Re: grep cannot find relative file names after save and re-open
emacs -q C-x C-f $YOUR_EMACE_SOURCE_TREE/lisp/ChangeLog RET M-x grep RET defface */*.el RET C-x C-w /tmp/TEST RET C-x k TEST RET now open /tmp/TEST C-x C-f /tmp/TEST RET now move the cursor to one of the calendar/calendar.el lines and press RET at that point you are prompted to enter the file name where the file calendar.el is found. The first line of the grep buffer (as well as the saved grep output) contains the variable `default-directory', e.g.: -*- mode: grep; default-directory: /usr/share/emacs/lisp/ -*- `grep-mode' uses the value of this variable to get the right absolute file name from the relative name on compilation or grep messages, and to store the absolute file name to the `message' text property. However, at the moment when it parses the grep output, the variable `default-directory' from the first line has no effect yet, because `grep-mode' from the first line is called before setting the value of `default-directory'. I see three solutions: 1. Set local variables specified in the -*- line before setting the mode from -*- mode tag. This might have an undesirable effect on existing files where reversing the order of processing -*- line will lead to unpredictable consequences. 2. Set only `default-directory' specified in the -*- line before setting the mode, and to set the remaining variables later. 3. Change only compile.el to search for `default-directory' in the first line and to set its value exlicitly before staring to parse the compilation/grep output. -- Juri Linkov http://www.jurta.org/emacs/ ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
build should require correct version of makeinfo
Hi, On my machine, CVS Emacs doesn't build if I have makeinfo 4.1 in my $PATH, but does if I have makeinfo 4.2. In the man/ subdirectory, I get: dired-x.texi:27: Unknown command `copying'. dired-x.texi:60: Unmatched [EMAIL PROTECTED]'. The configure should have warned me with a clear error message (saying version = 4.2 is required) or give a warning and disable the makeinfo target. -- Matthieu ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: IDO and TRAMP don't work well together
Matthieu Moy [EMAIL PROTECTED] writes: I got in a lot of bugs when using TRAMP + IDO, when I have to specify the user name. I guess the bugs are in IDO, not TRAMP (I'm CC-ing the author of IDO). For example: $ emacs -q M-x ido-mode RET C-x C-f /ssh:moy@ When I press @, Emacs tries to open a connection (tramp: Opening connection at ssh using nil...) and ends with an error message: tramp: Opening connection at ssh using nil It should have waited for the hostname. Sigh... The interaction between ido and tramp is very hard to get right. What ido tries to do is to ask tramp (indirectly via a call to file-name-all-completions) to provide a list of possible completions for each step of entering the tramp prefix. The problem is that tramp needs to get the partial string in certain formats for it to do the right thing, and I belive I got it working quite some time (years) ago, but some later changes to tramp must have broken it. I'll see what I can do about it... In the meantime, you can set ido-enable-tramp-completion to nil. Second bug: If I just do C-x C-f /moy@ then Emacs inserts a slash right after the @. I can't enter the hostname. This happens to me too. Thanks for noticing it. -- Kim F. Storm [EMAIL PROTECTED] http://www.cua.dk ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: ido shows wrong completions
Kim F. Storm said: Matthieu Moy [EMAIL PROTECTED] writes: I have a ~/local/ and a ~/local-mirror/ directory in my $HOME. With ido enabled, when I do C-x C-f ~/local/ TAB TAB, the list of completion is the list of files in my $HOME, with the 7th character highlighted, whereas it should list $HOME/local/. Actually, this is a feature rather than a bug. [...] When you have tabbed through all completions, the next TAB will go back to the first page, but this time showing you ALL the elements in the current directory. Since I've already typed a slash (the minibuffer shows ~/local/), I expect the current directory to be ~/local/, not ~/. What's even more confusing is that Emacs highlights the next caracter to type in the completion buffer, and in your case, the highlighted character is *not* the one to type. If you have some better idea how to handle this, please tell me. My advice would be to take the trailing slash into account. For example, make / equivalent to what is currently / RET. -- Matthieu ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: ido shows wrong completions
