newRPL - build 1255 released! [updated to 1299]
|
02-18-2018, 08:53 PM
Post: #141
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1016]
(02-18-2018 08:13 PM)Claudio L. Wrote:(02-18-2018 11:30 AM)martinot Wrote: How well does newRPL work on HP 39gs? As well as 50g (besides ROM size)? Thanks! If you send me a message with your adress I will be happy to sponsor you with a 39gs if that would help you with the great work you do with newRPL. HP 35s, HP 50g, HP Elite X3, SwissMicros DM42 |
|||
02-19-2018, 12:02 PM
Post: #142
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1016] | |||
02-28-2018, 11:45 PM
Post: #143
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
All ROMS were updated to build 1046. See first post for changelog.
|
|||
03-01-2018, 01:59 PM
Post: #144
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(02-28-2018 11:45 PM)Claudio L. Wrote: All ROMS were updated to build 1046. See first post for changelog. The simulator/emulator touch is nice. I think newRPL has quite some potential also on the desktop. Wikis are great, Contribute :) |
|||
03-01-2018, 10:27 PM
Post: #145
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(02-18-2018 08:53 PM)martinot Wrote: If you send me a message with your adress I will be happy to sponsor you with a 39gs if that would help you with the great work you do with newRPL. What I meant is because I don't have a 39, I don't use a 39 and don't know where a 39 user would expect to find things. Having a 39gs now won't really help me because I'm not used to it, I'd still be unable to get a good sense of where to move things in the keyboard. But now that the newRPL/39 community is growing, perhaps we can all coordinate efforts to decide how to reorganize the keyboard. I need ideas more than hardware, but I do appreciate your generous gesture. |
|||
03-02-2018, 07:02 AM
(This post was last modified: 03-02-2018 07:22 AM by The Shadow.)
Post: #146
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(02-28-2018 11:45 PM)Claudio L. Wrote: All ROMS were updated to build 1046. See first post for changelog. I'm not seeing the System/Lib submenu the wiki claims is there? EDIT: Is there any easier way to type a '$' than to use 36 ->UTFB? EDIT: Also, a library is apparently expecting a $VISIBLE list, and perhaps others, but they aren't documented in the wiki. |
|||
03-02-2018, 02:36 PM
Post: #147
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(03-02-2018 07:02 AM)The Shadow Wrote: I'm not seeing the System/Lib submenu the wiki claims is there?Ouch, me neither. I forgot to enable the menu!. Give me a couple of days and I'll release 1047. (03-02-2018 07:02 AM)The Shadow Wrote: EDIT: Is there any easier way to type a '$' than to use 36 ->UTFB?Yes, $ is where it's always been in the 50g: LS-4 in alpha mode. (03-02-2018 07:02 AM)The Shadow Wrote: EDIT: Also, a library is apparently expecting a $VISIBLE list, and perhaps others, but they aren't documented in the wiki.I've already started writing the wiki since creating libraries needs a lot of documentation, probably finish it over the weekend. But since this rom fixes a bug in SDRESTORE that could potentially cause a crash, I wanted it out before anyone found out the hard way. The worst thing in my book is a project perceived (or "branded" by its users) as unstable or unreliable. |
|||
03-02-2018, 03:20 PM
(This post was last modified: 03-02-2018 03:48 PM by Harald.)
Post: #148
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
I am still trying to get my head round how newRPL works, so please be patient with me
I understand that a trailing decimal point indicates that a number is aporoximate rather than exact. So if I devide 5 by 3, I get a trailing decimal point. I can also enter the trailing decimal point thus telling the calculater the number is aproximate by entering 5. rather then 5. If I devide 5. by 10, I get, as expected 0.5. . Now if I try the same with a complex number (0,5.) and dived that by 10, I get (0,0.5). Is that a bug? Or am I missing something? Next question: how do I enter single " ' "? Edit: I take that question back, that was stupid, it is in the same location it is on the 50g, that just wasn't obvious on my 39gs... And how do I enter commands? I have created a variable that I wanted to get rid of. On the 50g I would have entered: 'variable' purge enter How do I do this with newRPL? Purge seems to be treated as text... Edit: found that one too, I entered it lower case rather than upper case because alpha mode on the 50g is different from newRPL. Is there a manual, or is everything in the wiki? And last but not least, I wanted to download the emulator from sourceforge, but sourceforge appears to be down. Is that just me? Can anyone access sourceforge? Cheers, Harald Edit: and a new question, is "pi" (the symbol, not the two letters) not predifined to be 3.1415...? |
|||
03-02-2018, 04:29 PM
(This post was last modified: 03-02-2018 04:33 PM by The Shadow.)
