France it was easier to find a 49 than a 48. I guess HP dispatched the 48 depending on the demand, the fact that CAS are allowed in examination in France and not in all US exams (one reason why the 40 is sold in France and the 39 in the US) which make the 49 more attractive. > > I'm not sure I undestand why is so. The switch from 48SX to 48GX > required just higher clock for CPU which 5 years from the > introduction of SX model was probably very simple with never, > faster parts on the market. The software update was extremally > limited. It DID NOT reqired new tooling for the keyboard, and any > of the hardware inside. The colors in the keys probably didn't even > need new painting machine. > I believe you underestimate the cost of a CPU upgrade. And the software did really improve between the SX and GX, the G ROM is twice the size of the S ROM. Of course writing the +256K of the G is not the same as writing the 256K of the S but the S is not the first calc of the series, it is also a followup of the 28. > 49 model requires completely new tools for all plastic and rubber > parts, so tooling investment was significant. Painting on the rubber > is a different process than on the plastic, so the painting machines > changed. The hardware changed from IR port to flash memory, > so there was some electronic work involved. The software was > significantly more updated between 48GX to 49G than from 48SX > to 48GX. And also the fact remains that HP48GX replaced 48SX > where 49G was produced parallel with 48GX almost to date > which means, you needed new plant space. > I don't buy argument, that 49G was made on a shoestring budget. > Not at least in comparison to the 48SX to GX switch. Then you are simply wrong. Just consider that the 49 project was around 1 year, and it was in parallel to Xpander. > > For the student standpoint yes, but why then LU factorization > is not educational and does not require exact math is beyond me. > I don't know how linear algebra is teached elsewhere, but here in France we teach matrix inversion or linear system solving without LU. LU requires a higher level of Gauss-Jordan pivot mastering, including the notion of good pivot choice and the corresponding permutation (PA=LU). It was first introduced to solve numeric problems because it was more efficient than matrix inversion, but this efficiency has a price which is added complexity (try for example to program rref and compare with LU). >>Nothing, the char polynomial is irreducible. >>Go pick up any textbook in linear algebra if you want to pick up >>examples that will be solved by the 49. > > > I think, you misunderstood my point. I know, that neither TI nor HP > can solve general third or higher degree polynominal and > this matrix requires such capability. What I'm questioning, is that > HP49 returns empty list instead of either error message or reverting > to numerical solver. I consider such answer as a bug, because the > answer in fact exists, but algorithm just cannot find it. If I would use > such command in the program I would assume, that the command > would return either valid answer or throw exception. But EGVL > or EGV does neither. It returns to the stack invalid unexpected answer > instead. That is inconsistent with HP filosophy of consistent user > interface. This is your opinion, my opinion is that EGV/EGVL should return the list of eigenvectors/eigenvalues that were found. The reason is that for some matrices, the char. polynomial will have some eigenvalues that can be found exactly and some not. If I decide to return an error I will not return the set of eigenv. found which is not optimal. This is not inconsistent since the user can check the size of the matrix or vector returned and make his own test in a program if he requires all eigenvalues to be found. I find myself easier to check sizes than catching errors. > I believe you underestimate the cost of a CPU upgrade. Was the CPU completely new with completely new assembler or just faster ? > And the software did really improve between the SX and GX, the G ROM is > twice the size of the S ROM. Of course writing the +256K of the G is not > the same as writing the 256K of the S but the S is not the first > calc of the series, it is also a followup of the 28. I owned 28S, 48S, 48G and 49G. In my opinion the least improved switch was from 48S to 48G no matter how much larger ROM was. > significantly more updated between 48GX to 49G than from 48SX > to 48GX. And also the fact remains that HP48GX replaced 48SX > where 49G was produced parallel with 48GX almost to date > which means, you needed new plant space. > I don't buy argument, that 49G was made on a shoestring budget. > Not at least in comparison to the 48SX to GX switch. > Then you are simply wrong. Just consider that the 49 project was > around 1 year, and it was in parallel to Xpander. Most automotive manufacturers build completely new vehicle model within a year including tooling of the plants. Iy the switch from 48 to 49 required a year, than probably switch from 48S to 48G required about a month. Jack > > Was the CPU completely new with completely new assembler or just faster ? > I'm not the best to answer this question. The CPU was faster but they also had to handle covered ROM access. > It is just an opinion of the user. Exactly what so much better is between > G model and S model ? A little bit nicer interface and engineering library > that eats tons of ROM but itself is not that great anyway. > Hmmm, I think there are a lot of numerical instructions (mainly in linear algebra but also PROOT) that were implemented between S and G. If you look at the development environment that was available around year 90 (e.g. no emulator), I would certainly not describe the task as a 1 month task because it would require more now! > Iy the switch from 48 to 49 required a year, than probably > switch from 48S to 48G required about a month. > > It is just an opinion of the user. Exactly what so much better is between > G model and S model ? A little bit nicer interface and engineering library > that eats tons of ROM but itself is not that great anyway. > > Hmmm, I think there are a lot of numerical instructions (mainly in > linear algebra but also PROOT) that were implemented between S and G. > If you look at the development environment that was available around > year 90 (e.g. no emulator), I would certainly not describe the task as > a 1 month task because it would require more now! any change. The same applies to the most of the algebra, math and even engineering library which was always available for S model as an card extension anyway. Adding couple of numerical algorithms even over the period of a several years doesn't seem and manufacturing standpoint the changes were minimal or none. Jack Seen: > shortage of saturn CPU What an excellent opportunity! Since the HP49 runs vastly faster on an emulator these days, could any existing Saturn emulator be run on a low-power modern CPU, and then the emulator be hidden inside the original calc case? If speed was the main factor that stifled sales of the 49, then sales of this souped-up speed demon ought to skyrocket! ((o)) . Jack! (who has several times moderated my posts here, so we are not lovers, nor are we personal friends :) that ACO has been profitable - and I believe him !! You are just bashing the HP 49G here. True facts: The CPU should be faster (4000000Hz is not enough) The manuals are lacking. (Buy Urroz books) The release ROMs were buggy (ROMUPLOAD 1.19-6) Everything else is a matter of personal taste like color, ENTER size & location, rubber vs. plastic keys, which CAS is better TI, HP, Casio, Derive, Mable (= best! :) what programming language?, whether to code in a PC or using the calculator (ML/SysRPL only possible with the HP) etc. TI is certainly faster in Integration and Solving (except for some special cases) when doing matrix number crunching (or simple loops) The OS integration in the 49G is astonishing and it has Alarms (with text or program) and a beeper for melodies, games and feed-back, like the alarms. V200 corrects partly the lack of time keeping (I have a watch on every room, on my both Nokia mobiles but I don't were a wrist watch, but it's nice to have the time abilities: you can do series port data-analysis every hour) I think both calcs have they strength and weaknesses and some even like Casio (which also has step/step) Yes - step/step - that was the main reason for many students in Finland to buy the HP 49G. Maybe Bhuvanesh comes out with a step/step for the TI? Veli-Pekka X > HP49 is a part of the loosing money equation. > I wonder how you could know if the HP49 makes money or not. X > this lack of sales simply ruined investment made to produce > this calculator. X > You are just bashing the HP 49G here. You might say that. I simply belive, that both calcs are equal. Jack I fear I'm drifting off topic, but as this newsgroup's official, unofficial Casio calculator using interloper, I thought I'd respond to one of your comments: On 19 Sep 2002 10:30:02 -0700, timite_h@yahoo.fr (Timit.8e Hassan) ><handhelds showing up that undercut the rich customers. How many more >competitors you need in this niche market ?>> >HP has done a fatal error thus the HP49 can't really compete with the >TI89 and HP has given up(for now) anyway. >The Casio efforts to compete with the TI89 are pathetic. >Have you used the Algebra FX 2.0 ? >Besides,the very limited CAS it belongs to the TI83+ category. I've used (and use) both the Algebra FX 2.0 and the (very similar) 2.0 Plus, and I think you're being a little harsh. I should immediately confess that it's been years since I last used my TI-92 (and I never used it much then), and that I've never used a HP calculator (blushing with shame). That said, I've been quite pleased with FX 2.0/+. Although marketed as a high school companion, it has had no difficulty with college-level Calc 101 and I suspect it will find 102 less challenging than I will. It also kept me company during a very basic EE exam (AC ckts), and had no trouble there either. Its CAS (actually Saltire Software's CAS, http://www.saltire.com/) is extensive without being overwhelming, and it was fully debugged before it left the factory . Were it to ever come to mathematical blows with a high-end, CAS HP, I doubt my little oriental slugger would win. Not if shear mathematical prowess were the sole determinant of victory. (Were reliability & convenience factored in, it might stand a better chance.) Comparing it to a CAS-less TI is going a little far. Richard Kanarek << I've used (and use) both the Algebra FX 2.0 and the (very similar) 2.0 Plus, and I think you're being a little harsh. I should immediately confess that it's been years since I last used my TI-92 (and I never used it much then), and that I've never used a HP calculator (blushing with shame). That said, I've been quite pleased with FX 2.0/+. Although marketed as a high school companion, it has had no difficulty with college-level Calc 101 and I suspect it will find 102 less challenging than I will. It also kept me company during a very basic EE exam (AC ckts), and had no trouble there either. Its CAS (actually Saltire Software's CAS, http://www.saltire.com/) is extensive without being overwhelming, and it was fully debugged before it left the factory . >> Note that i have said that BESIDES THE CAS it belongs to the TI83+ category. Its memory management and its user language are as limited as the TI83+ ones. And its CAS behaves exactly like a TI83 Flash apps thus you can't use it in program or add new functions to it. And functions wise,its CAS absolutely can't compete with either the HP or the TI ones. Perhaps that Casio wanted to use the Dos version of Derive but was input because TI bought Soft Warehouse thus Casio had to find a less powerful CAS. If you are pleased with its CAS then it is good for you but i don't think that someone used to the TI or the HP CAS would be pleased with it. Btw,there is a third party CAS flash apps for the TI83+. It can't compete with the Algebra FX 2.0 CAS off course,but i think you must know that there is one availlable. << Richard Kanarek >> > Were it to ever come to mathematical blows with a high-end, CAS HP, I > doubt my little oriental slugger would win. Not if shear mathematical > prowess were the sole determinant of victory. (Were reliability & > convenience factored in, it might stand a better chance.) Comparing it > to a CAS-less TI is going a little far. Surely you're joking. CAS-less TI?? If anything, it's the Casio and Sharp calculators that are CAS-less. Also, I think TI calculators would win in terms of reliability and ease-of-use. -- Bhuvanesh compared the Casio Algebra FX 2.0(+), which has CAS, with the TI-83, which doesn't (http://education.ti.com/downloads/pdf/us/graphing.pdf). I certainly didn't mean to denigrate the CAS in TI calculators that have CAS; I'm sure -- in fact, I know (based upon my long distant TI-92 use) -- that it works, although I certainly wouldn't speculate which is best or worst. Sharp doesn't make a CAS-equiped calculator (see http://www.sharpusa.com/products/TypeLanding/0,1056,s6,00.html), so you're quite right in your admonishment. I also think you're quite right about TI's calculators being easier to use, but that is many a result (IMHO) of the high-quality documentation TI provides (or provided?). Once you get the hang of using a Casio -- if you do, and I don't know why you'd try -- the Casio's are quite nice too. Cordially, Richard Kanarek >> Were it to ever come to mathematical blows with a high-end, CAS HP, I >> doubt my little oriental slugger would win. Not if shear mathematical >> prowess were the sole determinant of victory. (Were reliability & >> convenience factored in, it might stand a better chance.) Comparing it >> to a CAS-less TI is going a little far. >Surely you're joking. CAS-less TI?? If anything, it's the Casio and >Sharp calculators that are CAS-less. Also, I think TI calculators >would win in terms of reliability and ease-of-use. I don't mean to nitpick your recent, excellent post, but I thought I'd comment on one of your points. I have read it several times recently (from other posters), and I think it's time it was challenged. >> At the end all graph calculators have been banned from the exams of >> their university. >I fully support it. One should be able to learn how to solve algebraic >problems without the help of CAS. I don't support it. Judging by the ghastly writing that is often demonstrated in this newsgroup, there is clearly much need and great room for educational reform. Eliminating CAS/Graphing calculators, however, need not be part of it. My trusty CAS/Graphing calculator (a Casio Algebra FX 2.0 Plus, sorry) is my friend. On those occasions when I'm in a expedient mood, it Level work (assuming you consider any part of Calc 101 higher level ). When I'm in a mood to go it alone, it waits patiently while I muddle through the maths. It checks my answers for me, and helps me to find my mistakes (and my text book's). During the course of my course , I did tons of extra problems (not extra-credit problems, just extra problems) for additional practice. I can't imagine my having done so if I had been restricted to using my Pickett Microline 120 (genuine slide rule). In summary, as far as cheating is concerned, the fault lies not in your calculators, but in yourselves. Cordially, Richard Kanarek > I don't mean to nitpick your recent, excellent post, but I thought I'd > comment on one of your points. I have read it several times recently > (from other posters), and I think it's time it was challenged. >I fully support it. One should be able to learn how to solve algebraic >problems without the help of CAS. > I don't support it. > Judging by the ghastly writing that is often demonstrated in this > newsgroup, there is clearly much need and great room for educational > reform. Eliminating CAS/Graphing calculators, however, need not be > part of it. In my opinion, this is exactly what you are going to get if you replace paper and pencil methods with machines on the exam. There is no need to know how to solve the problem if machine does it for us. At least majority of the people will never study the material if the can pass it with a help of the machine. > My trusty CAS/Graphing calculator (a Casio Algebra FX 2.0 Plus, sorry) > is my friend. On those occasions when I'm in a expedient mood, it > Level work (assuming you consider any part of Calc 101 higher level > ). When I'm in a mood to go it alone, it waits patiently while I > muddle through the maths. It checks my answers for me, and helps me to > find my mistakes (and my text book's). During the course of my course > , I did tons of extra problems (not extra-credit problems, just > extra problems) for additional practice. I can't imagine my having > done so if I had been restricted to using my Pickett Microline 120 > (genuine slide rule). You are probably the exception. Form you posting I deduct that you like math and are capable to study details for fun. But this is rather exception than the rule with the majority of students. Most of them are either neutral to math or highly dislike it. Peoples nature is such that they tend to choose the path of minimum resistance. Instead of learning the tedious details, how to integrate, differentiate etc. If they can use machine to do the dirty work they will. This is perfectly accaeptable, if you do it to solve problems at work or you are looking for the solution of the problem unrelated to math, but on the math exam the whole idea of the exam is to find out if you have learned the idea of algorithms and methods to solve problems that the machine is capable later to automate for you. The whole purpose of the exam is to find out if you can solve the problem using your brain, not to find out if you can work with advanced calculator. > In summary, as far as cheating is concerned, the fault lies not in > your calculators, but in yourselves. For the teacher it is still a concern to deal with. The easiest and fair way to address the concern is to ban advanced calculator during exam. In fact I don't see a need to have calculator during math exam. There might be a need to have numerical calculator when the exam is fro mphysics and involves solving numerical problems. Otherwise why you need it ? Jack > In my opinion, this is exactly what you are going to get if you > replace paper and pencil methods with machines on the exam. There > is no need to know how to solve the problem if machine does it for us. > At least majority of the people will never study the material if > the can pass it with a help of the machine. I don't know what sort of exams you are talking about. If you can pass with the use of a calculator CAS then you exams are too easy. > For the teacher it is still a concern to deal with. The easiest and fair way > to address the concern is to ban advanced calculator during exam. > In fact I don't see a need to have calculator during math exam. > There might be a need to have numerical calculator when the exam is > fro mphysics and involves solving numerical problems. Otherwise > why you need it ? Lecturers are not silly when they set exams. In my experience, the way math exams are set, the only advantage an advanced calculator gives is better proficiency in calculation time. The same amount of 'thinking power' is needed whether using a 49 or a 6s. Andrew > I don't know what sort of exams you are talking about. If you can pass with > the use of a calculator CAS then you exams are too easy. Everybody starts learning algebra and calculus some day. For those who just started, HP or TI CAS is still too difficult to grasp entirely. For those such powerfull CAS is an unfair help. By the way, many integrals that TI or HP solves in minute or so would take about half and hour or so to solve on paper. Just try to solve manually: Integrate(1/(1+a*cos(x)),dx) TI does it in 20 seconds. > For the teacher it is still a concern to deal with. The easiest and fair > way > to address the concern is to ban advanced calculator during exam. > In fact I don't see a need to have calculator during math exam. > There might be a need to have numerical calculator when the exam is > fro mphysics and involves solving numerical problems. Otherwise > why you need it ? > Lecturers are not silly when they set exams. In my experience, the way math > exams are set, the only advantage an advanced calculator gives is better > proficiency in calculation time. > The same amount of 'thinking power' is > needed whether using a 49 or a 6s. At certain level yes.But first years of algebra are pretty much in the power level of TI and HP and would give unfair advantage to the owner of such machine. Jack << Just try to solve manually: Integrate(1/(1+a*cos(x)),dx) TI does it in 20 seconds. >> I need less than 3 minutes to find the antiderivatives and i find the result returns by the TI92+ almost unusable and only partially true. The TI92+ assumes that (a-1)/(a+1) is >0 when it can be negative,equal to 0 or undefined ! In case (a-1)/(a+1) is positive(a>1 or a<-1) the antiderivative is: 2*tanh^-1(sqrt((a-1)/(a+1))*tan(x/2))/sqrt((a-1)*(a+1))+Constant In case (1-a)/(a+1) is positive(a>-1 and a <1) then the antiderivative is: 2*tan^-1(sqrt((1-a)/(a+1))*tan(x/2))/sqrt((1-a)*(a+1))+Constant In case a=1: tan(x/2) In case a=-1: -1/tan(x/2) > Everybody starts learning algebra and calculus some day. For those > who just started, HP or TI CAS is still too difficult to grasp entirely. > For those such powerfull CAS is an unfair help. True. I was never allowed a programmable calculator in high school. > By the way, many integrals that TI or HP solves in minute or so > would take about half and hour or so to solve on paper. > Just try to solve manually: > Integrate(1/(1+a*cos(x)),dx) > TI does it in 20 seconds. Which exam did you get this question? How would you justify your answer if your calculator does it all for you? Andrew > Just try to solve manually: > > Integrate(1/(1+a*cos(x)),dx) > > TI does it in 20 seconds. > Which exam did you get this question? This might be for example question related to the area of the pie section of the ellipse, parabola or hiperbola. In my country it was known as algebraic/calculus geometry and I did had exams altough it was prior to the calculator era. > How would you justify your answer if your calculator does it all for you? For the complete idiot this would probably help little. But there are many ways, the calc might help the underachiever to gain the edge over prepared person. The form of the answer might lead one to guess faster the method used to solve this particular integral. The final answer might also help such person to check his own answer and if they do not match, then the person might be able to track down what he/she did wrong and fix the answer. In case of deriviatives on HP48 or 49 it can also give you step by step solution that would eliminate completely need to learn. > Andrew Jack > In my opinion, this is exactly what you are going to get if you > replace paper and pencil methods with machines on the exam. There > is no need to know how to solve the problem if machine does it for us. > At least majority of the people will never study the material if > the can pass it with a help of the machine. > > I don't know what sort of exams you are talking about. If you can pass with > the use of a calculator CAS then you exams are too easy. The Math classes at the University I attend make thier tests and finals so that you can solve them without a calculator. What they want to know is if we know the processes and formulas for solving problems. answers are not to be in decimal forms and we have to show all of our work. Physics is the same. I forgot my graphing caculator(ti 83-this was before I saw the light) and had to use my old scientific that I had used on my chemistry final(no graphing or text entry allowed) and I got a high B. > For the teacher it is still a concern to deal with. The easiest and fair > way > to address the concern is to ban advanced calculator during exam. > In fact I don't see a need to have calculator during math exam. > There might be a need to have numerical calculator when the exam is > fro mphysics and involves solving numerical problems. Otherwise > why you need it ? Why ban the calculators? We will use them in the profesional world, so we need to learn how to use them proficiently now. Just do what professors are good at, make the test tricky. I have never had a math teacher in college who has made a test so simple that I could have aced it, or even passed it, just by having a great calculator. Its like basketball, good shoes dont make the player, they can just help the player to be better. Loug1806 X >> At the end all graph calculators have been banned from the exams of >> their university. > >I fully support it. One should be able to learn how to solve algebraic >problems without the help of CAS. > I don't support it. X You are both right! One should first learn to solve the problems BUT - later one could solve the same (or more applied) problems using a CAS (may it be Casio, HP, TI, iPaq+xcas, Zaurus+F1, ...) X > Advanced calculators should have: > -very powerful low power custom CPU with BCD coprocessor for lightning > fast integers and > floating point BCD computations.Eventually the CPU could execute > System RPL or equivalent language in hardware.A bit like CPU with Java > chip. Something like this: http://www.ptsc.com/products/images/ignitemicro.pdf OR http://www.mips.com/products/s2p7.html Where the System Coprocessor could be Saturn for compatibility. At 100MHz this chip consumes power about as much as Yorke. Opinions? X > If the HP49 had much more powerful hardware,it could have been the > precursor of such tools. > I don't think that it is too late for such tools. > However it would be too late when PDA will reach PC power or when > handheld full feature PC(such as TIQIT and OKO) would be widely > availlable and cheap enough. > Why ? First you would be using common algorithm and identical > matrix storage algorithm. Second, maintaining symbolic solution would > eliminte round-off error (with the exception of final calculations), third for > an educational tool TI engineers did not to be