B442 === Subject: Re: HPGCC installation issues on w98 w/ cygwin > HPGCC seems like it mixing versions between cygwin and itself and > getting confused. > I tried removing cygwin1.dll from arm-hp/bin but then it could not find the headers (and more). I thought about fiddling with the paths in the hpgcc.bat. > I just struggled with this a few days ago. You need to copy cygwin1.dll > from the cygwin installation into the hpgcc installation's bin > to that effect which I found with Google, and it works. > Here's some of the output: make hiworld.hp > arm-elf-gcc -mtune=arm920t -mcpu=arm920t -mlittle-endian > -fomit-frame-pointer -Wall -Os -Iinclude -Llib -mthumb-interwork > -mthumb -c hiworld.c > hiworld.c:13:43: error: hpgcc49.h: No such file or directory > hiworld.c: In function 'main': > hiworld.c:17: warning: implicit declaration of function 'clear_screen' > hiworld.c:19: warning: implicit declaration of function 'printf' > hiworld.c:19: warning: incompatible implicit declaration of built-in > function 'printf' > hiworld.c:21: warning: implicit declaration of function 'keyb_isON' > make: *** [hiworld.o] Error 1 Process Exit Code: 2 > Time Taken: 00:04 I noticed that the 'options/tools/project tool' list (command arguments > for make.exe and more) is blank on this machine but not on the w2k > ones. Perhaps I should copy these over? I placed set HPGCC=c:arm-hp in autoexec.bat it worked. === Subject: Re: 4 batteries vs. 3 batteries > Has anyone tried using NiMH (or NiCAD) cells in a 50g yet? That's all I use, having killed the included alkalines long ago. I use >1000 mAh cells from http://www.amondotech.com/ They work great. -Joe- The high turn-on voltage for the low battery indicator on the HP50 may be a good thing when using NiMH or NiCAD batteries. http://data.energizer.com/PDFs/nh12-850.pdf If the low battery indicator turned on at slightly more than 1 volt per cell (which is good when using alkalines), the NiMH cells would be on a very steep part of the discharge curve, only minutes away from exhaustion. But the HP50's low battery indicator comes on at 1.2125 volts per cell (my HP50 does so; I don't know about any others). Eyeballing the NiMH discharge curve, this looks to me like the cells would be about 70% discharged, so there is ample time to change them. === Subject: How to keep PI in symbolic form in a vector, a matrix or a complex ? I desperatly try to insert PI (the formal expression, not the approached value) in a vector or a matrix or a complex number, with no success (neither on 48G nor on 49g+ latest rom). I can bring some other formal expressions in, but PI resists. I tried several flags (exact mode on...), no success. Would any one have any idea ? Herve === Subject: Re: How to keep PI in symbolic form in a vector, a matrix or a complex ? I desperatly try to insert PI (the formal expression, not the approached > value) in a vector or a matrix or a complex number, with no success > (neither on 48G nor on 49g+ latest rom). I can bring some other formal > expressions in, but PI resists. I tried several flags (exact mode > on...), no success. Would any one have any idea ? Herve My RCLF command returns { # 300001028D210FFh # 0h # D8776302020428h #0h } Putting the symbol for pi, which I shall denote by pi, into a matrix with the matrixwriter works every time. So does pi and 1 on the stack and then using the ->ARRY command to make an array. What does fail is trying to do it with the [] key (at left_shift times) and then entering the pi symbol anywhere between the brackets. So I suggest either (1) using the matrixwriter (left shift ' ) or (2) entering all your matrix entries, then a list of one or two reals to define the desired size and shape of the matrix and then executing ->ARRY. For example, pi pi pi pi {2 2} ->ARRY should give you a 2 by 2 matrix with pi in each position. To get a symbolic pi into a complex number, I think you must be in exact mode. Then put the desired real an imaginary parts on the stack followed by the imaginary unit, i, on the stack, then do a times, *, followed by an add , +, and you will have it. The sequence pi pi i * + , for example, in exact mode, should result in 'pi+i*pi'. === Subject: Re: How to keep PI in symbolic form in a vector, a matrix or a complex ? What does fail is trying to do it with the [] key (at left_shift times) > and then entering the pi symbol anywhere between the brackets. For this you will need to surround PI with ' ' This is a system issue: PI is a RPL keyword which you can't add to a matrix, you can however add algebraic JY === Subject: A few more questions in How to keep PI in symbolic form in a vector, a matrix or a complex ? - what is the general signification of the - in there any way to access the more directly than through the CHARS table ? - in which manual can I find the complete liste of object types ? - and probably the most boring one : now that I have my nice [PI 1] array, how can I switch back to the numerical value ? For example, if I enter [PI 1] ABS, then I normally get SQRT(PIO +1), but fom then EVAL does nothing and ->NUM returns an Undefined Name error... Herve Jean-Yves Avenard a .8ecrit : >> What does fail is trying to do it with the [] key (at left_shift >> times) and then entering the pi symbol anywhere between the brackets. > For this you will need to surround PI with ' ' This is a system issue: PI is a RPL keyword which you can't add to a > matrix, you can however add algebraic > JY === Subject: Re: A few more questions in How to keep PI in symbolic form in a vector, a matrix or a complex ? > - what is the general signification of the > - in there any way to access the more directly than through the CHARS > table ? The backslash indicates that what follows in ASCII is different from what will appear as special symbols in the HP but without the backslash. When HP files are transferred to a desktop as text, those HP symbols not in the ASCII alphabet are represented by various combinations preceded by the backslash. So when we type pi here, it stands for the symbol for pi, but without a backslash in the HP. Similarly << and >>, for the program delimiters. Does anyone know where one can find the list of all the ASCII backslash codes for the HP characters? Not the numerical codes as one can get them from the HP itself, but the ones you get when transferring HP files to text files on the PC. === Subject: Re: A few more questions in How to keep PI in symbolic form in a vector, a matrix or a complex ? > - what is the general signification of the None... Just used when transferring files between the PC and the calculator. So you can type something like this: This is a quote people typing here pi is just a way to indicate the pi refers to the translation code used by Kermit, and this is equivalent to typing PI (the actual greek letter: Left Shift SPACE) on the calculator. You're not supposed to type here > - in there any way to access the more directly than through the CHARS > table ? CHARS will give you the equivalent shortcut, in this case: Alpha RS 5 > - in which manual can I find the complete liste of object types ? > - and probably the most boring one : now that I have my nice [PI 1] > array, how can I switch back to the numerical value ? ??? [ 'PI' 1 ] ->NUM [3.1415.... 1] > For example, if I enter [PI 1] ABS, then I normally get SQRT(PIO +1), > but fom then EVAL does nothing and ->NUM returns an Undefined Name error... > What do you want to use ??? JY === Subject: Re: A few more questions in How to keep PI in symbolic form in a vector, a matrix or a complex ? Got it. I was entering the on the calc itself... Now it works. Many thank's to all ! Herve Jean-Yves Avenard a .8ecrit : >> - what is the general signification of the > None... > Just used when transferring files between the PC and the calculator. > So you can type something like this: > This is a quote people typing here pi is just a way to indicate the pi refers to the > translation code used by Kermit, and this is equivalent to typing PI > (the actual greek letter: Left Shift SPACE) on the calculator. > You're not supposed to type here > - in there any way to access the more directly than through the >> CHARS table ? > CHARS will give you the equivalent shortcut, in this case: Alpha RS 5 >> - in which manual can I find the complete liste of object types ? >> - and probably the most boring one : now that I have my nice [PI 1] >> array, how can I switch back to the numerical value ? > ??? > [ 'PI' 1 ] > ->NUM > [3.1415.... 