D-119 ==== And of course you can't explain why you find this method superior . DUH I am a modular , FORTH type . I need not restrictions on piecing together my programs from general purpose fragments aka bottom up programming . User mode is slower . more key strokes . I place all fragments in HOME . It's messy only when you don't keep a note book to explain function of each . Don't short change this step ! Thomas Rast < Well, the book you mention isn't an official HP product as it seems you understood. And to start it is enough to read the manual, the advanced user guide (especially) and trying to experiment, solving your probs with the use of your machine. Also a look to www.hpcalc.org should be in your to do list cause here you can find lots of docs for free, both official and non (those ones are the most part). For programming you can refer to the HP48 documentation, cause large part of information concerning the HP48 could be applied to the ... Mr. Urroz; here you can find real problems solved with the HP49G, with the book itself divided in chapters each covering an aspect of the HP49. There's also another book, whose name is, as you probably can imagine: Science & Engineering Mathematics with the HP49G, vol. 2 This one covers more advanced topics (uni/multivariate calculus, O/PDE and statistics). The books are available in electronic and paper form at www.greatunpublished.com (search for E. Urroz books) Hope this can help Bye, Kickaha ok!! but where can i have a similar book freely? i'm a poor student! i can't pay them!! :-( Well, you can ask to Mr Urroz if him practices some sort of student facilitation... Or you can read the AUG trying to solve every university problem through the 49er, which is, I know, an uncomfortable solution, BUT the cheapest I can think of. Anyway $20 isn't too much for a large book like that. I've only a reserve to this books... they are full of errors, especially in exercises and some functions don't act exactly how the book describes them... maybe changes due to ROM upgrades, but I don't know... I wonder if the autor will release an erata corrige for free... it would be great! Bye, Kickaha Gents: FDP.exe is the flash update software posted on hpcalc.org. It is suposed to be just that without the other bells and wistles of HPcomm (and therefore smaller and easyer to email) The instructions that come with it direct you to drag and drop the flash rom onto it and then proceed normally. Problem is, when I do that I get an error: Can not open input File I ignore that, choose the correct com port and then press Talk to Calculator. But nothing happens. Any suggestions. Luis Hi Luis: Please make sure that the flash file you're trying to upload has the extension .flash If it doesn't, you'll have to rename it manually. If that doesn't work, try redownloading the file, just in case you have a corrupt version. Jason === Subject: Re: Help with FDP.exe - The flash rom file is corrupt, i.e. bad download, file system error, or HD failure (I hope it's not that one). - The flash rom file is already in use, e.g. by another instance of FDP.EXE - You don't have the necessary permissions. This is only a problem on the NT-based (NT, 2000, more to come) family of Windows. - You forgot to unzip the rom package. It must be a file with .flash extension I think. Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 Hi Brian! Well, mine was a separate metal plate, with adhesive back face. Nick === Subject: Re: HP 49G Nameplate Well it is separate, but then you stick it to the back of the calculator. It's got a special deepening (? *sigh* my english) so you can't take the plate back out easily. Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 Hurraaahh! Welcome to club (again), Jeff! Just for curiosity, if you don't mind, do you have the same enthousiasm like when you got the 41? Happy converting, Nick It would be good to have such a machine (especially for $20Aus ;-), but I doubt it (JYA was probably just teasing us :-) Hahahaha! But he said MacOS X, which is... well... different from MacOS classic ;-) Bhuvanesh. === Subject: Re: learning forum for hp49g users Oooh thank you! (I suppose you're talking about me? :-/ ) That was sort of a quick start - I bought my HP49 about a year ago. hpcalc.org already has a commenting & rating system for the programs. I like your idea, but maybe someone should just dig through the more recent questions and update the FAQ (that sounds like a hell lot of work, please excuse the expression). If the more regular posters can get together (through mail?) I can offer some voluntary help - whatever can be done besides school and all =) Agreed. Sometimes I dig for a function I know I've seen it before... Still I'd rather put 48{G,GX,G+} specific things in there. The 48 is in many things very similiar to the 49 (I've never used a 48 though. Never believe everything you hear or anything you say.). Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 It reminds me of my good old Conab: The Smallest Config Aborter! :) http://www.hpcalc.org/hp48/programming/libs/conab.zip Take a look at the accompanying documentation, especially the section REMARKS ABOUT THE LIBRARY ROMID (which was 14 not 4!) Bye. Jordi Hidalgo johil@tv3mail.com - There's something strange about those formulas you quoted: -------------------- (taken from the link) -------------------- Let X(i) be an array of i=1...N interval midpoints, and W(i) a corresponding array of weights (frequencies). Use arrays SW(i), SX(i), and M(i) to hold the calculations. Set SW(0)=SX(0)=M(0)=0, and then repeat SW(i+1)=SW(i)+W(i) d=[X(i+1)-M(i)]W(i) <<<<<<<<<<<<<<<< M(i+1)=M(i)+d/SW(i+1) SX(i+1)=SX(i)+d[X(i+1)-M(i+1)] SX(N) will contain N times the variance (SD squared), and M(N) the mean. You don't actually have to use arrays for SW,SX, and