2046. 225d E1h 4 { 41d} {29h} 612. 228d E4h 4 { 8d} { 8h} 156.5 230d E6h 4 { 13d} { Dh} 205. 256d 100h 23 { 6d 2d} { 6h 2h} 261. 257d 101h 4 { 24d} {18h} 427.5 788d 314h 4 { 1d} { 1h} 30. Note that the above is for only built-in ROM libraries, it doesn't include any external libraries, not even the equation library or periodic table library. 1-dimensional arrays and the array with the 2nd dimension being 1 are English language only, arrays with the 2nd dimension of 2 are English and French, and arrays with the 2nd dimension of 3 are English, French, and Spanish. If the calculator's LANGUAGE is set to French or Spanish and the particular library's message table doesn't have that language, then the English message is used instead. The library message tables are implemented as arrays of strings. Of course an array starts with the prologue address for it's own type, and then its total body length, but it also includes one copy of the prologue address for the type of elements that it contains, which are bodies of that object type, not complete objects per se. It also contains a count of how many dimensions it has, and the size of each dimension. A type 4 array is the normal (albeit not necessarily numeric) array, with all elements present and in order. A type 23 array is a linked array, with a series of pointers to the elements. The pointers must be present and in order, but the elements can be in any order, more than one pointer can point to the same element (object body), elements need not be present, and a null (00000) pointer is used for any missing elements. I'll probably dig into the arrays a bit more; I'm curious about those null messages. I believe that I see a way to convert a linked array of strings to a list array of strings, and it should be just a matter of actually writing the program. I'm assuming that the SysRPL commands for working with arrays are designed for normal arrays instead of linked arrays, but if I start out by using ->H on a linked array, I expect that I can deal with it. Note that a type 29 symbolic matrix is rather different from these arrays; It's composite object, and very much like a list array. > What I was kind of hoping was for a list somewhere of the indirection > pointers. In other words, I was expecting that somewhere in ROM was a > packed list of messages, and elsewhere would be a table that says > Message 12345 is at offset abcd in the message table. So I could > just page through the table looking at messages which actually exist. What you could do is store a plain text file of the messages and their addresses as a string on the calculator, perhaps on an SD card. As long as you know at least part of the message that you want, you could use the editor's FIND operation to search for it. -- James === Subject: Re: Error messages are useful strings -- which? <204f1$45759d75$4267eab7$10037@123.NET> Yes, my email is correct (not my main one, obviously, but a real one I check every day). Irl > First off, a correction: In the ASCII transfer header for that > long program for getting a range of messages, I reversed the colon > and semicolon. It should be %%HP: ;, not %%HP; :. I copied and > pasted the source code as transferred from my 49g+, but typed in > the transfer header, counted string delimiter, and comments, and I > got those reversed. > > Also, now that I've thought about it a bit more, for those message > numbers that are in a valid range but DOERR shows nothing after > Error:, and ERRM, GETTHEMESG, and JstGETTHEMSG return an empty > string, I think that it's more likely a matter of the linked array > having a null pointer and no element for that message, rather than > containing an empty string. I surmise that those commands return > an empty string for any non-existent message. > > James -- > SysRPL program (as a command-line entry, presumably) > > You're welcome. > > Either transfer the source code string or key in the source code > in the command line within a character string. You should have the > source code string in level 1 when you do the ASM command. > > Note that the source code string has to end with an @ character > all by itself as the last line. > > Note that most SysRPL commands don't do any argument checking at > all, so a buggy SysRPL program can corrupt memory and you could > lose everything that you have stored in home and it's > subdirectories and port 0. If you have anything that you don't > want to lose, then archive it first, at least to a port other than > 0, and preferably to an SD card and from there to your PC. > > The intention was that you could copy and paste the source code, > starting with the %%HP: ; transfer header and through the @ line, > to a plain text file, transfer the file to the calculator, recall > the source code string to the stack, and execute the ASM command > on it. > > to actually be a > program was useful; I have tried displaying the SysRPL stack (-85 SF) > with a tiny UserRPL program on the stack and then typing in the same > thing but couldn't get it to actually be a program. > Here's a presumably dumb question: there are two extable files at > hpcalc.org. One is for the meta kernel and is much shorter than the > other one. What are these, please? > > The Meta Kernel was written for the 48 series, and it, along with > some other very good applications written for the 48 series, was > incorporated into the system for the 49 series. I wasn't aware of > an extable library for the Meta Kernel at hpcalc.org, but I'd > assume that it's for the Meta Kernel as designed for the 48 > series, not for the 49 series. > > For the official HP extable for the 49 series with supported > entry points only, use the one at: > http://www.hpcalc.org/details.php?id=3245 > > For a better extable that includes unsupported but stable > entry points, use extable2. I don't find extable2 as a separate > downloadable file, but it's included in the Emacs package; see: > http://www.hpcalc.org/details.php?id=3940 > or: > http://staff.science.uva.nl/~dominik/hpcalc/ > > Regarding the original question: I've run my > exhaustive-search-for-all-messages routine (the IFERR n DOERR THEN ... > with n looping from 1 to infinity) and generated a list of everything > below 100000 decimal = #186A0h, but it does take a while. > > Then you should have all of them except #31401h, No Message > here. > > I guess I > could take advantage of your info and make my program advance by > multiples of #100h and skip to the next multiple if there is no message > at M*#100h+1. > > That should work, as long as none of the message arrays have a > null first element (and none of them actually do). > > What I did was first run a quick-and-dirty SysRPL program to look > for message tables: > > %%HP: T(3)A(R)F(.); > !NO CODE > !RPL > * Uses 1 unsupported but stable entry. > * Takes about 35 seconds to run. > * Checksum: # 2FDFh > * Size: 113. > :: > CK0NOLASTWD > NULL$ > # 800 > BINT0 > DO > INDEX@ > DUP > #>$ > CHR_d > >T$ > OVER > #>HXS > HXS>$ > CDR$ > &$ > APPEND_SPACE > SWAP > PTR 08130 (GETMSG_) > ITE > :: > ?ACCPTR> > DUP > xTYPE > a%>$, > APPEND_SPACE > SWAP > xBYTES > SWAPDROP > a%>$, > &$ > ; > :: > % 0. > a%>$, > ; > &$ > NEWLINE$&$ > &$ > LOOP > ; > @ > > This told me which libraries actually have message tables (as well > as which type of array they were and the size of the arrays). Then > I ran my program to get a range of messages for only those > libraries which actually have messages. > > Assuming that your e-mail address is valid, I'll send you a copy > of my text file of only the English messages. > > To satisfy my curiosity a bit more, I ran this next program for > each library that has a message table to get the array itself. > > %%HP: T(3)A(R)F(.); > !NO CODE > !RPL > * Uses 1 unsupported but stable entry. > * Checksum: # 8C98h > * Size: 42.5 > :: > CK1NOLASTWD > :: > CK&DISPATCH1 > BINT11 > :: > HXS># > ; > BINT1 > :: > COERCE > ; > ; > PTR 08130 (GETMSG_) > IT > ?ACCPTR> > ; > @ > > Now, even though the type 4 arrays can be displayed on the stack, > no UserRPL command is expecting an array to be an array of > character strings, so I wouldn't try doing much with one without > first doing ->STR to put it into a string that's safe to work > with. > > But I did try the SIZE command, and that worked and returned a > list of the dimensions. It turns out that all of the type 4 > message arrays are 1-dimensional. > > Obviously, the system has ways of working with linked arrays (type > 23), but I didn't find much information on working with them. But > RPLMAN.DOC describes them and HP 48S/SX Machine Language gives > the details of their structure. I used ->H to turn them into > strings of hex nibbles to look at the dimension count and > dimension sizes. So far, I've derived this information > about the message table arrays: > > > Library Array Array Object size > ID type dimensions in bytes > > 0d 0h 23 { 46d 3d} {2Eh 3h} 3178. > 1d 1h 23 { 89d 3d} {59h 3h} 4388.5 > 2d 2h 23 { 8d 3d} { 8h 3h} 618. > 3d 3h 23 { 5d 3d} { 5h 3h} 356. > 5d 5h 23 { 6d 3d} { 6h 3h} 428. > 6d 6h 23 { 47d 3d} {2Fh 3h} 3323. > 7d 7h 23 {238d} {EEh} 4558.5 > 8d 8h 23 {176d} {B0h} 3149.5 > 9d 9h 23 { 13d 1d} { Dh 1h} 8883. > 10d Ah 23 { 9d 3d} { 9h 3h} 375. > 11d Bh 23 { 2d 3d} { 2h 3h} 148. > 12d Ch 23 { 23d 3d} {17h 3h} 1364. > 13d Dh 23 { 4d 3d} { 4h 3h} 268. > 185d B9h 4 {224d} {E0h} 4190.5 > 186d BAh 4 { 81d} {51h} 1088. > 187d BBh 4 { 65d} {41h} 1127. > 188d BCh 4 { 59d} {3Bh} 735. > 189d BDh 4 { 39d} {27h} 703. > 190d BEh 23 {119d} {77h} 2540.5 > 191d BFh 4 { 86d} {56h} 1491.5 > 192d C0h 4 { 1d} { 1h} 34. > 222d DEh 23 {104d 2d} {68h 2h} 5188. > 223d DFh 23 { 50d 3d} {32h 3h} 2046. > 225d E1h 4 { 41d} {29h} 612. > 228d E4h 4 { 8d} { 8h} 156.5 > 230d E6h 4 { 13d} { Dh} 205. > 256d 100h 23 { 6d 2d} { 6h 2h} 261. > 257d 101h 4 { 24d} {18h} 427.5 > 788d 314h 4 { 1d} { 1h} 30. > > Note that the above is for only built-in ROM libraries, it doesn't > include any external libraries, not even the equation library or > periodic table library. > > 1-dimensional arrays and the array with the 2nd dimension being 1 > are English language only, arrays with the 2nd dimension of 2 are > English and French, and arrays with the 2nd dimension of 3 are > English, French, and Spanish. If the calculator's LANGUAGE is set > to French or Spanish and the particular library's message table > doesn't have that language, then the English message is used > instead. > > The library message tables are implemented as arrays of strings. > Of course an array starts with the prologue address for it's own > type, and then its total body length, but it also includes one > copy of the prologue address for the type of elements that it > contains, which are bodies of that object type, not complete > objects per se. It also contains a count of how many dimensions it > has, and the size of each dimension. > > A type 4 array is the normal (albeit not necessarily numeric) > array, with all elements present and in order. > > A type 23 array is a linked array, with a series of pointers to > the elements. The pointers must be present and in order, but the > elements can be in any order, more than one pointer can point to > the same element (object body), elements need not be present, and > a null (00000) pointer is used for any missing elements. > > I'll probably dig into the arrays a bit more; I'm curious about > those null messages. I believe that I see a way to convert a > linked array of strings to a list array of strings, and it > should be just a matter of actually writing the program. I'm > assuming that the SysRPL commands for working with arrays are > designed for normal arrays instead of linked arrays, but if I > start out by using ->H on a linked array, I expect that I can > deal with it. > > Note that a type 29 symbolic matrix is rather different from > these arrays; It's composite object, and very much like a list > array. > > What I was kind of hoping was for a list somewhere of the indirection > pointers. In other words, I was expecting that somewhere in ROM was a > packed list of messages, and elsewhere would be a table that says > Message 12345 is at offset abcd in the message table. So I could > just page through the table looking at messages which actually exist. > > What you could do is store a plain text file of the messages and > their addresses as a string on the calculator, perhaps on an SD > card. As long as you know at least part of the message that you > want, you could use the editor's FIND operation to search for it. > > > > -- > James === Subject: Re: Error messages are useful strings -- which? PS: I decided that it would be better to place the file on: http://www.i-is.com/users/jmprange/ The file name is: MSTBL.ZIP -- James === Subject: Re: Error messages are useful strings -- which? <204f1$45759d75$4267eab7$10037@123.NET> <62b4b$4575e26e$4267ea74$1722@123.NET> Now I'm wondering if there is a utility to do a translation from the display format used by the current Conn4x software. In other words, if I do an XModem transfer in ASCII mode from HP4X to PC, I get a file which is (a) readable without special fonts (and therefore emailable), and (b) can be reconstructed by simply transferring it via XModem to an HP-4X. The problem is that I may want to send some of my programs to my daughter, who uses only a Mac. I can email her the display format file and she can copy it onto her SD card, but to change that DadProg.txt file back to the original UserRPL (I don't aim higher than that at the moment) is not possible without said traslation program. I will check out HPConnect 0.9, which would be a good workaround if it works. --Irl > PS: > > I decided that it would be better to place the file on: > http://www.i-is.com/users/jmprange/ > > The file name is: MSTBL.ZIP > > -- > James === Subject: Re: Error messages are useful strings -- which? > James -- > SysRPL program (as a command-line entry, presumably) to actually be a > program was useful; I have tried displaying the SysRPL stack (-85 SF) > with a tiny UserRPL program on the stack and then typing in the same > thing but couldn't get it to actually be a program. > Here's a presumably dumb question: there are two extable files at > hpcalc.org. One is for the meta kernel and is much shorter than the > other one. What are these, please? You are using the 49g+/50G right? You should use extable2.lib which is large but has additional Entry Points Try www.hpcalc.org hp49 section Programming Entries Emacs (read the docs, too and see other useful stuff there) === Subject: Re: Error messages are useful strings -- which? <9u5dh.6$X24.1@reader1.news.saunalahti.fi> light reading. Looks like this is the same... Irl > James -- > SysRPL program (as a command-line entry, presumably) to actually be a > program was useful; I have tried displaying the SysRPL stack (-85 SF) > with a tiny UserRPL program on the stack and then typing in the same > thing but couldn't get it to actually be a program. > Here's a presumably dumb question: there are two extable files at > hpcalc.org. One is for the meta kernel and is much shorter than the > other one. What are these, please? > > You are using the 49g+/50G right? > You should use > extable2.lib > which is large but has additional Entry Points > Try www.hpcalc.org > hp49 section > Programming > Entries > Emacs > (read the docs, too and see other useful stuff there) === Subject: cmsg cancel Control: cancel Summary: EMP spam, BI=902 === === Subject: Data input: suggestions? I currently have a utility that will take point numbers, find them from a database, and calculate the area of the polygon. I'd like to expand it to allow the inclusion of curves in the figure as well. The problem I am running into is facilitating easy entry of curves with the build in command line parser. I really don't want to have to write my own for obvious reasons. These curves are defined by a beginning and end point, as well as a direction (curving right or left). I have a PDF made by a surveyor that is helping me try to figure out the best way to do this if anyone wants to take a look (www.pssllc.com/culdesac.pdf). It makes this discussion a little easier to visualize. In that paper, we came up with the idea of using - or + on the number to define L/R, and then if the value is less than 1, it is a radius point. So { 1 -.2 3 } would define a curve starting at point 1, curving to the left with radius at point 2, and ending at point 3. I thought this would work great until I remembered that if someone types in { 1 -.20 3 } to define a curve with radius point 20. . . So then I thought, I'll just switch to EXACT mode when they type in a list, but I'd really like to avoid that since I've been doing my best to hide exact integers from the people using my software. For a group of people who don't want to have to learn the difference between 1. and 1 , keeping it in approximate always is the best solution. So can anyone thing of an easier way to allow quick entry of a list like this without adding in a bunch of extra keystrokes to define the radius point? (for example, putting the radius point inside a list requires 2 keystrokes to put in the list, and an extra 2 R arrow presses to position the cursor after the list. I thought someone here might be able to see another way to get around this issue. Ideas? TW === Subject: Re: Data input: suggestions? Tim, The strategy that I've long used for this very situation is similar but slightly simpler. Use a negative real to signify radius point. In the example pdf this could be { 2. -3. 4 } where the -3. indicates radius point. There's no need to tell the routine + or - for left or right hand curve - the delta (central) angle should itself be + or - when the 2 azimuths are subtracted (3. to 2. and 3. to 4.). In the example, an odd case occurs at the cul-de-sac. If one inputs { 4. -5. 