Re: image.el doesn't associate image-mode with .JPG files
Richard Stallman wrote: Perhaps this can be solved by first doing a case-sensitive scan through auto-mode-alist, then if that fails to find a match do a case-insensitive scan. Brilliant! It would handle that one case, but it would still produce false matches. It would only produce false matches for cases where Emacs would have defaulted to Fundamental mode. Is that so bad, or am I missing something? ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: image.el doesn't associate image-mode with .JPG files
Jason Rumney [EMAIL PROTECTED] writes: Richard Stallman wrote: Perhaps this can be solved by first doing a case-sensitive scan through auto-mode-alist, then if that fails to find a match do a case-insensitive scan. Brilliant! It would handle that one case, but it would still produce false matches. It would only produce false matches for cases where Emacs would have defaulted to Fundamental mode. Is that so bad, or am I missing something? I don't see the problem either. It would make Emacs DTRT in a lot more cases, at the expense of _potentially_ doing TWT in a few obscure cases, such as entering changelog-mode for a file named cHANGElOG. I'm 100% in favour of your proposal (and consequently in 100% disagreement with RMS -- on this issue). -- 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: Emacsclient/server filename quoting error
On 12/10/06, Francis Wright [EMAIL PROTECTED] wrote:
I execute all the following commands from a cmd command
prompt (outside of Emacs).
emacsclient -n -a runemacs TO DO.txt
works correctly if Emacs IS already running, but if it is not
already running then Emacs does not see the filename correctly;
the quotes do not appear to be passed on to runemacs.
The following patch addresses the issue by allocating quoted copies of
any argument containing spaces before calling execvp.
Any objections to install this fix? As it stands, it affects also
non-Windows builds. Is requoting args the right behavior in these
environments?
/L/e/k/t/u
Index: lib-src/emacsclient.c
===
RCS file: /cvsroot/emacs/emacs/lib-src/emacsclient.c,v
retrieving revision 1.98
diff -u -2 -r1.98 emacsclient.c
--- lib-src/emacsclient.c 30 Nov 2006 22:49:38 - 1.98
+++ lib-src/emacsclient.c 15 Dec 2006 10:19:44 -
@@ -310,8 +310,20 @@
if (alternate_editor)
{
- int i = optind - 1;
+ int j, i = optind - 1;
+
#ifdef WINDOWSNT
- argv[i] = (char *)alternate_editor;
+ argv[i] = (char *) alternate_editor;
#endif
+
+ /* Arguments with spaces have been dequoted, so we
+have to requote them before calling execvp. */
+ for (j = i; argv[j]; j++)
+ if (strchr (argv[j], ' '))
+ {
+ char *quoted = alloca (strlen (argv[j]) + 3);
+ sprintf (quoted, \%s\, argv[j]);
+ argv[j] = quoted;
+ }
+
execvp (alternate_editor, argv + i);
message (TRUE, %s: error executing alternate editor \%s\\n,
___
emacs-pretest-bug mailing list
emacs-pretest-bug@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: gnus queries all mail folders on startup
From: Michael Welsh Duggan md5i at cs.cmu.edu Newsgroups: gmane.emacs.devel Date: 2006-12-14 07:15:03 GMT (18 hours and 20 minutes ago) From: Chris Moore dooglus at gmail.com To: emacs-pretest-bug at gnu.org Date: Thu, 07 Dec 2006 16:18:08 +0100 I'm using nnimap in gnus to read my email. Each time I start gnus, it polls every one of my hundreds of mail folders, which takes quite a long time. I've set gnus-activate-level to 5, and unsubscribed from the folders I don't want it to scan. I've checked, and those folders now have a rank of 6, and so they shouldn't be scanned by default. Since nobody seems to have followed up on this... Chris, what is the value of `gnus-read-active-file'? According to the documentation, it should be `some' by default. If it is t, try setting it to some to see if that solves the problem. (Gnus node: The Active File) Hi Michael. I'm amazed I saw your message. I reported this bug on emacs-pretest-bugs and don't read emacs-devel. I just happened to notice my name while looking through the archives for a Windows Emacs pretest binary. Please, anyone, when replying to messages forwarded from a different list, either reply to the original list or the original author. Otherwise your reply could well go unseen by the person you're attempting to address. To answer your question: gnus-read-active-file is a variable defined in `gnus-start.el'. Its value is some Chris. ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: Emacsclient/server filename quoting error
Date: Fri, 15 Dec 2006 11:31:18 +0100 From: Juanma Barranquero [EMAIL PROTECTED] Cc: emacs-pretest-bug@gnu.org, Emacs Devel emacs-devel@gnu.org I execute all the following commands from a cmd command prompt (outside of Emacs). emacsclient -n -a runemacs TO DO.txt works correctly if Emacs IS already running, but if it is not already running then Emacs does not see the filename correctly; the quotes do not appear to be passed on to runemacs. The following patch addresses the issue by allocating quoted copies of any argument containing spaces before calling execvp. Any objections to install this fix? As it stands, it affects also non-Windows builds. Is requoting args the right behavior in these environments? Quoting arguments passed to execvp is _definitely_ the wrong thing for Posix platforms. The fact that we need to do that for Windows is due to the broken implementation of exec* routines in the Microsoft libraries: they concatenate the arguments together without quoting special characters, and pass the result to CreateProcess, with predictably bad results. By contrast, Posix execvp passes the arguments directly into the argv[] array of the child process. So please make this change conditioned on WINDOWSNT. Actually, a cleaner way of fixing this would be to have a WINDOWSNT-only wrapper for execvp, called, say w32_execvp, that does TRT with quoting the arguments. Then you could say #ifdef WINDOWSNT #define execvp w32_execvp #endif in emacsclient.c, and leave the mainline code intact. ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: Emacsclient/server filename quoting error
On 12/15/06, Eli Zaretskii [EMAIL PROTECTED] wrote:
Actually, a cleaner way of fixing this would be to have a
WINDOWSNT-only wrapper for execvp, called, say w32_execvp, that does
TRT with quoting the arguments.
