Re: [Bug-apl] [patch] float scanning broken for certain numeric locales & suggestions

Hi Elias, actually what you see in the SVN repository is what "make dist" thinks should be distributed. It is exactly the files in the apl-1.X.tar.gz (and possibly some stale ones if I forgot to remove them). The thing with auto-generated files is that there are two steps: autoreconf and ./c

Re: [Bug-apl] Print width

Hi, that was on purpose because the default ⎕PW is 80 and on some 80 column terminals this causes an extra empty line to be printed. /// Jürgen On 01/28/2014 01:18 PM, Kacper Gutowski wrote: Hi, After lines are neatly wrapped in r109 in SVN, but I think there's some off by one error; it beha

Re: [Bug-apl] Transpose crashes when faced with empty array

Hi, thanks. Fixed in SVN 110. /// Jürgen On 01/28/2014 08:12 AM, Elias Mårtenson wrote: Very easy to reproduce this one: ⍉⍬ This gives the following error: == Assertion failed: !shape.is_empty() i

Re: [Bug-apl] Mismatched free( )/delete/delete [ ], invalid read of 4 bytes, & uninitialized values

Dear Fred, ad 1) the delete is actually correct. the Cells to which the pointer points have beed destructed before and delete[] would lead to double destruction. I have inserted a variable long_ravel in Value.cc which hopefully confuses valgrind enough to not complain anymore. ad 2) I have ch

Re: [Bug-apl] Build failure on OSX

Hi, thanks. fixed in SVN 112. /// Jürgen On 01/29/2014 08:45 AM, Elias Mårtenson wrote: In the function Tokenizer::scan_real(), there is a call to exp10(). This function is unique to glibc and does not exist on OSX (and likely not on FreeBSD either, although I haven't checked). Changing the

Re: [Bug-apl] Print width

Hi Kacper, I see. I have changed back to the old ⎕PW behavior and made 79 the default ⎕PW value. SVN version 113. /// Jürgen On 01/28/2014 08:55 PM, Kacper Gutowski wrote: On 2014-01-28 14:23:01, Juergen Sauermann wrote: that was on purpose because the default ⎕PW is 80 and on some 80 column

Re: [Bug-apl] Mismatched free( )/delete/delete [ ], invalid read of 4 bytes, & uninitialized values

Hello Freg, I checked that class Cell has no destructor, so delete and delete [] will not make a difference. I have changed to delete [] to make valgrind happy, see SVN 114. I will also look into your updated email but this will take a bit longer since I will be AFK for two weeks. Best Reg

Re: [Bug-apl] Inverse Tranfer Form fails to reconstruct variable

Hi Fred, I believe you should use the variable name instead of the variable value with ⎕TF, eg: 1 ⎕TF 'R' instead of 1 ⎕TR R and 2 ⎕TF 'R' instead of 2 ⎕TR R Also saw that you use -T -- cool! Best Regards, Juergen On 01/29/2014 06:01 AM, Frederick H. Pitts wrote: Gentle people,

Re: [Bug-apl] Inverse Tranfer Form fails to reconstruct variable

Regards, Fred On Wed, 2014-01-29 at 14:23 +0100, Juergen Sauermann wrote: Hi Fred, I believe you should use the variable name instead of the variable value with ⎕TF, eg: 1 ⎕TF 'R' instead of 1 ⎕TR R and 2 ⎕TF 'R' instead of 2 ⎕TR R Also saw that you use -T -- cool!

Re: [Bug-apl] Cleanup when SIGINT or SIGHUP is encountered

Hi Elias, thanks, included in SVN 116. /// Jürgen On 01/29/2014 03:07 PM, Elias Mårtenson wrote: Currently, the cleanup() function is never called when the process is killed. This causes the Emacs plugin to leave the socket file in /tmp. It would also prevent the terminal colour settings fro

Re: [Bug-apl] Near-real numbers not handled properly

Hi, thanks, fixed in SVN 117. /// Jürgen On 01/30/2014 10:23 PM, Kacper Gutowski wrote: Hi, It seems that functions requiring “near-real numbers” as arguments throw DOMAIN ERROR when given a value represented internally as complex even if imaginary part is exactly equal zero. 1 > 0J0

Re: [Bug-apl] Small workspace of an example from "APL in Exposition" corrupts keyboard input after reload

Hi, thanks, fixed in SVN 118. /// Jürgen On 02/02/2014 01:19 PM, ds.jaypee wrote: Environment: - GNU APL 1.2 or latest checkout from SVN, built from source - OS X Mavericks, MacBookPro 8 GB RAM, More than 10 GB free disk space Workspace, - Manually edited and tested on Dyalog APL, - Exported