6. }, your routine will probably default to the angle < 180 whereas one might desire the angle > 180. The easy way to handle this is to input { 4. -5. -6. } where the -6. tells the routine to follow the greater arc instead of the lesser one. Another advantage to this scheme is the resulting algebraic sign of the segment area, + if clockwise, - if anticlockwise (again determined by the sign of the delta). The disadvanatge to all of this is that it requires pre-calculated radius points, to which some users may object. But a major advantage is never having to tell the routine whether the chord is inside or outside the curve as one has to do with some predominant (and expensive) cad/cogo packages. :-) Chuck >I currently have a utility that will take point numbers, find them from > a database, and calculate the area of the polygon. I'd like to expand > it to allow the inclusion of curves in the figure as well. The problem > I am running into is facilitating easy entry of curves with the build > in command line parser. I really don't want to have to write my own > for obvious reasons. > > These curves are defined by a beginning and end point, as well as a > direction (curving right or left). I have a PDF made by a surveyor > that is helping me try to figure out the best way to do this if anyone > wants to take a look (www.pssllc.com/culdesac.pdf). It makes this > discussion a little easier to visualize. > > In that paper, we came up with the idea of using - or + on the number > to define L/R, and then if the value is less than 1, it is a radius > point. So { 1 -.2 3 } would define a curve starting at point 1, > curving to the left with radius at point 2, and ending at point 3. I > thought this would work great until I remembered that if someone types > in { 1 -.20 3 } to define a curve with radius point 20. . . > > So then I thought, I'll just switch to EXACT mode when they type in a > list, but I'd really like to avoid that since I've been doing my best > to hide exact integers from the people using my software. For a group > of people who don't want to have to learn the difference between 1. and > 1 , keeping it in approximate always is the best solution. > > So can anyone thing of an easier way to allow quick entry of a list > like this without adding in a bunch of extra keystrokes to define the > radius point? (for example, putting the radius point inside a list > requires 2 keystrokes to put in the list, and an extra 2 R arrow > presses to position the cursor after the list. I thought someone here > might be able to see another way to get around this issue. Ideas? > > TW > === Subject: cmsg cancel Control: cancel Summary: EMP spam, BI=929 === === Subject: Re: An Ancient Computer Surprises Scientists You could also visit the following link: http://www.antikythera-mechanism.gr/index.php?lang=en Alex Markatis Civil Engineer Athens, Greece A/ John H Meyers Ucoaaa: > An Ancient Computer Surprises Scientists: > > http://www.smh.com.au/news/science/scientists-decode-ancient-astral-computer/ 2006/11/30/1164777721260.html > http://www.abc.net.au/news/newsitems/200611/s1801051.htm > http://www.nzherald.co.nz/section/story.cfm?c id=5&ObjectID=10413251 > http://www.nytimes.com/2006/11/29/science/30computecnd.html?ex=1322456400&en= 4dd0f1ffb387158c&ei=5088&partner=rssnyt&emc=rss > http://www.chicagotribune.com/technology/chi-0611300255nov30,1,6121069.story? coll=chi-technology-hed&ctrack=1&cset=true > http://news.com.com/2100-11397 3-6139636.html > > Next thing you know, an HP50G (with working 2100-year-old keyboard) >A579 === Subject: Re: Slightly tinted dust cover on the new HP50G On Fri, 01 Dec 2006 23:03:42 +0100, John Torset >The second might be a keytime problem, but one of the first thing I >did was to put the Keytime to 0 and then to 150 and still I have the >missing keystrokes. Therefore I think there are a bug in the keyboard >buffer routine. the 'busy bug', I find that the missing keystroke problem is gone. I guess HP will come around and fix the 'busy bug' problem as well in a futere ROM revision. Here is a picture of the 50G with the dust cover from 49G+ http://home.broadpark.no/~jtorset/home/HP50G%20with%20HP49G+%20dust%20cover. jpg Nice execpt for the printed text. This is more in line with the old HP12C calculator design from the beginning of the 80's where the display also have a golden border around the display :-) -- John Torset === Subject: Re: Slightly tinted dust cover on the new HP50G <7891n257hf4isajedjjqc445qbdd100n5v@4ax.com> <8t94n21o4vn6p9fgospknhk58rvmvu2cr9@4ax.com> that inferior screen with the slightly reddish tint for the 50G is beyond me. I guess they don't want you to have a nice, tactile keyboard and a good contrast screen at the same time. (Perhaps I'll get up the courage up to do this operation (switching dust covers) on my 49G+ and 50G) . . . > On Fri, 01 Dec 2006 23:03:42 +0100, John Torset > >The second might be a keytime problem, but one of the first thing I >did was to put the Keytime to 0 and then to 150 and still I have the >missing keystrokes. Therefore I think there are a bug in the keyboard >buffer routine. > > the 'busy bug', I find that the missing keystroke problem is gone. > > I guess HP will come around and fix the 'busy bug' problem as well in > a futere ROM revision. > > Here is a picture of the 50G with the dust cover from 49G+ > > http://home.broadpark.no/~jtorset/home/HP50G%20with%20HP49G+%20dust%20cover.j pg > > Nice execpt for the printed text. > This is more in line with the old HP12C calculator design from the > beginning of the 80's where the display also have a golden border > around the display :-) > > -- > John Torset === Subject: cmsg cancel Control: cancel Summary: EMP spam === === Subject: Re: ARM debounce > There are my postings too ? :-) > Great, are those any good ? :-) Didn't you participate in those same threads? Yes, you seem to be most expert; are you (or do you want to become) a contractor to HP (or to Kinpo)? Then you could be as happy as JY or BP with their relationship to HP :) [r->] [OFF] === Subject: Re: ARM debounce > Yes, you seem to be most expert; > are you (or do you want to become) > a contractor to HP (or to Kinpo)? > > Then you could be as happy as JY or BP > wit will be dug up -- well, maybe only a pre-CAS version, > although of course some CAS functionality may already be known > to date back that far :) > > There is nothing new under any Sun > > [r->] [OFF] === Subject: Re: HP50G and UNIX Christophe Prevotaux, le Sun 03 Dec 2006 00:05:07 +0000, a .8ecrit : > Has anyone succeeded in connection its HP50G to Linux or FreeBSD > thru the USB port ? FreeBSD being the OS that I use :). Again, with linux it's easy with a recent enough kernel: just compile the hp4x module (USB_SERIAL_HP4X). Samuel === Subject: Re: HP50G and UNIX Christophe, If you can figure out how to get your 50g USB link working under BSD please let the rest of us know. I am running OpenBSD here and am at a loss. Damn HP and their lack of serial cable! What kind of company releases a product that has a data port but NO corresponding cable? Cup On Dec 2, 4:05 pm, Christophe Prevotaux > Hi All, > > Has anyone succeeded in connection its HP50G to Linux or FreeBSD > thru the USB port ? FreeBSD being the OS that I use :). > > -------------------- > Christophe > -------------------- === Subject: Re: HP50G and UNIX > Has anyone succeeded in connection its HP50G to Linux or FreeBSD > thru the USB port ? FreeBSD being the OS that I use :). This post from a few days ago. . . TW === Subject: Re: HP50G and UNIX Le Sat, 02 Dec 2006 20:29:21 -0800, TW a .8ecrit: >> Has anyone succeeded in connection its HP50G to Linux or FreeBSD >> thru the USB port ? FreeBSD being the OS that I use :). > > This post from a few days ago. . . > > > TW === Subject: Directories on the SD card I have been creating directories as outlined on page 18-4 in the User's manual and storing data files within them. These files are available to the HP50 and my PC (using a card reader). The problem begins when I try to verify the directories' existence using 3:TEST VTYPE. The result should be 15 and is for the first directory created in this manner, but the return for the rest of the directories has been -1. Only the first created directory returns the expected variable type of 15. Is anyone else having this problem, or am I misunderstanding how this should work. === Subject: Re: Directories on the SD card Sorry I can't be more detailed, but my experiments with SD card directories indicates that things are very, very strange! I couldn't delete directories, run a VLIST command, etc... I wouldn't be surprised if the VTYPE command isn't supported in this setting. > I have been creating directories as outlined on page 18-4 in the > User's manual and storing data files within them. These files are > available to the HP50 and my PC (using a card reader). The problem > begins when I try to verify the directories' existence using 3:TEST > VTYPE. The result should be 15 and is for the first directory > created in this manner, but the return for the rest of the > directories has been -1. Only the first created directory returns > the expected variable type of 15. Is anyone else having this > problem, or am I misunderstanding how this should work. === Subject: Re: Directories on the SD card Think the SD card as an external near-line tape-drive It's limited in functionality but ublimited as a storage device You can even start programs from it Manjo explained a lot more.. > Sorry I can't be more detailed, but my experiments with SD card > directories indicates that things are very, very strange! I couldn't > delete directories, run a VLIST command, etc... > > I wouldn't be surprised if the VTYPE command isn't supported in this > setting. > > >> I have been creating directories as outlined on page 18-4 in the >> User's manual and storing data files within them. These files are >> available to the HP50 and my PC (using a card reader). The problem >> begins when I try to verify the directories' existence using 3:TEST >> VTYPE. The result should be 15 and is for the first directory >> created in this manner, but the return for the rest of the >> directories has been -1. Only the first created directory returns >> the expected variable type of 15. Is anyone else having this >> problem, or am I misunderstanding how this should work. > === Subject: Re: Directories on the SD card Hello SD's root you will be able to test the type of those files by using: :3:directoryname VTYPE if you created DOS directories or subdirectories and put your files or HP directory in there you must supply the correct path :3:{directory subdir filename} VTYPE directory and subdir should be valid names of directories where your file 'filename' is stored. If you try to check the type (VTYPE) of DOS directory you get an error. DOS directories are not HP objects. At this time you can't browse trough HP directory objects on SD, like you would with DOS directories. i hope this helps manjo === Subject: Re: Directories on the SD card oh i forgot... (i was testing some directory names which had the same name as commands -that's why i was getting the error) BUT if you try to check DOS directory using path syntax : :3:{ directory } VTYPE -you should get : -1 for sure Maybe your other directories are realy not the *same* after all. (15. is type for HP directory, DOS directory is not HP object that's why -1) manjo === Subject: Logistic Regression on a 49G+. I'm aware of another discussion involving logistic regression that points to RAS 3.0: http://www.hpcalc.org/details.php?id=1301 However, it is only for the 48 series. I tried using hpconv: http://www.hpcalc.org/details.php?id=3799 To convert the 48 binary to a 49 one, but alas, it gave me numerous Address for entry xyz (abc) not found. errors. Unfortunately, I don't know SysRPL, so I have no clue what I'm doing. Would anyone be kind enough to help me out? Any suggestions for converting this binary? Ideas of where to start? --Joey C. http://joeyjwc.x3fusion.com/ === Subject: Re: Logistic Regression on a 49G+. > I'm aware of another discussion involving logistic regression > that points to RAS 3.0: > http://www.hpcalc.org/details.php?id=1301 > However, it is only for the 48 series. I tried using hpconv: > http://www.hpcalc.org/details.php?id=3799 > To convert the 48 binary to a 49 one, but alas, it gave me numerous > Address for entry xyz (abc) not found. errors. Which is odd, since there's nothing unsupported in it. > Unfortunately, I don't know SysRPL, so I have no clue what I'm doing. > Any suggestions for converting this binary? It's all *UserRPL* (author MacDonald R. Phillips' favorite calculator language :), except for one item 'PAD1' which is all supported SysRPL. So what I'd do is: o On HP48GX: Store dis-assembled text for 'PAD1' into 'PAD1' o Recall entire directory and do STD 64 STWS ->STR o Copy entire string to HP49 and do STR-> (in Approx. mode) o Store the directory. o Compile 'PAD1' and store it back into 'PAD1' Here's the source for 'PAD1' (to save time with the first step above): :: CK2NOLASTWD CK&DISPATCH0 # 31 :: DO>STR DUPLEN$ TWENTYTWO 4PICK LEN$ #- 2DUP#< ITE :: SWAP#- Blank$ SWAP&$ ; :: 2DROP ; &$ ; ; Don't have an HP48GX? So use Emu48 or Debug4x -- it's cheaper than eBay :) [r->] [OFF] === Subject: Re: Logistic Regression on a 49G+. Here are some things to watch out for when attempting to simply transfer this UserRPL directory from a 48 series calc to a 49 series calc: > logistic regression, RAS 3.0: > http://www.hpcalc.org/details.php?id=1301 There is one very small SysRPL program 'PAD1' (as mentioned earlier), which needs to be converted to source and recompiled after transfer. There is a variable named 'ANS' (referenced in two other places); however, ANS is a *command* in 49 series, so you'll have to change that name in all three places. It also contains *unsupported* #353ABh [HP48] SYSEVAL in two places; thus far, #2F110h SYSEVAL (or #2B7001h FLASHEVAL) does the same in 49G[+]/50G/48Gii, but these still aren't supported (works with most expressions, not at all with programs). This offers a few examples of why even UserRPL transfer needs close attention. [r->] [OFF] === Subject: Re: Logistic Regression on a 49G+. Original file for HP48GX: > logistic regression, RAS 3.0: > http://www.hpcalc.org/details.php?id=1301 Here's a downloadable Zip file for 49G[+]/50G/48Gii with source and binary (will remain posted for 7 days): File Name: ras30B.zip Size: 43.29 KB http://www.filefactory.com/file/775cc5/ [r->] [OFF] === Subject: Re: Good text editors for 50G? Maybe I am missing something, but I do not understand the built-in Text Editor. When I go to APPS then Text Editor, all I see is the Left Arrow cursor. I need to press ALPHA just so I can start writing text. This TEXT EDITOR is just a line editor. Am I correct ? > > I installed TED in my hp50, but the editor seems to work improperly: > two letters appear when pressing the key, the cursor moves twice > up/down,.... > > is there a prblem with the code or I made something wrong? > > > > GaaK ha escrito: > > > Glad you like it. > > > > If you are going to do any serious programming, use emacs. Otherwise, > > the built in editor is great. What exactly are you wanting to do with > > a text editor on the calc? Write a novel? :-) > > > > TW > > TED vs Editor built-in > > - TED makes no duplicate of the edited string if it is in temporary > object area, thus enabling editing very large strings. Note that this > implies that no backup of the original string is kept! > > - TED is very fast and supports the tabulator. > > - Most of the normal character keys are in their normal places, others > can be fetched via the special character browser or ASCII code: CHR?+64 > = @ > > - TED decompiles object to width 30. This can be changed by storing a > real number larger than 9 in variable TED.WID. > > - TED contains the FIND/REPLACE option very fast! > > - TED works with ALL display > > - TED contains the Marks support and Macros > > - GaaK - === Subject: Re: Good text editors for 50G? > Maybe I am missing something, > but I do not understand the built-in Text Editor. > When I go to APPS then Text Editor, > all I see is the Left Arrow cursor. > I need to press ALPHA just so I can start writing text. > This TEXT EDITOR is just a line editor. Am I correct ? This APPS entry has always confused everyone; it's only the command line! -- identical to pressing right-shift ENTRY (above Alpha), and then you have to enter something valid in UserRPL (although any quoted big long string will do). On the other hand, if you type (an empty string), press ENTER and then press cursor-down, the internal pure text editor (string editor) will start (as it will for any existing string on the stack). In either case, MODE DISP offers Edit: options, such as small text (minifont), full page (cursor can move beyond current end of any line), and indent (repeats a previous line's indentation). [r->] [OFF] === Subject: Re: Good text editors for 50G? > >> Maybe I am missing something, >> but I do not h their relationship to HP :) Well at this time i have a *day job* in a small private company programming is what i do, calculators (programming) is my hobby. I'm quite happy with things the way they are now, it would not be easy for me to change any of this. Since JY doesn't seem to be so much happy with his relationship with HP, although i *am* sure there were better times in that relationship -sad fact is everything changes (even the good old HP) (this would be my primary concern if i'd ever consider HP employment oportunities) As to my expetise : i've been around since early days of home computers, not only programing low level but also taking most things appart. Re-assembling again after inspection, some items sacrificed to the 'science' as well :-). I try to be clear when i'm hit with some new idea and develop it as i go, open for sugestions too. Being a telecomunications technitian and ever-lasting computer science student comes like a bonus :-) Of course that doesn't mean i'm right :-) that's where it comes frome (i guess) later... manjo === Subject: cmsg cancel Control: cancel Summary: EMP spam === === Subject: How to graph x(y)? How to graph this equation? x=y-3*y^3+y^5 Why solver cant solve for 'y'? Off topic: To graph x^2 for x>0 and x^3 for x<=0 in the same graph I graph: {'x^2*(x>0)' 'x^3*(x<=0)'} there is another way to do that? === Subject: Re: How to graph x(y)? Avatar e a ?crit : > How to graph this equation? > > x=y-3*y^3+y^5 > Graphing in RPN: - Set 'Y' as the independent variable (key MODE + menu CAS) - Press and hold RS and press F1 (Y=) - Type 'Y1(Y)=Y-3*Y^3+y^5' and press ENTER - Select ERASE and DRAW from the menu > Why solver cant solve for 'y'? > Solving in RPN (but without Solver): - Set 'Y' as the independent variable (key MODE + menu CAS) - Type 'Y-3*Y^3+y^5' in the stack and press ENTER - Type 'Y' in the stack - Select SOLVE from the Catalog - You get 5 solutions (coherent with graph) === Subject: Re: How to graph x(y)? > How to graph this equation? > x=y-3*y^3+y^5 Simpler if you make 'y' the independent variable :) > Why solver can't solve for 'y'? Analysis of your program would be easier if posted. But if you mean that you are using no program (to calculate 'y' given 'x'), then that's why :) The plotting engine (assuming FUNCTION plot) needs either an expression or a program which takes values of the independent variable and returns values to be plotted. When you supply an *equation* (containing =) to the plotter, it simply plots each side independently (ignoring the left side if it is only a variable name); it does not try to solve the equunderstand the built-in Text Editor. >> When I go to APPS then Text Editor, >> all I see is the Left Arrow cursor. >> I need to press ALPHA just so I can start writing text. > >> This TEXT EDITOR is just a line editor. Am I correct ? > > This APPS entry has always confused everyone; > it's only the command line! -- identical to pressing > right-shift ENTRY (above Alpha), and then > you have to enter something valid in UserRPL > (although any quoted big long string will do). > > On the other hand, if you type (an empty string), > press ENTER and then press cursor-down, the internal > pure text editor (string editor) will start > (as it will for any existing string on the stack). > > In either case, MODE DISP offers Edit: options, > such as small text (minifont), full page > (cursor can move beyond current end of any line), > and indent (repeats a previous line's indentation). Now I'm confused?! The Build-in editor has [TOOL]s on the [NXT] pages that allows to saerch&Replace, change STYLE and FONT better than other calculators that I have used.. === Subject: Re: Good text editors for 50G? > Now I'm confused?! ?! sure does look confusing :) > The Built-in editor has [TOOL]s on the [NXT] pages > that allows to search&Replace, change STYLE and FONT > better than