B29 Wishlist for HP 49G RPN programmable graphing calculator A. CAS related wishlist 1. Z and Fourier transforms and their inverses 2. support for integral and integro-differential equations 3. support for partial differential equations 4. arbitrary precision floating point math 5. the rest of the special functions (lngamma, bessels, beta, error, spherical harmonics, etc.) 6. the rest of the trigonometric functions [(A)SEC(H), (A)CSC(H), (A)CTN(H)] 7. symbolic/numeric products 8. vector calculus commands sensitive to the coordinate setting 9. vector (and complex) calculus commands using overriding/current angle units 10. complex solver using units (MSLV using units) 11. some kind of auto-simplify function (controlled by a flag) 12. more rewriting tools 13. advanced (linear) solver to handle non (linear) rational cases AND Wlamberts 14. symbolic tensor algebra 15. Heaviside step & Dirac delta as ordinary functions 16. support for hypergeometric functions 17. more Special Numbers (Stirling Numbers & so on & so forth) B. non-CAS related wishlist 1. more built-in curve-fitting options (including user-defined functions) 2. a generally faster plotter (for all graphs) with User Mark setting (grob) 3. combination of the wireframe plotter with the fast-3D plotter 4. optimization of the parametric surface plotter (P.S.P.) 5. possible combination of the P.S.P. with the fast-3D plotter 6. implicit plots 7. fast hidden-line removal algorithm 8. auto-setting of z-coordinates in the window parameters for 3D plots. 9. in the PICTURE environment, a way of slowing down the movement of the cursor 10. textbook display for more calculus functions 11. fractions (as a display mode) 12. thousands separator for all display modes (including integers) 13. timers and a IFTIMER...THEN...ELSE...END structure or similar 14. a programmable StopWatch with splits and all 15. control alarms that are run as subprograms (using Virtual Stack protection) 16. advanced statistical functions: MEAN, Q1, Q4, RELERROR, WMEAN 17. advanced financial solver: cashflows, bonds, combound int, etc. (19B III) 18. EQNLIB and enhance it with a user tailoring ability. 19. CHOOSE submenus 20. TMENU using shift & hold levels 21. ability to use hold-right/left shift keys in all calculator environments Best wishes, --Rahul Hor.8e ==== > B. non-CAS related wishlist 1. more built-in curve-fitting options (including user-defined functions) Isn't this part of the CAS? > 2. a generally faster plotter (for all graphs) with User Mark setting (grob) This too should be part of the CAS. 'CAS' doesn't mean only the symbolic part. Also applies to some items that follow (in the original post). > 16. advanced statistical functions: MEAN, Q1, Q4, RELERROR, WMEAN These aren't really advanced functions. Implemented in MathTools*. * One of my packages for the TI-89/92+: http://triton.towson.edu/users/bbhatt1/ti/MathTools.htm (I just wanted to give an idea of the sorts of things that are available.) -- ==== X > 15. Heaviside step & Dirac delta as ordinary functions Ordinary? What do you mean? The (I)LAP uses thes only inside the EQW I want to use them anywhere as any other function. X > These aren't really advanced functions. Implemented in MathTools*. AAAARRRGGGHHH!!!! > * One of my packages for the TI-89/92+: > http://triton.towson.edu/users/bbhatt1/ti/MathTools.htm > (I just wanted to give an idea of the sorts of things that are > available.) No, instead: How much for porting these tools to the Saturn based 49G? Could take maybe a whole week for you??? > -- > Hmm.. at least Voyage200 is bigger than v'ger BUT It looks smaller than 92 versions... ==== Does your beloved V200 have: -An advanced memory management which enables the built-in math functions to use the whole free RAM ? -Tree directory structure and long names ? -Soft keys support ? -A buzzer to use it as a clock alarm ? -Modular Integer arithmetic ? -Object oriented design ? -Number crunching ? -A user language as powerful as RPL ? -Built-in language equivalent to System RPL and a built-in SDK ? -A command line as powerful as the HP49 one ? -Optimised Home screen display,etc... If no,stop wanting to try to convert HP users to the TI68k or at least try before to convince TI to bring the V200 O.S to even the HP48 O.S level. > Wishlist for HP 49G RPN programmable graphing calculator > > A. CAS related wishlist 1. Z and Fourier transforms and their inverses > > Glenn Fisher and Lars Frederiksen have written Z and Fourier transform > routines for the TI-89/92+. 2. support for integral and integro-differential equations > > Might appear in an update of MathTools*. Volunteers? :-) 3. support for partial differential equations > > Implemented in MathTools* sometime back. 4. arbitrary precision floating point math > > Experimental support in the latest version of MathTools*. 5. the rest of the special functions (lngamma, bessels, beta, error, > spherical harmonics, etc.) > > BesselJ, Beta, Erf, Erfc, spherical harmonics etc. are part of > MathTools*. 7. symbolic/numeric products > > Built-in on the TI-89/92+. 8. vector calculus commands sensitive to the coordinate setting > > Well, in MathTools* I chose to have an extra argument that specifies > either a common coordinate system or the diagonal elements of the > metric. 10. complex solver using units (MSLV using units) > > Built-in on the TI-89/92+. 11. some kind of auto-simplify function (controlled by a flag) > > Built-in on the TI-89/92+ (sometimes it's a blessing, sometimes a > curse :-)). 12. more rewriting tools > > This is kind of vague. Rewriting how? 13. advanced (linear) solver to handle non (linear) rational cases AND > Wlamberts > > The Lambert W function? It's included in MathTools* (it basically > guesses the principal root of the W function and gives that to the > built-in solver to improve speed). 14. symbolic tensor algebra > > There's the Arrays package, by Doug and myself, for the TI-89/92+. 15. Heaviside step & Dirac delta as ordinary functions > > Ordinary? What do you mean? 16. support for hypergeometric functions > > Partially implemented in MathTools* (a lot of cases still need to be > added). 17. more Special Numbers (Stirling Numbers & so on & so forth) > > Implemented in MathTools*. 18. implementation and full integration of INTEGERASSUME > > Built-in on the TI-89/92+. 19. ability to work with and solve recurrence relations > > Implemented in MathTools*. 20. extending FACTORMOD to full factorization rather than square-free only > > This is something that needs to be added to MathTools. Volunteers? :-) 21. implementation of the Zeilberger algorithm in summation command > > Might be added to MathTools* in a future update. Volunteers? :-) 22. bring back XFRC for converting floats to quadratic irrationals > > The PSLQ algorithm also needs to be added to MathTools. Volunteers? > :-) 25. symbolic LU factorization > > Built-in on the TI-89/92+. > > B. non-CAS related wishlist 1. more built-in curve-fitting options (including user-defined functions) > > Isn't this part of the CAS? 2. a generally faster plotter (for all graphs) with User Mark setting (grob) > > This too should be part of the CAS. 'CAS' doesn't mean only the > symbolic part. Also applies to some items that follow (in the original > post). 16. advanced statistical functions: MEAN, Q1, Q4, RELERROR, WMEAN > > These aren't really advanced functions. Implemented in MathTools*. > * One of my packages for the TI-89/92+: > http://triton.towson.edu/users/bbhatt1/ti/MathTools.htm > (I just wanted to give an idea of the sorts of things that are > available.) ==== Does your beloved V200 have: > -An advanced memory management which enables the built-in math > functions to use the whole free RAM ? > -Tree directory structure and long names ? > -Soft keys support ? > -A buzzer to use it as a clock alarm ? > -Modular Integer arithmetic ? > -Object oriented design ? > -Number crunching ? > -A user language as powerful as RPL ? > -Built-in language equivalent to System RPL and a built-in SDK ? > -A command line as powerful as the HP49 one ? > -Optimised Home screen display,etc... If no,stop wanting to try to convert HP users to the TI68k or at least > try before to convince TI to bring the V200 O.S to even the HP48 O.S > level. There ARE areas where the Voyage software is better (49). The thing is just too BIG! The 89 is better (V200/92). TI 89 speed is generally better and the LCD sharper (49). If only would write a similar software package as the MathTools to the 49G, then we would enjoy these goodies, too. Where the 49>89 is: your list above. one more thing: the Matrix Writer in the TI 89 is better. ==== Ditto. Include: 1) Bag Algebraic mode--RPN only (HIST is a waste of valuable keyboard space). 2) alpha-LS-6 angle character, direct access from key (perhaps use the X key-leave alphas to the alpha key) 3) Customizable TOOL menu 4) Customizable APPS menu (without the use of WR's Appsman and the need for USR mode), or least get rid of stuff that's already keyboard accessible and make it easy to append. 5) Array<-->List of Lists mode switch from within MTWR 6) Yes, a stopwatch. 7) Clean up repetition and clutter in menus and menu structure (i.e. MAIN and SYMB) I am thinking, that if there is limited capacity for all these features, then let 19.6 stand as the official commercial release of the ROM and have the development of future ROMs be a departure from the commercial version. Let the new Enhanced ROM head back into the RPN world and get rid of features that were only intended for high school/TI market. Greg S B. non-CAS related wishlist 1. more built-in curve-fitting options (including user-defined > functions) > 2. a generally faster plotter (for all graphs) with User Mark setting > (grob) > 3. combination of the wireframe plotter with the fast-3D plotter > 4. optimization of the parametric surface plotter (P.S.P.) > 5. possible combination of the P.S.P. with the fast-3D plotter > 6. implicit plots > 7. fast hidden-line removal algorithm > 8. auto-setting of z-coordinates in the window parameters for 3D plots. > 9. in the PICTURE environment, a way of slowing down the movement of > the cursor > 10. textbook display for more calculus functions > 11. fractions (as a display mode) > 12. thousands separator for all display modes (including integers) > 13. timers and a IFTIMER...THEN...ELSE...END structure or similar > 14. a programmable StopWatch with splits and all > 15. control alarms that are run as subprograms (using Virtual Stack > protection) > 16. advanced statistical functions: MEAN, Q1, Q4, RELERROR, WMEAN > 17. advanced financial solver: cashflows, bonds, combound int, etc. (19B > III) > 18. EQNLIB and enhance it with a user tailoring ability. > 19. CHOOSE submenus > 20. TMENU using shift & hold levels > 21. ability to use hold-right/left shift keys in all calculator > environments Best wishes, > --Rahul Hor.8e ==== > Ditto. Include: 1) Bag Algebraic mode--RPN only (HIST is a waste of valuable keyboard > space). But I don't think that HP is likely to redesign the 49G at this point, much as we'd like that. You can always assign something else to the HIST key. > 2) alpha-LS-6 angle character, direct access from key (perhaps use the X > key-leave alphas to the alpha key) > 3) Customizable TOOL menu > 4) Customizable APPS menu (without the use of WR's Appsman and the need for > USR mode), or least get rid of stuff that's already keyboard accessible and > make it easy to append. Granted, it would be nice if some of this were built-in, but what's so bad about using Wolfgang's programs? I think that his biggest problems were having unsupported entry points move in new ROMs, and I think he's done a lot to correct those problems. What would no doubt be helpful would be having additional stable entry points (I don't think that we can expect any new officially supported entry points from HP), or at least advance notice of stable entry points that are expected to move in the next ROM. More information about unsupported but stable entry points wouldn't hurt, to put it mildly. And for those of us who want a customized calculator rather than having everything just as the developers think best, it's hard to imagine not using the user mode keyboard. One really can't expect any developer to come up with a design that everyone finds perfect. For that matter, it's quite useful to be able to change key assignments depending on your current task. > 5) Array<-->List of Lists mode switch from within MTWR > 6) Yes, a stopwatch. > 7) Clean up repetition and clutter in menus and menu structure (i.e. MAIN > and SYMB) I agree that the menu structure could use a lot of cleaning up, but some redundancy in the menus is highly desirable. What bothers me are commands that should be in a menu but apparently aren't, and menus that I can't seem to get to without knowing the menu number and using the MENU command. But with so many new commands in the 49G, it's hard to imagine a menu system that makes all commands easy to find and get to. On the 49G, I usually type in most commands. > I am thinking, that if there is limited capacity for all these features, Capacity is always limited; if it's available, then it will be filled. Too bad my bank account doesn't work like that; the principle there seems to be that if it's available, then it will be spent. > then let 19.6 stand as the official commercial release of the ROM and have > the development of future ROMs be a departure from the commercial version. The last official commercial release is 1.18; 1.19-6 is an unofficial Beta ROM. > Let the new Enhanced ROM head back into the RPN world and get rid of > features that were only intended for high school/TI market. I certainly wouldn't mourn the loss of certain features such as the ALG mode, and it can't hurt to have more ROM available for things that are really useful, but I have my doubts about how much flash memory would be recovered by removing ALG mode; it looks to me as if it's just a user interface front end to the real RPL system. I don't expect that anyone would care to take on the task of maintaining both an RPL professional ROM and an algebraic high school/TI ROM, but who knows? Maybe if and when the ROM goes GPL, we could have separate versions of unofficial ROMs (but I certainly wouldn't care to see several competing, similar, but incompatible ROMs). Desirable features can also be added with libraries that the user can install, at his option; this avoids using up system ROM on features that not everyone wants. The disadvantage of relying on a user installed library to fill an apparent gap in the features is that other libraries (and the system) can't count on it being available, so developers who'd like to use something from a different library have to either duplicate what they want or make it a requirement that the other library be installed. Things that can be expected to be used by user installed libraries are best placed in system ROM. -- ==== ==== having a wishlist for the HP 49G, but I guess the only way to find out is to post a semi-updated version. I have separated and starred those items that have been (partly) taken care of (with carriage returns) by either the 49G development team or independent programmers. Here is the list (note that it is categorized, but not yet prioritized): Wishlist for HP 49G RPN programmable graphing calculator A. Computer Algebra System 1. *Z and Fourier transforms and their inverses--(Z transforms, I believe,can be solved by Neopolys 8.1 by Apolonio Francisco Gonzalez Herrero; correct?)