A269 Does anyone have the installation software/driver ? http://www.hp.com/cposupport/swindexes/hpcd-write10789_swen.html > Does anyone have the installation software/driver ? You really shouldn't need a driver for it. If you really do, check HP's website under support. If what you need isn't listed there your OS probably has a driver for a generic model as HP's are pretty popular. I know that linux will support most HP's. . > Does anyone have the installation software/driver ? I guess it's a bit different, but I have a QIC-80-style HP tape drive. It is not supported anymore under Win2k, so I had to reinstall Win98, just to read my old tapes! So, if this rewriter is a bit old, maybe you should have a look at older Windows as well... > You really shouldn't need a driver for it. If you really do, check HP's > website under support. If what you need isn't listed there your OS probably > has a driver for a generic model as HP's are pretty popular. I know that > linux will support most HP's. > . > Does anyone have the installation software/driver ? > With Sys I achieved very good, speedy programs > > Yes how did you manage to make the library splitter of librerias so speedy? :-) > > Does somebody know so that it serves the command MERGE in the hp49? > > It doesn't do anything but it ensures backward capability for 48-userrpl programs Were thank you for that of bookstores, in three months devoted completely to the hp49g and is it for that reason that I manage the system with some easiness, but do I lack it a lot, did I learn reading the good programs as emacs, sdiag, among other, but their program LIBEX cannot read it, is it for that reason my interest of learning how to program in saturn machine language, can you help me? Library v1.1 in the www.hpcalc.org Excuses for the language, I speak Spanish. > Yes. I used an HP 82164A HP-IL to serial interface to hook it up to my > PC, and a simple BASIC program using PEEK$ in a loop to read the ROM > then print the strings over the serial. Gosh.. How long did it takes ? I remember doing something similar for an HP28S (modified to have a link with my Apple II serial port), it took me ages to get the ROM right... Yes. I used an HP 82164A HP-IL to serial interface to hook it up to my > PC, and a simple BASIC program using PEEK$ in a loop to read the ROM > then print the strings over the serial. Gosh.. How long did it takes ? I remember doing something similar for an HP28S (modified to have a link > with my Apple II serial port), it took me ages to get the ROM right... But this really depends on the hardware and software you use. I actually dumped a HP42S ROM (64KB) to the PC today to verify a receiver hardware in about 16 min. and that's near to the theoretically maximum. (3 + 12 * 2 + 3) * 427.25us = 12.8 ms / per character -> 780 baud I used an assembler program dumping the content in binary form with some block information and checksum similar to the Xmodem protocol to have the chance of detecting errors and get information which part must be redumped. I done the complete upload twice so far, but there was never a need of redumping parts. (65536 / 128) * 132 = 67584 number of bytes with protocol overhead -> 67584 * 10 / 780 = 866sec = 14.4min <-> 16min from top BTW, the batteries of the HP42S after two dumps are still alive. But in one case I agree, using BASIC or UserRPL for the main loop will slow down upload rapidly. The other big point is, that many receivers don't accept binary data, so you must send twice the data size. So where are the assembler freaks modifying the HP INPRT program for the HP48 that it's accepting binary data and transmitting the received data over the serial port to a PC? Christoph One of the original authors has just updated the website, with some explanations, links and also ideas for brave developpers out there... http://hpcomm.sourceforge.net/ Gerald. > Wartan> have sources, seem to no longer exist (www.hpcomm.org). Anyone > Wartan> has got the sources? I'd be very thankful. Wartan> Ah, found it. hpcomm on SourceForge. Surprise, surprise. Hmm. Too early to be happy. For some mysterious reason, sources > can't be downloaded from the site, only binaries.. > And what's even stranger, is that the sources are placed > exactly in the same place as binaries, on SF, but when I try to get > them, I get a Connection Refused. So, the question arises again: anybody have the sources? -- > Computers are not intelligent. They only think they are. Good, wanted to please know like I make a warmstart if to have to press [ON]+[C] Doesn't it exist some form somewhere around of making this with commands in programming SYSRPL? Gustavo Portales > Good, wanted to please know like I make a warmstart if to have to > press [ON]+[C] > Doesn't it exist some form somewhere around of making this with > commands in programming SYSRPL? You can always call PTR 00000 to get a warnstart. -- This message was written with 100% recycled electrons Pivo That he/she should make a 19 year-old person that he likes the world of the computers that possesses a hp49, does it also possess a great economic misery, but that it doesn't possess a pc? That person I am that should make to learn, to part of studying books? Gustavo Portales :-( translator globalink Wow, there's a HP9100B going for, currently, $1,500 USD on eBay: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=1352137422 (and, no, I'm not the one selling it and trying to get the price up even higher!!) That's more than even the Xpanders went for! Roger Metcalf > I called it Final madness lol :) > ...stupid, but it sounds really funny: > > ****************************** > > 1 > WHILE DUP 100 <= > REPEAT > 1 + > RAND 2000 * .05 BEEP > 100 .05 BEEP > 1000 .05 BEEP > 500 .02 BEEP > END > DROP > > ****************************** > Reminds me of when I had pac-man on my watch when I was a kid. Remember those watches? I think they had pole position and asteroids also. :o) > //frogfot I've managed to collect a few second hand hp48 series calcs (mostly thgouhg trades), and none of them have come with manuals. I have the complete manuals for my 28c, and have downloaded enough reference material on userRPL to get started, but I'm looking for actual 48 series user's guide type manuals. Does anyone know of a source online- so I don't end up having to spend 3x as much on manuals as I did on calculators? K >Does anyone know of a source online- so I don't end up having > to spend 3x as much on manuals as I did on calculators? K > In www.hpcalc.org you'll get what you are looking for. > In www.hpcalc.org you'll get what you are looking for. Yeah, I've browsed that site, even dowloaded a whole bunch of stuff. It's great for basic programming, software, and techinical data. I was sort of wondering about user's guide type stuff like navigating menus, using directories- hell, even finding directories. I haven't ever owned a 48 before and don't actually know the interface or basic operation of the calculator. I've found wuite a bit through trial and error, but it's not a veyr good way to do things. -K > Yeah, I've browsed that site, even dowloaded a whole bunch of stuff. It's > great for basic programming, software, and techinical data. I was sort of > wondering about user's guide type stuff like navigating menus, using > directories- hell, even finding directories. I haven't ever owned a 48 > before and don't actually know the interface or basic operation of the > calculator. I've found wuite a bit through trial and error, but it's not a > veyr good way to do things. -K You want the User's Guide, didn't you? http://www.hpcalc.org/hp48/docs/misc/hp48gug.zip > You want the User's Guide, didn't you? > > http://www.hpcalc.org/hp48/docs/misc/hp48gug.zip > I missed that one, then. tak! -K >Unfortunately, the name HOME is hard coded into the Filer and the stack >display engine. I never thought that people would have liked to change the >name so I didn't include the capability in my design. The header stack display is done entirely in ML, looking for the name of the >HOME command would have been quite difficult and would have slow down >everything. Displaying the content of VX was already slowing down enough the >display I don't think I understand you hear. You're saying that adding just a touch of generality makes the UI unbearably slow? Hmmm... Well the 48 uses the following for retrieving the name ( which you probably already know ): :: ' xHOME PTR 7E76 ( PTR>ROMPTR ) PTR 1616C ( ROMP>$ ) ; Using TIM from the Hack library I get a time of about 3.8 ms ( GX-R at 3.92 MHz ) . That's about 1/262 of a second. Even if you updated the display at grayscale frame rates the delay wouldn't be noticeable. If we simplify the above to : :: ' ROMPTR 2 22 PTR 81FB ( ROMPTRDECOMP ) ID>$ ; the time gets cut down to 2.7 ms or 1/370 of a second. Not what one would call slow. :) Perhaps you are referring to the slowdown incurred by the extra garbage collections due to the above. Well, we can write it in assembly to fix that. The following is adapted from PTR 81FB on the GX : ROMPHASH EQU #8479 C=0 A LC(2) 22 B=C A LC(2) 02 A=C A GOSBVL ROMPHASH ST=0 1 GONC + RTNSC + A=DAT0 A LC(5) =DOEXT1 ?C#A A GOYES + D0=D0+ 5 A=DAT0 A D0=D0+ 5 C=DAT0 A ?C=0 A GOYES ++ ST=1 1 D1=C GOSUB PC=C CD1EX ++ D0=A + RSTK=C D0=D0+ 5 C=DAT0 A D=C A CD0EX D=D+C A A=C A LC(5) #55 A=A+C A D0=A C=DAT0 A C=C+A A A=B A B=B+B A B=B+B A A=A+B A C=C+A A ?D<=C A GOYES error D0=C A=DAT0 A ?A=0 A GOYES error C=C-A A D0=C ?ST=0 1 GOYES + R0=C.F A C=0 A LC(2) #22 A=C A LC(5) #800F5 R1=C.F A C=RSTK P= 2 C=0 S GOSUB PC=C D0=(5) #800F5 RTNCC + C=RSTK RTNCC error C=RSTK P= 1 ?ST=0 1 GOYES + GOSUB PC=C + P= 0 RTNSC PC=C PC=C I get a time of , drum roll please, .53 ms or 1/1887 of a second. Slow? :D It's subjective whether the above is quite difficult . Maybe it's rather large but, in terms of effort I mean come on, I copied it out of the rom :). ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. > :: ' ROMPTR 2 22 PTR 81FB ( ROMPTRDECOMP ) ID>$ ; I hate to post such a silly remark but there should be a DROP after PTR 81FB. Bye. HPCC member #1046 - :: ' ROMPTR 2 22 PTR 81FB ( ROMPTRDECOMP ) ID>$ ; I hate to post such a silly remark but there should be a DROP after PTR 81FB. >Bye. HPCC member #1046 >- Oops ;) ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. On Mon, 13 May 2002 23:20:42 -0500, Jonathan Busby >I don't think I understand you hear. ^^^^ Loverly. Too bad they don't have a semantics checker ( yet ) ;) . I think my fingers have a mind of their own. ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. IÇm running out of paper for my HP-41Çs thermal printer (modell HP-82162A). While surfing through the net I found traders selling the original HP brand for about $20 the 6-pack . That seems to be a lot since I also have found traders of thermal paper rolls with identical specifications for about $50 the 100-pack. I just donÇt want to risk shopping 100 thermal paper rolls so they finally might not work or damage my printer. Now the questions: Has anybody made experiences with thermal paper of other brands than HPÇs? Does anybody know a dealer who sells them in small quantities (specially in the german room)? One place I have dealt with that has a huge selection of HP calculators and accessories is Calcpro www.calcpro.com. I have bought from them several times. In addition, I have had good dealing with its manager, Paul Nelson. I just talked to Paul Nelson today (Tues 5/14/02). Does anyone know if it is possible to get Fast3D's x/y/z parameters, so that I can copy them into WireFrame? That way I think it would be a lot easier to get a good perspective. Can be quite a hassle sometimes, especially because it responses quite bad to the Cancel key. Would be nice to make a draft with Fast3D and then plot it for real using WireFrame. Walter. Now, that is a good idea, the best I heard in ages! I will keep it in mind for the next time I touch this piece of the program. Does anyone know if it is possible to get Fast3D's x/y/z parameters, > so that I can copy them into WireFrame? That way I think it would be a > lot easier to get a good perspective. Can be quite a hassle sometimes, > especially because it responses quite bad to the Cancel key. Would be > nice to make a draft with Fast3D and then plot it for real using > WireFrame. Walter. > That really would be great! Most of the time I use Fast3D just for getting a quick mental picture of what a function looks like, but I don't want to spend too much time on it. Being able to let Wireframe redraw an interesting setting would be very nice. Walter. > > Now, that is a good idea, the best I heard in ages! I will keep it in mind > for the next time I touch this piece of the program. > Does anyone know if it is possible to get Fast3D's x/y/z parameters, > so that I can copy them into WireFrame? That way I think it would be a > lot easier to get a good perspective. Can be quite a hassle sometimes, > especially because it responses quite bad to the Cancel key. Would be > nice to make a draft with Fast3D and then plot it for real using > WireFrame. Walter. > I get only MySQL error messages ... normal (mostly). ttfn > I get only MySQL error messages ... > I get only MySQL error messages ... Works fine for me... - Carsten My problem is: I have a UserRPL program. Could you please describe me the steps to create an executable code, using ASMGX 1.1, or something like it, in a GX48? P. S.: I'm not shouting anymore. I guess that you'll have to learn ML ... DanieL. (Are User RPL programs, a problem?) > > My problem is: I have a UserRPL program. Could you please describe me > the steps to create an executable code, using ASMGX 1.1, or something > like it, in a GX48? > > P. S.: I'm not shouting anymore. Write an User program and compile it into executable code... is not posible (yet). Read Thomas' post: you must learn a new program lenguage. > > My problem is: I have a UserRPL program. Could you please describe me > the steps to create an executable code, using ASMGX 1.1, or something > like it, in a GX48? Well, if you have not already done that, you need to know the Saturn assembly language. There are a few documents around, you may want to start with Introduction to Saturn Assembly Language (F. H. Gilbert), which is available on hpcalc.org. The rest is easy ;-) You just write a string which contains your source code, and call the assembler. In Jazz it's called ASS, in MK ASM, whatever you use. 3 5 + > in UserRPL or - if at all - in SysRPL. Which is an entire chapter of its own. HTH Thomas > P. S.: I'm not shouting anymore. PS: I'm not joking anymore =) -- Thomas Rast If you cannot convince them, confuse them. -- Harry S. Truman Erico schrieb im Newsbeitrag > Why are you shouting? > Erico schrieb im Newsbeitrag Why are you shouting? Well, I have a simple program in UserRPL, <<5 3 +>, for example, and i need to compile that for assemble. I used compilers than jazz, RPLcomp, HPASMGX, but I can't make run. I need to know compile programs to make a project for academy. How to do? > Warning: Only works from fixed memory. I.e. store in a variable before executing. CODE GOSBVL PopASavptr R0=A.A D0=A GOSBVL SKIPOB AD0EX C=R0.A C=A-C.A RSTK=C C+10.A GOSBVL MAKE$N C=R0.A D1=C LC(5) DOCODE DAT1=C.A D1+10 $39 % LC(10) A=C.A A+10.A $8D % 10 nibs DAT1=C.10 D1+10 C=R1.A D0=C C=RSTK GOSBVL MOVEDOWN GOVLNG GPPushR0Lp ENDCODE No magic about it, is there? Run it on your UserRPL program of choice, and you get a Code object (that's what it looks like on the stack). Press EVAL, and it runs. Have a lot of fun =) Thomas PS: He who laughs last didn't get the joke. -- Thomas Rast If you cannot convince them, confuse them. -- Harry S. Truman Think about your question again.. And ask it again :) THANKS, ERICO How to use ASMgx? Please, give me a example or explain step by step. I found this to be quite interesting. I can't really imagine whipping it out during class, though! :-) http://education.ti.com/product/pdf/s02cares.pdf See page 8 of the document. Jason i can use my HP49G with an external PC keyboard. but! in my case, a PC keyboard can suck up to 250 mA and a wall transformer is needed for long periods of writing ( a rechargable 9V works OK too, but short life), thus adding volume and wiring. talking with a keyboard drains the batteries of the calculator fast because serial comms require extra power (remember the downloads?). the volume of a keyboard is not small either. portable external keyboards may not be a good idea for calculators if you plan to write long texts because of the battery issue. anyway, new gadgets are always good to have (more choices) :) > I found this to be quite