A157 MDI - Multiple document Interface. Basically the window layout where you get document windows inside an application window. So, a window per connection, with all the features of the particular model or emulator; and an application window tying it all together. About the auto-connection, it could be controlled by a (yet another) þag. Or at least to recognize a command to start server mode anytime a patent conection exists and the PC software requests it. And please, include ALL the UserRPL commands in the help system. > - Add support for comm ports > comm 4 Seems like it ought to support however many COM ports the operating > system makes available. It sounds easy to me, but I don¹t really know > what¹s involved. - add concurrent connections (i.e. two calculators, calc and emulator, etc.) > via MDI interface Good idea. I don¹t know what an MDI interface is, but now that we¹ve got > RPL calculators with I/O that can¹t communicate directly with each > other, it would be nice to be able to at least connect both to the PC > and use it to relay the transfer without having to tell it to store it > in a Þle while we switch connections. - Usb auto-detects the 49G+, so, can it autoconnect? (I guess this one is a > calculator issue) So far I haven¹t had any luck with getting it connected for anything > except ROM updates. Do you mean have it go into server mode when the > cable is connected? I think that I¹d really rather have the calculator > wait until I tell it to go into server mode. - >First: conn4x upgrade should have all of the program included, not just >>upgraded parts Agreed. I knew that there was an updated package available, so I Þrst > installed that one instead of the one on the CD-ROM, then uninstalled it > to install the old version so I could reinstall the new version over it. > Who would¹ve guessed that it wasn¹t the whole package? If what¹s on the > website isn¹t a complete package, at least make that very clear. >Second: it should be a zip package For downloadable packages, most deÞnitely. Or better yet, a > self-extracting compressed Þle. With what I downloaded for Conn4x from > the website, I just tried it both ways and it came out over 29% smaller > both times. Ok, I know it¹s not the biggest Þle, but it still would > seem to be worth the tiny bit of extra effort from HP. >Third: ASCII mode is needed Yes, ASCII mode is deÞnitely needed. Either by adding Kermit to the > Conn4x package (assuming that they get it working) or by adding ASCII > mode to Xmodem. For that matter, even ASCII mode for the SD card. All > the pieces are there -- decompiler, translation programs (both ways), > and compiler. They just need to be put together properly. I do it myself > for transfers with the SD card, but it adds the binary transfer header, > string prologue and so on that I strip off in the text editor, and if I > go so far as to build a counted string form (useful if the Þle is going > to end up on a 48 series), I haven¹t been able to Þgure a way to avoid > a string within a string until I get it off of the calculator. I don¹t > Þnd it that hard to do, but difÞcult to explain how and why and what > to do differently depending on the calculators involved and Þle > transfer methods available. It seems to me that it should be a piece of > cake for the developers. Which reminds me...the backslashed decompiled double-quote and > especially NUL certainly have their advantages on the command line, but > aren¹t always so nice for Þle transfers. The 49G and 49g+ understand > counted strings either from the command line or an incoming ASCII Þle > just Þne. I very much wish that I knew of a way to have them generate > them for Þle transfers at least (other that my clumsy UserRPL method, > that is). Ok, I know I could get an IrDA to RS-232 adapter, but I shouldn¹t need > one for connection to a PC and operating system with built-in USB that > works with everything else. and I expect just that for Xmas >>Opinions? I expected Kermit to be built-in and the USB to work with the Þrst > release of the communications package. I hope we get your wish. ==== > The CAS simpliÞes x^2-x-x^3 to -(x^3-x^2+x) (114 clear). Is there a > way to set it to simplify this to -x^3+x^2-x ? On the hp49/HP49+ use FDISTRIB. ==== > For some wierd reason, I have to press Eval twice on the SOLVEVX > result to get it to simplify... Weird isn¹t it. It¹s just because the CAS doesn¹t automatically simplify answers though. I don¹t see why I should have to do an EVAL beforehand though, all it does is make the LHS more cumbersome. If it wants to convert 2^x to EXP(X LN(2)) it should do that automatically rather than complaining about reduction errors. ==== > There is a message posted somewhere else in the newsgroup questioning > the validity of the result DOT([1+i],[1+i]) in the HP 49 G calculator. > I believe the HP 49G gives the correct result and here is why: The debate that will probbaly follow at this point might be cut off at its root by noting that different deÞnitions of the scalar product (or dot product) in vector spaces are possible, useful, and in use. > A dot product -- also called internal product Actually, it is called the inner product. > Thus, for vectors u = [u(1), u(2), ..., u(n)], > and v = [v(1), v(2), ..., v(n)], where the parentheses indicate > a sub-index, the dot product produces a scalar quantity and is deÞned > as u.v = u(1)*v(1) + u(2)*v(2) + ... + u(n)*v(n) This is one deÞnition of the dot product. Another one is u.v = u(1)*v¹(1) + u(2)*v¹(2) + ... + u(n)*v¹(n), where the prime denotes the complex conjugate. In fact, this latter deÞnition is by far the one most commonly used to deÞne the scalar product in C^n. > So, the HP 49G is right! The HP is not any more right, or wrong, than the TI in this case. The HP uses one deÞnition of the scalar product, whereas the TI uses another one. The TI¹s deÞnition is the one that is most commonly used for complex vectors. ==== > A dot product -- also called internal product Actually, it is called the inner product. Thus, for vectors u = [u(1), u(2), ..., u(n)], > and v = [v(1), v(2), .., v(n)], where the parentheses indicate > a sub-index, the dot product produces a scalar quantity and is deÞned > as u.v = u(1)*v(1) + u(2)*v(2) + ... + u(n)*v(n) This is one deÞnition of the dot product. I would say that this is the deÞnition of the dot product, which turns out to be an inner product if you are considering real vector spaces. > Another one is u.v = u(1)*v¹(1) + u(2)*v¹(2) + ... + u(n)*v¹(n), where the prime denotes the complex conjugate. In fact, this latter > deÞnition is by far the one most commonly used to deÞne the scalar > product in C^n. And this one is an inner product on complex vector spaces, whereas the dot product is not. So the HP49 seems more consistent with the fact that dot product is often referred to as an algorithm regardless of the Þeld - following the HP deÞnition, but the TI89 behaviour is more useful. Maybe TI should have chosen another name (scalar product or inner product) for its function. Camille ==== > Thus, for vectors u = [u(1), u(2), ..., u(n)], > and v = [v(1), v(2), .., v(n)], where the parentheses indicate > a sub-index, the dot product produces a scalar quantity and is deÞned > as > u.v = u(1)*v(1) + u(2)*v(2) + ... + u(n)*v(n) > This is one deÞnition of the dot product. > I would say that this is the deÞnition of the dot product, which turns > out to be an inner product if you are considering real vector spaces. > Another one is > u.v = u(1)*v¹(1) + u(2)*v¹(2) + ... + u(n)*v¹(n), > where the prime denotes the complex conjugate. In fact, this latter > deÞnition is by far the one most commonly used to deÞne the scalar > product in C^n. > And this one is an inner product on complex vector spaces, whereas the > dot product is not. So the HP49 seems more consistent with the fact that dot product is > often referred to as an algorithm regardless of the Þeld - following > the HP deÞnition, but the TI89 behaviour is more useful. Maybe TI > should have chosen another name (scalar product or inner product) for > its function. Is then this ŒINNER¹ << CONJ DOT >> ==== As it turns out, I just discovered that the HP-49G+ I bought a while back comes with a really nice, black leather case with a magnetic closure device (more sophisticated and professional than the Velcro used to close the þap on the case that came with my HP-48GX). Anyway, since the HP-49G+ case they can just sell me a couple of HP-49G+ cases for my two HP-48GXs. If they can¹t, maybe I can go directly to HP to buy the two cases. John Bonnett > www.calcpro.com (??) I know they have some hard cases for various models. -- > Does anyone know where I could buy a high quality, soft, black leather > case > for an HP-48GX calculator? ==== equipped with ALG48. I have created symbolic variables QD1 and QD2, where QD1= 2sin(x)+cos(3x) and QD2=-3sin(3x)+9cos(x). I then created a 2x2 symbolic matrix {{QD2^2 QD1*QD2}{QD1*QD2 QD1^2}}. I now need to substitute the contents of the two variables into the matrix so that I can evaluate the determinant, inverse, etc. How can I do this? ==== > equipped with ALG48. I have created symbolic variables QD1 and QD2, where > QD1= 2sin(x)+cos(3x) and QD2=-3sin(3x)+9cos(x). I then created a 2x2 > symbolic matrix {{QD2^2 QD1*QD2}{QD1*QD2 QD1^2}}. I now need to substitute > the contents of the two variables into the matrix so that I can evaluate the > determinant, inverse, etc. How can I do this? ŒQD1¹ DUP RCL 2 ->LIST MATCH^ @ where ^ is up-arrow character) ŒQD2¹ DUP RCL 2 ->LIST MATCH^ If you also have Erable, you can use EXEC instead: ŒQD1¹ DUP RCL 2 = EXEC ŒQD2¹ DUP RCL 2 = EXEC In the 49G series SUBST is the substituted instead of EXEC for execution of this (puns intended) There mifgt be other ways, like just EVAL to evaluate the variables (in 48) Much too lazy to test my proposals, sorry! ==== I tried the Þrst program (QD1¹ DUP RCL 2 ->LIST MATCH^). I don¹t have enough free space to use Erable. After putting the symbolic matrix on the stack and running the program, I get a list { QD1 Œ2sin(x)+cos(3x)¹ } on level one and then an error message ^MATCH error: bad argument type. > equipped with ALG48. I have created symbolic variables QD1 and QD2, where > QD1= 2sin(x)+cos(3x) and QD2=-3sin(3x)+9cos(x). I then created a 2x2 > symbolic matrix {{QD2^2 QD1*QD2}{QD1*QD2 QD1^2}}. I now need to substitute > the contents of the two variables into the matrix so that I can evaluate > the > determinant, inverse, etc. How can I do this? > ŒQD1¹ DUP RCL 2 ->LIST MATCH^ @ where ^ is up-arrow character) > ŒQD2¹ DUP RCL 2 ->LIST MATCH^ > If you also have Erable, you can use EXEC instead: > ŒQD1¹ DUP RCL 2 = EXEC > ŒQD2¹ DUP RCL 2 = EXEC > In the 49G series SUBST is the substituted instead of EXEC for execution of > this > (puns intended) > There mifgt be other ways, like just EVAL to evaluate the variables (in 48) > Much too lazy to test my proposals, sorry! ==== Looking for someone that can and will spread light on the following topic: The other day I was playing a bit with the units on my 49G+. As I converted 1_hp to W I reaslised that the relation between hp and W is another that the one I was tought back in school 18 years ago... I am sure that I was told that 1_hp equals 736_W... but now my 49G+ tells me that 1_hp equals 745.699871582_W... The Þrst question I asked myself was: What is going on here... has the deÞnition of one horse power changed within the last 18 years...??? Second I looked up in the most actual encyklopedia written in my mother-tounge (Den Store Danske Encyklop.bedi)... and see there: It conÞrmed my memory and claims that 1_hp equals 736_W... But this would mean that my 49G+ should be wrong... Even though the 49G and 49G+ have tought me not to expect everything written on the display of an HP calc to be right this was not where I was expecting errors... so looking at google I Þrst found the following page: http://www.web-cars.com/math/horsepower.html saying that the 49G+ is right... Another google search brought me to the following page http://www.simetric.co.uk/si_hij.htm which differ between several factors when converting horse power to watts speaking of boiler, metric or electric horse power... Yes I know the question is a bit OT but knowing the level of this group I am sure that somebody will help me... thanks a lot in advance. A merry christmas to all of you (who celebrates christmas) Martin P.S.: And to make this post a bit more on topic: Does anybody know why the Units Lilbrary of the HP calc only offer this one factor between horsepower and watt... ==== > Looking for someone that can and will spread light on the following topic: > The other day I was playing a bit with the units on my 49G+. As I converted > 1_hp to W I reaslised that the relation between hp and W is another that the > one I was tought back in school 18 years ago... I am sure that I was told > that 1_hp equals 736_W... but now my 49G+ tells me that 1_hp equals > 745.699871582_W... The Þrst question I asked myself was: What is going on > here... has the deÞnition of one horse power changed within the last 18 > years...??? Here¹s a Website that explains the differance: http://www.miatapower.net/w3archive/199903/msg00433.html ==== DeÞnitely off topic, but quite interesting nonetheless :-) I don¹t know what horses would have to say about it, but (please) let¹s not have an English horses v French horses battle! > Looking for someone that can and will spread light on the following topic: > The other day I was playing a bit with the units on my 49G+. As I > converted > 1_hp to W I reaslised that the relation between hp and W is another that > the > one I was tought back in school 18 years ago... I am sure that I was told > that 1_hp equals 736_W... but now my 49G+ tells me that 1_hp equals > 745.699871582_W... The Þrst question I asked myself was: What is going on > here... has the deÞnition of one horse power changed within the last 18 > years...??? Here¹s a Website that explains the differance: http://www.miatapower.net/w3archive/199903/msg00433.html ==== > DeÞnitely off topic, but quite interesting nonetheless :-) I don¹t know