Re: [Bug-apl] "SYSTEM_LIMIT (fun_oper)" error

Hi Fred, when designing derived functions I had the choice between a faster variant that uses pre-allocated memory for derived functions and a somewhat slower variant that uses dynamic memory allocation (i.e. new/delete directly or indirectly (eg. vector<>)). Since this happens on a performa

Re: [Bug-apl] Print width

Hi Kacper, I have rewritten the ⎕PW line break algorithm. A problem remaining is that there seems to be no precise description of how this should look like, in particular for nested arrays. The old algorithm tried to not break nested arrays (which caused the output to be longer than ⎕PW if a

Re: [Bug-apl] No errro message when defining functions using

Hi Elias, not on my machine: ⎕FX 'foo[x]' 'x' 1 )fns Could it be that you had foo defined earlier? The syntax error is then due to a niladic function called with an axis (which is no a valid pattern in APL). The return value of 1 is the index of the line that has caused the error

Re: [Bug-apl] Unique of nested arrays doesn't work

Hi, thanks, fixed in SVN 120. /// Jürgen On 02/07/2014 03:33 PM, Daniel H. Leidisch wrote: Hello! Unique with nested arrays results in a domain error: ∪1⍴¨1 2 3 4 1 DOMAIN ERROR ∪1⍴¨1 2 3 4 1 ^ ∪'foo' 'bar' 'baz' 'foo' DOMAIN ERROR ∪'foo' 'bar' 'baz' 'foo'

Re: [Bug-apl] Overbar not allowed in Gnu APL variable names but is in ISO/IEC 13751:2000(E)

Hi Fred, I have added overbar to the characters allowed in variable names. see SVN 121. I haven't seen ¯ in variable names either, but IBM APL2 supports them and so do we now. Note that ∆ and ⍙ are also valid characters (even as first character) in names. I would not use ⍺, ⍶, ⍵, and ⍹ beca

Re: [Bug-apl] Trouble with GNU APL 1.2

Hi Thomas, looks more like a signed/unsigned problem to me. Or order of constuctors. I have made the order of constructors more explicit and I have added some more information printed when this happens, see SVN 122. Please let me know if this helps. /// Jürgen On 02/11/2014 07:10 AM, Thomas

Re: [Bug-apl] A little APL plugin for Vim

Hi Thomas, yes that was it. Cool. Maybe its worthwhile to mention this in the .vim file. /// Jürgen On 02/14/2014 04:47 PM, baruc...@gmx.com wrote: Juergen Sauermann wrote: in my vim it merely shows :emenu APL.I Hummm... :set wildmenu is the key. I have it in my .vimrc but you can

Re: [Bug-apl] Potential performance optimisation tested

Hi, I have added the proposed optimisation, SVN 123. Hope it does not break anything. /// Jürgen On 02/02/2014 03:35 PM, Frederick H. Pitts wrote: Elias, This approach to optimizating Gnu APL sounds very reasonable to me. I hope it succeeds. May Conway's "Game of Life" go faster.

Re: [Bug-apl] Implementing realtime variable viewer support

Hi Elias, I believe deriving from class Symbol is not the right way because there can be multiple pointers to the original symbol around and when you simply replace a symbol in class Workspace then you have a mix of the original and the new derived symbol. I have therefore added a callback fu

Re: [Bug-apl] External tools acting on interpreter commands (nabla and )ED)

Hi, I dare to disagree. In my opinion the ∇ editor is very useful in APL scripts (not used as interactive editor, but to define functions used later on in the script). Functions specified with ∇ are IMHO much more readable than functions specified with ⎕FX, in particular when they are long.

Re: [Bug-apl] Unique of nested arrays doesn't work

Hi Kacper, thanks, should be fixed in SVN 125. /// Jürgen On 02/14/2014 06:16 PM, Kacper Gutowski wrote: On 2014-02-14 16:15:33, Juergen Sauermann wrote: fixed in SVN 120. In r122, it doesn't give domain error anymore, but I think that results are not correct: ⎕RL 1 8

Re: [Bug-apl] No errro message when defining functions using

where the error occurred now. Is there a way to get more detailed information as to what the error was? Right now I simply display "Parse error" but it would be nice if it was a bit more precise. Regards, Elias On 14 February 2014 01:35, Juergen Sauermann mailto:juergen.sauerm

Re: [Bug-apl] Small workspace of an example from "APL in Exposition" corrupts keyboard input after reload