other calculators that I have used.. Yes, that's always available (even while editing the command line :) What's different abouation (which is in the most general case often not possible). If you have not defined 'y' then of course an undefined name error occurs; storing a program under 'y' which solves for 'y' given 'x' (assuming that's the independent variable) might help. [r->] [OFF] === Subject: Re: How to graph x(y)? > storing a program under 'y' which solves for 'y' > given 'x' (assuming that's the independent variable) > might help. I neglected to add that you'd then need to plot 'y(x)' rather than your original equation. Alternatively, just store into 'EQ' the program for computing 'y' given 'x', then you plot the given *program's* output. Unfortunately, if you use the numeric root-finder, you might probably some of the time get an extremum, rather than a root, unless you also have programmed a good method of choosing one or more initial guesses; the following, for example, is not very satisfactory: << 'X=Y-3*Y^3+Y^5' 'Y' 0 ROOT >> STEQ FUNCTION 'X' INDEP ERASE DRAX DRAW PICTURE This may serve to illustrate why it's not reasonable to expect the calculator to figure out for itself how to solve any old equation. Now try instead: 'X=Y-3*Y^3+Y^5' STEQ FUNCTION 'Y' INDEP ERASE DRAX DRAW PICTURE Much better, isn't it? Just remember to view the calculator in a mirror, while lying down, to see 'Y' as a function of 'X' :) [r->] [OFF] === Subject: Re: How to graph x(y)? On Sun, 03 Dec 2006 02:19:06 -0600: [excuse my slow and disjointed thinking :] > Unfortunately, if you use the numeric root-finder, > you might probably some of the time get an extremum, > rather than a root. And you might miss the fact that sometimes there are as many as five different roots (for 'y') given some values of 'x' (such as for x=0) It is thus *impossible* to get this result via a FUNCTION plot (of a single equation) using 'x' as the independent variable, because you can give the plotting engine only one numeric result (per equation) to plot per distinct value of the independent variable. Of course, you could try to create a list of five different functions in EQ, and somehow try to get them to collectively cover all the possible roots -- but that's not very feasible, is it? So the way to graph 'y(x)' in this case is certainly to graph 'x(y)' instead (oh look, that's what you asked in the subject!) -- and the way to do that is simply to set the independent variable to 'y' instead of 'x' (perhaps I can claim that the question why can't the solver solve for 'y'? distracted me :) [r->] [OFF] === Subject: Re: How to graph x(y)? Generally I solve (with 'Y' SOLVE) the equation (for exampt editing a command line vs. editing a string is that a string isn't compiled after you edit it. When you invoke APPS > Text editor or Right-shift ENTRY, whatever you compose will get *compiled* and *executed* (just type two consecutive decimal points for a quick illustration), wheras when you do string EDITB the string is edited (without enclosing in quotes), and not re-compiled afterwards. The closest HP48 simulation of pure string editing would be: string INPUT (which you can also do on the 49G[+]/50G, if EDITB is too simple to use :) [r->] [OFF] === Subject: Re: Good text editors for 50G? Now I understand. I missed the [TOOLS] [NEXT] group. > > Now I'm confused?! > > ?! sure does look confusing :) > > The Built-in editor has [TOOL]s on the [NXT] pages > that allows to search&Replace, change STYLE and FONT > better than other calculators that I have used.. > > Yes, that's always available > (even while editing the command line :) > > What's different about editing a command line vs. editing a string > is that a string isn't compiled after you edit it. > > When you invoke APPS > Text editor or Right-shift ENTRY, > whatever you compose will get *compiled* and *executed* > (just type two consecutive decimal points for a quick illustration), > wheras when you do string EDITB the string is edited > (without enclosing in quotes), and not re-compiled afterwards. > > The closest HP48 simulation of pure string editing > would be: string INPUT > (which you can also do on the 49G[+]/50G, > if EDITB is too simple to use :) > > [r->] [OFF] === Subject: nonpareil HP-41 emulator under Mac-OSX? The website http://nonpareil.brouhaha.com/ indicates that the HP-41C and friends are simulated by nonpareil using the GTK+ toolkit. Has anyone already built nonpareil, from source with GTK+ for Mac-OSX, -- Chris. === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? Chris McDonald a ?crit : > The website http://nonpareil.brouhaha.com/ indicates that the HP-41C > and friends are simulated by nonpareil using the GTK+ toolkit. > > Has anyone already built nonpareil, from source with GTK+ for Mac-OSX, > > -- > Chris. You can find an HP 41 simulator for Macintosh at the fllowing URL: http://homepage.mac.com/mba/nsim/ Laurent === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? >Chris McDonald a =E9crit : >> The website http://nonpareil.brouhaha.com/ indicates that the HP-41C >> and friends are simulated by nonpareil using the GTK+ toolkit. >> >> Has anyone already built nonpareil, from source with GTK+ for Mac-OSX, >> >You can find an HP 41 simulator for Macintosh at the fllowing URL: >http://homepage.mac.com/mba/nsim/ but the webpages keeping saying older version of nsim... and I had wondered what I was missing out on. (?) -- Chris. === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? Chris, http://homepage.mac.com/mba/nonpareil/ (All credit goes to Eric Smith and Maciej Bartosiak (osx port). I'm just providing the link..) Pal > >Chris McDonald a =E9crit : > >> The website http://nonpareil.brouhaha.com/ indicates that the HP-41C >> and friends are simulated by nonpareil using the GTK+ toolkit. >> >> Has anyone already built nonpareil, from source with GTK+ for Mac-OSX, >> > >You can find an HP 41 simulator for Macintosh at the fllowing URL: > >http://homepage.mac.com/mba/nsim/ > > but the webpages keeping saying older version of nsim... > and I had wondered what I was missing out on. (?) > > -- > Chris. === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? >http://homepage.mac.com/mba/nonpareil/ >(All credit goes to Eric Smith and Maciej Bartosiak (osx port). I'm >just providing the link..) (see my original email). -- Chris. === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? > (see my original email). > Chris. Darn it. I hate it when I jump my guns. I looked too quickly. Sorry about that. Pal === Subject: connectivity kit I have downloaded at least two versions of the connectivity kit the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 I keep getting trapped in a circle of the software telling me to install Xmodem driver, calculator not found, timeout, etc. This latest version does not have the PC in the right pane as the older version did If I manage to drag or copy and paste a file over from the calc; it is usually unreadable by a PC with the proper software to open and use the file (specifically, a .cr5 file going to a PC with TDS software) Tried calling an 800 number for HP but the foreign person there informed that my Compaq computer is not in warranty*&^%$#@ Anyone know where to get some usable, printed help documentation? === Subject: Re: connectivity kit Note that for 48GX you need to download and install the XMODEM server library, and >>it must be in port 0 or 1<<. I tried it in other ports Irl > I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me to > install Xmodem driver, > calculator not found, timeout, etc. > This latest version does not have the PC in the right pane as the older > version did > If I manage to drag or copy and paste a file over from the calc; it is > usually unreadable by > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > Tried calling an 800 number for HP but the foreign person there > informed that my Compaq computer > is not in warranty*&^%$#@ > Anyone know where to get some usable, printed help documentation? === Subject: Re: connectivity kit now that we are talking about connectivity kit, is there anyway to use the serial conector with an USB adaptor? my laptop doesnt have serial plug. > Note that for 48GX you need to download and install the XMODEM server > library, and >>it must be in port 0 or 1<<. I tried it in other ports > Irl > I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me to > install Xmodem driver, > calculator not found, timeout, etc. > This latest version does not have the PC in the right pane as the older > version did > If I manage to drag or copy and paste a file over from the calc; it is > usually unreadable by > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > Tried calling an 800 number for HP but the foreign person there > informed that my Compaq computer > is not in warranty*&^%$#@ > Anyone know where to get some usable, printed help documentation? === Subject: Re: connectivity kit > now that we are talking about connectivity kit, > is there any way to use the serial conector > with a USB adaptor? > my laptop doesnt have serial plug. For the 49G+/50G, a serial cable (if it existed), pluged into a USB adaptor, would be very similar to the existing direct USB cable, except possibly that it might then appear (using the adapter's driver) as a COM port, rather than as a USB port, for use by some other software. I just bought some very small USB-to-serial adapters, and lo, they install fine but don't work with anything (neither on COM4, which some software recognizes, nor on COM5, which my applications have never heard of :) I think I'd better never upgrade to a computer without on-board serial, or else I'll have to throw out all my favorite organizers and calcs :) [r->] [OFF] === Subject: Re: connectivity kit so, would it work using a serial to USB adaptor with my 48gx? > > now that we are talking about connectivity kit, > is there any way to use the serial conector > with a USB adaptor? > my laptop doesnt have serial plug. > > For the 49G+/50G, a serial cable (if it existed), > pluged into a USB adaptor, would be very similar > to the existing direct USB cable, > except possibly that it might then appear > (using the adapter's driver) as a COM port, > rather than as a USB port, > for use by some other software. > > I just bought some very small USB-to-serial adapters, > and lo, they install fine but don't work with anything > (neither on COM4, which some software recognizes, > nor on COM5, which my applications have never heard of :) > > I think I'd better never upgrade to a computer > without on-board serial, or else I'll have to throw out > all my favorite organizers and calcs :) > > [r->] [OFF] === Subject: Re: connectivity kit > would it work using a serial to USB adaptor with my 48gx? Should work in theory, provided that USB adapter driver creates a serial COM port that your computer-side software can use, and that its hardware can talk to 48GX's UART (serial circuit), which won't be the calc's fault if it can't, because 48GX is engineered to be truly compatible with original serial hardware specs (using either no flow control or Xon/Xoff flow control, not hardware flow control). I hope that someone who has actually done it (on the same computer operating system that you use, and ideally with an adapter just like yours) will add something more encouraging, and perhaps useful. [r->] [OFF] === Subject: Re: connectivity kit adaptor. If it doesnt work, then it'd useless. And there's actually a custom cable sold on ebay for $40 which is kinda expensive > > would it work using a serial to USB adaptor with my 48gx? > > Should work in theory, provided that USB adapter driver > creates a serial COM port that your computer-side software can use, > and that its hardware can talk to 48GX's UART (serial circuit), > which won't be the calc's fault if it can't, > because 48GX is engineered to be truly compatible > with original serial hardware specs > (using either no flow control or Xon/Xoff flow control, > not hardware flow control). > > I hope that someone who has actually done it > (on the same computer operating system that you use, > and ideally with an adapter just like yours) > will add something more encouraging, and perhaps useful. > > [r->] [OFF] === Subject: Re: connectivity kit > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) This means you have a 48GX, correct? TW === Subject: Re: connectivity kit > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > > This means you have a 48GX, correct? > > TW Sorry about lack of specific info; yeah, HP 48GX with a TDS Cogo GX/SX card in it. === Subject: Re: connectivity kit >I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me to > install Xmodem driver, > calculator not found, timeout, etc. > This latest version does not have the PC in the right pane as the older > version did > If I manage to drag or copy and paste a file over from the calc; it is > usually unreadable by > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > Tried calling an 800 number for HP but the foreign person there > informed that my Compaq computer > is not in warranty*&^%$#@ > Anyone know where to get some usable, printed help documentation? Search this newsgroup the USB driver I downloaded here: http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareIndex.jsp?lang=en&c c=us&prodNameId=3235174&prodTypeId=215348&prodSeriesId=3235173&swLang=8&taskI d=135&swEnvOID=228 The HPConn4x version that I use is here: http://www.hpcalc.org/details.php?id=5411 in the HP web pages Training Modules http://h20331.www2.hp.com/hpsub/cache/383688-0-0-225-121.html I did not find Connecting to PC Funny enough the only official English web page (that I know of) giving some advice - is in Finland: http://www.dlc.fi/~radix/tuki-support.htm click on that 1st link on the center Lis.8atietoja englanniksi (starts Javascript pop-up windle from a Y^2 parabole) and save the answer (with 'EQ' STO) to do x(y) graphs, but in this case that does not work :( === Subject: Re: How to graph x(y)? > Generally I solve (with 'Y' SOLVE) the equation > (for example from a Y^2 parabole) > and save the answer (with 'EQ' STO) to do x(y) graphs, > but in this case that does not work :( Only a subset of all equations 'X=f(Y)' can be explicitly solved for 'Y=g(X)' It appears that you never even started graphing; the only issue was that SOLVE cannot perform a general solution for this equation. However, you can graph 'X=f(Y)' and flip the graph :) [r->] [OFF] === Subject: Re: How to graph x(y)? > Off topic: > To graph x^2 for x>0 and x^3 for x<=0 in the same graph 'IFTE(X>0,X^2,X^3)' [r->] [OFF] === Subject: Re: Debug4x supports User RPL X > Watch for Build 104, this coming week. It will remember separate fonts, > sizes, colors for SysRPL and UserRPL. > I see your email above. I will email you the patch file when it is ready > if you like (about 2-3 megs)? > -- - - - - - - - - - - - - - - - - > Bill Graves RKBA! > bgraves@ix.netcom.com Just let us all know where to DL the new v. when ready VP === Subject: Re: MXEval 2.4 > RESTACK belongs to the offcial ROM > It's SOOO...OOO usefull ! > <3 > of course! somehow I was expecting this comment;-) === Subject: Re: Error messages are useful strings -- which? > Raymond -- > guess I wasn't clear that I'm using an HP50g, so the binary libraries > won't work without disassembly and recompilation -- right? > Also, the list of messages gives only the error messages, as opposed to > all the useful other text strings which are accessible via the error > handler but which don't actually correspond to errors. I'm not Raymond, but I'll contribute to this discussion anyway. I posted the English language strings (but neglected the French and Spanish strings) in the message tables of ROM revision 1.19-6 for the 49G to hpcalc.org; see: http://www.hpcalc.org/search.php?query=message+tables That file is a bit dated; for example, some messages for the SD card have been added to the newer ROM revisions. My impression is that any unused messages are left in a ROM library message table when the ROM is revised, but each library has a maximum of 255 messages, so I wouldn't entirely rule out replacing unused messages with new ones, or otherwise changing old messages. That file includes messages from the equation library, but with the current ROM, the equation library and periodic table are optional, and their messages won't be available if you don't have those libraries installed. Some message in the tables are for errors, but some are for other purposes. The advantages of using a message table are, first, even if the message is used for more than routine, it can be stored in one place and called by number (a bit slower but usually reducing memory usage), and second, it's easier to arrange for multi-language messages. Message numbers are (in hex notation) of the form LLLMM, where LLL is the library number, from 000h through 7FFh, and MM is the message number within a library, from 01h through FFh. Another way of looking at it is that the message number is LLLh * 100h + MMh, or in decimal notation, LibNum * 256 + MsgNum, where LibNum is 0 through 2047 and MsgNum is 1 through 255. The message tables are arrays (or in some cases linked arrays) of strings. Some built-in message tables are English only; these are 1-dimensional arrays or 2-dimensional arrays with one of the dimensions being 1. Some have messages in French or both French and Spanish as well; these are 2-dimensional arrays with one of the dimensions being 2 or 3. If the LANGUAGE setting is French or Spanish and the message table doesn't have that language, then the English string will be used. A very few French or Spanish messages are empty strings. The UserRPL program: %%HP: ; @ Checksum: # 5C51h @ Memory usage: 22.5 @ Execution time: about .032 second << IFERR DOERR THEN ERRM END >> will get any message in the tables (using the current LANGUAGE setting), as well as the user-defined error message for #70000h (458752), except that an IFERR structure doesn't trap errors #123h (291) (used by the KILL command), #13Eh (318) (used by the CONT command), and #DFFh (3583) (used by control alarms). Messages #123h and #13Eh are empty strings anyway, and message #DFFh doesn't exist. Source code files for SysRPL alternatives are: %%HP: ; C$ $ !NO CODE !RPL * Checksum: # A78Fh * Memory usage: 35. * Execution time with a zint: about .004 second :: CK1NOLASTWD CK&DISPATCH1 BINT11 :: HXS># GETTHEMESG ; BINT1 :: COERCE GETTHEMESG ; ; @ and: %%HP: ; C$ $ !NO CODE !RPL * Checksum: # 3DDCh * Memory usage: 35. * Execution time with a zint: about .004 second :: CK1NOLASTWD CK&DISPATCH1 BINT11 :: HXS># JstGETTHEMSG ; BINT1 :: COERCE JstGETTHEMSG ; ; @ The difference between those SysRPL examples is that the one that uses GETTHEMESG will get the user-defined error message for an argument of #70000h (458752), and the one that uses JstGETTHEMSG doesn't check for #70000h. All of these programs return an empty string if the message doesn't exist. All three of the above programs take a binary integer, real number, or zint (exact integer), or any of these tagged, for the argument. A negative real or zint is treated as zero. A real with a fractional part is rounded to an integer value. A value over #7FFFFh (524287) is treated as modulus #80000h (524288), so that, foow) VP-Finland === Subject: Re: connectivity kit myankee ha escrito: > I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me to > install Xmodem driver, > calculator not found, timeout, etc. > This latest version does not have the PC in the right pane as the older > version did > If I manage to drag or copy and paste a file over from the calc; it is > usually unreadable by > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > Tried calling an 800 number for HP but the foreign person there > informed that my Compaq computer > is not in warranty*&^%$#@ > Anyone know where to get some usable, printed help documentation? I have problems with the latest version of HP connectivity kit, so I download and install the old 2.2 Build 2353 version from www.hpcalc.org, that work ok with my HP 49G and also with the HP 50G. I hope this help. === Subject: Re: connectivity kit > Anyone know where to get some usable, printed help documentation? > C:Program FilesHewlett-PackardConn4xConn4x.chm -- - - - - - - - - - - - - - - - - Bill Graves RKBA! bgraves@ix.netcom.com === Subject: Re: connectivity kit > I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for thr example, #1h and #80001h get the same message; I surmise that only the 19 least significant bits are used when getting messages from the tables. My above SysRPL examples should be faster when used with a binary integer or real for the argument. The argument checking sequence is, first, verify at least one object on the stack for the argument, then look for a binary integer, then look for a real, then strip any tags from the argument and convert a zint to a real, then look for a binary integer, then look for a real, so zints and tagged reals are slowest. A source code file for a SysRPL program that gets messages over a range of numbers and returns a string in the form: Nd Bb FirstMessage Nd Bb NextMesaage Nd Bb LastMessage where N is the decimal value, B is the value in the current binary integer display mode, and b is h, d, o, or b, depending on the display mode: %%HP; : C$ $ !NO CODE !RPL * Checksum: # A457h * Memory usage: 145. * Execution time with 1 and 255: about 2.98 seconds :: CK2NOLASTWD :: CK&DISPATCH1 # BB :: HXS>% SWAP HXS>% ; BINT17 SWAP BINT27 :: HXS>% ; # B1 :: SWAP HXS>% ; ; COERCE2 2DUP #> IT SWAP #1+ SWAP NULL$ UNROT GARBAGE DO ?ATTNQUIT INDEX@ DUP #>$ CHR_d >T$ OVER #>HXS HXS>$ CDR$ &$ APPEND_SPACE tokquote &$ SWAP JstGETTHEMSG &$ tokquote &$ NEWLINE$&$ &$ LOOP ; @ The above program requires 2 arguments, each of the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me > to install Xmodem driver, > calculator not found, timeout, etc. Does Conn4x see HPx9G+ [driver] in Connect using? On calc, what shows in APPS > I/O functions > Transfer next to Port: ? [r->] [OFF] === Subject: TED on HP50G Using TED on my new HP50G, cursor is moving two places on every key stroke. What I am doing wrong? Sorry for bed English. Damir === Subject: TED on HP50G Using TED on my new HP50G, cursor is moving two places on every key stroke. What I am doing wrong? Sorry for bad English. Damir === Subject: Oops -- please change title ... I just posted a reply to James Prange on the thread Error Messages are Useful Strings but it really should be a new thread -- it's about text translation and Mac Connectivity. I'd call it Translation from Conn4x ASCII display format How does one change the title of a thread, or move a post to a new thread? Irl === Subject: Re: Oops -- please change title ... > I just posted a reply to James Prange on the thread Error Messages are > Useful Strings but it really should be a new thread -- it's about text > translation and Mac Connectivity. I'd call it Translation from Conn4x > ASCII display format > How does one change the title of a thread, or move a post to a new > thread? Well, this is usenet, and I don't think that you can do anything to change it once it's posted. You could try to cancel a post, but it usually gets to several servers pretty darn quickly, and most of them don't honour cancel commands. I see that you're posting via Google, and I understand that Google does honour some sort of don't archive header, so such posts won't be stored there forever (although they may be on someone's PC). Other usenet servers don't archive posts, so eventually old posts are discarded from them to make room for new posts. The correct way to change a subject within a thread on usenet: New subject (was: Old subject) The better newsreaders will automatically discard the (was: Old subject) part and use just the New subject part instead of using New subject (was: Old subject) when they insert the subject header in a reply, but apparently a lot of newreaders don't. If your newsreader doesn't automatically discard the (was: Old subject) part when composing a reply, then you should edit it out yourself. programs to convert between the compiled object form (like what you'd see with a binary transfer), and the decompiled to a string and optionally translated form (like what you'd see with a Kermit ASCII or Conn4x Text transfer). They were intended to make it easier use the SD card for ASCII transfers, but they could be used for any binary transfer, or any other situation where you want to convert between these forms. You can find my programs at: http://www.hpcalc.org/search.php?query=%22ASCII+on+SD%22 Another (perhaps better) application that does something similar can be found at: http://www.hpcalc.org/search.php?query=ASCII+to+Binary+Converter+ Perhaps there are other at hpcalc.org. Regarding transfers with a Mac, try: http://hpconnect.sourceforge.net/ -- James === Subject: Translation from Conn4x,ASCII display format <9d00c$457604f9$4267eb85$28654@123.NET> James -- I will look at those links. Sounds like the solution to my problem. ... I'm feeling like I ought to contribute something to this group besides newbie questions, but I'm strictly a UserRPL optics/physics/math applications programmer. Irl > I just posted a reply to James Prange on the thread Error Messages are > Useful Strings but it really should be a new thread -- it's about text > translation and Mac Connectivity. I'd call it Translation from Conn4x > ASCII display format > How does one change the title of a thread, or move a post to a new > thread? > > Well, this is usenet, and I don't think that you can do anything > to change it once it's posted. You could try to cancel a post, but > it usually gets to several servers pretty darn quickly, and most > of them don't honour cancel commands. > > I see that you're posting via Google, and I understand that Google > does honour some sort of don't archive header, so such posts > won't be stored there forever (although they may be on someone's > PC). Other usenet servers don't archive posts, so eventually old > posts are discarded from them to make room for new posts. > > The correct way to change a subject within a thread on usenet: > > New subject (was: Old subject) > > The better newsreaders will automatically discard the > (was: Old subject) part and use just the New subject part > instead of using New subject (was: Old subject) when they insert > the subject header in a reply, but apparently a lot of newreaders > don't. If your newsreader doesn't automatically discard the > (was: Old subject) part when composing a reply, then you should > edit it out yourself. > > programs to convert between the compiled object form (like what > you'd see with a binary transfer), and the decompiled to a string > and optionally translated form (like what you'd see with a Kermit > ASCII or Conn4x Text transfer). They were intended to make it > easier use the SD card for ASCII transfers, but they could be used > for any binary transfer, or any other situation where you want to > convert between these forms. You can find my programs at: > http://www.hpcalc.org/search.php?query=%22ASCII+on+SD%22 > > Another (perhaps better) application that does something similar > can be found at: > http://www.hpcalc.org/search.php?query=ASCII+to+Binary+Converter+ > > Perhaps there are other at hpcalc.org. > > Regarding transfers with a Mac, try: > http://hpconnect.sourceforge.net/ > > -- > James === Subject: Translation from Conn4x,ASCII display format (was: Oops -- please change title ...) PS: I should've changed the subject myself, which I'll do with this post. -- James === Subject: Question about USB Cable of HP 50G Is the USB Cable that come with the HP 50G a normal USB cable? I can buy another cable in any store, or the USB cable have some special characteristics ? === Subject: Re: Question about USB Cable of HP 50G > Is the USB Cable that come with the HP 50G a normal USB cable? > I can buy another cable in any store, or the USB cable have some > special characteristics ? Yes it is a 'normal' USB to miniUSB cable, manjo === Subject: Re: Question about USB Cable of HP 50G manjo ha escrito: > Is the USB Cable that come with the HP 50G a normal USB cable? > I can buy another cable in any store, or the USB cable have some > special characteristics ? > > Yes it is a 'normal' USB to miniUSB cable, > > manjo What about the serial cable. Is there any place where I can get one ? Why is it so particular ? Daniel === Subject: Re: Question about USB Cable of HP 50G > What about the serial cable. Is there any place where I can get one ? > Why is it so particular ? Serial cable had to be different from USB (Although it was written in the manual on 48GII, there were 'problems' since users tried to connect it to USB, because connector was the same type) Further : Since CPU is delivering 3.3V logic (not RS232 levels) designers figured power should be available on the connector for 'level shifter' which would be bilt in the cable connector. I guess they choose that particular type because they wanted to use different type from the older models due to signal levels incompatibility, old cables would not work. -from what i said before -users don't pay much attention to users manual -this way users wouldn't get the idea to connect the old cable expecting it to work so much for now, manjo === Subject: Debug4x b104 UserRPL Additions Debug4x build 104 is now available from http://www.Debug4x.com The User RPL editor now has the Verb Completion ctrl-space listing of verbs, the stack pictures ctrl-j and the code completion (macros) alt-ins just like the SysRPL editors. Some other improvements for the User RPL system and some bug fixes. Have a nice Christmas everyone, I am vacationing! -- - - - - - - - - - - - - - - - - Bill Graves RKBA! bgraves@ix.netcom.com === Subject: Re: Debug4x b104 UserRPL Additions > Debug4x build 104 is now available from > http://www.Debug4x.com > > The User RPL editor now has the Verb Completion ctrl-space listing of > verbs, > the stack pictures ctrl-j > and the code completion (macros) alt-ins > just like the SysRPL editors. > > Some other improvements for the User RPL system and some bug fixes. > > Have a nice Christmas everyone, I am vacationing! > -- - - - - - - - - - - - - - - - - > Bill Graves RKBA! > bgraves@ix.netcom.com Downloading... === Subject: Re: Debug4x b104 UserRPL Additions > >> Debug4x build 104 is now available from >> http://www.Debug4x.com >> >> The User RPL editor now has the Verb Completion ctrl-space listing of >> verbs, >> the stack pictures ctrl-j >> and the code completion (macros) alt-ins >> just like the SysRPL editors. >> >> Some other improvements for the User RPL system and some bug fixes. >> >> Have a nice Christmas everyone, I am vacationing! >> -- - - - - - - - - - - - - - - - - >> Bill Graves RKBA! >> bgraves@ix.netcom.com > Downloading... Downloaded! === Subject: Re: Debug4x b104 UserRPL Additions > >> > Debug4x build 104 is now available from > http://www.Debug4x.com > > The User RPL editor now has the Verb Completion ctrl-space listing of > verbs, > the stack pictures ctrl-j > and the code completion (macros) alt-ins > just like the SysRPL editors. > > Some other improvements for the User RPL system and some bug fixes. > > Have a nice Christmas everyone, I am vacationing! > -- - - - - - - - - - - - - - - - - > Bill Graves RKBA! > bgraves@ix.netcom.com >> Downloading... > Downloaded! Unzipping... === Subject: Re: Debug4x b104 UserRPL Additions > >> > >> Debug4x build 104 is now available from >> http://www.Debug4x.com >> >> The User RPL editor now has the Verb Completion ctrl-space listing of >> verbs, >> the stack pictures ctrl-j >> and the code completion (macros) alt-ins >> just like the SysRPL editors. >> >> Some other improvements for the User RPL system and some bug fixes. >> >> Have a nice Christmas everyone, I am vacationing! >> -- - - - - - - - - - - - - - - - - >> Bill Graves RKBA! >> bgraves@ix.netcom.com > Downloading... >> Downloaded! > Unzipping... Unzipped! === Subject: Re: Debug4x b104 UserRPL Additions > >> > >> >>> Debug4x build 104 is now available from >>> http://www.Debug4x.com >>> >>> The User RPL editor now has the Verb Completion ctrl-space listing >>> of verbs, >>> the stack pictures ctrl-j >>> and the code completion (macros) alt-ins >>> just like the SysRPL editors. >>> >>> Some other improvements for the User RPL system and some bug fixes. >>> >>> Have a nice Christmas everyone, I am vacationing! >>> -- - - - - - - - - - - - - - - - - >>> Bill Graves RKBA! >>> bgraves@ix.netcom.com >> Downloading... > Downloaded! >> Unzipping... > Unzipped! Starting Installation... === Subject: Re: Debug4x b104 UserRPL Additions > >> > >> >>> >> Debug4x build 104 is now available from >> http://www.Debug4x.com >> >> The User RPL editor now has the Verb Completion ctrl-space listing >> of verbs, >> the stack pictures ctrl-j >> and the code completion (macros) alt-ins >> just like the SysRPL editors. >> >> Some other improvements for the User RPL system and some bug fixes. >> >> Have a nice Christmas everyone, I am vacationing! >> -- - - - - - - - - - - - - - - - - >> Bill Graves RKBA! >> bgraves@ix.netcom.com >>> Downloading... >> Downloaded! > Unzipping... >> Unzipped! > Starting Installation... Installed...testing..calc emulator works...replacing the old... continue testing.... === Subject: Re: Debug4x b104 UserRPL Additions >>> Debug4x build 104 is now available from >>> http://www.Debug4x.com >>> >>> The User RPL editor now has the Verb Completion ctrl-space listing >>> of verbs, >>> the stack pictures ctrl-j >>> and the code completion (macros) alt-ins >>> just like the SysRPL editors. >>> >>> Some other improvements for the User RPL system and some bug fixes. >>> >>> Have a nice Christmas everyone, I am vacationing! >>> -- - - - - - - - - - - - - - - - - >>> Bill Graves RKBA! >>> bgraves@ix.netcom.com >> Downloading... >>> Downloaded! >> Unziem the same as for the previous programs, and with the values in either order. Like the others, if the message doesn't exist, then an empty string is returned for that value. At each iteration of the loop, it checks whether CANCEL has been pressed, to allow the user to cancel out of a long-running execution, although actually, it will either finish or error out when it uses up all available memory within a few minutes anyway. These programs will get strings from the message table of any external library that you have stored in port 0, 1, or 2, even if it's not attached, as well from the ROM library message tables. Of course these programs get only message strings that are in library message tables. They don't get text elsewhere in ROM; try Tim's suggestion of using Emacs for that. To compile one of these SysRPL programs, first you have to have an extable library attached; see: http://www.hpcalc.org/details.php?id=3245 (The filename for the library is extable.HP.) Also, the Emacs package includes the Extable2 library, which also includes unsupported but stable entries; see: http://www.hpcalc.org/details.php?id=3940 After installing an extable library, put the source code string on the stack and execute ASM. To decompile the SysRPL program to a character string again, put the program on the stack and execute ->S2. ASM and ->S2 can be entered pping... > Unzipped! >> Starting Installation... > Installed...testing..calc emulator works...replacing the old... > continue testing.... Darm! Since my display is using 120dpi instead of 96dpi some windows overlap slighly also some text is not showing properly otherwise...ok (so far) VPN === Subject: Re: Debug4x b104 UserRPL Additions > Since my display is using 120dpi instead of 96dpi some windows overlap > slighly also some text is not showing properly otherwise...ok (so far) > VPN VPN Send me details if the Windows cannot be moved or sized to work properly. I thought I had fixed that for you some time back! (A picture or two would help as well). -- - - - - - - - - - - - - - - - - Bill Graves RKBA! bgraves@ix.netcom.coms === Subject: Re: Debug4x b104 UserRPL Additions > >> > >> >>> >> Debug4x build 104 is now available from >> http://www.Debug4x.com >> >> The User RPL editor now has the Verb Completion ctrl-space listing >> of verbs, >> the stack pictures ctrl-j >> and the code completion (macros) alt-ins >> just like the SysRPL editors. >> >> Some other improvements for the User RPL system and some bug fixes. >> >> Have a nice Christmas everyone, I am vacationing! >> -- - - - - - - - - - - - - - - - - >> Bill Graves RKBA! >> bgraves@ix.netcom.com >>> Downloading... >> Downloaded! > Unzipping... >> Unzippfrom the command line, as long as the development library 256 is attached (which installing extable does automatically), and are also available in menu 256.06. Actually, you can compile or decompile a SysRPL program without having extable installed, but in that case, most SysRPL commands are numeric pointers instead of mnemonic names. For example, an alternative source code file for the short SysRPL program that uses GETTHEMESG is: %%HP: ; C$ $ !NO CODE !RPL * Checksum: # A78Fh * Memory usage: 35. * Execution time with a zint: about .004 second :: PTR 26297 PTR 26328 # B :: PTR 05A03 PTR 04D64 : # 1 :: PTR 262F1 PTR 04D64 ; ; @ -- James === Subject: Re: Error messages are useful strings -- which? James -- SysRPL program (as a command-line entry, presumably) to actually be a program was useful; I have tried displaying the SysRPL stack (-85 SF) with a tiny UserRPL program on the stack and then typing in the same thing but couldn't get it to actually be a program. Here's a presumably dumb question: there are two extable files at hpcalc.org. One is for the meta kernel and is much shorter than the other one. What are these, please? Regarding the original question: I've run my exhaustive-search-for-all-messages routine (the IFERR n DOERR THEN ... with n looping from 1 to infinity) and generated a list of everything below 100000 decimal = #186A0h, but it does take a while. I guess I could take advantage of your info and make my program advance by multiples of #100h and skip to the next multiple if there is no message at M*#100h+1. What I was kind of hoping was for a list somewhere of the indirection pointers. In other words, I was expecting that somewhere in ROM was a packed list of messages, and elsewhere would be a table that says Message 12345 is at offset abcd in the message table. So I could just page through the table looking at messages which actually exist. Yours, Irl > Raymond -- > guess I wasn't clear that I'm using an HP50g, so the binary libraries > won't work without disassembly and recompilation -- right? > Also, the list of messages gives only the error messages, as opposed to > all the