HP-32 HP-32 Hope this hasn't already been mentioned and I just missed it, but I just downloaded the latest 49g+ User's Guide from www.hp.com and it's now bookmarked and about 2 MB bigger than before. I haven't looked yet to see what's changed though... ==== Here's a direct URL. http://h20015.www2.hp.com/en/taskPageSelect.jhtml?reg=&plc=&lc=en&cc=us&prod Id=hp49ggraph351775&pagetype=manual&docparent=manual > Hope this hasn't already been mentioned and I just missed it, > but I just downloaded the latest 49g+ User's Guide from > www.hp.com and it's now bookmarked and about 2 MB bigger > than before. I haven't looked yet to see what's changed though... > > Mike ==== I never had a 49G, so can't speak to a precise comparison (but I do have 48G's, and so have some idea). I say buy. I think the 49G+ stands on its own as a genuinely beautiful piece of work. They took the best of the 48G and 49G, wrapped it in a sleek, tight package, and added speed, inexpensive SD storage, & USB connectivity. The keyboard silk-screening is notably clear and sized to minimize clutter. The keypress issues appear to have been resolved. The display cover is flat, and has a slight anti-glare finish (not extreme examples on Joe Horn's website), I'd say the G+ has completely solved any display-related problems. (In fact, if I had a 49G, I think I'd follow Joe's advice and rip the plastic cover right off!) No, it's not going to satisfy everyone. No, it's not an HP-41, or a 15C, or a 42S. Yes, it is a really fine unit. FWIW, Paul B. ==== > Wait till hp updates the 49g software. NO Buy it now and learn to use it immediately The FlashROM is there for updates ==== > > Wait till hp updates the 49g software. > NO > Buy it now and learn to use it immediately > The FlashROM is there for updates My apologies for being unclear. I hope these other posts of mine explain what I meant. http://groups.google.com/groups?q=g:thl2179575005d&dq=&hl=en&lr=&ie=UTF-8&se lm=551a73ee.0311041953.75694487%40posting.google.com http://www-s.ti.com/cgi-bin/discuss/sdbmessageview.cgi?databasetoopen=calcul ators&topicarea=Graphing+Calculator+Comparisons&viewmethod=Thread&messageid=1 43206 ==== How do the function sets compare between the 49g and the 49g+? I understand that IR is restored so that is good. But are there any other changes, fixes, enhancements and/or new functions? Sincerely, Kevin Waite ==== > How do the function sets compare between the 49g and the 49g+? As far as I know, except for changes to deal with the SD card, USB, and IrDA, the same. > I understand that IR is restored so that is good. Good, but could be a whole lot better. It doesn't work for communicating with the old 48 series, just by IrDA and (sort of) printing to the 82240A/B printers. To the printers, the range is reduced to 2 or 3 inches instead of the 3 to 6 feet that I get from the 48SX and 48GX. > But are there any other changes, fixes, enhancements and/or new functions? I expect that some bug fixes that didn't make it into the 49G's 1.19-6 revision should be in the 49g+. And I view the SD card as a huge enhancement. But the new calculator introduces some new problems of its own. -- James ==== > How do the function sets compare between the 49g and the 49g+? > > I understand that IR is restored so that is good. > > But are there any other changes, fixes, enhancements and/or new functions? There's a LOCAL function which allows local variables to be defined that don't need to be pulled from the stack. Tom Lake ==== > > How do the function sets compare between the 49g and the 49g+? > > > > I understand that IR is restored so that is good. > > > > But are there any other changes, fixes, enhancements and/or new functions? > > There's a LOCAL function which allows local variables to be defined that > don't need to be pulled from the stack. That is also in my 1.19-6 ROM for the 49G so it is not new Jerry (-; ==== > > > How do the function sets compare between the 49g and the 49g+? > > > > > > I understand that IR is restored so that is good. > > > > > > But are there any other changes, fixes, enhancements and/or new > functions? try to plot: { 'X*Y^3-Y*X^3' 'SINX)*SIN(Y)' } in fast 3D mode... ==== > > > > > How do the function sets compare between the 49g and the 49g+? > > > > > > > > I understand that IR is restored so that is good. > > > > > > > > But are there any other changes, fixes, enhancements and/or new > > functions? > > try to plot: > { 'X*Y^3-Y*X^3' 'SINX)*SIN(Y)' } in fast 3D mode... > Aaah! A new enhancement on plotting or should I say a new operation A=Analytic function (SIN, LN, XROOT,...) F=Function (FP, MOD, FACT, <,... ) C=Command (ARCHIVE, CONT, CONIC,...) O=Operation (something non-programmable found on a menu label or a shortcut) ==== > > There's a LOCAL function which allows local variables to be defined that > > don't need to be pulled from the stack. > > That is also in my 1.19-6 ROM for the 49G > so it is not new Ah. I was looking at the 49G AUG which doesn't have that in it. Do you know how to use it? I tried putting it in a program but it doesn't seem to have any effect. The variables still show up in the Var menu. Does anyone have a short example of the use of this command? TIA Tom Lake ==== > Ah. I was looking at the 49G AUG which doesn't have that in it. Do > you know how to use it? I tried putting it in a program but it > doesn't seem to have any effect. The variables still show up in the > Var menu. Does anyone have a short example of the use of this command? > TIA LOCAL was created in order to create multiple local variables in ALG mode at once. The standard algebraic compile only allows to create one variable at a time doing something like: << -> A << -> B << -> C >> >> >> Which can quickly make program cumbersome. I don't think LOCAL will have much use in RPL mode. ==== > > Ah. I was looking at the 49G AUG which doesn't have that in it. Do > > you know how to use it? I tried putting it in a program but it > > doesn't seem to have any effect. The variables still show up in the > > Var menu. Does anyone have a short example of the use of this command? > > TIA > > LOCAL was created in order to create multiple local variables in ALG > mode at once. > > The standard algebraic compile only allows to create one variable at a > time doing something like: > << -> A << -> B << -> C >> >> >> > > Which can quickly make program cumbersome. > I don't think LOCAL will have much use in RPL mode. > Since in the 49 (and I presume the 49+) one can define multiple local variables in RPL mode with either of the formats -> A B C << @program >> or -> A B C '@algebraic', with the local variable creation preceding either a program or algebraic and valid to the end of the program or algebraic, I agree that LOCAL will not be much needed in RPL. ==== > LOCAL was created in order to create multiple local variables in ALG > mode at once. > > The standard algebraic compile only allows to create one variable at a > time doing something like: > << -> A << -> B << -> C >> >> >> > > Which can quickly make program cumbersome. > I don't think LOCAL will have much use in RPL mode. So what's the preferred way of defining a number of local variables in RPL mode without having to enter anything on the stack before the program is run? -- Tom Lake Experience keeps a dear school but fools will learn in no other - Poor Richard's Almanack ==== ...> So what's the preferred way of defining a number of local variables > in RPL mode without having to enter anything on the stack before the > program is run? > I was trying things like << 'A=0' LOCAL 5 'A' STO >> and got some weird results. When first run, variable A is created containing 5 which you can recover by pressing the key in the VAR menu. However if you press Alpha A Enter or 'A' EVAL, you get 0. Typing {A} PURGE will not purge the variable. Any entry of A in the command line seems to be resolved as a local variable A, not the one in the VAR menu. This local variable is global in the sense that it is available from any directory, exists outside a program environment, and seems to persist until a reset. Before the LOCAL command, local variables could exist during a HALTed program but disappeared when the program continued to completion. LOCAL could be useful for hidden global variables, if created by a startup program. The danger is: if another program containing a regular variable of the same name is recompiled, that program variable will now refer to the local variable. For that reason you should probably use the left-arrow local variable naming convention ('<-A'), since such variables are always compiled as local. Use the Filer to purge the A from the VAR directory and edit, save, and run the above program again. This time the 5 is saved in the local variable and no new variables appear in the VAR directory. Use ON+C to get rid of the local variable. Happy programming, Tom ==== > ... [LOCAL vars] seem to persist until a reset. Or until purged with an UNBIND command. Cf. the built-in help screen for the LOCAL command. -Joe- ==== > > ... [LOCAL vars] seem to persist until a reset. > > Or until purged with an UNBIND command. Cf. the built-in help screen > for the LOCAL command. On my HP49 (not HP49G+) I can't get UNBIND to work nohow. LOCAL is just fine, but no amount of fiddling produces anything from UNBIND but Can't unbind local var... I've tried it with and without the local name on the stack, with and without the last-stack enabled, and I dunno. Am I doing something wrong? Bill ==== >>Or until purged with an UNBIND command. Cf. the built-in help screen >>for the LOCAL command. > > On my HP49 (not HP49G+) I can't get UNBIND to work nohow. LOCAL is > just fine, but no amount of fiddling produces anything from UNBIND but > Can't unbind local var... > > I've tried it with and without the local name on the stack, with and > without the last-stack enabled, and I dunno. Am I doing something > wrong? I just tried this and also have trouble with UNBIND on my HP49G 1.19-5. On the 49g+, with the local name on the stack, I got UNBIND to work from the catalog or in a program but not from the command line. {UNBIND} OBJ-> DROP EVAL also works. Editing the command line may do something like start a new program environment from which local variables from the earlier LOCAL can't be accessed. LOCAL seems to work from the command line on both machines. Tom ==== > I just tried this and also have trouble with UNBIND on my HP49G 1.19-5. > On the 49g+, with the local name on the stack, I got UNBIND to work > from the catalog or in a program but not from the command line. haven't figured out what affects it and sometimes keeps it from working. You're right that it doesn't like the command line. A warmstart will do for clearing local vars, although it tends to clear some information I'd rather keep. But it's not too important an Bill ==== > > ... [LOCAL vars] seem to persist until a reset. > Or until purged with an UNBIND command. Cf. the built-in help screen > for the LOCAL command. This may cause misunderstanding. What means reset? Local variables created with the CAS command LOCAL do not survive the next warmstart. Hence, I do not see any striking application of this kind of local variables outside the context of certain CAS applications. Anyway, they are more stable as ordinary local variables. These are forgotten after a return to the default mode (no application active, no suspended programs, no edit line), no matter whether the ABND command was used or not. - Wolfgang ==== > >>>... [LOCAL vars] seem to persist until a reset. > > >>Or until purged with an UNBIND command. Cf. the built-in help screen >>for the LOCAL command. > > > This may cause misunderstanding. What means reset? Local variables > created with the CAS command LOCAL do not survive the next warmstart. > Hence, I do not see any striking application of this kind of local > variables outside the context of certain CAS applications. > > Anyway, they are more stable as ordinary local variables. These are > forgotten after a return to the default mode (no application active, no > suspended programs, no edit line), no matter whether the ABND command > was used or not. > > - Wolfgang I should have said warmstart instead of reset in the original post. My hope for LOCAL was that a STARTUP program in HOME could initialize some local variables which could act as hidden variables available from any directory. Unfortunately using LOCAL in STARTUP resulted in a lockup after a warmstart, and sometimes TTRM unless the backspace was held during the warmstart. Oh, well. I learned some things from this. (Be careful what you put in STARTUP, for example. :-) ) Tom ==== > X > I should have said warmstart instead of reset in the original post. > > My hope for LOCAL was that a STARTUP program in HOME could initialize > some local variables which could act as hidden variables available from > any directory. Unfortunately using LOCAL in STARTUP resulted in a > lockup after a warmstart, and sometimes TTRM unless the backspace was > held during the warmstart. Oh, well. I learned some things from this. > (Be careful what you put in STARTUP, for example. :-) ) Well why don't you just do this: << ...other stuff...MEM DROP DATE TIME .0001 + 'CREATE' 3 ->LIST STOALARM DROP >> 'STARTUP' STOand then STOre to 'CREATE' this: << { 'HIDDEN=777' '<-Secret=(1,1)' } LOCAL DROP2 >> and you're donw 1 sec after boot finishes ==== > > Well why don't you just do this: > << ...other stuff...MEM DROP > DATE TIME .0001 + 'CREATE' 3 ->LIST STOALARM DROP >> > 'STARTUP' STOand then STOre to 'CREATE' this: > << { 'HIDDEN=777' '<-Secret=(1,1)' } LOCAL DROP2 >> > and you're donw 1 sec after boot finishes > Great solution! But I'm afraid to try it now. How about 'dodgy'=666 LOCAL :-) ==== After the first STARTUP wiped out the memory I tested this delayed solution with no side-effects on 49g+ > > > > Well why don't you just do this: > > << ...other stuff...MEM DROP > > DATE TIME .0001 + 'CREATE' 3 ->LIST STOALARM DROP >> > > 'STARTUP' STOand then STOre to 'CREATE' this: > > << { 'HIDDEN=777' '<-Secret=(1,1)' } LOCAL DROP2 >> > > and you're donw 1 sec after boot finishes > > > > Great solution! But I'm afraid to try it now. > > How about 'dodgy'=666 LOCAL :-) > <87233f9e.0311151001.4e285e8c@posting.google.com> <3FB68C92.90E46669@math.fu-berlin.de> ==== > My hope for LOCAL was that a STARTUP program in HOME could initialize > some local variables which could act as hidden variables available > from any directory. I find the programming of LOCAL to be VERY dodgy and may have some weird effect on the behavior of the calculator. I can also be messed up by turning OFF and ON the LASTSTACK flag... <87233f9e.0311151001.4e285e8c@posting.google.com> <3FB68C92.90E46669@math.fu-berlin.de> ==== > > My hope for LOCAL was that a STARTUP program in HOME could initialize > > some local variables which could act as hidden variables available > > from any directory. > I find the programming of LOCAL to be VERY dodgy and may have some weird > effect on the behavior of the calculator. turn to explain why it doesn't work. There should be an official listing of what is *not* allowed to write in STARTUP, and why not :-) Wolfgang ==== >>... [LOCAL vars] seem to persist until a reset. > > > Or until purged with an UNBIND command. Cf. the built-in help screen > for the LOCAL command. > variables. For some reason I never thought of local variables outside the context of a (possibly HALTed) program. This raises new possibilities. The HELP examples use left-arrow variable names but they aren't required. As I mentioned earlier, however, regular names can lead to weird appearing behovior. Tom ==== > So what's the preferred way of defining a number of local variables in RPL > mode without having to enter anything on the stack before the program is > run? To create in an RPL program a,b and c as local variables with values of 1, 2 and 3 to be used in a following subprogram or algebraic use 1 2 3 -> a b c followed by the program or algebraic. Note that if there are more local variable names after the -> than values in front of it, there must be enough values on the stack to fill those extra local variable slots. ==== > > So what's the preferred way of defining a number of local variables in RPL > > mode without having to enter anything on the stack before the program is > > run? > > To create in an RPL program a,b and c as local variables with values > of 1, 2 and 3 to be used in a following subprogram or algebraic use > 1 2 3 -> a b c followed by the program or algebraic. understand now. -- Tom Lake Experience keeps a dear school but fools will learn in no other - Poor Richard's Almanack ==== >>LOCAL was created in order to create multiple local variables in ALG >>mode at once. >> >>The standard algebraic compile only allows to create one variable at a >>time doing something like: >><< -> A << -> B << -> C >> >> >> >> >>Which can quickly make program cumbersome. >>I don't think LOCAL will have much use in RPL mode. > > > So what's the preferred way of defining a number of local variables in RPL > mode without having to enter anything on the stack before the program is > run? You use the right-arrow (Right-Shift 0, not Cursor-Right) to define local variables to be used in the defining procedure, which can be either a program or an algebraic object. For example: -> x y z << x 2 ^ y 2 ^ + z 2 ^ + v/ >> or: -> x y z 'v/(x^2+y^2+z^2)' either on the command line or within a program will both take the three coordinates from the stack and return the distance from the origin. Usually you'll be using this structure within a program as in: << -> x y z << x 2 ^ y 2 ^ + z 2 ^ + v/ >> >> You can also put the object to be stored as a local variable within the program itself (before the right arrow), handy if you have a sub-program or constant value that you want to use repeatedly within the defining procedure. Or you can put a dummy object before the right arrow to define a local variable, and then within the defining procedure, use STO to replace it with a programmatically determined object. This is useful for when you want to use a local variable but don't know the value yet, and prefer to avoid nested local variable structures. -- James ==== > > So what's the preferred way of defining a number of local variables in RPL > > mode without having to enter anything on the stack before the program is > > run? > > You use the right-arrow (Right-Shift 0, not Cursor-Right) to define > local variables to be used in the defining procedure, which can be > either a program or an algebraic object. For example: > > -> x y z << x 2 ^ y 2 ^ + z 2 ^ + v/ >> > > or: > > -> x y z 'v/(x^2+y^2+z^2)' > > either on the command line or within a program will both take the three > coordinates from the stack and return the distance from the origin. Right, but I don't want to get values off the stack. I'm going to INPUT them during program execution. > > Usually you'll be using this structure within a program as in: > > << -> x y z << x 2 ^ y 2 ^ + z 2 ^ + v/ >> >> > > You can also put the object to be stored as a local variable within the > program itself (before the right arrow), handy if you have a sub-program > or constant value that you want to use repeatedly within the defining > procedure. > Could you give a short example of this and the paragraph below or point me to a Website that has examples? > Or you can put a dummy object before the right arrow to define a local > variable, and then within the defining procedure, use STO to replace it > with a programmatically determined object. This is useful for when you > want to use a local variable but don't know the value yet, and prefer to > avoid nested local variable structures. ==== >> > So what's the preferred way of defining a number of local variables in > > RPL > >> > mode without having to enter anything on the stack before the program > > is > >> > run? >> >>You use the right-arrow (Right-Shift 0, not Cursor-Right) to define >>local variables to be used in the defining procedure, which can be >>either a program or an algebraic object. For example: >> >>-> x y z << x 2 ^ y 2 ^ + z 2 ^ + v/ >> >> >>or: >> >>-> x y z 'v/(x^2+y^2+z^2)' >> >>either on the command line or within a program will both take the three >>coordinates from the stack and return the distance from the origin. > > > Right, but I don't want to get values off the stack. I'm going to INPUT > them during program execution. Ok, it doesn't matter how the values got on the stack, whether they were there when the program started, or the program put them there, or an INPUT command in the program put them there. For that matter, if a value to be used repeatedly is in a global variable that will take some time to retrieve, it may well be faster to retrieve it just once and store it in a local variable for use within the defining procedure; sort of a cache. What matters is that when the create local variables command executes, there is one object on the stack to be used with each name that occurs between the command and the defining procedure. For each local name, one object is removed from the stack. The object on level 1 goes with the last local name, the object on level 2 with the next to last name, and so on. For example: A B C D -> w x y z << w x y z >> Returns: A to level 4, B to level 3, C to level 2, and D to level 1. If there are more names after the arrow than objects on the stack, then it will error. If there are more objects on the stack than names after the arrow, then the unused objects will be left on the stack. If you're inputting them at the beginning of the program, just have INPUT commands put the values on the stack before the right-arrow, and the create local variables command will take them off the stack for use with the corresponding names as local variables. >>Usually you'll be using this structure within a program as in: >> >><< -> x y z << x 2 ^ y 2 ^ + z 2 ^ + v/ >> >> >> >>You can also put the object to be stored as a local variable within the >>program itself (before the right arrow), handy if you have a sub-program >>or constant value that you want to use repeatedly within the defining >>procedure. >> > > Could you give a short example of this and the paragraph below or point me > to a Website that has examples? Here's an example of a sub-program in a local variable: %%HP: T(3)A(D)F(.); @Checksum (48SX): # FC78h @Size (48SX): 268.5 @Checksum (49G): # 3F19h @Size (49G): 260.5 @Tested on 48SX and 49G (With the 49G in exact mode for the download). << @ Program to replace control codes @ corresponding translation codes. << @ Sub-program to accomplish SREPL on @ a 48, except that this @ modification doesn't return the @ count. OVER SIZE @ Get the size of the replacement @ string. @ Start the first part of the result @ as an empty string. 5. ROLL @ Move the original level 3 argument @ to level 1 as the remainder @ string. WHILE DUP 6. PICK @ Move the search string to level 1. POS DUP @ Is the search string in the @ remainder string? REPEAT ROT @ Move the first part down. 3. PICK @ Move the remainder down. 1. @ Beginning of the string. 4. PICK @ Location of the search string. 1. - @ 1 character before this. SUB @ Get the substring. + @ Add it to the first part. 5. PICK @ Replacement string. + @ Add it to the first part. ROT @ Old remainder. ROT @ Location of the search string. 4. PICK @ Size of replacement string. + @ Add them to specify the beginning. @ of the new remainder string. OVER SIZE @ Location of the end of the string. SUB @ New remainder string. END DROP @ Drop the 0 from the WHILE clause. + @ Add the remainder to the first @ part. 4. ROLLD @ Move the result string up. 3. DROPN @ Drop the search string, @ replacement string, and its size. >> -> s @ Store the sub-program with local @ name s. << @ Start of defining procedure. ->STR @ Verify that an argument exists and @ force it to be a string. 0 9 FOR n n CHR C$ 3 00 n + s EVAL @ Translate control codes 0-9. NEXT 11 31 FOR n n CHR C$ 2 0 n + s EVAL @ Translate control codes 11-31. NEXT 127 C$ 4 127 s EVAL @ Translate control code 127. >> >> For a 49G or 49g+, the subprogram wouldn't be needed; just SREPL DROP instead of each of the three s EVAL sequences. Of course I could just embed the sub-program 3 times instead of using a local variable, but just imagine what that would do to the size. Or if I knew for sure that the sub-program were in a global variable on the path, I could just call it by name 3 times. But if I were deep within the directory structure and the global variable were at the end of the home directory, it may be faster to have the program RCL the contents to the stack, store it as a local variable, and EVAL the local variable 3 times. And I expect that re-writing it in SysRPL would speed it up a lot. >>Or you can put a dummy object before the right arrow to define a local >>variable, and then within the defining procedure, use STO to replace it >>with a programmatically determined object. This is useful for when you >>want to use a local variable but don't know the value yet, and prefer to >>avoid nested local variable structures. Suppose that whether or which INPUT command is executed depends on what's happened previously within the defining procedure, by way of conditional structure. Well, you can nest another local variable structure within the defining procedure, but my preference is usually to create all of the local variables at one time, using a dummy value such as a zero for any to be replaced later. For a very contrived example, using the local variables even when using the stack manipulation commands may be better, and multi-byte local names at that: %%HP: T(3)A(D)F(.); @ Checksum: # F2D6h @ Bytes: 502. @ Tested on 49G. << 0. 0. -> response1 response2 << Do you have a question? (Y/N) INPUT 'response1' STO CASE response1 Y == response1 y == OR THEN What's the question? INPUT 'response2' STO END response1 N == response1 n == OR THEN Good, because I don't have an answer either. END What do you mean: response1 + ? Come back when you learn how to press Y or N! + END IF response2 0. =/ THEN Sorry, I don't have an answer for: response2 + END >> >> Was that a helpful example? How about this recursive one that doesn't quit asking for (Y/N) until you enter a proper response or CANCEL? Note the the subprogram uses the names for itself and the other local variables, so I use a dummy value for it's local variable, and don't STO the subprogram until after the local variables have been created so that the names are compiled as local rather than global. Well, I could've gotten around that by starting each name with a left-arrow, but this seemed more natural to me. I'll grant that it would seem more natural to me to use a loop until a Y or N were entered, but this is just an example of what can be done. %%HP: T(3)A(D)F(.); @ Checksum: # C0D2h @ Bytes: 585.5 @ Tested on 49G. << 0. 0. 0. -> response1 response2 subprogram << << Do you have a question? (Y/N) INPUT 'response1' STO CASE response1 Y == response1 y == OR THEN What's the question? INPUT 'response2' STO END response1 N == response1 n == OR THEN Good, because I don't have an answer either. END What do you mean: response1 + ? learn how to press Y or N! Try again! + 1. DISP 2. WAIT subprogram EVAL END >> 'subprogram' STO subprogram EVAL IF response2 0. =/ THEN Sorry, I don't have an answer for: response2 + END >> >> I expect that looking over various UserRPL programs at http://www.hpcalc.org/ or http://membres.lycos.fr/ekalin/index.php will show other examples of local variable uses. -- James ==== > ... > And I expect that re-writing it in SysRPL would speed it up a lot. > ... Why is the use of local variables in UsrRPL so slow? Its not their use itself but their *creation*, essentially done by ->. This command which separates the names of local variables from the delimiter << or ' (for user-defined functions) does a lot of work before making the actual binding. -> (or xRPN-> in SysRPL display) decompiles as follows: :: CKOATTNABORT (Abort in case CANCEL has been pressed in panic) PTR 3889D (Collect the chosen variables and makes them named LAMs) DEPTH OVER #2* (after which these form a metaobject on the stack) #>case (Check whether the input matches the number of LAMs) DOBIND (If it matches do finally the binding) NDROP (but if not remove the just created metaobject of LAMs SETSTACKERR (and trap the user error as a STACK error) ; The most involved of all this is PTR 3889D which is written in asm to make it as fast as possible. This is a runstream operation, i.e., it takes its arguments from the runstream, not from the data stack. Note that it must also compute how many arguments have to be taken. Thus, avoid using local variables whenever speed is important for your program. Just use your brain for using the enormous power of pure stack computation. - Wolfgang ==== > >>... >>And I expect that re-writing it in SysRPL would speed it up a lot. >>... > > > Why is the use of local variables in UsrRPL so slow? all, this was a program originally written (without any use of local variables) to translate control codes on the 49G, by using SREPL, into which, for purposes of demonstrating how local variables can be used, I substituted a program written (again without any use of local variables) for the 48 series to act much like SREPL, except modified to not bother keeping a count from another program which does keep a count. Even without local variables, neither of the original programs are quite as fast as I'd like. To be sure, I do have a faster program to replace control codes with translation codes on the 49G, but it's much larger, and I don't think that the modest increase in speed adequately compensates for the huge increase in size. > Its not their use > itself but their *creation*, essentially done by ->. This command which > separates the names of local variables from the delimiter << or ' (for > user-defined functions) does a lot of work before making the actual > binding. -> (or xRPN-> in SysRPL display) decompiles as follows: > > :: > CKOATTNABORT (Abort in case CANCEL has been pressed in panic) > PTR 3889D (Collect the chosen variables and makes them named > LAMs) > DEPTH OVER #2* (after which these form a metaobject on the stack) > #>case (Check whether the input matches the number of LAMs) > DOBIND (If it matches do finally the binding) > NDROP (but if not remove the just created metaobject of LAMs > SETSTACKERR (and trap the user error as a STACK error) > ; > > The most involved of all this is PTR 3889D which is written in asm to > make it as fast as possible. This is a runstream operation, i.e., it > takes its arguments from the runstream, not from the data stack. Note > that it must also compute how many arguments have to be taken. Ok, could it be speeded up significantly? Since you say that PTR 3889D is the most involved of all, am I correct in assuming that this is where most of the time is used? If so, considering that that part is already written in asm to make it as fast as possible, I guess that any speed improvement would be modest, and no doubt at the expense of a larger size. > Thus, avoid using local variables whenever speed is important for your > program. Just use your brain for using the enormous power of pure stack > computation. Execution speed is important of course, at least when the program takes a noticeable time to execute. But it's not the only consideration; program size and memory usage are also worth considering, but offhand, I'm not sure that there would be much difference between using local variables and just using the stack. Ease of modification or debugging is another consideration; certainly much easier when local variables are used. But for ordinary, everyday use, development time is a very important consideration. How much development time should be spent in reducing execution time? Unless the program is to be used many times, I'd say not much. Using the stack manipulation commands and avoiding local variables is good, as long as the stack is relatively easy to keep track of, but sometimes that gets complicated, and letting the calculator keep track of where an object, as a named local variable, is, can make sense. When does it make sense to use your brain to keep track of the stack instead of using local variables that the calculator keeps track of? Well, it's no doubt politically incorrect to say so, but some brains are better at the task than others. And experience helps; I don't think that the adage that Practice makes perfect. is entirely true, but I'm sure that practice does make better. And the calculator is usually a tool used to accomplish some other task, so your mind is likely to be occupied with the overall task, not just the details of using the tool. You may well be dealing with distractions and interruptions while writing a program, which make it easy to lose track of where things are on the stack. Being tired makes mental tasks more difficult. And of course there's the need to quickly write a program that accomplishes what it's intended to do. So yes, when writing a program as a hobby or for practice, when there isn't a need to get it written in a hurry, or when writing a program for public distribution and to be used many times and by many people, it very often is better to avoid the use of local variables. But for ordinary user programs to be used relatively few times as an aid to accomplishing another task, it makes sense to use local variables whenever you start having trouble keeping track of where things are on the stack. -- James ==== > > > There's a LOCAL function which allows local variables to be defined that > > > don't need to be pulled from the stack. > > > > That is also in my 1.19-6 ROM for the 49G > > so it is not new > > Ah. I was looking at the 49G AUG which doesn't have that in it. Do you > know how to use it? I tried putting it in a program but it doesn't seem to > have any effect. The variables still show up in the Var menu. Does anyone > have a short example of the use of this command? TIA > > Tom Lake Use -> in RPN mode, you need to«go to the dark side to use LOCAL... BUT if you insist: '<-compiled.local.name=100' LOCAL Veli+Pekka ==== > Veli+Pekka I think you're losing your command over RPN. That should have been Veli Pekka + ;-) -- Bhuvanesh ==== I dwell in the LOCAL Dark Side... > > Veli+Pekka > > I think you're losing your command over RPN. That should have been > Veli Pekka + ;-) > > -- > Bhuvanesh ==== > Use -> in RPN mode, you need to«go to the dark side to use LOCAL... > BUT if you insist: > '<-compiled.local.name=100' LOCAL Well, I'm getting there , slowly. When I try the following, n remains defined even after the program ends although g and ng don't! { '<-ng=0.' '<-g=0.' '<-n=0.' } LOCAL RAND 1000. * 1. + IP 'n' STO 5. 