Matthieu MOY [EMAIL PROTECTED] writes: Since I've already typed a slash (the minibuffer shows ~/local/), I expect the current directory to be ~/local/, not ~/. That is confusing, yes, but until you type RET, you don't switch directory. Consdier this: if you just type ~/cal/ rather than ~/local/ you will see the same possible completions, but it is now clear that you are not yet switched into local/. One possible enhancement would be to fontify the trailing / differently until you actually hit RET to enter the directory... What's even more confusing is that Emacs highlights the next caracter to type in the completion buffer, and in your case, the highlighted character is *not* the one to type. True -- it is probably better not to highlight anything in that case... If you have some better idea how to handle this, please tell me. My advice would be to take the trailing slash into account. For example, make / equivalent to what is currently / RET. Ido does that already if there are no other completions, see ido-enter-single-matching-directory. -- Kim F. Storm [EMAIL PROTECTED] http://www.cua.dk ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: mwheel-scroll = backtrace
This bug report will be sent to the Free Software Foundation, not to your local site managers! Please write in English if possible, because the Emacs maintainers usually do not have translators to read other languages for them. Your bug report will be posted to the emacs-pretest-bug@gnu.org mailing list. Please describe exactly what actions triggered the bug and the precise symptoms of the bug: backtrace, mods in the following code is nil when scrolling, so assoc fails. (defun mwheel-scroll (event) ... (mods (delq 'click (delq 'double (delq 'triple (event-modifiers event) (amt (assoc mods mouse-wheel-scroll-amount))) ;; Extract the actual amount or find the element that has no modifiers. In GNU Emacs 22.0.50.1 (i386-msvc-nt5.0.2195) of 2005-07-06 on SCARLATTI X server distributor `Microsoft Corp.', version 5.0.2195 configured using `configure --with-msvc (12.00)' Important settings: value of $LC_ALL: nil value of $LC_COLLATE: nil value of $LC_CTYPE: nil value of $LC_MESSAGES: nil value of $LC_MONETARY: nil value of $LC_NUMERIC: nil value of $LC_TIME: nil value of $LANG: DEU locale-coding-system: cp1252 default-enable-multibyte-characters: t Major mode: Makefile Minor modes in effect: partial-completion-mode: t erc-track-mode: t minibuffer-electric-default-mode: t which-function-mode: t auto-image-file-mode: t pc-selection-mode: t recentf-mode: t show-paren-mode: t encoded-kbd-mode: t tooltip-mode: t auto-compression-mode: t global-font-lock-mode: t font-lock-mode: t unify-8859-on-encoding-mode: t utf-translate-cjk-mode: t line-number-mode: t transient-mark-mode: t next-error-follow-minor-mode: Fol abbrev-mode: t Recent input: left C-x m c c c SPC down m w h e e l . e l SPC = S-SPC C o m p a n y C-c C-c q return C-backspace C-c C-g C-c C-c down down return return q up up q C-c C-g C-c C-c q C-backspace C-c C-g C-c C-c q return C-M-SPC C-insert left C-home M-delete M-delete C-x b _ e m return C-home C-s m w h e e l C-s C-s C-s C-s left home C-M-up M-x r e p o p r t tab backspace backspace backspace r t tab return Recent messages: Checking new news... nnml: Reading incoming mail from pop... nnml: Reading incoming mail (no new mail)...done Opening nntp server on News.CIS.DFN.DE...done Checking new news...done Generating the cache active file...done (No changes need to be saved) [2 times] Mark set Mark saved where search started Loading emacsbug...done ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: ido shows wrong completions