1] >> For example, if I enter [PI 1] ABS, then I normally get SQRT(PIO >> +1), but fom then EVAL does nothing and ->NUM returns an Undefined >> Name error... > What do you want to use ??? > JY === Subject: Re: A few more questions in How to keep PI in symbolic form in a vector, a matrix or a complex ? is used on the PC only -> trnsferred to the calculatrice (without the ) will change to a real right-arrow VPN > - what is the general signification of the > - in there any way to access the more directly than through the > CHARS table ? > - in which manual can I find the complete liste of object types ? > - and probably the most boring one : now that I have my nice [PI 1] > array, how can I switch back to the numerical value ? > For example, if I enter [PI 1] ABS, then I normally get SQRT(PIO > +1), but fom then EVAL does nothing and ->NUM returns an Undefined > Name error... > Herve Jean-Yves Avenard a .8ecrit : >> What does fail is trying to do it with the [] key (at left_shift > times) and then entering the pi symbol anywhere between the > brackets. >> For this you will need to surround PI with ' ' >> This is a system issue: PI is a RPL keyword which you can't add to a >> matrix, you can however add algebraic >> JY === Subject: Re: How to keep PI in symbolic form in a vector, a matrix or a complex ? > to insert pi [symbolically] Have you cleared flags -2 and -3 ? Then try [ 'pi' ] etc. (note that quotes are needed within an array, as with algebraic expressions in arrays). A complex number object (object type 1), by definition, can contain only real components; however, you can create a complex-valued algebraic expression (object type 9), e.g. '(0, pi)' [r->] [OFF] === Subject: Re: 50G Cable Survey > Certainly it would need to be a NULL-Modem (ie. cross-over) in addition > to a gender-changing adaptor, right? Since the calc port isn't any standard RS-232 connector, it has no meaning by itself as either DCE or DTE; only when a cable is attached to the calc, then the standard connector at the non-calc end of the cable, when the cable is plugged into the calc, will end up being wired as either DCE or DTE. The HP48/49 serial cables always made the calculators appear as DCE (like modems), and were ready to plug directly into a computer. For reasons why this was so, please see: Re: TX/RX/DCE/DTE... (Re: Trouble with modem) [1996/10/27] Re: PC cable on a MAC., null-modems, DCE, DTE, ... [1997/07/10] The makers of the proposed 50G cables can of course have their own thoughts on the matter -- ask them which way they intend to make it (and why, if different from the HP48/49 serial cables). [r->] [OFF] === Subject: Re: HP 50g + ROM 2.09 On Sun, 20 Aug 2006 19:33:36 -0500: [re Equation library won't SOLV] > You need to use soft-menu (-117SF) as described in the HP50 manual... Right on! But who wants to need a week's study to find out how to even be able to start using a major selling feature? Since menus are defaulted to choose when the calc comes out of the box, this means that the Equation library won't work out of the box! (even when RPN mode is set). Wouldn't it be a good idea to have the Equation library *automatically* set the necessary mode(s) [both RPN and SoftMenus], to give new purchasers a smoother experience? Surely any marketing types at HP would appreciate that, instead of having new customers take their new calcs right back to a flustered sales clerk at the store, and trade for other brands which work as delivered? After all, the CAS forces (often undesired) mode changes all the time; why not let the Equation library set a (usually more desirable) mode, expecially if there's no other way that it can work at all? [r->] [OFF] === Subject: Re: HP 50g + ROM 2.09 <4ksdavFdho9gU2@individual.net> how to even be able to start using a major selling feature? Exactly. I did read that part of the manual, but geez, why couldn't -117SF be included in the code for the solv button instead of making me do it? It's the same equation library I've been using for years, and I never had to set flag 117 before. It's stupid. Furthermore, as of right now, I STILL have not set flag 117, and it works. HOW DID IT GET SET? Don't know? Me either, but it works. === Subject: Re: HP 50g + ROM 2.09 > It's the same equation library I've been using for years, and I never > had to set flag 117 before. It's stupid. You never had to set the flag because there wasn't any EQL for the HP49G until about 1 year ago. So what are you on about? Furthermore, as of right now, I STILL have not set flag 117, and it > works. HOW DID IT GET SET? Don't know? Me either, but it works. > It will work if you're not using the APPS menu to access the EQNLIB but directly type it on the keyboard JY === Subject: Re: HP 50g + ROM 2.09 > On Sun, 20 Aug 2006 19:33:36 -0500: [re Equation library won't SOLV] You need to use soft-menu (-117SF) as described in the HP50 manual... Right on! But who wants to need a week's study to find out > how to even be able to start using a major selling feature? Since menus are defaulted to choose when the calc comes out of the box, > this means that the Equation library won't work out of the box! > (even when RPN mode is set). Wouldn't it be a good idea to have the Equation library *automatically* > set the necessary mode(s) [both RPN and SoftMenus], > to give new purchasers a smoother experience? Surely any marketing types at HP would appreciate that, > instead of having new customers take their new calcs > right back to a flustered sales clerk at the store, > and trade for other brands which work as delivered? After all, the CAS forces (often undesired) mode changes > all the time; why not let the Equation library > set a (usually more desirable) mode, > expecially if there's no other way that it can work at all? [r->] [OFF] I second the motion! === Subject: Re: HP 50g + ROM 2.09 > I second the motion! I estimate that it will take me a week to write and test a new UI for the solver so it works with the existing GUI... Will you and John fork the bill? :) JY === Subject: Re: HP 50g + ROM 2.09 > I estimate that it will take me a week to write and test > a new UI for the solver so it works with the existing GUI... Will you and John fork the bill? :) When SOLV starts (or even when xEQNLIB starts), insert: BINT117 SetSysFlag There -- took me less than a week :) By the way, is there any way to use either MES (MSOLVR) or 30 MENU in ALGebraic mode? (I may just be incompetent, but I can't seem to do either); if not, then please consider appending: BINT95 ClrSysFlag Actually, I've just discovered something completely different: [ROM 2.09 and new Equation libs] Solver fails to start properly (as observed) with: -117 CF then APPS > Equation Library > EQN Library > EQNLIB ... [SOLV] But NO problem at all with this more direct route: -117 CF EQNLIB ... [SOLV] Hey, how come that happen? Does the problem really lie somewhere else than with flag -117 ??? [r->] [OFF] === Subject: Re: HP 50g + ROM 2.09 > When SOLV starts (or even when xEQNLIB starts), insert: BINT117 SetSysFlag That's an ugly solution, because as soon as you start the the solver from the EQNLIB you're actually leaving the EQN library. Therefore the -117 flag will be set forever. What I'm suggesting is depending on the flag -117, I either start the old solver based on soft menus, or I start the new graphical solver filling all the Inform field... There -- took me less than a week :) Apple and Orange .. By the way, is there any way to use > either MES (MSOLVR) or 30 MENU in ALGebraic mode? > (I may just be incompetent, but I can't seem to do either); > if not, then please consider appending: BINT95 ClrSysFlag Actually, I've just discovered something completely different: [ROM 2.09 and new Equation libs] Solver fails to start properly (as observed) with: > -117 CF then APPS > Equation Library > EQN Library > EQNLIB ... [SOLV] But NO problem at all with this more direct route: > -117 CF EQNLIB ... [SOLV] Hey, how come that happen? You can't change soft menus unless you go back in the HOME view. Does the problem really lie somewhere else than with flag -117 ??? Yes, the EQNLIB is a program designed for the HP48 using all 48G entry points. It was quickly ported to the 49G and doesn't use any of the available new architecture... Jean-Yves === Subject: Re: HP 50g + ROM 2.09 <4ksdavFdho9gU2@individual.net> <4l1u2iFea0r7U2@individual.net> <4l2fi3FebqhiU1@individual.net> Just as a short reminider: There *is* a different equation library as an alternative. It uses the available new architecture and runs in algebraic mode or rpn, doesn«t need any flag settings and depending on system flag 30 the solver is chosen (SX or GX style). The MES is also supported and if the GX-Solver is chosen it behaves like the new inform box engine allowing rapid data input by entering all required data in the editline and then pressing enter will fill each field in the input form. It is ROM-independent running on any ROM since 1.19-6 on all machines (49G/49G+/50G). It«s fast and the usage is intuitive, the look and feel is based on the filer, so it nicly integrates in the whole system. A major advantage is the possibility to easily create your own collection of formulas dircetly on the calc, which is only possible otherwise if you know system-rpl. You only need basic knowledge of the calc (which you«ll need anyway if want to work with it) for the editor. And, of course, there is a written documentation. So, if this is a major selling feature of the 49G+/50G you might wanna give it a try and have a look at http://www.software49g.gmxhome.de All right, you might wanna say: that guy want«s something for it. But where the heck is the sense in spending money for something that doesn«t work? And sure JYA is right in saying that wants to get paid if working for HP and helping them to sell their calcs, we all have to pay our bills ;-) Andreas When SOLV starts (or even when xEQNLIB starts), insert: BINT117 SetSysFlag > That's an ugly solution, because as soon as you start the the solver > from the EQNLIB you're actually leaving the EQN library. > Therefore the -117 flag will be set forever. What I'm suggesting is depending on the flag -117, I either start the > old solver based on soft menus, or I start the new graphical solver > filling all the Inform field... There -- took me less than a week :) > Apple and Orange .. By the way, is there any way to use > either MES (MSOLVR) or 30 MENU in ALGebraic mode? > (I may just be incompetent, but I can't seem to do either); > if not, then please consider appending: BINT95 ClrSysFlag Actually, I've just discovered something completely different: [ROM 2.09 and new Equation libs] Solver fails to start properly (as observed) with: > -117 CF then APPS > Equation Library > EQN Library > EQNLIB ... [SOLV] But NO problem at all with this more direct route: > -117 CF EQNLIB ... [SOLV] Hey, how come that happen? > You can't change soft menus unless you go back in the HOME view. Does the problem really lie somewhere else than with flag -117 ??? > Yes, the EQNLIB is a program designed for the HP48 using all 48G entry > points. It was quickly ported to the 49G and doesn't use any of the > available new architecture... Jean-Yves === Subject: Re: HP 50g + ROM 2.09 [bug in menu 114 affects Equation Library] On Wed, 23 Aug 2006 02:48:18 -0500, >> BINT117 SetSysFlag > That's an ugly solution, because as soon as you start the the solver > from