A55 i just happen to read this message. water will ruin the leads to the components ( accelerates corrosion ) if \ left unattended and not dried inmediately. expeciallly the ones at the \ microprocessor (very tiny and close together). i've seen many radios ruined because of \ water penetration . wd40 displaces water and takes away moisture, but leave residue. ==== Please read the message attached. Is this really a CAS bug?! Well anyway: I got problems using XYGROBDISP when the grob excess the largest of the screen. That corrupt the screen and sometime i got very,very bad ttrm..( i had to reflash my rom last week!!) ==== The XYGROBDISP should enlarge the HARDBUFF grob to accomodate the new grob. But if you use grob (for example 20x20) and set his coordinates that XYGROBDISP enlarges ABUFF more than 5 pixel horizontaly then the display will be crashed. Ex: :: TOADISP BINT116 BINT1 BINT20 BINT20 MAKEGROB XYGROBDISP WaitForKey DROP DROP ; But there is no problem if you use GBUFF and then it works very well and GBUFF is enlarged correctly. Ex: :: CHECKPICT TOGDISP BINT116 BINT1 BINT20 BINT20 MAKEGROB XYGROBDISP WaitForKey DROP DROP ; RPLMAN.DOC, page 102, chapter 20.8.1: 2. In practice, it is best to use the word XYGROBDISP to place a grob into the display grob. The word XYGROBDISP is conservative in nature - if the graphic ==== Already a year ago a reported on the strange behaviour of the 49-XYDGROBDISP but there was no reaction. Since then I'm using CkGROBFITS in ABUFF which, of course, is not an elegant work-around and disagrees with RPLMAN, apart from that it cost some 20 bytes extra. I forgot I came along this while porting the EQ-lib, The bug is (was ?) actually RECLAIMDISP -- it didn't resize ABUFF to its default size as it does on the 48 (see also http://www.hpcalc.org/details.php?id=3249). The problem is that the 49 stack display routines expect ABUFF to be not wider than 132 pixel so normally a RECLAIMDISP at the end of a program should fix this behavior but at least for ROMs <= 1.10 it didn't. I discussed that a while w/ Cyrille 2 years ago and we came up w/ a workaround (see the link above). I added the fix to the EQL port but added ROM version detection, it ==== We can actually do this easily with good 'ole AWK which has been bundled with every version of Unix since the dawn of time ;). Here's the one line AWK script : awk '{ if(index($0,*)==1||!$0) next; printf %st%-13st%sn, $2, ---------------------------------------------------------------------------- ------ before replying. replace something like ^=([^ ]+) *EQU *#([0-9A-F]+)$ with EQU 2 1 Not tested though, and it'll be a bit slower than perl, awk, sed and ==== Well, I had many different experiences when it came to recovery of memory. Sometimes it works and after a long time I have everything back. Sometimes I have many directories named DIR1, DIR2 and so on in HOME after recovery. They contain my stuff and usually one of them is something like a copy of the hidden directory with the key ==== BTW: Is there a way to check if one is inside a second invokation of the build-in editor? I hate when the EXITED changes the Header back and I'm just exiting from a second invocation of TE. A text file is just a string stored in a variable. Recall the string to the stack and then hit the down arrow key to get the default text editor. You can also use the APPS menu to bring up the text editor or the command of whatever other text editors you have on your calculator. You can also use the up arrow or HIST ==== (I'd post this to the bugzilla thing, but it apparently requires a new When the ROM asks to turn flag -3 off, it says which probably has a typo in it somewhere, using a very flexible meaning of 'typo'. ==== How can I get a random number in ML ? Is there any ROM routine ? Or do I have to read the internal clocks ? HPThifu C=DAT0 2 LA #0023 GOSBVL =MUL# C=B A D0=C A=C A LC(5) MAXVALS GOSBVL =IntDiv A=C A LC(5) MAXVALS GOSBVL =MUL# CD0EX A=B A ==== I'm glad you asked that question: Naturally a total user control of the CAS or at least in the 48-way: Flag -1: Principal Value/General Solution DEG/RAF/GRAD stays on and rad is used _internally_ When the calc is in Real Mode and the Forbid Mode change is on and the all vars are reals is on and who knows what flag - the error should not be Mode Switch Cancelled, but eg: ASIN Error: Complex Result and an appropiate flag indicating a complex result should be risen maybe you could even say if only the real part is chosen and the imaginary part is ignored by setting a flag. At least the error text should be changed. AND I would like to see some engineering calculations: ==== I guess that would be my job as I was the one who was maintaining it in the first place. I did not know that people were still interested in it, so I just kept it under wraps. Maybe it's time to bring it back. Note that a lot of what people asked for has already been given to us either in the ROM or programs: auto-simplification, user-defined curve fitting, pretty