Hi Jean-Pierre, I have debugged the matter a little further. It seems like the readline library sometimes skips too few characters. For example, after entering ∇C[⎕]∇ and then recalling the line and deleting the last character with Backspace, the new line should be: ∇C[⎕] Howeve

Re: [Bug-apl] Re : Re: Trouble with GNU APL 1.2

Hi Thomas, your SVN checkout was correct. I was on vacation for two weeks and I am working on the backlog that has built up in the meantime. /// Jürgen On 02/16/2014 11:22 AM, Thomas Baruchel wrote: looks more like a signed/unsigned problem to me. Or order of constuctors. I have made the o

Re: [Bug-apl] Problems with shared variables

Hi, I believe the problems below are fixed in SVN 128. Ad 1) I changed the assertions Symbol.cc to short warnings visible in )MORE. The kill() should work now as well, looking at errno rather than just the return value of kill(). Ad 2) hopefully fixed. Ad 3) hopefully fixed. I could further

Re: [Bug-apl] A suggestion concerning the readline library

Hi Thomas, welcome to the world of APL! I did not know rlwrap before, but I just installed it and tried APL with it. Cool! As you may have seen there were some problems in GNU APL recently that seems to be readline related. The root cause of these problems is still unknown. After having GNU

Re: [Bug-apl] Unexpected DOMAIN ERROR when * is applied to conforming numeric vectors

Hi Fred, thanks, fixed in SVN 129. /// Jürgen On 02/14/2014 06:17 PM, Frederick H. Pitts wrote: Gentle people, Please exercise the attached SHALLIT0.tc.gz to confirm that a DOMAIN ERROR occurs on the last executable statement. I'm currently using Gnu APL, svn 117. It appear

Re: [Bug-apl] Re : Re: Trouble with GNU APL 1.2

Hi Thomas, I have found the root cause of your problem. It was non-ASCII characters (é) in the filename /home/thomas/Téléchargements/GNUAPL/trunk/workspaces/CONTINUE I have fixed this in SVN 130. There might be more such cases; pleas keep reporting them. /// Jürgen On 02/16/2014 11:22 A

Re: [Bug-apl] Gnu APL and ANSI escape sequences

Hi Fred, I believe you simply need to create the sequence and insert it into the other text as appropriate, eg: ESCSEQ←(⎕UCS 27),'[0;35;48m'⍝ RED foreground 'Hello',ESCSEQ,'World' HelloWorld Note that ⎕UCS is more portable than ⎕AV or ⎕AF. /// Jürgen On 02/14/2014 10:04 PM

Re: [Bug-apl] Small workspace of an example from "APL in Exposition" corrupts keyboard input after reload

et_LX(); } Regards, Jean-Pierre Le 15/02/2014 20:06, Juergen Sauermann a écrit : Hi Jean-Pierre, I have debugged the matter a little further. It seems like the readline library sometimes skips too few characters. For example, after entering ∇C[⎕]∇ and then recalling the line and deletin

Re: [Bug-apl] Re : Re: Trouble with GNU APL 1.2

, baruc...@gmx.com wrote: Juergen Sauermann wrote: I have found the root cause of your problem. It was non-ASCII characters (é) in the filename Thank you very much for all your work; it seems to work fine now. Do you already have an idea of the release date of the version allowing direct

Re: [Bug-apl] Implementing realtime variable viewer support

Hi Elias, The name of a Symbol uniquely identifies a symbol within one Workspace (and there is only one Workspace). The only cases where the pointer changes is when a new workspace with the same symbol name is loaded or when the symbol is explicitly )ERASEd or ?EXed. I all these cases you

Re: [Bug-apl] Dyadic / (replicate) does not work with ¨ (each)

Hi, I have changed the parser to degrade / and friends from operator to function in the examples below. One remaining case is (with / any of / ⌿ \ or ⍀): S//B where the left / is a function when S is a value and and an operator if S is a function (which is not known at parse time). The bind

Re: [Bug-apl] Unterminated strings does not result in error

Hi Elias, this is a feature. /// Jürgen On 02/18/2014 04:38 AM, Elias Mårtenson wrote: If I type the following: *'foo* It is parsed as: *'foo'* Instead of giving an error. Is this expected behaviour? To me, it would make much more sense if this was flagged as a syntax error.