Post: #149
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
I heartily approve of the $MENU option! That really makes things simple. What will happen if the menu accidentally doesn't include everything listed in $VISIBLE, though? (My guess is nothing, but it doesn't hurt to ask.)
Can the directory for use with CRLIB include subdirectories? Or is that more a case for using $MENU to call another library? EDIT: Purely out of curiosity, does NARGS still do anything if ALLOWINSYMB is 0? |
|||
03-02-2018, 07:33 PM
Post: #150
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(03-02-2018 03:20 PM)Harald Wrote: I am still trying to get my head round how newRPL works, so please be patient with meI'll have to investigate. On complex you should expect to get some unexpected approximated numbers (because of internal operations) but an approximated number becoming exact has to be a bug. (03-02-2018 03:20 PM)Harald Wrote: Next question: how do I enter single " ' "? Edit: I take that question back, that was stupid, it is in the same location it is on the 50g, that just wasn't obvious on my 39gs...The manual IS the wiki. A work in progress, but it's slowly growing. Since most commands are compatible with the 50g, the AUR helps a lot. Look at the detailed project status for a list of commands implemented, if marked green they are fully compatible with the 50g command so use the AUR, if not then it's the wiki or ask me in this forum. (03-02-2018 03:20 PM)Harald Wrote: And last but not least, I wanted to download the emulator from sourceforge, but sourceforge appears to be down. Is that just me? Can anyone access sourceforge? Sourceforge is having huge problems in the last few days with a DDOS attack apparently. It's down, way down. (03-02-2018 03:20 PM)Harald Wrote: Edit: and a new question, is "pi" (the symbol, not the two letters) not predifined to be 3.1415...? Constants are not implemented yet, you can use pi0 as a workaround for the time being. |
|||
03-02-2018, 07:38 PM
(This post was last modified: 03-02-2018 07:42 PM by Claudio L..)
Post: #151
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(03-02-2018 04:29 PM)The Shadow Wrote: I heartily approve of the $MENU option! That really makes things simple. What will happen if the menu accidentally doesn't include everything listed in $VISIBLE, though? (My guess is nothing, but it doesn't hurt to ask.)Nothing happens, the user won't see that command in the menu, but the command exists, can be typed on the keyboard and should even show up for autocomplete (when I finish implementing autocomplete, that is, as of today is not working). (03-02-2018 04:29 PM)The Shadow Wrote: Can the directory for use with CRLIB include subdirectories? Or is that more a case for using $MENU to call another library?It can, but CRLIB won't try to find anything in them, they will be ignored. (03-02-2018 04:29 PM)The Shadow Wrote: EDIT: Purely out of curiosity, does NARGS still do anything if ALLOWINSYMB is 0? No use at all, they are hints for the symbolic compiler, nothing else. EDIT: But having said that, please make sure you put the right number of arguments, even if ALLOWINSYMB is zero. For future expansion (think of a different compiler, for example, using the same commands to compile PPL instead of RPL). You may not allow it in a symbolic expression, but if the expression is evaluated at compile time that number may come in handy. |
|||
03-02-2018, 08:27 PM
Post: #152
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(03-01-2018 10:27 PM)Claudio L. Wrote:(02-18-2018 08:53 PM)martinot Wrote: If you send me a message with your adress I will be happy to sponsor you with a 39gs if that would help you with the great work you do with newRPL. Well, I am not a 39gs use either. In fact I never would have bought one if they hadn't become so cheap. €15 from China. And I thought newRPL would be a good way to make an otherwise useless machine (at least to me, as it is algebraic only) very usable. I'd much prefer an overlay than a dedicated 39gs layout. Keeping it close to the 50g is a good thing in my opinion Cheers, Harald |
|||
03-03-2018, 05:10 AM
(This post was last modified: 03-03-2018 05:41 AM by The Shadow.)