efficient. Keep in mind, > that working with 10x10 numerical matrix on the calculator is a proof > of the user being unable to choiose proper tool for the proper job. And > of course, I don't understand what you mean by converting to symbolic. > There is no need to convert to symbolic. You simply built expression by > not executing intermediate calculations till the end. > You can't do that. You must evaluate intermediate results for two reasons: it would be much too time consuming first, and you need to knwo which pivot you choose, especially you must find a non-zero pivot. I don't have a HP49 with me actually as i am in Murroco for hollidays. his is the reason why i haven't answer sooner as it is difficult to have decent connexion speeds here. However i can solve your problem in almost 2 minutes(including typing) on my twin brother TI92+ with this method: Step #1: [cos(aa),-sin(aa);sin(aa),cos(aa)]->m(aa)(Rotation matrix) Step #2: [x;y]-[a*cos(t);b*sin(t)]->v1(PP1) Step #3: [x;y]-[a*cos(t+2*pi/3);b*sin(t+2*pi/3)]->v2(PP2) Step #4: [x;y]-[a*cos(t-2*pi/3);b*sin(t-2*pi/3)]->v3(PP3) Step #5: expand(texpand(dotp(m(-pi/6)*v2,v1)))->eq1 Step #6: expand(texpand(dotp(m(pi/6)*v3,v1)))->eq2 Step #7: eq1-eq2->e1 Step #8: eq2+eq1->e2 Step #9: zeros(e1,y)[1]->yy Step #10: zeros(e2,x)|y=yy->lx Step #11: yy|x=lx->ly Step #12: tcollect(lx)->lx Step #13: tcollect(ly)->ly or: In a script the following operations take 46 s: Step #1: [cos(aa),-sin(aa);sin(aa),cos(aa)]->m(aa) Step #2: [x;y]-[a*cos(t);b*sin(t)] (PP1) Step #3: [x;y]-[a*cos(t+2*pi/3);b*sin(t+2*pi/3)] (PP2) Step #4: [x;y]-[a*cos(t-2*pi/3);b*sin(t-2*pi/3)] (PP3) Step #5: expand(texpand(dotp(m(-pi/6)*ans(2),ans(3)))) Step #6: expand(texpand(dotp(m(pi/6)*ans(2),ans(4)))) Step #7: ans(2)-ans(1) Step #8: ans(3)+ans(2) Step #9: zeros(ans(2),y)[1] Step #10: zeros(ans(2),x)|y=ans(1) Step #11: ans(2)|x=ans(1) Step #12: tcollect(ans(2)) Step #13: tcollect(ans(2)) Where: ans(2)={a*cos(t),(b/2-a/2)*cos(3*t)} or x solutions ans(1)={b*sin(t),(b/2-a/2)*sin(3*t)} or y solutions The first couple of solutions represents the case where P=P1 while the second couple are the solutions we are looking for. Something else to add about my ability to solve the problem as you are obviously questionning my mathematic skills ? There is also a similar method with dotProducts. Once you have the coordinates of P,it is a piece of cake to compute the angle between axis X and PP1 line. Unfortunately,i don't have a HP49 with me but i would have used a similar method. Btw this problem barely includes square roots and doesn't include complex at all so how does it relate to what i was talking of ? Also even if the TI92+ is faster than the HP49,it is dramatic to see that for a problem relying mostly on TI92+ strengths (much more CPU power and powerful autosimplification),it is just 3 times faster than the HP49. One could think that the TI92+ would be at least 10 times faster than the HP49 for such a problem. Anyway,the kind of problem where the autosimplification is nuisible are for example expansion of rationnal fractions in C,the domain of complex. Let say that you want the full expansion of: 1/(x^3-1) It is not possible on the TI92+ without 2 tricks that most users will never find. Also good luck in trying one of the following operations: determinant,inverse or gauss reduction on the following matrix: [sqrt(2)+I*sqrt(3),sqrt(3)+I*sqrt(5);sqrt(5)+I*sqrt(7);sqrt(7)+I*sqrt(11)] Where I is the complex number. The funny part is that it is possible to compute the determinant of such matrix with cross product very quickly. In general any problem including complex and/or square roots(especially in denominators) can quickly become very difficult or impossible to solve on the TI92+. And i don't even talk of the stupid habit that the TI92+ has to split complex expressions even when it is better not to. Also,i have found that the TI92+ is unable to completely simplify some square root or logaritms of integer try: -sqrt(65537^2*65539) which should be simplified in 65537*sqrt(65539) -expand(ln(65537^2*65539)) which should simplified in 2*ln(65537)+ln(65539) And to conclude as you seem to be so proud about integration abilities of the TI92+ just 3 examples where the TI92+ is unable to find the obvious antiderivative of the function because of autosimplification: (2*x+1)/sqrt(1-(x^2+x)^2) antiderivative:sin^-1(x^2+x) (2*x+1)/sqrt(1+(x^2+x)^2) antiderivative:sinh^-1(x^2+x) (2*x+1)/sqrt((x^2+x)^2-1) antiderivative:cosh^-1(x^2+x) The TI92+ automatically expands (x^2+x)^2 and is trapped. How to read the memory available in each port in the 49G in ML? , i.e. Alejandro Henao Loaiza. [ahl] > How to read the memory available in each port in the 49G in ML? , i.e. Here is the code for the PVARS function. It returns the amount of free memory and the list of variable. You should be able to modify it to return only the free space quite easely. ( % --> {..} ) :: CK1&Dispatch real :: COERCE DUP #2 #>case SETPORTNOTAV FLASHPTR XEQBVARS UNCOERCE %8 %* ; ; ********************************************************** * This is the Vars command. * the only difference with the RPL command is taht this one * takes a SB instead of a real. * This command is based on the AccessVars entry of the * access methodes. ********************************************************** ASSEMBLEM $(5) (*)+5 SAVE GOSBVL GARBAGECOL LOAD D-5.A SKNC { *.ins GOVLNG =memerrS } % Get the whole memory C=D.A C+C.A C+C.A C+D.A RSTK=C GOSBVL MAKE$N C=RSTK B=C.A GOSBVL D1=DSKTOP GOSBVL POP# CD0EX D1=C % the the port SB A-1.A GOC .P0 % Port0? A-1.A GOC .P1 % Port1? A=0.A R1=A.A % R1: free memory / 16 LC(5) ACCESSBank0 D=C.A C=0.S % for the 16 banks do { C=D.A P==AccessVars C=D.A GOSUB _PC=C *.ins2 GOC .ins % list vars P==AccessModuleInfo C=D.A GOSUB _PC=C % get module info (to compute free mem ) C=A.A ASR.W ASR.W ASR.W ASR.W ASR.W C+A.A CSR.A A=R1.A A+C.A R1=A.A % add the module free meme to the total free mem D+11.A C+1.S UPNC % Next access routine } C=R1.A % Ca: Free memory GOTO .End *.P0 GOTO .End *.P1 port 1 C+A.A *.End R2=C.A % Save free mem in R2 CD1EX D0=C GOSBVL Shrink$ % release some memory A=R0.A D1=A D0=A % ready to transfor this string in a list! LC(5)DOLIST DAT1=C.A D1+5 % list prolog D0+5 C=DAT0.A C-5.A D0+5 GOSBVL MOVEDOWN % shift all data by 5 nibbles LC(5)SEMI DAT1=C.A % write ending semi A=R2.A A+1.A SKNC { GOSBVL ROOM CSR.A R2=C.A } % get the free mem LOAD A=R0.A DAT1=A.A A=R2.A GOSBVL PUSH#ALOOP % and push the whole thing on the stack !RPL 2) Can anybody explain me the instruction GOSUB _PC=C. There isn't a subrutine in the code whit that name. I don't have some entry points in my extable like memerrS, AccessVars and AccessModuleInfo. Where can I find a more complete extable (www.hpcalc.org???). 3) Where can I find information about access methods. 4) Forgive about my English. Alejandro Henao Loaiza. [ahl] > Here is the code for the PVARS function. It returns the amount of free > memory and the list of variable. You should be able to modify it to return > only the free space quite easely. > > > ( % --> {..} ) > :: > CK1&Dispatch > real :: COERCE DUP #2 #>case SETPORTNOTAV FLASHPTR XEQBVARS UNCOERCE %8 > %* ; > ; > > > ********************************************************** > * This is the Vars command. > * the only difference with the RPL command is taht this one > * takes a SB instead of a real. > * This command is based on the AccessVars entry of the > * access methodes. > ********************************************************** > ASSEMBLEM > $(5) (*)+5 > SAVE GOSBVL GARBAGECOL LOAD > D-5.A SKNC { *.ins GOVLNG =memerrS } % Get the whole memory > C=D.A C+C.A C+C.A C+D.A RSTK=C GOSBVL MAKE$N C=RSTK B=C.A > GOSBVL D1=DSKTOP GOSBVL POP# CD0EX D1=C % the the port SB > A-1.A GOC .P0 % Port0? > A-1.A GOC .P1 % Port1? > A=0.A R1=A.A % R1: free memory > / 16 > LC(5) ACCESSBank0 D=C.A C=0.S % for the 16 banks > do > { > C=D.A P==AccessVars C=D.A GOSUB _PC=C *.ins2 GOC .ins % list vars > P==AccessModuleInfo C=D.A GOSUB _PC=C % get module info > (to compute free mem ) > C=A.A ASR.W ASR.W ASR.W ASR.W ASR.W C+A.A CSR.A A=R1.A A+C.A R1=A.A % add > the module free meme to the total free mem > D+11.A C+1.S UPNC % Next access > routine > } > C=R1.A % Ca: Free memory > GOTO .End > > *.P0 > GOTO .End > > *.P1 > port 1 C+A.A > > *.End > R2=C.A % Save free mem in R2 > CD1EX D0=C GOSBVL Shrink$ % release some memory > A=R0.A D1=A D0=A % ready to transfor this > string in a list! > LC(5)DOLIST DAT1=C.A D1+5 % list prolog > D0+5 C=DAT0.A C-5.A D0+5 GOSBVL MOVEDOWN % shift all data by 5 > nibbles > LC(5)SEMI DAT1=C.A % write ending semi > A=R2.A A+1.A SKNC { GOSBVL ROOM CSR.A R2=C.A } % get the free mem > LOAD A=R0.A DAT1=A.A A=R2.A GOSBVL PUSH#ALOOP % and push the whole thing > on the stack > !RPL Alejandro Henao Loaiza schreef: > 2) Can anybody explain me the instruction GOSUB _PC=C. There isn't a > subrutine in the code whit that name. the access routine needs a return address and with a gosub you provide this a piece from the code C=D A GOSUB _PC=C C=A A there should be a subroutine in your program called _PC=C (in the Vars command it is not present but it is in the ROM) your _PC=C subroutine is very simple: *_PC=C PC=C if you want to know more you could take a look at example 9.2 from HP49 Assembly Language Examples http://www.hpcalc.org/details.php?id=5007 > I don't have some entry points in my extable like memerrS, > AccessVars and AccessModuleInfo. These are aliasses for 1, 8 and 14 IIRC -- This message was written with 100% recycled electrons Pivo On Sat, 28 Sep 2002 10:48:56 +0200, Peter Geelhoed >> I don't have some entry points in my extable like memerrS, >> AccessVars and AccessModuleInfo. >These are aliasses for 1, 8 and 14 IIRC Slight correction: I decompiled PVARS and compared it to Cyrille's source to check the above values. I think the correct ones are memerrS = #39E3 AccessVars = 14 AccessModuleInfo = 8 -------------------------------------------------------------------------- -- --- Jonathan Busby - before replying. > > How to read the memory available in each port in the 49G in ML? , i.e. Not quite what you want but this should give you an idea. The program displays the usage of the userbanks in flash. Sorry that the CODE isn't commented but I lost the source in a crash. :: RECLAIMDISP TURNMENUOFF ' :: APPEND_SPACE APPEND_SPACE ROT #>$ DUPLEN$ BINT3 SWAP#- ZERO_DO APPEND_SPACE LOOP &$ SWAP #2/ #>$ DUPLEN$ BINT7 SWAP#- ROTSWAP ZERO_DO APPEND_SPACE LOOP SWAP&$ ; CODE 001B5 8FB97608F029623447A20144164AF281842613610C6260808249BB62D281A41CB4652381AF 1C 13434B21301441648F1276281AF10818F098DFE86281A40C812CAC6CAC6CA81B27D9F81AF1C1 3 4AF210A10910B1FA02048082083447A2014416434119201547169A0C5EE34B213014416481AF 1 413016914B31FF962153135962B016F16F1671771433404B209628016F1631C5146E61441691 4 3146C2144137C21356C9F81AF1434A5000C21341378185AF8185AF8185AF8185AF1441691368 1 AF0C6820808249BB62D281A41C812CAC6CAC6CA2181B277DF687E BINT0 { } BIND BINT0 BEGIN :: DUP BINT10 #= case :: #1+ BINT8 #/ DROP ; BINT15 #= case :: #1- BINT8 #/ DROP ; ; #1+ NTHCOMPDROP INCOMPDROP FREE: #40000 ROT#- #2/ #>$ DUPLEN$ BINT14 SWAP#- ROTSWAP ZERO_DO APPEND_SPACE LOOP SWAP&$ DISPROW4 BANK: #8+ #>$ &$ DISPROW1 # BYTES DISPROW3 DEL BAK DISPROW8 DEL LIB DISPROW7 BAK DISPROW6 LIB DISPROW5 GETKEY NOT_UNTIL ABND ; @ -- This message was written with 100% recycled electrons Pivo Hey All I have flashed my HP49G to HP39/40 and now i canĒt flash it back, kan anyone help me, Joakim http://www.epita.fr/~avenar_j/hp/49.html 2. On the 49G: Enter diagnostic mode by turning the calculator on, holding [ON], and pressing [F4]. While the Tests screen is showing, use your thumb to hold down [+] and [ENTER]. While holding down [+] and [ENTER], press and release [ON] to show the No System menu, and then release [+] and [ENTER]. Press [4] to select terminal mode. A blank screen appears. To abort at this time, hold down [ON], press [F3], and then release both keys. > Hey All > I have flashed my HP49G to HP39/40 and now i canĒt > flash it back, kan anyone help me, > Joakim > http://www.epita.fr/~avenar_j/hp/49.html > 2. On the 49G: > Enter diagnostic mode by turning the calculator on, > holding [ON], and pressing [F4]. > While the Tests screen is showing, > use your thumb to hold down [+] and [ENTER]. > While holding down [+] and [ENTER], > press and release [ON] to show the No System menu, > and then release [+] and [ENTER]. > Press [4] to select terminal mode. A blank screen appears. > To abort at this time, hold down [ON], press [F3], > and then release both keys. I doubt this will work. These are how to upgrade a HP49 On the HP49 you can switch to the No System Menu easily, because the main ROM (not the BIOS) also checks if you press [+] and [ENTER] at the same time. It's not the case on the HP40/39. You must boot the calculator on the boot sector , that is, must perform a Cold Start You must use the Reset button . Using a paper-clip, press the reset button on the back of the calculator. It should turn off. Hold [ENTER] and [+] then press ON . You should get to the No System Menu How to use: TIMERCTRL.1 , TIMERCTRL.2 , TIMER1 , TIMER2 , SHUTDN , DeepSleep and liteslp Alejandro Henao Loaiza [ahl] Have you checked www.hpcalc.org ML related documentss? Raymond Anybody have a 15C manual you'd like to sell? I just picked up a 15C and I'd like get the manual. Anybody have one you'd like to sell? I've the the HPMuseum CDs, so I have an ecopy of the manual. John they are available on Ebay Bob [Author unknown] 1/12 Love Bytes (or sex and the single computer) They say that computers do not think like human beings. But do they _feel_ like humans? A certain computer, to which (to whom?) we have guaranteed absolute anonymity, has supplied us with this piece of computer-style erotic-gossip that is going the rounds of the sophisticated number crunchers in the web. If you have an under-age computer around, on no account let it see this - it might be totally diskusted (!) and even corrupt its little CPU forever! 2/12 Micro was a real-time operator and protocol made it easy for him to inter- face with numerous input/output de- vices, even if it meant time-sharing. 3/12 One evening he arrived home just as the sun was crashing, and had parked his Merced in the main drive (he had missed the serial bus that morning), when he noticed an elegant piece of liveware admiring the daisy wheels in his garden. He thought to himself, She looks user-friendly. I'll see if she'd like an update tonight. 4/12 Mini was her name, and she was delightfully engineered with eyes like Ada and a Prime mainframe archi- tecture that set Micro's peripherals networking all over the place. 5/12 He browsed over to her casually, admiring the power of her twin, 128-bit floating point processors and enquired, How are you Honeywell? Yes, I am well, she responded, batting her optical fibres engagingly and smoothing her console over her curvilinear functions. 6/12 Micro settled for a straight line approximation. I'm stand-alone to- night, he said. How about computing a vector to my base address. I'll output a byte to eat, and maybe we could get offset later on. 7/12 Mini ran a priority process for 2.6 nanoseconds then transmitted 8K, I've been dumped myself recently, and a new page is just what I need to refresh my disks. I'll park my machine cycle in your background and meet you inside. She walked off, leaving Micro admiring her solenoids and thinking, Wow, what a global variable, I wonder if she'll like my firmware. 8/12 They sat down at the process table to a top of form feed of fiche and chips and a bucket of Baudot. Mini was in conversa- tional mode and expanded on ambi- guous arguments while Micro gave occational acknowledgements although, in real time, he was analysing the shortest and least critical path to her entry point. He finally settled on the old would you like to see my benchmark subroutine, but Mini was again one step ahead. 9/12 Suddenly she was up and stripping off her parity bits to reveal the full functio- nality operating system software. Let's Micro was loaded by this stage, but his hardware policing module had a pro- cessor of its own and was in danger of overflowing its output buffer, a hang-up that Micro had consulted his analyst about. Core, was all he could say. 10/12 Micro soon recovered, however, when she went down on the DEC and opened her device files to reveal her data set ready. He accessed his fully packed root device and was just about to start pushing into her CPU stack, when she attempted an escape sequence. 11/12 No, no! she piped. You're not shielded. Reset, Baby, he replied. I've been debugged. But I haven't got my current loop enabled, and I can't support child processes, she protested. Don't run away, he said. I'll generate an interrupt. No, that's too error prone, and I can't abort because of my design philosophy. 12/12 Micro was locked in by this stage though, and could not be turned off. But she soon stopped his trashing by introducing a voltage spike into his main supply, whereupon he fell over with a head crash and went to sleep. Computers, she thought as she compiled herself, all they ever think of is hex. As a LAST ARGument I say: I would have GROBbed her nibbles, ROLLD her over and then start pushing to her RPN stack, but my initial guess is that Micro was too AlgeBrained for that solution. @ > As a LAST ARGument I say: I would > have GROBbed her nibbles, ROLLD her > over and then start pushing to her RPN > stack, but my initial guess is that Micro > was too AlgeBrained for that solution. > @ I guess flashing her messes up her brains, not to mention the risk of leaving her on an unstable, irracional state? :-) (I couldn't resist...) Steve Sousa I already posted this once, but I didn't get an answer. If I store [1 2 3] in 'V1', why I can't evaluate '3*V1'? is there any way to work with vectors besides to do it directly on the stack? Is there any function that takes a vector and runs a program for each component of the vector? Let's say argument 1: vector and argument 2: << -> x 'x+2' >> Marturo UBI DUBIDUM IBI LIBERTAS > I already posted this once, but I didn't get an answer. If I store [1 > 2 3] in 'V1', why I can't evaluate '3*V1'? is there any way to work > with vectors besides to do it directly on the stack? > > Is there any function that takes a vector and runs a program for each > component of the vector? Let's say argument 1: vector and argument 2: > << -> x 'x+2' >> > > > Marturo > > > UBI DUBIDUM IBI LIBERTAS Your question has already been answered in this group, but here we go again. First *clear flag -124*! Then having matrix/vector in 'A' and 'B' and using EVAL to get the result of, for example 'B/2' or 'A*B', will work. But don't use EXPAND for this, EXPAND will error out. Nick. I just tried your example [1 2 3] ENTER 'V1' STO 3 ENTER 'V1' * EVAL and had the right vector [3 6 9] on the stack. I am using ROM 1.19-5 Martin > I already posted this once, but I didn't get an answer. If I store [1 > 2 3] in 'V1', why I can't evaluate '3*V1'? is there any way to work > with vectors besides to do it directly on the stack? > > Is there any function that takes a vector and runs a program for each > component of the vector? Let's say argument 1: vector and argument 2: > << -> x 'x+2' >> > > > Marturo > > > UBI DUBIDUM IBI LIBERTAS > I just tried your example [1 2 3] ENTER 'V1' STO 3 ENTER 'V1' * EVAL and > had the right vector [3 6 9] on the stack. > I am using ROM 1.19-5 > Martin Yes this works, but try to add any operation like '2*V1' or 'V1/2', and you get an error: non algebraic in expression. Of course when you put [1 2 3] on level 2 and 2 on level 1 you can multiply them. Marturo Ubi Dubidum Ibi Libertas in the example that I mentioned, I didn't have the vector itself on stack level 1... I had the name of the variable: V1... I think that you don't get me right... Here I have 2 examples (both requires that you have the [1 2 3] stored in the variable 'V1': Type in: what you see on the stack: 'V1' ENTER 'V1' 2 / V1/2 (pretty print) EVAL [1/2 1 3/2] and this one: 'V1' ENTER 'V1' 6 * V1*6 (pretty print) 'V1' ENTER - V1*6-V1 (pretty print) EVAL [5 10 15] doesn't your calc behave like this...??? Martin > >>I just tried your example [1 2 3] ENTER 'V1' STO 3 ENTER 'V1' * EVAL and >>had the right vector [3 6 9] on the stack. >>I am using ROM 1.19-5 >>Martin > > > Yes this works, but try to add any operation like '2*V1' or 'V1/2', > and you get an error: non algebraic in expression. Of course when > you put [1 2 3] on level 2 and 2 on level 1 you can multiply them. > > Marturo > > Ubi Dubidum Ibi Libertas > in the example that I mentioned, I didn't have the vector itself on > stack level 1... I had the name of the variable: V1... I think that you > don't get me right... > Here I have 2 examples (both requires that you have the [1 2 3] stored > in the variable 'V1': > > Type in: what you see on the stack: > > > 'V1' > ENTER > 'V1' > 2 > / > V1/2 (pretty print) > EVAL > [1/2 1 3/2] > > and this one: > > 'V1' > ENTER > 'V1' > 6 > * > V1*6 (pretty print) > 'V1' > ENTER > - > V1*6-V1 (pretty print) > EVAL > [5 10 15] > > doesn't your calc behave like this...??? > Martin No it doesn't, is there any flag that may affect this behaviour? I have ROM 19.6 installed as Nick already mentioned, be sure to have flag -124 cleared. This can be done by typing in 124 +/- CF in the command line. Martin > >>in the example that I mentioned, I didn't have the vector itself on >>stack level 1... I had the name of the variable: V1... I think that you >>don't get me right... >>Here I have 2 examples (both requires that you have the [1 2 3] stored >>in the variable 'V1': >> >>Type in: what you see on the stack: >> >> >>'V1' >>ENTER >> 'V1' >>2 >>/ >> V1/2 (pretty print) >>EVAL >> [1/2 1 3/2] >> >>and this one: >> >>'V1' >>ENTER >> 'V1' >>6 >>* >> V1*6 (pretty print) >>'V1' >>ENTER >>- >> V1*6-V1 (pretty print) >>EVAL >> [5 10 15] >> >>doesn't your calc behave like this...??? >>Martin > > > > No it doesn't, is there any flag that may affect this behaviour? > > I have ROM 19.6 installed It works for me ROM 1.19-6 ??? > I just tried your example [1 2 3] ENTER 'V1' STO 3 ENTER 'V1' * EVAL and > had the right vector [3 6 9] on the stack. > I am using ROM 1.19-5 > Martin > Yes this works, but try to add any operation like '2*V1' or 'V1/2', > and you get an error: non algebraic in expression. Of course when > you put [1 2 3] on level 2 and 2 on level 1 you can multiply them. > Marturo > Ubi Dubidum Ibi Libertas What are the common strategies for using the HP49g's 3 different Port memories? Here's what I've figured out so far: Port 0 is shared with the Home directory. To allow the max space in Home I wonder why you would ever use this port? Port 1 has the benefit of not being cleared on a full reset, but being ram it can still be lost it power fails (for more than a minute or two). I think I read that it's faster than Port 2, so maybe performance is a reason to use it. Port 2 is flash so that even in a power loss the data is retained, so I see a benefit to using it for data that can't be easily recovered. Steve > Port 0 is shared with the Home directory. To allow the max space in > Home I wonder why you would ever use this port? o It's compatible with old model 48 (allowed use of libraries with no card) o Faster library execution (per R.H.) o Some libraries might *only* work in this port o To deliberately reduce space in HOME, which has a use: > Port 1 o Is more protected than port 0 o But it's also partitioned; you can't store across the 128K boundary (this also applies to Port 2). > Port 2 o Data retained with no battery power. is full, the entire bank must be erased (higher current) and then and this could shorten your battery life if done very often. Fortunately this occurs automatically as needed. Has anyone ever lost some flash stuff if user memory was too low? (Is this possible?) [r->] [OFF] > Has anyone ever lost some flash stuff if user memory was too low? > (Is this possible?) student.tn.tudelft.nl -- This message was written with 100% recycled electrons Pivo One more reason to upgrade a very old ROM to the latest one. > Has anyone ever lost some flash stuff if user memory was too low? > (Is this possible?) student.tn.tudelft.nl > -- > This message was written with 100% recycled electrons > Pivo > One more reason to upgrade a very old ROM to the latest one. True enough, but it seems to me that in order to pack the objects in a flash bank, all the valid objects in that bank would have to be moved *somewhere*. Are they temporarily moved to TEMPOB while the bank is being cleared and then moved back to the same bank? Or maybe they're simply copied to whichever flash banks have enough room before the bank is cleared? It's an interesting problem; I can see where you might get into a situation where it looks like you have enough flash memory for an object, yet not be able to store it there. > > Has anyone ever lost some flash stuff if user memory was too low? > (Is this possible?) > > > > student.tn.tudelft.nl -- James > One more reason to upgrade a very old ROM to the latest one. > > True enough, but it seems to me that in order to pack the objects in a > flash bank, all the valid objects in that bank would have to be moved > *somewhere*. Are they temporarily moved to TEMPOB while the bank is > being cleared and then moved back to the same bank? Or maybe they're > simply copied to whichever flash banks have enough room before the bank > is cleared? It's an interesting problem; I can see where you might get > into a situation where it looks like you have enough flash memory for an > object, yet not be able to store it there. > No, the objects in a flash bank are copied to TEMPOB (only the ones that had not been erased), the bank is erased and formatted, then the objects are moved again to the bank. Werner > the objects in a flash bank are copied to TEMPOB (only the ones > that had not been erased), the bank is erased and formatted, > then the objects are moved again to the bank. 1. Is every bank examined to find one with room? 2. If more than one has room, which one is used? (these decisions matter to the best packing) 3. It's possible that re-distributing objects among banks could make room, even when no isolated bank seems to have enough room by itself, but I'd not