the EQNLIB you're actually leaving the EQN library. > Therefore the -117 flag will be set forever. Just like the ugly CAS changed flags? I had first blindly accepted that flag -117 had to be set (as the manual *incorrectly* states) to use those solvers (MES or menu 30) at all, in which case there was no alternative; however, it has later surfaced that flag -117 is a red herring; in fact you do NOT need flag -117 set to use the Equation Library! All you need to do is to invoke the EQNLIB command without using menu 114 (whose choose menu form is buggy), and you will see that everything works fine! Even if you do use the buggy choose menu form of menu 114 to invoke the EQNLIB command, when it aborts after you press the [SOLV] menu key, just execute 0 MENU (or left+hold+PREV) at that point, and you will be okay again! > What I'm suggesting is depending on the flag -117, > I either start the old solver based on soft menus, > or I start the new graphical solver, > filling all the Inform field... Oh heavens, I'm sure that no one had that in mind, particularly since MES (required by many equation sets) has *no* graphical equivalent at all, right? So making EQNLIB work with form-based solvers isn't even possible, and no one suggested doing that! The only issue that was raised is that EQNLIB (or ROMPTR E3 11) shouldn't simply *abort* when the SOLV menu key is pressed to start a [menu] solver, as it now does when invoked by starting from the APPS menu, which it appears might be really *simply* fixed (below). >> But NO problem at all with this more direct route: >> -117 CF EQNLIB ... [SOLV] >> Hey, how come that happen? > You can't change soft menus unless you go back in the HOME view. I don't think I got my point across; let me try again: Assume that flag -117 is currently cleared (for choose boxes); then: If you start from the APPS menu to finally get to the EQLIB command, then when you finally press a [SOLV] menu key to [exit to] a menu solver, that solver fails to initialize properly (this is what the complaint was about). However, when you simply *type* the EQNLIB command, or even do -117 CF 114 MENU then press the EQNLIB menu key, or invoke EQNLIB from the CATalog, or from a program, then continue exactly the same thru the [SOLV] menu key, the menu solver starts correctly! (with flag -117 still cleared!) My conclusion is that there is a problem somewhere between the APPS menu and the EQNLIB starting up, which causes the eventual final failure, in fact, we can isolate it to the choose menu which substitutes for soft menu 114: Do -117 CF then go to 113 MENU, now click the tabbed EQLIB key -- with soft menus, this would go to menu 114, but here it shows a choose menu; THIS choose menu is the culprit, because if from THIS menu you select the EQNLIB command, the final result will fail, but if you either select the EQNLIB command from 114 MENU (even if flag -117 remains cleared!) or type the EQNLIB command, or run it from a program, etc., then the last step (clicking [SOLV] to exit to a menu solver) WORKS NORMALLY! (yet flag -117 remains still cleared). Even if you let things fail by clearing flag -117 and then use the choose menu (114) to invoke the EQNLIB command, when that command aborts from the solver at the end, all you have to do is 0 MENU (or left+hold+PREV) and the solver again then works fine! I therefore conclude that the entire problem has been misunderstood; the Equation Library itself DOES WORK, even with flag -117 cleared -- the sole problem seems instead that ONE choose menu (the equivalent of menu 114) is buggy, and either does not properly start the EQNLIB command or it messes up the final menu after EQNLIB exits, while every other method of invoking the EQNLIB command works fine. > the EQNLIB is a program designed for the HP48 using all 48G > entry points. It was quickly ported to the 49G > and doesn't use any of the available new architecture... The only available new architecture which relates to this problem is the choose menu which corresponds to 114 MENU (before actually entering the Equation Library at all); simply bypass that single menu (or perform 0 MENU afterwards), and then you can use the Equation Library *regardless* of how flag -117 is set. It's simply a *bug* in a choose-format *menu* (114) !!! Easy to fix, no? Only this, and nothing more http://www.heise.de/ix/raven/Literature/Lore/TheRaven.html [r->] [OFF] === Subject: Re: HP 50g + ROM 2.09 [bug in menu 114 affects Equation Library] > I therefore conclude that the entire problem has been misunderstood; > the Equation Library itself DOES WORK, > even with flag -117 cleared -- the sole problem seems instead > that ONE choose menu (the equivalent of menu 114) is buggy, I disagree.... It is *not* buggy as such. The default choosebox menu always expect that what follows is a list of commands. Now what it gets is a program that sets special soft menu and exit immediately. So you now return to the original choose box which when you exit will restore the menu the way it was when you started, therefore overwriting the menu set up by the solver. One way to access it is to call the Previous Menu (with Left Shift and simultaneously NEXT) > Easy to fix, no? unfortunately no... Cause you would need a way to interrupt all the choose menu that have been started, and are still active, only waiting for you to return. But the equation liobrary returns immediately === Subject: Re: HP 50g + ROM 2.09 <4ksdavFdho9gU2@individual.net> <4l1u2iFea0r7U2@individual.net> ummm...do you work for HP? If so, didn't we already fork the bill by paying for one of your new machines? I mean, when I bought the thing, never did it say anywhere in the marketing material that I'd have to not only remember which flag to set, but also set it manually everytime I wanted to use the solver for the eqlib. Then, I also have to un-set it when I want choos? Based on my (several) past purchases of HP calculators that had this exact same eqlib, I just expected it to work. Also, as long as we're talking about bills, how about HP coughing up the $150 or so I paid for that POS HP49G+? I understand that the keyboard was fixed later on, but mine was nearly unusable, as I could never count on a button click getting the number put on the stack. I usually just ended up using my 10 year old calc from Radio Shack. It's painful (without RPN), but it's rock solid. You can take the softmenu code fix out of that pile of my money. > I second the motion! > I estimate that it will take me a week to write and test a new UI for > the solver so it works with the existing GUI... Will you and John fork the bill? :) JY === Subject: Re: HP 50g + ROM 2.09 > ummm...do you work for HP? If so, didn't we already fork the bill by > paying for one of your new machines? I mean, when I bought the thing, I certainly don't .. And I certainly didn't get pay for that either ! JY === Subject: Re: HP 50g + ROM 2.09 > ummm...do[es JYA] work for HP? No, HP dissolved the ACO group, and JYA went on (with others from former ACO) to found his own company: http://www.hydrix.com, which seems only to have had various contracts with HP: http://www.hydrix.com/pages/body/about/history_tiles.jsp So JYA can't possibly be responsible for anything wrong with HP49G+/50G (which deprives him as well of much due credit for what is right with it :) [r->] [OFF] === Subject: Re: HP 50g + ROM 2.09 Hi > Right on! But who wants to need a week's study to find out > how to even be able to start using a major selling feature? Since menus are defaulted to choose when the calc comes out of the box, > this means that the Equation library won't work out of the box! > (even when RPN mode is set). I'm looking at this from a different perpective: fighting illiteracy. People don't read these days when the information is right there on the first page of the Equation Library chapter... Flag [CapitalEth]117 should be set whenever you use the numeric solver to solve equations in the equations library. JY === Subject: Re: HP 50g + ROM 2.09 > I'm looking at this from a different perpective: > fighting illiteracy. > People don't read these days when the information is right there > on the first page of the Equation Library chapter... I don't even have a manual (neither printed nor CD :) > Flag [CapitalEth]117 should be set whenever you use the > numeric solver to solve equations in the equations library. $$$ Millions have been made (and lost) on issues of making things easier (and foolproof), even for dummies, and I think that the above approach is not that which has produced the most millionaires. You (and others) have done a wonderful job going this far, and I hope that anything which may lead it to be appreciated by more potential purchasers might be an advantage for you, too. for all the recent developments and improvements, and I hope it expands to receive its maximum possible appreciation and market share. [r->] [OFF] === Subject: Re: HP 50g + ROM 2.09 > I don't even have a manual (neither printed nor CD :) That's because you didn't buy your HP50G either > $$$ Millions have been made (and lost) > on issues of making things easier (and foolproof), > even for dummies, and I think that the above approach > is not that which has produced the most millionaires. Well Microsoft Windows sells more than Apple MacOS.... you need more than one manual to understand why Windows is acting the way it does (provided there's a logical explanation for some of its behaviour, which I doubt sometimes) Jean-Yves === Subject: Re: HP 50g + ROM 2.09 On Tue, 22 Aug 2006 21:48:37 -0500: > you need more than one manual to understand Windows No, I've never read a Windows manual! The idea, after all, is not to *understand* but to *use* it -- just to press a button and start using, without mysteriously just exiting and saying nothing, like a brand new car which won't start. Unlike all other cars on market, which you can just get in and drive, this car you have to first reset a circuit breaker, which is right on first page of Chapter 12 of owner's manual Oh well, there's no interest in a slight update; I give up! [r->] [OFF] === Subject: Re: HP 50g + ROM 2.09 > Oh well, there's no interest in a slight update; I give up! Unfortunately, that update isn't an easy one. I had a look last night and while you could easily call the graphic solver for simple equations, there's no GUI for the multiple solver. It would need to be written from scratch if you want a nice and easy to use UI. That would be more than one week JY === Subject: Re: Local vs Global variables These are the valid syntax forms of UserRPL local variable structures, which initially take stuff off the stack, store it into newly created local variables, and then evaluate *one* program or formula which can use (and re-use) those local variables: -> x y z ... << program using x y z ... >> -> x y z ... 