A37 ==== No News, Good News. Normally users tend to post if they find that something is wrong. When it's all right, there's non need to post something... By the way, I did'n know that the flag part was finished, and I have find it on hpcalc.org just today (I check every day hpcalc.org to see if there's something new and interesting). I will see the document as soon as possible and I will try to say there anymore?) My problem now is the time, but more documents I'll find about HP49, more happy I'll become. Ciao, Marco /-----/ ==== (SNIP) Yes, I know... but cause I'm not an HP49 guru, I'd like to hear something from other users and, if I'm particulary lucky, from the gurus themselves! Although I'm not writing a book on SysRPL or Assembler or on any other particularly complex aspect of the calc, I think that the pointofview of a really experienced user of HP49G can be the plus for a good documentation, especially if you want to write something *really* complete, that can help the newbies and be useful also for experienced users. I post here the URL to download the document, because lot of people - after person asking me about the doc, I prefer discussions here on topic. BTW The URL is: http://www.hpcalc.org/hp49/docs/misc/flags10.zip perch216 noi un'idea precisa della realt210 delle calcolatrici HP nel mondo, come hai appunto scritto tu in un msg. Questo NG invece 217 una vera miniera di risorseed a mio avviso 217 assurdo che qualcuno pensi solo di acquistare una HP49 senza conoscere l'inglese, come ho letto in alcuni messaggi di disperati utenti. L'inglese 217 un requisito base per chi si occupa di materie tecniche ed in generale per chi vuol uscire dall'orizzonte culturale nazionale: 217 pi235 facile (ed utile) che gli utenti italiani imparino l'inglese o che io (ma soprattutto tu, che sei davvero attivo sulla ML) insegnamo loro ad usare l'HP? Qui possiamo parlare delle stesse cose con molte pi235 persone, molte delle quali molto pi235 esperte di me; non mi ritengo sufficientemente esperto per offrire molto supporto, anche didattico, ai nuovi utenti italiani di HP4x! :o) (SNIP) As all of us! ==== (snip) ==== dgeve: Have you talk about your project with W. Rautemberg? He said in a ==== I have released my first ever contribution to the HP world. I have written a guide that is geared towards the college precalc class at my ==== (snip) EVAL when in approx mode will work, if you don't mind such things as 2*3* being simplified to 6.* Curiously, why do you want this? (snip) ==== exhibits the same behavior. It doesn't respond to hitting the ON button intially, I have to hit it many many times (30+) before I can even get a response. It then reboots, and if I turn it off and then on again quickly it works fine, however, if I let it sit for a moment, or don't hit the ON button exactly right, I have to go through the whole process again. I'm assuming I damaged it somehow. Is there anything else I should/could try? The calc is no longer under warranty (though I don't think the warranty would cover accidental damage on my part anyways) - is it still worth sending it to HP to get ==== ==== Oops, the program is in part two. ==== ==== ==== Look where they are made... Indonesian made have a black screen, Singapore one have a blue screen ==== He is a spammer. This is not one of his spam runs, but a punishment run. He shares his code (for generating random Markov text, finding open Proxies and using them to post hundreds of thousands of messages to USENET, etc.) with his and spammer king. Those who do not accept his rule are targetted for punishment. Among his favourite targets are NANAE (news.admin.net-abuse.usenet) He used just to send hundreds of thousands of harassing posts to the groups. They block him based on the path in the header. messages to various groups with the Followup-To header set to the group he is currently targetting for punishment (currently NANAE). If you reply or post an irritated followup you will be doing his in one of his targetted groups). The Followup-To header will have your message redirected to flood the group he wants to punish (which will be unable to block messages based on the path since the followups come from all over). The current message is a Mime Attachment, HTML formatted, JavaScript message to write (in a new browser window) text pretending to format your hard drive. Two weeks ago the harassing messages were obscene JPGs of Bin Laden buggering Bush. Who knows what they will be two weeks from now? Maybe just plain random Markov text. The headers from his post are: ==== ==== (SNIP) I have 1.19-6. * Cartesian Coord: doing EXPAND on a complex like '3+6*i' doesn't change anything * Polar/Spherical: doing EXPAND on the same complex returns '3*SQRT(5)*EXP(i*ATAN(2))' exactly how Nick (and me) like :o). Why with complex the CAS behaviour its not the same? More and more experience I take on the HP49G, more I think this machine is a sort of exceptions collection!! For problems which seem to be similar, the calc can give two differents solutions. I