A56 ==== Is there a trick to pull of the protection cover from the HP49G? I really have to pull hard and i always scratched my display. Is ther a special spot to push? ==== I'm glad to see you've brought this up to date with L1791. Perhaps with a bit of work, we could merge our efforts (add the extprg caller, expand to non-numbered items, etc). I must disagree. The author's choice of whether or not to offer HELP is the *author's* choice. However, that often has a tendancy (and we as programmers encounter this often enough) to frustrate the user either with its lack or its incompleteness. Giving the user the ability to provide their own marginalia is the sole reason Sola and myself, among others, were writing our own HELP commands (such as mine in L1791). Mine's more of an extended browser, and could/should probably be re-written using something less clunky, but the fact remains there will always be a need for, at the least, adding annotations to external libraries. ==== It self reboot when open the calc sometime while using it but it's not happen every time.I have reinstall ROM but it doesn't fix this problem.Is this bug? ==== You appear to be complaining that the calc warmboots sometimes when you open it (the case?) while using it. OK. Leave it off when it's open? ==== Try leaving it off when it's open. ==== The only problem I've ever encountered with unusual IDs is that when one is used as a variable name, you can't conveniently press the key for it when entering something (like if you wanted a list of two variable names, you'd press {} and then the two f-keys -- this won't work for '1A'. This also isn't a big deal, or difficult to handle.) ==== What is the key sequence in the kbd2 test (ON-C), in the 49? I tryed searching the old messages but I just found the sequence for the 48. ==== I continued improving the MINEHUNT game to make it particularly attractive for children at Christmas. The new version on my site below is about 1 KB smaller than the original (now less than 2.5 KB), and generates a corresponding sound if running on a mine. I excercised it with some children. They where exited and want nothing else than getting a HP49 :-) ftp://ftp.math.fu-berlin.de/pub/usr/raut/HP49/games/ ==== Gee, Wolfgang -- you sure that's political correct ? Somehow the terms MINEHUNT and children don't fit very well in my mindset... Huh, training children to run into a mine ? Makes me wonder why that game was named MINEHUNT in the 1st place -- MINEAVOID or SNEAKTHEM seems more appropriate. Be careful the BPjS (http://www.bpjs.bmfsfj.de/) doesn't get ahold of the game. http://mein.hamburg.de/homepage/grendel ==== Children do quite well distinguish between a game and reality. Of course, I took the occasion to mention that there are children in the world being in real danger of running into real mines. But his made no particular impression on them. They were fascinated by the game, first exercising on the open mine field which is quite interesting in itself, with about 25 mines scattered around, to find a way through the labyrinth. Do you know how fast the kids learned to store into or recall from the variable Nmines? When playing with covered mines I was again surprised how quickly children learn drawing right conclusions from the numbers of adjacent mines. Particularly clever was an 10 year old girl (which shows that girls in that age are not yet fixed on roles in the later life :-) ==== The screen cover of the 49G manages in some probably complex and half-magical way to acquire dust underneath it, which is probably a design feature of screen covers since every cover I've ever seen gets dust under it. I was OK with this, since I didn't see the dust in most lights, but a few days ago it appeared -- there's no way to miss such a thing, and no light (except too little light?) exists to help me see the screen beyond it. After a few minutes of just... noticing the black might be a good idea to pry the cover off. After all, I've read reports here of people who've pulled it off, say that it goes back on fine, and reports from one person who removed the cover and left it off, leading to bad damage in a minor fall. So, after I was annoyed enough I got a little screwdriver and managed to remove the screen cover, with some minor scratching of the case in my clumsiness. It comes off with a little effort (but isn't much easier to take off, after it's already been removed once). I then cleaned the LCD screen and the inside of the cover with a cloth, and put the screen cover back on. Now it looks *fine*. Really clear. ==== *If* the source code is released... JYA has indicated that that will probably not happen. ==== I believe he means a partial differential equation, with spatial and temporal dimensions. Of course Mathematica and Maple can do it. ==== for all of you that wanted a Handheld PC with keyboard in, here are some options: HP Jornada 565/568 with its new keyboard accesory: http://www.jpn.hp.com/db01/pressrelease/fy2002/img/cbo02keyboard.jpg Sharp SL5500, a beast, with linux in, keyboard in..: http://www.the-gadgeteer.com/zaurus-sl-5000d-review.html http://www.trolltech.com/products/qtopia/sharpimg.html At least, with Sharp macihe there is the option of running Giac/Xcas soft from B.Parisse. One bad thing for scince use of these machines is that their keyboard is not designed for numerical input as main use. -- ==== says... No, not at all... We are well aware that it does not rank among the 'good beers' ! ;-) Werner Huysegoms xwerner_huysegoms@freegates.bex (delete the leading and trailing x) <3c154422.1512374@nntp1.onemain.com> <9v3jc4$bda$1@diana.bcn.ttd.net> <76ea4fd3.0112110509.1137d071@posting.google.com> <9v5rdh$4bd$1@nreada.inwind.it> <9v6a75$nb$1@diana.bcn.ttd.net> <9v93ro$9v1$1@diana.bcn.ttd.net> <9vauji$hih$1@nreada.inwind.it> <52RT7.11414$zB6.11735@newsfeeds.bigpond.com> <9vpeoq$p64$1@nyytiset.pp.htv.fi> <3C2060FB.FB9CBA85@student.tnw.tudelft.nl> ==== What about Carlsberg and Tuborg (and Albani and many other Danish brands whose names escape me right now :-) ). - hans kristian - ==== Sure, easy for you to say, you're Belgian. who can forget Duvel, Brugse Tripel en Straffe Hendrik. Personally I'm not fond of Leffe but I hear some people are. I mentioned Heineken because most people know it, I drink Hertog Jan myself -- ==== I was wondering if any HP calculator users are Tolkien fans too? Harold A. Climer Dept. of Physics,Geology and Astronomy U. Tennessee at Chattanooga ==== Oh, most definitely! I first read The Lord of the Rings in about 1972 and have re-read it (and The Hobbit, The Silmarillion, etc.) more times over the years than I can remember. -- rs, Silverlock ==== If I was clever enough, I'd try to write an Angband port to the HP49 :) Greetings, Marco ==== I read the Hobbit and LOTR for the first time in the early 80's. It is hard to say how many times I have read them since then. I know I have listened to the BBC audio versions at least 50 times. It is what I listen to when driving or flying an great distance. Charles Perry P.E. ==== Tolkien, Yes. Star Trek, No. I wish they never named it vger. G. Savage ==== The best person for a job is one with a brain. Definetly not you..... -- http://193.145.89.239/~rblasco/ Un222os, hermanos linuxeros 640 Kb de memoria son m207s que suficientes (IBM AT Designers - 1982) Bueno, espero que esto no aparezca en la versi227n final (Bill Gates en la presentaci227n de Windows Crash Debug '98) ==== I beg to differ! Ricardo has something here: the best man for the job is indeed always a man -- for the best man couldn't really be a woman, could to give birth to a baby is always a man. Mrs. Fiorina, though, is neither best man nor best woman. ==== I'm sure the intention was to say sthg like the best applicant or person If not, which I really doubt, as It was sesxist comentary, sorry. Ricardo -- http://193.145.89.239/~rblasco/ 640 Kb de memoria son m207s que suficientes (IBM AT Designers - 1982) Bueno, espero que esto no aparezca en la versi227n final (Bill Gates en la presentaci227n de Windows Crash Debug '98) ==== Since nobody has been able to develop a WindowCE version of Giac/Xcas (the most powerfull GPL CAS I've ever seen), Emu48 is the best free math option for PocketPCs. I use Emu48 in my Jornada525, the cheapest color mahcine nowdays, and it runs fine. There is some problems with color palette (it turns red several times) but could be repair de-selecting Always display KML Compilation Result. PS:I would like to conect my hp49 with my J525 using emu48, but it seems not posible :-((( ==== More or less, twice the real hp49 speed.. more or less.. ==== Which one? V13N3? Or V13N4? - where the '42S version appeared ;-) http://groups.google.com/groups?hl=en&selm=9404210502004074%40nybble.com ==== I went to my local electronics supplier and tried to find a match for the plug that goes into the top of the HP38G. They tried various different ones and couldn't get a good enough match for the pin spacing. In the end we concluded that HP had used a proprietary version that was not used by anyone else. What did you use? Is it commonly available? I ask because I often get enquiries via my website from HP38G & HP39G users about whether they can make their own and I've had to say to them I can't help. (My knowledge of hardware and soldering is very limited - I'm a software sort of person.) ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== ==== i couldn't find a connector with the HP spacing either but you could make one!! you can make a very small PCB (printed circuit board) with only 3 pins that maches the spacing. in fact, if you take apart a HP49G connector, there is a tiny PCB inside! i modified my connector so that i can also use the pin for the 5 volts power (used for the external projector). but i don't think many people would go as far as me. he he ==== Abrevations mil = milli-inches mm = milli-meter available in the HP-Units catalog. The HP serial connector dot pitch has the European format 2 mm in opposite to standard device dot pitches from 100 mil = 2.54 mm 75 mil = 1.905 mm 50 mil = 1.27 mm The 2 mm dot pitch is nearly often use by companies for connectors when they only have little room. I got my 10 pin 2 mm dot pitch connectors from used PCB's. I cutted the outer pins on the bottom side (only the three in the middle are use), so I was able to solder the connctor on a 100 mil dot pitch PCB. A paper clip, a cable, a 9 SUB-D female connector and some glue and ready was my own cable. Sorry I haven't a camera yet to show you the result. ==== Be sure to also read (and mind the linebreak) http://cgi-bin.spaceports.com/~hpkb/view.php?code=5e1ca6b67d16fcb7b82e60f97d 517b94 which can shed some light on why some people won't just fork money over to professionals like those people of www.samsoncables.com As will *no one* else, you fool. ==== How can I use the HP 48GX to calculate dates? I used to do it easily w/41cv but the 48 has me stumped. ==== Well, calculating with dates is easy on the 48. on your settings. The command DATE returns the actual date. The command ->DATE takes a number in the format given above and sets the current date accordingly. The command DATE+ takes a date and a number of days from the stack, adds the number of days to the date, and returns the new date. The command DDAYS takes two dates from the stack and returns the number of days between the two days. The command TSTR takes a date and a time (format hh.mmss) and returns a string that contains the day of the week, the date and the time. (Time stamp) ==== [Beep] That's not me, that's my HP49G wanting the case badly! ;-) [Beep] OK HP, I ask about the case. In case I want the case how can I buy it? What should I... [Beep] I told you HP, don't interupt me! What should I do? [Beep] Useless! ;-) convinced. :) Hopefully I can find the PPC and CCD modules, over time. One other quick question: What's the difference between the math/stat module and the individual Mathematics and Statistics modules? Is the math/stat a feature-by-feature combination of the two, or is it more of an all-season radial, maybe with the greatest hits of each one? ==== last week i finished my tests and now i have time to play. now, it is possible to read and write to a flashcard!!!!!(both HP49s & HP48s) not only to a flashcard (conpact flash card=CFC), but also to the IBM microdisk and also PC hard-drives !! => NO MORE MEMORY LIMIT!! i've just tested it (CFC). everything works!! the chip cost $27 + i have to make a small PCB (printed circuit board) and the connector (rigth now i am using a PCMCIA where i tested the flashcard. the chip manual says it also works with microdrive and harddrives) i works at 9600 bps (TX & RX). it is very easy to operate. it took me only 10 minutes to program the commands and i was on the road writing and reading. also it uses reverse polarity, like the HP49G. so you don't have to invert the signal. it uses minimum power (except if you would use a laptop harddrive .5A) 5V. I haven't tested the current yet, but i guess is around 20mA. i don't have a PCB layout sofware and i would need it in order to design it to a very small size using the connector for the CFC which has a lot of pins!(i want to make it a little bit bigger than a match box, enough for the flash card and 2 AAA bateries). anybody has this sofware? once the PCB is designed, i can place an order for the PCB in small quantities (10 pieces). PS: i am doing more interesting things (a lot more than this one!), but you ==== HP48s) Hmm, that sounds interesting. A pocket-size backup device? Let us know when it is finished, will you? ==== you can already buy one. this is where i got it from: http://www.star.net/people/~mvs/SDRVSPEC.TXT info. note: the $27 applies to 1000 units! they have a PCB with the chips and a DB9 connector already made out. it sells for $99. all it needs is a +5V suply (the name is 'STAMPDRIVE' or SDRV-1, there is a picture on the main page at the bottom). the DB9 connector is female, so you would need a double male adpater to connect to the HP49G. so far, i write and read sectors using commands (i.e. 'xmit' the command followed by the data). you can move around the sectors and read and write them. the minimun write block is 512 bytes, but you can increase it (so you would not have to use a lot of R/W). what i do is i make a string ->STR and i send it to a sector. if the string is bigger than the sector, it automatically appends the rest to the next sector. so far, i made a menu with all the commands (example for write: << w ' XMIT DROP XMIT >>). you can also send binary (not tested yet, but it should work). the type of data does not matter. it only has a total of 13 commands. ==== You mean it is just plug'n'play for my 49G You must be heavily overestimating my HW talents. What I mean is that I'm interested in buying a complete product once you get the mass production flowing. , ready for128MB HP49GX^X ==== Well, with a whooping 9600 Baud transfer rate you cannot really call this a RAM extension, it's just an external disk drive. :) ==== What we really need is a card that: 1) is self powered (important!) 2) connects like the existing memory cards (importan too! i don't want things hanging from my calc!) 3) has a stand alone program for retrieval/storage of data (much like while a processor can address only soandso GB of data, hard drives have way more than that) quoting the entire previous message >:| some people here have really limited connections, you know >:| Let's just say that my calc isn't much slower at serial transfers.) 5) has so much storage that we won't ever need to transfer anything ever again (in other words, what is the size of Eric's hpcalc.org archive...? 276MB.. ok, lets make it 200 MB because some of them are pictures/pc files.) 6) makes Carly Fiorina give more money for devel on the calc department >:-] 7) gives me enough motivation to study my physics :((8) can withstand a little abuse. i am not talking that it should pass a 1e6g test, but it should not fail while i use my calc normaly. you think a hard drive can do that? way to go, rcobo! What about a module that: has the volume of TWO memory cards (this way we can take advantage of the room lost between two ram cards when putting them in the GX/SX) has two connectors, so this way the gx thinks that there are two cards. The module has: (pick anything you like) memory, AD/DAC, modem, USB, power supply, wireless network capability (i've seen that someone has done something like this somewhere) now, if we can get clement and stephane to work more on their os's and ask our help... so we can make proper support of this GXX module...teeheeeee ==== I simply PURGE the libraries from the filer and as you say, all is fine. 2. For this question, I don't have an answer. ==== something in the process. As far as the speed difference for running Libraries from Ports 1 or 2, I haven't noticed any difference on my 49G. I'll assume there is no difference (or that the difference is insignificant). Steve S ==== Abaout speed, I can tell you what happens on 48GX, but I don't know if it applies to 49G. I don't know its memory distribution Àok? I've have read about that speed difference in the documentation of some libs that recommend you install in ports 0 or 1 because speed and I know that ALG48's doc., says something like a special programming technique was used for avoiding the slowdown when running in ports I have never make time tests, but I like to think that this delay is, And I wonder, why it is so popular to pick on poor Bill Gates. Why people are so jelaous of his success ? In fact HE IS the creator of the most successful software company in the world and most used piece of software in the history of the computer industry whether we like it or not. Is it because he saw an opportunity, where smart people didn't notice anything ? Sure, there will be thousand of excuses, how unfairly he did it, and how underdeveloped his software is :-) Assumption of course is, that if WE WERE HIM, our software would so much better. What a shame that we were born too late and missed the train of opportunity altogether. In fact many of his critics are older than him :-) Undoubtfully it would be great to improve the units capabilities, but until then and especially for Micas' problem, what about something like: << RCLF SWAP ->LST DUP 4 GET '1_ft' CONVERT 4 SWAP PUT ->ALG EXPAND SWAP STOF With all necessary changes of course ;-) ==== Do erable have all the function of alg48? I have Hp48GX I don't think there is enough memory to install both. ==== If you do decide to kill your mother, please try to make it look like an obvious and tragic accident -- then you can get to using your 49G early. If it looks like a murder, or is mysterious, or is obviously your doing, you may not get to using your calc much earlier than Christmas, and may not even get to use it *at all*. Less violent possibilities might be hypnosis (you *will* give me the calculator early), a great speech, or the holding of household pets hostage -- but be careful with this last, as she may think of holding your 49G hostage for the pets! It may be possible also to claim longterm abuse of her upon you, and get social services on your side (and then use the 49G while healing in a hospital from self-inflicted wounds). There are indeed lots of possibilities, but you should look both on the immediate consequences and the distant consequences if you are to get maximal pre-Christmas use out of your wonderful calculator. ==== I wonder if any of you knows about an EMU48 version for Windows CE, where the menu bar of the EMU is in the Title Bar and not in the window body. The reason why I need it, is that this menu bar is so much in the way, and I don't even use it. The best solution would be, not to have this menu bar at all. I am doing a 39G/40G bitmap and kml for my Jornada. The config of my Jornada: HP 680, SH3 processor, WinCE 2.11 Core, display 640*220. I have tried the below EMUs (actually the only one, I found) and it isn't the one I need: HP48.exe 131,072 bytes 12/01/2001 V1.23 (I think) Budapest, Kristof ==== I wonder if any of you knows about an EMU48 version for Windows CE, where the menu bar of the EMU is in the Title Bar and not in the window body. The reason why I need it, is that this menu bar is so much in the way, and I don't even use it. The best solution would be, not to have this menu bar at all. I am doing a 39G/40G bitmap and kml for my Jornada. The config of my Jornada: HP 680, SH3 processor, WinCE 2.11 Core, display 640*220. I have tried the below EMUs (actually the only one, I found) and it isn't the one I need: HP48.exe 131,072 bytes 12/01/2001 V1.23 (I think) Budapest, Kristof ==== Why LAP('1/(X+1/a)') returns 2*exp(-1/a*X) and not exp(-1/a*X) ? ==== Looks like it - read this: http://www.semiconductor.agilent.com/cgi-bin/morpheus/promotions/promotionsD ==== Ok, Today I used for third or fourth time emu48 at work, and I always begin from new (I never save, so it always ask me for Try to recover memory? and I always answer no). I enter in EQW, type e^, and I get EXP(), the same in the stack. My flags settings: { # 204010FF0h # 0h # 8000000002000000h # 0h } ==== J.Manrique Lopez de la Fuente schrieb im Newsbeitrag do to course schrieb Sorry I see absolute no difference between Emu48 (v1.29) running with 1.19-6 and the original machine. Always library 258 attached, but this doesn't matter. is a calculator key left shift right shift <- input cursor I used your flag settings: Typing Result <- e Typing 8 Result 8 e Typing Result <- e Typing 8 Result 8 e Typing <'> 8 Result 'EXP(8<-)' Typing <'> 8 Result 8 e Have you active libraries in port2 on Emu48? I never got EXP() in the EQW environment. ==== And when you type VERSION, what do you get ? Are you sure it says 1.19-6 ? Or in fact, are you sure you're using the official beta 1.19-6 that you can find on my beta ? If you type VER you should have a date after September, 1st I've sent a couple of pre-version of 1.19-6, it seems that you got your hand on one of them.. ==== I am sure it is the last one.. the oficial , not a pre-release. I use it for my hp49 usage classes, using a emu48 in a PC with a proyector, to show in a wall what I do, and after some use, (normal use for a basic calculator usage course), the EQW begans to display e^ as EXP(). J.Manrique ==== program Fonts without assigning it to a key? When I imported it into Lin ==== Did you transfer the program to your HP using ASCII transfer? If so, then try again using binary transfer this time. Greetings, ==== with 110V power adaptor. works just fine $60 shipped to USA ==== That's incredible! Who made it and what model is it? Sincerely, James ==== Except this is false advertising as this device is definitely not an IrDA device ==== my bad!!! I