Kim F. Storm said: My advice would be to take the trailing slash into account. For example, make / equivalent to what is currently / RET. Ido does that already if there are no other completions, see ido-enter-single-matching-directory. How about this? --- ido.el.orig 2005-08-16 15:44:53.0 +0200 +++ ido.el 2005-08-16 15:45:48.0 +0200 @@ -696,6 +696,14 @@ (other :tag Always t)) :group 'ido) +(defcustom ido-enter-matching-directory-anyway t + *Automatically enter sub-directory even if it is not the only match. + +If non-nil, typing \/\ enters the directory corresponding to the +first match, even if it is not the only match. + :type 'boolean + :group 'ido) + (defcustom ido-create-new-buffer 'prompt *Specify whether a new buffer is created if no buffer matches substring. Choices are 'always to create new buffers unconditionally, 'prompt to @@ -3976,7 +3984,8 @@ (when (and ido-enter-single-matching-directory ido-matches - (null (cdr ido-matches)) + (or ido-enter-matching-directory-anyway + (null (cdr ido-matches))) (ido-final-slash (car ido-matches)) (or try-single-dir-match (eq ido-enter-single-matching-directory t))) (I leave the default value up to you ;-). The problem with this patch is that it inhibit one feature of ido: typing '/' to be able to cycle through directories. Well, *I* prefer the behavior of my patch, but I don't know about other users). -- Matthieu ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: mwheel-scroll = backtrace
Christoph Conrad wrote: Please describe exactly what actions triggered the bug and the precise symptoms of the bug: backtrace, mods in the following code is nil when scrolling, so assoc fails. (defun mwheel-scroll (event) ... (mods (delq 'click (delq 'double (delq 'triple (event-modifiers event) (amt (assoc mods mouse-wheel-scroll-amount))) ;; Extract the actual amount or find the element that has no modifiers. In GNU Emacs 22.0.50.1 (i386-msvc-nt5.0.2195) of 2005-07-06 on SCARLATTI Please describe exactly what actions triggered the bug, and the precise symptoms of the bug. Please also update to a more recent version of CVS, to make sure it hasn't already been fixed there. ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: mwheel-scroll = backtrace
Hello Jason, (defun mwheel-scroll (event) Please describe exactly what actions triggered the bug, and the precise symptoms of the bug. Please also update to a more recent version of CVS, to make sure it hasn't already been fixed there. Solved. I misinterpreted the doc-string of `mouse-wheel-scroll-amount', but it is clearly expressed there what is allowed. My fault - sorry. Best wishes, Christoph ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
GUD node names
Two GUD-related nodes in the Emacs manual have too general names: `Layout' and `Other Buffers'. Unlike other GUD subnode names, after looking at these nodes names you can't get an idea that they document only GUD aspects. Any objections to adding the word `GUD' to these names? Index: man/building.texi === RCS file: /cvsroot/emacs/emacs/man/building.texi,v retrieving revision 1.68 diff -c -r1.68 building.texi *** man/building.texi 14 Aug 2005 08:06:03 - 1.68 --- man/building.texi 16 Aug 2005 11:37:05 - *** *** 773,787 way, then you can use @kbd{M-x gdba} to invoke GDB in graphical mode. @menu ! * Layout:: Control the number of displayed buffers. * Breakpoints Buffer:: A breakpoint control panel. * Stack Buffer:: Select a frame from the call stack. * Watch Expressions::Monitor variable values in the speedbar. ! * Other Buffers::Input/output, locals, registers, assembler, threads and memory buffers. @end menu ! @node Layout @subsubsection Layout @cindex GDB User Interface layout --- 773,787 way, then you can use @kbd{M-x gdba} to invoke GDB in graphical mode. @menu ! * GUD Layout:: Control the number of displayed buffers. * Breakpoints Buffer:: A breakpoint control panel. * Stack Buffer:: Select a frame from the call stack. * Watch Expressions::Monitor variable values in the speedbar. ! * Other GUD Buffers::Input/output, locals, registers, assembler, threads and memory buffers. @end menu ! @node GUD Layout @subsubsection Layout @cindex GDB User Interface layout *** *** 917,923 Since this does not work for variables defined in compound statements, the default value is @code{nil}. ! @node Other Buffers @subsubsection Other Buffers @table @asis --- 917,923 Since this does not work for variables defined in compound statements, the default value is @code{nil}. ! @node Other GUD Buffers @subsubsection Other Buffers @table @asis -- Juri Linkov http://www.jurta.org/emacs/ ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Emacs Lisp reference: [EMAIL PROTECTED] vs body...