simply think it's not good symbolic EXPAND does something different from numeric EXPAND. And isn't '3+4*i' an algebraic as 'x+i*y'? ==== I disagree. It would be a significant improvement to have multitasking if you pick the right scheduling algorithm. The fact of the matter is that almost any computing device you use is *doing nothing* 90% of the time because it is waiting for the slow user. Unix type multilevel feedback queue algorithms can take advantage of this and give good responsiveness even if the system is heavily loaded. This is true if you include the stipulation that it only works for assembly programs and furthermore those assembly programs must conform to a strict api. While it is *theoretically* possible to do this with RPL programs it would require a major reworking of many parts of the OS. Anything that could cause race conditions (such as the memory management system) would have to rewritten or else the scheduler would have to check if the saved PC was in a critical section of code and block any other process that tried to access it concurrently. What I'm planning isn't. ;) ---- ------ ==== Well you certainly got my attention. :) If anyone wants to know, yes George, Pagala, and I were originally involved in a project to wright a modular/microkernel multitasking OS for the 48 (and eventually everything else since it was supposed to be modular). It was going to jump to the next level where other OS's (such as ShellOS) had left off. But, alas, the fact that the burden of writing most of the core of the OS was placed on me and the change of my ambitions really kind of killed it before it had a chance to take its first breath. Originally, I was interested in writing a totally new OS for the pure pleasure of innovation or its aesthetic value but eventually it dawned on me that this would just turn out as yet another niche system that, while elegant, would fade into oblivion because of the lack of applications and support. ( If I could name every example of this it would fill a few pages :) What I'm interested in now is improving the built in 48/49 RPL OS. Yes, although you might think the 49 OS is optimized to the limit, the core has changed very little since the 48 (even SX). There is a lot of room for improvement that could potentially result in dramatic speed increases. One example of this would be a rewrite of the garbage collector and memory management system which, contrary to popular belief, would not necessitate a total rewrite of the rest of the OS. In fact, you could maintain backward compatibility for the most part. Even more surprising is that you could accomplish this without burning a new rom chip! :D (although on the 48(GX) you would need a ram card in port 1). This is just one of the many changes that I've had in mind. I have all the information I need to implement this now although with the recent announce of the possible GPL of the 48G/49 source code it would make it that much easier. :) Now, I'm not making any promises (lest I be accused of vaporware ;) but if it advances to the point where it is presentable then we/I plan on releasing everything under GPL so that the development can advance much more rapidly. As George said if anyone is interested in helping please contact him or I. ---- ------ ==== On 4 Nov 2001 04:35:20 -0800, Lz8B@free.fr (Stephane Cocquereaumont) Indeed I have been contemplating this idea for quite some time. Contrary to popular belief, it is actually possible to *emulate* the whole calc on a card ( running on an SA-110 say ) and interface with the internal Yorke peripherals while keeping the Saturn in low power mode most of the time. Me too. ;) ==== I totally agree with you, Jonathan. It is simply untrue that the Saturn-chip is a dead end, mainly because we cannot affort wasting the huge intellecual capital invested in the 48/49. The 49 OS is by no means optimized to the limit. I know a lot of ROM examples where sagacious programming would not only ==== With the bug fixing, the HELP menu key works excellent. Now the proposed assignment on the CAT key works in just the same way as the whole CAT, but MUCH FASTER. help for the stack diagram) would work in just the same way: Only present when SDIAG really has its diagram in ==== This would be really nice and I have thought about this and even tried it. But I think this will be too slow, so that selection in the CHOOSE box would be slowed down too much. It is already slowed down by checking if Help is available. Fortunately Jean Yves implemented this in a way, that checking for Help is only done when the cursor keys do not repeat. That is why you can scroll extremely fast in the ==== HLDRAW seems to be a nice and fast program. Unfortunately, it has been made for the 49G only so far (AFAIK). Does anyone have interest in porting it to the HP-48? Or at least the Masd sources to standard SASM & RPLCOMP format? ==== Hey! :-) I know that if you simply put the DIR names in the CST-menu, then they will automatically appear as directories... But what if I want to execute several commands before entering the subdir, is it then still possible to make the menu tab look like a dir? thanx in advance, -- . Artur Meinild . . . . . : . 