what horses would have to say about it, but (please) let¹s not > have an English horses v French horses battle! One could deÞne your own horsepower using ŒconstNAME¹ to store the value. WR has a niec UNITMAN program to automate unit construction. ==== The value used on the hp calcs is derived from the conversion of ft * lbf to N * m (Remember also that 1 N*m = 1 J 1 J/s = 1 W 1 N = 1 kg m / s2 ) The following hp intrinsic constant are used: g = (9.80665 * 0.3084) ft/s2 = 32.1740485564 ft/s2 1 kg = 2.20462262185 lb 1 m = 0.3048 ft and yes 1hp = 550 ft * lbf /s. So, what it boils down to is a multiplication of 32.1740485564 * 0.3048 * 0.3048 *550 / 2.20462262185 which leads to the 745.699... ==== Furthermore: You can create your own Custom Menu (CST). For instance, my Energy entries contain { Btu Œ1_Btu¹ } { J Œ1_J¹ } { ftlbf Œ1_ft*lbf¹ } { kWh Œ1_kW*h¹ } { Nm Œ1_N*m¹ } { therm Œ1_therm¹ } { |> } { kJ/kg Œ1_kJ/kg¹ } { Btu/lb Œ1_Btu/lb¹ } { |> } { B/lF Œ1_Btu/(lb*^oF)¹ } { kJ/kC Œ1_kJ/(kg*^oC)¹ } { cl/gC Œ1_cal/(g*^oC)¹ } { |> } { J/m3 Œ1_J/m^3¹ } } and my Power directory has { Btu/h Œ1_Btu/h¹ } { B/min Œ1_Btu/min¹ } { Btu/s Œ1_Btu/s¹ } { þf/s Œ1_ft*lbf/s¹ } { þf/m Œ1_ft*lbf/min¹ } { hp Œ1_hp¹ } { |> } { kW Œ1_kW¹ } { Nm/s Œ1_N*m/s¹ } } ==== > Well... I have just changed my batteries and the hex number went up to > 0x3FE... just like I said it should if it Þts my formula! I¹m gonna > track it for a while, just to make sure it works perfectly. My last set > of batteries lasted for a month! > > Now, since we know what the trigger is, we need to Þnd out how to get > at this number without having to reboot the thing. It is a real PITA to > have to ON-F (actually, the resulting font change) each time I want to > see how much battery is left. My Idea: Might the number mean that it is 3.5 + XX/256 volts, with XX > being replaced by the value of the last two hex digits? That would make > 0x3FF be right at 4.5 volts, and it also Þts my current battery voltage > (3.98 VDC). Anyone else out there with a VOM wanna test this hypothesis? Also, I guess my 49g+ has quite good battery life, since after three > weeks of medium use, it is only down to about 0x37A, probably meaning I > have quite a bit of life in it if the trigger is 0x302. If this is the > case, I am amazed because, according to many posts already, it should > ALREADY be at 0x302 :-D. >> -=[ Sat, 6.12.03 8:09 p.m. +1300 (NZDT) ]=- >> in message ID <122b2fe.0312052047.477adffd@posting.google.com> : >Pressing ON+F then 8 launches a power test which shows a hex number >in the middle of the screen and BATTERY NORMAL at the bottom of the >screen. >[...] >Does anybody know what the hex number refers to? >>I don¹t know what they refer to, but I watched them for >a while and it looked to me as if the number went down by >one about every sixty seconds. >> Same here. Since Joe mentioned it I kept an eye on that number >> - here I see BATTERY WARNING when the number gets down to: >> 0X00000302, whereas when my NiMH batteries are fairly freshly >> charged I see ...3B0. Looks like we have an indicator with a >> 256 step scale, where the last 2 digits possibly go from 00 up >> to FF. Possibly fresh alkalines would give a reading up near >> FF. this formula also works for me getting a hex number of 3AA corresponding to 4.164 reads 4.14V on my voltmeter ==== > Huh? Oh I get it, because the USB directly between the calculator and > the PC still doesn¹t work. But if they ever get it working, whatever > would I want USB to IrDA for? 49g+ has IrDA, maybe *IT* works... ==== >>If you were going to change the 49G+ manual, what would you do? > 3¹ THEN OVER + SWAP @ press ENTER here DUP 2 DISP Value 1 DISP @ this line was autointended to be in line with the previous line ==== Just one, single, unique, mono, solo, lonely sugestion: Two manuals: Algebraic Edition, and RPN Edition. I read so much stuff that I don¹t want to read... This reminds me of bilingual product labels here at home... > If you were going to change the 49G+ manual, what would you do? How about a numbered list of ideas, things missing, etc. Positive comments > only. Slams won¹t help. I already have a few ideas myself, which I¹ll share later. Get the juices > þowing and think of what should be done to help them from their current > state! Gene > P.S. I AM of the opinion that the 49G+ manuals are much better than the 49G, > but not as good as the documentation for the 48GX. > -- > * These statements and opinions are mine alone and do not reþect my > employer¹s views. * ==== > Just one, single, unique, mono, solo, lonely sugestion: Two manuals: > Algebraic Edition, and RPN Edition. I read so much stuff that I don¹t want > to read... This reminds me of bilingual product labels here at home... I would rather have all the _even_ pages showing RPN stuff and all the *odd* pages showing that odd algebraic system examples... ==== > IMO (comparing the size of port¹s content and huge sizes SD ca come > with), an option to backup entire port would have been welcome. This has been realized in a new Filer5 version 2.2004 (1685 bytes). Menu option ->SD-> preserves selective saving on the SD-card if being a port. Note that from outside a port, one can save on the card with the normal COPY, perhaps in multiselection. Therefore, ouside a port, ->SD-> Þrst politely asks whether all standard ports should be saved on the card. Answering YES starts the saving process. That may take up to a minute if your ports are full of libraries and backups. I tested it with more than 25 libraries in the ports, among them such huge libs like Emacs, SDIAG and extable. Everything was nicely saved on the card. In contrast to the standard COPY, ->SD-> does not overwrite. If an object was saved already on the card it remains there until purging it. Objects with long names (containing the space character) aren¹t tranfered to the card. Such names are created by Filer5 only in form of dated HOME backups. Not necessary to save such backups from standard ports to the card since dated HOME backups can directly be created with Filer5. Would be nice if Filer5 users would test this amazing function which permits a return of the 49+ to the dealer, keeping the SD-card at home http://page.mi.fu-berlin.de/~raut/WR49/index.htm#General ==== I need help. I didn«t get splitter the periodic table of the Marcelo in my HP49 g+. I am using OT49 and it show insufÞcient Memory. What«s happening? ==== Is easy you can graph y = root(-x1^2 +1) its the same eq ok ;) ==== You should put in the eq : {Œ(1-x^2)^0.5¹ Œ-(1-x^2)^0.5¹ } In that way you draw both the + and the - of the function. > Still, I think because of resolution problems you won¹t see a full > circle. In the horizontal quadrat it doesn¹t touch the x axis. Don¹t > know why. Bye Idan If you set the resolution to 1 pixel, you get the full graph here and with CONIC graphs, but with the default setting, you may miss a vertical line of pixels at each end of the graph. But you lose speed in graphing. ==== Are you there ?? ( ? ) -- Frank Bachman Grumpy Aero Guy > I¹m writing to ask if someone