M: scalars will do if you update them. -------------------- (end) -------------------- d uses X(i+1), and this implies i = 0...N-1, but this conflicts with d using W(i), since W(0) is out of range. a basic loop-formula is (same starting values, same vars) i = 1...N SW += W(i) // obvious d = W(i)*X(i) SX += d*X(i) M += d at any stage, mean = M/SW, variance = (SX/SW - (M/SW)^2 ), where variance is (how do you call that in English?) uncorrected (?), that is divided by N and not by N-1 (this allows to get 0 when N=1) -- The set of solutions is never empty. Two solutions together form a new problem. -- Mycroft Holmes I know, even the early R rev calcs have molded keys. And, my first SX (Rev A, Serial #3003A.....) from 1990 is still alive. You don't want to compare apples with peaches, do you;-) Raymond KC7CC@hotmail.com 49 will not last , painted on . Apples and peaches have nothing to do with the many here who have compalined their 49G key marking is wearing off . H.P. quality sucks . Raymond Hellstern < I don't know why everyone has so many problems...I have an hp comp, hp monitor, hp calc...NONE of them have ANY problems whatsoever... === Subject: Re: New HP48G+ And I have a non-HP comp, monitor etc and they all DO have problems. Except the calc - which is HP. =) Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 KC7CC@hotmail.com Good keys ? They are all painted on . I read here INdonesian GX has painted keys ! I have purchased my last HP . back a Does mentioned. I need a chemistry program for my hp49. Which periodic table program would you suggest? Eric. Tabla Periodica -- This message was written entirely with recycled electrons Pivo === Subject: Re: periodic table program. I'm using Labo. Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 You could always try them all and see which one you like best. Using can decide yourself. :-) TW Palmtops, March 1995, page 7: 82243-60001 HP 17BII, 10B, 20S, 32SII, 42S replacement vinyl case. For ordering extra owner's manuals or soft replacement cases call 800-227-8164. Maybe you prefer an elegant leather case: HP-17B and HP-27S Cases Black (HP 92169K) Brown (HP 92169L) Burgundy (HP 92169M) Similar cases for the HP-12C and IR printers are (were?) available too. ------------------------- By the way, John, have you given up trying to insert a null char in the command line? You were very near: Hmm, *very* near... Another hint? Look at this: http://www.hpcc.org/datafile.html#V20N4 What's on page 17? ;-) Bye. Jordi Hidalgo johil@tv3mail.com - Hello guys ! Can anybody explain the price difference between the HP40 and the HP49 ?? For me the hardware is almost the same... Thx in advance and keep up supporting HPs greats Calcs ! PS: i own a hp48GX and very satisfied of it ! -- Nobody is perfect I am nobody Therefor I am perfect I think the main difference in price is because of memory. The HP40G has 250? KB of memory whereas the 49G has 1.5 MB. However, there is another thing to consider. It seems to me that since the 40G was released AFTER the 49G, that it is possible that its CAS is better integrated into the OS. Of course, I have never even used a 40G, so I cannot be sure of that. Hope this helps. Jason The 40G has 234 Kb of usable memory. The CAS is integrated into the system very effectively and it makes it very easy to use. Another consideration for the price difference is for the intended target market (high school/early college). A cheaper calculator with a high mathematical ability is more likely to be bought than a calculator with about the same abilities but costs higher than anything else (aka:49G). How many high school students are really flowing with cash anyway? ;-) In terms of power to price ratio, the 40G wins hands down compared to any other calculator making it a great value. TW ~The enemy's gate is down. === Subject: Re: price difference between the HP40 and 49 It's the almost that make the whole difference. Makes a big difference on the price Jean-Yves KC7CC@hotmail.com The proper way is to send no CRC nor checksum They don't work ! Send each small packet twice to test the connection . If error rate is good , increase packet size . +++++++++++++++++++++++++++++++++++++++ escaping. === Subject: Re: Protocol What's wrong with checksums? I never got a corrupt file transfer so far. And obviously your error checking would halve the transfer rate. Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 === Subject: Re: Protocol And that's at best. If a packet is incorrect, you would have to send each packet twice again... Jean-Yves KC7CC@hotmail.com There is no way to correct compressed data w/. a tiny addition ( checksum/CRC) at the end of a packet . It can not contain enough info to correct a compressed packet . Best way to correct it is to send small packets twice and check them for errors . If they are good , send larger packets twice . Then only once . At this point you can calc a small fast checksum , but if it is bad , you must go back to the 1st packet sent once . Of course at this point a human is watching and will realize hardware must be fixed. Keep data on nodes that send packets so you can start at last good rate . BTW Another arguementative data sending method is RLL . RLL ( vigorously denied ) is now used in almost all applications . It is very adjustable . Systematic error correcting codes do exactly this. One famous example is the Reed-Solomon block code, which is used among other things for error correction on data CDs. Have a look at http://people.qualcomm.com/karn/code/fec/ for source code in C to do this encoding/decoding. When using the so called (255,223) code, a block of length 223 bytes is appended with 32 bytes of error correction data to form a 255 byte block. The decoding algorithm can recover the original 223 bytes even if 16 of the 255 