232lge R ud237r . . . ___________________________________________________ Eásite : 6paq.messages.to : Wásite : 6paq.go.to (I've only tested this on my 49, but I assume it's the same on the 48.) Instead of a string label, put a GROB into the menu list. E.g. { { GROB ... TEST } } The MakeDirLabel entry converts a string into a directory style GROB. The addresses are #2E1EB (49G) and #3A3EC (48G), so to use it from UserRPL you'd do the following: 48G: #3A3ECh SYSEVAL 49G: #2E1EBh SYSEVAL Like most SysRPL entries, it does not have any error protection. DO NOT feed it anything else than a string. You've been warned. Thomas -- ==== thanx Thomas! However, I've got a better idea, I want to draw the labels myself... I've tried storing a 21x8 pixel grob, and putting it in the menu list, like this: { { GROB1 COMMAND1 } { GROB2 COMMAND2 } } but it's not working, it just displays the name GROB1 in the menuline... any help? thanx again, Artur ==== Artur Meinild schreef: put the grob *itself* in the menu list, instead of a varaible name -- ==== ok... makes sense, but it unfortunately yields the stupid question of the day: How is *that* done? I really hope you're not losing your patience yet! :-) thanx in advance, Artur ==== Artur Meinild schreef: CONGRATULATIONS! You have won today's jackpot!!! :-) step-by-step mode on RPL on 1. make the grob a) MyDirLabel b) #3A3ECh SYSEVAL (on 48 only) c) leave on stack 2. put the command on the stack b) leave on stack 3. make a list 4. repeat for other keys a) goto 1 5. make one big list and there it is send click, netscape close, out log wooops, forgot to turn step-by-step and RPN off :-) -- ==== I see! :-) So when the list is finished, you can't edit it like any other menu list.. and edit them one by one! correct? I've just tested it, and this is too cool! HP RULES! :-D Thanx - Artur -- . Artur Meinild . . . . . : . 232lge R ud237r . . . ___________________________________________________ Eásite : 6paq.messages.to : Wásite : 6paq.go.to Suppose you have a variable containing your current menu. Recall it. On your stack you now have something like 1: { { NAME CMD1 } {HELLO CMD2} } obtain this: 3: { NAME CMD1 } 2: { HELLO CMD2 } 1: 2. Now do 3 ROLL (or use the interactive stack) to get the first list down 5: { HELLO CMD2 } 4: 2. 3: CMD1 2: 2. 1: NAME Here MakeDirLabel comes in: Do #3A3ECh SYSEVAL (49G: #2E1EBh SYSEVAL). The NAME turns into the label you want. Now you have to reassemble the 3: { HELLO CMD2 } 2: 2. 1: { GROB CMD1 } Note that, as Pivo pointed out, GROB is a graphical object, not 'GROB' obtain 1: { { GROB CMD1 } { HELLO CMD2 } } Store this to a variable. You can also test it by doing DUP TMENU - DUP so you don't lose all your hard work - without overwriting your CST. Phew. I hope that helps... Do you think I should consider writing ==== ==== ==== ==== In this newsgroup you'll find some postings relative to the HP-12C, from time to time. But perhaps the most relevant site is: http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/forum.cgi Here are more links: http://www.hpcc.org/links.html I recommend you the September/October issue of Datafile: http://www.hpcc.org/datafile.html#V20N5 It's especially dedicated to the HP-12C. It will surprise you! Bye. HPCC member #1046 ==== ==== for a product, one could be assured that one was buying quality. Unfortunately, HP has lost their vision. Like many companies, they are on a quest for the Holy Grail, which is profits, no matter what the cost. Although many of us have denigrated TI's products, website, etc., at least their calculator division is not in the process of disbanding and throwing there highly skilled programmers and engineers out into the street. I, for one, am more than disenchanted with HP's cavalier attitude relative to their formally valued customers. Because of this, I will most likely not purchase another HP product. For those of you who are predisposed to masochism, continue purchasing HP products and you will not be disappointed. ==== Well Said John, I would also like to take this opportunity to thank the many HP engineers who brought us excellent products like the HP-41, HP-42, HP-48. And although I do not consider the HP49 a high quality (hardware) product I enjoy using it and would like to thank the ACO team and everyone who contributed for their efforts. The great pity is that the expertise and development effort that went into the HP49 should have gone into a new machine with a new, faster CPU and modern hardware, not an outdated, slow CPU. I must also support you in that I will also probably never buy a HP product, or anything associated with HP again, unless they as a company go to great lengths to convince me that they can supply anything except a high-priced mediocre quality product. In future I will prefer to buy a mediocre quality product from one of HP's competitors at a low price. Perhaps