interesting. I can't really imagine whipping > it out during class, though! :-) > > http://education.ti.com/product/pdf/s02cares.pdf > > See page 8 of the document. > > Jason I just finished a small utility for the HP48 that some may find useful. I'm not sure if it already exists, but I didn't find anything when I did a quick search on hpcalc.org The program simply removes the carriage returns that are on text files you type on the PC. So if you type a source file in a text editor, and then move it to the emulator / calculator, it strips off the CR characters at the end of each line. I know in my experience, it's incredibly annoying to have a PC text file mucked up with CR's at the end of each line (at least that's how TED displays them). The program is pretty fast since it's written in assembly. It took .42 seconds to fix an 11k source file with a CR at the end of each line. You can download it from my site: http://leviathan.orblivion.com/hp48/software/index.shtml -- Aaron AW> I just finished a small utility for the HP48 that some may find useful. AW> I'm not sure if it already exists, but I didn't find anything when I did AW> a quick search on hpcalc.org AW> The program simply removes the carriage returns that are on text files AW> you type on the PC. So if you type a source file in a text editor, and AW> then move it to the emulator / calculator, it strips off the CR AW> characters at the end of each line. I know in my experience, it's AW> incredibly annoying to have a PC text file mucked up with CR's at the AW> end of each line (at least that's how TED displays them). The program AW> is pretty fast since it's written in assembly. It took .42 seconds to AW> fix an 11k source file with a CR at the end of each line. There is a supported SystemRPL entry on both the 48 and the 49: KINVISLF $ -> $' 48G: PTR 3016B 49G: PTR 2F34D This entry removes the CR from the end of lines. It also translates the digraphs like -> into characters. So if this is what you want, or if your file does not contain those special escape sequences, this entry may be an alternative. Don't know how the speed compares, though. - Carsten CD> There is a supported SystemRPL entry on both the 48 and the 49: CD> KINVISLF $ -> $' CD> 48G: PTR 3016B CD> 49G: PTR 2F34D Correction: The stack diagram is different, and the description was KINVISLF $ -> $' $'' 48G: 3016B 49G: 2F34D String translation for transfer from PC to HP. Translates digraphs in the string to characters and removes (character 12) in front of newline characters. Which translations are actually made depends upon the current translation mode (the last number in the IOPAR variable) which can be set using DOTRANSIO. 0: No translation 1: CRLF translation 2: CRLF and characters 128-159 (80h-9Fh) 3: CRLF and characters 128-255 (80h-FFh) So in order to get only CRLF translation without touching the backslash escape stuff, use something like :: %1 DOTRANSIO KINVISLF DROP ; - Carsten That works perfectly. I tried it on the same file and it's a bit slower (about 2x as long), but it's still fast enough by all means. As I was already written. =) -- Aaron CD> There is a supported SystemRPL entry on both the 48 and the 49: CD> KINVISLF $ -> $' > CD> 48G: PTR 3016B > CD> 49G: PTR 2F34D Correction: The stack diagram is different, and the description was KINVISLF $ -> $' $'' > 48G: 3016B > 49G: 2F34D String translation for transfer from PC to HP. Translates digraphs in > the string to characters and removes (character 12) in front of > newline characters. Which translations are actually made depends upon > the current translation mode (the last number in the IOPAR variable) > which can be set using DOTRANSIO. > 0: No translation > 1: CRLF translation > 2: CRLF and characters 128-159 (80h-9Fh) > 3: CRLF and characters 128-255 (80h-FFh) So in order to get only CRLF translation without touching the > backslash escape stuff, use something like :: %1 DOTRANSIO KINVISLF DROP ; - Carsten AW> That works perfectly. I tried it on the same file and it's a bit slower AW> (about 2x as long), but it's still fast enough by all means. As I was AW> already written. =) Sure. Just trying to spread some knowledge about little known entries... And it might still be useful if you write UserRPL code this way and need the special characters... - Carsten I forgot to add that the program is called nocr. Here's a direct link: http://leviathan.orblivion.com/hp48/software/nocr.zip -- Aaron << 13 CHR SREPL DROP > should do it.... > I forgot to add that the program is called nocr. Here's a direct link: > http://leviathan.orblivion.com/hp48/software/nocr.zip -- > Aaron > I don't have an HP49 though.. and if I recall, the 48 string operations are rather slow (even sysrpl ones). I didn't even bother to try them. I don't even think there is a systemrpl equivalent for SREPL on the HP48. -- Aaron > << 13 CHR SREPL DROP > should do it.... I don't have an HP49 though.. and if I recall, the 48 string operations are > rather slow (even sysrpl ones). I didn't even bother to try them. I don't > even think there is a systemrpl equivalent for SREPL on the HP48. Yes there is. Just install String Writer, it's one of the command anyway, I'll just use it. I'm pretty accustomed to TED and it's built-in to Jazz so I'd hesitate to switch to a new editor. Besides, 24k is a lot of space for just one tool (SREPL) tha I would want. -- Aaron I don't have an HP49 though.. and if I recall, the 48 string operations > are > rather slow (even sysrpl ones). I didn't even bother to try them. I > don't > even think there is a systemrpl equivalent for SREPL on the HP48. Yes there is. > Just install String Writer, it's one of the command > Is it possible to make an input form in the hp49 such as when you pick an element from a choose box, the elements on a diferent choose box change? If it is, can any body tell me how should I set the message handler to do this? Marturo I think the main problem of the HP48GX ** at that time ** was the course of development of the new graphic engine (very very slow choose boxes etc). This can be read in some development journal (I think available at hpcalc.org) Caspar Timit.8e Hassan schreef in bericht > Could an EE very knowledgeable about computer technology of the early > 90's,tell me if it was possible by the time of the HP48GX release,to > release a much more powerful calculator(with FPU,faster CPU,higher > resolution screen,more memory and memory expansions > possibilities,etc...) at the price of the HP48SX ? > Because if it was possible than the fall of the HP calculators have > begun at the release of the HP48GX. > Let me be clear,the HP48GX was a great calculator but HP was targeting > both students and professionnals with the HP48G series,thus HP had to > keep these calculators cheap. > The consequencies are that the HP48GX was less powerful than it could > have been thus certainly pushing pottential buyers to look for > alternative like PC with more horsepower and which better fit their > needs. > What do you think ? > I think the main problem of the HP48GX ** at that time ** was the course of > development of the new graphic engine (very very slow choose boxes etc). > This can be read in some development journal (I think available at > hpcalc.org) Then the best/easiest solution for all these speed problems would have been to redesign the SATURN to make it faster while keeping software compatibility. A better SATURN with some kind of BCD FPU would have been incredible. > Caspar > > Timit.8e Hassan schreef in bericht > Could an EE very knowledgeable about computer technology of the early > 90's,tell me if it was possible by the time of the HP48GX release,to > release a much more powerful calculator(with FPU,faster CPU,higher > resolution screen,more memory and memory expansions > possibilities,etc...) at the price of the HP48SX ? > Because if it was possible than the fall of the HP calculators have > begun at the release of the HP48GX. > Let me be clear,the HP48GX was a great calculator but HP was targeting > both students and professionnals with the HP48G series,thus HP had to > keep these calculators cheap. > The consequencies are that the HP48GX was less powerful than it could > have been thus certainly pushing pottential buyers to look for > alternative like PC with more horsepower and which better fit their > needs. > What do you think ? > I think the main problem of the HP48GX ** at that time ** was the course of > development of the new graphic engine (very very slow choose boxes etc). > This can be read in some development journal (I think available at > hpcalc.org) Then the best/easiest solution for all these speed problems would have > been to redesign the SATURN to make it faster while keeping software > compatibility. > A better SATURN with some kind of BCD FPU would have been incredible. Caspar It's not too late to downscale the Saturn/Yorke CPU process from 2 micron to .25 micron (the cheapest current CMOS) with an eigth-fold increase in speed The new parts might be 3V, too. > It's not too late to downscale the Saturn/Yorke CPU process > from 2 micron to .25 micron (the cheapest current CMOS) > with an eigth-fold increase in speed > The new parts might be 3V, too. And who will do it? Who will finance this project ? There's no R&D left at hp for any handheld devices. This cost of project cost a fortune and takes about 18 months It's not too late to downscale the Saturn/Yorke CPU process > from 2 micron to .25 micron (the cheapest current CMOS) > with an eigth-fold increase in speed > The new parts might be 3V, too. And who will do it? Who will finance this project ? > There's no R&D left at hp for any handheld devices. > This cost of project cost a fortune and takes about 18 months > I guess you're so right about this one. The Carly-Fiorina Co. has no such intends (formerly:Hewlett-Packard, W.Hewlett was kicked out) We will never see the HP..err..CF 50G Cyclops... :Ç-( but maybe I get a SS 50G one nice day PS: If I manage to win 4,500,000 ?uros in lottery next Saturday here in Finland, I may do something.... >They could then take the existing integrated Saturn CPU chip, >Yorke, and downscale and downvoltage it >To go to 3V requires redesign on the power supplies >and some other components thus forcing a motherboard test >and some batch runs. Surely the CPU would go 8x >and use less juice on batteries. >I would rather have one AA alcaline than three AAA. yeap, i'll tell her it is for a printer. it may go through. there technologies now that all you got to do is 'inject code' into a program and a microprocessor design comes out :) they even recommend not to know anything about how it works and let the machine do its work. there are also processors that change their configuration! on demand (wuau, this is the future). some companies offer you to design your processor on the web and test it on the web without much knowledge of the internals (they even guess for you) and then you can try things out for free. after you have 'design' it and test it, you could buy (i don't know the price). these implementations are made in modern micros (fast). It's not too late to downscale the Saturn/Yorke CPU process > from 2 micron to .25 micron (the cheapest current CMOS) > with an eigth-fold increase in speed > The new parts might be 3V, too. And who will do it? Who will finance this project ? > There's no R&D left at hp for any handheld devices. > This cost of project cost a fortune and takes about 18 months > I guess you're so right about this one. > The Carly-Fiorina Co. has no such intends > (formerly:Hewlett-Packard, W.Hewlett was kicked out) > We will never see the HP..err..CF 50G Cyclops... > :Ç-( > but maybe I get a SS 50G one nice day > PS: If I manage to win 4,500,000 ?uros in lottery > next Saturday here in Finland, I may do something.... > there technologies now that all you got to do is 'inject code' into a > program and a microprocessor design comes out :) they even recommend > not to know anything about how it works and let the machine do its > work. there are also processors that change their configuration! on > demand (wuau, this is the future). some companies offer you to design > your processor on the web and test it on the web without much > knowledge of the internals (they even guess for you) and then you can > try things out for free. after you have 'design' it and test it, you > could buy (i don't know the price). these implementations are made in > modern micros (fast). X > It's not too late to downscale the Saturn/Yorke CPU process > from 2 micron to .25 micron (the cheapest current CMOS) > with an eigth-fold increase in speed > The new parts might be 3V, too. X They could then take the existing integrated Saturn CPU chip, Yorke, and downscale and downvoltage it To go to 3V requires redesign on the power supplies and some other components thus forcing a motherboard test and some batch runs. Surely the CPU would go 8x and use less juice on batteries. I would rather have one AA alcaline than three AAA. IIRC the poor guy programming that EQW in SysRPL (no ML) got a burn-out and left the company. PS: It's still not too late to release the Saturn CPU in 3V or less and use .20 micron process instead of 2.0 then release a HP 58GX being compatible with the 48GX but with 10x speed. or a 59G with true eraser keys ;-) > I think the main problem of the HP48GX ** at that time ** was the course of > development of the new graphic engine (very very slow choose boxes etc). > This can be read in some development journal (I think available at > hpcalc.org) Caspar >The HP48 GX has been released almost 5 years after the HP48SX letting >enough time to develop software for this newer architecture in higher >level language off course. HP48SX released March 13, 1990 HP48GX released June, 1993 Elapse time = 3 yrs, 3 mos. ---- not almost 5 years The HP48G's were essentially HP48S's with more memory capabilities, Equation Library card internal to the Calc, & forms for user-friendliness. HP49G released July, 1999 Elapse time = abt. 6 years. One of the comments that has been mention several times with the latest heated discusions has been that people wish that ACO had concentrated on a all-new calculator rather than coming up with the HP49. BUT - in 1998, for a new HP division, ACO needed to show that they could get a new product to the market for both the corporate folks and the customers. How long would've ACO lasted if they took to 2001 or 2 for a new product to the market? The big cheeses might have decided that they weren't doing anything and axe them. Or the customers may have decided to by TI since TI had the best calculator for the forseeable future? The HP49G was a smart product that required relatively small development effort. If it delayed the Expander any at all, it was only a few months - perhaps about 3 months. The problem was with the hardware changes. To this day, I can't get used to the feel of the buttons, compared to the HP48GX. John Edry > What do you think ? > > Why do people think that more { memory MHz pixels } means better > calculator? Because it does. More MHz mean a calculator faster to get the job done and sometimes even more usable thus better. > The GX is the best calculator that will ever be even though it is > old. It has proven it's value over a decade with an architecture > proven over many years. Well,its architecture is archaic now and i feel that it was even a bit outdated by the time of the HP48GX release as it was several years old. Don't forget that the HP48GX is basically an ehanced and faster HP48SX. I even see at least one significant inefficiency in the HP48G/GX design:the hardware memory management as even the old TI85 uses the > JY (and the rest of ACO): The 49G had the potential to become even > people fucked it up. You did a great job with the 49G... the rest of > HP did not... They did a great job with the HP49G but it is unfortunate it has to have such a outdated hardware. Anyway it didn't seem to be the main focus of the ACO. > > The HP48 GX has been released almost 5 years after the HP48SX letting > enough time to develop software for this newer architecture in higher > level language off course. > Backward compatibility could have been achieved with use of a SATURN > as co-processor just for this purpose a bit like the Playstation 2 > uses a Playstation CPU as IO processor but also for backward > > But if you're using two cpus how can you then achieve the same price as an > HP48GX alone then ? I haven't said as expensive as the HP48GX but as expensive as the HP48SX. One of the reason the HP48GX is as cheap was because it was an And the HP48G targets students. > You