Re: [Bug-apl] Calling Value::print() with different ⎕PW

Hi Elias, normally you do something like this: // const Value & value; PrintContext pctx(style, Workspace::get_PP(), Workspace::get_CT(), Workspace::get_PW()); PrintBuffer pb(value, pctx); UCS_string ucs(pb, value.get_rank(), pctx.get_PW()); ... You can use your own ⎕PW value instead of Work

Re: [Bug-apl] Question regarding the power operator (as in Dyalog APL)

Hi Thomas, currently the power operator is not on my TODO list. I may implement it at some point in time, but the mere fact that it is described in "Mastering Dyalog APL" - which is a very good book - does not suffice. In most cases the power operator can be replaced with a simple loop (I kno

Re: [Bug-apl] Novice question regarding APL

Hi Thomas, you can use ⎕INP like this: TEXT←⎕INP 'END-OF-TEXT' ⍝ the subsequent lines until 'END-OF-TEXT' 1 1 1 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 1 1 1 1 1 1 0 0 0 1 0 0 0 0 0 0 1 0 'END-OF-TEXT' MATRIX←⊃⍎¨TEXT /// Jürgen On 02/19/2014 07:28 AM, baruc...@gmx.com wrote: Hi Jürgen, I can now so

Re: [Bug-apl] How to get SEV_ERASED event?

Hi, A←1 2 3 ⎕EX 'A' /// Jürgen On 02/19/2014 10:15 AM, Elias Mårtenson wrote: I have a hard time testing the Emacs mode's treatment of SEV_ERASE, as I can't seem to get such event. Nothing I tried (including calling )ERASE on it) causes this to happen. When is this event sent? Regards, El

Re: [Bug-apl] Lambda functions are not loaded from a workspace

Hi, thanks, fixed in SVN 134. /// Jürgen On 02/19/2014 08:04 AM, Elias Mårtenson wrote: When loading a workspace that was previously saved using the )SAVE command, any functions defined using name←{...} are not loaded. The functions are there, as they can be seen in the XML file, but they

Re: [Bug-apl] Problems with shared variables

Hi, thanks, should work in SVN 136. /// Jürgen On 02/20/2014 11:57 AM, Kacper Gutowski wrote: On 2014-02-16 18:09:06, Juergen Sauermann wrote: Ad 1) I changed the assertions Symbol.cc to short warnings visible in )MORE. I wouldn't guess to check )MORE upon getting VALUE ERROR on s

Re: [Bug-apl] Problems with shared variables

sufficient. /// Jürgen On 02/20/2014 12:16 PM, Elias Mårtenson wrote: On 20 Feb 2014 18:57, "Kacper Gutowski" <mailto:mwgam...@gmail.com>> wrote: > > On 2014-02-16 18:09:06, Juergen Sauermann wrote: > > Ad 1) I changed the assertions Symbol.cc to short war

Re: [Bug-apl] Bug when assigning variable to result of function call on self

Hi Elias, yes. Fixed in SVN 137. /// Jürgen On 02/21/2014 04:02 AM, Elias Mårtenson wrote: On 20 February 2014 17:24, Kacper Gutowski > wrote: In this function, next is 2 (a label) rather than function you defined earlier. So you literally have [3] S←2 S.

Re: [Bug-apl] Problems with shared variables

cs when it contains something. Would this be the right approach? Regards, Elias On 20 February 2014 23:44, Juergen Sauermann mailto:juergen.sauerm...@t-online.de>> wrote: Hi, the normal 3-line error printout is documented by IBM and shall remain as is for compatibility.

Re: [Bug-apl] More "Mismatched free() / delete / delete []" valgrind errors

Hi Fred, thanks, fixed in SVN 140. /// Jürgen On 02/21/2014 03:18 AM, Frederick H. Pitts wrote: Gentle people, In Gnu APL file file_io.cc, lines 597, 617, and 643, `[]' should be inserted between the `delete' and `del' to match the `new char [ bytes ... ]' on lines 591, 612, and 636,

Re: [Bug-apl] Another problem with direct-functions

Hi, yes. fixed in SVN 142. /// Jürgen On 02/21/2014 10:04 PM, baruc...@gmx.com wrote: Is this another bug: R ← { ⍵ }¨⍳15 works in Dyalog APL and NGN APL but not in GNU APL. Regards,

Re: [Bug-apl] APL line editor: proof of concept (c code)

Hi Thomas, could be a cool thing. One thing that I am missing is a bit more information of what it exactly does and how it should integrate with GNU APL (for example). Currently I am trying to keep keyboard input and fonts out of GNU APL as much as possible. The reason is that different platform

Re: [Bug-apl] No ubuntu packages

Hi Akiva, not sure if I don't understand the subject? You can create Debian packages with make DEB in the top-level directory (provided that you have the necessary debian tools installed). /// Jürgen On 02/24/2014 09:29 AM, Akiva wrote: Also, may be interested in contributing on non-core stu