I noticed a small inconsistency in the Emacs Lisp Reference manual. Some functions have their argument `body' renamed to `forms' in the documentation, some not. Some use the Texinfo command [EMAIL PROTECTED]' to produce dots, some use `...' literally. Even very similar commands have different argument names in the manual. For example: @defmac while-no-input body... @defmac with-local-quit [EMAIL PROTECTED] @defspec save-excursion [EMAIL PROTECTED] @defspec save-restriction [EMAIL PROTECTED] Should these arguments be fixed in the manual to be more uniform? I guess [EMAIL PROTECTED]' would be more preferable. A list of `body' and `form' arguments to check: buffers.texi:205:@defspec save-current-buffer body... buffers.texi:218:@defmac with-current-buffer buffer-or-name body... buffers.texi:230:@defmac with-temp-buffer body... commands.texi:2437:@defmac while-no-input body... commands.texi:2661:@defmac with-local-quit [EMAIL PROTECTED] compile.texi:473:@defspec with-no-warnings body... control.texi:80:@defspec progn [EMAIL PROTECTED] control.texi:100:@defspec prog1 form1 [EMAIL PROTECTED] control.texi:124:@defspec prog2 form1 form2 [EMAIL PROTECTED] control.texi:151:@defspec if condition then-form [EMAIL PROTECTED] control.texi:176:@defmac when condition [EMAIL PROTECTED] control.texi:192:@defmac unless condition [EMAIL PROTECTED] control.texi:425:@defspec while condition [EMAIL PROTECTED] control.texi:477:@defmac dolist (var list [result]) [EMAIL PROTECTED] control.texi:493:@defmac dotimes (var count [result]) [EMAIL PROTECTED] control.texi:591:@defspec catch tag [EMAIL PROTECTED] control.texi:1175:@defspec unwind-protect body-form [EMAIL PROTECTED] display.texi:405:@defmac dotimes-with-progress-reporter (var count [result]) message body... display.texi:953:@defspec with-output-to-temp-buffer buffer-name [EMAIL PROTECTED] edebug.texi:847:@defmac edebug-tracing string [EMAIL PROTECTED] files.texi:628:@defmac with-temp-file file body... frames.texi:1322:@defspec track-mouse [EMAIL PROTECTED] intro.texi:415:@defspec count-loop (@var{var} [EMAIL PROTECTED] @var{to} [EMAIL PROTECTED]) @[EMAIL PROTECTED] macros.texi:209:@defspec defmacro name argument-list [EMAIL PROTECTED] modes.texi:978:@defmac define-derived-mode variant parent name docstring [EMAIL PROTECTED] [EMAIL PROTECTED] modes.texi:1157:@defmac delay-mode-hooks body... modes.texi:1367:@defmac define-minor-mode mode doc [init-value [lighter [keymap]]] keyword-args... body... os.texi:1443:@defmac with-timeout (seconds [EMAIL PROTECTED]) [EMAIL PROTECTED] positions.texi:832:@defspec save-excursion [EMAIL PROTECTED] positions.texi:947:@defspec save-restriction [EMAIL PROTECTED] searching.texi:1597:@defmac save-match-data [EMAIL PROTECTED] streams.texi:666:@defmac with-output-to-string body... syntax.texi:515:@defmac with-syntax-table @var{table} @var{body}... text.texi:4270:@defmac combine-after-change-calls body... variables.texi:190:@defspec let ([EMAIL PROTECTED]) [EMAIL PROTECTED] variables.texi:220:@defspec let* ([EMAIL PROTECTED]) [EMAIL PROTECTED] windows.texi:463:@defmac save-selected-window [EMAIL PROTECTED] windows.texi:478:@defmac with-selected-window window [EMAIL PROTECTED] windows.texi::@defspec save-window-excursion [EMAIL PROTECTED] -- Juri Linkov http://www.jurta.org/emacs/ ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: grep cannot find relative file names after save and re-open
C-x C-f /tmp/TEST RET now move the cursor to one of the calendar/calendar.el lines and press RET at that point you are prompted to enter the file name where the file calendar.el is found. That is no bug. Emacs can't find the files unless you tell it where to look. If you use M-x cd to specify the proper default dir, does that make it work? ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