simply can't put two cpu like this on the same board, especially using > let say a 32 bits CPU and a 4 bits one. It should be possible as the SATURN and the other processor are not supposed to work as bi-processors. The other processor would be used as the main processor.The SATURN would be used only to execute SATURN code when needed,for backward compatibility. I don't see why it couldn't be done for a calculator if SONY was able to do it for the PS2. > The Saturn is one of the most expensive component (above $11). The calculator was not supposed to be as cheap as the HP48GX,you remember ? > Otherwise, you don't seem to realize the job, I doubt 5 years is enough to > achieve something as complete as the HP48GX Here we disagree. Isn't that you who said that at least 5 years was needed to built a calculator from scratch ? Wasn't the time needed to design both the TI92 and the HP48SX ? Moreover as the main processor would be much more powerful than a SATURN and as the O.S wouldn't have to be coded mainly in ASM,than 5 years should be enough to design such a calculator as or even more complex than even the HP48GX. > I don't see why it couldn't be done for a calculator if SONY was able > to do it for the PS2. Because the two sony's CPU are using the same architecture despite being different on the instruction. I don't say it's not possible the saturn CPU as a second one, just inot practicable. Also, if you go for a better CPU, I believe you will gain more speed using a software emulator, than using a spare Saturn. There are other problems that makes things difficult: The saturn runs in 5V, almost all other CPU in 3V. Using a 2nd Saturn almost means having everythign in double: power-supply controller, memory, data bus etc... Expensive that is. > The calculator was not supposed to be as cheap as the HP48GX,you > remember ? If the HP48SX was sold at the time the HP48GX was sold, it would have been priced like an HP48GX. A HP48SX was not more expensive hardware wise than the HP48GX. They just made more money out of it... Profit was huge on this machine. > Here we disagree. > Isn't that you who said that at least 5 years was needed to built a > calculator from scratch ? I didn't say so, I said TI took 5 years to design the TI92. And the TI92 is mainly based on the Derive engine. Just as the HP48SX is mainly based on the HP28S. > Wasn't the time needed to design both the TI92 and the HP48SX ? > Moreover as the main processor would be much more powerful than a > SATURN and as the O.S wouldn't have to be coded mainly in ASM,than 5 > years should be enough to design such a calculator as or even more > complex than even the HP48GX. 5 years is possible of course, it depends how much money you're willing to put in it. The HP48SX team was a big team compare to the HP48GX team... > > I don't see why it couldn't be done for a calculator if SONY was able > to do it for the PS2. > > Because the two sony's CPU are using the same architecture despite being > different on the instruction. I don't think so. The PS CPU is a relatively normal 32 bits Risc. The PS2 CPU is a 128 bits RISC with a very weird design. So there is at least as much differences(if no more)between these 2 CPU than between the SATURN and a 32 bits RISC CPU. > I don't say it's not possible the saturn CPU as a second one, just inot > practicable. Ok,i misread,sorry. > Also, if you go for a better CPU, I believe you will gain more speed using a > software emulator, than using a spare Saturn. Well,it shouldn't be easy at the time to find a reasonnable power consumption CPU able to emulate the SATURN. However there should have been several CPU able to highly outperform it and with FPU coprocessors or built-in. My main bet would be on a 68020 or 68030 with a 68882 FPU. > There are other problems that makes things difficult: The saturn runs in 5V, > almost all other CPU in 3V. I have read in this thread that the SATURN use an adapter so it shouldn't be a problem. > Using a 2nd Saturn almost means having everythign in double: power-supply > controller, memory, data bus etc... > Expensive that is. > The calculator was not supposed to be as cheap as the HP48GX,you > remember ? > > If the HP48SX was sold at the time the HP48GX was sold, it would have been > priced like an HP48GX. A HP48SX was not more expensive hardware wise than > the HP48GX. They just made more money out of it... Profit was huge on this > machine. Well,the HP48SX wasn't as cheap as the HP48GX even at the release of the HP48GX. I remember one of my friend complaining because he has bought a HP48SX a few weeks before the HP48GX release. He talked of a price difference of over $100. > Here we disagree. > Isn't that you who said that at least 5 years was needed to built a > calculator from scratch ? > > I didn't say so, I said TI took 5 years to design the TI92. And the TI92 is > mainly based on the Derive engine. > Just as the HP48SX is mainly based on the HP28S. There a lot of differences between the HP28S and the HP48SX. So many in fact that the HP48SX can be considered a whole new calculator. And even if the TI92 CAS was based on Derive,TI had to: -Design the hardware -Design the OS and apps other than the CAS and Geometry -Port the Derive engine and design the CAS which is somehow different than Derive -Port Cabri Geometre And they have contacted both Soft Warehouse and the Cabri Geometre designers in 1992 thus they have had less than 4 years to work on the ports. > Wasn't the time needed to design both the TI92 and the HP48SX ? > Moreover as the main processor would be much more powerful than a > SATURN and as the O.S wouldn't have to be coded mainly in ASM,than 5 > years should be enough to design such a calculator as or even more > complex than even the HP48GX. > > 5 years is possible of course, it depends how much money you're willing to > put in it. The HP48SX team was a big team compare to the HP48GX team... HP has alway had deep pockets(Even if it could changed in the close future if Carly continued her policy;)),if the managers wanted it could have been done. > I don't think so. > The PS CPU is a relatively normal 32 bits Risc. > The PS2 CPU is a 128 bits RISC with a very weird design. > So there is at least as much differences(if no more)between these 2 > CPU than > between the SATURN and a 32 bits RISC CPU. It may be a 128 bits CPU, it doesn't mean it's using a 128 bits external bus. It's a 32 bits one as well, so they can still share the same bus. > Well,it shouldn't be easy at the time to find a reasonnable power > consumption CPU able to emulate the SATURN. > However there should have been several CPU able to highly outperform > it and with FPU coprocessors or built-in. > My main bet would be on a 68020 or 68030 with a 68882 FPU. You have much better, and cheaper CPU than this. A 32 bits RISC SH cpu is cheaper than using a motorola one. You have to remember that the price of the CPU is not everything. You have to look at all the external devices you're going to plug to it. Using modern technology always put the price down. > I have read in this thread that the SATURN use an adapter so it > shouldn't be a problem. Yes it does as all the components plug to it have to use 5V technology. It's a reason why the flash on the HP49 costs so much. It's pretty hard these days to find any components running in 5V. > There a lot of differences between the HP28S and the HP48SX. > So many in fact that the HP48SX can be considered a whole new > calculator. In which way? All the the mathematical tools and libraries are the same, it's the main issue in a calculator, that's also where all the innovative software and IP is. It's not too difficult to design a user interface, re-doing a whole mathematical library is another thing > > I don't think so. > The PS CPU is a relatively normal 32 bits Risc. > The PS2 CPU is a 128 bits RISC with a very weird design. > So there is at least as much differences(if no more)between these 2 > CPU than > between the SATURN and a 32 bits RISC CPU. > > It may be a 128 bits CPU, it doesn't mean it's using a 128 bits external > bus. > It's a 32 bits one as well, so they can still share the same bus. In fact the PS chips has its own 32 bits memory. > cheaper than using a motorola one. I was talking of useing another CPU at the place of the SATURN when the HP48GX was released in 1993 or at the limit in 1994/1995. I don't think that there were sh3 at the time. If a CPU has to be used today for an advanced calculator than a Xscale or a ARM10* based CPU would be the best choice i think. I bet on the 68020 or 68030 and a 68882 FPU because this FPU supports BCD floating point. It would have been really interesting to see how faster a calculator is with hardware floating point BCD. > You have to remember that the price of the CPU is not everything. You have > to look at all the external devices you're going to plug to it. Using modern > technology always put the price down. I agree. > I have read in this thread that the SATURN use an adapter so it > shouldn't be a problem. > > Yes it does as all the components plug to it have to use 5V technology. It's > a reason why the flash on the HP49 costs so much. It's pretty hard these > days to find any components running in 5V. Damn,this SATURN is really an archaic processor and i am even too kind with it. Wasn't it possible to use an adapter just for the saturn and an intermediate chip to prevent other components to be plugged in this adapter ? Thus you could have used other components running at lower voltage. Btw,is it because of the Flash memory cost that you haven't used it in the HP40 ? > There a lot of differences between the HP28S and the HP48SX. > So many in fact that the HP48SX can be considered a whole new > calculator. > > In which way? All the the mathematical tools and libraries are the same, > it's the main issue in a calculator, that's also where all the innovative > software and IP is. It's not too difficult to design a user interface, > re-doing a whole mathematical library is another thing Sorry,i was comparing the HP48SX to the HP28C(i had one). There are much less differences between the HP28S and the HP48SX than between the HP28C and the HP48 SX,even if there are significant differences such as EQW,bigger screen,expansions slots,etc... X > Yes it does as all the components plug to it have to use 5V technology. It's > a reason why the flash on the HP49 costs so much. It's pretty hard these > days to find any components running in 5V. Damn,this SATURN is really an archaic processor and i am even too kind > with it. > Wasn't it possible to use an adapter just for the saturn and an > intermediate chip to prevent other components to be plugged in this > adapter ? > Thus you could have used other components running at lower voltage. > Btw,is it because of the Flash memory cost that you haven't used it in > the HP40 ? Yes, the cost would have be around 15$ more and the price would have above the sweet spot sayz the marketdroids (or where they bean counter clones?) > There a lot of differences between the HP28S and the HP48SX. > So many in fact that the HP48SX can be considered a whole new > calculator. In which way? All the the mathematical tools and libraries are the same, > it's the main issue in a calculator, that's also where all the innovative > software and IP is. It's not too difficult to design a user interface, > re-doing a whole mathematical library is another thing Sorry,i was comparing the HP48SX to the HP28C(i had one). > There are much less differences between the HP28S and the HP48SX than > between the HP28C and the HP48 SX,even if there are significant > differences such as EQW,bigger screen,expansions slots,etc... X Now I don't get it all ??? Have a look at the http://www.hpcalc.org/details.php?id=3012 where the features are compared. PS: I have not upgraded this again since the FILER in the 1.19-6 has so many new functions that there was no room for detailed explanations anymore (plus the new flag choose boxes - BASE as an example and...) > > There are other problems that makes things difficult: The saturn runs in 5V, > almost all other CPU in 3V. Then how come the HP49, for example, only uses 3 1.5V batteries? Doesn't seem like a big margin when the voltage starts to drop... / Magnus > > > There are other problems that makes things difficult: The saturn runs in 5V, > almost all other CPU in 3V. > > > Then how come the HP49, for example, only uses 3 1.5V batteries? Doesn't > seem like a big margin when the voltage starts to drop... > > / Magnus Ever here of DC-DC boost switchers for a power supply. example. input 0.9-4.5V output 5V. That is how you can run a 5V CPU on 1 to 3 1.5V batteries. Or even generate 5V 3.3V 1.5V etc from a lithium ion cell for a strongarm PDA. Lion batt typ 3.8V. Colin Because includded in the saturn is a switched regulator that generate 5V out of everything from 3v to 4.5v... this is the main reasnon for the big cap and self on the pcb... > There are other problems that makes things difficult: The saturn runs in 5V, > almost all other CPU in 3V. > Then how come the HP49, for example, only uses 3 1.5V batteries? Doesn't > seem like a big margin when the voltage starts to drop... / Magnus -Anonymous Would you mind telling us who you are? Thomas -- Thomas S. Strathmann http://pdp7.org RMS: NOOOOOObody expects the GNU Inquisition! Confess! Confess! Confess! > > -Anonymous > > Would you mind telling us who you are? > > Thomas Actually I am a little bit shy, and I really wish to stay at least sort of anonymous. I realize that people out there could track me down if that were their interest, but then they have to at least spend the time, and I am not worth the effort. However, to give a drop of my background. I have been programming since a very young age. Young meaning somewhere just under the age of 10. That gives me about 25 years of programming experience. I am what I would consider a good programmer, though I am not in any way one of the ultimate best. I have spent most of my programming career doing mindless dribble code, meaning I have been creating games. -Anonymous here it is: http://www.brighthand.com/newsite/features/jornada.html Have fun, J.Manrique Can not seem to assign this menu to a user key? Would like to do this for easy > Can not seem to assign this menu to a user key? Would like to do this for easy Assign this program <<14 MENU>> to the desired key where can you find a list of the menu #s? > where can you find a list of the menu #s? Navigate to select the desired menu and run RCLMENU. http://www.hpcalc.org/details.php?id=3026 > where can you find a list of the menu #s? Navigate to select the desired menu and run RCLMENU. where can you find a list of the menu #s? The mine (48) is in the Advanced User's Guide. I don't know if 49's menus are just the same... I think youu can download the 49AUR from hpcalc or from Area48 It seems that my new HP49G is slower than the 48GX. When I execute one program (userRPL) on 48 it runs faster.(up to 3x) Is there any reasonable axplanation? > It seems that my new HP49G is slower than the 48GX. > When I execute one program (userRPL) on 48 it runs faster.(up to 3x) > Is there any reasonable axplanation? Every number without a dot is treated as an integer. The integer in the HP49G has various lengths and is slower than the real. You must put the dot after every number (2 -> 2.) or open the program to edit in approx mode and save. slawek39@idea.net.pl It seems that my new HP49G is slower than the 48GX. >When I execute one program (userRPL) on 48 it runs faster.