Re: [Bug-apl] Crash when forgetting to close lambda function

Hi, thanks, fixed in SVN 145. /// Jürgen On 02/25/2014 05:17 PM, Elias Mårtenson wrote: Very easy to reproduce this one. Simply type a single *{* character. This will crash GNU APL with the following stack trace: Token: 0x4C03001B { at Token.cc:490 (void *)get_Class(): '0x1b' at Toke

Re: [Bug-apl] Regression: Lambda functions can no longer be assigned to a variable

Hi, thanks. Should work again in SVN 146. Unfortunately the rules governing the syntax of {} are not standardised and still somewhat unclear to me. /// Jürgen On 02/25/2014 05:19 PM, Elias Mårtenson wrote: This used to work a few versions back, now I'm just getting a SYNTAX ERROR: *foo ← {

Re: [Bug-apl] Strange text corruption when sending a list to ⍞

Hi, thanks, fixed in SVN 147. /// Jürgen On 02/27/2014 11:02 AM, Elias Mårtenson wrote: When I type the following: *⍞←1,'x'* I get a strange result: 1#x Were # is a non-existing Unicode character (codepoint U+). Regards, Elias

Re: [Bug-apl] Bug with lambda functions

Hi, thanks, fixed in SVN 148. /// Jürgen On 02/28/2014 01:35 PM, baruc...@gmx.com wrote: Hi, I have a bug with: ( { ⍵ }¨⍳10 ), ( { - ⍵ }¨ ⍳10) two lambda on the same line??? Regards,

Re: [Bug-apl] Getting list of quad-commands and system commands

Hi Elias, for commands, you can #define macro cmd_def() and then #include Command.def. For Quad-vars you can #define macros ro_sv_def() and rw_sv_def and then #include SystemVariable.def. For example, to get the strings for all commands: #define cmd_def(cmd_str, code, _arg) #cmd_str #include

Re: [Bug-apl] Getting list of quad-commands and system commands

to release 1.3? Regards, Elias On 1 March 2014 23:50, Juergen Sauermann mailto:juergen.sauerm...@t-online.de>> wrote: Hi Elias, for commands, you can #define macro cmd_def() and then #include Command.def. For Quad-vars you can #define macros ro_sv_def() and rw_sv_d

Re: [Bug-apl] Problems with shared variables

Hi Kacper, thanks for reporting this. I believe I have fixed this in SVN 149. There is now a system limit of 65kByte for the size of shared variables, caused by the UDP protocol used for the communication between processors. /// Jürgen On 03/02/2014 07:42 AM, Kacper Gutowski wrote: Moreover,

Re: [Bug-apl] Commute compress

Hi, I believe this relates to an earlier discussion on this list (around Feb. 24, 2014). You can put the left arg of /⍨ in parentheses which marks it as a value at ⎕FX time (the default interpretation is function) like this: (v)/⍨2|v←⍳4 1 3 /// Jürgen On 03/02/2014 11:31 AM, Dani

Re: [Bug-apl] Each first

Hi Daniel, thanks for reporting thisd. Fixed in SVN 149. /// Jürgen On 03/02/2014 11:37 AM, Daniel H. Leidisch wrote: Hello! Each first of a nested lists behaves weirdly. ↑¨(1 2) (3 4) 1 3 ≡↑¨(1 2) (3 4) 1 So far so good, but: 8⎕cr ↑¨(1 2) (3 4) ┌→──┐ │┌─┐ ┌─┐│ ││

Re: [Bug-apl] Multiply and add with axis

Hi, thanks, fixed in SVN 150. /// Jürgen On 03/02/2014 11:50 AM, Daniel H. Leidisch wrote: Hello! Multiply and add doesn't seem to work with axis specifications: (3 5⍴1) ×[1] ⍳3 INDEX ERROR (3 5⍴1)×[1]⍳3 ^ ^ INDEX ERROR (3 5⍴1)+[1]⍳3 ^ ^

Re: [Bug-apl] D-fn binding leackage, Nested d-fns

Hello Daniel, I would like to explain my point of view regarding Lambdas aka. D-funs. Initially I considered them as not needed; they are not defined in the APL standard and the rules related to them seem to be rather unclear. After seeing some requests to support them I started thinking about

Re: [Bug-apl] Patch: gnu-apl-ansi-colors

Hi David, thanks, included in SVN 151. /// Jürgen On 03/03/2014 04:13 AM, David B. Lamkins wrote: See attached.