GUD node names
Two GUD-related nodes in the Emacs manual have too general names: `Layout' and `Other Buffers'. Unlike other GUD subnode names, after looking at these nodes names you can't get an idea that they document only GUD aspects. Any objections to adding the word `GUD' to these names? They are not subnodes of GUD but of GDB Graphical Interface. All the other (GUD) nodes listed with GDB Graphical Interface apply to a range of debuggers (GDB, DBX, PDB, PERLDB etc). The nodes within GDB Graphical Interface such as Layout and Other Buffers only apply to GDB, so I think that this change would be wrong. The introduction explains that this mode uses different Emacs windows. I guess Layout could be changed to Window Layout but I think the other node names are self-explanatory within their context. Nick ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: grep cannot find relative file names after save and re-open
Richard M. Stallman [EMAIL PROTECTED] writes: C-x C-f /tmp/TEST RET now move the cursor to one of the calendar/calendar.el lines and press RET at that point you are prompted to enter the file name where the file calendar.el is found. That is no bug. Emacs can't find the files unless you tell it where to look. See also Juri's answer, Emacs has the directory information: -*- mode: grep; default-directory: ~/src/emacs/lisp/ -*- it seems that it's just not used in the right order. When `grep-mode' is turned on `default-directory' is not bound, so it does not parse the buffer correctly. For some reason all the directory information is stripped when parsing. Everything works just fine for grep results that are of the form: FILENAME:LINENUMBER: etc etc the problem occurs only when the results are of the form: DIRNAME/FILENAME:LINENUMBER... etc etc If you use M-x cd to specify the proper default dir, does that make it work? No, it does not. ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
pop-to-buffer with new frames is slow when custom faces have been installed
pop-to-buffer is very slow (1 sec, sometimes more!) when a color theme (color-theme package) is chosen that sets a couple of faces and when pop-up-frames it t. Test case: (require 'color-theme) (color-theme-arjen) ;; or any other theme (setq pop-up-frames t) (pop-to-buffer test nil nil) To directly reproduce, you will need to have the color-theme package available. But to give you an impression of what's installed by color- theme-arjen, its code follows below. When I define my own one without the faces, I don't get the above problem, so I'm inclined to think it's due to the faces defined. This can be consistently reproduced on a vanilla Carbon Emacs on Mac OS X. I traced the problem down to pop-to-buffer. I think tests on another system are warranted. Please do something about it - it's a real show-stopper to have to wait for seconds before a new frame opens. Thanks Dave === (defun color-theme-arjen () Color theme by awiersma, created 2001-08-27. (interactive) (color-theme-install '(color-theme-arjen ((background-color . black) (background-mode . dark) (border-color . black) (cursor-color . yellow) (foreground-color . White) (mouse-color . sienna1)) ((buffers-tab-face . buffers-tab) (cperl-here-face . font-lock-string-face) (cperl-invalid-face quote underline) (cperl-pod-face . font-lock-comment-face) (cperl-pod-head-face . font-lock-variable-name-face) (vc-mode-face . highlight)) (default ((t (:background black :foreground