'formula using x y z ...' If the << program using... >> contains yet another internal -> w (a *nested* local variable structure and scope) then the syntax would again require either -> w << a program here >> or -> w 'a formula here' Each of the above local variable structures always creates *new* local variables; it never stores into already-existing variables, so to re-use existing local variables, use new value 'name' STO instead. The scope of these local variables is only that single object (program or formula) which follows the variable names. When the end of the program (or formula) defining the structure is reached, that's when all the local variables created by that structure cease to exist, so if the same name is used *after* that point, it's interpreted in the original (usually global) context again. If you nest local variable structures (which can be done only in the -> << program >> form) then any inner environment creates *new* variables. Although variables in higher levels remain visible to inner/lower levels, if any names are duplicated, the innermost (most recently created) temporarily hides any others, much as any variable 'X' stored into the current directory will be found and used before any other 'X' that had been stored into a higher directory. While SysRPL syntax clearly distinguishes in UserRPL only the context distinguishes whether 'x' is a global name (object type 6) or a local name (object type 7), which is often slightly confusing (see the first line above), as is practically anything which we encounter for the first time, especially when it resembles other stuff which we already know, yet which has its own, completely different nature. [r->] [OFF] === Subject: Re: Local vs Global variables > I'm a fairly new user with a 49G+. I have started writing some >> programs using matrices as a lookup table and some other items. I >> have come to realize that I really don't quite understand the nature >> of local variables. For instance, I don't understand why the >> following program doesn't result in the value of local variable X >> being loaded onto the stack. Instead I get 'X' on the stack. >> << -> X >> << X sin 1 X SQ + / >> X >> It seems to me that if X was defined on the highest (first) level of >> the program then its value should be available when control returns >> to the highest level. No! Local variable so defined by the << X -> << construction are > only available inside the next innermost level. and sin is written in capitals, eg: SIN > I also don't see how I change the value of a local variable >> within a program. I've tried using -> within a program but keep >> getting errors. Maybe I'm not supposed to change the variable's >> value. Once the local variable is defined, you store values to it just as > you would to a global variable. For example, executing the program << 2 -> X > << 3 'X' STO > X > >> Why also does the following give an error about invalid syntax? >> << -> X >> << X sin 1 X SQ + / ->W >> W > > Each -> W must be followed by either << or by an algebraic in > single quotes, such as 'W' if you start a new structure -> W then you must have at least << >> they belong together There is also an algebraic syntacs: << ->n 'IFTE(n>1,Fn(n-1)+Fn(n-2),n)' >> 'Fn' STO will calculate Fibonacci numbers using recursion >> I realize that the User's guide tried to explain the local variables >> and so does the AUR, but apparently not in a way which I can >> understand. >> I have ROM version 2.00, if that makes any difference. As far as I know, this works the same way all the way back to the > earliest of the HP48's HP-28C had the Algebraic version HP-28S added the program syntac of User Functions === Subject: Re: Local vs Global variables > I'm a fairly new user with a 49G+. I have started writing some >> programs using matrices as a lookup table and some other items. I >> have come to realize that I really don't quite understand the nature >> of local variables. For instance, I don't understand why the >> following program doesn't result in the value of local variable X >> being loaded onto the stack. Instead I get 'X' on the stack. >> << -> X >> << X sin 1 X SQ + / >> X >> It seems to me that if X was defined on the highest (first) level of >> the program then its value should be available when control returns >> to the highest level. No! Local variable so defined by the << X -> << construction are > only available inside the next innermost level. and sin is written in capitals, eg: SIN I also don't see how I change the value of a local variable >> within a program. I've tried using -> within a program but keep >> getting errors. Maybe I'm not supposed to change the variable's >> value. Once the local variable is defined, you store values to it just as > you would to a global variable. For example, executing the program << 2 -> X > << 3 'X' STO > X > > >> Why also does the following give an error about invalid syntax? >> << -> X >> << X sin 1 X SQ + / ->W >> W > > Each -> W must be followed by either << or by an algebraic in > single quotes, such as 'W' if you start a new structure -> W then you must have at least << > they belong together > There is also an algebraic syntacs: > << ->n 'IFTE(n>1,Fn(n-1)+Fn(n-2),n)' >> 'Fn' STO > will calculate Fibonacci numbers using recursion But for any but small n, the multiple recursion involved is horribly Try << -> n '(((1+SQRT(5))/2^n - (( 1-SQRT(5))^n)/SQRT(5)' EVAL >> which avoids that multiple recursion. Enter, compile and run it in exact mode, and it will give exact values. If you compile it or run it in approximate mode it will only give approximate answers, but rounding to the nearest integer will restore exactness, at least for n <= 55. at least until n gets quite large. > I realize that the User's guide tried to explain the local variables >> and so does the AUR, but apparently not in a way which I can >> understand. >> I have ROM version 2.00, if that makes any difference. As far as I know, this works the same way all the way back to the > earliest of the HP48's HP-28C had the Algebraic version > HP-28S added the program syntac of User Functions === Subject: Re: Local vs Global variables > X >> There is also an algebraic syntacs: >> << ->n 'IFTE(n>1,Fn(n-1)+Fn(n-2),n)' >> 'Fn' STO >> will calculate Fibonacci numbers using recursion But for any but small n, the multiple recursion involved is horribly Try > << -> n '(((1+SQRT(5))/2^n - (( 1-SQRT(5))^n)/SQRT(5)' EVAL > which avoids that multiple recursion. Enter, compile and run it in exact mode, and it will give exact > values. If you compile it or run it in approximate mode it will only give > approximate answers, but rounding to the nearest integer will restore > exactness, at least for n <= 55. > at least until n gets quite large. Quite right, Virgil! It was not about speed, but an example If you want to avoid exponential explode of recursion you can use a simple loop in RPL A hint: the loop is OVER + SWAP what is outside START NEXT is an exercise for blenno === On Mon, 21 Aug 2006 22:06:54 -0500: > I actually cannot see the current graphing calculator as expedient. > It is Weird, Huge, overinvolved, esoteric, foreign, enigmatic, > and incomplete... Old story (when marketing takes over): http://www.uoregon.edu/~ftepfer/SchlFacilities/TireSwingTable.html http://home.insightbb.com/~gregorym/it.jpeg [r->] [OFF] === Subject: Re: Is my HP49G+ from a good batch?? On Mon, 21 Aug 2006 22:35:55 -0500: >> I think the SD card slot looks like that in the recent posting by >> ... with the following link to a picture: > They all look basically like that, but when you have the two versions > next to each other you can see the difference. Though I can't see the original photo (due to stupid firewall), I think there's a very noticeable difference that we can describe: The older 49G+ SD slot had its top 1/3 or 1/4 in the black belt area, and lower 2/3 or 3/4 in the wider gold belt area, with the circular thumb depression going all the way across the entire SD slot. The third generation keyboard slot which I've seen has moved distinctly down (towards the back of the calc), with rectangular opening now entirely in the widest (gold) body strip, has a thin extra (gold-edged on 49G+) insert where the original upper 1/4 of the slot was, and appears to have a much less wide remnant of the original thumb depression, as if that original thumb depression were left exactly where it originally was when the rectangular slot was simply drilled down lower, which eliminated the thumb slot's original outer edges, and in fact has made the thumb slot almost disappear. The thumb slot is handy for removing the SD card via the thumbnail groove on the label side of the SD card (there's actualy no need to use the fancier spring ejector) Couldn't $$$ have therefore been saved on the spring ejector, and put into the keyboard, say? But All's well that ends well, I suppose: http://www.oldielyrics.com/lyrics/the_kingston_trio/the_unfortunate_miss_bai ley.html http://users2.ev1.net/~smyth/linernotes/thesongs/UnfortunateMiss.htm [r->] [OFF] === Subject: Re: Is my HP49G+ from a good batch?? > The older 49G+ SD slot had its top 1/3 or 1/4 in the black belt area, > and lower 2/3 or 3/4 in the wider gold belt area, with the circular > thumb depression going all the way across the entire SD slot. I have probably the oldest 49g+ out there (and what's for sure older than any 49g+ that ever hit the market), as well as the newest HP50g However, the newer 50g has a little black strip at the top of the SD slots. Newer 49g+ also has this plastic strip, except than it's gold... So I don't see how the SD card slot looks like can actually define which series of calculator you have... JY === Subject: Re: Is my HP49G+ from a good batch?? <4l2ioeFecju5U1@individual.net The older 49G+ SD slot had its top 1/3 or 1/4 in the black belt area, > and lower 2/3 or 3/4 in the wider gold belt area, with the circular > thumb depression going all the way across the entire SD slot. > I have probably the oldest 49g+ out there (and what's for sure older > than any 49g+ that ever hit the market), as well as the newest HP50g > However, the newer 50g has a little black strip at the top of the SD > slots. Newer 49g+ also has this plastic strip, except than it's gold... So I don't see how the SD card slot looks like can actually define which > series of calculator you have... JY Hmmm. I just got my 50g today and the SD card does appear to sit lower. The bottom is virtually flush with the bottom of the band. On both 49g+ calcs, the bottom of the SD card rides almost in the middle of the band. In any event, the 50g is a lot nicer. Scott Chapin === Subject: Re: Is my HP49G+ from a good batch?? > Hmmm. I just got my 50g today and the SD card does appear to sit lower. > The bottom is virtually flush with the bottom of the band. On both 49g+ > calcs, the bottom of the SD card rides almost in the middle of the > band. Only the curve is different between the model. Not where the hole is located. it's where the strips are located JY === Subject: Re: Is my HP49G+ from a good batch?? <4l2ioeFecju5U1@individual.net So I don't see how the SD card slot looks like can actually define which > series of calculator you have... JY Look again. This time look below the SD slot, not above it. The older calcs/keyboards have a deeper, wider, concave cut out, that is as wide as the SD slot . The newer calcs/keyboards have a much, much smaller concave cut out, that is about 2/3 of the width of the SD slot. hth, mk === Subject: Re: Is my HP49G+ from a good batch?? <4wtgmy16lix9$.cfyxv2pc8i9y.dlg@40tude.net CNAXXXXXX53 is the good batch with fixed keyboard. Uh, no. The 49g+ and 50g follow this scheme for the serial number: First digit (after CN or CNA) = Year (mod 1000). Next two digits = Week (1-53) within that year. Remaining digits (4 or 5) = incremented throughout that week. This HP50g returns CNA6212986 at the end of the SERIAL command. That means that it was made in 2006, 21st week, the 2986th machine that week. -Joe- === Subject: Re: Is my HP49G+ from a good batch?? <4wtgmy16lix9$.cfyxv2pc8i9y.dlg@40tude.net > Uh, no. The 49g+ and 50g follow this scheme for the serial number: First digit (after CN or CNA) = Year (mod 1000). > Next two digits = Week (1-53) within that year. > Remaining digits (4 or 5) = incremented throughout that week. This HP50g returns CNA6212986 at the end of the SERIAL command. That > means that it was made in 2006, 21st week, the 2986th machine that > week. -Joe- What is the difference between the number returned via the SERIAL command the the number on the back of the case? They're close, but not the same. === Subject: Re: Is my HP49G+ from a good batch?? <4wtgmy16lix9$.cfyxv2pc8i9y.dlg@40tude.net First digit (after CN or CNA) = Year (mod 1000). Learn math, dummy. It's mod 10, not mod 1000. Sheesh. === Subject: HP50g problems I have just purchased an HP50g from Amazon. I have a major problem with the CAS system in the original ROM version rev 2.08. My problem relates to solving electrical network loop equations using the inverse matrix. If I use real numbers plus a variable, I get ridiculous results in the inverse matrix. It could be that this problem is fixed with the ROM revision, however I cannot get conn4x to run on either of my XP machines, so I have been unable to download the new operating system. The USB driver seems to install correctly, however the exe will not run without performing an access violation. I have noticed that worldwide other people are having this problem. It appears that older versions of conn4x do work with xp, but how can I get hold of the older version? I have version 2.3 build 2434. Unfortunately, HP Customer Care in the States are not interested in supporting a user in New Zealand. The 50g is not yet available in NZ. === Subject: Re: HP50g problems Check your email. === Subject: Re: HP50g problems <57udnYavBbwCLXbZnZ2dnUVZ_oOdnZ2d@giganews.com Check your email. show version 2.09 after entering the version command. Unfortunately the CAS problem is still with me! I wonder if anyone else can confirm a problem with inverse matrices or determinants when the matric contains expressions containing a single variable and real numbers. For instance if I have a matrix A containing A11=s*1e-3+1/(s*1e-6), A12=A21= -1/(s*1e-6) A22= s*1e-3+1/(s*1e-6)+1e3. Version 2.09 calculates the inverse quite quickly but the results are obviously wrong containing terms of s^8, s^9 ....s^14 I would very much appreciate some help here. HP (US) are not interested as I am in NZ. === Subject: Re: HP50g problems > I wonder if anyone else can confirm > a problem with inverse matrices or determinants when the matric > contains expressions containing a single variable and real numbers. > For instance if I have a matrix A containing A11=s*1e-3+1/(s*1e-6), > A12=A21= -1/(s*1e-6) > A22= s*1e-3+1/(s*1e-6)+1e3. > Version 2.09 calculates the inverse quite quickly but the results are > obviously wrong containing terms of s^8, s^9 ....s^14 The result is obviously correct, just not fully simplified (as this calc does not autosimplify anything). Steen === Subject: Re: HP50g problems > The result is obviously correct, just not fully simplified (as this > calc does not autosimplify anything). And on that note, can anyone tell me why it wouldn't be desireable to have the calc automatically return something like six and a half instead of thirteen over two? Colin === Subject: Re: HP50g problems > The result is obviously correct, just not fully simplified (as this >> calc does not autosimplify anything). And on that note, can anyone tell me why it wouldn't be desireable to > have the calc automatically return something like six and a half > instead of thirteen over two? To that I want a flag Why? I hate 6¸ Also the simplification should take a flag A) Automatic B) No simplification HPQ should put the non-User flags in hidden memory and release all those grazy flags (not listed in MODE FLAGS) to a better use === Subject: Selling = Developing <44d7aa86$0$847$ba4acef3@news.orange.fr> <44D8EDC9.708@wildopensource.com> This is my first post... so a bit of me. 23 y/o, using hp48g since 13. Chem. Eng. from Venezuela. I've been thinking about this topic almost since I first started programming. And my general idea of the best way of selling programs is if the HP had an electronic serial number (ESN) wich you can relate to the hard serial number of your machine. So that the buyer gives me the hard serial number and with the algoritm I have already developed I give him he password in order to complete the instalation. Of course, since the algorithm is in the program itself, after decompiling it you will have the algortim and release a Crack. But thats the way every program on the net works. You will have some buyers... and then someone releases the crack. Then you develope a better program, and better security and so on... But I am sure that at least with the profit of your first sales you'll be happy, it's better than nothing, and you are still sharing your knowledge very chip. C.J. P.S. If there is such thing as an ESN, please let me know!!!! === Subject: Re: Selling = Developing hello, there is an internal serial number in the 50g, you can get it by using the command SERIAL (lib 256 needs to be attached). cyrille > This is my first post... so a bit of me. 23 y/o, using hp48g since 13. > Chem. Eng. from Venezuela. I've been thinking about this topic almost since I first started > programming. And my general idea of the best way of selling programs is > if the HP had an electronic serial number (ESN) wich you can relate to > the hard serial number of your machine. So that the buyer gives me the > hard serial number and with the algoritm I have already developed I > give him he password in order to complete the instalation. Of course, since the algorithm is in the program itself, after > decompiling it you will have the algortim and release a Crack. But > thats the way every program on the net works. You will have some > buyers... and then someone releases the crack. Then you develope a > better program, and better security and so on... But I am sure that at least with the profit of your first sales you'll > be happy, it's better than nothing, and you are still sharing your > knowledge very chip. C.J. P.S. If there is such thing as an ESN, please let me know!!!! > === Subject: Re: Selling = Developing On Wed, 23 Aug 2006 08:25:29 -0500: > there is an internal serial number in the 50g[and 49G], > you can get it by using the command SERIAL > (lib 256 needs to be attached). Or #10001Dh LIBEVAL (always works). However, the usual issues arise: o What happens if calc is replaced, either for service or trade-up? [similar to Microsoft's early XP issues], as has been common among 49G+ owners who have first and second generation keyboards. o What about the owner wanting to also use an emulator for his calc, on his computer? [if you enable running on no serial calc, then it will run in emulator -- this might even be of use for