C2 ==== when attempting to simplify '4^(x-1)/3^(1-x)' i was unable to get the response of (4/3)^(x-1) that i needed. it was a step for solving a sumation to break it down to a familiar geometric series form. the hp gave me 3e^(2x*ln(2))/(4e^(x*ln(3)) which i figure probably results from the x in the exponent, but i do not know how or why that is what the expression is changed to. i tried a variety of algebraic commands to manipulate it to something different, but never managed to remove the e^ln(x) setup within it. my ti89 returns 3(4/3)^x/4, which i am guessing is supposed to just be a kind of rationalizing the denominator, and although actually appearing closer to what i wanted, it is not what i wanted. is there a flag or command that i missed that could help? i was looking for a form that would help me match to a*r^n which in this case, i already had the a and needed to get it in the form of (4/3)^(x-1) so that r would be 4/3 and x-1 would be n, which is going to infinity in the series. thanks, Ed Sutton ==== > when attempting to simplify '4^(x-1)/3^(1-x)' i was unable to get the > response of (4/3)^(x-1) that i needed. They're not the same - try graphing or calculating a few points. The exponents are not the same either so you can't use normal index rules on them (I think that's what you were trying to do??). You can quite easily rearrange it so that y=e^(x(ln3 + 2ln2))/12 This simplifies to .0833333(12)^x If you enter the original expression onto the stack and press evaluate twice you'll get .083333(4)^x(3)^x which simplifies to the above... >the hp > gave me 3e^(2x*ln(2))/(4e^(x*ln(3)) which i figure probably results > from the x in the exponent, but i do not know how or why that is what > the expression is changed to. That's not right either... i tried a variety of algebraic commands > to manipulate it to something different, but never managed to remove > the e^ln(x) This is just x... >my ti89 returns 3(4/3)^x/4, which i am > guessing is supposed to just be a kind of rationalizing the > denominator, and although actually appearing closer to what i wanted, > it is not what i wanted. Nor is this right... I think you rely too much on algebraic calculators :) Here is how I got my answer (which matches that of my HP) Let y = 4^(x-1)/3^(1-x) ln(y)=ln(4^(x-1)/3^(1-x)) =ln(4^(x-1)) - ln(3^(1-x)) =(x-1)ln(2^2) - (1-x) ln3 =2(x-1)ln2 - (1-x)ln3 now e(lny)=e(2(x-1)ln2 - (1-x)ln3) ....(sorry I'm too tired to type the rest) You'll get y = .083333(12)^x Hope this helps?? Brody ==== > when attempting to simplify '4^(x-1)/3^(1-x)' <<...> my ti89 returns 3(4/3)^x/4 How did you get this? I get 12^x/12 when I type in your expression and press [Enter]. -- ==== > -Are the keyboard/screen protector problems that bad? Have you ever used a HP48? If you haven't, then the keyboard, screen cover, and hard plastic shell are absolutely fine. I can't complain at all, even though I own an ID93 calc with the extra stiff keys and extra rainbow effects. I have access to and have played with the 48, so I will acknowledge that the 49 could be improved, but I love my calculator and wouldn't dream of switching to the 48. > -How well have your HP 49 aged so far? I don't know for certain what you mean by 'aged', but I can say that I carry my HP49G around with me literally everywhere outside my home (stored in my front left pocket). Inside the home, it moves around from various locations, and since it is used as my alarm clock, it is the last and first things I touch when in bed. I've had the calculator slide out of my pocket several times when opening my car door and getting out, so it's hit asphalt several times. Damage wise, over the 2.3 years I've owned it, there are a couple of minor scratches in the plastic at one corner, two of the rubber feet on the because the design seals up well enough. The screen cover itself does not have any visible scratches at all. The keyboard has maintained its original feel (i.e. none of the keys require noticeably more force to press than the others, nor are any of them sticky). Electrical functionality wise, I can notice no damage whatsoever. Howard Henry Schlunder ==== I had a 48G, and I bought a 49G when the former's screen got crushed. Like other HP fans, I wasn't happy with a lot of things they did with the 49G. At one point, I thought I would just use the 48 and keep the 49 to the side. When I tried that, I realized the 49G can do a lot of stuff the 48 can't. One thing I really like about the 49 is it's ability to do symbolic matrix algebra. The CAS of the 49 also is more powerful. I use the 49G. Despite the things I don't like about it, it's an overall better calculator. Tom > Sorry for bringing up such a dangerous topic again. As the subject implies, i'm considering whether to buy a 48g* or and > 49g. Now, i just saw the rumours for a 49G+ (or whatever), so, > obviously, this changes a lot. Still, i'd be interested in everyone's opinion: -Are the keyboard/screen protector problems that bad? > -How well have your HP 49 aged so far? > etc. Money is an issue, but not that much; if i knew that the 49G was more > interesting than an 48G+ or X(upgraded, etc), i'd definitely go for > it. In fact, the only thing that has stopped me from reaching a > decision are the keyboard problem. A keyboard's tactile feedback is > extremely important to me (and so is a screen protector that isn't as > ridiculously fragile as some say it is). I'll try to use this as an > indicator of the quality of the upcoming 49G+. I am going to enter my 2nd year of CEGEP(sort of like 1st year of > college for non-quebecois), where i will take Calc III, Optic & > Astrophysics. Afterward, i have yet to know where i will go, but it > will probably be between med or engineering. Paul PS, for anyone looking at the headers, i'm in Cambridge, MA for the > summer - thus my posting through google :) ==== Well I got a 48GX and then the 49 about 3 months after (instead of RAM cards for the 48). Now I'll selling the 49 on eBay. I can't bring myself to use its keyboard after the 48 and I've got the 'made in China' 49 with the 'good' keyboard. I'd hate to see what the others are like. You can download emulators for your PC and test the functionality of both calculators to see how they will suite your studies. I don't think a fancy calculator is 'that' much of a help during exams. I never took the 49 into an exam, the keyboard is just too horrible. It is easier to use the 48 quickly even though the software is slower. For things outside exams you have Maple or Matchcad, etc... Ollie > Sorry for bringing up such a dangerous topic again. As the subject implies, i'm considering whether to buy a 48g* or and > 49g. Now, i just saw the rumours for a 49G+ (or whatever), so, > obviously, this changes a lot. Still, i'd be interested in everyone's opinion: -Are the keyboard/screen protector problems that bad? > -How well have your HP 49 aged so far? > etc. Money is an issue, but not that much; if i knew that the 49G was more > interesting than an 48G+ or X(upgraded, etc), i'd definitely go for > it. In fact, the only thing that has stopped me from reaching a > decision are the keyboard problem. A keyboard's tactile feedback is > extremely important to me (and so is a screen protector that isn't as > ridiculously fragile as some say it is). I'll try to use this as an > indicator of the quality of the upcoming 49G+. I am going to enter my 2nd year of CEGEP(sort of like 1st year of > college for non-quebecois), where i will take Calc III, Optic & > Astrophysics. Afterward, i have yet to know where i will go, but it > will probably be between med or engineering. PS, for anyone looking at the headers, i'm in Cambridge, MA for the > summer - thus my posting through google :) ==== > Well I got a 48GX and then the 49 about 3 months after (instead of RAM cards > for the 48). Now I'll selling the 49 on eBay. I can't bring myself to use > its keyboard after the 48 and I've got the 'made in China' 49 with the > 'good' keyboard. I'd hate to see what the others are like. You can download emulators for your PC and test the functionality of both > calculators to see how they will suite your studies. I don't think a fancy > calculator is 'that' much of a help during exams. I never took the 49 into > an exam, the keyboard is just too horrible. It is easier to use the 48 > quickly even though the software is slower. For things outside exams you have Maple or Matchcad, etc... No, in exams, i only use (and always will, it seems) pen, paper and my good ol' EL-506L (sharp) i've had since grade 6. Yes, i have Maple at home; in class, though? I just don't think i'll be bringing a laptop for a while. It seems the 49's keyboard is good enough, as long as i don't try the 48's superior one. Interesting. Still, does anyone know if the non-chinese ones are still sold? I guess I will have to wait a couple of weeks for the new 48G and 49. ==== >Sorry for bringing up such a dangerous topic again. As the subject implies, i'm considering whether to buy a 48g* or and >49g. Now, i just saw the rumours for a 49G+ (or whatever), so, >obviously, this changes a lot. Still, i'd be interested in everyone's opinion: > I own a 48GX and I wouldnt upgrade to a 49G. It does everything i need and the build is solid. I dont like the look and feel of the 49 at all - cheap and nasty. The 49 does do fancy stuff like symbolic matrices - So it depends how much sophisticated maths you plan on doing. Mark. -- Mark Richardson mDOTrichardsonATutasDOTeduDOTau Member of S.M.A.S.H. (Sarcastic Middle aged Atheists with a Sense of Humour) ----------------------------------------------------- ==== Still, i'd be interested in everyone's opinion: > Compared, the 48 has got some things better, ie keyboard. The 49 is a more powerfull calculator. Look at my post 7 July ( Take an objective stand ), that expresses my opinion. Download the emulator, rom etc. (I've you experience difficulty with then decide what to buy once the 49G+ and the 48GII is available. Kaspaas ==== from what ive learned, the hp48g is more of an engineering type calculator and a 49g is more of a calculator for a mathematician. im sure that real world apps of each could drift between those areas a bit, as different people have different needs and desires from clculations. ive got my complaints about the keyboard, but i generally am kinda centered over a key i go to press so i dont have the problem i usually hear people having with it. my complaint is that i noticed (mainly, and first, with the arrow keys) that i can press a key down so that it moves down to a position that 'feels' pressed down but is really still not making electrical contact. that problem goes away if i just make sure i give the keys a decent push since i am usually light on controls until i find myself in a spot where i have to press hard. if u apply a force to press down a key, keep it there if u want to hold the key and you should be fine. you will probably be irritated if you like to press the sides of keys a lot, but i would suggest trying one out in a store or from a friend before making a yes or no on that. i have a 'made in china' model with a CNxxxx serial number. i dont notice a rainbow on the screen, but a few minor surface scratches (like u see in the glare off of most wristwatches that arent brand new) on the screen protector because i bought mine used. they dont seem to block or distort my view of it; im just anal(istic about detail to a paranoid level) about some of those things. as for fragile in terms of breaking; dont try to take it off and your calculator would serve you quite well. my old teacher had a damaged 48(some variant) with a damaged screen because the soft case doesnt have hard protection and has had it get bumped to the on position too. i suggested to her that she save the case from the new ti89 i traded to her and cut out a section of plastic to placein her case, but only apparently suggested it as a screen protection rather than to protect the screen and keys all with a bigger piece. without additional software, the 49g is quite a nice thing to have for symbolic manipulation; im pretty sure the 48 calcs have symbolic math, but it may not be quite as powerful (until you go get solike erable and alg48 for it). ive heard there is a speed difference between the 48x and 49g, but havent had much of a chance to test it. the loss of the infrared port on the 49g saddens me, but i didnt have any need for it and i think i can make an external one if i really want anyways. the 49g comes with the flash memory which can have advantages, but it does not have room for expansion cards