Re: [Bug-apl] [Feature request] "]boxing on" analogue

Hi, I have added command ]BOXING which takes an argument OFF, 2, 3, 4, 7, 8, or 9. See SVN 152. /// Jürgen On 03/02/2014 12:06 PM, Daniel H. Leidisch wrote: Hello! When experimenting at the REPL, I often find myself typing n⎕CR repeatedly at the beginning of lines. As an Emacs user, it wou

Re: [Bug-apl] D-fn binding leackage, Nested d-fns

sense that you can also nest them. I'd also like to present my previous email as an example to a situation where I wanted nested lambdas. Regards, Elias On 3 March 2014 20:13, Juergen Sauermann mailto:juergen.sauerm...@t-online.de>> wrote: Hello Daniel, I would like to expla

Re: [Bug-apl] Should a native function be expungeable?

Hi, thanks, fixed in SVN 154. /// Jürgen On 03/04/2014 01:26 AM, David B. Lamkins wrote: Once one defines a native function, should the function be expungeable? Attempting to )erase the function fails silently; applying ⎕ex to the native function's name causes a crash.

Re: [Bug-apl] Missing Doxyfile

Hi, thanks, added in SVN 154. /// Jürgen On 03/04/2014 05:46 AM, David B. Lamkins wrote: The doxygen tool used by the DOXY make target looks for a configuration file name Doxyfile. The attached Doxyfile retains all defaults except for INPUT = . src SOURCE_BROWSER = YES which is the minimum

Re: [Bug-apl] )COPY and )PCOPY

Hi, true. fixed in SVN 154. /// Jürgen On 03/03/2014 07:32 PM, David B. Lamkins wrote: )COPY and )PCOPY should not change wsid.

Re: [Bug-apl] Rank operator and function assignment

Hi Elias, I assume by spec you mean ISO/IEC 13751. They give the following syntax: Z ← f ⍤y B (deriving monadically) and Z ← A f ⍤y B (deriving dyadically). They say that f ⍤y is a function which is then called with arguments B and possibly A. Thus Z is a value and not a function; the functi

Re: [Bug-apl] Possible improvement to ]BOXING

Hi Elias, I believe it would be more confusing if ⎕CR is sometimes computed and sometimes not. The example below might be easy to detect but what about cases like Z←8 ⎕CR xxx ◊ Z or some ⎕CR returned by a function? /// Jürgen On 03/04/2014 05:35 AM, Elias Mårtenson wrote: I love ]BOXING a

Re: [Bug-apl] Rank operator and function assignment

Hi Daniel, how does NARS2000 define f⍤B and A f⍤B (the standard does not)? /// Jürgen On 03/04/2014 07:52 PM, Daniel H. Leidisch wrote: Hello! Juergen Sauermann writes: What you probably intended to do is: * {'foo',⍵}⍤1 (4 5⍴⍳10)* foo 1 2 3 4 5 foo 6 7 8 9 10 foo 1 2 3

Re: [Bug-apl] Rank operator and function assignment

Hi Tobia, I see, thanks for the explanation. I will look into this, but can't promise too much short term because it probably requires some rework of the parser. /// Jürgen On 03/06/2014 01:27 AM, Tobia Conforto wrote: Hi Juergen how does NARS2000 define f⍤B and A f⍤B (the standard does no

Re: [Bug-apl] Getting list of quad-commands and system commands

expansion of both system commands and quad functions in the Emacs mode. Regards, Elias On 2 March 2014 01:52, Juergen Sauermann mailto:juergen.sauerm...@t-online.de>> wrote: Hi, providing descriptions of functions is a lot of work. Most people know A

Re: [Bug-apl] Suggestions needed: Editing of named lambdas

Hi, in GNU APL you can have lambdas without result (eg. {}) and they don't start with *λ*←. As Kacper has pointed out you can also have user-defined functions starting with *λ*←. I have no plans to prevent that, but I have also no idea where the lambdas will bring us in GNU APL. Right now {

Re: [Bug-apl] Implementation of ∩ - Intersection

Hi Elias, yes, it would make sense. I have added dyadic ∪ (union) and diadic ∩ (intersection), see SVN 157. /// Jürgen On 03/06/2014 10:27 AM, Elias Mårtenson wrote: Would it make sense to implement the ∩ function in GNU APL? In Dyalog, it represents the intersection function which returns

Re: [Bug-apl] Configure wrongly reports Readline missing

Hello Jan-Pieter, I would expect that the check for curses (a little higher/before) the check for readline has indicated a problem. Could you please send me your config.log? Thanks, Jürgen On 03/06/2014 03:48 PM, Jan-Pieter Jacobs wrote: Hey, I noticed (in GNU APL 1.2) that I did not get