a demo] An occasional patched boot loader (especially of older 49G+) might also have no serial. [r->] [OFF] === Subject: Re: Selling = Developing > This is my first post... so a bit of me. 23 y/o, using hp48g since 13. > Chem. Eng. from Venezuela. I've been thinking about this topic almost since I first started > programming. And my general idea of the best way of selling programs > is if the HP had an electronic serial number (ESN) wich you can > relate to the hard serial number of your machine. So that the buyer > gives me the hard serial number and with the algoritm I have already > developed I give him he password in order to complete the instalation. Of course, since the algorithm is in the program itself, after > decompiling it you will have the algortim and release a Crack. But > thats the way every program on the net works. You will have some > buyers... and then someone releases the crack. Then you develope a > better program, and better security and so on... But I am sure that at least with the profit of your first sales you'll > be happy, it's better than nothing, and you are still sharing your > knowledge very chip. C.J. P.S. If there is such thing as an ESN, please let me know!!!! At least HP 49G had it... === Subject: Re: Selling = Developing > Of course, since the algorithm is in the program itself, after > decompiling it you will have the algortim and release a Crack. But > thats the way every program on the net works. You will have some > buyers... and then someone releases the crack. Then you develope a > better program, and better security and so on... Just a side comment; you are being overly pessimistic here. Unless you're in the entertainment market selling mainly to unsupervised kids, the majority of your potential users will not use stolen copies of your software and will continue to buy it. The exact percentage probably depends on your market, and on how much you are charging. -- Chris Smith === Subject: Re: HP-50g Polynomial Coefficient Array Functions > What you want to look at is FPTR2 ^VXXLext For safety try: :: CK2&Dispatch SYMID :: FPTR2 ^VXXLext SWAPDROP ; ; With args 'X^2-5*X+6' 'X' [in Exact mode] you'll get a coefficient list of exact integers, but with args 'X^2.-5.*X+6.' 'X' [real-valued coeffs] you'll get a coefficient list of %% (double-precision reals), which need further digesting for normal output (AXL aborts). If there is *mixed* (integer and real) input, then the resulting list is mixed (integer and double), and CAS annoyance mode also sets in :) These threads may also be of interest: HELP WITH COEFFICIENTS [1999/11/09] We have convolution!! [2000/05/17] (real inputs to 'CONV' also produce double results). [r->] [OFF] === Subject: Re: HP-50g Polynomial Coefficient Array Functions > HI > WARNING, my hp just went plain dead. Perhaps you should not do as I > suggested. I doesn't look good. I have changed batteries, tried Turned out I had to reflash the ROM of the hp49g+. It only reacted to the sequence for updating the rom. Never thought I could wreck the ROM by accident... Maybe somebody has any comments? Gjermund Skailand === Subject: Re: HP-50g Polynomial Coefficient Array Functions > I did have one that worked fine on the HP-48G Here are more Polynomial -> Array functions for HP48 *only* (because TAYLR lost its original finesse in the new CAS :) Array -> Polynomial are trivial, e.g. for HP48G/49G/50G: Args: { 2 3 5 } Coefficients (symbolic okay), use AXL on array input 'X' Variable name << -3. CF 0. ROT 1. << 3. PICK ENDSUB NSUB - ^ * + >> DOSUBS SWAP DROP >> On the 49G/50G, [array] 'X' << PEVAL EVAL >> works (but the CAS springs into annoyance mode if any reals are seen :) [r->] [OFF] === Subject: Re: HP-50g Polynomial Coefficient Array Functions On Tue, 22 Aug 2006 17:27:09 -0500: > On the 49G/50G, [array] 'X' << PEVAL EVAL >> works > (but the CAS springs into annoyance mode if any reals are seen :) Could also be: [array] 'X' << PEVAL EXPAND >> (but now the CAS gets annoying even for integers :) Extract Coeffs of polynoms [by Beto] 2000/09/18 [r->] [OFF] === Subject: Re: HP 50G - Connectivity software problems? <44db9ab4_1@news.bluewin.ch> <44e0aab6$0$10455$ba620e4c@news.skynet.be mats schrieb I had the same troubles with this version (some sort of Windows > problem I guess). Just try the older one > (http://h10025.www1.hp.com/ewfrf/wc/softwareList?lc=en&cc=us&dlc=en&product=3 51775&dest_page=product&os=228) Paul, > please check help about on your con-kit and report the build number. > Mats linked con-kit, give Version 2.2 Build 2353. > This con-kit works fine. Note the Setup_ReadMe.txt from the above download... if it applies to > your problem. Heiko The earlier version will load and work, but trying the newer version on several PC's with WinXP Home or Pro gives the same result of access violation. See also recent posting in this group by Ivo. === Subject: Re: HP 50G - Connectivity software problems? <44db9ab4_1@news.bluewin.ch> <44e0aab6$0$10455$ba620e4c@news.skynet.be> The linked con-kit works, minus screencapture. Build 2353, the 49g+ one. === Subject: Re: Local Language - Printed Manuals X > Yes, it takes time and effort and money to localize a product and of > course will only be done if their is a significant demand for it. X There is demand in Finland for a localized manual, but HPQ diesn't seem to spend money on it...??? > Beside the fact that the translation process was quite time consuming > and that Hewlett-Packard is making money with there calculators I see > no reason why I should give away the localized software for free and > helping them to boost their sales with no benefit for me. > Like my economic prof. told me and as you all know: Things for free > don«t have a value and TANSTAAFL. WINTW? > I will, however, share the core which takes care of the translation > process, because I would like to see some localized 49G/49G+/50G > arrive in other languages, e.g. Spanish, French, Finnish (go on VPN), > Catalan, Portuguese, Italian, Russian, Chinese, you name it... that uses port 0 - a slight problem with HP 48gII and JYA (from ACO - then) helped on a technical point I translated on about 15KB of messages and there is an even smaller, less translated version > limitations: SETMESG only accepts that the alternative messages are > usage. This is not so much a problem with 256kb of ram but still > nonessential. My program works with 49G, 49g+, 50G, 48gII and the 48gII is the most sold in Finland (my guess) > As a circumvention to this, my program runs in an indefinite loop (in > any port, 0, 1, 2 and if in covered memory it takes care of it«s own > adress in case something that is stored before the lib is purged) > having total control over each keypress and checks if the messages are > still attached and if not reinstalls them. Nice What I would like to see (in C ?) is a program so that messages could be fit somewhere else than port 0 === Subject: Re: Local Language - Printed Manuals <0XSGg.9609$Kv1.4880@reader1.news.jippii.net> I don«t even see a reason why they crippled the 49G+ and made a 48GII. What«s the reason that it sells this good in Finland? > What I would like to see (in C ?) is a program > so that messages could be fit somewhere else than port 0 there don«t fit any message arrays in. > that uses port 0 - a slight problem with HP 48gII > and JYA (from ACO - then) helped on a technical point > I translated on about 15KB of messages > and there is an even smaller, less translated version Did you check the STO-Bug ? Concerning the price: I«m sure you didn«t do it for free, you had to invest something. I did learn a lot from this newsgroup and I hope that I made some valuable contributions back and hope that I«ll still learn more. But, call it what ever you want, you«ll always have to pay some way or the other for things in life. So I believe it«s only fair to be rewarded (that doesn«t necessarily means money) if you offer opportunities for other people to help them solve their task. Just my two cents Andreas === Subject: Re: Local Language - Printed Manuals You are certainly right pointing to the impact of the local law conc. the requirement of shipping a localized manual. What I critisize is, that the currently shipping localized documentation is neither complete (look at the index for example ...) nor correct! I'm certain that this is not compliant (for instance) with German law ... In my opinion it's up to HP & it's distributors to change that. That's the reason I've posted my proposal to how it eventually could be changed. And I stress it again: The basic English manual has to be of excellent quality first! > I think it is obligatory to deliver printed manual in local languege > for any item you want to sell inside the european union, same thing for > 2 years warranty; so nothing to thank for or to as for. It's only the > law inside the european union, and I think that's the reason why hp 50g > doesn't exist in all the EU hp webs. === Subject: Re: Local Language - Printed Manuals On Tue, 22 Aug 2006 16:28:54 -0500: > local law... the requirement of shipping a localized manual. Kinda funny, given France's world-leading Minitel (eliminating the phone book), but evidently fairer to people without computers. [r->] [OFF] === Subject: Re: Local Language - Printed Manuals Sorry, I forgot: (As an example) it is absolutely ok according to German law, to offer a customer to ship English docs with a product - only if you don't negotiate beforehand, then you are required to ship localized docs! So HP would be able to offer two different SKUs. Peter A. Gebhardt Disclaimer: I'm not a lawyer, so don't take my words as legal advice! > You are certainly right pointing to the impact of the local law conc. > the requirement of shipping a localized manual. What I critisize is, > that the currently shipping localized documentation is neither complete > (look at the index for example ...) nor correct! I'm certain that this is not compliant (for instance) with