* 2. support for integral and integro-differential equations 3. support for partial differential equations 4. *arbitrary precision floating point math--(Werner Huysegoms was working on adding this ability to the 49G; I'm not sure if he's still involved/interested in the project)* 5. *the rest of the special functions (lngamma, bessels, beta, error, spherical harmonics, zeta, etc.)--(Steven Ahlig's Bessel 1.2 gives us some of this functionality)* 6. *the rest of the trigonometric functions [(A)SEC(H), (A)CSC(H), 7. symbolic/numeric products 8. vector calculus commands sensitive to the coordinate setting 9. vector (and complex) calculus commands using overriding/current angle units 10. complex solver using units (MSLV using units) 11. *some kind of auto-simplify function (controlled by a interested)* 12. more (specialized) rewriting tools 13. advanced (linear) solver to handle non (linear) rational cases AND Wlamberts 14. symbolic tensor algebra 15. Heaviside step & Dirac delta as ordinary functions 16. support for generalized hypergeometric functions 17. more Special Numbers (Stirling Numbers, etc.) 18. implementation and full integration of INTEGERASSUME 19. ability to work with and solve recurrence relations 20. extending FACTORMOD to full factorization rather than square-free only 21. implementation of the Zeilberger algorithm in summation command 22. bring back XFRC for converting floats to quadratic irrationals 23. the ability to work with continued fractions (kind of like CFR by Joe Horn) 24. if possible, implement faster, yet full precision, numerical integration 25. symbolic LU factorization B. Graphics/Plotting -- (for this section, the only major improvement in this area that I can recall is the release of Jurjen N.E. Bos' HLDRAW 2.6 that gives us fast, 3D wireframe plotting with hidden line removal; please tell me if I'm missing anything else significant) 1. a generally faster plotter (for all graphs) with User Mark setting (grob) 2. combination of the wireframe plotter with the fast-3D plotter 3. optimization of the parametric surface plotter (P.S.P.) 4. possible combination of the P.S.P. with the fast-3D plotter 5. fast hidden-line removal algorithm 6. implicit plots 7. auto-setting of z-coordinates in the window parameters for 3D plots. 8. in the PICTURE environment, a way of slowing down the movement of the cursor C. Non-CAS Math (Statistics, Finance, Science, etc.) 1. *more built-in curve-fitting options (including user-defined functions)--(Steven Ahlig's Moda 1.52 is an excellent tool for this)* 2. advanced statistical functions: MEAN, Q1, Q4, RELERROR, WMEAN 3. advanced financial solver: cashflows, bonds, compound interest, etc. (19BIII) 4. bring back EQNLIB and enhance it with a user tailoring ability--(many EQNLIBs exist as add-ons; maybe this is something that does not need to be officially added to the ROM--opinions?) D. Operating System 1. *textbook display for more calculus functions--(we have textbook display for limits and exp as of 1.19-6; perhaps indefinite integration, Laplace transforms, etc. should be next?)* 2. fractions (as a display mode) 3. thousands separator for all display modes (including integers) 4. CHOOSE submenus 5. TMENU using shift & hold levels 6. ability to use hold-right/left shift keys in all calculator environments 7. timers and a IFTIMER...THEN...ELSE...END structure or similar 8. a programmable StopWatch with splits and all 9. control alarms that are run as subprograms (using Virtual Stack protection) 10. full hiding capabilities in FILER integrated--(has this been implemented?) 11. bring back STAT soft key menu much like RS-hold + 9 (soft key TIME menu) E. Editor and Debugger -- (here's the place that I know the least about; in my opinion we are very fortunate that we have EMACS to work with as many of these items can be taken care of by that; also JAZZ is available for debugging purposes and much more, but I'm not too sure if it is fully functional yet) 1. clever reintroduction of interruption handler in ObEdit command 2. RPLCPL used as replacement for CAT key 3. ability to create link to HALT command in editor mode as it is deep in menu 4. increase practicality of FIND option 5. generally, a more intelligent editor like in JAZZ or EMACS 6. a built-in SysRPL/Assembly debugger just as we have one for User RPL As you can see we have a pretty lengthy list, albeit a good deal of our goals for this calculator have already been attained. To me, this shows great promise for the HP 49G in the future as well. As I always say, please, if you have time, discuss some of these entries as it may give the HP 49G development team a better picture of what priorities the users of this calculator have in mind (for example, in the CAS section, describe what special functions are most needed and why). I'm not sure of the situation, but as far as I know, there is about 30K left for CAS improvements, and the rest for general improvements (minus what we got in 1.19-6). I'm not sure about the general consensus, but if the 49G team decides to continue improving this already wonderful product, I would not mind sacrificing another user bank (other opinions on this matter?). I cannot possibly do justice to this list all by myself, but I will help in any way I can. If you this thread) and I will add (or subtract, if that be the case) them to (from) the list. I hope to hear your comments/opinions soon. a satisfied user of the HP 49G, --Rahul Hor.8e ==== > having a wishlist for the HP 49G, but I guess the only way to find out > is to post a semi-updated version. I have separated and starred those > items that have been (partly) taken care of (with carriage returns) by > either the 49G development team or independent programmers. Here is > the list (note that it is categorized, but not yet prioritized): I see that http://bugs.hpcalc.org is down at the moment, but I think that you may want to add these (or at least the ones that you still feel are useful) there as Enhancement requests when it's available again. I wouldn't have the foggiest notion of how to use many of these features, but they may indeed be useful to those with more knowledge of mathematics. > Wishlist for HP 49G RPN programmable graphing calculator A. Computer Algebra System 1. *Z and Fourier transforms and their inverses--(Z transforms, I > believe,can be solved by Neopolys 8.1 by Apolonio Francisco Gonzalez > Herrero; correct?)* 2. support for integral and integro-differential equations > 3. support for partial differential equations 4. *arbitrary precision floating point math--(Werner Huysegoms was > working on adding this ability to the 49G; I'm not sure if he's still > involved/interested in the project)* 5. *the rest of the special functions (lngamma, bessels, beta, error, > spherical harmonics, zeta, etc.)--(Steven Ahlig's Bessel 1.2 gives us > some of this functionality)* 6. *the rest of the trigonometric functions [(A)SEC(H), (A)CSC(H), Yes! For those of us who learned trig with pencil, paper, and a trig table, it seems much more natural to use these where that lets us use application provides them, and they can even be implemented in UserRPL, but it seems ridiculous that they aren't built-in commands on the more advanced calculators. > 7. symbolic/numeric products > 8. vector calculus commands sensitive to the coordinate setting > 9. vector (and complex) calculus commands using overriding/current > angle units > 10. complex solver using units (MSLV using units) 11. *some kind of auto-simplify function (controlled by a > interested)* 12. more (specialized) rewriting tools > 13. advanced (linear) solver to handle non (linear) rational cases AND > Wlamberts > 14. symbolic tensor algebra > 15. Heaviside step & Dirac delta as ordinary functions > 16. support for generalized hypergeometric functions > 17. more Special Numbers (Stirling Numbers, etc.) > 18. implementation and full integration of INTEGERASSUME > 19. ability to work with and solve recurrence relations > 20. extending FACTORMOD to full factorization rather than square-free > only > 21. implementation of the Zeilberger algorithm in summation command > 22. bring back XFRC for converting floats to quadratic irrationals > 23. the ability to work with continued fractions (kind of like CFR by > Joe Horn) > 24. if possible, implement faster, yet full precision, numerical > integration > 25. symbolic LU factorization Did you forget? If a command must change a mode setting, first save the current mode, and then restore the starting mode at the end! What's so difficult about that? I even use that technique in my own personal use UserRPL programs; why in the world can't a CAS command manage to do such a simple, obvious thing? > B. Graphics/Plotting -- (for this section, the only major improvement > in this area that I can recall is the release of Jurjen N.E. Bos' > HLDRAW 2.6 that gives us fast, 3D wireframe plotting with hidden line > removal; please tell me if I'm missing anything else significant) 1. a generally faster plotter (for all graphs) with User Mark setting > (grob) > 2. combination of the wireframe plotter with the fast-3D plotter > 3. optimization of the parametric surface plotter (P.S.P.) > 4. possible combination of the P.S.P. with the fast-3D plotter > 5. fast hidden-line removal algorithm > 6. implicit plots > 7. auto-setting of z-coordinates in the window parameters for 3D > plots. > 8. in the PICTURE environment, a way of slowing down the movement of > the cursor C. Non-CAS Math (Statistics, Finance, Science, etc.) 1. *more built-in curve-fitting options (including user-defined > functions)--(Steven Ahlig's Moda 1.52 is an excellent tool for this)* 2. advanced statistical functions: MEAN, Q1, Q4, RELERROR, WMEAN Umm..., isn't MEAN already in there? I mean, it was available as far back as the 28C if you know what I mean. Or do you mean some other kind of MEAN? ;-) > 3. advanced financial solver: cashflows, bonds, compound interest, > etc. (19BIII) > 4. bring back EQNLIB and enhance it with a user tailoring > ability--(many EQNLIBs exist as add-ons; maybe this is something that > does not need to be officially added to the ROM--opinions?) D. Operating System 1. *textbook display for more calculus functions--(we have textbook > display for limits and exp as of 1.19-6; perhaps indefinite > integration, Laplace transforms, etc. should be next?)