'g' STO 10. 'ng' STO Any idea why? -- Tom Lake Experience keeps a dear school but fools will learn in no other - Poor Richard's Almanack ==== > > Use -> in RPN mode, you need to«go to the dark side to use LOCAL... > > BUT if you insist: > > '<-compiled.local.name=100' LOCAL > > Well, I'm getting there , slowly. When I try the following, n remains > defined even after the program ends although g and ng don't! > > { '<-ng=0.' '<-g=0.' '<-n=0.' } LOCAL RAND 1000. * 1. + IP 'n' STO @ 'n' is global now, use '<-n' instead 5. 'g' @ jada-jada-jada > STO 10. 'ng' STO @ ...... > > Any idea why? Veli+Pekka ==== > > > Use -> in RPN mode, you need to«go to the dark side to use LOCAL... > > > BUT if you insist: > > > '<-compiled.local.name=100' LOCAL > > > > Well, I'm getting there , slowly. When I try the following, n remains > > defined even after the program ends although g and ng don't! > > > > { '<-ng=0.' '<-g=0.' '<-n=0.' } LOCAL RAND 1000. * 1. + IP > 'n' STO @ 'n' is global now, use '<-n' instead > 5. 'g' @ jada-jada-jada > > STO 10. > 'ng' STO @ ...... > > > > Any idea why? > Veli+Pekka <- ??? Whats that for?? never seen backward arrows b4. TYIA ==== > > > > Use -> in RPN mode, you need to«go to the dark side to use LOCAL... > > > > BUT if you insist: > > > > '<-compiled.local.name=100' LOCAL > > > > > > Well, I'm getting there , slowly. When I try the following, n remains > > > defined even after the program ends although g and ng don't! > > > > > > { '<-ng=0.' '<-g=0.' '<-n=0.' } LOCAL RAND 1000. * 1. + IP > > 'n' STO @ 'n' is global now, use '<-n' instead > > 5. 'g' @ jada-jada-jada > > > STO 10. > > 'ng' STO @ ...... > > > > > > Any idea why? > > Veli+Pekka > > > > <- ??? > > Whats that for?? never seen backward arrows b4. Even if LOCAL would not need the left arrow I would still use it. First introduced in the HP 48 G[+]/GX <-local is a compiled local variable meant for using in between separate subprograms (therefore distinguisshed from true subroutines) Example: << -> <-compiled << IF TEST THEN MAX ELSE OTHER END >> >> 'MAIN' STO << <-compiled GLOBAL > >> 'TEST' STO << <-compiled 'GLOBAL' STO >> 'MAX' STO << <-compiled was smaller + 1 DISP >> 'OTHER' STO Without the <- you need to pass the value through the stack: << -> local << IF local TEST THEN local MAX ELSE local OTHER END >> >> * OR BETTER YET * << IF DUP TEST THEN MAX ELSE OTHER END >> >> 'MAIN' STO << GLOBAL > >> 'TEST' STO << 'GLOBAL' STO >> 'MAX' STO << was smaller + 1 DISP >> 'OTHER' STO OR in order to document the needed arguments << -> name << name GLOBAL > >> >> 'TEST' STO << -> maybe.max << maybe.max 'GLOBAL' STO >> >> 'MAX' STO << -> local << local was smaller + 1 DISP >> >> 'OTHER' STO Naturally one could use another global name for parameter passing: << 'PARAM' STO IF TEST THEN MAX ELSE OTHER END >> 'MAIN' STO << PARAM GLOBAL > >> 'TEST' STO << PARAM 'GLOBAL' STO >> 'MAX' STO << PARAM was smaller + 1 DISP >> 'OTHER' STO It's up to you to deside if you want to use ordinary local variable for better documentation or pure stack for speed or any other presented way or perhaps use a mixture of methods or even make new innovations it's all up to you - the versatility is provided by the RPL system! ==== > How do the function sets compare between the 49g and the 49g+? > > I understand that IR is restored so that is good. > > But are there any other changes, fixes, enhancements and/or new functions? I think only some bug-fixes, a better Filer, but no new commands some SysRPL entry points are NULLed and some new added a few new Virtual-Saturn op-codes added (ML coders be aware) + adds 25% to screen estate (header always on) and 3-4x speed plus ~12x I/O speed and the SD card. It's almost the same, but on steroids ==== I want a program to change directories so that all of its vars will be stored there? I have not found a CHDIR command or similar. Sincerely, Kevin Waite ==== > > I want a program to change directories so that all of its > vars will be stored there? > > I have not found a CHDIR command or similar. Say you have under HOME DIRS TEST and you want to change to HOME DIRS MATH EX.1 {HOME DIRS MATH EX.1} EVAL will do it You can save your current PATH via command PATH ==== I'd like to know if there's a place where I can find a complete list of every command used in the UserRPL programming language, with their description (not like the HP 49G+ Manual, that have only names...). SystemRPL is too difficult for me, now (I must to have more experience...), so I'd like to work with UserRPL, first. Daniele ==== > I'd like to know if there's a place where I can find a complete list > of every command used in the UserRPL programming language, with their > description (not like the HP 49G+ Manual, that have only names...). > SystemRPL is too difficult for me, now (I must to have more > experience...), so I'd like to work with UserRPL, first. I use SDIAG (without ML & SysRPL) from http://www.hpcalc.org/details.php?id=3940 So it stays with the calc. If you need more information, then there is no [real] AUR for the 49G(+) ==== I am thinking of buying a printed copy of the 49G AUR, but am wondering if it contains a full, alphabetically sorted and well indexed command reference similar to the 48G Series AUR that I already own? If anyone has both these AURs and could provide feedback on a comparison, I would much appreciate it. Mike Mander ==== > I am thinking of buying a printed copy of the 49G AUR, but am > wondering if it contains a full, alphabetically sorted and well > indexed command reference similar to the 48G Series AUR that I already > own? If anyone has both these AURs and could provide feedback on a > comparison, I would much appreciate it. No, save your money. The printed 49G Advanced User's *Guide* has the CAS commands, but most commands fall into the other commands category, so you need the PDF Advanced User's Guide for them. Maybe HP thinks that any advanced user already know everything about every one of the other commands. -- James ==== 49G AUR was similar to the 49G PDF Advanced User's Guide, but obviously it is not. Yup, I will save my money then... Mike Mander > > I am thinking of buying a printed copy of the 49G AUR, but am > > wondering if it contains a full, alphabetically sorted and well > > indexed command reference similar to the 48G Series AUR that I already > > own? If anyone has both these AURs and could provide feedback on a > > comparison, I would much appreciate it. > > No, save your money. The printed 49G Advanced User's *Guide* has the > CAS commands, but most commands fall into the other commands category, > so you need the PDF Advanced User's Guide for them. Maybe HP thinks that > any advanced user already know everything about every one of the > other commands. ==== > I am thinking of buying a printed copy of the 49G AUR, but am > wondering if it contains a full, alphabetically sorted and well > indexed command reference similar to the 48G Series AUR that I already > own? If anyone has both these AURs and could provide feedback on a > comparison, I would much appreciate it. > > Mike Mander Save yourself some money. This is directly from HP. References for non-CAS commands The help facility contains entries for all the commands developed for the CAS (Computer Algebraic System). There is a large number of other functions and commands that were originally developed for the HP 48G series calculators that are not included in the help facility. Good references for those commands are the HP 48G Series Userâs Guide (HP Part No. 00048-90126) and the HP 48G Series Advanced Userâs Reference Manual (HP Part No. 00048-90136) both published by Hewlett-Packard Company, Corvallis, Oregon, in 1993. HP49G+ user's manual Page C-13 Virgilio ==== I am still curious though about the printed 49G AUR (not 49g+) that Samson Cables is listing. I have downloaded (from hpcalc.org I think) an indexed 49G AUR that is split into many individual PDF files. It is missing a handful of new functions that were introduced with the last 49G ROMs, but still seems to be a decent and fairly complete reference. However, I always like having a printed manual as well and don't feel like cranking out the 800+ pages myself. I am wondering if anyone can compare the published 49G AUR with the online one I have described? Mike Mander > > > I am thinking of buying a printed copy of the 49G AUR, but am > > wondering if it contains a full, alphabetically sorted and well > > indexed command reference similar to the 48G Series AUR that I already > > own? If anyone has both these AURs and could provide feedback on a > > comparison, I would much appreciate it. > > > > Mike Mander > > Save yourself some money. This is directly from HP. > > References for non-CAS commands > > The help facility contains entries for all the commands developed for the > CAS (Computer Algebraic System). There is a large number of other > functions and commands that were originally developed for the HP 48G > series calculators that are not included in the help facility. Good > references for those commands are the HP 48G Series User?s Guide (HP Part > No. 00048-90126) and the HP 48G Series Advanced User?s Reference Manual > (HP Part No. 00048-90136) both published by Hewlett-Packard Company, > Corvallis, Oregon, in 1993. > > HP49G+ user's manual Page C-13 > > Virgilio ==== A while back there was some discussion about LAST MENU which can be obtained by LS (hold) NXT. That's quite handy. A step better would be a means of tracking changes and allowing the user to roll back through those changes, much like Undo in MS Office Aps (at least starting a few years ago). The current HP LAST... functions are behind Microsoft by a few years. :-) I would love an application for my 49G+ that tracked changes over several iterations, as it does currently with LAST CMD. The best for me would be two USER Keys: one that cycles the stack backwards in time (UNDO on steroids), and another that cycles the menu bar backwards in time (LAST MENU on steroids). In a perfect world these features would have functionality in Edit Mode as well. I suspect that the logging activity could slow down the calculator rather substantially (especially with a heavy stack), but this new G+ is pretty darn quick! Perhaps user-adjustable limits could be set as to how many iterations to log, and how large the stack memory would be to cause logging to be suspended. Chris ==== > > A while back there was some discussion about LAST MENU which can be obtained > by LS (hold) NXT. That's quite handy. A step better would be a means of > tracking changes and allowing the user to roll back through those changes, > much like Undo in MS Office Aps (at least starting a few years ago). The > current HP LAST... functions are behind Microsoft by a few years. :-) > > I would love an application for my 49G+ that tracked changes over several > iterations, as it does currently with LAST CMD. The best for me would be > two USER Keys: one that cycles the stack backwards in time (UNDO on Someone has programmed a multiple LASTARG using vectored enter and a list IIRC, but I can't find it on the www.hpcalc.org THe same could be used for storing the menus, the path, etc... For more UNDO levels I use HALT perhaps another HALT (& CLEAR to start on a fresh table, unless I need to use the old values) and then later when results of that calculation is stored CLEAR CONT UNDO Veli+Pekka > steroids), and another that cycles the menu bar backwards in time (LAST MENU > on steroids). In a perfect world these features would have functionality in > Edit Mode as well. > > I suspect that the logging activity could slow down the calculator rather > substantially (especially with a heavy stack), but this new G+ is pretty > darn quick! Perhaps user-adjustable limits could be set as to how many > iterations to log, and how large the stack memory would be to cause logging > to be suspended. > > Chris > > ==== I have found that everytime I meke my HP49G draw a function, it assign a nummeric value to X. Usually a real number. I also found that it's only when i use the Auto in plot window that this happens. Let me take an example: * I plot the function Y1(X)=X^2 * I change to plot window, sets H-View to 0 and 5. Next, I press F4 (AUTO), wich makes V-View -3.75 and 25. * Now, i don't let the calc draws the graph. Instead i cancel. * After pushing the VAR button, I can see that the new variable X is being created. * The variable X now contains the number 5.125. So I have two questions about this weird behaviour: * What does the number 5.125 mean? Can I use it to something, or is there a bug in my HP, so that X as an temporary variable while V-View is calculated, is not removed when the calculation is completed? * Is there any posibility that I can use another variable name than X, Ex QZZX! as i probably never will use myself. Of course I still want to write Y1(X)=X^2 and not Y1(QZZX)=QZZX^2. That's what happens if I in plot setup change Indep form 'X' to 'QZZX'. ----------------- Another thing: My calc can't solve 'SIN(X)-1/4*X' using SOLVEVX. Instead it gives an error. When i plotted the function, I saw that the graph had three posible solutions. Is there any way I can get the calc returning those three answers? Another guy in my class have a much cheaper calc, an texas i think. He's calc was able to return all three answers. That's was quite frustrating, because my HP49G is indeed ment to be the greatest one. Being more expensiv and advanced than both texas and casio calcs. I know I can draw the graph and find any extrema, but that's not my goal. So my question is (probably as expected): Is there any built-in function on my HP49g that I can use to find all values for X that makes 'SIN(X)-1/4*X' return 0 within an spesific range? Ps: Have upgraded to ROM 1.19-6. ==== > I have found that everytime I meke my HP49G draw a function, it assign a > nummeric value to X. Usually a real number. I also found that it's only when > i use the Auto in plot window that this happens. Let me take an example: > * I plot the function Y1(X)=X^2 > * I change to plot window, sets H-View to 0 and 5. Next, I press F4 > (AUTO), wich makes V-View -3.75 and 25. > * Now, i don't let the calc draws the graph. Instead i cancel. > * After pushing the VAR button, I can see that the new variable X is being > created. > * The variable X now contains the number 5.125. AUTO samples maybe 20 values of the function in order to determine the suitable plotting ranges the last X used is in X I suggest that you use CleanSolve from www.hpcalc.org or get used to that idea of X needs to be purged > So I have two questions about this weird behaviour: > * What does the number 5.125 mean? Can I use it to something, or is there a > bug in my HP, so that X as an temporary variable while V-View is calculated, > is not removed when the calculation is completed? > * Is there any posibility that I can use another variable name than X, Ex > QZZX! as i probably never will use myself. Of course I still want to write > Y1(X)=X^2 and not Y1(QZZX)=QZZX^2. That's what happens if I in plot > setup change Indep form 'X' to 'QZZX'. can't you just use X1 ? > ----------------- > > Another thing: > My calc can't solve 'SIN(X)-1/4*X' using SOLVEVX. Instead it gives an error. > When i plotted the function, I saw that the graph had three posible > solutions. Is there any way I can get the calc returning those three > answers? > Another guy in my class have a much cheaper calc, an texas i think. He's > calc was able to return all three answers. That's was quite frustrating, > because my HP49G is indeed ment to be the greatest one. Being more expensiv > and advanced than both texas and casio calcs. > I know I can draw the graph and find any extrema, but that's not my goal. > So my question is (probably as expected): Is there any built-in function on > my HP49g that I can use to find all values for X that makes 'SIN(X)-1/4*X' > return 0 within an spesific range? > > Ps: Have upgraded to ROM 1.19-6. 'SIN(X)-1/4*X' 'X' 1.5 ROOT (do it with -1.5 and 0.) Veli+Pekka ==== > 'SIN(X)-1/4*X' > 'X' > 1.5 > ROOT > (do it with -1.5 and 0.) > Veli+Pekka Whats the 1.5 (or 0 or -1.5 )?? TY-IA ==== > > > 'SIN(X)-1/4*X' > > 'X' > > 1.5 > > ROOT > > (do it with -1.5 and 0.) > > Veli+Pekka > > > Whats the 1.5 (or 0 or -1.5 )?? They are your best guesses where to seek for a root In case of multiple roots you need give the ROOT a hint If know where only a single root is, then you can bracket it like this {lower upper} and ROOT will then *always find it !! (or in case of rounding errors claim a local maxima or minima is found) ==== Veli-Pekka Nousiainen skrev i melding > > > I have found that everytime I meke my HP49G draw a function, it assign a > > nummeric value to X. Usually a real number. I also found that it's only > when > > i use the Auto in plot window that this happens. Let me take an example: > > * I plot the function Y1(X)=X^2 > > * I change to plot window, sets H-View to 0 and 5. Next, I press F4 > > (AUTO), wich makes V-View -3.75 and 25. > > * Now, i don't let the calc draws the graph. Instead i cancel. > > * After pushing the VAR button, I can see that the new variable X is being > > created. > > * The variable X now contains the number 5.125. > > AUTO samples maybe 20 values of the function > in order to determine the suitable plotting ranges > the last X used is in X > I suggest that you use CleanSolve from www.hpcalc.org > or get used to that idea of X needs to be purged I hope that will be fixed in next ROM. Now I was getting a little disorientated. The short description of CleanSolve says that it can be used to solve an equation, but you say (as i understand) that it can be helpfull due to the variable 'X' that apears. ==== > > > I have found that everytime I meke my HP49G draw a function, it assign a > > nummeric value to X. Usually a real number. I also found that it's only > when > > i use the Auto in plot window that this happens. Let me take an example: > > * I plot the function Y1(X)=X^2 > > * I change to plot window, sets H-View to 0 and 5. Next, I press F4 > > (AUTO), wich makes V-View -3.75 and 25. > > * Now, i don't let the calc draws the graph. Instead i cancel. > > * After pushing the VAR button, I can see that the new variable X is being > > created. > > * The variable X now contains the number 5.125. > > AUTO samples maybe 20 values of the function > in order to determine the suitable plotting ranges > the last X used is in X > I suggest that you use CleanSolve from www.hpcalc.org > or get used to that idea of X needs to be purged it's here http://www.hpcalc.org/details.php?id=3141 ==== > > if you do 256 ATTACH, the ASM function should work flawlessly... > > > > > I would like to begin SysRPL programming on my HP49G+, but MASD (library > > 257) won't attach correctly. Has anyone out there gotten MASD attached on > > a -49G+? > > > > MrM > > Make sure you have flag -92 SET, so that MASD works in SysRPL mode rather than assembly mode. ==== Folks, I have an 8th ed. version of this guide for sale. (December 1994) Very fine condition. $30.00 shipped USA/Canada Bill ==== By any chance, would you have also a hp48sx user manual(s) to sale ? Herve a .8ecrit dans le message de > > Folks, > > I have an 8th ed. version of this guide for sale. (December 1994) > Very fine condition. > > $30.00 shipped USA/Canada > > Bill > ==== > Perhaps it just my paranoid nature left over from 9-11, but the name and > the repeated off subject posts make me wonder if there is more than just > garbage in these odd patterns in his signature. Anybody else think the > FBI should take a look at these? > KW It is just ASCII art. Stenographed ascii art? I don't think.. :) ==== Greetings Is it OK to use the following libraries on the HP49G+ (all from HPCalc.org): Emacs version 1.11 by Carsten Dominik CQIF version 1.7.7F by Pierre Tardy Nosy version 4.0 by Jurjen Bos I do not want to load them if they are just going to crash the machine. Gordon Pegue ==== GP> Greetings GP> Is it OK to use the following libraries on the HP49G+ (all from HPCalc.org): GP> Emacs version 1.11 by Carsten Dominik GP> CQIF version 1.7.7F by Pierre Tardy GP> Nosy version 4.0 by Jurjen Bos Emacs 1.11 largely works but has problems with the longpress functionality. Better to get Emacs 1.12 from my site for the time being http://zon.astro.uva.nl/~dominik/hpcalc/emacs/ You need to create a global variable (in HOME) 'lht', and store a number like 900. into it. This will also slow down KEYMAN's long-hold time correctly. I don't have a 49G+ yet (still waiting for better news about the keyboard), but from WR's posts I read that Nosy does not work on rom 1.20, but it works fine on rom 1.22. I don't know about CQIF. - Carsten GP> I do not want to load them if they are just going to crash the machine. GP> Gordon Pegue ==== > Emacs 1.11 largely works but has problems with the longpress > functionality. Better to get Emacs 1.12 from my site for the time > being. Long-press is not the only problem with Emacs. There are problems with the header size change in Emacs' Search and Replace option, manageable only by those who know its mechanisms, hardly by the normal user. In addition, RPLED has an old bug which easily leads to a TTRM, namely when a control alarm arrives during a session with RPLED. Who does not know what a control alarm is should read the old HP48 docus. It is not mentioned in the 49 docus, even not in the huge User's Guide from the internet where it should have been explained. For this and other reason I made a tool Topkeys+ containing a serie of useful key assignments for the HP49G+, in particular a bug-free and faster variant of RPLED on longhold DownArrow. It can still jointly be used with the application memu of Emacs. Clearly, Topkeys+ includes also a TEVAL option, just on the longhold EVAL key. A very nice assignment can be made at the RS DownArrow key which yields details on the current menu. Clearly, the assignment preseres this function in default mode. However, in edit mode it moves the cursor down only if DownArrow is normally pressed. If longer hold details on the current menu are displayed. The edit mode is not left or affected in any way by this intermediate display. I wunder why such an option is not included in the built-in Edit menu. At any rate, I included the stable PTR 258D6 (display details on the menu) in my personal extable and called it DETAIL. This useful pointer should also be added to the expected new extable, at least to Thomas Rast's extable version. Wolfgang http://page.mi.fu-berlin.de/~raut/WR49/Topkeys+.htm ==== WR> At any rate, I included the stable PTR 258D6 (display details on the WR> menu) in my personal extable and called it DETAIL. This useful pointer WR> should also be added to the expected new extable, at least to Thomas WR> Rast's extable version. The official name of this pointer is DoReview. It has been the my list for a long time but it did not make it into extable because I did not have a stack diagram for it. Now I do - so it will be in the next - Carsten ==== > WR> At any rate, I included the stable PTR 258D6 (display details on the > WR> menu) in my personal extable ... > The official name of this pointer is DoReview. It has been the my > list for a long time but it did not make it into extable because I did > not have a stack diagram for it. Now I do - so it will be in the next Indeed amazing that the pointer DoReview works so phantastically also in Edit mode. And this is extremely useful. Those who use Keyman should give the Rightshift DownArrow key the following extended functionality: :: TakeOver NoEdit?case DoReview (Review the menu in default mode) ROMPTR 4B 8 :: (execute this if DownArrow is shortly pressed) CODE GOSBVL SAVPTR GOSBVL CMDSIZE (fast computation of the edited object's size) GOVLNG PUSH#ALOOP ENDCODE STO_CURS_POS2 (puts cursor at the very end) SetDA2aNoCh ; DoReview (but review the menu if DownArrow is longpressed. ; @ With the Keyman tools this assignment may be realized as follows: Put PTR 258D6 on level 2, Standard assignment of Righshift DownArrow on Level 1 and press IfL. Leave this in Level 1 and put PTR 258D6 again on Level 2 and press IfE. Process the result with ->TO and assign it with A?D to Righshift DownArrow. Then Rightshift DownArrow is acting as follows: << If in default mode I review the menu. But while in edit mode shortly pressed I shift the cursor to the end whereas if I'm hold a bit longer I again review the menu without doing any harm to the open editor >> This all has nothing to do with the dispensable Rightshift-hold DownArrow key which may be used for RPLED of Emacs, for instance. - Wolfgang http://page.mi.fu-berlin.de/~raut/WR49/Keyman.htm ==== > > WR> At any rate, I included the stable PTR 258D6 (display details on the > > WR> menu) in my personal extable ... > > > The official name of this pointer is DoReview. It has been the my > > list for a long time but it did not make it into extable because I did > > not have a stack diagram for it. Now I do - so it will be in the next > > Indeed amazing that the pointer DoReview works so phantastically also in > Edit mode. And this is extremely useful. Those who use Keyman should > give the Rightshift DownArrow key the following extended functionality: And that goes to the Keyman examples, right? ==== > Greetings > > Is it OK to use the following libraries on the HP49G+ (all from HPCalc.org): > > Emacs version 1.11 by Carsten Dominik > CQIF version 1.7.7F by Pierre Tardy > Nosy version 4.0 by Jurjen Bos Nosy works, but the cursor moves too fast unless you're careful. No idea about the others, but I recall someone saying Emacs works. Al > > I do not want to load them if they are just going to crash the machine. > > Gordon Pegue ==== I have posted this earlier. Is anyone working on a translation for Crazy cars to the 49G? Someone said they were and that they were going to release it soon. Its been almost 2 months and I still can't see anything posted? Thanxz in Advance, CID ==== Put a message (in English) on http://www.hp-network.com/forum > I have posted this earlier. Is anyone working on a translation for > Crazy cars to the 49G? Someone said they were and that they were going > to release it soon. Its been almost 2 months and I still can't see > anything posted? > > Thanxz in Advance, > > CID ==== I have a sneaking suspicion that there are a lot of people in this NG that are good in many areas of mathematics. I work in construction, and have found basic math, geometry, algebra, and trigonometry useful, as well as I know them. I would like recommendations of good books or websites to brush up on and learn more about these. For me, it's fun. Also, I've never done calculus, and would like to know useful applications for it, as well as good books, etc. as above. GC ==== Well, thank you all for the suggestions. I'm not going to get all of these, but they all look like good books. I'm sure I will enjoy them. GC > I have a sneaking suspicion that there are a lot of people in this NG that > are good in many areas of mathematics. > > I