German law > ... In my opinion it's up to HP & it's distributors to change that. That's > the reason I've posted my proposal to how it eventually could be > changed. And I stress it again: The basic English manual has to be of excellent > quality first! I think it is obligatory to deliver printed manual in local languege > for any item you want to sell inside the european union, same thing for > 2 years warranty; so nothing to thank for or to as for. It's only the > law inside the european union, and I think that's the reason why hp 50g > doesn't exist in all the EU hp webs. === Subject: Re: HPGCC > Well I have been spurred on by the introduction of the new 50G, does anybody > have any good suggestions for good C programs that I might consider? My > background is electronics engineering. > mickpc@bigpond.com Big mistake, I was not clear in my heading. I was after suggestions for new programs that I might consider writing. DUH! If anybody has suggestions please. === Subject: C programs I was unfortunatly not clear in my previous post that I was looking for suggestions for *new* C programs that might be worth considering writing. Looking forward to any suggestions, the more the better. I need a good challenge. mickpc@bigpond.com === Subject: Re: C programs > I was unfortunatly not clear in my previous post that I was looking for > suggestions for *new* C programs that might be worth considering > writing. Looking forward to any suggestions, the more the better. I need > a good challenge. Ahhhhhh, I see. Now I understand what you meant, I thought you were looking for existing programs, not ideas for new programs. Here you have a couple of ideas to get you thinking: 1) Fully C-implemented filer. Tim's SDFiler is a hybrid C/sysRPL, therefore it's a little slow. A full C implementation would give users a lot more power (faster browsing, multiple partitions in the same card, saving entire USEROB directory trees as separate files/directories on the card, and viceversa, etc). This is a *big* challenge, since you need to program the screen interface from scratch (grayscale to make it more impressive :-). 2) A snapshot save/restore library: A C program that can save a concept can be used to create virtual calculators, pretty much like EMU48 does. You could hot-swap different calculator setups saved in your SD card. You would need to save/restore all saturn registers and perhaps modify some system variables to keep the date and time but other than that, it's a simple program. Wouldn't it be cool if you can choose a calculator setup from the filer and get it running in a fraction of a second? After a total memory loss, it would take you just a second to get your calc up and running again. What I picture is a hot-key that let you swap the two most-recently used calcs, it would work like a virtual-stack but you would have also swap the entire HOME directory. Maybe you should start with 2) since that feature has never been seen in the calculator. Think about it... Claudio === Subject: Re: C programs > I was unfortunatly not clear in my previous post that I was looking for > suggestions for *new* C programs that might be worth considering > writing. Looking forward to any suggestions, the more the better. I need > a good challenge. > mickpc@bigpond.com Oh yes and would anybody be interested in forming a group of C programmers that can network on this? HP has given us a new machine, we should all show some faith and get this off the ground for them (said thought clenched teeth lol) === Subject: Re: hpgcc decnumber matrix solver call for betatesters It seems Gilbert Fernandes' and Eric Reichlin's Introduction to Saturn Assembly Language contains the data I was looking for. It might be feasible to create a stack to C parser, which could handle all the data types. I'd guess that would speed up development of HPGCC software. Ville Koskinen > You should read the available SysRPL and ASM documentation for this > info. Steen === Subject: Re: hpgcc decnumber matrix solver call for betatesters > I'd like to try that. Much faster and requires less memory. Instead of sending the complete .c file, I'll just explain to you what to do (faster and less cluttered). An example with a zint: //reserve Saturn stack space in nibbles, remember room for prologue, size field and sign int sat_address=sat_createtemp(zint_size); If sat_address==0 there wasn't enough room! //copy address, you'll need it later int sat_addr=sat_address; //push prologue, 5 nibbles sat_poke(sat_address,0x2614,5); sat_address+=5; //push size field, 5 nibbles sat_poke(sat_address,zint_size-5,5); sat_address+=5; Push the zint nibbles from back to front with sat_poke(). Remember to update sat_address along the way. //push sign, 1 nibble if(zint.sign) sat_poke(sat_address,0x9,1);//negative else sat_poke(sat_address,0x0,1);//positive //push start address onto Saturn stack sat_pushptr(sat_addr); Done! Now, I've been looking at your Longfloat lib, and was thinking it wise to adopt the same object type in Saturn land for my ExtraFunc lib. Meaning that when I need to return floating point numbers - say the result of AiryAi(z) - I return it as a longfloat, adhering to a global or local variable DIGITS as your do. I just need to understand the format. It's type 27, but you use two different prologues, depending on the number being real or complex. Can you explain the structure and the limits of the fields to me? I do think that we should dispose of the limit of DIGITS<10000 - it's totally feasible to calculate even hundreds of thousands of digits, especially if these apps get ported to run on calcs after the 50G with an integer part exceeding 10000 digits (calculated in about 4 seconds). What do you think? Steen === Subject: Re: hpgcc decnumber matrix solver call for betatesters Hi Steen, > I'd like to try that. Much faster and requires less memory. Instead of sending the complete .c file, I'll just explain to you what > to do (faster and less cluttered). An example with a zint: //reserve Saturn stack space in nibbles, remember room for prologue, > size field and sign > int sat_address=sat_createtemp(zint_size); If sat_address==0 there wasn't enough room! //copy address, you'll need it later > int sat_addr=sat_address; //push prologue, 5 nibbles > sat_poke(sat_address,0x2614,5); > sat_address+=5; //push size field, 5 nibbles > sat_poke(sat_address,zint_size-5,5); > sat_address+=5; Push the zint nibbles from back to front with sat_poke(). Remember to > update sat_address along the way. //push sign, 1 nibble > if(zint.sign) > sat_poke(sat_address,0x9,1);//negative > else > sat_poke(sat_address,0x0,1);//positive //push start address onto Saturn stack > sat_pushptr(sat_addr); Done! In the above, change the header, then repeat for two zints. Now, I've been looking at your Longfloat lib, and was thinking it wise > to adopt the same object type in Saturn land for my ExtraFunc lib. > Meaning that when I need to return floating point numbers - say the > result of AiryAi(z) - I return it as a longfloat, adhering to a global > or local variable DIGITS as your do. look at the hidden source directory at the files CheckDigits, PutDigits and DIGITS (86008) i believe. You can read (peek/poke)that directly. I just need to understand the format. It's type 27, but you use two > different prologues, depending on the number being real or complex. Can you explain the structure and the limits of the fields to me? zints you know for longreal the structure is actually one longfloat header and two zints, one for mantissa and one for exponent for longcomplex is actually one longcomplex header, then 4 zints zint: header(5 bints) size_of_integer(5 bints) integer, sign(1 nibble) longreal: header(5 bints) size_of_mantissa (5 bints) body_ of_ mantissa, sign(1 nibble) size_of_exponent_(5 bints) body_of_exponent sign(1 nibble) ie longreal: header(5 bints) zint_mantissa zint_exponent longcomplex: header(5bints) longreal, longreal Thus it is only limited by free memory. use ->H and you will see it. look at the functions in the hidden (source) directory zz->f, zz->F , f->zz ff->c , c->ff , difference between zz->f and zz->F is that uppercase denotes a normalizing. All the intricate ( to me at least) memory handling is to take care of what may happen during garbage collections. If you know or perhaps someone will tell? which saturns registerflags are safe during garbage, I'd really appreciate to know that. Note that 0. is special. Please implement it with a mantissa which obeys the DIGITS Flag for backwards compatibility. Also look at the INCLUDE file. You should consider to use same flags for rounding (FLOOR, CEIL, TRUNC, ROUND) if you want to implement it. In my library is governs directly the normalize function. further you may know the issues with integer 0, Test like ^QIsZero, only tests the size field this for integers need to be trimmed. Presently the version of hpgcc returns a zint 0 which need to be trimmed. I do think that we should dispose of the limit of DIGITS<10000 - it's Actually it has been disposed of for basic arithmetic. Try e.g. with repeated squaring. The longfloat arrays are fixed mantissa size according to DIGITS with an fixed exponent size of 15 digits. Is makes conversions faster but we need to consider if that should be of matrix type instead lists of lists, as there are better support for that in the current OS. However, If you chose to dispose of it you can then no longer read DIGITS The limit of <10000 is due to Mika Heiskanens old library, he used a bint for the exponent, resulting in 5 digits. Some of the sysrpl , I don't remember exactly #>Z or #>Z or something similar has check for 5 digits. I did try to remove those calls. > totally feasible to calculate even hundreds of thousands of digits, > especially if these apps get ported to run on calcs after the 50G with > an integer part exceeding 10000 digits (calculated in about 4 seconds). > What do you think? based on Bernoulli numbers? Steen I'd say it would fit straight into the existing system. And I have a ln function which calculates with an accuracy of 1200 digits in about 5 seconds. 