* 2. fractions (as a display mode) > 3. thousands separator for all display modes (including integers) This would be useful, but perhaps every four digits for hexadecimal and binary displayed user binary integers? > 4. CHOOSE submenus I'm not entirely sure what you mean. For example, if I clear flag -117, then PRG lets me choose BRANCH... and so on, BRANCH... lets me choose IF... and so on, which lets me choose IF, THEN, ELSE, END, or go back to BRANCH.... Maybe you had something else in mind? But it would be nice if I could be in BRANCH..., highlight IF..., and get the nice IF THEN ELSE END structure in the command line by pressing RightShift ENTER or RightShift OK, similar to using soft-key style menus. And as long as we have choose box style menus available, why not have user variables and library commands available in a choose box too? As it is, the choose box menu scheme seems rather incomplete and hard to use. If it were improved, then it might be useful, especially for menus that take more than one page of a soft-key style menu. > 5. TMENU using shift & hold levels > 6. ability to use hold-right/left shift keys in all calculator > environments > 7. timers and a IFTIMER...THEN...ELSE...END structure or similar > 8. a programmable StopWatch with splits and all > 9. control alarms that are run as subprograms (using Virtual Stack > protection) Comments on this in a separate post. > 10. full hiding capabilities in FILER integrated--(has this been > implemented?) > 11. bring back STAT soft key menu much like RS-hold + 9 (soft key TIME > menu) Definitely. Right now, the only way that I know of to get to this is 96 MENU. > E. Editor and Debugger -- (here's the place that I know the least > about; in my opinion we are very fortunate that we have EMACS to work > with as many of these items can be taken care of by that; also JAZZ is > available for debugging purposes and much more, but I'm not too sure > if it is fully functional yet) It would be nice to be able to use Jazz from other ports. Is anyone currently maintaining Jazz? > 1. clever reintroduction of interruption handler in ObEdit command > 2. RPLCPL used as replacement for CAT key > 3. ability to create link to HALT command in editor mode as it is > deep in menu > 4. increase practicality of FIND option It would be nice to be able to use FIND in the command line, instead of just in the editor. Is that what you mean? > 5. generally, a more intelligent editor like in JAZZ or EMACS > 6. a built-in SysRPL/Assembly debugger just as we have one for User > RPL Add: If I start to mark a block with BEGIN, nothing is highlighted until the block is complete (that is, the END key is pressed). If I start to mark a block with END, then the beginning of the object is assumed to be the beginning of the block until I press BEGIN somewhere else. I'd prefer that highlighting start when either the BEGIN or the END key is pressed and follow the cursor until the other key is pressed. And if I use CUT, the cursor jumps back to where the block was. I'd prefer that the cursor stay on the character where I left it, or failing that, at least stay in the same position on the screen. I know, that's probably easier wished for than done, and these are rather minor points. > As you can see we have a pretty lengthy list, albeit a good deal of > our goals for this calculator have already been attained. To me, this > shows great promise for the HP 49G in the future as well. As I always > say, please, if you have time, discuss some of these entries as it may > give the HP 49G development team a better picture of what priorities > the users of this calculator have in mind (for example, in the CAS > section, describe what special functions are most needed and why). > I'm not sure of the situation, but as far as I know, there is about > 30K left for CAS improvements, and the rest for general improvements > (minus what we got in 1.19-6). I'm not sure about the general > consensus, but if the 49G team decides to continue improving this > already wonderful product, I would not mind sacrificing another user > bank (other opinions on this matter?). Well, maybe. I'd prefer to have non-essential applications (the Equation Library and games come to mind) in optional user-installed libraries. Perhaps even some of the things currently in system ROM could be moved to libraries. Essential things, that is, basic RPL operations and operating system type things like entry points that are called from other parts of the system or would likely be useful to optional libraries should be kept in system ROM; use optional libraries for the extras. > I cannot possibly do justice > to this list all by myself, but I will help in any way I can. If you > this thread) and I will add (or subtract, if that be the case) them to > (from) the list. I hope to hear your comments/opinions soon. That's certainly a full plate for the developers, but I don't know of any good reason for it to be done by the ACO team alone. Others are helping with useful add-in applications, and if and when the source code is released under the GPL, they can help more with the system ROM as well, and I expect that the applications should improve with the source code available. The 49G isn't finished (in either sense of the word) yet. > a satisfied user of the HP 49G, -- ==== X > I see that http://bugs.hpcalc.org is down at the moment, but I think > that you may want to add these (or at least the ones that you still feel > are useful) there as Enhancement requests when it's available again. Could you do so, please! I have had trouble with using it and now I have trouble with my Outlook & outlook Express AND my HD has failed sectors... X > Did you forget? If a command must change a mode setting, first save the current mode, > and then restore the starting mode at the end! What's so difficult about > that? I even use that technique in my own personal use UserRPL > programs; why in the world can't a CAS command manage to do such a > simple, obvious thing? RIGHT !! X > 2. advanced statistical functions: MEAN, Q1, Q4, RELERROR, WMEAN Umm..., isn't MEAN already in there? I mean, it was available as far > back as the 28C if you know what I mean. Or do you mean some other kind > of MEAN? ;-) Oops, I did it again! (using Britney's voice :) X > 3. thousands separator for all display modes (including integers) This would be useful, but perhaps every four digits for hexadecimal and > binary displayed user binary integers? YESSS !!! > 4. CHOOSE submenus I'm not entirely sure what you mean. For example, if I clear flag -117, X Take a look at a HP 39/40 > 11. bring back STAT soft key menu much like RS-hold + 9 (soft key TIME > menu) Definitely. Right now, the only way that I know of to get to this is 96 > MENU. Good tip! > E. Editor and Debugger -- (here's the place that I know the least > about; in my opinion we are very fortunate that we have EMACS to work > with as many of these items can be taken care of by that; also JAZZ is > available for debugging purposes and much more, but I'm not too sure > if it is fully functional yet) It would be nice to be able to use Jazz from other ports. Is anyone > currently maintaining Jazz? Avenard is porting this behavior to the MASD, I think. > 4. increase practicality of FIND option It would be nice to be able to use FIND in the command line, instead of > just in the editor. Is that what you mean? No, but doesn't SREPL do that? I want to have a HELP for pattern options and those patterns, too!! X ==== I would like to contribute to this daydreaming session called Wishlist by proposing some extensions to the PLOTTING facilities of the HP48/49. I have provided my HP48 with user- defined commands (in userRPN) similar to the MoveTo LineTo commands which are familiar to everyone who has looked at plotting and graphing under Windows. My plotting is elementary, moving towards consecutive points (x+Dx, y+Dy) etc.(Dx and Dy small increments), and looping, for instance 1 200 START ...NEXT. Two examples: 1. Mapping. You may wish to add (and map) a point (xM,yM), POINT-WISE, to each point (x,y) of a curve y=f(x), in which xM, yM, x, y, are related by functions xM=g(x,y) and yM=h(x,y). A special case of this is conformal mapping, well known to our friends the mathematicians, and also to every-one who has had the patience to produce Julia fractals on the HP. As far as I can see you cannot do that with the standard HP plotting and graphing modes. 2. Consider the x,y relationship defined by the simultaneous differential (DE) equations: dx/dt=x(b-y), dy/dt=y(x-a); a and b positive, for the time being. Famous set, named after Volterra and Lotke (VL), describing the cyclic variations of prey-fish and hunting-fish in certain parts of the Mediterranean. Hunting-fish eat prey-fish; prey will dwindle; hunting-fish will starve because no prey: prey fish will start to multiply again because hardly hunting-fish; prey again for the remaining hunting-fish, etc. Not necessarily similar to the corporate world. You cannot solve and plot that in parameter plotting mode because the equations are D.E.'s. But if you select a suitable increment of t: Dt and a starting point (x0,y0) you can nicely calculate Dx and Dy, MoveTo the next position, and loop, loop. Similar time dependent D.E. 's will describe ballistic problems in which a projectile is chasing a target, or in celestial mechanics, where the equations are in fact Newton's law in 2 or 3 dimensions. In the case of the VL equations you could of course eliminate dt and obtain something like dy/dx= y(x-a) / x(b-x) and MoveTo the D.E. plotting mode of your calculator. Losing t means you lose the dynamics of the system. (you might like to know the duration of the period with a low concentration of both types of fish). But, do it! You will find that you will not get the full picture; DE-plotting won't get you around the extreme values of x, logical but disappointing. So, what is needed is: - extension of parametric plotting towards D.E. 's with t as independent variable - extension of D.E. plotting towards more general x,y relationships, not just neat single-valued y=f(x) functions. I may well have overlooked work of others along similar lines. If so, sorry. Yeb Veli-Pekka Nousiainen schreef in bericht > having a wishlist for the HP 49G, but I guess the only way to find out > is to post a semi-updated version. I have separated and starred those > items that have been (partly) taken care of (with carriage returns) by > either the 49G development team or independent programmers. Here is > the list (note that it is categorized, but not yet prioritized): Wishlist for HP 49G RPN programmable graphing calculator A. Computer Algebra System 1. *Z and Fourier transforms and their inverses--(Z transforms, I > believe,can be solved by Neopolys 8.1 by Apolonio Francisco Gonzalez > Herrero; correct?)* 2. support for integral and integro-differential equations > 3. support for partial differential equations 4. *arbitrary precision floating point math--(Werner Huysegoms was > working on adding this ability to the 49G; I'm not sure if he's still > involved/interested in the project)* > > 5. *the rest of the special functions (lngamma, bessels, beta, error, > spherical harmonics, zeta, etc.)--(Steven Ahlig's Bessel 1.2 gives us > some of this functionality)* 6. *the rest of the trigonometric functions [(A)SEC(H), (A)CSC(H), 7. symbolic/numeric products > 8. vector calculus commands sensitive to the coordinate setting > 9. vector (and complex) calculus commands using overriding/current > angle units > 10. complex solver using units (MSLV using units) 11. *some kind of auto-simplify function (controlled by a > interested)* 12. more (specialized) rewriting tools > 13. advanced (linear) solver to handle non (linear) rational cases AND > Wlamberts > 14. symbolic tensor algebra > 15. Heaviside step & Dirac delta as ordinary functions > 16. support for generalized hypergeometric functions > 17. more Special Numbers (Stirling Numbers, etc.) > 18. implementation and full integration of INTEGERASSUME > 19. ability to work with and solve recurrence relations > 20. extending FACTORMOD to full factorization rather than square-free > only > 21. implementation of the Zeilberger algorithm in summation command > 22. bring back XFRC for converting floats to quadratic irrationals > 23. the ability to work with continued fractions (kind of like CFR by > Joe Horn) > 24. if possible, implement faster, yet full precision, numerical > integration > 25. symbolic LU factorization B. Graphics/Plotting -- (for this section, the only major improvement > in this area that I can recall is the release of Jurjen N.E. Bos' > HLDRAW 2.6 that gives us fast, 3D wireframe plotting with hidden line > removal; please tell me if I'm missing anything else significant) 1. a generally faster plotter (for all graphs) with User Mark setting > (grob) > 2. combination of the wireframe plotter with the fast-3D plotter > 3. optimization of the parametric surface plotter (P.S.P.) > 4. possible combination of the P.S.P. with the fast-3D plotter > 5. fast hidden-line removal algorithm > 6. implicit plots > 7. auto-setting of z-coordinates in the window parameters for 3D > plots. > 8. in the PICTURE environment, a way of slowing down the movement of > the cursor C. Non-CAS Math (Statistics, Finance, Science, etc.) 1. *more built-in curve-fitting options (including user-defined > functions)--(Steven Ahlig's Moda 1.52 is an excellent tool for this)* 2. advanced statistical functions: MEAN, Q1, Q4, RELERROR, WMEAN > 3. advanced financial solver: cashflows, bonds, compound interest, > etc. (19BIII) > 4. bring back EQNLIB and enhance it with a user tailoring > ability--(many EQNLIBs exist as add-ons; maybe this is something that > does not need to be officially added to the ROM--opinions?) D. Operating System 1. *textbook display for more calculus functions--(we have textbook > display for limits and exp as of 1.19-6; perhaps indefinite > integration, Laplace transforms, etc. should be next?)