as a result. i havent gotten into the use of much memory dependent uses of my 49g yet, so memory limitation isnt something i have a good perspective of on here yet. my best suggestions would be to research them, get the emulators for free testing of what they are capable of, try out the real thing in person, and allow that to help you figure out what you (and your pocketbook, if you must) find best suits you for school, hobbies, career, etc. hope this helps and happy shopping. =) Ed Sutton > Sorry for bringing up such a dangerous topic again. As the subject implies, i'm considering whether to buy a 48g* or and > 49g. Now, i just saw the rumours for a 49G+ (or whatever), so, > obviously, this changes a lot. Still, i'd be interested in everyone's opinion: -Are the keyboard/screen protector problems that bad? > -How well have your HP 49 aged so far? > etc. Money is an issue, but not that much; if i knew that the 49G was more > interesting than an 48G+ or X(upgraded, etc), i'd definitely go for > it. In fact, the only thing that has stopped me from reaching a > decision are the keyboard problem. A keyboard's tactile feedback is > extremely important to me (and so is a screen protector that isn't as > ridiculously fragile as some say it is). I'll try to use this as an > indicator of the quality of the upcoming 49G+. I am going to enter my 2nd year of CEGEP(sort of like 1st year of > college for non-quebecois), where i will take Calc III, Optic & > Astrophysics. Afterward, i have yet to know where i will go, but it > will probably be between med or engineering. Paul PS, for anyone looking at the headers, i'm in Cambridge, MA for the > summer - thus my posting through google :) ==== It was simple to tell that I was getting a Chinese model - you can read the serial number on the back of the package. The Chinese model starts with a CN. It's funny because when I asked at the shop that I wanted to know the serial number they knew exactly what I was after and told me they only stocked later model calcs because of all the problems they had with earlier ones. I really hope that HP keep going with the 48 style hardware it is really good to use. The 49 is a damn shame as the software for it IS very good. Personaly I reckon HP should jam the brains of the 49 into a 48 and forget about the 'better' functionality that the new keyboard layout allegedly provides. Ollie > Well I got a 48GX and then the 49 about 3 months after (instead of RAM cards > for the 48). Now I'll selling the 49 on eBay. I can't bring myself to use > its keyboard after the 48 and I've got the 'made in China' 49 with the > 'good' keyboard. I'd hate to see what the others are like. > > You can download emulators for your PC and test the functionality of both > calculators to see how they will suite your studies. I don't think a fancy > calculator is 'that' much of a help during exams. I never took the 49 into > an exam, the keyboard is just too horrible. It is easier to use the 48 > quickly even though the software is slower. > > For things outside exams you have Maple or Matchcad, etc... > No, in exams, i only use (and always will, it seems) pen, paper and my > good ol' EL-506L (sharp) i've had since grade 6. Yes, i have Maple at > home; in class, though? I just don't think i'll be bringing a laptop > for a while. It seems the 49's keyboard is good enough, as long as i don't try the > 48's superior one. Interesting. Still, does anyone know if the > non-chinese ones are still sold? I guess I will have to wait a couple of weeks for the new 48G and 49. ==== > -Are the keyboard/screen protector problems that bad? > -How well have your HP 49 aged so far? > etc. Money is an issue, but not that much; if i knew that the 49G was more > interesting than an 48G+ or X(upgraded, etc), i'd definitely go for > it. In fact, the only thing that has stopped me from reaching a > decision are the keyboard problem. A keyboard's tactile feedback is > extremely important to me (and so is a screen protector that isn't as > ridiculously fragile as some say it is). I'll try to use this as an > indicator of the quality of the upcoming 49G+. I am going to enter my 2nd year of CEGEP(sort of like 1st year of > college for non-quebecois), where i will take Calc III, Optic & > Astrophysics. Afterward, i have yet to know where i will go, but it > will probably be between med or engineering. > I wouldn't bother to wait for the 49G+. I feel that Hp will just take the standard 49G and replace the RS232 with USB, perhaps add more RAM (2 or 4 Meg, but it already has plenty, more of a marketing coup to bash Ti) and maybeee, just maybeee increase the clock speed to 6 MHz (not likely, but maybe). To do anything else would require a new CPU and then HP would have called it an Hp5#G something. I like my 48GX better than my 49G, but I will try to be fair in my critique of both. Neither is the perfect calculator (I prefer a pocket calculator and neither quite fills that niche). First you will not be making a mistake with either, just giving up some of the strengths of the one you did not choose. On paper the choice is obvious and simple: the 49G. But in the real world there are reasons to go with the 48G. While it is true that the 49G is faster for many operations, it isn't any faster for plain old number crunching (90% of your use and more like 95%+ after you are out of school). The speed shows when you use your calculator for symbolic operations (which you may have to do by hand in Math courses anyway). But the 48G series has plenty of software available via downloads. Why do I even suggest looking at the 48G? 1. Better keyboard feel and as an old timer, the enter key is in the RIGHT Place! 2. It has everything a physics major needs. And you could make good use of the equation library that is included in the 48, but not in the 49. Sure you can download this eq. library for the 49, but the units features are not yet integrated vs on the 48's eq. lib THEY ARE. Vector operations are good, and great vector programs are available (that are small, and handle symbolics). 3. One small + feature for the 48G (that Irks me that it isn't included in the 49G!): the 48G's clock has SECONDS shown vs the 49G only shows minutes (I know there are probably programs available, but you will have to find them to make your 49G work as an accurate timer in a lab). 4. IR port: I don't really consider this a great plus, but as a student you may, IF there is anyone else with a 48 to share programs with, then no cable is needed. However I will quickly offer the obvious reasons to choose a 49G: 1. Flash ROM (upgradeable to latest and greatest SW ver. whenever it is available) 2. CAS 3. Memory (1.5 Meg vs 128K and is cheap compared to upgrading a 48 to 1.25 Meg) 4. Is a damn tough calculator (with both a LCD screen cover and a slide on cover) This LCD cover may scratch, but can take abuse vs a bare LCD would break, in fact that is one of the 48's weaknesses. 5. Keyboard layout 6. Comes with a free cable, if it doesn't, YOU DON'T WANT IT! Older stock calculator's came with poor LCD screen cover (rainbow effect) and supposed poorer keyboard and may not work with certain RS 232 ports. If you haven't used an Hp earlier, you probably won't complain about the newer keyboard ie you won't be spoiled. Many experts like the layout better and I am not in a position to argue with them. I do suggest the Hp's over a comparable Ti-89. They are simply better. The Ti-89 does have a few pluses over the Hp's: Speed and nice canned answers to textbook type problems. But the available sophisticated programs that are available and more appropriate for your field are much better with the Hp's. In numerical matrix calculations the Hp's usually perform better that the Ti. The Hp's also come with more functions than the Ti (even the older 48G series still has more functions sans CAS). The 49G shames the Ti for features and functions (though to be fair, the Ti's can be downloaded and if you bought a 48G series, you would also download many as well, ie Laplace transforms for example). Another nice feature the Hp's both have is Step-by-Step solutions. Not that it is always useful, but you can follow through to some solutions (and it can be slow). The techniques are not always the way you would solve by hand, but the solutions would hold up for a physics major (not likely for a math major though) to show solutions or proofs for a report or paper. I also like the way the Hp's handle different coordinate systems, but I can't say that the Ti-89 is better or worse (but You should find out, if you even consider the Ti-89). I also can read my Hp's screens better as you have more control over font size, but younger eyes may not care (and may even prefer the Ti). ==== Veli-Pekka Nousiainen :: >i've a function like this: 100*(1-i) >>________ >>(1+i)^3 The result must be -180¡ , but hp 49g give me a result of 180. This is not right because ph(100)=0, ph(1-i)=-45, ph(1+i)^3=45*3=135 but at >denominator it is -135 -45 -135 = -180 > That's grazy, Chris! I get -50 > Yes, it should be -50 :) By hand: (1+i)^3=1+3*i-3-i=-2*(1-i); 100*(1-i)/[-2*(1-i)]=-50 Robert Tiismus. ==== > Veli-Pekka Nousiainen :: > > i've a function like this: > 100*(1-i) >>> ________ >>> (1+i)^3 > The result must be -180¡ , but hp 49g give me a result of 180. > This is not right because ph(100)=0, ph(1-i)=-45, ph(1+i)^3=45*3=135 but >> at > denominator it is -135 > -45 -135 = -180 That's grazy, Chris! I get -50 Yes, it should be -50 :) By hand: (1+i)^3=1+3*i-3-i=-2*(1-i); > 100*(1-i)/[-2*(1-i)]=-50 Which is equal to complex number (-50, 0). The ***phase*** of this complex number is 180¡ modulo 360¡, thus 180¡, -180¡ (...) are solutions. ==== Hi, you could try on www.palmtop.net, THE place for the 200LX. Maybe there is the suitable driver. However, it would be much easier to transfer data via the serial line using a cable;-) Raymond > I am trying to transfer text files like the following (2 lines in this > case) from an HP 200LX palmtop to an HP 48SX calculator via their > infrared I/O ports. Both understand Kermit, and I would like to take > the shortest route possible to figuring out all required settings on > both machines -- 200LX Datacomm application: File, Protocol, Kermit; > Connect, Settings, Baud ?, Interface Infrared, Parity?, Data?, Stop?; > Connect, Kermit Commands, server mode?. 48SX I/O settings (many)?) %%HP: T(3)A(D)F(.); > << -> n 'IFTE(n <=1,n,FIB1(n-1)+FIB1(n-2))' > > Any help? Rick > ==== > [AC] OK mister Calvo, so you say there's not a HP49G+ on its way, or what is your story? The golden rule, when there's smoke, there's a fire. When the dealers and officials start talking about it independantly, around the world, then it need not be official to be true. Or I suppose the dealer lied to me. Or maybe different people around the globe had similar dreams. Or maybe you have got a better explanation?? It's on it's way - and you can bet your money it. Also I doubt if it will be any later than end of September. ==== As far as I am concerned, I won't deny that a 49G+ is coming. I'll just stand there, in complete denial ... I WANT A SUCCESSOR TO THE 41!!! Or at least to the 42! And I want to know more about the 17BII+ and the 19BII+! (if such things do exist, of course, because IBIWISI) > > Not Official yet > [AC] OK mister Calvo, so you say there's not a HP49G+ on its way, or what > is your story? The golden rule, when there's smoke, there's a fire. > When the dealers and officials start talking about it independantly, > around the world, then it need not be official to be true. Or I suppose the dealer lied to me. Or maybe different people around > the globe had similar dreams. Or maybe you have got a better > explanation?? It's on it's way - and you can bet your money it. Also I doubt if it > will be any later than end of September. ==== I already checked it out. But I only got a couple of addresses from it. The Nosy program that John told me works. Right now, I am only missng these forms: * CAS mode input form * DISP mode input form * prog choose box * math choose box * symb choose box * sslv choose box * calc choose box * alg choose box * arith choose box * cmplx choose box * exp&ln choose box * trig choose box * matrices choose box * convrt choose box * units choose box * base choose box Dave ==== I did NOT say or mean to imply that it was useless. *All* I asked was what it did that was unusual (ie beyond what was normally available on the calculator)? The documentation supplied with it (quoted in its entirety below) is very minimal. It could be the most fantastic program in creation but if