Hi,
I really appreciate the discussions on this because I believe
there is a open issue (libraries and how to provision them).
1. First of all I would like to share my current point of view regarding
⎕-functions and
-variables. The (only ?) good thing about them is that they are ready to
be us
Hi Blake,
I believe your 1, 2, 3 below is exactly the current priorities of GNU
APL and 3. is already
present.
We use shared libraries for extending APL so you don't need an extra
developer lib
for that. The only thing needed is a header file declaring the functions
available in
GNU APL (may
Hi ,
don't wait any longer for ⎕host. Its already there - called popen() in
FILE_IO.
see man popen how to use it.
/// Jürgen
On 04/23/2014 03:33 PM, enz...@gmx.com wrote:
On Tue, 22 Apr 2014 22:14:21 -0400
Peter Teeson wrote:
Blake said:
"This means creating a quad function to load, exe
discontinued using gcc a while back and it is no longer installed as part
of Xcode.
For gcc external builds Apple uses llvm-gcc.
respect….
Peter
On 2014-04-22, at 12:39 PM, Juergen Sauermann
wrote:
Hi Peter,
man gcc ?
/// Jürgen
Hi David,
thanks, applied in SVN 222.
I also added base64 encoding and decoding according to RFC 4648.
/// Jürgen
On 04/21/2014 08:11 PM, David B. Lamkins wrote:
Diff attached.
Hi David,
thank you very much for all the flowers from you and the others in the
last days.
Not my birthday yet, but I'll keep them until then.
I agree that dependency tracking could be a good thing. Maybe we can use
the dependency
system of RPM or Debian packages for that (GNU APL has make t
Hi,
just to mention it, cells are not allocated by their constructor because
for cells "placement new" is always used. The allocation of all ravel
cells is
done by the Value constructor.
So the 2.2 billion "allocations" are actually 2.2 billion ravel cell
initializations
(without involving m
ng to the Böhm collector is
something to investigate...
Regards,
Elias
On 25 April 2014 18:36, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi,
just to mention it, cells are not allocated by their constructor
because
for cells "placement new"
Hi,
much simpler: edit a (UTF8 encoded) text file and )COPY it.
vi(m) is perfect for this.
Variables can be assigned like in APL, i.e.
VAR←1 2 3
Functions can be created via ⎕FX (hardly readable) or (may favorite)
using ∇:
∇Z ← AVE B
Z←(+/B)÷⍴B
∇
The file must start with #! so )COPY knows
Hi,
OK seems we have some agreements so far:
1. namespaces will be delayed at least until we have so many libs that
we need it,
2. RPM or debian packaging shall not be used for dependencies, but an
alternative has
not been found yet.
I personally would prefer a solution that also works
Hi,
the )COPY of text files is pretty new, so we have not much experience
with it yet.
I could make the ∇-editor aware that a )COPY is in progress and that
functions shall be
deleted automatically by the ∇-editor.
/// Jürgen
On 04/26/2014 03:08 AM, enz...@gmx.com wrote:
I delete the emai
files
and not for scrippt (.apl) files. This was mainly for compatibility with
IBM APL2 - I have always
considered that as not very handy.
/// Jürgen
On 04/25/2014 08:30 PM, enz...@gmx.com wrote:
On Fri, 25 Apr 2014 19:44:21 +0200
Juergen Sauermann wrote:
Hi,
much simpler: edit a (UTF8
Hi Peter,
thanks, I have changed the ?-editor.
My interpretation of eg [2?10] is "display lines 2-10 and edit line
10 (or 10.1 or 11 if 10 or 10 and 11 alreadyexists). See SVN 224.
/// Jürgen
On 04/25/2014 04:01 PM, Peter Teeson wrote:
As of svn 221
I expected the editor to allow editing of
also IBM make some assumptions about the editor used in their
)EDITOR command (and neither vi
nor emacs meet these assumptions).
/// Jürgen
On 04/26/2014 03:13 PM, enz...@gmx.com wrote:
On Sat, 26 Apr 2014 14:52:49 +0200
Kacper Gutowski wrote:
On 2014-04-26 12:08:14, Juergen Sauermann wrote
Hi Peter,
great!
I believe we can't do much about the remaining -rdynamic because your
compiler
apparently pretends to be g++ but then warns about a valid g++ argument
being (un-)used.