* 2. fractions (as a display mode) > 3. thousands separator for all display modes (including integers) > 4. CHOOSE submenus > 5. TMENU using shift & hold levels > 6. ability to use hold-right/left shift keys in all calculator > environments > 7. timers and a IFTIMER...THEN...ELSE...END structure or similar > 8. a programmable StopWatch with splits and all > 9. control alarms that are run as subprograms (using Virtual Stack > protection) > 10. full hiding capabilities in FILER integrated--(has this been > implemented?) > 11. bring back STAT soft key menu much like RS-hold + 9 (soft key TIME > menu) E. Editor and Debugger -- (here's the place that I know the least > about; in my opinion we are very fortunate that we have EMACS to work > with as many of these items can be taken care of by that; also JAZZ is > available for debugging purposes and much more, but I'm not too sure > if it is fully functional yet) 1. clever reintroduction of interruption handler in ObEdit command > 2. RPLCPL used as replacement for CAT key > 3. ability to create link to HALT command in editor mode as it is > deep in menu > 4. increase practicality of FIND option > 5. generally, a more intelligent editor like in JAZZ or EMACS > 6. a built-in SysRPL/Assembly debugger just as we have one for User > RPL As you can see we have a pretty lengthy list, albeit a good deal of > our goals for this calculator have already been attained. To me, this > shows great promise for the HP 49G in the future as well. As I always > say, please, if you have time, discuss some of these entries as it may > give the HP 49G development team a better picture of what priorities > the users of this calculator have in mind (for example, in the CAS > section, describe what special functions are most needed and why). > I'm not sure of the situation, but as far as I know, there is about > 30K left for CAS improvements, and the rest for general improvements > (minus what we got in 1.19-6). I'm not sure about the general > consensus, but if the 49G team decides to continue improving this > already wonderful product, I would not mind sacrificing another user > bank (other opinions on this matter?). I cannot possibly do justice > to this list all by myself, but I will help in any way I can. If you > this thread) and I will add (or subtract, if that be the case) them to > (from) the list. I hope to hear your comments/opinions soon. a satisfied user of the HP 49G, > --Rahul Hor.8e ==== I have the 49g and would like to use my infrared printer from my old hp 48g. Any hardware hacks for using the cable recommended or how to ==== > I have the 49g and would like to use my infrared printer from my old > hp 48g. Any hardware hacks for using the cable recommended or how to If you mean for wiring directly to the printer, it's on my wish list too! Actually, I've always wished that they'd designed the printer to accept either wire or infrared input; all too often I've broken the IR connection by picking up the calculator or putting something in the IR path, and there's no way for the printer to ask for a resend or tell the calculator when its buffer is almost full or it's ready to receive more data. On the other hand, with a wire, I'd be connecting and disconnecting a lot, and I'm not sure how well the connections would hold up over the years. Presumably, if one had the schematics for the printer, one could come up with a way to make a wired connection. Of course, there's a possibility that internal changes may have been made to the printer over the years. Marcel Flipse was working on an IR adapter for use with the 49G, but I don't know that he ever developed it to the point where it would send in the format required by the printer. Search for his name on www.hpcalc.org. For information on the signals used by the calculator for ordinary serial I/O, see www.hpcalc.org/details.php?id=1722 or www.hpcalc.org/details.php?id=4741. Those are for the 48 series, but I expect that the 49G would be very similar. For information on the IR signals needed by the printer, see www.hpcalc.org/details.php?id=4742. If you're willing to use your 48G for a printer adapter, you can certainly do it that way using just the cable and adapter that came with your 49G. But it's no surprise that the 48 will error out on the new 49G object types, ZINTs will be converted to reals, and the new command names will be treated as variable names. My initial approach was to send such objects as character strings, using ->STR, DECOMP$, or EDITDECOMP$ to change the object type to something that the 48 can handle. The problems with that include that in ASCII transmission mode, the 49G translates the NUL character, , and to 00, , and , and the 48 doesn't understand that any such translation has been made. The 00 and can be corrected on the 48 easily enough, but, for example, ab will be sent as ab and the 48 will treat this as the string a, followed by b, followed by a beginning string delimiter, and helpfully supply the missing ending string delimiter, so now you have three objects where you wanted only one; it's a bit strange when a variable on the 48 contains what looks sort of like 3 objects that TYPE says is a program. This makes it rather difficult to send lists or programs as strings if they contain strings within them, as well as character strings that include a literal within them. I can get around this by sending a string in binary transmission mode and chopping off the header (but don't do this with just any object, even though it works for character string objects). Too bad that there isn't a way to force the 49G to send character strings that contain any literal as counted strings. This would improve readability in an editor on the PC too, in my opinion. The control sequences for style and font changes in a string are a bit problematic too. Ralf Fritzsch mentioned using the ->GROB commands to make an object that the 48 understands. This avoids the problems that may come up with using strings. Using the ->GROB command seems to be a lot easier, even though GROBs tend to be rather large and so need more memory on both calculators and take longer to send and are printed more slowly. My approach with strings was to send everything to the 48 and let me get back to work on the 49G while the 48 handled the printing and purging, but the memory requirements for GROBs suggest that it may be better to send them one at a time, waiting for the 48 to finish sending it to the printer and get back to an Awaiting Server Cmd. status before sending the next GROB. Or maybe check that the 48 still has enough memory to handle the next object before sending it? That sounds feasible, and preserves the get back to work as soon as possible philosophy, but checking when possible actually occurs. ->GROB has the advantage that styles, fonts, and EquationWriter format can be preserved. For objects that won't fit on one line, before the ->GROB, you can use DECOMP$, EDITDECOMP$, or, I believe that there are other SysRPL commands that wrap or wordwrap strings. When printing programs, algebraics or lists intended as procedural objects, or objects intended as input, be sure to use STD display mode and wordsize 64 to avoid having some digits rounded off or ignored; note that EDITDECOMP$ always uses STD and wordsize 64 when making the string. Note that the SEND command works with local variables too; you don't have to use a global variable to temporarily store the object on the sending calculator. That makes things a bit easier. It will be stored on the receiving calculator as a global variable though, so you'll have to take care of purging it when it's no longer needed, or you could just can recycle the variable. Assuming that the 48 is in Kermit server mode, you can use the PKT command to control the 48 from the 49G. For example: 't' DUP RCL SWAP PURGE PR1 C PKT on the 49G purges the variable 't' on the 48 and prints its contents. Note that the command string has to fit in one single Kermit packet; a bit less than 90 characters when you allow for Kermit's overhead, if I recall correctly. If you want to do something more, you can SEND a program in a variable (or have a program already stored on the 48), and then use the PKT command to execute the variable. Hmm.... Since I abandoned my print server project, I've been using a 48SX to print out 49G programs one at a time, but writing this makes it sound a lot easier (or maybe I've forgotten something), so maybe it's about time for getting back to it. -- ==== > X > let's hope KEEP, KEEP2, KEEPN are commands in the next ROM. Good idea; would you add it to http://bugs.hpcalc.org as an > Enhancement request? I'd thought of KEEP as being used as KEEPN would > be, but having all three available as commands may have some merit. The > KEEP operation in the interactive stack seems to be very fast, so I > guess that the underlying code is already in the ROM and it's a matter > of making it available as user commands. X While waiting one can use: << -> x << CLEAR x >'KEEP' << -> x y << CLEAR x y >'KEEP2' << -> n << n ->LIST -> l << CLEAR l >>'KEEPN' One could easily ask for SWAP2, SWAPN, OVER2, OVERN maybe after NIP is introduced once could want NIP2, NIPN and perhaps TUCK, TUCK2, TUCKN (While NIP is SWAP DROP, TUCK is SWAP OVER) > It depends on what we mean by memory efficiency; do we mean the most > memory actually required at any place in the program, or do we count > memory used even temporarily and available for reuse after a garbage > collection? On how small user memory will it work? In other words: minimize even any temporary large memory usage. ==== > X > let's hope KEEP, KEEP2, KEEPN are commands in the next ROM. Good idea; would you add it to http://bugs.hpcalc.org as an > Enhancement request? I'd thought of KEEP as being used as KEEPN would > be, but having all three available as commands may have some merit. The > KEEP operation in the interactive stack seems to be very fast, so I > guess that the underlying code is already in the ROM and it's a matter > of making it available as user commands. > X > While waiting one can use: > << -> x << CLEAR x >'KEEP' > << -> x y << CLEAR x y >'KEEP2' > << -> n << n ->LIST -> l << CLEAR l >>'KEEPN' Good point; I wanted to avoid local variables for reasons of speed, but in this case, saving part of the stack in a local variable while clearing the rest of it may indeed be the fastest way do it. Of course, this does have the problem of arbitrarily discarding whatever may have been on the stack above the program's arguments. One could save this in a list, and later restore it. > One could easily ask for SWAP2, SWAPN, I take it that SWAP2 would change the stack from ob4 ob3 ob2 ob1 to ob2 ob1 ob4 ob3, and SWAPN from ob2n ... obn+1 obn ... ob1 to obn ... ob1 ob2n ... obn+1 (not counting the argument for SWAPN)? > OVER2, OVERN You lost me there. I've always thought of OVER as having the same effect as 2 PICK. On the 49G we already have OVER, PICK3, and PICK; what would OVER2 and OVERN do? > maybe after NIP is introduced once could want NIP2, NIPN > and perhaps TUCK, TUCK2, TUCKN > (While NIP is SWAP DROP, TUCK is SWAP OVER) Sometimes my use of the calculator is rather nip and tuck :-). One could wish for these, but for now those variations on NIP can be done with variations on ROLL DROP, and TUCK and its kin with variations on DUP ROLLD. Note that SWAP is similar to either 2 ROLL or 2 ROLLD, ROT matches 3 ROLL, and UNROT matches 3 ROLLD (or ROT ROT). I'm glad to have NIP; my 48 series programs use SWAP DROP sequences fairly often. How did they ever come up with the name NIP though? When I first saw UNPICK, I had my doubts as to whether I'd ever use it, but it turns out to be very useful indeed, and much faster than rolling the stack, dropping the unwanted object, and unrolling the stack. > It depends on what we mean by memory efficiency; do we mean the most > memory actually required at any place in the program, or do we count > memory used even temporarily and available for reuse after a garbage > collection? > On how small user memory will it work? > In other words: minimize even any temporary large memory usage. Hmm... For example, if a program uses REVLIST, then, in addition to whatever memory is already in use before REVLIST, at least enough memory for the reversed list is needed before the original list is discarded. But the memory used by the original list isn't available for use (assuming that there are no other references to it) until a garbage collection is done. Garbage collection slows drastically when you have several pointers to list elements on the stack. I'm not sure why it slows down so much. Note that when you have an element from a list on the stack, unless you use NEWOB on it, you really have just a pointer into the list on the stack, which uses just another 2.5 bytes, but for the sake of the pointer, the list has to be kept in memory and marked as being referenced (in use) as long as any pointer to it is on the stack. It follows that when several pointers to a list are on the stack, it would be referenced several times. But it seems to me that the garbage collection would just have to check whether or not an object is referenced, and not be concerned at all about how many times it's referenced. One possible explanation would be that instead of (or perhaps in addition to) the list itself being marked as referenced, the individual elements within the list are marked as referenced; I can see how that would slow down garbage collection quite a bit. Or maybe I'm confused about how garbage collection really works. I believe that other composites, and maybe arrays, are handled similarly to lists in this respect. -- ==== X > One could easily ask for SWAP2, SWAPN, I take it that SWAP2 would change the stack from ob4 ob3 ob2 ob1 to ob2 > ob1 ob4 ob3, and SWAPN from ob2n ... obn+1 obn ... ob1 to obn ... ob1 > ob2n ... obn+1 (not counting the argument for SWAPN)? OVER2, OVERN You lost me there. I've always thought of OVER as having the same effect > as 2 PICK. On the 49G we already have OVER, PICK3, and PICK; what would > OVER2 and OVERN do? copy 2 or N object above 2 or N objects. Look at your own SWAP2, SWAPN explanations. Avenard sayz that any new Lib 2 commands are impossible so we just have to use what we have... The Wolfgang's KEYMAN is a blessing. PS: The names NIP and TUCK are old FORTH command names. ==== > The KEEP operation in the interactive stack seems to be very fast, so I > guess that the underlying code is already in the ROM and it's a matter > of making it available as user commands. > While waiting one can use: > << -> x << CLEAR x >'KEEP' With the latest Keyman even someone who doesn't have the slightest idea on programming can program this kind of Keep the level 1 object and drop anything else. He has only to record the 4 key presses necessary to realize this normally: UpArrow (Interactive stack), then NXT and F4 (KEEP), and finally ENTER. Do longpress K&SA. Then press the 4 keys as given above, and finally CANCEL to end the recording. The result is equivalent to the KEEP above. It is the SysRPL-program 10d 13d 4d 51d 4d XLIB 1200 13 I omitted here the bint symbol 164. There is no special code for the KEEP operation of the interactive stack. It doesn't pay since this operation, with input %n from outside and error-protected, is simply :: CKN PushMetaVstack&Drop PopVstack ; This has 3 commands only and is very fast. And it would cost JYA only 12.5 bytes to add this to UsrRPL :-) . Wolfgang ==== > realize this normally: UpArrow (Interactive stack), > then NXT and F4 (KEEP), and finally ENTER. That won't work as KEEP there is using the current stack level, in your situation: level 1 > :: CKN PushMetaVstack&Drop PopVstack ; Well, there's actually an entry point for the KEEP which is not using the virtual stack at all. =KEEP EQU #28335 It's in the stable area. Full ML so