there's no documentation then it's going to be pretty hard to use. > I will not > enumerate all these features, but HPPAINT allow a full screen editing, a > dot +/- options, a copy & past function... and it has been writing entirely > in ASM, so it's faster. WHY will you not enumerate all these features ? Surely that's why we do documentation? The only point I was making was that IN YOUR DOCS there is no mention of anything that is not already available. Dot +/-? Already there. Copy & paste? Already there. Full screen editing? Already there - all except for area covered by the command line at the bottom of the screen. What I'm asking (still) is why is yours better? I am not willing to download it onto my calculator and simply press keys at random in the hopes that something will happen. > HPPAINT V1.3 - PICT editor > -------------------------- > For the HP38G, HP39G & HP40G > 07/07/2003 Keys: > øøøøø > ENTER Commands menu > Cursor keys Move cursor > SHIFT + Cursor keys Move cursor to border > * Set mark > + Set pixel > - Clear pixel > DEL Cancel key in menus / magnification Some tips : > øøøøøøøø [OSl ash]øø > This programm can only edit the graphic of the plot view. > You can clear it with the ERASEPLOT command. > You can store this graphic in a variable with the ->PLOT command. > You can also recall a graph with the PLOT-> command. > Authors : > øøøøøøøø [OSl ash] > Original version by : Fran .8dois Darras (Dar) > Erwan Ancel (Burd) > Laurent Mallet (Ellis) > Translated to English by Fin > HP38 & HP39/40 versions by HP-Poska (hposka@free.fr) > (http://hposka.free.fr/hpcalc/) > >Could you tell me how this ap is different to the normal editing >>available via the built in functions? By this I mean, pressing >>ON+PLOT to capture the screen to G0 and then using the VARS menu to >>paste into the Sketch view. All the normal sketch tools then apply >>such as pixel editing as well as boxes, circles etc. What does yours >>offer that's extra to this process? > Colin, I think you could try it before to say it's useless. > Even if this app isn't basicly very different, it adds some little things > that make it more pleasant to use. I agree with you that the built in > sketch editor of the HP38/39/40 is sufficient to draw sketches (to > illustrate math problems for instance), but to draw more complex pictures, > (especially for games), this programm becomes essential. I will not > enumerate all these features, but HPPAINT allow a full screen editing, a > dot +/- options, a copy & past function... and it has been writing entirely > in ASM, so it's faster. >Just a couple of notes to HP-Poska: >>Try it with my ticking clock active: the hourglass starts blinking! :) > Yes, I've seen that, it's strange... > It's certainly due to the interruptions, I will correct that... > Done, you can now download the new version at the same adresse: > http://hposka.online.fr/hpcalc/hppaint3839.zip > In fact, I've just disallow completely the interruptions (INTOFF & INTON > replaced by GOSBVL =DisableIntr & GOSBVL =$jAllowIntr) it seems to work > fine now with the ticking clock active. >>There's no way for other aplets to use the modified grob. On the 48 >>version, it was left in GBUFF (PICT). I suggest you use the aplet's >>sketch view (LastBut0) or implement a Store-to-Grob option. > In this version too the modified grob is left in GBUFF, you can get access > to it with the PLOT commands, for instance just after to have work on a > picture, you can save it in G1 by doing PLOT-> G1, if you want to re-edit > it, do ->PLOT G1. > But if you want to improve the programm by adding a Store-to-Grob option, > I'll send you the source code. HP-Poska ==== Hi Anyone know why L2-> X*e^3X L1-> X^2-6X+9 LDEC gives infinite result error? This is a simple constant coefficient diff eq. ( HP49G - ROM #1.19-6) ==== Hell-0 Dave Did you mean undefined result ? You probably work in approximate mode. Try using exact mode and then you should get the right answer like I just did! !Demeter! ==== It appears that the problem is with reals/ZINTS. The LDEC command should only be used in exact mode, and if the coefficients are entered as ZINTS LDEC works ok. If any of the coefficients are entered as reals ie X^2*6.0X+9.0 the CAS asks for approx mode and LDEC fails. Can this be right? Dave H > Hi Anyone know why L2-> X*e^3X > L1-> X^2-6X+9 LDEC gives infinite result error? This is a simple constant coefficient diff eq. ( HP49G - ROM #1.19-6) ==== Sorry Demeter, didn't see your post until I'd written the 2nd. You are right it should be in exact mode. The trouble was I entered the coefficients as reals and so the CAS kept switching back to approx mode?!? Dave > Hell-0 Dave Did you mean undefined result ? > You probably work in approximate mode. > Try using exact mode and then you should get > the right answer like I just did! > !Demeter! ==== Sad to know there no new 42 remake... AFAIK it was better than 32S. Am I wrong? Besides that, I'm glad that we'll have new developments on the 4x series and eager to see the improvements this community suggested, in order to have another valid argument to keep posting. Carlos -- > Here's a link to an order form (Excel spreadsheet) that has all the > new HP calculators, including a 48GII and 39G+ (as well as the 33S, > 17BII+, 19BII+, and 49G+). No other technical details: http://www.dstewart.com/newdsc/c2es_03_orderbook/excel/Calculators_2003.XLS Matt ==== 48GII? When they also introduce a 49G+? Does it mean a 48-type model with decent case and keyboard and a 49-type model with crappy case and squishy keyboard? On the other hand, I'd like to know more about the 17BII+ and the 19BII+ ... > Here's a link to an order form (Excel spreadsheet) that has all the > new HP calculators, including a 48GII and 39G+ (as well as the 33S, > 17BII+, 19BII+, and 49G+). No other technical details: http://www.dstewart.com/newdsc/c2es_03_orderbook/excel/Calculators_2003.XLS Matt ==== So do I. On the other hand, I'd like to know more about the 17BII+ and the 19BII+ ... > ==== > Here's a link to an order form (Excel spreadsheet) that has all the > new HP calculators, including a 48GII and 39G+ (as well as the 33S, > 17BII+, 19BII+, and 49G+). No other technical details: Hmm... aren't you screwing up someone's relation with HP by posting this? Greetings Steve Sousa ==== i would need regular expressions for my HP49. is there a build in possability or any libary i can insall. till now i didn't found a way. greetings max ==== MK> i would need regular expressions for my HP49. is there a build in MK> possability or any libary i can insall. till now i didn't found a way. there are (as far as I know) two different implementations for regular expressions on the HP49. One is part of the Emacs library. It can be used in the Editor, for search and replace in the edited file. http://www.hpcalc.org/details.php?id=3940 The other implementation is part of Martin Langs MAWK library which implements many string operations, including a regexp engine. http://www.hpcalc.org/details.php?id=4254 For a comparison of the two engines, see my earlier post http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=qzuzo579mau. fsf%40sand.science.uva.nl The bug with expression grouping in MAWK mentioned in that post has by now been fixed in the mean time. Hope this helps - Carsten ==== thanks a lot. that helped! i will download MAWK. greetings max > he bug with expression grouping in MAWK mentioned in that post has > by now been fixed in the mean time. ==== hallo I need a textviewer for .doc and .txt files for my calculator HP48GX. This viewer dos't be change the form of the textoutput. Tanks Patrick ==== I know that this is a very sensitive question in this group but I try to ask as kind as possible. told us you solved this bug and that bug. This is really very kind of you especially because you do it in your free time. Will you ever release another (beta) version? Does HP let you release beta-versions? Perhaps after the new generation calcs are released? Perhaps you are again involved in the engineering process of the new generation calcs and the HP49+ or HP49II or what's its name and the ROM works with the old 49 too. Matthias Bunte -- mbunte@gmx.net ==== > Hi gurus, I wonder if there is any chance of buing the 4 pin end connector used > in the PC <-> HP-48 serial interface, and if so where and at what > price? I am thinking of a way to communicate faster between my calcs > SX and GX by wire instead of using IR. Thanx in advance > Matti, Sweden It's just a 4 pin, 2mm pitch connector. I used one from an old serial mouse as it had the DB9 on one end and the 2mm on the other. I some heat shrink tubing and used a short piece of a plastic cable tie as a strain relief. Look around at your local electronics suppliers for the connector. Jeff Birt ==== Is there anyone out there who still makes/sells 128K cards? I already have a 1Mb TDS card, and a 128K HP card, (82215a). I don't think that TDS makes any ==, I have a hp49G calc with serial CN125..., in this calc ALL INDICATORS ARE ALWAYS SET (LS RS ALPHA BATTERY BUSY I/O). I have Another with serial CN113... in which the LS RS and ALPHA INDICATORS ARE ALWAYS SET. Somebody knows How to clear them ? I tried a warm start, upgrade the calc to rom 1.19-6, but the indicators continue set. ==== I've the equation (s^2+1) . If i use solve function i obtain a infinite numerical result. >I would like to obtain (0,-1) (0,1) or +-i I've already activated the complex flag. Use Zeros Put the equation and the variable after you will to solve in the stack. 2: (s^2+1) 1: 's' Then press [s.slv] [Zeros]. The Result is (0,-1) and (0,1). When the complex Flag is set. ==== can i solve 2nd order Differential Equations with the built in progs?I think it's only for first one. > any program for solving differential equations, at least 2nd order. I own a > HP 49G. Do you mean besides the one that's built in? Check www.hpcalc.org. ==== > can i solve 2nd order Differential Equations with the built in progs?I think > it's only for first one. Let's say I want the solution to the following second-order ordinary differential equation: X*Y + 2*Y' + X = 1 First, I have to create 'Y' as a function of 'X'. I can't do this from the stack, apparently parenthesis have a definite purpose on the stack, and it is not functional notation. I have to create this in equation writer. When I was in DiffEq class, I had Y(X) stored in a variable so I wouldn't have to do this all the time. Anyway, open equation writer and type 'Y(X)' and hit enter. Hit enter again to make a copy of it. Hit 'DERVX' enter 'DERVX'. Now, on the stack, you should have: Y(X) d1Y(X) d1d1Y(X) Typing X * RightArrow 2 * X + + 1 = should give you the rest of the equation. Hit the right arrow (switching the differential equation and Y(X)) then hit 'DESOLVE'. This will give you: { 'Y(X)=-1/6*X^2+1/2X-cC0/X+cC1' } Which is the solution to the second order ordinary differential equation. This was a basic linear example, but your calculator will solve other kinds as well. I haven't found one yet that I could solve by hand, that the calculator couldn't do, though sometimes the calculator takes longer than me. A word of caution: Sometimes the HP's answers to differential equations are very difficult to read. There are all sorts of commands (FDISTR EXPAND FACTOR SIMPLIFY) you might have to use to make sense out of them. For instance, in the above example the calculator gave us -cC0/X as part of our solution. We know that in terms of arbitrary constants, -cC0=cC0, but the calculator won't make substitutions like that. You can also get the calculator to approximate the answer numerically, but that's usually not what the math teachers want until you get to that section. ==== I got mine at http://www.calcpro.com a few months ago. > Does anyone know where to purchase a copy of the HP 48G AUR? HP > indicates that it is no longer in print. > jzaring@attbi.com > >The HP 48G Series >Advanced User's Reference Manual >is available in print only. >No PDF or other elctronical document. >The User's Guide is available in www.hpcalc.org >> gr >> Is there an Advanced Users Guide for the HP49G or a reference for the >> equation library? >> GR >> ==== > Hi there I made the comments into my program on HP49G using @ characters. There were no problem, but when I run the program and wanted to correct that, I found out that the comments disappeared. Is there some way to have the comments, which will be in the body of program and will not dissapear? > Store the program *source* as a *string*, like << SIN @ some comment here > 1 + > Then use STR -> on this string to compile it. ==== Does anyone have a list of addresses used with LIBEVAL for calling an HP 49 application such as single var statistics inform menu or TVM solver inform menu? I can't find any list at hpcalc.org or by Dave ==== Is there an HP non programable calculator thats really good? Preferably RPN and similar to the HP49g (NONE Programable). ==== http://www.finseth.com/~fin/hpdata.html http://www.finseth.com/parts/hpdata.zip ==== > Is there an HP non programable calculator thats really good? > Preferably RPN and similar to the HP49g (NONE Programable). Why? (Please tell me because this is interesting!) I thought that the programmability of HP calcs was what made them really good. So I'd have to say no. ==== > Here's a conventional sys-rpl solution. Still falls 5 bytes short of > Werner's though. ( just 2 commands ;) sections, UsrRPL and SysRPL, is Werner Huysegoms with his programs << 6. ALOG / OVER MANT IP 2. SAME ROT OVER NEG DATE+ ROT + SWAP DATE+ >> (50 bytes) :: %6 %ALOG %/ SWAPDUP %MANTISSA %CEIL %2 %MOD DUP4UNROLL %CHS DATE+DAYS %+ SWAP DATE+DAYS ; (40 bytes) The small UsrRPL program belongs to the best ones I have ever seen. The Ceil -trick in the second program was proposed by Werner and myself independently. It can be replaced by the unsupported but stable HP49-PTR 3CCB4 which is the already mentioned :: EQUAL COERCEFLAG ; This pointer is very useful and should be included in TR's extable2. Any good name for it? ==== WR> It can be replaced by the unsupported but stable HP49-PTR 3CCB4 which is WR> the already mentioned :: EQUAL COERCEFLAG ; This pointer is very useful WR> and should be included in TR's extable2. Any good name for it? SAME, obviously. ==== Just use binary mode in kermit transfer to transfer just about anything over to HP49G that will automatically be enclosed in quotes just like a string; then use the SCROLL command. !Demeter! ==== > http://zon.astro.uva.nl/~dominik/hpcalc/emacs/extable2.bz This update is recommended for all users of Thomas' expanded extable. I think, I have an important hint for the compressed file. I made some bad experienced with the emulator, when having not sufficient memory before decompressing!! You should have at least 115 000 Bytes before decompressing. This means about 205 000 Byte, before loading extable2.bz puting into the stack and clear the global variable. The dangerous amount of free memory seem to be between 112000 and 114 600 Byte for decompressing. Below the lower value the BZ-Toggler from OT49 simply errors, so do not harm the memory, but in the dangerous region corruption of the flash seem to be possible? Heiko ==== Today Thomas has sent me a new EXTABLE which corrects this error. I have made it available on my website - an emacs update (which contains extable2.lib) will follow on hpcalc.org. http://zon.astro.uva.nl/~dominik/hpcalc/emacs/extable2.lib For slightly faster calculator upload, you can also take the BZ compressed library http://zon.astro.uva.nl/~dominik/hpcalc/emacs/extable2.bz This update is recommended for all users of Thomas' expanded extable. It seems that a few people have problems with the compressed version - I therefore no longer recommend using it and have removed it from the website. Just get the uncompressed library from http://zon.astro.uva.nl/~dominik/hpcalc/emacs/extable2.lib Make your home directory as empty as possible and transfer with xmodem. ******* ==== Just a little disclaimer about this program.. it only tests the data lines, not the address lines. Basically, it cycles through RAM and address line, it may think that it is writing to one address, but actually be writing to another due to the bad address line. The program won't see it when it reads that data back from RAM, because it will be grabbing it from the same (wrong) address in memory. Should you feel compelled to extend the test capability of the program, you might clear (write all 0's to) the memory, then write a 0xF to 0x00000. Then scan the rest of memory to verify there are no stray (binary) 1's at any other address (in the case that one address line is held high). Then clear everything again, and write a 0xF to 0xFFFFF, and again verify there are no stray 1's anywhere in the rest of RAM (in case one address line is held low). I think that should be an adequate memory address line test, but if anyone has any corrections, additions, etc. please add your input! > I re-controlled 1000 times all the components and all the wiring and it > seems to be all ok. > At this point my suspect is that the RAM chip is not appropriate: > partnumber > on the this chip is: Try http://www.hpcalc.org/hp48/utils/memory/ckport.zip alternate address: > http://www.exzodia.net/hp48/software/CkPort.zip and post results. ==== > I'm trying to understand the algorithm... It's very simple once you have > already read it > Basically you're multiplying your number by one in decimal mode. ==== >I'm trying to understand the algorithm... It's very simple once you have >>already read it ;-) > Basically you're multiplying your number by one in decimal mode. Oups ! Excuse my bad english expression ;-) I was trying to say that I have understood your program, but only after reading it more than a few times (maybe because it was late, so that i was a littele bit tired ;-) ==== > How can I use the command ChooseViewUI [...] I have seen the code > but I find it quite difficult to use the command in my own sRPL > aplets. That's because the choose box is part of my aplet's structure, and also because I did not comment the code :-) Here's a sample modeless choose box. It shows three options and the RUN softkey. The ENTER, DEL and SHIFT-DEL keys are enabled. I assume you are familiar with standard (modal) choose boxes :-) Binary ********** * Example of Modeless Full Screen Choose Box for the HP39/40G * ********** INCLUDE common39.h ASSEMBLE Dir_Head 10,Choose Box,601 RPL * This is to disable the SKETCH view EXTERNAL view7_CAPLET { topic_CAPLET view01_CAPLET view23_CAPLET view45_CAPLET xREADNOTE view7_CAPLET } INCLUDE share39.h ASSEMBLE Vfield L34,IntTable CON(5) =DOHSTR REL(5) endLink Link _Main Link _Dispatch Link _DoOption1 Link _DoOption2 Link _DoOption3 endLink RPL DEFINE Main INT_00 DEFINE Dispatch INT_01 DEFINE DoOption1 INT_02 DEFINE DoOption2 INT_03 DEFINE DoOption3 INT_04 ASSEMBLE Vfield IntTable,DirVar RPL ASSEMBLE CON(5) =DOHSTR REL(5) endCode RPL NAMELESS _Main :: { MINUSONE { { $ Start :: ZERO INT_00 ; } { $ About NotesViewUI } } StndVar StndReset StndEntry StndExit } BinLookup case EVAL DROP ' :: ' :: * Message handlers. Choose Box Size. TRUE: Full screen choose box SIXTY #=casedrop TrueTrue * Choose Key Ob. This is to enable the ENTER key. Otherwise, it would * be like the List or Matrix catalogs where the ENTER key is not * active. SIXTYEIGHT #=casedrop :: DUP ONE #=case ( Primary key ) :: SWAP { FIFTYONE ( ENTER key ) :: Dispatch ; ( If ENTER pressed, run Dispatch, see below ) BINT21 ( DEL key, to provide for del/reset ) :: TakeOver TOLVar8@ ( Returns the highlighted option as a bint ) LESetItem ( Takes a bint: returns that element from item list) CARCOMP $ Reset SWAP&$ ? &$ DoQueryBox ( Displays a NO/YES choose box: Reset Option x? ) DROP ( Does nothing ) ; } #=Lookup case :: SWAPDROP TrueTrue ; SWAP FALSE ; DUP TWO #=case ( Shifted Keys ) :: SWAP { BINT21 ( Shift Del ) :: TakeOver $ Reset all options? DoQueryBox DROP ; } #=Lookup case :: SWAPDROP TrueTrue ; SWAP FALSE ; FALSE ; * Choose Box Menu FIVETHREE #=casedrop :: ' :: NoExitAction { NullMenuKey NullMenuKey NullMenuKey NullMenuKey NullMenuKey { $ RUN :: Dispatch ; } } ; TRUE ; DROPFALSE ; $ Modeless Choose Box SEVENTEEN * Item list { { $ Option 1 DoOption1 } { $ Option 2 DoOption2 } { $ Option 3 DoOption3 } } ONE ChooseVEntry ; ' :: ChooseVExit DROP ; ChooseViewUI ; * Dispatch just EVALs the corresponding DoOption_x routine NAMELESS _Dispatch :: TakeOver TOLVar8@ LESetItem TWO NTHCOMPDROP EVAL ; NAMELESS _DoOption1 :: NOP ; NAMELESS _DoOption2 :: NOP ; NAMELESS _DoOption3 :: NOP ; ASSEMBLE endCode RPL ASSEMBLE Vfield DirVar,DirCode RPL { } ASSEMBLE Vfield DirCode,DirAlt RPL { } ASSEMBLE Vfield DirAlt,DirNot RPL $ Sample Aplet. By Jordi Hidalgo ASSEMBLE Vfield DirNot,DirEnd RPL { } ********** That's all. You may find this ill-formatted post somewhat helpful: http://groups.google.com/groups?selm=7fcf10a1.0303071129.5f61bf18%40posting. google.com ==== www.hpcalc.org. ==== > It was listed in the list of supported entry points Although the CAS commands are declared in the =head.h file, their addresses are not included in =Entry39.h (because of the telltale TESTEUROPE, I guess :-) ==== Does anyone know where to purchase a copy of the HP 48G AUR? HP indicates that it is no longer in print. >The HP 48G Series >Advanced User's Reference Manual >is available in print only. >No PDF or other elctronical document. >The User's Guide is available in www.hpcalc.org >> Is there an Advanced Users Guide for the HP49G or a reference for the >> equation library? Unless you don't want to use this one... http://www.contrib.andrew.cmu.edu/~drury/hp48gaur/ ==== You might begin with http://www.hpcalc.org/details.php?id=1716. > Since it is no longer in print, wouldn't it be nice to finally have it > available on-line? Can anyone ask for it? Toby > No PDF or other elctronical document. Unless you don't want to use this one... > http://www.contrib.andrew.cmu.edu/~drury/hp48gaur/ ==== I was wondering what people have to say about the Klotz ram cards, good or bad. Also anyone know if you can pick these up somewhere in North America? ==== I have bought a 2MB ram card from Klotz (1999). It's good.I have no problem with it. And don't forget to change the battery every year, are you lost the nibbles. Pthfndr A.R. a .8ecrit : > I was wondering what people have to say about the Klotz ram cards, good or bad. Also anyone know if you can pick these up somewhere in North America? ==== I made the comments into my program on HP49G using @ characters. There were no problem, but when I run the program and wanted to correct that, I found out that the comments disappeared. Is there some way to have the comments, which will be in the body of program and will not dissapear? Store the program *source* as a *string*, like << SIN @ some comment here 1 + > Then use STR -> on this string to compile it. ==== >> << DUP 2 MOD 2 INV == - CEIL >> This rounds to even, and works for negative integers. >>X >>Have you tested it on positive integers? How embarrassing! It's not the positive integers that are problematical. It's the >positive (and negative) numbers with fractional part OTHER than .5 that don't work right. >This isn't going to be easy to fix without adding bytes. Especially if it is to work with >both negative and positive numbers. I was so busy getting the right behavior for the special case of fractional parts of >exactly .5 and both positive and negative numbers that I forgot the easy stuff!! Well, after the weekend and some thought, here is the best I can do. Both of the routines SEEM to work properly with integers with fractional parts of .5 and otherwise. Both routines work for positive and negative numbers, and are 35 bytes long. Round to even: << 2 INV OVER 2 MOD OVER == - + FLOOR >> Round to odd: << -2 INV OVER -2 MOD OVER == + - FLOOR >> C'mon beta testers; break 'em or improve 'em!! ==== Well, after the weekend and some thought, here is the best I can do. Both of the routines > SEEM to work properly with integers with fractional parts of .5 and otherwise. Both > routines work for positive and negative numbers, and are 35 bytes long. Round to even: << 2 INV OVER 2 MOD OVER == - + FLOOR Round to odd: << -2 INV OVER -2 MOD OVER == + - FLOOR C'mon beta testers; break 'em or improve 'em!! The first one fails for 999999999997, as it rounds that whole number to 999999999998. The last one fails for 999999999997 as it runs into the aforementioned problem with handling a 12 digit number and the addition/subtration of 0.5. ==== Does anybody know how to program the 49G to go into ALPHA mode in User RPL (so one could enter alpha characters on input without having to press ALPHA)? For that matter, how do you program a left or right shift key? I'm afraid you cannot have anything special on keys when you still on standard keyboard and have just pressed the USER on. BUT you may set the ALPHA to work with single keystroke (& USER 2) by pressing |MODE| [FLAGS] and scroll using left-shift down-arrow to flags 60 and 61 and set both. Alternately you may simply use { -60 -61 } SF in RPN Mode. << program >> 42.2 @ .2 = left shift ASN .1 or .0 = standard key .2 = left shift .3 = right shift .4 = ALPHA standard key .5 = ALPHA left shift .6 = ALPHA right shift By adding .01 to these (except for the standard key assignment) you declare Shift&Hold feature Try it out! You may also find KEYMAN very useful! Search for it in www.hpcalc.org It is written by professor Wolfgang Rautenberg ==== Does anybody know how to program the 49G to go into ALPHA mode in User RPL (so one could enter alpha characters on input without having to press ALPHA)? For that matter, how do you program a left or right shift key? In what context do you want to do this? If it is for INPUT, you can set ALPHA via one of the flags in the list on level 1, for example Prompt { Ga} INPUT where Ga stands for the alpha character. To have alpha mode active at the end of a UserRPL program, I can only think of << #25EE9h SYSEVAL >> @ HP49G, doublecheck the number before exec!!! ==== Does anybody know how to program the 49G to go into ALPHA mode in User RPL (so one could enter alpha characters on input without having to press ALPHA)? For that matter, how do you program a left right shift key? In what context do you want to do this? If it is for INPUT, you can > set ALPHA via one of the flags in the list on level 1, for example Prompt { Ga} INPUT where Ga stands for the alpha character. To have alpha mode active at the end of a UserRPL program, I can only > think of << #25EE9h SYSEVAL >> @ HP49G, doublecheck the number before exec!!! That is what I was looking for - the option to INPUT. ==== Hi, > When I first got one I showed to a colleague at work and the first thing > he did was type in 69! (69 factorial) and compared the speed with his > Casio. It was at least twice as fast. Wow was the reaction. Then he > typed in 70! Double wow! He'd never used a calculator that could handle > exponents of more than 2 digits. By that stage (I was mid way through university) I'd almost forgotten > that at school the 69! benchmark was the way all new calculator models > were 'tested'. :-) > Yes, but we used a small synthetic program for the HP-41 in school which could iterate the factorial to as many exponents as the ALPHA register could handle, including the mantissa;-) Raymond ==== so why I made Emu28, the HP28C emulation. sentimental remembrance maybe one part, my 2nd HP calc was a HP28S, but most I think it was, I had time and more important all imformation to make it. In short words, I made it because it was there. ;-) > Hi, When I first got one I showed to a colleague at work and the first thing > he did was type in 69! (69 factorial) and compared the speed with his > Casio. It was at least twice as fast. Wow was the reaction. Then he > typed in 70! Double wow! He'd never used a calculator that could handle > exponents of more than 2 digits. By that stage (I was mid way through university) I'd almost forgotten > that at school the 69! benchmark was the way all new calculator models > were 'tested'. :-) Yes, but we used a small synthetic program for the HP-41 > in school which could iterate the factorial to as many exponents > as the ALPHA register could handle, including the mantissa;-) > HP48 to calculate numbers > 253! in the early 90'ties. But it was very inaccurate, so I converted it in my SystemRPL learning phase into SystemRPL. The following program is quite old and from a time where I used German as documention language (pre Emu48, pre internet). <--------------cut here-----------------> NIBASC /HPHP48-E/ RPL * * Programmbeginn * :: CK1NoBlame CK&DISPATCH1 ( teste Stack1 auf real ) real :: DUP %0< IT :: #305 DO#EXIT ; ( < 0, Infinite Result ) COERCE ( ganze Zahl ) %%10 %%LN ( %%LN[10] vorausberechnen ) %%0 ( f .9fr Ergebnis ) ROT ( Z .8ahler ) #1+_ONE_DO (DO) INDEX@ UNCOERCE%% ( Zahl in Long Real erstellen ) %%LN 3PICK %%/ ( %%LOG, nicht %%LN wegen Genauigkeit ) %%+ ( zu Ergebnis ) LOOP DUP %%FLOOR ( Exponent holen ) DUP4UNROLL %%- ( Nachkommaanteil ) %%* %%EXP ( %%ALOG ) 2%%>% a%>$ ( in String umwandeln ) CHR_E >T$ ( E anh .8angen ) SWAP a%>$ &$ ( Exponent anh .8angen ) ; ; <--------------cut here-----------------> remember, some lines in the source, especially with remarks, are wrapped by the newreader. Have fun with it. Christoph ==== 1. What is this that I've been reading about the Metakernel? Is that a type of card? Program? 2. There are two options for turning a decimal number into a fraction. Q and QPi. The latter calculates the fractional equivalent of a decimal number, AND calculates the fractional equivalent divided by pi and gives the one with the smallest denominator. What I want to know is why would one want QPi? What kinds of situations would call for it? Also, I have an idea of how to calculate the fractional equivalent of a decimal number (Q). Let d be that decimal number, and let GCD be the greatest common denominator. Then I'd compute GCD (1,d) and I'd let 1/GCD(1,d) be the denominator and d/GCD(1,d) be the numerator. For example, 0.428571428571 = 3/7. But let's say that you didn't know that.Then, GCD(1,0.428571428571) = 0.142857142857 (1/7), the denominator =1/(1/7) = 7, and the numerator = 0 .428571428571/GCD(1/d) = (3/7)/(1/7) = 3. But how would you compute QPi? I use Pi in place of the Pi symbol because I don't know where it is on my computer right now. 3. Where can I buy a belt clip for my HP48GX? 4. How many levels will the stack hold? I mean how many items can you have on the stack? How many levels of paranthesis can you have? 5. I ran across this program in the AUR. Ç (0,0) (2,3) LINE { # 0d # 0d } PVIEW 7 FREEZE È I'm new to the FREEZE command. I know that the 7 in this case is the display area that is being frozen until a key is pressed. I've tried values less than 7 only to have the image disappear as soon as it appeared. What exactly does the 7 mean? How do I determine what other values to use wtih the FREEZE command? to a size of 20X10, (or perhaps 21X11 if (0,0) is counted as a point) and draws a triangle with vertices (0,0), (20,0), (5,10). But when I run them, the triangle won't fit the screen, and in one program. I have to scroll around to see all of it.Why? I want to write a program which will take 3 coordinates, and draw a triangle, but I want the triangle to just take up the whole screen. No less, no more. I don't ask that you write the program for me, I just want to know why the below programs only showed less than the whole triangle.Try them out, if you're not sure what I mean. Ç ERASE (0,0) (20,10) PDIM (0,0) (20,0) LINE (20,0) (5,10) LINE (5,10) (0,0) LINE { # 0d # 0d } PVIEW 7 FREEZE È Ç ERASE (0,0) (20,10) PDIM (0,0) (20,0) LINE (20,0) (5,10) LINE (5,10) (0,0) LINE PICTURE È The second program is the one where I can't scroll. Now, I've noticed that if I delete the file PPAR and run these programs, I get a triangle that won't fit in the screen, and a PPAR file that looks like this. { (0,0) (20,10) X 0 (0,0) FUNCTION Y } The above is created if I delete PPAR, and run either of the programs. The triangle won't fit. But, if I do a PLOT with y=X^2 and dimension the screen to -20 to 20 for X and -10 to 10 for Y, and then run the above programs, the triangles fit. And yet, PPAR is the same as the above after I run the programs with the triangles now able to fit. So, what's happening here? 7. Is there more than ONE format for libraries? I mean, I know that you can write a library who's items show up in the softkeys, but can you write one which gives a list of those options the way the built-in equation library does? Or, is it always with the softkeys? -- ------------------------------- Patrick D. Rockwell ==== Hi. 1. What is this that I've been reading about the Metakernel? Is that a > type of card? Program? It used to be that you could only buy it on a ROM card. I believe now you can download it from hpcalc.org and install it on a 128k RAM card. > 2. There are two options for turning a decimal number into a fraction. > Q and QPi. The latter calculates the fractional equivalent of a > decimal number, AND calculates the fractional equivalent divided > by pi and gives the one with the smallest denominator. What I want > to know is why would one want QPi? What kinds of situations would call > for it? Also, I have an idea of how to calculate the fractional > equivalent of a decimal number (Q). Let d be that decimal number, > and let GCD be the greatest common denominator. Then I'd compute > GCD (1,d) and I'd let 1/GCD(1,d) be the denominator and d/GCD(1,d) > be the numerator. For example, 0.428571428571 = 3/7. But let's say > that you didn't know that.Then, GCD(1,0.428571428571) = 0.142857142857 > (1/7), the denominator =1/(1/7) = 7, and the numerator = > 0 .428571428571/GCD(1/d) = (3/7)/(1/7) = 3. But how would you compute > QPi? I use Pi in place of the Pi symbol because I don't know where it is > on my computer right now. Even better, there is a program (you can also find it on www.hpcalc.org) that is better (faster?) than the built-in one. It's called QPI. > 4. How many levels will the stack hold? I mean how many items can you have > on the stack? How many levels of paranthesis can you have? As many (stack / parenthesis) levels as you can fit in [free] memory. > 7. Is there more than ONE format for libraries? I mean, I know that you can > write a library who's items show up in the softkeys, but can you write > one which gives a list of those options the way the built-in equation > library does? Or, is it always with the softkeys? You could have a softkey that calls a GUI similar to the equation library, which contains the choices in the softmenu. Since the GUI is so slow, you may be better off with the softkeys. At any rate, it doesnt do that by default , you'd have to write it yourself. Aaron ==== > 1. What is this that I've been reading about the Metakernel? Is that a > type of card? Program? It's a ROM card and it is also available as a free program, which fits on a RAM card. I would by a RAM card for the HP 48GX just to get the MetaKerenel. Check this page: http://etud.epita.fr:8000/~avenar_j/hp/48.html > 2. There are two options for turning a decimal number into a fraction. > Q and QPi. The latter calculates the fractional equivalent of a > decimal number, AND calculates the fractional equivalent divided > by pi and gives the one with the smallest denominator. What I want > to know is why would one want QPi? When you know that pi is involved. Trigonometrics, etc.. > 3. Where can I buy a belt clip for my HP48GX? Beats me. > 4. How many levels will the stack hold? I mean how many items can you have > on the stack? How many levels of paranthesis can you have? Memory pending, no limit. > 5. I ran across this program in the AUR. Ç (0,0) (2,3) LINE { # 0d # 0d } > PVIEW 7 FREEZE È I'm new to the FREEZE command. I know that the > 7 in this case is the display area that is being frozen until a key is > pressed. I've tried values less than 7 only to have the image disappear > as soon as it appeared. What exactly does the 7 mean? How do I determine > what other values to use wtih the FREEZE command? 