white (blue ((t (:foreground blue (bold ((t (:bold t (bold-italic ((t (:bold t (border-glyph ((t (nil (buffers-tab ((t (:background black :foreground white (calendar-today-face ((t (:underline t (cperl-array-face ((t (:foreground darkseagreen (cperl-hash-face ((t (:foreground darkseagreen (cperl-nonoverridable-face ((t (:foreground SkyBlue (custom-button-face ((t (nil (custom-changed-face ((t (:background blue :foreground white (custom-documentation-face ((t (nil (custom-face-tag-face ((t (:underline t (custom-group-tag-face ((t (:underline t :foreground light blue (custom-group-tag-face-1 ((t (:underline t :foreground pink (custom-invalid-face ((t (:background red :foreground yellow (custom-modified-face ((t (:background blue :foreground white (custom-rogue-face ((t (:background black :foreground pink (custom-saved-face ((t (:underline t (custom-set-face ((t (:background white :foreground blue (custom-state-face ((t (:foreground lime green (custom-variable-button-face ((t (:underline t :bold t (custom-variable-tag-face ((t (:underline t :foreground light blue (diary-face ((t (:foreground IndianRed (erc-action-face ((t (:bold t (erc-bold-face ((t (:bold t (erc-default-face ((t (nil (erc-direct-msg-face ((t (:foreground sandybrown (erc-error-face ((t (:bold t :foreground IndianRed (erc-input-face ((t (:foreground Beige (erc-inverse-face ((t (:background wheat :foreground darkslategrey (erc-notice-face ((t (:foreground MediumAquamarine (erc-pal-face ((t (:foreground pale green (erc-prompt-face ((t (:foreground MediumAquamarine (erc-underline-face ((t (:underline t (eshell-ls-archive-face ((t (:bold t :foreground IndianRed (eshell-ls-backup-face ((t (:foreground Grey (eshell-ls-clutter-face ((t (:foreground DimGray (eshell-ls-directory-face ((t (:bold t :foreground MediumSlateBlue (eshell-ls-executable-face ((t (:foreground Coral (eshell-ls-missing-face ((t (:foreground black (eshell-ls-picture-face ((t (:foreground Violet (eshell-ls-product-face ((t (:foreground sandybrown (eshell-ls-readonly-face ((t (:foreground Aquamarine (eshell-ls-special-face ((t (:foreground Gold (eshell-ls-symlink-face ((t (:foreground White (eshell-ls-unreadable-face ((t (:foreground DimGray (eshell-prompt-face ((t (:foreground MediumAquamarine (fl-comment-face ((t (:foreground pink (fl-doc-string-face ((t (:foreground purple (fl-function-name-face ((t (:foreground red (fl-keyword-face ((t (:foreground cadetblue (fl-string-face ((t (:foreground green (fl-type-face ((t (:foreground yellow (font-lock-builtin-face ((t (:foreground LightSteelBlue (font-lock-comment-face ((t (:foreground IndianRed (font-lock-constant-face ((t (:foreground Aquamarine (font-lock-doc-string-face ((t (:foreground DarkOrange (font-lock-function-name-face ((t (:foreground YellowGreen (font-lock-keyword-face ((t (:foreground PaleYellow (font-lock-preprocessor-face ((t
local var evaluation order
Dan Nicolaescu [EMAIL PROTECTED] writes: -*- mode: grep; default-directory: ~/src/emacs/lisp/ -*- it seems that it's just not used in the right order. maybe this is related to another bug: w/ recent (this week) emacs from cvs, this small file: --8 - level 1 - level 2 ## Local Variables: ## mode: outline ## outline-regexp: \\([ ][ ]\\)*- ## End: ---8- does not fontify, whereas w/ older emacs versions it does. it also does not fontify after reversing the order of the `mode:' and `outline-regexp:' lines. if i set-default the value of `outline-regexp', however, it fontifies. (emacs -q --no-site-file and global font lock mode enabled.) thi ___ Emacs-pretest-bug mailing list Emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug