A99 ==== (http://perso.wanadoo.fr/bernard.parisse/english.html) at hpcalc.org, I could not find any link there to HP France. I tried going to HPs French site but, not speaking french, did not manage to get very far. Could you please give me a more specific address? -- Colin Croft Old mathematicians never die; they just lose some of their functions. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== ==== http://www.hp.com/calculators/france/prod/casenglish.zip ==== http://www.hp.com/calculators/france/ http://www.hp.com/calculators/france/prod40.html#05 but there are only French manuals... You can find the manuals for the HP40G in French, or for the HP49 in English there: http://www-fourier.ujf-grenoble.fr/~degraeve/ Hope that helps ==== the other language versions may be useful for the guy in Europe who had contacted me to see if I knew where to find the CAS manual for him and his friends. How come they (so he says) don't supply the manual with the HP40G in Europe? I would have thought that it was quite necessary to put it mildly! -- Colin Croft Old mathematicians never die; they just lose some of their functions. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== ==== My homepage is not maintainted at perso.wanadoo.fr, the maintained one is www-fourier.ujf-grenoble.fr/~parisse/english.html ==== is there a chance to access the CPU speed by using a little program in UsrRPL or SysRPL. I would like to make some experiments, where I need direkt access to the CPU-speed. The access via ON-hold D is not sufficient for my requirements. Heiko ==== Don't be afraid about Unix installation. If I can do it, then anyone can. The Unix distributions nowadays are really very easy. Not to speak about the already installed MacOS X, which *has* the IDIOT type of interface that you talked about. And what an interface! Looks great! Greetings, Nick. ==== Well, Thomas, even the classic MacOS isn't totally automatically installed. You have to answer some questions. But my fears about unix installations just disappeared when I tried to install Solaris on a PC. It wasn't difficult at all. And just worked right. I guess that the difficult installation of unix is more remembering what it used to be than real difficulty. Grretings, nick. ==== That is exactly my point. It is Windows that has been accused of monopoly. Office still has viable competition. I can't see how you eliminate Windows monopoly. You just raised interesting argument. Are you suggesting that we should split evil TI for it's monopolistic practices int the calculator area, where they seem to own more and more of the market due to their marketing strategy over supposedly better competition of HP ? :-) It doesn't work this way. The file format of the document is not required to be public domain and is not required to be guaranteed in the future. Threfore all parties are guilty of using their own, non published formats that do change with application upgrades. You cannot pick just Microsoft and blame them for doing what everybody else is doing. But even common format does not eliminate practical monopoly. Images are typically stored in public domain formats of TIFF, JPEG, GIF etc. But Adobe with it's Photoshop has practically eliminated all other competition for professionals. There are some toys for amatours, don't even mention them :-) You are mistaken. What is better is a matter of opinion. What is sold on the market is mainly Microsoft Office :-) and can be measured in a real statistics. Generally it is assumed, that better product wins over worse, and better here is defined in economical-technical relationship, not purely technological marvel but expensive. I cannot use Mac. Many of the applications I use are not available Windows that crash on me much more rarely than Sun Unix at work. Nooo. Evil Microsoft help Mac to be better ?!?! ;-) But you see, World Standard CAD is not there. I'm sure, there are many such programs that you can't get for Mac and almost 99.99% Mac applications are now available for Windows or in fact were originally written for Windows. You might not need them today so you think you have a good computer. But when you will find out, that you need software not available for Mac then you will understand, what I meant by sacrifice :-) Last time I've checked it was Windows who first come out with preemptive multitasking on the market. Mac was second. Same with multiple CPU support. It was used toward Mac new versions of OS that suppose to be as good as new versions of Windows. Jack <3C20DE69.46A4A85@home.com> <3C20EBF5.AF48E490@wildopensource.com> <3C21464A.674C212F@home.com> <3C222F7C.B2DB9788@wildopensource.com> <3C2243F9.266620D9@home.com> <9vthjf$nts$1@nyytiset.pp.htv.fi> <3C22BF74.163FB503@home.com> <3C34D460.4E2B484D@home.com> <3C36381D.4B3260DF@home.com> <3C387425.4AB765CD@home.com> ==== GIF is not public domain, some company (IIRC IBM) holds a patent on the used compression technology. Maybe you should fire your sysadmin. Solaris is considered even more crashes throughout 3 years of use. And where do you think do these ideas come from? AFAIK the very first Unix had a preemptive multitasking kernel (else it wouldn't be Unix). In the meantime Mac OS has become a true Unix flavour system. Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== Nooo! What about Mathematica?! :-) It tries to do everything (scriptable Java interface, XML/MathML, and of course the usual CAS stuff), and will no doubt continue to improve in that regard. Bhuvanesh. ==== It is in a sense, that it's file format is widely published and available for other applications to use. I think, the only part of GIF that is patented is LZW compression algorithm used by GIF, which initially was thought by Compuserve as a public domain and later created all the GIF controversy when UNISYS decided to take advantage of it. The network was setup by Sun. It would be difficult to fire Sun :-) Considered is a key word here. What is considered does not reflect the truth necessarily. Solaris crashing once per couple of months is considered stable. My Windows did not crash for the last year although working time was probably similar due to the fact, that one uses home computer generally less at home than at work so clearly Windows instability is for me a street hype and myth. As of firing sysadmin, it would be a little difficult to fire such people in a 300,000 employee + company by me. It is not a single person, but entire department that heavily tests hardware and software before even committing change to the production floor :-) Did I said that Unix didn't have it ? You must missed something. I said Mac OS did not have it till very late years after Windows. Really ? Are you claiming, that Mac OS is in fact Unix ? How come then any of the Unix applications available for Solaris DO NOT RUN on Mac ? ==== I am referring to the *fact* that Mac OS X (the earlier versions don't) has a Unix kernel. You'd most likely only need to recompile your programs on Mac OS to make them work. (Granted, Unix is not exactly the same everywhere, so your programs might be using Solaris-only features.) ==== I know some people who do in fact prefer to have tasks separated into multiple applications. For example some programmers prefer separate compiler, linker and debugger. But it is sufficient to go to nearby software store to find out that the opposite trend is in fact the mainstream. For example Visual Basic, Visual C++, Borland Delphi, Borland C++ Builder etc. all offer at least all those applications combined into the common graphical interface that looks to the user as a single application. Those who prefer to use them separately are in fact left to run DOS style command interface. They of course consider theselves GURUS but it rarely makes them more effective :-) Most of the CAD/CAM/CAE applications are evolving int othe entire modeling/drafting/analysis/manufacturing packages. Graphics packages are now evolving into the single application etc. Almost every type of software is now evolving into the huge packages capable of doing every little need that