should be quite fast. This has 3 commands only and is very fast. And it would > cost JYA only 12.5 bytes to add this to UsrRPL :-) It would cost much more. A few bytes for the name, there's the hash-table and link table. Also, this new command would have to be added in a covered library (e.g. not in library 2) and you won't be able to call this command with an address but with a ROMPTR only (like HEAD, TAIL etc...) ==== > UpArrow (Interactive stack), then NXT and F4 (KEEP), and finally ENTER. > That won't work as KEEP there is using the current stack level, in your > situation: level 1 It works perfect (but as I said, only for keeping level 1 :-) > Well, there's actually an entry point for the KEEP which is not using the > virtual stack at all. > =KEEP EQU #28335 entry list. It isn't error-protected. The level 1 bint must be smaller than the stack depth. Thus, the shortest and fastest ERROR-PROTECTED KEEP macro which takes a real or zint as input seems to be :: CKN PTR 28335 ; To my experience the KEEP command isn't really needed for programming. But since page 4 of the STACK menu has 4 empty places, these could be used for the following additonal stack-operations, for the benefit of UsrRPL: SWAP2 (1 2 3 4 -> 3 4 1 2) OVER2 (1 2 3 4 -> 1 2 3 4 1 2) SWAPDUP (1 2 -> 2 1 2) There is still a place left which I suggest to fill with the single quoter x' (PTR 389EF), to be able to quote names and other objects inside lists. That this is impossible in UsrRPL is a serious, and by no means necessary restriction of UsrRPL, compared to SysRPL. It is not difficult to adapt the builtin compiler so that it can handle also this UsrRPL expansion as has been demonstrated in the library ExtUsrRPL. - Wolfgang ==== > SWAPDUP (1 2 -> 2 1 2) Hmmm...If you mean SWAP OVER it has already a name in Forth: TUCK Lets use it! PS: magnificent suggestions, oh Professor! X > for programming. But since page 4 of the STACK menu has > 4 empty places, these could be used for the following > additonal stack-operations, for the benefit of UsrRPL: SWAP2 (1 2 3 4 -> 3 4 1 2) > OVER2 (1 2 3 4 -> 1 2 3 4 1 2) > SWAPDUP (1 2 -> 2 1 2) There is still a place left which I suggest to fill > with the single quoter x' (PTR 389EF), to be able to > quote names and other objects inside lists. That this > is impossible in UsrRPL is a serious, and by no means > necessary restriction of UsrRPL, compared to SysRPL. > It is not difficult to adapt the builtin compiler so > that it can handle also this UsrRPL expansion as has > been demonstrated in the library ExtUsrRPL. - Wolfgang ==== > entry list. It isn't error-protected. The level 1 bint > must be smaller than the stack depth. Thus, the shortest you can do some testing on it... It comes back saying you're on holidays. Why would anybody take holidays in winter ???? ==== > entry list. It isn't error-protected. The level 1 bint I'm pretty sure I've given this entry point before. > must be smaller than the stack depth. Thus, the shortest > and fastest ERROR-PROTECTED KEEP macro which takes a > real or zint as input seems to be > :: CKN PTR 28335 ; If it's a SysRPL keep that you want, why would you want it to be protected ? To my experience the KEEP command isn't really needed > for programming. But since page 4 of the STACK menu has > 4 empty places, these could be used for the following > additonal stack-operations, for the benefit of UsrRPL: SWAP2 (1 2 3 4 -> 3 4 1 2) > OVER2 (1 2 3 4 -> 1 2 3 4 1 2) > SWAPDUP (1 2 -> 2 1 2) > It is very easy to add commands, however, none of these commands will be part of the existing library 2 (like SWAP, DUP etc...) as otherwise you would change some fixed entry points. You would also break all compatibility between different version of the HP49G. ==== Too bad that ANS is not a function in RPN mode but I circumvent it by starting the 'Algebraic' with shift held down to get `Algebraic` Example: I'm reusing an Algebraic from stack level 4 I key in `ANS(4)+X` to get X added to the Algebraic. I find this behavior useful. > PS: There's no such thing as ANS in RPN mode as the historic is not saved in the > stack, so you can't easily retrieve it. It seems to me that using ANS in ALG mode is sort of like using > something from the stack in RPN mode. Actually, it turns out that ANS does indeed do *something* in RPN mode, > although pressing LeftShift ANS does LASTARG instead. ANS in RPN mode > seems to be similar to PICK, except that the argument to ANS is counted > as a stack level if the argument is on the stack, but not if the > argument for ANS is still in the command line. So if we want to put a > copy of level n on the stack we could use n PICK, or n 1 + ANS, or n (in > the command line, and while the command line is still active) ANS. I'm > fairly certain that I'll never use it. > > > ==== > Too bad that ANS is not a function in RPN mode > but I circumvent it by starting the 'Algebraic' > with shift held down to get `Algebraic` > Example: > I'm reusing an Algebraic from stack level 4 > I key in `ANS(4)+X` > to get X added to the Algebraic. Ok, I'd just do 4PICK'X'+, which seems at least as easy to me. > I find this behavior useful. Whatever works for you. -- ==== Don't they ALL do it? I'm almost always in degrees mode, while using my 49G for physics. I use units, but EQW49 doesn't support them so I use the RPN stack a lot. The Math is done in RAD mode. One could set the Vectored ENTER flag and the User Mode the write a betaENTER at home << { -17 } CF DROP >@ { } is optional with just one flag where you can clear (or set using SF) several flags OR << { # 60A00B8685000FF1h } STOF DROP >{ } opt with one # One may want use the full flag set provided by the RCF Then I would still use only the first three # in order to have some user flags which don't change AND to perhaps use the user flags to control the STOF argument. Say: << CASE 101 FS? THEN :F:flags1 STOF END 102 FS? THEN :F:flags2 STOF END ... END DROP > Note that when you have to use the DROP like this you loose the LASTARG contents. You may want to (disable LASTARG and) use UNDO instead. I hope this helps everyone using this calculator. > Why the HP49 Change the mode from Deg to Rad when I use SUBST or > SOLVE? > Is there any way to avoid this? I know can set the flag 120, so the calculator lets me know before > change the setting, but this is really annoying. Maybe I can use MATCH (up arrowMATCH, I mean) but this doesn't help > with SOLVE, is there any list of the functions that change some > settings in the calculator? > Marino Arturo ==== Avenard! Is it possible to get back the Vectored ENTER behavior in the Algebraic Mode. At least the BetaENTER. Many Finnish HP 49G customer are now satisfied with their calculators by setting flags: -62, -63, -120 AND then using betaENTER as follows: << { -17 -105 } CF DROP>> BUT The Algebraic users are NOT happy. Even sacrificing the Autoquote comes in mind... > Don't they ALL do it? I'm almost always in degrees mode, while using my 49G > for physics. I use units, but EQW49 doesn't support them > so I use the RPN stack a lot. The Math is done in RAD mode. One could set the Vectored ENTER flag and the User Mode > the write a betaENTER at home > << { -17 } CF DROP >@ { } is optional with just one flag > where you can clear (or set using SF) several flags > OR > << { # 60A00B8685000FF1h } STOF DROP >{ } opt with one # X ==== I have just down loaded the latest development version of Erable 3.024 from http://perso.wanadoo.fr/bernard.parisse/english.html . After I run the install program, the hp48GX comes up in high resolution (little characters ) display mode. After using the HP48GX for awhile, the screen goes back to it's normal resolution. I am sure that I hit some key sequence the changes the display mode. Yesterday during class, it came back for awhile and then went away again. Does any one know what is going on here? Any information that you can nice day. ==== I need help on the following: For my finals I was trying to put all the more complicated formulae into my HP49G calculator...first as simple functions, but then I realized that there's no way to see what variables need to be given as arguments unless you memorize the function. So I tried to make a little program using the INPUT command that prompts the user for all the needed values and then does the calculation. So far so good...BUT what I really want is a program that only prompts for the values of needed variables if the variable does not yet exist. Also my program is pretty BIG with all those INPUT commands for each variable...but I have no clue what else to do. Does anybody have an idea and/or a nice and easy little program to perform the aforementioned tasks? Any help is appreciated! Gunnar ==== It's not exactly what you asked for but use the numeric solver. It will display a form with all the variables in the equation and their value if they exist. It has the added benefit that you can then solve the equation for any variable not just evaluate the right hand side of the equation. I need help on the following: For my finals I was trying to put all the more complicated formulae > into my HP49G calculator...first as simple functions, but then I > realized that there's no way to see what variables need to be given as > arguments unless you memorize the function. So I tried to make a > little program using the INPUT command that prompts the user for all > the needed values and then does the calculation. > So far so good...BUT what I really want is a program that only prompts > for the values of needed variables if the variable does not yet exist. > Also my program is pretty BIG with all those INPUT commands for each > variable...but I have no clue what else to do. Does anybody have an idea and/or a nice and easy little program to > perform the aforementioned tasks? > Any help is appreciated! Gunnar ==== If have 'a.x+b' defined in the variable 'y' then you can use the numeric solver and EXPR= to drop the result on the stack. Or you can change the format of your equation to 'y=a.x+b' stored in some variable fy. The numeric solver can then be used to solve for any of the variables. The numeric solver CHOOSE option allows you to use the filer to select anything you have stored and you can retrieve anything you like from the stack with the CALC option. Have fun. > It's not exactly what you asked for but use the numeric solver. It will > display a form with all the variables in the equation and their value if > they exist. It has the added benefit that you can then solve the equation > for any variable not just evaluate the right hand side of the equation. > I need help on the following: For my finals I was trying to put all the more complicated formulae > into my HP49G calculator...first as simple functions, but then I > realized that there's no way to see what variables need to be given as > arguments unless you memorize the function. So I tried to make a > little program using the INPUT command that prompts the user for all > the needed values and then does the calculation. > So far so good...BUT what I really want is a program that only prompts > for the values of needed variables if the variable does not yet exist. > Also my program is pretty BIG with all those INPUT commands for each > variable...but I have no clue what else to do. Does anybody have an idea and/or a nice and easy little program to > perform the aforementioned tasks? > Any help is appreciated! Gunnar ==== I installed Eclib on my 49G, but it does not work, it gives me a STO Error: Undefined XLIB Name. Someone tolled me that I probably had the wrong compressor installed, but I have tried all of these and still NO GO. I have tried BZ49, BZu49, BZ, bz49lib & Bzlib will not install because its for 48. Has anybody had success installing Eclib on their 49G? Is so, Please Help. ==== I had the same problem with eclib at first. And yes the problem has probably something with the (wrong) compressor to do. If you get the Error: Undefined XLIB Name you also should get a number on the stack. Note this number. Compare the number you get with the library-listing from Carsten Dominic on www.hpcalc.org ; this way I figured out, which compressor I need to use. the XLIB-Error and I will tell you which compressor you need. (Assuming the problem is related to the compressor at all) BTW I don't have eclib anymore on the calc and I also can't remember which compressor I installed. Sorry about that. But BZLIB is also for the HP49G avaliable as well. Hope that helps Roman PS: Which ROM-version are you using, 1.19-6? Jorge G. schrieb im Newsbeitrag > I installed Eclib on my 49G, but it does not work, it gives me a STO > Error: Undefined XLIB Name. Someone tolled me that I probably had the > wrong compressor installed, but I have tried all of these and still NO > GO. I have tried BZ49, BZu49, BZ, bz49lib & Bzlib will not install > because its for 48. > > Has anybody had success installing Eclib on their 49G? Is so, Please > Help. ==== > I had the same problem with eclib at first. And yes the problem has probably > something with the (wrong) compressor to do. If you get the Error: Undefined > XLIB Name you also should get a number on the stack. Note this number. > > Compare the number you get with the library-listing from Carsten Dominic on > www.hpcalc.org ; this way I figured out, which compressor I need to use. > > the XLIB-Error and I will tell you which compressor you need. (Assuming the > problem is related to the compressor at all) > > BTW I don't have eclib anymore on the calc and I also can't remember which > compressor I installed. Sorry about that. > But BZLIB is also for the HP49G avaliable as well. > Hope that helps > Roman > > PS: Which ROM-version are you using, 1.19-6? Jorge G. schrieb im Newsbeitrag > I installed Eclib on my 49G, but it does not work, it gives me a STO > Error: Undefined XLIB Name. Someone tolled me that I probably had the > wrong compressor installed, but I have tried all of these and