mean Infrared! sorry..... ==== I will appreciate it if you could help me to handle ON-A-F in my simple password program (that act as a NUMLOCK) the users? %%HP: T(3)A(D)F(.); @ RS232 goes down for 2 seconds to -4V or -3V across 1k. << -33 CF @ I/O directed to serial port. Enter the password CLLCD 1 DISP 0 0 0 0 82 83 84 72 0 -> a b c d e f g h z << WHILE 1 REPEAT DO UNTIL KEY END 'z' STO b 'a' STO c 'b' STO d 'c' STO z 'd' STO a e == b f == AND c g == AND d h == AND IF THEN OPENIO OPEN CLLCD 3 DISP 2 WAIT CLOSEIO OFF END END >> >> Tal P.S. NUMLOCK program is a non-stop program that catchs a series of 4 (or more) digits and compares them to other known digits, when the digits are the same, the calculator outputs logic 1 for 2 seconds via RS232 port for activate an electrical lock. ==== I think that ON-[SomeOtherKey] kombinations just can't be handled with UserRPL. Greetings, ==== for Ni-Cd batteries try www.partsexpress.com www.mcmelectronics.com/welcome.jhtml ==== A true engineer would have found the raw materials, designed and built the batteries themselves! :-) ==== Well, I didn't want to show off..... ;-) Roger ==== You should mention in such questions what calculator you're talking about. It sometimes matters. If you have the HP49G, library 256 has a command ->PRG that does what you want. ==== COMP-> is redundant. ==== On the UserRPL side, you can use the command sequence OBJ-> ->PRG to do this. Greetings, ==== which few ever give and which I've perhaps wrongly gotten tired of seeking. ==== It would also be nice if you uploaded that to my site, at http://move.to/hpkb . For the site, only text and html formats are allowed. -- Eduardo M Kalinowski http://move.to/hpkb ==== Well, yes. Libraries evaluate to themselves. To install a library, you must store it in a port and then attach it -- the simplest way to do this for you would be to go into the filer, use the MOVE command in the filer to move the libraries to port2 (flash), and when this is finished press ON and C (f3) at the same time. This will effect a warmboot and your port-resident libraries will have their $CONFIGs evaluated. You should then be able to use the documented commands of chemlib and UFL. ==== one question - can I calculate leasing on HP17BII using so called rule 78 (interest is loaded up front)? I could not find any solution for that :-( Many thanks for all comments BR Johny ==== Ah, I am both gladdened and saddened -- that you have the joy of using the most excellent 32sii user interface re fractions, and that I with my HP49G am not using such an interface. Truly such great divisions in the world of calculators determine the fates of many, and turn the winds to blow in strange directions. If only there were some to way to bind all the disparate features and excellencies to convenient union, or perhaps rather to divide them in important ways, so that specialized tools may attack each in its time. pawlrevere2, or whoever you really are, I congratulate you in your own step, your passage to greatness as you enter the spheres of greatness. When you work with fractions using the particular user interface of the 32ssi, work with them in beauty and pride, and by implication of your message also work with your 48 or 49, and know the calculators by their differences. Having said that, could you please describe for everyone here who does not share your quality, that of having a 32sii, exactly how that great calculator handles fractions, so that we may know its greatness and perhaps even create feeble imitations of it for our own benefit? ==== Does any one knows the name of a program who works as microsoft Excel??? I'm really needing one. Andre Ribeiro ==== Try Xcell, it is very good. You can even import/export files to Excel. http://www.hpcalc.org/hp48/apps/matrix/xcellv20.zip Otherwise, look into http://www.hpcalc.org you'll find others that may better meet your needs. Diego ==== True, but ti's emphasis is on C, not assembly and c on the calc does not make sense to me. And there is more to copyto_h than just a rom call. Besides most people do not know that the rom call exists and most are not assembly programmers. Most are just users and don't give a rip how something was programmed. huh? How is basic trying to emulate assembly? This is probably true. Ams 1.00 exports about 1,000 functions and the new ams version about 1,500 rom calls. could somebody tell me how much the hp exports? This probably is not a true measure of how open a system is. And How about tsrs (terminate and stay resident) programs on the 49g? The tios is very extendable with tsrs. Would it be possible to build an auto completion feature into the entry line like i have done with my program complete? I also have plans to impliment a user-mode for the 89 by the user pressing the on button. Don't get me wrong as a die-hard 89er. My 49g is in rpn mode and I like it a lot because the interface and useability is very well thought out. I am just a newbie. -Samuel ==== But why not use functions? ==== the hp means nothing, but since I have an HP49G I will talk about it. I've heard 2,000-something, which may be true -- I don't know how to ask extable how much more there are. hpcalc.org has old entry lists, and the Entry Database (not on hpcalc) might be better. Of course, there are as many rom calls possible as there is ROM on the calc; the Entry Database describes a number of unsupported (read: not in extable) but useful entry points. Most SysRPL programmers will just use what extable has. Actually, a SysRPL tutorial by someone said this (very modified): So, you heard of all those commands in UserRPL? Well, the HP49G has *thousands* of unnamed commands available to the SysRPL programmer ... which HP had in-house names for. Auto completion as you describe is certainly possible; you might want to look at the interface and stack sections of hpcalc to see some of what people've done. user-mode means nothing to me. TSRs -- cooperative multitasking as in DOS? I don't know of anything like that, though I doubt that this gives you any power not available under the HP49G. ==== Yes :-) ==== This means that I can copy some code to ram, lock it (the handle) so that garbage collection does not move it and modify some global pointer to point to that code in ram so that when something specific happens the code in ram is executed. Tios makes it very easy to watch some of the stuff it is doing and what keys are being pressed. ==== But we have TIGCC. That is unlikely, at least for non-Flash programs. This is only one way... there is always TI-Basic, in which also you can write pretty advanced stuff. But we generally write in C, not in assembly. There could still be memory leaks, but software should always be tested before releasing it anyway. It is slow, but that is because it is an interpreted language. Then you write it. ==== Please don't supply silly reasons in lieu of real information. The FAQ (see hpcalc.org) talks about the HP49G's infrared port, such as it is (it isn't). ==== I assure you, they do it also without IR. ;-) Greetings, ==== How can I make grayscales in 64 shades, and then view them in my calc? What software do I need (for the calculator and for the PC)?. I tried with Xnview (pc), but I cannot save the files in 64 shades, only in 32. I have an ignorant question: what is a grob? Is the same as a grayscale image? How can they be viewed in the calc?. (As you can see, I don't know anything about graphics or images in the calculator, any information or explanation will help!.) Sorry for the english, please correct me if there's any mistake, I'm learning!. Camilo ==== Software for making grayscales on the 49 is at http://www.hpcalc.org/hp49/graphics/editors/ and also at http://www.hpcalc.org/hp49/graphics/misc/ but I don't know how many grays are supported by these programms. A GROB is the GRaphics OBject of HP49. It existed long before the grayscale image and it supports only black and white pictures. You can view a GROB by putting in on the stack and pressing the down arrow key. This takes the GROB in the built-in GROB-Editor of the HP49G, where you can view it and do some elementary editing. Your english is just fine. :-) Greetings, ==== 32 shades is pretty good. What are you looking at, that you want 64 shades for? ==== inspiring c00l :) now I only need a pic from Wolfgang... so I could also join rcobo's t-shirt competition :) My T-shirt would be ideally made of tiled pics with faces of all the people Greetings, Marco ==== http://www.math.fu-berlin.de/~raut/ BUT the picture is perhaps from his thesis day... :-D X rcobo's people ==== I think this are more recent: http://www.math.fu-berlin.de/events/math/festkolloquium_raut/page_01.htm ==== Am I mistaking or do I see there a bunch of Germans people drinking ... wine ?? What is going on in this world if the basic rules are not respected anymore !!?? Germans are supposed to drink beer and only beer (and good one)! ==== Well,.. Picture N.8 should be Professor Wolfgang drinking wine, if I'm not wrong ==== You don't need to turn off the clock display. Just updgrade your calculator's ROM to 1.19-6 and you will get a lot of goodies (and no problem with garbage collection). Try this link http://www.epita.fr/~avenar_j/hp/49.html to get 1.19-6 and instructions for upgrade. Saludos Jorge M. Valenzani ==== Waste less money at www.samsoncables.com and please stop whining. ==== I have two questions regarding Libraries and Ports: 1. Is it really neccesary to detach a library before purging it? I've updated a number of libraries lately, and accidentally purged one before detaching it. It seemed to purge just fine. I tried purging some more and they all purged just fine. Is this a feature of the HP49G only? 2. Is there really any difference in storing a library in Port 2 rather than Port 0 or 1 on the 49G? The instructions for the HP48 say the library will operate slightly faster if stored in Port 0 or 1. Does this also apply to the 49G? ==== Once you get used to it, it is very useful. You mean the HP49 Informbox engine from debug2? Did you try http://hpmad.zoy.org/eng/informbox.html on Cyrille«s side? Also search fo IfMain. Pieter Kuyck explained it to me once and from ther on I could make my way throug. Try this: http://groups.google.com/groups?hl=de&threadm=8vc6gs%24p9pg%241%40reader4.wx s.nl&rnum=1&prev=/groups%3Fhl%3Dde%26selm%3D8vc6gs%2524p9pg%25241%2540reader 4 .wxs.nl (everything in on line) Hope that helps. AM ==== Don't try this at home - your memory will be lost! In RPN mode go to the filer and create a new object of no type: [LEFT SHIFT] [FILES] [RIGHT] [NXT] [F3] [DOWN] [ALPHA] [A] [ENTER] [ENTER] If you hadn't already an object named A this created a PROG but without the programm delimiters. Now edit the object to put some (silly) content: [NXT] [F3] [+] [ENTER] An error message too few arguments will appear shortly. Now quit the filer. [ON] Now your calculator is loaded. Press some keys (e.g. [ENTER]) and it will fire a TTRM. I don't know whether this was reported before. Thomas 19:17:45 -0600) ==== Thomas Becker found a bug (present in 1.19-6, perhaps earlier). It is unrelated to the Filer, and unrelated to the original (valid) variable content. E.g. (backup memory before trying): 123 'X' STO CLEAR 'X' VISIT While editing, replace 123 with + then press ENTER (an expected error occurs, but apparently isn't well trapped). Trying then to recall or use 'X' won't be healthy. If this were the Microsoft Security Notification Service by this sort of additional information: Mitigating Factors: ==================== User Error: - The vulnerability could not be exploited if the user did not try to replace the object by an invalid operation. http://www.microsoft.com/technet/security/notify.asp [r->] [OFF] ==== That would be HP-6S not HP-65 -- Tom Lianza Technical Director Sequel Imaging Inc. 25 Nashua Rd. Londonderry, NH 03053 ==== I'm looking for an HP 41CX calculator really in mint condition, provided with its original box, case, AC adapter, manuals and a Card Reader with cards. Have you heard from someone interested to sell? William Tomaselli Harold A. Climer ha scritto nel messaggio ==== I don't know what the highest price ever paid for an HP is, but your comparison with Hewlett Dackart, errr, Harley Davinson, is just nice. Not so fast, but trusty. Despite my complaining about speed, I'm going to keep buying Harley Packinson, errr Hewlett Packard calculators. Better, the devil you know! ;-) Or was it, better the devil, you know? ;-) Greetings, Newckett Kackard. ==== I've encountered the same problem and also my memory stick slot will now not recognize any memory stick inserted. I'm taking mine back to the store I bought it from ==== In this week's isue of EETimes (isue 1197) they have a special report on HP on the front page.it is on page 1, 17, 18 and 20. i went to their web-site, but they haven't put it on the web yet (at least i didn't see it, probably next week). The title is hp's lost way: a special report. it is good stuff. http://www.eetimes.com/story/OEG20011214S0063 http://www.eetimes.com/story/OEG20011214S0062 http://www.eet.com/story/editorial/OEG20010917S0055 23:58:17 -0600) ==== One suggestion was: And deletes all variables, leaving user memory empty; as an alternative, try ON+A+F Yes This invokes a Recovery Memory procedure, which will delete corrupted variables, hence may recover your lost memory; however, it may also delete non-corrupted variables (it may not even find all your variables, even if nothing is corrupted); finally, it will almost certainly delete your alarms and user key definitions (and anything else stored in the hidden directory), and also it will unhide stuff that was hidden by any variable hiders. It might be prudent to make a backup first (at least you can go back and recall one object at a time, if it later turns out that Recover memory loses your stuff). [r->] [OFF] ==== This has never worked for me, I regularly loose about 70 kB (Jazz) after a crash and the only way I can get them back is No -- ==== another way, maybe using some software (like the handy OT49 lib). The reason I asked is because I'll be traveling and I might not always have access to a PC (for BACKUP/RESTORE) and if this happens again I probably won't have (an easy) a way of restoring the lost memory. Fortunately I had a recent backup that I used to restore my work. This time there was no harm done. Usually I'm not very diligent in making backups. This time I glad I was. I'll try to remember this for the future. There is nothing like being able to type RESTORE and everything goes back to normal. Life, unfortunately, does not have this feature :-). Diego <3C1B9617.3A0112ED@iname.com> <3C1EDDD9.50A85153@miu.edu> ==== You don't need it. Just save a backup in port 2, like this (RPN mode): :2:filename ARCHIVE Then you can restore again with :2:filename RESTORE I've never managed to lose any of my flash memory. Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== When a library or any object stored in port0 crashes the calculator by writing on itself this object will be marked as unvalid and the message Invalid Card Data will appear. Unfortunately, it also means that the space used by this object is now lost. If you look at how the HP48 memory organized you will see that the there's a pointer located in 80711 (USEROB) which contain the address of the HOME directory. In 80716 (ROMPARTS) there's a pointer to the beginning of port0. If you want to recover the memory lost after the corruption of an object you will have to move the data contained between these two pointers and update the pointer. It could be done by a simple program which I'm sure somebody will write one day :) <3C1B9617.3A0112ED@iname.com> <3C1EDDD9.50A85153@miu.edu> <3C1F0BF3.CC407A7C@student.tnw.tudelft.nl> 19:43:32 -0600) ==== Perhaps the HP49 Recover memory doesn't work as well as on HP48? Or is it just bad luck of the draw re memory contents at the time? [r->] [OFF] ==== results, this NG has become one of my dialy stops for news. ==== such a simple (for you) program. Mr. Nousiainen X you one This message was written with 50% of positrons and 50% of electrons It will automatically annihilate when you read this. The resulting energy burst will clean up any viruses on your HD (including any data, programs, sector info, coating, magnetic material...) ;-) ==== Just use ARCHIVE and backup to your flash memory, so you don't need a PC handy. Greetings, Marco <3C1B9617.3A0112ED@iname.com> <3C1EDDD9.50A85153@miu.edu> <3C1FC718.919ECD5E@iname.com> ==== You're welcome =) As an addition to this, I always save backups as :2:Byyyymmdd, where yyyy is the year, mm the month and dd the day. Furthermore, my standard flag setup is in :2:FLAGS. Then I have the following program stored in :2:R: << :2:FLAGS RCL STOF 2. PVARS DROP 1. << IF DUP VTYPE 15. =/ OVER ->STR 1. 5. SUB :2: B =/ OR THEN DROP END >> DOSUBS REVLIST HEAD RESTORE >> It automatically restores the most recent backup. When I get a TTRM, I just hit NO, then later - now in ALG mode - type EVAL(:2:R) ENTER to call the program. If you have OT49, prepend a D dash to gain some speed. Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== i hope i am not late try purging any object you had in port 0, especialy any libraries that you do NOT see in the port 0 menu, but you HAD installed (that's correct). the calc thinks that the library is still there. of course, ON-A-F [YES] works too, but you lose some settings. not very important, but it is not a good solution. i even think that on www.hpcalc.org there is a program to do an even better recovery. ==== i _think_ there is something wrong with jazz 6.8 and its VV. I got a heavy crash today and i thought i'd give it a try... it froze up really nasty in VV, removed batts, lost jazz. retransfered to make the checksums somewhere i wouldn't have to download it again and again tell me your story. what have YOU been tinkering with ;) ==== Explanation: I've made some programs which share the same vars. I've also made a program which asks for this vars, so that the user doesn't have to deal with the vars. I don't want the vars to be vissible (I dont't want some subroutines to be vissible either). So I thought that making a library is my best solution. The problem is that I want menu keys to do different actions if they are pressed with right or left shift (like in CST menus). Questions: Is this possible? Is there any program which will make this for me in a non-painfull way? If not, how can I do it myself? What number do I assign to the library? ==== Sure, it is in fact an absolutely trivial process, which requires not even SysRPL. See the TMENU UserRPL command. To make some variables invisible, put them in $HIDDEN instead of $VISIBLE when you create your library =) I'd suggest that you get OT49 and use its D<-->L to look at a few small libraries and see how they're constructed -- and also to create yours. ==== Well, when you enter the library menu you see the visible variables. Why would you want to create a temporary menu for this? And, if you're entering a 'library menu', you're running a command that could just as easily been one to bring up a temporary menu from the library. Not AFAIK. ==== I know I could have done the program in UserRPL, but I'm using SysRPL to make it faster. About using TMENU, you mean I will have to make a program (the only thing visible in the whole library) which will create the temporary menu. Is that, or can i make the temporary menu appear automatically when you enter the library menu? Are there any problems using hidden vars where I will STO things (cause I read USRLIB's documentation and it said something about this which I didn't really understood)? I still would like to know which library id can I assign to my library. (It will probably be used by other people). ==== So that different things happen if you press the menu keys followed by right shift or left shift. Anyway what I will do is what I thought you meant: the only thing vissible in the library will be a program which sets up a temporary menu from which you can access all the different commands of the program. ==== The following table is based on in-game experimentation and simple mathematical analysis. Armor Armament Range Cost Trike 0 0 3 1 Char leger 1 4 4 1.17 Mobile 2 2.17 Moisonneuse 3 1.67 Char d'Assaut 2 8 5 1.67 Destroyer 4 10 4 2.50 Lance-flammes 1 16 7 2.17 Tourelle 4 16 5 2.02 Tourelle-R 5 22 6 2.33 1Cost=192Spice deltaH(strikee) = 8 - Armor(strikee) + aRmament(striker) Portants provide 10 transportations and then are destroyed. At a cost of 448 a portant, you get approx 44spice per transport. Armor Chantier 6 Piege a vent 4 Silo d'epice 4 Usine Legere 8 Usine Lourde 8 Raffinerie 6 Reparation 6 There are significant speed differences among the units, which I do not care to measure. These would be important if, say, you intended to approach a stationary target with a unit. How many times would a trike get hit by the tournelle-R before the trike gets in range to fire, and how much damage could the trike do before being destroyed? Not cost-effective answers, I'll tell you that. The range value above may be a little confusing. To see how it really applies in the game, imagine a unit at (0,0) on a cartesian coordinate plane, on which x=r, x=-r, y=r, and y=-r are drawn. Any unit on or in the square these lines form can be fired upon. Note, finally, that distance, relative healths, and many other possible and concievable variables in the consideration of deltaH are all irrelevant. In this area, Dune is a very simple game =) ==== runspeed shootspeed trike 3 .75 char leger 3 1 mobile 1.5 moisonneuse 1.5 char d'assaut 3 1 destroyer 1.5 6/7 lance-flammes 1.5 .5 tourelle .75 tourelle-R .75 These should be read as a trike can move x squares or shoot y times in the time it takes a char leger to shoot once. For instance, the destroyer can shoot six times in the same span that a char leger shoots seven times, and a char d'assaut can move six squares closer to a lance-flammes in the time it takes the latter to fire -- note that this would be firing *again*, since there is no delay in an initial attack or initial move. Perhaps shootspeed should be recoiltime? ==== I've been reading the information on ^IfMain(2)'s message handlers that came along with Debug2. There were a few messages that weren't very well explained. I'm asking for some enlightment on these, if anyone here has more information on them. Message 12 - IfMsgGet3KeysMenu What does this do, what are its inputs? Message 13 and 15 - IfMsgCancel and IfMsgOK How does these work, and what need the new handlers do, at a mininum? Message 17, 18 and 19 - IfMsgChoose, IfMsgType and IfMsgCalc What are the parameters for these message? -- Eduardo M Kalinowski http://move.to/hpkb ==== Hey George Mmm... Nasty. This shouldn't happen. The message invalid... is a left-over from the 48. On the 48(s/sx/g/g+/gx) where you could install memory cards, if one of them had invalid data, this error occured. The 49 doesn't have ports like the 48, so this message is somewhat wrong. Anyway. No. PINIT does not destroy any data. More than that, its use on the 49 is a mystery to me, since there are no ports to initialize, as on the 48 :-| Go read the FAQ, where it mentions recovering from crashes. Do not worry... noone understands nothing anyway, life is a big mess. George Tsiros ==== oh, fine. 1) what kind of library is it? If nobody's going to muss with its rompointers (and even if people do), your ROMID need not be constant. That is, if someone complains you can change it; if someone has a problem they can change it. In other words, PICK ANYTHING, IT DOESN'T MATTER. 