bytes have been corrupted. If you are satisfied with being able to correct at most 2 bytes per packet, you can use the (255,251) code, which contains 251 data bytes and a 4 byte checksum. -- Peter 205sterlund peter.osterlund@mailbox.swipnet.se Sk232ndalsv212gen 35 http://home1.swipnet.se/~w-15919 S-128 66 Sk232ndal +46 8 942647 Sweden KC7CC@hotmail.com This is not rocket science . It's just common sense . A 1024 byte compressed packet can not , with a 32 bit CRC or checksum or whatever ,correct the data in that packet . Not REED SOLOMON not anything . Nothing can correct it . You are trying to put 23 lbs of s in a 1 lb box . There simply is not enough bits in the CRC/checksum/parity to correct it . ------------- -------------- Uh DUH , only 16 ! When sending twice you can correct all 255 bytes , and no C routine needed to do any math ! And you don't need to send the whole 256 bytes , you can send 1 byte twice , send 2nd byte twice .. at the 2nd or 3rd or 4th you don't need to double send anymore 'cause confidense is so high . I'm talking common sense , something that is not so common in computer nerds/C programmers . === Subject: Re: Protocol True. Who said we want to transfer ?? When sending it twice and they don't match you don't even know which one is right. So (as JYA already said, and you didn't believe him) you have to resend the whole 256 bytes. Twice. There's also a slight (and I mean *slight*) chance that both blocks are - though wrong - the same. OTOH, with checksum routines, you only have to append a few (2, 4, 8, whatever) bytes to each block sent. This can save nearly all of the double-send-overhead. The chance that a checksum is matches, even though the data is wrong, is a bit bigger, but you can usually (and we're if an error is detected, you have to resend the whole block including checksum. I hope that was clear enough for somebody speaking what you call common sense. Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 === Subject: Re: Protocol No I'm not playing with fire ;) That should've read The chance that a checksum matches, i.e. no is. Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 So, are you suggesting that compact disks are not really working and the whole world is just imagining that they do work? Well, no, if you are unlucky and get the same error at the same place in your two packets, you will not even be able to detect two errors. It is becoming evident that you are not intelligent enough to have a meaningful conversation with. -- Peter 205sterlund peter.osterlund@mailbox.swipnet.se Sk232ndalsv212gen 35 http://home1.swipnet.se/~w-15919 S-128 66 Sk232ndal +46 8 942647 Sweden === Subject: Re: Protocol OMG. How can people be *so* clever ;) I'll have to learn those tricks. It's simply amazing. Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 Welcome to the Matrix...:-) === Subject: Re: Protocol -0500) TVD spewed: As to error *correction* vs. error *detection*, it has already been pointed out that the CRC is intended for the latter purpose only; compared to transmitting every packet twice, appending the error detection code is much shorter than a complete re-transmission, hence saves bandwidth, unless the error rate is exceedingly high. When error *correction* codes are used, all that they need to correct is what is being transmitted, not what the transmission represents, so it doesn't matter whether the transmission represents compressed data or not. http://www.newsfeeds.com - The #1 Newsgroup Service in the World! -----== Over 80,000 Newsgroups - 16 Different Servers! =----- === Subject: Re: Protocol ______TVD______ said: That's pretty much the truth, I think. Common sense is really knowing the stuff everybody knows, or being able to figure out the stuff anybody ought to be able to figure out. Some nerd/programmer types do have difficulty with this. So did Einstein, from what I've read. I think it has something to do with the ability to concentrate and focus and the need to find out what really is and the unwillingness to just know what everybody knows or to just accept appearance as reality. It's kind of the need to dig below the surface. That's not really a common trait. And it probably interferes to some extent with common sense. Are you telling us that you rely on common sense? Barry http://www.newsfeeds.com - The #1 Newsgroup Service in the World! Check out our new Unlimited Server. No Download or Time Limits! -----== Over 80,000 Newsgroups - 19 Different Servers! ==----- === Subject: Re: Protocol You're so full of it if I may say so. Checksum and/or CRC in a communication protocol such as Kermit or Xmodem, have never been used to correct the data, but just to make sure the data was correctly sent. The probability for a packet to be corrupted with one inccorect byte only and still have the same CRC is null. And that's the whole point of a CRC. The data will be corrected in these protocols by re-sending the packet until the CRC comparison says: the packet is correct. It's that simple. So why are you talking about error correction and so on... Maybe just to show off, in which case you've failed anyway Jean-Yves Thomas Rast KC7CC@hotmail.com Mute point since any system can do this . You should think before posting . Thomas Rast OTOH, with checksum routines, you only have to append a few (2, 4, 8, whatever) bytes to each block sent. This can save nearly all of the double-send-overhead. The chance that a checksum is matches, even though the data is wrong, is a bit bigger, but you can usually (and we're if an error is detected, you have to resend the whole block including checksum. KC7CC Those few bytes ( checksum) cost more than sending twice at start of session to get stats' , so that double sending is no longer needed . You will checksum on all packets , i will send twice only to get performance stats . The way it works is you push the envelope , then back off till a wide margin in safety is achieved . Then only occasional checking . You must a C++ programmer ! Argue argue argue , no room for simple improvements , after all it's your paycheck you are protecting ( got to keep it complicated , and low performance to make a life long job of it ) KC7CC: John H Meyers As to error *correction* vs. error *detection*, KC7CC Unfortunately it does not get the job done ( saves bandwidth ) !! You must be a govt official , focus on anything but get job done Mine works , yours just detects . John H Meyers KC7CC Brilliant ! How profound ! ..... not what it represents IDIOT ! Non compressed data is cake ! 