still NO > GO. I have tried BZ49, BZu49, BZ, bz49lib & Bzlib will not install > because its for 48. Has anybody had success installing Eclib on their 49G? Is so, Please > Help. it was not giving me a number on my stack, but I did research that list, and it ended up being that the Library I needed was (1788 BZ+49) not (1700 BZLIB) I guess I should have just read the Eclib instructions better. [CCedilla]ççIMPORTANT!!! THIS PROGRAM WILL ONLY WORK IF YOU HAVE BZ+49 COMPRESSOR LIBRARY INSTALLED. Unfortunately I didn't know the difference between one BZ from the other BZ. BZ49, BZu49, BZ, bz49lib, Bzlib , & BZ+49 they all look alike to me. Well thanks a lot for leading me in the right direction. It seems to be working know. ==== Does anyone know how to obtain these communication settings with HP8GX : baud rate 1200 data bits 8 stop bits 2 parity even flow control none ? I made these settings and when I try to transfer some bytes to Hyper Terminal different characters came out (XMIT command). ==== > Does anyone know how to obtain these communication settings with HP8GX : > baud rate 1200 > data bits 8 > stop bits 2 > parity even > flow control none ? See pages 27-15 through 27-19 in the User's Guide for information on changing I/O parameters. If you also have the Advanced User's Guide, then look in appendix D for information on the reserved variable IOPAR. As far as I know, these calculators always send two stop bits, and expect to receive at least one. I don't think that there's any way to change that. You always have 8 data bits when using the 48, but I believe that if you have parity on, the most significant data bit is used for parity, limiting you to characters in the range of 0 through 127. Maybe that's where you have the problem. Why not use 9600 baud? You might as well turn software flow control on. On the 48, change IOPAR so that the third and fourth parameter are non-zero. Your PC can very likely handle anything that the 48 sends it, but I doubt that the 48 will be able to handle data as fast as your PC can send it, without using software flow control or the built-in XModem or Kermit protocols. See www.hpcalc.org/details.php?id=1722 or www.hpcalc.org/details.php?id=4741 for more information on I/O on the 48 series. > I made these settings and when I try to transfer some bytes to Hyper > Terminal > different characters came out (XMIT command). personal use), and have never had a problem with it; it even receives correctly no matter what I set HyperTerminal's speed to. I guess it must adjust to the incoming data automatically. But I've never tried it with parity set to anything but none. Visit www.hilgraeve.com for more information about this product. -- ==== Me gustar.92a saber si alguien en este grupo de noticias habla espa.96ol. De se Asi seri bueno intercambiar informaci.97n. ==== 1.19-6, which made a large difference in the slowdowns I was experiencing. I will save your letter, in case it comes back! Steve Smith > >>The calculator seems to do garbage collection quite frequently. > > >>Very often I try to simply add two numbers >>and have to wait several seconds > > >>Is there anything that can be done to improve this? > > >Actually the GCs will be much less noticeable (shorter, that is) >if they are caused to occur *more* frequently, akin to how >each cleanup of a household is quite a bit less of a task >if it is performed more often, rather than waiting >until there's a huge amount of reorganizing to do, >which could then keep you busy for an entire weekend, >instead of just a quick tidy-up before going partying. > >To cause GC's to be more frequent but much shorter, >choose a smaller amount of memory to leave free in the calc, >e.g. 30000 FMEM -- now the calc behaves like a good old HP48G (32K), >which never kept anybody waiting very long :) > >Whenever you want your free memory expanded again, >to store or download more or bigger stuff, >1E7 FMEM will expand it to its maximum >(i.e. all the free mem that's left). > >Of course you can keep re-using FMEM to adjust the free memory >upwards or downwards, whenever you want, up to a maximum of all >the user memory (for variables + port 0) that's available. > >You can also use it to find out how much free memory is needed >to run various applications, by lowering free memory until >the application complains of Insufficient Memory. > >The input to FMEM is how much memory you want to leave free; >the output is how much memory is actually left free. >FMEM for HP49: >(do HOME 256 ATTACH before downloading or typing the program, > because MAKESTR needs built-in library 256 to compile properly) << :0:GCJUNK DUP PURGE MEM ROT - > IF DUP 0. > THEN MAKESTR SWAP STO > ELSE DROP2 END MEM > >FMEM for HP48 (try it if you've got 128K or merged 256K): << :0:GCJUNK DUP PURGE MEM ROT - > IF DUP 0. > THEN SWAP 2. * > :COERCE: #18CEAh SYSEVAL :EXPAND: #61C1Ch SYSEVAL > SWAP STO ELSE DROP2 END MEM > > >The standard precaution applies: >Backup your memory before trying any new program! [r->] [OFF] ==== > Me gustar.92a saber si alguien en este grupo de noticias habla espa.96ol. De ser > Asi seria bueno intercambiar informaci.97n. Hola, el grupo de noticias sobre las Calculadoras HP en espa.96ol es: es.comp.sistemas.hp48 - Dante Aron.92 C. ==== pero ese grupo esta casi muerto Dante Aron.92 C. escribi.97 en el mensaje > Me gustar.92a saber si alguien en este grupo de noticias habla espa.96ol. De ser > Asi seria bueno intercambiar informaci.97n. Hola, el grupo de noticias sobre las Calculadoras HP en espa.96ol es: > es.comp.sistemas.hp48 > - Dante Aron.92 C. ==== > pero ese grupo esta casi muerto > Muerto??? Bueno, considerando que es verano es normal que no tengas muchos posts.. pero yo no lo llamar.92a muerto ;-) Un saludo, J.Manrique Club de Usuarios de Gij.97n #1077 HPCC Member > Dante Aron.92 C. escribi.97 en el mensaje > Me gustar.92a saber si alguien en este grupo de noticias habla espa.96ol. De > ser > Asi seria bueno intercambiar informaci.97n. Hola, el grupo de noticias sobre las Calculadoras HP en espa.96ol es: > es.comp.sistemas.hp48 > - Dante Aron.92 C. ==== PD: seme avia ulvidao. Mira lo que puesto en el otro foro haber si tu sabes algo Andreu Cerver ==== > pero ese grupo esta casi muerto > Muerto??? > > Bueno, considerando que es verano es normal que no tengas muchos > posts.. pero yo no lo llamar.92a muerto ;-) > > Un saludo, > J.Manrique > Club de Usuarios de Gij.97n > #1077 HPCC Member Hombre muerto muerto lo que se dice muerto no est.87, por lo menos goza de mejor forma que el del club de usuarios de Gij.97n :( Andreu Cerver ==== The subject says it all. ==== Hey ! I want to know if is possible to give 2 equations and multiply them together... Ex1. : (2 + 4x) and (1 - 3x) to give (2 - 2x - 12x^2). Ex2. : Or giving (2 + 4x)*(1 - 3x) to get the right answer... ==== > Hey ! > > I want to know if is possible to give 2 equations and multiply them > together... > > Ex1. : (2 + 4x) and (1 - 3x) > > to give (2 - 2x - 12x^2). > > Ex2. : Or giving (2 + 4x)*(1 - 3x) to get the right answer... What you have are not _equations_. An equation must, by definition, contain an equal sign, =, and your expressions have no equal signs. In HP-ese, you should call '2+4*X' and '1-3*x' _algebraics_. In RPN mode, just put them on the stack and press the multiply key. ==== I am in the market for a new HP49. One vendor has a package that includes preinstalled ROM with certain engineering formulas, routines, etc. that I am indicates ROM 1.18 and a copy right of 2000. The readout also included a number like 1.7453299e-7 (or similar). Does this sound like an up-to-date machine with respect to keyboard-feel problems and screen complaints I have read about? I guess I should have asked the serial number, but it was prefixed with a CN, and the supplier has a fairly high turnover, I believe. I saw another post complaining about garbage collection with the 1.18. I am sure I will be a lightweight user in terms of pushing the boundaries of this machine. Is this GC problem only of concern to the intensive user? ==== I believe that from what I have researched, all of the 49G's with the SN starting with CN were of the type with the good screen and keyboard. Perhaps others can confirm or deny... > I am in the market for a new HP49. One vendor has a package that > includes preinstalled ROM with certain engineering formulas, routines, > etc. that I am indicates ROM 1.18 and a copy right of 2000. The > readout also included a number like 1.7453299e-7 (or similar). Does > this sound like an up-to-date machine with respect to keyboard-feel > problems and screen complaints I have read about? I guess I should > have asked the serial number, but it was prefixed with a CN, and the > supplier has a fairly high turnover, I believe. I saw another post complaining about garbage collection with the > 1.18. I am sure I will be a lightweight user in terms of pushing the > boundaries of this machine. Is this GC problem only of concern to the > intensive user? > ==== > I believe that from what I have researched, all of the 49G's with the SN > starting with CN were of the type with the good screen and keyboard. > Perhaps others can confirm or deny... > I saw some 49Gs made in Indonesia and I own one made in China. The Indonesian had two types of screens, some with a good visible rainbow-effect and some with a non disturbed display. Mine, which is the only Chinese model, has a non disturbed screen. By testing the keyboards I couldn't notice any reproducable difference. It is hard to use it if you have been used to a HP 48 before. But all 49Gs have difficulties by using the 5-key. very often you press the key once and you get two 5es. In normal cases I write numbers without loking on the display and so I had some 'surprises' in my exams. Because of the keyboard I need much more time for calculating. (The keyboard-layout is much different to the 48 and so I don't switch the calcs) But there are a lot of nice features in the 49G and so I use this calc. Perhaps there is a possibility to transfer the 49er Heart into an 48 corps. Stefan ==== > > But all 49Gs have difficulties by using the 5-key. very often you > press the key once and you get two 5es. In normal cases I write > numbers without loking on the display and so I had some 'surprises' in > my exams. > > Because of the keyboard I need much more time for calculating. > But there are a lot of nice features in the 49G and so I use this > calc. Great! I'm investing a small fortune in a machine and devoting months of study for a test and the 5 key is not going to work all the time? Why am I getting nervous and thinking about cancelling this purchase and either buying a used HP48G or sticking with a drugstore model? ==== But all 49Gs have difficulties by using the 5-key. very often you > press the key once and you get two 5es. In normal cases I write > numbers without loking on the display and so I had some 'surprises' in > my exams. > I'm investing a small fortune in a machine and devoting months of > study for a test and the 5 key is not going to work all the time? I never heard about this before. If you get double keypresses you can simply increase the keybounce time with ->KEYTIME. Do not worry, you will not have wasted your money, the 49 is customizeable to the extreme. -- This message was written with 100% recycled electrons Pivo ==== > I believe that from what I have researched, all of the 49G's with the SN > starting with CN were of the type with the good screen and keyboard. > Perhaps others can confirm or deny... I got an HP 49g from HP a few days ago, with a CN serial, and the keyboard and screen are the good ones. ==== you can update the HP49G due updateble Flash-ROM to a more recent ROM-version (at the moment version 1.19-6). You only need the connectivity-kit a pc and the connecting-software to do this. So make sure you get the connectivity-kit along with the calc, so you can update your calc to the newest ROM-version. CN would indicate that the calc was made in China and offers the better keyboard (less stiff) than the ones made in Indonesia. Garbage collection on some early ROM's wasn't optimised and took quite some time but on the newer ROM-versions this seems to work better. Roman john cuam schrieb im Newsbeitrag > I am in the market for a new HP49. One vendor has a package that > includes preinstalled ROM with certain engineering formulas, routines, > etc. that I am indicates ROM 1.18 and a copy right of 2000. The > readout also included a number like 1.7453299e-7 (or similar). Does > this sound like an up-to-date machine with respect to keyboard-feel > problems and screen complaints I have read about? I guess I should > have asked the serial number, but it was prefixed with a CN, and the > supplier has a fairly high turnover, I believe. I saw another post complaining about garbage collection with the > 1.18. I am sure I will be a lightweight user in terms of pushing the > boundaries of this machine. Is this GC problem only of concern to the > intensive user? ==== > Garbage collection on some early ROM's wasn't optimised and took quite some > time but on the newer ROM-versions this seems to work better. > Not at all. There's no difference between the Garbage collection process between the version 1.05 and version 1.19-6 The only thing that make it look a bit faster was for people who displayed the clock in the header. Displaying the clock used to be made in SysRPL and was creating a few temporary objects. Then after a few hours the memory would get full requiring a GC to occur. In the latest ROM, displaying the clock is made entirely in ML and do not create any temporary object, so GC happens less often. ==== Steve Smith >Not at all. >There's no difference between the Garbage collection process between the >version 1.05 and version 1.19-6 > >The only thing that make it look a bit faster was for people who displayed >the clock in the header. Displaying the clock used to be made in SysRPL and >was creating a few temporary objects. Then after a few hours the memory >would get full requiring a GC to occur. >In the latest ROM, displaying the clock is made entirely in ML and do not >create any temporary object, so GC happens less often. ==== Your answer was exactly what I was looking for...for some strange reason (a friend told me so) I thought the best and fastest way to work with complex formulae was to define a bunch of functions. The numeric solver is great! I feel kind of stupid now 'cause of course it's all in the manual...oh well. Now my rockets & orbit mechanics final will be a piece of cake! :-) Have a nice day, Gunnar ==== That's why I keep an eye on this group. There are many and useful things I've picked up here. Reading the manual isn't fun and you just don't make the connection with some features until much later. 