/// Jürgen
On 04/24/2014 02:35 PM, Peter Teeson wrote:
/bin/sh ../../libtool --tag=CXX --mode=compile g++
Hi Kacper,
very good explanation, thanks. Fixed in SVN 226.
/// Jürgen
On 04/26/2014 03:43 PM, Kacper Gutowski wrote:
On 2014-04-26 07:13:21, enz...@gmx.com wrote:
btw if you can take a look at the FILE_IO code in
http://lists.gnu.org/archive/html/bug-apl/2014-04/msg00269.html
and give me a
...@gmx.com wrote:
On Sat, 26 Apr 2014 12:34:06 +0200
Juergen Sauermann wrote:
Hi,
correct. Since saving an entire workspace is usually fast (unless it
contains huge amounts of data)
that should be ok. Think of a workspace as an analog of a C/C++ source
file where you typically
have several
Hi,
actually it should not. )HOST is a command so its arguments
are not computed (in GNU APL, but probably also others).
Thus 'workspaces/',a is passed verbatim to popen().
/// Jürgen
On 04/25/2014 11:10 PM, enz...@gmx.com wrote:
presently )host takes direct code
)host vi workspaces/
in the middle of the line is somewhat complicated
because we have
different input options (with readline, without readline, testcase files
etc.) that are
difficult to bring in line.
/// Jürgen
On 04/26/2014 05:12 PM, enz...@gmx.com wrote:
On Sat, 26 Apr 2014 15:15:03 +0200
Juergen Sauermann
ed if readline isn't a required lib as you say below - ? maybe
any right side number moves cursor to end of the line?
[⎕2-4] displays lines 2-4 and opens up line 5
enztec
On Sat, 26 Apr 2014 19:47:00 +0200
Juergen Sauermann wrote:
Hi,
I changed it to what the example in the IBM langua
Hi,
at some point in time I started writing a dynamic ]keyb. But then people
started
to complain about me using xmodmap (too old, too static, etc) and we now
have
several other methods as well (see README-3-keyboard).
The downside is that it has become almost impossible to figure the
current
help in a separate window,
automatically updated to correspond the the current active keymap. :-)
Regards,
Elias
On 27 April 2014 18:02, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi,
at some point in time I started writing a dynamic ]keyb. But then
peopl
Hi,
it is actually a feature. You are the second to wonder about it, so shall
I make that ./configurable ?
/// Jürgen
On 04/27/2014 08:05 AM, David B. Lamkins wrote:
A newline terminates a string constant. Shouldn't that be a syntax
error?
Hi David,
thanks. IBM APL provides only 1-4, but I have added 5 and 6 in SVN 227.
/// Jürgen
On 04/27/2014 01:18 AM, David B. Lamkins wrote:
quad-NC should return 5 for system variables and 6 for system functions
(Ref: ISO 13751, pg. 175).
In GNU APL, quad-NC returns negative 1 for these cas
Hi Blake,
0x86 is the culprit because the first byte of a non-ASCII UTF8 should be
11xx in binary.
I have changde the code so that the sequence is truncated after printing
the Bad UTF message and the stack trace and no assertion is thrown, SVN 227.
/// Jürgen
On 04/27/2014 04:28 AM, Bl
Hi Blake,
I have added the files, see SVN 228.
I also added one of the photos (which blew up the GNU APL tar file from a
slim 1.6 MB to 3.7 MB).
BTW how did you get the Unicomp keyboard? I sent a query to them last
year on their web-page and
never got a response?
/// Jürgen
On 04/27/2014 0
Hi David and all,
thanks for the link below.
Attached is a quick summary of my thoughts regarding libraries.
Looks like David and myself are on the same page, just the focus is a
bit different.
I wouldn't rule out ⎕CP for GNU APL in the long run, but I feel like not
having fully
understood wh
oards. They allow you to upload a
file containing the symbol design when you order. Making an APL
keyboard would be trivial.
http://www.wasdkeyboards.com/
Regards,
Elias
On 28 April 2014 00:40, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi Blake,
I hav
, although the files are drastically smaller, they're fine.
Thanks.
Blake
On Sun, Apr 27, 2014 at 11:40 AM, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi Blake,
I have added the files, see SVN 228.
