Work at SourceForge, help us to make it a better place! We have an immediate need for a Support Technician in our San Francisco or Denver office.

Close

Site Xml Log


Commit Date  
[9150b6] (version-590rc2) by lbutler lbutler

This commit was manufactured by cvs2svn to create tag 'version-590rc2'.

2002-10-28 21:28:51 Tree
[2a925a] by amundson amundson

ChangeLog updated by cvs2cl.pl.

2002-10-28 21:28:50 Tree
[04a086] by amundson amundson

Last minute packaging problem discovered and fixed: Elliptic.texi was
not included in the distribution tarball. Now it is.

2002-10-28 18:22:54 Tree
[a96d69] by amundson amundson

Increment version to 5.9.0rc2.

2002-10-28 02:53:49 Tree
[5d4a45] by amundson amundson

Fix courtesy Sam Steingold:

* init-cl.lisp (user::run): use `with-simple-restart' for
CLISP, just like for CMUCL.
(sys::commands) [CLISP]: do not redefine the internal function

2002-10-28 02:50:41 Tree
[c1fff2] by amundson amundson

Fix courtesy Sam Steingold:

(step7): call `hyp-algv' with a correct number of arguments

2002-10-28 02:47:07 Tree
[0b6e90] by amundson amundson

Make double precision numbers default under clisp. This change was
suggested by Sam Steingold.

Precision in Maxima needs revisiting.

2002-10-28 02:43:00 Tree
[81ddc3] by amundson amundson

Make sure binary-cmucl directories exist before invoking cmucl. This
change is necessary since maxima.system started calling truename for
binary-cmucl.

2002-10-28 02:41:12 Tree
[11a2c6] by amundson amundson

Update package handling for gcl.

2002-10-28 02:39:51 Tree
[e834ae] by amundson amundson

Add contents of win32 directory to distribution tarball.

2002-10-28 02:33:13 Tree
[0530b9] by amundson amundson

Fixes to restore the functionality of xmaxima-local, at least under
Unix.

2002-10-28 02:31:02 Tree
[057222] by amundson amundson

Small update on GCL versions.

2002-10-28 02:29:00 Tree
[e96d3a] by wjenkner wjenkner

ITERATE-OVER-BKPTS: Deleting disabled breakpoints didn't work and
enabling already enabled breakpoints effectively deleted them.

2002-10-27 02:20:59 Tree
[0dc281] by amundson amundson

Very important bug fix. The old default version of pid was -1. This
was explosively dangerous -- if the maxima subprocess failed to start,
xmaxima will call "kill -1", which results in killing *all* user
process on (at least) Linux.

The new default version for pid is "none".

2002-10-25 17:03:29 Tree
[85f2c9] by billingd billingd

Fix typo in disabled test, then enable it.

2002-10-24 00:03:30 Tree
[b10d8f] by billingd billingd

Add tests for ode2() with method:xformtoconstcoeff

2002-10-23 23:57:16 Tree
[d5a357] by rtoy rtoy

Recent versions of CMUCL need truename to get the right location of
binary-cmucl. Don't know if this is a bug in CMUCL or defsystem.

2002-10-17 22:32:25 Tree
[6a2ab5] by rtoy rtoy

Fix typo in loop expression.

2002-10-17 21:32:43 Tree
[d2a3a5] by amundson amundson

Restore large clisp banner at startup.

2002-10-17 21:08:38 Tree
[33ec71] by belanger belanger

A "$" was added to a skip-chars-backwards command, to deal with
commands ending in "$".

2002-10-16 15:39:22 Tree
[9ae2eb] by rtoy rtoy

Bug 623620: -log(-1.0b0) unnormalized

Fixed as suggested by Stavros Macrakis (with minor editing).

2002-10-15 17:34:27 Tree
[f373df] by rtoy rtoy

Bug 623384: acoth(-2.0) was computing the wrong thing because the
formula was wrong. Fix from Stavros Macrakis.

2002-10-15 17:11:05 Tree
[5655c8] by rtoy rtoy

Rev 1.4 incorrectly lifted the multiplication by 2 outside the loop.
Make it right.

2002-10-14 16:44:36 Tree
[a69ff5] by amundson amundson

This change fixes the "socket bug." Error conditions often result in a
(throw 'maxima-quit) which results in macsyma-top-level being
restarted. The way the batch mode worked, restarting macsyma-top-level
caused the input to revert to the batch string (used to start the
server) which had now reached its end. Since there was an eof on the
input stream, maxima would simply exit. Now, if we are using a socket
connection and if the input is not standard input, we switch the input
to standard input. (The socket connection points standard input to the
socket.)

The very fact that this commit message is terribly confusing means
that the entire input-stream/socket/batch-mode scheme needs to be
rethought.

2002-10-10 14:06:32 Tree
[68a661] by rtoy rtoy

o Fix bug in fpshift wherein right shifts of negative numbers were not
truncated properly
o Add some comments on the algorithm used by fplog.
o Move the multiplication by 2 that was inside the loop to outside the
loop.

2002-10-08 17:06:25 Tree
Older >