(up to 3x) >Is there any reasonable axplanation? The most probable explanation is that you perform operations in exact mode, which heavily involves CAS (computer algebra system) and so in some circumstances slows down the HP49G like you observed it. If you choose numeric mode, the run time should be comparable on both models. One of my standard benchmark problems for programmable calculatord (sum 1/n for n=1 to 1000) runs in UserRPL on the 48 in 13.96 seconds, whereas the 49 does it in 11.38 seconds, thus beeing 22% faster ;-) Hope that helps, ------------------------------------------------------------------------- Ralf Fritzsch Bundesanstalt fuer Wasserbau Federal Waterways Engineering and Research Dienststelle Kueste Institute - Department Hamburg ------------------------------------------------------------------------- Unix _IS_ user friendly - it's just selective about who its friends are. ------------------------------------------------------------------------- > One of my standard benchmark problems for programmable calculatord (sum > 1/n for n=1 to 1000) runs in UserRPL on the 48 in 13.96 seconds, whereas > the 49 does it in 11.38 seconds, thus beeing 22% faster ;-) How exactly do you run this program? Because when I let my 49 calculate this sum, it takes about 3 seconds. EVAL NUM->. Without the EVAL it takes a lot longer, likely the 11.38 seconds. Or do I miss the point? New to this forum (and to to 49), so please don't shoot if this is somehow a stupid question :-). Walter. >... > One of my standard benchmark problems for programmable calculatord (sum > 1/n for n=1 to 1000) runs in UserRPL on the 48 in 13.96 seconds, whereas > the 49 does it in 11.38 seconds, thus beeing 22% faster ;-) How exactly do you run this program? Because when I let my 49 >calculate this sum, it takes about 3 seconds. EVAL NUM->. Without the >EVAL it takes a lot longer, likely the 11.38 seconds. >Or do I miss the point? >... To have comparable results between various types of calculators (CAS and non-CAS), I forced evaluation in numeric mode (i.e. not exact), which yields a longer runtime (and, by the way, a less precise result). With HP49 & exact, one has [sum above]= -(Psi(1)-Psi(1001)) which after ->NUM is exact to all 12 decimal places, and without eval, where pure summation is executed, only 11 significant figures are obtained. ------------------------------------------------------------------------- Ralf Fritzsch Bundesanstalt fuer Wasserbau Federal Waterways Engineering and Research Dienststelle Kueste Institute - Department Hamburg ------------------------------------------------------------------------- Unix _IS_ user friendly - it's just selective about who its friends are. ------------------------------------------------------------------------- What if you give more memory to your Win XP Pro? Also; it's possible to start a process in high priority. AND there are other tweaks... > So what is it ? I use a dual-cpu Windows 2000 machine, and I find the dual CPU support > pretty good. > As a matter of fact, my dual CPU pentum 530 compile almost 10 times > faster > than the HPUX machine with dual 300Mhz PA-RISC CPU (HP9000 server). It > was > one of the most expensive HP machine at that time. And trust me, we sure > did > get angry at that Unix machine when I was the only user and it took over > 40 > minutes to compile the HP49 ROM while my poor Windows NT box with cygwin > could do it in 8 minutes. > I was doing calculus in MuPAD for Unix. I had an integral that took 3 days > to solve in Unix on an P4 1.4 with 768 MB Ram. The same running Windows > 2000 solved in 8 hours. Using Windows ME, 10 hours and Windows 95, 13 > hours. > Trying to make an gigantic plot. Unix: 3 hours(2x P600, 128 MB). Windows > 2000: 1 hour. Windows XP: 1 hour+20 minutes. When I suggested disabling the IR-led in a 39G in order to make it a 40G PS: I think that Helen is actually Carly Fiorina!!! classify him as the politest person I know. Even at his worst however How was I at my worst then ? :) Comments from different posts below. > X > I read about Control Data in the 60-70s where the VPs would run > very good people out of the company because those people would tell the > VP his idea won't work. Then the VP gets canned after a year or two. > Result=no good people in the company any more. This is exactly what > Carly is doing. X I have never ever fired even one person while acting as a VP in a company. VP Nousiainen PS: I think Helen is actually Carly Fiorina!!! > Your comment concerning JYA is interesting, by the way. It is > obviously true that the same people that complain about me being > rude or even insulting excuse the same and worse behavior by JYA > without even hesitating. A fascinating example of double standards. I've had my differences with JYA in the past. In particular, I was not impressed by the direction HP calculators have taken in general or with the 49G in particular. He hasn't always been very happy with my criticism of his baby. You, however, have been far more rude and insulting than I have ever seen JYA behave toward me or anyone else in this newsgroup. I respect him and his knowledge, even when I don't agree with him. As for your demeaning remarks about his age (referring to him as child), I see far more credibility in his opinions about HP than yours -- and I'm 47. -- Wayne Brown | When your tail's in a crack, you improvise fwbrown@bellsouth.net | if you're good enough. Otherwise you give | your pelt to the trapper. e^(i*pi) = -1 -- Euler | -- John Myers Myers, Silverlock LAST == LASTARG RATIO exists and is / eg. division CLMF is oppisite to 7 FREEZE EXSUB cannot be easily done what else? Give me your programs and I'll give you the translations (if possible) Anyone to program EXSUB, OBSUB ? Vice President of PKT, VP > How to port programs from one to other? So I will have 2x programs(48+28). > There are things like CLMF, RATIO, LAST, EXSUB and so on... CLUSR is > replaced by CLVAR. > Anyone to program EXSUB, OBSUB ? > Vice President of PKT, VP Well, could you describe what they exactly do? Nick. ''Algebraic' n @ level number of the algebraic hierarchy exploded to RPN of either A) an OBject or @ use OBGET B) an EXpression @ use EXGET The corresponding OBSUB and EXSUB made an SUBstitution instead of a GET with an extra input of an substitute.8f You should really have the HP 28 documentation (which was/is good) to get a grip 'cause I'm not good in this kind of explanations without the calc or the book at hand. Anyone to program EXSUB, OBSUB ? > Vice President of PKT, VP Well, could you describe what they exactly do? > Nick. Only one -> variable << at a time > I changed to RPN mode, and the problem disappeared... Is there a C compiler for the HP 49g? > Is there a C compiler for the HP 49g? The only one I know: http://sourceforge.net/projects/hp48xgcc/ As the name implies, it's gcc-based, and under GPL. Have fun, Gerald. To just to add: And the EQNLIB is missing from MSOLVR Fortunately we have an even better one: MSLV PS: I almost forgot the GREDUCE I would be quite interested about this units detail, the use of > units is one of my number one functionalities on the 48 so if the 49 > does not do them, it would be a No! > The HP49 Equation Writer doesn't handle units for editing, only viewing. > Otherwise, all the tools present in the HP48 (including the numeric solver) > are the same than in the HP49. > The new SOLVE commands handle units as well. > You can also use the old HP48 numeric solver (75 MENU) PS: I almost forgot the GREDUCE > GSIMP of Alg48 I would appreciate it if you could give me tip & trick in this remote control program to optimize it. I am sure that there is room for improvement and maybe this would speed up the program also. %%HP: T(3)A(D)F(.); << GROB 131 7 EFFFF8FFFF3EFFFF8FFFF3EFFFF8FFFF30FFFFFDFFFF7FFFFFDFFFF7FFFFFDFFFF70FFFFFDFF F F7FFFFFDFFFF7FFFFFDFFFF70FFFFFDFFFF7FFFFFDFFFF7FFFFFDFFFF70FFFFFDFFFF7FFFFFD F FFF7FFFFFDFFFF70FFFFFDFFFF7FFFFFDFFFF7FFFFFDFFFF70EFFFF8FFFF3EFFFF8FFFF3EFFF F 8FFFF30 -> bl << 7 RCLF -> j flags << { @ picture refresh (for arrow keys). 1 SF DROP @ key number drop. # 0d j R->B 2 ->LIST PICT OVER bl GXOR PICT ROT bl GXOR } 2400 BAUD # 130d # 63d BLANK { # 0d # 1d } ---TRAINS REMOTE CONTROL--- 1 ->GROB GOR { # 25d # 8d } FAST STOP SLOW DIR 1 ->GROB GOR { # 0d # 15d } LOC1 LOC2 LOC3 LOC4 LOC5 LOC6 1 ->GROB GOR { # 0d # 22d } LOC7 AUTO SIG1 SIG2 1 ->GROB GOR { # 5d # 29d } SW1 SW2 SW3 SW4 SW5 SW6 1 ->GROB GOR { # 5d # 36d } SW7 SW8 FAST2 STOP SLOW DIR2 1 ->GROB GOR { # 0d # 43d } LIGHT BELL TOYS WIND UTILS 1 ->GROB GOR { # 0d # 50d } 2ND QUIT UNDO RESET INIT. 1 ->GROB GOR PICT STO { # 0d # 0d } PVIEW { # 0d # 7d } PICT OVER bl GXOR WHILE 1 REPEAT #41F65h SYSEVAL @ WaitForKey (for a very low power consumption, accept [ON] key also). #1950Bh SYSEVAL @ UNCOERCE2 (convert to real number). DROP @ Now, the key number is in the stack. @ y position: 7; 14; 21; 28; 35; 42; 49; 56. { << > @ position=0, without matching number. << flags STOF 9600 BAUD HOME 2 MENU DROP2 DROP CLOSEIO KILL > @ [ON] key. << CLOSEIO OFF > @ [<-] key. << j 7 > << 'j' 7 STO- > << 49 'j' STO > IFTE 3 PICK EVAL > @ 11 = up arrow; 3 PICK Returns a picture refresh. << j 43 < << 'j' 7 STO+ > << 7 'j' STO > IFTE 3 PICK EVAL > @ 17 = down arrow. } { 45 29 11 17 } 3 PICK POS 1 + GET EVAL @ 3 PICK = key number. 1 FC?C << { @ = position0, without matching number. K 0 Q P @ FAST STOP SLOW DIR. A B C D E F @ LOC1 LOC2 LOC3 LOC4 LOC5 LOC6. G Y Z  @ LOC7 AUTO SIG1 SIG2 (= ASC(27)). 1 2 3 4 5 6 @ SW1 SW2 SW3 SW4 SW5 SW6. 7 8 + . - @ SW7 SW8 FAST2 STOP SLOW DIR2. O X U V W @ LIGHT BELL TOYS WIND UTILS. S N T H @ 2ND QUIT UNDO RESET ( = ASC(13)). @ not used. } { (7,1) (7,2) (7,3) (7,4) (7,5) (7,6) (14,1) (14,2) (14,3) (14,4) (14,5) (14,6) (21,1) (21,2) (21,3) (21,4) (21,5) (21,6) (28,1) (28,2) (28,3) (28,4) (28,5) (28,6) (35,1) (35,2) (35,3) (35,4) (35,5) (35,6) (42,1) (42,2) (42,3) (42,4) (42,5) (42,6) (49,1) (49,2) (49,3) (49,4) (49,5) (49,6) (56,1) (56,2) (56,3) (56,4) (56,5) (56,6) } j 4 PICK R->C POS 1 + GET @ 4 PICK = key number. DUP DUP + + XMIT DROP2 CLOSEIO > IFT END > Tal that you can change the languages on non English hp48gx's to english. I don't see any reference to this on hpcc, hpcalc, or the manuals. Anyone know about this? -K > that you can change the languages on non English hp48gx's to english. I > don't see any reference to this on hpcc, hpcalc, or the manuals. > > Anyone know about this? > > -K i don't think there are non-english hp48gx's > that you can change the languages on non English hp48gx's to english. I > don't see any reference to this on hpcc, hpcalc, or the manuals. Anyone know about this? Yes and its easy. use the ->LANGUAGE 1 is for English 2 for French These are the most complete languages. 3 is for Spanish if there ever be another ROM release, would it be possible to add the messages which are hardcoded into the calc (e.g. messages from IfMain, Filer) to $MESSAGE so that they can be translated to another language? TIA Andreas if there ever be another ROM release, would it be possible to add the > messages which are hardcoded into the calc (e.g. messages from IfMain, > Filer) to $MESSAGE so that they can be translated to another language? What does that mean? No messages are hardcoded, they are all into arrays, so you can always change them with an external library. There's not enough space left in the ROM to fully translate all the messages JY > if there ever be another ROM release, would it be possible to add the > messages which are hardcoded into the calc (e.g. messages from IfMain, > Filer) to $MESSAGE so that they can be translated to another language? > > What does that mean? No messages are hardcoded, they are all into arrays, so > you can always change them with an external library. > There's not enough space left in the ROM to fully translate all the messages Not all messages are in the message arrays. For example: (F:6DE19) :: FILE MANAGER TOTEMPOB TRUE PTR 6BC87 <-FPTR F F ClrDAsOK ; FILE MANAGER is not in a message array. Also the messages from the TYPES softkey in ^IfMain are not in a message array. If I translate the messages for the mode screen to german, this works well. But if I supply a wrong argument for the fraction mark and then press TYPES, I get a choose box [Valid objects Type: RealNum]. So my question is, if this messages can also be placed to the message array, so that one can translate them. Another questions: If I replace the message of the various inform boxes and store something in a port, this messages are lost and I have to rerun my 'SETMSG-Program'. Can you explain why this happens? For example: The mode-screen is empty, only the fields appear. This does not happen to the error-messages, for example DROP on an empty stack results in the translated error for 'To Few Arguments'. My solution at the moment is a loop which monitors the ports, but of course this slows down the calc (not very much, but noticable if there are a lot of objects on the stack). TIA Andreas Rip them out of the ROM !!! There! Now you have more room for improvements. A separate library like extable should do. Or maybe I should say: like the DEMO library because it should be pre-installed (naturally) The French students could then eagerly update their 40G models to the new French-speaking 49G. PS: Just to convince more: Showing a message is usually done when user attention is needed thus speed is not that important. > if there ever be another ROM release, would it be possible to add the > messages which are hardcoded into the calc (e.g. messages from IfMain, > Filer) to $MESSAGE so that they can be translated to another language? What does that mean? No messages are hardcoded, they are all into arrays, so > you can always change them with an external library. > There's not enough space left in the ROM to fully translate all the messages JY Even better: release them as a separate Library Data which is then installed in UserFlash just like the extable and with the OT49 anyone can split the data and do the necessary translations. Maybe even so that the language numbers should be handled so that one could have more than one language installed, but the both the ->|LANGUAGE|-> will work. Naturally French could get their calculatrice preloaded with the French Library Data only. :-D if there ever be another ROM release, would it be possible to add the > messages which are hardcoded into the calc (e.g. messages from IfMain, > Filer) to $MESSAGE so that they can be translated to another language? TIA > Andreas > that you can change the languages on non English hp48gx's to english. I > don't see any reference to this on hpcc, hpcalc, or the manuals. Anyone know about this? Yes and its easy. > use the ->LANGUAGE > 1 is for English > 2 for French > These are the most complete languages. > 3 is for Spanish I have a feeling that he was talking about the 48, what tipped me off was his mention of hp48gx :-) There are language libraries on hpcalc.org i seem to remember Spanish, Caltalan, Portuguese Italian and German. also try the hyde library They can be found in the util/misc/ section -- This message was written with 100% recycled electrons Pivo I expanded yesterday my equation-collection. Then I checked the new equations on the calc. I'm using the eqution-library from James D. Purdey (BTW: I dont't blame him or his libraray for this). My own equations have been stored in a list ('USEREQ'). By checking (solving) one of the equations the calc crashed. ON&C didn't work. A Reset wich the reset-button on the backside didn't work either. ON&F1&F2 works but either if I choose YES or NO when asked to recover memory the calc didn't came back to a usable state. If I reboot it with ON&D-Q I can see the Metakernel-logo (as usualy) but then there is nothing on the screen although the calc is running (I can change the contrast with ON&+ or ON&-). Pressing the backspace key by rebooting (with ON&D-Q) didn't help either. If I keep pressing the keys for some time the calc responds with a beep (Bufferoverflow?). But again there is nothing to see on the screen. ON&D was the only working thing so I downloaded the ROM again (same as I was using before: version 1.19-6). Download went fine but calc was still in same unusable state. Has anybody an idea what went wrong. It think it must be something software-related (probably some bad instructions in my equation-list). I didn't droped th calc or spoiled water over it. But as far as I know it isn't possible to destroy the calc with software. Roman This will happen if something went wrong with flashrom, so, You'll have to reinstall the flash. Try to do this: ON + + after reseting the calc and before you see the MK logo. Then try to update the flashrom. Something like that has happened with me ( I was playing with Debug49... ) so, the calc, when the MK logo has shown, restarts. Re-flashing the machine has solved the problem. I think that this could help you. DanieL. (Sorry for the bad english) > I expanded yesterday my equation-collection. Then I checked the new > equations on the calc. I'm using the eqution-library from James D. Purdey > (BTW: I dont't blame him or his libraray for this). My own equations have > been stored in a list ('USEREQ'). By checking (solving) one of the equations > the calc crashed. ON&C didn't work. A Reset wich the reset-button on the > backside didn't work either. > ON&F1&F2 works but either if I choose YES or NO when asked to recover memory > the calc didn't came back to a usable state. If I reboot it with ON&D-Q I > can see the Metakernel-logo (as usualy) but then there is nothing on the > screen although the calc is running (I can change the contrast with ON&+ or > ON&-). > Pressing the backspace key by rebooting (with ON&D-Q) didn't help either. If > I keep pressing the keys for some time the calc responds with a beep > (Bufferoverflow?). But again there is nothing to see on the screen. > > ON&D was the only working thing so I downloaded the ROM again (same as I was > using before: version 1.19-6). Download went fine but calc was still in same > unusable state. > > Has anybody an idea what went wrong. It think it must be something > software-related (probably some bad instructions in my equation-list). I > didn't droped th calc or spoiled water over it. But as far as I know it > isn't possible to destroy the calc with software. > > Roman I tried this of course before posting here. I reinstalled the flash-rom several times without getting the calc to work. I even changed the computer to make the update and installed some old rom-versions too (1.18). Didn't help. But anyway it works now. Have a look at my other posting. I had to erase the flash-rom completely. BTW my calc wasn't rebooting all the time it just booted after a reset with [ON]&[F4] -Q showed the metakernel-logo and showed then a complete empty screen although it was running (I could change the contrast). The only sign of live was the RS,LS and alpha-key-indicators if pressed for longtime. I couldn't turn the calc off also. But again it works now! have a lot of fun Roman Luiz DanieL schrieb im Newsbeitrag This will happen if something went wrong with flashrom, so, You'll > have to reinstall the flash. Try to do this: ON + + after reseting > the calc and before you see the MK logo. Then try to update the > flashrom. > Something like that has happened with me ( I was playing with > Debug49... ) > so, the calc, when the MK logo has shown, restarts. Re-flashing the > machine has solved the problem. > I think that this could help you. DanieL. (Sorry for the bad english) I expanded yesterday my equation-collection. Then I checked the new > equations on the calc. I'm using the eqution-library from James D. Purdey > (BTW: I dont't blame him or his libraray for this). My own equations have > been stored in a list ('USEREQ'). By checking (solving) one of the equations > the calc crashed. ON&C didn't work. A Reset wich the reset-button on the > backside didn't work either. > ON&F1&F2 works but either if I choose YES or NO when asked to recover memory > the calc didn't came back to a usable state. If I reboot it with ON&D-Q I > can see the Metakernel-logo (as usualy) but then there is nothing on the > screen although the calc is running (I can change the contrast with ON&+ or > ON&-). > Pressing the backspace key by rebooting (with ON&D-Q) didn't help either. If > I keep pressing the keys for some time the calc responds with a beep > (Bufferoverflow?). But again there is nothing to see on the screen. ON&D was the only working thing so I downloaded the ROM again (same as I was > using before: version 1.19-6). Download went fine but calc was still in same > unusable state. Has anybody an idea what went wrong. It think it must be something > software-related (probably some bad instructions in my equation-list). I > didn't droped th calc or spoiled water over it. But as far as I know it > isn't possible to destroy the calc with software. Roman Solved the problem. I had to erase the USER BANKS (8-15) manually in terminal mode. Lost all my stuff in port 2 of course but the calc seems to work again. It seems that some objects in port 2 had freeze the calc. But how can this happen? (And yes I pressed the backspace key during reboot!) Anyway my calc is back from death have a lot of fun (and no crash like this) Roman PS: what if something like this happens to a complete newbie? Roman Hartmann schrieb im Newsbeitrag > I expanded yesterday my equation-collection. Then I checked the new > equations on the calc. I'm using the eqution-library from James D. Purdey > (BTW: I dont't blame him or his libraray for this). My own equations have > been stored in a list ('USEREQ'). By checking (solving) one of the equations > the calc crashed. ON&C didn't work. A Reset wich the reset-button on the > backside didn't work either. > ON&F1&F2 works but either if I choose YES or NO when asked to recover memory > the calc didn't came back to a usable state. If I reboot it with ON&D-Q I > can see the Metakernel-logo (as usualy) but then there is nothing on the > screen although the calc is running (I can change the contrast with ON&+ or > ON&-). > Pressing the backspace key by rebooting (with ON&D-Q) didn't help either. If > I keep pressing the keys for some time the calc responds with a beep > (Bufferoverflow?). But again there is nothing to see on the screen. ON&D was the only working thing so I downloaded the ROM again (same as I was > using before: version 1.19-6). Download went fine but calc was still in same > unusable state. Has anybody an idea what went wrong. It think it must be something > software-related (probably some bad instructions in my equation-list). I > didn't droped th calc or spoiled water over it. But as far as I know it > isn't possible to destroy the calc with software. Roman On Wed, 15 May 2002 14:59:10 +0200, Roman Hartmann >Solved the problem. >PS: what if something like this happens to a complete newbie? A newbie never would have installed such adventurous libraries ;-) > I expanded yesterday my equation-collection. Then I checked the new > equations on the calc. I'm using the eqution-library from James D. Purdey > (BTW: I dont't blame him or his libraray for this). My own equations have > been stored in a list ('USEREQ'). By checking (solving) one of the equations > the calc crashed. ON&C didn't work. A Reset wich the reset-button on the > backside didn't work either. How many times will we have to answer this problem. Press the backspace key when booting, then delete the corrupted software and reboot . How many times will we have to answer this problem. Press the backspace key when booting, then delete the corrupted software and reboot . Quote from the original post: > Pressing the backspace key by rebooting (with ON&D-Q) didn't help either. If > I keep pressing the keys for some time the calc responds with a beep > (Bufferoverflow?). But again there is nothing to see on the screen. So it seems he tried this already. - Carsten BTW: I forgot to mention that I tried also the program P0erase (from you and CdB). Also without succes (I could erase Port 0 and installed the OS again but the calc is still unusable). Hope you have a more closer look at my post again and hopefully know a solution. Roman I expanded yesterday my equation-collection. Then I checked the new > equations on the calc. I'm using the eqution-library from James D. Purdey > (BTW: I dont't blame him or his libraray for this). My own equations have > been stored in a list ('USEREQ'). By checking (solving) one of the > equations > the calc crashed. ON&C didn't work. A Reset wich the reset-button on the > backside didn't work either. How many times will we have to answer this problem. > Press the backspace key when booting, then delete the corrupted software and > reboot . does something like that exist? And I need a function or a program to interpolate between two values from a steam table. greetings, bernie You can find several equation-librarys (like the one on the hp48) on www.hpcalc.org have a lot of fun Roman Bernie schrieb im Newsbeitrag does something like that exist? > And I need a function or a program to interpolate between two values from a > steam table. what about: << + 0.5 * > greetings, bernie > Is it normal that when transfering using XModem (PC->Calc) it takes about ~10 seconds before the computer starts sending? I'm using HyperTerminal, because HPComm doesn't work at all (sometimes it does, but most of the time it crashes, just after autodetecting the com port). I've noticed that HyperTerminal by default uses CRC error checking, but when it finally starts sending it changes to Checksum. Maybe that's the reason? BTW, has anyone noticed that when using small fonts for editing the height of the status area gets reduced by one pixel? Not that I'm worried about it, just found it curious. Walter. Is it normal that when transfering using XModem (PC->Calc) it takes > about ~10 seconds before the computer starts sending? I'm using > HyperTerminal, because HPComm doesn't work at all (sometimes it does, > but most of the time it crashes, just after autodetecting the com > port). I've noticed that HyperTerminal by default uses CRC error > checking, but when it finally starts sending it changes to Checksum. > Maybe that's the reason? You got the answer yourself. When sending with Xmodem on the HP49/48, it first tries three times to use Xmodem CRC, when it all fails, it switches back to checksum. That's why it takes a while. Have a look at the Xmodem server software for the HP49G on hpcalc.oirg. It's really fast and don't have this problem. It's also faster as it's using Xmodem 1K BTW, has anyone noticed that when using small fonts for editing the > height of the status area gets reduced by one pixel? Not that I'm > worried about it, just found it curious. Yes, and that's normal. The header height is calculated so the command line can take all the space. If the header was using one more screen line, the top text line of the command line would be truncated. X > The header height is calculated so the command line can take all the space. > If the header was using one more screen line, the top text line of the > command line would be truncated. > Q: why it can't be hard-coded to be the smaller size all the time? > Q: why it can't be hard-coded to be the smaller size all the time? Think about it, it's for the same reason the header have to be reduced if you use a smaller font. If you use a 8 pixels height font, header normal (2 text lines) Up to 5 line in the command line: 8*5 = 40, that leaves 16 lines for the header (screen is 56 lines) If you use 7 pixels height font, up to 6 lines in the command line: 7*6 = 42, that leaves 14 lines for the header It just looks better that way Q: why it can't be hard-coded to be the smaller size all the time? Think about it, it's for the same reason the header have to be reduced if > you use a smaller font. > If you use a 8 pixels height font, header normal (2 text lines) > Up to 5 line in the command line: > 8*5 = 40, that leaves 16 lines for the header (screen is 56 lines) If you use 7 pixels height font, up to 6 lines in the command line: > 7*6 = 42, that leaves 14 lines for the header It just looks better that way > ok but I would leave it at 14 lines hard-coded PS: My father still uses the 48GX with 10 pixels high font PPS: Are there any stack enhancers for the 49G to show more than 8-pixel font? Not even with ML graphics? pity...we have only one 48GX in the family and I can't play with it anymore... oh well - back to my 48SX+128KB at least the function key colors are not too Gay for a strait guy like me. I wish I could persuade my mom to use 10B and my daughter the HP 6S hmmm maybe Nina will use the 10B... > but I would leave it at 14 lines hard-coded If I did that, you would see half a text line in the command line and you would be the first one to complain about it ! but I would leave it at 14 lines hard-coded If I did that, you would see half a text line in the command line and you > would be the first one to complain about it ! > ó##%/&% You're right - again !!! I didn't know that it's best to start communications in this doesn't allow to change the initial error checking setting (maybe because I'm using the free version). Does anyone know of a good alternative? The XModem software from HPCalc.org (I mistakenly called it HPComm in my previous post, sorry for that) constantly locks up. Walter. > Does anyone know of a good > alternative? I always use Emu48 for transfers. Is like to own a second (virtual) calculator for backup. Both calculator have a personal I/O menu with the option Send and Recive: the transfer is automatically done from level 1 of one stack to the other. If desired, any of the files installed in Emu, can be saved to the PC... I used HPComm (slow Kermit) only at the beginning and no other software, so I can't compare my way with other: only say it's nice. that sounds like a good suggestion... I'll give it a try. Walter. > Is it normal that when transfering using XModem (PC->Calc) it takes > about ~10 seconds before the computer starts sending? I'm using > HyperTerminal, because HPComm doesn't work at all (sometimes it does, > but most of the time it crashes, just after autodetecting the com > port). I've noticed that HyperTerminal by default uses CRC error > checking, but when it finally starts sending it changes to Checksum. > Maybe that's the reason? BTW, has anyone noticed that when using small fonts for editing the > height of the status area gets reduced by one pixel? Not that I'm > worried about it, just found it curious. This is exactly the reason. The PC tries to send in CRC mode, and the 49 does not support the CRC mode (it supports the checksum and the HPCRC (another type of CRC)). So your PC retries 10 times, with 1s interval before switching to CHK mode. If you configure it in CHK from the start, you wont have the problem. Also, always start the Sender first, and then the receiver it will speed up the transfer start. Is it normal that when transfering using XModem (PC->Calc) it takes > about ~10 seconds before the computer starts sending? If I'm not wrong, when using XModem, the sending must start before the other starts to recive. Just the other way when using Kermit. > BTW, has anyone noticed that when using small fonts for editing the > height of the status area gets reduced by one pixel? I've seen that when I use 7 font instead the normal 8 font. (I alwais talk about my metakerneled and erabled 48GX) Can the HP 48 or 49 be programmed for data collection of number of events in an interval (ie. # of times a key is pressed during 8 hrs at 15 minutes interval), and the the data transfered to a PC as an ASCII file? TIA, Carlos M. Contreras Both can be programmed to do so. PS: How $£? is the budget? > Can the HP 48 or 49 be programmed for data collection of number of events in > an interval (ie. # of times a key is pressed during 8 hrs at 15 minutes > interval), and the the data transfered to a PC as an ASCII file? TIA, Carlos M. Contreras Is 9600 the fastet speed to transfer data/programs? A speeded-up 49G can do 19,200 and a standard 49G to 49G can go to 15,360 > Is 9600 the fastet speed to transfer data/programs? > I would like to know if there are some programs to test the divergence, convergence of a serie (numerical or function). As well interval and radius and convergency! I know i ask perhaps too much! vince0905@yahoo.fr For those who are interested in what exactly Calypso was supposed to be, this is the product brief HP sent me back in April 2001. http://leviathan.orblivion.com/downloads/brief.pdf After I took a look at the brief, I sent HP some questions and they replied with the following: Here are the answers to your questions, from one of our software engineers: of a PDA-type thing. Will it have functionality that could / would replace the HP48 / 49 ? A: No, but you are quite welcome to develop a HP48/49 emulator for it! Q: I know Xpander was cancelled. Is this a compromise between the Xpander and something like the HP49? A: No, this is a completely new product line that the ACO is starting, not mathematically or financially oriented. Q: The Xpander didn't have hardly any keys... will there be many keys on this product, or will most of the interface be via the touch screen? A: It will be mainly a touch screen interface, however, enough key have been included in order to allow fast access to important application, as well as action games. Calypso has 9 keys: 4 direction keys (included in a directional keypad) 4 function keys (on the right of the keypad) (they can be programmatically reassigned for gaming purpose) 1 ON/OFF key These keys are hard keys, not rubber keys I was sort of under the impression that it was going to be mathematically oriented, so I was somewhat disappointed to read these answers. After receiving this, I promptly declined their offer to be on the developer's community (I don't think I would have been able to contribute anyway) as well as the invitation to attend some sort of conference (called the Kick Start Conference) I actually have the April 26, 2001 Hewlett Packard, through its Appliance & Calculator Operation (ACO), is releasing the first of a new range of products viz. a handheld personal appliance. As a highly talented member of the developers community you have been recommended