I also added one of the photos (which blew up
Hi Blake,
it is definitely saving space, but it most likely taking more time.
Consider Z←A ⍴ IDX← ⍳ N to represent the creation (IDX ← ⍳ N) and use (Z
← A ⍴ IDX) of IDX.
At ravel cell level a cell is created with constructor IntCell() and
used with get_ravel(r). The
argument of IntCell is a
Hi,
I have added 18 ⎕CR and 19 ⎕CR to convert from Unicode strings
(like 'Hello ⍴ ∇' in APL) to byte vectors (character vectors with values
0-255) in UTF8 encoding
and back. See SVN 233.
/// Jürgen
On 04/28/2014 06:17 AM, David B. Lamkins wrote:
I can use lib_file_io to read a sequence of b
Hi,
generally speaking unnecessary clone() of values should of course be
avoided.
In GNU APL 1.0 and 1.1 there was a flag-based system of value ownership
where the last owner
would delete the value when giving up its interest in the value. This
system began like the
tmp flag, but then caused
Hi Elias,
I used ⎕CR instead of ⎕UCS because I believe it is more "standard" and also
has a different semantics in GNU APL (mix of integer and char, etc).
⎕UCS was
more for IBM APL2 compatibility.
Also, I try to keep the number of libraries to a minimum. Almost every
library I used so
far in
Hi,
I have initialized current_char in SVN 234, but can't see either how
this would
make a difference. I would be interested in the workspace file as well
if the error persists.
/// Jürgen
On 04/28/2014 06:04 AM, Elias Mårtenson wrote:
I've analysed the problem using Valgrind, and it seems
Hi,
thanks, fixed in SVN 235.
/// Jürgen
On 04/25/2014 05:40 AM, Elias Mårtenson wrote:
To reproduce, load the attached APL file that provides a definition
for the function io∆readfile. Then try to read the attached file
called "services":
* foo ← io∆readfile '/path/to/services'*
Thi
Hi Thomas,
a log of the input (without the responses from APL) is contained in the
.apl.history
file created by readline (length and position settable in the
preferences file).
For a mix of input and output you could use eg. xterm logging facility
(CTRL-left-mouse-button
and select "Log to f
Hi Peter,
the current line editor of GNU APL is in Nabla.cc. It is pretty simple.
The editor gets its line input from Input::get_user_line_nabla().
My concern with [a?b] is this: currently we print a prompt and then get
a user line
from either readline or from the operating system (in cooked m
Hi,
I have added a tiny new feature called "Probe" that allows more precise
benchmarks and simpler benchmark programs.
A probe is a small compile-time history of the CPU cycle counter.
Currently two probes are installed in SkalarFunction.cc to measure one
iteration of the inner loop of skalar fu
Hi Peter,
good question, I'm really happy to explain it.
GNU APL is almost completely Doxygen documented. Almost conmpletely
means that new functions
take a while before I update the Doxygen documentation (it makes little
sense to do that while functions
come and go).
Doxygen is a terrific t
.xml (if it is a different one) and the vid printed on stderr.
/// Jürgen
On 04/28/2014 05:11 PM, Blake McBride wrote:
Here is the workspace.
On Mon, Apr 28, 2014 at 10:06 AM, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi,
I have initialized current_c
Hi David,
I am glad that you found a workaround. ⎕CP would have been rather tricky
because
)COPY supports .xml and .apl files. The .xml files can be decoded
directly, but the .apl files
involve the APL interpreter to execute them (including APL commands)
which is somewhat
difficult to do (and
Hi,
I believe there is no reasonable way how a script that uses the ∇-editor
can know how a possibly already existing function looks like. In other
words, appending in a script is, in practice, always a mistake.
I have therefore changed the ∇-editor so that it always clears a function
if execute
Hi Peter,
there are already a number of facilities in GNU APL that can be switched
on and off via ./configure, for example Assert(). See the macros in
configure.ac
that end in _WANTED to see how thjis works.
I guess what you are after is ∇-editor command [A⎕B] . I would rate the
lack of it
as
I'd be glad to do that.
Thanks.
Blake
On Tue, Apr 29, 2014 at 8:45 AM, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi,
I managed to reproduce this once with )LOAD Devices (and without
emacs), but after turning Archive related
logging on and of
know because
that would mean you've come across a bug.