this user group should consider applying their efforts into developing software for TI's calculators that would turn them into a decent RPN calculator. After all, I believe their CPU's are faster (I am no expert on TI so I apologise if I am wrong on this one!) and I could live with less memory on the calc. Who knows, if the guys at TI see that they are attracting a more serious user group they might actually develop something decent for us. The point I am trying to make is that I do not think anyone in this user group has any great loyalty towards HP. What I do not like is that this group will continue developing great software for HP products and HP will sit like a parasite making money selling HP-49's while someone else does the development work for free. (That is assuming they are not going to simply discontinue the HP48 and HP49 because they are'nt profitable enough) And while HP could care less about one or two people who say they will not buy HP products anymore, I am also an engineer, and as such frequently have to make decisions on the purchasing of various items of equipment, and I must admit what we have seen recently from HP in the line of test gear and such like has also failed to impress. There are several European and Japanese companies out there who are offering well designed high quality products, so one does not really miss HP. (In fact when our 15 year old HP logic state analyser eventually breaks there will be no more HP equipment in a lab where HP once reigned supreme.) Once again thanks to all the dedicated and capable people who brought us the great HP products through the years. I will continue using my HP calculators for as long as they work and hope that by then someone else will have realised that there is a need for a powerful, programmable calculator in the ==== ==== ==== I just got 49g and I can't find how to input complex n. in rpn mode, in 48 it use to be: bracket; real; space; complex; What program can I use to make a program for 49g on my PC and where can I get it. I downloaded connectivity kit from HP site but there is no communication, the instructions deal only comm. between 49 or 48. Any help would be greatly appreciated. ==== After these days of sadness, thinking about the present and future of handheld mathematical devices, I have made my own thoughts. With ACO we could dream with a wake up from HP after years, after now, we are at the same point that 3 or 4 years ago. After all, nothing has changed much (despite the new hp49, a great change). What will happen now? Ok, probably calculators as we know nowdays aren't a great tool for engineers, scientifics,.. They have great lacks like speed, screen capabilities, memory,.. So, is there any other option? Ok, handheld pc's market is growing fast and the prices are becoming lower everyday.. Perhaps good software and better math interfaces for them would make the rest. Perhaps HP has caught the idea, and that's the reason of the new keyboard developed by them: http://www.hp.com/jornada/products/560/acc_pocket_kbd.html http://discussion.brighthand.com/showthread.php?s=721b37828954367d932076cfdc 94275c&threadid=31508&highlight=hp+keyboard http://discussion.brighthand.com/attachment.php?s=721b37828954367d932076cfdc 94275c&postid=188056 http://www.cepolis.com/noticias/graficos/02102001hpkeymax.jpg And after seeing these photos.. do you remember Xpander looks? Yes, these keyboards are very similar to hp49/40 shape... Yes, I am agree that HP has made a mistake with JYA, CdB, Gerald and all the people that could make it better, but I think they have leave an open door for soft development for his PDA world. We hope to have a PocketPC version of Bernard Parisse's Xcas/Giac soon, it could be a begining... who knows the end of the story... Calculators were a great engineer tool when computer couldn't do the calculus job they do. Nowdays, calcs are beaten for laptops and handhelds in capabilities, is a question of time that the soft for this kind of machine make calculators a simple learning tool for school and university. Probably calculators as engineer tools have to be redefined.. and looks like.. read his posts.. Anyway, I love my hp machines (20s, 32sii, 48g, 40g, 49g and jornada525). And I love this comunnity, and all the people that share his knowledge here are aunthentic jedi masters.. MAY THE FORCE BE WITH YOU... ==== ==== ==== Earlier I had mentioned that it was likely that the program PowerPlot for the HP48 would be ported to the HP49. Due to recent events, I am sorry to report that this has changed, and there are currently no plans to port PowerPlot to the HP49. However, should anyone be interested, my offer still stands. If you are ==== Does anyone know of a computer software editor to use to write programs for the HP48 on a computer to later be downloaded to the HP48 ? The absolute least would be one that you can enter the hex code equivalent of hp code, but preferably one that recognizes hp commands and automatically enters the correct numerical equivalent. And what would