would like to help me taking my library in a > website dedicated to HP calculators until hpcalc.org will be updated. > This because I don¹t own a website of my own and I¹m feeling a bit strange > developing new versions of the library and uploading it on hpcalc.org > without results. :oD > intermediate versions, since he will put only the last one on the site when > he¹ll have time to do the upload... ==== What happen whit emulator for hp49+? ==== I just got my new HP49G+ It is already running ROM 1.22. The keyboard is Þne, at least compared to my now old HP49G. Well, after 15 minutes my index Þnger that I usually use with calcs became painful, so I changed for the thumb. Olso, once a key didn¹t register. A part of that, it¹s great. HP, please count me (back in your happy customers ! ==== I posted lately my review about my hp49g+. I am not sure but I think I have what you call : unregistering keys. Sometimes, when i press a key not at its middle but in its left or right upper corner, the key clicks but nothing happends. Is that what you mean ? I didn¹t quite understand the solution that was suggested. I tried to push the button like I do normally but very Þrmly this time. so that the button will do all the way inside. then, I moved it from side to side in a rubbing way.. I haven¹t noticed a change. Can you please be kind and explain to me again the solution ? If I press in the center - Usually there is no problem, but it is very annoying to look all the time if I pressed it well or not. Please help cause I have a bad feeling that I will make mistakes in tests and it will cost me... I don¹t want to think what is going to happend to this poor calc if it will fail me in a test.... Please Please Please help and explain again, those of you who tried and succeed the solution. ==== > Did you know that there is even an implementation on TI calcs, RPN Calc > v1.01 : > http://www.ticalc.org/archives/Þles/Þleinfo/103/10381.html IMHO, Lars Frederiksen¹s RPN is much better: http://www.paxm.org/symbulator/download/rpn.html Besides Doug¹s extensive documentation, the program itself is more functional. -- > Don¹t you read this NG? :) Obviously I don¹t... Actually I did not assosiate your name with the only known failure... Too much post to read (and it¹s past midnight local time again) zzzzzz.... > I tried this, it didn¹t work on mine and may have made the backspace > key worse, and I posted it already... Pete: did you get a replacement unit from HP? How are those keys? ==== I have recently purchased a 49g+ and I would like to know what are the best Þles to add to it to make it purr like a kitten in my electrical engineering and physics courses. I have had and used a 48GX for years and still use it, and I have purchased a lot of the plug in software (i.e. Spice 48, EE pro, Physics pac and others), but I would like to retire it and go with the 49. I have read this NG for a year or two, but I am confused as to what would be the best software to add. I know a lot of you either teach, or work in these Þelds, and I am hoping that you can help. ==== IF YOU DON¹T OPEN IT, HOW CAN CURIOSITY BE CURED?? The hp49G+ was made to be opened, it is very easy. Open your battery compartment. You will see two round plastic rivits. Use a razor blade and cut them off þush. They can be easily replaced with screws later, and are not even mandatory. Next use a small knife, or Þngernail, and seperate the dark plastic band from the bottom half of the calculator. Do this starting up near the USB port, and working your way around the calculator. You will probably notice a slight gap in the trim piece up near the USB if you look closely. The calculator is snapped together during assembly, with the bottom half snapping to the top half. I have had mine apart 20 times and the snaps still grip and release like new. There is nothing more to opening the calculator. With dissambly this far, you can install a switching type DC power jack (I have), and even connect a LED (I have)in parallel with the pizo buzzer. This is also all the further you need to go if you are one of the people that bent a contact inside of your battery compartment, and you want to make a repair. Also, if your SD card socket would ever wear out from heavy use, you can order the part on the internet, and solder it in yourself. I absolutly love my hp, and it seems to me that someone took some well planned design considerations regarding repairability. These units could be repaired by most any dedicated amature, and most deÞnatly anything could be handled by the factory. But quite honestly... at $150.. just buy a new one. If you desire, after this stage of disassembly... there are many more rivits that you can razor blade away if you want to see the other side of the board!!! ==== > Does anyone know how to open a HP 49G+? Carefully drill out the melted-over heads of the two gray heat stakes inside the battery compartment. (Don¹t use a power drill -- I used a counter-sink bit with a sharp ~3/16 bit mounted & twisted by hand.) Pry out the IR lens with a small screwdriver, working gently across the USB connector opening. Then, working in one direction or the other, carefully pry the gold bumper strip away from the black BACK case. (The bumper is part of the front case -- the black upper edge is applied and easily damaged -- don¹t pry against that.) You¹ll Þnd pictures of what to expect inside at the Museum of HP Calculators Articles Forum [Article # 408 -- ] ==== >> What character set and Þle encoding must be used so that a text Þle >> can be written in notepad or word pad, and saved to the 48/49 without >> having to go in and edit the special characters? ie. sqrt and pi I >> know the ascii codes, and I can save as a text and encode as Some 49 hints: >1) [APPS] 2.I/O functions..[ENTER] 5.Transfer..[ENTER] Xlat: |CHOOS| >*note: the number on the left are the corresponding codes I put there* >0 None >1 Newline (Ch 10) >2 Chr 128-159 >3 Chr 128-255 >Select the last one >You could also do on the command line: >HOME ŒIOPAR¹ 6 3. PUT >2) >This goes for all models: >http://www.Þnseth.com/~Þn/hpdata/chars.html Some 48 hints: >1) >SX > [Left-Shit] [I/O] |SETUP| > I/O setup menu >IR/wire: wire >ASCII/binary: ASCII >baud: 9600 >parity: none 0 >checksum type: 3 >translate code:0 > To change the last one press [ 3 ] |TRAN| > For explanation look the codes on the GX example > eg. 0->None; 1->Newline, 2->Chr 128-159; 3->Chr 128-255 >GX > [Right-Shift] [I/O] Transfer..