1= status area 2=middle section 4=menu keys any combinations add these together to freeze parts of the display so that program stop and show exactly what you want and does not revert to standard display. I leave this to somebody else...too tired....prone to error... > 7. Is there more than ONE format for libraries? I mean, I know that you can > write a library who's items show up in the softkeys, but can you write > one which gives a list of those options the way the built-in equation > library does? Or, is it always with the softkeys? Likewise zzzzzzzzzzzzzzzz ==== >No. if i press 30 and then sin it write: 30 sin (with a space >between) and if I press ENTER I have Sintax error. To be sure we are >tolking abaut the same thing: I am on SOLVE(right shift+7)/Solve >poly.../Coefficient(field) >In that field I cannot put sin30 or > __ >V2 (i.e. 2^ 0.5). also if I put quotes. You need to enter an array. Press right-shift x (multiply) to get the array brackets [ ] in the display. You then enter the coefficients separated by spaces. There is no easy way to get an array from the stack into the Coefficients field. There is no easy way to do a calculation e.g. sin(30) and get that entered as one of the coefficients. The best solution is to press [EDIT] which starts the matrix editor. In the matrix editor you can type 30 SIN ENTER and see 0.5 (in degrees mode) appear. Press ENTER again to return to the Polynomial solver. You could experiment with left+shift 7 instead. > is ON+SPC still available for coma mode, like the HP28/48? Although there is no equivalent COMA mode in 49G, you can make a memory backup (ARCHIVE) to flash (port 2) -- also copy any vital objects from RAM ports 0 and 1 -- then either remove batteries or just insert an insulator between any one battery end and its spring contact. This tends to make the 49G pretty comatose, but you can usually revive it after giving CPR via a RESTORE command -- don't forget to get out of Algebraic mode; might as well actually save and restore all flags, which a STARTUP program can help automate. > I'm trying to solve for x (real values) in the equation: > 0=-25000+5000/(1+x)+5000/(1+x)^2+5000/(1+x)^3+5000/(1+x)^4+5000/(1+x)^5+5000 > /(1+x)^6 As far as I'm aware, there is only two (real) solutions (-1.688 and 0.0547). > The function SOLVEVX (where x is the default independent variable) produces > 4 (real) solutions. Why is this? I'm not real sure what the other two > solutions (-1.2125 and -0.8162) are. It is quite easy on the HP49 to simplify the problem of finding real roots to your original equation to finding zeros of the polynomial 5*x^6+29*x^5+69*x^4+85*x^3+55*x^2+15*x-1. The real zeros to this, as found by SOLVEVX, are aproximately the -1.688 and 0.0547 values you were aware of without any extraneous solutions that SOLVEVX gives for the original equation. Extraneous solutions may be artifacts of the method used by the HP for finding numerical solutions, probably found one at a time and then using reduced equations for subsequent solutions, if any. Making obvious simplifications of your equation should minimize the danger of such extraneous solutions appearing. Merely dividing out the common numerical factor of 5000 eliminated one extraneous solution. Then eliminating denominators to get a polynomial equation eliminated all extraneous solutions you found. And checking the solutions found by substituting back into the original equation should should identify such extraneous solutions as still might still happen to occur. Other interesting results: -20. CF -21. CF -22. SF @ no underflow/overflow/infinite errors 1. 'X' STO 'INV(1.-X)' ->NUM Result on 48G: 9.99999999999E499 Result on 49G: 9.99999999999E499 -1. 'x' STO '-25000.+5000./(1.+x)+5000./(1.+x)^2.+5000./(1.+x)^3. +5000./(1.+x)^4.+5000./(1.+x)^5.+5000./(1.+x)^6.' ->NUM Result on 48G: 9.99999999999E499 Result on 49G: Error: Bad Argument Value 'X' PURGE 'INV(1.-X)|(X=1.)' ->NUM Result on 48G: 9.99999999999E499 Result on 49G: '1./-(1.-1.)' Is this a number? === Has anybody built a 128K RAMCARD for HP48GX? I tried to build one following the instructions of Olaf (see hpcalc.com for the files), but when I try to use it, this is what I obtain: If I turn on the HP, it says Invalid Card Data . OK. I call PINIT . OK. If I watch the contents in port 1, there's nothing. OK. If I call 1 PVARS , the result is 2: {}, 1: 131072 . OK. I store an object in the RAMCARD. OK. If I watch the contents in port 1 now, it says Invalid Card Data . NO! If I try to call MERGE1 to merge the RAMCARD with user memory, I get a system block and I have to reset. I re-controlled 1000 times all the components and all the wiring and it seems to be all ok. At this point my suspect is that the RAM chip is not appropriate: partnumber on the this chip is: USA 9545 HM628128LFP-7 0102QQQG Where I'm wrong? What can be the problem? ==== I bought a HP42S, still in the box, with manual, no wear & tear, > for R400, approx $50. Is that a reasonable price? I searched eBay for recently completed sales (30 days, I believe) > of HP-42S , and found lowest price US $162.50 and highest $255.00 That highest price was for a used 42S with no manuals, > while another described as mint, with manual, went for $245.25 So let us know your secret source, and stand by for a stampede! I reckon it has got something to do with the exchange rate. (R1= one South African Rand) R8 = ±$1, but with R8 you can buy more in SA than with $1 (In most cases except cars, houses), especially food and locally produced goods. Thats why foreigners enjoy SA holidays: accommodation is absurdly cheap (for them) and they can get totally pissed on decent liquor for ±$15. That can explain why somebody would not buy a hp42s, an at least five year old calculator with very strange input (RPN as seen by the average idiot), for R1296 (=$162, a new 48GX cost about $225/R1800 around here) when R1296 is what 3 meals a day for 2-3 months add up to. But I did not realise the 42s is so expensive on e-bay. We generaly avoid e-bay as everything is sold in dollars, its immediately expensive for us. I will keep an I on the classifieds and post on comp.sys.hp48 if there is something available. I myself can inspect the state of the merchandise, but how we will manage mutual security when I post it is here. Ok, thats all for now ==== I used to own an HP49, and it used to display numbers according to the following rules: 1. If the number was a fractional or approximate number (a real ), the number would be displayed on the stack in engineering notation (or whatever display mode was set to). Eg: 20.2321E-3 2. If the number was an exact integer, the number would be displayed without any exponent or fractional part. Eg: 5 The number could be converted to eng notation my hitting ->num. I have got mk+erable+alg48 installed on the 48, and when I enter any integer onto the stack or equation writer, the calculator formats it in engineering notation. For example, if I enter 1 2 I get: 1.0000E0 -------- 2.0000E0 which is rather cumbersome. The equation writer does a similar thing. The equation f(x) = 1 - x comes out looking like: f(x) = 1.0000E0 - x This uses up lots of screen space. Does anyone know of a way to make the 48 display only floating point numbers in engineering notation? ==== > I used to own an HP49, and it used to display numbers > [reals with decimal points etc., per current ENG/SCI display mode, > exact integers as plain integers, always the same] > I have got mk+erable+alg48 installed on the 48 > [where both reals and integers display only like reals in ENG/SCI] That's because all individual numbers on the 48 are reals; there never was any separate type for integers. MK for the 48 did not create a new exact integer type; the 49G's MK is a significant extension of the original 48 MK, but when you go back to an HP48, you've still only got an HP48. . ==== > Does anyone know of a way to make the 48 display only floating point > numbers in engineering notation? > Now that's a funny one. If you knew one many times I've received request so integers on the hP49 would be displayed following the sci,eng or fix mode ... ==== > Does anyone know of a way to make the 48 display only floating point > numbers in engineering notation? > Now that's a funny one. If you knew one many times I've received request so integers on the hP49 > would be displayed following the sci,eng or fix mode ... > Then how about the HP39/40 style Fractional Display Mode? Have you counted how many times you have been asked that? Are you using a zint to count all the requests? (-; ==== Just a quick followup. In my google searches, I did see a few other people asking how to make the 49 act like a 48 :-). I definately think that the 49 behaviour is an improvement over the 48 in this case. John: I was afraid that the answer might be something like this. Many thanks to both of you, and to anyone else who replies. ==== >To connect two computers directly via ethernet, >you'll have to use crossover cables. >In these cables the receive and transmit lines are crossed. Just to confuse things even further :-) some of the more recent network cards can autodetect whether a straight or cross-over cable is being used and change themselves automagically. So it's worth trying to connect directly with one of your existing cables just in case. You don't say what version of Windows you are using so my terminology might not exactly match what you see. But... 1) You need to give each machine an IP address. (Most Windows versions can do this automatically but it's just as simple to used fixed addresses.) You set the IP in the network control panel. Use 192.168.1.5 in one machine and 192.168.1.6 in the other. (Don't use .1 at the end because if you ever get a router or ADSL eventually then it will take this by default.) Start a command window and from the .5 machine type c: > ping 192.168.1.6 and you should get responses that look like: Reply from 192.168.1.6: bytes=32 time<10ms TTL=128 and vice versa from the other machine. The IP address 127.0.0.1 is always the current machine so a good test is to type: C: > ping 127.0.0.1 If you get nothing back then it means that the networking hardware is not working on that machine. It's always worth checking this first for each machine. 2) Configure your internet dial up on the internet machine just as before. Note that your ISP might give you a completely different IP address to use or they will say something about getting one automatically from their DHCP server. This is fine: Windows keeps the two separate so when dialled-up you use the ISP one and when not dialled-up it uses your 192.168.etc one. 3) On both machine you need to turn on networking and put them into a workgroup . Search through help for workgroup to find the relevant details. 4) Install the printer on one of the machines. On the properties set it to be shared. Again, search help for printer sharing for details. 5) On the other machine start the install new printer wizard but choose the networked printer option. If the machines are wired together and turned on at this stage then the name of the other machine might appear automatically. If not you can always use an ip address directly e.g. 192.168.1.6 printer_name in the address field. (Where 192...6 is the address of the other machine.) Finally, don't be surprised if you can't network to the internet machine while it is dialled-up to the internet. This can be fixed but that's best left for another time if it becomes a problem. That's about it. Other things to look at in the help are file sharing so as to be able to transfer files from one machine to the other. (Always worth doing as a backup so each machine acts as a backup of the other.) ==== > This post is NOT targeted to start a dicussion on the merits of the > product versus the 48/49, only to get some feedback if anyone actually > has one of these beasts? I'm interested in where you purchased it (I > find it only at the Casio webite), what you paid, and what you think > of it as a tool for learning math. Simon www.cynox.de www.classpad300.de www.casio.com/index.cfm?fuseaction=products.detail&Product=CLASSPad300 ClassPAD300 is very similar to the TI89/92plus/Voyage200 have approximately 430 commands/functions, approximately 150 have the same name, and make the same operation. Examples: ( arcLen colDim colNorm CubicReg Define GetKey Goto i polyEval ...) Approximately 50 make the same function but they defer in the name Ejemplos: ( combine-CASIO comDenom-TI DelFolder-CASIO DelFold-TI dSolve-CASIO deSolveöTI ExpToStr-CASIO string-TI listToMat-CASIO list>mat-TI NumToChr-CASIO char-TI ·) The ClassPAD300 cannot solve 1:solve(3^x=x^2,x) TI if it can 2: cannot solve Nonlinear system of equations solve(3*b^2*x^2=4*a*^2*b*y and (b*x)/2=(a*y)/3, {x,y}) In casio to replace (AND) by (,) 3:it cannot factor((x+y)^2-4*x*y with respect to x Comparaciones TI-CASIO www.geocities.com/jaimezacalcs/casio/TIvsCASIOclassPad300.htm ==== X > -1. 'X' STO 'INV(1.