work in construction, and have found basic math, geometry, algebra, and > trigonometry useful, as well as I know them. > > I would like recommendations of good books or websites to brush up on and > learn more about these. For me, it's fun. > > Also, I've never done calculus, and would like to know useful applications > for it, as well as good books, etc. as above. > > > GC > > ==== >I have a sneaking suspicion that there are a lot of people in this NG that >are good in many areas of mathematics. > >I work in construction, and have found basic math, geometry, algebra, and >trigonometry useful, as well as I know them. > >I would like recommendations of good books or websites to brush up on and >learn more about these. For me, it's fun. > >Also, I've never done calculus, and would like to know useful applications >for it, as well as good books, etc. as above. > > >GC > It's not a pure mathematical book, but I'll order it quickly: A New Kind of Science by Stephen Wolfram The author is the author :-) of Mathematica, the best simbolic calculation program (for me). It talks about all sciences, from mathematics to finance, fractals, geology etc. There are hundreds of beatiful images created with Mathematica, and it's huge!!! about 1200 pages. Daniele ==== GC, >[...] >I would like recommendations of good books or websites to brush up on and >learn more about these. For me, it's fun. >[...] Here are some books I really enjoyed because they offer some great (and not at all boring) history as well. But they're definitely more about how math developed rather than teaching how to do it. So if you want text books, skip these! Livio: The Golden Ratio - Phi Maor: e - The story of a number Others recommended to me that I haven't gotten to yet: Blatner: The Joy of Pi Bodanis: e=mc^2 Bunch: The Kingdom of Infinite Number Maor: To Infinity and Beyond Nahin: The Story of (sqrt)-1 Seife: Zero Have fun! Mike ==== If you can find a copy of these old books by Isaac Asimov, they are really good too! Realm of Numbers Realm of Algebra (often called the best book for learning algebra ever written) Realm of Measure They prop up on the prominent internet auction site with some regularity. That's where I got my copies. FWIW, Gene -- * These statements and opinions are mine alone and do not reflect my employer's views. * > GC, > > >[...] > >I would like recommendations of good books or websites to brush up on and > >learn more about these. For me, it's fun. > >[...] > > Here are some books I really enjoyed because they offer some > great (and not at all boring) history as well. But they're > definitely more about how math developed rather than teaching > how to do it. So if you want text books, skip these! > > Livio: The Golden Ratio - Phi > Maor: e - The story of a number > > Others recommended to me that I haven't gotten to yet: > > Blatner: The Joy of Pi > Bodanis: e=mc^2 > Bunch: The Kingdom of Infinite Number > Maor: To Infinity and Beyond > Nahin: The Story of (sqrt)-1 > Seife: Zero > > Have fun! > Mike ==== i have a soft spot for the Stroud books http://www.amazon.com/exec/obidos/search-handle-form/103-1191043-6838251 I havent used them as a student but i have recomended them to my own engineering students who have had great succcess with mr Strouds very particular way of teaching mathematical problem solving. > I have a sneaking suspicion that there are a lot of people in this NG that > are good in many areas of mathematics. > > I work in construction, and have found basic math, geometry, algebra, and > trigonometry useful, as well as I know them. > > I would like recommendations of good books or websites to brush up on and > learn more about these. For me, it's fun. > > Also, I've never done calculus, and would like to know useful applications > for it, as well as good books, etc. as above. > > > GC > > ==== There are thousands and thousands of excellent Math books, and I'm sure plenty of people will offer their suggestions, but trust me, you can't do better than getting this one first: What Is Mathematics?: An Elementary Approach to Ideas and Methods by Richard Courant and Herbert Robbins Here's a link to it (if copying to the the URL field of your web browser, remember to place it all in a single line) http://www.amazon.com/exec/obidos/tg/detail/-/0195105192/qid=1068810774/sr=1 -1/ref=sr_1_1/102-6918850-4792151?v=glance&s=books If you visit the link, you'll be able to read up to 18 customer's reviews, as well as a number of editorial reviews and comments, including a comment by Albert Einstein himself: A lucid representation of the fundamental concepts and methods of the whole field of mathematics ... Easily understandable. This book has it all: it's extremely comprehensive, dealing with all the most important mathematical fields, plus it's very clear, easy to understand, and most importantly, will appeal to anyone's imagination and thus it makes excellent, very enjoyable reading. I could recommend you at least a dozen additional excellent Math books, but honestly, I think this is the one you ought to see first. Welcome to the wonderful world of Mathematics. ==== > I have a sneaking suspicion that there are a lot of people in this NG that > are good in many areas of mathematics. > > I work in construction, and have found basic math, geometry, algebra, and > trigonometry useful, as well as I know them. > > I would like recommendations of good books or websites to brush up on and > learn more about these. For me, it's fun. > > Also, I've never done calculus, and would like to know useful applications > for it, as well as good books, etc. as above. > > > GC It quite off topic, but give Calculus ... author Adams a look. it«s thich, heavy and full of examples. I RCL that there is even an solutions book availible. ==== Use Advanced Search on www.hpcalc.org for Nick as author You will find complex, calculus, trigonometric, etc Marathons They are free and show everything on a 49G Veli+Pekka > I have a sneaking suspicion that there are a lot of people in this NG that > are good in many areas of mathematics. > > I work in construction, and have found basic math, geometry, algebra, and > trigonometry useful, as well as I know them. > > I would like recommendations of good books or websites to brush up on and > learn more about these. For me, it's fun. > > Also, I've never done calculus, and would like to know useful applications > for it, as well as good books, etc. as above. > > > GC > > ==== I live in Ontario Canada. Can anyone tell me where I can buy an HP 49G+ nearby? Jacques ==== Same here. After I created a library with the built-in command (no problems when using the one in OT49!) and a worm start my HP49G+ is completely stuck. It passes ok all standard tests(!!!), but doesn't work. It's ON, I can change the display contrast and when pressing LS RS or Alpha indicators flash. Neither reset nor ON+A+F works. Left without batteries for long hours also doesn't seem to help :( Please someone, help us! > My new calculator, a hp 49g+, is completely stuck. When i switch it > or No then only leads to beeping, no other response. > I experienced similar problems with my former hp, a 49g. The solution > then was either holding down the backspace button when starting, or > setting the device in terminal mode and activating the Flash Download > Program -I then got the option to manually erase user banks i > suspected to be corrupt. > > The problem is that i can't figure out how to do it on my 49g+. > Do anyone of you know? ==== I haven't done it - but isn't that what the pinhole on the back about 1.5 inches up from the bottom is for? ==== Leaving the batteries out won't help if you load each time a bad library. Try, while soft resetting, and keep pressed the backspace key but be very quick immediately following the release of the ON-C combination. You will, hopefully, start a new session with no library loaded at all. Then, move the bad lib from the associated port into main RAM and work on it! !Demeter! ==== I've got *bad* library in port 2 for sure. Whatever I do I always get the Try to Recover memory messgage and no matter what the answer I get the blank screen after short display ... CAS by ... and Powered by ... Still passes all tests :- > Leaving the batteries out won't help if you > load each time a bad library. Try, while soft > resetting, and keep pressed the backspace key > but be very quick immediately following the > release of the ON-C combination. You will, > hopefully, start a new session with no library > loaded at all. Then, move the bad lib from the > associated port into main RAM and work on it! > > !Demeter! ==== > I've got *bad* library in port 2 for sure. Whatever I do I always get the > Try to Recover memory messgage and no matter what the answer I get the > blank screen after short display ... CAS by ... and Powered by ... > Still passes all tests :- > > > > > Leaving the batteries out won't help if you > > load each time a bad library. Try, while soft > > resetting, and keep pressed the backspace key > > but be very quick immediately following the > > release of the ON-C combination. You will, > > hopefully, start a new session with no library > > loaded at all. Then, move the bad lib from the > > associated port into main RAM and work on it! > > > > !Demeter! > > IIRC, there is a way on the 48 and 49 to turn on the calculator without loading any libraries that might also work on the 49+. Something like holding down ON and ENTER, or maybe the ON and + keys, simultaneously when turning the HP on. Once turned on in this fashion, one can purge offending libraries in the usual way. Does anyone recall the details of this technique? It was posted here a month or so ago. ==== > IIRC, there is a way on the 48 and 49 to turn on the calculator > without loading any libraries that might also work on the 49+. > > Something like holding down ON and ENTER, or maybe the ON and + > keys, simultaneously when turning the HP on. Once turned on in this > fashion, one can purge offending libraries in the usual way. > > Does anyone recall the details of this technique? It was posted here > a month or so ago. I'm not sure on the 48 series, but on the 49G, hold down the backspace key while doing a warmstart. I haven't tested this on the 49g+ yet. -- James ==== At least paper-clip reset works on hp48... not here! :( cheers > > > > > IIRC, there is a way on the 48 and 49 to turn on the calculator > > without loading any libraries that might also work on the 49+. > > > > Something like holding down ON and ENTER, or maybe the ON and + > > keys, simultaneously when turning the HP on. Once turned on in this > > fashion, one can purge offending libraries in the usual way. > > > > Does anyone recall the details of this technique? It was posted here > > a month or so ago. > > I'm not sure on the 48 series, but on the 49G, hold down the backspace > key while doing a warmstart. I haven't tested this on the 49g+ yet. > > -- > James > ==== | Same here. After I created a library with the built-in command (no problems | when using the one in OT49!) and a worm start my HP49G+ is completely stuck. | It passes ok all standard tests(!!!), but doesn't work. It's ON, I can | change the display contrast and when pressing LS RS or Alpha indicators | flash. Neither reset nor ON+A+F works. Left without batteries for long hours | also doesn't seem to help :( | Please someone, help us! Did you remove the backup battery (button cell) as well? ==== WinXP Pro USB 1.0.4.2, installed on two front ports conn4x: 1689, 1783, both nehave the same ROM 1.22 1) Calc connected and recognized 2) conn4x startted, x9+ device selected 3) XSERVER started on calc 4) connection made, window shows calc files 5A) a library from the PC moves, but doesn't show on calc B) a text fil warns first then also shows the bytes trnasferred and ends but theres is nothing on calc 6 calc disconnected and checked - no files 7 calc reconnected as above steps 1..5 8 a file from calc moves to PC side and shows there and can be opened there 9 Call William Graves ==== I have the same problem. Seems to depend on the file to be transferred. Some programs show up on the 49G+, others don't. Could it be something with a wrong header? Were you able to read the message that flashes up in conn4x just before the unsuccessful end of the transfer? I am too slow. :-) Holger > WinXP Pro > USB 1.0.4.2, installed on two front ports > conn4x: 1689, 1783, both nehave the same > ROM 1.22 > > 1) Calc connected and recognized > 2) conn4x startted, x9+ device selected > 3) XSERVER started on calc > 4) connection made, window shows calc files > 5A) a library from the PC moves, but doesn't show on calc > B) a text fil warns first then also shows the bytes trnasferred and ends > but theres is nothing on calc > 6 calc disconnected and checked - no files > 7 calc reconnected as above steps 1..5 > 8 a file from calc moves to PC side and shows there and can be opened > there 9 Call William Graves > ==== > I have the same problem. > Seems to depend on the file to be transferred. Some programs show up on the > 49G+, others don't. Could it be something with a wrong header? > > Were you able to read the message that flashes up in conn4x just before the > unsuccessful end of the transfer? I am too slow. :-) You PC is too fast, try some hogging application and look again there is at least a windows button saying [ CANCEL ]... I have a 2GHz P4 with 1/2MB cache and 1/2 GB memoey and WinXP Pro Radeon 9100 is too fast in this simple case... During a text send that message waits there to warn about possible harmful content, but the transfer is very very fast with the flashing message at the end. If I remember (which I doubt) I try while photoshop filters something... ==== I remember a nice equation library on the 48G. Where can I find this on the 49g+? ==== > I remember a nice equation library on the 48G. Where can I find this on the 49g+? http://www.hpcalc.org/details.php?id=3181 ==== ==== > On a related note, does anyone know the font's that were used in the documentation? It's difficult to read because it's pulling Wing Dings in as a font for some headers. Dave ==== i have the hp49g+ with rom version 1.22, even if the rom is the last the life of batteries is very short, about 2 weeks.Strangely i verified the charge of batteries by a battery test and the charge was still good (at 90% of their max)...(the batteries were 3 AAA duracexx........M3..., a good product i think!!Or not?!!), but they are discharge for the calculator, i don't understand the way. I have forgotten..... with these i updated the rom and i installed only the library: solvesys!! Somebody can help me??? Somebody have the same problem?? peppe ==== Are they rechargeable? The low voltage triggering mechanism in the HP49G+ gets rightfully activated because they are considered good when they are at 1.2 volts which is the low for the regular alcaline (nominal 1.5 volts). Try your test machine (HP49G+) with alcaline or regular AAA batteries. !Demeter! ==== Demeter ha scritto nel messaggio > > Are they rechargeable? The low voltage triggering mechanism > in the HP49G+ gets rightfully activated because they are > considered good when they are at 1.2 volts which is the low > for the regular alcaline (nominal 1.5 volts). Try your test > machine (HP49G+) with alcaline or regular AAA batteries. > > !Demeter! No....... they aren't rechargeable, ihave tested their voltage and current, current = about 4.5 A and voltage about 1.4 V. Their are alcaline batteries!!!!! peppe ==== > > Demeter ha scritto nel messaggio > > > > Are they rechargeable? The low voltage triggering mechanism > > in the HP49G+ gets rightfully activated because they are > > considered good when they are at 1.2 volts which is the low > > for the regular alcaline (nominal 1.5 volts). Try your test > > machine (HP49G+) with alcaline or regular AAA batteries. > > > > !Demeter! > > No....... they aren't rechargeable, ihave tested their voltage and current, > current = about 4.5 A That's an awful lot of current. You didn't by any chance test them by connecting an ammeter *across* them, instead of in series with them? If so, the batteries will now be very flat... James ==== James Lothian ha scritto nel > > > > Demeter ha scritto nel messaggio > > > > > > Are they rechargeable? The low voltage triggering mechanism > > > in the HP49G+ gets rightfully activated because they are > > > considered good when they are at 1.2 volts which is the low > > > for the regular alcaline (nominal 1.5 volts). Try your test > > > machine (HP49G+) with alcaline or regular AAA batteries. > > > > > > !Demeter! > > > > No....... they aren't rechargeable, ihave tested their voltage and current, > > current = about 4.5 A > > That's an awful lot of current. You didn't by any chance test them by > connecting > an ammeter *across* them, instead of in series with them? If so, the > batteries > will now be very flat... > > James sorry, i have mistaken to type the number... the current is about 1.5A, the ammeter was in series with them!!! peppe ==== I live in Ontario Canada. Can anyone tell me where I can buy an HP 49G+ nearby? Jacques ==== Jacques, I looked at the HP Web Site and selected Canada as the country. It then lists the retail outlets which includes Grand & Toy, and Office Depot. I am visiting Toronto this week, and shall be looking to see who has them. There are some specialist electronics outlets on Yonge Street who often have high-end calculators, and sometimes are prepared to do a little haggling. This is very similar to Tottenham Court Road in London (England). Fun, but only if you know exactly what you are looking for, and at what price :-) Based on some of the experiences reported recently in this group, it seems a you can test the unit before you leave the store. Mark. > > I live in Ontario Canada. Can anyone tell me where I can buy an HP 49G+ > nearby? > > > Jacques > > > ==== When you use Find at Emacs with the full screen it happens a bug. The word Find: is covered by some code. What am I supposed to do? ==== Using the HP 49g+, when I type Find at Emacs with the full screen it happens a bug. The word Find: is covered by some code. What am I supposed to do? ==== > Using the HP 49g+, when I type Find at Emacs with the full screen it > happens a bug. The word Find: is covered by some code. What am I > supposed to do? Cry? [Until it is fixed by the author as a XMas present to all of us] ==== I saw a USB mini-b cable at the department store today. It looked a lot like the USB cable that came with my HP-49G+. It would be an additional plus for the 49G+ (bad pun that should have been avoided) if its USB cable is not some exotic, HP-only-source design. Can anyone confirm that the HP-49G+ calculator-end socket is a USB mini-b? ==== > Can anyone confirm that the HP-49G+ calculator-end socket is a USB > mini-b? Well, it's the same cable that my Olympus digital camera uses, so I suppose it's fairly common. Bye, Andrea The graveyards are full of indispensable men. Charles de Gaulle (1890-1970) ==== I confirm. Due to recent outbreaks in this NG, if you want to receive my wonderfull 200 lines signature fixed font, please contact me directly :-) > I saw a USB mini-b cable at the department store today. It looked a > lot like the USB cable that came with my HP-49G+. It would be an > additional plus for the 49G+ (bad pun that should have been avoided) > if its USB cable is not some exotic, HP-only-source design. > > Can anyone confirm that the HP-49G+ calculator-end socket is a USB > mini-b? ==== > > I confirm. > > > Due to recent outbreaks in this NG, if you want to receive my wonderfull 200 > lines signature fixed font, please contact me directly :-) Naaah! I think we could allow the only remaining ACO member *once* send that sig in this NG. Only from you, Wolwerine Can we folks? Veli+Pekka ==== > > Due to recent outbreaks in this NG, if you want to receive my wonderfull 200 > lines signature fixed font, please contact me directly :-) I do I do!! BTW: what exactly does fixed font mean?? ==== > BTW: what exactly does fixed font mean?? Well, not fixed as in repaired, but fixed as in not changeable, and opposed to a proportional font. Fixed fonts are what you get when using a typewriter; every character has the same width as every other character. With a proportion font some characters are wider than others; for example, a W is wider than an l. Looking at tables and program listings with a proportional font generally puts things out of alignment, assuming that they were written with a fixed font or even a different proportional font. I suppose that ASCII art wouldn't appear as intended either. -- James ==== > > > > > BTW: what exactly does fixed font mean?? > > Well, not fixed as in repaired, but fixed as in not changeable, and > opposed to a proportional font. Fixed fonts are what you get when > using a typewriter; every character has the same width as every other > character. With a proportion font some characters are wider than > others; for example, a W is wider than an l. What is here called fixed is more commonly called fixed width meaning that every character is displayed in a space of the same width as every other character. This is like the old mechanical typrewriters which mored the platen over the same distance after each character before typing the next one. Fixed width is also useful in financial text, when you want numbers to line up nicely in columns. > > Looking at tables and program listings with a proportional font > generally puts things out of alignment, assuming that they were written > with a fixed font or even a different proportional font. I suppose that > ASCII art wouldn't appear as intended either. Since each different proportional font tends to use different width patterns than others, vertical alignments will differ from one to another, whereas with fixed width fonts, changing fonts does not affect vertical alignment at all (unless too wide a font causes lines to overflow and get broken up). ==== Just received a new 49g+; my first upgrade since purchasing two 11Cs years ago. What a difference! The programming chapter of the Users Guide refers to sample programs provided on the CD that came with the calculator. Has anyone found these? KB ==== Amazon.com have this calculator for $141 with Free Shipping. ==== > Amazon.com have this calculator for $141 with Free Shipping. I think they are no longer selling them. Amazon now just links to other retailers. http://www.amazon.com/exec/obidos/tg/stores/offering/list/-/B0000DK662/all/r ef=dp_pb_a/102-8184222-8217734 ==== > Amazon.com have this calculator for $141 with Free Shipping. I bought mine with them... :) ==== Does anyone know where I can find an hp49g+ anywhere in South Carolina? If so, which store and city is it in in South Carolina. I teach mathematics in Goose Creek South Carolina near Charleston, SC. I live in Summerville SC. Wilber Tindall > > Amazon.com have this calculator for $141 with Free Shipping. > > I bought mine with them... :) ==== I am working on the 2nd-order ODE example found on 16-64 of the HP49G+ User's Manual (the 10MB one). It uses the numerical differential eqn solver to solve the following: x(t)= -18.75x(t)-1.962x'(t) with x'(0)=6 and x(0)=0. The goal is to find x and x' at t=2. I entered it into the solver as shown in the example, but receive the following error message: NON-ALGEBRAIC IN EXPRESSION. PLEASE NOTE: that I am not running a G+; rather, it is a 49G using 1.19-6 (the most recent build for that calculator). Is there an error in the manual, or can this problem only be done on a 49G+? Any help would be appreciated. NOTE: on the follow-up example (graphical solution to the same problem on p. 16-66), the calculator doesn't seem to like [0 6] as the initial solution. ==== > I entered it into the solver as shown in the example, but receive the > following error message: NON-ALGEBRAIC IN EXPRESSION. This was a bug fixed in ROM 1.20 of the 49G+ .. ==== > > I entered it into the solver as shown in the example, but receive the > > following error message: NON-ALGEBRAIC IN EXPRESSION. > > This was a bug fixed in ROM 1.20 of the 49G+ .. > the 49G (which is what I own), or do I need to get myself a G+? Jeff ==== > > > I entered it into the solver as shown in the example, but receive the > > > following error message: NON-ALGEBRAIC IN EXPRESSION. > > > > This was a bug fixed in ROM 1.20 of the 49G+ .. > > > > > the 49G (which is what I own), or do I need to get myself a G+? Get a g+, + 1.20 ROM => +1.22 ROM => + 25% LCD + plastic keys + better LCD cover + better hard case + SD card + better contrast on the shifted key labels + [EVAL] and [ ' ] tick back to primary key level + IRDA at 115200 baud + USB at 115200 baud, cable included + 2 to 7* speed + modern 75MHz ARM9 CPU for future enhancements - battery life (Û&%Û&%Û 75MHz and all that speed) ==== > > I entered it into the solver as shown in the example, but receive the > > following error message: NON-ALGEBRAIC IN EXPRESSION. > > This was a bug fixed in ROM 1.20 of the 49G+ .. > Could you kindly release in your web-pages the bug fixes & enhancements since 1.19-6 to 1.20 and from 1.20 to 1.22 and later up to 1.24, too Please make them separate for 48GII & 49G+ where SD is involved. ==== > I am working on the 2nd-order ODE example found on 16-64 of the HP49G+ > User's Manual (the 10MB one). It uses the numerical differential eqn > solver to solve the following: > > x(t)= -18.75x(t)-1.962x'(t) > with x'(0)=6 and x(0)=0. The goal is to find x and x' at t=2. > > I entered it into the solver as shown in the example, but receive the > following error message: NON-ALGEBRAIC IN EXPRESSION. You need to store 't' to VX either using the filer of directly via 't' STOVX OR better yet, do it in a standard way: d1d1Y(X) is Y'' [ d1d2d3F(X,Y,Z) would diff. variable 1, 2, and 3 ] You better aquire the new hp 49g+ manuals from the www.hp.com Veli+Pekka > PLEASE NOTE: that I am not running a G+; rather, it is a 49G using > 1.19-6 (the most recent build for that calculator). ==== Can anyone tell me if programs written for the HP48GX and G+ programs will run directly on the new HP48GII? I have searched the web but have not been able to find an answer to this question. Many thanks Grant Nicklin ==== Only very simple programs will work when switching platforms. If the program made use of and of the following functions and/or features the program will probably not work properly. 