[ENTER] XLAT: |CHOOS| >*note: the number on the left are the corresponding codes I put there* >0 None >1 Newline (Ch 10) >2 Chr 128-159 >3 Chr 128-255 >Select the last one 2) Character Translations > SX: Chapter 33 (Volume II if 2 volumes) > GX: Page 27-16 HP48 G Series User¹s Guide Finally: (By John H Meyers) >http://membres.lycos.fr/ekalin/view_category.php?id=1 >How to Use the Backlash Sequences to Create Files in the PC > unicode, the formatting will work on the pc, but the calculator >> doesn¹t recognize it. I¹ve loaded a HPGCalc.ttf into my font >> listing for windows, so I can type using the correct type set....just >> fails when I upload. I¹ve already tried DOS text and windows text..no >> luck there either. >> I can¹t help you with a HP48/49 (don¹t have one) but I think you¹re >> using my HPGCalc.ttf font. That font¹s Þne, and as you say you¹re >> saving in Unicode, but the Þles still need to be translated. (If it¹s not my font ignore me) That font is designed for the HP38/39/40G calculators and special >> application software to convert from/to the calculator format and >> Unicode. I don¹t know of such software for the HP48/49 but that doesn¹t >> mean it doesn¹t exist. Also, I don¹t think ANY of the calculators would natively support >> Unicode so that probably will never work. I don¹t think there¹s a >> matching encoding unfortunately. ==== > That works for standard UserRPL.... > What about SysRPL/Asm? > You need the source and that¹s ASCII by default when i have the program in hp Þles menu, this is a string type Þle and when i open it starts whith: hphp48.......... dir. and after each line command there is a square point. ==== Exellent library. Just a little thing that would make it better - HDF doesn¹t act as DISP, 22 chrs long string will not Þt the display as expected. We have created a new head manager. This 2-page library is only for > the HP49+ and makes it easy to use the new header or to integrate it > into the display for various environements. HeadMan is perfectly > usable but still under work. Note that some commands may still be > modiÞed in future. > X > http://page.mi.fu-berlin.de/~raut/WR49/Headman.htm > X > http://page.mi.fu-berlin.de/~raut/WR49/index.htm#General > ==== > Exellent library. > Just a little thing that would make it better - HDF doesn¹t act as DISP, 22 > chrs long string will not Þt the display as expected. > http://page.mi.fu-berlin.de/~raut/WR49/Headman.htm > http://page.mi.fu-berlin.de/~raut/WR49/index.htm#General HDF (Header Freeze) is similar to PROMPT but doesn¹t suspend (there will be no HLT indicator). To use the 2nd header line the input string should contain in both cases a linebreak before the line gets too long. I added this information to Headman.htm a few minutes ago :-) ==== > I¹m thinking about submitting a program to HPCalc.org. > It will require a library, but I don¹t want to use a library number for > anything that¹s already out there. Is there a list of current libraries, and if > not, how can I be sure that I won¹t be using a number that¹s already been used. > you did not say which calculator, but there are lists on hpcalc.org for HP48 and HP49. I am not sure when these lists have last been updated, but they should be sort of complete... Check out http://www.hpcalc.org/details.php?id=4532 http://www.hpcalc.org/details.php?id=4533 ==== sorry to chase but I still have the same issue (I installed the font you told). The problem seems to be related to a HP48GOS1 font... > I have updated the web page to include: * Word and zipped Word version of the instruction document > * hp49font.zip - a Þle with hp 49 fonts I hope this solves the problem of the missing fonts. ==== detachable 48, I¹d be willing to buy something like what you are > describing ... or at least try it out. You bring up another interesting idea based on a slight alteration of my original concept. Rather than sawing up a G series unit (dead or otherwise - did any dead G series unit die a natural death??) as a donor for the keyboard/PDA project, perhaps a different project exists for a docking module that has legacy support? Instrument interface, secondary color display, deep archive (when is enough memory enough?), <Þll in your need>. Prior to the 48 series, HP had a dizzy number of gadgets and items available for their calculators. All necessitated by the lack of cheap large memory and cheap large PCs in that day. I do not propose going that route (way to much wire and gadgetry to go wrong) nor does the eXpander/TI 200 seem like an ideal direction. Essentially the intelligent choice comes down to the correct tool for the job. My table saw is Þne by itself. But if I want to do tenon joints, I need a tenon jig. Dados are best with my dado blade set. The correct tool for the job/need as determined by the worker. I admit to remembering the days when people lamented over the feel of their slider and quality of the line on slide rules. [sigh] Of course, you could not plot anything other than a straight line with that tool. [grin] ==== > This would appear to be the future for the professional world of > number crunching. Any thoughts? I try to be as open-minded as possible, and I¹ve tried a whole slew of > handheld devices; for most purposes, when I need to make calculations, I > reach for my 41CV. I have yet to Þnd a more efÞcient device for getting > the job done. Just the same!!! I have a PDA, powerfull, color, memory, etc, but when I want to do my Field Work (no electricity in rainforest) I use my HP, back in the ofÞce, the DeskTop. Finally I am going to chage my PDA for a brand new HP49G+ very soon. ==== [Blush] Right. I was trying to describe a general shape, not a particular function. I¹m too often typing myself into verbal corners from which I don¹t quite escape. Must read carefully and revise! In fact, I actually was at the hobby shop just yesterday, looking for a plastic model kit for a Œ57 (Œ59?) Cadillac. (Couldn¹t Þnd one.) That image of an HP-49G+ with tail Þns simply MUST be realized somehow . . . ==== > the secret apparently is that one has press both [ON] and [F3] for a > LOOOOOOOOONG time. Then the library appears with > [Right-Shift] [ 2!!!!!!!!! ] eg. [LIB]. You may have experienced a case of an unrecognized key. Some keys have to be pressed harder or they won¹t be read - it¹s a well known problem. ==== > Then press Þrst [ON] keep it down and press [F3] release both The time does not need to be long as long as you keep the [ON] pressed along with the [F3] - longing to get my long lost HP-15C back before long > After that you Þnd the new library using: > [Right-Shift] [ 3 ] eg. [LIB] > and thanks. It worked even without Filer. Once 360 appears in port 2, > the secret apparently is that one has press both [ON] and [F3] for a > LOOOOOOOOONG time. Then the library appears with > [Right-Shift] [ 2!!!!!!!!! ] eg. [LIB]. That one really has to press *FAIRLY LONG*, should be stressed by > Rautenberg. I now have the OT49+ on both of my HP49+¹s. ==== I saw some references to rom 1.23 in some messages. Another message indicated that this would be available some time in January to the general public. Patience is a virtue, but I am a sinner. Some people have it included in new calculators. Is it possible (and legal) for this to be downloaded by these lucky þicker-deprived individuals, and then published somewhere on the web? ==== > I saw some references to rom 1.23 in some messages. Another message > indicated that this would be available some time in January to the > general public. Patience is a virtue, but I am a sinner. Some people > have it included in new calculators. Is it possible (and legal) for > this to be downloaded by these lucky þicker-deprived individuals, and > then published somewhere on the web? If allowed and legal I¹m willing to. However, I have no idea how to dump a 49+ rom. -- ==== >>http://h20000.www2.hp.com/bizsupport/TechSupport/ SoftwareDescription.jsp?l > ocale=en_US?=English&pnameOID=351776&prodSeriesId=33568& prodTypeId=215348&ba > sePartNum=COL4344&locBasepartNum=ca-14082-3&os=Microsoft+ Windows+XP&tech=App > lication > > This is the new version of Con4x with help Þles and ASCII > support. It is 1.5 mb and works much better than previous. Also includes the new USB package. Ignore the old USB package on the web > site. I¹ve been watching the home and home ofÞce support page for a > revised Conn4x or USB driver package, but it hasn¹t shown up there; only > under business support. It works on my PC. > Finally! 8-D HELP HELP HELP I can not get this to install. it extracts but does not tell me where. i have searched all hard drives with no luck. i double clicked the downloaded Þle it takes all but .5 of a second and is done. any help would be greatlly appriciated. thank you. TW ==== if that is the problem on my machine or not, since drive d is a > separate physical drive, and the install worked on drive c, just Þne. The original Conn4X resides on the d drive, and I just sort of Þgured > the installer would upgrade existing Þles. I also have hp ofÞcejet > printer software on drive c. Maybe you have the answer though, I¹ll (or can I just call you 22/7 for short) >i had a little trouble getting it to go into a w98 pc, turns out I had two >different directories called Hewlett Packard in the c:program Þles >directory, because I also have an HP printer and jetdirect card as well as >the previously installed calc connectivity software, and this mightily >confused the uninstaller that tries to strip out the previous installation. >Once it was all tidied up into hewlet~1 though it all works Þne >C My PC has one HD: 120GB IBM The Multiboot PC has WinXP on logical drive J: where the conn4x nicely went - just like before ==== I have a problem with my hp49g+. When I want to start a program that I have transfered, I get a warning message : UndeÞned XLIB Name What does it mean ? What must I do ? Where can I Þnd this library ? ==== There ars so many sd card maker. Is there any list for the compatible brand? ==== > When I set Connect > the program says Looking for Calculator for a few seconds > and then says: I assume you put the calculator in server mode? I know it does that if you don¹t. ==== > I assume you put the calculator in server mode? I know it does > that if you don¹t. WELL Its very interesting..... Sometimes when activating the Server Mode, the HP49G+ says: Awaiting server Cmd when it does, Connectivity Kit can«t work. But sometimes, when I play with the IOPAR var, NO CHANGES, JUST manipulate it gets into: Xmodem Server, Waiting for commad and IT WORKS !!!!. What I really don«t know is when I can do the HP49G+ to turn on the XMODEM SERVER. Is it related to the IOPAR ? I really do not change anything. Is someone else having this trouble, or its just because I am learning how to? ==== > Sometimes when activating the Server Mode, the HP49G+ says: > Awaiting server Cmd That means that you held down the shift key while pressing the right-cursor key. It¹s not XMODEM Server mode; it¹s Kermit Server mode (same as obtained by the SERVER command). > when it does, Connectivity Kit can«t work. Correct, because the Kit requires XMODEM Server mode. > But sometimes, when I play with the IOPAR var, NO CHANGES, JUST > manipulate it gets into: Xmodem Server, Waiting for commad That means that you released the shift key before pressing the right-cursor key, and obtained XMODEM Server mode. > What I really don«t know is when I can do the HP49G+ to turn on the > XMODEM SERVER. Don¹t hold down the shift key. Just press it and release it before pressing the right-cursor key. Hope this helps! ==== > What I really don«t know is when I can do the HP49G+ to turn on the > XMODEM SERVER. Is it related to the IOPAR ? I really do not change > anything. Is someone else having this trouble, or its just because I am learning > how to? The biggest problem is that HP doesn¹t give clear instructions on how to get this to work, and I think the USB drivers on the provided CD are buggy (they BSOD¹ed my Windows 2000 laptop on numerous occasions). Ok, for starters, go in and remove the connectivity kit you installed from the CD and visit http://www.hp.com/go/calculators/ and download the latest build. You¹ll need to do this anyway if you want to upgrade your calculator¹s OS to the most recent version (which yes, you do want to do that eventually). The new build has better USB drivers and, more importantly, accurate instructions on how to set your calculator into the proper server mode. I was in the same position you¹re experiencing until I was told by the lastest connectivity kit that I didn¹t want to hold the right-shift button and then press the right-arrow key (which is the way of the old 49), I wanted to simply right-shift and then right arrow. The difference is that the former puts the calculator into kermit mode, and the latter puts it into xmodem server mode, which the connectivity kit wants to see. Once you do that, you¹ll be in business. ==== > I assume you put the calculator in server mode? I know it does > that if you don¹t. WELL Its very interesting..... Sometimes when activating the Server Mode, the HP49G+ says: > Awaiting server Cmd when it does, Connectivity Kit can«t work. When you hold the shift key down > But sometimes, when I play with the IOPAR var, NO CHANGES, JUST > manipulate it gets into: Xmodem Server, Waiting for commad and IT WORKS !!!!. When you release the shift key 1st > What I really don«t know is when I can do the HP49G+ to turn on the > XMODEM SERVER. Is it related to the IOPAR ? I really do not change > anything. Is someone else having this trouble, or its just because I am learning > how to? ==== Do you know where I can Þnd a list of all the serial port commands? And how do you select the IR port instead of the serial or USB port (using UserRPL)? Thomas Tuttle > The AUG says that in the Xon/Xoff mode of SRECV the calc will send an XOFF > signal when the receive buffer is almost full. What does almost mean here? > Does it send the signal before or after the 255th byte starts loading? In the case of the 48SX or 48GX, and presumably the 48S, 48G, and 48G+: With receive pacing enabled, the XOFF (ASCII control code 19 decimal) > has to be sent before the input buffer is full (255 bytes), so that > the other device has time to respond to the XOFF and stop sending > before the calculator¹s input buffer overþows. If I recall correctly, > the XOFF is sent if the calculator has at least 223 bytes in the input > buffer and receives another one. After an XOFF has been sent, when the > input buffer has less than 25 bytes in it, the calculator sends an XON > (ASCII control code 17) to tell the other device to resume sending, You can Þnd out the exact behaviour by using a terminal (or terminal > emulator), or