every user might have. It doesn't matter that the underlying technology is separated into the blocks of separate code. The reason of it is more of a manufacturing and code maintenance. It is obviously not the customer' demand. Jack ==== You must be kidding. I can't recompile code that I don't have and own. Don't even compare amount of applications available for Windows and for Mac even with Unix kernel. IMHO Unix kernel only proof how desperate Mac become. Jack ==== Yikes! Been playing with buggy svgalib programs? I think that Unix is almost entirely outside Jack's scope. ==== I own a HP48G and I have a problem I need to solve. I have 3 equations with 3 variable, how do I punch it into the calculator to get the results of the 3 variables? I have looked and I have looked in the manual, but I cannot seem to find it. Can anyone here help me, PLEASE????? Pernille ==== with the The keywords should be solve linear equation matrix. To solve linear equations you simply divide two matrizes. HP48G is well documented, you should find that. or have a look/search on www.hpcalc.org ..Heiko ==== Which is the best way to load computer written files into the calculator and how to watch them later? How can i create grob pictures? Grettings Richard ==== ==== But seriously. Of course making a new company is not an easy thing, but knowing what excellent talents there are out there, I would find it a pity if all those invented things would remain unknown to the wide mass. You already have a customer. I order right now a calc and you can tell me what it costs when it is ready. Come on guys, enforce the revolution! Greetings, Nick. P.S. Rcobo has great knowledge about electronic devices. Wouldn't you take him in your team in Denmark? A multinational multitalented crew would be marvelous. ==== As much as I would like to include you all in this adventure, time must tell. The deal is that I and a couple of my friends/colleagues have started a business now, and we are in full development. We do have the knowledge in all areas necessary, so I don't think it'll be much of a problem at first. If and when production starts, and the product line needs to be further developed, we maybe aren't able to manage it ourselves. We are three EE's, one datamatician and one CNC custom tools designer on the side. This means that we are highly qualified in working with all aspects of hardware and software development. This started out as an idea, but turned into a company. ==== I really need to work with exal but i don't know why I can't put the program in my calculator. I have a HP48Gx. I've placed the program in my library but then the exal don't run. I would like that someone could help me to solve this problem. may someone send me information about the program, the file or instructions or how I can start the exal. ==== Another Xpander has shown up on eBay: http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?ViewItem&item=1318175911 Roger == If the key handling routines are ML and *very* fast, you can use =POPKEY or =Debounce. Thomas -- Hmm, I don't know if one could notice the speed difference. Perhaps a little bit slower but really only a little bit. Sorry that I can't help you with SysRPL and ML. Exactly, although only XYGROBDISP is used, no freezing. Sorry, but I do not know which display status flag you're talking about. Anyway, the problem has nothing to do with the TurnOff-command. Already the simple program ; does not refresch the display if stored in STARTOFF. not turned off the calculator by hand ... you could use the display status flag ClrDAsOK... This signals that the whole DA (Display Area) has to be redrawn. This one should not be fixed! I'm gonna use this as an identification feature. So if someone NOT familiar with the calc forever!!!! ==== Is this a bug or a feature: pressing Backspace in the Matrix Writer causes an item to be dropped from the outside stack. I was really surprised when I left the Matrix Writer and some items from the stack were lost. BTW: pressing left-shift + backspace makes the current matrix element zero. Are there any more hidden features? ==== Haa! This one should be the other way around! A simple BS to drop an element and the DEL to DROP a stack element... . ==== . ==== There are these books - Science and Engineering Mathematics with the HP 49 G - Volume 1 and 2 by Gilberto Urroz that you can find in www.greatunpublished.com I don't own a 49, neither those books. But I've heard a lot about them, and they seem to be good. ==== Hmm. Of course ClrDAsOK itself is *not* a DA flag;-) Instead, it's a word that clears the OK status of the DA1, DA2a, DA2b, and DA3 display status flags... ==== My Hp shows me for differential sin(x)=0 and not cos(x), which mode is wrong?? ==== Thomas Kremer schrieb: Check your VX setting (or execute X STOVX). If VX is something else than X, then SIN(X) is a constant and therefore its derivative is 0. -- ==== I'm not talking about converting an existing ROM over to reference counting. Or, for that matter, even writing a new HP49 ROM. My main question was would it be possible to implement RPL style semantics atop a reference counting garbage collector. I know that in Lisp, it's pretty easy to get circular references in lists, and couldn't come up with a similar case in RPL. They did well not to. http://www.mschaef.com ==== .0 ==== It is located at address #00101h and is two nibbles wide You can PEEK at that address or, you can use this code. Assemble it with Jazz :: CODE GOSBVL =SAVPTR D0=(5) #101 A=0 A A=DAT0 B GOVLNG =PUSH#ALOOP ENDCODE UNCOERCE ; -- : Try a search at http://www.hpcalc.org/. James ==== ==== There are some packagaes on hpcalc.org that do just that. Just out of curiosity, why? Does the 17BII lack something you need? I was considering augmenting my 48GX with a 17 to get the cleaner interface. http://www.mschaef.com ==== I haven't touched my 49G for a long time. The user key assignments I had seem to have disappeared, and I can't find nor remember how I did them! I wish to switch the functions MTH and SYMB. Someone once told me it had to do with using a minus sign in front of the key designations. If someone would be so kind as to describe the process to me, ==== For key assignments, please take a look into the user manual;-) The program you mean is KEYMAN (or alike) Both available on www.hpcalc.org ==== None of them, it must in the source from which the informboxes are called. (via xNAME or NULLNAME) Read http://groups.google.com/groups?hl=en&selm=8vc6gs%24p9pg%241%40reader4.wxs.n l from Pieter, it explains it in detail. So after you have read the post from Pieter it is obviously that ASSEMBLE =PTR39251 EQU #39251 RPL must be placed in start.s in his example. Then the use of PTR39251 works. Don't forget the ' before your messagehandler! Didn«t know that, thanks for the info. ==== : Even pointers from 25EC to B3C7 seems to be unsupported but stable (always JYA docet) o /-----------------------/ ==== I have a remote control program that based on some other programs. The part of the keyboard manager is mine, and I believe that there is room for improvement. The problem is the slowness, from the moment that the ENTER key is pressed until the 3 characters being transmitted (to increase the reliability in the IR) it takes nearly 1 second. I would appreciate it if you could give me another alternative or tips to improve the task: %%HP: T(3)A(D)F(.); @ With 10 minutes auto power-off. @ [DEL] key = Esc. @ [ON] key stop the program; [Backspace] key to OFF the calculator. @ The program based on dterm program. @ [ENTER] key for confirmation. << GROB 18 10 FFFF30FFFF30FFFF30FFFF30FFFF30FFFF30FFFF30FFFF30FFFF30FFFF30 -> black << 89 1 0 0 rc48.gro RCLF 0 -> i j k c menu flags tmp inln << CLLCD { IR { } { } { } { } WIRE } TMENU [IR] or [WIRE] ? 1 DISP -1 WAIT @ select IR-control or WIRE-control. IF 11.1 SAME THEN -33 SF ELSE -33 CF END @ IR-control ELSE WIRE-control. 2400 BAUD CLLCD menu PICT STO { # 0d # 0d } PVIEW { # 89d # 1d } PICT OVER black GXOR OPENIO TICKS 'tmp' STO IFERR WHILE 1 REPEAT BUFLEN DROP DUP IF THEN TICKS 'tmp' STO SRECV DROP 'inln' SWAP STO+ ELSE DROP inln DUP DUP SIZE SWAP DUP 13 CHR 10 CHR + POS DUP @ crlf. IF THEN SWAP OVER 1 - 1 SWAP SUB PICT { # 0d # 59d } { # 130d # 63d } SUB @ scroll. PICT { # 0d # 59d } ROT GXOR 1 ->GROB PICT { # 0d # 59d } ROT GXOR 2 + SWAP SUB 'inln' STO ELSE DROP DROP DROP DROP END IF KEY THEN TICKS 'tmp' STO @ for IR control. 'k' STO k 55 == IF THEN OFF OPENIO TICKS 'tmp' STO ELSE k 36 == k 34 == OR k 25 == OR k 35 == OR IF THEN @ x position: 1; 23; 45; 67; 89; 111. @ y position: 1; 16; 31; 46. CASE k 36 == i 90 < AND THEN i 22 + 'i' STO END k 34 == i 22 > AND j 32 < AND THEN i 22 - 'i' STO END @ without [ENTER] key. k 34 == i 66 > AND j 46 == AND THEN i 22 - 'i' STO END k 25 == j 15 > AND THEN j 15 - 'j' STO END k 35 == j 17 < AND THEN j 15 + 'j' STO END @ without [ENTER] key. k 35 == j 32 < AND i 44 > AND THEN j 15 + 'j' STO END END i R->B j R->B 2 ->LIST PICT OVER black GXOR PICT ROT black GXOR ELSE k 51 == IF THEN @ [ENTER] key for confirmation. j 1 == IF THEN CASE i 1 == THEN 21 'k' STO END i 23 == THEN 22 'k' STO END i 45 == THEN 23 'k' STO END i 67 == THEN 24 'k' STO END i 89 == THEN 25 'k' STO END i 111 == THEN 26 'k' STO END END END j 16 == IF THEN CASE i 1 == THEN 31 'k' STO END i 23 == THEN 32 'k' STO END i 45 == THEN 33 'k' STO END i 67 == THEN 34 'k' STO END i 89 == THEN 35 'k' STO END i 111 == THEN 36 'k' STO END END END j 31 == IF THEN CASE i 1 == THEN 41 'k' STO END i 23 == THEN 42 'k' STO END i 45 == THEN 43 'k' STO END i 67 == THEN 44 'k' STO END i 89 == THEN 45 'k' STO END i 111 == THEN 46 'k' STO END END END j 46 == IF THEN CASE i 45 == THEN 52 'k' STO END i 67 == THEN 53 'k' STO END i 89 == THEN 54 'k' STO END @ Esc. i 111 == THEN CLOSEIO OFF 89 'i' STO 1 'j' STO END @ off. END END END @ keyboard table: AAABBBCCCDDDEEEFFF GGGHHHIIIJJJKKKLLL NNNOOOPPPQQQRRR SSSTTTUUUVVVWWWXXX YYYZZZ 777888999/// 444555666*** 111222333--- 000... +++ k 3 * DUP 2 + SUB IF XMIT NOT THEN END END END END END TICKS tmp - B->R 8192 / 600 >= @ after 10 minutes(600 seconds). IF THEN OFF OPENIO TICKS 'tmp' STO END END THEN flags STOF 9600 BAUD -33 CF 2 MENU CLOSEIO KILL @ Error: [ON] key trap. END >> >> >> Tal ==== I haven't analized your program, but looking at it, it seems well structured. However, you make intensive use of global variables, this slows down running time. Try to use local variables whenever possible. Stack handling only would be great, but it's harder to do . ==== I think I've seen this same problem in ver 1.10 where I did not see the problem in 1.18? ==== I am looking for an HP48G that is disfunctional but has a good LCD. If anyone has an broken calc due to ram upgrade gone sour, then I would purchase it for a reduced price as long as the LCD is good. I would also like to know if anyone has any experiences replacing a broken LCD on HP48G/GX. Is this more difficult to do than ram upgrade modification ? ==== You may go that route if you find a cheap enough 48 to work with. However Hp will still service your unit for a flat fee of $45. in the USA. You send it to their Hp repair and they will just send you a new or refurbished Hp48G+ if you have a 48G or G+. For a 48GX the repair might be higher. Ph#:970-392-1001 ext: 0 Their address is: Hp Service Center 1030 NE Circle Blvd. Bld#11 Corvallis, OR 97330 ==== I think my HP48 GX, complete with all of the cards I've purchased for it for use in my engineering work, is the greatest thing ever. With the demise of the HP calculator line and the use of handhelds on the rise, it only makes sense to create a full-featured emulator for Palm OS. While I already utilize an RPN pop-up calculator on my Visor, the functionality of the HP 48 is sorely missed. The Palm platform could offer a great deal of benefits to the current HP 48 user -- an infrared port, a serial port, a faster processor, larger screen with superior resolution complete with color, and, in the case of the Visor, an expansion slot where HP ROM cards could be used. Finally, it would allow those of us who would like to retain the HP's quality in an all-in-one handheld option. I couldn't help but notice that EMU48, an excellent HP 48 emulator for the Windows Platform, comes complete with source code (in C++, I believe). Has anyone considered porting it to Palm OS? Has it already been done, and if so, where can I find it? ==== I would know how to read and write in port 1 or 2 of the HP49G with the ASM langage ? ==== You can access any flash banks with ACCESSBankX, where X is 0 to 15. IIRC 8 to 15 are user banks (aka port 2). If you set P=0 before calling these routines, they'll switch to the bank; if you set P=1, they'll switch back. You can access the bank in memory at #40000h-#7FFFFh. For port 1 there are ACCESSERAM1 and ACCESSERAM2, I don't know about these but I suppose it's the same. Writing is a bit more difficult, especially for the flash ROM... Thomas ==== who can show me the equation where I can get all solution of an root funktion. real and irreal regeons for exampel 5 th root -31 greetings Reiner ==== Damit du alle L232sungen bekommst, musst du den komplexen Modus anschalten, am einfachsten mit LS&TOOL (Blaue Shift-Taste gedr237ckt halten und TOOL dr237cken). Ein C (statt R) im Header zeigt den komplexen Modus an. Du kannst die Gleichung mit SOLVE aufl232sen: Algebraischer Modus: SOLVE('X^5=-31',X) RPN Modus: X 5 ^ -31 = X SOLVE Auf meinem Rechner (1.19-6) ergibt das f237nf L232sungen. ==== If the definition of Y1(X) is 5 + X^2, alors trying to solve Y1(X) is similar. Do not worry about what is displayed ==== 1) I know that the last ROM revision for the 48gx is R. The one i found is a P. There is something i should know about the differences between P and R ? -- on www.hpcalc.org should be a description of the differences between the ROM revisions. The 'P' should have the newer GX display with better contrast, not the SX display. The latest built 'R' models from China have a different (better) display again. at least the port STO bug has been fixed. But more info will be on hpcalc.org;-) I see no problem there. If it's new (or unused), and the batts were outside the calc, there should be no problem. The only thing I noticed on my very first SX (Serial # 3003A00131) is that the rubber feet are slowly melting... But the keys are still very good. Chances are good that you won't use your SX again, ==== Has anyone thought of basing new hardware on the old hardware? A big part of why I like HP is the keys. What if you took a 48GX and used only the keypad, interfacing that to a target board, perhaps StrongARM based? You could write or find a Saturn emulator and then use the 48GX ROM image as a PC emulator does. You would need to convert keypresses into whatever input the ROM expects and also need to interface the display output to a new LCD. This all sounds difficult but I think it is still easier than developing a whole new code base for a calculator. A bonus is that a modern development board probably already has heaps of RAM (by calculator standards), serial port and flexible power requirements, perhaps even flash memory, USB and a PCMCIA or CF card interface. I'd really like a calculator with rechargable batteries that sat in a cradle like a Pocket PC, for syncing and for recharging. Perhaps there are better choices for the target hardware as far as power conservation go, but a LiIon battery should be able to power a device without a color LCD for a long time. == Whats the difference between 1.18 official ROM and 1.19 Beta ROM??? ==== Read this to see all the additions and bug fixes since 1.18. There are substantial improvements from the 1.18 to the most recent 1.19-6 beta ROM, especially when it comes to the CAS and SysRPL programming. http://www.epita.fr/~avenar_j/hp/49.html ==== ==== Run CASCFG before EVALuating the limit. The flag setup on EMU48 is: { #11800B8285010FF0h #0h #801000000A028088h #0h } which should be the default one. ==== you as my professor know that I'm not a newbee anymore :-) Of course, the first I tried long ago was to fix it ClrDAsOK. That doesn't help at all since, I guess, the ROM encapsulates