2 other guys now have 49's in the office and after 3 months one said he thought he only understood 25% of what the calculator could do and was a bit shocked when I said that after 2 years I think I only understand 10% of what it can do. It seems that as your understanding grows so does the depth of the area you realise you haven't explored and may never get to explore. Stephen Your answer was exactly what I was looking for...for some strange > reason (a friend told me so) I thought the best and fastest way to > work with complex formulae was to define a bunch of functions. The > numeric solver is great! I feel kind of stupid now 'cause of course > it's all in the manual...oh well. > Now my rockets & orbit mechanics final will be a piece of cake! :-) Have a nice day, > Gunnar ==== Communication settings are determined by devices which I'm going to control. It seems to be the problem with 8 data bits and even parity. When I modify Hyper Terminal to receive 7 bit data or without parity everything is OK, but this is not solution for my problem. I was trying to find some programs for HP48 to correct this feature. I read all the files you suggested me and I think I'll have to learn (finally) Sys RPL and ML to solve this. very grateful. Milan ==== > Communication settings are determined by devices which I'm going to control. > It seems to be the problem with 8 data bits and even parity. When I modify > Hyper Terminal to receive 7 bit data or without parity everything is OK, but > this is not solution for my problem. OK, now I see; you're just using HyperTerminal to find out what's going correct, I should've written 8 bits between the start bit and the stop bits, that is, 8 data bits with no parity or 7 data bits with a parity (or mark or space) bit, but you figured that out. If there's no way to change the settings on your other devices to match the above, that is a problem. > I was trying to find some programs for HP48 to correct this feature. I read > all the files you suggested me and I think I'll have to learn (finally) Sys > RPL and ML to solve this. If there's a software solution, it would be using one of these I/O, will volunteer some help. It's way out of my league. Maybe try doing an advanced search at www.hpcalc.org for I/O or comm; who knows? Maybe someone has already solved your problem. My guess is that it's a hardware limitation. Maybe a serial buffer black box that could communicate with the 48 on one serial port and your other devices on another port? > very grateful. You're welcome. -- ==== I am buying a HP49G with a third-party ROM card (as I understand it) with civil engineering formulas and such on it. The version command returns ROM v 1.18 and copyright 2000. There is a CN prefix to the serial no., and all this preceded by a number like 1.7453299E-7, as near as I could copy it as it was read to me. Is this an up-to date machine and viable for my purposes, which will be undergraduate-level civil engineering apps. ==== > I am buying a HP49G with a third-party ROM card (as I understand it) > with civil engineering formulas and such on it. > The version command returns ROM v 1.18 and copyright 2000. There is > a CN prefix to the serial no., and all this preceded by a number like > 1.7453299E-7, as near as I could copy it as it was read to me. > > Is this an up-to date machine and viable for my purposes, which will > be undergraduate-level civil engineering apps. A third Party ROM card? How can you connect it to the 49G ? There is no slot either (or didn't I find it yet?). perhaps it's a cd-rom with apps you can load to the HP Besides that I think, that the 49G is viable for nearly all - mathematical of course - purposes. I am studying civil engeneering too and my opinion is, that the calc is more capable as I ever will need. Greetings Stefan ==== > A third Party ROM card? > How can you connect it to the 49G ? There is no slot either (or didn't > I find it yet?). > perhaps it's a cd-rom with apps you can load to the HP Right you are. The info has been downloaded to the ROM by the vendor. > ==== > Since I have had no response explaining the problem you found I assume that > you have found the error to be in your model and not in the FEM49 > application. for the reason why for that error because I'm in midterm examination. So, I just want to talk to some one who use this great FEM program and I'm glade to hear that FEM49 no error but that time I had already check with other fem program on pc. If I can't find the reason of error I will ask for help again. sorry for my english not good. lomt ==== I have a question regarding symbolic evaluation of an equation: Whenever I try to rearrange an equation using the symbolic solver, it will only do so if the variable I want to solve/rearrange for does not exist yet. So once I have solved for a variable with the numeric solver, I can't solve it symbolically anymore unless I purge the variables first. Instead, I get a Bad argument type error. I hope this makes sense to you... It'd be great if anyone could tell me how to work around this, since I often have several equations using the same variables, and I want to be able to solve one equation numerically and after that maybe another equation symbolically without losing any numerical solutions! Gunnar ==== Hey everyone, My beloved 20-year-old HP41CX finally died this past week. While I'm deciding whether or not to pursue having the 41CX repaired (which I probably will), I picked up an HP48S and an HP48G on eBay. The 48G came with manuals, but the 48S did not. So, the question is, is there a source for manuals for the 48S (electronic and/ hardcopy)? BTW, I did scan thru www.hpcalc.org and the HP48faq for this info, but didn't see it. TIA, Chuck -- Chuck Pierce NAR 78629, TRA 9308, Level 2 Huntsville Area Rocketry Assoc (NAR 403, TRA 80) cpierce_AT_knology.net ==== | Hey everyone, | My beloved 20-year-old HP41CX finally died this past week. While I'm deciding whether or not to pursue | having the 41CX repaired (which I probably will), I picked up an HP48S | and an HP48G on eBay. The 48G came with manuals, but the 48S did not. | So, the question is, is there a source for manuals for the 48S | (electronic and/ hardcopy)? BTW, I did scan thru www.hpcalc.org and | the HP48faq for this info, but didn't see it. The 48S is a 48SX without expandability. The 48S isn't that much different from the 48G except that the menus are a little different. ==== > Hey everyone, > My beloved 20-year-old HP41CX finally died this past week. rolling down cheeks> While I'm deciding whether or not to pursue > having the 41CX repaired (which I probably will), I know that HP 41 users are deeply in love with their calculators; I hope that it gets better soon. Are you familiar with www.hpmuseum.org? It has a Repairs & Batteries page and it's own forum where you might ask questions. Even if you decide not to repair it, please don't sent it to a landfill; many others would be glad to have it either to repair or as an organ donor. Also see the Classified Ads at the museum. > I picked up an HP48S > and an HP48G on eBay. The 48G came with manuals, but the 48S did not. > So, the question is, is there a source for manuals for the 48S > (electronic and/ hardcopy)? BTW, I did scan thru www.hpcalc.org and > the HP48faq for this info, but didn't see it. The 48 series will take a little getting used to, but I expect that you'll like them. I have a 16C, and always having four levels for a stack, the lack of named variables and so forth, does seem just a bit limiting. Anyway, the 48SX manuals are on one of the Museum's CDs. See www.hpmuseum.org/software/swcd.htm. I think that the 48S came with a 48SX manual, but I'm not sure. The 48S was a low-cost follow-up to the 48SX; just ignore anything about plug-in cards and ports 1 and 2. Also see Google's archive of this newsgroup at http://groups.google.com/groups?q=comp.sys.hp48, and you can search the newsgroup archive from http://groups.google.com/advanced_group_search?group=comp.sys.hp48. You can use www.google.com/help/customize.html to customize Google to your preferences. -- ==== >I know that HP 41 users are deeply in love with their calculators; I >hope that it gets better soon. Are you familiar with >www.hpmuseum.org? It has a Repairs & Batteries page and it's >own forum where you might ask questions. know about batteries/repairs page. I'll check it out. >If you decide not to >repair it, please don't sent it to a landfill; many others would be glad >to have it either to repair or as an organ donor. Also see the >Classified Ads at the museum. Landfill? That's heresy, man! My N cell killer will never suffer an undignified death. International Calculator in Orlando does a fine job repair/restoration of older adding devices, at least that's what I've been told. I plan to give them a shot at pumping new life back into my 41CX. It may end up costing a little more than buying one on eBay, but that's okay. >Anyway, the 48SX manuals are on one of the Museum's CDs. See >www.hpmuseum.org/software/swcd.htm. I think that the 48S came with a >48SX manual, but I'm not sure. The 48S was a low-cost follow-up to the >48SX; just ignore anything about plug-in cards and ports 1 and 2. it, but had assumed that it was akin to C/CV/CX for the HP41. I'll definitely get the CD from HPMUSEUM. >Also see Google's archive of this newsgroup at >http://groups.google.com/groups?q=comp.sys.hp48, and you can search the >newsgroup archive from >http://groups.google.com/advanced_group_search?group=comp.sys.hp48. You >can use www.google.com/help/customize.html to customize Google to your >preferences. Ah, yes, ye ole Google search tool. I've used it many times with the NG of my primary hobby/addiction: high-power model rocketry, rec.models.rockets. BTW, I'm always happy to find usenet NG's still thriving, like this one. This is a great communication medium, when folks are civilized Chuck -- Chuck Pierce NAR 78629, TRA 9308, Level 2 Huntsville Area Rocketry Assoc (NAR 403, TRA 80) cpierce_AT_knology.net ==== HP-48SX manuals are available on the HP Museum CD set. See www.hpmuseum.org for info. Gary. ==== Hay Everybody, I just recently bought a 49G, that has a serial of ID00503*** which came with rom 1.16. Does this mean it is a very old machine? I don't see the rainbow affect that you all are talking about, but the keys are stiff compared to my old reliable 48GX. I have only had it for about 6 months and I love its power especially after upgrading to Rom 1.19-6, but I was wondering if I am going to start having problems since it has a Serial of ID*** instead of CH***. Should I be worried? Is this a very old machine? ==== > Hay Everybody, I just recently bought a 49G, that has a serial of ID00503*** which > came with rom 1.16. Does this mean it is a very old machine? I don't It's not the oldest, but it's not the newest either. It's very wierd that it came with ROM 1.16, I don't recall this version to be used in manufacturing. If it comes with a ROM higher than 1.10 anyway, then at least you have the good screen cover which doesn't get scratch easily. ==== Is it possible to get a good screen cover anywhere and install it without having to send it back to the factory? > came with rom 1.16. Does this mean it is a very old machine? I don't > > It's not the oldest, but it's not the newest either. > It's very wierd that it came with ROM 1.16, I don't recall this version to > be used in manufacturing. > If it comes with a ROM higher than 1.10 anyway, then at least you have the > good screen cover which doesn't get scratch easily. > ==== > Is it possible to get a good screen cover anywhere and install it > without having to send it back to the factory? That would be nice. The last that I heard, sending a calculator back to HP for repair gets you a new replacement calculator, not getting your old one back repaired. But I wouldn't count on getting one new from the factory, just new off the shelf. Joe Horn reported cutting a piece from an 8 floppy disc to make a nice trim piece to cover the holes and surround the display. I think that he meant the disc cover, not the disc itself. I don't think that a 5-1/4 disc would be big enough, and it's been a long time since I've had my hands on an 8 floppy; I suppose that they're probably collectible by now. But it shouldn't be all that hard to find something that would work well. Obviously, the screen itself is more vulnerable to damage if you do that, but you won't have to look through a chunk of plastic to read it. Also, I remember reading that someone said that the screen cover adds to the structural integrity of the calculator case, but I doubt that it makes all that much difference. See http://pweb.jps.net/~joehorn/HP49.htm#nowindow for his, shall we say, invisible screen cover calculator. Someone mentioned using a small screwdriver or something like that through the serial port to pop the cover off. Of course, be careful not to damage the pins, and it might be a good idea to do an ARCHIVE and remove the batteries first if you choose this route, just in case you short-circuit the pins. I managed to loosen