Regards,
Elias
On 30 Apr 2014 20:04, "Juergen Sauermann"
mailto:juergen.sauerm...@t-online.de>> wrote:
Hi Blake,
if the problem is that stderr gets lost, then it might help if
you c
Hi,
Hmm - below is what I get. Elias, can you reproduce the fault easily?
I guess we can rule out the archive (]log 39) then.
I see that there is a ⎕FX so I would do ]log 32 and ]log 33 next.
/// Jürgen
eedjsa@server64:~/projects/juergen/apl-1.3/src$ ./apl
__ _ __ _
Hi,
maybe some ⎕SYL limit could work. Currently we have such limits on the
depth of the SI,
on the number of values, and on the number of ravel bytes. This was to
limit infinite recursion of
user-defined functions. When such a limit is reached then an ATTENTION
is thrown and you can decide
to
to be implemented?
Regards,
Elias
On 30 April 2014 23:29, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi,
maybe some ⎕SYL limit could work. Currently we have such limits on
the depth of the SI,
on the number of values, and on the number of ravel bytes. This
Hi Peter,
sorry as well. I believe i was mislead by your earlier question for the
source of the Nabla editor.
I cannot give you a full answer because the question touches many aspects
of GNU APL, like how to do things, who does and maintain things, what do
we like,
and so on.
But I can share
Hi Peter,
there is very little emacs related code in the GNU APL core. The --emacs
command line option
replaces the escape sequences normally used for colors by sequences
that are easier to detect
by emacs so that emacs (which intercepts the GNU APL output) knows if a
line printed by the
int
Hi David,
from what I hear Dyalog APL seems to be a good interpreter and I have no
problem with it.
I am only a little more conservative when it comes to new and
non-standard APL features.
But Peter Teeson had ideas going into a similar direction.
One question that I have is how the implement
ed libraries I mentioned?
Is there something those shared library authors need to change?
Thanks!!
Blake
On Sat, May 3, 2014 at 12:33 PM, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi,
changed in SVN 243. I also added the caller argument in
close_fu
Hi Akiva,
I have removed the extra debug printout, see SVN 245.
Monadic ? is undefined as indicated by 'Error' in your table;
I believe the last element of X is ¯1?X and not ?X.
/// Jürgen
On 05/04/2014 03:20 PM, Akiva Avraham wrote:
Thought I'd point out an inconsistency.
http://i.imgur.co
more, would you like them to be
forwarded?
" ?B does not work. I think it should grab the last element."
That was a mistake on my part; should have said,
Shouldn't monadic ?B drop the first element?
--
On 14-05-04 07:05 AM, Juergen Sauermann wrote:
Hi Akiva,
I have re
Hi Blake,
thanks. fixed in SVN 249.
/// Jürgen
On 05/05/2014 02:45 AM, Blake McBride wrote:
)clear
CLEAR WS
∇test
[1] [∆2]
execute_oper() failed at No/bad edit_to at Nabla.cc:621
[1]
Hi Blake,
thanks. fixed in SVN 249.
/// Jürgen
On 05/05/2014 03:14 AM, Blake McBride wrote:
I know this code doesn't make sense but it is part of a function that
was crashing APL. I narrowed it down to the following:
⍎'→0⍴0'
-- Stack trace at
mands "missing"
in GNU APL.
See ]HELP or 'info apl'. SVN 250.
/// Jürgen
On 05/03/2014 08:08 PM, David B. Lamkins wrote:
On Sat, 2014-05-03 at 15:02 +0200, Juergen Sauermann wrote:
Hi David,
from what I hear Dyalog APL seems to be a good interpreter and I have no
problem with it
the list of ]-commands
(for tab-expansion). Clearly this will not be enough if there is a
facility to add new ones at runtime.
What method should I use to get a full list of these commands from
within a native function?
Regards,
Elias
On 6 May 2014 23:07, Juergen Sauermann <mailto:juerg
Hi David.
thanks. Removing first and printing then was maybe not so smart. SVN 251.
/// Jürgen
On 05/06/2014 06:47 PM, David B. Lamkins wrote:
Thank you, Jürgen! That'll do what I need.
FYI, I'm seeing a bit of cruft on the REMOVE:
]usercmd remove ]foo
User-defined command \371\
definitions probably shouldn't survive )load
and )clear.