1) KEY functions (keys have moved location) 2) Raw serial port (RS-232) operations 3) Programs that use SYSEVAL functions 4) Library functions 5) Flags -2 and -3 have changed some operational results between the platforms 6) Programs that use lots of memory (the GII has less user memory) 7) Programs that utilize the card slot of the GX 8) Programs that still need to communicate (via IR) to G+ units (different protocol) 9) Programs that use formatted numeric output to the screen or printer (decimal point added in all reals in 48GII) 10) Programs that print and use a fixed delay for the 82440B Bob ==== >Only very simple programs will work when switching platforms. If the >program made use of and of the following functions and/or features the >program will probably not work properly. > ... > >Bob I think some fairly complex programs have been written with-in those limitations... Pete M. Wilson Gamewood, Inc. wilsonpm@gamewood.net ==== > I think some fairly complex programs have been written with-in those > limitations... > > Pete M. Wilson > Gamewood, Inc. > wilsonpm@gamewood.net I guess it would depend on your perspective. What some might consider complex, others might consider simple. Bob ==== Generally speaking, yes. There may be a few things that would not work, but I can't think of any off the top of my head. Gene -- * These statements and opinions are mine alone and do not reflect my employer's views. * > Can anyone tell me if programs written for the HP48GX and G+ > programs will run directly on the new HP48GII? I have searched > the web but have not been able to find an answer to this > question. > > Many thanks > Grant Nicklin ==== > Can anyone tell me if programs written for the HP48GX and G+ > programs will run directly on the new HP48GII? I have searched > the web but have not been able to find an answer to this > question. > > Many thanks > Grant Nicklin I am talking baout UserRPL programs here. Grant Nicklin ==== > > > Can anyone tell me if programs written for the HP48GX and G+ > > programs will run directly on the new HP48GII? I have searched > > the web but have not been able to find an answer to this > > question. > > > > Many thanks > > Grant Nicklin > > I am talking baout UserRPL programs here. To ensure functionality do ->STR on the sending end and use RPL/Approx. Mode on the new model before receiving Finally do STR-> (or OBJ->) Veli+Pekka ==== > > > > > Can anyone tell me if programs written for the HP48GX and G+ > > > programs will run directly on the new HP48GII? I have searched > > > the web but have not been able to find an answer to this > > > question. > > > > > > Many thanks > > > Grant Nicklin I've just ordered a 48GII and 49G+ here in Australia and will see how simple UserRPL Surveying programs and libraries work with the new calculators. I suspect that libraries may have to be recompiled, but will see. As soon as I've done that I'll put up some comments on the web that should be of some use to guys like yourself. (p.s. I should stress that I'm a kiwi too ! Must be lots of emotional turmoil over there at the moment after last weekend) Richard Stanaway ==== Having played with the 48's equation library on the 49 I found it quite good but too big so I compressed it. The library is now 40% smaller. I would like to post it on hpcalc.org but I would first like to have James authorisation, address he gives in the documentation is not valid. indicating hp to sort it out from spam. Arnaud ==== The entry point DispStsBound doesn«t display a horizontal line at y=14, normally the separation between header and stack. ==== In <26a67dc4.0311141321.75d3cf2d@posting.google.com> Pedro Liberato > The entry point DispStsBound doesn?t display a horizontal line at > y=14, normally the separation between header and stack. > If I recall correctly this entry point doesn't exist on the 49G ==== > > The entry point DispStsBound doesn't display > > a horizontal line at y=14, normally the > > separation between header and stack. > > If I recall correctly this entry point doesn't > exist on the 49G Actually it does, but it's not very useful on the 49g+ since it draws the line in the stack area instead of where the status line ought to be. To see it in action, run this program: << #2C2F9h SYSEVAL 0. FREEZE >> -Joe- <2d624c12.0311131716.469a9404@posting.google.com> ==== > I would also consider going to slow speed faster, like immediately > after the display refresh if there is no activity on keyboard. > Changing the clock frequency of a CPU has some consequences. There's always a bit of lag between frequency speed and you also need to re- synchronize all external devices: screen, I/O, timers etc... Some weird stuff will happen when you change frequency very often. Like due to the lag some keypresses could be missed and so on.. Everything is always done for a reason. Especially in a device where you can tell they've spent a lot of time working on the CPU frequencies... why bother with a specific wait after a keypress otherwise? ==== > > I would also consider going to slow speed faster, like immediately > > after the display refresh if there is no activity on keyboard. > > > > Changing the clock frequency of a CPU has some consequences. There's > always a bit of lag between frequency speed and you also need to re- > synchronize all external devices: screen, I/O, timers etc... > Some weird stuff will happen when you change frequency very often. Like > due to the lag some keypresses could be missed and so on.. > > Everything is always done for a reason. Especially in a device where you > can tell they've spent a lot of time working on the CPU frequencies... > why bother with a specific wait after a keypress otherwise? OK I was wondering... Thnx for xplanation, Jean+Yves Veli+Pekka PS: Any possibility to still lower the slow mode speed? ==== Now that the 49G+ is in the hands of the user is it worth buying over the 49G or 48GX. What are the 49G users using for support, e.g. ROM updates? Thx. ==== My first -49g+ failed on 10/26. My second -49g+ I received as a replacement, kept it for two weeks to make certain it was good, and gave it to a friend as a gift. My third -49g+ I just opened and, yep, it is defective also. Anybody else out there have a 66 2/3% failure rate? ==== In Jim > My first -49g+ failed on 10/26. > > My second -49g+ I received as a replacement, kept it for two weeks to > make certain it was good, and gave it to a friend as a gift. > > My third -49g+ I just opened and, yep, it is defective also. > > Anybody else out there have a 66 2/3% failure rate? > What do you mean by it failed? There have been issues about dead batteries and high power consumption. Upgrading to ROM 1.22 fixed all problems. ==== : What do you mean by it failed? : : There have been issues about dead batteries and high power consumption. : Upgrading to ROM 1.22 fixed all problems. Yes, all true. However, in my case I put in brand new Duracell Alkaline AAA Ultra cells. The machine blinked on and then off three times. And then it stayed off. No matter how many battery removals/reinsertions I performed, no matter how many reset operations I attempted. This machine was dead. ==== In Jim > Yes, all true. However, in my case I put in brand new Duracell > Alkaline AAA Ultra cells. The machine blinked on and then off three > times. And then it stayed off. No matter how many battery removals/ > reinsertions I performed, no matter how many reset operations I > attempted. This machine was dead. Did you also remove the backup battery? Internal memory could have been seriously corrupted and unable to recover it ==== battery? Internal memory could have been seriously corrupted and unable to recover it. _____________________ Yes. The backup battery was also removed. No change. Still dead. ==== It has been my experience that some batteries do not have a long (or pointed) enough tip at the positive terminal to make contact with the metal fixed in the hp49g+ battery compartment. As a result the batteries would not work or a slight motion or rotation of some batteries was enough to power on momentarily the unit! I fixed the problem for good by inserting narrow copper strips between the battery head and the battery compartment so that electrical contact is secured. !Demeter! ==== : : It has been my experience that some batteries do : not have a long (or pointed) enough tip at the : positive terminal to make contact with the metal : fixed in the hp49g+ battery compartment. As a result : the batteries would not work or a slight motion : or rotation of some batteries was enough to power on : momentarily the unit! I fixed the problem for good by : inserting narrow copper strips between the battery head : and the battery compartment so that electrical contact : is secured. : : !Demeter! I TRIED IT ON THE VERY FIRST -49G+ TO FAIL -- YES IT'S STILL HERE ON MY DESK -- AND IT NOW WORKS PERFECTLY!!!! DEMETER HAS DISCOVERED, I BELIEVE, THE REMEDY TO ALL THESE FAILURES! Interestingly when the store and I tried to power on the number four calculator I insisted that we use a brand-new set of Duracell AAA Ultras. It failed. The store people insisted that we use the Panasonic batteries that came with the unit. It worked. I am nearly certain that what HP is facing, at least on the 49g+s, is a molding tolerance vs. battery + end terminal forming tolerance that does not guarantee that all batteries will fit in all battery slots, applying only to the plus end. I just tried inserting new Duracell AAA Ultras into my first machine. No matter what I did, pushing, pulling, twisting, turning, the machine will not come on using Duracells. I took them out and put in the Panasonic cells, and my first machine resumed working perfectly! The problem is mechanical. Although I couldn't find any copper strips in my ham radio junk box, that is the next thing to try: making a conductive spacer so that the plus terminals of the batteries will always make contact, no matter how shallow they have been formed. I will report tomorrow. ==== > > : > : It has been my experience that some batteries do > : not have a long (or pointed) enough tip at the > : positive terminal to make contact with the metal > : fixed in the hp49g+ battery compartment. As a result > : the batteries would not work or a slight motion > : or rotation of some batteries was enough to power on > : momentarily the unit! I fixed the problem for good by > : inserting narrow copper strips between the battery head > : and the battery compartment so that electrical contact > : is secured. > : > : !Demeter! > > I TRIED IT ON THE VERY FIRST -49G+ TO FAIL -- YES IT'S STILL HERE ON MY > DESK -- AND IT NOW WORKS PERFECTLY!!!! > > DEMETER HAS DISCOVERED, I BELIEVE, THE REMEDY TO ALL THESE FAILURES! > > Interestingly when the store and I tried to power on the number four > calculator I insisted that we use a brand-new set of Duracell AAA Ultras. It > failed. The store people insisted that we use the Panasonic batteries that > came with the unit. It worked. > > I am nearly certain that what HP is facing, at least on the 49g+s, is a > molding tolerance vs. battery + end terminal forming tolerance that does not > guarantee that all batteries will fit in all battery slots, applying only to > the plus end. > > I just tried inserting new Duracell AAA Ultras into my first machine. No > matter what I did, pushing, pulling, twisting, turning, the machine will not > come on using Duracells. I took them out and put in the Panasonic cells, and > my first machine resumed working perfectly! > > The problem is mechanical. Although I couldn't find any copper strips in my > ham radio junk box, that is the next thing to try: making a conductive > spacer so that the plus terminals of the batteries will always make contact, > no matter how shallow they have been formed. > > I will report tomorrow. Duracell's are working fine in my 49G+. I popped them in and first try it turned on and has been working ever since. Travis ==== I can confirm this battery compartment issue. After my Panasonic batteries started dying, I switched to Duracell Ultras. Guess what... no power... totally dead. I did a little testing and found that only the top battery slot was to blame (at least for my G+). I had no probelm with Duracells in the bottom two slots, so long as there was a Panasonic battery in the top slot. After playing with the flat wire spiral on the right side of that top slot in the calculator--using a safety pin, in effect trying to pry the wire spiral out a bit--I got my calculator to work just fine with Duracells in all three slots. I'm actually surprised that so much manufacturing tolerance can exist in batteries. I would be curious to see how battery specifications read. Is anyone here in the battery industry, and can you quote us the specs? I found one specification on-line that calls for the overall length of a AAA battery to be 44.2 +/- 0.3 mm, and the barrel length to be 42.6 +/- 0.3 mm. That suggests that the button should be 1.6 mm nominally; however, if the overall battery length is short (at spec limit) and the barrel length is long (at spec limit), then the button could be just 1 mm tall. Conversely, in the opposite circumstance, the button could be as long as 2.2 mm. That's over twice as long! I sure hope that the specification I saw is not what manufacturers really go by for QC. Chris > > > > : > > : It has been my experience that some batteries do > > : not have a long (or pointed) enough tip at the > > : positive terminal to make contact with the metal > > : fixed in the hp49g+ battery compartment. As a result > > : the batteries would not work or a slight motion > > : or rotation of some batteries was enough to power on > > : momentarily the unit! I fixed the problem for good by > > : inserting narrow copper strips between the battery head > > : and the battery compartment so that electrical contact > > : is secured. > > : > > : !Demeter! > > > > I TRIED IT ON THE VERY FIRST -49G+ TO FAIL -- YES IT'S STILL HERE ON MY > > DESK -- AND IT NOW WORKS PERFECTLY!!!! > > > > DEMETER HAS DISCOVERED, I BELIEVE, THE REMEDY TO ALL THESE FAILURES! > > >> ...snip... > > > > I will report tomorrow. > > > Duracell's are working fine in my 49G+. I popped them in and first > try it turned on and has been working ever since. Travis ==== Yep! Chris, you definitely are the man! I reformed the right-side flat spiral to bring the center of it out just the tiniest amount, and now any kind of battery can be used. A W O R D O F CAUTION If you want to try this, be sure not to reform the center of the flat spiral very much. I am nearly certain that the battery compartment's + power lead comes from the center of the spiral rather than the edge. If you break it, the machine will probably be very seriously dead. : I can confirm this battery compartment issue. After my Panasonic batteries : started dying, I switched to Duracell Ultras. : : Guess what... no power... totally dead. : : I did a little testing and found that only the top battery slot was to blame : (at least for my G+). I had no probelm with Duracells in the bottom two : slots, so long as there was a Panasonic battery in the top slot. : : After playing with the flat wire spiral on the right side of that top slot : in the calculator--using a safety pin, in effect trying to pry the wire : spiral out a bit--I got my calculator to work just fine with Duracells in : all three slots. ==== I had the same problem with the batteries (Energizer). I checked the voltage when the batteries were in the HP. My voltmeter said 4,5 V; That's ok. But the HP was dead. After some trials I cut off a bit of the plastic cover at the plus-end of the batterie - an that worked. I tried other combinations, it was always the same. When I cut off the plastic the HP worked. So I hadn't to bow the contact. Greetings Stefan ==== I had the same problem with the batteries (Energizer). I checked the voltage when the batteries were in the HP. My voltmeter said 4,5 V; That's ok. But the HP was dead. After some trials I cut off a bit of the plastic cover at the plus-end of the batterie - an that worked. I tried other combinations, it was always the same. When I cut off the plastic the HP worked. So I hadn't to bow the contact. Greetings Stefan ==== No kidding!!! I have to pull my 49G+ out of the drawer (currently focusing on 32SII and 48G+ for some work at hand) to see this for myself, that there may be a noncontact area on in the battery compartment! > > It has been my experience that some batteries do > not have a long (or pointed) enough tip at the > positive terminal to make contact with the metal > fixed in the hp49g+ battery compartment. As a result > the batteries would not work or a slight motion > or rotation of some batteries was enough to power on > momentarily the unit! I fixed the problem for good by > inserting narrow copper strips between the battery head > and the battery compartment so that electrical contact > is secured. > > !Demeter! > ==== memory on a 49g+ (system level)? in this NG. And if you or someone else can, please explain what exactly the reset (paper clip hole @ the back) button does? IMHO it doesn't wipe-out what's in port 2 (flash) so it's useless in many occasions. > In Jim > > Yes, all true. However, in my case I put in brand new Duracell > > Alkaline AAA Ultra cells. The machine blinked on and then off three > > times. And then it stayed off. No matter how many battery removals/ > > reinsertions I performed, no matter how many reset operations I > > attempted. This machine was dead. > > Did you also remove the backup battery? > > Internal memory could have been seriously corrupted and unable to > recover it > ==== By any chance did you get it from the same retailer? Or were their serial numbers close together? > My first -49g+ failed on 10/26. > > My second -49g+ I received as a replacement, kept it for two weeks to make > certain it was good, and gave it to a friend as a gift. > > My third -49g+ I just opened and, yep, it is defective also. > > Anybody else out there have a 66 2/3% failure rate? > > > ==== : By any chance did you get it from the same retailer? Or were their : serial numbers close together? Nope, different retailers. It looks as if the fourth one--I just got back from swapping out the defective third one--is going to be OK. Interestingly, though, we put batteries in it at the store, and it would not power on either (#4!). I used the borrowed paper clip reset tool and attempted to reset the machine. No luck. As a last resort, we took out the batteries, for about 30 seconds or so, put them back in, and it now is working. I am taking bets on how long it will live. It has a very early second production run serial number: CN33400845. Next to my first one, #4 here has the earliest serial number. ==== ... It has a very early second production run > serial number: CN33400845. Next to my first one, #4 here has the earliest > serial number. > You're at 50% success rate now. I have the same 50% rate with 2 early units ... with a happy ending. After trying it for a week, I exchanged my first 49+ at the local college bookstore where I bought it. It exhibited the missed-keypress behavior discussed here (on the + and +/- keys). We tried out the second in the store and it was fine. Both had CN334... serial numbers very close the the number of your fourth machine. Neither machine had battery problems. I updated both to the latest rom using usb on the first night that I had each. I am extremely impressed with the 49+. I like the light-weight but solid feel, the legibility of the screen and keyboard, and most of all the speed and responsiveness. I'm hopeful the hp calculator team will have soon resolved the quality control issues apparent in early production runs. After all, I had a 50% rate on my first two HP41C calculators years ago. I had to exchange the first due to a key problem (no tactile click on one of the keys) at the very same bookstore. Tom ==== I am sitting at my computer at this moment with Programming in SystemRPL Eduardo Kalinowski's book online, trying to understand informs and message handlers. Message Handlers have me very excited as I am a Delphi programmer and know all about events and handling them. I am convinced that if I am to write quality programs on my 49G+ I must master informs and choose boxes and ALL their intricacies in SystemRPL. Yet, I do not get the entire picture reading these book. They tease you and then leave you hanging. For example. On page 276 it says and I quote 35.4 Message Handlers. As with other input/output applications of the HP49, input forms use message handlers to allow the programmer to have more control over the input form. There is one message handler for each field, and one for the input form itself My questions on the above lines are:- As with which other input/output applications?. Which other applications for the 49G use message handlers? Immediately following, it again says on the same page, As with other message handlers, the program you provide is called with a message number (a bint) in level one, and sometimes other parameters I know what a bint is but could someone explain to me, if anyone cares to, exactly what is meant here? The program you provide! How exactly do I provide that program. Do I put it's name as a program object on the stack where the message handler must be put or do I between the program's << >> write out the entire program and place it in the message handler's place on the stack? Or ??? Exactly how do I provide the program? and put the bint and other parameters on the stack (yes, I do know how to put things on the stack normally, but remember I am putting information for an input form) how do I put these things on the stack. Could someone explain? The program is called with a message number (a bint) in level one and sometimes other parameters How exactly do I call the program? Sometimes these little bits of information could make all the difference between one understanding fully or being left hanging. And right now, I am almost passing out. As I said in my opening, I have reognized (for my part) that Informs and Choose boxes in System RPL are the gateway for spreading one's wings as it were in programming the Hp49G+. They are like forms in a program like Delphi. Could someone remove the cobwebs from my eyes and let me really smell this Inform coffee or point me in the direction of a definitive source of information on these Choice subjects. Earl Cenac ==== I can't seem to get the 49.bin to convert User rpl. in the emu. how exactly is this done? I am new to this whole thing as I always had it easy with 48gx ==== In order to better understand the new permanent header part and to try to program it, I started to study entry points that use to deal with the header. I started with DispStsBound that used to draw the line between the header and the stack. Unfortunately the function seems to be no longer valid, as mentioned in a previous post, because it does not draw the line at 15 rows from the real top of the screen. I then had a look at FlashMsg that displays a message in the header part of the screen. Fortunately, this function has been modified to deal with the new screen size and here what I observed : 1) Second, there is an entry point to check for screen size (kind of CheckForLargeScreen) that returns True/False. That could indicate that the same code is present in both the 49G+ and the 48GII. The entry point in ROM 1.22 is PTR 2F3C1 ; 2) FlashMsg uses a function that seems to extract the grob from the new header area with PTR 2F3CF and, at the end, a function to put a grob in the new header area with PTR 2F3D0. Both function uses the CheckForLargeScreen function ; 3) DISPSTATUS2, used in FlashMsg has been adapted to handle the new screen and also uses the CheckForLargeScreen ; 4) Two FPTRs, FPTR 2 C8 and 2 C9 are used to write text in first and second line of the header (rows 1-8 and rows 8-15 of the new screen) ; 5) As mentioned in a previous post, a lot of the functions that deal with writing to the new screen area use BUSCC (BUSCx) instructions, like in the PTR 6C251 code (that seems to replace HARDBUFF_X_Y) ; But, the new Clr8 code, doesn't the BUSCx instruction and looks like : GOSBVL SAVPTR D1= 7E014 LC 00110 GOSBVL =WIPEOUT GOSBVL =GETPTRLOOP So there is some hope to be able to develop libraries that can use the full screen of the new calc. That's it for the moment ... Alain ==== > I then had a look at FlashMsg that displays a message in the header part > of the screen. Fortunately, this function has been modified to deal with > the new screen size and here what I observed : IMHO, the classical FlashMsg command flashes a message definitely too short in the new Header area. Therefore, I modified it to the command Flash in the mini-library Msgman which can be found below. And this very useful command is now finally available also to the normal user. - Wolfgang http://page.mi.fu-berlin.de/~raut/WR49/Msgman.htm ==== > So there is some hope to be able to develop libraries > that can use the full screen of the new calc. Extending *graphing* to the full screen (as falsely advertised in the photos from HP) would be great. Wolfgang's OT49+ and Keyman+ libraries have been using the 49g+ status area since they were released. He mostly seems to use DISPSTATUS2, but there are also some texts overlaid onto certain sections of the status area in a small font, and that's pretty cool. -Joe- ==== > Wolfgang's OT49+ and Keyman+ libraries have been using the 49g+ status > area since they were released. He mostly seems to use DISPSTATUS2, > but there are also some texts overlaid onto certain sections of the > status area in a small font, and that's pretty cool. Good news for those who wish to use the new header on the HP49G+. Everybody can do it now, provided he uses the small library Msgman below (only 736 bytes). Msgman.htm contains an elaborated example for writing something in minifont in the new header. The main concern of Msgman is to make the built-in message management (normally applied only by SysRPL programmers) available to the normal user who is not familiar with SysRPL. Perhaps most useful is the message searcher which displays all built-in messages containing a given word or word fragment in a multi-selective browser. No need anymore to look up huge message tables on hpcalc.org which are incomplete anyway... Using also graphics on the full 49+ screen will not be realized in Msgman because this is an entirely different task. - Wolfgang http://page.mi.fu-berlin.de/~raut/WR49/Msgman.htm PS. As soon as a new extable for the HP49+ will be released some nacked pointers occurring in the Msgman rompointers will hopefully appear with nice new SysRPL names :-) ==== Wolfgang... YOU ROCK!!! Your lib is now implemented into my surveying programs as a very nice little indicator of the name of my current data file. John Evers ==== > Good news for those who wish to use the new header on the HP49G+. > Everybody can do it now, provided he uses the small library Msgman below > (only 736 bytes). Msgman.htm contains an elaborated example for writing > something in minifont in the new header. Sorry, for a few hours a wrong file was on the site. Who downloaded Msgman.zip should repeat it. Msgman has actually 738 bytes, CRC 1655h. - Wolfgang http://page.mi.fu-berlin.de/~raut/WR49/Msgman.htm ==== I have noticed no reply to your post from more experienced users, so I will attempt to help... ;-) Regarding the digit seperators, no you cannot have them in standard mode - only in fixed. I too wish it were an option in all the different numeric display modes. However, you should be able to change the fraction mark setting using the FM, softkey in the Modes menu on your 48G. On my 48SX, it is in Left-Shift MODES NXT NXT NXT - not sure, but your 48G might be slightly different. As far as memory, or register storage goes, I understand exactly what you are saying. I came from an 11C when I got my 48SX and also missed the easy STO 1 type of sequence of the older calculators. Many years ago now, I created 3 small programs for my 48SX (and now also use them with my 49g+) which help with this - STOx, RCLx and EVALx. I have these in the first page of my Custom (CST) menu so it is only 2 keystrokes to access. This how you use STOx: With the object to be stored (number, program, list etc.) on stack level 1, press the STOx softkey and a prompt will appear asking you to press any key on the keyboard. Press any key and the object is then stored into that key but also still remains on stack level 1 - just like on my old 11C. This is what actually happens inside my STOx program: 1) The current directory path is saved (using PATH). 2) The path is switched to my {HOME PREFS} directory. I created this PREFS directory to store many miscellaneous variables and preferences in my programs that I never need direct (menu softkey) access to. 3) The prompt is displayed and the program then logs the next keystroke. Let's say you press the MTH key near the top left of your keyboard - the program then logs this keypress as 21.1. 4) The program then assembles this keypress into a variable name - in this example it would create the variable 'K21' in {HOME PREFS} and stores the level 1 object in it. 5) The directory is then switched back to the one saved at the beginning and the last menu is restored using 0 MENU. So, now you have a total of 45 possible direct-access memory locations, and no cluttering of your current directory with a bunch of temporary variables that need purging later! Also, you can store and recall the exact same registers, no matter what directory you happen to be in currently. The way I have these programs set up, they only require 3 keystrokes... eg: CST STOx [key]. As well, since the current directory and menu is saved, you don't lose context of where you were and what you were doing, so to speak, and won't need any additional keystrokes to bring your last menu back etc. For comparison, the first 10 registers on my 11c only require 2 keys to access, but the second 10 also require 3 keystrokes. Your 32S has 26 direct access registers that require 2 keystrokes (I think? - never used one myself) so this 3-keystroke solution for the 48/49 calcs is not too bad, is it? Only the alpha, left and right-shift and ON keys won't work directly. I could have left the shift indicator decimal point (.1 or .2 etc.) on the keypress number to gain many more registers but I thought 45 were plenty. Frankly, I have a hard enough time keeping track of far less than 45... Finally, RCLx simply recalls the value - opposite of STOx. EVALx will evaluate the register - run the program for example if a program object is stored instead of a number. Obviously, these registers would be somewhat awkward to use in a program (and have poor non-descriptive names!) but I did not create STOx/RCLx for that. I use STOx/RCLx strictly when I am performing manual calculations and want a quick and easy way to save and recall objects (usually numbers) to/from the stack. It might give you some useful UserRPL experience on your new 48G if (and think STOx, RCLx and EVALx might be useful to you) let me know and I will post them. (I am assuming you are new to the 48 series calculators) There may already be some other ready-made utilities for this type of thing on hpcalc.org - I have never checked. I created these programs long before I ever had Internet access... ;-) BTW, hopefully there are no SysRPL commands already called STOx/RCLx/EVALx that I might be confusing people with these UserRPL program names! Mike Mander > > I have just replaced my old HP32S with an HP48G. Now, after a little testing > and typing and reading the manual I have 2 questions about the HP48G. > > My HP32S always used digit separators. I could choose between , and . as > fraction mark: > 3.800.000,5 or 3,800,000.5 > The HP48G uses digit separators only in the fixed mode but not in standard > or scientific mode: > 3800000,5 or 3800000.5 > the use of digit separators helps me to see the number with one view without > counting digits ;-) > Is ist possible to change this ? > > To put a number in memory was more easy with the HP32S. I pressed STO + > A...Z and the number was stored. Restore was as easy as this, i have just > pressed RCL + A...Z. This is more complicated on the HP48G. Can this be > changed ? > > Many thanks for your answer > > Christoph Deppe > dc1ecd@darc.de ==== Sorry, not sure why this post didn't get attached to the original thread... Mike ==== Manel. I am using the version 1.19-7 on my old hp49G. OOOPS! Excuse me. THIS ROM VERSION IS NOT AVAILABLE. What is happening with 1.19-7 or 1.2X for OLD HP49G. Any information about this?. No body speak about it! ------------------------------------------------------ The BUG? The version I am using is 1.19-6. If you enter in RPN mode 4: [[1 2 3][6 5 3][3 5 3]] 3: [1 2 5] 2: [7 3 4] 1: 3 ROW-> (or COL->) HP RETURNS: COL-> Error - n.bc 514 Argumento Incorrecto <<--- this is OK 5: [5 2 5] 4: [7 3 4] 3: [1 2 5] 2: (character 164) 2h <<--- what is that? 1: 3 I think the algorith does not check the arguments at the begining. Goooood-bye! ==== > Manel. > > I am using the version 1.19-7 on my old hp49G. > > OOOPS! Excuse me. THIS ROM VERSION IS NOT AVAILABLE. > What is happening with 1.19-7 or 1.2X for OLD HP49G. > > Any information about this?. No body speak about it! > ------------------------------------------------------ > > > The BUG? > The version I am using is 1.19-6. > > If you enter in RPN mode > > 4: [[1 2 3][6 5 3][3 5 3]] > 3: [1 2 5] > 2: [7 3 4] > 1: 3 > ROW-> (or COL->) > > HP RETURNS: > > COL-> Error - n.bc 514 > Argumento Incorrecto <<--- this is OK > 5: [5 2 5] > 4: [7 3 4] > 3: [1 2 5] > 2: (character 164) 2h <<--- what is that? > 1: 3 > > I think the algorith does not check the arguments at the begining. > > > Goooood-bye! Most certainly a bug.. probably introduced in the extra code for the new symbolic matrix type. The 2h is probably the number of dimensions of the array in level 4 - the arguments should all have but one dimension for COL-> and ROW-> Werner ==== > > Manel. > > > > I am using the version 1.19-7 on my old hp49G. > > > > OOOPS! Excuse me. THIS ROM VERSION IS NOT AVAILABLE. > > What is happening with 1.19-7 or 1.2X for OLD HP49G. > > > > Any information about this?. No body speak about it! > > ------------------------------------------------------ > > > > > > The BUG? > > The version I am using is 1.19-6. > > > > If you enter in RPN mode > > > > 4: [[1 2 3][6 5 3][3 5 3]] > > 3: [1 2 5] > > 2: [7 3 4] > > 1: 3 > > ROW-> (or COL->) > > > > HP RETURNS: > > > > COL-> Error - n.bc 514 > > Argumento Incorrecto <<--- this is OK > > 5: [5 2 5] > > 4: [7 3 4] > > 3: [1 2 5] > > 2: (character 164) 2h <<--- what is that? > > 1: 3 > > > > I think the algorith does not check the arguments at the begining. > > > > > > Goooood-bye! > > Most certainly a bug.. probably introduced in the extra code for the > new symbolic matrix type. > The 2h is probably the number of dimensions of the array in level 4 - No, it's the index to that level on stack. Add one more column and you see! This error seems to occur only ahen a square matrix is the top most argment. Could not read the SysRPL good enough to find the problem. > the arguments should all have but one dimension for COL-> and ROW-> > > Werner ==== ha scritto nel messaggio > Manel. > > I am using the version 1.19-7 on my old hp49G. > > OOOPS! Excuse me. THIS ROM VERSION IS NOT AVAILABLE. > What is happening with 1.19-7 or 1.2X for OLD HP49G. > > Any information about this?. No body speak about it! > ------------------------------------------------------ Well, with the new 49+ coming and all the problems people are reporting I think the hp efforts are now on the new machines! > If you enter in RPN mode > > 4: [[1 2 3][6 5 3][3 5 3]] > 3: [1 2 5] > 2: [7 3 4] > 1: 3 > ROW-> (or COL->) > > HP RETURNS: > > COL-> Error - n.bc 514 > Argumento Incorrecto <<--- this is OK > 5: [5 2 5] > 4: [7 3 4] > 3: [1 2 5] > 2: (character 164) 2h <<--- what is that? > 1: 3 My stack is: 4: [7 3 4] 3: [1 2 5] 2: (character 164) 2h <<--- what is that? 1: 3 NOTE: nothing on stack level 5... Well, I think it's a bug... someone knows if it is already reported? ==== For example: <<@ userRPL program gathers input from user -> local var << @ simple math crunches user's input with results to stack @ usrRPL program does: 0 WAIT (I want to view results) @ usrRPL program returns to calling program >> >> Unless I use a HALT or stop program execution, the display does not update to show stack contents. Is there a method to force the display to update? I've tried: TEXT, 7 FREEZE, -1 WAIT, 0 WAIT and a few other things all of which did not do the trick. I would like to stay within usrRPL command set, it possible. Can anyone help please? ==== > For example: > > <<@ userRPL program gathers input from user -> local var > > << @ simple math crunches user's input with results to stack > @ usrRPL program does: 0 WAIT (I want to view results) > @ usrRPL program returns to calling program > >> > > >> > > Unless I use a HALT or stop program execution, the display does not > update to show stack contents. Is there a method to force the display > to update? I've tried: TEXT, 7 FREEZE, -1 WAIT, 0 WAIT and a few > other things all of which did not do the trick. I would like to stay > within usrRPL command set, it possible. Can anyone help please? Why not use {CONT} TMENU Press CONT PROMPT ==== approach, but this is just as well. I hated to ask, but nothing I tried and none of my research on the popular net sites revealed anything useful for this problem. >> For example: >> >> <<@ userRPL program gathers input from user -> local var >> >> << @ simple math crunches user's input with results to stack >> @ usrRPL program does: 0 WAIT (I want to view results) >> @ usrRPL program returns to calling program >> >> >> >> >> >> >> Unless I use a HALT or stop program execution, the display does not >> update to show stack contents. Is there a method to force the display >> to update? I've tried: TEXT, 7 FREEZE, -1 WAIT, 0 WAIT and a few >> other things all of which did not do the trick. I would like to stay >> within usrRPL command set, it possible. Can anyone help please? >Why not use >{CONT} TMENU Press CONT PROMPT > ==== There is more about this new behaviour of some display commands. Suppose you like to see what's left in the stack after a program finishes while on top rows of the display result still remains shown. For example I like to see angle (as proper DMS-format string) and distance on the display (for check what entered is correct) and the pair of resulting coordinates in the stack for future use. That works on the 48, only frozen part of the display is not regenerated. Unlike the 49g+. I know there is workaround but why should use it? Or that's yet another bug? > approach, but this is just as well. I hated to ask, but nothing I > tried and none of my research on the popular net sites revealed > anything useful for this problem. > > > > >> For example: > >> > >> <<@ userRPL program gathers input from user -> local var > >> > >> << @ simple math crunches user's input with results to stack > >> @ usrRPL program does: 0 WAIT (I want to view results) > >> @ usrRPL program returns to calling program > >> >> > >> > >> >> > >> > >> Unless I use a HALT or stop program execution, the display does not > >> update to show stack contents. Is there a method to force the display > >> to update? I've tried: TEXT, 7 FREEZE, -1 WAIT, 0 WAIT and a few > >> other things all of which did not do the trick. I would like to stay > >> within usrRPL command set, it possible. Can anyone help please? > >Why not use > >{CONT} TMENU Press CONT PROMPT > > > ==== > There is more about this new behaviour of some display commands. > Suppose you like to see what's left in the stack after a program finishes > while on top rows of the display result still remains shown. For example I > like to see angle (as proper DMS-format string) and distance on the display > (for check what entered is correct) and the pair of resulting coordinates in > the stack for future use. That works on the 48, only frozen part of the > display is not regenerated. Unlike the 49g+. I know there is workaround but > why should use it? Or that's yet another bug? Give the exact example so that I understand you (I read RPL much easier than English) ==== ok, Veli+Pekka, check that on the 48: 'Vinput' %%HP: T(3)A(D)F(.); << CYLIN SWAP HMS-> ->V2 DUP mkvstr 1 DISP 1 FREEZE @ 2 DISP for 49g+ RECT 3 FIX >> 'mkvstr' %%HP: T(3)A(D)F(.); << DEG CYLIN V-> 360 MOD ->HMS 4 RND 4 FIX HMS-> ->HMS ->STR DUP . POS SWAP OVER ^o REPL DUP2 1 ROT 2 + SUB ' + SWAP ROT 3 + 8 SUB 34 CHR + + SWAP 3 FIX 3 RND ->STR SWAP + + + RECT >> cheers > > There is more about this new behaviour of some display commands. > > Suppose you like to see what's left in the stack after a program finishes > > while on top rows of the display result still remains shown. For example I > > like to see angle (as proper DMS-format string) and distance on the > display > > (for check what entered is correct) and the pair of resulting coordinates > in > > the stack for future use. That works on the 48, only frozen part of the > > display is not regenerated. Unlike the 49g+. I know there is workaround > but > > why should use it? Or that's yet another bug? > > Give the exact example so that I understand you > (I read RPL much easier than English) > > ==== I have a simple subroutine that fails when RAND is executed. RAND is to generate a random number, but does not. The program fails due to lack of arguments. This is the routine: <> RAND executes from the soft key menu, but not from within a program. What am I doing wrong, or is this a bug? RANGE and LOW are user variables. Scott ==== > I have a simple subroutine that fails when RAND is executed. RAND is to > generate a random number, but does not. The program fails due to lack of > arguments. > > This is the routine: > > <> > > RAND executes from the soft key menu, but not from within a program. What am > I doing wrong, or is this a bug? RANGE and LOW are user variables. You may have same value for RANGE and LOW The above code works correctly, check your data using single step ==== > > I have a simple subroutine that fails when RAND is executed. RAND is to > > generate a random number, but does not. The program fails due to lack of > > arguments. > > > > This is the routine: > > > > <> > > > > RAND executes from the soft key menu, but not from within a program. What > am > > I doing wrong, or is this a bug? RANGE and LOW are user variables. > > You may have same value for RANGE and LOW > The above code works correctly, check your data using single step > > Weirdness. Maybe because I transfered the file from a 48GX. Finally I got the RAND function to work by evaluating {RAND} enterd from the soft key menu. It worked like clockwork until I executed the stored code. I again evaluated {RAND} and entered the code by hand and sored it to its variable. Now it appears to be working......strange. RSC ==== X > Weirdness. Maybe because I transfered the file from a 48GX. Finally I got Ha-haa! There it is! Somehow you did not get a valid code. Can you translate it again? Could you compare with tools from 256 MENU 49g+ RAND examined by {RAND} HEAD DUP 256 MENU ->H hexacode = D9D200B262DA903B2130 H-> gives: External External ->S2 gives :: CK0 %RAN ; @ Reverse hexacode = 0312B 309AD 262B0 02D9D AND here it ends - you go on with your studies of the phenomen [I think I'm 2nd time on wrong tracks here] > the RAND function to work by evaluating {RAND} enterd from the soft key > menu. It worked like clockwork until I executed the stored code. I again > evaluated {RAND} and entered the code by hand and sored it to its variable. > > Now it appears to be working......strange. > > RSC ==== > 49g+ RAND examined by > {RAND} HEAD DUP 256 MENU ->H > hexacode = D9D200B262DA903B2130 > H-> gives: External External > ->S2 gives > :: > CK0 > %RAN > ; > @ I got ::PTR 262b0 PTR 309AD; I presume that the G+ doesn't have extable 'in there' and so doesn't decode the pointer names? I also guess that the 'usual' 49G version of extable will do the required job with the G+? JasonG PS: My G+ is working perfectly so far; batteries lasting fine, keyboard working well, no flickering and no paint falling off! ==== > > > 49g+ RAND examined by > > {RAND} HEAD DUP 256 MENU ->H > > hexacode = D9D200B262DA903B2130 > > H-> gives: External External > > ->S2 gives > > :: > > CK0 > > %RAN > > ; > > @ > > I got ::PTR 262b0 PTR 309AD; > > I presume that the G+ doesn't have extable 'in there' and so doesn't >>>> Oops! One of the first thing s to do is to put that big extable2 there first then other tools - sorted according to their sizes. > decode the pointer names? I also guess that the 'usual' 49G version of > extable will do the required job with the G+? >>>> Only WR, who is a die-hard SysRPL coder has founf some missing entry points otherwise it's the most compatible upgrade ever. > > JasonG > > PS: My G+ is working perfectly so far; batteries lasting fine, keyboard > working well, no flickering and no paint falling off! >>>> How about USB connection? ROM updates from flash? ==== > X > > Weirdness. Maybe because I transfered the file from a 48GX. Finally I got > Ha-haa! There it is! > Somehow you did not get a valid code. > Can you translate it again? Could you compare with tools from 256 MENU > 49g+ RAND examined by > {RAND} HEAD DUP 256 MENU ->H > hexacode = D9D200B262DA903B2130 > H-> gives: External External > ->S2 gives > :: > CK0 > %RAN > ; > @ > Reverse hexacode = 0312B 309AD 262B0 02D9D > AND > here it ends - you go on with your studies of the phenomen > [I think I'm 2nd time on wrong tracks here] > > > the RAND function to work by evaluating {RAND} enterd from the soft key > > menu. It worked like clockwork until I executed the stored code. I again > > evaluated {RAND} and entered the code by hand and sored it to its > variable. > > > > Now it appears to be working......strange. > > > > RSC > > variable (code) seems to have solved it. RSC R=Robert ==== > > I have a simple subroutine that fails when RAND is executed. RAND is to > > generate a random number, but does not. The program fails due to lack of > > arguments. > > > > This is the routine: > > > > <> > > > > RAND executes from the soft key menu, but not from within a program. What > am > > I doing wrong, or is this a bug? RANGE and LOW are user variables. > > You may have same value for RANGE and LOW > The above code works correctly, check your data using single step > > Might be a defect, or might reset solve it? Single step with RANGE=100 and LOW =1 is: 1. 100 2. 1 3. 99 4. 1 5. 100 6. 100 (with RAND at top) 7. Too Few Arguments. with <> in the stack single step yields: 1. Nothing in stack (with RAND at top). :-( RSC ==== After changing dead batteries my 42S will not turn on. Are there any remedies for this,i.e. contact cleaner sprayed next to the key, or do I have a paper weight? thx ==== > After changing dead batteries my 42S will not turn on. Are there any > remedies for this,i.e. contact cleaner sprayed next to the key, or do I have > a paper weight? thx > Try the forum at www.hpmuseum.com There are a bunch of people who know quite a lot about older models Arnaud ==== > After changing dead batteries my 42S will not turn on. Are there any > remedies for this,i.e. contact cleaner sprayed next to the key, or do I have > a paper weight? thx I assume your not dealing with a battery corrosion problem... It's rare, but not totally un-heard of to get a bad 'new' battery. Can you check them with a volt meter? I'd do that first... or pull a set known of good batteries out of a working calculator for a quick test of your 42S... I wouldn't spray contact cleaner on my keyboard. Good Luck! Erik ==== There is a guy who should be able to service the 42S. The website is: www.fixthatcalc.com in case you aren't going to continue to go at it by yourself. > After changing dead batteries my 42S will not turn on. Are there any > remedies for this,i.e. contact cleaner sprayed next to the key, or do I have > a paper weight? thx > > > > > ==== This is a thread to note connectivity success and problems with the HP 49+ and Apple Mac computers. (A very small portion of the HP 49+ user group community.) If you are such a user, would you mind leaving a post to this thread of your experience to date. It would be helpful to list what Mac (eg eMac, iMac, powerbook, ibook) and OS (eg 9.2.2, OSX10.3) and connectivity software you are using, along with HP 49+ ROM version, etc. I am hoping this thread will build over time as a handy reference for other Mac users on this NG. Cameron Downunder. ==== [...] James> But I don't see that the 48gII will have any worthwhile James> advantage over a 48GX or 49G, and may in fact be a James> downgrade from the calculators that I already own, so James> that'll be somebody else's concern. I would have to agree. With the reviews of poor construction and documentation of the newer HP's coming out here daily, I'll stick with a 48GX. Maybe a silly side question, but what on earth do ya'll need more than 128K for anyway? ==== I just got my HP49G+ and I upgraded the ROM to 1.22 but now I just can not figure out how to transfer files from my PC to the 49G+. I have been searching thru all the menus and I have not found a window to brose my hard disk, select a file form there and send it to the HP49G+. Does anyone know how to do that? Dante, ==== > > I just got my HP49G+ and I upgraded the ROM to 1.22 but now I just can > not figure out how to transfer files from my PC to the 49G+. I have > been searching thru all the menus and I have not found a window to > brose my hard disk, select a file form there and send it to the > HP49G+. > Does anyone know how to do that? Just start windows explorer, and drag'n'drop the files directly into the pane on the conn4x software. I personally prefer the way the old kit worked. cheers, Al > > > Dante, ==== > > > > I just got my HP49G+ and I upgraded the ROM to 1.22 but now I just can > > not figure out how to transfer files from my PC to the 49G+. I have > > been searching thru all the menus and I have not found a window to > > brose my hard disk, select a file form there and send it to the > > HP49G+. > > Does anyone know how to do that? > > Just start windows explorer, and drag'n'drop the files directly into the > pane on the conn4x software. I personally prefer the way the old kit > worked. BUT I more often once used the old X-modem kit (www.hpcalc.org) which also worked that way - used windows window from PC side There are much more versatile search etc tools on the Windows explorer than on any connectivity kit and you can have many windows open for different types of files or different sources (I am quite organized) I prefer the conn4x way So, I bought a Casio fx-115MS algebraic calculator as a beater. I went to Fry's (where I tried to convince a man to buy his son an HP49G+ instead of a TI, by the way) and picked the calculator that had the most features that I wanted. Pros: Small, true algebraic, and only US$17! Cons: Wont let use e^x on complex numbers, wont return complex results for arcsin, no RPN, and not programmable. Well, on the 32SII (and hence the 33S) you can use e^x on complex numbers, but you can't get a complex result from arcsin. Plus, you have RPN and programmability. The BIG question is this: Are the features of the HP33S (other than RPN) worth another US$43? More than three times the price, and you don't get that much more. Sure, I really like programmability, but at that price I just don't think that enough people will buy them to make the bean counters at HP happy. Well, let's wait and see, I guess. -Joshua -- -Joshua Belsky jjbelsky@yahoo.com http://belsky.net ==== Here are some of the deficiencies with the CASIO model: Calculate 1 divided by 9 and then multiply the result by 9. Subtract 1 and you get 0. This is on a, presumably, BCD system. This strikes me as very peculiar. In general the accuracy does not come anywhere near matching the accuracy of the current line of HP calculators. Register memory storage for statistical results is independent of register storage for mathematical operations on the 32/33 series. This is not true of the CASIO. The exponential range is significantly smaller on the CASIO. The number of memory storage registers is smaller on the CASIO. CASIO also has a peculiar algebraic bug. It can handle 2(3 + 4), but has problems with (3 + 4)2. I'm suspicious that there are other similar bugs in their S-VPAM truly algebraic interpreter. Feature for feature, the CASIO may match up, but I don't think they match up in the details. Unfortunately, right now, there is nothing even close to the 32SII available in the market, RPN or Algebraic. I would pay the price differential for the 33S, but I would have to be convinced of its reliability and durability. > So, I bought a Casio fx-115MS algebraic calculator as a beater. I went to > Fry's (where I tried to convince a man to buy his son an HP49G+ instead of > a TI, by the way) and picked the calculator that had the most features that > I wanted. > > Pros: > Small, true algebraic, and only US$17! > > Cons: > Wont let use e^x on complex numbers, wont return complex results for arcsin, > no RPN, and not programmable. > > Well, on the 32SII (and hence the 33S) you can use e^x on complex numbers, > but you can't get a complex result from arcsin. Plus, you have RPN and > programmability. > > The BIG question is this: > Are the features of the HP33S (other than RPN) worth another US$43? > > More than three times the price, and you don't get that much more. Sure, > I really like programmability, but at that price I just don't think that > enough people will buy them to make the bean counters at HP happy. > > Well, let's wait and see, I guess. > > -Joshua ==== Oh, there are tons more deficiencies. But I think that it is pretty good for $17. Really, to suit my needs, I don't think I can handle anything less than the HP48S. I should just carry my 48G around with me, but it is so big. -Josh > Here are some of the deficiencies with the CASIO model: > Calculate 1 divided by 9 and then multiply the result by 9. Subtract 1 > and you get 0. This is on a, presumably, BCD system. This strikes me > as very peculiar. In general the accuracy does not come anywhere near > matching the accuracy of the current line of HP calculators. > Register memory storage for statistical results is independent of > register storage for mathematical operations on the 32/33 series. This > is not true of the CASIO. The exponential range is significantly > smaller on the CASIO. The number of memory storage registers is > smaller on the CASIO. > CASIO also has a peculiar algebraic bug. It can handle 2(3 + 4), but > has problems with (3 + 4)2. I'm suspicious that there are other > similar bugs in their S-VPAM truly algebraic interpreter. > Feature for feature, the CASIO may match up, but I don't think they > match up in the details. > Unfortunately, right now, there is nothing even close to the 32SII > available in the market, RPN or Algebraic. I would pay the price > differential for the 33S, but I would have to be convinced of its > reliability and durability. >> So, I bought a Casio fx-115MS algebraic calculator as a beater. I went to >> Fry's (where I tried to convince a man to buy his son an HP49G+ instead of >> a TI, by the way) and picked the calculator that had the most features that >> I wanted. >> >> Pros: >> Small, true algebraic, and only US$17! >> >> Cons: >> Wont let use e^x on complex numbers, wont return complex results for arcsin, >> no RPN, and not programmable. >> >> Well, on the 32SII (and hence the 33S) you can use e^x on complex numbers, >> but you can't get a complex result from arcsin. Plus, you have RPN and >> programmability. >> >> The BIG question is this: >> Are the features of the HP33S (other than RPN) worth another US$43? >> >> More than three times the price, and you don't get that much more. Sure, >> I really like programmability, but at that price I just don't think that >> enough people will buy them to make the bean counters at HP happy. >> >> Well, let's wait and see, I guess. >> >> -Joshua -- -Joshua Belsky jjbelsky@yahoo.com http://belsky.net ==== : : Calculate 1 divided by 9 and then multiply the result by 9. Subtract 1 : and you get 0. _______________________ Isn't this the right answer? ==== > : > : Calculate 1 divided by 9 and then multiply the result by 9. Subtract 1 > : and you get 0. > _______________________ > > Isn't this the right answer? Not if the calculator keeps a finite number of decimal places. -- Tom Lake Experience keeps a dear school but fools will learn in no other - Poor Richard's Almanack ==== Calculate 1 divided by 9 and then multiply the result by 9. Subtract 1 and you get 0. ______________________ I asked: Isn't this the right answer? ______________________ Not if the calculator keeps a finite number of decimal places. ______________________ Oops. Moral: Never let an English Lit. major ask math questions in public. ==== : Calculate 1 divided by 9 and then multiply the result by 9. Subtract 1 : and you get 0. : ______________________ : : I asked: : Isn't this the right answer? : ______________________ : : Not if the calculator keeps a finite number of decimal places. : ______________________ Oops. Moral: Never let an English Lit. major ask math questions in public. Except that back in 1976/77 on the National Semiconductor-7100, we displayed ten decimal digits but carried all computations out to twelve decimal digits internally. I believe that in this case my NS-7100 would have displayed just a 0. Perhaps the Casio is doing something similar? ==== It would be nice if that were so, but the CASIO has the same accuracy as the HP-6S and several other common calculators. They seem to have been made with the philosophy that if the result is sufficiently close to an integer, we will make it an integer. I don't know the exact mechanics of these algorithms, but the TI-83 uses a similar methodology. Some calculators use some form of binary storage and the 1/9 test would not be valid. The one that comes to mind is the HP-30S. This calculator has excellent accuracy. The TI-30XIIS and related calculators also use some form of binary storage of numbers, but their accuracy is on par with the CASIOs if not worse. I have not been impressed with the accuracy of the integration routine. It may use some form of Simpson's rule. I haven't done a careful analysis, however. What is impressive is the internal stack size. Surprisingly, CASIO documents the workings of the internal algebraic stack fairly comprehensively. I have rarely seen this sort of thing on the paper documents that you get with such calculators. So it doesn't get lost in translation -- I do think that the 115 series is impressive value for the money. I llike the older model 115 over the newer models because on the two latest editions, the STO key is a shifted function. Why the STO would be a shifted function when you have the total waste of space with a cube root key is not clear. I would also like to see the comparable model with built-in conversions and constants. And if KINPO and SPECTRA can get 24 digit accuracy or whatever, why can't TI, CASIO, and SHARP? Also why can't I get a numeric display traditional algebraic with more than 3 memory register s, like the HP-20S, but WITH fraction capability? ... > : Calculate 1 divided by 9 and then multiply the result by 9. Subtract 1 > : and you get 0. > : ______________________ > : > : I asked: > : Isn't this the right answer? > : ______________________ > : > : Not if the calculator keeps a finite number of decimal places. > : ______________________ > > > > > > > Oops. > > Moral: Never let an English Lit. major ask math questions in public. > > > > > > Except that back in 1976/77 on the National Semiconductor-7100, we displayed > ten decimal digits but carried all computations out to twelve decimal digits > internally. I believe that in this case my NS-7100 would have displayed just > a 0. Perhaps the Casio is doing something similar? ==== > >Except that back in 1976/77 on the National Semiconductor-7100, we displayed >ten decimal digits but carried all computations out to twelve decimal digits >internally. I believe that in this case my NS-7100 would have displayed just >a 0. Perhaps the Casio is doing something similar? > That would still be the wrong answer - 0.111111111111 * 9 = 0.999999999999 - 1 does not equal zero. Pete M. Wilson Gamewood, Inc. wilsonpm@gamewood.net ==== --------------------------------------------------------------------- : : : > : >Except that back in 1976/77 on the National Semiconductor-7100, we displayed : >ten decimal digits but carried all computations out to twelve decimal digits : >internally. I believe that in this case my NS-7100 would have displayed just : >a 0. Perhaps the Casio is doing something similar? : > : : That would still be the wrong answer - 0.111111111111 * 9 = : 0.999999999999 - 1 does not equal zero. : : Pete M. Wilson : Gamewood, Inc. : wilsonpm@gamewood.net But on the other hand, as my lawyer would say, who ever decided that a calculator should lie to us? It is perhaps the single most important architectural decision, next to algebraic/RPN, to decide that a calculator should keep 2 secret digits which it never shows us and use them in a half-a**ed, er I mean half-hearted, attempt to fake an EXACT answer, when we all know or sort of remember when we think about it that it is really only an APPROXIMATE mode machine, and that we will have to wait 25 more years before we can completely trust an exact-mode calculator. I do not like the effects of hidden digits. If you are a calculator, tell us what you computed, not what you guess the exact answer to be. I do not like the answer my NS-7100 gave me, which was: 0. It should have given me: 0.000000001 which is the correct APPROXIMATE mode answer. It guessed, well not really--it tried its best to compute-- the correct EXACT mode answer: 0. which it happened this time to get right, but which it had no business messing with anyway! ==== --------------------------------------------------------------------- : : : > : >Except that back in 1976/77 on the National Semiconductor-7100, we displayed : >ten decimal digits but carried all computations out to twelve decimal digits : >internally. I believe that in this case my NS-7100 would have displayed just : >a 0. Perhaps the Casio is doing something similar? : > : : That would still be the wrong answer - 0.111111111111 * 9 = : 0.999999999999 - 1 does not equal zero. : : Pete M. Wilson : Gamewood, Inc. : wilsonpm@gamewood.net No, it doesn't equal zero, not exactly. But it does equal: 0.00000000001 , which my NS-7100 would have displayed as 0. It would have rounded the 0.00000000001 to ten significant figures: 0.000000000 , and displayed it as 0. , which is correct, rounded to 10 significant figures At least is does according to Keats, Byron, and Shelly, my favorite poets. ==== No, they aren't strictly programmable. They do have quite a lot of room for writing simple algorithms though. It has 9 variables, and assignment operator, a multi-statement separator, ANS variable, evaluate button and numeric calc functions. Not too limited for a pocket scientific! I've found several compact routines for root-finding etc, just to see what could be done with it. I also like being able to make complex number display default to polar. My g+ beats the pants off it of course, but it's rather absent-minded when it comes to registering keystrokes, and regularly switches itself off. > So, I bought a Casio fx-115MS algebraic calculator as a beater. I went to > Fry's (where I tried to convince a man to buy his son an HP49G+ instead of > a TI, by the way) and picked the calculator that had the most features that > I wanted. > > Pros: > Small, true algebraic, and only US$17! > > Cons: > Wont let use e^x on complex numbers, wont return complex results for arcsin, > no RPN, and not programmable. > > Well, on the 32SII (and hence the 33S) you can use e^x on complex numbers, > but you can't get a complex result from arcsin. Plus, you have RPN and > programmability. > > The BIG question is this: > Are the features of the HP33S (other than RPN) worth another US$43? > > More than three times the price, and you don't get that much more. Sure, > I really like programmability, but at that price I just don't think that > enough people will buy them to make the bean counters at HP happy. > > Well, let's wait and see, I guess. > > -Joshua ==== > I've found several compact routines for root-finding etc, just to see > what could be done with it. I also like being able to make complex > number display default to polar. My g+ beats the pants off it of > course, but it's rather absent-minded when it comes to registering > keystrokes, and regularly switches itself off. It might be saying, I'm tired, I want to rest ;-) -- Bhuvanesh ==== Great, cheap calc. My kids carry those to school. They have a simple solver app, calculus functions, complex num mode, hex mode, two line, solar powered, and more. I am also waiting for the 33s to try to convert my kids to hp. Toby > So, I bought a Casio fx-115MS algebraic calculator as a beater. I went to > Fry's (where I tried to convince a man to buy his son an HP49G+ instead of > a TI, by the way) and picked the calculator that had the most features that > I wanted. > > Pros: > Small, true algebraic, and only US$17! > > Cons: > Wont let use e^x on complex numbers, wont return complex results for arcsin, > no RPN, and not programmable. > > Well, on the 32SII (and hence the 33S) you can use e^x on complex numbers, > but you can't get a complex result from arcsin. Plus, you have RPN and > programmability. > > The BIG question is this: > Are the features of the HP33S (other than RPN) worth another US$43? > > More than three times the price, and you don't get that much more. Sure, > I really like programmability, but at that price I just don't think that > enough people will buy them to make the bean counters at HP happy. > > Well, let's wait and see, I guess. > > -Joshua > > > > -- > -Joshua Belsky > jjbelsky@yahoo.com > http://belsky.net > ==== X > Well, on the 32SII (and hence the 33S) you can use e^x on complex numbers, > but you can't get a complex result from arcsin. Plus, you have RPN and > programmability. > > The BIG question is this: > Are the features of the HP33S (other than RPN) worth another US$43? 