print for limits, faster wireframe plotter, just to name a few. This is great; now we can ask for the really hard stuff: symbolic tensor algebra, symbolic products (come OMVHO: Without the Holy Spirit (or Holy Ghost, if you prefer that) there can be no wisdom. That is: You think you have an excellent scientist brain without religion, but in fact that is a lie from the devil (or perhaps it comes from your flesh eg. human mind) I used to be one of those up to age 35 when God interfered my life in such a way that everything is different now! Now I think I understand you. You have first to know the number of the numeric solver menu. You get it by pressing after its setting a key to which is assigned the command RCLMENU (this assignment is basic for any user). Asume this number is nn (I forgot my HP49 today). Now add the following item-list to your list in APAR: Then, pressing the item MySolveMenu in the new APPS choose box, you'll see your expanded Solver menu :-) Here LMN is from OT49. Easy, isn't it? - Wolfgang PS for those who use an own assignment on the APPS key: Sure, but APPSman allows to insert standard APPS submenus easily in your customized menu, just put 2 in your APAR to have the I/O submenu displayed just as if you were in the real APPS menu. I'm sure there's a way to do the same without using APPSman but it's surely faster to use it right away. Libraries interacting with the APPS menu are a great idea, unless you don't like the menu structure they use. But I usually have better things to do than keep editing the menu to add/remove every library I feel like trying... ==== thanx for the link, I've downloaded the program for my 48, and read the text file... However, I'm still unsure to which commands I can use to move the cursor... A little more help would be greatly appreciated! :-) -- . Artur Meinild . . . . . : . 232lge R ud237r . . . ___________________________________________________ Eásite : 6paq.messages.to : Wásite : 6paq.go.to any key in any shifted plane. For example, to assign the left-unshifted action also to alpha-RS-left, you can use SA left A?D (short) alpha-RS-left I have this really big equation f(s,t,u,v) that I want to derive wrt s. This is no problem. But after playing around with the answer trying to simplify it, I end up with 12s - (2t - (24u - (12v + 50))) This always happens when I try to simplify equations. How can I get from there to 12s - 2t + 24u - 12v - 50 which is the expression I want? Are there any flags in this respect? Ray pictures on hp are coded in hexa using 16 symboles (0123456789ABCDEF). since they are 255 of them why dont we code grobs on 255 chars so that grob size is divided by 256/16=16 ? PS: please excuse my poor english... GROBs are already stored in a very compact format: One pixel per bit, or four pixels per nibble, or eight per byte. It's just that to make the GROB readable (if you call that readable), the decompiler shows each ==== 3BD73420.E72250C7@iname.com... ==== Write down your equation and then *fully* parenthesize it. (That means that *every* binary operator should have an enclosing pair of parens... just like operator precedence did *not* exist). Now move the binary ==== I understand you very well. And to make the decision even more difficult, it is not only that the stack provides such a huge variety of ways to solve something. In addition you have a huge number of commands that can be used. So the choise of one way for the solution is more difficult. But I think that this is also what this group is meant for. You have a solution? Post it and ask the people what they think. You get 100 tips and tricks how to do it better/faster. And with the time you accumulate knowledge that you can use in future. For example, some years ago I wanted to write a routine that finds the smallest of a list of numbers. I tried many possible things until some guy asked me Why don't you store your list of numbers as an array in ==== You mean RPN (Reverse Polish Notation), of course :-) Bhuvanesh. Whenever I believe I am having issues w/hardware or an M$ product I boot preconfigured rescue/demo disks. I believe I remember at least a few having minicom. Even if they do not have minicom you'll still have ==== Greetings, I have an HP49G version 1.19-5 fw. I was hoping that someone could tell me how to get unit conversion to work as per the 48S-GX? Currently I get this: 1 on stack [RS][6] (units) scroll down to length press enter scroll down to feet press enter get 1_ft on stack get back to length submenu scroll down to inches press enter get 1_(ft*in) on stack WTFO?!! This is a simple example I mean I can convert a foot to 12 inches in my head but it's just an example. Whoever thought this would be the most desireable of default settings should be beat to death with an HP49G! I've look around and can't find any data explaining how to get it back to the swell-in-all-ways HP48 series behavior. Please help! Gary ==== [MODE] |FLAGS| Use up-arrow until you reach System Flag 117: 117 CHOOSE Boxes Press [F3] which is labeled |V CHK| The display now reads OK, I found out! However, if you try to enter +OO, the calc will reduce it to only OO... What you have to do is enter OO in algebraic, and then evaluate the expression, so you get '+OO' on the stack. Then type the number of the key you want it assigned to, and use the ASN command to manually assign this one key! When you recall your list of keys, it will in fact appear as '+OO', which is what I wanted... :-) Now I only need to find out how to remap the arrow buttons... (see other topic) -- . Artur Meinild . . . . . : . 