Post: #153
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
Does LIBCLEAR erase the entire library directory? Or do you specify a list of targets like PURGE?
I gather that it doesn't matter what directory you're in when you ATTACH a library, unlike oldRPL? I never did see the point in ATTACHing anywhere but HOME. LIBSTO and LIBRCL look fun. EDIT: Just made my first library, and some of the names in the menu are filled with extraneous characters. They still work as programs, but don't display help. Oh, and the RS-hold-menu key thing to get the library ID doesn't seem to be working. |
|||
03-03-2018, 02:16 PM
Post: #154
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
hello,
I am having trouble to upgrade FW on the calculator. Errror: USB connection cannot be established. Windows recognizes the calculator (notification sound after connecting the calculator), but the HP Conn3x doesn't see the calculator. Device: hp39gs and hp40gs, firmware version 1016. Behavior: calculator is not recognized by conn3x application (windows 7 and windows 10). While I put calculator to flash mode (reset & + -, then update), "R" button shortly appears and calculator gets connected. In 2-3 seconds it disconnects. Note: I did not have issue to flash 1001 and 1016 firmware on the same PC with the same OS. Anyone else having this issue? Can you advice how can I flash new version or revert to stock firmware? Thanks for help. Jan |
|||
03-03-2018, 02:19 PM
(This post was last modified: 03-05-2018 11:21 AM by compsystems.)
Post: #155
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
On the demo, I recommend calling it "newRPL SDK" = new hp50 connectivity kit + simulator
PD: NewRPL spanish forum: https://www.honradoshp.com/f49-newrpl . |
|||
03-04-2018, 04:22 AM
Post: #156
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(03-03-2018 02:16 PM)osg Wrote: hello, This issue is not related to newRPL, since you are trying to connect the original bootloader with the original connectivity kit. If you were successful before, it should work again. One thing to test: USB drivers in Windows are bound to a device attached to a certain USB port. If you plug the device later on a different port, it will try to install a new instance of the driver for the other port. Perhaps you are connecting it on a different port than the one you installed the driver on? It could've happened that you disabled signed drivers, installed them on a certain USB port, then windows re-enables the signed drivers lock (behind your back), and when you plug it in a different port it is unable to install drivers there. I'm not certain this is your problem, but it's one conjecture that fits your symptoms. Try different ports, trying to find which one you used when it worked. |
|||
03-04-2018, 04:34 AM
Post: #157
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(03-03-2018 05:10 AM)The Shadow Wrote: Does LIBCLEAR erase the entire library directory? Or do you specify a list of targets like PURGE?You provide the library ID you want to clear. It only clears the data associated with the given library, does not uninstall the library. This is in case you need to "factory reset" a library if you messed up the settings badly. It also cleans up if the library leaves data stored after you detach it. The system does not delete library private data when you detach it, since you might want to attach it back later and keep using it, so to completely clean uninstall you need to DETACH and LIBCLEAR the same library ID. (03-03-2018 05:10 AM)The Shadow Wrote: I gather that it doesn't matter what directory you're in when you ATTACH a library, unlike oldRPL? I never did see the point in ATTACHing anywhere but HOME. Correct, I could never find any practical use for libraries attached to directories, so newRPL attaches libraries to the system, not to any particular directory. They become part of the system just as much as any built-in library. LIBSTO and LIBRCL look fun. (03-03-2018 05:10 AM)The Shadow Wrote: EDIT: Just made my first library, and some of the names in the menu are filled with extraneous characters. They still work as programs, but don't display help. Can you email me your entire source directory? I need to build your library with the debugger on to see what's going wrong with those names. The RS-hold-menu key works in Algebraic mode only. Similar to the variables menu, where RS means RCL, using RS will RCL the library itself, but if you put the quotes to go into A mode, RS will put the variable name (in this case library ID) on the command line. |
|||
03-04-2018, 06:19 PM
Post: #158
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
I sent you the directory, Claudio.