32KB RAM? Much more than three times! Perhaps also more speed... > More than three times the price, and you don't get that much more. Sure, > I really like programmability, but at that price I just don't think that > enough people will buy them to make the bean counters at HP happy. > > Well, let's wait and see, I guess. We just have to do that. My guess is this: the uglies keyboard looks (V-shape) ever from HP ==== I guess this is sort of like peer review, and a good place for it. You can check if results can be duplicated by others, and have others check for errors. GC > > > Yes, you are right - I screwed up. Apparently I had the emulator in > degrees mode and the real 49G in radians - oops. This accounted for > the differences in the numerical results of the integrals on Daniel's > page. I've been too busy to try all the integrals or plotting > examples, but I imagine I will see similar results. > > Therefore, I retract my statement that the ...emulator is screwed up > results. I now have faith in the emulator again. :-) > ==== >I had thought it plausible that the emulator's implementation of some >underlying Saturn math routines might be ever so slightly different >than a real Saturn processor, where this difference might only be >obvious for extreme cases when the last significant digit of precision >has a large bearing on the outcome of a calculation or plot. Of >course, I am no expert on emulation software or on the low-level >details of HP's math routines, so that assumption may have been naive. EMU48 emulates the Saturn/Yorke at the machine code / hardware level ie. , it's not a simulation. If there was an error in the emulation of the basic arithmetic instructions, then I seriously doubt anything else would function at all. ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. ==== I have been a 49G user for a few years now, and I now own two 49g+ models. I upgraded from my old 48G when the first 49G came out, and I never turned back. I realize now that I will not use my 48G again. I have a trusty 48G (not GX), complete with the manuals. Rather than try to scrape a few dollars out of it, I would like to donate it to a worthy cause. I'm sure someone on this newsgroup might like to have the 48G, but be limited on funds. I recently donated my 49G to someone who will make good use of it, and I am happy with this decision. So, here are the rules... 1) Best reason wins... 2) I decide the best reason. Please let me know why I should give this fine machine to you. If you don't Send to << mitch dot thompson at ieee dot org >> Mitch ==== I am also a poor college student! I have used the 38/39Gs in highschool and am keen to give back to the HP community (see www.members.optusnet.com.au/~the_grosvenors) I am also a regular reader and contributor to this forum (see HP48SX for pre-calculus student??). If your 48 found its way to me, I would like to use it to learn SysRPL (which is difficult on the 39G) in order to write a set of tutorials about sysRPL programming for 39G users. It would also be of great benifit to my education in Mechatronic engineering. Regardless of what I have to say here, I trust > I have been a 49G user for a few years now, and I now own two 49g+ models. > I upgraded from my old 48G when the first 49G came out, and I never turned > back. I realize now that I will not use my 48G again. I have a trusty 48G > (not GX), complete with the manuals. Rather than try to scrape a few > dollars out of it, I would like to donate it to a worthy cause. I'm sure > someone on this newsgroup might like to have the 48G, but be limited on > funds. I recently donated my 49G to someone who will make good use of it, > and I am happy with this decision. So, here are the rules... > > 1) Best reason wins... > 2) I decide the best reason. > > Please let me know why I should give this fine machine to you. If you don't > > Send to << mitch dot thompson at ieee dot org >> > > > Mitch > > ==== >I would like to use it to learn SysRPL (which is > difficult on the 39G) in order to write a set of tutorials about sysRPL > programming for 39G users. You will need more memry than that to do proper sysRPL. The mnemonic tables alone would already fill a 48G Arnaud ==== I heard from a friend (Colin Croft) that the 48G was a good place to start by learning the usrRPL and sysRPL commands for the 48 and then moving on the to the 38/39 how valid it is I do not know, but that is whst I heard! M@ > >I would like to use it to learn SysRPL (which is > > difficult on the 39G) in order to write a set of tutorials about sysRPL > > programming for 39G users. > > You will need more memry than that to do proper sysRPL. The mnemonic > tables alone would already fill a 48G > > Arnaud ==== > I heard from a friend (Colin Croft) that the 48G was a good place to start > by learning the usrRPL and sysRPL commands for the 48 and then moving on the > to the 38/39 how valid it is I do not know, but that is whst I heard! > If you want to enjoy sysRPL you need at least a 48G+ or 48GX or a 49. The best development environment on the calc (which is better as you can learn a lot by decompiling ROMroutines) is Jazz light to which you need to add the entry points tables which brings you to a good 50-60 k of RAM. To this I have to add that you will certainly loose your RAM quite often while learning sysRPL so a GX with expension card or a 49 with flash are really must haves. But a 48G with 32k just doesn ot have enough RAM to develop on. Arnaud ==== way, but I simply cant afford to buy any of these calculators! I guess that leaves me stuck with EMU48 and the good old HP Goodies Disks! M@ > > I heard from a friend (Colin Croft) that the 48G was a good place to start > > by learning the usrRPL and sysRPL commands for the 48 and then moving on the > > to the 38/39 how valid it is I do not know, but that is whst I heard! > > > > If you want to enjoy sysRPL you need at least a 48G+ or 48GX or a 49. > The best development environment on the calc (which is better as you > can learn a lot by decompiling ROMroutines) is Jazz light to which you > need to add the entry points tables which brings you to a good 50-60 k > of RAM. > > To this I have to add that you will certainly loose your RAM quite > often while learning sysRPL so a GX with expension card or a 49 with > flash are really must haves. But a 48G with 32k just doesn ot have > enough RAM to develop on. > > Arnaud ==== MG, I hope you get the 48G, as you can then have people (here in the 'States, in Germany, etc.) upgrade it to G+ level RAM (128 kb) and then you'll have a serviceable machine for your sysRPL needs. (E.g., Randy Sloyer at www.fixthatcalc.com) > way, but I simply cant afford to buy any of these calculators! I guess that > leaves me stuck with EMU48 and the good old HP Goodies Disks! > > M@ > >> > >>>I heard from a friend (Colin Croft) that the 48G was a good place to >>> > start > >>>by learning the usrRPL and sysRPL commands for the 48 and then moving on >>> > the > >>>to the 38/39 how valid it is I do not know, but that is whst I heard! >>> >>> >>If you want to enjoy sysRPL you need at least a 48G+ or 48GX or a 49. >>The best development environment on the calc (which is better as you >>can learn a lot by decompiling ROMroutines) is Jazz light to which you >>need to add the entry points tables which brings you to a good 50-60 k >>of RAM. >> >>To this I have to add that you will certainly loose your RAM quite >>often while learning sysRPL so a GX with expension card or a 49 with >>flash are really must haves. But a 48G with 32k just doesn ot have >>enough RAM to develop on. >> >>Arnaud >> > > ==== > I have been a 49G user for a few years now, and I now own two 49g+ models. > I upgraded from my old 48G when the first 49G came out, and I never turned > back. I realize now that I will not use my 48G again. I have a trusty 48G > (not GX), complete with the manuals. Rather than try to scrape a few > dollars out of it, I would like to donate it to a worthy cause. I'm sure > someone on this newsgroup might like to have the 48G, but be limited on > funds. I recently donated my 49G to someone who will make good use of it, > and I am happy with this decision. So, here are the rules... > > 1) Best reason wins... > 2) I decide the best reason. > > Please let me know why I should give this fine machine to you. If you don't *cue in random cello song* I'm a poor (right..aren't we all?) college student who has to carry around 2 calculators. I've my father's 41-C for quick calcs, and an 83+(won it in a math contest - can't argue with $0) for when i need graphing. Only carrying around one instead of two (three if you count the sorry excuse for a calculator i have to use in exams) calculators would be a big win. I'm sure a good developper will need one to check his programs for backward compatibility, but that's definitely not me. ==== I used 2 weeks my HP 49g+ and it displays a low battery, what«s a problem? ==== > I used 2 weeks my HP 49g+ and it displays a low battery, what«s a problem? Time for new batteries. ==== Pedro Liberato schrieb: > I used 2 weeks my HP 49g+ and it displays a low battery, what«s a problem? update to SW 1.22 done? Toralf ==== Toralf Schumann ha scritto nel messaggio > Pedro Liberato schrieb: > > I used 2 weeks my HP 49g+ and it displays a low battery, what«s a problem? > > update to SW 1.22 done? > > Toralf > batteries and they were good, they were alcaline!!! ==== > Toralf Schumann ha scritto nel messaggio > > Pedro Liberato schrieb: > > > I used 2 weeks my HP 49g+ and it displays a low battery, what«s a > problem? > > > > update to SW 1.22 done? > > > > Toralf > > > > batteries and they were good, they were alcaline!!! Yes, I updated. Maybe.. the original battery has a BUG. :o) Please, we need urgent a new ROM. ==== Toralf Schumann ha scritto nel messaggio > Pedro Liberato schrieb: > > I used 2 weeks my HP 49g+ and it displays a low battery, what«s a problem? > > update to SW 1.22 done? > > Toralf > batteries and they were good, they were alcaline!!! ==== I'm proud to annunce the new release of my gasdynamics library. I made lot of efforts to ensure stability via a more robust code... This version correct bugs found trough intensive tests and add also a pair of smart things... I'll be happy if someone like to test it or want to download it. Because I've not my own web site and hcalc.org is not updating in a month you can request the library directly at kickaha@tiscali.it Here's the extract from the readme (history): * Text below revised. New section added: INTERPRETING BAD ARGUMENT VALUE ERRORS. * Inputline procedure related with Gamma STOring greatly enforced; now it should be able to resist to any attempt of circumvention. * Oblique shock waves table computational methods revised; the outputs and the available commands have been changed too. This release also fix bugs found through intensive tests. * Cosmetic bugs: - Now founding strong solution in oblique shock waves table prompts the correct string. - No more arguments swapping if error in any of the table. * System flags settings bugs: - Fixed overflow error due to system flags settings (flag -22 set) when using the infinity symbol. - Computations involving angle units - Prandtl-Meyer angle and Oblique shock waves tables - now work correctly independently on the current settings for the angle system. * Prandtl-Meyer angle table bugs: Check against w(M)_max now does not error. * Oblique shock waves table bugs: Added a check against the minimum Sigma angle (Delta angle equal to zero) when using the command LS+'M,Sigma' searching for a weak solution. Federico 'Kickaha' Marziali ==== Sounds interesting... but one question: For which calc did you write the LIB (48G*, 49G or even 49G+)? Martin > I'm proud to annunce the new release of my gasdynamics library. I made lot > of efforts to ensure stability via a more robust code... > This version correct bugs found trough intensive tests and add also a pair > of smart things... > I'll be happy if someone like to test it or want to download it. Because > I've not my own web site and hcalc.org is not updating in a month you can > request the library directly at kickaha@tiscali.it > > Here's the extract from the readme (history): > * Text below revised. New section added: INTERPRETING BAD ARGUMENT VALUE > ERRORS. > * Inputline procedure related with Gamma STOring greatly enforced; now it > should > be able to resist to any attempt of circumvention. > * Oblique shock waves table computational methods revised; the outputs > and the > available commands have been changed too. > This release also fix bugs found through intensive tests. > * Cosmetic bugs: > - Now founding strong solution in oblique shock waves table prompts > the correct string. > - No more arguments swapping if error in any of the table. > * System flags settings bugs: > - Fixed overflow error due to system flags settings (flag -22 set) > when using the infinity symbol. > - Computations involving angle units - Prandtl-Meyer angle and Oblique > shock waves tables - now work correctly independently on the current > settings for the angle system. > * Prandtl-Meyer angle table bugs: Check against w(M)_max now does not > error. > * Oblique shock waves table bugs: Added a check against the minimum Sigma > angle (Delta angle equal to zero) when using the command LS+'M,Sigma' > searching for a weak solution. > > Federico 'Kickaha' Marziali > > > ==== http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3059534043&category=50577 &rd=1 ==== > I recently got my 32MB sd card for my hp49g+. I've noticed that the > SD card dosen't appear in the library, which the manual says. I also > wanted to store library files (with titles such as Library 756) on > the SD card, but I get an error, Error: Invalid DOS Name. How can I > use my SD card just like another port, or Home? Is there a program I > can use? Do I need to format my card or what? Now what can I do to > either get the SD card to show up in the library or be able to store > library files on the SD card? > > > Nathan Format(FAT 16,not FAT32) SD card by PC. ==== > > > The manual says you can only access > > > the SD card through the Files key and > > > will not ever show up as a port. > > > > > Huh? > > > > It should say the exact opposite! > > What they hopefully *meant* to say is: > > Pressing the FILES key shows the SD card as Port 3 and lets you browse > it, whereas the LIB key (which shows your ports) never shows it at > all. > > -Joe- OK, where am I confused? Is this not the same thing? I thought the LIB key shows the ports, but won't show port 3. The FILES key is used to access the SD card. Scott ==== Is there a program that lets ou create Apps-like menus, so you can make and edit an Apps-like menu. You could taker the libs, put them in a directory, then move the directory to the sd card, then you make a shortcut, or menu to the sd card. Will this work? ==== > Is there a program that lets ou create Apps-like menus, so you can > make and edit an Apps-like menu. You could taker the libs, put them > in a directory, then move the directory to the sd card, then you make > a shortcut, or menu to the sd card. Will this work? :3: Program :3: {directory sub directory Program} Try to EVALuate those in your 49g+ Veli+Pekka ==== help me please!. I can't compile programs in sysrpl, my calculator flags are -85 sysrpl stk disp -92 MASD sysrpl mode {256 257} attach i suppose that the hp49g+ doesn't need any special program to do this. If I being a mistake please correct me. i.e. :: CK2&Dispatch # 11 :: %+ SUMA TERMINADA FlashMsg ; ; @ to compile this, the hp49g+ throw me the next error: asm error: invalid file What I doing wrong?,My calc is bad? or simply i haven't idea about program in sys :) TIA A. Badillo sorry by my english! ==== says... > help me please!. I can't compile programs in sysrpl, my calculator flags are > -85 sysrpl stk disp > -92 MASD sysrpl mode > {256 257} attach > > i suppose that the hp49g+ doesn't need any special program to do this. > If I being a mistake please correct me. > i.e. > :: > CK2&Dispatch > # 11 > :: > %+ > SUMA TERMINADA > FlashMsg > ; > ; > @ > to compile this, the hp49g+ throw me the next error: > asm error: > invalid file > > What I doing wrong?,My calc is bad? or simply i haven't idea about program in sys > :) Do you have extable installed? ==== TCO> :: TCO> CK2&Dispatch TCO> # 11 TCO> :: TCO> %+ TCO> SUMA TERMINADA TCO> FlashMsg TCO> ; TCO> ; TCO> @ TCO> to compile this, the hp49g+ throw me the next error: TCO> asm error: TCO> invalid file TCO> What I doing wrong?,My calc is bad? or simply i haven't idea about program in sys :) Make sure that the @ is really the last character in the source string - i.e. no newline after it. Go to the end of the file and hit backspace until the cursor is right after @. - Carsten ==== > Help... > > I have seen a few posts recently on this subject, so I know I'm not the only > one. > > I am using Windows 98 (not SE) and trying to connect to my 49G+. I have > downloaded the latest USB driver and Connectivity Kit from the HP web site. > When the calculator was first connected to the PC via the USB cable, Windows > 98 identified the new connection so all was fine to that point. I updated > the drivers as instructed and that seemed to be fine too. > > However, when I run Conn4x, it just doesn't seem able to see the 49G+. Conn4x should displays HPx9g+ in divice list if win98 identified HPx9g+ divice. Then choose HPx9g+, press right shift then right arrow then [enter]. Then press connect button. > In the Conn4x instructions, it says I am supposed to press right shift then > right arrow, which turns on the Xmodem Server. The very first time I ran > Conn4x after upgarding the USB driver, I got a message from Conn4x saying > the Xmodem was the wrong kind of server and it failed to connect. > Thereafter, I get no error message other than Conn4x fails to find the 49G+. > > I have looked through the PDF manual but have not found any help there > either, I'm stuck. > > Conn4x is set to Auto - so I have not played with the connection speeds. > When I click on the menu Help in Conn4x, nothing happens - i.e. there is no > > Can you help please? > > [PS - I think the 49G+, is, on the whole, a very nice calculator. Big, big > improvement on the 49G+. ] > > > Mark. ==== > > Conn4x is set to Auto - so I have not played with the connection speeds. > > When I click on the menu Help in Conn4x, nothing happens - i.e. there is no Build 1783 on the HP web is JUST the Conn4x.exe file. The help files and many support files must be installed from the original CDROM then copy the build 1783 file. This probably will not solve your connection problem. Windows 98 (plain) and to a less extent Windows 98 SE have some internal problems with certain USB devices. -- - - - - - - - - - - - - - - - - Bill Graves RKBA! bgraves@ix.netcom.com ==== >>>Conn4x is set to Auto - so I have not played with the connection speeds. >>>When I click on the menu Help in Conn4x, nothing happens - i.e. there is no > > > Build 1783 on the HP web is JUST the Conn4x.exe file. The help files and many support files must be installed from the original > CDROM then copy the build 1783 file. This probably will not solve your connection problem. Windows 98 (plain) and to a less extent > Windows 98 SE have some internal problems with certain USB devices. Yes, on my computer, the certain USB devices are the 49g+, and maybe, just maybe, the Prolific USB to RS-232 converter. The USB converter doesn't work for screen captures with HPComm, and doesn't work at all with that Xmodem HP Connectivity kit for the 49G, but it does work fine for file transfers with HPComm. With HyperTerminal, it works perfectly for Kermit and Xmodem file transfers, capture text from the calculator's print commands and XMIT, and sending text for the calculators' SRECV. I can also use the converter with various software for printing to the RS-232 ports of various Epson impact dot matrix printers, communication with my CyberPower UPS, and to put it simply, everything else that I've tried to use it with. The HPComm capture image and the Xmodem HP Connectivity Kit do work fine with a real RS-232 port. But the USB does work flawlessly with, let's see now, A Canon BubbleJet printer, an Epson Stylus inkjet or whatever printer, a Stratitec hub, a Visioneer scanner (a warranty replacement for an earlier model with mechanical problems, which also worked fine as far as USB went), a Seagate tape drive, and the SanDisk MMC/SD card drive. Maybe I've left something out, but in any case, the USB on this computer works fine with everything else that I've tried it with. Somehow, whatever problems MS Windows 98SE on this computer may have with USB, I've a pretty strong hunch that there are viable work-arounds for them. I hope that HP doesn't give up until they have everything working. -- James ==== Can anyone comment on the build quality of the 49g+ specifically as it compares to the TI-89? I've read many posts about the stiff keys on the 49g+ and the light, hollow sound it produces when those keys are pressed. To my amazement, most users report that this is an *improvement* over the 49g. I've never used a 49g, so I'll take everyone's word for it on that. But how does this compare to the TI-89? FWIW, I am a long time HP user (12C, 17BII, 48GX), but I am also familiar with the TI's. I have always found the TI's to have a very mushy keyboard, with no tactile feedback. So, for those of you that have experience using both of these calc's, which is the lesser evil: the stiff keyboard and hollow sound of the 49g+, or the mushy keyboard of the 89? ==== PS I wanted to comment on the comparison website. The following qutoe appears on there: Doug Burkett says The 49G+ development tools are free and open. Developing flash apps for the 68K calculators requires using TI's proprietary and expensive development system, with its cumbersome 'certificates'. On the 68Ks, ASM programs (C and assembler) are limited to 24K through hardware protection. This can be overcome with a hack, which shouldn't be necessary to begin with. >Developing flash apps for the 68K calculators requires using TI's proprietary >and expensive development system TI Flash Studio is and has been freely available for download at http://education.ti.com/us/resources/developer/8992/download/download.html. Proprietary? The compiler uses standard K&R style C. Sure, the compiler does leave some things to be desired but if you know C you shouldnt have any trouble programming with it. Maybe he is talking about the FLASH App APIs. Of course those are proprietary. The TI-89 OS is proprietary, just like HP's. >with its cumbersome 'certificates'. Ya, I agree. it is a pain having to send your source code off to TI for two weeks befoe it is compiled. I dont really like the idea of them not letting me compile a program because they disagree with the content of it. It used to be like this for the TI-83 as well but TI released something that let people sign their own apps. Maybe they will do this for the 68k series. >ASM programs (C and assembler) are limited to 24K through hardware protection. >This can be overcome with a hack, which shouldn't be necessary to begin with. Just as a side note, does any one know why it is limited to 24k? Aleph Mobius ==== > >Developing flash apps for the 68K calculators requires using TI's > proprietary > >and expensive development system > > TI Flash Studio is and has been freely available for download at > http://education.ti.com/us/resources/developer/8992/download/download.html. OK, I'd been hoping to keep this between Doug and me. Oh, well... Yes, TIFS is free, and one can always request a developer key, which allows him/her to sign the app for his/her own calculator for testing. > Proprietary? The compiler uses standard K&R style C. Sure, the > compiler does leave some things to be desired but if you know C you > shouldnt have any trouble programming with it. Maybe he is talking > about the FLASH App APIs. Of course those are proprietary. The TI-89 > OS is proprietary, just like HP's. I, too, think Doug's talking about the Flash app APIs. > >ASM programs (C and assembler) are limited to 24K through hardware > protection. > >This can be overcome with a hack, which shouldn't be necessary to > begin with. > > Just as a side note, does any one know why it is limited to 24k? Supposedly to prevent users from pirating for-pay flash apps and/or releasing them as plain RAM programs. -- Bhuvanesh ==== > TI Flash Studio is and has been freely available for download at > http://education.ti.com/us/resources/developer/8992/download/download.html. > Proprietary? The compiler uses standard K&R style C. Sure, the > compiler does leave some things to be desired but if you know C you > shouldnt have any trouble programming with it. I have never used flash studio, but have chatted with some who have. I know of one person complaining about when porting from tigcc to a flash-app the compiler would not compile the origional c code (just algorythms no API stuff). Also the compiler ti uses is not the sharpest crayon in the box. > >ASM programs (C and assembler) are limited to 24K through hardware > protection. > >This can be overcome with a hack, which shouldn't be necessary to > begin with. > > Just as a side note, does any one know why it is limited to 24k? It used to be limited to 8k. Then ti realized just how oppresive they where being and quickly upped it to 24k. The official reason is that they do not want to have flash apps pirated. I wouldn't be supprised if it was possible. -Samuel S http://www.nyall.net ==== > PS I wanted to comment on the comparison website. The following qutoe > appears on there: I have taken the liberity of quoting you... if you'd prefer I didn't, please let me know and I'll remove your comments ASAP. Al ==== By the way, Al, TI-Basic does have a comment marker. It looks like the copyright symbol, C with a circle around it. -- Bhuvanesh ==== I dont understand everyone's comments about the TI-89's mushy keyboard. I own an HP-48GX and a TI-89 and from what I've heard the 49g+ keyboard is similar to the 48's. It took me a while to get used to the 48GX keyboard after using the TI-89 for over two years and a Casio AFX 2.0 for a little less than that. The clicking keys dont do anything for me. I just have to press harder, which makes typing slower. I suppose it is just a matter of personal preference but from the momment I took my TI-89 from its package I had no trouble telling if a keypress was regsitered. The keys press back against my fingers plenty hard enough. Maybe its just me. Maybe its all those years of Game Boy (origional) playing. Aleph Mobius ==== > Can anyone comment on the build quality of the 49g+ specifically as it > compares to the TI-89? I can, kinda. I've got a TI92+, which has an almost identical key-feel to the TI89 my colleague has. The 49G+ keyboard is rather like the 48G I've got, but feels a bit more 'hollow'. It is not dissimilar from the 89's keys, albeit the HP has clicky-feedback whch the TI does not. I've never had a problem with the TI keyboard, which is actually quite good for fast data entry; I just don't like that they are all algebraic entry, which I'm not keen on. The calculator construction is likewise a bit 'hollow'-feeling (on both the HP and the TI) but is generally fine. I'll probably upset a number of people on this group by saying that the two calculators are probably about evenly matched, quality-wise, which is IMHO a good thing. However, the TI's aesthetics are a bit black, if you get my meaning; the HP certainly looks nicer! I have no idea how the g+ will fare in the future, obviously, and I'm really not sure how it would respond to mechanical abuse...my colleague's TI89 went down the stairs on the way to coffee and was still running the program just fine when it was recovered! ttfn JasonG ==== oN 16-Nov-03, bmclaurin said: > So, for those of you that have experience using both of these calc's, > which is the lesser evil: the stiff keyboard and hollow sound of the > 49g+, or the mushy keyboard of the 89? I find the 49G+ to be as good or better than the TI-89, as far as the keyboard is concerned. In other matters, it's too soon to tell, as I'm well down the slope on the learning curve. -- Bill ==== I have used calcs of all kinds, TIs, mostly HPs, and various off brand ones I picked up only because I was either in a financial bind or they were irresistably cheap for the advertised features. To me there is NO contest- the HP-49G+ key feel and click is preferable to all manner of mushy keys, including TI keys. As I posted OT (I think on this very board), I even prefer my PC keyboards clicky. > oN 16-Nov-03, bmclaurin said: > > >>So, for those of you that have experience using both of these calc's, >>which is the lesser evil: the stiff keyboard and hollow sound of the >>49g+, or the mushy keyboard of the 89? >> > > I find the 49G+ to be as good or better than the TI-89, as far as the > keyboard is concerned. In other matters, it's too soon to tell, as I'm > well down the slope on the learning curve. > > ==== Although my hp49g+ is on my way to me from SAMSON CABLES and should arrive next week I think you should prefer the hp49g+. There is no contest.hp wins without a fight. I am an 49g user for 4 years now. Look at this link. I do have some comments about this comparison but first read it : http://alpage.ath.cx/hptute/compare.htm Keep something in mind : The hp49g+ and 49g highlight is the equation writer. after 4 years of use I dont remember myself use more than one or two baracts. when I have a big expression I have to work with - equation writer is the tool ! you cannot imagine how simple it works. after you will get used to it you cnnot go back to simple calculator anymore - it will be just as going back in time. look at the example in this link I gave you and try to understand how it works. about the auto simplify of the TI - there is a program you can download from hpcalc.org that does the autosimplification. I use a lot of usefull programs from hpcalc.org like key assignment and numerical solvers and editors. Some people will say that its hard to learn to use the hp - I think its not true. After you catch the idea of RPN and how hp works - its very easy and very friendly. Notice that the hp equation writer has also been used on the hp48g and it works fine but the hp49g+ is so much faster than the hp48g that the equation writer becomes a real time editor. I heard that TI hands better with complex numbers - dont know