expect this famous problem to have been worked out in the 49 :) 4. How similar is this to the problem of designing a flash memory telephone answering machine? (((o))) . > > the objects in a flash bank are copied to TEMPOB (only the ones > that had not been erased), the bank is erased and formatted, > then the objects are moved again to the bank. > > 1. Is every bank examined to find one with room? > > 2. If more than one has room, which one is used? > (these decisions matter to the best packing) > > 3. It's possible that re-distributing objects among banks > could make room, even when no isolated bank seems to > have enough room by itself, but I'd not expect this > famous problem to have been worked out in the 49 :) > > 4. How similar is this to the problem of designing > a flash memory telephone answering machine? > > (((o))) AFAIK, storing an object to the flash goes like this: - find the first bank that can hold the object - if there is no such bank, find the first bank that can hold the object after compressing that bank (=copying al non-deleted obs to TEMPOB, erase the bank, format the bank, copy the obs back) - if there is no such bank - display error message The order in which the banks are examined is (1.19-6:) F,E,D... I have the impression it used to be the other way round in some earlier rom revision. Werner > No, the objects in a flash bank are copied to TEMPOB (only the ones > that had not been erased), the bank is erased and formatted, then the > objects are moved again to the bank. Do you guys want an undelete program for the Flash? I have one lying about. It's very crude and I expect that Thomas' version will be more sophisticated but if you want it it can be found at http://www.ahd.tudelft.nl/~peter/getbank.zip It recalls the contents of a user bank (real/zint on the stack) It obviously only works for banks 8 through 15 and you need 128 kB of free mem if the real on the stack is positive it extracts the object from the bank and when it is negative it gets the purged objects the objects will be in a list on the stack (could be the other way around , I don't know which version this is :-) -- This message was written with 100% recycled electrons Pivo > True enough, but it seems to me that in order to pack the objects in a > flash bank, all the valid objects in that bank would have to be moved > *somewhere*. Are they temporarily moved to TEMPOB while the bank is > being cleared and then moved back to the same bank? Or maybe they're > simply copied to whichever flash banks have enough room before the bank > is cleared? It's an interesting problem; I can see where you might get > into a situation where it looks like you have enough flash memory for an > object, yet not be able to store it there. The amount of memory actually used in the bank is calculated, and reserved how much you need. if the memory is present, the objects are copied from new object is copied. If there is an error while erasing or programming the flash, the list of the object that where in flash is pushed on the stack with a critical error message.. > If there is an error while erasing or programming the flash, > the list of the objects that were in flash is pushed on the stack > with a critical error message... What if there is then not enough free memory to explode that list and store the individual objects back, after a battery change -- can that happen? If so, probably there is some way to deal with it, but which might need a small program to be mistake free? This contingency might be one reason to store user memory backups both in port 2 and port 1; if batteries get low and some backup vanishes before it can be stored back into port 2, then have another in port 1 for good luck! Don't count your chickens if they are all in one basket :) (((o))) . > What are the common strategies for using the HP49g's 3 different Port > memories? Here's what I've figured out so far: > > Port 0 is shared with the Home directory. To allow the max space in > Home I wonder why you would ever use this port? > > Port 1 has the benefit of not being cleared on a full reset, but being > ram it can still be lost it power fails (for more than a minute or > two). I think I read that it's faster than Port 2, so maybe > performance is a reason to use it. > > Port 2 is flash so that even in a power loss the data is retained, so > I see a benefit to using it for data that can't be easily recovered. > > > Steve Execution is fastest from port 0, because every romptr executed from either port 1 or port 2 will be copied to TEMPOB and executed from there. Werner Some programs don't work from bank-switched ports For example my FinLib49 must be installed in port 0. The size is about 15KB and people can live with that. When my Flash fills up I have to use the Port1. I use it also for scratch things, which a TTRM will certainly wipe out of the HOME, but usually not from port 1. I keep all my libs (if possible) in port 2. Also any useful stuff goes there and I run them via a CST menu using :2: tag and EVAL It's also good idea to ARCHIEVE your system to Flash (remember to save flags and keys first to a variable) Anything else? Was it this? You may have other habits. > What are the common strategies for using the HP49g's 3 different Port > memories? Here's what I've figured out so far: > Port 0 is shared with the Home directory. To allow the max space in > Home I wonder why you would ever use this port? > Port 1 has the benefit of not being cleared on a full reset, but being > ram it can still be lost it power fails (for more than a minute or > two). I think I read that it's faster than Port 2, so maybe > performance is a reason to use it. > Port 2 is flash so that even in a power loss the data is retained, so > I see a benefit to using it for data that can't be easily recovered. > Steve > It's also good idea to ARCHIEVE your system to Flash > (remember to save flags and keys first to a variable) > Is there something I should know about flags and keys with regard to ARCHIVE? Steve I'm looking through my ebooks from Urroz and I don't see this. Further I don't see it in my regular manuals. Let's make this very simple. I'm plotting an equation, say x^2. I'd then like to find the area under the curve, to the x axis, from 0 to 2. The only way I can find is to use the cursor to position at almost zero, click area, then position at at about 2 and click area again. Unfortunately I'm not sure if this is really the area or not on the interval 0 to 2 or if it's .01 to 1.9 etc.... Is there someway on the plot I can get exact upper and lower bounds? On my TI-89 it asks me for upper and lower bounds and I can supply a number. ce > I'm looking through my ebooks from Urroz and I don't see this. Further I > don't see it in my regular manuals. Let's make this very simple. I'm > plotting an equation, say x^2. I'd then like to find the area under the > curve, to the x axis, from 0 to 2. The only way I can find is to use the > cursor to position at almost zero, click area, then position at at about 2 > and click area again. Unfortunately I'm not sure if this is really the area > or not on the interval 0 to 2 or if it's .01 to 1.9 etc.... Is there someway > on the plot I can get exact upper and lower bounds? > > On my TI-89 it asks me for upper and lower bounds and I can supply a number. > ce While in the plot, you can press the key [+] to see the x-y-coordinates at the bottom of the screen. But even this is only an approximation. For numeric values of integrals with exactly the endpoints that you want, you should use the integral function. Nick Nick said, While in the plot, you can press the key [+] to see the x-y-coordinates at the bottom of the screen. But even this is only an approximation. For numeric values of integrals with exactly the endpoints that you want, you should use the integral function. had several with this calculator. It doesn't really 'clean up' problems like the TI does. It isn't as quick and responsive on the keyboard either, sigh, and it's slow. I've had instances where I've set up an integral and while I was waiting for the answer to be calculated I've picked up my TI put in the integral and got the answer before my HP returned the answer! I know some of it is my ignorance on using it. However as in this case and the simple way to substitute on the TI 89 I have to say I prefer my 89. Unfortunately for this class it's not allowed. I can't really see why as if you can't set up the problem it doesn't matter how nicely it's integrated it won't be right. Ah well I do appreciate your answer and I will keep learning this calc (at least until December 15th LOL). Carmen > Nick said, While in the plot, you can press the key [+] to see the > x-y-coordinates at the bottom of the screen. But even this is only an > approximation. For numeric values of integrals with exactly the > endpoints that you want, you should use the integral function. > now > had several with this calculator. It doesn't really 'clean up' problems > like the TI does. It isn't as quick and responsive on the keyboard either, > sigh, and it's slow. I've had instances where I've set up an integral and > while I was waiting for the answer to be calculated I've picked up my TI put > in the integral and got the answer before my HP returned the answer! I know > some of it is my ignorance on using it. However as in this case and the > simple way to substitute on the TI 89 I have to say I prefer my 89. > Unfortunately for this class it's not allowed. I can't really see why as if > you can't set up the problem it doesn't matter how nicely it's integrated it > won't be right. Ah well I do appreciate your answer and I will keep > learning this calc (at least until December 15th LOL). > Carmen Hmmm, the HP49G is *not* a simple push-button-get-answer-calc. It is a complex system whose depths can't be mastered easily, but which allows to tailor it to the specific needs of the user. You could (have to?) program it, to respond like you want it to respond, when you press a special key while you see the plot of a function. It is possible to do that already with UserRPL. The point here is, what do you prefere? If you have a system that offers a fixed set of capabilities, then you depend on what the makers thought you will need. If you prefer a system that doesn't offer what somebody has thought you will need, but a rich set of tools (commands etc) instead, then you are not so dependend. You are free(er). For example I have made some programs (that will be posted with the second part of the sequences marathon) that find generating functions of recurrences and transform recurrences to their analytic closed forms. Now, if we prefer a somebody-has-thought-of-all-what-you-need calc, then in my case the calc should have some functions built-in, that turn recurrences to analytic closed forms and find generating functions. OK, let's suppose the Professor puts such functions in the CAS. (He did partially, SIGMA). Now, Carmen wants to integrate from the plotting window, and Bhuvanesh wants to solve partial diffeqs, and everybody has specific needs. Whose needs are more important? Nobody can say. Any solution to this problem that goes the way add more functions for me is a very unsatisfying solution. It is better to give a more or less complete set of tools, of elementary things with which other more complex things can be composed. (Of course, then we must answer the question: What is elementary?) I don't want to start yet another which is better-competition. But think of the following: How long do you think that the TI89 will be able to give you the answers that you want? Don't you think that the built-in functions will not suffice some day? Will you always do only differentiating, drawing tangential lines and calculating integrals between two points? Is that all maths you will ever need? If it is not, what happens in 1 year? Or in 2? Will you be able to use the built-in functions combined to create new, non built-in things? About allowing the HP49G in class. The most probable thing is that the average student will not use the real powers of the HP49G, because he/she simply doesn't have the time to learn even more than the school demands. (We can't eb always learning, we must also just live sometimes ;-)) The average student will expect that TI or any other company has exactly thought of his/her specific needs and just expect that there is a key labeled with his/her problem. So it is probable that the HP49G will not be used in its real *immense* power. If the average student had the will/time to think about programming what he/she needs in class (and about the structure of the underlying CAS), then we would have a very different situation concerning what is allowed and what not. The (implied) rule It is not allowed in class, so it must be better is not only not logical, but also too dangerous in the sence that it influences students to expect that the solutions to their problems all lye in front of them and that they only need to buy them. The HP49G schools us in a different direction. Thinking about how this and that could be (ab)used to solve some particular problem, we learn to think about mathematics by the by. We learn to find relations between unrelated parts of the CAS, we learn to seek solutions in non-traditional ways. Or would somebody expect right from the start that LINSOLVE of the HP49G can be used in the process of finding a generating function? Ending this rather lengthy reply, I must admit that I for myself also wish that I had some of the comforts of the TI calculators and also their speedy hardware. But for the theoretical, hardware-and-inteface-non-dependend part, I wouldn't replace my HP49G with anything else, except perhaps with casiopeia with maple, or with mini-mathematica. (Bhuvanesh, any plans at Wolfram-Research?) Nick. > Hmmm, the HP49G is *not* a simple push-button-get-answer-calc. ...... snip Neither is the TI-89. You can program the TI-89 as well. Further it's documented by the manufacturer so you have a chance of actually adopting it. Putting out this type of complicated machinery and then requiring people to hunt down obscure documentation (how readily available is Urroz's material) and you can see there are great obstacles to adoption of this hardware. HP admitted it by cancelling production. This was the most sophisticated of it's machine's but it did the poorest. It was also the poorest documented, coincidence, possibly but I'd be willing to bet it wasn't. You can't hand this out, say guess how to use it and expect people to adopt it, when there is a cheaper, more accessible, better hardware product out there. Will this calculator do great things? I don't know but the fact I have to pay 60 bucks for some documentation to figure out if it will is angering. Further when you can get exact integrals on the plot, symbolic integration by simply omitting the bounds and other simple accessible functions, on the 89, it's hard to adopt a deliberately obfuscated piece of machinery. Will it do these things? Maybe, but you have to program them. I'm a software engineer by profession, to design a program for something you have to have a better understanding than the rank beginner (assuming you actually want your program to do what you say it will do). Anyway to have to understand these functions to the level of being able to design programs for them sort of puts the cart before the horse doesn't it? You are taking the class to understand these algorithms but you have to program a calculator to do them, that requires a complete understanding of the algorithms you are attempting to learn. You can't really argue that HP isn't targetting the academic market because that's why they removed the beloved equation library from it isn't it, or is it? That seems to be another problem with this calc they didn't define the target market. Anyway I believe in HP calcs. I'm an older person and have owned them for years and years. I purchased this calc based upon my experience with my 48GX, and all my older ones. Suffice to say the last TI I owned was the SR 56 before the 89 came out. I didn't adopt it lightly but only when my 48GX wouldn't do the job quickly enough. Anyway I will learn this calc and use it. But I won't pretend it's something it isn't. ce > > Hmmm, the HP49G is *not* a simple push-button-get-answer-calc. ...... snip > > Neither is the TI-89. You can program the TI-89 as well. Because my understanding of the TI89 is quite limited and only based on what I read on this newsgroup and in its (very nice) manual, I have one question: Can the TI89 be programmed to find for example the already mentioned (and also snipped ;-)) generating functions of sequences? Or anything else that lies totally outside the domain of mathematics which its built-in commands cover? How could that be done? Is the command set of the TI89 capable of doing this? > Further it's > documented by the manufacturer so you have a chance of actually adopting it. Yep, that's completely right. The TI89 *is* documented, while this can't be said for other machines ;-) > Putting out this type of complicated machinery and then requiring people to > hunt down obscure documentation (how readily available is Urroz's material) > and you can see there are great obstacles to adoption of this hardware. HP > admitted it by cancelling production. This was the most sophisticated of > it's machine's but it did the poorest. Sophisticated? Well, I would rather say complex, but for sophisticated I would expect some things to be considerably better. But it doesn't do poor, if doing poor means what problems it can solve. Quite the contrary, it does very good, but only if one had 1000000 years time to play with it and learn its tricky behaviors. > It was also the poorest documented, > coincidence, possibly but I'd be willing to bet it wasn't. You can't hand > this out, say guess how to use it and expect people to adopt it, when there > is a cheaper, more accessible, better hardware product out there. So it is. Not many people will buy a product only to find out, that its usage has to be learned by gathering information bit by bit from newsgroups. > Will this calculator do great things? I don't know but the fact I have to > pay 60 bucks for some documentation to figure out if it will is angering. Oh yes, it is very much so! That's why I make the marathons ;-) > Further when you can get exact integrals on the plot, symbolic integration > by simply omitting the bounds and other simple accessible functions, on the > 89, it's hard to adopt a deliberately obfuscated piece of machinery. Will > it do these things? Maybe, but you have to program them. This type of interactive operations are limited on the HP49G compared to the TI89. The HP49G puts more weight on the algebraic manipulation. Find some special points of a function, solutions, extrema, whatsoever, symbolically or numerically, and then use them as integration limits in the functions provided for integration. > I'm a software engineer by profession, to design a program for something you > have to have a better understanding than the rank beginner (assuming you > actually want your program to do what you say it will do). Not only this. For mathematics software you also must have good understanding of the underlying mathematics. This seems to frustrate many students because they must master maths stuff that they didn't yet take (and perhaps they'll never do, because they don't need that directly). > Anyway to have > to understand these functions to the level of being able to design programs > for them sort of puts the cart before the horse doesn't it? front of the question: How much math knowledge is required for programming the HP49G? Even further, how much math should one learn? The answer is: Learn as much as you can. It helps and not only for programming the HP49G or any other machine. But most of the world nowadays preferes to gather knowledge about the color of the new shoes of some celebrity and other useless stuff, instead of focusing on the queen of all sciences. Have anybody wondered why A beautiful mind was such a success? Why has almost nobody knew about nash prior to the movie? ;-) > You are taking > the class to understand these algorithms but you have to program a > calculator to do them, that requires a complete understanding of the > algorithms you are attempting to learn. Oh no! If you carefully track your path through the jungle of mathematics you can achieve a situation where (more or less) the new stuff bases on the already learned stuff. > You can't really argue that HP > isn't targetting the academic market because that's why they removed the > beloved equation library from it isn't it, or is it? Did I say that? > That seems to be > another problem with this calc they didn't define the target market. Yes, something like that. To me it seems that HP just thought that the whole market will (must?) buy the HP49G. This was a failure if we consider how big and this market is and how big the differences between its sub groups are. > Anyway I believe in HP calcs. I'm an older person and have owned them for > years and years. I purchased this calc based upon my experience with my > 48GX, and all my older ones. Suffice to say the last TI I owned was the SR > 56 before the 89 came out. I didn't adopt it lightly but only when my 48GX > wouldn't do the job quickly enough. > > Anyway I will learn this calc and use it. But I won't pretend it's something > it isn't. What is it? And what is it not? I think it is a quite complex and quirky but also very flexible machine, that allows to program it to do what you want it to do. Sometimes I get the impression that the HP-users, especially the older ones (like me, that is ;-)) expect the time to stop somewhere at the HP41 or the HP48. But this will never work. A calculator that is mainly for carrying some operations with numbers and perhaps a bit of symbolic manipulations is something completely different to a CAS-calculator. The known simplicity *will* vanish when we move from number crunchers to CAS-machines. Anyway, as said before, I also find many things on the HP49G quite unexpected, but for the time being we have nothing else than finding such things and trying to work around them using real existing means. Anyway, still dreaming of the HP58. Nick. ......snip > Because my understanding of the TI89 is quite limited and only based > on what I read on this newsgroup and in its (very nice) manual, I have > one question: Can the TI89 be programmed to find for example the > already mentioned (and also snipped ;-)) generating functions of > sequences? Or anything else that lies totally outside the domain of > mathematics which its built-in commands cover? How could that be done? > Is the command set of the TI89 capable of doing this? snipped..... Nick the TI-89 supports use of assembly language, initially an addon but eventually ti sponsored, There is also a C compiler, third party, so it's fairly extendable, just different.... why am I using the 49 if I'm such a fan of the 89? I'm not a fan of either. These are both tools. I only code when paid and I have other hobbies, mostly ancient textile arts so I'd rather spend my spare time there than programming around arcane and poorly supported hardware. The instructor of the class I'm taking won't allow use of the 89 but will allow use of the 49. That certainly says something. But whatever that statement is is open for interpretation. If you are interested in programming for the 89 here's a good site to gain information from http://www.technoplaza.net/tigcc-asm/. A great example of the inconsistency of the 49 is putting 4x+3 on the stack then click on solvex This gets you the correct answer -3/4. Open the numeric solver put in the expression 4x+3 and you get -.750000000001! It really seems the answer to such a basic expression should be consistently correct..... anyway... As for me I just want the tools to work with a reasonable investment of time for the level of sophistication of the tool. So far it seems to me the 49 requires to great a learning curve and to many unnecessary hurdles to perform some items that should seemingly be included. But that's just my opinion I am sure everyone elses will vary. However I will use it and appreciate the help I get here. If it weren't for this newsgroup this calculator would be virtually useless. So you guys are an incredible resource and very much appreciated. Carmen > ......snip > Because my understanding of the TI89 is quite limited and only based > on what I read on this newsgroup and in its (very nice) manual, I have > one question: Can the TI89 be programmed to find for example the > already mentioned (and also snipped ;-)) generating functions of > sequences? Or anything else that lies totally outside the domain of > mathematics which its built-in commands cover? How could that be done? > Is the command set of the TI89 capable of doing this? > > snipped..... > > Nick the TI-89 supports use of assembly language, initially an addon but > eventually ti sponsored, There is also a C compiler, third party, so it's > fairly extendable, just different.... I see. It is only that ( I think ) most people out there who buy a programmable calculator expect some more or less easy understandable programming language, one that resembles what they do when they do calculations by pressing keys. I can't imagine many people buying a programmable calculator that can be programmed for doing more complicated things but with assemply or C. It is that the underlying math gets hidden for most people then. (Note: I don't say that math disappears, only gets hidden.) Example: Let's mess up with generating functions. (They impressed me quite strongly, I guess ;-)) We want to make a program to find such a thing. Now, if the calculator has a rich set of commands that allows us to somehow accomplish the job by using our more or less known commands like LIMIT, SOLVE or whatsoever, then two things are more focused and appear transparent: The underlying mathematics and the clarity of the program. It is easier to think in terms of find the characteristic polynomial, solve it, make a general linear combination of the solutions, find the linear combination that satisfies the inital conditions,... and so on and at the same time having