even another calculator (that uses XON/XOFF þow > control), to send bytes to the calculator and paying attention to when > BUFLEN stops telling you that more bytes have accumulated (don¹t > remove bytes from the buffer with SRECV just yet). For example, send > 220 bytes to the calculator, and then send bytes one at a time, using > BUFLEN after each one to see how many bytes are in the input buffer. > When the value from BUFLEN stops increasing, you¹ll know that the > calculator must have sent an XOFF. Then use SRECV to remove bytes from > the input buffer; for example, remove, let¹s say, 195 bytes, and then > remove bytes one at a time, using BUFLEN after each SRECV to see how > many bytes are in the input buffer. When the number of bytes in the > input buffer stays the same or increases after an SRECV, you¹ll know > that the calculator must have sent an XON. But be careful that the > other device doesn¹t time out while you¹re experimenting. I don¹t know of a way to adjust these trigger points on the > calculator, but perhaps it could be done with SysRPL or assembler > language. But for everything that I¹ve tried it with, the built-in > XON/XOFF behaviour works þawlessly. The only problem is that you have to be careful not to have the XOFF > or XON signals in the data. If they might occur in the data, then > you¹ll have to implement some sort of encoding/decoding scheme to get > them through disguised as something else. With transmit pacing enabled, for XMIT and the various printing > commands (when printing via wire), the calculator stops sending when > it receives an XOFF, and resumes sending (assuming that it¹s waiting > to send anything) when it receives an XON. I don¹t know how long it > takes for the calculator to get around to sending these signals, but > the other device should assume that it won¹t be instantaneous. The above applies to a connection via wire. XON/XOFF þow control > doesn¹t work with Serial IR because that¹s half-duplex. Full-duplex > isn¹t used with IR because reþected IR signals would be > misinterpreted as signals from the other device. Note that with the Kermit protocol, the input buffer should never have > enough bytes in it to trigger an XOFF response. Presumably, the same > would be true of most (all?) other devices that Kermit might be > communicating with or through. I¹ve never experimented with whether > either receive or transmit pacing has any effect when using Kermit or > Xmodem. In the case of the 49G, XON/XOFF þow control doesn¹t work. It doesn¹t > send XOFF or XON signals when receiving, nor does it heed the XOFF when > transmitting. I expect that the 48gII will behave like the 49G, and if the 49g+ can > use the serial I/O or printing commands (other than sending to the > 82240A/B printer in Red-Eye mode), then I guess that it will behave as > the 49G does. ==== > Do you know where I can Þnd a list of all the serial port commands? > And how do you select the IR port instead of the serial or USB port > (using UserRPL)? -33 SF ==== > But the best which could be done has now been done. In the back-transfer > from the card, you may choose in a browser the port to which the > selected Þles have to return. Also other small details in Filer5 have > been modiÞed (but no version change). > Wow! Why not a version change?! There was still a small bug in the yesterday version of Filer5. Hence, I updated the Þxed version to 3.2004. There is still a little bug which steams from the buggy operation system ROM 1.22. If a running Þler is suspended so that the HLT indicator is seen in the new header, HLT remains there after return to the Þler with the CONT key. This clearly confuses the user. I don¹t make bug reports anymore because in all the years I never got a single word of thanks from JYA or CdB. http://page.mi.fu-berlin.de/~raut/WR49/index.htm#General ==== Check: http://www.engineering.usu.edu/cee/faculty/gurro/Software_ Calculators/HP48_4 9G_Docs/ProgrammingInThePC.htm > I am wondering if some program exists, for pc that could safe the health > of my poor eyes, and my patience and mood when I write this crap in > user-rpl. > I have just dev4x, but it doesn¹t seem good: > I can¹t understand how to write or edit any program! > I remember there was something like an editor with a few character > substitution for special charactes, like programming brackets, but it was > a lot of time ago, developed for hp48sx. ==== > Richtiges Klicken [F1 EIN] [DEL], Gleichungen zu l.9aschen. Richtiges > Klicken [F1 EIN] [HINZUF.86GEN] zu hinzuzuf.9fgen neue Gleichungen., Sie. > >wer kann mir erkl.8aren wie man eine kurvendiskussion mit dem hp49g >machen kann und wie man die komplette history des plottens l.9aschen >kann? obwohl ich die history l.9asche, k.9annen die alten plotfunktionen >trotzdem aufgerufen werden. sorry aber das geht nicht, wenn du das ganze auf dem weg l.9ascht, (wo das .9fber umschaltung + clear, wenn da mehr drin steht schneller geht) steht trotzdem unter choose die alten funktionen drin und genau diese m.9achte ich l.9aschen. also kurz, ich m.9achte nirgends eine funktion mehr haben, wenn ich alle gel.9ascht habe. ==== Wenn Sie die Gleichung zu einem var Sie sparen, k.9annen zur.9fckrufen es sp.8ater. Beispiel: X^2 + 2X + 4 ŒEQN¹ STO (RPN) Dann sp.8ater: RECHT KLICKEN EQN und f.9fgt es Ihrem Grundst.9fck hinzu. >> Richtiges Klicken [F1 EIN] [DEL], Gleichungen zu l.9aschen. Richtiges >> Klicken [F1 EIN] [HINZUF.86GEN] zu hinzuzuf.9fgen neue Gleichungen., Sie. >> >>wer kann mir erkl.8aren wie man eine kurvendiskussion mit dem hp49g >>machen kann und wie man die komplette history des plottens l.9aschen >>kann? obwohl ich die history l.9asche, k.9annen die alten plotfunktionen >>trotzdem aufgerufen werden. sorry aber das geht nicht, wenn du das ganze auf dem weg l.9ascht, (wo >das .9fber umschaltung + clear, wenn da mehr drin steht schneller geht) >steht trotzdem unter choose die alten funktionen drin und genau diese >m.9achte ich l.9aschen. also kurz, ich m.9achte nirgends eine funktion mehr >haben, wenn ich alle gel.9ascht habe. ==== >>>I believe the 49G+ has a 16 bit processor while the 89 is an 8 bit. >>The 49g+ has a 32-bit CPU which emulates the Saturn CPU, >>which is a 4-bit serial computer with 64-bit registers, 20-bit addr >>>(subtract 1 from 0 when in hex mode and count the F¹s) I don¹t how >>>useful this is for you, but the TI¹s can¹t do large number >>>computations (such as 1000 factorial). >>The MC64000 used in the TI is a 16/32-bit CPU >>The registers are 32-bit, but I¹m not wheter the addr is 24 or 32 bit? It uses 24 bit addressing For a calculator who¹s memory space is measured in KB, this leaves lots of headspace. > www.nyall.net ==== > It uses 24 bit addressing For a calculator who¹s memory space is measured in KB, this leaves lots > of headspace. BUT there seems to be a lot of 64KB restrictions on the TI AND the Saturn(+) on the 49G(+) needs bankswitching. The rest of the world seems to be eventually going to 64-bit addressing so it¹s time to give up emulating and go for a simulation and full 4GB maximum memory for the