be fantastic is there would be available ==== The matrix editor can only handle numerical matrices. However, you can create a symbolic matrix on the HOME screen like [ S1 , S2 ] You can use the CAS matrices commands with symbolic matrices, it just has to ==== Colin Croft a 216crit : answer. You can manipulate symbolic matrices on the 40 from HOME, but that's a little bit tricky. The best is to use list of lists to represent matrices (like on the 48), you can enter lists as elements of a list in the list editor. Then you must use S1-S5 to represent parameters. I have done all I could so that the CAS matrix commands understand list of lists as symbolic matrices, for example rref, JORDAN, ... (see the 49G CAS instructions) will accept L1 as argument if L1 is a list of ==== ==== Visit the thread: ==== It does not really matter what HP does, the HP49 is outdated and slow. What the guys at ACO did was amazing, but no amount of genius can keep squeezing more performance from an outdated hardware platform. Face facts guys, it (the HP49) is not going to get better and eventually the majority of people will loose interest and move onto something better. Andre Claassens aim ;-) the ==== AMEN. The root of the problem can be probably traced back to the point of time, when couple years ago someone at HP made educated, executive decision (may be even made by a team of highly paid experts) during some Decision Making Workshop Meeting. The outcome of such must have been that the customer will accept ancient hardware packaged around hype of most advanced in the world advertisement and HP will save money and software expertise - read compatibility with old 28/48 series. I can even envision software developers and engineers actively cheering such effort as typically it is human nature to defend one's own work and tend to miss signs of product deterioration. Also most people involved were by such move kept in the comfort zone of the hardware/software they were working and familiar for years. Such makes those people automatically experts over anybody brought to the program later, and it also saves them from rewriting simple and boring parts of OS into the new platform. This is also human nature to support such comfort zones. It is managment responsibility to force new ideas and solutions over resistnat corporate culture. Clearly HP lost it's vision in this case in the worst possible time, when it started to have recognizable competition in the field. The idea of saving money by pushing repackaged hardware junk was possible at the time of HP 48GX model when Saturn was already outdated but had no competition whatsoever. But such decision at the time when even such weak competitior as TI is capable to built much more advanced machine in the same price brackets was simply a nail to the coffin of the entire HP advanced calculator enterprise as anybody can clerly see now, and I stated here at the time when the first info about HP49 was available on this discussion group. Don't think I'm happy. I wish, I was wrong, but it is not the case. I don't blame anybody personally simply because I don't know who actually made such stupid decisions years ago. Probably nobody admits today as the old saying says Success has many fathers, failure is an orphan. Dead hardware can not be reanimated even by heroic effort of software developers. At some point of time managment of the company has to face a decision, that no matter how good their emplyees are in squeezing every bit of performance from the old junk, the buil in performance limits will simply render their effort just a waste of money and might kill the entire enterprise with total loss of expertise, market etc. when the intention of the experts was clearly the opposite. It is additionally sad, that HP was such a wimp, when one considers, how ==== ==== i think (heh... i think sounds dumb on its own... heh heh heh) we (the 48 community. please don't mention the 49) could (want, like, should) look for what the GX lacks and improve on those we don't need no fancy RISC cpu or high res lcd. Or , at least, i see it that way. And, definetly, keep the GX housing and keys. DEFINETLY. Anything else is of lower quality. a graphing calc. That cannot be remedied with a Saturn/Yorke spin-off, so we might as well just pick another processor. The housing and keys is the big question, much of the other stuff is in the mold already. 