2) if there is a conflict, it will be from a library that uses specifc other libraries -- ie, libraries whose natures are documented, and the person probably has these auxilary libraries anyway. The most likely result is that the person trying to install your library would be told that it was already extant on the port. Fine, big deal. OT49 anyone? 3) hpcalc.org has a list of library IDs, but I don't know how recent it is. 4) if you post your intended ID here, or a list of IDs, people could tell you if they noticed a problem. ==== How do I decide which ID number I assgin to a library I make? I've done a search here in comp.sys.hp48 and I found a set of rules to The rules have not helped me. My program is a program which will be used by other people at my university, so I wouldn't like to get a an ID number which will make things crash. ==== them to run on my HP48 ?? Est il possible de coder en C sur mon pc linux puis de crosscompiler pour que ca tourne sur ma HP48 ?? Pied ==== Yes it is. http://www.hpcalc.org/hp48/pc/programming/ Ricardo -- http://193.145.89.239/~rblasco/ Un222os, hermanos linuxeros 640 Kb de memoria son m207s que suficientes (IBM AT Designers - 1982) Bueno, espero que esto no aparezca en la versi227n final (Bill Gates en la presentaci227n de Windows Crash Debug '98) ==== I'm interested in using this command, but can find no reference for it. Does anyone know how it works? Will ==== The online help of the HP49 says that this command creates local variables in algebraic mode. Example: LOCAL({'<-A=1','<-B=2'}) These local variables must then be purged later in the program, before the program exits with the command UNBIND. I tried the command also in RPN mode and it worked only for compiled local variables but not for normal local variables. In RPN mode the list {'<-A=1','<-B=2'}) must be on the stack before LOCAL is used. By the way, does anybody know why the command doesn't work with normal local variables in RPN mode? (Compiled) local greetings, ==== online manual, I've done so many searches, and not found anything that has the latest commands in it. Will ==== Well, the only manuals I know of, are: http://www.hpcalc.org/hp49/docs/misc/49g_aug.zip http://www.hpcalc.org/hp49/docs/misc/augerr.zip http://www.hpcalc.org/hp49/docs/misc/49g_ug.zip http://www.hpcalc.org/hp49/docs/misc/ugerrata.zip You can find more documentation at http://www.hpcalc.org/hp49/docs/misc/ And of course you can ask here, if you want to know about something. (Kind of documentation stored in the collective memory of the 49BORG ;-) ) Greetings, ==== thanks that helps. I looked earlier, but I must have missed it. thanks again Ken ____________________________________________________________________________ __ http://www.binaries.net ==== the implicit source (well, distribution point) for all code and documentation is hpcalc.org ==== If you mean milli and pico as prefixes for other units, then I think the HP48GX and the HP49G can do that. Example: Enter '1_mm' 1'_pm' and then use the command CONVERT to turn 1_mm to picometers. Other prefixes are also possible. (k, P etc) Greetings, ==== And yet another thing to add: 113 MENU will take you to a menu with four things: [EQLIB][COLIB][ MES ][UTILS] [EQLIB] doesn't do anything. It was the old EQ-Library of the 48 [COLIB] is the constants librars [ MES ] is the Multi Equation Solver and its commands [UTILS] some additional commands for specific purposes. Greetings, ==== Please belive to me, my calc is still UNUSABLE. I«ve tried every kind of things that everyone has told me. 1.The famous paper clip in the back hole 2.The same paper clip holding the backspace key 3.ON+C 4.ON+A+F 5.I took the batteries off for two days. 6.I put the batteries in inverse polarity for 15 sec. 7.I called up to HP Service in my town, but they weren«t able to give any answer. BELIVE: NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING NOTHING has worked. Please belive to me, NOTHING has worked. It still displays the same message: TRY TO RECOVERY MEMORY?, but when I press any key (A,F,ON+A+F) it doesn«t work. P.D: I don«t know what for some people told me about erase the port 0, if the calc would allows it, I think that this message would not exist. ==== That was a very very bad (and stupid) thing to do. Do not do it again. It isn't funny. Ask them where you can take your calc for a replacement, if the warranty still holds... ==== One of the engineers in the office has just got a new HP49. I have to say that the keyboard on it is very nice. Nice action and light pressure getting fairly close to the feel of my old HP41. He also has a HP41. Also came with ROM 1.18 installed. My HP49 is also a chinese model about 18 months old with a somewhat stiff keyboard. Don't know if HP has been doing some work improving the keyboards or they vary from calc to calc. I've had nasty thoughts wondering if I'd get away with swapping calcs during a lunch break. Other thatn the keyboard and some wear on the keys they're the same. ==== Is there any program that can perform Newton's method,Bisection method,Secant method or any thing in field of numerical analysis in English.I found only in spanish at hpcalc.org ==== The built-in numerical solver perhaps? Greetings, ==== You could pay an agent to buy one at clearance for you =) ==== Interesting... the 49G costs less than the 48GX? ==== I have a new hp49 from china and an old 48gx with the cable that connects to the PC. Can I use the old 48 cable to connect my 49 to the computer?thanks ==== to computer?thanks Only if you want to fry your 49G's LCD display. :) No, the cables are not compatible (at least mine's not, I tried it ;-). ==== Go to http://www.hpcalc.org/hp49/docs/faq/ and read item 5.9 (Link cables). GWB. ==== take a look at this page: http://pweb.jps.net/~joehorn/cable.htm and decide for yourself -- ==== Top marks to Joe for a very useful page. I've added a link to it on my page. Joe - any chance of adding to it to acknowledge the existence of the HP39G/HP40G? As I understand it this should be simply a matter of putting HP38/39/40G wherever HP38G appears. -- Colin Croft Old mathematicians never die; they just lose some of their functions. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== ==== It depends which old cable it is. If the old cable has a *non-removable* 4-pin HP48 plug, with a slip-over adapter which then fits the HP49G, then it's a compatible cable made for HP48 and HP38G, but if the non-removable plug has 10 pins, watch out for calc-frying potential! For more complete info, see Joe Horn's Cable Table: http://www.HolyJoe.Net/cable.htm Holy Joe! Did you move away from Abbey road? [r->] [OFF] ==== I think, the members said it all. I have a Jornada 680. On paper it is great. But it is too heavy and difficult to carry it around. For just an organizer, it is too big, for a true pocket PC it is not enough. If I could redo my purchase, I would definetely go with a palm size, that fits into the pocket of my jacket and does not pull it down with 3-5 inches. Kristof ==== I'll try to describe a strange behaviour of two (!) programs, looking for someone that give me an answer. I'm really confused. They are the well-known ALG48v4.2 and Erable3.2 running in my 48GX with MK. other answers too ;-) I was preparing some exercises for my pupils :-( and all seems work fine BEFORE and AFTER the poltergeist X This simple function was the subject: ----------- 4 X - 16 1. INTGR from ALG48 was unable to integrate it, giving me INT(The same funct.) 2. PF from ALG48, gave me the unmodified function At the same time, the Erable commands INTVX and PARTFRAC, gave me strange results: 1. INTVX, gave 0 (zero!) 2. PARTFRAC, gave me two fractions with 0 above (sorry my engish). I can't remember them well but they had something like 2*X+2*SQRT(2)+..., below. I hope you understand something. And as I said above, many others calculations were made after and before, getting rigth results. The final: I try doing the same in my PC backup (Emu48) and the four commands, worked fine!! 1. Flags problem? No: I transfered standard system and user flags giving bad answers for THIS function. 2. Corrupted Erable and Alg48 in my calc?: I don't think so, but I transfered them again. Bad answers... 3. I restored my Home backup from Port 9: Bad answers. And then... ON A-F and restore Home backup again, AND THE FOUR COMMANDS WORK NOW as well as in my PC backup... Worst of all: the little devil in my mind telling This is a nonsense with an obvious reason... but you won't see it till you press the Post message button... and everybody is going to think you are a poor beginer... ha, ha! Please... some ideas? (...and of course, happy Xmas to you all) ==== To plot x^2+y^2+z^2=R^2 you could use the program: << RCLF -103 SF SWAP Z SOLVE SWAP STOF 1 << EQ-> NIP >> DOSUBS STEQ You enter x^2+y^2+z^2=R^2 (with known radius R and let the program run. It finds the tho solutions Z=SQRT(R^2-X^2-Y^2) and Z=-SQRT(R^2-X^2-Y^2) and stores them as a list in EQ. You can then set plot type to wireframe, adjust plotting ranges and plot the two equations together. Of course you may add commands to the program, that automatically set plot type and make adjustments etc. Greetings, ==== Last week I bought a HP 49G and played with it for some days before upgrading the ROM. With ROM 1.18, which was the version when I bought it, I could do things like this: A=B*X^2; X; SOLVE; and get the answer { 'X=-(SQRT(B*A)/A)' 'X=SQRT(B*A)/A' }. (The ';' are ENTERs, SQRT(...) is the square root-symbol, I used Exact-mode.) This is about what I expected, and Maple 7 agrees that this is the answer. After upgrading the ROM to 1.19-6 I tested solving the same equation. And got an error, SOLVE Error: Not reducible to a rational expression. I changed mode to Approx and tried again. The mode change helped, I got the solution { 'X=-(2*SQRT(A*B))/(2*B)' 'X=2*SQRT(A*B)/(2*B)' }. Using EXPAND on this result gave me the same answer as ROM 1.18, no unnecessary '2's in the result. Being puzzled about why this had to be solved in Approx-mode, I changed back to Exact and tried solving for the mother of all irrational numbers, 2=X^2. This was apparently a rational expression, no error and the solution { 'X=SQRT(2)*-1' 'X=SQRT(2)' }. Again I had to EXPAND the solution to get rid of the '*-1' to get 'X=-SQRT(2)' as the first part of the answer, and had a solution Maple 7 agreed to. Why do I have to use Approx-mode to solve the symbolic equation? I guess SQRT(2) is just an irrational number, not an expression, but still, why can I solve this in Exact-mode and not the first expression I tried? And why the need to EXPAND the solutions, is it because HP 49G wants to give me all the factors in the solution, not simplifying the answer right away? Is the changed behavior after I changed the ROM because I haven't set some system flags? Another question: Is the SIMPLIFY function on some other menu/key than in the EQW menus? 2300+ functions and 128 system flags is enough to confuse me a bit, so I am sorry if I am asking silly questions, I am a newbie and don't know better :) Feels like when I was in high school and bought my first HP (a 12C), that was quite confusing in the beginning too. Good Karma to you all, Jon ==== Look at this: http://groups.google.com/groups?hl=en&selm=1900%40seq.uncwil.edu Wasn't that the first one? Bye. ==== for WR, JKH or JM JM? Jesus and Mary? (better ask JKH, then :) See 'Creating User-Defined Derivatives' (Chapter 20, HP48G Series Users Guide), e.g.: 'derSEC(x,dx)=SEC(x)*TAN(x)*dx' HOME DEFINE then -3 CF RAD 'SEC(x)' 'x' [derivative symbol] gives the answer: 'SEC(x)*TAN(x)' and 'SQ(SEC(x))' 'x' COLCT gives the answer: '2*SEC(x)^2*TAN(x)' etc. But the HP49 CAS works differently: 'd1XXXX(x)' takes the place of 'derXXXX(x,dx)' E.g (assuming RPN mode and Exact mode): 'd1SEC(x)=SEC(x)*TAN(x)' HOME DEFINE then -3 CF RAD 'SEC(x)' 'x' EVAL [derivative symbol] gives the answer: 'SEC(x)*TAN(x)' and 'SQ(SEC(x))' 'x' EVAL gives the answer: '2*TAN(x)*SEC(x)^2' --------------------------------------------------------- The lyf so short, the craft so long to lerne - Chaucer ==== I want to know your opinion : - what do you consider best, RPL or SYS-RPL ? - what is the best manual available ? (include Inet address) TIA aresWALL ============================================================ ^----^ aresWALL / |--------| www.upcnet.es/~pag X | _ | X |__/ __| d d ============================================================ ==== SysRPL is naturally better although I personally try to use UserRPL so that newbie would have one quick answer from me (or from Nick the GreeK) to a simple UserRPL question. There are so many SysRPL gurus with so much knowledge that I personally cannot see how I ever could be of any help to anyone if I'd try to cope with it. So, SysRPL is it. http://www.hpcalc.org/details.php?id=4443 ==== a> I want to know your opinion : a> - what do you consider best, RPL or SYS-RPL ? This question is impossible to answer without extra information. As a calculator user who just solves problems, I would always say the UserRPL is vastly superior. Much easier, it will never crash your calculator, but still very flexible, and fast enough for most things. Only if you want to write applications for the calculator yourself, it is worth going to SysRPL, and, for specific applications, to ML. The HP48 Advanced User Guide is the best manual for UserRPL. Eduardo Kalinowski's book is the best documentation for programming SystemRPL, but the addresses of entries and some other details apply only for HP48, not HP49. http://www.hpcalc.org/hp48/docs/programming/ For a list of entries on the HP49G, you can go to http://zon.astro.uva.nl/~dominik/hpcalc/entries - Carsten ==== It depends. If you want to extend the basic functionality of the calulator, there's no way around SysRPL. If you want to write a simple program for solving a mathematical problem, using the existing facilities, then UsrRPL is better because it's easier to come up with a working solution quickly. Kind of like the dichotomy between C and scripting languages. Check www.hpcalc.org and http://cgi-bin.spaceports.com/~hpkb/index.php . Everything you need is there (and more ;). ==== I'm going to pretend that you said UserRPL instead of RPL, because otherwise your question makes little sense... not that it makes much more sense when I do this =) Gosh, SysRPL or UserRPL? The latter is a subset of the former, using very safe commands -- so if you're not interested in being very careful with your programming, or don't care for SysRPL's conceptually heavier control flow, you could just use UserRPL. I don't consider this a really strong division, though: I use UserRPL commands in some of my SysRPL programs; I've written UserRPL programs and then optimized them slightly with SysRPL. hpcalc.org has good manuals for UserRPL, SysRPL, and ML. <76ea4fd3.0112190515.5902f589@posting.google.com> 20:05:48 -0600) ==== The supported objects made available through UserRPL are indeed a subset of all the supported objects you can name in SysRPL, but there are other distinctions, in part because when we take the terminology out of context, we neglect to distinguish between the objects themselves and the languages which use them: The *source*languages* for UserRPL and SysRPL are very different (and SysRPL has more than one variant, sort of, given the variety of sometimes incompatible compilers). Both the UserRPL compiler and the execution of UserRPL functions provide handy training wheels to make programming easier (by hiding many internal details) and also much safer (by automatically preventing crashes). The UserRPL compiler (built into the calculator) helps you by adding lots of necessary things to your program that you can't even see, and also disguises many things to look simpler than they really are; it also prevents combining the objects in invalid ways. The UserRPL functions (except VISIT, apparently :) always trap errors and recover gracefully (at a cost of some little overhead), while SysRPL/ML provides vastly more freedom to cause warmstarts and TTRMs :) For reasons of personal taste, I like to use the highest-level approach which can get a task done well; thus I try UserRPL first, but when its smaller toolkit can't hack the job, then I turn to some SysRPL; even then, if the required SysRPL is very minor, I may sprinkle an occasional SYSEVAL, LIBEVAL or FLASHEVAL into a UserRPL program, and only when that still isn't good enough, then ... probably I drop the project! (I specialize in programs of twenty-five words or less, which are about as long as I can get interested in these days :) [r->] [OFF] ==== I hope that this does not sound like an ignorant question (but truth be known, I am rather ignorant concerning RPL being new to the subject.) How do people develop their programmes? I remember in the old days with the x, y, z, t, and last x and keystroke programming we used to use coding sheets to see where we were going. Now a days in c or perl I simply code using a coding tool or a fancy text editor with copious comments. This RPL however is a whole new proposition. I seem to get completely lost with the multi level stack. How do experienced programmers code in RPL or do you all just visualise the stack as you go along? ==== The old RPN calculators such as the 41C series and the 15C, just to name two, had a stack of x, y, z and t registers and a last x register for the last argument. As such the stack manipulation commands were x<>y, rill down, roll up, and last x. On these calculators the program was simply a series of keystrokes that would be typed on the keyboard, but stored in memury to be reproduced later, thus keystroke programming. RPL seems to have an element of this but with much more, together with a topless stack. ==== What do these letters mean? What is keystroke programming? I don't understand your use of 'or' here; I'll assume that it's a mistake. Yes, you have to know what's on the stack as you program -- which should be perfectly easy, and if it isn't you know that you need to start working to simplify your usage of the stack. It's helpful to divide your programs into discrete units of code and give these stack comments -- then you manage the units, can more easily check that they take and leave the stack correctly, and can more easily optimize the 'real' parts of your program. Forth people call this 'factoring', this breaking up of programs -- and people using other languages have 'factoring' too, and they often pay lipservice to it, but they often don't do it anything like Forth people do it -- which is OK, because their languages don't demand it: you can write a pages-long C function and your only problem will be that you can't easily fit it all in your head at one time -- a pages-long UserRPL 'function' will *kill* you, because in order to grok the stack you *must* have it in your head, and since you can't do this you'll find little mistakes messing up the stack and snowballing horribly. Well, pages-long has different meanings to people on a computer and people on the 48/9, but you see what I mean. I don't know what else will help you right now; managing the stack really isn't hard. ==== You certainly have to know what you have in the stack. My comments (in the code) are of what each group of commands do. I find easy remembering what the stack holds. If not I just use paper and pencil. But, well, I'm not an experienced programmer at all. Mith. ==== the way it does. ==== I suspect this is a FAQ (and an old one, too), but I have no time to delve into manuals and groups searching for the light. Why is the (menu 30) solver so inaccurate and how do I get around this problem? i.e. solving this equation in=r+p with in=11 p=9. yields: r:1.999999999996 (!) Greetings, Marco ==== testing 1 2 2.5 3 3.14 ==== Symphonie 1.3 ist a very good program. Dimitri Missoh. ==== You guys might find this to be humorous. I was browsing the Texas Instruments discussion groups yesterday, and I came across this unique message: [12:26, 12/17/01] Why me??? by Morton G. I was upgrading my TI-83 plus with the new ROM upgrade (1.14). Anyway, I was really excitied, because this TI-Connect software is so good. Anyway, my hands were kind of shaking, and when the upgrade process was about 7 percent done, I dropped the calculator off my desk onto the hard tile floor. Now, several things happened. 1. The graph link cable snapped off, and the metal bit is still stuck in the port. 2. The picture cracked, and there's a really colorful, kind of violet looking splotch that covers about 80% of the picture. 3. The case cracked (There's a huge crack by the graph link port on the bottom of the calc, and a big piece actually cracked off and flew off somewhere. I can't find it, either). 4. The On button cracked and fell out of its socket. I can sort of turn the calc on if I stick my keys in the socket and press hard. 5. Almost all of the lower row of keys kind of got sunk in. By that, I mean that when I press them, they hardly move at all, and they make a cracking sound if I press hard. 6. The battery cover flew off and the latch part got snapped off, and now the battery cover doesn't sit in place anymore. I can still keep the batteries in, but it looks bad because I'm using some brown duct tape. 7. The two halves of the calculator don't fit together anymore. There's a big gap between the two halves, and when I tried to push them together, there was a loud cracking sound, and now the two halves are really, really loose. Good thing I kept that duct tape out! 8. The bit of plastic that is in the middle of the four arrow keys cracked and flew out somewhere. The four arrow keys stayed in, but they fall out really easy, if I move the calculator. 