calculators I believe WinCE.NET has a 256MB limit for the PDA¹s. Why all these Gates in our memory for the computing devices? ==== > If I recall, the address is 32-bit internally but the 89 only uses 24 > bits for its memory, so the most signiÞcant byte is ignored and/or > zero. The most signiÞcant byte is ignored. To be precise, all motorola 68000 chips use 24 bit addressing. This is not something speciÞc to the 89. ==== > Actually, the 89 keeps integers and þoats as different data types. > Integers never have decimal points and never have exponents--if you do > something huge like 2^32, it shows all the digits, or inÞnity if it > overþows. Sometime I will Þnd the highest integer it can hold... > stay tuned. The CAS of the 68k calculators use a variable length format format for integers. Integers can range from 0 bytes to 255 bytes long. As a quick example: the CAS of the ti89 can do 299! (factorial) but not 300! On the home app string(299!) displays about twice as fast 299!. I suspect that the ti89s slow integer to string conversion is being used twice when 299! is calculated. (the Þrst time when the calculation is done and the dimensions of the object are needed when adding the expression to the list of entry/answer pairs. The second time when it is actually being displayed. ==== > Even an > 4-bit processor can calculate 1000! if the software designers represent > numbers internally to allow such a thing. > But if it has no way to store the answer, then it doesn¹t work... It stores the answer in as many wo (of whatever size) as is necessary. A 32-but machine could store a single-precision þoating point number in one word while a 4-bit machine would use 8 wo to store it. The size of the CPU registers does NOT limit the size of the data you can work with. I it had approximately the same dynamic range as modern day C++ on a 64-bit processor. Of course the 32- or 64-bit processor would be much faster at it (especially since they have a þoating point processor built into the hardware) but the 4-bit machine could get the job done. ==== > Because of that and thanks to the work of its authors,the HP49G+ CAS > is a more powerful and more complete CAS. > Sweet! > You can fully customise the calculator from the keyboard to the stack > size. > You can take screenshot of the screen. > You can redeÞne the virtual size of the graphic screen. > And last but not list you have access to other applications such as > the Equation writer,the matrix writer,the text editor,etc... during > the execution of a RPL program. > Sweet! > Sweet! > Sweet! > Something completely impossible on the TI89. > X > Did I mention I think the HP-49G+ is sweet? > Better see your dentist, soon...(-; Oh, don¹t worry, I¹ve already had enough christmas candy to outweigh that. > I like very much of the task switching ability of the 49G OS - sweet... ==== X > And last but not list you have access to other applications such as > the Equation writer,the matrix writer,the text editor,etc... during > the execution of a RPL program. X > I like very much of the task switching ability of the 49G OS - sweet... Manual Task switching = You can start an application witheen an application Like start Matrix Writer then use Equation Editor and press [HIST] to use the Stack items (or use COPY & PASTE) etc... You may naturally |EDIT| a subexpression in the EQW and... I think you got it! (These are not usually possible on a TI89/V200) Also you can suspend even your stack by HALT Then you can CLEAR and start another calculation, maybe save the result(s) Finally you do CONT and [UNDO] and continue the calculation that was interrupted ==== > I want see like this, 123,456.1235 for the number 123456.12345 It is obvious that FIX alone doesn¹t solve the problem. Hence, a little Mini Challenge for the New Year¹s Eve: Write a smallest UsrRPL program NRD (Nice Real Display) which, applied from STD mode to a real which is not too large (should not contain the exponential character E), yields the wanted display, no matter how many decimals the real has after the separation Here an error protected 37.5 bytes solution in SysRPL: :: CKREAL MINUSONE OVER BEGIN SWAP#1+ SWAPDUP %FP %0<> WHILE %10* REPEAT DROP DOFIX ; Roughly speaking, Þrst a suitable index for FIX is computed. The above is very fast on the 49+. Maybe somebody gets a better solution :-) ==== >> I want see like this, 123,456.1235 for the number 123456.12345 It is obvious that FIX alone doesn¹t solve the problem. Hence, a little >Mini Challenge for the New Year¹s Eve: Write a smallest UsrRPL program NRD (Nice Real Display) which, applied >from STD mode to a real which is not too large (should not contain the >exponential character E), yields the wanted display, no matter how many >decimals the real has after the separation > NRD: << DUP FP STD ->STR SIZE 1 - FIX >> ==== I¹m just moving from using a blue HP49g to a gold 49g+. I¹m having comms problems getting Þles out of my old calculator and onto the new one. Can I restore a save from the HP49g to the new HP49g+? I¹ve tried saving directories from the 49g onto my PC and then copying them to the new machine. But on the new machine they appear as big strings (containing all my variables) as opposed to being a directory. Anyone know how one is meant to copy directories around? How do I convert an ascii string into a directory? They start with %%HP: which I though was enough info to tell the calculator how to deal with the data. ==== > I¹m just moving from using a blue HP49g to a gold 49g+. I¹m having > comms problems getting Þles out of my old calculator and onto the new > one. Can I restore a save from the HP49g to the new HP49g+? If you have a good Conn4x connection, you should be able to do a Backup to the PC (Edit / Backup) using the serial communications. Then a Restore to the 49G+ (Edit / Restore) using USB. I do this all the time for testing and it works Þne. This does not deal with the port memory though, just the user variables. YMMV, I would be interested in knowing anything that can¹t be handled this way. ==== Edit the big string, delete the %%HP... line and if there is some garbage at the end of the big string clean it up too. Then OBJ-> should do the trick. I don¹t know how you¹ve generated the big string, I did it by putting the whole directory in the stack, ->STR, store it as a variable, send it to PC, then to HP49G+ etc. BTW are you using the new version 2.1 build 2118 of the communication software? I¹m just moving from using a blue HP49g to a gold 49g+. I¹m having > comms problems getting Þles out of my old calculator and onto the new > one. Can I restore a save from the HP49g to the new HP49g+? I¹ve tried saving directories from the 49g onto my PC and then copying > them to the new machine. But on the new machine they appear as big > strings (containing all my variables) as opposed to being a directory. > Anyone know how one is meant to copy directories around? How do I > convert an ascii string into a directory? They start with %%HP: which > I though was enough info to tell the calculator how to deal with the > data.