useful other text strings which are accessible via the error > handler but which don't actually correspond to errors. > > I'm not Raymond, but I'll contribute to this discussion anyway. > I posted the English language strings (but neglected the French > and Spanish strings) in the message tables of ROM revision 1.19-6 > for the 49G to hpcalc.org; see: > http://www.hpcalc.org/search.php?query=message+tables > > That file is a bit dated; for example, some messages for the > SD card have been added to the newer ROM revisions. My impression > is that any unused messages are left in a ROM library message > table when the ROM is revised, but each library has a maximum of > 255 messages, so I wouldn't entirely rule out replacing unused > messages with new ones, or otherwise changing old messages. > > That file includes messages from the equation library, but with > the current ROM, the equation library and periodic table are > optional, and their messages won't be available if you don't have > those libraries installed. > > Some message in the tables are for errors, but some are for other > purposes. The advantages of using a message table are, first, even > if the message is used for more than routine, it can be stored in > one place and called by number (a bit slower but usually reducing > memory usage), and second, it's easier to arrange for > multi-language messages. > > Message numbers are (in hex notation) of the form LLLMM, where LLL > is the library number, from 000h through 7FFh, and MM is the > message number within a library, from 01h through FFh. Another way > of looking at it is that the message number is LLLh * 100h + MMh, > or in decimal notation, LibNum * 256 + MsgNum, where LibNum is 0 > through 2047 and MsgNum is 1 through 255. > > The message tables are arrays (or in some cases linked arrays) > of strings. Some built-in message tables are English only; these > are 1-dimensional arrays or 2-dimensional arrays with one of the > dimensions being 1. Some have messages in French or both French > and Spanish as well; these are 2-dimensional arrays with one of > the dimensions being 2 or 3. > > If the LANGUAGE setting is French or Spanish and the message table > doesn't have that language, then the English string will be used. > A very few French or Spanish messages are empty strings. > > The UserRPL program: > > %%HP: ; > @ Checksum: # 5C51h > @ Memory usage: 22.5 > @ Execution time: about .032 second > << > IFERR > DOERR > THEN > ERRM > END >> > > will get any message in the tables (using the current LANGUAGE > setting), as well as the user-defined error message for > #70000h (458752), except that an IFERR structure doesn't trap > errors #123h (291) (used by the KILL command), #13Eh (318) (used > by the CONT command), and #DFFh (3583) (used by control alarms). > Messages #123h and #13Eh are empty strings anyway, and message > #DFFh doesn't exist. > > Source code files for SysRPL alternatives are: > > %%HP: ; > C$ $ > !NO CODE > !RPL > * Checksum: # A78Fh > * Memory usage: 35. > * Execution time with a zint: about .004 second > :: > CK1NOLASTWD > CK&DISPATCH1 > BINT11 > :: > HXS># > GETTHEMESG > ; > BINT1 > :: > COERCE > GETTHEMESG > ; > ; > @ > > and: > > %%HP: ; > C$ $ > !NO CODE > !RPL > * Checksum: # 3DDCh > * Memory usage: 35. > * Execution time with a zint: about .004 second > :: > CK1NOLASTWD > CK&DISPATCH1 > BINT11 > :: > HXS># > JstGETTHEMSG > ; > BINT1 > :: > COERCE > JstGETTHEMSG > ; > ; > @ > > The difference between those SysRPL examples is that the one that > uses GETTHEMESG will get the user-defined error message for an > argument of #70000h (458752), and the one that uses JstGETTHEMSG > doesn't check for #70000h. > > All of these programs return an empty string if the message > doesn't exist. > > All three of the above programs take a binary integer, real > number, or zint (exact integer), or any of these tagged, for the > argument. A negative real or zint is treated as zero. A real with > a fractional part is rounded to an integer value. A value over > #7FFFFh (524287) is treated as modulus #80000h (524288), so that, > for example, #1h and #80001h get the same message; I surmise that > only the 19 least significant bits are used when getting messages > from the tables. > > My above SysRPL examples should be faster when used with a binary > integer or real for the argument. The argument checking sequence > is, first, verify at least one object on the stack for the > argument, then look for a binary integer, then look for a real, > then strip any tags from the argument and convert a zint to a > real, then look for a binary integer, then look for a reaed! > Starting Installation... Installed...testing... === Subject: Re: Debug4x b104 UserRPL Additions <2jDdh.1125$O8.54@reader1.news.saunalahti.fi> >>> Downloading... >> Downloaded! > Unzipping... >> Unzipped! > Starting Installation... > Installed...testing... ...? - GaaK - === Subject: SD Capacity wrong in Files I have a HP49G+ with version 4.20060602 (VER) and 2.09 (VERSION). I have installed a 1 Gbyte SD card. The card has been formatted both on a PC and the HP, however when I look in files it shows the capacity as being 55842kB or around 55Mbytes. Does anybody have an explanation. === Subject: Re: SD Capacity wrong in Files if you mean the free capacity on the SD card, maybe you are experiencing the slack space effect. For more info on this: http://www.webopedia.com/TERM/S/slack_space.html HTH Raymond schrieb im Newsbeitrag >I have a HP49G+ with version 4.20060602 (VER) and 2.09 (VERSION). I > have installed a 1 Gbyte SD card. The card has been formatted both on a > PC and the HP, however when I look in files it shows the capacity as > being 55842kB or around 55Mbytes. Does anybody have an explanation. > > === Subject: Re: SD Capacity wrong in Files >I have a HP49G+ with version 4.20060602 (VER) and 2.09 (VERSION). I > have installed a 1 Gbyte SD card. The card has been formatted both on a > PC and the HP, however when I look in files it shows the capacity as > being 55842kB or around 55Mbytes. Does anybody have an explanation. your card most probably has 955842kB or around 955 MBytes (maybe even 1055842kB) but... leftmost digits are probably not visible due to limited display space available. Don't worry, your SD card is in there, your space will be used, *and* it will be displayed correctly once the size reduces do displayable figure. either size should be displayed as MB when more than 90000 kB is available *or* mem sizes should be moved at least one space to the right so larger sizes could be displayed. manjo l, so > zints and tagged reals are slowest. > > A source code file for a SysRPL program that gets messages over a > range of numbers and returns a string in the form: > > Nd Bb FirstMessage > Nd Bb NextMesaage > Nd Bb LastMessage > > > where N is the decimal value, B is the value in the current binary > integer display mode, and b is h, d, o, or b, depending on the > display mode: > > %%HP; : > C$ $ > !NO CODE > !RPL > * Checksum: # A457h > * Memory usage: 145. > * Execution time with 1 and 255: about 2.98 seconds > :: > CK2NOLASTWD > :: > CK&DISPATCH1 > # BB > :: > HXS>% > SWAP > HXS>% > ; > BINT17 > SWAP > BINT27 > :: > HXS>% > ; > # B1 > :: > SWAP > HXS>% > ; > ; > COERCE2 > 2DUP > #> > IT > SWAP > #1+ > SWAP > NULL$ > UNROT > GARBAGE > DO > ?ATTNQUIT > INDEX@ > DUP > #>$ > CHR_d > >T$ > OVER > #>HXS > HXS>$ > CDR$ > &$ > APPEND_SPACE > tokquote > &$ > SWAP > JstGETTHEMSG > &$ > tokquote > &$ > NEWLINE$&$ > &$ > LOOP > ; > @ > > The above program requires 2 arguments, each of them the same as > for the previous programs, and with the values in either order. > Like the others, if the message doesn't exist, then an empty > string is returned for that value. > > At each iteration of the loop, it checks whether CANCEL has been > pressed, to allow the user to cancel out of a long-running > execution, although actually, it will either finish or error out > when it uses up all available memory within a few minutes anyway. > > These programs will get strings from the message table of any > external library that you have stored in port 0, 1, or 2, even if > it's not attached, as well from the ROM library message tables. > > Of course these programs get only message strings that are in > library message tables. They don't get text elsewhere in ROM; try > Tim's suggestion of using Emacs for that. > > To compile one of these SysRPL programs, first you have to have an > extable library attached; see: > http://www.hpcalc.org/details.php?id=3245 > (The filename for the library is extable.HP.) > Also, the Emacs package includes the Extable2 library, which also > includes unsupported but stable entries; see: > http://www.hpcalc.org/details.php?id=3940 > > After installing an extable library, put the source code string on > the stack and execute ASM. To decompile the SysRPL program to a > character string again, put the program on the stack and execute > ->S2. ASM and ->S2 can be entered from the command line, as long > as the development library 256 is attached (which installing > extable does automatically), and are also available in > menu 256.06. > > Actually, you can compile or decompile a SysRPL program without > having extable installed, but in that case, most SysRPL commands > are numeric pointers instead of mnemonic names. For example, an > alternative source code file for the short SysRPL program that > uses GETTHEMESG is: > > %%HP: ; > C$ $ > !NO CODE > !RPL > * Checksum: # A78Fh > * Memory usage: 35. > * Execution time with a zint: about .004 second > :: > PTR 26297 > PTR 26328 > # B > :: > PTR 05A03 > PTR 04D64 > : > # 1 > :: > PTR 262F1 > PTR 04D64 > ; > ; > @ > > > > -- > James === Subject: Re: Error messages are useful strings -- which? First off, a correction: In the ASCII transfer header for that long program for getting a range of messages, I reversed the colon and semicolon. It should be %%HP: ;, not %%HP; :. I copied and pasted the source code as transferred from my 49g+, but typed in the transfer header, counted string delimiter, and comments, and I got those reversed. Also, now that I've thought about it a bit more, for those message numbers that are in a valid range but DOERR shows nothing after Error:, and ERRM, GETTHEMESG, and JstGETTHEMSG return an empty string, I think that it's more likely a matter of the linked array having a null pointer and no element for that message, rather than containing an empty string. I surmise that those commands return an empty string for any non-existent message. > James -- > SysRPL program (as a command-line entry, presumably) You're welcome. Either transfer the source code string or key in the source code in the command line within a character string. You should have the source code string in level 1 when you do the ASM command. Note that the source code string has to end with an @ character all by itself as the last line. Note that most SysRPL commands don't do any argument checking at all, so a buggy SysRPL program can corrupt memory and you could lose everything that you have stored in home and it's subdirectories and port 0. If you have anything that you don't want to lose, then archive it first, at least to a port other than 0, and preferably to an SD card and from there to your PC. The intention was that you could copy and paste the source code, starting with the %%HP: ; transfer header and through the @ line, to a plain text file, transfer the file to the calculator, recall the source code string to the stack, and execute the ASM command on it. > to actually be a > program was useful; I have tried displaying the SysRPL stack (-85 SF) > with a tiny UserRPL program on the stack and then typing in the same > thing but couldn't get it to actually be a program. > Here's a presumably dumb question: there are two extable files at > hpcalc.org. One is for the meta kernel and is much shorter than the > other one. What are these, please? The Meta Kernel was written for the 48 series, and it, along with some other very good applications written for the 48 series, was incorporated into the system for the 49 series. I wasn't aware of an extable library for the Meta Kernel at hpcalc.org, but I'd assume that it's for the Meta Kernel as designed for the 48 series, not for the 49 series. For the official HP extable for the 49 series with supported entry points only, use the one at: http://www.hpcalc.org/details.php?id=3245 For a better extable that includes unsupported but stable entry points, use extable2. I don't find extable2 as a separate downloadable file, but it's included in the Emacs package; see: http://www.hpcalc.org/details.php?id=3940 or: http://staff.science.uva.nl/~dominik/hpcalc/ > Regarding the original question: I've run my > exhaustive-search-for-all-messages routine (the IFERR n DOERR THEN ... > with n looping from 1 to infinity) and generated a list of everything > below 100000 decimal = #186A0h, but it does take a while. Then you should have all of them except #31401h, No Message here. > I guess I > could take advantage of your info and make my program advance by > multiples of #100h and skip to the next multiple if there is no message > at M*#100h+1. That should work, as long as none of the message arrays have a null first element (and none of them actually do). What I did was first run a quick-and-dirty SysRPL program to look for message tables: %%HP: T(3)A(R)F(.); !NO CODE !RPL * Uses 1 unsupported but stable entry. * Takes about 35 seconds to run. * Checksum: # 2FDFh * Size: 113. :: CK0NOLASTWD NULL$ # 800 BINT0 DO INDEX@ DUP #>$ CHR_d >T$ OVER #>HXS HXS>$ CDR$ &$ APPEND_SPACE SWAP PTR 08130 (GETMSG_) ITE :: ?ACCPTR> DUP xTYPE a%>$, APPEND_SPACE SWAP xBYTES SWAPDROP a%>$, &$ ; :: % 0. a%>$, ; &$ NEWLINE$&$ &$ LOOP ; @ This told me which libraries actually have message tables (as well as which type of array they were and the size of the arrays). Then I ran my program to get a range of messages for only those libraries which actually have messages. Assuming that your e-mail address is valid, I'll send you a copy of my text file of only the English messages. To satisfy my curiosity a bit more, I ran this next program for each library that has a message table to get the array itself. %%HP: T(3)A(R)F(.); !NO CODE !RPL * Uses 1 unsupported but stable entry. * Checksum: # 8C98h * Size: 42.5 :: CK1NOLASTWD :: CK&DISPATCH1 BINT11 :: HXS># ; BINT1 :: COERCE ; ; PTR 08130 (GETMSG_) IT ?ACCPTR> ; @ Now, even though the type 4 arrays can be displayed on the stack, no UserRPL command is expecting an array to be an array of character strings, so I wouldn't try doing much with one without first doing ->STR to put it into a string that's safe to work with. But I did try the SIZE command, and that worked and returned a list of the dimensions. It turns out that all of the type 4 message arrays are 1-dimensional. Obviously, the system has ways of working with linked arrays (type 23), but I didn't find much information on working with them. But RPLMAN.DOC describes them and HP 48S/SX Machine Language gives the details of their structure. I used ->H to turn them into strings of hex nibbles to look at the dimension count and dimension sizes. So far, I've derived this information about the message table arrays: Library Array Array Object size ID type dimensions in bytes 0d 0h 23 { 46d 3d} {2Eh 3h} 3178. 1d 1h 23 { 89d 3d} {59h 3h} 4388.5 2d 2h 23 { 8d 3d} { 8h 3h} 618. 3d 3h 23 { 5d 3d} { 5h 3h} 356. 5d 5h 23 { 6d 3d} { 6h 3h} 428. 6d 6h 23 { 47d 3d} {2Fh 3h} 3323. 7d 7h 23 {238d} {EEh} 4558.5 8d 8h 23 {176d} {B0h} 3149.5 9d 9h 23 { 13d 1d} { Dh 1h} 8883. 10d Ah 23 { 9d 3d} { 9h 3h} 375. 11d Bh 23 { 2d 3d} { 2h 3h} 148. 12d Ch 23 { 23d 3d} {17h 3h} 1364. 13d Dh 23 { 4d 3d} { 4h 3h} 268. 185d B9h 4 {224d} {E0h} 4190.5 186d BAh 4 { 81d} {51h} 1088. 187d BBh 4 { 65d} {41h} 1127. 188d BCh 4 { 59d} {3Bh} 735. 189d BDh 4 { 39d} {27h} 703. 190d BEh 23 {119d} {77h} 2540.5 191d BFh 4 { 86d} {56h} 1491.5 192d C0h 4 { 1d} { 1h} 34. 222d DEh 23 {104d 2d} {68h 2h} 5188. 223d DFh 23 { 50d 3d} {32h 3h} 2046. 225d E1h 4 { 41d} {29h} 612. 228d E4h 4 { 8d} { 8h} 156.5 230d E6h 4 { 13d} { Dh} 205. 256d 100h 23 { 6d 2d} { 6h 2h} 261. 257d 101h 4 { 24d} {18h} 427.5 788d 314h 4 { 1d} { 1h} 30. Note that the above is for only built-in ROM libraries, it doesn't include any external libraries, not even the equation library or periodic table library. 1-dimensional arrays and the array with the 2nd dimension being 1 are English language only, arrays with the 2nd dimension of 2 are English and French, and arrays with the 2nd dimension of 3 are English, French, and Spanish. If the calculator's LANGUAGE is set to French or Spanish and the particular library's message table doesn't have that language, then the English message is used instead. The library message tables are implemented as arrays of strings. Of course an array starts with the prologue address for it's own type, and then its total body length, but it also includes one copy of the prologue address for the type of elements that it contains, which are bodies of that object type, not complete objects per se. It also contains a count of how many dimensions it has, and the size of each dimension. A type 4 array is the normal (albeit not necessarily numeric) array, with all elements present and in order. A type 23 array is a linked array, with a series of pointers to the elements. The pointers must be present and in order, but the elements can be in any order, more than one pointer can point to the same element (object body), elements need not be present, and a null (00000) pointer is used for any missing elements. I'll probably dig into the arrays a bit more; I'm curious about those null messages. I believe that I see a way to convert a linked array of strings to a list array of strings, and it should be just a matter of actually writing the program. I'm assuming that the SysRPL commands for working with arrays are designed for normal arrays instead of linked arrays, but if I start out by using ->H on a linked array, I expect that I can deal with it. Note that a type 29 symbolic matrix is rather different from these arrays; It's composite object, and very much like a list array. > What I was kind of hoping was for a list somewhere of the indirection > pointers. In other words, I was expecting that somewhere in ROM was a > packed list of messages, and elsewhere would be a table that says > Message 12345 is at offset abcd in the message table. So I could > just page through the table looking at messages which actually exist. What you could do is store a plain text file of the messages and their addresses as a string on the calculator, perhaps on an SD card. As long as you know at least part of the message that you want, you could use the editor's FIND operation to search for it. -- James === Subject: Re: Error messages are useful strings -- which? <204f1$45759d75$4267eab7$10037@123.NET> Yes, my email is correct (not my main one, obviously, but a real one I check every day). Irl > First off, a correction: In the ASCII transfer header for that > long program for getting a range of messages, I reversed the colon > and semicolon. It should be %%HP: ;, not %%HP; :. I copied and > pasted the source code as transferred from my 49g+, but typed in > the transfer header, counted string delimiter, and comments, and I > got those reversed. > > Also, now that I've