9. When I turn the calculator on, there some sort of message on the screen, but I can't read the screen because it's cracked really badly. It looks like the message starts with a Pr, but I could be mistaken, since I can barely make it out. Is there anything I can do to get the calculator working by tomorrow? I really need it because my school has only three 82s and it's first come, first serve. I would go out and buy another calculator, but the office depot where I live is out of TI calcs (Probably Christmas gifts.) They did have some funny-looking calculators made by the printer company (hewlett-packert) for $29, but I'm not sure if I could figure those out or not. Anyway, to make a long story short, I'd be really grateful if anyone could ==== aaaaaHAHAHAHAHA!!! excuse me. that was uncalled for. Anyway the other day I had my trusty 48g in bookbag. I was pulling the bookbag out of my car and heading for class. Suddenly the naked 48g (no protective case) slips out of the pocket in the bookbag and falls corner first onto solid pavement. It heartily tumbles a few times before coming to a stop face down on the hard parking lot surface. I pause for a moment to think about what just happened and immediately pick it up and toss it back into my car, knowing full well it's just fine. Upon later inspection one corner of the calculator had a small ding that could easily be buffed out with my girlfriend's nail file, that is if it was noticable enough for me to even care. <9vmnl5$okv$06$1@news.t-online.com>: -- ==== sorry I just noticed that last bit.... hewlett-packert HAHAHAHAHA!! ok. I'm done now. :-) ==== My question is: How high was the table? -- ==== I have dropped my HP 25A from about 7 ft (>2_m) to a concrete floor while I was serving in the Finnish army around 1980 AND it didn't work anymore... at the next vocation I took to the university where my friend was working as an assistant in the physics department. We took it a part and couldn't find no physical damage BUT the calculator did work when twisting the case a little Finally we noticed that the connector connecting the display from the upper case to the Mainboard in the lower case (IIRC) was somewhat fold-in, that is, the connector fork sliding into a tunnel or cylinder connector was - well- not bend anymore as it should be. The connector fork looked like if you were touching your pointing finger to the first joint with your middle finger. So we opened up the fingers of the connector a little (each one of them, perhaps a dozen or two) so that a proper connection to the cylinder walls could be achieved AND it worked again!! AND it has been working ever since (never mind the batteries...) Still trusting on HP quality!!! ==== ... My wife has a 15C she bought in high school (circa 1982). Last week she accidentally put it through the washing machine. We took it out, let it dry, and replaced the batteries. Still works fine. Amazing. Glen Henshaw Space Systems Laboratory University of Maryland ==== Don't you know that everything is bigger in the US ???!! I find it very suprising that a calculator would be so damage after a fall from a desk. There's no doubt that TI won't release a calculator if it didn't pass the 1.2m fall test. ==== . . . I am wondering why anyone would find this funny. BTW: I download three newsgroups, when I do download them, and this one and comp.sys.ti are two of them. I didn't see this post there or in four others that AOL gave me in a search for ti calculators. Maybe AOL just doesn't let me see the one you read it on. The only reason I ask is that I'd be interested in reading replies to the post. Please let us know where you read it. Robert Pecherek De Kalb, IL 20:41:34 -0600) ==== This is all part of the protective crash dynamics built into the unit, just like the crumple zones found at the front and rear ends of better-engineered vehicles, designed to absorb crash energy and protect the much more vital inner brain of the calc, much as you would want your own irreplaceable brain to be protected! Rest assured, despite all the superficial parts which snapped off, all your stored programs and variables (and the latest ROM version, in the case of models with flash memory) remain safe and snug inside -- even if you can't see them (or their results) any longer. Just bring the intact precious internal cargo back to TI, and so long as they can see that the calc was not seriously abused (which it wasn't, because all these parts were designed to fly off to prevent that from happening), TI will replace the cheap outer shell, and everything will be as good as new again. Don't buy cheap TI imitations -- HP doesn't know how to design proper break-away parts (in fact, it's almost impossible to even take them apart at all), and it's a proven fact that one-piece, solid construction can be fatal to internal cargo during crashes. That, most sadly, is why NASCAR driver Dale Earnhardt died, while his reinforced car, meanwhile, suffered very little damage: http://www.carclicka.com/carclick/stories/010511/5055618.html Sorry to end seriously, but stay healthy, bundle up this winter, and don't drink egg nog and drive at the same time (unless protected by the force field of 800_MHz - 2.5_GHz waves radiated by a cell phone held to your ear) [r->] [OFF] ==== I believe the fellow who posted about his Ti-83 was a hoax. I would have slammed the post hard except, people who own the Ti-83 are overprotective of this low end casterated Ti-86. I believe it was someone that was in an indirect way attacking the intellegience of Ti-83 users. I can't blame him and actually enjoyed the post. Next he mentioned that he noticed at Office depot a funny looking calculator from the people who make printers. Nice jab at Hp as well. This guy got in a couple of good shots all the way around. Yes, he appearred inept, but so do 80% of the posters on that board (and I feel I am being conservative). I don't think highly of the Ti-83, but for the majority of users who follow the keystrokes of a teacher who has one, it is their only choice. I can't believe how ignorant the majority of people are in the field of mathematics (and I mean basic algebra). ==== Good Lord, he must have thrown it out of an airplane window or something! TI calculators are very durable. ==== Yep. Modern calculators generally do. ==== Very. To think that, today, people *expect* such devices to be fragile! ==== A writer asked earlier why anyone would laugh at this kind of thing. I'd like to speak to that. I started my study of electronics and math just a year ago, in my mid forties. Being required to purchase a scientific calc for a course called Alg/Trig for technology I purchased the recommended Sharp EL-531R at the college bookstore. About four weeks into the course I realized that I wasn't satisfied with its capabilities. Looking ahead in the textbook I could see that matrices were coming up in use of superposition circuit analysis, as well as quadratic equations, complex numbers, etc. I wanted to be able to check my graphs as well. So I began looking into a graphing calc, the TI89 & the HP48 or 49 quickly became my finalists, and I encountered for the first time the RPN debate. I was warned by several co-students and one professor that it would ruin me! There was ONE professor in the department who used an HP48 and loved RPN, a guy I really liked & respected. I decided upon the HP, and upon mentioning this to a professor he told me he had an HP48 he never used, he preferred the algebraic entry method, and he would sell it me for $30, but warned me that I would be on the path less travelled. I bought it on the spot. Being nervous that I may have made a huge mistake and put my career in terrible jeopardy, I put aside everything and studied the owners manual for over a week. As the power and ease of use of this magnificent tool was unfolded to me I became filled almost with awe. Soon on tests I was able to check my answers to systems of equations, arrived at through the required tedious processes, by simply dividing the vector of constants by the matrix of coefficients, which we all know takes about two seconds in RPN. Then ac circuit analysys came along! I have no doubt that the TI apologists who haunt our pages will laugh this off as nonsense, but the fact is that EVERYONE in my class used a TI-type calc, including three or four 89's, and they ALL believed, including the teacher, that they had to convert to rectangular coordinates to add and subtract complex numbers! Whether or not it was true I do not know, all I know is that THEY believed this to be the case, and as a result they avoided use of the clearly superior equations for solving RCL series-parrallel combos such as Zt=1/[(1/Z1)+(1/Z2)+(1/Z3)+ ...(1/Zn)], which we all know takes about two seconds to do in RPN, all in polar form of course. In lecture the professor would put a circuit on the board to be solved by the class, I'd set to work with my 48 and a few minutes would have the complex impedence, voltage, current, and power for each individual component as well as the circuit as a whole. Then I would sit there for half an hour as the class stumbled and bumbled their way through the circuit with their seemingly archaic TI's, constantly converting back and forth from polar to rectangular, constantly making their quotient of ticky-tack calc entry mistakes which necessarily comes from such a tedious process. When no one else could I would give the answer so we could move along. One fellow saw how easy it was for me on my 48 and borrowed his sister-engineer's 48SX, and he went from pulling a C at midterm to pulling an A for the class. So forgive me when I laugh about an 89 exploding after falling three feet, it's just that I feel bad that TI has the educational process in a strangle hold and I feel sorry for people. dan ==== At least TI didn't fire all their calculator engineers, and tell the public everything is OK. Its like the skipper of the Titanic telling the passengers we are not in trouble, we are just stopping for ice. don ==== in a previous post, i had the impression that the processor for the HP49G was the same as for the 19B (the 19B being a lot faster). Is this true? i ask because where i work, we are going to get a hot air tool that would allow me to replace the processors. ==== Well, *superficially* it would appear that the new Agilent CPU wouldn't be compatible with the current Yorke CPU as a drop in replacement. I say this because the 19B(II) *used* to use a Lewis CPU which was different in many ways compared to the Yorke. Specifically, it had 64 KB of on-board ROM, on-board display RAM ( the exact amount is unknown to me ) and no UART among many other minor differences. Also, the Lewis was housed in a totally different package with a different pin out as compared to the Yorke. Besides these differences AFAIK the Saturn CPU in the Lewis and Yorke is pretty much identical and I would suppose the on-board peripherals that they share would have very similar software based interfaces due to the fact that the Yorke is almost identical to the Clarke ( on the software side ) and the Clarke is a highly leveraged redesign of the Lewis. But, don't take my word on this. :D ( I'm not in any way familiar with the Lewis hardware ) Anyway, if indeed the new Agilent CPU is identical to the Lewis on the software side, then they're not much farther from a full blown Yorke. It is mystery to me what would be stopping them in this case. ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. ==== Nope -- the 19B runs on a Lewis 1MHz CPU (same as in the 28S), the predessor of the Clarke 2MHz chip used in the 48SX.The reason the 19B seems faster is the smaller and a lot less powerfull OS (64k ROM, 8k RAM !). Lewis and Clarke or Yorke aren't pin or function compatible. Bye. -- `What a depressingly stupid machine' Detlef Mueller -- Marvin Detlef[DOT]M[AT]hamburg[DOT]de http://mein.hamburg.de/homepage/grendel ==== Hmmm...I never noticed this until now, but if you change the spelling of two of those names just slightly you have the names of three members of the Lewis and Clark expedition (Meriwether Lewis, William Clark, and York -- Clark's slave), who spent the winter of 1805-1806 about 100 miles from Corvallis. Maybe this means the next CPU should be called Sakakawea (or Sacagawea, depending on spelling preference). Eric Rechlin Bismarck, ND, USA gro.clacph@cire (reverse to respond) http://www.hpcalc.org ==== Theese days I've made some tests of v3.1 on Emu48 and I have got any TTRM with REF command, one of the most frecuently used by me, and other minor problems... but is difficult to me comparing the actual speed on Emu, without many of my own programs and key assigments that use 3.2 commands. years) ==== I remember her. What a beauty! Some years ago, when I had no web access at all (1995 or 96) I rang Daniel to try to get RPLMAN.DOC - **He gave me his printed copy**, folks. I knew him because I had studied some of his works (especially TTT4D, which I printed with my 82240B, as I did with Rapisarda's Smartkeys). A few months later, he invited me to his home and gave me copies of the Goodies Disks I missed, which were assimilated that very night! the HP39G/40G. Otherwise, I'll try it. But be patient, since it could be long. Sorry. HPCC member #1046 - ==== Continuing the marathon with the third part, I see that there is stuff left having to do with substitutions and programming, which only indirectly relates to trigonometry. I think that it would be a pity, not to mention this stuff, so let's start where we ended last time. At the possiblities for replacements and general manipulations of algebraics. It may seem that we lost the path of the trig marathon, but what we see here will be very valuable for later, when we make our way though the jungle of trigonemetry. We will return in part 4 to the maine route of the trigonometric marathon. some odd power with a -1 would be incorrect when s1 represents a 1. So we must find a way to distinguish the cases where s1=1 from the cases where s1=-1. One possible solution would be to take advantage of the fact that the command IFTE is actually a function which can be included in algebraics. So we could substitute s1 raised to some odd power with IFTE(s1==1,1,-1). The replacement list should also had a third item that checks if the power is odd, like for example FP(&n/2)!=0. , where != denotes the character not equals of the 49. The replacement program would look like: << { 's1^&n' 'IFTE(s1==1,1,-1) 'FP(&n/2)!=0.' } ^MATCH DROP If you store this in '^s1odd->' you can use it together with ^s1even->1 in another program that replaces s1 raised to even *and* to odd powers. This could be something like: << ^s1even->1 EXPAND ^s1odd-> EXPAND Now, for small expressions the programs are nice but for bigger expressions that also contain many odd powers of s1, you can quickly come to very ugly looking results with many IFTE, which don't contribut to the overall readability of the expression. Another problem is that the replacement programs will match anything of the form s1^n but *not* a single s1 that isn't raised to some power. (Remember? No algebraic replacement, just patterns.) We could of course check the whole algebraich for existence of s1 not raised to any power, but that would be cumbersome. In other words, I am too lazy to do that. ;-) advantage of MATCH is that is can find and replace patterns. Now, we made the convention that s1 is always the name of the arbitrary sign, so we don't need to look for patterns. The command SUBST seems to fit better here, as the code 's1=1' SUBST or 's1=-1' SUBST would substitute every occurence of s1 with 1 or -1 respectively, even when s1 is not raised to any power at all. But the problem is that we can't do both substitutions with only one SUBST. Remember, when the HP49G wants to tell us that there are more than one possible results, it packs them in a list. Imitating this behavior we can make a program like: << DUP 's1=-1' SUBST EXPAND SWAP 's1=1' SUBST EXPAND 2 ->LIST which makes both substitutions and returns the two results in a list. We see here, that the choise of the right tool can make life easier. Since the 49G has so many commands, it is sometimes not easy to decide which one should be used. But using some method for a long time often shows the disadvantages and suggests another method to be used. Continuing about replecements: As M207ximo Casta226eda Riloba has pointed out, the replacement of the pattern COS(&X) with s1*SQRT(1-SIN(&X)^2) will work for an expression that only contains one occurence of the pattern COS(&X). When we have an expression with multiple occurences of COS(&X), as in 'COS(X)+COS(2*X)' then each of the COS patterns should be replaced in a way that each sign is independent from the other, because one sign can be 1 or -1, no matter what the other sign is. Putting simply s1 for every replaced pattern makes them both the same. If we only somehow could use a numbering system that distinguishes the signs writing, say, s1 for the first and s2 for the second. MATCH can't do that, because one pattern is replaced by one pattern and there is no numbering or other distiguishing capabilities. Neither SUBST nor | (where) can be used for this purpose. But this doesn't mean that it isn't possible. ( A somehow well known phrase when using the HP49G ;-) ) What we need is to write a program that not only checks occurences of COS (or other trig functions) but also keeps track of the value of some iterator variable, which then can be used to construct the signs s1,s2,... and so on. A fantastic property of the HP49G is that functions (like COS) are also objects, which can be used not only to perform calculations but for also for other purposes. Type for example { COS }, enter this, and then use the command HEAD, which extracts the first element of a list. A naked COS function sits now on stack level 1. It can be used as argument for tests and other things. Another command made available to the user on the HP49G is the command ->LST (Menu 256, second page). This beauty takes an algebraic object, translates it to the RPN command sequence that corresponds to the algebraic, and returns this sequence as a list. Enter for example '(COS(X)-1)*SIN(2*X)' and press ->LST. The result is the list representation of the algebraic: { X COS 1 - 2 X * SIN * } The opposite command is ->ALG (Menu 256, second page) which would take this list and build up the original algebraic object. Having this two things in combination allows as to turn an algebraic to a list, check for occurencis of COS (or any other command) keeping track ot the number of the occurence and replace each occurence with something that contains an numbered arbitrary sign. For a replacement of COS with sn*SQRT(1-SIN(X)^2) where sn is the numbered arbitrary sign, enter the program: << ->LST 1 << IF DUP { COS } HEAD SAME THEN DROP { SIN SQ NEG 1 + SQRT } s NSUB R->I + S~N { * } + + OBJ-> DROP END >> DOSUBS ->ALG and STOre it in 'C->S' It simply checks if each object in the list that ->LST created is COS, and if it is, it constructs the list { SIN SQ NEG 1 + SQRT sn * } and then explodes it dropping the item count of the list. The program may look like one that uses a local variable procedure but it isn't. The inner program is placed on the stack at run time and is used by the command DOSUBS. DOSUBS needs a list at stack level 3, the number of arguments that the program on level 1 needs, and a program on level 1. It applies this program to each group of n arguments of the list in level 3, n being the number of arguments in stack level 2. The result of the program replaces each group of n arguments in the list. NSUB returns the number of the current group of items used as arguments for the program on level 1, as a real number. R->I transforms this real to an integer and adding this integer to the string s returns the numbered arbitrary sign as a string, which the command S~N then transforms to a name. To try it, enter: 'COS(X)*SIN(X)+COS(X)^2 and press C->S. You get: 'SQRT(-SQ(SIN(X))+1*s2*SIN(X)+(SQRT(-SQ(SIN(X))+1)*s7)^2 where s2 and s7 are two distinct arbitrary signs. You could now be impressed of the ease with which such things are achievable with the HP49G, but there are things even more impressing. Consider the algebraic object: /Pi | | COS(X)*SIN(X) dX | /0 Applying C->S to this returns the same object unchanged. Why? If you reenter the algebraic and apply the command ->LST to it, the result is: { 0 Pi 'COS(X)*SIN(X)' X S } (where S is the integrate command ) Checking if each item is the same as COS doesn't work here because the integrant is not transformed into an elementary command sequence that includes COS at some point. Some of the functions that can be in algebraics behave this way, when we apply ->LST. One of these functions is the integrate function. So we could of course think, that we muct first find all that functions and do something special if we encounter them. But what if such functions are nested? How many passes should we explilitely program to catch all of those functions? Fortunately we don't need to do that. Recall that C->S splits an algebraic to its elements, checks each element for a COS function and if it is, it replaces this element. Exactly the same procedure can be used for the case, when one of the elements of the list is itself an algebraic object. Since algebraic objects and built-in functions have different types (9 and 18 respectively), we can do: << ->LST 1 << IF DUP TYPE 9. == THEN C->S END IF DUP { COS } HEAD SAME THEN DROP { SIN SQ NEG 1 + SQRT } s NSUB R->I + S~N { * } + + OBJ-> DROP END >> DOSUBS ->ALG STOre this in 'C->S'. Note that the program uses itself recursively to split any number of nested algebraics and replace occurences of COS (!) To test it, enter for example: /Pi | | /Pi | | | | COS(X)*COS(Y) dX dY | | | /0 | /0 Then press C->S . Isn't that nice? (It is a recursion, so Nick *must* find it beautiful. ;-) ) Finishing for today, there is a problem left, that arises because we use numbered arbitrary signs. How could one write a program, that checks for any arbitrary sign, like s1,s2 and so on, and then builds up a list with algebraics in all possible combinations of all arbitrary signs? That's all for today. Take care and keep on HP49Ging. Till next time, recursive greetings, ==== Has anyone ported Tunnel to the 49? ==== Are you aware of a problem involving the speed of the emu in response to mouse events and speed emulation settings? In particular, mouse events will speed up the emu for their duration, causing the clock to speed up and fall out of synch with the computer's clock. Also, setting the emu to emulate real HP speed speeds things up even faster (to the point that auto-off will execute within seconds). I noticed this while programming on the emu. Keith Farmer ==== i am looking to buy an HP48SX. i got some equipment that likes this model and only this model. Not sure if this is the right place but this place has address below. thank you, ARQ. ARQDENNIS@EUDORAMAIL.COM ==== Visit http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/adforum.cgi Also you might try eBay. James ==== is there a way to run 39/40 aplets on the 49 with a minimum effort? Merry Xmas, Marco ==== If you don't have a 39G and just want to check it out then you can always use the emulator. A package set up to emulate the 38G, 39G and 40G is available from my Utilities page. -- Colin Croft Old mathematicians never die; they just lose some of their functions. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== ==== You can install 39/40's rom in your HP49. But you'll lose all the 49 facilities. You can download the rom from hpcalc. Ricardo -- ==== Have you seen DevNet from HP? It's a site for developers in hp machines, I've surfing throw it these days and I've found some interesnting points, two games for jornada linux pda (the famous jx25): https://devnet.hp.com/projects/xmame/ https://devnet.hp.com/projects/frodo-jx25/ It seems that there is too activity ChaiVM related.. we'll see.. Have fun, J.Manrique CdU de la ETSIG ==== These are clearly misplaced, as they belong under the JX25 DevNet. I'm glad I never uploaded anything on that account. It may be the right time to delete these Java source files now? I guess so... No new Java plotter at the release, although the big display was perfect for it :-/ ==== Obviously Dell and Compaq did not insist, that they want to destroy Microsoft :-) Did IBM skipped the deal and than later insisted on the same price or was rejected at the time of deal ? Deal is a deal. What you expect, special rain check for IBM because you don't like Microsoft ? If I remember correctly OS/2 used new, advanced disk format different from Windows old DOS based FAT and simply they could not exist together because of the technical reasons. That is why you had to choose OS you want to live with. Don't consider technical limitations created by the advanced OS/2 disk format as sinister action of Microsoft. Again, Microsoft was not obligated to create special IBM version of Windows to operate on the OS/2 formatted partition. Quite frankly I don't understand even why Microsoft at all suppose to be obligated to provide dual boot ? You want Windows use Windows, you want OS/2, use OS/2. What is wrong with it ? OS/2 at the time was more advanced than Windows. All you had to do is to choose OS/2 and don't even bother with Windows. What monopolistic ? Clearly OS/2 was a direct competitor to Windows at the time and IBM had exactly the same chance to extort Microsoft by simply not installing Windows on it's machines at all. Additionally IBM also owned manufacturing that Microsoft lacked. What I don't understand is, why IBM never stroke a deal with Dell and Compaq to sell OS/2 at the same price or cheaper to them that Microsoft did with Windows. IBM can only congratulate themselves to be that stupid. What a joke. David back stabbed Goliath. IBM was a hundred times bigger company at the time with much more powerful financial and programming resources not even mentioning giant PC manufacturing base. Microsoft was at the time little company with very limited chance to stand against IBM. Microsoft realized that IBM want to get rid of them by killing Windows with it's commonly owned OS/2 and then getting rid of Microsoft by extorting Microsoft from the ownership of the evolving OS/2 they would control. Microsoft did the best it could to survive. Allowed IBM THE FULL OWNERSHIP of OS/2 and getting involved into the direct competition with it's Windows. It was purely IBM stupidity and complete lack of competitive skills, that Microsoft won despite being so late with it's competitive version of Windows, and survived IBM attempt to destroy them. Level field with who ? Even today Microsoft is a smaller company than IBM. Jack ==== You can't apply ln(x+y)=ln(x)+ln(y). The correct formula is ln(x*y) = ln(x) + ln(y). K.K. ==== It works on the '49 v1.18, too. Must be a recently introduced bug! Greetings, Marco ==== This is easily explained. The two writings are actually equivalent, but it looks like that the '49 and the '48 have different tastes when it comes to choosing how to display it. the '49 will display: 5^x+7^x^2=15 the '48 will display: 5^x+7^(x^2)=15 (tried on an emulated 48GX with ROM rev R) Setting Flag -53 (keep extra parenthesis) has no effect on this one, they are always discarded when assembling it on the stack; they are retained if the expression is built in the EQW. However it does not matter whether the parenthesis are in place or not, 1.19-6 will TTRM on that expression anyway. My 1.19-6 (real) 49 takes about 3.5 secs to find one of the solutions. The NUM.SLV won't find both solutions but only one), thus the 39 solver must be completely different: I must take a look at it :) An emulated 1.18 49 takes about 6.5 secs to find both solutions symbolically. This one was calculated using Authentic machine speed. Merry Xmas, Marco ==== Sorry, you misunderstood. I meant why is there a 5^x in the original but a 5^A in all the discussion since? I wondered if this was a typo or significant. I'm not sure how the 48/49 works but on the 39/40 when you use the Solve aplet you have to supply a starting value and it moves to the zero closest to that. Usually this is so fast that the 'distance' from the start to the zero is irrelevant. In this case the time from a starting value (as I recall x=6) to the zero at about 1.07 was quite a long time. It also seems to have been increased by the fact that I had chosen to display the iteration values as the process occurred. When I solved just now without displaying the time was only 10 seconds. Still, this is not fast by normal standards. -- Colin Croft Old mathematicians never die; they just lose some of their functions. ====================================== Applications in Mathematics ccroft@iinet.net.au http://members.iinet.net.au/~ccroft/ ====================================== ==== a 5^A Whoops, sorry! I must be still under the effect of this xmas wine :) if 5^x is entered instead of 5^a, the 49 doesn't ttrm. He probably made a typo because the solution is correct (the same returned by a 1.18 49) aplet that. The same holds true for the 48/49, if you want to find both roots you have to repeat the process with two different initial guesses (ie: one positive and one negative) 10 seconds for one solution or for both? in the latter case it's close to the 49 Merry Xmas, Marco <2ZSzlbEgeJZQ@opalo.etsiig.uniovi.es> ==== Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== Perhaps because there is no linux for Jornada.. and I prefer Giac over MuPAD for some reasons: -It's free, open source, GPL.. -It comes from Bernard Parisse, and I trust in his CAS knowledge.. -Giac seems to be faster than MuPAD The only CAS that could run in PocketPC are: -Maxima (slooooooow): http://www.rainer-keuchel.de/wince/maxima-ce.html -JACAL (uses scheme as language and not integration features): http://www-swiss.ai.mit.edu/~jaffer/JACAL.html For JACAL, PocketScheme is needed: http://www.angrygraycat.com/scheme/pscheme.htm It would be great a real C++ CAS using PocketPC eVC++ features.. Have fun these Christmas.. J.Manrique CdU de la ETSIG ==== sign. The equal sign is displayed when you first enter the lim function in the eqw; if you put the limit on the stack and then EDITB it again, the pretty print will be displayed instead and the equal sign is replaced with the little arrow. This happens even if you didn't put the identifier specification in the first session and add it later. For the UNDO/inversion bug, there's a problem with the pretty print limit: if you press COPY on the limit the display will be garbled (but the data is another bug :) Merry Xmas Marco <4HEV7.3304$aS.418385@news010.worldonline.dk> <76ea4fd3.0112241711.5176604d@posting.google.com> ==== The data is corrupted, too. Try this: 1) Enter lim(INV(X),X=oo) in EQW. 2) Press UpArrow until the whole expression is selected. 3) COPY 4) ENTER Now it's become lim(X=oo,INV(X))! You can turn it around again with DownArrow COPY ENTER. Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== It has been stated that the official name of the new company will stay Hewlett-Packard. ==== -- Thierry Morissette FXXV7.14922$HW3.12145@newsfeeds.bigpond.com... ====

You may excuse the command XQ after you solve the equation.  This gives you a rational answer.
K.K.

have you tried without the dot appended and in exact mode?
...
>
> I suspect this is a FAQ (and an old one, too), but I have no time to delve
> into manuals and groups searching for the light.
>
> Why is the (menu 30) solver so inaccurate and how do I get around this
> problem?
>
> i.e. solving this equation
>
> in=r+p
>
> with in=11 p=9.
>
> yields: r:1.999999999996 (!)
>
> Greetings,
> Marco
>
>
>
==== You may create an m by (m+1) matrix and use RREF command. Remember to negate the constants. For example [ [1 2 0] 2 1 -3] RREF gives [[1 0 -2] 0 1 1] which means x=-2, y=1. K.K. henar2 274g¤J¡G ==== I thought that I posted this a few hours ago, but it's not showing up so I'm posting it again. My apologies to all if this is a duplicate. First, I should mention that those numbers are my mistake, they'd be for @AB instead of ABC. To change ABC to <65><66><67> on a 28 or 48, try this program: << OVER SIZE 1 SWAP FOR s < + OVER s DUP SUB NUM ->STR + > + NEXT SWAP DROP >> For a 49G, try this program instead: << OVER SIZE 1 SWAP FOR s < + OVER s DUP SUB NUM R->I ->STR + > + NEXT NIP >> Note that << and >> represent the left and right program delimiters, and -> represents the right arrow character. Again, these programs aren't optimised; I don't doubt that they could be improved. James ==== www.hpcalc.org ... Eu schrieb im Newsbeitrag ==== Bokoo1 escreveu na mensagem <2rc22uov5r0tiudubbk0hue6a4inckm9ae@4ax.com>... No. Get an 128 K ram card and use Alg48 in port 0 and Erable in 1. ==== Is there an equivalent of the HP48 EQlib (for the hp49)? I got Eqlib v0.5 for the hp49 of hpcalc.org but it doesn't have all the stuff that lived in the hp48. taa -- -------- It is the least I can do, which, as a matter of policy, is the most I can do. ==== Andre escreveu na mensagem ... exAL. ==== Carlos Marangon escreveu na mensagem <2cfc228.0108261700.7f7ab26c@posting.google.com>... Tacio reconstruced it @ http://www.eportateis.com.br/hpclub. Look @ it! ==== an lame people said that they wanted it for Handhelds & Portateis Club (www.hpclub.com.br) = for the lamers , hp should be Handhelds and Portables. now @ www.eportateis.com.br/hpclub. ==== It's revolutionary under the respect it has no wheel group because it's fascist. Microsoft puts its focus on usability, and under that respect it beats the blood out any X-based user interface. Some Open Source OSes puts their focus on speed and security (namely *BSD) because there are enterprises which are funding their development and rely on those products to run their web businesses. A MS CEO told the world that 640k ought to be enough for anyone, but we don't believe this just because he said, true? A consultant is someone who's selling himself to you, and considering most security consultants: a) don't want to sound offbeat (so they won't tell you that unices are unsecure) b) have a background in unix c) like to sell unix based services because: 1. there are far more MS based service providers/consultants you can switch to if you think your consultant isn't good enough/is too expensive 2. unix based consultants request higher fees 3. a MS based system -once installed- is likely to cease being a cash cow for the consultant because it's so easy to just take your employees and tell em to service the system, and only call the consultant when the system is so messed up that employees can't figure it out in reasonable time Anyway, may I suggest to read the BSD security advisories yourself? You're likely to be surprised. PDAs are not mission critical systems, so stability isn't really an issue... it's much more important to have a highly usable system which doesn't waste your valuable time just to enter a phone number into your personal directory. Well, an OS which is claimed to be better than MS's stuff surely captures people imagination... at least until they figure out that X is far slower than Windows' GUI and also technically inferior (unless you want to easily access your desktop from remote) and that they ultimately can't surf the Net with a decent degree of comfort. I'll get roasted, too, but who cares? I only hope I'm not going to heard that linux common belief bullshit again (a 486 running linux is faster than your brand new Athlon 1.4 running w2k) Greetings, Marco ==== I'm surprised... I have sometimes seen that special MK for Emu, but I have never need it. When I wanted install MK in Emu, I simply followed the same steps that I followed when I installed it in my calc. All was OK... and still working, of course. ==== maybe you could use CONT to return to the MK shell... Then, you could replace HALT by 0 WAIT or KILL. The action of HALT from inside the MK should be documented somewhere;-) HALT is mainly used for debugging purposes, so in a normal program, there is no need for using HALT. Raymond Eu schrieb im Newsbeitrag ==== Your program seems to find out primes from 7 to 6001. There is a bug in your program -- missing next. Try this: << 1 1000 FOR X '6*X+1' EVAL DUP IF PRIM? 0 = THEN HALT ELSE DROP END NEXT >> or this: << {} 1 1000 FOR X '6*X+1' EVAL DUP IF PRIM? 0 = THEN + ELSE DROP END NEXT >> Hope this helps. K.K. ==== so there is no symbolic solution :-( Anyway, many thanks for your help. I wish you good luck in the new year 2002 :-) BR Johny ==== 2002 :-) Best for you this year too. ==== Nope, it works fine. It's hard to find a built-in operation that will cause the TI-89/92+ to reboot or memory-clear (C/assembly programs, if not written correctly, can do these things). If built-in operations were unstable, students (and I ;-) would not be very happy. ==== Just a quick question. As I understand it, the HP48's RPL interpreter uses mark and sweep garbage collection, as opposed to reference counting. Is there any RPL (SysRPL words are fine) use case that would create a circular chain of pointers? It seems like reference counting could be faster and I was wondering why it wasn't picked. Mike -- http://www.mschaef.com ==== Nice try Mike! AND Who will do the job? You? ==== I didn't want to say anything more specific, since I'm probably not going to take my project to completion. But since you ask, here's why I'm asking my question. At this point, it appears safe to say that the 48/49 series is dead to new development from HP. Also, it seems unlikely that HP will see fit to GPL or otherwise release the source to the firmware. Perhaps I'm wrong, hopefully I'm wrong, but I think hedging my (our?) bets is appropriate. What this means from my point of view is a nice C (maybe C++, but I don't trust the compilers enough) implementation of an RPL-like interpreter. I'd like to be able to use HP48 like command sequences from my PC keyboard. Conceivably, if done right, it could be portable to WinCE and perhaps Palm devices. So... what do I have now? I spent a few hours hacking up a little command line RPN calculator emulator. A command c 4 2 / dup 2 * 4 pi * * will result in the correct output and an option to enter an interactive mode. The interactive mode preserves the contents of the stack and enables the user to issue more commands. Usually, when I run a calculation, I want to do more than I initially thought, so this is a nice way to allow that. Currently, I use IEEE 64-bit floats (double in C) to do my math (I'd prefer arbitrary precision BCD), and have functions equivalant roughly to a non-programmable HP10C. My next step is to add support for different data types. My initial priority is an integer type for binary math, but that'll soon extend to programs, lists, complex numbers, etc. Of course, like all free-time projects, this is contingent on the marginal cost of developing the next feature being both lower than the marginal benefit and able to fit into my other commitments. In other words, don't hold your breath. :-) I've thought about some other options I have, and wasn't really happy with any of them. * calc mode for emacs - Nice idea, but I don't think I want to have to run emacs to do math. Also, not very portable to smaller platforms, or alternative interfaces. * OKit/OKEval (http://sourceforge.net/projects/libokit) - The looks like a nice piece of work. I'm not entirely sure I want to use their 'virtual machine', and I'm more interested in source compatibility with UserRPL HP48 software. * HP48/49 ROM emulator (X48, whatever) - This doesn't solve the HP48 source code availability problem. It also is somewhat difficult to use the interface, and even worse to improve the interface. * Lesser HP emulator - Does it really make sense to worry about keystroke programming and unnamed variables on a PC? In a portable device, I think the tradeoffs make a lot of sense. On a PC, more flexibility would be nice. * Some fancy lisp-based calculator environment. - Nice, powerful, but there are already plenty of high-powered environments for doing that kind of work. A simple little threaded interpreter will probably suffice for the level of functionality I feel is appropriate. Anyway, that's my thought process. I don't know where it will go, but wherever it stops, I hope I'll end up with a useful tool. Closing the circle, I asked my initial question when trying to decide on a GC mechanism for my little program. If I pick reference counting, but am unable to support some key RPL feature, I'd rather not have to go back and rewrite a more complex GC algorithm. -- http://www.mschaef.com ==== Well, I guess that mark and sweep would be better suited for a system with a small memory arena. Also, using reference counting doesn't solve everything, you still have to compact the heap from time to time to prevent memory fragmentation. This is pure speculation, of course, I haven't seen any discussion of GC design issues for the HP48 in the publicly available docs. Concerning the circular pointer chains, I don't think that you can do this in sysrpl, but it should be easy in ML, so this would also complicate matters for a ref counting gc. ==== Try to press firmly the area just below the screen and above the first row of button. That usually fixes the problem. If it doesn't and your calculator still under warranty then you should call HP. HP has ceased the new calculators development, it doesn't mean that they won't support the machine they've previously sold... So don't worry, warranty will work HP ==== I need to calculate Complex values and I haven't an Idea how to. I want to calculate in degree mode with the exp expresion. Somtimes I need to convert this values in the two parts, angle and the value. If I try it in the complex mode I only get sensless answers. ==== It is easy. Turn on Degrees mode by entering DEG and then enter some complex number as (x y) where x and y are the real and the imaginery parts. The function ABS ( blue-shift / ) will give you the magnitude and the function ARG will give you the angle (red-shift / ) If you enter (x y) and turn on the polar mode by entering CYLIN or SPHERE then all complex numbers will be automatically displayed as (r < theta) where < is the symbol for angle on the HP49G. Conversely you can also enter a complex number as (r > I need to calculate Complex values and I haven't an Idea how to. may I add: try the command V-> ==== Recently, you posted that the HP-17BII calculator had undergone a partial re-design. Does this mean that the keyboard assembly and liquid crystal display can no longer be used to repair a broken HP-42S Larry ==== The redesign was just a change in the keyboard layout. That's all ==== I`m trying to find out a shop in Europe which can deliver the rechargeable ==== Years ago I made one by opening up the existing one examining it and replacing the NiCd batteries. Just check that proper contacts are made everywhere. ==== On Thu, 3 Jan 2002 09:19:41 +0200, Nousiainen There is also a place in Florida USA that will re- build them. $30.00 US with old battery pack and $35.00 US without.if you do not feel like doing it yourself. 1-800-535-5692: (Toll Free USA and Canada, I think) Internationaal Calculator& Computer 2916 Corrine Dr. Orlando, FL 32803 (407)898-0081: Voice 1-800-535-5692: Toll Free This Site will not accept credit card orders. To place an order please They also repair HP 41 Calculators Harold A. Climer Dept. of Physics,Geology and Astronomy U. Tennessee at Chattanooga ==== How DID you open it up ?? My dead one seems to have been bonded with super glue or something. I tried prying but stopped when I could hear the plastic begin to crack. Anyway, if anyone knows were I could buy either a rechargeable pack or just the battery holder for the regular batteries I would gladly pay much gold or silver for these triffles!!! Santos Lucero ÀProblemas con las news? Webnews, el acceso m207s r207pido y fiable http://webnews.aforo.com User IP: 207.212.230.68 ==== Try if you have not already. Should just be the four screws under the rubber feet. Yani ==== First the easy question: Does anyone have or know where I can get some printable templates for HP48G's? You know how some of the ROM cards come with their own keyboard plates? Well, I would like to make one of my own. Now the less easy question: How suitable is the HP48 as an embedded computer? I mean it has everything I need specwise (keypad, LCD, serial, easy programming language, etc.) but is the hardware capable of running 24/7? (assuming a DC power supply) Vimal ==== HP48 (embeded computer):i guess it would work fine. no overheating :) at less than 20mA. i just pick up a laptop for $20 (toshiba T3100SX). if you need something like that go to your state surplus. they sell old laptop for a dime a dozen. and also tons of calculators and printers and old junk. ==== week my HP48G died so its time for the big decision - to hp49 or not. I am an engineering student who mainly uses the hp for structural analysis. Does the hp 49g offer anything above its magical, and widly stated CAS? I am passed my first year calculus courses and I wouldnt even bother using it to solve integrals/difs. Ryan ==== Personally, if I am not impressed at all with the 49G. I have used the 48G for seven years and just graduated from Engineering and am now working in an Engineering firm. Mostly everyone there used the 48G, a few are TI weirdoes. A few of the problems are speed of simple things (Adding, subtracting, multiplying), moving of buttons (ENTER, undo, cst), and no infrared port. If I could sell mine back to HP, I would in a second. My recommendation is the HP48G+. Hope this was helpful. Last analysis. it ==== I have both and I'm partial to the 49G. Since you had a 48G, it shouldn't be a very hard transition. The keyboard is not as good on the 49G though. :-) don Last analysis. it ==== The 1MB Flash for storing your programs and data Last analysis. it ==== A FAQ - switch the 49 to approx mode. Isn't this just a matter of taste and being used to it? I don't have any problems with the 49's keyboard layout, maybe because I haven't used a HP before. For some optimization you can reassign keys (Keyman is a must). Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== On deciding between 48 or 49. 48. I'm a happy owner of a HP48GX and I have recently been using HP49G for about two months. Trust me if you use HP49G you will work SLOWER. The first thing you think when you start using 49 is this keyboard sucks wildly. When you have been using the calc for two weeks you begin to curse whoever decided to put that soft and disapointing keys. Just that, makes you work considerably slower. You sometimes even waste seconds staring at the calc waiting for it to do something until you realize that you had not pressed the key hard enough. Another thing against the keyboard, is it's layout. Ok, it has two good points about it, it's more friendly, less overwhelming. When my young brother saw the calc he said it has less keys. Well, in fact, it has more (one or two). And also you can use the arrows when in alpha mode. But, and there is an important BUT, the layout is thought to work in algebraic mode. If you are gonna work in RPN mode (and let's all say a prayer to thank HP for having made RPN calculators) you're gonna end up loosing time with right shift and left shift keys and going into menus for commands that are accesible in the 48 with one keystroke. About speed. Ok, the 49 is faster for many things, like plotting or loading scroll menus, but i have found that 48 is MUCH, MUCH faster for some programs which were initially written in UserRPL for the 48 and should work fine in 49. Then there is the IR issue. It really is very handy to have an IR port if you're going to exchange things with other 48s. The answer is 48. Unless you want to avoid RPN (which I do not reccomend). At least that how I see it. On deciding between GX, G+ or G. This is a more difficult decission. For the use that I give to the calculator, G is not an option. So far I have not needed to buy a memory card, but well, I thought that it's better to be able to expand the memory in the future. This is something that you have to decide. What programs are you going to use? 50k ones or 3k ones? Do you think this could change in the future? Are you going to use the calc for SysRPL programming? Then you'll need GX if you want to install Jazz, UFL and other programs. If the programs you're gonna use are not big ones and you don't plan to program SysRPL in your calc (I use a computer emulator for that), then I would buy G+. If not GX. But well, it's just my opinion, hope it's useful. Mithrandir. -------------------------- La honestidad no es una virtud, es una obligaci227n A.C. mizrandir@terra.com ==== I agree with Mithrandir... and if you buy a GX and two RAMcards (Klotz or you will get a 48GX Super, as fast as 49 in editing, CHOOSE2ing, with a CAS near 49, etc, etc but with better keyboard... Mithrandir escribi227 en el mensaje ==== And by the way: one of the best things of the 49G is its flash ROM, but now, who is going to upgrade it (officialy)? (I hpe you understand my bad english) ==== Maybe it did not die. Take out the batteries for a day. Reinstall batteries and it might be resurrected! I had to do this recently; for what reason ... who knows? ==== Marco Polo ha scritto nel messaggio The HP48GX seems to be really expensive here in Italy. You can buy the HP49G for less you'll pay the HP48GX. Or at least this refers to Rome. Here in Rome you can find all the models of the HP company (included the discontinued HP40) at Ingegneria2000, a technical bookshop near engineering of the La Spaienza university (Colle Oppio; chi conosce Roma sa di cosa detailed info about reaching this bookshop Bye, Kickaha ==== Here's me buying bits to make a transfer cable for my (then) impending HP49. It came with a 49-49 cable, an adaptor to D-9 for one end and a hp49-hp48 adaptor too. sweet. -- -------- It is the least I can do, which, as a matter of policy, is the most I can do. ==== kennt irgendjemand ein gutes Statik-Prog (sowas wie Stab2D) f237r den HP 49G? Thanxx ==== Was wollen Sie? http://www.hpcalc.org/hp49/math/statistic/ 49G? ==== programs. Better suited links are: http://www.hpcalc.org/hp49/science/civil/ and http://www.hpcalc.org/hp49/science/physics/ P.S. Advanced knowledge of Spanish is highly recommended ;-) ------------------------------------------------------------------------- Ralf Fritzsch Bundesanstalt fuer Wasserbau Federal Waterways Engineering and Research Dienststelle Kueste Institute - Department Hamburg ------------------------------------------------------------------------- Unix _IS_ user friendly - it's just selective about who its friends are. ------------------------------------------------------------------------- ==== I recently bought a 49 with this Id# 94705417 Is this a recent or old model? larry ==== I take it that you mean ID94705417. The ID means that it was manufactured in Indonesia, the 9 is for 1999, the 47 is for the 47th week of that year, and the 05417 means the 5,417th unit of that week. It's new enough that it doesn't have the serial port hardware bug, but the CN (for China) models have been reported to have various improvements. Note that whatever the serial is, you can update the ROM. Visit http://www.epita.fr/~avenar_j/hp/49.html for the latest ROM. James ==== PS: After I bought mine (in August of 2000), I called the phone number on the inside back cover of the user guide, and they sent me a CD with the latest ROM and communications software, a PC to 49G cable, an Advanced User's Guide, and an updated Pocket Guide. I don't know whether they're still being that generous, but it's probably worth a try. Please let us know how it works out. James ==== James , I made a call and they will send a cable. I'll wait and see what comes... thanks larry inside and an generous, ==== I'm glad to hear that; buying a cable would be a pain in the wallet. The CD isn't very important; everything on it is downloadable for free. I've read that the Advanced User's Guide has been improved - the one I have doesn't include the other commands. My old Pocket Guide has a sort of grey rectangular with rounded corners calculator on the cover, but the updated one has a blue 49G shaped calculator pictured. Note that all of the guides are available in PDF format at hpcalc.org, but it's nice to have them printed and bound. James ==== James I downloaded the hpcomm connectivity program but it will not install in win nt4.0 but it does install in win 95. I get a Dr. Watson message that the uninstaller failed to initialize. Has anyone experienced this problem before. larry ==== X X The User interface has still some major flaws I will name one: Pop-up windows with choose buttons Why? While writing a program or a document a pop-up window pops up AND too many times I hit the key corresponding to either a shortcut on a button or simply hit the space key and thus select the default answer. Sometimes this happens so fast that I can't even flash-red the message box title, so I don't no what it was all about. I guess that occasionally the pop up passes by totally unnoticed. Do there is some user-friendliness in modern GUIs. To correct this one could offer something new to the poor user: Error pop-ups that stop your application and need a selection before continuing. Warning pop-ups that your current application can ignore and so you can look at them a little later... The best place for these pop-ups could be a separate page or screen that you must select in case of error pop-up in order to continue. A docking bar on the right side of the screen might be another place to show these pop-ups (unless your application is in a full-screen mode) Opinions? AHNY2002 ==== What is important is the date of purchase. Regarding the keyboard, I don't see why HP would replace a working calculator for a new one. They have no obligation to do so. Sure, it could improve their image, but is it really what matters now ? ==== Just one small question... Does HP have an image that could get any worse ;-) Andre Claassens ==== Yes...they are even not good at making their image worse... They had call COMPAQ to help them... ==== Huh, why ??? Bought new but manufactured over a year ago, let's say bought in Nov 2001, manufactured in Dec 1999. What is the magic formula that determines whether you can get a replacement or not? Buying something new which has been sitting on the shelf for 2 years without any one warning you that this is an old model, you know, the one without all the improvements, but we are going to charge you the same price as one with all the improvements sounds shady at best. As a matter of fact it cost you more than the latest model because you have to spend YOUR time downloading the latest ROM, YOUR $$ to buy the serial cable, YOUR time to order the serial cable, installing, etc., etc., etc. Yes, I am one of the naive ones who paid the same price for an almost two years-old HP49 that I could have paid for one manufactured in China in late 2001. Santos Lucero ÀProblemas con las news? Webnews, el acceso m207s r207pido y fiable http://webnews.aforo.com User IP: 207.212.230.68 ==== Hum, maybe simply the purchase's receipt ?? Again, it's the date of purchase that matters, it's actually the same for every product you can buy on the market. Check with your dealer for more information. For example, I just bought a new motorbike (Honda VFR800), it's been manufactured in October 2001. Does that mean my warranty (that is 2 years) When you send your calculator for repair, you have to include the copy of the receipt. Well, at least that's the normal procedure, it's not always the case, in my experience it seems that HP often replace the calc without checking. ==== X formula X X Just to add something: Yes, but the calculator better have a real manufacture fault... old model with a flash is the best one could get! I have an older model of 48SX with no changes ever to trade in with one with a better ROM e.g. less bugs. I'm still using it occasionally. The colors are business like unlike the GAY Models recently introduced.. (except the 39/40G series, which are strait business!) ==== Buying Sorry for the misunderstanding I have created! I am not questioning the regular warranty - It brok, You fik (once I got this message with an instrument that came in for repair). What I meant to ask is if HP replaces calculators which are bought brand new but have been sitting on the shelf for umtpeen number of months (> 12). I would mind if the calculator was clearly marked as one not up-to-date, sold as is for let's say half price. What bugs me is that the calculator I bought for full price thinking it was the latest and greastes was 23 months old and I have had to spend time and $$ to bring it up-to-date and I am still saddled with a lousy keyboard, a lousy display cover, don't know if I have the serial problem (haven't used the to someone in Victoria, Australia (as suggested in the card that came with the anyone. Anyway, I am mad as hell and I will not take it anymore!! (famous last words). Santos Lucero ÀProblemas con las news? Webnews, el acceso m207s r207pido y fiable http://webnews.aforo.com User IP: 207.212.230.68 ==== 1.19-6 One of the ideas behind the supported entries expression is that addresses will never change, and that as few new entry points as possible should be added. In this way, most ROM updates shouldn't require any change at all of the suprom.a file. However, some new RPL words were introduced (I'm thinking about new CAS functions, which handle locals and assumptions about variables). They might be XLIBs, though. David Haguenauer ==== Happy New Year for all, I just bought new HP48G+, very nice for what I need. I've seen that I could add more memory to 48G+. Where can I find all the details how to do it and what memory cheeps to use? Could somebody point me to the resources on the www. Janek ==== Try here: http://ca-on.hpcalc.org/hp48/docs/opening/ Jan Piaszczynski escribi227 en el mensaje could ==== Janek and ==== I do not as yet own an hp 49g, but am considering the purchase of one. The two following capabilities are important to me, and I'd like to confirm that the 49g can do these things: 1) Simultaneously display the graphs of multiple 2D or 3D functions, such as displaying 2 or 3 different conic equations simultaneously. Can the 49g do this for 2D? For 3D? 2) Use a symbolic variable name as an element of a matrix (or a vector) and be able to use the matrix in a calculation that, when evaluated, will end up with symbolic results. I do see the 49g supports symbolic eigenvectors and was wondering if this symbolic capability can be applied to more general matrix or vector elements. --tom-- ==== Sorry about the last post - hit a wrong key. I do not as yet own an hp 49g, but am considering the purchase of one. The two following capabilities are important to me, and I'd like to confirm that the 49g can do these things: 1) Simultaneously display the graphs of multiple 2D or 3D functions, such as displaying 2 or 3 different conic equations simultaneously. Can the 49g do this for 2D? For 3D? 2) Use a symbolic variable name as an element of a matrix (or a vector) and be able to use the matrix in a calculation that, when evaluated, will end up with symbolic results. I do see the 49g supports symbolic eigenvectors and was wondering if this symbolic capability can be applied to more general matrix or vector elements. --tom-- ==== Right, and it works with most other 2D plot modes as well. Fast3D simply ignores all expressions of EQ except the first one. However, other types of 3D plots (wireframe, parametric, etc) can be superimposed manually after plotting (using GOR), because faces are transparent. David Haguenauer ==== Look at: ftp://fourier.ujf-grenoble.fr/pub/hp48/vger.flash It should fix 7^x^2=15 x solve bug... Have fun, J.Manrique CdU de la ETSIG ==== how do I use this? -- Andreas Korinek ********************************** 'We tried to establish the expansion of sheep after rain started, but with the techniques available to us and due to the fact that sheep tend to move, the detected expan- sion was statistically not significant (within 3 sigma).' Prof. T. M. Klap232tke in the 'Saturday Times' August 10th,1996 about his research on sheep getting bigger in the rain ==== [...] I believe you should use binary format on both ends, if you want to transmit a string without quotes. ASCII means that what is transmitted to the calculator will be turned into binary on the fly (this can be used to have a user-RPL program displayed as text on the computer). David Haguenauer ==== re: Binary mode ... I suppose that would work but that defeats the translation capability, i.e. translating CRLF's and '' sequences, etc ... I was looking for a way to send a SysRPL program to the calculator without it being interpretted by the calculator as it is received ... that way I could then compile it using MASD which is resident on the calculator ... Mark Sawyer ==== I changed the subject to see if there are replies. I wonder if after all this time, HP still has the exclusive right to produce saturn cpu's, it is well over 10 years old... Any idea? Steve Sousa ==== I don't know whether these sites are relevant to the Saturn, but visit http://www.brouhaha.com/~eric/hpcalc/patents/ and http://users2.ev1.net/~sebastia/patents/patents.htm for information on various patents. James ==== The only reason that they would have exclusive rights to do so would be if they had patents covering it. AFAIK, they don't. You wouldn't be allowed to call your implementation Saturn if HP had a trademark on that, but AFAIK they don't. You probably can't copy the mask or layout of an HP Saturn chip, since those are most likely copyrighted. Similarly for the HP documentation, although some of that is already released under licenses that permit redistribution (subject to some conditions). But if you just want to build your own processor that happens to execute Saturn object code, you should be OK. Of course, IANAL, so YMMV. ==== Don't say this Jordi, especially not for Bin Laden. What kind of Satan can be still introduced to him? Just hope that this misery will end sometime. Wish you all the best for the new year and a better world to all of us for the future. Greetings, ---snipped rest--- ==== You misunderstood, Nick! it's the Satan's punishment to meet ObL in Hell, not the other way around ;-) PS: The probably most efficient way (not yet tested) to limit terrorism is... Announce them pigs and with surgery make it real (on nose & tail & ears at least), see : http://64.4.32.248/l/wc/msnbc/-20103?um=_lang%3DEN&site=http://www%2emsnbc%2 ==== Nousiainen escribi227 en el mensaje http://64.4.32.248/l/wc/msnbc/-20103?um=_lang%3DEN&site=http://www%2emsnbc%2 ...but I can«t go there: unresolved host name error ==== http://www.msnbc.com/news/681021.asp?cp1=1 http://64.4.32.248/l/wc/msnbc/-20103?um=_lang%3DEN&site=http://www%2emsnbc%2 ==== Nousiainen escribi227 en el mensaje character) ==== some day we will be able to grow meat like grass. the set up will have a raditor-like lung and a fish-tank heart-pump. if you cut a piece, it will grow back to its original form. you will be able to untwist your head off your body if you get a headake or insert a module with all existing math & science. then, without knowing how, your body parts will be incorporated into a great ball of breathing meat that has many eye-balls. you will be part of a great meat-ball that puffs and farts. you will hear voices and air noises, not knowing where they come from because you cannot move. then you will wake up. then you will forget about the meat-dream. ==== Is this the assimilation procedure? ;-) Don't you tell me! You made such a module, you crazy electronics master, didn't you? ;-) Considering the last two properties, not much difference to the current state. ;-) Yes, it's awful! Let's have more implants, mechanic arms and robot hands. And implant calculators. ;-) Greetings, ==== And, soon after that, the HP 50G will be released. Martin Cohen ==== Since this concerns all calcs i removed the [ 4 9 ] from the subject. As promised here it goes: The following extract is taken from: HP48 Insights PartII: Problem-Solving Resources By William C. Wickes 1992 Larken Publications Page 622 (hope he doesn't mind...) The current time is encoded in HP48 memory by means of a 52-bit RAM register and a 32-bit electronic counter. The register stores the scheduled time of the next event--the next time-dependent action that the HP 48 must take, such as the next alarm to come due, the next one-second display clock update, or the ten-minute inactivity calculator turn-off. (There is a second time counter that is used for cursor blink, which is therefore independent of the rest of the time system.) The counter is a countdown timer that represents the time remaining until that next event. The current real time then is the difference between the values in the next event register and the counter. That time can be anywhere in the range January 1, 1990 to Dec 30, 2089. The counter decrements by one for each four crystal oscillations, meaning that the fundamental clock tick is actually 1/8192 second. With 32 bits for the counter, the longest the HP48 can go without a time event is 2^31 ticks, slightly more than 3 days. That occurs when the HP48 is turned off, and no alarms are scheduled within that time--then the calculator wakes itself up just long enough to reschedule the next event time. Usually, all of this processing happens so fast that it is not noticeable. The only exception is when you have a large number of alarms stored in the alarm catalog (section 19.2.5);the time required to scan through the list of alarms to find the next due alarm could make the digital clock display miss an occasional one-second display update. The time encoded by the next event register and the timer is deliberately designed to be protected from accidental reset. Even system halts and memory resets (section 5.8) do not affect the time data, even if a reset is caused automatically by the detection of a memory fault. A special checksum is maintained for the next event register; only if that checksum indicates a problem with the time register content itself is the clock reset to the default of midnight, January 1, 1990. Now you all know why sometimes your calc wakes up at night and screams: Bip! Greetings! Steve Sousa ==== Reposting with a subject: I'll take his word for it, but what I've always wondered is, with 32 bits, couldn't it have been designed to go 2^32 ticks without a time event? James ==== Well, I think this was done for two reasons : simplicity and time system integrity. Having the high order bit of the TIMER2 register tied to a ( level sensitive ) cpu interrupt line is simpler than keeping a separate carry bit that behaves differently than the rest of the 32 bits in the register. Also, time system integrity is ensured because a time event can go unserviced for up to 72 hours before the system time becomes inaccurate. This might happen when interrupts have been disabled for some reason and the clock hits zero . As long as they are re-enabled in a reasonable amount of time no harm is done. Hope this answers your question... ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. ==== Off topic: My ISP's news feed has been down for the last couple weeks, so I've been using newsranger.com. I've found that newsranger stops filling in the subject line of a reply when it gets to a quotation mark. Ah, that makes sense! Hmm, are you saying that as long as the high order bit of the TIMER2 register hasn't toggled again all will be well? Well, at least it answers that question. James jmprange@i-is.com ==== The copyright notice is primarily to prevent people from copying the book to avoid buying it (the bane of all authors, print or software), and from plagiarizing sections for use in other books. --Bill Wickes http://groups.google.com/groups?hl=en&selm=17980000%40hpcvra.CV.HP.COM Bye. HPCC member #1046 ==== I've often wondered whether the self-test CPU speed is in terms of this independent cursor blink counter. sentence. I find that for my 48SX version D the valid dates are January 1st, 1989 through December 31st, 2088, and in my 48GX version R and 49G ROM 1.19-6 the valid dates are January 1st, 1991 through December 31st, 2090. The 48s do a warmstart (System time is corrupt.) when midnight of the last valid date arrives, but the 49G simply jumps back a century (no warmstart) to celebrate that New Year. Hmm, I wouldn't mind jumping back a few decades myself, but a full century is out of the question for a while. I wonder how this ties in with TOFF on the 49G. I know that setting a value of #100000000h (2^32) can result in an instant off behaviour that apparently requires a memory clear to break out of and also changes the system time. Suppose I set TOFF to a value of #FFFFFFFFh (2^32-1); I take it that the calc would turn off after only 2^31 ticks of idle time? That is, assuming that I've hooked up an external power supply so that it doesn't shut down due to low batteries first. Well, I've never noticed even a b'p from it when it's off (unless I've set an alarm). Someday I'll have to watch and see if I can catch it in the act. James ==== Exactly. :) ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. ==== C'mon James, there's no need to repeat things every so often :-) http://groups.google.com/groups?hl=en&selm=2ae0fa42.1935comp.sys.hp48%40hpcv bbs.cv.hp.com BTW, any news about HHC 2002? I plan to go to the HPCC 20th Anniversary Conference in London this September. Will they be the same thing? Or on consecutive weekends? Bye. HPCC member #1046 - ==== Oops, again reposting with a subject. ==== If your mobile uses serial protocol for communications with other equipment, then you can use the command XMIT of the 49G to send the AT commands. For example, you place AT0 at stack level 1 and the execute XMIT, which sends the command. Of course, the mobile and the HP49G have to be configurured to the same transfer speed, mode, etc, first. Greetings, P.S. By the way, do you want to transfer solutions over a mobile network during exams? ;-) ==== TEST ==== Greeks once again? Boy, we seem to infuence everybody. ;-) Greekkings, ==== Why? The 49G is itself in a tunnel of updates, corrections of bugs that generate new bugs, hopes and fears and every possible impossible dream for the future. We are in a tunnel without end. Underground calculator, so to speak. ;-) Greetings, ==== I would also like to have a dead 49G. PS: a live one will also do, I'm gonna kill it anyway... ==== I'm trying to convert a program into a library. It has different sub-programs which share the same data, so I created some Global vars to hold this data. I compile the library keeping the global vars hidden (i have the same problem if i don't). And the programs which USE the vars work fine, but the programs which STORE data on the vars don't work, they give me an Undefined Local Name error. How can I get around this? ==== this topic should be covered in the documentation of the Library maker you use. If it's not, then the docs are lacking something very important;-) At least the documentation for the D>LIB nad the <-LIB-> package by Detlef Mueller provides sufficient information on this topic. Search for the word $VAR. If used, it should contain a list of which variables should stay outside the library. Here 'Variables' means one or more arbitrary objects which are referenced by their name in the $VAR list. In general a library is read-only. Raymond Software available on www.hpcalc.org Mithrandir schrieb im Newsbeitrag ==== You can't - it's just that way, no variables can be stored *in* libraries. However your libraries may create/use/store to variables in the current directory, or you can use lambda variables. Thomas -- Thomas Rast If you cannot convince them, t.rast@iname.com confuse them. ==== What I wanted is to keep the varibles IN the librry so i could use them but they weren't accesible to the user. ==== OK, it's a pity. What I will do is create a variable in HOME whith a list containig all the necessary data. My question now, is if I will have to be on the HOME directory to use that data. If everything is how I think, the answer is no. ==== Mithrandir schrieb im Newsbeitrag You can access data variables which are located in the HOME directory from any sub-directory, even from the hidden directory. And libraries attached to HOME have access to those data variables, too. There's one thing that doesn't work directly: Executing programs which are in HOME from a sub-directory. But there's a workaround. Get the current PATH, switch to HOME temporarily, execute the object, evaluate the recalled PATH. Raymond ==== for the original problem with the added notes from Rain-Man below ==== Ok, two questions about the hidden directory. 1. Isn't the hidden directory a sub-directory of HOME? In that case, will I have to temporarily switch to the hidden dir? 2. I read the the hidden directory is created on startup. What does that mean? Normal ON starts, ON-C warmstarts, the fist time you put on your calculator? Does that mean that if I store data there I will loose when one of those happen? Anyway, I'll test these things. Mith. ==== I thought so, but wasn't quite sure. never realized that the escape character had any special meaning to the calculator except in the sense that it keeps track of whether it should be formatting for double-wide to the IR printer. See all of the mug shots at http://pweb.jps.net/~joehorn/hhc2001/people.htm James ==== I wonder if 0 WAIT command have the same function as #41F65h SYSEVAL (WaitForKey) for a very low power consumption. Tal ==== I think so. At least the manual says something similar about low power consumption. But how much lower it is, I don't know. Greetings, ==== Don't forget that WaitForKey returns two bints which represent the key and the plane 0 WAIT only returns one real. -- ==== Power consumtion is just the same apart from a few million eV for additional processing of 0 WAIT. Functionality is different, not speaking here on the different output. If pressing CANCEL, 0 WAIT errors, WaitForKey does not but returns the system key code of the CANCEL key. PS. WaitForKey returns the same key code for shifted keys whether the shift is hold or not. This was a decision of ACO to maintain backward compatibilty (hahaha :-) which makes the life hard for SysRPL programmers. -0600) ==== Yes, 0 WAIT itself executes WaitForKey. To explore the ROM code in HP49G, install Extable and Nosy and perform { WAIT } Nosy; for HP48 install Jazz and use ED. [r->] [OFF] ==== On 9 Jan 2002 02:09:08 -0800, nk@imos-consulting.com (Nick Well, when you call WaitForKey the calculator goes into a low(er) power state via the SHUTDN instruction. This halts the Saturn processor clock but the rest of the Yorke ( eg. the display controller ) continues to function normally. According to http://www.hpcalc.org/hp49/docs/misc/49power.zip , and http://www.hpcalc.org/hp48/docs/misc/drain.htm the calculators draw ~ 4 / 7 mA when the processor is sleeping as compared to ~ 15 / 20 mA when it is running normally for the 48GX(M) and 49 respectively. These numbers assume the keyboard and display are enabled and there are no wire/ir transfers or flash writing going on. :) ( in which case they will be substantially higher. In the case of writing to flash something like 8 (!) times higher. :) ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. ==== Well, i m not a boss with the HP 40G... I dont know if somebody has got 40G here... You can try to ask this question here http://www.hp-network.com/forum Eric 3C3B9B6F.BE7021D8@iinet.net.au... ==== hi there, is there somewhere on the net a site where i can find an advanced user manual 4 the HP48G? I want to know more about programming the HP. In the standard manual this chapter is really small. So if someone knows where to find an online manual or pdf it sure would be appreciated. ==== Wich lanageg do you want to learn ? RPL ? www.HPcalc.org There are some good docs about this langage on this web site :) Eric Mickeysoft a 216crit dans le message de cE__7.15245$DG4.3315245274@hestia.telenet-ops.be... ==== M--- Not that I've been able to find. AFAIK, you can get them directly from HP. I know that a replacement user's manual for the G/GX is $28 through them. Good luck, and if you do happen to find it PDF'd somewhere, let us know. Henry C. Gernhardt, III ==== i recently migrated form my old 200Mhz PC to a much faster one (450MHz). On my old linux box the x48 worked just fine, but on the faster one the emulators backspace <- and cursor keys will not work on a character on character basis but i.e. delete an arbitrary number of characters. I already tried an Windows emulator with the same results and also a different ROM. Probably this is a frequently asced question, can anyone give me a hint? TIA, Oliver -- Theoretische Elektrotechnik 2-06 TU Hamburg-Harburg Harburger Schlossstr. 20 21079 Hamburg ==== On 8 Jan 2002 08:21:27 -0800, mochoa@exatec.itesm.mx (Manuel Ochoa) Not to my knowledge. All of the compression programs I know of for the 48/49 used some type of LZ77 based algorithm. Are you planning on writing a GIF viewer for the calculator? ;) ---------------------------------------------------------------------------- --- Jonathan Busby - before replying. ==== That is probably very true, but I have no need for a 25000 $ machine to run couple of small applications :-) I just noticed, that the CAD my company uses has a potential to crash Solaris to the point that it needs reboot, because machine does not respond neither to user input nor to remote commands from different workstation. And this happens several times a year. It is not a big deal, but my Windows 2000 do not freeze that often :-) I just find stories about Windows 2000 instability greatly overstated. If the machine has good hardware and decent software it might crash. I once had faulty memory DIMM back then when I was using NT. Every GURU was blaming my NT for blue screen of death. When the DIMM was repalced, the problem was gone. I have no reason to complain about Windows instability since I was using NT about 3-4 years ago. So the GUI must be from the old OS, not X Windows. Therefore Sun Unix applications need much more than just recompilation to the new Mac CPU or I must miss something. I'm sure, that old Mac GUI API are not mirror copy of X Windows API. Unless you emulate X Windows API into the calls into the Mac GUI (probably doable but not efficient) you cannot use Unix applications on Mac so easily. Of course I don't have in mind command line type applications. I disagree. Switching from native OS to Unix is in a certain sense admit of defeat wither technical or economical. would that be still Windows or admit of technical defeat ? I don't deny. Unix was a professional OS long before Windows, and did had all the bells and whistles that everybody later copied, but Unix was created for the expensive mainframes, not home PC. In fact in the Home PC area the first preemptive multitasking OS was Commodore Amiga altough it did not had protected memory space. What Mac introduced was GUI but that WAS NOT APPLE invention. In fact Apple COPIED GUI idea from Xerox. So did Microsoft later from Mac :-) The point here is, that there is nothing really wrong with implementing good competition ideas into one's own product (unless you are stealing a patent). In any other business like for example automotive copying from good ideas of competition is everyday acceptable practice and nobody really is outraged about it. And shouldn't be. Good ideas are for the customer advantage. Jack ==== You may have just one application that does everything that you want. But this can't be said to be true for everyone. There are many many people that *have to* work with many smaller software packages from different vendors. Sounds strange, doesn't it? My knowledge about the inner workings of unix systems are not very good, but I thought that such systems are supposed not to be influenced by an application that crashes. How can this happen? But I have reasons to complain. Probably because I didn't follow the blessed by M$ way of working with NT at the office. The GUI of MacOS X is not X Windows. It's Aqua and it is not from the old OS. The GUI used for the old (classic) Mac applications is the GUI of the latest classic OS. (System 9 I think) If some program solely uses libs to construct its interface, it should be easily compiled on a Mac running MacOS X. Of course this has several problems. Source codes are not always available, applications do not use only libs, but have thir own quirks and tricks and the like. This must be a joke, isn' it? Again, libs! The unix running on Macs nowadays *is* native. If by *native* you mean what used to be then please consider the switch from Win 3.whatsoever to Win95 and all later versions also to be like admiting a defeat. For the majority of the users it would be windows, if what they see and how they interact with the system, is like they are used to. For tech freaks it would be perhaps a defeat. This again must be a joke. Unis created for the expensive mainframes? Yes. So I can blame Apple for copying the idea from Xerox. But there is a difference. Apple behaved enthousiastically right from the start. They didn't say that a GUI is only for children, but they tried to propagate the idea of easy interacting with the computer. On the contrary the whole rest of the world was just naming a GUI a toy which isn't good for professionals. And then, a couple of years later they started telling me about ergonomy and how nice their new systems are. You are considering the economic aspects and from this point of view you are right. But I think about technology and inventors. And I see that Apple has brought so many ideas which then were presented as brand new by other companies ==== I( and all the people i know both professional and no professionals) find MS office very easy to use and ergonomic. Even for doing weird such as those you suggest. So with a better/easier/more intuitive OS,Apple has not even able to win over MS ? Then they should be complete asses and they deserve what happen to them. MS is far from being a dictator. Btw do you imagine how worse it would have been if Apple was at the place of MS ? At least with Windows,you have almost complete hardware freedom. Something you don't have at all with Mac computers. ==== An application freeze - like X seizing up - is not an operating system issue. Solaris cannot be con- sidered to be unstable unless it actually panics. Otherwise it's still running, still managing your machine as best it can under the current load. A slow, unresponsive machine is not one with a buggy OS, it's one with too much to do. The other symptoms you are describing are common to machines that have had privileged applications use up nearly all available physical and virtual memory. That, too, is not the OS's fault. It can't manage resources you don't have, all it can do is try to juggle the ones you do have - and what's more, such problems are entirely the user's fault. You are either running an app that leaks memory, or you are simply asking the machine to do more than you have resources for. Again the OS is not to blame for overloading the machine. Unices never crash. They panic, with a full diagnostic dump so the manufacturer can tell what happened, and can frequently still flush the file buffers and shut down gracefully. If your machine isn't panicing, then you have a stable OS. This doesn't guarantee the machine can always do whatever you want it to, and it is still possible to screw it up one way or another - but it's not the OS's fault that you can do so. It's not there to protect you from yourself, it's there to manage the machine and to keep you from shooting yourself in the foot accidently. If you insist on DELIBERATELY shooting yourself in the foot Solaris, like any good Unix, will do its utmost to assist you. You want to see REAL instability, try doing what you are doing to that Solaris box to one running NT. Windows systems do not panic, they crash most ungracefully leaving all sorts of damage to the software environment in the process. -- ==== ha ha. nicely put ;) ==== On the HP49, VARS and TVARS are fully written in ML, and I doubt that you would notice a big time difference even if your directory contains over 1000 variables. ==== I appreciate it if you could guide me how to turn the contrast down in User RPL (HP48). Tal ==== You can't. Use the ON, - key combination to do it -- ==== The only way I know is to hold down the Key [ON] and at the same time press the key [-]. Greetings, ==== It's possible in ML or by using the PEEK/POKE commands : For 48s/sx : http://20020109-ng-cshp48.courbis.com/hp48ml-241.html (CONTRAST program) http://20020109-ng-cshp48.courbis.com/hp48ml-220.html (PEEK) http://20020109-ng-cshp48.courbis.com/hp48ml-222.html (POKE) For 48g/gx (in French) : http://20020109-ng-cshp48.courbis.com/v48g-330.html (CONTRAST) http://20020109-ng-cshp48.courbis.com/v48g-301.html (PEEK) http://20020109-ng-cshp48.courbis.com/v48g-304.html (POKE) contrast. Paul ==== You can do it also with the ASM, you must change a value at #001?? ( i dont remember ) Eric 3C3C2AA9.2080707@courbis.net... contrast. ==== Well, I know these ones: http://groups.google.com http://newsone.net http://www.squeakycleannews.com http://easyusenet.com http://www.news2web.com http://www.etin.com (down) And two or three more places I've forgotten. And try this: http://www.newsreaders.com/link/link.cgi?group=comp.sys.hp48 Please, add more links!! Bye. HPCC member #1046 - ==== thanks. which one do you think is the best if i may ask? newsranger was great. says... ==== I appreciate it if you could guide me how to create a picture in 4 level gray. Tal ==== go to http://www.xnview.com/ and download XNView you can convert any picture (including TI92 format :-) to a 2,4,8,16 & 32 shade picture Then send it to your 48 ==== It is not exactly what I need. I need to CREATE and not to convert. ==== What do you mean create. Just drawing them? That you can do with the standard PICT environment on your 48, take a look at some greyscales to see how. ==== Welle thre is also grob edit you can download hhttp://www.hp-network.com/tests/grobedit/grobedit.shtml and maybe on HPcalc.org ==== Yoann says: Exact, that's because integers aren't coded the same way as the real numbers. This has already been fixed. ==== I postet a few days ago«, that I need a method to enter Complex Values in R Teta Mode. I got the following Answer I tried it out but it didn't work. Here my complete Keys 5 + SPC + ALPHA + RS + 6 + 45 Then I got 5 <45, but if I enter this equation I get the Error Message: Invalid Syntax and the < Is marked. Thx for help Andreas ==== i think you forgot to put the parenthesis: (5 <45) complx numbers use parenthesis on the HP49G don't be afraid to ask anything. the manual sucks. says... ==== Your key sequence is almost OK. Just add LS + - to put the whole number in parenthesis. Complex numbers are like (5 <45) on the HP49G. Greetings, ==== On 7 Jan 2002 16:46:13 -0800, NeyT@dnr.qld.gov.au (Timothy Ney) Well, it depends what you mean by simple ;). The following is surely much simpler than writing a real number parser yourself :). You need to call the rpl PCO realPAcode using, for example, the code supplied here : http://groups.google.com/groups?selm=sl6ottg61bn46231s3piplssarv9obvulk%404a x.com Using the above, it might look like this : CON(5) =DOCODE REL(5) end . . . GOSBVL =GPPushA * A contains address of string * prologue. STARTRPL tokexponent tok. realPAcode NOTcsdrpfls NOTcsdrpfls TRUE ENDRPL GOSBVL =popflag GONC error * If we just popped FALSE then the * string was not parsable. * Otherwise the prologue of the real number will be on the top * of the stack. GOSBVL =POP1% * RPL pointers saved. Real number in * A.W . ==== Just a little =) If you change your mind, it's pretty easy to just stick some SysRPL right after your Code object in a secondary. ==== My first thought was to minimize the operations inside of the loop, and then after looking at the problem, it occurred to me that it might be easier to build a new list for comparison rather than step through the existing list. I came up with the following: (For 49G) << DUP HEAD OVER SIZE OVER + 1 - {} UNROT FOR i i + NEXT == NOT >> For 48G, replace UNROT with ROT ROT. For 48S, also replace HEAD with 1 GET. Not tested on the 48s. James ==== First, the NOT at the end is in error. Second, on the 49G, it turns out that the == test returns 0 if one list is of reals and the other is of ZINTs, so it needs an I->R inserted as the first command in the program. Third, for the 48G and 49G, the SEQ command is faster. (For 48G or 49G) << I->R DUP 'n' DUP ROT DUP HEAD SWAP SIZE OVER + 1. - 1. SEQ == >> James ==== Rats! Of course, for the 48G, I->R is neither needed nor avalable, so the 48G: << DUP 'n' DUP ROT DUP HEAD SWAP SIZE OVER + 1 - 1 SEQ == >> I hope that I've gotten it right at last. James ==== << 1. SWAP 2. << < * >> DOSUBS >> Werner Huysegoms werner-huysegoms@freegates.bex omit the trailing x ==== What about << DLIST PLIST 1 == >> where D is the greek capital Delta and P is the greek capital pi ? Greetings, ==== That doesn't work for the goven examples. This one works: << 1 SWAP 2 << - -1 == AND >> DOSUBS >> On my real HP49, I tested it on a list of 50 reals, and it took 20895 ticks. -- Erwann ABALEA erwann@abalea.com ----- ``We're operating from a knowledge base that is not very dense.'' Jim Skeen Explaining how to say that we don't know what we are doing. ==== Interesting, but try it on the list { 3 2 1 }. James ==== Actually, I haven't figured out what character < is supposed to be. I guess it's a typo. Nice, but for the list { 1. 2. ... 50. }, I got 27071 ticks for your program, compared to 9627 ticks for my program: << I->R DUP 'n' DUP ROT DUP HEAD SWAP SIZE OVER + 1. - 1. SEQ == >> For a list of 100 reals, { 1. 2. ... 