40-100x faster than the existing in longfloat. Generally for functions in the old longfloat calculations were becomming unpractical when digits were >100, basic aritmetic excluded. And consider this: with hpgcc and e.g. decnumber with arbitrary precision may be equal or faster than existing functions on the hp if we can get rid of the overhead. It would be quite an impressive hand-calculator. Gjermund study it. === Subject: Re: hpgcc decnumber matrix solver call for betatesters > I just need to understand the format. It's type 27, but you use two > different prologues, depending on the number being real or complex. Can you explain the structure and the limits of the fields to me? Just replying to my own post; Eric Rechlin pointed out to me that these (infinite precision real and complex numbers) are in fact new built-in object types of the HP49/50 - I didn't know that :-) object types. Steen === Subject: Re: v92 ROM and the Equation Library <4k2b7eFa83h1U1@individual.net> <4ksnamFdg48hU1@individual.net > just copy ROMG+.49G (the ROM-file for an emulated 49G+) to ROM.49G (the > ROM-file for an emulated 49G) in your emu48 directory. Emulate an 49G (type VERSION and you«ll notice that your emulated 49G > now has the latest ROM), > make sure you activated the serial port in EMU48, plug in your cable > and run ROMUPLOAD. Follow the update procedure and voil.88: a 49G with ROM 2.09. Andreas Yes, they will work. You should upgrade your HP49G to a later ROM though ... > (the 2.09 ROM provided with Emu49 will work on the original HP49G > provided you can extract it) Jean-Yves Hello Jean Yves, how can we do that? Paul Estepan Paul === Subject: Re: v92 ROM and the Equation Library just copy ROMG+.49G (the ROM-file for an emulated 49G+) to ROM.49G (the > ROM-file for an emulated 49G) in your emu48 directory. You can't use a 49G+ ROM in a 49G ... As it uses some of the new opcodes only available on the emulated Saturn CPU Emulate an 49G (type VERSION and you«ll notice that your emulated 49G > now has the latest ROM), > make sure you activated the serial port in EMU48, plug in your cable > and run ROMUPLOAD. ROMUPLOAD will not work on newer ROM Follow the update procedure and voil.88: a 49G with ROM 2.09. Did you actually try this ? === Subject: Re: v92 ROM and the Equation Library <4k2b7eFa83h1U1@individual.net> <4ksnamFdg48hU1@individual.net> <4l36t6F1tjiU1@individual.net> Yes, I did and transfered ROM 2.01 it to my old 49G. ROMPUPLOAD only checks the machine it is running on (look in the source, don«t know if they now change it for the next update). :: CK0 IsApple case :: Not Available TOTEMPOB PTR 13BEFUDO$EXIT ; DOCLLCD TURNMENUOFF Current version: BIGDISPROW1 xVERSION DROP . . . ; See: HP49B-2.05-4 Emulator ROM? And just checked it before I send this message with emu48 1.40+ emulating a 49G with ROM 2.09, but didn«t transfered it «cause I«m having no 49G around here at the moment. Andreas just copy ROMG+.49G (the ROM-file for an emulated 49G+) to ROM.49G (the > ROM-file for an emulated 49G) in your emu48 directory. > You can't use a 49G+ ROM in a 49G ... As it uses some of the new opcodes > only available on the emulated Saturn CPU Emulate an 49G (type VERSION and you«ll notice that your emulated 49G > now has the latest ROM), > make sure you activated the serial port in EMU48, plug in your cable > and run ROMUPLOAD. > ROMUPLOAD will not work on newer ROM Follow the update procedure and voil.88: a 49G with ROM 2.09. Did you actually try this ? === Subject: Re: HP50g Poster joga ha scritto: > Hi hp.eople, Does any body know where can I get a HIGH resolution picture of hp50g? > I want it to print in large format to do a poster. Tank You so much joga Hi Joga look at this page: http://mikemander.servehttp.com:8008/hp/HP50g/ I hope that are will be enought for your aim Ciao Antonio === Subject: Re: HP50g Poster : Hi hp.eople, : Does any body know where can I get a HIGH resolution picture of hp50g? : I want it to print in large format to do a poster. : Tank You so much How high a resolution do you need and how soon do you need it? A little work with my 4x5 and scanner and I can get you a 2400dpi scan of a 4x5 transperancy. The file size about a gig. -- ------------------- Keep working millions on welfare depend on you === Subject: Re: HP50g Poster <4tWdnXDQAI8D_XbZnZ2dnUVZ_r6dnZ2d@giganews.com How high a resolution do you need and how soon do you need it? A > little work with my 4x5 and scanner and I can get you a 2400dpi scan > of a 4x5 transperancy. The file size about a gig. That resolution is not real, is a software interpolation or is in BMP format :-D === Subject: Re: HP50g Poster : > How high a resolution do you need and how soon do you need it? A : > little work with my 4x5 and scanner and I can get you a 2400dpi scan : > of a 4x5 transperancy. The file size about a gig. : That resolution is not real, is a software interpolation or is in BMP : format :-D The file size is for that of a photoshop file. I'm hoping that a tiff file will be a bit smaller. :-) I'll try to do this on saturday. Unfortunatly if I don't get it done then it'll be another week or so. Is there some special I should put in the display? -- ------------------- Keep working millions on welfare depend on you === Subject: Re: HP50g Poster > How high a resolution do you need and how soon do you need it? A > little work with my 4x5 and scanner and I can get you a 2400dpi scan > of a 4x5 transperancy. The file size about a gig. > That resolution is not real, is a software interpolation or is in BMP >> format :-D The file size is for that of a photoshop file. I'm hoping that a tiff > file will be a bit smaller. :-) I'll try to do this on saturday. > Unfortunatly if I don't get it done then it'll be another week or so. Is there some special I should put in the display? Perhaps Vertical SIN curve? It is explained in Parametric Plot Learning Module === Subject: Re: HP50g Poster <4tWdnXDQAI8D_XbZnZ2dnUVZ_r6dnZ2d@giganews.com> Hi Frank, It could be very good, please send me by email to: j o g a 5 2 0 0 1 @hotmail.com > : Hi hp.eople, : Does any body know where can I get a HIGH resolution picture of hp50g? > : I want it to print in large format to do a poster. : Tank You so much How high a resolution do you need and how soon do you need it? A > little work with my 4x5 and scanner and I can get you a 2400dpi scan > of a 4x5 transperancy. The file size about a gig. -- > ------------------- > Keep working millions on welfare depend on you === Subject: Re: HP50g Poster > Hi Frank, > It could be very good, > please send me by email to: j o g a 5 2 0 0 1 @hotmail.com 1 gig email? don't think so ... === Subject: Re: HP50g Poster >> Hi Frank, >> It could be very good, >> please send me by email to: j o g a 5 2 0 0 1 @hotmail.com > 1 gig email? > don't think so ... OK, how about an FTP transfer Home page with file? Perhaps a User name and password before access Maybe Hydrix could host it? One could have it on different resolutions ~ < 1Gig ~ < 100 megs ~ < 10 megs ~ < 1meg === Subject: Re: HP50g Poster <4tWdnXDQAI8D_XbZnZ2dnUVZ_r6dnZ2d@giganews.com> <4l1db6Feb6tvU1@individual.net> Frank I am sorry, I did not account the huge file size and the small space in my hotmail inbox, Please can You attach this hotmail account: s o l o h p @hotmail.com as a CONTACT to transfer the file,? jose > Hi Frank, > It could be very good, > please send me by email to: j o g a 5 2 0 0 1 @hotmail.com > 1 gig email? > don't think so ... === Subject: Re: HP50g Poster 2 things that might help ... 1. To print @ A1 (poster size) you probably won't see any difference between a 2400 dpi (1gb) and a 1200 dpi (250MB) scan. 2. To send such a big file just use www.sendthisfile.com (sign up for a free account) - we use them all the time and they're faultless. === Subject: Re: HP50g Poster >> Hi hp.eople, > Does any body know where can I get a HIGH resolution picture of >> hp50g? I want it to print in large format to do a poster. > Tank You so much How high a resolution do you need and how soon do you need it? A > little work with my 4x5 and scanner and I can get you a 2400dpi scan > of a 4x5 transperancy. The file size about a gig. Send me that one, please vpn at dlc dot fi === Subject: HP 50g Manual on-line Has any one been able to find the HP50g Manuals on-line in the HP website?? I see the training modules are there but I can find no reference to the User Guides and Manual. Gc === Subject: HP 50g Manual on-line Has any one been able to find the HP50g Manuals on-line in the HP website?? I see the training modules are there but I can find no reference to the User Guides and Manual. Gc === Subject: Xserver Can xserver48 work from port 2 of the HP 48Gx? The xserver48 seems to lock up when used from port 2. When used from port 0, there seems to be no problems. Is there a xserver library that works with HP 48Sx? === Subject: Can Metakernal be temporarily disabled? Is it possible to disable/enable Metakernal temporarily, without clearing the memory and reinstalling Metakernal? The softkey Kernal indicates version 2.29, not 2.30. Why is there a version discrepancy? I thought the HP 48GX was slow because of the microprocessor when I purchased it back in 1993. It was a frustrating experience learning to use the calculator because of the slow response. Now, after using Metakernal, I know this is not the case. Was the interface of the HP 48GX written in sRPL? It is unfortunate that there were no further improvements to the ROM, especially since the operating system was being developed for the HP 49g while the HP 48GX was still in production. === Subject: Re: Can Metakernal be temporarily disabled? > Is it possible to disable/enable Metakernal temporarily, without > clearing the memory and reinstalling Metakernal? 64 SF, then reboot It's MetaKernel