* Eli Zaretskii (2006-04-08) writes: >> From: Ralf Angeli <[EMAIL PROTECTED]> >> Date: Sat, 08 Apr 2006 21:23:21 +0200 >> >> When starting Emacs with >> >> emacs -Q -f toggle-debug-on-error test.c >> >> the following error occurs: >> >> Debugger entered--Lisp error: (void-variable c-subword-mode) >> >> Ater quitting the debugger with `q' the error can be provoked as well >> by clicking with the mouse on a menu. (That's in fact the way I >> noticed this at all.) > > I cannot reproduce this with the current CVS. Neither of your two > recipes causes an error.
Yesterday I had the chance to update the repository for building on Windows and compiled a new version. Trying the recipe above with this build does not provoke the error. So this seems to be a problem only affecting the build on GNU/Linux (maybe even specific to one configured with GTK). > Is there something special about test.c? is it an existing file? (I > tried with both non-existing file and one of the Emacs source files, > but perhaps you have something in test.c that triggers this problem.) "test.c" is supposed to be a non-existing file. Unrelated, but annoying: With the new build on Windows AUCTeX cannot be configured. I got the following error in config.log: configure:1539: checking if c:/Programme/Emacs/bin/emacs is XEmacs c:/Programme/Emacs/bin/emacs -batch -no-site-file -eval (let* ((x (if (featurep (quote xemacs)) "yes" "no"))) (write-region (if (stringp x) x (prin1-to-string x)) nil "./conftest-140")) Wrong type argument: arrayp, nil => configure:1557: error: Unable to run c:/Programme/Emacs/bin/emacs! Aborting! Trying the above (properly quoted) in an MSYS shell succeeds. So I am a bit stumped. If anybody has an idea what the cause could be, I am all ears. (The carriage return after nil looks suspicious.) The build from beginning of March did not have this problem. -- Ralf _______________________________________________ emacs-pretest-bug mailing list [email protected] http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