4 of my friends (three EE's and one datamatician) have wanted to join in - that was the four I've asked until now. We have not found an lcd yet, but we will use the XScale or nothing. It runs in sleepmode at 50 MHz / 10 mW (not 1 GHz is tested, but the CPU warranty would be on us then, so that won't be implemented. If anyone is interested in the progress, I'll probably post it here. The XScale @ 733 MHz performs roughly 2000 times better than the Saturn. Even at 50 MHz it'll be more than 100 times as fast, and use less power I guess. It'll also be a bit more expensive, and that may be the only thing to throw us off - Intel haven't returned my inquiry yet. We seem to have ==== Yes, This sounds like our dreamed machine.. How could we help? It's time to play a diferent game, because the game is over.. J.Manrique CdU dela ETSIG PS: Ep II Teaser Trailer has given me new forces... ==== ==== I think this is an excellent idea. Might I make a suggestion at this early stage. If you are going to develop such a machine, write the operting system in modular format. What I mean is develop a 'core' operating system that gives the calc it's standard functionality - basic scientific functions and programming language. All the other features (CAS, Statistics, Formulas, Graphing etc. ) should then be downloadable modules, much like the libraries on the HP48/9. This way a user can decide what he needs on the calc. It would also permit students to 'downgrade' the functionality of the calc for examination purposes by deleting all the add-ons, thus making it more widely acceptable in examinations. As I have said elsewhere, I do not like the idea of this group developing software and refining an hP product free of charge, while HP, after having ditched it's development engineers, sits like a parasite making money selling calculators while someone else does the work for them. Good luck and I hope you can get this project working. Andre Claassens for we the we (not 1 to PCMCIA much like Java in structure - library modules with standard and well defined interfaces. This is possible because of the huge performance of the XScale - we do not need to push the last percent out of the CPU, we can focus on maintainability. for I'm not sure this will be so much a school calc as a powerhouse in the ==== of course you mean the 48 *processor* lacks speed. agreed. But for what applications? For the existing applications it is more or less acceptable. But if we are to be using c compiler and a heavier cas, oh yeah, we need the power. Definetly in the 200MHz area, with fluctuations. As i see it, you (well, to be honest, i do too, maybe even more than you) want the power of a '95 era pc in your palm. Yes, i agree on that. I did not ask you. What do you want this machine to do? Now for some brainstorming. Get your umbrellas... The 48gx. Is it a good machine? It is fscking awesome, of course! is it the perfect machine? most probably not. But this didn't stop Mika to write Jazz, and all the others to write equaly impressive code. The fact that this machine had little inneficiencies were somehow the factor for its success. It was hackable, not perfect. If it was perfect, we wouldn't strive to make it better outselfs. Hope you are getting me because i can not analyse this better. What imperfection in the name of god do you find in the gx housing!??! i am not talking about the keyboard layout, i am talking about the quality! Ok, so you know who i am : a physicist, 4th year. Next summer i was supposed to be getting my bsc, but i will have to delay it a year or so. Many things took time from me (guitar, my thoughts) so i owe a lot of classes. So far i went very good at programming (C++) and math (almost everything). I am not interested in the progress... i am interested in participating. whoa...that's a lot of oomph! ==== Lo people ! It is good to see how passionate this community is ... I have an idea, a thought .... This new calc, should be designed the way this community wants it to be, now most of you are GX lovers (i prefer 49 though :) ). The question i am asking, is why do we have to re-make an hp48gx with more power ?? Why not make a calculator from scratch ? Where on a web-site, people could choose the housing and keyboard, and propose their ideas that would be discussed .... Of course the final I (and probably others) dont want a supra powerful HP48, but an easy to use powerful calculator but also very customizable (for 48 fans). hardest and best calculator available. Isnt that a proof of their competence ? ==== Lo people ! It is good to see how passionate this community is ... I have an idea, a thought .... This new calc, should be designed the way this community wants it to be, now most of you are GX lovers (i prefer 49 though :) ). The question i am asking, is why do we have to re-make an hp48gx with more power ?? Why not make a calculator from scratch ? Where on a web-site, people could choose the housing and keyboard, and propose their ideas that would be discussed .... Of course the final I (and probably others) dont want a supra powerful HP48, but an easy to use powerful calculator but also very customizable (for 48 fans). hardest and best calculator available. Isnt that a proof of their competence? Why not let them give ideas for their new favorite toy ? I have discovered hp calcs 2 months (in just 16 y/o) ago and i am fond of them ... ==== I have a probably heretical suggestin. Given that hand held, and for that matter, laptop computers are getting cheaper, why not build a calculator for small computations. I would like to see something like an upgraded HP27. Perhaps with a ==== Isn't that what happened over the