to us to be invited to become part of HP's developers community for this new product. Should you choose to accept this invitation, as part of this group you will receive:- * A Developer Unit of the new product * Invitation to the Developers' Kick Start Conference - an intensive workshop in Cupertino, to be held June 1 to June 3, 2001 * Free membership Registration on the Developers Portal and participate in various technical forums * Access to SDK and technical information * HP Certificates and souvenir as recognition for being the pioneers of this development group * Invitation to participate in software development for future products * Opportunity to visit ACO headquarters to participate in future ACO product development programs If you would like to participate in this program and attend our Kick Start Conference on June 1-3 to be held in Hewlett Packard , Mountain View, please time we will be able to provide you with details of the product itself. The June 1-3 Kick Start Conference Agenda is attached. Maybe that will quench some of the thirst for knowledge about Calypso. Recently I've seen a lot of posts here about it. I haven't been keeping up with anything related to this, so maybe this information is already available. Anyway, I hope it was insightful. BTW, my NDA protection period ended August 2001. =) -- Aaron >A project for a talent like you is to write a new >SysRPL-debugger for the 49; that should be easier >than a ML-debugger I guess :-) > We already have one - it's called SDB :). We just need more people > working on fully porting it to the 49. The xSDB command could be > broken off from the rest of the bulk of Jazz. In fact, this could be > done with DB also; then the 49 development tool set would be complete. I agree. That's exactly what I proposed several times in this NG month ago. In vain, there was no reaction. It is obvious that this project should be based on the achievements of JAZZ. However, one shouldn't underestimate the problems. Several pointers used in JAZZ have to be recovered or rewritten on the 49 (for JYA this should not be too difficult since he probably remembers his own ROM-shiftings :-) The best solution would probably be to integrate a debugger in the builtin lib 257 for the next ROM-version. That would make the 49 an ideal instrument for high-quality development. with Emacs is nearly indispensable. A few bugs will hopefully be fixed in the very next future. I propose to turn XLIB 258 0 (= nop, does nothing at present) into a SysRPL-toggler like SDA~ from OT49. This frees a OT49-rompointer from which I could make a Header-toggler, for instance. Here is the shortest such toggler (which works also in Edit mode if assigned with a TakeOver to a key: :: GetHeader #1+ BINT3 #/ DROP SetHeader ; - Wolfgang WR> However, one shouldn't underestimate the problems. WR> Several pointers used in JAZZ have to be recovered WR> or rewritten on the 49 Why? Danial Lidstroem did all this. Jazz49 and in particular SDB are restart from the original Jazz sources and do all this work over again. I believe it is only decompilation/entry lookup and display which have to converted to HP49 native stuff. But you are right of course that some of those pointers discovered by Daniel may not be stable on the HP49G, so a solution for this would have to be found. Integrating into lib258 would not help since this is a separate library which is not part of the Rom. JYA, I hope very much you will have the time and patience to do this. WR> with Emacs is nearly indispensable. A few bugs WR> will hopefully be fixed in the very next future. The bug is identified (the addresses of unsupported rompointers in the table was wrong, blame me not Thomas...), will be fixed in the next version. - Carsten > JYA, I hope very much you will have the time and patience to do this. Ok, I now have a version of Jazz working with the extable and the built-in font. Will try to take sdb out now... Maybe this week-end .. It's actually fun to code :) Ok, I now have a version of Jazz working with the extable and the built-in font. Also this should of course be released! Will try to take sdb out now... Maybe this week-end .. It's actually fun to code :) It sure is. Welcome back! - Carsten > Ok, I now have a version of Jazz working with the extable and the built-in > font. You wouldn't happen to need beta testers would you? :-) > Will try to take sdb out now... Maybe this week-end .. It's actually fun to > code :) And it's a long week-end, so perhaps you can take DB out as well? :-) -- This message was written with 100% recycled electrons Pivo > And it's a long week-end, so perhaps you can take DB out as well? :-) > Yes, but Star Wars is out this week... I have to get my priority right first ! > But you are right of course that some of those pointers discovered by > Daniel may not be stable on the HP49G, so a solution for this would > have to be found. Integrating into lib258 would not help since this > is a separate library which is not part of the Rom. Please read my message more carefully. I'd not speak on integrating into lib 258 but into lib 257 which makes a huge difference. - Wolfgang > ... I'd not speak on integrating into lib 258 > but into lib 257 which makes a huge difference. > Right, my mistake. Sorry Why should the debugger be in library 257? Because this is a useful supplement to lib 256 concerning assembling/reassembling of SysRPL and ML. The two unvisible rompointers of lib 257 can be shifted without any problems for CdB who that ACO swiped in ROM version 19-5 or so :-) and there should still be enough room in block 0 for a complete debugger. It is obvious that such a tool need only be consistent with the ROM from version 19-6 on. - Wolfgang > ... I'd not speak on integrating into lib 258 > but into lib 257 which makes a huge difference. Right, my mistake. Sorry Why should the debugger be in library 257? > Because this is a useful supplement to lib 256 > concerning assembling/reassembling of SysRPL > and ML. The two unvisible rompointers of lib 257 > can be shifted without any problems for CdB who > that ACO swiped in ROM version 19-5 or so :-) > and there should still be enough room in block 0 > for a complete debugger. It is obvious that such > a tool need only be consistent with the ROM from > version 19-6 on. - Wolfgang Excellent idea !!! Like your new programmable filer Can you give us macros for the filer Maybe just the keyboard redefinitions could be seen through when there is no Filer assignment present... PS: You mean version 1.19-6 (or 1.19-7 :) JB> We already have one - it's called SDB :). We just need more people JB> working on fully porting it to the 49. The xSDB command could be JB> broken off from the rest of the bulk of Jazz. In fact, this could be JB> done with DB also; then the 49 development tool set would be complete. That would be fantastic. In fact, isolating SDB should not be too difficult, mainly - avoid using the Jazz SysRPL stack - using ->S2 to decompile code snippets and display them in some fashion independent of the Jazz special display routines. I have no time to do it, also I am not familiar with the PC tools and with the Jazz source code. But I would be extremely grateful if someone could get SDB working as a separate program on the HP49. I would guess that for someone like Daniel, this should be a relatively simple thing - correct me if I am wrong. - Carsten > That would be fantastic. In fact, isolating SDB should not be too > difficult, mainly > > - avoid using the Jazz SysRPL stack > - using ->S2 to decompile code snippets and display them in some > fashion independent of the Jazz special display routines. - using extable's routines, in order to save the 30kb of the JAZZ tables. -- Eduardo M Kalinowski http://move.to/hpkb > I have no time to do it, also I am not familiar with the PC tools and > with the Jazz source code. But I would be extremely grateful if > someone could get SDB working as a separate program on the HP49. I > would guess that for someone like Daniel, this should be a relatively > simple thing - correct me if I am wrong. I'm not too busy for the next 2 days, maybe I could have a look at it. The source code is available on hpcalc.org ? I'm not too busy for the next 2 days, maybe I could have a look at it. The source code is available on hpcalc.org ? yes JA> I'm not too busy for the next 2 days, maybe I could have a look at it. The > JA> source code is available on hpcalc.org ? I'll have a deeper look tomorrow. I've identified where the search in the entries table is done. Hopefully it shouldn't be too difficult to change to use the extable library. I want to change it so it uses the built-in font as well. Then I'll look into making SDB independent... I'm not promissing anything though > > Then I'll look into making SDB independent... I'm not promissing anything > though > Luis > What remained is the command ->XU (to extended UsrRPL) > from OT49 which discards all unnecessary UsrRPL > delimiters and quoters recursively. That speeds > up a UsrRPL consicerably and makes it shorter. Something like that for the old 48? i've seen that WINHP sourcecode has been released! http://membres.lycos.fr/chabanoi/ would it be possible to compile that delphi program to be used in linux? > i've seen that WINHP sourcecode has been released! > http://membres.lycos.fr/chabanoi/ would it be possible to compile that delphi program to be used in linux? -- ----- Ce ne sont que des propositions. Je ne veux pas les faire passer en force. Je pense que si mes id.8ees doivent .90tre reprises, elles ne doivent pas passer au vote, pour plusieurs raison : -+- BC in : http://neuneu.ctw.cc - Neuneu sans vote et sans forcer -+- i've spoken with the author and told me it wasn't so simple because it uses some libraries for window$ i don't now much about programing, could you try to compile it and let us know about it?? vicent > > i've seen that WINHP sourcecode has been released! > http://membres.lycos.fr/chabanoi/ would it be possible to compile that delphi program to be used in > linux? > > uses some libraries for window$ Then a port should be considered. > i don't now much about programing, could you try to compile it and let > us know about it?? You're kidding, right? I don't have Kylix, I don't know anything about this WinHP tool, I don't like Windows only programs, and I have too much spare time to waste it... > i've seen that WINHP sourcecode has been released! > http://membres.lycos.fr/chabanoi/ would it be possible to compile that delphi program to be used in > linux? > -- ----- It works better if you plug it in where it should be. COL DROP SWAP ->COL DROP DOT > kkng > I get dimension error when I try to do a dot product of two column > vectors. On the TI's I know that it doesn't matter if the vectors are > rows or columns, they both work. Help please, You may try this with two column vectors(matrices) at level 1 and 2: << TRAN ->ROW DROP SWAP TRAN ->ROW DROP DOT > EVAL kkng > I get dimension error when I try to do a dot product of two column > vectors. On the TI's I know that it doesn't matter if the vectors are > rows or columns, they both work. Help please, > I get dimension error when I try to do a dot product of two column > vectors. On the TI's I know that it doesn't matter if the vectors are > rows or columns, they both work. Help please, Well, first of all if we are talking about vectors, there is only one type of vectors on the HP49G. A vector like [x1 x2 ...] is always a vector, be it column or row. For two vectors the command DOT will *always* work, provided of course, the two vectors have the same number of elements. Now, if you have two *matrices* [[x1 x2 ...]] and [[y1] [y2]...] then you of course can use the command * to get the DOT product but the first matrix must be a row matrix and the second a column matrix, and the first must have as many columns as the second has rows. (Definition of matric product) Nick. Make sure that you are using the sqare brackets [] for the vectors on the HP 49g. I belive on the TI's they use {} instead. That could be the only thing I could imagine that would cause such an error. Micah > I get dimension error when I try to do a dot product of two column > vectors. On the TI's I know that it doesn't matter if the vectors are > rows or columns, they both work. Help please, Can you please give the exact example? > I get dimension error when I try to do a dot product of two column > vectors. On the TI's I know that it doesn't matter if the vectors are > rows or columns, they both work. Help please, Is it possible to set a choose box such as when i pick an element, the elements in another choose box change? For example i have a choose box with a list of countries and another with cities, and i want that when i pick a country the cities choose box displays only the cities in that country. Anybody could give an idea about how to do this? Marturo UBI DUBIDUM IBI LIBERTAS > Is it possible to set a choose box such as when i pick an element, the > elements in another choose box change? This is an example of an element << Countries {{Spain scities} {Finland fcities} ....} 1 CHOOSE IF THEN EVAL END>> When you select Spain option in the CHOOSE box, scities will be evaluated. ...Yes, scities is the choose menu of spanish cities. Hope this helps > Is it possible to set a choose box such as when i pick an element, the > elements in another choose box change? > > > This is an example of an element > << Countries > {{Spain scities} > {Finland fcities} > ....} > 1 CHOOSE > IF THEN EVAL END> > When you select Spain option in the CHOOSE box, scities will be > evaluated. > ...Yes, scities is the choose menu of spanish cities. > > Hope this helps > > One coould even have the choose cities menu program of each country saved under the name of that country: << Countries {Spain Finland} ....} 1 CHOOSE IF THEN EVAL END > where the spanish cities choose program is saved as 'Spain', etc. I think Marturo asked about a Choose Box from inside an input form, and he wanted the contents of some other Choose Box in the same Input form to change according to what he chooses for the first Choose Box. Or did I made things more complex than they already are? ;-) Nick. > > Is it possible to set a choose box such as when i pick an element, the > elements in another choose box change? > > > This is an example of an element > << Countries > {{Spain scities} > {Finland fcities} > ....} > 1 CHOOSE > IF THEN EVAL END> > When you select Spain option in the CHOOSE box, scities will be > evaluated. > ...Yes, scities is the choose menu of spanish cities. > > Hope this helps > > > > One coould even have the choose cities menu program of each country > saved under the name of that country: > > << Countries > {Spain Finland} > ....