the STARTOFF-program in a very specific way, with its own screen handling. Unfortunately, even NOSY is unable to dig deep enough in the ROM to discover where the STARTOFF program is handled :-) The problem remains. Even the banal program does not refresch the display if stored in STARTOFF. ==== Try SetDAsBad instead (ClrDAsOK only changes the Valid, Temp and NoCh flags). If that doesn't work it's a flaw in the STARTOFF handling -- I'd suspect it anyway, normally your sample doesn't need the Clr..., SysDisplay in the ParOuterLoop calls it just after restoring the display (so any program starts w/ the display area not OK). -- http://mein.hamburg.de/homepage/grendel ==== So far I've been looking into STARTUP, so of course I didn't see any problems :) I don't think there's any flaws in the way STARTOFF works. STARTOFF will be called if it exists instead of TurnOff after the idle time have reached a specified time (usually 5 minutes). STARTOFF will be called by GETKEY* inside GetKeyOb. Until you've pressed a key the display will not be refreshed as you're still inside GetKeyOb. For memory, the System Loop is like this: BEGIN AtUserStack ( signal a valid user stack; this word must be here at the top of the loop to insure non-standard entrances via restartol are considered ) SysMenuCheck ( do menu maintenance ) SysDisplay ( do the system display ) GetKeyOb ( wait for a key and convert it to a key object ) ERRSET ( prepare to trap key object evaluation errors ) DoKeyOb ( try to do the key object ) ERRTRAP ( if DoKeyOb errors... ) SysErrorTrap ( ...then clean up and continue ) AGAIN As you can see, you will never reach the SysDisplay until you exit GetKeyOb. And in the case of STARTOFF, you are NOT exiting it. It would be the same with a user alarm... The best way you can do, to refresh the screen, and that will be done ONLY once a key have been pressed, is by calling SetDA1Bad, SetDA2aBad, SetDA2bBad, SetDA3Bad at the end of STARTOFF which will force the system to redisplay the screen as soon as SysDisplay is called. If you know which area you are modifying, then call just the area that you want to update. If you want the screen to be redisplayed before a key is pressed, then manually call SysDisplay (but you will have to be careful, as it may not be enough inside a ParOuterLoop) SetDAsBad is not supported and is located in an area that can move easily. .. ==== 0posting.google.com&prev=/groups%3Fhl%3Den%26group%3Dcomp.sys.hp48 made me try such summations as SUM(n=0,2,SUM(m=0,n,a(m+1)*b(n-m+1))) where the list {2 4 0} is stored in variable a and the list {1 5 0} is stored in variable b. The HP49G can use indices for retrieving list or array elements, like a(1), a(2) etc. It does this slow, but it does it. But the sum above, doesn't get completely evaluated. You get an expression: 2.*b(n+1.)+4.*b(n-1.+1.)+.... As you can see, the elements of a for which the index of summation runs from m=0 to n, could be retrieved from the list a. But the elements of b, for which the index of summation contains n, whose value comes from the outer sum, couldn't be retrieved. At first I thought that the problem was, that the values of the index n, which runs at the outer sum, can't be used for the inner sum. But this can't be right. The sum: SUM(n=0,2,SUM(m=0,n,(m+1)*(n-m+1))) gets evaluated without problems. (Though very slow) Could it be that the outer index n can't be used as an argument of a function in the inner sum? Well, storing for example << -> m << m 1 + >> in a, and << -> n << n 2 >> in b, and then evaluating SUM(n=0,2,SUM(m=0,n,a(m+1)*b(n-m+1))) also works. When a and b are defined using algebraic objects, as in << -> n 'n+1' >> and << ->n 'n^2' >> the evaluation of SUM(n=0,2,SUM(m=0,n,a(m+1)*b(n-m+1))) also works. This shows that the value of the summation variable n of the outer sum is correctly retrieved in the inner sum. What is the problem then, when a and b contain lists and in the summation we access the list elements through indices like a(m) etc? Can anybody figure this out? ==== I am discovering that the HP has a bit more numerical functionality than the TI83. But I am wondering how the HP (like a 48g or 49g) compare to the TI92+? Specifically, I wonder if there is anyone who has used both a TI92 and a HP48 or 49, using both to almost their full capabilities? If so I wonder about these things: 1. A limitation of TI83 that bugs me is that I can't define a function to call a program for its evaluation. Say I want to evaluate some series to N terms. On TI83 I must write a program to generate a list of coefficients (say its L1) and powers of X (L2), then I can define a function like Y1=sum(L1*X^L2) This works, but it is clumsy. I want to be able to write a program which returns the Y-variable after doing its thing. 2. Also, for series, I wish I could do something like I can do in Mathematica: CoefficientsFunction[i_]= PowersFunction[i_]= NumTerms= Y1[x_]=Sum[ CoefficientsFunction[i] x^PowersFunction[i], {i,0,NumTerms}] I suspect I could do this on a TI92, but not sure. How about a HP? 3. Piecewise linear and nonlinear function definitions? I can do it on TI83 using logical tests like this: Y1=(X<0)X^3+(X>=0)X generates a cubic joined to a line at X=0. Any more elegant ways on TI92 or HP? Also, the TI then is limited to a value of zero when no tests evaluate to 1, but if you have parts of the function that are undefined, you can't really represent this. 4. Differential equations: Out of time. -- _____________________ As you say, the problem with multiplying by the domain is that the function goes to zero outside the domain, causing problems with vertical lines joining ends of functions to the next function or to the x axis. For a function like yours where the total function is continuous this is not a problem but for non-continuous functions it is. A better way is to divide by the domain. This means that the function is undefined outside the domain and not graphed at all. The only drawback is that you have to do each section as a separate function. eg. Y1=X^3/(X<0) Y2=X/(X>=0) I would imagine that this would work on a TI too. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== Many possible ways to do that on the HP49G: You could define: L1(n)=n/n! (Takes n, returns coefficient) L2(n)=n-1 (Takes n, returns power ) Y1(N,var)= sum(n=0.,N,L1(n)*var^(L2(n))) This is a bit slow when it executes. Y1 could also be defined as: << -> N var << 0 0 N FOR n n L1 var n L2 ^ * + NEXT >> which is much faster when it executes. Using the same definitions as in (1) with some little differences: CoefficientsFunction(n)=n/n! (Remains the same) PowersFunction(n)=n-1 (Remains the same) Store a number in N. Y1(var)= sum(n=0.,N,CoefficientsFunction(n)*var^(PowersFunction(n))) (No N argument in definition) This is also a bit slow when it executes. Alternatively you can define Y1 as: << -> var << 0 0 N FOR n n CoefficientsFunction var n PowersFunction ^ * + NEXT >> For the HP49G: First, you can include IFTE in algebraic objects. The piecewise definition would then be: Y1(X)=IFTE(X<0.,X^3,X) IFTE can also be nested in algebraics: Y1(X)=IFTE(X<-1.,X^2,IFTE(X>=-1. AND X<1.,-X^2,3*X)) The definition Y1=(X<0)*X^3+(X>=0)*X is also possible on the HP49G. The comparison X<0 is simply evaluated to 1 (true) when x is less than 0. When x is greater than 0, it evaluates to 0 (false). The result of this operation is multiplied with X^3 then. The same way is (X>=0)*X evaluated. Definition through a program is also possible: << -> X << IF X 0. < THEN X 3 ^ ELSE X END >> Do that :-) Greetings, Nick. ==== You do not really need to do each seperately. You could do: Y1=((X^3*(X<0))+(X*(X>=0)))/(not((X<0) and (X>=0))), but I want to do something with a discontinuity, so: Y1=(X^3*(X<0)+X*(X>0))/(not((X<0) and (X>0))). I have a tendancy to use parenthesies _very_ liberaly, to ensure that my equations are interpreted correctly. The 'not(and)' statement at the bottom of the fraction makes the function undefined whenever neither conditional is met, and the multipliers on the top make only the correct statement