'cause you can use the patterns to more easily CORRECT data . DUH ! Or have you forgotten what the objective is !!! KC7CC@hotmail.com Tutorial : C++ programmers ,computer nerds and chess players and math types ,& women are left brained . Meaning they can memorize so easily those things that are arbitrary, tiny and not connected to anything . Ex: 36 char strings that must be typed in to make DOS smoke . C programmers w/ inference engines . What genius !! These types lack analytical ability . Any slight ability is actually just a recital of something they memorized . This must of course be popular so as to be the truth . Left brained must be strongly guided . R' brained are the analytical types . These type have common sense . Com' sense here means as the whole world goes off into computer outer space , the R' brained zero in on truth , simplify it and get the honor of having common sense . Common sense is not recital of memorized buzz words , it's being objective and explaning this to the left brained masses . Why R' brain ? Because here is where true analysis is done . It is ussually thicker in males and allways less developed in females . This explains why there has never been a world class female scientist . Medical science has proven NO female has the thicker R side developement that the top men have . Einstein had trouble with his proffessors . All thinkers have trouble with their tutors ! Tutors are to recite to you , they call it teaching . Now allow the govt to own 95% of all schools and you really sheepskin w/o much help from your teacher ! Univ' T.A's ( Teachers aids ) are the real teachers ! They can actually explain a concept ! Govt teacher = oxymoron Barry KC7CC@hotmail.com You need to school yourself in English . I stated a way to better CORRECT data . Now you say same thing . The data will be corrected in these protocols by re-sending the packet until the CRC comparison says: the packet is correct. It's that simple. If you can not see how my method of initially sending twice (to get stats so you know how far to back down in sending speed) to obtain 100% integrity , you need help !! My system is faster than XMODEM et al . Jean-Yves Avenard <... was until Then just program your system and post it so that we can test it. Who knows? Perhaps we will use yours instead of XMODEM. Nick. Well...I've had my HP49G for about 2 weeks and now it's useless...I let some scratch resistant fluid seep into the insides of the calculator and now it's dead...now I have no calculator to use in Algebra 2 or Chemistry...I'm in a very bad situation...If anyone of you know where I can get another one for very cheap, please tell me...my dad won't buy me another new one... -- --Matt Bryant If at first you don't succeed, try, try again. Then quit. No use being a damn fool about it. - W.C. Fields (1880 - 1946) Never try to teach a pig to sing. It wastes your time and annoys the pig. - Proverb Were you trying to coat the scrren to increase scratch resistance? Hopefully it will work when dry. Exactly what part did you let it seep into? The keyboard? around the screen? TW ~The enemy's gate is down. it got down to the F keys and seeped down between the keys and the actual calc...so its down in the electronics... Hopefully it some it's a for hello Matt, you'll probably be able to fix it. 1- remove the blue plastic rivets at the battery compartment. for this, use a drill and drill the heads of the rivets (they'll come easy). 2- pry apart the calc in half (with moderate effort). use your hands. 3- untwist the four metal brackets that hold the main board. 4- separate the board (motherboard) 5- now you'll have access to the main keyboard and screen connectors. 6- use a cute-tip (a cotton stick) and some alcohol to clean the connectors and any fluid left. 7- put the calculator together GOOD LUCK says... === Subject: Re: pushed my luck I have an old 48S that just sits and gathers dust. If you cant find another solution let me know and I'll let you have it. It's not a 49G but it's still a pretty powerful calculator. But first see if you can fix it. I live in the country and I avoid town like the plague but I'll have to go to town to ship it. Barry useless...I let some and now it's Chemistry...I'm in a another one for one... use being the pig. http://www.newsfeeds.com - The #1 Newsgroup Service in the World! Check out our new Unlimited Server. No Download or Time Limits! -----== Over 80,000 Newsgroups - 19 Different Servers! ==----- === Subject: Re: pushed my luck What was the serial number of your HP49G, because it may have been useless anyway to put any scratch resistant liquid on it) ? In anyway, I don't see why it would have a bad effect on your calculator once it's dried, should still work fine if the quantity is not too important Jean-Yves some it's a pig. It was an ID. He just got excieted with his new calc and wanted to improve it a bit. We had a long disussion and he made an appointment with some wax remover. Anyway, the keys unstuck themselve, and everything seems to work now that it's dried. The screen is fine now with