commands that correspond more or less directly to our involved steps. Using assempler or C makes things a bit more difficult, not because the mathematics disappear, but because we go deeper and deeper in the inner world of bits and bytes, which of course are just another way to express what the mathematics that we learn in school express, but harder to decode for our brains. When we take this to the endth degree, then we end binary code, which (for me) is more difficult to understand, than a couple of commands like LIMIT or SOLVE. If I remember well, Bhuvanesh has said that for some reason the built-in commands of the TI89/92 can't be used when someone programs that in C. Or am I mistake here? Now, if most people would program the calc with its own programming language (the one that contains the set of built-in commands and functions) rather than with C or assempler, one question rises: Is it OK to (indirectly) limit this programming possibility by restricting the set of built-in commands to a less flexible level? Shouldn't a high degree of flexibility be offered to the normal-user who doesn't care that much about bits and bytes? > why am I using the 49 if I'm such a > fan of the 89? I'm not a fan of either. These are both tools. Yes, they are. That's why they should be flexible and their flexibility should be easily available for anyone that uses them. > I only code > when paid and I have other hobbies, mostly ancient textile arts so I'd > rather spend my spare time there than programming around arcane and poorly > supported hardware. The instructor of the class I'm taking won't allow use > of the 89 but will allow use of the 49. That certainly says something. It says that the HP49G is allowed while the TI89 isn't. That's all. No more, no less. If wee trusted the opinions of class instructors back in the eighties or even in the early nineties, we would not be allowed to use anything today. Opinions anybody can have. But knowledge is sparse. Had the (majority of)instructors undestood what math is about, they would not have any problem with the 89 or the 49. The rest of the sentence you are allowed to use this but not that is only interpretation and speculation. > But > whatever that statement is is open for interpretation. I should have read further, me stupid person with strictly serial interfaces to the world ;-)) > If you are interested > in programming for the 89 here's a good site to gain information from > http://www.technoplaza.net/tigcc-asm/. Ahh, thank you very much. I'll try to understand the contents of the site, though assempler always causes an allergic reaction in my soul...;-) > A great example of the inconsistency of the 49 is putting 4x+3 on the stack > then click on solvex This gets you the correct answer -3/4. Open the > numeric solver put in the expression 4x+3 and you get -.750000000001! It > really seems the answer to such a basic expression should be consistently > correct..... anyway... It is not inconsistent at all. SOLVEX is (mainly) an exact analytic solver, while the other is just what you said, a *numeric* solver. The mathematical methods involved are completely different. The first solves analytically (if possible) and doesn't care about particular values of results. You could also solve a*x+b and get the result -b/a with SOLVEX, which is the analytic closed form of the solution no matter what a and b are. But the numeric solver solves only for one particular case of a and b (your case being b=3 and a=4) and returns the solution that holds only for this case. > As for me I just want the tools to work with a reasonable investment of time > for the level of sophistication of the tool. So far it seems to me the 49 > requires to great a learning curve It indeed does. But so does mathematics. ;-) > and to many unnecessary hurdles to > perform some items that should seemingly be included. Seemingly included, seemingly for you, Carmen. Your wish for included is important, but so are the wishes of anybody else. Add all these wishes and make them includes, then you end up with presumably thousends of functions. It is not elegant and also not sufficient this way. It is better to analyze the wishes and search for a set of commands that allows most wishes to come true. The question then is where the optimum between flexibility and amount of typing is, but it is really not at the extremal point of adding each and every wish as a built-in command. > But that's just my > opinion I am sure everyone elses will vary. However I will use it and > appreciate the help I get here. If it weren't for this newsgroup this > calculator would be virtually useless. So you guys are an incredible > resource and very much appreciated. Yes, this group is a real treasure. I also would still be looking for answers if the people here weren't that helpful. Not to mention how many HP49Gs would have been executed if the answers and the help found here were not available. Nick. > However as in this case and the simple way to substitute on the TI 89 I have > to say I prefer my 89. Unfortunately for this class it's not allowed. That is strange... the TI-89 is not allowed but the HP49G is? -- Bhuvanesh > That is strange... the TI-89 is not allowed but the HP49G is? Simply amazing eh? :) I was wondering, could someone explain to me how to make a library (with Jazz) that uses absolute addresses (like Erable beta on hpcalc.org). Caspar Caspar Lugtmeier & Eva Skotarczak schrieb im > I was wondering, could someone explain to me how to make a library (with > Jazz) that uses absolute addresses (like Erable beta on hpcalc.org). You could take one (or more;-) look(s) at the sources of the <-RPL-> package by Detlef Mueller and me. http://www.hpcalc.org/hp48/programming/asm/rpl48_30.zip It's one of the first software packs for the HP-48 which used absolute addressing. Only the 41 emulator card and the OS itself were earlier in this aspect;-) Raymond > > Caspar Lugtmeier & Eva Skotarczak schrieb im > > I was wondering, could someone explain to me how to make a library (with > Jazz) that uses absolute addresses (like Erable beta on hpcalc.org). > > You could take one (or more;-) look(s) at the sources of the <-RPL-> package > by Detlef Mueller and me. > > http://www.hpcalc.org/hp48/programming/asm/rpl48_30.zip > > It's one of the first software packs for the HP-48 which used absolute > addressing. > Only the 41 emulator card and the OS itself were earlier in this aspect;-) > > > Raymond Jazz doesn't support ABS xyztu. I have not been able to create an absolute lib using Jazz. I have been able to do it with MASD (on the 48/49) Werner First, is there anything worth-while in this manual beyond the commands reference that is on hpcalc.org (hp48gaur.zip)? If so, where can you buy the manual? I found it on hp.com, but their price is $70. Steve Follow this Link and you will see, that there is an offer for the AUR for 33 Euros ! http://www.stolte-edv.com/Laden/shop.cgi HTH Bernd > First, is there anything worth-while in this manual beyond the > commands reference that is on hpcalc.org (hp48gaur.zip)? If so, where > can you buy the manual? I found it on hp.com, but their price is $70. > Steve > > Follow this Link and you will see, that there is an offer for the AUR for 33 > Euros ! > > http://www.stolte-edv.com/Laden/shop.cgi Or www.calcpro.com lists the 48G AUR for US$29.95. -- James I paid more...that's it. For the 48G series, it's good - for the 49G: Buy Urroz book for the 49G Vol.I > First, is there anything worth-while in this manual beyond the > commands reference that is on hpcalc.org (hp48gaur.zip)? If so, where > can you buy the manual? I found it on hp.com, but their price is $70. > Steve how can I receive a screen capture made with [On]+[Up] from the cat < /dev/ttyS1, but I wasn't successful. Christoph -- I am a HP calculators user, used to be electrical engineer. I had HP-11C, 15C some 13 yrs ago, and used to RPN logic. Recently I aquired HP 49G for my daughter. Her complaint is that the keystroke response is too slow as compared to some of her friend's TI calculators ( BTW, the teacher in her class seems to like TI) . I've checked some folk who recommended typing something like KEYTIME command with some numeer that corresponds to wait time for a key before it responds to the stroke. Does any one have any idea how to improve the keystroke response on HP 49G? Terry Mission Viejo,CA As you use a shorter time for Keytime you increase the possiblity that a short keystroke will be interpreted as a double keypress. I think the default is 1500 and Ive reduced my HP49 to 300 using the 300 /->KEYTIME command. This seems to feel a lot more responsive but is still slower than simpler HP's like the HP15 and HP41 or cheap 4 function calculaters. Keep in mind KEYTIME is only the delay in displaying keypresses so even with long KEYTIME delays the HP49 is still registering your keystrokes and will catch up when you pause, it only feels slow. The only exception I've found to this is when you enter a list. There appears to be a pause in the OS after entering the opening {, I often lose the first digit of the list entry. KEYTIME appears to have no effect on this. Very annoying and hope it's fixed one day. > I am a HP calculators user, used to be electrical engineer. I had > HP-11C, 15C some 13 yrs ago, and used to RPN logic. Recently I aquired > HP 49G for my daughter. Her complaint is that the keystroke response > is too slow as compared to some of her friend's TI calculators ( BTW, > the teacher in her class seems to like TI) . I've checked some folk > who recommended typing something like KEYTIME command with some numeer > that corresponds to wait time for a key before it responds to the > stroke. Does any one have any idea how to improve the keystroke > response on HP 49G? > Terry > Mission Viejo,CA Not sure if anyone noticed the price discounts HP placed on pretty much ALL their calculators on their web site, but I'm wondering if thats any indication that something is in the works. With HP making an announcement this weekend at the HPCC conference in London, it kind of makes you wonder what they have (if anything) up their sleeve. Doug They have something on the works, but I think that they are some Chinese cheapos. According to different sources there should at least one, perhaps even two new calculators, probably scientific, very likely Algebraic only, released before Xmas. What I want is a turbo-charged Saturn/Yorke in a modern process of 0.15 microns instead of 2 micros and a voltage of 1,8V down from 5V and a speed of 33.3333MHz (not 4MHz) That would make a HP 50G/58GX run rings around anything. AND That I'm not going to get !! >:-( Unless...some one of you at the HPCC *demand* this. It's time for HP to realize that in order to sell they need to be the best AND have a top-model !! (with a proper keyboard !!!) > Not sure if anyone noticed the price discounts HP placed on pretty > much ALL their calculators on their web site, but I'm wondering if > thats any indication that something is in the works. With HP making > an announcement this weekend at the HPCC conference in London, it kind > of makes you wonder what they have (if anything) up their sleeve. > Doug Anything at all, no matter how small, would be nice in that it would be a step towards the sort of re-committment to the market that we've all been hoping to see. > > They have something on the works, > but I think that they are some Chinese cheapos. > According to different sources there should at least one, > perhaps even two new calculators, probably scientific, > very likely Algebraic only, released before Xmas. > > What I want is a turbo-charged Saturn/Yorke in a modern > process of 0.15 microns instead of 2 micros and a voltage > of 1,8V down from 5V and a speed of 33.3333MHz (not 4MHz) > That would make a HP 50G/58GX run rings around anything. > AND > That I'm not going to get !! > >:-( > Unless...some one of you at the HPCC *demand* this. > It's time for HP to realize that in order to sell > they need to be the best AND have a top-model !! > (with a proper keyboard !!!) > >>Not sure if anyone noticed the price discounts HP placed on pretty >>much ALL their calculators on their web site, but I'm wondering if >>thats any indication that something is in the works. With HP making >>an announcement this weekend at the HPCC conference in London, it kind >>of makes you wonder what they have (if anything) up their sleeve. >> >>Doug >> > > Afraid to say, I'll believe it when I see it. Remember, a good number of us didn't believe the HP-49G was real for a while but thought it was an April Fool's stunt. It just doesn't make any sense at all to kill off ACO if you're going to jump back into calculators. I don't believe it. Finance and marketing people, often the brunt of many jokes here (and in finance guy. Gene Finance guy AND HP calculator collector! -- * All opinions herein expressed are mine and * * mine alone. You may choose to ignore them * * but I own them. * > Anything at all, no matter how small, would be nice in that it would be > a step towards the sort of re-committment to the market that we've all > been hoping to see. > > They have something on the works, > but I think that they are some Chinese cheapos. > According to different sources there should at least one, > perhaps even two new calculators, probably scientific, > very likely Algebraic only, released before Xmas. > > What I want is a turbo-charged Saturn/Yorke in a modern > process of 0.15 microns instead of 2 micros and a voltage > of 1,8V down from 5V and a speed of 33.3333MHz (not 4MHz) > That would make a HP 50G/58GX run rings around anything. > AND > That I'm not going to get !! > >:-( > Unless...some one of you at the HPCC *demand* this. > It's time for HP to realize that in order to sell > they need to be the best AND have a top-model !! > (with a proper keyboard !!!) > >>Not sure if anyone noticed the price discounts HP placed on pretty >>much ALL their calculators on their web site, but I'm wondering if >>thats any indication that something is in the works. With HP making >>an announcement this weekend at the HPCC conference in London, it kind >>of makes you wonder what they have (if anything) up their sleeve. >> >>Doug >> > > >What I want is a turbo-charged Saturn/Yorke in a modern >process of 0.15 microns instead of 2 micros and a voltage >of 1,8V down from 5V and a speed of 33.3333MHz (not 4MHz) >That would make a HP 50G/58GX run rings around anything. >AND >That I'm not going to get !! And don't forget the eight pound battery to get it to run for more than ten minutes at a time ... -- john R. Latala jrlatala@golden.net The program works fine except that the maximum speed is 5 revolution to second(that mean ~39Km/h). I was wondering if someone could help me to transfer this loop to SysEVAL commands in order to speed-up the program(speeding it in 2 factor should be enough). 0 STIME TICKS DO WHILE KEY NOT REPEAT 1 SRECV DROP2 TICKS DUP ROT - B->R 8192 / .00218 SWAP / 3600 * 2 RND 2 FIX SWAP + Km/h. + 3 DISP END UNTIL 13 == END Tal > The program works fine except that the maximum > speed is 5 revolution to second(that mean ~39Km/h). > I was wondering if someone could help me to transfer > this loop to SysEVAL commands in order to speed-up > the program(speeding it in 2 factor should be enough). > 0 STIME > TICKS > DO > WHILE > KEY NOT > REPEAT > 1 SRECV > DROP2 > TICKS DUP ROT - B->R 8192 / .00218 SWAP / 3600 * 2 RND 2 FIX > SWAP + Km/h. + > 3 DISP > END > UNTIL > 13 == > END Even without using SYSEVALs or SysRPL, you could speed it up by taking 2 FIX out of the loop, and since you're only going to add the number to a string, I don't see any need to also use 2 RND. Also, you can save some steps by simplifying the arithmetic. Of course, a 12-significant digit multiplier seems a little silly under the circumstances; round it off as you see fit. If you're willing to let any keypress (except CANCEL) break out of the loop instead of requiring [C], then you can eliminate the DO... loop. You could eliminate the NOT by using a DO... loop instead of a WHILE... loop. To save on batteries, don't forget a CLOSEIO. If you're using a 49G, use reals instead of ZINTs. Try the following: 0. STIME TICKS 2. FIX DO 1. SRECV DROP2 TICKS DUP ROT - B->R 1.55543211646E-5 * SWAP + Km/h. + 3. DISP UNTIL KEY END CLOSEIO DROP I don't know how fast that will work, but it should help. By the way, for anyone to convert this to SYSEVALS, he'd very likely need to know which calculator you're using. But I'm not at all sure that simply substituting SYSEVAL commands would speed it up; if you're just invoking a UserRPL command with a user binary integer plus SYSEVAL, then it seems to me that it would be slower. I think that what you'd want would be SysRPL. -- James Oops! > 0. STIME > TICKS > 2. FIX > DO > 1. SRECV > DROP2 > TICKS DUP ROT - B->R > 1.55543211646E-5 * > SWAP + Km/h. + > 3. DISP > UNTIL > KEY > END > CLOSEIO > DROP Make that: 0. STIME TICKS 2. FIX DO 1. SRECV DROP2 TICKS DUP ROT - B->R 64290.816 SWAP / SWAP + Km/h. + 3. DISP UNTIL KEY END CLOSEIO DROP -- James PS: Sorry, I see that you did have HP48 in the subject line. I suppose that you mean something in the 48G series. You might want to make sure that there's a CLOSEIO before the loop so that the first time through, SRECV opens the port and therefore flushes the input buffer. Otherwise, you might start out going pretty fast. ;-) Also, 2 FIX gives a lot of jitter at high speed, so -2 RND may be better. It looks as if: 0 STIME CLOSEIO TICKS DO 1 SRECV DROP2 TICKS DUP ROT - B->R 64290.816 SWAP / -2 RND SWAP + Km/h. + 3 DISP UNTIL KEY END CLOSEIO DROP should be good up to about 83 Km/h. If you can settle for about 73 Km/h, then the following will let you break out of the loop (after a timeout) even when stationary: 25.4 STIME @ Maximum to wait before timeout. CLOSEIO @ To make sure buffer gets flushed. TICKS DO ERR0 @ Clear last error. 1 SRECV DROP2 IF ERRN # C02h =/ @ If we didn't have a timeout... THEN TICKS DUP ROT - B->R 64290.816 SWAP / -2 RND @ Round to 2 significant digits. SWAP + Km/h. + 3 DISP END UNTIL KEY @ Was any key (other than CANCEL) @ pressed? END CLOSEIO @ Close port to save power. DROP @ Drop the key number. Play around with different values for STIME; lower values (greater than zero) will let you break out of the loop faster, but may introduce additional timing variation at very low speeds. I expect that any garbage collections may well cause a few false speed readings too. -- James my ON, right shift, left shift and alpha keys malfunction I can not shutdown the HP, the on key works only in the ON-C combinaison the shift and alpha works somtimes, somtimes not my HP is 8 years old, so no garanty anymore and I was told it's very expensive to repair it anyway to fix it myself ? -- Tom Press slightly just below the LCD and above the menu keys. If it works then you have to somehow apply constant pressure to that area. I'm afraid you have to open your calculator in order to that. How? please refer to the www.hpcalc.org for that. > my ON, right shift, left shift and alpha keys malfunction > I can not shutdown the HP, > the on key works only in the ON-C combinaison > the shift and alpha works somtimes, somtimes not > my HP is 8 years old, so no garanty anymore > and I was told it's very expensive to repair it > anyway to fix it myself ? > -- > Tom thank you for your help effectively if I press just below the LCD I can use the keys I'm gonna try to open the calculator and put something there to make pressure -- Tom Veli-Pekka Nousiainen a .8ecrit dans le message de > Press slightly just below the LCD and above the menu keys. > If it works then you have to somehow apply constant pressure > to that area. I'm afraid you have to open your calculator > in order to that. > How? > please refer to the www.hpcalc.org for that. > > my ON, right shift, left shift and alpha keys malfunction > I can not shutdown the HP, > the on key works only in the ON-C combinaison > the shift and alpha works somtimes, somtimes not > my HP is 8 years old, so no garanty anymore > and I was told it's very expensive to repair it > > anyway to fix it myself ? > > -- > > Tom > > > > This morning, at work, I could see that hpcalc has been updated. Now, at home, I only see the famous july 21st... > This morning, at work, I could see that hpcalc has been updated. Now, at > home, I only see the famous july 21st... It depends on which server you happen to get. -- James > It depends on which server you happen to get. Yes, I know but some time ago I changed the mirror pressing F5. Today this seems not work :-( is it possible to embed User RPL code into a HP Basic program? I found out that you can do it the other way, as shown in the following snippet: %%HP: T(3)A(R)F(.); << -> A B << `->STR(A+B)` >> MSGBOX >> Has anyone tried the opposite? Christoph -- Neurotics build castles in the sky, Psychotics live in them, And psychiatrists collect the rent. RPL> > is it possible to embed User RPL code into a HP Basic program? I > found out that you can do it the other way, as shown in the following > snippet: > %%HP: T(3)A(R)F(.); > << -> A B > << `->STR(A+B)` > >> MSGBOX >> > Has anyone tried the opposite? > Christoph > -- > Neurotics build castles in the sky, > Psychotics live in them, > And psychiatrists collect the rent. #51191833 > is it possible to embed User RPL code into a HP Basic program? I > found out that you can do it the other way, as shown in the following > snippet: > > %%HP: T(3)A(R)F(.); > << -> A B > << `->STR(A+B)` > >> MSGBOX >> > > RPL> found out how it worked, but it worked only at the beginning of a program: RPL> << prog... >> I intended to ask the following: How can you embed RPL code anywhere in a Basic program, like this: << -> A B << `MSGBOX(eval_this_as_rpl(<< A B + ->STR >>))` >> >> Christoph -- There was a young lady named Bright, / Whose speed was faster than light. She went out one day, / In a relative way, And returned the previous night. X found > out how it worked, but it worked only at the beginning of a program: > RPL> > << > prog... >> > I intended to ask the following: How can you embed RPL code anywhere in a Basic > program, like this: > << -> A B > << `MSGBOX(eval_this_as_rpl(<< A B + ->STR >>))` >> >> ********************************************* How about using a separate subprogram for that? RPL> << -> A B << A B + ->STR > Main: << -> A B << MSGBOX(subprogram(A,B)) >> >> You need to add the parameter passing to the subprogram OR you have a change to use 'compiled' local variables. They start with left arrow: <-A, <-B and they keep they values until they are purged at the end of a main program. You can even use a list as a program: << -> <-A <-B << MSGBOX(EVAL(sublist)) >> >> sublist: RPL> { <-A <-B + ->STR } Ofcourse there are many more ways to skin the cat with the HP 49G's extensible RPL/HPBasic language. and am in the process of migrating to the HP's from a TI-89. I was wondering if anyone knew of an online tutorial or manual for the 48. I spend a lot of time trying to figure out simple functions, and would like a little explanation for stuff. Things can be somewhat mystical with this calculator. BTW, I like many of the features of the 48, but is there a reason why it is so *SLOW* to do stuff like integrals and derivatives? I have the PC cable on the way so I can try out software from hpcalc.org... are the programs there a large improvement over the out-of-the-box functions? jeff Jose: Download the manual at www.hpcalc.org or at http://www.hpcalc.org/hp48/docs/misc/hp48gug.zip It is about 16MB so good luck if u have 56k ee4fun > and am in the process of migrating to the HP's from a TI-89. I was > wondering if anyone knew of an online tutorial or manual for the 48. I > spend a lot of time trying to figure out simple functions, and would > like a little explanation for stuff. Things can be somewhat mystical > with this calculator. > BTW, I like many of the features of the 48, but is there a reason why > it is so *SLOW* to do stuff like integrals and derivatives? I have the > PC cable on the way so I can try out software from hpcalc.org... are > the programs there a large improvement over the out-of-the-box > functions? > jeff I got several manuals/books as well as plug in cards on Ebay. Byron > and am in the process of migrating to the HP's from a TI-89. I was > wondering if anyone knew of an online tutorial or manual for the 48. I > spend a lot of time trying to figure out simple functions, and would > like a little explanation for stuff. Things can be somewhat mystical > with this calculator. > BTW, I like many of the features of the 48, but is there a reason why > it is so *SLOW* to do stuff like integrals and derivatives? I have the > PC cable on the way so I can try out software from hpcalc.org... are > the programs there a large improvement over the out-of-the-box > functions? > jeff Jeff, If you're in the US, you can try to buy a set of manuals (getting started and user's guide) from www.calcpro.com for $25. Daniel >and am in the process of migrating to the HP's from a TI-89. I was >wondering if anyone knew of an online tutorial or manual for the 48. I >spend a lot of time trying to figure out simple functions, and would >like a little explanation for stuff. Things can be somewhat mystical >with this calculator. >BTW, I like many of the features of the 48, but is there a reason why >it is so *SLOW* to do stuff like integrals and derivatives? I have the >PC cable on the way so I can try out software from hpcalc.org... are >the programs there a large improvement over the out-of-the-box >functions? >jeff 1)I think you can download the user's manual from hpcalc 2)YES. The software can improve the features VERY MUCH. Try Alg48 and Erable if you are interested in Maths, Stat48Pro for Statistics. If you have a Look for software in your field: it will be worth. > 2)YES. The software can improve the features VERY MUCH. Of course. I have just gone back. Port >=2: Alg48 & add-on,UFL,QPI, TED(VV command),Neopolys,Bode-Routh,Bodepack, Solvesys and much more.. Port 1: Erable,Java Port 0: Completly free. Maybe there's a better configuration with the MK and the port 0 free. > Maybe there's a better configuration with the MK and the port 0 free. I have used many tricks for getting as more free memory