graph. In essence, at X=2, mine reads '(X^3*0+X*1)/(not (0 and 1))'. The 'not(0 and 1)' evaluates to 'not(0)', or 1. On the top, 'X^3*0' results in 0, and therefore, the expression simplifies to '(0+X)/1', which is, of course, the same as 'Y=X'. Anyway, the function you originaly gave matches up with both methods, but mine graphs one function and yours takes longer to graph 2. Mine also lets you trace along one continuous line. Yours is _a lot_ simpler to explain, though. Boolean logic is fun! Dylan Stewart ==== Depending on how sparse the non-zero coefficients are, it may be even faster to use Horners scheme, which obviates the need for calculating powers. (The drawback is that all coeffecients of 1, x, ..., x^n must be calculated, even those that are zero) << -> x n << 0 n 1 FOR i x * i L + -1 STEP >> where L is the name of the function f(n)-> c_n - hans kristian - ==== I have used the TI-89/92+/92 for more than five years. I have used the HP49G emulator, and that too not very frequently (simply because I don't need to :) I don't recall whether the TI-83 supports functions (similar to programs), but the TI-89/92+ do. In fact, most of the things I have written are functions (in both TI-Basic and C). Yes, of course you can do this. One way is to use the part() function. You can also find the coefficients by differentiating. when(x<0,x^3,x) The fourth argument of when() is what is returned when the condition does not evaluate to either true or false. So, to graph: x^3, x>1 x, x<0 undef, otherwise use: when(x>1,x^3,when(x<0,x,undef)) If you prefer using the method of multiplying by conditionals, use division instead of multiplication to represent undef. For example, x/(x>1) will be undef when x<=1. :) Use the great DiffEq by Lars Frederiksen: http://www.perez-franco.com/symbulator/download/am.html Bhuvanesh. ==== You're confusing me :) What's 'n'? The polynomial? It can't be. Bhuvanesh. ==== Yours is an interesting variation but it has a number of disadvantages. Firstly it is quite difficult for a student to understand. Secondly, and more importantly, it doesn't work as well for non-continuous functions. For example, if you try to use it for X/(X<1) and (X^2-1)/(X>=1) then you'll find that the discontinuous ends of the function are joined by a jagged vertical line which should not be there if it is a proper graph. (At least, they are on my HP39G) If you use my method then the graph looks exactly as it should. -- Colin Croft Old mathematicians never die; they just lose some of their functions. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== ==== n is the series term index or number. ____________________________________ Christopher R. Carlen Principal Laser/Optical Technologist Sandia National Laboratories CA USA crcarle@sandia.gov ==== this is the manual for the CAS (pdf): http://www-fourier.ujf-grenoble.fr/~degraeve/cas49_1.pdf it is writen for algebraic mode, but it will give you an idea of what the commands do. best of all, it is free! ==== It is not difficult to write a program to find an IRR. (Use Newton's method of finding a root of a polynomial.) Sometimes, when cash flows are a mixture of positive and negative values, there may be multiple IRRs. The number of possible IRRs is bounded by the number of sign changes in the series of payments. I am interested in how the 17BII handles multiple IRRs. Does it report all IRRs, one IRR, or none at all? On 15 Jan 2002 04:53:46 -0800, jm_dl.blenner@bigpond.com (Martin ==== Don Phillips' FIN 4.2.1 doesn't do the job? http://www.hpcalc.org/viewzip.php?id=191&file=FIN421B.TXT To quote the document I link to above: The 11 programs are EQV, TVM, capital budgeting (IRR, NPV, NUS, NFV), generalized IRR and NPV calculations, bonds, stocks, depreciation, options, inflation adjustments, simple interest, and discounted securities using the GDP or CPI indexes. Doh... I guess FIN doesn't work with a 49? Maybe it's easily portable? ==== Don, your statement about Cashflows potentially producing more than one IRR is correct however, the 17BII does not have the feature built in to produce MIRR's. Personally, I have not used the 17BII to do MIRR's in my work apart from varying the input Cashflows (using sensitivity variables), but don't doubt the capability of the HP17BII as a Financial Calculator. It has the Solver capability, which allows the operator to customise your own functions. I've used this capability in the past for a range of functions. In the beginning, it seems a little tricky, but once you've worked on it for a while, the power of the 17BII comes to the fore. I've found the manual really helpful. Only complaint is the Button Cell batteries don't last long when you're spending heaps of time on the Calculator. Martin. ==== Mike, thanks heaps. I missed this program on previous searches. The text file talks about heaps of features that I would love to use. Will try to load into the HP49G. Martin ==== Mike, thanks heaps for this. I have missed it on previous seaches. Will try to install on the 49G. Martin ==== No problem. Please let the group know if it works on the 49, and of course, make a backup before trying. :-) -Mike -- http://www.mschaef.com ==== Sorry, when I said root directory I meant the first page of your custom menu. LL ==== I am still puzzled why Wolfgang's problem would prevent the screen to be updated. The call to TurnOff calls itself ClrDAsOK. Also, running a *normal* RPL program on my machine that is doing something similar as what Wolfgang is doing has no problem with the screen update. I really don't understand here. I didn't spend too much time either, but still ==== Is there someone who could tell me what derXROOT(_,_,_,_) means? If it's possible, answer in spanish (As you can see, I'm spanish). ==== Cuando la 48 no sabe derivar algo, expresa el resultado de esa forma... El comando DERIV, en la biblioteca INTGR de ALG48 v4.2, deriva las ra222ces de 222ndice mayor que 2. Escribe con lo que sea. Trataremos de ayudar. Ra234l ==== When the HP48 must has to find the derivative of some function, for which it doesn't know what the derivative is, it denotes this with derFunction(Arguments). Arguments means here not only the arguments that you supplied to the original funtion, but also additional arguments. There is one additional argument for each argument of the original function. If you, for example, would like to take the derivative of the built-in function %(Y,X) with respect to X, you would have the result: der%(Y,X,0,1). The 0 is the inner derivative of the initial argument Y with respect to X. The 1 is the inner derivative of the initial argument X with respect to X. So der%(Y,X,0,1) means a function that depends on the initial arguments and the derivatives of these arguments with respect to X. If you want, you can define such derivatives. For XROOT you could do the following: Enter: derXROOT(Y,X,a,b)=INV(X)*XROOT(Y,X)/Y Press DEF. Now, everytime you take the derivative of XROOT your definition will be evaluated and you will get the defined function instead of derXROOT. If I remember well, this is also well documented in the manuals of the 48. Hope it helped you a little, Nick. ==== I recently updated to the 1/4/02 vger.flash ROM that updates a SOLVE bug. Every now and then the FILER refuses to open and hangs w/ the hourglass that stays forever. Only an ON+C would fix it. I'm just wondering if anyone experienced this using this or any other ROM version. With the original 1.19-6 ROM I never had this problem. I'm unable to determine what triggers this. I believe it happen after I save a variable using the shortcut LS+F# but I can't verify this for all the cases. Sometimes after a PINIT, the Filer starts working again, sometimes it doesn't (unfortunately this is currently the case). It is very strange, can anyone validate this? Diego. ==== Ooops.. really ugly, I have confirmed it in my emulator.. I think we need come back to 1.19-6 (old one) and wait to 1.19-7 that Jean Yves and Bernard Pairsse are working in.. J.Manrique CdU de la ETSIG ==== i vote for chucking the algebraic mode once for all. with less weight it'll run faster (he he). the algebraic mode was for marketing purposes and nobody uses it. so, why have it? (for algebraic V1.18 works great!) PS: if they made public the source, it would eventually end up being pure RPN (where HP shines). may be the V1.20 will be :) ==== Funny :-) I am at work now, where I have emu48 too.. and FILER works right :-) Have fun, J.Manrique CdU de la ETSIG ==== Yeeeeeeeesssssssssssssss. Most people that use algebraic mode don't know how to change the ROM version ;) A new age ;) Ricardo -- http://iespgirona.xtec.es/~rblasco/ Un222os, hermanos linuxeros 640 Kb de memoria son m207s que suficientes (IBM AT Designers - 1982) Bueno, espero que esto no aparezca en la versi227n final (Bill Gates en la presentaci227n de Windows Crash Debug '98) ==== I have a PEEK program which is a part from a contrast program. When I run #00101h PEEK I get a very long string. I would appreciate it if you could help me to get the contrast level in UserRpl from this stage since I don't know enough in ML. Tal ==== If the PEEK you refer to is mine, you need to enter 2 parameters : Address # of nibbles to peek Thus #00101h #1h PEEK should reply the proper data Paul ==== Here we go. I am an old HP-fan and I am looking for a well-preserved and working HP01 clock, but at no astronomic price, please. Can someone help me? I didn't know there was such a thing as a HP01 clock. -- David Haguenauer ==== There was never an HP01 clock... it was actually a watch. The limits of the language is not Uhr problem. Try: http://www.hpmuseum.org/hp01.htm Greg S. ==== You're sick, sick, sick!! :-) -- Colin Croft Old mathematicians never die; they just lose some of their functions. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== <3c43810a@news.svn.net> <20020115031740.7eba6827.hsamplon@lemel.fr> ==== David--- If I remember my German class correctly, the term for a wristwatch is ``armbanduhr''. It is likely a simple abbreviation---for instance, we no longer refer to our watches as wristwatches in common parlance. In any case, Marcus should probably look on eBay first and foremost. If not there, perhaps do searches for calculator and/or wristwatch collectors. Mit Vielen Glueck, Henry C. Gernhardt, III ==== some days ago I started using the program BKUP from Bob Apodaca for the HP 48. I have some minor problems, which I don't understand. I tried to I have a 48G which is upgraded to 768 kB RAM. I configured BKUP to back up everything, so I have: BACK UP PORT(S): { 1 2 3 4 5 } SAVE STACK & LASTARG: YES USE DEFAULT FILENAME: YES SAVE FLAGS: YES ARCHIVE HOME: YES PROTOCOL: HP48 KERMIT BKUP on the HP and it starts to back up the objects in the upper ports first. Some days ago ports 4 and 5 were empty, so it started with Port 3, then 2. Meanwhile I added some libraries to port 4. When I did a backup again yesterday, it started at Port 3 again and 'forgot' Port 4. I started the backup a second time, and entered the Port list again, but with the same result. The second problem is: It backs up the Ports first. Then, before backing up HOME (i.e. HPJAN14.BAK or some other date) it stops with an error message 'ARCHIVE error: Too few arguments'. Then I deleted the port list in the CNFG and tried again. Now it did back up the HPJAN14.BAK file. The file transmission itself works. -- Ralf Kleineisel ==== Are there any of those people who did so much great upgrade work in the early days still reading this newsgroup? (Or any electronics expert?) I still have my original SX, and I have no desire to get anything newer...but I do want to upgrade the memory. I have read all the memory upgrade material that I can find and have successfully added a 128K chip to either port. Now, I would like to have 128K on each port, without using two chips. I have an idea, but I would like to have an expert tell me if it would work. I have a 256K chip, therefore it has eighteen address lines A0..A17. would drive line A17 from CE2 through a tri-state driver, e.g. 74244, something like this: CE1---------------+--CE | | Vcc-------| >-----+ |/ | CE2--------| | | Vcc-------| >--------A17 |/ Now, my only knowledge of HP48 and memory chips comes from reading old postings, but from what I see this looks like it would work. Can anyone tell me for sure before I waste my time on it? The things that worry me are: 1) When CE2 is driven high, is CE1 driven low? If so, this would create high current from Vcc to CE1 through the driver. Could that be fixed with resistors? I don't see how. 2) What happens when CE2 is low, so A17 floats? Can an address line float or must it be pulled low? I guess it wouldn't be too hard to add a pull-down resistor, but if it can float, that's better. 3) That there's a problem with the whole scheme that doesn't even occur to me because I'm inexperienced. Karl. ==== what I'm writing here was newer tested. Please use this information here at your own risk. CE1 and CE2 (both are active high) are normal Chip Enable lines, that means that it isn't possible that both lines are going high at the same time. For the RAM chip A17 line you can either use the CE1 or the CE2 line, it doesn't matter which one you use (at the RAM chip access CE1 and CE2 have an inversed signal). The CE of the RAM depends if the RAM chip-CE is active low or high (some RAM chips have both). Active high: CE = CE1 OR CE2 Active low: /CE = /(CE1 OR CE2) = /CE1 AND /CE2 Please remember to set both CDT1 and 2 (Card detection) input lines to high to indicate present and writable cards in both slots. Christoph Karl Nelson schrieb im Newsbeitrag ==== I have just uploaded the document to www.hpcalc.org until it shows up there you can download it at http://www.sv-kats.nl/MLTUT.zip -- This message was written with 100% recycled electrons Pivo ==== Mike I've looked at the programs on hpcalc.org and there are some basic Financial capabilities such as Present Value and Future Value of Annuities but, they do not extend to what I'm looking for. I would like to enter a series of Cashflows and be able to analyse them against a discount rate of interest for Net Present Value and similarly to obtain the Internal Rate of Return, etc as mentioned above. Yes, the 17BII does provide these functions quite adequately, but for convenience, it would be great to perform these types of analysis with the 49G. Martin. ==== I just recieved my 49G and although I like it I noticed that the key labels seem like they are going to wear off really fast, so I thought I'd ask someone who has had the calculator for a longer time. Do the key labels wear off (by sweat etc) or are they sturdy and just seem fragile ? ==== Petr Novak schrieb: I've had mine (ID93 but soft keyboard) for about a year and a half now, the key labels (and everything else) are still in perfect condition. Except the dust under the display cover. Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== Well my ENTER key is starting to wear off and I just noticed that today. I wonder if someone has found some sort of lacquer that can be painted over the keys to protect them. Luis ==== Mines about the same age, 18 mths. Lost most of the F on the F6 key, R is worn on the Enter and O on the ON key. Some of this wear maybe form sliding the cover on and off as I take the calc home every night. Also have a fine crack in the body above the screen. No dust under the screen though. ==== Mine is ID93 but the hardest possible keyboard. I have it now for 3 years. Key labels F1, F6 and ENTER are almost completely unreadable. I hope I'll be able to memorize the position of all keys, because if not, then I'll have also the game key hunt built-in in addition to the current menu hunt (thanks to J.H. Meyers who discovered this. ;-) ) Greetings, Hunting Nick. ==== I have to lists, called a and b. a = {2, 4, 0} b = {1, 5, 0} And this is what I wanted to calculate: SUM(i=0, 2, SUM(j=0, i, GET(a, j+1)*GET(b, i-j+1))) I tried it in the eqw, but it didn't allow the GET commando. I am new to programming with the hp49g and I tried several things, but nothing worked. Can anybody help? Thanx in adanvace! cu, Thomas Leitner ==== It should work if you use for example: SUM(n=1,3,SUM(m=1,n,a(m)*b(n-m))) But it doesn't! The evaluation of the index of the list elements in the inner sum, doesn't work. Be happy, you just entered the heavens of the bug finder! ;-) Use a program instead. Example: << 0 1 3 FOR I 1 I FOR J a J GET b I J - GET * + Greetings, Nick. ==== cu Thomas ==== True, GET isn't a function, and so can't be used in an algebraic object. Also, SUM (upper case sigma) doesn't like the function i as an index; just a guess, but maybe the index for a summation is a global variable rather than a local variable. Or maybe it's a flag setting? Also note that (n-m) evaluates to 0 in the first iteration, which doesn't work for the index of a variable. I used 'GS(m=0,2,GS(n=0,m,a(n+1)*b(m-n+1)))' which correctly EVALs to 36 on my 48SX and 48GX, but EVALs to '2.