if its true since I dont use complex numbers but all the electrical engineers here in my university use the TI - dont know if its not more than tradition... In my opinion - for now TI belongs to the old age of calculators. if TI will come out with a new model than the 92 that has equation writer and the speed and memory of the 49g+ - than hp will have a problem. but for now - hp is ahead. I dont know what about the software archive of TI has to offer but if you are looking for usefull programs that very talented people has written look at : www.hpcalc.org As I see my friends who use the TI - They download usually games. hpcalc has much more to offer almost in any subject of studying and soon i am sure that there will be a nice place for hp49g+. Those people who write those programs are engineers and mathematicians and programers who get no profit from their work and sometimes I want to come and shake their hands gratefully. Let us know what was your decision... Ok. Bye. Idan. ==== > I heard that TI hands better with complex numbers - dont know if its > true since I dont use complex numbers but all the electrical engineers I use complex numbers a lot at uni, the HP beats any other calculator hands down in usability. If I want to do a polar->rectangular conversion for instance, all I need do is have RECT and CYLIN in my CUSTOM menu. On my TI-92, I have to change this in the mode setup. Apparently the 92/89 doesn't offer a way to enter complex numbers as ordered pairs, ie (x,y) either , which means more keystrokes. > In my opinion - for now TI belongs to the old age of calculators. if > TI will come out with a new model than the 92... about as powerful as the 49g+ except with linux underneath. Probably just a rumour though, this is getting a bit silly for a calculator, but hell I'd buy one =) ==== > > I heard that TI hands better with complex numbers - dont know if its > > true since I dont use complex numbers but all the electrical engineers > > I use complex numbers a lot at uni, the HP beats any other calculator > hands down in usability. If I want to do a polar->rectangular > conversion for instance, all I need do is have RECT and CYLIN in my > CUSTOM menu. On my TI-92, I have to change this in the mode setup. No, you don't. If you just want to convert, you can use >Rect or >Polar or ... It's kind of like the Exact/Approx/Auto mode setting: you can set the mode to whatever format you generally want the answer to be in, and if you occasionally need the answer in another format, you can use built-in commands to do that. > Apparently the 92/89 doesn't offer a way to enter complex numbers as > ordered pairs, ie (x,y) either , which means more keystrokes. Why are you guys so obsessed with the number of keystrokes? :-) Even if you were entering hundreds of x[k]+y[k]*i pairs, you could just type in the vectors x[k] and y[k] and then do x+y*i. > > In my opinion - for now TI belongs to the old age of calculators. if > > TI will come out with a new model than the 92... > > about as powerful as the 49g+ except with linux underneath. Probably > just a rumour though, this is getting a bit silly for a calculator, > but hell I'd buy one =) If they do, I'm not sure it would be worth it, with all the extra development cost and not much in terms of improvements for users. -- Bhuvanesh ==== > No, you don't. If you just want to convert, you can use >Rect or > >Polar or ... This only works with vectors, which can't really be used as complex numbers - you can't divide or multiply so aren't any use to me. What's more >Polar and >Rect only work to convert one quantity (ans(n) or some other prefixed input)- they don't change the display format of everything in the stack/history. If I want to do this on my TI, it takes me ten keystrokes. On my HP, it takes 3 at max, only one if my custom menu is already shown. The TI doesn't let me use the angle button to enter polar complex numbers either (only works with vectors). This means I have to muck around with e and i. > Why are you guys so obsessed with the number of keystrokes? :-) Hmm, let me enlighten you! :-) Less keystrokes make for: -Faster completion of task at hand -Reduced input error Subsequently less feelings of frustration and sore fingers from navigating endless menus et cetera, making Joe Calculator User a happier man! Something tells me you don't use your calculator for doing a lot of numeric math. I may be wrong but I think you'd better understand where I'm coming from, why for me and many others the general HP interface (keyboard and software) is so much more conventient and intuitive than anything else available. I could never go back to a casio or TI (both of which I've used for a long time before discovering HP). They just don't even come close to cutting it. Casio interfaces are so horribly rigid (NO means of customisation, no object-oriented working environment...) that I won't bother to compare them to TI or HP. TI on the other hand: none of the scientific functions (which I use a lot) are closer than two keystrokes away. If TI had had the same sort of intuitive ideas about how professionals, not just high-school students, like to use their calculators as HP has they'd probably have produced a machine I'd be happy to use by now. XYZ ==== > > I heard that TI hands better with complex numbers - dont know if its > > true since I dont use complex numbers but all the electrical engineers > > I use complex numbers a lot at uni, the HP beats any other calculator > hands down in usability. If I want to do a polar->rectangular > conversion for instance, all I need do is have RECT and CYLIN in my > CUSTOM menu. On my TI-92, I have to change this in the mode setup. > Apparently the 92/89 doesn't offer a way to enter complex numbers as > ordered pairs, ie (x,y) either , which means more keystrokes. Please see the >rect and >polar commands On the 89 complex numbers are symbolic expressions involving i. If you want to enter complex numbers as ordered pair then use vectors: [[x, y]] [[r,rect and >polar commands work with vectors and with complex arguments. -Samuel S www.nyall.net ==== > Notice that the hp equation writer has also been used on the hp48g and > it works fine but the hp49g+ is so much faster than the hp48g that the > equation writer becomes a real time editor. Can you still feel the delay when doing things like copy/paste in the 49g eqw? -Samuel ==== >about the auto simplify of the TI - there is a program you can >download from hpcalc.org that does the autosimplification. Wich one is it? Does it the job as well as the TI? Gru§ Alexandra ==== hi Alexandra I think you want AutoSimp49 http://www.hpcalc.org/hp49/math/misc/asv11.zip Gruss Achim Alexandra Schneider schrieb: > > > >>about the auto simplify of the TI - there is a program you can >>download from hpcalc.org that does the autosimplification. >> >> > >Wich one is it? >Does it the job as well as the TI? > >Gru§ >Alexandra > > ==== 1. You can get the TI-89 for much less than the Amazon price. Try searching froogle.com or shopper.cnet.com. I think I saw the TI-89 somewhere for $124.99. 2. The HP49G+ is still pretty much equivalent to a TI-89, besides the SD card (in my opinion, of course). 3. The TI-89 has an equation writer, including two free ones written by EW and Samuel Stearley. 4. You can do custom key assignments using Samuel Stearley's Shortcuts program. 5. RPN is not really the major barrier in learning how to use HP calculators. 6. I'm not sure how the TI handles complex numbers better. Please elaborate. 7. As for software written for the TI-68k, there is currently no one good place to get software, but I'm working on a (hopefully somewhat comprehensive) list of useful math/science/engineering programs. 8. Regarding games, that's unfortunately what many students buy these calculators for, although some of them are quite good. -- Bhuvanesh ==== > 1. You can get the TI-89 for much less than the Amazon price. Try > searching froogle.com or shopper.cnet.com. I think I saw the TI-89 > somewhere for $124.99. > > 2. The HP49G+ is still pretty much equivalent to a TI-89, besides the > SD card (in my opinion, of course). IrDA and USB at 115200 baud beeper, alarm system, all of the key can be tailored cursor keys work in the alpha mode, you can run RPL program in the ALG mode and vice versa. EQW is totally integrated into the system [HIST] key can pick stuff sfrom stack and you can naturally do CUT/PASTE Matrix writer can do CUT/COPY/PASTE Taylor SERIES is build in and mich, much more without Bhuvanesh TI would suck (now it blows) ==== > > 2. The HP49G+ is still pretty much equivalent to a TI-89, besides the > > SD card (in my opinion, of course). > IrDA and USB at 115200 baud The 68k also have USB cables to connect to a computer. As for the IR, it seems to have a very low range. > beeper, alarm system, These two are very minor things (in my opinion, of course). Would be nice to have, but I don't really miss them. > all of the key can be tailored See my post above. There is a program to do this on the 68k. > cursor keys work in the alpha mode, This is a big one. However, I now use my V200, so this is not an issue for me. > you can run RPL program in the ALG mode and vice versa. RPL does not exist on the 68k, so this is not an issue. > EQW is totally integrated into the system Same on the 68k if you get the Flash app version. > [HIST] key can pick stuff sfrom stack and > you can naturally do CUT/PASTE > Matrix writer can do CUT/COPY/PASTE Same on the 68k, if I understand you correctly. > Taylor SERIES is build in and mich, much more Taylor series is built-in on the 68k too. I guess you mean Laurent series? > without Bhuvanesh TI would suck (now it blows) -- Bhuvanesh ==== Veli-Pekka Nousiainen meinte >[... some Discussion snipped...) >without Bhuvanesh TI would suck (now it blows) What a waste of talent, shouldn't we get Bhuvanesh away from the dark side of the world? He should follow his inner dedication and convert to HP. Always stay on the bright side of life. Imagine what he could add to the software pool of HP48*/HP49* etc. SCNR :-) Gru§ G.9fnter from Germany -- ==== > 1. You can get the TI-89 for much less than the Amazon price. Try > searching froogle.com or shopper.cnet.com. I think I saw the TI-89 > somewhere for $124.99. > ive seen $109.99... > 2. The HP49G+ is still pretty much equivalent to a TI-89, besides the > SD card (in my opinion, of course). hahahaha! [quick reply] > > 3. The TI-89 has an equation writer, including two free ones written > by EW and Samuel Stearley. Two FREE ones?? Does that mean that getting TI programs can cost you?? > > 4. You can do custom key assignments using Samuel Stearley's Shortcuts > program. nice! > > 5. RPN is not really the major barrier in learning how to use HP > calculators. > agreed. > 6. I'm not sure how the TI handles complex numbers better. Please > elaborate. > I dont know either so... no comments > 7. As for software written for the TI-68k, there is currently no one > good place to get software, but I'm working on a (hopefully somewhat > comprehensive) list of useful math/science/engineering programs. > Thats just what TI needs. > 8. Regarding games, that's unfortunately what many students buy these > calculators for, although some of them are quite good. HAve you played Babal?? It rokz. ==== > > 3. The TI-89 has an equation writer, including two free ones written > > by EW and Samuel Stearley. > > Two FREE ones?? Does that mean that getting TI programs can cost you?? > > > > > > 4. You can do custom key assignments using Samuel Stearley's Shortcuts > > program. > > nice! of only mapping one program the origional eqw by EW. I aught to revamp part of it, but between school and work I'm 50/60 hour weeks. The Hail Expression Writer and shortcuts can be found on my site: http://www.nyall.net -Samuel ==== >>3. The TI-89 has an equation writer, including two free ones written >>by EW and Samuel Stearley. > > > Two FREE ones?? Does that mean that getting TI programs can cost you?? Yes. Actually, the TI's have a FlashApp capability, some of which can require `signing' for each calculator it can run on. -MrM ==== > > Two FREE ones?? Does that mean that getting TI programs can cost you?? > > Yes. Actually, the TI's have a FlashApp capability, some of which can > require `signing' for each calculator it can run on. That depends on whether you want to release the app as freeware (as I do, in which case you don't need a signing per calculator) or a commercial (for-pay) app. It's the developer's choice. -- Bhuvanesh ==== > > Although my hp49g+ is on my way to me from SAMSON CABLES and should > arrive next week I think you should prefer the hp49g+. There is no > contest.hp wins without a fight. I am an 49g user for 4 years now. > Look at this link. I do have some comments about this comparison but > first read it : > > http://alpage.ath.cx/hptute/compare.htm Wow, my webpage has become famous :-) Please note that it is being hosted by my main PC, and has frequent downtime - eg whenever I need the bandwidth for online games... I'll try and get it hosted by my uni later. If you have any feedback, let me know and I'll adjust the page. Al http://alpage.ath.cx/hptute ==== > > Al > http://alpage.ath.cx/hptute > Forgot the trailing slash - it should be http://alpage.ath.cx/hptute/ Al ==== I just wanted to know if there is an HP49G+ emulator already. Neo33, ==== > > I just wanted to know if there is an HP49G+ emulator already. > > > Neo33, Not at this time, no. But I do not know if it is being actively worked on. Does anyone in this NG know of any active projects? Matthew F. G. ==== > Not at this time, no. But I do not know if it is being actively > worked on. Does anyone in this NG know of any active projects? I'm not sure, but the key bit of information a simulator developer would need is that the CPU is a Samsung S3C2410, and the Flash is an SST 39VF160. Information is available from the manufacturer's web sites: http://www.samsung.com/Products/Semiconductor/SystemLSI/MobileSolutions/Mobil eASSP/MobileComputing/S3C2410X/S3C2410X.htm http://www.sst.com/products.xhtml/parallel_flash/39/x16/SST39LF160 There are photos of the innards of an HP 49G+ prototype on my web site: Select the HP 49G+ innards album. Click on a thumbnail for a larger view, and you can click on that for a full size (2272x1704 pixel, 2 MB) image if you really need to inspect the details. Sorry the pictures aren't of better quality, I didn't have a tripod or good lighting with me at the time. ==== 1. I have a new HP49G+ and would like to transfer variables etc. from my old HP49G. I keep getting connect time outs. I am using the Get ==== > 1. I have a new HP49G+ and would like to transfer variables etc. from > my old HP49G. I keep getting connect time outs. I am using the Get Um, the HP49G does not have an IR port :-) > > 2. I am also having trouble getting the USB Conn4x to work. When I > try to connect I get get a message that says cannot connect to the > HP49G+, yet, I can do Xmodem transfers and have successfully updated > the ROM in the calculator. Is there any documentation for the HP SW? You're running win98, right? If so, join the club. I just use IrDA under win98, and the old software. Al > > Any help would be appreciated. ==== > > 1. I have a new HP49G+ and would like to transfer variables etc. from > > my old HP49G. I keep getting connect time outs. I am using the Get > > Um, the HP49G does not have an IR port :-) Yep, I forgot about that -- one of the big complaints about it if I remember correctly. I found my old cable and uploaded the ingo using conn4x which worked find. I was also able to get the download to work this time into the G+. It didn't work at first. > > > > > > 2. I am also having trouble getting the USB Conn4x to work. When I > > try to connect I get get a message that says cannot connect to the > > HP49G+, yet, I can do Xmodem transfers and have successfully updated > > the ROM in the calculator. Is there any documentation for the HP SW? > > You're running win98, right? If so, join the club. I just use IrDA under > win98, and the old software. > > Al > > > > > Any help would be appreciated. ==== This week end, I wanted to transfer something to my PC from the 49g+ and suddenly it wasn't working anymore. I reinstalled the connection kit to no avail. Then ON-A-F and it was working again. This is quite bad when you want to upload something. Luckily flash is here. I guess that it is one of those instabilities I have found in the ROM... but I can't say what caused it at the moment. Maybe some connection problems reported here are linked to this. Arnaud ==== In lurking around here I have seen past posts indicating the problem may be a defective USB cable, no personal experience though. I know I have read at least one thread where the OEM cable was bad. I understand the USB cable is nothing special, you can pick one up at any electronics store. KW >This week end, I wanted to transfer something to my PC from the 49g+ >and suddenly it wasn't working anymore. I reinstalled the connection >kit to no avail. Then ON-A-F and it was working again. This is quite >bad when you want to upload something. Luckily flash is here. >I guess that it is one of those instabilities I have found in the >ROM... but I can't say what caused it at the moment. Maybe some >connection problems reported here are linked to this. > >Arnaud > > ==== > In lurking around here I have seen past posts indicating the problem may > be a defective USB cable, no personal experience though. I know I have > read at least one thread where the OEM cable was bad. I understand the > USB cable is nothing special, you can pick one up at any electronics store. > KW I wasn't looking for help in commecting, it was just a statement. Now we can say that connection problems come from either: The PC The cable The calc I hadn't seen any report saying it comes from the calc itself, that is why I posted. Arnaud ==== I never could get USB or connX to work on my XP laptop. It is just plain flaky. Everything needs to be done via SD chip for now until a dependable USB driver and connectviity software can be made dependable across platforms. For some it seems to work, but for others it is useless. Keep watching for new versions. >This week end, I wanted to transfer something to my PC from the 49g+ >and suddenly it wasn't working anymore. I reinstalled the connection >kit to no avail. Then ON-A-F and it was working again. This is quite >bad when you want to upload something. Luckily flash is here. >I guess that it is one of those instabilities I have found in the >ROM... but I can't say what caused it at the moment. Maybe some >connection problems reported here are linked to this. > >Arnaud ==== I've owned a 49G for a few years now, I use it lightly for various purposes (mostly simple numeric and matrix calculations, and smallish UserRPL programs), and I'm weighing the pros and cons of upgrading to a 49G+. I am seeking other people's opinions on whether or not the upgrade is worth it: Pros: Plastic vs rubber keys Faster computation (2-4x on average) Brighter screen, better overall appearence new toy value Cons: No new math functions Much worse battery life (my first set of 49g batteries lasted over 1 year) Poor compatibility with older programs. Don't really care about: SD cards (1 meg flash is way more than enough for me) IR transfer Are there any other factors, either pro or con, that I should take into consideration? With the current balance, I'm just not sure its worth the $140. ==== On Mon, 17 N > Are there any other factors, either pro or con, that I should take into > consideration? With the current balance, I'm just not sure its worth > the $140. It may be because I'm getting to be an old guy (52), but I found the 49G keyboard to be practically unreadable, and the display is pretty bad, too. With the 49G+ I can read the keyboard much more easily, and the display is better, too. - Ed ==== >On Mon, 17 N >> Are there any other factors, either pro or con, that I should take into >> consideration? With the current balance, I'm just not sure its worth >> the $140. > >It may be because I'm getting to be an old guy (52), but I found the 49G >keyboard to be practically unreadable, and the display is pretty bad, >too. With the 49G+ I can read the keyboard much more easily, and the >display is better, too. > >- Ed I agree. At 54, with astigmatism, the 49G screen was very hard for me to read. The choice of background color and the colors chosen for the shifted functions was horrible too, Also not having the ability to change from a five level stack to a four level stack( as in the 48) was a minus too. The 49G+ is much better in this respect. Harold A. Climer Dept.Of Physics,Geology,and Astronomy University of Tennessee at Chattanooga Chattanooga TN USA 37403 ==== > > >On Mon, 17 N > >> Are there any other factors, either pro or con, that I should take into > >> consideration? With the current balance, I'm just not sure its worth > >> the $140. > > > >It may be because I'm getting to be an old guy (52), but I found the 49G > >keyboard to be practically unreadable, and the display is pretty bad, > >too. With the 49G+ I can read the keyboard much more easily, and the > >display is better, too. > > > >- Ed > I agree. At 54, with astigmatism, the 49G screen was very hard for me > to read. The choice of background color and the colors chosen for the > shifted functions was horrible too, Also not having the ability to > change from a five level stack to a four level stack( as in the 48) > was a minus too. The 49G+ is much better in this respect. [EVAL] and tick [ ' ] are back to primary keys now if only the [ENTER].... The small speedup make it snappier and thus so faster the ultimate HP caluclator in user interface speed (using EQW, Plotter, Choose, Inform) ==== it works perfectly on linux :) a friend is now working in a beautiful gui... we accept ideas -- debian gnu/linux 22:07:25 up 1 day, 9:43, 2 users, load average: 0.30, 0.62, 0.49 www.elwater.net ==== > > it works perfectly on linux :) > a friend is now working in a beautiful gui... we accept ideas > I've got it working: bash# modprobe usbserial vendor=0x3f0 product=0x121 Also install module safe-serial (or safeserial or safe_serial, i forgot which) for kernel 2.6.x To send: bash$ sx filename < /dev/ttyUSB0 > /dev/ttyUSB0 To receive: bash$ rx filename < /dev/ttyUSB0 > /dev/ttyUSB0 I've set up some shell scripts so that i have to just `hpsend filename` Please not that # signifies a root console and $ signifies normal user. -MrM ==== escribiÌ°å°Ì[Hyphen][ ARing]°Ì°å[Hyphen]Ì[Hy phen]åÒ: what about upgrade rom program? could it be posible to use the upgrade program used with the 49g? -- debian gnu/linux 14:28:27 up 11:42, 3 users, load average: 0.20, 0.10, 0.03 www.elwater.net ==== > I've owned a 49G for a few years now, I use it lightly for various > purposes (mostly simple numeric and matrix calculations, and smallish > UserRPL programs), and I'm weighing the pros and cons of upgrading to a > 49G+. I am seeking other people's opinions on whether or not the > upgrade is worth it: > > Pros: > Plastic vs rubber keys Gene: Now that the key-press-but-nothing-happens problem really appears to have been fixed, this difference is very important. I can't stand to use my 49G any longer. > Faster computation (2-4x on average) > Brighter screen, better overall appearence Gene: The bigger screen shows 7 stack levels + the header information. This is really nice for big expressions that are on the stack. You see more of them with the headers displayed. > new toy value > > Cons: > No new math functions GENE: True, but the ROM will continue to evolve so that might change. > Much worse battery life (my first set of 49g batteries lasted over 1 year) Gene: Battery life depends on use to an extent. ROM 1.22 removed the really bad battery life bug. > Poor compatibility with older programs. Gene: This depends. I have these libraries installed and they all work fine: EquationLibrary 2.7 (Purdy's), StatPro49, XCELL48, Poker and Yahtzee. > > Don't really care about: > SD cards (1 meg flash is way more than enough for me) > IR transfer > > > Are there any other factors, either pro or con, that I should take into > consideration? With the current balance, I'm just not sure its worth > the $140. > Gene -- * These statements and opinions are mine alone and do not reflect my employer's views. * ==== I have just noticed what may be a bug on the hp-49g+. To duplicate it try the following. 1.) Navigate to an empty directory. If your calculator does not have an empty directory, then create one. You can purge it later. 2.) Place a number on the stack, a 5 for instance. This number is not actually needed, it will only be used as a place holder. 3.) Execute VARS, this will place an empty pair of curly braces on the stack, { } 4.) Place a 1 on the stack. 5.) Place this program on the stack, << DTAG >> 6.) Execute this command, DOLIST 7.) The stack should now contain: 2: 5.00 1: { } 8.) Execute OBJ-> 9.) The stack should now contain: 3: 5.00 2: 1: 1.00 10.) According to the result the 'empty list' contained one item. That one item is on stack level 2. However what is shown on level two is a 'blank' item. The 5 is on level 3. 11.) Normally executing OBJ-> on an empty list results in a zero on level one with the list having been deleted. If the above is tried on an hp-48gx, the execution of DOLIST causes an error. Any comments? rdb. ==== This bug can be reproduced more easily: 5 {} 1. << >> DOLIST OBJ-> That is, run an empty program on an empty list with DOLIST. As in your example, 5 is just a place holder so you can easily see the invisible object created. On my 49g 1.19-5 rom the DOLIST above gives an Invalid Dimension error. On the 49g+ it produces a list with an invisible object of type 8. (Program). It seems to act like an empty program. If EVALuated on the stack, it disappears. ->STR produces an empty string. I have often wished that DOLIST would work on empty lists, but the correct result would be another empty list, not a list with one invisible object. Tom > I have just noticed what may be a bug on the hp-49g+. To duplicate it > try the following. > > > 1.) Navigate to an empty directory. If your calculator does not have > an empty directory, then create one. You can purge it later. > > 2.) Place a number on the stack, a 5 for instance. This number is not > actually needed, it will only be used as a place holder. > > 3.) Execute VARS, this will place an empty pair of curly braces on the > stack, { } > > 4.) Place a 1 on the stack. > > 5.) Place this program on the stack, << DTAG >> > > 6.) Execute this command, DOLIST > > 7.) The stack should now contain: > > 2: 5.00 > 1: { } > > 8.) Execute OBJ-> > > 9.) The stack should now contain: > > 3: 5.00 > 2: > 1: 1.00 > > 10.) According to the result the 'empty list' contained one item. That > one item is on stack level 2. However what is shown on level two is a > 'blank' item. The 5 is on level 3. > > 11.) Normally executing OBJ-> on an empty list results in a zero on > level one with the list having been deleted. > > > If the above is tried on an hp-48gx, the execution of DOLIST causes an > error. > > Any comments? > > rdb. ==== > ... run an empty program on an empty list with DOLIST. > > On the 49g+ it produces a list with an invisible object of type 8. > (Program). It seems to act like an empty program. If EVALuated on the > stack, it disappears. ->STR produces an empty string. > In fact the invisible object seems to be identical to the object created by {} ->PRG in the hackers menu. ==== > > ... run an empty program on an empty list with DOLIST. > > > > On the 49g+ it produces a list with an invisible object of type 8. > > (Program). It seems to act like an empty program. If EVALuated on the > > stack, it disappears. ->STR produces an empty string. > In fact the invisible object seems to be identical to the object created > by {} ->PRG in the hackers menu. Tom, OT49 was made to help the normal user in such a situation. The odd object in the empty list is PTR 05607 and this is the empty SysRPL program. You will see it on the stack if applying 3tog on the empty looking list, provided flag -85 is set and extable is on your 49+. That << {} 1 << >> DOLIST >> creates on the HP49+ a list containing the empty SysRPL program is definitely a bug. An additional example that 1.22 is only partly backward compatible with the HP49 at the user level. Logic demands that applying DOLIST with anything on the empty list should yield the empty list (as does DoL from OT49), or errors out provided backward compatibility to the 49 is taken seriously. - Wolfgang http://page.mi.fu-berlin.de/~raut/WR49/OT49.htm ==== > Tom, OT49 was made to help the normal user in such a situation. The odd > object in the empty list is PTR 05607 and this is the empty SysRPL > program. You will see it on the stack if applying 3tog on the empty > looking list, provided flag -85 is set and extable is on your 49+. > Logic demands that applying DOLIST with anything on the empty list > should yield the empty list (as does DoL from OT49), or errors out > provided backward compatibility to the 49 is taken seriously. I assume the empty program object will be banished soon, probably going back to the 48g/49g behavior of erroring out when confronted with empty list input. But, if DOLIST were made to handle an empty list properly on the 49+ (returning an empty list result), would that really be an impermissible breach of backward compatibility? I would see it as an improvement with little downside, although maybe I'm missing something. Tom ==== >Instruct students to use the STO key and store intermediate decimal values and >only round at the end to the required place... >Eva Demyen >NYSED Scoring Guide and Math B Committees The New York Department of Education has this policy for high school exams. They are crying for RPN. I am thankfull I have an RPN calc (HP-49g plus) for the state exams, or else I would have to use store my intermediate results using the mery key registars. On anther topic New York bans the TI-89 on state exams but allows the HP-49g plus. ==== > On anther topic New York bans the TI-89 on state exams but allows the HP-49g plus. Maybe because they aren't aware of the existence of the HP49G+? -- Bhuvanesh ==== > Maybe because they aren't aware of the existence of the HP49G+? There in no mention of the older 49g in state exam instuctions as a banned calculator. I know it's ue is against the spirit of the tests'; but it does break the rules. ==== >On anther topic New York bans the TI-89 on state exams but allows the HP-49g plus. So how long do you think it will be till it is banned? -Samuel ==== Anyone knows if such a guide is in preparation? I mean, for sure the new guides for the HP49G+ are better than the User Guide available for the HP49G ( I think I can say they are