Re: [Bug-apl] Plus sign but no )MORE

Hi, thanks, fixed in SVN 158. /// Jürgen On 03/07/2014 05:08 PM, Elias Mårtenson wrote: If I type →2 on the prompt, I get an error message with a plus: SYNTAX ERROR+ But When I type )MORE, I get the answer: NO MORE ERROR INFO Is this correct? Regards, Elias

Re: [Bug-apl] svn update conflicts on *.gmo files

Hi David, you could try: svn update --accept theirs-full Unfortunately the gmo files are put into the repository even though they are generated. The Makefile doing this was installed by the NLS package and I don't know how to change it. Simply removing the .gmo files from the repo could

Re: [Bug-apl] Rank operator and function assignment

Hi Kacper, see some comments inline below. On 03/04/2014 10:52 PM, Kacper Gutowski wrote: On 2014-03-04 18:53:29, Juergen Sauermann wrote: They say that f ⍤y is a function which is then called with arguments B and possibly A. Thus Z is a value and not a function; the function is created

Re: [Bug-apl] Rank operator and function assignment

first (throwing the rank error because of argument 1 which is expected to be a vector). /// Jürgen On 03/04/2014 07:52 PM, Daniel H. Leidisch wrote: Hello! Juergen Sauermann writes: What you probably intended to do is: * {'foo',⍵}⍤1 (4 5⍴⍳10)* foo 1 2 3 4 5 foo 6 7 8 9 10

Re: [Bug-apl] Common IO operations

Hi Elias, sounds like a case for FILE_IO and partition: 1. get a line with fgets() in FILE_IO 2. call partition (dyadic ⊂) For example (FILE_IO part omitted): TEXT←"Hello,World,xyz" (TEXT≠",")⊂TEXT Hello World xyz ⍴(TEXT≠",")⊂TEXT 3 ⊃((TEXT≠",")⊂TEXT)[1] Hello ⊃((TEXT≠",")⊂TEXT)[2] World ⊃((

Re: [Bug-apl] Ideas for an online demo-version of GNU APl

Hi Thomas, nice idea, we should try that. I believe that it should already be possible to start GNU APL from eg. an apache web-server; the missing piece could be the backward direction (user input -> webserver). Security should not be too hard on linux systems. A problem remaining is where to

Re: [Bug-apl] Proposal for initial multicore support

Hi David, I think your ideas are correct. I have planned multicore support for GNU APL 1.4 and every help is welcome. Actually parallel processing was one of the main reasons for me to write GNU APL. I published a Ph.D thesis "A parallel APL computer" (in German) back in 1990. We had built a

Re: [Bug-apl] Ideas for an online demo-version of GNU APl

"safe mode"? This would restrict the availability of shared variables and native functions? Regards, Elias On 10 Mar 2014 19:19, "Juergen Sauermann" mailto:juergen.sauerm...@t-online.de>> wrote: Hi Thomas, nice idea, we should try that. I believe

Re: [Bug-apl] Assertion error when typing ()

Hi Elias, thanks, fixed in SVN 160. /// Jürgen On 03/09/2014 05:46 PM, Elias Mårtenson wrote: After some playing around, I just decided to type an opening and closing paren: *()* I then got the following error (trying again after restarting will not reproduce the error): Assert

Re: [Bug-apl] Ideas for an online demo-version of GNU APl

add a flag that enables "safe mode"? This would restrict the availability of shared variables and native functions? Regards, Elias On 10 Mar 2014 19:19, "Juergen Sauermann" mailto:juergen.sauerm...@t-online.de>> wrote: Hi Thomas, nice idea, we should try th

Re: [Bug-apl] What is the daemon mode?

Hi, seems like the parent closes stdin/cin. Have to find out why. /// Jürgen On 03/10/2014 05:01 PM, baruc...@gmx.com wrote: Hi, I tried to figure out how to enable the daemon mode and how to use it, but with no success; first of all, does it work? Each time I try to enable it, I have an inf

Re: [Bug-apl] What is the daemon mode?

Hi Thomas, I fixed the infinite lopp when the interpreter has no input. It now stops after 10 attempts within short time (to distinguish that from ^D typed by the user). Deamon mode (-d) simply fork()s the process of the APL interpreter with the parent process exit()ing and the child continui

Re: [Bug-apl] Experimental OpenMP patch

Hi David, looks good! Some comments, though. 1 .you could adapt src/testcases/Performance.pt with some longer skalar functions in order to get some performance figures. You can start it like this: ./apl -T testcases/Performance.pt 2. I believe we should not bother the user with specifying p

Re: [Bug-apl] Experimental OpenMP patch

user-selected (or automatically based on cores) number of threads as default is important for this reason. Regards, Elias On 11 March 2014 23:07, Juergen Sauermann mailto:juergen.sauerm...@t-online.de>> wrote: Hi David, looks good! Some comments, though. 1 .you could ada

Re: [Bug-apl] Experimental OpenMP patch

g able to choose a compile-time value is good too. Regards, Elias On 11 March 2014 23:20, Juergen Sauermann mailto:juergen.sauerm...@t-online.de>> wrote: Hi, we could do it similar to the LOG macro where you can choose between more efficie

Re: [Bug-apl] Building on OS X 10.8.5 Mountain Lion - partly resolved….