-X)|(X=1.)' X> Good news and so you won't have to complain anymore. > Bernard and I worked together and it's now working just as on the HP48 > (except of course that the result is not evaluated: it just shows the > replacement)... > PS: How about the CAS flag changes? Can you make them invisible to the user and restore the flags before result display? * A new error might then be needed: Error: Complex Result! (When in real mode.) * Note: Angle functions may need pi/180 or pi/200 (or 1) multiplier for arguments (and divider for result) [internal calculations in radians] depending on current mode: degrees, grads, radians. Hi, everybody I've just written a small User-RPL program (48G, GX) to return the day of the week (as a three character string), given a valid date. This is very trivial task, I think many of you in this NG have already done that, but I have figured out a way to make the program independent of flag -42 setting. That is, no matter the date format you use, the program will work just the same. For example, if your date format is mm/dd/yy, and you enter 6,212003 (June, 21, 2003) the program will return Sat . On the other hand, if your date format is dd.mm.yy and you enter 21,062003 (21 de junho de 2003), the program will return Sat also. The program is just 76.6 bytes long and should work on the 49G too. Tip: no need to check, save or restore any flags. This is not a mini-challenge. I just curious to know if someone has a better solution. If you think the program is worth posting, I would do it later. I've just written a small User-RPL program (48G, GX) to return the day of the week (as a three character string), given a valid date. This is very trivial task, I think many of you in this NG have already done that, but I have figured out a way to make the program independent of flag -42 setting. That is, no matter the date format you use, the program will work just the same. For example, if your date format is mm/dd/yy, and you enter 6,212003 (June, 21 2003) the program will return Sat . And if your date format is dd.mm.yy and you enter 21,062003 (21 de junho de 2003), the program will return Sat also. The program is just 76.5 bytes long, including the 21-character string, and should work on the 49G too. (No need to check any flags!). This is not a mini-challenge. I am just curious to know if someone has a better or shorter solution. If you think the program is worth posting, ==== > These appear to be lengthy and specialized calendar programs (moving holidays, > such as Easter, for instance). I am interested in programs shorter than mine, > which just return the days of the week for any valid date. > GWB. Well, since the system TSTR does this also, you may want to use Nosy on it to find out how one would do it in SysRPL in order to make it very short. I did it and there was this: Maybe this is woth looking for? ==== > make the program independent of flag -42 setting. TSTR takes flag -42 into account; otherwise, if you are calculating the day of the week yourself, use a reference date which has mm equal to dd Some user programs, for example: << 1.0119 SWAP DDAYS 7. MOD 1 + >> 'WN' STO << 0. DATE+ MonTueWedThuFriSatSun OVER WN 3. * DUP 2. - SWAP SUB ->TAG >> 'WD' STO E.g. 7.07 WD -> :Mon: 7.072003 . ==== make the program independent of flag -42 setting. TSTR takes flag -42 into account; Yes it does and I like it! << TSTR 1. 3. SUB >> ==== He meant << 0. TSTR 1. 3. SUB >> . ==== >Well, since the system TSTR does this also, >you may want to use Nosy on it >to find out how one would do it in SysRPL >in order to make it very short. I did it and there was this: >Maybe this is woth looking for? Here is my wonderful program: << SatSunMonTueWedThuFri 1.012 ROT DDAYS 7 MOD 3 * 1 + DUP 2 + SUB >> It seems very very lengthy now :-) It should be useful only if you want the week day in your native language. ==== > << SatSunMonTueWedThuFri 1.012 ROT DDAYS 7 MOD 3 * 1 + DUP 2 + SUB >> X > It should be useful only if you want the week day in your native language. X modified text below: > French SamDimLunMarMerJeuVen > Italian SabDomLunMarMerGioVen > Spanish SabDomLunMarMieJueVie > Portuguese SabDomSegTerQuaQuiSex Don't know what the string looks like in Finnish ... :-) Finnish LauSunMaaTiiKesTorPer You may get more languages from http://www.hpcalc.org/details.php?id=4674 ==== For a cute Windows desk accessory, which has choice of Algebraic with precedence, algebraic as entered, or RPN, and many extras, such as time arithmetic, stopwatch/timer, complex numbers, finance, stat, base conversion, matrices, and even Roman numerals!: http://www.calculator.org HP receives mention among their favorite real calculators: http://www.calculator.org/calcs.html . ==== variables and get back a matrix filled with values depending on A,E,L. For example: Kp1(A,E,L) = [[A*E/L,-A*E/L] [A*E/L,-A*E/L]] I have tried to do this with the Define function but it dosn't work.I entered: Define(Kp1(A,E,L) = [[A*E/L,-A*E/L] [A*E/L,-A*E/L]]) Enter but it got me a error mesage.How can i do it? P.S.:Is there any good guide to HP49G programming and creating functions for newbies? ==== MAP > variables and get back a matrix filled with values depending on A,E,L. > For example: Kp1(A,E,L) = [[A*E/L,-A*E/L] [A*E/L,-A*E/L]] I have tried to do this with the Define function but it dosn't work.I > entered: Define(Kp1(A,E,L) = [[A*E/L,-A*E/L] [A*E/L,-A*E/L]]) > Enter > but it got me a error mesage.How can i do it? ==== I found it in a guide for cas programming.The syntax sholud be for those who work in Algebraic mode: << -> A E L <<[[A*E/L,-A*E/L] [A*E/L,-A*E/L]]>> STO Kp1 ==== Hi does anyone have a program to make the hp48 work as a PID or a powerdistributor by controlling on and off times for a relay? Michael ==== > my HP48G's ON button stopped working. > the printed circuit board (PCB) for the 'ON button is worn out. > There are clearly visible bars etched in the carbon (or whatever > conducting material HP used) film. The minute bars match the bars > on the contact point. I would think that high-force keys would have this problem more than low-force keys, and that very thin films would have this problem more than thicker contact elements. At any rate, keyboard failure and LCD connection strip detachment seem to be the most often reported causes of failure, with keyboard failures in HP calcs going back as far as 20 years. Neither has yet occured on my emulator, but then, my emulator doesn't fit in my pocket :) . ==== Through ordinary matrix algebra, given: Ma (1x3) = [ J, K, L] and Mb (3x3) = [ A, B, C] [ D, E, F] [ G, H, I] the product is: Ma*Mb = [JA + KD + LG BJ + KE + LH JC + KF + LI] BUT in the HP49G, when I enter: Ma Mb * I get an invalid dimension error. If I reverse the stack order of Ma and Mb, then it works, but I get a convoluted answer: Mb*Ma = [LC + K B + JA LF + KE + JD LI + KH + JG] According to the 48G manual, entering Ma first is the correct order. Am I doing something wrong here? ==== 073Ja.8819$cp5.4961@news02.roc.ny.frontiernet.net, JohnnyCashFan > According to the 48G manual, entering Ma first is the correct order. Am I > doing something wrong here? This is the same on the HP48... * on matrix is not commutative (is that an English word?) ==== > Through ordinary matrix algebra, given: Ma (1x3) = > [ J, K, L] Use Matrix Writer and check off VEC to get [[J,K,L]] > and Mb (3x3) = > [ A, B, C] > [ D, E, F] > [ G, H, I] the product is: Ma*Mb = > [JA + KD + LG BJ + KE + LH JC + KF + LI] [GL+DX+AJ HL+EK+BJ IL+FK+CJ] Only the order is different > BUT in the HP49G, when I enter: Ma I get an invalid dimension error. If I reverse the stack order of Ma and > Mb, > then it works, but I get a convoluted answer: Mb*Ma = [LC + K B + JA LF + KE + JD LI + KH + JG] According to the 48G manual, entering Ma first is the correct order. Am I > doing something wrong here? That's the way to handle it with HP symbolic calculators ==== [second try, better explanation] > Ma (1x3) = [ J, K, L] That's a vector, not a matrix; a vector is a 1-dimensional list of values, while a matrix is a list of vectors, which has two dimensions (how many columns, how many rows). On the stack, the number of square brackets distinguishes [ ...vector... ] from [[ ...matrix... ]] In the Matrix Writer, the VEC menu key shows a square dot for a vector, not for a matrix. > Mb (3x3) = > [ A, B, C] > [ D, E, F] > [ G, H, I] > Ma*Mb = invalid dimension Vector * Vector is not allowed, it's in fact ambiguous, because [ 1 2 3 ] can not distinguish between: A = [[ 1 2 3 ]] a 1-row, 3-column matrix B = [[1] [2] [3]] a 1-column, 3-row matrix and you could multiply either A*B or B*A, giving very different results. Vector * Matrix (what you tried to do) is also not allowed. Matrix * Vector is allowed as a special case, in which the vector is interpreted like a 1-column matrix, and the result is likewise delivered in vector form, as a convenience for expressing equation systems like 'A*X=B' where 'X' and 'B' may both be expressed as vectors. . ==== I can't figure >out how to change the contrast. Hold down the ON button and press + or - for darker/lighter contrast. -- ==== I don't know if any of you remember me, but I'm the original author of HP-IDE, HPKermit, HP-IR and a misc. number of other tools. My original managed to drop off the planet. I'm pleased to announce that I've located the original source to my HP-IDE application and that I'll be releasing it under the GPL. If anyone would send it. Otherwise, I'll send an announcement to the list when I'm ready to release the code and let everyone know. I know that someone within the community has managed to write another IDE. With any luck, some of the work I put into HP-IDE can be integrated into this other IDE or the bugs can be fixed. I'll also be releasing the source to any other HP-48 applications I can find. ==== > What infrared printers were produced (I see you did get a HP82240A > & Hp82240B), are still produced, are they compatible with the 42S I use my HP82240B with my 42S; I don't recall whether it makes any difference between A and B printer models for the 42S (and 17B etc.), because IIRC the difference is only in regard to printing some of the special characters in the 128-255 range, which are needed for HP28/48/49, but I don't think are used by earlier calculator series. I'm sure these are still sold and/or auctioned in many places (I haven't even looked to check at HP shopping). > and can you use them with the HP49G that has no IR port? The only input device for an HP82240A/B printer is its IR receiver (there is no wire/cable input), so a 49G, which has no IR transmitter of its own, can not directly send anything to it. However, a 49G can send *ascii* text (not binary objects) to a 48 by wire (flag -33 clear on each calc) using Kermit in ascii mode (flag -35 clear on 49G), and a 48 can print that in turn to an HP82240 by IR (flag -34 *must* be clear on the 48 to print to the HP82240). . ==== About sending strings from 49G to 48 to HP82240 IR printer: It occurs to me that sometimes you can't even send some string in ascii mode from 49G to 48; for example, the way any embedded quotes are internally escaped using on the 49G is not understood by the 48. If this problem comes up, try sending the string in *binary* mode from the 49G to the 48; then on the 48, remove the first 13 characters of the string that you receive, via << 14 1E7 SUB >> This fixes all incompatibilities when sending any string object from a 49G to a 48 calculator. If what you want to send from the 49G isn't already a string object, first do STD 64 STWS ->STR on the 49G to make it a string. ==== Now, all I need is a 2 Meg card for my GX. Suggestions? Price is an object, as I'm decidely *not* in def 1 above. ==== >> doesn't work for negative numbers >> But there is an escape clause, which is that >> Christian originally asked for a program >> to deal only with positive numbers >> (and many programmers never deal with >> anything more general than they're asked to, >> which is why so many narrowly conceived systems are so poor, >> and why there were so many Y2K bugs, too :) >> or for x > 1E20 >> Change it to: << DUP 0 RND DUP 2 MOD NOT ROT FP 2 INV == AND - 45 bytes, checksum # D2h >Still doesn't work for negatives (didn't think we were doing them), >but does work for 1E20. the 19th . Still trying to recover data. ) Here's a solution that I believe works for all non-negative reals : << DUP 2 MOD 2 INV - DUP FP XOR - 0 RND >> Size : t40 bytes Checksum : # 8EDCh (48) Jonathan ==== > Here's a solution that I believe works for all non-negative reals : << DUP 2 MOD 2 INV - > DUP FP XOR - 0 RND Size : 40 bytes > Checksum : # 8EDCh (48) > D'oh! Yeah, the reason for my original program's failure is the impact of how the HP handles the subtraction of 0.5 from the input number--it doesn't have enough digits to give an accurate result for 12 digit inputs and so rounds the result inappropriately. By doing the 2 MOD first you get around that issue. Nifty. How about: << DUP 2 MOD 2 INV 3 * == - 0 RND >> 37.5 bytes, checksum # FEEEh ==== X > How about: > << DUP 2 MOD 2 INV 3 * == - 0 RND > 37.5 bytes, checksum # FEEEh IF this 0K, THEN I'd suggest: << DUP 2. MOD 3. 2. / == - 0. RND >> 35 bytes, checksum # D6AEh END ==== How about: >> << DUP 2 MOD 2 INV 3 * == - 0 RND >> 37.5 bytes, checksum # FEEEh >IF this 0K, THEN I'd suggest: ><< DUP 2. MOD 3. 2. / == - 0. RND >35 bytes, checksum # D6AEh END > I've been waiting to see if my earlier comments about the nature the HP48's representation of integers would inspire anyone; I hope they did. Now, Busby, Bachman, and Nousiainen's latest are getting progressively smaller and they do work for all positive integers. But here is one that is smaller yet and works for negative integers as well: << DUP 2 MOD 2 INV == + FLOOR >> This rounds to odd as requested by the original poster. With a couple of changes we get: << DUP 2 MOD 2 INV == - CEIL >> This rounds to even, and works for negative integers. Both are 30 bytes long. ==== > X > How about: > << DUP 2 MOD 2 INV 3 * == - 0 RND > 37.5 bytes, checksum # FEEEh > IF this 0K, THEN I'd suggest: > << DUP 2. MOD 3. 