Do variables not in the $VISIBLE list count as being in the library's private directory? Using the MyHW example from the wiki, would 'WORLD' LIBRCL return "world!"? Likewise, LIBSTO would overwrite it? If not, how does one initialize library settings? Also, I'm assuming that library variables don't conflict in any way with variables on the calculator, correct? So if I had attached MyHW, I could have a variable called WORLD elsewhere and it wouldn't be a problem? |
|||
03-04-2018, 09:31 PM
Post: #159
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
(03-04-2018 06:19 PM)The Shadow Wrote: I sent you the directory, Claudio. No, data your library stores with LIBSTO is not included in the library, but in a separate directory, created automatically when the first variable is stored. 'WORLD' LIBRCL will error 'undefined variable'. Invisible commands and data included with the library are unnamed, cannot be RCL'd in any way. If you want the user to be able to get some read-only data, you need to make it visible or provide a program that returns it to the stack. (03-04-2018 06:19 PM)The Shadow Wrote: If not, how does one initialize library settings? I'm still debating between 2 methods: a) Library commands must detect if the data exists (by using IFERR with LIBRCL) and either recreate with LIBSTO or at least provide a default value. b) Use a $HANDLER program that will be called on installation, removal and have it recreate the variables. The first solution is more robust, but will add some burden on the programmer. The second solution also adds some burden (to write that handler), and doesn't address the case where the data is not there (for any reason, could've gotten corrupted and removed by the memory checker). (03-04-2018 06:19 PM)The Shadow Wrote: Also, I'm assuming that library variables don't conflict in any way with variables on the calculator, correct? So if I had attached MyHW, I could have a variable called WORLD elsewhere and it wouldn't be a problem? In newRPL variables NEVER conflict with commands. The only issue when you have a variable named the same as a command, is that you cannot use it in ways where the command would be valid. For example 3 'DROP' STO is fine, 'DROP' RCL works as expected, but obviously, if you write just DROP, it will be compiled as a command, not as an identifier. In a symbolic, 'DROP(a)' will fail since it will recognize the command DROP as not allowed in a symbolic. So you can always create these variables, as long as you refer to them by quoted name. The WORLD case you present is completely different, though. WORLD becomes an internal library command, not exposed to the user, therefore the word WORLD will never be compiled to the library command WORLD (which doesn't exist by the way, only commands in the $VISIBLE list receive a name, all others are null-named. On the other hand, if you do create a variable with LIBSTO, it will never conflict with other variables. Physically it's stored in a directory '.Settings/LIBDATA/MyHW', that should clarify the effect for your use. |
|||
03-04-2018, 10:09 PM
(This post was last modified: 03-05-2018 05:48 AM by The Shadow.)
Post: #160
|
|||
|
|||
RE: newRPL - build 1001 released! [update:build 1046]
Thanks, that was very informative.
(03-04-2018 09:31 PM)Claudio L. Wrote:(03-04-2018 06:19 PM)The Shadow Wrote: If not, how does one initialize library settings? I can think of refinements for both of these, though I'm not sure how practical they are to code: a) Have LIBRCL take two arguments instead of one. The second one is a default value - if the recall fails, the default gets both returned and LIBSTO'd. b) Instead of, or in addition to, $HANDLER, there is $DEFAULTS. This is a simple list of names and content, which are to be stored into the directory when the library is attached. There might also be a command LIBDEF which would restore the directory to the default values in case things get borked up. I personally think that either version of b) is more elegant, provided there's a way to re-run the $HANDLER or restore the $DEFAULTS. Otherwise, programmers will effectively have to implement a) anyway. |
|||
« Next Oldest | Next Newest »
|
User(s) browsing this thread: 7 Guest(s)