as possible (converting my programs in libraries, hiding in libraries the MK system programs like STARTED, etc). Before a warmstart I get near 33000 bytes free: enougth for the everyday work. If I need more space (for some transfers, spliting libraries, for instance) I purge Erable from port 0. At the end, I copy the backup from port 9 to port 0 and warmstart. I used Java years ago, and really it's good, but MK is MUCH better: the editors for text, matrix, equations, fonts, are fast and they let you things like cut, copy and paste. Does anybody out there remember at which price levels the HP-15C was sold in Germany in its time (1982-1989)? I know that in U.S. the calculator was introduced at $135, and later was $80. TIA, ------------------------------------------------------------------------- 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. ------------------------------------------------------------------------- I have a HP15C bill dated at 10/28/87 over 258,90 DM (266,90 DM - 8,00 DM rebate) from meisterknecht gmbh, b.9fromaschinen + computersyteme. Christoph Ralf Fritzsch schrieb im Newsbeitrag > Does anybody out there remember at which price levels the HP-15C was sold > in Germany in its time (1982-1989)? I know that in U.S. the calculator > was introduced at $135, and later was $80. > TIA, > ------------------------------------------------------------------------- > 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. > ------------------------------------------------------------------------- web pages that say if you save a user binary integer into a variable named 'TOFF' in the HOME directory, you can adjust the delay before the calculator shuts down. I've done this and the calculator isn't shutting down. For example, I set the TOFF variable to #491520d (one minute) and at 60 seconds, the calculator's hour-glass pops up for a brief second and nothing happens. I'm using ROM 1.19-6. Am I making a mistake somewhere? Doug > web pages that say if you save a user binary integer into a variable > named 'TOFF' in the HOME directory, you can adjust the delay before > the calculator shuts down. I've done this and the calculator isn't > shutting down. For example, I set the TOFF variable to #491520d (one > minute) and at 60 seconds, the calculator's hour-glass pops up for a > brief second and nothing happens. > I'm using ROM 1.19-6. Am I making a mistake somewhere? I use 'TOFF' with ROM 1.19-6, and have only found problems when using ridiculously high values for TOFF; strange things may happen if you use an integer that takes over 32 bits, that is, higher than # FFFFFFFFh. But what you've described sounds like it should work. By any chance do you also have a reserved variable 'STARTOFF'? If you do, then the calculator won't turn off unless the program in 'STARTOFF' contains the OFF command. I suspect that that's where the problem is. -- James Keyman was intendet to be updated for the 48, but this was errouneously done in the HP49 section. Untill this is corrected (which may take weeks :-) use my own site. This should always be done to be sure to get actual versions of my programs. - Wolfgang I understand u can put a text file into 48GX as long as u put a string for the content and header, but I was wondering can I change the font size in order to fit the txt fit in one screen instead I need to use a arrow key move around. I know people are good. I hope some once here can help. tsli > I understand u can put a text file into 48GX as long as u put > a string for the content and header If you mean downloading from a computer, the following gimmick allows even embedded quotes [], and without needing to wrap the text inside quotation marks, either: %%HP: T(3); C$ $ [here put any text at all, any number of lines] The compiler recognizes C$ followed by a decimal character count to make a string, or $ to represent an infinite count (to the end of the actual computer file, that is). > can I change the font size You can download HP48 text viewers that display in a small font (VV or TED or Stringwriter or ???) from www.hpcalc.org (((o))) . Prior submitted message not seen? > Prior submitted message not seen? > Trying again HP48Gx has a RCIJ command which allows one to add to a row a multiple of another row. I need a CRIJ command which allows one to add a multiple of a column to another column. How to program? Is it available anywhere? > I need a CRIJ command which allows one to add a multiple of a column to > another column. > How to program? Is it available anywhere? .82 4 ROLL TRAN 4 ROLLD RCJI TRAN é Use it in the same arguments than RCJI, but columns instead rows. Is this what you need? > .82 4 ROLL TRAN 4 ROLLD RCJI TRAN é Sorry: TRN, not TRAN. I often forget which are standard 48 commands and which added... To do column reductions equivalent to the row reduction command RREF, simply prefix and postfix the RREF command by TRN transpose commands. > What happens when you click reload while holding the shift key? Use http://hpcalc.beachnet.org/ directly, but you should randomly get either one or the other site: glen@glens(2)~$ host www.hpcalc.org www.hpcalc.org has address 216.138.211.243 www.hpcalc.org has address 130.161.43.47 > What happens when you click reload while holding the shift key? I get the same again:from Electronics Laboratory - Neural Networks and Nanoelectronics Group > Use http://hpcalc.beachnet.org/ directly, but you should randomly get > either one or the other site: > glen@glens(2)~$ host www.hpcalc.org > www.hpcalc.org has address 216.138.211.243 > www.hpcalc.org has address 130.161.43.47 And the same again I solved the ODE below with the HP49G (Solve diff eq..) and the calc had more than 10 minutes to solve it. I solved the same ODE on the HP48G and it only had 63s to solve it. Can anyone confirm this behaviour. ODE: dX/dY=(1-Q^2/g*((b+2*m*Y)/((b+m*Y)*Y))^3)/(so-(Q*n/Cu)^2*(b+2*Y*sqrt(1+m^2 )) ^(4/3)/((b+m*Y)*Y)^(10/3) (The ODE is from a hydraulic equation by G. Urroz) information): b=2 [m] m=0 g=9.81 [m/s^2] Q=7 [m^3/s] n=0.013 Cu=1.0 [m^(1/3)/s] so=0.0003 This are the settings: Indep: Y, Init: 2.7, Final: 3.1 Soln: X, Init: 0 Tol: 0.0001, Step: Dflt The solution is: -8127.801.. (which is given by both calcs) Why does the HP49G take so much time to solve this ODE? I thought the algorithms for the numerical Differential equation solver haven't been changed (much) from the 48 to the 49? Roman Hmmm. I don't know... AND I didn't even manage to get that ODE into the solver correctly Error: Bad Argument Type BUT since you can do that - have you tried to set the CAS MODES *both* Numeric and Approx. typing in the values and the formula _after_ that change Did it help in anyway?? Veli-Pekka X> ODE: dX/dY=(1-Q^2/g*((b+2*m*Y)/((b+m*Y)*Y))^3)/(so-(Q*n/Cu)^2*(b+2*Y*sqrt(1+m^2 )) > ^(4/3)/((b+m*Y)*Y)^(10/3) > (The ODE is from a hydraulic equation by G. Urroz) > information): > b=2 [m] > m=0 > g=9.81 [m/s^2] > Q=7 [m^3/s] > n=0.013 > Cu=1.0 [m^(1/3)/s] > so=0.0003 > This are the settings: > Indep: Y, Init: 2.7, Final: 3.1 > Soln: X, Init: 0 > Tol: 0.0001, Step: Dflt > The solution is: -8127.801.. (which is given by both calcs) > Why does the HP49G take so much time to solve this ODE? I thought the > algorithms for the numerical Differential equation solver haven't been > changed (much) from the 48 to the 49? > Roman also evaluated the whole expression on the stack before and pasted the resulting expression in the diff equation solver. It's still a lot slower than the 48. It needs still more than 5 minutes. The 48G needs with the preevaluated expression about 40s. BTW you get the Wrong Argument Type error if you mistake X and Y. But the expression I gave was also not correct. An ')' was missing. I made a copy and paste, but something went wrong obviously. Sorry about that. But in case you should try again don't forget you won't be able to use your HP49G for the next 10minutes;-). PS: Below I give the now correct ODE again (with a few comments). I would really like if someone could explain why the 49 takes so much more time to solve it than the 48. Or maybe someone knows how to speed things a bit up (without changing the tolerance;-)). The ODE: F: '(1-Q^2/g*((b+2*m*Y)/((b+m*Y)*Y)^3))/ (So -(Q*n/Cu)^2*((b+2*Y*sqrt(1+m^2))^(4/3)/ ((b+m*Y)*Y)^(10/3))) (where sqrt has to be replaced with the squar-root symbol of course) In case you don't like to type: %%HP: T(1)A(R)F(.); '(1-Q^2/g*((b+2*m*Y)/ ((b+m*Y)*Y)^3))/(So-( Q*n/Cu)^2*((b+2*Y*f(1 +m^2))^(4/3)/((b+m*Y) *Y)^(10/3)))' Settings: Indep: Y, Init: 2.7; Final: 3.1, Soln: X, Init: 0, Final: [to solve for], Tol: 0.0001, Step: Dflt (Don't mistake Y, and X or you get a Bad Argument Type error-message) Variables: Q: 7, g: 9.81, m: 0, b: 2, n: 0.013, So: 0.0003, Cu: 1 (The variables need to have the values befor you try to solve the ODE in the diff equation solver) The result for the ODE with this settings is (or should be): -8127.801... Peter Geelhoed schrieb im Newsbeitrag > The solution is: -8127.801.. (which is given by both calcs) > Why does the HP49G take so much time to solve this ODE? I thought the > algorithms for the numerical Differential equation solver haven't been > changed (much) from the 48 to the 49? > Perhaps your calc is in exact mode, and the 49 tries to solve it > algebraicly The numeric diff eq solver ([right-shift]+[7],2. Solve diff eq..) gives only numeric answers anyway (Runge-Kutta?, at least this is my understanding of it) so I don't think this should have any influence. But just to be sure I checked it and it doesn't seem to matter if the calc (49 of course) is in exact or approximate mode. It takes int both modes more than 10 minutes to solve the ODE. Roman > -- > This message was written with 100% recycled electrons > Pivo > Peter Geelhoed schrieb im Newsbeitrag > > The solution is: -8127.801.. (which is given by both calcs) > Why does the HP49G take so much time to solve this ODE? I thought the > algorithms for the numerical Differential equation solver haven't been > changed (much) from the 48 to the 49? > > Perhaps your calc is in exact mode, and the 49 tries to solve it > algebraicly > > The numeric diff eq solver ([right-shift]+[7],2. Solve diff eq..) gives > only numeric answers anyway (Runge-Kutta?, at least this is my understanding > of it) so I don't think this should have any influence. But just to be sure > I checked it and it doesn't seem to matter if the calc (49 of course) is in > exact or approximate mode. It takes int both modes more than 10 minutes to > solve the ODE. > > Roman numerically, that is, for example not Y+2*X but rather Y+2.*X. And if it contains some variable name where a number is stored, it could help to store the number with decimal point in that variable. Hope that this speeds things up a bit. Nick. I have a HP95LX with a broken serial port (just the serial port)...but i can comunicate the HP95LX with my HP48GX throught the IR port....but i can't comunicate my PC with my HP95LX (i don't have an IR port on my PC). I am just wondering if there is any software that connects throught the serial port of my PC to the serial port of the HP48GX, and use the HP48GX IR port.... the HP48GX will act like an interface to the 95LX...so i can transfer software form my PC to the HP95LX throught my 48GX. OBS>I can transfer files from the PC to the calculator and then to the HP95LX, but that way, i am limited to 128Kb... :-( Diego My mother is going on a buisness trip to Stockholm, and I treat this as a great chance to add a loveley GX to my 49G :) (I've never seen 48's here, in Russia). Now, the question is -- someone knows a place in Stockholm to buy a 48GX+link kit and a memory card, perhaps? hpcalc.org says BB Marketing is in Sweden, but their site is not accessible for us mortals (not knowing the password). -- TIA, Wartan. Computers are not intelligent. They only think they are. Prastui Wartan, One place is Svanstr.9ams at City Sveav. 17, Box 3236, 103 64 STOCKHOLM 08-24 59 00. ItĒs right in the middle of the centre at the Concert Hall. I phoned them and they have the48G+ at stock (1995 SEK incl VAT minus 375 SEK taxfree IF REQUESTED at purhace , well given back at customs) and will recieve the48GX(2650 - 375tax off at customs) on monday 23rd of sept. The mem cards are out of stock and - I donĒt know the precise term - but kind of out of stock at the supplier as well. They donĒt have the AUR either, seldom sell stuff like that although itĒs quite good when youĒre in to it. I dont assume its the cheapest but might be easiest to find - time might be precious to your mom. BTW 1 USD apprx 9 SEK /Matti > > My mother is going on a buisness trip to Stockholm, and > I treat this as a great chance to add a loveley GX to > my 49G :) (I've never seen 48's here, in Russia). > Now, the question is -- someone knows a place in Stockholm > to buy a 48GX+link kit and a memory card, perhaps? > hpcalc.org says BB Marketing is in Sweden, but their site > is not accessible for us mortals (not knowing the password). > One place is Svanstr?ms at City Sveav. 17, Box 3236, 103 64 STOCKHOLM > 08-24 59 00. It?s right in the middle of the centre at the Concert > Hall. Hope she would have enough time to find it :) -- Computers are not intelligent. They only think they are. I have a 41CX in excellent condition for sale. No corrosion on battery and extension port connectors. Original soft carrying case. But, no manuals. Interested ? Make me an offer ! Pascal Mass.9fn Oostende, Belgium lets say I want to input: 4x + 6 = 7x so that I can solve for x Can it be done? > lets say I want to input: > > 4x + 6 = 7x > > so that I can solve for x > > > Can it be done? Let's not get crazy here (yet). You can obtain a numeric answer by going to the solver menu by pressing [RS] [SOLVE] then choose Solve equation... and then type in EQ the algebraic '4*X+6=7*X'. In the field X: type an initial value for guess, press OK, highlight the X: field again and press SOLVE. You will get the approximate solution. You may also use the polynomial menu to get exactly 2 but you will have to supply the terms of x powers, that is, you will have to collect terms first (eg. 7x-4x=3x etc). You can also use the multi solver menu by pressing [LS] [SOLVE] and then ROOT and then SOLVR. Check it out in the user's guide on how it works. You can also install a very good general purpose solver, namely the solvesys program found in www.hpcalc.org that also handles systems of equations with nonlinear terms. It is very small (less than 5k) and it gets the job done nicely especially for polynomials. You don't have to install erable or alg48 to do these simple things unless you want symbolic answers :-) !Demeter! > lets say I want to input: > 4x + 6 = 7x > so that I can solve for x > Can it be done? Sure. Type in: '4*x+6=7*x' 'x' ISOL -- ----- Je suis a la recherche de la recette d'un space cake. Je ne trouve que des recettes a base d'herbes. Je voudrais une recette avec du shit (haschish), j'aurais de plus une preference pour un gateau au chocolat. -+- fred in : Ah ben cake qui space mon neuneu? > Sure. Type in: > '4*x+6=7*x' > 'x' > ISOL ... and you'll get unable to isolate. Better use QUAD. And much better, instal Erable and use SOLVE. > Sure. Type in: > '4*x+6=7*x' > 'x' > ISOL > ... and you'll get unable to isolate. Better use QUAD. > And much better, instal Erable and use SOLVE. I hadn't tried on my real HP48GX, I just thought such a simple expression would be something easy for the built-in CAS of the HP48G... It works better if you reorder the expression: '4*x+6-7*x=0' COLCT 'x' ISOL And you get the answer: 'x=2'. IMHO, installing Erable for such simple expressions is a little bit 'overkill'. You can get the answer faster by *not* using your calculator... -- ----- surtout avec les pentium 3. Merci. -+- CA in : Guide du Neuneu Usenetien - Souriez vous .90tes film.8es -+- > IMHO, installing Erable for such simple expressions is a little bit > 'overkill'. You can get the answer faster by *not* using your > calculator... Completely agree with you. But I hope you believe me when I say that I was not thinking ONLY in first degree equations when I recommended install Erable. By the way, there's no room in standard G+ for installing all Erable libraries and I use them frecuently. Regrads > > IMHO, installing Erable for such simple expressions is a little bit > 'overkill'. You can get the answer faster by *not* using your > calculator... > Completely agree with you. > But I hope you believe me when I say that I was not thinking ONLY in first > degree equations when I recommended install Erable. Of course. I just think that if someone has some problems to solve such a simple equation, then Erable will be of no use for him ;) I'm not talking about you, but about the original poster. > By the way, there's no room in standard G+ for installing all Erable > libraries and I use them frecuently. It's been a while since I installed Erable (and ALG48), I don't remember all these details... -- ----- ``Numbers talk, bullshit walks.'' Anonymous Speaking as the original poster, I didn't realize the number of cynics that felt it necessary to post completely useless replies to a curious question. I actually feel sorry for know if the calculator could do it without trying to decipher the manual. in reply to: > Of course. I just think that if someone has some problems to solve such a > simple equation, then Erable will be of no use for him ;) > I'm not talking about you, but about the original poster. I could probably solve equations in my sleep that would drive you insane. I was putting a simple equation like that in my post only to make it easier to understand for you. The numbers I'm working on would probably make you blind just by looking at them. Have a nice day! :) > > IMHO, installing Erable for such simple expressions is a little bit > 'overkill'. You can get the answer faster by *not* using your > calculator... > > Completely agree with you. > But I hope you believe me when I say that I was not thinking ONLY in first > degree equations when I recommended install Erable. > Of course. I just think that if someone has some problems to solve such a > simple equation, then Erable will be of no use for him ;) > I'm not talking about you, but about the original poster. > By the way, there's no room in standard G+ for installing all Erable > libraries and I use them frecuently. > It's been a while since I installed Erable (and ALG48), I don't remember > all these details... > -- > ----- > ``Numbers talk, bullshit walks.'' > Anonymous > I could probably solve equations in my sleep that would drive you insane. > I was putting a simple equation like that in my post only to make it easier > to understand for you. The numbers I'm working on would probably make you > blind just by looking at them. And so, thinking in those equations, I recomended you install Erable. > Speaking as the original poster, > I didn't realize the number of cynics that felt it necessary to post > completely useless replies to a curious question. I actually feel sorry for to > know if the calculator could do it without trying to decipher the manual. Well. I think you have now one more answer: read the manual to be able to get the most of your HP48. It's sad that the HP48G manual is not as good as older ones from HP (for example the ones that come with the HP48S, HP42S, HP28S, HP15/16C, ...). The HP49G manual is even worse. I think you can read the HP48S manuals, most of them is still valid with the HP48G. I'd bet these manuals are available on one of the CDs sold by www.hpmuseum.org. One more answer: the HP48 CAS is somewhat limited, and as you have probably noticed, you can greatly enhance it with additional libraries, such as Erable or ALG48. > in reply to: > Of course. I just think that if someone has some problems to solve such a > simple equation, then Erable will be of no use for him ;) > I'm not talking about you, but about the original poster. > I could probably solve equations in my sleep that would drive you insane. Probably, but that's not the point. > I was putting a simple equation like that in my post only to make it easier > to understand for you. The numbers I'm working on would probably make you > blind just by looking at them. Well. I manipulate big numbers everyday, it's my job. I'm a cryptographer. We can go on with such 'mine is bigger than yours' arguments, but I doubt we'll go far ;) > Have a nice day! :) -- ----- C'est pas pour dire, mais j'ai d.9e mal avoir .88 quel objet pr.8ecis se rapportent ce forum dans les newsgroup du web. -+- JP in GNU : un peu de pr.8ecision dans les newsgroup du web -+- Well, *heh heh* you win. a cryptographer I am not. ~C > Speaking as the original poster, > I didn't realize the number of cynics that felt it necessary to post > completely useless replies to a curious question. I actually feel sorry for to > know if the calculator could do it without trying to decipher the manual. > Well. I think you have now one more answer: read the manual to be able to > get the most of your HP48. It's sad that the HP48G manual is not as good > as older ones from HP (for example the ones that come with the HP48S, > HP42S, HP28S, HP15/16C, ...). The HP49G manual is even worse. I think you > can read the HP48S manuals, most of them is still valid with the HP48G. > I'd bet these manuals are available on one of the CDs sold by > www.hpmuseum.org. > One more answer: the HP48 CAS is somewhat limited, and as you have > probably noticed, you can greatly enhance it with additional libraries, > such as Erable or ALG48. > in reply to: > Of course. I just think that if someone has some problems to solve such a > simple equation, then Erable will be of no use for him ;) > I'm not talking about you, but about the original poster. > > I could probably solve equations in my sleep that would drive you insane. > Probably, but that's not the point. > I was putting a simple equation like that in my post only to make it easier > to understand for you. The numbers I'm working on would probably make you > blind just by looking at them. > Well. I manipulate big numbers everyday, it's my job. I'm a cryptographer. > We can go on with such 'mine is bigger than yours' arguments, but I doubt > we'll go far ;) > Have a nice day! :) > -- > ----- > C'est pas pour dire, mais j'ai d.9e mal avoir .88 quel objet pr.8ecis se > rapportent ce forum dans les newsgroup du web. > -+- JP in GNU : un peu de pr.8ecision dans les newsgroup du web -+- > Of course. I just think that if someone has some problems to solve such a > simple equation, then Erable will be of no use for him ;) > I'm not talking about you, but about the original poster. And I hope that if someone has problems to solve such a simple equation, learn Maths instead playing with expensive toys. Sorry... just the view of a Maths teacher ;-) > I hadn't tried on my real HP48GX, I just thought such a simple expression > would be something easy for the built-in CAS of the HP48G... > It works better if you reorder the expression: To the totally confused I add two commands before COLCT 'X' ISOL: EQ-> and - Veli-Pekka > '4*x+6-7*x=0' > COLCT > 'x' > ISOL > And you get the answer: 'x=2'. Help again guys; I dug through my Urroz book and can't find it and then through that worthless manual. I am having the toughest time here. Could someone please tell me how to copy an equation from the x register and from the command line to the equation writer please? Jo If you are using the HP 49G you can press HIST key while already in the EQW and ECHO the desired line to your EQW Also note that it's possible to EDIT and then select via BEGIN and END a portion to COPY to clipboard for later PASTE into you EQW equation. Note: The UNDO works, too! > Help again guys; > I dug through my Urroz book and can't find it and then through that > worthless manual. I am having the toughest time here. Could someone please > tell me how to copy an equation from the x register and from the command > line to the equation writer please? > Jo > If you are using the HP 49G you can press HIST key while > already in the EQW and ECHO the desired line to your EQW > Also note that it's possible to EDIT and then select via > BEGIN and END a portion to COPY to clipboard for later > PASTE into you EQW equation. Note: The UNDO works, too! Great ideas thank you very much! As usual you guys help me out. I can't tell you how much I appreciate this. I had the same problem, if I understand you correctly, I.e. how to edit an equation (or matrix) after it's been put on the stack. Answer is very simple - select the view option (in the area of the cursor keys). The HP48 seems to know what the best viewer is for the object on the stack. Hope this helps... , Daniel >Help again guys; >I dug through my Urroz book and can't find it and then through that >worthless manual. I am having the toughest time here. Could someone please >tell me how to copy an equation from the x register and from the command >line to the equation writer please? >Jo instead of speaking of x register, I prefer to speak of stack register no. 1...I think this is acchived by switching to RPN mode... having a equation on the command line you only have to press ENTER to move the equation to stack register no. 1. If you want to keep a copy of the equation before editing it in the equation writer first press ENTER to have two identical equations in stack level 1 and 2. Now pressing down-arrow.... and voila... you now have the equation in the equation writer ready to edit... and when you have finished editing, all you have to do is press ENTER... now you have the original equation in stack level 2 and your edited equation in stack level 1. Pressing down-arrow always open the most appropriate editor to edit the contents of stack level 1: If you have an equation or algebraic expression on stack level 1 it will be moved to the equation writer for editing... if you have a matrix in stack level 1 it will be moved to the matrix writer for editing... after pressing ENTER the edited object will be returned to stack level 1. I don't hope that the above explanation is too confusing...:-) Martin > Help again guys; > I dug through my Urroz book and can't find it and then through that > worthless manual. I am having the toughest time here. Could someone please > tell me how to copy an equation from the