100. }, I got 54253 ticks for yours compared to 16745 ticks for mine. On the other hand, yours is only 45.5 bytes compared to 63 bytes for mine. ==== what is the < in the fourth line? pleaze ==== Yes, I didn't choke on it, and thought it was a simple '<'... Sorry, I only got my HP49 a couple of days ago (I was previously a happy HP28S/42S/48SX/48GX user), so I didn't take into account that my list was composed of integers rather than reals... The program works slower when used with reals... Nice... Yes, it seems faster to build a new list composed of the same number of elements than processing the entire list. I also tested something with STREAM, but it's even slower than using DOSUBS... And yours creates a new list with the same number of elements. The calc should be able to handle twice the size of the list in memory. ==== That explains why I didn't recognize it in the first message. What does I->R do, exactly? -- http://www.mschaef.com ==== I guess it's my fault for not specifying that the numbers in my list are never <0 and the list is always sorted too. Timing your method on my GX showed it to be about twice as fast as other methods so that's what I decided on so I'm a happy camper. ==== elements [...] What about: << DUP TAIL SWAP 1. OVER SIZE 1 - SUB - DUP NOT NOT == >> strictly I don't know how well this program will behave for long lists, but it uses some built-in list-processing functions instead of a FOR loop, which should speed up operations at least a little. The idea, with my program is: o subtract the list from itself shifted, yielding a list of ones if the list consisted of consecutive integers o then check whether the list was indeed full of ones, which means that NOT NOT won't change it ==== When HP was still subsidizing the calculator classes offered through the Math Learning Center (Portland, OR), I took an HP 39G class and a 49G class and was able to play around with the box of Xpanders that the instructor had brought to the classes. The Xpanders were assigned to his high school math students for evaluation purposes. IIRC, they learned the Xpander couldn't be recharged while it was being operated, because the blue translucent cover covered-up the charging terminals when slid onto the back of the Xpander. As you know, it was basically a Windows CE machine. I thought it was pretty cool, but the instructor said he didn't have a good feeling about it in it's present state of development (in other words, it was pretty buggy). In any case, I sure do wish I could've gotten one 'em. Matt ==== You're right - the NDA doesn't state that the items are to be returned to HP. I believe they still own the devices though - if they can enforce it is another issue. At least 4000 Xpanders are not acounted for, so I guess you're right again ==== Thats not the 1st time thre s a Xpander for Sale 6months ago, there was the same model (but the seller was differrent....) What did arrive for the last saler ? ==== Forced is a strong word. The HP legal department surely requested it. But eBay basically will remove anything that anyone complains about. I tried to auction an Apex AD-600A DVD player which I purchased at a local electronics store. I clearly owned the player and had the right to sell it. eBay terminated my auction because Macrovision complained about it. ==== In the absence of some other legal agreement, then, the device is a gift. HP has no lawful claim to it. That won't prevent them from *claiming* they own it, though. ==== What is the way to do interrupt, when my program waits for key? If I don't press any key it should jump to subroutine every some time then return to main program. Is it possible in SysRPL or only in ML? The subroutine should to be executed exactly every the same time, so it have to be interrupt. Piotr Kowalewski. ==== If it has to be SysRPL or ML then drop this message. But if it can also be UserRPL, then the following code snippet would do: .... The zeros will be used by UNTIL to continue the loop. If you would like to exit the loop when some particular key has been pressed, then some code has to be added to check for that key and return 1 if it has been pressed: .... DO IF KEY THEN IF 105. == THEN 1 ELSE thingsToDoIfKeyPressed 0 END ELSE thingsToDoIfNoKeyPressed 0 END UNTIL END ==== OK, but... when I press a key, the subroutine will be executed later, but it has to be executed ALWAYS every the same time. Because for example if I use this subroutine for animation, the animation will be slower when I press key. So it should be done by interrupt (in my opinion), but maybe there is another solution. Piotr Kowalewski. ==== ==== I have an extra copy in brand new condition if you wan to buy it from me. I can sell it to you for $38 including shipping. I have seen it sell for $48 on the internet. ==== As a thankyou I thought I'd post this. It was the reason I had the query in the first place %%HP: T(3)A(D)F(.); @ Vmenu is a basic UserRPL Menu program using Choose to @ select a directory and it's equations and programs. It will also @ recognize a List set up with an equation and custom menu's for @ use with the 30 MENU solver. @ If an equation is selected then the 30 MENU Equation @ solver is started. Selection of a program will run @ the program. @ The program caches the contents of the directory in a @ variable $VMENU on its first scan of a directory. $VMENU @ can be refressed at any time with a menu selection. @ I've only used UserRPL for safety and imagine SysRPL @ could speed things up and maybe elliminate the need for @ a cache. @ Most of this was done while travelling for Christmas and @ is a tribute to the portability of the HP48/49 series. @ Anybody is free to improve this I only request you send @ me a copy of your efforts. @ This program needs OT49 Library by Wolfgang Rautenberg @ and a small utility Here by John H. Myers loaded. @ and support for the HP49. @ Stephen Nash ssnash&bigpond.net.au @ OT49 can be found on www.hpcalc.org @ The 'Here' utility is created with the following D9D201DF604CE52B2130 #100001h LIBEVAL HOME 'Here' STO << << {} SWAP 1 << IF DUP TYPE 6 =/ THEN IF DUP TYPE 5 == THEN 1 ->LIST END + ELSE DUP RCL -> X1 X << CASE X TYPE 5 == THEN IF X HEAD TYPE 9 == THEN X1 DUP {} SWAP + SWAP 2 ->LIST 1 ->LIST + END END X TYPE 8 == THEN X1 DUP <> SWAP + SWAP 2 ->LIST 1 ->LIST + END X TYPE 9 == THEN X1 DUP '' SWAP + SWAP 2 ->LIST 1 ->LIST + END X TYPE 15 == THEN X1 DUP _ SWAP + SWAP 2 ->LIST 1 ->LIST + END END >> END >> DOSUBS >> -> Process << DO Choose Variable to solve IF '$VMENU' Here VTYPE -1 > $VMENU TYPE 5 == AND THEN IF $VMENU Refresh POS 0 > THEN $VMENU ELSE {Refresh} $VMENU + END ELSE {Refresh} { {Change Directory CD} } + 15 TVARS + {..updir} + {CST IOPAR GSPAR PPAR PRTPAR TPAR USEREQ VPAR ZPAR $HIDDEN VMENU } @ Remove some common variables IF $HIDDEN TYPE 5 == THEN $HIDDEN + END @ Remove optional hidden items 173 Process EVAL DUP '$VMENU' STO END 1. UNTIL IF CHOOSE THEN -> X << CASE X TYPE 2 == THEN CASE X CD == THEN CD 0 END X ..updir == THEN UPDIR 0 END X Refresh == THEN Refresh '$VMENU' STO 0 END 0 END END X RCL TYPE 5. == THEN X RCL STEQ 30. MENU 1 END X RCL TYPE 9. == THEN X RCL STEQ 30. MENU 1 END X RCL TYPE 8. == THEN X EVAL 1 END X RCL TYPE 15. == THEN X EVAL 0 END 1. END >> ELSE 1. END END >> >> 'Vmenu' STO ==== It may be stable, but it has brought its own ration of problems. One example, when trying to open a word or excel spreadsheet by either double clicking, or right shift open, the communication channel to Office 97 results in an attempt to open the file twice. For Word97, you have to decline the second attempt to open; for Excel, you have to acknowledge that you were stupid and tried to open the document twice. Everybody in my facility with Win2000 has this problem with their install, and MIS has not made the jump to office 2000 yet, which theoretically is more compatible. Bill alternate E-dress wtstorey@ieee.org.no.spam.please (Use the obvious) ==== Ok the last time i was mad because of family problems but now i feel better. Let me answer as i should have. Perhaps ergonomic is not the good word.I should have use the words:consistent through apps,doing exactly what is expected and very intuitive for doing things which as generally a pain in the ass to do on other O.S for the average user. Last time i checked it wasn't that easy to install an apps on And some thought that it is me which is a baby duck. that Windows doesn't have ? And remember that we are talking about the average user not about computer wizzards. No, only Windows ME. But perhaps you haven't seriously used any Windows OS besides Windows ME ? If so i would understand your dislike of Windows. Windows 98 SE was significantly better than Windows ME even if Windows ME has brought a very interesting feature to the Windows familly. MS should have never skipped Windows ME and wait to directly release XP as the successor of Win98 SE. You don't know windows 98 SE a.k.a Windows 98 second edition ? What real problem do you have with Windows 2000 ? Oh,i have already use X windows but it is quite a poor GUI in comparaison of Windows or Mac OS. And i haven't found that it has any significant speed advantage over Windows on the same machine. or Unix. Anyway,KDE is a much better GUI and is on par with Windows and Mac OS.Too bad it is significantly slower. Oh i am talking about all those MS bashers not enough objective to admit the good things that MS achieved or the good aspects of Windows. I have said it before and i will say it again,i am not a MS lover or drone. I don't like many thing things about MS like for example what they have done to the internet browsers market. However objectivity enables me to admit what they do right and what they do wrong. Can you say the same ? ==== I'm currently using Office premium 2000 FI and I'm planning to upgrade to Office XP later this year If there is a change that it's compatible down to Office97 example, or attempt attempt to open problem which ==== Well Win95, Win 98, WinME (almost) worked fine and NT40 also worked, WinXP so far (almost) ok AND the only serious problem with the W2K is the lost paging file Is NOT a HW problem (new IBM 75GB) Intel (inside, idiot outside) mainboard, which I later will upgrade keyboard is fully functioning, I try to upgrade the driver... Device drivers belong rather to the OS than GUI (except perhaps display card/monitor drivers ;) XP refers to the WinXP OS. Hmmm. Maybe AMD will release also a JackHammer CPU ;-) ==== OK, they are coming up in the next days. Many thanks in advance. Of course I know that it is absolutely no pain to install an application on a previous version of MacOS, because I do that all the time. Both Apple and M$ targeted the rest of us-market. Oh yes, they are. Well, I wouldn't like to see M$ completely stop. Not because I believe in their innovations but because I find it always a pity, when some enterprize vanishes. No, it isn't. It is about the same. The incompatible Mac is a myth. See the specifications of a Mac to see what I mean. It seems now that both frienly co-operate more than ever. That means, beware Apple or the monster will eat you up. ;-) Anyway, what happened, just happened. Now the rest of the world should be aware not to be eaten up by M$. I think they are more aggresive in their strategy than any other company. Greetings, ==== On the HP 49G, is it possible to create sub directories in the CST menu (because in fact I have many buttons and I would like to sort them)? Fir£fly /////// ( o o ) ooO-( )-Ooo X-no-archive=yes ==== Save the following as CST: { {MATRX <<{RREF DET {UP <>}} TMENU>>} {SUB2 <<{command1 command2 ...{UP <>}} TMENU>>} } {UP <>} will bring you to the root directory. Hope this helps. Lin ==== hi, a 'menu' (call) is very different from a 'dir'. a menu is not a structure, therefore it has nothing inside but calls (to a program, and object, etc). you could use a custom menu that evaluates the path {} to different directories example { { dos <<{home uno dos} eval>>} { tres <<{home uno dos tres} eval>>} } cst sto ==== http://www.hpcalc.org/details.php?id=4360 ==== I have for an auction on eBay an HP-41CX. Included in the auction is a card reader, Advantage Advanced Solutions Pac, IR Printer module, all manuals, two third party books and a few odds and ends. Please check it out at: http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?ViewItem&item=1321837090 If you have any questions, please let me know. Bill McMannis PS: Look for my upcoming auctions of HP 65 software and an HP 95LX ==== i am using the FEM48/49-program in the version 4.4 on a HP 49G. When i am trying to use the Wizzard, the Query or the Print-function, there is the failure-screen: Error: FEM49 Not Compatible. Can you tell me please, what is wrong? Thanxx a lot... ==== Does anybody know if there is a way to use one of those Gameboy Snake Lamps on a HP49G? This lamps use the gb-socket for power - maybe there's a way to use it on a calc... ==== you could use a mechanic snake lamp and put it around the neck, he he. (and a scarf on top to disguise it) i'ver seen that lamp. if the lamp has its own power (batteries) then it is OK if not, then it is not OK. PS: i've seen glasses with two light bulbs on the frame. those will work, he he ==== you have to EVAL (red + p) the program. or if you have it on a menu key (F1 to F6), just press the Key press the VAR key to see your directory programs. ==== Hallo, bin ein alter HP-Fan und suche eine gut erhaltene und funktionsf212hige HP01-Uhr, aber bitte nicht zu Mondpreisen. Kann mir jemand helfen? Gru§ Marcus e-Mail.: marcus.herold@gmx.de ==== Program gives a syntex error at < NEXT > program runs fine till it hits the START NEXT loop......I'm about ready to give up will someone please help me... %%HP: T(3)A(D)F(.); << (L) Lap ? INPUT OBJ-> FD DUP IF 1.625 == THEN 1.625 .1354 - ELSE .0104 - END (R) Lap ? INPUT OBJ-> FD DUP IF 1.625 == THEN 1.625 .1354 - ELSE .0104 - END -> a1 a2 << Bay Width ? INPUT OBJ-> FD -> a3 << a1 a2 + a3 + -> pl << Rows Of Sag Angles ? INPUT OBJ-> 1 + -> a4 << a3 a4 / -> sp << a1 sp + -> sp1 << CLLCD sp1 DF Key 1 0 FIX + = + ->TAG 4 FIX 1 DISP 1 FREEZE a4 2 - -> a4 << 2 -> a5 << START STD sp1 sp + -> sp1 << sp1 DF Key a5 + = + ->TAG 4 FIX a5 DISP a5 FREEZE a5 1 + -> a5 << NEXT pl DF purlin len + = + ->TAG 4 FIX a5 1 + DISP a5 1 + FREEZE >> ==== Maybe I'm missing something, but I think you should review a lot more of lines, because several local programs end delimitators missing. Take a view: 11 opening delimitators just 1 closing delimitator Gregorio Rodr222guez ==== It looks to me as if it wouldn't even download, let alone run at all. The reason is that after the START, you've begun a program, and a nested program within that, and neither program has been ended before the NEXT, so the syntax error occurs. What would it mean to begin a program within a loop and finish it after the loop is all done? Also, it's hard to tell what this program will do without some idea of what FD and DF are supposed to do. See additional comments below. <%%HP: T(3)A(D)F(.); <<< <(L) Lap ? < INPUT OBJ-> FD DUP FD DUP a1 a2 << < FD -> a3 << pl << < 1 + -> a4 << sp << sp1 << < TAG 4 FIX <1 DISP <1 FREEZE < a4 << 2 -> a5 << You need two real numbers on the stack for START; are they there? sp1 << TAG 4 FIX < a5 << sp1 << sp1 DF Key a5 + = + ->TAG 4 FIX a5 DISP a5 FREEZE a5 1 + -> a5 << >> >> NEXT and I expect that it would successfully download, but it looks as if you mean to change the value of sp1 and a5 with each iteration of the loop, and that won't do it. The sp1 and a5 created would be new local variables that went into the bit bucket when their defining programs ended, so the next iteration would use the old local variables. Instead of creating new local variables, use STO to replace the contents of the existing local variables, like this: START STD sp1 sp + 'sp1' STO sp1 DF Key a5 + = + ->TAG 4 FIX a5 DISP a5 FREEZE a5 1 + 'a5' STO NEXT TAG 4 FIX < > Note that you can avoid the nested local variable structures by using dummy values to create all of the local variables at once, and then using STO to replace their values as desired. You might do well to minimize your dependence on local variables and use the stack manipulation commands instead. James ==== Where i can find electrician library for HP 49 g? Thanck's to all user! ==== Stefan schrieb im Newsbeitrag I have two of them and none ever caused problems (except of the clock...) I was told (by someone that should know) that this mod doubles the clock-speed... The increase of speed seems to depend on the function evaluated (dont ask me why). For example the filer doesn't seem to be faster but graphics are... ==== Peter Geelhoed schrieb im Newsbeitrag Hell!!! Never heard of backups??? ==== Please don't report any bug with this vger.flash file. It's not up to date, and I don't want things to be messed up between version ! It will be impossible to sort things out after that. I've entered a new bug #68 into the bug tracking system. This bug manif= ests in the newest vger.flash of 1/4/02, involving a equation entry loss in = the plot functions app. -- Andreas Korinek ********************************** Viel hilft viel und mehr bringt mehr. ==== Actually I have had this problem for quite awhile. If you look at your plot window it shows Y(10) but with no equal. I posted this problem to the group awhile ago and no one else could duplicate it. BUT!!! I changed the two things today: I unchecked pixel in the 2d/3d window and changed the numbers in the v/h ticks and lo and behold it works the right way all of a sudden. Hmmmm I haven't tried changing it back cause I'm just happy that it displays the plot correctly. BTW I am not running the 1/04/02 roms but the 1st 19.6 betas that came out. Maybe this will help.. Very Confused I've entered a new bug #68 into the bug tracking system. This bug manif= ests in the newest vger.flash of 1/4/02, involving a equation entry loss in = the plot functions app. -- Andreas Korinek ********************************** Viel hilft viel und mehr bringt mehr. ==== Sorry, but I just though of something more... You have to delete the Y(X) without the = and then select the soft key [choose] and it will bring up all your equations. If you do not delete the blank Y(X) then it will draw (plot) the last equation. But you can't edit or do anything else to it. I've entered a new bug #68 into the bug tracking system. This bug manif= ests in the newest vger.flash of 1/4/02, involving a equation entry loss in = the plot functions app. -- Andreas Korinek ********************************** Viel hilft viel und mehr bringt mehr. ==== which one is up to date then? Perhaps I could flash this one instead. -- Andreas Korinek ********************************** Viel hilft viel und mehr bringt mehr. ==== Andreas Korinek a 216crit : to stop the distribution of my CAS updates during that time. ==== I am solving a second-order system of differential equations on my 49G using the (fourth-order) RKF function, and I am having trouble...When I use the input form (non-programmable) approach, everything works well; but when I set up the problem by storing the necessary function, initial conditions, etc. and invoke the RKF function (with the proper arguments), the calculator errors. Interestingly, the RKF function works on the problem perfectly if I use my 48GX instead! Is there some flag or setting in the 49G that I might be overlooking? I am using OS version 1.19-5. I have written the differential function as a program, as it is a complicated vector function, but that shouldn't cause any problems as the input syntax is correct, and the 48 likes it just fine, as does the 49 input form...Any ideas? ==== Sorry to hear this. FEM is a very impressive achievment for the HP49. Very usable, respectable speed and interface well suited to the 49. I never thought I'd use a frame analysis package on a calculator. All the best Caspar. message some of ==== J.Manrique larry ==== The same that 1.19-6 ROM... J.Manrique CdU de la ETSIG ==== i have been testing the speed difference between the HP48GX and the HP49G, and the HP48GX is quite faster! run this to hear the difference (if you have both): << DO RAND 1000 * .13 BEEP UNTIL 1 2 == END >> it runs at least *3* times faster on the HP48. why? (i cleared the memory on the HP49G before running it to go even faster) ==== Because HP48 doesn't have ZINTs (integers) and everything is done using reals. You should add . after every number in HP49 program - it will be much faster: << DO RAND 1000. * .13 BEEP UNTIL 1. 2. == END >> Piotr Kowalewski ==== Kowalewscy escribi227 en el mensaje de noticias a1q06h$gho$1@news.tpi.pl... or edit the program and then press ENTER in aprox mode ==== yeap THANKS i read this before, but things don't stick on me unless it happens to me! ==== this small program runs at least *5* times slower on the HP49G than on the HP48GX. why? << -> N 'IFTE(N<=1., N, FIBO(N-1.)+FIBO(N-2.))' >> FIBO STO *the program needs to be stored as 'FIBO' (it calls itself). then put a number on the stack. example 10 FIBO (wait a little: on the HP48GX i counted 6, on the HP49G i counted 40!!!) 55 ==== Algebraic expressions seem to be slower in hp49 than in hp48, but this doesn't happen when all the program is written in rpn: << -> N << N 1. <= << N >> << N 1. - FIBO N 2. - FIBO + >> IFT >> >> This take only 4.6 sec in the hp49 (3.9 sec with IF...THEN...ELSE). I don't have a hp48, but I suppose that take the same time. rcobo escribi227 en el mensaje de noticias 639cb51.0201130513.73392141@posting.google.com... ==== When you're using algebraic expression like this, the CAS will be called, so it can be MUCH slower. However, you can do things you couldn't even dream of on the HP48 ==== Waterloo Maple 7: 63245.553155933421717687945514831 (32 digits precision) <9thc5r$3s3$1@news.wanadoo.es> <3c406b3d$0$75157$ba620e4c@news.skynet.be> ==== hp49g in exact mode, after EXPAND: 63245.5531559 GNU bc with scale=100: 63245.5531559334217176879455148307848234606062699224835562064121292337049831 666120370564244401257846332191 -- Erwann ABALEA erwann@abalea.com ----- (A)bort, (R)etry, (S)mack the @#$&*! thing! ==== So? What's the point? Here's Mathematica's result to 100 digits: 63245.5531559334217176879455148307848234606062699224835562064121292337049831 66 61203705642444012601300 If you want it to 100000 digits, you can get that in less than 1/2 second. ==== with my program i can choose till how much digits i want to go, even 1000000 than the one given by the win98 calculator, here the result till 100 digits, again with Waterloo Maple: Bhuvanesh schreef in bericht precision) 63245.5531559334217176879455148307848234606062699224835562064121292337049831 66 ==== with my program i can choose till how much digits i want to go, even 1000000 than the one given by the win98 calculator, here the result till 100 digits, again with Waterloo Maple: 63245.5531559334217176879455148307848234606062699224835562064121292337049831 6661203705642444012601300 Bhuvanesh schreef in bericht precision) 63245.5531559334217176879455148307848234606062699224835562064121292337049831 66 ==== When I said 32 digits precision in w98 calc, I did reference to an arithmetic of 32 digits with floating point, for that reason the 32 digits obtained are not all correct, since in this example there is cancellation of digits. needless to say escribi227 en el mensaje de noticias 3c40d780$0$33512$ba620e4c@news.skynet.be... 1000000 digits, 63245.553155933421717687945514830784823460606269922483556206412129233704983 1 digit 63245.553155933421717687945514830784823460606269922483556206412129233704983 1 second ==== OK, but I still don't get the point of this. The calculators do not have arbitrary-precision arithmetic, so this arbitrary-precision stuff doesn't seem to make sense to me in this context.