mine with a lollypop; the end of the stick that is, not the candy. I expect that some covers are stuck on tighter than others. Just in case you choose to put it back on, pushing it out from the back seems less likely to cause obvious disfigurement than using pliers on it. Search on www.hpcalc.org for other information on removing the screen cover. I've found that the SNAP TREATMENT FOR GLASS AND PLASTIC that I bought from EduCalc years ago works well for cleaning the screen cover, but I wouldn't expect it do do much for deep scratches. I've noticed that the rainbow effect is more obvious when the calculator is subjected to a sudden change in temperature. I suppose that the temperature change stresses the plastic more until its temperature stabilizes. -- ==== > Hay Everybody, I just recently bought a 49G, that has a serial of ID00503*** which > came with rom 1.16. Does this mean it is a very old machine? I don't > see the rainbow affect that you all are talking about, but the keys > are stiff compared to my old reliable 48GX. I have only had it for > about 6 months and I love its power especially after upgrading to Rom > 1.19-6, but I was wondering if I am going to start having problems > since it has a Serial of ID*** instead of CH***. Should I be worried? > Is this a very old machine? The serial number indicates your HP49 was manufactured in Indonesia in May 2000. Rejoice, I bought my HP49 in November last year and got one made in 1999. If you can live with the stiff keys and the plastic display cover I wouldn't... er, I mean YOU shouldn't worry about anything. The only other problem (besides scratches in the plastic display cover) I have experienced is that the painted lettering in one of the most used keys (F6) started to wear out, something that never happened in any of my other HP calculators. Santos Lucero ==== I guess that you are using the Fast 3D Feature... in this case, try with the following parameters: X: - 1 1 Y: -1 1 Z: -1 1 setp indep: 9, depnd 9 draw it and rotate it, you should see your 'standard' saddle.... For z^2=x^2+y^2, you can draw 1 1/2 by drawing: SQRT(x^2+y^2) with the same parameters If you had the new ROM version (which you don't because HP never released it :-) you could even enter: { 'SQRT(x^2+y^2)' '-SQRT(x^2+y^2)' } and get your hour glass > I just got an HP49G and I can't seem to get good 3D graphs. For > example, I tried graphing x^2-y^2 and the graph it gave me was not > really good and didn't really look like it was supposed to. Are there > any parameters I can change to make 3D graphing better? Also, is it > possible to graph a function like z^2=x^2+y^2 so that I get two ==== Hey, Cyrille! How about the hidden-line-removal for a Slow-3D What about animation support (Y-slice) in Fast-3D I suppose that your wireframe to Fast-3D support and other enhancements will be ready in 1.19-7 X > If you had the new ROM version (which you don't because HP never released it > :-) you could even enter: > { 'SQRT(x^2+y^2)' '-SQRT(x^2+y^2)' } and get your hour glass X ==== > If you had the new ROM version (which you don't because HP never released it > :-) you could even enter: > { 'SQRT(x^2+y^2)' '-SQRT(x^2+y^2)' } and get your hour glass That's so mean .... :) ==== > > If you had the new ROM version (which you don't because HP never released > it > :-) you could even enter: > { 'SQRT(x^2+y^2)' '-SQRT(x^2+y^2)' } and get your hour glass > > That's so mean .... :) > How do I find out which ROM version I have. When I executed VERSION() it said something like build #1.18 - is that it? Also, how do I get ==== > If you had the new ROM version (which you don't because HP never released > it :-) you could even enter: > { 'SQRT(x^2+y^2)' '-SQRT(x^2+y^2)' } and get your hour glass That's so mean .... :) > I know, but I feel better now :-) BTW, here is a picture of the hourglass :-) > How do I find out which ROM version I have. When I executed VERSION() > it said something like build #1.18 - is that it? Also, how do I get HP never released the version I am talking about, but HP did release newer versions of the ROM you can download the ROM version 1.19-6 from http://www.hpcalc.org/hp49/pc/rom/ it does not contain this new feature, but it does contain a good set of bug fixes and a couple of extra features... ==== Yes, use VERSION. To upgrade to 1.19-6 use: http://www.epita.fr/~avenar_j/hp/49.html X > How do I find out which ROM version I have. When I executed VERSION() > it said something like build #1.18 - is that it? Also, how do I get ==== Here is the SysRpl program that I am trying to compile. Does anyone have any suggestions? Jeff !NOCODE !RPL : : 0LASTOWDOB! CK2NOLASTWD CK&DISPATCH0 2REAL : : %MAXorder SWAP2DUP DUP %* % 1.25 %* SWAPDUP %* % .25 %* %- %SQRT ID RD1b % .15625 %+ Insod TAGOBS ID PT1b ; ; @ > SysRpl program and it is not working. When I compile a HP-49G program the > external file is left on the stack. With the HP-48GX, only a portion of the > entire file is left on the stack and it is not an external. What am I doing > wrong? > Jeff > --- > Checked by AVG anti-virus system (http://www.grisoft.com). --- Checked by AVG anti-virus system (http://www.grisoft.com). ==== I assume that you have the entry point library installed on both you 48 and your 49 (be carefull, they are different) and that the 48 and the 49 are in ASM mode, not in RPL mode for MASD? If this is the case (as it is on my calculators), you should have no problem compiling on both calculators... > Here is the SysRpl program that I am trying to compile. Does anyone have any > suggestions? > Jeff > !NOCODE !RPL : : 0LASTOWDOB! CK2NOLASTWD CK&DISPATCH0 2REAL : : %MAXorder SWAP2DUP DUP %* % 1.25 %* SWAPDUP %* % .25 %* %- > > %SQRT ID RD1b % .15625 %+ Insod TAGOBS ID PT1b ; ; @ > a > SysRpl program and it is not working. When I compile a HP-49G program the > external file is left on the stack. With the HP-48GX, only a portion of > the > entire file is left on the stack and it is not an external. What am I > doing > wrong? > Jeff > --- > Checked by AVG anti-virus system (http://www.grisoft.com). > --- > Checked by AVG anti-virus system (http://www.grisoft.com). ==== Can you try to be more precise? I do not really understand the problem... Example would be good. > SysRpl program and it is not working. When I compile a HP-49G program the > external file is left on the stack. With the HP-48GX, only a portion of the > entire file is left on the stack and it is not an external. What am I doing > wrong? > Jeff > --- > Checked by AVG anti-virus system (http://www.grisoft.com). ==== Does anybody know where to buy one of those acrylic stands that elevate the face of the 48GX? I know that in the dark ages Educalc did, but I don't know who does now. TIA. ==== > Does anybody know where to buy one of those acrylic stands that elevate the > face of the 48GX? I know that in the dark ages Educalc did, but I don't > know who does now. TIA. I don't know whether it's the one that you're looking for, but www.calcpro.com does have an Angle View Calculator Stand. -- ==== There was one on Ebay the other day. I wasn't sure what it was, but I've got some scans of old Educalc catalogs and it shows it in them. Bob > Does anybody know where to buy one of those acrylic stands that elevate the > face of the 48GX? I know that in the dark ages Educalc did, but I don't > know who does now. TIA. ==== While fooling around with repeating control alarms on a 49G, I was trying to edit an alarm entry and had my editing session aborted by a control alarm. Oh yeah, I've got to be careful that I have time to do something about a repeating control alarm before the next repeat comes due. It's a problem if you don't even have time to do n DELALARM (where n is the alarm index) before the next repeat aborts your command line. I remembered that there was a way to break out of this on a 48SX, so I Even a system halt (section 5.8) won't help, because that (by design) does not affect the rescheduling of alarms. You can escape from this predicament by pressing [ON]-[4] (together). This disables rescheduling of the next alarm, which allows yot to activate the alarm catalog and change the repeat interval for the alarm, or to delete it altogether. Turning the calculator off, then on, re-enables rescheduling in case you used [ON]-[4] unnecessarily. Good, just what I was looking for, but some changes to the alarm system were made in the 48G series. No matter, the [ON]-[4] safety net still works on a 48GX. But [ON]-[4] doesn't help on a 49G. No surprise that a warmstart doesn't help either, and neither does holding down backspace while warmstarting. Is there a way to break out of a repeating alarm loop short on the 49G short of a memory clear? If not, then perhaps holding down backspace while warmstarting would be a good place to add this safety feature. -- ==== X > But [ON]-[4] doesn't help on a 49G. No surprise that a warmstart doesn't > help either, and neither does holding down backspace while warmstarting. Is there a way to break out of a repeating alarm loop short on the 49G > short of a memory clear? If not, then perhaps holding down backspace > while warmstarting would be a good place to add this safety feature. I suggest a similar logical place: [ON]&[ 9 ] [ 9 ] has TIME above it... ==== > X > But [ON]-[4] doesn't help on a 49G. No surprise that a warmstart doesn't > help either, and neither does holding down backspace while warmstarting. Is there a way to break out of a repeating alarm loop short on the 49G > short of a memory clear? If not, then perhaps holding down backspace > while warmstarting would be a good place to add this safety feature. I suggest a similar logical place: [ON]&[ 9 ] > [ 9 ] has TIME above it... That makes sense to me. I expect that it's [ON]&[4] on the 48s because TIME is above the [4] on them. That's a good reminder of what the special key is. Just in case it was already implemented that way on the 49G, I checked, and it's not. -- ==== [snip] Interesting that you fond all that old info... > Is there a way to break out of a repeating alarm loop short on the 49G > short of a memory clear? If not, then perhaps holding down backspace > while warmstarting would be a good place to add this safety feature. What about using the reset hole on the back side? Greg S ==== > [snip] Interesting that you fond all that old info... Actually, except for things that are obviously 49G only, like new commands, flags, and reserved variables, the Insights books are usually the first place that I look when I have a question about these calculators. What is interesting is that I haven't memorized them. > Is there a way to break out of a repeating alarm loop short on the 49G > short of a memory clear? If not, then perhaps holding down backspace > while warmstarting would be a good place to add this safety feature. What about using the reset hole on the back side? Good suggestion. Tried it, but the next alarm occurred as expected. No surprise there; I think that a hardware reset just forces a warmstart the next time you turn on the calculator. That's certainly useful if things get so fouled up that you can't even force a warmstart by using the keyboard, but other than that, I don't think that it accomplishes anything that warmstarting by using the keyboard doesn't. Of course a complete memory clear gets rid of any alarms, but most of us would usually rather not get rid of everything that a memory clear takes with it. -- ==== This is weird, cause I just got it in the last week under warranty... So if you do that, how do you know that you aren't going to get another > one > with the rainbow screen and scratchy and hard to press keys? You don't. > But this has been fixed a long time ago now. > You may not get the new improved keyboard, but chances are you will get a > better screen cover ==== Some people have mentioned that if you are buying a HP49 then it is better to get one manufactured in China rather than one from Indonesia. I am planning to buy a HP48GX. Can somebody tell me if there are any quality differences for the HP48 calculators manufactured in different countries, or do they all come from the same place? John ==== > I am thinking of buying an HP49G with a pre-installed ROM featuring > civil engineering formulas, etc. I will be doing undergrad-level > work. > Customer service at the vendor checked a sample unit at her desk and > the version command returned a ROM version no. 1.18 and copyright > 2000. I have been reading complaints about this version. Is this unit > defective or are talking about minor beefs. I have used HPs before, Just upgrade it to the 1.19-6 Flash-ROM to benefit from all the new improvements made so far. http://www.epita.fr/~avenar_j/hp/49.html ==== > I am thinking of buying an HP49G with a pre-installed ROM featuring I have been an HP user for the last 6 years and the list is comprised of an HP19BII, HP20S, HP28S, HP32SII, HP38G, HP48G, HP49G plus a couple of palmtops. The HP49G is, without doubt in MY mind, the most mature do-it-all calculator that HP has produced as far as software and customizability! Never mind about the stiff keyboard, the scratchy screen-cover and the rainbow effect of the first units. If you use algebra, calculus AND you want a 'clever' (RPL) way of binding it all together then this is the machine you want to have :-) However, before you buy it, try and use it first! It is a VERY demanding device in the sense that it requires a great deal of dedication and persistence. One of the problems is the multitude of ways it offers you to reach a result! It is hard to choose from so many options! I think I am just a bit biased :-) !Demeter! ==== X > Strange. When I use my HP48GX to solve this, I enter the equation, then > 's', and then QUAD. I get the symbolic solution: > 400+s1*400 > s=---------- > -1240 X > What have I missed in the HP49 to not be able to get the symbolic form of > the solutions? Hmmm..You get a list of solutions instead one using 's1'