x register and from the command > line to the equation writer please? > Jo > > a bit when I get my manuals. I did buy those manuals from Booksurge and am anxiously awaiting their arrival. Is anyone else having a problem with Allmem 2.0 crashing their 48GX when looking at a covered port with no object stored in it? After invoking Allmem 2.0, the calc crashes on the next startup (128kb card in slot 1, 1 meg card in slot 2). Is there a better memory browser or a later version of Allmem that isn't so buggy? John > Is anyone else having a problem with Allmem 2.0 crashing their 48GX when > looking at a covered port with no object stored in it? After invoking Allmem > 2.0, the calc crashes on the next startup (128kb card in slot 1, 1 meg card > in slot 2). > Is there a better memory browser or a later version of Allmem that isn't so > buggy? > John Have you tried Filer48? It seems a good browser very similar to the MK or 49 Filer. calc (or was it the ports of the calc?). Go to hpcalc.org and search for my name as author - it is the only program that I have submitted so it shouldn't be that hard to find...:-) Martin > Is anyone else having a problem with Allmem 2.0 crashing their 48GX when > looking at a covered port with no object stored in it? After invoking Allmem > 2.0, the calc crashes on the next startup (128kb card in slot 1, 1 meg card > in slot 2). > > Is there a better memory browser or a later version of Allmem that isn't so > buggy? > > John > > now I found the program... it is not a browser but just a program to show, how much free memory you have in the different ports (covered and uncovered. It is called PMEM (for Port Memory) and library has the ID no. 1269. Martin > calc (or was it the ports of the calc?). Go to hpcalc.org and search for > my name as author - it is the only program that I have submitted so it > shouldn't be that hard to find...:-) > Martin > > >> Is anyone else having a problem with Allmem 2.0 crashing their 48GX >> when >> looking at a covered port with no object stored in it? After invoking >> Allmem >> 2.0, the calc crashes on the next startup (128kb card in slot 1, 1 meg >> card >> in slot 2). >> >> Is there a better memory browser or a later version of Allmem that >> isn't so >> buggy? >> >> John >> >> > > now I found the program... it is not a browser but just a program to > show, how much free memory you have in the different ports (covered and > uncovered. It is called PMEM (for Port Memory) and library has the ID > no. 1269. > Martin nice to hear that someone can (or could) use the program... aside of the compliment it is to have a program on hpcalc.org, your comment is the first positive comment about it...:-) Martin > >>now I found the program... it is not a browser but just a program to >>show, how much free memory you have in the different ports (covered and >>uncovered. It is called PMEM (for Port Memory) and library has the ID >>no. 1269. >>Martin > > > > > > > nice to hear that someone can (or could) use the program... aside of the > compliment it is to have a program on hpcalc.org, your comment is the > first positive comment about it...:-) > Martin Yes, Martin: it was installed doing (wiyhout problems) the work that the 48 can't do... untill MK with its Filer... What is MK and what is it used for? > now I found the program... it is not a browser but just a program to > show, how much free memory you have in the different ports (covered and > uncovered. It is called PMEM (for Port Memory) and library has the ID > no. 1269. > Martin > > > I retract the stupid question. I answered too fast before realizing it was Metakernel. > What is MK and what is it used for? > now I found the program... it is not a browser but just a program to > show, how much free memory you have in the different ports (covered and > uncovered. It is called PMEM (for Port Memory) and library has the ID > no. 1269. > Martin > > > http://miaif.lip6.fr/~squelart/english/mk.html > What is MK and what is it used for? I am a new user of any type of HP calculators. I bought a 49g about 4 months ago and have been trying to get used to it in time for classes this fall. One problem that I have not been able to solve is how to enter text and equations and then be able to save it and later look at it again to read it. I used to do this for my Caulc I class for givens (derivatives of trig and such) that we needed for assignments and tests. I only have the users guide that came with the 49, so, if any one could help me, or point me in the right direction, I would appriciate it. Loug1806 this is in fact quite easy. I assume you are in RPN mode. You have to put your text between ... and save them as a string. Example: This is just a test [ENTER] 'Test' [ENTER] [STO] The variable 'Test' holds now the string. If you want to put your formulas also in the string you have them to put between ' .. '. Example: Pythagoras 'a^2+b^2=c^2' kineticE 'Ek=1/2*m*v^2' [ENTER] 'Pyth' [ENTER] [STO] But keep in mind you have to copy and paste your formulas in order to use them. Roman Loug1806 schrieb im Newsbeitrag > I am a new user of any type of HP calculators. I bought a 49g about 4 > months ago and have been trying to get used to it in time for classes > this fall. One problem that I have not been able to solve is how to > enter text and equations and then be able to save it and later look at > it again to read it. I used to do this for my Caulc I class for > givens (derivatives of trig and such) that we needed for assignments > and tests. I only have the users guide that came with the 49, so, if > any one could help me, or point me in the right direction, I would > appriciate it. > Loug1806 Roman, trying to figure it out. Stan > this is in fact quite easy. I assume you are in RPN mode. > You have to put your text between ... and save them as a string. > Example: > > This is just a test [ENTER] > 'Test' [ENTER] > [STO] > > The variable 'Test' holds now the string. > If you want to put your formulas also in the string you have them to put > between ' .. '. > Example: > > Pythagoras 'a^2+b^2=c^2' kineticE 'Ek=1/2*m*v^2' [ENTER] > 'Pyth' [ENTER] > [STO] > > But keep in mind you have to copy and paste your formulas in order to use > them. > > Roman > > > Loug1806 schrieb im Newsbeitrag > I am a new user of any type of HP calculators. I bought a 49g about 4 > months ago and have been trying to get used to it in time for classes > this fall. One problem that I have not been able to solve is how to > enter text and equations and then be able to save it and later look at > it again to read it. I used to do this for my Caulc I class for > givens (derivatives of trig and such) that we needed for assignments > and tests. I only have the users guide that came with the 49, so, if > any one could help me, or point me in the right direction, I would > appriciate it. > Loug1806 I used this loop for a bicycle speedometer. The loop works fine except that the maximum speed is 5.6 revolution to second(that mean ~44Km/h). I would appreciate it if someone could help me to achieve 60Km/s ~7.7Hz. 0 STIME TICKS DO DO 1 SRECV DROP2 TICKS DUP ROT - B->R 64290.816 SWAP / SWAP + Km/h. + 3 DISP 'j' INCR DROP UNTIL KEY END UNTIL 13 == END Tal machine-language programming. There are also some applications with source code. This site will very soon have to be discontinued and I am looking for someone to put these things on their own homepage. If you are interested, contact me as soon as possible (the site goes down on 29/9 but I'll keep everything on my hd). This is the page: http://www.ite.mh.se/~danli97/hp48.html -- Daniel Lidstr.9am danli97@_DELETE_ite.mh.se or (after 29/9): dali9800@_DELETE_student.mh.se http://www.ite.mh.se/~danli97/ 62 23' 25 N 17 15' 31 E Hey All: I was reading Riemann's paper On the Number of Primes Less Than a Given Magnitude and was checking the values in a table that he calculated (actually, he did a very impressive job considering he calculated by hand, so long ago). The definite integral I was looking at is: 1/log(x) and the limits are 2 to 500,000...1,000,000...1,500,000.... For 2 to 500,000: Mathematica 4 gives the approximation 41,605.2 (and WON'T give more decimal places) Derive 5.05 gives 41,605.24366 Maple V gives 41,605.24373 and my HP49 gives 95,799.6137 (and takes about 30 min. to do so). What sort of algorithm would be used for integral of (1/log(x)) and why is my HP's answer so different from the others? I ran it twice and I *think* I entered everything correctly..... Roger Metcalf Arlington, Texas, USA > Hey All: I was reading Riemann's paper On the Number of Primes Less Than a > Given Magnitude The PrimePi function? > and was checking the values in a table that he calculated > (actually, he did a very impressive job considering he calculated by hand, > so long ago). The definite integral I was looking at is: > > 1/log(x) and the limits are 2 to 500,000...1,000,000...1,500,000.... > > For 2 to 500,000: > Mathematica 4 gives the approximation 41,605.2 (and WON'T give more decimal > places) In[1]:= NIntegrate[1/Log[x], {x,2,500000}, WorkingPrecision->50, MaxRecursion->20] Out[1]= 41605.24362265250075725157712483776450994 In[2]:= Precision[%] Out[2]= 40 The integration took about 1.48 seconds. -- Bhuvanesh > The integration took about 1.48 seconds. > -- > Bhuvanesh How the heck did Riemann do this integral? (I guess stuff like that was why he WAS Riemann....) Roger > Hey All: I was reading Riemann's paper On the Number of Primes Less Than a > Given Magnitude and was checking the values in a table that he calculated > (actually, he did a very impressive job considering he calculated by hand, > so long ago). The definite integral I was looking at is: > > 1/log(x) and the limits are 2 to 500,000...1,000,000...1,500,000.... > > For 2 to 500,000: > Mathematica 4 gives the approximation 41,605.2 (and WON'T give more decimal > places) > Derive 5.05 gives 41,605.24366 > Maple V gives 41,605.24373 > and my HP49 gives 95,799.6137 (and takes about 30 min. to do so). > > What sort of algorithm would be used for integral of (1/log(x)) and why is > my HP's answer so different from the others? I ran it twice and I *think* I > entered everything correctly..... I would bet that log is natural logarithm on Mathematica, Derive and Maple and decimal log on the HP. > I would bet that log is natural logarithm on Mathematica, Derive > and Maple and decimal log on the HP. As usual, you are correct, sir. Actually I DID check in Derive using Log(x) and Ln(x) notation and got the same result for both, not realizing that I should have used Log(x,10) for the decimal log. Of course, the answer from the 49 was correct--I was asking the wrong question! Roger Metcalf Arlington, Texas, USA Roger: > 1/log(x) [Riemann's asymptotic prime counting estimator] Use 'INV(LN(X))' [ *not* decimal LOG ] > For 2 to 500,000: > Mathematica 4 gives the approximation 41,605.2 Set your calc to 4 SCI display mode before integrating; it will get a close answer pretty soon. ((o)) http://www.mum.edu . I am still learning a few things about my 49, and needing to learn a whole lot more. I downloaded a physics constants library to my cacl, but I cant access it. I also have the same problem with other programs that I have downloaded. How do I access them and be able to see the constants in the library, or to use the other programs. When I do find them and have their name in the input line and press enter, it only re-inputs the name. Loug1806 Loug: A library object (which displays as Library NNN on stack) must be stored in a port to work, rather than in a variable. With library on stack, type the digit 1 or 2 and press STO [that's for RPN mode; use STO(1) or STO(2) in ALG mode] A library generally contains multiple functions, which add themselves to those that are built into the calc (actually, all functions that are built into the calc are also in libraries, but just located in ROM instead). After your library is stored, its name generally appears in the right-shift LIB menu, and its command menu appears when you press its name, or when you do nnnn MENU (specifying the actual library number); the commands are also included in the 49G CATalog. ((o)) http://www.mum.edu . John and Roman, lot less stressful. Also, I have been able to use some useful for all the help. Loug1806 > Loug: > > A library object (which displays as Library NNN on stack) > must be stored in a port to work, rather than in a variable. > > With library on stack, type the digit 1 or 2 and press STO > [that's for RPN mode; use STO(1) or STO(2) in ALG mode] > > A library generally contains multiple functions, which add themselves > to those that are built into the calc (actually, all functions > that are built into the calc are also in libraries, > but just located in ROM instead). > > After your library is stored, its name generally appears > in the right-shift LIB menu, and its command menu appears > when you press its name, or when you do nnnn MENU (specifying > the actual library number); the commands are also included > in the 49G CATalog. > > ((o)) http://www.mum.edu > > > > > > > > > > > > > > > > > > > > > > > > > > . > > hi, I assume your calc is in RPN mode. I assume further you know how to connect your calc with the PC and send the files (libs) to the calc. Press on your calc the [VAR]-key to see the file you downloaded. In case you can't see it, press the [NXT]-key till you see it. Then press the corresponding soft key for the file [F1-F6] to put the file on the stack. Store the Lib now in a port 0, 1 or 2. To store the lib in port 2 put the numer 2 on the stack and press the [STO] key. Now make a warmstart [ON]&[F3]. The libs should now be attached to your calc. You can find your libs now by pressing [R-SHIFT]-[2/LIB]. Most libs have also some documentation how to use them. Roman Loug1806 schrieb im Newsbeitrag > I am still learning a few things about my 49, and needing to learn a > whole lot more. I downloaded a physics constants library to my cacl, > but I cant access it. I also have the same problem with other > programs that I have downloaded. How do I access them and be able to > see the constants in the library, or to use the other programs. When > I do find them and have their name in the input line and press enter, > it only re-inputs the name. > Loug1806 Will these cards work properly in the 48gx? Is there a better card to buy? Do any of the Cynox cards (2M, 1M, 512k, 256k, 128k) work capacity perspective, the 1MB seems optimal, but I wanted to get some advance, Steve hi Steve, i bought two ramcards (128Kb,1Mb) for my 48gx from O.Klotz in Germany. Really i'm happy of my choice for quality 'n price (http://uuhome.de/oklotz/index_e.html) see ya, Acrux, Italy > Will these cards work properly in the 48gx? Is there a better card to > buy? Do any of the Cynox cards (2M, 1M, 512k, 256k, 128k) work > capacity perspective, the 1MB seems optimal, but I wanted to get some > advance, > Steve A created a multitasking OS for hp49. It currently only beta. Can somebody upload it somewhere ? I can send > > A created a multitasking OS for hp49. > It currently only beta. Can somebody upload it somewhere ? I can send > OK. It is @ cooldemo.miesto.sk I'm warning you it has nothing. it is kinda just demo. Can anyone how to obtain a pointer to HOME dir ??? Sorry if this gets posted twice but I got a timeout error on the first try. >Can anyone how to obtain a pointer to HOME dir ??? In sys-rpl : :: CONTEXT@ ( Get current directory ) SYSCONTEXT ( Change to home directory... ) CONTEXT@ ( ...and get address of it ) SWAP CONTEXT! ( Restore original directory ) ; or in assembly : CODE LC(5) =USEROB * #80711 , pointer to home dir CD0EX A=DAT0 A * A.A contains address of home dir D0=C GOVLNG =DOBINT * Push address to stack ENDCODE or a shorter version of the above : CODE LC(5) =USEROB GOVLNG =SysPtr@ * Push address of pointee to stack ENDCODE -------------------------------------------------------------------------- -- --- Jonathan Busby - before replying. >Can anyone how to obtain a pointer to HOME dir ??? > In sys-rpl : > :: > CONTEXT@ ( Get current directory ) > SYSCONTEXT ( Change to home directory... ) > CONTEXT@ ( ...and get address of it ) > SWAP CONTEXT! ( Restore original directory ) > ; I think the shortest SysRPL program to recall (the pointer to) the HOME directory from any position is :: SaveVarRes HOMEDIR CONTEXT@ RestVarRes ;-) - Wolfgang memory for my hp48gx and hp48s? i search for schemes and pcb for upgrade the memory for the gx model i have found a 2 mega card for 70 euro ,it's a good price ? roberto cardascia studente del politecnico di bari you can find something (about schemes,etc.. on hpcalc.org). For your hp48gx IMHO the best choice are two memcards (a 128kb for slot1 and a 1Mb for slot two). As you can experimented by Emu48 they are enough. see ya, Acrux, Italy http://it.groups.yahoo.com/group/lista_hp48/ > memory for my hp48gx and hp48s? > i search for schemes and pcb for upgrade the memory > for the gx model i have found a 2 mega card for 70 euro ,it's a good price > ? > roberto cardascia studente del politecnico di bari > for the gx model i have found a 2 mega card for 70 euro ,it's a good price > ? Possibly the best one. But, will you really need 2Mb? I have a lot of free mem in my 1Mb card: and believe me: with a lot of stuff! Is it possible to do this in hp49g ? 1. Conformal Transformation of regions 2. Calculate Function Series where the function is defined from C-->C Matias > Is it possible to do this in hp49g ? > 1. Conformal Transformation of regions Well, if you mean that the HP49G should check if the mapping is conformal, then you have to program a little. If you mean just showing what the mapping lloks like, then you have the plot type GRIDMAP, which will draw what the transformation looks like. > 2. Calculate Function Series where the function is defined from C-->C The command SERIES does this in case of an explicitely defined function like Z-->e^Z . If you have something else (parametric functions etc), then you must convert it to its explicit form (if possible). Nick. After many doubts I'll be there from Saturday morning. Who else be there? Would be good to see as many people as possible. ...and many of us, will be here waiting for your posts... Please and Alas, I will be stuck over the pond here in the states. Maybe I should arrange a meeting in Nashville Tennessee next year? I'm sure I could make that one. ;-) Hopefully, Jake will be videotaping the event. That's the way I know most of you guys...you've been on my TV many times. Gene -- * All opinions herein expressed are mine and * * mine alone. You may choose to ignore them * * but I own them. * > ...and many of us, will be here waiting for your posts... Please and someone put a webcam! hi hi >Do you think it is possible to make a connection between a HP-48 and >the modem of a Mac (which has no serial port...) to send/receive file? There was a post about a year back from someone who claimed to have done this. Google is your friend. You'll need a second modem though - you can't connect the Mac's built-in modem directly to the 48's serial port. However, you can connect the Mac's modem to a second modem [1]. Then you'll need to change the settings to turn off a load of default stuff: e.g. *don't* wait for a dialtone; answer immediately as you won't get any rings; turn off all error correction; turn off line quality monitoring and I'm sure there'll be loads more. Choose a slow baud rate - the max the 48 can manage is 9600 so there's no point going any higher. You may even need to establish the connection from both ends: i.e. set up the Mac to connect trying to see what's out there and do the same on the HP48 - you'll need to write a small program to send the appropriate AT commands to the modem attached to the 48 and hopefully they'll meet in the middle, so to speak. There's no way it'll work if you try and have one answer automatically since the rings etc. are generated by the telephone company and you don't have one of those in the middle. [2] [1] It needs to be mains or battery powered. A line powered one won't work. [2] Unless of course you have free local phone calls and two lines. In which case you can just dial yourself! -- Bruce Horrocks Hampshire England bh@granby.demon.co.uk I have very good news!, next week I will release version 2.0, it's about time to make a new release, plus on September Filer48 became 2 years old!! THANK YOU very much for all your suggestions, if it wasn't for you Filer48 wouldn't be what it is today, so please keep sending in your letters!!! What's new in version 2.0? For now, I can tell you this: - 5 Flavors to choose from!!! - Windows' Explorer style Interface!!! - New API for customizing the Filer!!! - Much more... I can assure you this is the best version ever! I think you're going to love it! at least I hope so ;) Nuevo explorador de archivos para la hp48 Un gran librer.92a para la hp48. Me gustar.92a que alg.9cn d.92a FILERhp49 sea similar a FILERhp48 Tienes alguna pregunta? Alguna sugerencia para el programa? Encontraste un Bug que se me escap.97? Raoolio@yahoo.com Ra.9cl del Cid L.97pez No tienes una hp48real descarga emu48 con mejores im.87genes para la emulaci.97n en: www.geocities.com/jaimezacalcs/hp/pc/skinhp3-4x.htm www.geocities.com/hp4x www.geocities.com/jaimezacalcs > Nuevo explorador de archivos para la hp48 > Un gran librer.92a para la hp48. Me gustar.92a que alg.9cn d.92a FILERhp49 sea > similar a FILERhp48 > Tienes alguna pregunta? > Alguna sugerencia para el programa? > Encontraste un Bug que se me escap.97? > Raoolio@yahoo.com > Ra.9cl del Cid L.97pez Que? Translated of Spanish to English with Power translator v7.0 Pardon for incorrect syntax New browser of files for the hp48 in: www.geocities.com/jaimezacalcs/hp/filerhp48.zip A great library for the hp48. I would like that some day FILERhp49 is similar to FILERhp48 Jaime Fernando Meza Do you have some question? Some suggestion for the program? Did a Bug that I was escaped find? Raoolio@yahoo.com Ra.9cl del Cid L.97pez You don't have a hp48real it discharges emu48 with better images for the emulation in: www.geocities.com/jaimezacalcs/hp/pc/skinhp3-4x.htm www.geocities.com/hp4x www.geocities.com/jaimezacalcs I've got OS 1.19.6. I have approximate unchecked in the cas. Yet when I put in a matrix and use the LU command on it I get decimals instead of fractions. How can I make sure it returns the exact answer instead of the approximation? > I've got OS 1.19.6. I have approximate unchecked in the cas. Yet when I put > in a matrix and use the LU command on it I get decimals instead of > fractions. How can I make sure it returns the exact answer instead of the > approximation? > > LU and many of the other matrix commands are carryovers from the HP48 which did everything in approximate mode, so the 49 also does these commands in approximate mode regardless of exact/approximate settings. You can often get the correct exact expressions by using the conversion commands: XQ or ->Q or ->Qpi on your resulting matrices.( -> is the right arrow symbol and pi is the pi symbol) WARNING: You should, however, check the exact results produced by these commands to see if they are correct, since roundoff errors may cause conversions to incorrect exact matrices. The first two of the above commands also switches to exact mode when executed, but the ->Qpi command leaves the exact/approximate mode unchanged. XQ and ->Qpi will also produce rational multiples of pi when appropriate but ->Q will only produce rationals (or integers). The next program is the standard assignment of RS MENU for going back to the last menu: :: MenuMaker ROMPTR A9 0 DoMenuKey ; @ When I run this program, I get the last menu, but I need press ON for refreshing the screen... What do I need for getting the autorefreh? The next program is the standard assignment of RS MENU for going back to the last menu: :: MenuMaker ROMPTR A9 0 DoMenuKey ; @ When I run this program, I get the last menu, but I need press ON for refreshing the screen... What do I need for getting the autorefreh? > When I run this program, I get the last menu, but I need press ON for > refreshing the screen... > What do I need for getting the autorefreh? I think that maybe ClrDA3OK help you. I don't have the calc here so I can't test. Hope this helps Luis > I think that maybe ClrDA3OK help you. I don't have the calc here so I can't test. Adding this command to thr prg, I still need press ON for refreshing the stack :-( > I think that maybe ClrDA3OK help you. I don't have the calc here so I > can't test. > Adding this command to thr prg, I still need press ON for refreshing the > stack :-( The strongest and perhaps fastest command to refresh the display completely is SysDISPLAY - Wolfgang I think you should use ClrDAsOK, which does :: ClrDA1OK ClrDA2OK ClrDA3OK ; (stack area 1 and 2 and menu area) Caspar If this does not work I have to check my Donnely book... > I think that maybe ClrDA3OK help you. I don't have the calc here so I > can't test. > Adding this command to thr prg, I still need press ON for refreshing the > stack :-( Caspar Lugtmeier & Eva Skotarczak escribi.97 en el > I think you should use ClrDAsOK, which does > :: ClrDA1OK ClrDA2OK ClrDA3OK ; > (stack area 1 and 2 and menu area) > Caspar > If this does not work I have to check my Donnely book... Don't check your book :-) THANKS If you type in RPN: 2 sqrt (the button, not typing) enter EVAL then it says (sqrt(2))^2 does anyone know why it doesn't answer only 2? If you then try to -> NUM it answers 1.99999999999 and im having exact (=) mode set. Is their some kind of setting that i maybe have set accidently? Rom 1.19-6 Well, now i have downgraded to rom 1.18 and now everything works fine again. Don't know what the problem was unforunately. The 1.19-6 worked fine in the emulator, but not on my real hp49g. Maybe i give it a chance somewhere in the future. // Jocke Please friend! It does not get any easier that this. Use EVAL instead for ->NUM to get the exact answer 2. !Demeter! > If you type in RPN: > 2 > sqrt (the button, not typing) > enter > EVAL > then it says (sqrt(2))^2 Are you sure you're not missing something in your description? 