232lge R ud237r . . . ___________________________________________________ ==== Well the fact is that the pocket PC has available tech specs, bout when I searched for the GB IR transmission rate it said it was 512 kbs!!! congratu- lations Nintendo!! XDDD So, as I can't change hardware or software of the GB, I have to tweak the 48GX. I don't have time nor knowledge at the moment, so I was thinking about ==== I want to map a key to echo the sequence: -OO that's: [minus][infinity] I've tried to map this program: but it puts the sequence directly on the stack, and I don't want that. I just want the two characters to be echoed in a row, just if I pressed [minus][alpha][right][cst] and this should work in any entry mode... Any bright ideas??? ==== ? ==== Could someone give me a little more info about KEYEVAL ? I can't find anything in my User's Guide. thanx a lot... ==== I've redesigned my site -- The HP Knowledge Base. For those that aspects of HP calculators: programming, usage, tips, etc. The biggest difference is that now anyone can post their own I welcome you all to take a look at the site, and, perhaps, add something to it. The address is http://move.to/hpkb or, directly, I would love to, but I don't have the $time$... :-) - Artur - ==== or: << { 1. 3. } PICK3 TYPE POS GET EVAL provided the list of types is exhaustive, otherwise you'd have to test for POS to be non-zero. Werner Huysegoms xwerner_huysegoms@my-deja.comx (delete the leading and trailing x) ==== X Bernard is right here: A) You can add the extra time that EVAL uses if, and only if you prefer the TI chosen answer B) There is no option for the TI to give the HP answer, in case you happen to prefer that one. C) As always - The HP 49G let you have the cake and eat it, too (was this the correct phrase?) You, The User, have the choise: Use Algebraic or RPN, EVAL or not, etc. With the TI-89/92(+) it is always: what you're given is what you've got :-( Veli-pekka Nousiainen PS: If Professor Parisse want the CAS to be even better than it already is, then one thing is definately needed: PLEASE CHANGE TO RAD INSIDE ONLY The User says Degrees/Grads/Radians and the CAS obeys. That's the rule 1. If the algorithm needs rads, then do it internally, please. (Remember the pi/180 or pi/200 multiplication) The result should be back in User Units, not rads... (Multiplied by 180/pi or 200/pi) I have the custom menu(s) the way I like them. In the way the HP 49G does it (used directly on chosen subexpression(s)) it was impossible to do in the TI-89, I did a little User-RPL implementation of the simplex method, and i wanted it to work with symbolic matrix .It doesn«t really iterate, it calculates an iteration in order to help with exams. The problem came when using it with a symbolic M in the C row (the upper one) I want to simplify the upper row of the matrix expressions, but without EVAL (because of the process there is a var called M which is worth a big real number). At the moment i extract the first row , AXL and COLLECT, AXL and ROW- This is slow, and my friends with older ROM claim that it doesn't work with I tried this but it didn't work (The expression was evaluated). However, by writing ==== = should not create a test clause (like it really does), but a system of equations (or a vectorial equation, if you like this form the better) If = is applied to two scalar expression, I get in return an equation... so why does it return a value when the same command is applied to a vector? there's also SysRPL CHANGETYPE to do the same: :: CK1&Dispatch list :: #02AB8 ( DOSYMB ) CHANGETYPE ; ; Of course you need a SysRPL compiler (e.g. Jazz) to make this an executable. HTH ==== when It works great, thanks :) Do you know why it works this way? this case, figure out if the two vectors are equal; they aren't. However what you want is '[X Y]=[0 0]', or in RPL syntax (the HP stores all algebraics as RPL objects), ['X' 'Y'] [0 0] =. Lists happen to be very similiar to algebraics; they only have a different prologue, but the actual contents are the same. So you can also put whatever belongs in the algebraic in a list. But you have to use RPL syntax! Then the ==== Me, too! Twise! Though more than year ago (the last centennial it was... ;-) Later, after the STO to Flash problems have been corrected I have had no flash problem - not even when I try! Good work JYA! CoefficientsFunction(n)=n/n! (Remains the same)