thought about it a bit more, for those message > numbers that are in a valid range but DOERR shows nothing after > Error:, and ERRM, GETTHEMESG, and JstGETTHEMSG return an empty > string, I think that it's more likely a matter of the linked array > having a null pointer and no element for that message, rather than > containing an empty string. I surmise that those commands return > an empty string for any non-existent message. > > James -- > SysRPL program (as a command-line entry, presumably) > > You're welcome. > > Either transfer the source code string or key in the source code > in the command line within a character string. You should have the > source code string in level 1 when you do the ASM command. > > Note that the source code string has to end with an @ character > all by itself as the last line. > > Note that most SysRPL commands don't do any argument checking at > all, so a buggy SysRPL program can corrupt memory and you could > lose everything that you have stored in home and it's > subdirectories and port 0. If you have anything that you don't > want to lose, then archive it first, at least to a port other than > 0, and preferably to an SD card and from there to your PC. > > The intention was that you could copy and paste the source code, > starting with the %%HP: ; transfer header and through the @ line, > to a plain text file, transfer the file to the calculator, recall > the source code string to the stack, and execute the ASM command > on it. > > to actually be a > program was useful; I have tried displaying the SysRPL stack (-85 SF) > with a tiny UserRPL program on the stack and then typing in the same > thing but couldn't get it to actually be a program. > Here's a presumably dumb question: there are two extable files at > hpcalc.org. One is for the meta kernel and is much shorter than the > other one. What are these, please? > > The Meta Kernel was written for the 48 series, and it, along with > some other very good applications written for the 48 series, was > incorporated into the system for the 49 series. I wasn't aware of > an extable library for the Meta Kernel at hpcalc.org, but I'd > assume that it's for the Meta Kernel as designed for the 48 > series, not for the 49 series. > > For the official HP extable for the 49 series with supported > entry points only, use the one at: > http://www.hpcalc.org/details.php?id=3245 > > For a better extable that includes unsupported but stable > entry points, use extable2. I don't find extable2 as a separate > downloadable file, but it's included in the Emacs package; see: > http://www.hpcalc.org/details.php?id=3940 > or: > http://staff.science.uva.nl/~dominik/hpcalc/ > > Regarding the original question: I've run my > exhaustive-search-for-all-messages routine (the IFERR n DOERR THEN ... > with n looping from 1 to infinity) and generated a list of everything > below 100000 decimal = #186A0h, but it does take a while. > > Then you should have all of them except #31401h, No Message > here. > > I guess I > could take advantage of your info and make my program advance by > multiples of #100h and skip to the next multiple if there is no message > at M*#100h+1. > > That should work, as long as none of the message arrays have a > null first element (and none of them actually do). > > What I did was first run a quick-and-dirty SysRPL program to look > for message tables: > > %%HP: T(3)A(R)F(.); > !NO CODE > !RPL > * Uses 1 unsupported but stable entry. > * Takes about 35 seconds to run. > * Checksum: # 2FDFh > * Size: 113. > :: > CK0NOLASTWD > NULL$ > # 800 > BINT0 > DO > INDEX@ > DUP > #>$ > CHR_d > >T$ > OVER > #>HXS > HXS>$ > CDR$ > &$ > APPEND_SPACE > SWAP > PTR 08130 (GETMSG_) > ITE > :: > ?ACCPTR> > DUP > xTYPE > a%>$, > APPEND_SPACE > SWAP > xBYTES > SWAPDROP > a%>$, > &$ > ; > :: > % 0. > a%>$, > ; > &$ > NEWLINE$&$ > &$ > LOOP > ; > @ > > This told me which libraries actually have message tables (as well > as which type of array they were and the size of the arrays). Then > I ran my program to get a range of messages for only those > libraries which actually have messages. > > Assuming that your e-mail address is valid, I'll send you a copy > of my text file of only the English messages. > > To satisfy my curiosity a bit more, I ran this next program for > each library that has a message table to get the array itself. > > %%HP: T(3)A(R)F(.); > !NO CODE > !RPL > * Uses 1 unsupported but stable entry. > * Checksum: # 8C98h > * Size: 42.5 > :: > CK1NOLASTWD > :: > CK&DISPATCH1 > BINT11 > :: > HXS># > ; > BINT1 > :: > COERCE > ; > ; > PTR 08130 (GETMSG_) > IT > ?ACCPTR> > ; > @ > > Now, even though the type 4 arrays can be displayed on the stack, > no UserRPL command is expecting an array to be an array of > character strings, so I wouldn't try doing much with one without > first doing ->STR to put it into a string that's safe to work > with. > > But I did try the SIZE command, and that worked and returned a > list of the dimensions. It turns out that all of the type 4 > message arrays are 1-dimensional. > > Obviously, the system has ways of working with linked arrays (type > 23), but I didn't find much information on working with them. But > RPLMAN.DOC describes them and HP 48S/SX Machine Language gives > the details of their structure. I used ->H to turn them into > strings of hex nibbles to look at the dimension count and > dimension sizes. So far, I've derived this information > about the message table arrays: > > > Library Array Array Object size > ID type dimensions in bytes > > 0d 0h 23 { 46d 3d} {2Eh 3h} 3178. > 1d 1h 23 { 89d 3d} {59h 3h} 4388.5 > 2d 2h 23 { 8d 3d} { 8h 3h} 618. > 3d 3h 23 { 5d 3d} { 5h 3h} 356. > 5d 5h 23 { 6d 3d} { 6h 3h} 428. > 6d 6h 23 { 47d 3d} {2Fh 3h} 3323. > 7d 7h 23 {238d} {EEh} 4558.5 > 8d 8h 23 {176d} {B0h} 3149.5 > 9d 9h 23 { 13d 1d} { Dh 1h} 8883. > 10d Ah 23 { 9d 3d} { 9h 3h} 375. > 11d Bh 23 { 2d 3d} { 2h 3h} 148. > 12d Ch 23 { 23d 3d} {17h 3h} 1364. > 13d Dh 23 { 4d 3d} { 4h 3h} 268. > 185d B9h 4 {224d} {E0h} 4190.5 > 186d BAh 4 { 81d} {51h} 1088. > 187d BBh 4 { 65d} {41h} 1127. > 188d BCh 4 { 59d} {3Bh} 735. > 189d BDh 4 { 39d} {27h} 703. > 190d BEh 23 {119d} {77h} 2540.5 > 191d BFh 4 { 86d} {56h} 1491.5 > 192d C0h 4 { 1d} { 1h} 34. > 222d DEh 23 {104d 2d} {68h 2h} 5188. > 223d DFh 23 { 50d 3d} {32h 3h} 2046. > 225d E1h 4 { 41d} {29h} 612. > 228d E4h 4 { 8d} { 8h} 156.5 > 230d E6h 4 { 13d} { Dh} 205. > 256d 100h 23 { 6d 2d} { 6h 2h} 261. > 257d 101h 4 { 24d} {18h} 427.5 > 788d 314h 4 { 1d} { 1h} 30. > > Note that the above is for only built-in ROM libraries, it doesn't > include any external libraries, not even the equation library or > periodic table library. > > 1-dimensional arrays and the array with the 2nd dimension being 1 > are English language only, arrays with the 2nd dimension of 2 are > English and French, and arrays with the 2nd dimension of 3 are > English, French, and Spanish. If the calculator's LANGUAGE is set > to French or Spanish and the particular library's message table > doesn't have that language, then the English message is used > instead. > > The library message tables are implemented as arrays of strings. > Of course an array starts with the prologue address for it's own > type, and then its total body length, but it also includes one > copy of the prologue address for the type of elements that it > contains, which are bodies of that object type, not complete > objects per se. It also contains a count of how many dimensions it > has, and the size of each dimension. > > A type 4 array is the normal (albeit not necessarily numeric) > array, with all elements present and in order. > > A type 23 array is a linked array, with a series of pointers to > the elements. The pointers must be present and in order, but the > elements can be in any order, more than one pointer can point to > the same element (object body), elements need not be present, and > a null (00000) pointer is used for any missing elements. > > I'll probably dig into the arrays a bit more; I'm curious about > those null messages. I believe that I see a way to convert a > linked array of strings to a list array of strings, and it > should be just a matter of actually writing the program. I'm > assuming that the SysRPL commands for working with arrays are > designed for normal arrays instead of linked arrays, but if I > start out by using ->H on a linked array, I expect that I can > deal with it. > > Note that a type 29 symbolic matrix is rather different from > these arrays; It's composite object, and very much like a list > array. > > What I was kind of hoping was for a list somewhere of the indirection > pointers. In other words, I was expecting that somewhere in ROM was a > packed list of messages, and elsewhere would be a table that says > Message 12345 is at offset abcd in the message table. So I could > just page through the table looking at messages which actually exist. > > What you could do is store a plain text file of the messages and > their addresses as a string on the calculator, perhaps on an SD > card. As long as you know at least part of the message that you > want, you could use the editor's FIND operation to search for it. > > > > -- > James === Subject: Re: Error messages are useful strings -- which? PS: I decided that it would be better to place the file on: http://www.i-is.com/users/jmprange/ The file name is: MSTBL.ZIP -- James === Subject: Re: Error messages are useful strings -- which? <204f1$45759d75$4267eab7$10037@123.NET> <62b4b$4575e26e$4267ea74$1722@123.NET> Now I'm wondering if there is a utility to do a translation from the display format used by the current Conn4x software. In other words, if I do an XModem transfer in ASCII mode from HP4X to PC, I get a file which is (a) readable without special fonts (and therefore emailable), and (b) can be reconstructed by simply transferring it via XModem to an HP-4X. The problem is that I may want to send some of my programs to my daughter, who uses only a Mac. I can email her the display format file and she can copy it onto her SD card, but to change that DadProg.txt file back to the original UserRPL (I don't aim higher than that at the moment) is not possible without said traslation program. I will check out HPConnect 0.9, which would be a good workaround if it works. --Irl > PS: > > I decided that it would be better to place the file on: > http://www.i-is.com/users/jmprange/ > > The file name is: MSTBL.ZIP > > -- > James === Subject: Re: Error messages are useful strings -- which? > James -- > SysRPL program (as a command-line entry, presumably) to actually be a > program was useful; I have tried displaying the SysRPL stack (-85 SF) > with a tiny UserRPL program on the stack and then typing in the same > thing but couldn't get it to actually be a program. > Here's a presumably dumb question: there are two extable files at > hpcalc.org. One is for the meta kernel and is much shorter than the > other one. What are these, please? You are using the 49g+/50G right? You should use extable2.lib which is large but has additional Entry Points Try www.hpcalc.org hp49 section Programming Entries Emacs (read the docs, too and see other useful stuff there) === Subject: Re: Error messages are useful strings -- which? <9u5dh.6$X24.1@reader1.news.saunalahti.fi> light reading. Looks like this is the same... Irl > James -- > SysRPL program (as a command-line entry, presumably) to actually be a > program was useful; I have tried displaying the SysRPL stack (-85 SF) > with a tiny UserRPL program on the stack and then typing in the same > thing but couldn't get it to actually be a program. > Here's a presumably dumb question: there are two extable files at > hpcalc.org. One is for the meta kernel and is much shorter than the > other one. What are these, please? > > You are using the 49g+/50G right? > You should use > extable2.lib > which is large but has additional Entry Points > Try www.hpcalc.org > hp49 section > Programming > Entries > Emacs > (read the docs, too and see other useful stuff there) === Subject: cmsg cancel Control: cancel Summary: EMP spam, BI=902 === === Subject: Data input: suggestions? I currently have a utility that will take point numbers, find them from a database, and calculate the area of the polygon. I'd like to expand it to allow the inclusion of curves in the figure as well. The problem I am running into is facilitating easy entry of curves with the build in command line parser. I really don't want to have to write my own for obvious reasons. These curves are defined by a beginning and end point, as well as a direction (curving right or left). I have a PDF made by a surveyor that is helping me try to figure out the best way to do this if anyone wants to take a look (www.pssllc.com/culdesac.pdf). It makes this discussion a little easier to visualize. In that paper, we came up with the idea of using - or + on the number to define L/R, and then if the value is less than 1, it is a radius point. So { 1 -.2 3 } would define a curve starting at point 1, curving to the left with radius at point 2, and ending at point 3. I thought this would work great until I remembered that if someone types in { 1 -.20 3 } to define a curve with radius point 20. . . So then I thought, I'll just switch to EXACT mode when they type in a list, but I'd really like to avoid that since I've been doing my best to hide exact integers from the people using my software. For a group of people who don't want to have to learn the difference between 1. and 1 , keeping it in approximate always is the best solution. So can anyone thing of an easier way to allow quick entry of a list like this without adding in a bunch of extra keystrokes to define the radius point? (for example, putting the radius point inside a list requires 2 keystrokes to put in the list, and an extra 2 R arrow presses to position the cursor after the list. I thought someone here might be able to see another way to get around this issue. Ideas? TW === Subject: Re: Data input: suggestions? Tim, The strategy that I've long used for this very situation is similar but slightly simpler. Use a negative real to signify radius point. In the example pdf this could be { 2. -3. 4 } where the -3. indicates radius point. There's no need to tell the routine + or - for left or right hand curve - the delta (central) angle should itself be + or - when the 2 azimuths are subtracted (3. to 2. and 3. to 4.). In the example, an odd case occurs at the cul-de-sac. If one inputs { 4. -5. 6. }, your routine will probably default to the angle < 180 whereas one might desire the angle > 180. The easy way to handle this is to input { 4. -5. -6. } where the -6. tells the routine to follow the greater arc instead of the lesser one. Another advantage to this scheme is the resulting algebraic sign of the segment area, + if clockwise, - if anticlockwise (again determined by the sign of the delta). The disadvanatge to all of this is that it requires pre-calculated radius points, to which some users may object. But a major advantage is never having to tell the routine whether the chord is inside or outside the curve as one has to do with some predominant (and expensive) cad/cogo packages. :-) Chuck >I currently have a utility that will take point numbers, find them from > a database, and calculate the area of the polygon. I'd like to expand > it to allow the inclusion of curves in the figure as well. The problem > I am running into is facilitating easy entry of curves with the build > in command line parser. I really don't want to have to write my own > for obvious reasons. > > These curves are defined by a beginning and end point, as well as a > direction (curving right or left). I have a PDF made by a surveyor > that is helping me try to figure out the best way to do this if anyone > wants to take a look (www.pssllc.com/culdesac.pdf). It makes this > discussion a little easier to visualize. > > In that paper, we came up with the idea of using - or + on the number > to define L/R, and then if the value is less than 1, it is a radius > point. So { 1 -.2 3 } would define a curve starting at point 1, > curving to the left with radius at point 2, and ending at point 3. I > thought this would work great until I remembered that if someone types > in { 1 -.20 3 } to define a curve with radius point 20. . . > > So then I thought, I'll just switch to EXACT mode when they type in a > list, but I'd really like to avoid that since I've been doing my best > to hide exact integers from the people using my software. For a group > of people who don't want to have to learn the difference between 1. and > 1 , keeping it in approximate always is the best solution. > > So can anyone thing of an easier way to allow quick entry of a list > like this without adding in a bunch of extra keystrokes to define the > radius point? (for example, putting the radius point inside a list > requires 2 keystrokes to put in the list, and an extra 2 R arrow > presses to position the cursor after the list. I thought someone here > might be able to see another way to get around this issue. Ideas? > > TW > === Subject: cmsg cancel Control: cancel Summary: EMP spam, BI=929 === === Subject: Re: An Ancient Computer Surprises Scientists You could also visit the following link: http://www.antikythera-mechanism.gr/index.php?lang=en Alex Markatis Civil Engineer Athens, Greece A/ John H Meyers Ucoaaa: > An Ancient Computer Surprises Scientists: > > http://www.smh.com.au/news/science/scientists-decode-ancient-astral-computer/ 2006/11/30/1164777721260.html > http://www.abc.net.au/news/newsitems/200611/s1801051.htm > http://www.nzherald.co.nz/section/story.cfm?c id=5&ObjectID=10413251 > http://www.nytimes.com/2006/11/29/science/30computecnd.html?ex=1322456400&en= 4dd0f1ffb387158c&ei=5088&partner=rssnyt&emc=rss > http://www.chicagotribune.com/technology/chi-0611300255nov30,1,6121069.story? coll=chi-technology-hed&ctrack=1&cset=true > http://news.com.com/2100-11397 3-6139636.html > > Next thing you know, an HP50G (with working 2100-year-old keyboard) > will be dug up -- well, maybe only a pre-CAS version, > although of course some CAS functionality may already be known > to date back that far :) > > There is nothing new under any Sun > > [r->] [OFF] === Subject: Re: HP50G and UNIX Christophe Prevotaux, le Sun 03 Dec 2006 00:05:07 +0000, a .8ecrit : > Has anyone succeeded in connection its HP50G to Linux or FreeBSD > thru the USB port ? FreeBSD being the OS that I use :). Again, with linux it's easy with a recent enough kernel: just compile the hp4x module (USB_SERIAL_HP4X). Samuel === Subject: Re: HP50G and UNIX Christophe, If you can figure out how to get your 50g USB link working under BSD please let the rest of us know. I am running OpenBSD here and am at a loss. Damn HP and their lack of serial cable! What kind of company releases a product that has a data port but NO corresponding cable? Cup On Dec 2, 4:05 pm, Christophe Prevotaux > Hi All, > > Has anyone succeeded in connection its HP50G to Linux or FreeBSD > thru the USB port ? FreeBSD being the OS that I use :). > > -------------------- > Christophe > -------------------- === Subject: Re: HP50G and UNIX > Has anyone succeeded in connection its HP50G to Linux or FreeBSD > thru the USB port ? FreeBSD being the OS that I use :). This post from a few days ago. . . TW === Subject: Re: HP50G and UNIX Le Sat, 02 Dec 2006 20:29:21 -0800, TW a .8ecrit: >> Has anyone succeeded in connection its HP50G to Linux or FreeBSD >> thru the USB port ? FreeBSD being the OS that I use :). > > This post from a few days ago. . . > > > TW === Subject: Directories on the SD card I have been creating directories as outlined on page 18-4 in the User's manual and storing data files within them. These files are available to the HP50 and my PC (using a card reader). The problem begins when I try to verify the directories' existence using 3:TEST VTYPE. The result should be 15 and is for the first directory created in this manner, but the return for the rest of the directories has been -1. Only the first created directory returns the expected variable type of 15. Is anyone else having this problem, or am I misunderstanding how this should work. === Subject: Re: Directories on the SD card Sorry I can't be more detailed, but my experiments with SD card directories indicates that things are very, very strange! I couldn't delete directories, run a VLIST command, etc... I wouldn't be surprised if the VTYPE command isn't supported in this setting. > I have been creating directories as outlined on page 18-4 in the > User's manual and storing data files within them. These files are > available to the HP50 and my PC (using a card reader). The problem > begins when I try to verify the directories' existence using 3:TEST > VTYPE. The result should be 15 and is for the first directory > created in this manner, but the return for the rest of the > directories has been -1. Only the first created directory returns > the expected variable type of 15. Is anyone else having this > problem, or am I misunderstanding how this should work. === Subject: Re: Directories on the SD card Think the SD card as an external near-line tape-drive It's limited in functionality but ublimited as a storage device You can even start programs from it Manjo explained a lot more.. > Sorry I can't be more detailed, but my experiments with SD card > directories indicates that things are very, very strange! I couldn't > delete directories, run a VLIST command, etc... > > I wouldn't be surprised if the VTYPE command isn't supported in this > setting. > > >> I have been creating directories as outlined on page 18-4 in the >> User's manual and storing data files within them. These files are >> available to the HP50 and my PC (using a card reader). The problem >> begins when I try to verify the directories' existence using 3:TEST >> VTYPE. The result should be 15 and is for the first directory >> created in this manner, but the return for the rest of the >> directories has been -1. Only the first created directory returns >> the expected variable type of 15. Is anyone else having this >> problem, or am I misunderstanding how this should work. > === Subject: Re: Directories on the SD card Hello SD's root you will be able to test the type of those files by using: :3:directoryname VTYPE if you created DOS directories or subdirectories and put your files or HP directory in there you must supply the correct path :3:{directory subdir filename} VTYPE directory and subdir should be valid names of directories where your file 'filename' is stored. If you try to check the type (VTYPE) of DOS directory you get an error. DOS directories are not HP objects. At this time you can't browse trough HP directory objects on SD, like you would with DOS directories. i hope this helps manjo === Subject: Re: Directories on the SD card oh i forgot... (i was testing some directory names which had the same name as commands -that's why i was getting the error) BUT if you try to check DOS directory using path syntax : :3:{ directory } VTYPE -you should get : -1 for sure Maybe your other directories are realy not the *same* after all. (15. is type for HP directory, DOS directory is not HP object that's why -1) manjo === Subject: Logistic Regression on a 49G+. I'm aware of another discussion involving logistic regression that points to RAS 3.0: http://www.hpcalc.org/details.php?id=1301 However, it is only for the 48 series. I tried using hpconv: http://www.hpcalc.org/details.php?id=3799 To convert the 48 binary to a 49 one, but alas, it gave me numerous Address for entry xyz (abc) not found. errors. Unfortunately, I don't know SysRPL, so I have no clue what I'm doing. Would anyone be kind enough to help me out? Any suggestions for converting this binary? Ideas of where to start? --Joey C. http://joeyjwc.x3fusion.com/ === Subject: Re: Logistic Regression on a 49G+. > I'm aware of another discussion involving logistic regression > that points to RAS 3.0: > http://www.hpcalc.org/details.php?id=1301 > However, it is only for the 48 series. I tried using hpconv: > http://www.hpcalc.org/details.php?id=3799 > To convert the 48 binary to a 49 one, but alas, it gave me numerous > Address for entry xyz (abc) not found. errors. Which is odd, since there's nothing unsupported in it. > Unfortunately, I don't know SysRPL, so I have no clue what I'm doing. > Any suggestions for converting this binary? It's all *UserRPL* (author MacDonald R. Phillips' favorite calculator language :), except for one item 'PAD1' which is all supported SysRPL. So what I'd do is: o On HP48GX: Store dis-assembled text for 'PAD1' into 'PAD1' o Recall entire directory and do STD 64 STWS ->STR o Copy entire string to HP49 and do STR-> (in Approx. mode) o Store the directory. o Compile 'PAD1' and store it back into 'PAD1' Here's the source for 'PAD1' (to save time with the first step above): :: CK2NOLASTWD CK&DISPATCH0 # 31 :: DO>STR DUPLEN$ TWENTYTWO 4PICK LEN$ #- 2DUP#< ITE :: SWAP#- Blank$ SWAP&$ ; :: 2DROP ; &$ ; ; Don't have an HP48GX? So use Emu48 or Debug4x -- it's cheaper than eBay :) [r->] [OFF] === Subject: Re: Logistic Regression on a 49G+. Here are some things to watch out for when attempting to simply transfer this UserRPL directory from a 48 series calc to a 49 series calc: > logistic regression, RAS 3.0: > http://www.hpcalc.org/details.php?id=1301 There is one very small SysRPL program 'PAD1' (as mentioned earlier), which needs to be converted to source and recompiled after transfer. There is a variable named 'ANS' (referenced in two other places); however, ANS is a *command* in 49 series, so you'll have to change that name in all three places. It also contains *unsupported* #353ABh [HP48] SYSEVAL in two places; thus far, #2F110h SYSEVAL (or #2B7001h FLASHEVAL) does the same in 49G[+]/50G/48Gii, but these still aren't supported (works with most expressions, not at all with programs). This offers a few examples of why even UserRPL transfer needs close attention. [r->] [OFF] === Subject: Re: Logistic Regression on a 49G+. Original file for HP48GX: > logistic regression, RAS 3.0: > http://www.hpcalc.org/details.php?id=1301 Here's a downloadable Zip file for 49G[+]/50G/48Gii with source and binary (will remain posted for 7 days): File Name: ras30B.zip Size: 43.29 KB http://www.filefactory.com/file/775cc5/ [r->] [OFF] === Subject: Re: Good text editors for 50G? Maybe I am missing something, but I do not understand the built-in Text Editor. When I go to APPS then Text Editor, all I see is the Left Arrow cursor. I need to press ALPHA just so I can start writing text. This TEXT EDITOR is just a line editor. Am I correct ? > > I installed TED in my hp50, but the editor seems to work improperly: > two letters appear when pressing the key, the cursor moves twice > up/down,.... > > is there a prblem with the code or I made something wrong? > > > > GaaK ha escrito: > > > Glad you like it. > > > > If you are going to do any serious programming, use emacs. Otherwise, > > the built in editor is great. What exactly are you wanting to do with > > a text editor on the calc? Write a novel? :-) > > > > TW > > TED vs Editor built-in > > - TED makes no duplicate of the edited string if it is in temporary > object area, thus enabling editing very large strings. Note that this > implies that no backup of the original string is kept! > > - TED is very fast and supports the tabulator. > > - Most of the normal character keys are in their normal places, others > can be fetched via the special character browser or ASCII code: CHR?+64 > = @ > > - TED decompiles object to width 30. This can be changed by storing a > real number larger than 9 in variable TED.WID. > > - TED contains the FIND/REPLACE option very fast! > > - TED works with ALL display > > - TED contains the Marks support and Macros > > - GaaK - === Subject: Re: Good text editors for 50G? > Maybe I am missing something, > but I do not understand the built-in Text Editor. > When I go to APPS then Text Editor, > all I see is the Left Arrow cursor. > I need to press ALPHA just so I can start writing text. > This TEXT EDITOR is just a line editor. Am I correct ? This APPS entry has always confused everyone; it's only the command line! -- identical to pressing right-shift ENTRY (above Alpha), and then you have to enter something valid in UserRPL (although any quoted big long string will do). On the other hand, if you type (an empty string), press ENTER and then press cursor-down, the internal pure text editor (string editor) will start (as it will for any existing string on the stack). In either case, MODE DISP offers Edit: options, such as small text (minifont), full page (cursor can move beyond current end of any line), and indent (repeats a previous line's indentation). [r->] [OFF] === Subject: Re: Good text editors for 50G? > >> Maybe I am missing something, >> but I do not understand the built-in Text Editor. >> When I go to APPS then Text Editor, >> all I see is the Left Arrow cursor. >> I need to press ALPHA just so I can start writing text. > >> This TEXT EDITOR is just a line editor. Am I correct ? > > This APPS entry has always confused everyone; > it's only the command line! -- identical to pressing > right-shift ENTRY (above Alpha), and then > you have to enter something valid in UserRPL > (although any quoted big long string will do). > > On the other hand, if you type (an empty string), > press ENTER and then press cursor-down, the internal > pure text editor (string editor) will start > (as it will for any existing string on the stack). > > In either case, MODE DISP offers Edit: options, > such as small text (minifont), full page > (cursor can move beyond current end of any line), > and indent (repeats a previous line's indentation). Now I'm confused?! The Build-in editor has [TOOL]s on the [NXT] pages that allows to saerch&Replace, change STYLE and FONT better than other calculators that I have used.. === Subject: Re: Good text editors for 50G? > Now I'm confused?! ?! sure does look confusing :) > The Built-in editor has [TOOL]s on the [NXT] pages > that allows to search&Replace, change STYLE and FONT > better than other calculators that I have used.. Yes, that's always available (even while editing the command line :) What's different about editing a command line vs. editing a string is that a string isn't compiled after you edit it. When you invoke APPS > Text editor or Right-shift ENTRY, whatever you compose will get *compiled* and *executed* (just type two consecutive decimal points for a quick illustration), wheras when you do string EDITB the string is edited (without enclosing in quotes), and not re-compiled afterwards. The closest HP48 simulation of pure string editing would be: string INPUT (which you can also do on the 49G[+]/50G, if EDITB is too simple to use :) [r->] [OFF] === Subject: Re: Good text editors for 50G? Now I understand. I missed the [TOOLS] [NEXT] group. > > Now I'm confused?! > > ?! sure does look confusing :) > > The Built-in editor has [TOOL]s on the [NXT] pages > that allows to search&Replace, change STYLE and FONT > better than other calculators that I have used.. > > Yes, that's always available > (even while editing the command line :) > > What's different about editing a command line vs. editing a string > is that a string isn't compiled after you edit it. > > When you invoke APPS > Text editor or Right-shift ENTRY, > whatever you compose will get *compiled* and *executed* > (just type two consecutive decimal points for a quick illustration), > wheras when you do string EDITB the string is edited > (without enclosing in quotes), and not re-compiled afterwards. > > The closest HP48 simulation of pure string editing > would be: string INPUT > (which you can also do on the 49G[+]/50G, > if EDITB is too simple to use :) > > [r->] [OFF] === Subject: nonpareil HP-41 emulator under Mac-OSX? The website http://nonpareil.brouhaha.com/ indicates that the HP-41C and friends are simulated by nonpareil using the GTK+ toolkit. Has anyone already built nonpareil, from source with GTK+ for Mac-OSX, -- Chris. === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? Chris McDonald a ?crit : > The website http://nonpareil.brouhaha.com/ indicates that the HP-41C > and friends are simulated by nonpareil using the GTK+ toolkit. > > Has anyone already built nonpareil, from source with GTK+ for Mac-OSX, > > -- > Chris. You can find an HP 41 simulator for Macintosh at the fllowing URL: http://homepage.mac.com/mba/nsim/ Laurent === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? >Chris McDonald a =E9crit : >> The website http://nonpareil.brouhaha.com/ indicates that the HP-41C >> and friends are simulated by nonpareil using the GTK+ toolkit. >> >> Has anyone already built nonpareil, from source with GTK+ for Mac-OSX, >> >You can find an HP 41 simulator for Macintosh at the fllowing URL: >http://homepage.mac.com/mba/nsim/ but the webpages keeping saying older version of nsim... and I had wondered what I was missing out on. (?) -- Chris. === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? Chris, http://homepage.mac.com/mba/nonpareil/ (All credit goes to Eric Smith and Maciej Bartosiak (osx port). I'm just providing the link..) Pal > >Chris McDonald a =E9crit : > >> The website http://nonpareil.brouhaha.com/ indicates that the HP-41C >> and friends are simulated by nonpareil using the GTK+ toolkit. >> >> Has anyone already built nonpareil, from source with GTK+ for Mac-OSX, >> > >You can find an HP 41 simulator for Macintosh at the fllowing URL: > >http://homepage.mac.com/mba/nsim/ > > but the webpages keeping saying older version of nsim... > and I had wondered what I was missing out on. (?) > > -- > Chris. === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? >http://homepage.mac.com/mba/nonpareil/ >(All credit goes to Eric Smith and Maciej Bartosiak (osx port). I'm >just providing the link..) (see my original email). -- Chris. === Subject: Re: nonpareil HP-41 emulator under Mac-OSX? > (see my original email). > Chris. Darn it. I hate it when I jump my guns. I looked too quickly. Sorry about that. Pal === Subject: connectivity kit I have downloaded at least two versions of the connectivity kit the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 I keep getting trapped in a circle of the software telling me to install Xmodem driver, calculator not found, timeout, etc. This latest version does not have the PC in the right pane as the older version did If I manage to drag or copy and paste a file over from the calc; it is usually unreadable by a PC with the proper software to open and use the file (specifically, a .cr5 file going to a PC with TDS software) Tried calling an 800 number for HP but the foreign person there informed that my Compaq computer is not in warranty*&^%$#@ Anyone know where to get some usable, printed help documentation? === Subject: Re: connectivity kit Note that for 48GX you need to download and install the XMODEM server library, and >>it must be in port 0 or 1<<. I tried it in other ports Irl > I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me to > install Xmodem driver, > calculator not found, timeout, etc. > This latest version does not have the PC in the right pane as the older > version did > If I manage to drag or copy and paste a file over from the calc; it is > usually unreadable by > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > Tried calling an 800 number for HP but the foreign person there > informed that my Compaq computer > is not in warranty*&^%$#@ > Anyone know where to get some usable, printed help documentation? === Subject: Re: connectivity kit now that we are talking about connectivity kit, is there anyway to use the serial conector with an USB adaptor? my laptop doesnt have serial plug. > Note that for 48GX you need to download and install the XMODEM server > library, and >>it must be in port 0 or 1<<. I tried it in other ports > Irl > I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me to > install Xmodem driver, > calculator not found, timeout, etc. > This latest version does not have the PC in the right pane as the older > version did > If I manage to drag or copy and paste a file over from the calc; it is > usually unreadable by > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > Tried calling an 800 number for HP but the foreign person there > informed that my Compaq computer > is not in warranty*&^%$#@ > Anyone know where to get some usable, printed help documentation? === Subject: Re: connectivity kit > now that we are talking about connectivity kit, > is there any way to use the serial conector > with a USB adaptor? > my laptop doesnt have serial plug. For the 49G+/50G, a serial cable (if it existed), pluged into a USB adaptor, would be very similar to the existing direct USB cable, except possibly that it might then appear (using the adapter's driver) as a COM port, rather than as a USB port, for use by some other software. I just bought some very small USB-to-serial adapters, and lo, they install fine but don't work with anything (neither on COM4, which some software recognizes, nor on COM5, which my applications have never heard of :) I think I'd better never upgrade to a computer without on-board serial, or else I'll have to throw out all my favorite organizers and calcs :) [r->] [OFF] === Subject: Re: connectivity kit so, would it work using a serial to USB adaptor with my 48gx? > > now that we are talking about connectivity kit, > is there any way to use the serial conector > with a USB adaptor? > my laptop doesnt have serial plug. > > For the 49G+/50G, a serial cable (if it existed), > pluged into a USB adaptor, would be very similar > to the existing direct USB cable, > except possibly that it might then appear > (using the adapter's driver) as a COM port, > rather than as a USB port, > for use by some other software. > > I just bought some very small USB-to-serial adapters, > and lo, they install fine but don't work with anything > (neither on COM4, which some software recognizes, > nor on COM5, which my applications have never heard of :) > > I think I'd better never upgrade to a computer > without on-board serial, or else I'll have to throw out > all my favorite organizers and calcs :) > > [r->] [OFF] === Subject: Re: connectivity kit > would it work using a serial to USB adaptor with my 48gx? Should work in theory, provided that USB adapter driver creates a serial COM port that your computer-side software can use, and that its hardware can talk to 48GX's UART (serial circuit), which won't be the calc's fault if it can't, because 48GX is engineered to be truly compatible with original serial hardware specs (using either no flow control or Xon/Xoff flow control, not hardware flow control). I hope