last ~2 years? ;-) Without the brain power of many paople in the c.s.hp48 community, the 49 OS would not be what it is. just my 0.01 cents (discounted) Raymond As usual, OMHO, and maybe tomorrow I forgot what I said today;-) Exactly. What the '49 couldn't, and what I expected from a real successor to the '48 series. I want Mathematica performance in a machine the size of the '49. The XScale supports, and will be equipped with if everything works out, SDRAM. That means that in some far fetched fantasy, one could shell out US$ 50 extra to get 512 MB RAM instead of 32 MB in the machine. It could even be a custom design like Dell applies - order on the net, and receive within 10 days your personalized math machine. I'm getting you. The system will be open, so anyone will be able to create software for it - we won't install some protection against running some self-made ARM code to take control of the hardware underneath the OS. the The quality is good, the layout of the keys is not. I'm saying that it will be tough to reproduce that type of quality. hehe, yes I can understand that ;-) I would not keep such a fun hobby all to myself. I couldn't even do it by myself. There will be a (probably long) period before any development can be distributed. If I need to shell out US$ 100.000 on this project, I'll need to be somewhat in control. Yes, and the IO companion chip (Intel 80312) does have an 64 bit 100 MHz interface to the XScale, and it even supports PCI bus @ 32/64 bits, 33/66 MHz operation. Intel Flash RAM at 0.13 u does not use up much power, neither does the XScale, and neither does alot of new display technologies. in this new machine. If anything it should be a livense of Mr. Parisse' CAS ;-) We have already discussed several ways of mudularity, which means for example that we will be able to change to another display within hours. Adding RAM or changing types will be a matter of hours too. The OS has with 95% certainty been decided to be OO. We will probably make a C/C++ compiler for it too. the competence I'd love ideas, and whereever possible, I'd be glad for help. I just need to make the calls, since it'll be my money in the production. I'm sure everybody here has many great ideas. That was the reason I posted this at the first place. You do not need to make excuses because you're French - that's something to be proud of :-) Your spelling is close to immaculate in my eyes. Whoa, definitly sounds like a worthy project guys. Just thinking here, we're going to be drawing a decent more battery power than any other calc out there and batteries are not going to last too long. So why not use a Lithium Ion battery? It could ether be replacable or build right into the housing. We *should* be able to get a decent amount more runtime out of it. Just an idea. O - Be sure that you do not use the same keyboard type as the 49, the keys are much too hard to press. I much prefer the Casio and newer TI keyboards. Andrew itself will run off one AA battery if needed. I'm thinking about 600-1200 mAh Lithium Ion (to use old and inexpensive technology), and then a power plug that will take a charger. The machine will be able to run off that charger too - it will easily be able to catch up with the power consumption ==== Already been done. HP200LX palmtop. Smaller than a 48 when closed. Press the calculator button and you get an implementation of the HP- 19BII calc. Just as you describe it. RPN. 4-level stack, 10 registers, all visible. HP solver, plotting, stats, trig, financial, unit & currency conversions, etc. Unusually for a palmtop it has a numeric keypad which makes it slightly more calculator-like. Mine's also running Derive v4.11 (factors a 15 digit number as two 8 digit primes in 22 seconds) and Borland's C compiler just in case, etc., etc. In fact, any software that will run on a 8086, CGA screen PC. Mine also has my diary, phone book, a Collins dictionary, book reader, mobile if I'm stuck for a phone line. ==== It's nice to read some much enthousiasm here. But please, be realistic about the amount of work involved in creating a new calculator. it took 5 years, 5 people to write it) and Erable (and Bernard worked on it for over 5 years too). Still, it took over three years to get ROM 1.19-6 It's been estimated by HP, that creating the HP48GX has required over 150 engineer years. Not that I want to sound too pessimistic, but the task is huge ==== ==== ==== And say what. . . 1) you're making a big mistake! 2) please release the code for the calculators! 3) may the fleas of 1000 camels infest your armpits! ==== Jean Yves, with some technical queries. To date they have not even had the good manners to bother sending a reply. (Not even and automated reply to acknowledge receipt and neither do they appear to care what we, apparently a minority group in their eyes, think. I have had enough to say about HP recently so I am not going waste everyone else's time repeating myself. Once again, thank you (and all the other guys) for all you efforts and enthusiasm. Andre Claassens