Index: slime/ChangeLog diff -u slime/ChangeLog:1.1776 slime/ChangeLog:1.1779 --- slime/ChangeLog:1.1776 Thu May 28 15:01:09 2009 +++ slime/ChangeLog Thu Jun 4 05:08:06 2009 @@ -1,3 +1,13 @@ +2009-06-04 Helmut Eller heller@common-lisp.net + + * slime.el (slime-dispatch-event [:emcas-rex]): Don't clutter + the main code-path with confusing error handling. + +2009-06-04 Helmut Eller heller@common-lisp.net + + * swank-openmcl.lisp (*known-processes*, mailbox): Use a weak + hashtable to plug the memory leak. + 2009-05-28 Tobias C. Rittweiler tcr@freebits.de
* slime.el (def-slime-test): Forgot to remove debugging code.
Helmut Eller heller@common-lisp.net writes:
Index: slime/ChangeLog diff -u slime/ChangeLog:1.1776 slime/ChangeLog:1.1779 --- slime/ChangeLog:1.1776 Thu May 28 15:01:09 2009 +++ slime/ChangeLog Thu Jun 4 05:08:06 2009 @@ -1,3 +1,13 @@ +2009-06-04 Helmut Eller heller@common-lisp.net
- slime.el (slime-dispatch-event [:emcas-rex]): Don't clutter
- the main code-path with confusing error handling.
What about Mark Cox' report (2009-05-26) that `slime-net-send' can signal an error in which case the rex-continuation which was pushed in :emacs-rex will not be removed again? Not removing it will make Slime think it's busy all time.
-T.
* Tobias C. Rittweiler [2009-06-11 10:05+0200] writes:
+2009-06-04 Helmut Eller heller@common-lisp.net
- slime.el (slime-dispatch-event [:emcas-rex]): Don't clutter
- the main code-path with confusing error handling.
What about Mark Cox' report (2009-05-26) that `slime-net-send' can signal an error in which case the rex-continuation which was pushed in :emacs-rex will not be removed again? Not removing it will make Slime think it's busy all time.
Note that the continuation gets now pushed after calling slime-send. If slime-send fails then nothing gets pushed. That solves the problem, right?
Helmut.
Helmut Eller heller@common-lisp.net writes:
- Tobias C. Rittweiler [2009-06-11 10:05+0200] writes:
+2009-06-04 Helmut Eller heller@common-lisp.net
- slime.el (slime-dispatch-event [:emcas-rex]): Don't clutter
- the main code-path with confusing error handling.
What about Mark Cox' report (2009-05-26) that `slime-net-send' can signal an error in which case the rex-continuation which was pushed in :emacs-rex will not be removed again? Not removing it will make Slime think it's busy all time.
Note that the continuation gets now pushed after calling slime-send. If slime-send fails then nothing gets pushed. That solves the problem, right?
Yes, I'm sorry, I missed that in the diff.
-T.
-
Helmut Eller -
Tobias C. Rittweiler