2. / == - 0. RND > 35 bytes, checksum # D6AEh END Well I guess there is no reason to '2 INV 3 *' when you can simply '3 2 /'. :) ==== X > << DUP 2 MOD 2 INV == - CEIL > This rounds to even, and works for negative integers. X Have you tested it on positive integers? ==== >X >> << DUP 2 MOD 2 INV == - CEIL >> This rounds to even, and works for negative integers. >X >Have you tested it on positive integers? Yes. Have you? ==== >X >> << DUP 2 MOD 2 INV == - CEIL >> This rounds to even, and works for negative integers. >X >Have you tested it on positive integers? > How embarrassing! It's not the positive integers that are problematical. It's the positive (and negative) numbers with fractional part OTHER than .5 that don't work right. This isn't going to be easy to fix without adding bytes. Especially if it is to work with both negative and positive numbers. I was so busy getting the right behavior for the special case of fractional parts of exactly .5 and both positive and negative numbers that I forgot the easy stuff!! ==== > On Sun, 22 Jun 2003 05:10:37 +0300, Veli-Pekka Nousiainen X >> << DUP 2 MOD 2 INV == - CEIL >> This rounds to even, and works for negative integers. >X >Have you tested it on positive integers? > How embarrassing! It's not the positive integers that are problematical. It's the > positive (and negative) numbers with fractional part OTHER than .5 that don't work right. > This isn't going to be easy to fix without adding bytes. Especially if it is to work with > both negative and positive numbers. I was so busy getting the right behavior for the special case of fractional parts of > exactly .5 and both positive and negative numbers that I forgot the easy stuff!! ==== > ... want to write a text on notepad or Word on the PC and transfer it to > the calc.How can i do it? ... > how am i going to view it on the calculator?Is there an application on > the calc to view the text? Scribe's pretty useful:-- http://www.hpcalc.org/details.php?id=4411 Have to format your text for upload something like this:-- %%HP: T(1)A(D)F(.); { I want to write a text on notepad or Word on the PC and transfer it to the calc.How can i do it? ... how am i going to view it on the calculator?Is there an application on the calc to view the text? ==== Why the curly brackets? It seems that it should work without them... le'me test it. Toby > ... want to write a text on notepad or Word on the PC and transfer it to > the calc.How can i do it? ... > how am i going to view it on the calculator?Is there an application on > the calc to view the text? Scribe's pretty useful:-- http://www.hpcalc.org/details.php?id=4411 Have to format your text for upload something like this:-- %%HP: T(1)A(D)F(.); > { > I want to write a text on notepad > or Word on the PC and transfer it > to the calc.How can i do it? ... > how am i going to view it on the > calculator?Is there an > application on the calc to view > the text? > } ==== I'm learning sys-rpl these days (well... trying it). Someone could explain me which is the difference between a command a its PTR reference? Do I save memory using it? Should I use them always I can? PD: when I finish learning sys-rpl I will begin with english, sorry. ==== On 21 Jun 2003 18:43:07 -0700, antoine.jaj@worldonline.es (antonio >Hi, >I'm learning sys-rpl these days (well... trying it). Someone could >explain me which is the difference between a command a its PTR >reference? Do I save memory using it? Should I use them always I can? I don't master SysRPL neither the innerworkings of the calculator, but a PTR is a pointer (un puntero) to the actual memory address of the command. You should not use this, you must look for the SysRPL supported command to perform the operation you want to perform, because a PTR ussualy changes between different ROM versions and your program will only work in the ROM version you're using. That's why you must use the supported commands. -- ==== > what is the difference between a command and its PTR reference? None. Commands within your programs are either the direct memory address of the ROM code which implements them (a PTR, 2.5 bytes), or an object which refers to a library number and a function number within that library (a ROMPTR, 5.5 bytes), or else, in the 49G, a similar reference in a FLASHPTR. The stable, supported commands have a fixed address (or library/function numbers) in all ROM versions of the same product, and also tend to have a symbolic name to which you can refer when you compile or decompile using standard symbol tables, or when the calculator itself compiles or decompiles UserRPL, so that you don't have to use PTR xxxxx to refer to them. If you are exploring ROM with a decompiler which has access to the standard symbol tables, and yet encounter PTR xxxxx in the listing, this generally means that the address is not one of those whose location and function can be relied upon to be the same in all versions; however, there are *some* of those which are asserted to be stable and which yet were not included in the supported list. User commands can usually be represented in SysRPL by a symbol prefixed with x ; e.g. xVARS for VARS, although there are also some user symbols which actually have multiple variants, depending on where they appear in user syntax (e.g. xIFEND vs. xWHILEEND, and multiple different forms of program delimiters); there are also commands within user programs which are not displayed at all when normally viewed in the calculator (e.g. xSILENT' and xFCNAPPLY, :: and ; ). . ==== > This may be a little off topic, but how would I program a random number > generator without using any RAND functions? Anyone attempting to produce random numbers by purely arithmetic means is, of course, in a state of sin. Some introductory material... http://plus.maths.org/issue24/features/keeling/ http://csep1.phy.ornl.gov/rn/node6.html http://www.cs.sunysb.edu/~skiena/jaialai/excerpts/node7.html ==== I use 1.19-6. I have been trying to customize my user keyboard. I have been using the S?UK utility of the OT49 library to get the code behind keys such as the equation writer on level one of the stack, then reassigning them to different keys, typing, for example, 41.1 2. ->LIST STOKEYS That method has worked for reassigning the equation writer to key 41.1 and the double-tick mark (where the cursor is placed between the ticks) to key 43.1. I also tried using it to switch the functionality of the MODE key to 22.2 and the custom menu to 22.1. At first, that seems to work without any problems. Pressing [MODE] produces the custom menu, and pressing Lshift [MODE] produces the CALCULATOR MODES dialogue box. However, the glitch happens when I press Lshift-HOLD-[MODE]. Instead of producing that convenient menu in which one can quickly change modes and see some settings-related commands, it simply opens the CALCULATOR MODES dialogue box. Lshift-HOLD-[MODE] works perfectly fine when i delete the key assignment to 22.2, and it also works when I switch out of user mode. Can anybody explain why this occurs? ==== > ... I have been using the > S?UK utility of the OT49 library to get the code behind keys such as the > equation writer on level one of the stack, then reassigning them to > different keys ... > I also tried using it to switch the functionality of the MODE key to 22.2 > and the custom menu to 22.1. At first, that seems to work without any > problems. Pressing [MODE] produces the custom menu, and pressing Lshift > [MODE] produces the CALCULATOR MODES dialogue box. However, the glitch > happens when I press Lshift-HOLD-[MODE]. Instead of producing that > convenient menu in which one can quickly change modes and see some > settings-related commands, it simply opens the CALCULATOR MODES > dialogue box. This is a kind of a bug in the operating system. As soon as you assign a shifted key, the corresponding shift-hold key will do the same, just overriding its standard functionality. Can prevented in your case only by explicitly reassigning Lshift-hold 22.21 with its standard function (the invoke of Menu 63 in all modes, only ROM 19-6). ==== I have been using the S?UK utility of the OT49 library to get the code behind keys such as > This is a kind of a bug in the operating system. As soon as you assign > a shifted key, the corresponding shift-hold key will do the same, just > overriding its standard functionality. Can prevented in your case only > by explicitly reassigning Lshift-hold 22.21 with its standard function > (the invoke of Menu 63 in all modes, only ROM 19-6). ==== >the new edition of Timeman contains also a fast and comfortable alarm >setter. It allows repeating an alarm after precisely one (or several) >year(s), to remind on birthdays, say. one year later means here: the >alarm repeats at the *same day of the same month* one year later. Alarms >set on the 29th of a February follow the Birthday rule , i.e., a year >later is the 1st of March next year. This type of alarm setting cannot >be done on the 49 in the standard way. Hence the following CHALLENGE. Write a shortest program in (a) UsrRPL program (b) SysRPL >whose input is a whole number n in Level2 and a DATE real in Level1, and >whose output is the correct date n years later as explained above. CONDITIONS: The program should be independent on the date flag -42. Only >supported pointers and named rompointers are allowed. The input DATE can >be any real accepted by ->DATE. The output must be correct as long as << >DATE n 365. * DATE+ >> does not error. Argument checking is not >required. The number in Level2 is a positive real with fractional part >0. TIME is not considered in the challenge. Assuming my interpretation of the above requirements is correct, then here are two solutions that should work. Each accepts a positive or negative year argument in level 2 and a date argument in level 1. User-rpl Bytes : 48 Checksum : # 1A3h (48) << 1.012 DUP ROT DDAYS SWAP ROT 6 ALOG / + SWAP DATE+ >> System-rpl Bytes : 35.5 Checksum : # 7595h (48) :: % 1.012 DUPROT DDAYS SWAPROT %6 %ALOG %/ %+SWAP DATE+DAYS ==== On Sat, 21 Jun 2003 18:56:19 -0500, Jonathan Busby >Assuming my interpretation of the above requirements is correct, then >here are two solutions that should work. Each accepts a positive or >negative year argument in level 2 and a date argument in level 1. User-rpl >Bytes : 48 >Checksum : # 1A3h (48) << 1.012 DUP ROT > DDAYS SWAP ROT 6 > ALOG / + SWAP DATE+ > System-rpl >Bytes : 35.5 >Checksum : # 7595h (48) :: > % 1.012 DUPROT DDAYS SWAPROT > %6 %ALOG %/ %+SWAP DATE+DAYS >; Nth note to self: think before posting.... ==== any program for solving differential equations, at least 2nd order. I own a HP 49G. ==== Press |EQW| |NXT| [HELP] |ALPHA| | L | Scroll to LDEC and press [ ON ] Read the help and then press [SEE1] for DESOLVE > any program for solving differential equations, at least 2nd order. I own a HP 49G. ==== > a new ROM function CASCOMPEVAL, instead of the original COMPEVAL Well, I'm wrong -- in the 49G, both of those give the same result > on David Davies' example: -1. 'X' STO 'INV(1.-X)|(X=1.)' > followed by either CASCOMPEVAL or COMPEVAL > results in the same answer 0.5 > (also regardless of the original state of flag -3) > Good news and so you won't have to complain anymore. Bernard and I worked together and it's now working just as on the HP48 (except of course that the result is not evaluated: it just shows the replacement)... ==== > Good news and so you won't have to complain anymore. I'm just documenting, JYA, not complaining; I'm not a bug -- I'm a feature :-) > Bernard and I worked together and it's now working > just as on the HP48 > (except of course that the result is not evaluated: > it just shows the replacement)... I'm sure that David Davies, who is the one who complained (to whom I was trying to _ex_plain) will be delighted that it's now working on his 49G! I hope you do get through the obstacles at HP to be able to continue to distribute your fine work, and best wishes. ==== I'm just documenting, JYA, not complaining; Is there an upgrade for this newsgroup? ==== http://www.coastaltech.com/monty.htm http://www.io.com/~kmellis/mh-math-faq.html http://www.math.toronto.edu/mathnet/games/montymath.html http://www.rdrop.com/~half/Creations/Puzzles/LetsMakeADeal/explanations.html http://www.rdrop.com/~half/Creations/Puzzles/LetsMakeADeal/ http://stevenlehrer.com/explorers/chapter_4.htm