2 SQRT EVAL return SQRT(2) > Are you sure you're not missing something in your description? > 2 SQRT EVAL return SQRT(2) > Yes i did, sorry. I missed the multiplication sign. I try one more time: RPN-mode: 2 sqrt (the button) enter (duplicates sqrt(2), so you have two identical terms) * (multiplication sign) EVAL Now my 49g answers (sqrt(2))^2 If i do a ANS -> NUM it says: 1.99999999... e.g. Can't 49g tell sqrt(2) * sqrt(2) exact? // Jocke > e.g. Can't 49g tell sqrt(2) * sqrt(2) exact? Try EXPAND > Try EXPAND > Done that now, and it still results in (sqrt(2))^2 :-( // Jocke to do VERSION and 64 STWS HEX RCLF Here's the result from my hp49g VERSION Copyright HP 2001 and... 64 STWS HEX RCLF { # 38205010FF1h # 1000000000000001h # 1004400002000800h # 0h } Hope this will help me and you... =) It appears to be flag -125 being set which is causing 'v/2^2' not to simplify using either EVAL or EXPAND. Why? ((o)) . > to do VERSION and 64 STWS HEX RCLF > > Here's the result from my hp49g > > VERSION > Copyright HP 2001 > > and... > > 64 STWS HEX RCLF > { # 38205010FF1h > # 1000000000000001h > # 1004400002000800h > # 0h } > > Hope this will help me and you... =) I tried this on my HP 49. The result of the multiplication is exactly 2 (i.e., there is no need to do EVAL). I also tried entering (SQRT(2)*SQRT(2))^2 directly, and doing EVAL; the result is 2 exactly. Both numeric and approximate are not checked. The binary word size (result of RCWS) is 64. My ROM version is 1.18. Is it possible this issue only arises with certain ROM versions? Geoff Garner > Try EXPAND > > Done that now, and it still results in (sqrt(2))^2 > :-( I have a 48GX with Erable: I get 2 in exact mode (the same in my 49 emu): 2 sqrt ENTER * EXPAND Sorry for can not help more... Interested in the great book Mastering the HP68G/GX book? look at my auction in ebay, the book is in very good condition: This is a book in like new condition. Mastering the HP 48G/GX: A step by step, easy to read introduction to operating and programming the HP48G/GX. Great for students who own an HP calculator, it provides a great introduction and mid level treatment of the RPL programming in the HP calculators. http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2053646105 Mariano I bought an HP49 for the younger brother of my girlfriend to replace a TI. He was thrilled and instantly proud of it. I spent time giving him a crash course in the machine, and how to do the ROM updates. A bright young guy, he noticed that the ROM updates were almost 2 years old. He asked aren't there any newer ones? I answered no, and briefly explained what I knew about HP's calculator division problems. He then said this machine is dead in its tracks, huh? He was then visibly depressed, as if his great new gift was now a tarnished cast-off. There was little I could say, except that I found the HP49 of great value in my engineering studies, and it would surely help him in high school. As I drove home, it made me wonder the same thing: is the HP49 truely dead in its tracks, are HP calcs all dinosaurs, relics of a by gone era? J.C. Randerson Pueblo, Colorado As long as there is support the currnet calculators will be current. If it breaks you can get it fixed or get a new one. Who cares if the rom updates are 2 years old. The ROMS are updated to get rid of bugs. no more bugs ... no more updates. Newer is not better all the time. I still wish the 41 series was in production. I still think that was my favorite calulator - inspite of some relliability issues. If somone gave me an 49 in highschool I'd be jazzed ('course back then I used a slide rule). Heck if somone gave me one now I'd be. Heck it's the brother of your girlfriend. The little snot should be happy you gave him anything! As long as you can get your calc replaced/repaired by HP it's no DINO. Even then ... after that it's no dino as long as it keeps working! On 21 Sep 2002 10:14:33 -0700, jrand@aculink.net (J.C. Randerson) >I bought an HP49 for the younger brother of my girlfriend to replace a >TI. He was thrilled and instantly proud of it. I spent time giving him >a crash course in the machine, and how to do the ROM updates. A bright >young guy, he noticed that the ROM updates were almost 2 years old. He >asked aren't there any newer ones? I answered no, and briefly >explained what I knew about HP's calculator division problems. He then >said this machine is dead in its tracks, huh? He was then visibly >depressed, as if his great new gift was now a tarnished cast-off. >There was little I could say, except that I found the HP49 of great >value in my engineering studies, and it would surely help him in high >school. As I drove home, it made me wonder the same thing: is the HP49 >truely dead in its tracks, are HP calcs all dinosaurs, relics of a by >gone era? >J.C. Randerson >Pueblo, Colorado > the HP49 ROM updates are almost 2 years old. 1.19-6 is about 11 months old. Of course, HP appears to have dissolved its organization for producing any more such products, but the products still exist, and will keep working anyway, with or without HP's further interest. Former HP employees may or may not produce future firmware enhancements (legally blocked at present), but contributors who participate in this newsgroup do not seem to stop producing more software themselves, also with or without HP's further interest, and they also offer far more ongoing support to users than HP ever did. > He asked aren't there any newer ones? I answered no, and briefly > explained what I knew about HP's calculator division problems. > He then said this machine is dead in its tracks, huh? > He was then visibly depressed, as if his great new gift > was now a tarnished cast-off. > There was little I could say, except that I found the HP49 > of great value in my engineering studies, > and it would surely help him in high school. > As I drove home, it made me wonder the same thing: > is the HP49 truly dead in its tracks, are HP calcs all dinosaurs, > relics of a bygone era? The HP48GX hasn't been updated since 1993, the HP12C (still selling widely) not since before anyone ever heard of the internet, and it seems that very few of the lesser, ordinary calculators of all manufacturers have changed a bit in far longer than that, yet they still seem to be sold to people who find them very useful. Paper, pencils and ink, the most archaic of scholarly tools, are still more helpful (and cheaper) than any calculator in many cases, and the most powerful of all analytical tools apparently hasn't undergone hardware or firmware changes in all of recorded human history, which makes me wonder how your girlfriend's young brother can stand to still be using only his own utterly outmoded brain. With best wishes from http://www.mum.edu . > > He asked aren't there any newer ones? I answered no, and briefly > explained what I knew about HP's calculator division problems. > He then said this machine is dead in its tracks, huh? > He was then visibly depressed, as if his great new gift > was now a tarnished cast-off. > There was little I could say, except that I found the HP49 > of great value in my engineering studies, > and it would surely help him in high school. > As I drove home, it made me wonder the same thing: > is the HP49 truly dead in its tracks, are HP calcs all dinosaurs, > relics of a bygone era? > > The HP48GX hasn't been updated since 1993, the HP12C (still > selling widely) not since before anyone ever heard of the internet, > and it seems that very few of the lesser, ordinary calculators > of all manufacturers have changed a bit in far longer than that, > yet they still seem to be sold to people who find them very useful. > > Paper, pencils and ink, the most archaic of scholarly tools, > are still more helpful (and cheaper) than any calculator in many cases, > and the most powerful of all analytical tools apparently hasn't > undergone hardware or firmware changes in all of recorded > human history, which makes me wonder how your girlfriend's young > brother can stand to still be using only his own utterly outmoded brain. > > > With best wishes from http://www.mum.edu > > If someone had given me a new 49g, even if the last rom update was 11 months ago, disappointment wouldnt have been my reaction. Inversly, if the young lad is so disappointed, slap him with an old ti 81 and tell him to have fun with it. There is wisdom in the passage about not casting your pearls before swine. John H Meyers schrieb im Newsbeitrag > the HP49 ROM updates are almost 2 years old. > 1.19-6 is about 11 months old. >[..] > The HP48GX hasn't been updated since 1993, ...since there was no need;-) It was nearly perfect. Raymond Raymond Hellstern escribi.97 en el > The HP48GX hasn't been updated since 1993, > > ...since there was no need;-) > It was nearly perfect. > Raymond and... many many features that are not in the 48 out of the box and I need/want for my work. HP did not too much for it, but the users wich prgs I'm using, (and myself) did. And EVERYDAY, this calculator amazes me: how it can grow up and be better. And so perhaps you are right: it's nearly perfect ;-) > Raymond Hellstern escribi.97 en el > > The HP48GX hasn't been updated since 1993, > > ...since there was no need;-) > > It was nearly perfect. > > Raymond > Erable > and... many many features that are not in the 48 out of the box and I > need/want for my work. > HP did not too much for it, but the users wich prgs I'm using, (and myself) > did. > And EVERYDAY, this calculator amazes me: how it can grow up and be better. > And so perhaps you are right: it's nearly perfect ;-) > Isn't the 49 such an updated 48? > Isn't the 49 such an updated 48? Yes. I think so. > Isn't the 49 such an updated 48? It was supposed too. Unfortunatly, those who did the 49 wheren't as smart as the 48's developpers where :-( > Paper, pencils and ink, the most archaic of scholarly tools, are > still more helpful (and cheaper) than any calculator in many cases, > and the most powerful of all analytical tools apparently hasn't > undergone hardware or firmware changes in all of recorded human > history, which makes me wonder how your girlfriend's young brother > can stand to still be using only his own utterly outmoded brain. Hmm.. Come to think of it, there is some software you can add to your brain, which helps it quite a bit to think more clearly, to feel far more positive, and to manage its entire body more effectively for its energy, health and longevity. The lack of knowledge of its existence, and lack of interest in its installation, seems as perplexing to me as to why many folks don't discover and update their 49G rom, (or even install all of WR's further enhancements :) Fortunately, unlike HP, the distributor of this brain software is not packing up and going away; we're staying the course: http://www.mum.edu http://www.maharishischooliowa.org/about_maharishi_school.htm http://www.permanentpeace.org/multimedia http://www.permanentpeace.org/multimedia/Hagelin_12.ram http://www.permanentpeace.org/research http://www.tm.org http://www.alltm.org . X > Fortunately, unlike HP, the distributor of this brain software > is not packing up and going away; we're staying the course: > http://www.mum.edu > http://www.maharishischooliowa.org/about_maharishi_school.htm > http://www.permanentpeace.org/multimedia > http://www.permanentpeace.org/multimedia/Hagelin_12.ram > http://www.permanentpeace.org/research > http://www.tm.org > http://www.alltm.org John - to you it's TM - to me it's Jesus Christ. Don't you think you're off topic?! > Don't you think you're off topic?! The primary topic of all education, which is what all sharing of information is for, is to develop the whole of life, both for individuals, at one end of focus of the scale, and of the whole of humanity at the other; if an educator does not keep this in mind, becoming so lost in the details that (s)he forgets to connect them to what we're learning things for, then it's a baseless and useless education that we're dispensing, much as R. Buckminster Fuller has so well pointed out. > to you it's TM - to me it's Jesus Christ Then you must have very little comprehension about at least one of these topics. Students following the instructions for one of these throughout their educational career are scoring in the top 1% of the general public student population; other people following the instructions are spending so much less on medical care that they can get cheaper health insurance for doing it -- and the list goes on and on. Try achieving that with anything that you can get from any church, from any religion, or from anything else available to anyone, anywhere on this planet. One of the things that made present human civilization possible was developing the knowledge of how to culture plants, how to grow vastly more on the same original land. It is equally as significant to know how to get much more out of your own nervous system -- more energy and intelligence, more creativity, finer health, and most of all, feeling finer, under all circumstances. If this isn't *the* topic of greatest importance to civilization, what is? -[]- . > If this isn't *the* topic of greatest importance to civilization, > what is? Embracing the love of God and transferring it to the world -- not contemplating our belly buttons and savoring our egos. Let's get back to the news group at hand. Gregory Warsewicz Savage a.k.a. Greg S X > One of the things that made present human civilization possible > was developing the knowledge of how to culture plants, > how to grow vastly more on the same original land. > It is equally as significant to know how to get much more > out of your own nervous system -- more energy and intelligence, > more creativity, finer health, and most of all, > feeling finer, under all circumstances. > If this isn't *the* topic of greatest importance to civilization, > what is? You seem to have good goals and the means to get there don't sound bad, but they still do to me! Why? I don't believe that man himself can build a better world. You have to co-operate with God. To my experience it's pointless to go on with this so why don't we just agree to disagree. BUT if you insist you can clarify your point Veli-Pekka Perhaps the next best one will be the iPaq50....ugggh... > I bought an HP49 for the younger brother of my girlfriend to replace a > TI. He was thrilled and instantly proud of it. I spent time giving him > a crash course in the machine, and how to do the ROM updates. A bright > young guy, he noticed that the ROM updates were almost 2 years old. He > asked aren't there any newer ones? I answered no, and briefly > explained what I knew about HP's calculator division problems. He then > said this machine is dead in its tracks, huh? He was then visibly > depressed, as if his great new gift was now a tarnished cast-off. > There was little I could say, except that I found the HP49 of great > value in my engineering studies, and it would surely help him in high > school. As I drove home, it made me wonder the same thing: is the HP49 > truely dead in its tracks, are HP calcs all dinosaurs, relics of a by > gone era? > J.C. Randerson > Pueblo, Colorado Hopefully, it is not a fossil in time, but I would say it suffered from the meteoric rise and crash of the 'new era economy'. Large corporation discarded tried and true principles of producing wealth for the get rich quick scheme of mergers and .con's and other paper shuffling scams. Well at least the CEOs made out ... like bandits. be sure to understand http://www.anti-matrix.net i wanna make a link cable but i dont know the order of the wires within the cable.... any one know where i can find out? > i don't know the order of the wires within the [hp48] cable... http://www.hpcalc.org/viewzip.php?id=3517&file=c_hp-pc.txt http://www.engr.uvic.ca/~aschoorl/faq/48faq-12.html#ss12.1 Etc. ((o)) . HP48+ > i wanna make a link cable but i dont know the order of the wires within the > cable.... > any one know where i can find out? everybody requesting it. Now, I have setup a web site to cover the time before HpCalc.org gets updated. Even if it has all the library I ported to the HP49G, it is not intended to replace HpCalc.org. It is just a temporary repository with a faster refresh rate ;-). The URL is http://www3.sympatico.ca/france.vallieres/ Alain Robillard http://www.softintegration.com/ ************************************************************************** ** *** We checked sharp zaurus at http://www.sharp-usa.com/products/FunctionPressReleaseSingle/0,1080,205-29 ,0 0.html It looks that they use Lineo's Embeddix, QT, and Java, Opera Web browser. PDA such as Sharp Zaurus. We will appreciate it if you can ask Sharp Zaurus and Lineo to run Ch. It will be easy for us to work them out. Xiaodong Zhou ************************************************************************** ** cool! i'll see how it turns out. did you guys see the comparison with mathematica they have on the site mentioned bellow? > http://www.softintegration.com/ I have an hp49. I am having touble attaching the fumo library from the hp49 electrical engineering section of the hpcalc.org website. Has anyone had any luck attaching this library? I am able to copy it but my calculate fails to recognize it as a library. I was looking at the manual but i can't find it! It's there a possiblity to solve Equation with letters ? Is there a possiblity to solve it with the calc ? Jani try the SOLVE command ! e.g : 1: 'X' SOLVE HTH Bernd jani schrieb im Newsbeitrag > I was looking at the manual but i can't find it! > It's there a possiblity to solve Equation with letters ? > Is there a possiblity to solve it with the calc ? > Jani thank you very much!!! greets jani thank you so much!!!!! greets jani be sure to understand http://www.anti-matrix.net > >be sure to understand http://www.anti-matrix.net I am starting up a project to develop a 39/40G Shell / Operating System Application to be written in System RPL. I am very new to this language but understand it sort of well. Anyway I want to make this a group project so I am going to need sample code from willing authors. I want this application to adhere to the following standards: - Programmable API for future Programs - Multi-Tasking (like linux with seperate terminals) - Gui Interface - Filing System , File Explorer - Connectivity with other infrared devies If you are willing to assist with this project reply here Help in Need, Rob Morgan Why? > I am starting up a project to develop a 39/40G Shell / Operating System > Application > to be written in System RPL. I am very new to this language but understand > it sort of well. > > Anyway I want to make this a group project so I am going to need sample code > from willing > authors. > > I want this application to adhere to the following standards: > - Programmable API for future Programs > - Multi-Tasking (like linux with seperate terminals) > - Gui Interface > - Filing System , File Explorer > - Connectivity with other infrared devies > > If you are willing to assist with this project reply here > > Help in Need, > Rob Morgan > > > > Why? - To improve on the flaws of the current system - Provide Better security - Faster for Equations or Games - Why was ShellEX developed for the 48G? - More intregration with other infrared devices - A easier system to develop games or apps for the current RPL OS - Fun! Rob Morgan <3d8f00c5@dnews.tpgi.com.au> > Why? > - To improve on the flaws of the current system What flaws? > - Provide Better security Hard to achieve without the help of an MMU in the CPU. > - Faster for Equations or Games I doubt you'll be able to do faster... And as you're talking about games, I don't think that adding a suplementary layer will enhance the speed (mainly because actual games talk directly to the hardware). > - Why was ShellEX developed for the 48G? Why did it stop? > - More intregration with other infrared devices That's the first good reason. Unfortunately, you won't be able to do IrDA faster than 2400 bauds with the HP39 hardware... And if I remember correctly, the lowest available speed for IrDA is 9600 bauds. > - A easier system to develop games or apps for the current RPL OS Well... The actual OS is really good for RPL apps. > - Fun! Just add another one: - Lose time! ;) -- ----- RECHERCHONS HOMMES pr.8esentant peau sensible utilisant d.8eodorant sans alcool, pour tester produits cosm.8etiques, r.8emun.8eration en fin d'essai. -+- in Guide du Neuneu d'Usenet-Halte .88 l'exp.8erimentation neuneutale -+- answer and no idea how to do. I want to install the scribe and todo list in the 1.1, i read there're in the 1.01 ? How i can bring it to run ? Can somebody help me ? Thx Ellum if you have Organizer 1.1 properly installed, you only have to install the Scribe library. Then, from Organizer 1.1, press the Todo softkey (C) and you should have Scribe invoked with a Todo Manager database automatically setup in port 0 for the HP48 and in port 1 for the HP49G. Alain Robillard aml2t3$20km$1@news.imp.ch... > answer and no idea how to do. > I want to install the scribe and todo list in the 1.1, i read there're in > the 1.01 ? How i can bring it to run ? > Can somebody help me ? > Thx > Ellum on sept 30 you will be able to go to MIT for free!! well, almost :) http://news.bbc.co.uk/1/hi/technology/2270648.stm viva the internet! and here it is (initial): http://ocw.mit.edu/index.html not as much as i thought :( and being free, no complaint from me :) > on sept 30 you will be able to go to MIT for free!! well, almost :) > > http://news.bbc.co.uk/1/hi/technology/2270648.stm > > viva the internet! >Another Baltimorean, >Bhuvanesh great! did you go to Townson University? a physics teacher of mine lectures there too: his name is Dr. Harry Olson. Do you know him by any chance? BTW the initial courses from MIT are: Anthropology Biology Chemical Engineering Chemistry Civil and Environmental Engineering Earth, Atmospheric, and Planetary Sciences Economics Electrical Engineering and Computer Science Linguistics and Philosophy Management Mathematics Mechanical Engineering Ocean Engineering Political Science Urban Studies and Planning more info here: http://web.mit.edu/ocw/ > on sept 30 you will be able to go to MIT for free!! well, almost :) > > http://news.bbc.co.uk/1/hi/technology/2270648.stm > > viva the internet! >Another Baltimorean, >Bhuvanesh > > great! did you go to Townson University? Yep! I was there for about six years. My physics classes were the most fun. > a physics teacher of mine lectures there too: his name is Dr. Harry > Olson. Do you know him by any chance? No. Did he join recently? I moved to Champaign, IL, about two years back. > BTW the initial courses from MIT are: > > Anthropology > Biology > Chemical Engineering > Chemistry > Civil and Environmental Engineering > Earth, Atmospheric, and Planetary Sciences > Economics > Electrical Engineering and Computer Science > Linguistics and Philosophy > Management > Mathematics > Mechanical Engineering > Ocean Engineering > Political Science > Urban Studies and Planning Argh, no physics?? :-( Bhuvanesh. > No. Did he join recently? i don't know, 2 years maybe? > BTW the initial courses from MIT are: > > Anthropology > Biology > Chemical Engineering > Chemistry > Civil and Environmental Engineering > Earth, Atmospheric, and Planetary Sciences > Economics > Electrical Engineering and Computer Science > Linguistics and Philosophy > Management > Mathematics > Mechanical Engineering > Ocean Engineering > Political Science > Urban Studies and Planning > > Argh, no physics?? :-( they may have made a typo. i read down the page that they have a physics pilot course already made. Courses were selected from the following departments and groups for the pilot phase: Biology Electrical Engineering and Computer Science Linguistics and Philosophy Ocean Engineering Physics Sloan School of Management Speech Communications Urban Studies and Planning so, you no longer live in Baltimore? > Bhuvanesh. > they may have made a typo. i read down the page that they have a > physics pilot course already made. Oh, good :-) > so, you no longer live in Baltimore? No, I'm now at Wolfram Research (www.wolfram.com). -- Bhuvanesh > No, I'm now at Wolfram Research (www.wolfram.com). great, congrats! hope i'll see one of these days a Bhuvanesh-Research.com or a Bhuvafram-research.com or a wolfvanesh-research.com or ... :) i've been reading S. Wolfram's book The New Kind of Science. interesting. he was in Greenbelt-MD the other day (where i got the book). good luck > on sept 30 you will be able to go to MIT for free!! well, almost :) > > http://news.bbc.co.uk/1/hi/technology/2270648.stm Another Baltimorean, Bhuvanesh. I just got my copy of the URROZ books. Like 'em so far, but..... One of the first examples (actually a psudo-example) is t^2-3t+5=0. so you enter this on the stack: 2: t^2-3t+5=0 1:'t' enter QUAD and I get QUAD Error: No soluntions found I have tried this several times and used the same key pattern out of the book. If you see what I am fouling up, I'd appreciate it. Matt > > I just got my copy of the URROZ books. Like 'em so far, but..... > > One of the first examples (actually a psudo-example) is t^2-3t+5=0. > so you enter this on the stack: > > 2: t^2-3t+5=0 > 1:'t' > enter QUAD > > and I get QUAD Error: No soluntions found I'll bet you are trying to find complex solutions while in real mode. This does not work very well! If you are in real mode, your example above,t^2-3t+5=0, has no solutions, as you may easily verify by hand. When I am in ->complex mode<- and exact mode I get { 't=(3-i*sqrt(11))/2' 't=(3+-i*sqrt(11))/2'} and when in ->complex<- and approximate modes I get {'t=(1.5,-1.65831239158)''t=(1.5,1.65831239158)'} What more could anyone want of QUAD? > What more could anyone want of QUAD? That it was backward compatible with the HP48 series? If some commands are not, some are somewhat, and some are completely backward compatible, then backward compatibility should be ditched alltogether. I disagree. While it was not entirely clear or intuitive to get this particular problem to work on the 49G, I blame myself for not understanding how the new 49G works as opposed to the 48GX, which I have. I'm still trying to kick my old 48GX habits to understand everything about the 49G, not an easy task, but at the same time I realize that most of the mistakes I'm making are my own fault, not the calculators. Doug > What more could anyone