the wax off and only a little remains on the keys. TW ~The enemy's gate is down. here : http://perso.atsat.com/pigallio/Hp48/Lil_hp48.htm Have fun. RPIT or IT: stack,so, in system I want to do the same but I cant get the string on the stack after put the real number 0 , the code is following: In sys :: %0= $ Hello RPIT ; I dont get anything when I use the debbuger the real number 0 is comparade with the flag and get FALSE. What is my mistake? Here is a small prog that does that. :: %0= IT $ Hello ; If you are using a 49, your problem may be that it is an integer, not a real number. If the 0 has a period after it, then it is a real number. If it doesn't, it is an integer. Change to approximate to get real numbers. ~The enemy's gate is down. just wondering, but just say after graphing a function in power plot 1.34,5,6, you want to return to the normal stack, so i press the 'exit' button to go back to the stack. when it is going back to the normal stack, does anyone see the menu bar redrawing itself? what i mean by redrawing is there will be a few dots and specks around the menubar area, and then the taskbar will finish forming quickly after that. it seems to have caused my calculator to crash on an older version of power plot. i just upgraded to 1.36... does the same redrawing of the menubar, but so far no crashes. this program is cool.... but i dunno about the crashing. just wondering if its my calculator or if other people have this problem. The artifacts you see where the menu bar is after pressing EXIT or the ON/CANCEL button is just the screen redrawing itself. When a reclaimdisp command is done, it redraws the screen to the correct (normal) size and has to redraw the menu, which causes this effect. If you try the short sysrpl program you will see the effect: :: TURNMENUOFF RECLAIMDISP %1 dowait TURNMENUON ; If it bothers people (or causes them to worry), I can just put a TURNMENUON before the RECLAIMDISP and you won't see the effect. As far as the crash you mentioned goes, it was probably some problem not related to the glitches you see in the screen. If it does crash, then you should definitely email me so I can fix it before it happens to someone else. Hopefully you won't have to though. =) -- Aaron The 49G User's Guides are not integrated in the redesigned HP web site. Maybe they will add them later. At the moment, at least the old pages are still accessible: User's guide: http://www.hp.com/calculators/techsupport/graphing/49g_userguide/english_ug. html Advanced user's guide: http://www.hp.com/calculators/techsupport/graphing/49g_userguide/english_aug .html Beta ROM: http://www.hp.com/calculators/graphing/rom/beta.html At least, there is a link to commercial ROM now, but hard to find: http://www.hp.com/cgi-bin/cposupport/swindex.pl?p=hp49ggraph24252&l=en&c=BIO S where ROM 1.18 is downloadable. By the way, for now I resist commenting the new web design ;-) ------------------------------------------------------------------------- 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. ------------------------------------------------------------------------- === Subject: Re: redesigned hp.com web site I see. What threw me was that I tried: http://www.hp.com/calculators/graphing/49g_tech.html which redirected me to: =en&h_cc=us&h_tool=prodhomes which left me rather lost. I'd like it better if http://www.hp.com/cposupport/docindex/hp49ggraph24252_setup.html didn't say no online manuals. http://www.hp.com/calculators/techsupport/graphing/49g_userguide/english_ug. html http://www.hp.com/calculators/techsupport/graphing/49g_userguide/english_aug .html At the moment, those get me file not found errors. I suppose that it would be overly optimistic to hope that this means that updated documentation will be posted. http://www.hp.com/cgi-bin/cposupport/swindex.pl?p=hp49ggraph24252&l=en&c=BIO S BIOS? Well, I suppose that that sort of makes sense, if I think of the 49G as a computer. James What new site are you referring to? The HP calculator site looks exactly the same to me. Jason === Subject: Re: redesigned hp.com web site exactly Hi Jason, Not as different as I had thought, but try to navigate from http://www.hp.com/calculators/graphing/49g_info.html to the 49G Guides. James How do you plot the regression line from a scatter plot. I have tried the STATL Button, but it seems to only do linear lines. Is there a way to have Scott Hellewell Lines are, by nature, linear. There are regression models possible other than the linear regression model. When your regression model is one of those, the STATL button will plot as other than a straight line. In the statistics choose box{ STAT or right-shift 5} , choose 3.Fit Data, then under model press the choose key. -- --Matt Bryant If at first you don't succeed, try, try again. Then quit. No use being a damn fool about it. - W.C. Fields (1880 - 1946) Never try to teach a pig to sing. It wastes your time and annoys the pig. - Proverb ~The enemy's gate is down. === Subject: Re: resolution You can actually change the resolution for plotting. With a theorical maximum of 2048 by 2048. Simply create a graphic of let say GROB 256 256, save it into PICT with PICT STO Then do your drawing as usually. Jean-Yves isn't visible, use pig. Alright, this may sound very funny, but will this speed up graphing? I really was asking how to speed up graphing (especially of truth plots which just takes too long compared to a TI)...is there anyway to change settings to speed *graphing* up...in terms of resolution or the sort? PICT being Hi again, Matt! You can speed up the plot by inputing the step value in the input field Step: in the Plot Window screen. If you set this value to be greater, then the plot is faster but but not so rich in details as it could be. You can also use the command RES for this