*b(m+1.)+(4.*b(m-1.+1.)+2.*b(m+1.))+(4.*b(m-1.+1.)+2.*b(m+1.))' on my 49G. I tried 'GS(m=1,2,GS(n=1,2,m+n))', and got the correct result of 12 on all of the calculators, so apparently the inner summation can access the index of the outer summation, but the evaluation of the variable index indeed fails to find the index of the outer summation on the 49G. Bug? New Limitation? Wrong flag settings? Maybe: << 0 1 3 FOR i 1 i FOR j a j GET b i j - 1 + GET * + NEXT NEXT >> James ==== True also, but you can use 'a(I)' instead. It is a local variable, but 'i' is a reserved word (can't be declared as a local variable). ==== board. Up until now to remove a library I simply go to ERAM or FLASH in the files menu select the library I want to remove and hit the purge button. Is this sufficient, and if not why not. Aubrey ==== It is. -- This message was written with 100% recycled electrons Pivo <3C4545EA.69DE3E2E@student.tnw.tudelft.nl> ==== Peter Geelhoed schrieb: Except if the library stores some information in the hidden directory and you want to get rid of it. Usually there's a Purge Configuration command inside the library just for that purpose. Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== I forgot about that. My hidden directory is very clean though, since TTRM's clear it regularly :-) -- This message was written with 100% recycled electrons Pivo ==== Are all the HP Web Rings die ? Reagards Eric ==== Oh no! Keep living somewhere in the twilight zone. Greetings, Nick. ==== No because, i would launch a HP webring and i would ask to some persons if that s a good idea ? If ther are a lot thats not useful, but i can have the help of HP Network to launch it :) Eric Nick Karagiaouroglou a 216crit dans le message de ==== Eric schrieb im Newsbeitrag The HP48Ring is still alive, use http://www.webring.org/cgi-bin/webring?ring=hp48&;list to view the list of sites in the ring. Christoph ==== Had a HP sale burst here in the Office. 3xHP32S, HP49 and a HP48G+. Any chance of a kickback. Apparently there are no more 32's in Australia, the only HP RPN scientific calculator model. Also the 48 was bought because the retailer insisted there was no 49. ==== I know that HP at one point used a two color mold process to form key lables that were pretty much indelible. I was wondering if HP's newer calculators were similarly made, as I'm perhaps planning on buying a 17BII or a 12C. I'm sure the 12 at one point used the molded keys, but am not so sure about the more recently produced versions. Also, if it did change at some point in time, is there an easy way to see if a machine was made with molded keys? If I get stuck trying to buy an older machine on ebay, that might become important. Mike -- http://www.mschaef.com ==== you might check www.hpmueum.org about molded keys or not;-) AFAIK all HP calcs are made with painted keys now, I have a NIB 12C and a NIB 32SII, bought some years ago, and even these machines have painted keys;-((( Ot the 32SII the legends/printings are so bad compared to the older version w/molded keys, I would not buy this model again with painted keys. I have both, so I can compare easily;-) The 12C kbd & legends are not much better, and the keys 'feel' not as solid as on my U.S. made 12C... But I have a NIB 42S made in 1999, which seems to have molded keys. The models with rubber keys all have the symbols painted on. My suggestion: Try to get a used one with molded keys. It's simply easier to read the symbols. Raymond MSCHAEF.COM schrieb im Newsbeitrag ==== The HP49 has a whole new set of entries for drawing text in the screen (or elsewhere). These are entries such as LEFT$3x5, LEFT$3x5Arrow and others. LEFT$3x5 seems to work like CENTER$3x6 already existing from the '48. Is that true? Is there any difference? And what about the other entries? I couldn't figure out their use. Could someone explain them? -- Eduardo M Kalinowski http://move.to/hpkb ==== They do work in a similar way. I've sent the whole description of the entries to Carsten and it should be on his entry points web site .. ==== I have created a library using the neat program LBMKR of Marco Ciolfi. Is it possible to decompose a library back into it's original components ? I suppose I don't really understand what the CRLIB command is physically doing, hence this question. Aubrey. ==== Yes, you can convert libraries into directories. There are a number of programs that do it. Take a look at this page: http://www.hpcalc.org/hp49/programming/libs/ Since you are a newbie, OT49 probably suits you best. -- This message was written with 100% recycled electrons Pivo ==== Well, i don't use the InformBox Generator of debug2 anymore, because i had a lot of problems with it. One of them was the handling of errors. So i use the text version (IfMain) and that works fine. Doesn't matter if you choose Mini font, because the font is handled in the message IfMsgGetFieldGrob. Also no problem, the field is a text field so no Choose related message send. That's right, the general messagehandled where to put a new menu, and to handled a OK and Cancel. I will consider it, if the time is there Pieter ==== When I ran a program (Library) which had never given me problems before, I had a memory reset. I reloaded the library and had a memory reset again at the same point. So I had a look at the directory from which the library had been created and at the point where the problem occurred there were three Externals which I do not remember placing there. I used John Meyers PACK-> to discover which sysevals had generated these Externals and they were merrily converted back to the format: #XXXX h SYSEVAL (three of them) Unfortunately I have already edited the program to its original form so I cannot tell you what these entry points were, but I looked in the XREF table and they do not exist. The only thing I can tell is that they had taken the place of a local variable. This local variable is used in other places in the program and I don't think the same substitution has occurred. I have V 1.18 installed in my machine. Does anybody have the foggiest idea what is going on? Luis ==== You must have run a byggy program prior to this one. A buggy program running wild can easily change something in memory - also stored objects. ==== I need to convert a program that was written for a HP48GX to my HP49G. It is mainly written in User RPL with a little System RPL. The program is Don Phillips' Finance 4.2.1 B which is on the hpcalc.org site. Can someone assist please? Martin Blennerhassett. ==== Sorry for the delay - one week without web access. I guess we've got to try and play. Bye. HPCC member #1046 - ==== since i don't know (if it is possible at all) how to operate the x48-emulator at HP-original speed, i'd like to ask if the autorepeat function of the calculator can be turned off so that i could delete one character at a time? TIA, Oliver -- Theoretische Elektrotechnik 2-06 TU Hamburg-Harburg Harburger Schlossstr. 20 21079 Hamburg ==== Well, the HP48 has no available option to turn the autorepeat of the backspace key off. Greetings, Nick. ==== i'd turn off autorepeat for all keys - that would work for my purpuses. TIA, Olivier -- Theoretische Elektrotechnik 2-06 TU Hamburg-Harburg Harburger Schlossstr. 