Hi Peter, I believe you should do something along the lines of: CXX=llvm-gcc4 ./configure /// Jürgen On 03/11/2014 10:29 PM, Peter Teeson wrote: On 2014-03-11, at 4:30 PM, Peter Teeson wrote: Mac Pro Desktop dual quad core CPU's, Mountain Lion OS X 10.8.5 I DL'd the 2014-01-13 APL 1.2 tar

Re: [Bug-apl] OpenMP performance: first result

Hi David, I guess the circle functions and ⋆/⍟ might do a better job in raising your motivation! If I remember correctly then in 1990 we got a speedup of 5-6 on our 32 processor machine, which means that the break-even point is at about 6 cores. Unfortunately my own machine has only 2 cores

Re: [Bug-apl] Test case generation?

Hi, I'm afraid cut-and-paste is the only tool around. For performance testin it is probably simpler to )DUMP the Perfornmance.pt testcase at the end and then modify/start the dumped workspace with apl -f. Then you don't need to mess around with the matching of the interpreter output. Or just

Re: [Bug-apl] "Largest integer" isnt' really largest?

Hi Elias, thanks, fixed in SVN 162. In that range it can still happen that "small" differences occur because an operation may be internally performed as double and then converted to integer. The double has 48-1 bit precision and the integer 64-1 bit precision. It depends a little on how the actua

Re: [Bug-apl] Crash upon throwing an error from quadSYL setters

Hi David, thanks, fixed in SVN 163. /// Jürgen On 03/12/2014 06:51 AM, David Lamkins wrote: Any error thrown from a quadSYL setter triggers causes a heap error and backtrace. For example: ⎕syl[5;2]←1 INDEX ERROR ⎕SYL[5;2]←1 ^^ *** Error in `/usr/local/bin/apl': double f

Re: [Bug-apl] Building on OS X 10.8.5 Mountain Lion - SOLVED and How to setup your keyboard.

Hi Peter, my plan was to include your information in our README-6-porting in the next release. I am currently preparing that release, but have not reached that point yet because I am working on keyboard input related proposals and methods. Please feel invited to send an updated README-6-por

Re: [Bug-apl] Building on OS X 10.8.5 Mountain Lion - SOLVED and How to setup your keyboard.

that they understand that they don't have to download it themselves. Regards, Elias On 14 March 2014 21:52, Juergen Sauermann mailto:juergen.sauerm...@t-online.de>> wrote: Hi Peter, my plan was to include your information in our README-6-porting in the next release

Re: [Bug-apl] Performance optimisations: Results

Hi David, sounds like something is wrong. OpenMP states that their parallisation overhead is 4000 or so cycles which is much less than your measurements. Maybe some unintended sync between the threads? Could you send me the latest patch? /// Juergen On 03/14/2014 05:18 PM, David Lamkins wro

Re: [Bug-apl] "Largest integer" isnt' really largest?

that demonstrates this issue is attached. I respectfully submit that the documentation should reflect the narrower range for integers until such time this issue is corrected. Regards, Fred On Wed, 2014-03-12 at 13:24 +0100, Juergen Sauermann wrote: Hi Elias, thanks, fixed in SVN 16

Re: [Bug-apl] Performance optimisations: Results

ht of synchronization. That'd suggest that OpenMP infers the presence of shared state. I'll refresh the patch for you very soon. Basically, though, the only difference should be another pragma for the loop in check_value(). On Mar 14, 2014 10:22 AM, "Juergen Sauermann" m

Re: [Bug-apl] Error with the commute operator

Hi Thomas, you can put I in parentheses so mark it as a value: I ← 1 2 3 1 2 3 / ⍨ 1 1 0 1 2 I / ⍨ 1 1 0 SYNTAX ERROR I/⍨1 1 0 ^ ^ (I) / ⍨ 1 1 0 1 2 See also an earlier discussion on bug-apl regarding this topic. /// Jürgen On 03/16/2014 09:23 PM, baruc...@gmx.c

  1   2   3   4   5   6   7   8   9   10   >