A-209 KEYTIME works for me That's what I though a few months ago, and I was also convinced Keytime was the solution. We will give you a week or two, and you will see that the problem will still persist. Anyhow, the default Keytime setting is 0, on the HP49G+ & HP48GII. The keyboard problem needs more than Keytime to fix it. Noel Causerano Registered Engineering Surveyor, SBQ Foundation Member, Spatial Sciences Institute Surveyor - Survey Services/Cairns City Council Director - GEOCALC SOFTWARE Registered Reseller HP Invent Cairns, QLD, Australia Email: noel@geocalc.net WEB: www.geocalc.net posting-account=CdmerAwAAAA_5trBbttE1dOcxC9Z5qXZ I've noticed from recent posts that HP49 series owners who aren't familiar with the HP48 have trouble finding basic commands (which, like a lot of things, are not in the manuals). There's a wealth of old 48-style softkey menus in the 49s. If you want a complete list, you can get it from www.hpcalc.org/hp49/docs/misc/menulibs.zip. Here are a few of the useful ones. Follow the number with the MENU command (which is in PRG-MODES-MENU or LS+MODE-MENU). 28 EDIT menu. Most commands don't work in stack mode, but it might be useful in programs. 81 PLOT main menu. Lots of good submenus. 93 old SYMB. Commands like Quad, Match, etc. 104 I/O menu - lots of stuff that isn't in APPS, including a comprehensive Print menu. 111 LIBS - same as RS-2. 113 old EQLIB - four useful submenus. 256 - the hacker's toolkit. Type 256 ATTACH first, unless you have a program like OT49 that does it for you. 258 Extable menu, if you have Extable installed. 1792 Programming commands. That last one is really useful for UsrRPL. If you assign it to a key, it's more convenient than looking through the PRG menus. And if you need an explanation of older commands, www.hpcalc.org/hp49/apps/misc/hlp49v2.zip. Soon to be followed by a major upgrade if the lazy author ever gets himself going. :-) Bill posting-account=Kc05ZA0AAADq7YTEHoMU9Aer5AhqUt9P leg up once in a while. Excellent info. Especially HLP49. X-RFC2646: Format=Flowed; Original WARNING! I would be extremely delighted if I were appertized by small news about Qonos say, every weekend by Hydrix personnel Example 1: Hardware is finished now and we found a keyboard that has the good old tactile feedback plus a light touch Example 2: The software is still under construction, but both emulators are working 100% it's the USB host that still needs some work Example 3: eCos details for the emulators will be announced to the HPgcc project members, thus facilitating the development of XScale software in C language for the emulated 49G On possible usage would be to enhance the display size to 80 lines just like on the 49g+, since the LCD has beutifully room for that. Example 4: Intel PXA 263 XScale CPU runs normally at 400MHz but will be downclocked to 100MHz when running eCos The Flash chip is ... Example 5: The emulators come without the ROM but the users are adviced to visit comp.sys.qonos where advanced users can tell how to move the ROM image(s) to the emulator space Please, Jean-Yves! Anything? Maybe a history of the project development, like this: we used a 3¸\ color LCD first but since it turned out to be a major power drainer we promptly decided to use a 3\ greyscale LCD instead (HxV: 320x240) During the development a 128MB Compact Flash Card is used and it's quite sufficient for holding the needed binaries The Host Development PC is Linux based and a cross-compiler produces the required binaries [VPN] Personally I wouldn't hand something like that over to HP. It's open source or I get paid for doing it. Steen Why do you even bother to reply to answer somebody with the mind and intelligence just slightly higher than a dog on drugs? plus it makes me receive his bs... Jean-Yves X-RFC2646: Format=Flowed; Original Maybe someone in this NG is close to him geographically? Maybe we can offer to help him.... I have no clue where he resides, but if it is close to where I am, I would be happy to assist in any way possible. just a thought -- Frank Bachman (Grumpy Aero Guy) X-RFC2646: Format=Flowed; Response His Parish is St. Thomas More in Irvine, California so he probably lives at \ the Rectory there. Tom Lake I thought that somewhere I'd seen a message from Joe, indicating that he'd had some serious physical difficulty, and that he'd left fulfillment of the calc sales in someone else's care at his abbey meanwhile, but that they'd not taken care of it for him. I hadn't saved that message or its URL, however, and I've been wondering why no one else has posted that they'd seen it too, perhaps with better recollection than I. Anyway, apparently the reason that I could not find that message again is that it got deleted from Joe's web site (holyjoe.net) by some hackers, as he just told me -- yes, I finally spoke with Joe for the first time ever (Joe is #14 here: http://abbeynews.net/contact.htm) Joe just told me that he had been deliberately run down by a malicious motorist, and sustained a serious back injury; obviously there are situations in which we need all our energies to recover, and simply can't take care of anything else, even though we would never fail to do so under normal circumstances. No doubt you'll be hearing from Joe again directly in due course, as soon as he can spare the effort it must take. Best wishes to Joe for a speedy and full recovery from http://www.mum.edu and http://www.maharishischooliowa.org go to holyjoe.org and click on the Catholic forums. he left a message there \ a while back. He has had dual disc failures, both of his computer's hard disk \ and his back. He was flat on his back for several months, it seems. He had left the 'fulfillment' of the orders to 'the school' and apparently they \ have not been able to either fill the orders or communicate with the people who have placed orders. I know nothing more than that. I found the message because of browsing his \ site for other info, and am not one of the ones who ordered. Lee never trust a priest... \With malice toward none, with charity for all...