} > 1 CHOOSE > IF THEN EVAL END > where the spanish cities choose program is saved as 'Spain', etc. > > I think Marturo asked about a Choose Box from inside an input form, > and he wanted the contents of some other Choose Box in the same > Input form to change according to what he chooses for the first Choose > Box. Yes this is what i want i don't want to open a new input box. And i wanna do it in sysRPl. Any Ideas? Marturo UBI DUBIDUM IBI LIBERTAS > > I think Marturo asked about a Choose Box from inside an input form, > and he wanted the contents of some other Choose Box in the same > Input form to change according to what he chooses for the first Choose > Box. > > > Yes this is what i want i don't want to open a new input box. And i > wanna do it in sysRPl. Any Ideas? > > > Marturo > > UBI DUBIDUM IBI LIBERTAS I don't think it can be done. Once you open a choose box, I think your only options are to choose something (and exit) or to cancel (and exit). But what is wrong with nesting choose boxes? How would one using a HP 49G convert a decimal number (i.e. 7.8) into a fraction (i.e. 7 4/5)? If you have an answer, please reply to the group and, if not too much trouble, also send the reply to: mattkbryant@cox.net itsawombat@mac.com response(s). --Matt Bryant You can type XQ when you have the decimal on the first level of the stack. > How would one using a HP 49G convert a decimal number (i.e. 7.8) into a > fraction (i.e. 7 4/5)? If you have an answer, please reply to the group and, if not too much > trouble, also send the reply to: mattkbryant@cox.net > itsawombat@mac.com response(s). --Matt Bryant Very nice Micah, thank you. But how would I go about converting that new fraction into a mixed number? Please reply as fast as you did the first time :-) > You can type XQ when you have the decimal on the first level of the stack. > How would one using a HP 49G convert a decimal number (i.e. 7.8) into a > fraction (i.e. 7 4/5)? If you have an answer, please reply to the group and, if not too much > trouble, also send the reply to: mattkbryant@cox.net > itsawombat@mac.com response(s). --Matt Bryant > Very nice Micah, thank you. > > But how would I go about converting that new fraction into a mixed number? Use the PROPFRAC command. Matt, use Left Shift(blue) ARITH(1) NXT PROPF(F2) for 1.19-6 this should give you the mixed number you are looking for HM > Very nice Micah, thank you. But how would I go about converting that new fraction into a mixed number? Please reply as fast as you did the first time :-) > You can type XQ when you have the decimal on the first level of the stack. > How would one using a HP 49G convert a decimal number (i.e. 7.8) into a > fraction (i.e. 7 4/5)? If you have an answer, please reply to the group and, if not too much > trouble, also send the reply to: mattkbryant@cox.net > itsawombat@mac.com response(s). --Matt Bryant I have finished now the long float delta version (the Pi and the Exp functions are done). I don't have much time to work in this lib but at any free time that I have, I translate the Alg48v401 Long Float code for the Long Float 49 ;] If anyone could test the speed of Alg48 (I don't know how) post the time here (test the FPI function with 100 digits), I'm getting 4.9316 s for PI with 100 digits... DanieL. > I have finished now the long float delta version (the Pi and the >Exp functions are done). I don't have much time to work in this lib >but at any free >time that I have, I translate the Alg48v401 Long Float code for the >Long Float 49 ;] > If anyone could test the speed of Alg48 (I don't know how) post >the time here (test the FPI function with 100 digits), I'm getting >4.9316 s for PI with 100 digits... DanieL. > It takes mine 4.05 s to compute this (average of three time measurements). (A minha calcula isso em 4.05 s (m.8edia de tr.90s medidas). By the way, once I intended to write a User-RPL program to compute pi to n places. I chose a Brent-Salamin AGM method which then seemed easy to implement, I just had to write my own User-RPL long float square root, multiplication and division routines (which I never did). Instead, I used Mika Heikanen's LongFloat library to test the algorithm. Of course, a User-RPL only code would be much slower, but I wanted to know how fast this could be done compared to other previous HP calculators and programs... I don't know which method Mika used but surely it was a very efficient one. However, I wonder if this algorithm would have a similar performance if converted to machine language... PS.: don't forget to make your LongFloat translation available (just in case I get a 49G - if they don't discontinue it... ) (n.8bo se esque.8da de deixar sua tradu.8d.8bo de LongFloat para a gente - s.97 para o caso de algum dia eu comprar uma 49G, isto .8e, se eles n.8bo pararem de fabric.87-la antes...) Gerson W. Barbosa ------------------------------------------------------------- Usage: 1: number of desired decimal places (requires LongFloat library to run) Returns: 2: running time (h.mmss) 1: pi to n places 100 places: 62.59 s (should take much more time with User-RPL long float routines...) %%HP: T(2)A(D)F(,); .82 1 + DUP 'DIGITS' STO TIME SWAP . 1 ROT START 0 + NEXT 1 + 0.5 1 2 FSQRT 2 FDIV 1 DO DUP2 OVER FADD 2 FDIV ROT ROT FMUL FSQRT OVER 4 ROLL FSUB DUP FMUL 4 ROLL 2 FMUL DUP2 FMUL 6 ROLL SWAP FSUB ROT 5 ROLLD 4 ROLLD SWAP ROT 5 ROLL 6 PICK FCMP UNTIL 1 =/ END FMUL 2 FMUL ROT FDIV SWAP ROT DROP2 TIME ROT HMS- SWAP é > I don't have much time to work in this lib but at any free > time that I have, I translate the Alg48v401 Long Float code for the > Long Float 49 ;] Has Mika said that's ok? > Has Mika said that's ok? Isn't long float in GPL ?? (77 secs...) Whoa! I'm getting 8Minutes and 58 secs with 1000 digits... How can I contact Mika? DanieL > I don't have much time to work in this lib but at any free > time that I have, I translate the Alg48v401 Long Float code for the > Long Float 49 ;] > > Has Mika said that's ok? > Yes - to me - about two years ago when I asked him if he could do the porting... > I don't have much time to work in this lib but at any free > time that I have, I translate the Alg48v401 Long Float code for the > Long Float 49 ;] Has Mika said that's ok? I have finished now the long float delta version (the Pi and the > Exp functions are done). I don't have much time to work in this lib > but at any free > time that I have, I translate the Alg48v401 Long Float code for the > Long Float 49 ;] > If anyone could test the speed of Alg48 (I don't know how) post > the time here (test the FPI function with 100 digits), I'm getting > 4.9316 s for PI with 100 digits... > > DanieL. Mine does it in 3 secs flat, and 77 secs for 1000. digits. What method are you using? Werner 77 secs... Do you have written the program in 100% ML? or the algorithm that you're using is much effective than Ramanujan's? I saw a PI generator for TI that was built in 100% ASM, well it calculates the pi a quite fast (a lot fast), well, if a Pi Generator could be done in Z80 asm, why it can't be done in Saturn's ML? DanieL. (Sorry for the english) > I have finished now the long float delta version (the Pi and the > Exp functions are done). I don't have much time to work in this lib > but at any free > time that I have, I translate the Alg48v401 Long Float code for the > Long Float 49 ;] > If anyone could test the speed of Alg48 (I don't know how) post > the time here (test the FPI function with 100 digits), I'm getting > 4.9316 s for PI with 100 digits... > > DanieL. > > Mine does it in 3 secs flat, and 77 secs for 1000. digits. > What method are you using? > Werner WH> Mine does it in 3 secs flat, and 77 secs for 1000. digits. WH> What method are you using? WH> Werner Have you ever released this? Why are you keeping it from us??? - Carsten Another thing: My calc is loaded with lots of stuff (194kb in home, 97kb in port1, and 298kb in flash), this could slow down the code... DanieL. > I have finished now the long float delta version (the Pi and the > Exp functions are done). I don't have much time to work in this lib > but at any free > time that I have, I translate the Alg48v401 Long Float code for the > Long Float 49 ;] > If anyone could test the speed of Alg48 (I don't know how) post > the time here (test the FPI function with 100 digits), I'm getting > 4.9316 s for PI with 100 digits... > > DanieL. > > Mine does it in 3 secs flat, and 77 secs for 1000. digits. > What method are you using? > Werner Ramanujan's formula, the code is based in Alg48 Long float, the speed problem could be in the conversions... I'll try to change something here for a speed up (I have tested using the Emu48, and I see a very little delay between two answers - the Alg48 and Long float 49). DanieL. > I have finished now the long float delta version (the Pi and the > Exp functions are done). I don't have much time to work in this lib > but at any free > time that I have, I translate the Alg48v401 Long Float code for the > Long Float 49 ;] > If anyone could test the speed of Alg48 (I don't know how) post > the time here (test the FPI function with 100 digits), I'm getting > 4.9316 s for PI with 100 digits... > > DanieL. > > Mine does it in 3 secs flat, and 77 secs for 1000. digits. > What method are you using? > Werner Can anyone help me how get the primitiv function of f(x)=1/x^(n/2) Joakim > Can anyone help me how get the primitiv function of > f(x)=1/x^(n/2) > > Joakim Assuming RPN mode: Just type 'f(x)=1/x^(n/2)' in the EQW, press ENTER and then press DEF. ([blue-shift], then [3]). A new variable f will be created in the current directory, which works much like the built-in commands. That means, enter for example X or 2 or what you like, press the soft key for f, and you get the result f(x)=1/x^(n/2) or f(x)=1/2^(n/2) and so on. Hope it helped, Nick. Start EQW key in [ 1 ] then divide [ / ] then the variable [ X ] you should see 1 --- X Continue with power key [Y^x] then [left-shift] [ - ] which gives the ( ) written above the minus key [ - ] then press [ALPHA] [left-shift] [ N ] and now you have (n) as the exponent of X continue with divide [ / ] and [ 2 ] Select all by pressing [right-shift] [up-arrow] finally key [left-shift] [ 4 ], which reads CALC above the key [ 4 ] and then select the INTVX command To EVALuate you need to press [right-shift] [SYMB] ... Unfortunately EXP>EXP/LN is not handled so you have to ask more help from the CAS programmers here or go and buy a TI 89 ! > Can anyone help me how get the primitiv function of > f(x)=1/x^(n/2) Joakim Being the proud winner of an ebay auction for a 15c (I've been after one of these for _years_), I now have it! Excellent condition, excellent price... A quick question for y'all...there is a little asterisk flashing, when the calc is on, in the left-hand side of the display...what is this trying to tell me? Before someone says 'RTFM@, I have ordered the 'FM' CDs from hpmuseum.org but it'll be a little while before I get them...I was hoping someone here could help me out a little. :-) WigglePig > Being the proud winner of an ebay auction for a 15c (I've been after one > of these for _years_), I now have it! Excellent condition, excellent > price... Surely not less that what I paid for mine a few weeks ago (60 euros, about US$ 50)... ;) I even got an HP16C for 30 euros, some days earlier... Of course, that was not on eBay. > A quick question for y'all...there is a little asterisk flashing, when > the calc is on, in the left-hand side of the display...what is this > trying to tell me? Change the batteries. -- ----- If there' s another French on this list, will he be kind enough to give me some addresses of places that still sells the Neo (and games) ? -+- SB in : Guide du Neuneu d'Usenet - French spoken on fufe -+- Erwann, On Thu, 16 May 2002 12:42:40 +0200, Erwann ABALEA >I even got an HP16C for 30 euros, some days earlier... > Do you know where I can get one for that price???? As stated on HPMuseum.org, it was an off-line sale. Ask your friends, relatives, collegues, ... -- ----- Juste un truc, ca te ferait mal au cerveau de lire les messages auxquels tu reponds ? -+-RMD in : Guide du Neuneu d'Usenet - Neuneu apprend .88 lire -+- Surely not less that what I paid for mine a few weeks ago (60 euros, about > US$ 50)... ;) > Indeedy. :) I even got an HP16C for 30 euros, some days earlier... > Ah, now my brother is looking for a 16c...no luck yet... Of course, that was not on eBay. > I've seen some horrendous prices and equipment conditions on ebay but, in my case, neither were a problem with the calc I won. Now I just need to sell my 49G to begin to enjoy HP calcs again! > Change the batteries. > Will be done, just as soon as I can get some silver oxide ones...they seem to be a bit difficult to get ATM; lithium ones are everywhere... WigglePig That you've got a stolen HP 15C and it's sending it's location information constantly !!! OR You batteries need to be replaced by a HPQ technician. (HWP+CPQ=>HPQ officially!!!) Pick your choice ;-) Being the proud winner of an ebay auction for a 15c (I've been after one > of these for _years_), I now have it! Excellent condition, excellent > price... A quick question for y'all...there is a little asterisk flashing, when > the calc is on, in the left-hand side of the display...what is this > trying to tell me? Before someone says 'RTFM@, I have ordered the 'FM' CDs from > hpmuseum.org but it'll be a little while before I get them...I was > hoping someone here could help me out a little. :-) WigglePig That you've got a stolen HP 15C > and it's sending it's location information constantly !!! > OR > You batteries need to be replaced by a HPQ technician. > (HWP+CPQ=>HPQ officially!!!) > Pick your choice > ;-) > AH, that's it then...I've put it in a metal biscuit tin to prevent them from finding me and taking it away. :-) And I've ordered some new batteries from the stores... WigglePig I'm interested in upgrading my 48g to 128k ram. Having difficulties locating the 128k chip. The unoffical hp upgrade page lists electronics retailers who no longer stock the chip, or any like it. I'm all ready to Matt. Try www.digikey.com or search Ebay for sram only in the title. John Znalezniak > I'm interested in upgrading my 48g to 128k ram. Having difficulties > locating the 128k chip. The unoffical hp upgrade page lists electronics > retailers who no longer stock the chip, or any like it. I'm all ready to > upgrade to 128k, even 256k! Just need the chip. Any suggestions? > Matt. I have two HP 48GX's. I have a engineering ROM card for one, and was the other. I guess it's not a big deal, but it would be awfully convenient. I remember seing a program that allowed you to move a library from a sure, but I think it was in one of the goodies disks (possibly disk 11). Maybe someone else can throw a little more light into this. Luis > I have two HP 48GX's. I have a engineering ROM card for one, and was > the other. > > I guess it's not a big deal, but it would be awfully convenient. > > I remember seing a program that allowed you to move a library from a > sure, but I think it was in one of the goodies disks (possibly disk > 11). > Maybe someone else can throw a little more light into this. Luis I saw it in hpcalc. slashdot (http://slashdot.org ) is conducting a test, and RPN is in it! Go and vote for you favorite equation entry method! I've written a new 49-filer which is somewhat easier to use than the original. Main differences: You first drive into the scanned directory. The TREE option is available on the convenient F6. The application menu has only *one* page, with the most important options, not three pages as the (overfull) original appl menu. All other options are on convenient keys (Don't worry, the keys work only while the filer is active :-). The name of the filer is presently Filer2. It has only 319 bytes and is a beta-version, of course. Here it is ftp://ftp.math.fu-berlin.de/pub/usr/raut/HP49/tools/Filer2.zip - Wolfgang PS. Clearly, Filer2 uses essential parts of the original filer, one of the best things of ACO. I plan to write a library, Fileman, which enables the *normal* user to create his individual filer, with his own appl menu and his own key assignments. I do this only if asked by at least three people because this will be a fairly involved project ... > I've written a new 49-filer which is somewhat easier > to use than the original. Main differences: You first > drive into the scanned directory. The TREE option is > available on the convenient F6. The application menu > has only *one* page, with the most important options, > not three pages as the (overfull) original appl menu. But all the functions of the original filer have been mapped to keys... So you barely have to use the menu Quite pleased that somebody finally decided to use it's Filer's infrastructure > ... The application menu > has only *one* page, with the most important options, > not four pages as the (overfull) original appl menu. > > But all the functions of the original filer have been mapped to keys... So > you barely have to use the menu Since the menu is one-page now, F1 -F6 are as good as any key. I still revised it somewhat this night (cannot sleep if excited :-) The appl menu of Filer2 is now |EDITB| |NEW| |HALT| |SORT| |VIEW| |TREE| EditB (= best possible editing) edits like DownArrow in default mode. Hence, leftshift EditB should edit as does leftshift DownArrow normally, and this is realized now. Since the important VIEW option is integrated, the V-key has become free. It now shows a complete version info: Internal Serial number, ROM version, CAS version. Thus, repeating questions in this NG about this topics may disappear ... All alpha-hold options of the original filer are now on a nonshift key, e.g., key X for Sending with XMODEM, not alpha-hold X. Clearly, key O (Oh!) is used for ORDER, RCL and EVAL options are on shifted keys. I orgininally planned to integrate a ARCHIVE/RESTORE choose box (as I have it in port 2). But the RESTORE option would be useless after a crash which regularly distroys key-assignments. Clearly, the name Filer2 or its content have to be assigned to leftshift APPS, or perhaps to rightshift APPS. For the latter use Keyman to save the key functionality in Edit mode. Have fun, Wolfgang ftp://ftp.math.fu-berlin.de/pub/usr/raut/HP49/tools/Filer2.zip I need these for another project ... I wish that you continue with your project! Anyone else? ;) Eduardo > PS. Clearly, Filer2 uses essential parts of the original > filer, one of the best things of ACO. I plan to write a > library, Fileman, which enables the *normal* user to create > his individual filer, with his own appl menu and his own > key assignments. I do this only if asked by at least three > people because this will be a fairly involved project ... Counting JYA, it's now five already, dear Wolfgang!l Go, Wolfgang, Go :-D > 1) Veli-Pekka Nousiainen > 2) Carlos Lacroze > 3) Stefan Wei¤enmayer gute Idee Gruss Stefan > I wish that you continue with your project! > Anyone else? > ;) Eduardo this will be a fairly involved project ... I wish you to continue with your project! Anyone else? ;) Eduardo > PS. Clearly, Filer2 uses essential parts of the original > filer, one of the best things of ACO. I plan to write a > library, Fileman, which enables the *normal* user to create > his individual filer, with his own appl menu and his own > key assignments. I do this only if asked by at least three > people because this will be a fairly involved project ... 