20 21079 Hamburg ==== Sorry, I didn't make that clear. The HP48 does not have the option to turn autorepeat off for any key. This is true for the machine itself, but I can' remember if the x48 emulator has this feature. Greetings, Nick. ==== ok, now i got it; a already had a look at the x48 source-code, but i hoped to get around modifying it. Since the x48-emulator-package has not changed since years (fast computers too), the job is probably not that easy. But maybe i'll try if i find the time. -- Theoretische Elektrotechnik 2-06 TU Hamburg-Harburg Harburger Schlossstr. 20 21079 Hamburg ==== Interesting concept... may be theoretically possible, though perhaps with some functionality missing (e.g., card support sounds especially questionable). However, despite the faster processor and larger screen, Palm OS has some serious restrictions of its own - 256 KB heap, regardless of total memory; 4 KB stack, and 16-bit integers. Some of the heap space is used by the system so the full amount is in fact never available. Palm development tools like CodeWarrior and prc-tools are capable of compiling C++, but the C library is severely crippled and some functions just don't work the same on Palm OS (e.g., sprintf). Would definitely require a great deal of work. -- Andre Schoorl http://www.engr.uvic.ca/~aschoorl/ ==== it for the OS. <> I too own both an HP48GX and a Visor Platinum. I would love to merge these into one device, and I've researched a lot of software to to see the feasibility of this. Sadly, the most powerful RPN program I can find is RPN. There are other powerful calculators (Power One and CplxCalc), but none can remotely touch the power of my HP48 with the Erable libraries. The HP emulator can not be ported to the existing PalmOS because the ROM is so much larger than the Palm's stack space. Much as I dislike the interface of the PocketPCs, they do have the hardware and memory-addressing to handle the usual bloat-ware that accompanies it. So, there is an HP48 emulator for the PocketPC. Oddly enough, there seems to be more requests for a Matlab emulator than an HP48 emulator. The text interface eats less of the stack memory, and the cut&paste abilities of the PalmOS make this easy to use. Lyme (www.calerga.com) has the best rendition of this, complete with user-defined functions and plotting. In the meantime, perhaps we'll be able to look again at the HP emulator when the new PalmOS5 comes out with the new Arm processor. Until then, my HP is faithfully pushing me through my Master's degree! ==== Please excuse my ignorance if I'm making an obvious mistake here as I've never programmed in Java. But my understanding is that Java is supposed to be able to run on any machine with minimal software modification - so why has no-one ever written the emulator in Java and solved the problem of platforms for it once and for all? -- Colin Croft Old mathematicians never die; they just lose some of their functions. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== ==== You are quite correct in theory; however, in my opinion, the main issue in a Java emulator would be performance. (Unix OS, RISC CPU), the overhead of an emulator written in C roughly slows down the CPU clock speed by an order of magnitude. For example, my emulator runs a saturn CPU at an apparent speed of 8-10MHz on a 100MHz Alpha CPU, assuming there is no paging; RISC CPU in mind. Simple, interpreted implementations of the Java virtual machine like the ones to be found on portable devices are likely to slow down the emulation process by an additional order of magnitude; at least, these are the figures I obtained when I attempted to implement the main CPU ISA execution loop using JDK 1.0.2, Digital Unix platform. Things should be better if the Java VM supports advanced execution techniqus, like code mutation, just-in-time compilation, and so on, but I have never checked this, so I cannot estimate how much speed could be gained and I don't know if these optimizations are feasible on a portable device. Does anyone have any clue on this? Ivan -- ------------------------------------------------------------------------ Ivan Cibrario Bertolotti / phone: +39-011-3919246 IRITI - National Research Council / fax: +39-011-341882 ==== Thx Thomas Aber ich mochte schreiben im port 2.... Where did you learn tahr is ther a doc ? Eric 3C432C9F.5520B667@iname.com... ASM <3C432C9F.5520B667@iname.com> <3c446e0d$0$192$626a54ce@news.free.fr> ==== Eric schrieb: I've read somewhere on hpcalc.org that HP has not released any information about *writing* to the flash. Other Ps in the ACCESSBankX routines have other meanings, but I don't know them, maybe somebody else can tell us? In general, for documentation you should check hpcalc.org. There's also the new entry database, compiled by Carsten Dominik, which lists a lot of entries (both RPL and ML). The URL is: http://zon.astro.uva.nl/~dominik/hpcalc/entries Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== Hallo Thomas Rast, danke f237r die schnelle Antwort. Genau das war die Frage wie man z.B. bei einer Wurzel s212mtliche L232sungen im komplexen Zahlenbereich erh212lt. So erh212lt man z.B. bei einer 5. Wurzel immer 5 L232sungen. Wo ist denn diese TOOL Taste ? mfG Reiner Oesterle Thomas Rast schrieb: <3C44B774.91617E63@t-online.de> ==== Reiner-Oesterle schrieb: Angenommen, du hast einen 49G, dann ist sie gleich neben den Pfeiltasten. F237r die 48er brauchst du wahrscheinlich zus212tzliche Software, damit du solche Gleichungen l232sen kannst, und 237ber die 39/40er weiss ich zuwenig. Es kommt also darauf an welchen Rechner du hast. Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== If you keep using your RAM card into Port 1 and not into Port 2, then you will not see any difference. I still have my P version bought many years ago. The *new* R version that have been made in China (from last year) have the same screen as the HP49G. This screen has a much better contrast, however, I think for gaming purposes the old blue screen of the earlier HP48G was better, and the SX screen (still for games or pictures moving quickly) was even better. The HP49G screen is very readable however ==== Well, i forgot two more informations: 1) i have an Equation Library Card. can i use it on a GX to add the Periodic Table and, eventually, what is non included in the gx's library? 2) just today, i found a shop which is waiting for a stock of GXs. Assuming they will be the latest revision (i.e. R rom, new display (China?), etc.) what is your suggestion? Should i buy the old P (kept in its box in the shop for years, but having the _old_ HP quality) for 191 euro (about 212 US $) or the new R (i heard something about the painting on the keys and, in general, about poor quality of plastics and assembly) for 226 euro (about 251 US $). Could i have some warranty problems buying an old calculator? -- Marco Polo marco.polo@katai.cina ==== 191 euro (about 212 US $) Are you sure about this exchange? I thought 1¥=0,90$ (approx.) ==== Marco--- Huh! Here in the US, I can order a GX direct from HP for something like $150 or so. Good luck, Henry C. Gernhardt, III ==== ordering the calculator from an european dealer. The price of a new is lower than the average price of most italian shops. are greater and at least until 1999 there was an additional custom tax of 6.75% for foreign (outside EEC) electronic goods imported into Italy. I am in no way affiliated with them but, for example, I had good experience with both Cynox (http://www.cynox.de) and Dynatech (http://www.dynatech.de). Have a look to http://www.hpcalc.org/buying.php for more information. Ivan ==== Oops!!!!!!!!!!! I badly converted Euro to $ (1 Euro=.9 $.....i made the contrary :-))) ) So 191 Euro = 171.9 $ and 226 Euro = 203.4 $ Sorry!! -- Marco Polo marco.polo@katai.cina Marco Polo ha scritto nel messaggio Should i buy the old P Could ==== We all (europeans) are newbies with Euro!! ;-) ==== UTool 2.5 is up on hpcalc.org. This version fixes a bug in the USIMP command which affected simplification of electric and radiation units. - Carsten