want of QUAD? > That it was backward compatible with the HP48 series? > If some commands are not, some are somewhat, and some are completely > backward compatible, then backward compatibility should be ditched > alltogether. > I disagree. While it was not entirely clear or intuitive to get this > particular problem to work on the 49G, I blame myself for not understanding > how the new 49G works as opposed to the 48GX, which I have. QUAD does not work like on the HP48GX. Period. That was the intent - why keep it otherwise? > I'm still > trying to kick my old 48GX habits to understand everything about the 49G, > not an easy task, but at the same time I realize that most of the mistakes > I'm making are my own fault, not the calculators. Yeah, right. That's non-intuitive for ya. Regrads I think you're missing the point. Either way, it can be done on both calculators. > I disagree. While it was not entirely clear or intuitive to get this > particular problem to work on the 49G, I blame myself for not > understanding > how the new 49G works as opposed to the 48GX, which I have. > QUAD does not work like on the HP48GX. Period. That was the intent - why > keep it otherwise? > I'm still > trying to kick my old 48GX habits to understand everything about the 49G, > not an easy task, but at the same time I realize that most of the mistakes > I'm making are my own fault, not the calculators. > Yeah, right. That's non-intuitive for ya. > Regrads > I think you're missing the point. No, *you're* missing the point. I know it can be done - I know it can be done a zillion ways on the HP49G, but that was not the point. The point is that QUAD does not behave like on the HP48G series, which was the intent of the former HP49G programmers. That's the point, and with all these commands trying to emulate the HP48G, they simply shouldn't, as half-emulation is much much worse than no emulation at all. I know QUAD can solve a quadratic, it simply does not return the same result as on the HP48G. I don't care if the output is mathematically correct, it's not the same as on the HP48G. That's the sole purpose of QUAD, to deliver the exact same result as on the HP48G - not to imitate the HP49G SOLVE command. That's my point. > I think you're missing the point. > > No, *you're* missing the point. I know it can be done - I know it can be > done a zillion ways on the HP49G, but that was not the point. The point is > that QUAD does not behave like on the HP48G series, which was the intent of > the former HP49G programmers. That's the point, and with all these commands > trying to emulate the HP48G, they simply shouldn't, as half-emulation is > much much worse than no emulation at all. > > I know QUAD can solve a quadratic, it simply does not return the same result > as on the HP48G. I don't care if the output is mathematically correct, it's > not the same as on the HP48G. That's the sole purpose of QUAD, to deliver > the exact same result as on the HP48G - not to imitate the HP49G SOLVE > command. > > That's my point. > same *names* like on the HP48 but behave, well, almost(?) like on the HP48. It is superfluous and inellegant to have them when they don't guarantee backwards compatibility but are just side offsprings of other commands. In this particular case, SOLVE would be completelly sufficient. Nick. ISOL is another command which could have been very useful if left as it was on the HP48; since the new 49G SOLVE command seems to include the ISOL function, I don't see why there was any need to modify the original ISOL command, but it was modified anyway, and old programs which used ISOL to rearrange equations and solve quickly may therefore not be portable to the 49G (one example was my triangle solver TRI1, which I did finally port, but runs at only 1/16 the original speed, solely because of what happened to ISOL). The multiple equation solver (MES), which is really only a *numeric* solver, also got changed into the CAS category some while back, which caused it to fail in many cases (especially for the entire original HP48G equation library), but in 1.19-6 the MES was changed back and now works again. I think that there could have been far greater compatibility if certain design elements (for ALG mode) hadn't been adopted, and if various original 48G commands hadn't been changed to CAS commands when there was no need to, because the CAS, in most cases, has its own new commands anyway. There was also a tendency to introduce functions which were not careful to affect only variables in the current directory, and leave higher directories alone; there still exists a significant flaw in that the CAS can't function if some variable name exists in memory (that's why it asks you to delete it, and won't proceed if you don't), and if you have that same name existing in two directories (say both HOME and another directory that you're currently using), then these commands can't work at all. The CAS also assumes that every CAS command uses VX, but they don't (some use a variable named on level 1 instead, and others don't need any variable to be deleted), which leads to other problems like that above. ACO must have rushed this along too fast for a complete and thorough job to be done, and now that they have also killed the project entirely, it's unlikely that these things will ever be straightened out. So the 49G does a lot, but it wasn't as well executed as HP projects used to be in all past history. I wonder whether this finds a parallel or not in regard to the rest of HP's products? ((o)) . > ACO must have rushed this along too fast for a complete > and thorough job to be done, and now that they have > also killed the project entirely, it's unlikely > that these things will ever be straightened out. Exactly. > I know QUAD can solve a quadratic, it simply does not return the same result > as on the HP48G. I don't care if the output is mathematically correct, it's > not the same as on the HP48G. That's the sole purpose of QUAD, to deliver > the exact same result as on the HP48G - not to imitate the HP49G SOLVE > command. Then go back to the 48. For my part, vive le difference! > Then go back to the 48. For my part, vive le difference! I don't give a flying f*** about the '48 (and soon neither the '49 for that matter). The point is - and read this carefully - there is *no* need for QUAD on the '49, was it not for backward compatibility. That purpose does QUAD not serve, hence it's buggy! Buggy, like so many other commands on the '49 - they'll never get fixed, and so what? My *point* is that QUAD does not behave like it should. If you don't get my point, fine. Don't tell me which calculator to to use - that decission is long since made. While I am not crazy about learning new stuff with the 49g (I own a 48gx as well) I do not blame the calculator in this case. I blame the book. This is a typo, plain and simple. On the calc front, it might be nice if the calc popped up a window which said Hey, can I switch to Complex mode? instead of just dumping. The operator should have been smart enough to figure that out also, but the operator was up till 2:30 studying, which was dumb. I'll remember that next time I feel like a marathon 49g lesson. Matt > I blame the book. This is a typo, plain and simple. Or the book was written for another ROM version? You wouldn't believe how much this type of behaviour has changed throughout the last 30 or so ROM revisions. > On the calc front, it might be nice if the calc popped up a window which > said Hey, can I switch to Complex mode? instead of just dumping. If your expression was 't^2-3*t+i' instead of 't^2-3*t+5', it would. I don't think it should even then (if QUAD imitates SOLVE), since one has definetely stated real mode. On the other hand, I don't think that QUAD should imitate SOLVE - it should function exactly as on the HP48G series, and return the exact same result, or not exist on the HP49G at all. It's an obvious error/typo. If you try to do it manually you'll see that it indeed does not factor down. Doug > I just got my copy of the URROZ books. Like 'em so far, but..... > One of the first examples (actually a psudo-example) is t^2-3t+5=0. > so you enter this on the stack: > 2: t^2-3t+5=0 > 1:'t' > enter QUAD > and I get QUAD Error: No soluntions found > I have tried this several times and used the same key pattern out of the > book. If you see what I am fouling up, I'd appreciate it. > Matt > It's an obvious error/typo. If you try to do it manually you'll see that it > indeed does not factor down. And? It does have two numeric solutions (as all quadratics have). It works fine on the HP48GX (returning 't=(3+s1*(0,3.31662479036))/2'), so this is another incompatibility on the HP49G. Why expect it to work anyway? I've stopped to. I can get the same answer on the 49G. I apologize for not being more clear. What I should have said was that it wasn't a straight forward problem like normal quadratics in that you can factorialize the equation out to get the factors of 't'. You are right in that all quadratics have solutions, sorry for not being clear. In order to get the same thing on the 49G, you have to make sure the calculator is in complex mode and make sure flags 2 and 3 are clear. My machine was also in exact mode. The answer I got was: { 't=(3-i*SQRT(11))/2' 't=(3+i*SQRT(11))/2' } Doug > It's an obvious error/typo. If you try to do it manually you'll see that > it > indeed does not factor down. > And? It does have two numeric solutions (as all quadratics have). It works > fine on the HP48GX (returning 't=(3+s1*(0,3.31662479036))/2'), so this is > another incompatibility on the HP49G. > Why expect it to work anyway? I've stopped to. I'm a dufus. Last night when I looked at it I didn't even think to do it myself. Thats why I bought such a great calc! :) Matt > It's an obvious error/typo. If you try to do it manually you'll see that it > indeed does not factor down. > > Doug > > >> >>I just got my copy of the URROZ books. Like 'em so far, but..... >> >>One of the first examples (actually a psudo-example) is t^2-3t+5=0. >>so you enter this on the stack: >> >>2: t^2-3t+5=0 >>1:'t' >>enter QUAD >> >>and I get QUAD Error: No soluntions found >> >>I have tried this several times and used the same key pattern out of the >>book. If you see what I am fouling up, I'd appreciate it. >> >>Matt >> > > > I had an example in calc where we had to do some regression. No big deal right? Wrong, I keep getting an error back from the Calc. I hit right-shift and STAT, then go to fit data. I enter the data by going to EDAT and hit edit. I enter into the matrix editor the following: 0,540 50,390 100,390 150,430 200,425 250,360 300,275 350,125 400,0 And there it sits in EDAT. Then I used Exponential FIT, Power FIT, and BEST FIT. With Expo. AND Power FIT I get the following error: ERROR: Invalid E(summation) Data LN(0) AND Best Fit gives me a linear formula. The answer is a quartic (which I forgot to copy off the board!!!) This is disturbing as the TI83 had no problems with it. I would assume that either with additional software or the default stuff I could get the right answer. Matt > > I had an example in calc where we had to do some regression. No big deal > right? Wrong, I keep getting an error back from the Calc. > > I hit right-shift and STAT, then go to fit data. > > I enter the data by going to EDAT and hit edit. > > I enter into the matrix editor the following: > 0,540 > 50,390 > 100,390 > 150,430 > 200,425 > 250,360 > 300,275 > 350,125 > 400,0 > > And there it sits in EDAT. > > > Then I used Exponential FIT, Power FIT, and BEST FIT. > With Expo. AND Power FIT I get the following error: > ERROR: Invalid E(summation) Data LN(0) If any of the x values are zero, the LOG fit wont work because it requires calculation of LOG(0), similarly if any y values are zero the EXP fit reqires calculation of LOG(0). > > AND Best Fit gives me a linear formula. > > The answer is a quartic (which I forgot to copy off the board!!!) > > This is disturbing as the TI83 had no problems with it. I would assume > that either with additional software or the default stuff I could get > the right answer. > > Matt > There are a variety of polynomial-fitting libraries available at http://www.hpcalc.org/ FITn is probably the simplest of them. > 0,540 >......... > 400,0 You can't use neither log fit nor power fit if x=0, as in you first data. The y=0 in the last one, makes you can't use neither exp nor power (again) fits. Clear these data and you'll be able... Hope this helps Does anyone have ny experience of using these AAA NiMH rechareable batteries in an HP48. The adertising blurb says they last twice as long as alkaline but what about their voltage? I am using Rechargable Energizer they are pretty good, and yes you have to keep a spare set because the low battery sign last a few minutes. Marturo Ubi Dubidum Ibi Libertas > Does anyone have ny experience of using these AAA NiMH rechareable batteries > in an HP48. The adertising blurb says they last twice as long as alkaline > but what about their voltage? I've used them in the 48 and in other devices. It's not true that they have twice the life on a charge as alkaline. It's more like 2/3 the life. Alkaline holds a charge quite a bit longer. Of course NiMH can be recharged. The reason I didn't keep using them is there are times I don't use my calculator for a month and NiMH self discharges. They're gone in a couple of weeks whether you use them or not. The problem with not getting enough notice when they're low isn't as severe as people make it sound. I've gotten warnings when I was away from the house and couldn't change the battery till I got home. I didn't use it as much till I got home later but I did use it some. It wasn't a problem. Barry > > Does anyone have ny experience of using these AAA NiMH rechareable batteries > in an HP48. The adertising blurb says they last twice as long as alkaline > but what about their voltage? They are great, the lower voltage doesn't affect the calc. -- This message was written with 100% recycled electrons Pivo > > Does anyone have ny experience of using these AAA NiMH rechareable batteries > in an HP48. The adertising blurb says they last twice as long as alkaline > but what about their voltage? > > They are great, the lower voltage doesn't affect the calc. read this: http://www.greenbatteries.com/documents/Battery_FAQ.htm#substitue It explains why - the lower voltage is not a problem - we have less time after the 'low battery' sign comes on the rest of the FAQ (and Myths) is highly recommended, too. Werner Except that a very careful user should have a spare set available, since there is only one minute or so left when the battery annuciator goes on. The alkalines have a straight line voltage curve and the NiCd/NiMH have a flatcurve with a quick drop at the end. Be careful out there! This message was written with 100% new positrons and it will self-annihilate in a picosecond :) > > Does anyone have ny experience of using these AAA NiMH rechareable batteries > in an HP48. The adertising blurb says they last twice as long as alkaline > but what about their voltage? > They are great, the lower voltage doesn't affect the calc. > -- > This message was written with 100% recycled electrons > Pivo Is there a way to find the one-sided limit of a function? Our math class wants us to graph each equation and detemerine it my looking at it, but there must be a quicker way, as it becomes tedious to wait for the graph to finish. By a one-sided limit, I mean finding the limit of a function as you approach from the left or the right. --Andrew > Is there a way to find the one-sided limit of a function? Our math class > wants us to graph each equation and detemerine it my looking at it, but > there must be a quicker way, as it becomes tedious to wait for the graph to > finish. Take '1/X' as an example (try it out). '1/X' 'X=0' LIMIT -> unsigned inf '1/X' 'X=0+0' LIMIT -> +inf '1/X' 'X=0-0' LIMIT -> -inf Does the above make sense on its own? If not, please write again. > Take '1/X' as an example (try it out). > > '1/X' 'X=0' LIMIT -> unsigned inf > '1/X' 'X=0+0' LIMIT -> +inf > '1/X' 'X=0-0' LIMIT -> -inf > > Does the above make sense on its own? If not, please write again. Yes, it makes sense to me. However, when I try out the examples, I still get an unsigned infinity symbol each time. Are there certain CAS settings with must be set? > Yes, it makes sense to me. However, when I try out the examples, I still get > an unsigned infinity symbol each time. Are there certain CAS settings with > must be set? Clear flag -120 (Silent mode off) Clear flag -106 (Simp. in series) > Clear flag -120 (Silent mode off) > Clear flag -106 (Simp. in series) --Andrew Huey By the way, I realise I can use the table of values, but that is also quite slow. I'm not sure I worded my question properly. By a one-sided limit, I mean: as you approach a point from each side, you are approaching different points, and a limit does not exist. Is there a way to find the point you are approaching from each side? In case anyone has a link to my HP38/39/40G site (The HP HOME view) please note that I now have a new address of http://www.hphomeview.com/ with a new look too. It's still under construction so if you notice any broken links etc please let me know. The content is pretty much the same as before, just organised a bit differently and with some outdated material removed. I am a newbie to writing programs and I am trying to write a program for both the Hp49 and the Ti89, and I need some help. I have looked all over the net and can't find an answer to my question. Here it is: I have a list stored in variable 'l1'as well as a list in 'l2', 'l3', and 'l4'. How do write the program to randomly pick a value out of list 'l1'? How about list 'l2', etc.? I have tried rand(), but that doesn't work. Any ideas, or am I going about it all wrong? EG (89): :{1,2,3,4}->l1 :{House, car, boat, bike}->l2 :{CO,AZ,UT,NY}->l3 :{Amy, Scott, Mike, Susan}->l4 :Disp ??????? > I am a newbie to writing programs and I am trying to write a program > for both the Hp49 and the Ti89, and I need some help. I have looked > all over the net and can't find an answer to my question. Here it is: > I have a list stored in variable 'l1'as well as a list in 'l2', 'l3', > and 'l4'. How do write the program to randomly pick a value out of > list 'l1'? How about list 'l2', etc.? I have tried rand(), but that > doesn't work. Any ideas, or am I going about it all wrong? > > EG (89): > > :{1,2,3,4}->l1 > :{House, car, boat, bike}->l2 > :{CO,AZ,UT,NY}->l3 > :{Amy, Scott, Mike, Susan}->l4 > :Disp ??????? > Plain user RPL : In order to get any element out of a list : 1: yourlist --> 1: random element out of yourlist << DUP SIZE RAND * CEIL GET >> store this and run it... That's a start, it's not very clear what end result you want to achieve with the other lists. G.8erard > I am a newbie to writing programs and I am trying to write a program > for both the Hp49 and the Ti89, and I need some help. I have looked > all over the net and can't find an answer to my question. Here it is: > I have a list stored in variable 'l1'as well as a list in 'l2', 'l3', > and 'l4'. How do write the program to randomly pick a value out of > list 'l1'? How about list 'l2', etc.? I have tried rand(), but that > doesn't work. Any ideas, or am I going about it all wrong? > > EG (89): > > :{1,2,3,4}->l1 > :{House, car, boat, bike}->l2 > :{CO,AZ,UT,NY}->l3 > :{Amy, Scott, Mike, Susan}->l4 > :Disp ??????? > > > Plain user RPL : > > In order to get any element out of a list : > > 1: yourlist --> 1: random element out of yourlist > > << DUP SIZE RAND * CEIL GET >> > > store this and run it... > > That's a start, it's not very clear what end result you > want to achieve with the other lists. > > > G.8erard appreciate it! Have a nice day! :o) Aaron > I am a newbie to writing programs and I am trying to write a program > for both the Hp49 and the Ti89, and I need some help. I have looked > all over the net and can't find an answer to my question. Here it is: > I have a list stored in variable 'l1'as well as a list in 'l2', 'l3', > and 'l4'. How do write the program to randomly pick a value out of > list 'l1'? How about list 'l2', etc.? I have tried rand(), but that > doesn't work. Any ideas, or am I going about it all wrong? > > EG (89): > > :{1,2,3,4}->l1 > :{House, car, boat, bike}->l2 > :{CO,AZ,UT,NY}->l3 > :{Amy, Scott, Mike, Susan}->l4 > :Disp ??????? > How about l1[rand(dim(l1))] ? (Same for l2 and l3.) -- Bhuvanesh It really was just about perfect for programming with the extra alphabetic keys. What made them go to a traditional form-factor for the 48? Britt I have a 28S as well as a 48GX and I prefer the single keyboard. The 28S is nice in some ways, but is a pain in the butt to hold. I think th 28S was and is a great calculator though! > It really was just about perfect for programming with the extra > alphabetic keys. What made them go to a traditional form-factor for the > 48? > Britt > The 28S is a pain in the butt to hold. With its double-hinged case (same as 18C/28C/19B[ii]), you can open it 360 degrees, in which case it looks (and holds) like other vertical calculators, except that the alpha keys (and any associated functions on the HP28 models) are on the back side, so that you have to flip it over, more or less, to use those (but the display and all the numeric keys and ENTER are in front). As to a computer interface, mentioned in another post, this clamshell series has only IR *HP*printer* output (not readable by any PC unless some HP palmtop can read it), and of course no input but the keyboard, so there is no computer interface, hence no programming other than by manually typable user command programs. ((o)) . > > It really was just about perfect for programming with the extra > alphabetic keys. What made them go to a traditional form-factor for the > 48? I had a 28S and it was good for keying in...but I couldn't get software for it. HP told me they had only released the 28s to give programmers a worktool but that they themselves didn't write software, and to check Grapevine Publishing. But when I did, they said, We don't have any programs for the 28S, but if you write one, we'll sell it for you. and intersection program, and then the 48 came out (and SMI already had a good cogo prog) so I returned the 28 and purchased the one I still own. I'm not sure but I don't remember an interface on the 28. /ramalane -- Offical Alt.Hackers.Malicious Survival Guide, FAQ, and Who's Who (Or as rikijo sez, Who dat iz?) http://www.ramalane.com when it comes to an interface there were actually a possibility to connect the 28S to the PC transferring files (programs) forth and back between PC and calc. As the 28S was still top of the product line a german guy named Frederich Schr.9ader (at that time living in Denmark) developed a cable connection to the PC. Several years ago I contacted him because I wanted him to modify one of my 28s but (I think he told me) it would be too much work starting this project again... and unfortunately I haven't ever seen such a modified 28 for sale on ebay... which means I am still waiting...:-) Martin > >>It really was just about perfect for programming with the extra >>alphabetic keys. What made them go to a traditional form-factor for the >>48? > > > I had a 28S and it was good for keying in...but I couldn't get software > for it. HP told me they had only released the 28s to give programmers a > worktool but that they themselves didn't write software, and to check > Grapevine Publishing. But when I did, they said, We don't have any > programs for the 28S, but if you write one, we'll sell it for you. > and intersection program, and then the 48 came out (and SMI already had > a good cogo prog) so I returned the 28 and purchased the one I still > own. > I'm not sure but I don't remember an interface on the 28. > > /ramalane > ...they DRAG ME BACK IN! :o) I just blew the dust off of my 48SX and 48GX calculators. I use to use the hell out of them back in my land surveying days so I was familiar with some surveying software that converted the 48 into a data collector. Now I'm in a different field and I need a portable data collector which can d/L to a PC (Stanley Trent's SMI program comes to mind) ...but now I need a customizable program with this feature: It must be able to print a paper receipt in the field. And I doubt they even still sell the old infrared-link thermal printers (and even if they didI'd think that would be a rather cumbersome solution). Does anybody have any ideas? BTW, if you have any experience with Portable DataLab I'd welcome your comments as well. /ramalane -- Offical Alt.Hackers.Malicious Survival Guide, FAQ, and Who's Who (Or as rikijo sez, Who dat iz?) http://www.ramalane.com > ...they DRAG ME BACK IN! :o) > > I just blew the dust off of my 48SX and 48GX calculators. > I use to use the hell out of them back in my land surveying > days so I was familiar with some surveying software that > converted the 48 into a data collector. Now I'm in a > different field and I need a portable data collector which can The hp 82240B printers have been going for $60ish on ebay lately. and I don't see them as cumbersome at all. I've got a 48sx, manual, and printer all carried in a small (like, the old small 386 era) notebook pc envelope stuffed with foam cutouts to give protection. total weight around a kilo, size of a 3 ring binder for A4 paper. Actually, I've added an hp95lx to the case lately, and theres' room for batteries. > d/L to a PC (Stanley Trent's SMI program comes to mind) > ...but now I need a customizable program with this feature: > > It must be able to print a paper receipt in the field. And I doubt > they even still sell the old infrared-link thermal printers (and even > if they didI'd think that would be a rather cumbersome solution). > > Does anybody have any ideas? > > > BTW, if you have any experience with Portable DataLab I'd > welcome your comments as well. > > /ramalane > -- while E <> This is not exactly a 48/49 question, but I was wondering if any of the HP collectors out there know when HP changed the 17Bii's packaging from a box to the plastic bubble pack? Also, based on the serial number or other info can one tell when the unit was manufactured. Steve