On May 6, 2014 10:57 AM, "Juergen Sauermann"
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi David.
thanks. Removing first and printing then was maybe not so smart.
SVN 251.
/// Jürgen
On 05/06
Hi Peter,
thanks, should be fixed in SVN 250.
/// Jürgen
On 05/03/2014 07:41 PM, Peter Teeson wrote:
Hi Jürgen:
..GNUAPL/apl-svn/src/Logging.cc:32:11: Enumeration values 'LID_NONE' and
'LID_MAX' not handled in switch
..GNUAPL/apl-svn/src/Logging.cc:42:11: Enumeration values 'LID_NONE' and
'
nd libintl.
Following the same steps as above the terminal log shows
…
checking whether NLS is requested… yes
…
checking for GNU gettext in libintl... no <<<<=
checking whether to use NLS… no <<<===
…
So no problem because no NLS
On 2014-05-04, at 7:13 AM, Juergen Sauerma
Hi,
could you please print the value of start_input just before line 223 in
Input.cc like:
*I**nput::get_user_line(const UCS_string * prompt)
{
Q(start_input)
if (start_input) (*start_input)();
...*
On my machine it looks OK (even if I don't set it):
* This program is free softwar
Hi David,
thanks, fixed in SVN 252.
/// Jürgen
On 05/07/2014 08:53 PM, David B. Lamkins wrote:
'lib_file_io' ⎕fx 'pkg⍙fileio'
pkg⍙fileio
)erase pkg⍙fileio
'lib_file_io' ⎕fx 'pkg⍙fileio'
pkg⍙fileio
pkg⍙fileio⍬
VALUE ERROR
pkg⍙fileio⍬
^
ented?
Regards,
Elias
On 30 April 2014 23:29, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi,
maybe some ⎕SYL limit could work. Currently we have such limits on
the depth of the SI,
on the number of values, and on the number of ravel bytes. This
s separator for
internal functions.
Regards,
Elias
On 8 May 2014 20:56, Juergen Sauermann <mailto:juergen.sauerm...@t-online.de>> wrote:
Hi David,
thanks, fixed in SVN 252.
/// Jürgen
On 05/07/2014 08:53 PM, David B. Lamkins wrote:
'
Hi,
after having more and more problems with it, I decided
to remove gettext() and ⎕NLT from GNU APL. SVN 257.
/// Jürgen
Hi,
GNU APL is following the IBM APL2 language reference manual as much as
possible (so that we can
use it as GNU APL reference and need no write our own one).
It says:
"For example:
)WSID
THISWS
)SAVE
1992-03-27 21 .51.09 (GMT-4) THISWS
Note: The system response includes the wor
Hi,
I fixed the printouts of )COPY, )SAVE and )ERASE. See SVN 258.
/// Jürgen
On 05/09/2014 06:15 AM, Blake McBride wrote:
I'm not sure of this but - if you try to )ERASE something that isn't
there, shouldn't it display a not found message on the ones not found?
The problem is, if you erase
rhaps the time they are showing is local and the "(GMT-4)" is
showing its relationship to GMT.
Blake
On Fri, May 9, 2014 at 11:03 AM, Juergen Sauermann
<mailto:juergen.sauerm...@t-online.de>> wrote:
Hi,
GNU APL is following the IBM APL2 language reference manual
.
Regards,
Elias
On 8 May 2014 18:51, Juergen Sauermann <mailto:juergen.sauerm...@t-online.de>> wrote:
Hi,
could you please print the value of start_input just before line
223 in Input.cc like:
*I**nput::get_user_line(const UCS_string * prompt)
{
Q(sta
nd of trouble.
Regards,
Elias
On 10 May 2014 21:02, Juergen Sauermann <mailto:juergen.sauerm...@t-online.de>> wrote:
Hi Elias,
I believe this relates to an earlier discussion about removing of
callbacks.
Currently a library can decide how it wants to be handled
Hi Elias,
I see. So when would you like to load the shared lib? We could do it by
)COMMAND or by a --command-line-option.
Sounds like you want it rather early, even before APL's immediate
execution loop starts. That would be a
--command-line-option then. We could keep the rest (function names i
Hi Blake,
thanks, fixed in SVN 259. I had removed ⎕NLT together with gettext() but it
can still be around in workspace files.
