>>>>> "Romain" == Romain Francoise <[EMAIL PROTECTED]> writes:
Romain> "Andrew M. Scott" <[EMAIL PROTECTED]> writes:
>> (locate-library "cc-mode.el") erroneously returns nil, not
>> noticing that the cc-mode.el.gz exists.
Romain> Is that a problem in practice?
Yes, to the extent that the syntax (locate-library "cc-mode.el")
*used* to work up to the point that the *.el's were compressed. Here
are two cases I can think of:
1. My fingers sometimes add the .el extension without my asking. I
often load .el's, as they are easier to debug. That's what prompted
this original bug-report.
2. If someone has legacy ~/.emacs code like the following
example, but without the optional 2nd arg of nil, a move from
Emacs-21.3 -> Emacs-22 would unexpected not load/execute their code
fragment:
(when (locate-library "specman-mode" nil)
(progn
(autoload 'specman-mode "specman-mode" "Specman code editing mode" t)
(add-to-list 'auto-mode-alist '("\\.e\\'" . specman-mode))
(add-to-list 'auto-mode-alist '("\\.e3\\'" . specman-mode))
(add-to-list 'auto-mode-alist '("\\.load\\'" . specman-mode))
(add-to-list 'auto-mode-alist '("\\.ecom\\'" . specman-mode))
(add-to-list 'auto-mode-alist '("\\.etst\\'" . specman-mode))
))
Thanks,
Andy Scott
_______________________________________________
Emacs-pretest-bug mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