\ [A. Lincoln] http://www.bartleby.com/124/pres32.html \Never entertain negativity, never denounce anyone, let us be radiating the light of life.\ [Upanishads] X-RFC2646: Format=Flowed; Response Put a 1 gig card in there and see if it works not sure if it is easy like that. i want it to boot from the flash without metakernel anyhow involved. should be possible just regarding the size of tiny linux-distros. asmutils used gains further flash space (or is it simple to execute sd-card data?) i don't think mmu usage makes much sense. a simple search resulted in: http://www.linuxdevices.com/news/NS5794540327.html where \A whitepaper by the Miniket's system architect, Hyok Choi, is also available. The paper describes the rationale for using uClinux rather than Linux on ARM9 processors with MMUs...\ i'm yet very sceptical about my ability to see every detail of such an operation. but i'm working on it... helluvadrug to point ou the important parts... \...however, the Miniket's uClinux 2.6 kernel does not use the chip's MMU. This is because research at Samsung showed that context switches and IPCs (inter-process communications) are faster under uClinux on processors that have virtually indexed caches and a TLB (translation lookaside buffer) without address space tags.\ at the very moment trying to understand) follow this link: \According to Choi, a vanilla version of the uClinux/ARM 2.6 kernel can boot to a sash shell in 80msec. The Miniket can boot to its main() function in 0.8 seconds, Choi adds. The Miniket's firmware features an extremely small footprint, with uncompressed sizes listed by Choi as follows: - uClinux Kernel -- 403KB - FS for NAND module -- 110KB - device driver modules -- 337KB - main app -- 588KB - UI resources -- 1.6MB \ HelluvaDrug and another one. i'm moving this thread to \root level\ for better access and due to off topic analytic continuations for nonintegral n, but I haven't heard about them, and Abramowitz and Stegun is at the office), so it's missing only one value (n=-1). Much more disappointing, for me, is that the digamma doesn't return explicit values for integer arguments (-eulergamma+rational number), while the polygammas do (when they work!). I guess it's related to the presence of Euler's constant. Also, the relationship between the polygammas and the digammas isn't implemented; it can be fixed by the user by defining 'd1Psi(x)=PSI(1,x)' etc., but why should we have to bother? As for the keyboard, well, I was making the point that the 49G didn't have this problem (I have one and there's no problem with it), but still many people dismissed it as a toy because it's blue and has a rubber keyboard (other things as well, but this is what struck me). The 49G+ would appear to have much more serious problems with the keyboard (I don't actually have one, so I'm basing this on posts here). Anyway, I basically do agree with you. Oh, and you're right about the contour plot! posting-account=p96ZUA0AAACXwuRub9Hyvsg2HJdsq2Lr Yep, this is even worse than I thought, I just verified. The polygamma-digamma derivative relationship is not implemented at all! Don't get me wrong, I like HP. I only wish they would start making a quality product again. They know about the broken alarms (I contacted them about that 7 months ago). They have not updated the ROM since shipping the same flimsy design, which will eventually break, and are giving customers a 90 day replacement guarantee. I heard that the manufacturing of the new keys are supposedly environmentally more friendly, however (according to Eric Rechlin, who sold me my 49g+). 8 posting-account=TNcafQwAAAChTv9Mf9hLBl-3-uNSbRwn All, Each of the following problem parts asks you to give an algorithm to determine two distinct numbers x and y in the set S that satisfy a stated condition. Your algorithm must be specified using pseudocode. You must justify that your alogorithm has the requested run time. for all w, z in S. b) In O(n lg n) time, determine x, y = S such that x!=y and |x - y| =< |w - z| for all w, z in S such that w !=z Bob. This reminds of college questions. Hang on a minute.... ... we never had trick questions though. And this must be a trick question because there is no solution in O(n) time unless you know something about the data and unfortunately all we know is that it is unsorted. Are you sure that you don't mean O(n^2)? Of course, you could always give them a trick answer, which is to define parallel processes in your pseudocode, and then you could solve it in O(n). (Think of the extra credit you'd get. ;-)) Hmm, looks tricky. ;-) -- Bruce Horrocks Surrey England Knowing that S is a finite subset of the reals of n elements, can't we find the max and min of S in O(n)? There are (n-1)! absolute differences of distinct values of S, and the miminum of these can be found in O((n-1)!) time. to point ou the important parts... \...however, the Miniket's uClinux 2.6 kernel does not use the chip's MMU. This is because research at Samsung showed that context switches and IPCs (inter-process communications) are faster under uClinux on processors that have virtually indexed caches and a TLB (translation lookaside buffer) without address space tags.\ at the very moment trying to understand) follow this link: \According to Choi, a vanilla version of the uClinux/ARM 2.6 kernel can boot to a sash shell in 80msec. The Miniket can boot to its main() function in 0.8 seconds, Choi adds. The Miniket's firmware features an extremely small footprint, with uncompressed sizes listed by Choi as follows: - uClinux Kernel -- 403KB - FS for NAND module -- 110KB - device driver modules -- 337KB - main app -- 588KB - UI resources -- 1.6MB \ HelluvaDrug Hehe, you would, would you? ;-) There are too few hours in a day unfortunately... Steen I would like to know some time that it already happened in your HP 48 and why that happened in mine. I need to change the part of the code that manipulates KML bitmap.