purpose. It must be feeded with one binary integer or a real. Example: If you want to plot every, say, every second pixel, then type #2d RES. If you want to plot every .5 coordinate units, then type .5 RES. Hope this is what you wanted. Nick Try www.grafikrechner.com ; HP 49G for 329,- DM incl. PC-link. Detlef WiSCy99 v4.26 for (Windows'9x/NT/2000) is the complete and-to-use scientific calculator http://www.simtel.net/pub/pd/17596.shtml. The results of calculation can be visualization, printing as graphic, as text or saving to disk. Unit Converter is pre-configured to convert over 500 units in 30 categories and editor for custom units conversion.Periodic table of the elements utility provides basic and extended properties of the elements. Complex and MATRIX operations is available. - Arithmetic and logical operators and functions - Common functions such as exp, ln, sqrt, sqr, bnml etc. - Common, trigonometric, hyperbolic complex functions - Trigonometric, Hyperbolic functions - Numerical Integration - Equations can be solved - Special functions (Gamma, Bessel's, Si, Ci, erf, erfc, Fresnel's) - Statistic functions (Average, Standard deviation, Sum, Random, Gauss random, statistical variance, etc ) - FOR-type loop - if (...) then (...) else (...) function - Tape of results - Assistant and debug: error position fixed - Plot f(X), Contour Plot f(X,Y), Color Shading f(X,Y), real 3D-Plot f(X,Y), Derivative, Fit. - Print results, graphics and print preview - Save graphics to BMP, WMF, EMF formats - Matrix Operations(A+B=C, A-B=C, A*B=C, inverse(A)=C, Power(A,n)=C, det |A|=C[1.1], Solve A(X)=C) - Decimal, Hexadecimal and Binary bases - Fixed point, Scientific, Engineering and Sexagesimal notations - Radian and Degree modes for trigonometric functions - Precision: 10-12 significant digits. - Range: _(3.4E-4392 to 1.1E+4392) - 10 pre defined variables, user define variables - User define functions - 30 user defined constants (up to 16000), search and edit file with constants. - Stack for expressions (up to 16000) - Stack for results (up to 16000) - Unit Converter - Custom unit converter - Evaluate expressions from file - Simple tape calculator - Periodic table of the elements Special requirements: None. Changes: Added Periodic table of the elements More than 400 units in 20 categories Igor Evsikov ievsikov@flowpath.com Very nice thing. Now, if it had a black plastic shell, and tall plastic keys, and it could be cut away from the screen and taken in my pocket ... ;-) Good. Good. Good. Especially that hyperbolics and arc-hyperbolics are on the keys. We already had this, or am I missing something? Works fast. But why no symbolics? (Future improvement?) Again, only numeric roots. No symbolic solutions. Each one on a separate key. Nice! Only for univariate arrays. Good for programming, implementing summations and the like. Good. Well, not the real stack, but easy to copying and pasting previous results in the input line. But the calc is algebraic. And help text is shown about the error in the command line. Very good. It works fast and you can modify many things about how the graph should look like. Nice! That's good! I repeat myself saying that only numeric matrices are allowed. Good. Good. Good. Good. Wow! OK. But again, only numerics. Nice! But without units. This is no real stack. But you can easily copy/paste from the history to your input line. Good. Type a value for some unit and see how it is converted to all possible units with same dimensions at once. Good for storing previous work. OK. Nice! Very nice!!! The program somehow reminds me of IGOR. (No, not you Igor ;-), I mean IGOR of the software company WaveMetrics.) Nick We'll I've written another comparison, this time of enty level scientific calculators (because most everyone needs a cheap one at some point in time ;-). I compared the HP-30S with two almost equivalent TI models. I posted it to hpcalc.org but it hasn't been updated, so if anyone wants to get one right now. Email me for it at timwessman@yahoo.com or timwessman@calc.org. Oh, and there is a programming contest coming up shortly at calc.org (within a week or so). Last time the challenge was to move text around like a bouncing ball. There were awards for smallest size, most features, nicest looking, etc. There are points awarded for winning the different areas and they can be added up to get stuff (like a new 49 or something). Efficency...check. Nice looking...check. Most Features...check. Hey! I'm envisioning a sweep here by the HP crowd ;-) TW ~The enemy's gate is down. Does anyone know of a reputable source of secondhand HP calculators? Jason You mean, you would like to buy a secondhand 42S? Sometime you see several 42S' listed on eBay. Sometime you don't see any. Anyway, a mint condition 42S seems to cost well over US$100 nowadays. You may want to check that out by searching Completed Items on eBay. Yoichi I was wondering if it is possible to send an command RC6, codeset 4, key 235#D command with the HP48 (using Remote 3.4). See this for the why: http://www.home.zonnet.nl/jeroenverhaar/en/vervolgindex/dvdregiovrij/dvdregi ovrijall.htm Ok, I got my program to compile correctly and it appears to be working great, but I'm having trouble getting it to send data. I thought that maybe the XRECV, XSEND commands would work in the same way as on the 49, but they don't seem to be working. Oh, and one other question. How would I go about storing and recalling from varibles Like L0-9, c0-9 and M0-9. There are xC0-9 commands etc. What input would I use for that or is there another way to do it? TW ~The enemy's gate is down. how do i solve inequalities on my calc? every time i try to use the solver it gives me some bogus error. and also i would like to know how i solve solve normal equations and inequalities with absolute value. thank you for all your help. Hi Tristan! Would you please tell a little more about your inequation (what it looks like), flag settings, what error you get and so on? Some examples for the syntax: Example for solving an inequation. 