that someone who has actually done it (on the same computer operating system that you use, and ideally with an adapter just like yours) will add something more encouraging, and perhaps useful. [r->] [OFF] === Subject: Re: connectivity kit adaptor. If it doesnt work, then it'd useless. And there's actually a custom cable sold on ebay for $40 which is kinda expensive > > would it work using a serial to USB adaptor with my 48gx? > > Should work in theory, provided that USB adapter driver > creates a serial COM port that your computer-side software can use, > and that its hardware can talk to 48GX's UART (serial circuit), > which won't be the calc's fault if it can't, > because 48GX is engineered to be truly compatible > with original serial hardware specs > (using either no flow control or Xon/Xoff flow control, > not hardware flow control). > > I hope that someone who has actually done it > (on the same computer operating system that you use, > and ideally with an adapter just like yours) > will add something more encouraging, and perhaps useful. > > [r->] [OFF] === Subject: Re: connectivity kit > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) This means you have a 48GX, correct? TW === Subject: Re: connectivity kit > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > > This means you have a 48GX, correct? > > TW Sorry about lack of specific info; yeah, HP 48GX with a TDS Cogo GX/SX card in it. === Subject: Re: connectivity kit >I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me to > install Xmodem driver, > calculator not found, timeout, etc. > This latest version does not have the PC in the right pane as the older > version did > If I manage to drag or copy and paste a file over from the calc; it is > usually unreadable by > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > Tried calling an 800 number for HP but the foreign person there > informed that my Compaq computer > is not in warranty*&^%$#@ > Anyone know where to get some usable, printed help documentation? Search this newsgroup the USB driver I downloaded here: http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareIndex.jsp?lang=en&c c=us&prodNameId=3235174&prodTypeId=215348&prodSeriesId=3235173&swLang=8&taskI d=135&swEnvOID=228 The HPConn4x version that I use is here: http://www.hpcalc.org/details.php?id=5411 in the HP web pages Training Modules http://h20331.www2.hp.com/hpsub/cache/383688-0-0-225-121.html I did not find Connecting to PC Funny enough the only official English web page (that I know of) giving some advice - is in Finland: http://www.dlc.fi/~radix/tuki-support.htm click on that 1st link on the center Lis.8atietoja englanniksi (starts Javascript pop-up window) VP-Finland === Subject: Re: connectivity kit myankee ha escrito: > I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me to > install Xmodem driver, > calculator not found, timeout, etc. > This latest version does not have the PC in the right pane as the older > version did > If I manage to drag or copy and paste a file over from the calc; it is > usually unreadable by > a PC with the proper software to open and use the file (specifically, a > .cr5 file going to a PC with TDS > software) > Tried calling an 800 number for HP but the foreign person there > informed that my Compaq computer > is not in warranty*&^%$#@ > Anyone know where to get some usable, printed help documentation? I have problems with the latest version of HP connectivity kit, so I download and install the old 2.2 Build 2353 version from www.hpcalc.org, that work ok with my HP 49G and also with the HP 50G. I hope this help. === Subject: Re: connectivity kit > Anyone know where to get some usable, printed help documentation? > C:Program FilesHewlett-PackardConn4xConn4x.chm -- - - - - - - - - - - - - - - - - Bill Graves RKBA! bgraves@ix.netcom.com === Subject: Re: connectivity kit > I have downloaded at least two versions of the connectivity kit > the latest is Ver 2.3 Build 2439 for the 48, 49 & 50 > I keep getting trapped in a circle of the software telling me > to install Xmodem driver, > calculator not found, timeout, etc. Does Conn4x see HPx9G+ [driver] in Connect using? On calc, what shows in APPS > I/O functions > Transfer next to Port: ? [r->] [OFF] === Subject: TED on HP50G Using TED on my new HP50G, cursor is moving two places on every key stroke. What I am doing wrong? Sorry for bed English. Damir === Subject: TED on HP50G Using TED on my new HP50G, cursor is moving two places on every key stroke. What I am doing wrong? Sorry for bad English. Damir === Subject: Oops -- please change title ... I just posted a reply to James Prange on the thread Error Messages are Useful Strings but it really should be a new thread -- it's about text translation and Mac Connectivity. I'd call it Translation from Conn4x ASCII display format How does one change the title of a thread, or move a post to a new thread? Irl === Subject: Re: Oops -- please change title ... > I just posted a reply to James Prange on the thread Error Messages are > Useful Strings but it really should be a new thread -- it's about text > translation and Mac Connectivity. I'd call it Translation from Conn4x > ASCII display format > How does one change the title of a thread, or move a post to a new > thread? Well, this is usenet, and I don't think that you can do anything to change it once it's posted. You could try to cancel a post, but it usually gets to several servers pretty darn quickly, and most of them don't honour cancel commands. I see that you're posting via Google, and I understand that Google does honour some sort of don't archive header, so such posts won't be stored there forever (although they may be on someone's PC). Other usenet servers don't archive posts, so eventually old posts are discarded from them to make room for new posts. The correct way to change a subject within a thread on usenet: New subject (was: Old subject) The better newsreaders will automatically discard the (was: Old subject) part and use just the New subject part instead of using New subject (was: Old subject) when they insert the subject header in a reply, but apparently a lot of newreaders don't. If your newsreader doesn't automatically discard the (was: Old subject) part when composing a reply, then you should edit it out yourself. programs to convert between the compiled object form (like what you'd see with a binary transfer), and the decompiled to a string and optionally translated form (like what you'd see with a Kermit ASCII or Conn4x Text transfer). They were intended to make it easier use the SD card for ASCII transfers, but they could be used for any binary transfer, or any other situation where you want to convert between these forms. You can find my programs at: http://www.hpcalc.org/search.php?query=%22ASCII+on+SD%22 Another (perhaps better) application that does something similar can be found at: http://www.hpcalc.org/search.php?query=ASCII+to+Binary+Converter+ Perhaps there are other at hpcalc.org. Regarding transfers with a Mac, try: http://hpconnect.sourceforge.net/ -- James === Subject: Translation from Conn4x,ASCII display format <9d00c$457604f9$4267eb85$28654@123.NET> James -- I will look at those links. Sounds like the solution to my problem. ... I'm feeling like I ought to contribute something to this group besides newbie questions, but I'm strictly a UserRPL optics/physics/math applications programmer. Irl > I just posted a reply to James Prange on the thread Error Messages are > Useful Strings but it really should be a new thread -- it's about text > translation and Mac Connectivity. I'd call it Translation from Conn4x > ASCII display format > How does one change the title of a thread, or move a post to a new > thread? > > Well, this is usenet, and I don't think that you can do anything > to change it once it's posted. You could try to cancel a post, but > it usually gets to several servers pretty darn quickly, and most > of them don't honour cancel commands. > > I see that you're posting via Google, and I understand that Google > does honour some sort of don't archive header, so such posts > won't be stored there forever (although they may be on someone's > PC). Other usenet servers don't archive posts, so eventually old > posts are discarded from them to make room for new posts. > > The correct way to change a subject within a thread on usenet: > > New subject (was: Old subject) > > The better newsreaders will automatically discard the > (was: Old subject) part and use just the New subject part > instead of using New subject (was: Old subject) when they insert > the subject header in a reply, but apparently a lot of newreaders > don't. If your newsreader doesn't automatically discard the > (was: Old subject) part when composing a reply, then you should > edit it out yourself. > > programs to convert between the compiled object form (like what > you'd see with a binary transfer), and the decompiled to a string > and optionally translated form (like what you'd see with a Kermit > ASCII or Conn4x Text transfer). They were intended to make it > easier use the SD card for ASCII transfers, but they could be used > for any binary transfer, or any other situation where you want to > convert between these forms. You can find my programs at: > http://www.hpcalc.org/search.php?query=%22ASCII+on+SD%22 > > Another (perhaps better) application that does something similar > can be found at: > http://www.hpcalc.org/search.php?query=ASCII+to+Binary+Converter+ > > Perhaps there are other at hpcalc.org. > > Regarding transfers with a Mac, try: > http://hpconnect.sourceforge.net/ > > -- > James === Subject: Translation from Conn4x,ASCII display format (was: Oops -- please change title ...) PS: I should've changed the subject myself, which I'll do with this post. -- James === Subject: Question about USB Cable of HP 50G Is the USB Cable that come with the HP 50G a normal USB cable? I can buy another cable in any store, or the USB cable have some special characteristics ? === Subject: Re: Question about USB Cable of HP 50G > Is the USB Cable that come with the HP 50G a normal USB cable? > I can buy another cable in any store, or the USB cable have some > special characteristics ? Yes it is a 'normal' USB to miniUSB cable, manjo === Subject: Re: Question about USB Cable of HP 50G manjo ha escrito: > Is the USB Cable that come with the HP 50G a normal USB cable? > I can buy another cable in any store, or the USB cable have some > special characteristics ? > > Yes it is a 'normal' USB to miniUSB cable, > > manjo What about the serial cable. Is there any place where I can get one ? Why is it so particular ? Daniel === Subject: Re: Question about USB Cable of HP 50G > What about the serial cable. Is there any place where I can get one ? > Why is it so particular ? Serial cable had to be different from USB (Although it was written in the manual on 48GII, there were 'problems' since users tried to connect it to USB, because connector was the same type) Further : Since CPU is delivering 3.3V logic (not RS232 levels) designers figured power should be available on the connector for 'level shifter' which would be bilt in the cable connector. I guess they choose that particular type because they wanted to use different type from the older models due to signal levels incompatibility, old cables would not work. -from what i said before -users don't pay much attention to users manual -this way users wouldn't get the idea to connect the old cable expecting it to work so much for now, manjo === Subject: Debug4x b104 UserRPL Additions Debug4x build 104 is now available from http://www.Debug4x.com The User RPL editor now has the Verb Completion ctrl-space listing of verbs, the stack pictures ctrl-j and the code completion (macros) alt-ins just like the SysRPL editors. Some other improvements for the User RPL system and some bug fixes. Have a nice Christmas everyone, I am vacationing! -- - - - - - - - - - - - - - - - - Bill Graves RKBA! bgraves@ix.netcom.com === Subject: Re: Debug4x b104 UserRPL Additions > Debug4x build 104 is now available from > http://www.Debug4x.com > > The User RPL editor now has the Verb Completion ctrl-space listing of > verbs, > the stack pictures ctrl-j > and the code completion (macros) alt-ins > just like the SysRPL editors. > > Some other improvements for the User RPL system and some bug fixes. > > Have a nice Christmas everyone, I am vacationing! > -- - - - - - - - - - - - - - - - - > Bill Graves RKBA! > bgraves@ix.netcom.com Downloading... === Subject: Re: Debug4x b104 UserRPL Additions > >> Debug4x build 104 is now available from >> http://www.Debug4x.com >> >> The User RPL editor now has the Verb Completion ctrl-space listing of >> verbs, >> the stack pictures ctrl-j >> and the code completion (macros) alt-ins >> just like the SysRPL editors. >> >> Some other improvements for the User RPL system and some bug fixes. >> >> Have a nice Christmas everyone, I am vacationing! >> -- - - - - - - - - - - - - - - - - >> Bill Graves RKBA! >> bgraves@ix.netcom.com > Downloading... Downloaded! === Subject: Re: Debug4x b104 UserRPL Additions > >> > Debug4x build 104 is now available from > http://www.Debug4x.com > > The User RPL editor now has the Verb Completion ctrl-space listing of > verbs, > the stack pictures ctrl-j > and the code completion (macros) alt-ins > just like the SysRPL editors. > > Some other improvements for the User RPL system and some bug fixes. > > Have a nice Christmas everyone, I am vacationing! > -- - - - - - - - - - - - - - - - - > Bill Graves RKBA! > bgraves@ix.netcom.com >> Downloading... > Downloaded! Unzipping... === Subject: Re: Debug4x b104 UserRPL Additions > >> > >> Debug4x build 104 is now available from >> http://www.Debug4x.com >> >> The User RPL editor now has the Verb Completion ctrl-space listing of >> verbs, >> the stack pictures ctrl-j >> and the code completion (macros) alt-ins >> just like the SysRPL editors. >> >> Some other improvements for the User RPL system and some bug fixes. >> >> Have a nice Christmas everyone, I am vacationing! >> -- - - - - - - - - - - - - - - - - >> Bill Graves RKBA! >> bgraves@ix.netcom.com > Downloading... >> Downloaded! > Unzipping... Unzipped! === Subject: Re: Debug4x b104 UserRPL Additions > >> > >> >>> Debug4x build 104 is now available from >>> http://www.Debug4x.com >>> >>> The User RPL editor now has the Verb Completion ctrl-space listing >>> of verbs, >>> the stack pictures ctrl-j >>> and the code completion (macros) alt-ins >>> just like the SysRPL editors. >>> >>> Some other improvements for the User RPL system and some bug fixes. >>> >>> Have a nice Christmas everyone, I am vacationing! >>> -- - - - - - - - - - - - - - - - - >>> Bill Graves RKBA! >>> bgraves@ix.netcom.com >> Downloading... > Downloaded! >> Unzipping... > Unzipped! Starting Installation... === Subject: Re: Debug4x b104 UserRPL Additions > >> > >> >>> >> Debug4x build 104 is now available from >> http://www.Debug4x.com >> >> The User RPL editor now has the Verb Completion ctrl-space listing >> of verbs, >> the stack pictures ctrl-j >> and the code completion (macros) alt-ins >> just like the SysRPL editors. >> >> Some other improvements for the User RPL system and some bug fixes. >> >> Have a nice Christmas everyone, I am vacationing! >> -- - - - - - - - - - - - - - - - - >> Bill Graves RKBA! >> bgraves@ix.netcom.com >>> Downloading... >> Downloaded! > Unzipping... >> Unzipped! > Starting Installation... Installed...testing..calc emulator works...replacing the old... continue testing.... === Subject: Re: Debug4x b104 UserRPL Additions >>> Debug4x build 104 is now available from >>> http://www.Debug4x.com >>> >>> The User RPL editor now has the Verb Completion ctrl-space listing >>> of verbs, >>> the stack pictures ctrl-j >>> and the code completion (macros) alt-ins >>> just like the SysRPL editors. >>> >>> Some other improvements for the User RPL system and some bug fixes. >>> >>> Have a nice Christmas everyone, I am vacationing! >>> -- - - - - - - - - - - - - - - - - >>> Bill Graves RKBA! >>> bgraves@ix.netcom.com >> Downloading... >>> Downloaded! >> Unzipping... > Unzipped! >> Starting Installation... > Installed...testing..calc emulator works...replacing the old... > continue testing.... Darm! Since my display is using 120dpi instead of 96dpi some windows overlap slighly also some text is not showing properly otherwise...ok (so far) VPN === Subject: Re: Debug4x b104 UserRPL Additions > Since my display is using 120dpi instead of 96dpi some windows overlap > slighly also some text is not showing properly otherwise...ok (so far) > VPN VPN Send me details if the Windows cannot be moved or sized to work properly. I thought I had fixed that for you some time back! (A picture or two would help as well). -- - - - - - - - - - - - - - - - - Bill Graves RKBA! bgraves@ix.netcom.coms === Subject: Re: Debug4x b104 UserRPL Additions > >> > >> >>> >> Debug4x build 104 is now available from >> http://www.Debug4x.com >> >> The User RPL editor now has the Verb Completion ctrl-space listing >> of verbs, >> the stack pictures ctrl-j >> and the code completion (macros) alt-ins >> just like the SysRPL editors. >> >> Some other improvements for the User RPL system and some bug fixes. >> >> Have a nice Christmas everyone, I am vacationing! >> -- - - - - - - - - - - - - - - - - >> Bill Graves RKBA! >> bgraves@ix.netcom.com >>> Downloading... >> Downloaded! > Unzipping... >> Unzipped! > Starting Installation... Installed...testing... === Subject: Re: Debug4x b104 UserRPL Additions <2jDdh.1125$O8.54@reader1.news.saunalahti.fi> >>> Downloading... >> Downloaded! > Unzipping... >> Unzipped! > Starting Installation... > Installed...testing... ...? - GaaK - === Subject: SD Capacity wrong in Files I have a HP49G+ with version 4.20060602 (VER) and 2.09 (VERSION). I have installed a 1 Gbyte SD card. The card has been formatted both on a PC and the HP, however when I look in files it shows the capacity as being 55842kB or around 55Mbytes. Does anybody have an explanation. === Subject: Re: SD Capacity wrong in Files if you mean the free capacity on the SD card, maybe you are experiencing the slack space effect. For more info on this: http://www.webopedia.com/TERM/S/slack_space.html HTH Raymond schrieb im Newsbeitrag >I have a HP49G+ with version 4.20060602 (VER) and 2.09 (VERSION). I > have installed a 1 Gbyte SD card. The card has been formatted both on a > PC and the HP, however when I look in files it shows the capacity as > being 55842kB or around 55Mbytes. Does anybody have an explanation. > > === Subject: Re: SD Capacity wrong in Files >I have a HP49G+ with version 4.20060602 (VER) and 2.09 (VERSION). I > have installed a 1 Gbyte SD card. The card has been formatted both on a > PC and the HP, however when I look in files it shows the capacity as > being 55842kB or around 55Mbytes. Does anybody have an explanation. your card most probably has 955842kB or around 955 MBytes (maybe even 1055842kB) but... leftmost digits are probably not visible due to limited display space available. Don't worry, your SD card is in there, your space will be used, *and* it will be displayed correctly once the size reduces do displayable figure. either size should be displayed as MB when more than 90000 kB is available *or* mem sizes should be moved at least one space to the right so larger sizes could be displayed. manjo