1) Veli-Pekka Nousiainen 2) 3) I've written a new 49-filer which is somewhat easier > to use than the original. Main differences: You first > drive into the scanned directory. The TREE option is > available on the convenient F6. The application menu > has only *one* page, with the most important options, > not three pages as the (overfull) original appl menu. > All other options are on convenient keys (Don't worry, > the keys work only while the filer is active :-). The name of the filer is presently Filer2. It has only > 319 bytes and is a beta-version, of course. Here it is > ftp://ftp.math.fu-berlin.de/pub/usr/raut/HP49/tools/Filer2.zip - Wolfgang PS. Clearly, Filer2 uses essential parts of the original > filer, one of the best things of ACO. I plan to write a > library, Fileman, which enables the *normal* user to create > his individual filer, with his own appl menu and his own > key assignments. I do this only if asked by at least three > people because this will be a fairly involved project ... 1) Veli-Pekka Nousiainen 2) Carlos Lacroze 3) > 1) Veli-Pekka Nousiainen > 2) > 3) > I've written a new 49-filer which is somewhat easier > to use than the original. Main differences: You first > drive into the scanned directory. The TREE option is > available on the convenient F6. The application menu > has only *one* page, with the most important options, > not three pages as the (overfull) original appl menu. > All other options are on convenient keys (Don't worry, > the keys work only while the filer is active :-). The name of the filer is presently Filer2. It has only > 319 bytes and is a beta-version, of course. Here it is > ftp://ftp.math.fu-berlin.de/pub/usr/raut/HP49/tools/Filer2.zip - Wolfgang PS. Clearly, Filer2 uses essential parts of the original > filer, one of the best things of ACO. I plan to write a > library, Fileman, which enables the *normal* user to create > his individual filer, with his own appl menu and his own > key assignments. I do this only if asked by at least three > people because this will be a fairly involved project ... > 1) Veli-Pekka Nousiainen > 2) Carlos Lacroze > 3) Stefan Wei¤enmayer gute Idee Gruss Stefan Does anyone know where to download the HP48 SX manuals (in PDF or PS)? Cristiano They are available for ale on one of the cd's that hpmuseum.org puts out. Other than that, I haven't found anything for my 48s (still hunting, wish some would turn up). hpcalc.org also has a 48g series manual, but that's different enough to be really annoying. -K > > Does anyone know where to download the HP48 SX manuals (in PDF or PS)? > > Cristiano Gene > They are available for ale on one of the cd's that hpmuseum.org puts out. Cristiano Heheheh I guess it's just a calcgeek thing. ;) > Gene > They are available for ale on one of the cd's t not to think about calcgreeks? ;-) Heheheh I guess it's just a calcgeek thing. ;) Gene They are available for ale on one of the cd's t from HPCALC.org. As far as I can tell everything was installed correctly, BUT now every time I replace the batteries I lose the ps- anyone know how I can merge ports one and two to get 256K combined? Any draw backs to doing this? on such a low supply voltage that the 48G might use. OR which should have a backup battery ??? > from HPCALC.org. As far as I can tell everything was installed > correctly, BUT now every time I replace the batteries I lose the ps- anyone know how I can merge ports one and two to get 256K > combined? Any draw backs to doing this? The FAQ metions a program called Remote-BG. I cannot seem to find it anywhere on the net. The authors ftpsite does not seems to accept anon-ftp logins. Does anyone have a copy of this? TIA rich -- I will not be pushed, filed, stamped, indexed, briefed, de-briefed, or numbered...My life is my own. I am not a number. I am a free man. No. 6 Heyo- I'm looking for manuals for the S model specifically, as well as either a 19bii manual or a 19bii with manual at a reasonable price. scans, downloads, anything readable counts! K > [...] anything readable counts! Then you'll enjoy this: http://www.geocities.com/tuesdayspuzzle My HP19BII came with a '19B manual (a pleasant reading experience!) and a booklet about new features on the 19BII (mainly RPN) which was later included as appendix D in the '95LX User's Guide (also recommended). Bye. HPCC member #1046 - cinderella can now be use on PDA :) http://www.cinderella.de/en/ Looking at my newly acquired HP15c, my mind recalls a discussion on this list (some while ago) where the power supply of the 48/49 was discussed; these calcs have a ?charge-pump? voltage step-up to run the Saturn/Yorke off three AAA batteries. Is there a similar arrangement in the Voyager series calcs? If so, then it must be incredibly efficient as my 15c only just requires its first new set of batteries (calc made wk 33 1985)! WigglePig Does somebody know if LR44 batteries are valid for the 15c? > Does somebody know if LR44 batteries are valid for the 15c? Indeed, the LR44 is a drop-in replacement for the SR44, but is a lithium cell rather than silver oxide. They work fine but don't have quite as much capacity or shelf-life. WigglePig > Does somebody know if LR44 batteries are valid for the 15c? W> Indeed, the LR44 is a drop-in replacement for the SR44, but is a lithium W> cell rather than silver oxide. They work fine but don't have quite as W> much capacity or shelf-life. ... which is really no problem for the HP15C, because it uses so extremely little power. - Carsten I'm waitng for potsman... he brings the hp15c I've got exchanging. > W> cell rather than silver oxide. They work fine but don't have quite as > W> much capacity or shelf-life. ... which is really no problem for the HP15C, because it uses so > extremely little power. - Carsten But I don't want to be having to change the batteries again in 8 years...I want the next set to last as well as the last (and first) set...17 years. It takes _soooo long_ to install the new set that I really want to do it properly. ;-) It's just one of those environmental concerns...don't want to use _too many_ batteries if I don't have to. It's also nice to be able to say that the batteries lasted 17 years before you changed them. :-) WigglePig I'm currently working on Jazz for the HP49 and there are a lot of things that are bothering me. In order to debug the program in an easier way, I've first modified the source so they could be compiled with Debug 2. That was the first annoying step as debug2 put the comment column at 500 instead of the normal default of 30. Not having the source code of Debug 2 I had to patch the exe file to remove the nasty call with -c 500. After compiling it (and getting the exact same library as using the makeg.bat call (CRC are the same)) I realized that there are MANY things very wrong with the library on www.hpcalc.org First of all, there are hundred of undefined labels, errors in the source code etc... For example, running ASS on an empty file will reset the calculator... etc.. Is it just me or it has happened to other people ? (I'm not running 1.19-6 so maybe it comes from there). Before I can do any more work I need to fix the basic errors. Any comments? > I'm currently working on Jazz for the HP49 and there are a lot of things > that are bothering me. > First of all, there are hundred of undefined labels, errors in the source > code etc... > For example, running ASS on an empty file will reset the calculator... etc.. Nope it doesn't, it says ASS error: No Program > Is it just me or it has happened to other people ? (I'm not running 1.19-6 > so maybe it comes from there). Maybe you should upgrade to it :-) > Before I can do any more work I need to fix the basic errors. Any comments? I do not think that many people use Jazz on the 49 the same way they did on the 48. I can tell you that all I use it for is SDB and especially DB. For hacking we have Nosy and CQIF, the SRPLstack is redundant as well. Personally I have trouble with ED; I cannot turn alpha mode off as editor Emacs is a bit better I think. -- This message was written with 100% recycled electrons Pivo X > Any comments? I do not think that many people use Jazz on the 49 the same way > they did on the 48. I can tell you that all I use it for is SDB and > especially DB. For hacking > we have Nosy and CQIF, the SRPLstack is redundant as well. X Exactly! Only source level debugging is needed! 1) See Start Wars II 2 times 2) give us only the debugger(s) 3) later (after 1.19-7) a new mini-Jazz - maybe ?? > Nope it doesn't, it says ASS error: No Program Well, once I get this message it freezes. > Maybe you should upgrade to it :-) Downgrade you mean ? :) > I do not think that many people use Jazz on the 49 the same way > they did on the 48. I can tell you that all I use it for is SDB and > especially DB. For hacking In that case, I'm surprised that works... There are some very wrong thing with DB, especially with some pointers that have been incorrectly translated from the HP48. > we have Nosy and CQIF, the SRPLstack is redundant as well. Personally I have trouble with ED; I cannot turn alpha mode off > as editor Emacs is a bit better I think. Yeah, I've found this problem and fixed it. The problem is that I know nothing about Jazz and its tools, for example, in ED, pressing ON gives me a 3, not very useful... I think with Jazz it would be useful to remove the SysRPL stack, it is incomplete and doesn't support the new HP49 objects (won't be too difficult to improve though) I guess we could make some flavours of Jazz as Java did > Maybe you should upgrade to it :-) > Downgrade you mean ? :) Of course! Since we are not in the possession of your ROM 19-7 (or maybe ROM 20 ??) and there is probably no hope to get it soon, all what concerns developement should be backward compatibel to the present 19-6. PS. Maybe I prized MASD too much :-) There are some intelligent ideas in JAZZ which MASD unfortunately lacks. In particular the following: There is a flag setting which allows the user to call DIS on a SysRPL object *with or without* decompiling possibly included code objects. In MASD you've first to isolate these codes and treat them separatly - for not everything can be done on very large SysRpl objects in the editor. You could use, e.g., the unused flag -4. Still better is the somewhat childish flag -92. This one is indeed dispensable if ->S2 would include the possibly decompiled ML-subcodes always into the CODE/ENDCODE delimiters. Most developers probably do this anyway to make compilation independent on flag -92. On the other hand, flag -85 is IMHO an ingenious ACO idea. It helps very much in modifying or rearranging very large SysRPL-object on the stack. I'm mot surprised that this flag is redundant for Pivo. He's a good ML programmer but not a Sys-RPL programmer I believe ... > On the other hand, flag -85 is IMHO an ingenious ACO idea. > It helps very much in modifying or rearranging very large > SysRPL-object on the stack. I noticed, your default is to set -85 to display SysRPL. > I'm not surprised that this > flag is redundant for Pivo. I haven't said that, I meant that the Jazz SSTK is redundant > He's a good ML programmer but > not a Sys-RPL programmer I believe ... by the beauty of ML :-) Try my Oil game for the 48. I think it is one of the few programs that uses the POL. Raemen97 was my first SysRPL prog. Darts was a quest for bytes, I tried to get it as small as possible (should appeal to your taste) -- This message was written with 100% recycled electrons Pivo WR> PS. Maybe I prized MASD too much :-) There are some WR> intelligent ideas in JAZZ which MASD unfortunately lacks. WR> In particular the following: WR> There is a flag setting which allows the user to call WR> DIS on a SysRPL object *with or without* decompiling WR> possibly included code objects. In MASD you've first to WR> isolate these codes and treat them separatly - for not WR> everything can be done on very large SysRpl objects in WR> the editor. You could use, e.g., the unused flag -4. WR> Still better is the somewhat childish flag -92. This one WR> is indeed dispensable if ->S2 would include the possibly WR> decompiled ML-subcodes always into the CODE/ENDCODE WR> delimiters. Most developers probably do this anyway WR> to make compilation independent on flag -92. WR> On the other hand, flag -85 is IMHO an ingenious ACO idea. WR> It helps very much in modifying or rearranging very large WR> SysRPL-object on the stack. I'm mot surprised that this WR> flag is redundant for Pivo. He's a good ML programmer but WR> not a Sys-RPL programmer I believe ... Seems to me that your book update is finished, since you have time to judge everybody and everything like an Oberlehrer again :-) My guess with flag 92 would be that MASD was initially developed for ML alone and only later got a SystemRPL mode added to it. Flag 92 would then have been introduced for backward compatibility, in order to keep compiling those older files which were ML only. Anybody knows if this is how it went? And, while Pivo may certainly be more at home in ML than in SystemRPL, he has written some significant SystemRPL programs, including the incremental search in Emacs and his DARTS library, which I use every day :-) - Carsten I'm currently working on Jazz for the HP49 and there are a lot of things that are bothering me. After compiling it (and getting the exact same library as using the makeg.bat call (CRC are the same)) I realized that there are MANY things very wrong with the library on www.hpcalc.org First of all, there are hundred of undefined labels, errors in the source code etc... For example, running ASS on an empty file will reset the calculator... etc.. Is it just me or it has happened to other people ? (I'm not running 1.19-6 so maybe it comes from there). I think Daniel search for and found a number of unsupported pointers used by Jazz. Many of these may not be supported on the HP49G, I have no idea. But I seem to remember that Jazz had to be updated for Rom 1.19-6. So it is quite possible that you need to get ROM 1.19-6 into the emulator for it to work. Daniel, are you reading this? Can you help out? - Carsten > I'm currently working on Jazz for the HP49 and there are a lot of things > that are bothering me ... > Before I can do any more work I need to fix the basic errors. or xDIS for the 49. Meanwhile MASD is in a nearly perfect state. What we urgendly need are the debuggers. As long posters know, I was formerly a JAZZ-fan. But JAZZ doesn't handle arrays which MASD meanwhile does (except symbolic arrays, helas). Therefore, and since Emacs is meanwhile better than ED, I never loaded JAZZ49, also because of the unpleasant port restrictions. I heard from CD and others that they use JAZZ49 only for debugging. - Wolfgang > or xDIS for the 49. Meanwhile MASD is in a nearly > perfect state. What we urgendly need are the debuggers. > I understand that.. However, if I release the Jazz I've modified to use the extable and the system minifont, I want to fix the existing problems as well. I'm not going to release something with my name on it if it contains obvious bugs, question of personal principles Nevertheless, I'm pretty sure that by fixing the existing bugs (which are mainly due to the incorrect use of unsupported entry points) will also make the Jazz debugger works. For example, in the unsupported entries list, many of them (over 50%) the new HP49 address are incorrect. What they used to do on the HP48 are not the same on the HP49. Having access to the source code of both machines, it's easy for me to fix these problems Can u check with the 49 if a serie (or sum) is convergent or not. (with d'alembert criteria, Cauchy...) vince0905@yahoo.fr