/// Jürgen
On 05/10/2014 11:53 PM, Blake McBride wrote:
This was fine prior to my SVN update and rebuild. Now I get:
)load Devices
===
Hi Elias,
thanks, already interesting. If you could loop around the core count:
*for ((i=1; $i<=80; ++i)); do**
** ./Parallel $i**
** ./Parallel_OMP $i**
**done*
then I could understand the data better. Also not sure if something
is wrong with the benchmark program. On my new 4-core with OMP I
Hi Blake,
the asterisk marks an own SI entry for immediate execution, so
it does not belong to fun and should be on a separate line, see below.
/// Jürgen
]SI
Depth:3
Exec: 0x98373a8
Pmode:∇ fun[1]
PC: 1 ENDL
Stat: fun2
err_code: 0x30001
thrown: at Symbol.cc:679
e
Hi Blake,
I have changed the cope so that ⎕SVE is not )COPYd.
However, this may only hide some other fault. I was not able to
reproduce the fault on my machine because I did not have the workspace
file gg.xml.
If you could send me that file then I can have a deeper look.
Thanks,
/// Jürgen
hi David,
thank you very much for your feedback. See some answers inline below...
/// Jürgen
On 05/10/2014 08:40 PM, David B. Lamkins wrote:
First, my thanks to Jürgen for all the time and effort that went into
this proposal.
I found some of the specification regarding components to be mildl
Hi Blake,
if I remember correctly then I was equally surprised when I saw it first.
But for me as an APL designer, the IBM reference manual was by far the
most valuable, so I took it as the authoritative reference for GNU APL.
/// Jürgen
On 05/12/2014 04:13 PM, Blake McBride wrote:
Your repres
Hi Blake,
maybe what you are after is ⊃¨ instead of ⊃:
⊃'333' '5'
333
5
⊃¨'333' '5'
333 5
I guess they thought 'why do something that already exists by other
means (ie. ⊃¨) and do something different (ie. ⊃)
that could be useful elsewhere'.
/// Jürgen
On 05/12
Hi Blake,
thanks, found the fault. Fixed in SVN 261.
/// Jürgen
On 05/12/2014 03:57 PM, Blake McBride wrote:
Here is is. Thanks.
On Mon, May 12, 2014 at 6:44 AM, Juergen Sauermann
mailto:juergen.sauerm...@t-online.de>>
wrote:
Hi Blake,
I have changed the cope so that ⎕
Hi,
looks like a bug, I will look into it.
/// Jürgen
On 05/12/2014 05:28 PM, Jay Foad wrote:
⍬≡⊃⊂⍬
0
That looks like a bug in GNU APL. It's true in IBM APL2 and Dyalog APL.
Jay.
Hi,
looks like a bug, I will look into it.
/// Jürgen
On 05/12/2014 07:35 PM, David B. Lamkins wrote:
I'm forwarding this post to the list under a new topic so it doesn't get
buried in the other discussion:
Hi Blake,
I am really sorry to hear that. I am also coming from the old APL times
where empty values were just that - empty.
It took a while to grasp how APL2 handles empty values and as we have
seen the implementation of empty values
in GNU APL is not always correct. So I hope that your frus
Hi Jay,
thanks, fixed in SVN 262.
/// Jürgen
On 05/12/2014 06:26 PM, Juergen Sauermann wrote:
Hi,
looks like a bug, I will look into it.
/// Jürgen
On 05/12/2014 05:28 PM, Jay Foad wrote:
⍬≡⊃⊂⍬
0
That looks like a bug in GNU APL. It's true in IBM APL2 and Dyalog APL.
Jay.
Hi Elias,
thanks, fixed in SVN 263.
/// Jürgen
On 05/12/2014 05:23 PM, Elias Mårtenson wrote:
I noticed that my error messages were not coloured anymore, and I
discovered that the special control sequences were not sent anymore.
The solution is to set use_curses to false where the special s
Hi Akiva,
could not reproduce it with ^C but with ^D (which normally means
end-of-input).
Instead of throwing a segfault we will no )OFF after the 5'th ^D (to
avoid )OFF when
hitting ^D accidentally. SVN 264.
Thanks,
/// Jürgen
On 05/11/2014 03:29 AM, Akiva Avraham wrote:
I don't know if yo
bols lifecycle.
I don't really see any need to even have the ability to unload an
extension, so if that's complicated to implement it can be ignored.
Regards,
Elias
On 11 May 2014 00:33, Juergen Sauermann <mailto:juergen.sauerm...@t-online.de>> wrote:
Hi Elias,
I s
Hi Chris,
thank you!
/// Jürgen
On 05/03/2014 07:15 AM, Chris Jones wrote:
I finally wrote a GNU/apl XCompose file that can provide a quick, no
headaches :) alternative to more orthodox methods such as xmodmap or
setxkbmap + a third-level modifier.
This hack may come in handy for new users o
could find the code already in their caches. So the < 1
cycles cases
shown for OMP are most likely coming from these cases.
I will update the benchmark to do some real work - no point to repeat the
measurements in a loop before that.
/// Jürgen
On 05/11/2014 05:02 PM, Juergen Sauerm
Hi,
I thought I did in SVN 262?
/// Jürgen
On 05/13/2014 03:16 PM, Elias Mårtenson wrote:
Yes, that's the bug that was mentioned earlier. The result should be
the same as ''. Jürgen confirmed this bug, so I'd expect it to be
fixed soon. :-)
Regards,
Elias
On 13 May 2014 21:14, Blake McBr
Hi Blake,
thanks, fixed in SVN 265.
/// Jürgen
On 05/13/2014 05:07 PM, Blake McBride wrote:
In GNU APL, when you list a function via:
∇fun[⎕]∇
the format of the returned output has several problems
(inconsistencies with respect to the IBM documentation). Some of the
differences are small
Hi David,
thanks, fixed in SVN 266.
/// Jürgen
On 05/14/2014 01:56 AM, David B. Lamkins wrote:
SVN Revision: 264
{⍵}¨''
SEGMENTATION FAULT
-- Stack trace at main.cc:122
-
Hi Jay,
thanks, fixed in SVN 267.
/// Jürgen
On 05/14/2014 05:44 PM, Jay Foad wrote:
That's because of a bug in GNU APL:
x←(1 2)(3 4)
(a b)←x
a≡1 2
0
:-(
Jay.
On 14 May 2014 15:24, Blake McBride wrote:
Your unbox doesn't work. The following does:
(s r)←⊃x ⋄ z←(⊃s)
Hi,
thanks, hopefully fixed in SVN 268. Also the debug printouts.
/// Jürgen
On 05/14/2014 07:01 PM, David B. Lamkins wrote:
I think Elias is correct.
>From the IBM manual, pg. 110 (check the source; I think I've transcribed
accurately, but make no guarantee):
-- begin
Empty Argument: If R
Hi Elias,
thanks, should work now.
/// Jürgen
On 05/14/2014 09:19 AM, Elias Mårtenson wrote:
To reproduce, load a native function:
* 'libemacs' ⎕FX 'EE'*
Then type *]OWNERS* and the interpreter will crash with the following
dump:
Value #0: Flags = COMPLETE, ⍴⊏21⊐ ≡1:
SystemVariabl
Hi Blake,
thanks. should work now.
/// Jürgen
On 05/15/2014 03:37 PM, Blake McBride wrote:
Greetings,
GNU software index points to:
http://www.gnu.org/software/apl
That page is blank.
---
The page at: http://www.gnu.org/software/apl/apl.html
talks about downloading but doesn
Hi,
first of all, the search path for )LOAD, )SAVE, and friends is
configurable in the GNU APL preferences file
as LIBREF-0 ... 9.
Shared libraries are installed in an own directory which is usually
/usr/lib/apl or /usr/local/lib/apl.
The exact location is determined by (and can also be chang
Hi David,
not sure if it contradicts.
(1) First of all the ISO standard says this about *A f¨ B* and *f¨ B:*
*9.2.6 Each*
...
For both forms:
If B is not empty, Signal domain-error.
...
I believe this is a typo and they mean "if B is empty". Thus according
to the ISO standard +¨'' is undef
Hi Elias,
I have changed the code so that libemacs is loaded automatically
when GNU APL is started with --emacs. SVN 269.
I believe that other libraries shall unload themselves on )CLEAR,
at least that is what I would expect.
/// Jürgen
201 - 300 of 1721 matches
Mail list logo