'X^2-1<=3' 'X' SOLVE Result: But if the inequation contains other symbolic variables than the variable to solve for, the HP49G says: SOLVE Error: Parameters not allowed That means that 'X^2-a<=3' 'X' SOLVE will not work. Example for solving an equation: 'SIN(Y)^2-3*SIN(Y)=3' 'Y' SOLVE Example for solving an equation containg absolute values: '|COS(X)|=SQRT(2)' 'X' SOLVE Example for solving an inequation containg absolute values: 'X' SOLVE Did that help a little? Nick. implement for *my* inequality solver. :) Bhuvanesh. I get the error message SOLVE Error: Non unary operator. Using = instead of <= works fine. Are the inequalities a feature of the 1.19 beta-roms (I use 1.18) or is there a special flag that needs to be set? I don't know about flags or antyhing, but you should upgrade to 1.19-5 (it works for me by the way). it is well proven to be stable and much better by now after a year of use ;-) TW ~The enemy's gate is down. Hmm, can you tell us your inequation? I remember getting this error in the past and I don`t think that it is because of flag settings. Unfortunately I don`t remember exactly what I did, so if you would post the inequation that you are trying to solve, then I could tell you more. Nick I used the inequation you were using as an example ('X^2-1<=3'). With the only (obvious?) difference that I don't type <= but Left (blue) shift and X. :) I tried it in algebraic mode also with SOLVE(X^2-1<=3, 'X'). I also tried using SQ(X) instead of X^2 to no avail. Strange! When I do: 'X' SOLVE I get: I think as Tim said before that it could be because your ROM is too old. I can`t see any other reason. Which ROM do you have, anyway? Nick. P.S. Does anybody have another idea about the behavior of Jesper`s HP49G? Today I upgraded from 1.18 to 1.19-5 and now the inequations are solved correctly. Well, sometimes we can find the reason why something is going wrong. ;-) Nick. I seem to remember somwhere in my mind a remember a thread dealing with sorting paired items in a list that had several small and fast sysRPL/assembly programs I wonder if I imagined it. If not, can someone help me. What I mean by sorting paired items in a lists is that if a list had { ID QWE {1 2 3} ID ASD {4 5 6} ID HJK (7 8 9}} The first items, ID's in this case, would be sorted alpabetically but bringing along the next item with it so the final result would be { ID ASD TW ~The enemy's gate is down. Hi Tim! In User RPL we could use something like: << {} 1 list SIZE FOR n list n DUP 1 + SUB 2 STEP SORT Perhaps you can use the idea for writing a System RPL algorithm ? Nick === Subject: Re: Sorting paired items in list -0500) Curiously enough, the built-in user SORT command, when the items in the list to be sorted are themselves lists, sorts those items according to the first element in each list, so it already has the capability you seek, although it could be made faster. Note that in your example, you neglected to enclose each whole item group in its own list; then it would be sortable by [x]SORT. Werner Huysegoms has written a fast replacement of his own, however: http://www.hpcalc.org/viewzip.php?id=2828&file=Lsort.txt [description] http://www.hpcalc.org/hp48/utils/misc/lsortv04.zip [download] rs=huysegoms&filter=0 uthors=huysegoms Perfection is achieved, not when there is nothing left to add, but when there is nothing left to remove. - Antoine de Saint-Exupery Music, in performance, is a type of sculpture. The air in the performance is sculpted into something. - Frank Zappa Architecture is frozen music. - Goethe, Schelling Donald Swann's music has often been compared with defrosted architecture. - Michael Flanders http://members.tripod.com/~TimothyPlatypus/FaS/lyricsframeset.html http://members.tripod.com/~TimothyPlatypus/FaS/hatintro.html http://members.tripod.com/~TimothyPlatypus/FaS/Cover.jpg ----------------------------------------------------------- http://www.newsfeeds.com - The #1 Newsgroup Service in the World! -----== Over 80,000 Newsgroups - 16 Different Servers! =----- To my port 2 have entered a great number of unknown variables that can not be purged by the file manager. They don't let me have access to the port 2 using Right Shift lib; if I do so it shows me Insufficient Memory. Also in case of any warmstart or ON-C it shows me a message of Warning:Invalid Card Data. I want to erase all of my ports especially the port 2, please tell me how!!!!. My calculator has an error it can not connect to the computer in any way that's why I can't upgrade my ROM Version :(. The variables have extremely long names and with weird characters, they occupy 2 bites. suspect that they got there when I tryed to pass form another hp 49g the JAZZ,extable, UFL or programs in SysRPL; I didn't see when such thing happened they just appeared suddenly. I took the bateries out for two days in order to see if the files in memory (port 2) would erase, but nothing happened. Hello Stig, do a PINIT 2 and look what happens. This should help to solve your problem. Port 2 is the flash memory so it does not help to take out batteries ;-) Hope this helped. Matthias -- MBunte@gmx.net Hi Stig try PINIT command to remove Invalid Card Data error message. For purging from 2nd port try blue shift + files, then select port 2:FLASH Finally - did you test hard reset using paper clip? HTH John Hi and don't panik! Use the command PINIT. I had the same problem as you (to a lesser extend though, as I only had only two such variables on port 2, which contained garbage). PINIT has worked for me. After issuing this command, these variables were