MUCH better), but it will be also useful to have an Advanced User Guide like the one for the 49 with ALL commands (either CAS ones and not) alphabetically listed (I personally prefer a total aphabetic indicization, without distinguishing between CAS functions and other commands as it is now). Then it will be good to have an index either alphabetic and by function type And there will be a Pocket Guide? PS: If someone is doing something like this, even at hp, I offer my free colaboration. Kickaha -- Per rispondere rimuovere il SiPAriuM To reply remove the SiPAriuM ==== Apropos the subject of the HP49G+ manual - in particular a printed copy... I went to a print store in town today (near Toronto, Ontario) and asked the cost of a black and white single copy of an 865 page document. I was quoted 30 cents a page - so that would come to about 230 Canadian Dollars, plus tax. What a bargain, eh? Not. Mark. > Anyone knows if such a guide is in preparation? I mean, for sure the new > guides for the HP49G+ are better than the User Guide available for the HP49G > ( I think I can say they are MUCH better), but it will be also useful to > have an Advanced User Guide like the one for the 49 with ALL commands > (either CAS ones and not) alphabetically listed (I personally prefer a total > aphabetic indicization, without distinguishing between CAS functions and > other commands as it is now). Then it will be good to have an index either > alphabetic and by function type > And there will be a Pocket Guide? > > PS: If someone is doing something like this, even at hp, I offer my free > colaboration. > Kickaha > > -- > Per rispondere rimuovere il SiPAriuM > To reply remove the SiPAriuM > > ==== >Apropos the subject of the HP49G+ manual - in particular a printed copy... > >I went to a print store in town today (near Toronto, Ontario) and asked the >cost of a black and white single copy of an 865 page document. I was quoted >30 cents a page - so that would come to about 230 Canadian Dollars, plus >tax. What a bargain, eh? Not. > >Mark. > >> Anyone knows if such a guide is in preparation? I mean, for sure the new >> guides for the HP49G+ are better than the User Guide available for the >HP49G >> ( I think I can say they are MUCH better), but it will be also useful to >> have an Advanced User Guide like the one for the 49 with ALL commands >> (either CAS ones and not) alphabetically listed (I personally prefer a >total >> aphabetic indicization, without distinguishing between CAS functions and >> other commands as it is now). Then it will be good to have an index either >> alphabetic and by function type >> And there will be a Pocket Guide? >> >> PS: If someone is doing something like this, even at hp, I offer my free >> colaboration. >> Kickaha >> >> -- >> Per rispondere rimuovere il SiPAriuM >> To reply remove the SiPAriuM >> >> > Do they have Kinkos in Canada? Here in Chattanooga TN it is about 6 cents a page and usually for large documents you can negotiate a slightly lower price. However if you ever need a document scanned to PDF or Word ,and have you own scanner, do it yourself. Their first quote was $9.00 a page. When I told them it was a 64 page document, they kindly lowered it to $2.00 a page, This price was if you provided your own CD-RW or CD-R disk. Harold A. Climer Dept.Of Physics,Geology,and Astronomy University of Tennessee at Chattanooga Chattanooga TN USA 37403 ==== I printed mine in landscape mode, two pages per side. With duplex printing, you get 4 pages for the price of one. Also, when the manual lies flat on your desk, you can see four pages of text side by side. I printed it at work (free), and had Kinko's bind it with a large coil for less than $5.00. > > >Apropos the subject of the HP49G+ manual - in particular a printed copy... > > > >I went to a print store in town today (near Toronto, Ontario) and asked the > >cost of a black and white single copy of an 865 page document. I was quoted > >30 cents a page - so that would come to about 230 Canadian Dollars, plus > >tax. What a bargain, eh? Not. > > > >Mark. > > > >> Anyone knows if such a guide is in preparation? I mean, for sure the new > >> guides for the HP49G+ are better than the User Guide available for the > >HP49G > >> ( I think I can say they are MUCH better), but it will be also useful to > >> have an Advanced User Guide like the one for the 49 with ALL commands > >> (either CAS ones and not) alphabetically listed (I personally prefer a > >total > >> aphabetic indicization, without distinguishing between CAS functions and > >> other commands as it is now). Then it will be good to have an index either > >> alphabetic and by function type > >> And there will be a Pocket Guide? > >> > >> PS: If someone is doing something like this, even at hp, I offer my free > >> colaboration. > >> Kickaha > >> > >> -- > >> Per rispondere rimuovere il SiPAriuM > >> To reply remove the SiPAriuM > >> > >> > > > Do they have Kinkos in Canada? Here in Chattanooga TN it is about 6 > cents a page and usually for large documents you can negotiate a > slightly lower price. However if you ever need a document scanned to > PDF or Word ,and have you own scanner, do it yourself. Their first > quote was $9.00 a page. When I told them it was a 64 page document, > they kindly lowered it to $2.00 a page, This price was if you provided > your own CD-RW or CD-R disk. > Harold A. Climer > Dept.Of Physics,Geology,and Astronomy > University of Tennessee at Chattanooga > Chattanooga TN USA 37403 ==== --------------------------------------------------------------------- This is the messsage that my laptop shows when it detects IR from my hp 49G+. Hmmm... Toby ==== I am new to saturn assembler, so I apologize if this is too elementary a question. I read various places about having your code object wiped by a garbage collection when you run from a library in a covered port. Under what circumstances does this happen? Does this only happen if you have a merged memory card? I am guessing that if I don't do anything that will call for some free memory or moves memory around, then I should be safe. But some of my routines return to RPL mode leaving a new BINT on the stack. They are structured something like: :: blah blah CODE blah blah GOSBVL =PUSH#ALOOP ENDCODE blah blah ; Would PUSH#ALOOP being the last instruction before returning to RPL be spared should a garbage collection take place to make room for the BINT? ==== > I am new to saturn assembler, so I apologize if this is too elementary > a question. I read various places about having your code object wiped > by a garbage collection when you run from a library in a covered port. > Under what circumstances does this happen? Does this only happen if > you have a merged memory card? I am guessing that if I don't do > anything that will call for some free memory or moves memory around, > then I should be safe. But some of my routines return to RPL mode > leaving a new BINT on the stack. They are structured something like: > > :: > blah > blah > CODE > blah > blah > GOSBVL =PUSH#ALOOP > ENDCODE > blah > blah > ; > > Would PUSH#ALOOP being the last instruction before returning to RPL be > spared should a garbage collection take place to make room for the > BINT? > Yes, that is safe. The gc, if needed, will be called from ROM and the LOOP part will return to SysRPL code correctly because the RPL pointers are saved and updated by the system in the event of a gc. Remark that GOSBVL =PUSH#A GOVLNG =LOOP is *not* safe. It's only when a Push is combined with a LOOP that it is. ==== I am taking calculus, and I was asked to graph: ((x^2-1)^2)^(1/3) I graphed it and thought it was continuous everywhere. I tried it on my hp49g to check. I typed it in with RPN, it simplified, and I stored to EQ and then number is raised to the 2/3 which should be equal to squaring and then raising to 1/3? I then tried on ti-89. It does the same as the 49 in complex mode. In real mode, (-1)^(2/3)=1 and it graphs it everywhere. Does ((x^2-1)^2)^(1/3)=(x^2-1)^(2/3)? If they are not always equal, why do CAS systems simplfy? If I type it in unsimplified in the plot equation editer I get a continuos graph. Will someone please explain the math to me or direct me to a good (-1)^(2/3)=-1/2+(3^.5/2)i instead of 1. I assume they are doing the ^(1/3) before the ^2 but why? Also I have one algebra question. With imaginary numbers, I know (-1)^.5=i, and know how to factor i out of other square roots. So ((-1)^(1/3))^2=(i^(2/3))^2, but what do I do from here? an algebra question in addition to a calculator question. ==== I don't know what you need it for but how about accepting it as a complex number function and go for the Parametric Plot: X(t)+i*Y(t) > I am taking calculus, and I was asked to graph: ((x^2-1)^2)^(1/3) > I graphed it and thought it was continuous everywhere. > > I tried it on my hp49g to check. > I typed it in with RPN, it simplified, and I stored to EQ and then > number is raised to the 2/3 which should be equal to squaring and then > raising to 1/3? > > I then tried on ti-89. > It does the same as the 49 in complex mode. In real mode, (-1)^(2/3)=1 > and it graphs it everywhere. > > Does ((x^2-1)^2)^(1/3)=(x^2-1)^(2/3)? If they are not always equal, > why do CAS systems simplfy? If I type it in unsimplified in the plot > equation editer I get a continuos graph. > > Will someone please explain the math to me or direct me to a good > (-1)^(2/3)=-1/2+(3^.5/2)i instead of 1. I assume they are doing the > ^(1/3) before the ^2 but why? > > Also I have one algebra question. With imaginary numbers, I know > (-1)^.5=i, and know how to factor i out of other square roots. So > ((-1)^(1/3))^2=(i^(2/3))^2, but what do I do from here? > > an algebra question in addition to a calculator question. ==== Why do some commands in the catalog have help available (such as LOCAL) and most don't? -- Tom Lake Experience keeps a dear school but fools will learn in no other - Poor Richard's Almanack ==== TL> Why do some commands in the catalog have help available (such as LOCAL) and TL> most don't? Only CAS commands have help. LOCAL is a CAS command. - Carsten ==== > > TL> Why do some commands in the catalog have help available (such as LOCAL) and > TL> most don't? > > > Only CAS commands have help. LOCAL is a CAS command. > > - Carsten Hmm? Local variables don't have much to do with the CAS. I would classify LOCAL as a programming construct. -- Bhuvanesh ==== >> Only CAS commands have help. LOCAL is a CAS command. >> >> - Carsten lb> Hmm? Local variables don't have much to do with the CAS. I would lb> classify LOCAL as a programming construct. Sure. But it has been introduced into th HP49G as part of the CAS. Thats why there is help for it. - Carsten lb> -- lb> Bhuvanesh ==== > > Only CAS commands have help. LOCAL is a CAS command. > Hmm? Local variables don't have much to do with the CAS. I would > classify LOCAL as a programming construct. If you don't believe that LOCAL is a CAS command run 222 LofRP from Libman below. That puts you the huge list of 467 visible and unvisible CAS rompointers on the stack. Explode this list and go in the interactive stack to position 408. There is our LOCAL :-) - Wolfgang http://page.mi.fu-berlin.de/~raut/WR49/Libman.htm ==== > If you don't believe that LOCAL is a CAS command run 222 LofRP from > Libman below. That puts you the huge list of 467 visible and unvisible > CAS rompointers on the stack. Explode this list and go in the > interactive stack to position 408. There is our LOCAL :-) > Despite not being a CAS command, LOCAL is indeed part of the CAS source code. It was created at the request of Renee de Graeve to be able to create more than one local variable at a time in algebraic mode (variables that are not arguments). ==== >> If you don't believe that LOCAL is a CAS command run 222 LofRP from >> Libman below. That puts you the huge list of 467 visible and unvisible >> CAS rompointers on the stack. Explode this list and go in the >> interactive stack to position 408. There is our LOCAL :-) >> > >Despite not being a CAS command, LOCAL is indeed part of the CAS >source code. It was created at the request of Renee de Graeve >to be able to create more than one local variable at a time >in algebraic mode (variables that are not arguments). Excuse my ignorance, but is the idea of a local variable for the 49/49+ similar to a local variable in C++ and JAVA. with which I am familiar? Does the idea of scope also apply to the local variable? Harold A. Climer Dept.Of Physics,Geology,and Astronomy University of Tennessee at Chattanooga Chattanooga TN USA 37403 ==== Tom, It is a very good question, and I am sure lots of people will ask it when they try to use the HELP facility on the 49G+. The non-CAS commands are covered by the 48G AUG (it refers to this in the printed User Manual), but it would seem to me that with the greater memory capacity, improved screen and better speed - noticeable when scrolling through the CATalog[ue], it would be ideal to have help available on every command. It is definitely achievable, but would some effort of course. I have been playing with my 49G+ over the last few days and I feel that it really is a significant step forward from the 49G. I'm not quite happy with the keys, but I prefer the machine to the 49G, that's for sure. Mark. > > TL> Why do some commands in the catalog have help available (such as LOCAL) and > TL> most don't? > > > Only CAS commands have help. LOCAL is a CAS command. > > - Carsten ==== > The non-CAS commands are covered by the 48G AUG (it refers to this in the > printed User Manual), but it would seem to me that with the greater memory > capacity, improved screen and better speed - noticeable when scrolling > through the CATalog[ue], it would be ideal to have help available on every > command. It is definitely achievable, but would some effort of course. Is it? How much effort? What is your proposed strategy? Not better thinking before posting? The problem is that a library's $EXTPRG handles in message handler 9 only the rompointers of its own lib. These then may or may not get a HELP text by means of message handler 10. But UsrRPL commands are rompointers from different internal libs. Thus, I see no other way as to recreate these libs and adding a SEXTPRG to each of it. But this may have serious consequences for speed ... For CAS this was no problem because the 65 visible CAS commands live all in the same house, library 222. - Wolfgang ==== >> The non-CAS commands are covered by the 48G AUG (it refers to this in the >> printed User Manual), but it would seem to me that with the greater memory >> capacity, improved screen and better speed - noticeable when scrolling >> through the CATalog[ue], it would be ideal to have help available on every >> command. It is definitely achievable, but would some effort of course. WR> Is it? How much effort? What is your proposed strategy? Not better WR> thinking before posting? How is there any thinking missing? Of course this is feasible. It may only be difficult to integrate this help directly into the command browser of the catalog. However, one could rewrite the CAT to make this possible, or one could use SDIAG in order to access information from the command line. Main problem with SDIAG is that it is already big, so in order to put additional information, one would have to change the library, probably making it access files on the SD card instead of library rompointers. - Carsten ==== > > The non-CAS commands are covered by the 48G AUG (it refers to this in the > > printed User Manual), but it would seem to me that with the greater memory > > capacity, improved screen and better speed - noticeable when scrolling > > through the CATalog[ue], it would be ideal to have help available on every > > command. It is definitely achievable, but would some effort of course. Use Emacs and SDIAG from Dominik: http://zon.astro.uva.nl/~dominik/hpcalc/emacs/emacs112.zip The SDIAG version without ML and SysRPL is smaller ==== > Use Emacs and SDIAG from Dominik: thankful to the authors of SDIAG for that). But this is clearly less than CAS HELP in general. If SDIAG's { -> } option would comtain here and there some additional information I would be completely happy. There is another more serious problem with the CAT display. Rompointers from external libs are displayed in italics and are simply unreadable if choose boxes are used in minifont. If would be much better and more informative to print these rompointers names normally and to add after the name just the lib ID of the corresponding external library. - Wolfgang ==== > > I posted this originally in the TI-89/92+/V200 discussion group. I'm > > posting this here also, in case Dr. Parisse or someone else can > > explain this. > > > > I'm curious: Isn't it possible to factor a (univariate) polynomial > > modulo a composite integer? I know one can find roots mod a composite > > integer by finding roots mod each of its constituent primes and then > > using Hensel lifting (to make the roots correct mod p^k) followed by > > the Chinese Remainder algorithm, but how would one find the > > multiplicities? > > > > hmm, actually it is possible to lift a root of a polynomial > from Z/p to Z/p^k, but not with different primes. For example > P:=x^7+x^3+1 has one root modulo 17 but none modulo 11. > The same happens for higher order factors. This is used > to compute a table of possible degrees for factors in > factoring programs using distinct degree factorization > and a few primes. > > > Am I missing something obvious? > > > > Also, as an add-on to my original post, is it possible to determine > > how many roots a polynomial should have modulo an integer p (possibly > > composite) without doing the actual factorization? > > > Yes, for example using xcas > P:=x^7+x^3+1; > p:=11; > gcd(P%p,(x^p-x)%p) > p:=17; > gcd(P%p,(x^p-x)%p); > > On the 49 > x^7+x^3+1 STO> P > MODSTO(11) > GCDMOD(P,X^11-X) > MODSTO(17) > GCDMOD(P,X^17-X) > > Hmmm, it is probably a bit harder to test this on the TI89, > maybe using mathtool? Yes, one can do this using the PolyGCD and PolyMod functions from MathTools. Bhuvanesh. ==== On the TI-89, when in trace mode, you can enter a value for X. Is there a way to do so on the HP 49G+? -- Bill ==== > On the TI-89, when in trace mode, you can enter a value for X. Is there > a way to do so on the HP 49G+? There is a debug in the 48 and 49 and 49+. Is that what you mean by trace? It allows single stepping through programs and their subroutines ( in RPL mode at least for the 49s), and you can do anything you like between steps. ==== > > > On the TI-89, when in trace mode, you can enter a value for X. Is there > > a way to do so on the HP 49G+? > > > There is a debug in the 48 and 49 and 49+. Is that what you mean by > trace? > > It allows single stepping through programs and their subroutines ( > in RPL mode at least for the 49s), and you can do anything you like > between steps. No, Bill means tracing a plot and jumping to the point (x,f(x)) after you enter a value for x. -- Bhuvanesh ==== I believe you are referring to 2-D function graphing, correct? On the TI-89, you may key in a value for X, and your cursor will move to that point, showing you the Y-value for that X. The 49x doesn't do that, but you can easily switch to the table view to obtain this data. The table also gives you a good idea of the behavior of the function near your point. > > > On the TI-89, when in trace mode, you can enter a value for X. Is there > > a way to do so on the HP 49G+? > > > There is a debug in the 48 and 49 and 49+. Is that what you mean by > trace? > > It allows single stepping through programs and their subroutines ( > in RPL mode at least for the 49s), and you can do anything you like > between steps. ==== > I believe you are referring to 2-D function graphing, correct? On the > TI-89, you may key in a value for X, and your cursor will move to that > point, showing you the Y-value for that X. > > The 49x doesn't do that, but you can easily switch to the table view to > obtain this data. The table also gives you a good idea of the behavior of > the function near your point. > Now if only the 49 and/or 49+ allowed one to export table values to the stack somehow! ==== I just bought a new 12C and I noticed there is no serial number on it. Does it mean they sold me a refurbished unit? ==== > I just bought a new 12C and I noticed there is no serial number on it. Does > it mean they sold me a refurbished unit? Rather someone at the hp forgot to mark the serial to a new other cell A collectible item. Keep it. ==== I took a cursory glance in Future Shop yesterday at San Disk cards, with the intention of getting one for my new 49G+. They had them in only two capacities, 128MB and 256MB which were 99 and 129 Canadian Dollars respectively. I did see a 1GB unit but it looked like it was a larger piece, and I wasn't sure if it would fit the calculator slot. I don't think I was confuing it with a different kind of memory card, but it's possible. Any advice or guidelines from those in the know, especially those who have bought some SD RAM already? Mark. ==== > I took a cursory glance in Future Shop yesterday at San Disk cards, with the > intention of getting one for my new 49G+. They had them in only two > capacities, 128MB and 256MB which were 99 and 129 Canadian Dollars > respectively. I did see a 1GB unit but it looked like it was a larger > piece, and I wasn't sure if it would fit the calculator slot. I don't think > I was confuing it with a different kind of memory card, but it's possible. > > Any advice or guidelines from those in the know, especially those who have > bought some SD RAM already? The memory type is Flash, not RAM Sometimes we refer the 49G(+) ROM are as FlashROM and the user area as FlashRAM or simply flash I tried to refer the SD card as external Flash and the FlashRAM as internal Flash but that seemed to be confusing. SD cards are not searched for libraries thus working (eg. attached) libraries much reside inside the calculator either in FlashRAM port 3, or in lower (RAM) ports. Since the HP 49G+ supports both SD/MMC cards and apparently any size using DOS FAT16 format, it's upt o you which size to use. I currently have one 64MB MMC in one 49g+ and a 8MB MMC in the other The smaller card is sufficient for both the original Nokia 9110i Communicator and the hp 49g+ and there is still room for large text file and some htm files If you plan to fit digicamera pictures in, get a CF flash model instead and use the largest possible (4GB?) size of Compact Flash. For caluclator usage only, use the cheapest. You may also consider a reader for your card. a small USB SD reader migh be a good idea, unless you need those pictures in which case you want to look Sundisk 6-in-1 or similar devices with multi-slots. Warning, if you manage to by a 512MB SD card it might be formatted to FAT32 and dies not work unless you format it to FAT16, which can be done in your 49g+ More questions? ==== HP Xpander for sale on ebay http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3059419223&category=3734 Mint condition complete and boxed. UK sale but ship worldwide ==== I've seen in various places that the HP49G+ runs ROM 1.22. Software: 49G ROM 1.22 with bugfixes and some minor added features http://www.hpcalc.org/hp49gplus.php Can the 49G be upgraded to 1.22? Or is 1.19-6 the highest I can go with the 49G? Jeff ==== mmbjeff ha scritto nel messaggio > I've seen in various places that the HP49G+ runs ROM 1.22. (SNIP) > Can the 49G be upgraded to 1.22? Or is 1.19-6 the highest I can go with the 49G? Yes, it is the highest you can have..... now Kickaha ==== I use my hp49g for four years now and the hp49g+ is on his way by my problem is that i dont know how to program in UserRPL and the examples are not so good. so i guess that if i will challenge you to write for me some programs (very simple ones) so i could learn the syntax and go on from that point. 1. so the first program should do this : take a function which look like this : f(t) = C1*sin(wt)+C2cos(wt)+Ysin(wt) now it takes initial cindition as : f(t=0)=0 and df/dt(t=0)=0 and finds the conficians c1 and c2. it is only an example for me so i can learn a bit. you dont have to do it for any number of conficians - just c1 and c2. 2. another nice program that will be helpful to me is that if i could enter a function and the hp will calculate symbolically the deruvation till 4 derivation but i want it to display them all. for example : the function is : x^3 the display should be : 3x^2 6x 6 0 it would be better if the program will ask me how many times to derivate the function. nad to what variable to derivate. everything of couse after simplification. it will really help me a lot cause i really dont know where to start. thanks very much Idan ==== you'll have now four port entries at the end of the LIB menu on key 93.3, :0; :1: :2: :3: - the last one is just an SD-card browser similar to the built-in Port3 browser on the filer. This is the best one can have. You should read at least the Fast Instruction for a first Use before customizing your LIB menu. It may be that in the next ROM for the HP49+ this function will be built in. In this case I could regain the 70 bytes I spent for this new functionality :-) Have fun Wolfgang http://page.mi.fu-berlin.de/~raut/WR49/Libman.htm ==== > > you'll have now four port entries at the end of the LIB menu on key > 93.3, :0; :1: :2: :3: - the last one is just an SD-card browser similar > to the built-in Port3 browser on the filer. This is the best one can > have. > > You should read at least the Fast Instruction for a first Use before > customizing your LIB menu. It may be that in the next ROM for the HP49+ > this function will be built in. In this case I could regain the 70 bytes > I spent for this new functionality :-) > > Have fun > Wolfgang > http://page.mi.fu-berlin.de/~raut/WR49/Libman.htm Right! Now how to get the LFont? The URL is broken...better check all links... ==== I have discovered what may be an anomaly with 'CR' on the HP-49G+. In order to demonstrate this anomaly I will describe six different procedures, please bear with me. You may want to hold your 49g+ close to your ear while you do the following. 1.) Enter CR then press and hold the enter key for about a second. Did the calculator beep? 2. Enter CR then press and release the enter key as quickly as possible. Did the calculator Beep? 3.) Open the CATalog and scroll to CR. Press and hold the OK key for about a second. Did the calculator beep? 4.) Again find CR in the catalog. This time press and release the OK key as quickly as possible. Did the calculator beep? 5.) Create the following program and duplicate it a number of times on the stack: << CR >> Now press and hold the EVAL key for about a second. Did the calculator beep? 6.) Then press and release the EVAL key as quickly as possible. Did the calculator beep? 7.) The answers I get are: Yes, No, Yes, No, Yes & Yes. Should the calculator really beep when it executes a CR? And in any event why does it react differently depending on how the executing key press is applied? rdb. ==== hi! me and two friends are working in a gui for transfering files between and hp(i own 48gx,49g and 49g+ and it would work with all of them) first of all using kermit and if everything works ok maybe also using xmodem. i'm not an expert with kermit so any help would be apreciated.... how it's posible to send the list of vars in the current directory in the calc to the computer? i've tried in my calc 3:D 2:G 1:PKT and puting kermit in recieve put it's not possible to recive anithing does anybody have any idea? thanks! :) -- debian gnu/linux 22:45:53 up 20:00, 4 users, load average: 0.42, 0.36, 0.24 www.elwater.net ==== I only have a little experience with Kermit, but from what I know, it must be in Server mode, and must have CRC-3 checking enabled. BTW, did you write the kermit interface youself? If so is there any chance that I could get hold of the source for a similar project on the the 38/39/40G's. If not, M@ > hi! > > me and two friends are working in a gui for transfering files between and > hp(i own 48gx,49g and 49g+ and it would work with all of them) first of > all using kermit and if everything works ok maybe also using xmodem. > > i'm not an expert with kermit so any help would be apreciated.... > > how it's posible to send the list of vars in the current directory in the > calc to the computer? i've tried in my calc > 3:D > 2:G > 1:PKT > and puting kermit in recieve put it's not possible to recive anithing > > does anybody have any idea? > > thanks! :) > > -- > debian gnu/linux > 22:45:53 up 20:00, 4 users, load average: 0.42, 0.36, 0.24 > www.elwater.net > ==== > I only have a little experience with Kermit, but from what I know, it must > be in Server mode, and must have CRC-3 checking enabled. i put the calc in server mode and the computer in client mode > BTW, did you write the kermit interface youself? no. we want to use the gkermit o ckermit packages from debian >If so is there any chance that I could > get hold of the source for a similar project on the the 38/39/40G's. If > advance. if 38/39/40G's work in the same way that 4xg i think it would be easy to make it work. -- debian gnu/linux 00:39:34 up 21:53, 3 users, load average: 0.48, 0.31, 0.25 www.elwater.net ==== If I read this right, my hp49g+ is going to calculate x and y rectangular spherical coordinatesexactly the same: x = y = p*sin(phi) *cos(theta) Is this one of those, don't believe everything you read, tricks? I could get an inferiority complex trying to learn from this document. How's a kid supposed to learn anything from elders like this? I guess the old RTFM is taking a beating with this guide. Comments? ==== I use my hp49g for four years now and the hp49g+ is on his way by my problem is that i dont know how to program in UserRPL and the examples are not so good. so i guess that if i will challenge you to write for me some programs (very simple ones) so i could learn the syntax and go on from that point. 1. so the first program should do this : take a function which look like this : f(t) = C1*sin(wt)+C2cos(wt)+Ysin(wt) now it takes initial cindition as : f(t=0)=0 and df/dt(t=0)=0 and finds the conficians c1 and c2. it is only an example for me so i can learn a bit. you dont have to do it for any number of conficians - just c1 and c2. 2. another nice program that will be helpful to me is that if i could enter a function and the hp will calculate symbolically the deruvation till 4 derivation but i want it to display them all. for example : the function is : x^3 the display should be : 3x^2 6x 6 0 it would be better if the program will ask me how many times to derivate the function. nad to what variable to derivate. everything of couse after simplification. it will really help me a lot cause i really dont know where to start. thanks very much Idan