gone. So perhaps it works also for you. If it doesn't work, then perhaps you could use the following method: Type 2 PVARS. This returns a list of the variables in port 2 on level 2 and the free memory on port 2 on level 1. DROP the number on level necessary to get one of the garbage variables on stack level 1. Now press PURGE. Repeat to purge all these variables from port 2. Hope this works and good luck. Nick. P.S. If you want to purge everything from port 2, you could issue the cammand purge using the list of variables that is returned from the command PVARS. But then you must re-install everything. P.P.S. You took the batts out for 2 days and the vars are still there, you say? How is this possible? I believe that you did, but I just can't imagine how the HP49G preserved power without batteries for 2 days. RCOBO, any ideas? Try this : PINIT It should solve your problem. Bye, Yoann. === Subject: Re: S.O.S: Help me with port 2!!! (PLEASE!!! everybody) There's also another way: 1. Hit ON-D, hold down + and ENTER, then hit ON. This should bring you to the No system menu. 2. Hit 4, Terminal mode. Your screen will go blank. 3. Use another calc (ROMUPLOAD) or a PC (the flash upload program) to send a special 4-bytes code. The calculator with the bad memory will show the Download System menu. 4. Choose Erase bank from this menu. Banks 8-15 are the user flashrom banks; 0 is also one of those in releases before 1.19-5. Erase all user banks. 5. Reboot. HTH Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 Hi it«s me again help to my calc, i try with PINIT and just disapeare the Invalid Card Data message, the vars are still there, i«m really in panic... are still there, i can copied to HOME but i never could ERASE it. They have names like e#216A.3tS226...216 and inside just have an External. this external dissemble is: PTR 8C050. By the way, i can«t store nothing in 2port. PLEASE HELP ME!!! Sounds pretty bad. Have you tried Mr. Rast's suggestion with the No system menu? Have you tried re-flashing the ROM from a computer? There's just no way that this could be a no solution situation, since the boot code isn't corruptable? [Hopefully] Perhaps you could also try to e-mail HP. Or, if absolutely nothing works, you could send it to HP, and they'll probably send you a new one. (If it's still under warranty). Just a suggestion, try P0ERASE from hpcalc.org. Jason Hi everybody I'm very pleased, I tried many times with PINIT and it worked, you have saved my calculator, I was very scared, now I'm so happy :) , you're my heroes, I wish to thank M Bunte, John, Nick, Thomas and Yoann for helping me again. Stig R. Nielsen P.S.S. Nick, i agree with you, I can't understand why flash ROM can work without Batteries. Flash rom is non-volatile. In other words, it doesn't need battery power in order to maintain its contents. That's what it's there for. James === Subject: Re: S.O.S: Help me with port 2!!! (PLEASE!!! everybody) Hello Port 2 is the Flash memory, so removing the battery will not help. Simply run the command PINIT, and it will fix all your problems in port 2 (as well as port1 and port 0 if you had any) Jean-Yves === Subject: Re: S.O.S: Help me with port 2!!! (PLEASE!!! everybody) -0500) I once subscribed to a trade newspaper called Electronic News. One week they covered some convention where somebody talked about potential devices of what was then the future, and when this speaker spoke such weird and contradictory phrases as electrically alterable read-only memory (EAROM), well, the reporter said, in effect, that nobody quite understood what this nut could possibly have been saying :) I believe that the speaker's name was Ovshinsky, about whom more can be found here: http://web.mit.edu/invent/www/inventorsI-Q/ovshinsky.html http://www.time.com/time/reports/environment/heroes/heroesgallery/0,2967,ovs hinsky,00.html Stan Ovshinsky seems to have popped up more often than Leonard Zelig :) And all this without benefit of a college dis-education! -[]- http://www.newsfeeds.com - The #1 Newsgroup Service in the World! -----== Over 80,000 Newsgroups - 16 Different Servers! =----- === Subject: Re: S.O.S: Help me with port 2!!! (PLEASE!!! everybody) I remember some post by JYA, with essentially the following point: Bank 0 was designed to replace the boot code in case there is any major bug. So if your bank 0 is ever seriously corrupted, it might refuse to boot. .bigpond.net.au Thomas -- Thomas Rast t.rast@iname.com ICQ# 103670088 i think a better bet would be the HP computer museum (www.hpmuseum.com) I remember seeing in the discussion forum a PIC-based module replacement far cheaper and smaller. **vp Is it me or has spam begun to show up here more and more frequently? I have noticed it on all other groups I subscribe to as well. Is there any recourse for it? Is it possible to filter it out with the use of subject indicators like cs48, hp48, hp49, etc? -- -Al Arduengo ------------ M$: Our software of tommorrow will make sure that on your hardware of tomorrow everything will work at the usual speed. -- -Al Arduengo ------------ What you do not want done to yourself, do not do to others. -- Confucius Noticed it to but only thru one of the ISP's I use. The Spams not appearing in the newsgroup feeds I get at home. It's not just in this group either. Hello folks, I have and love an HP-41CV calculator that I want to take out of mothballs. I am missing the rectangular plastic battery holder which holds 4 N batteries. Does anyone have a suggestion of how I can Bob I'm sure this is easy but I can't seem to find it in the manual. How do I enter two symbolic expressions and multiply them giving a symbolic answer? In other words, I'd like to enter 'x + 2' and '2x - 3' and get back '2x^2 + x - 6'. Can this be done on the 48? David Betz dbetz@xlisper.mv.com