You like this one better, then?
/L/e/k/t/u
Index: lib-src/emacsclient.c
===
RCS file: /cvsroot/emacs/emacs/lib-src/emacsclient.c,v
retrieving revision 1.98
diff -u -2 -r1.98 emacsclient.c
--- lib-src/emacsclient.c 30 Nov 2006 22:49:38 - 1.98
+++ lib-src/emacsclient.c 15 Dec 2006 11:49:57 -
@@ -299,4 +299,35 @@
+#ifdef WINDOWSNT
+
+/*
+ execvp() wrapper for Windows.
+ Quotes arguments with embedded spaces.
+*/
+int
+w32_execvp (path, argv)
+ char *path;
+ char **argv;
+{
+ int i;
+
+ argv[0] = (char *) alternate_editor;
+
+ for (i = 0; argv[i]; i++)
+if (strchr (argv[i], ' '))
+ {
+ char *quoted = alloca (strlen (argv[i]) + 3);
+ sprintf (quoted, \%s\, argv[i]);
+ argv[i] = quoted;
+ }
+
+ return execvp (path, argv);
+}
+
+#undef execvp
+#define execvp w32_execvp
+
+#endif /* WINDOWSNT */
+
/*
Try to run a different command, or --if no alternate editor is
@@ -311,7 +342,5 @@
{
int i = optind - 1;
-#ifdef WINDOWSNT
- argv[i] = (char *)alternate_editor;
-#endif
+
execvp (alternate_editor, argv + i);
message (TRUE, %s: error executing alternate editor \%s\\n,
___
emacs-pretest-bug mailing list
emacs-pretest-bug@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
text.texi typos
cvs diff -u -- text.texi (in directory D:\download\emacs-gbk\man\)
Index: text.texi
===
RCS file: /cvsroot/cvs.savannah.gnu.org/emacs/emacs/man/text.texi,v
retrieving revision 1.73
diff -u -r1.73 text.texi
--- text.texi 6 Dec 2006 01:29:49 - 1.73
+++ text.texi 15 Dec 2006 13:23:32 -
@@ -1844,7 +1844,7 @@
Run a shell command (which you must specify) to validate the current
buffer as SGML (@code{sgml-validate}).
[EMAIL PROTECTED] C-x TAB
[EMAIL PROTECTED] C-c TAB
@kindex C-c TAB @r{(SGML mode)}
@findex sgml-tags-invisible
Toggle the visibility of existing tags in the buffer. This can be
* CVS exited normally with code 1 *
___
emacs-pretest-bug mailing list
emacs-pretest-bug@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: text.texi typos
From: Zhang Wei [EMAIL PROTECTED]
Date: Fri, 15 Dec 2006 21:27:34 +0800
cvs diff -u -- text.texi (in directory D:\download\emacs-gbk\man\)
Index: text.texi
===
RCS file: /cvsroot/cvs.savannah.gnu.org/emacs/emacs/man/text.texi,v
retrieving revision 1.73
diff -u -r1.73 text.texi
--- text.texi 6 Dec 2006 01:29:49 - 1.73
+++ text.texi 15 Dec 2006 13:23:32 -
@@ -1844,7 +1844,7 @@
Run a shell command (which you must specify) to validate the current
buffer as SGML (@code{sgml-validate}).
[EMAIL PROTECTED] C-x TAB
[EMAIL PROTECTED] C-c TAB
@kindex C-c TAB @r{(SGML mode)}
@findex sgml-tags-invisible
Toggle the visibility of existing tags in the buffer. This can be
Thanks, I installed this.
___
emacs-pretest-bug mailing list
emacs-pretest-bug@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: Emacsclient/server filename quoting error
Date: Fri, 15 Dec 2006 13:07:01 +0100 From: Juanma Barranquero [EMAIL PROTECTED] Cc: [EMAIL PROTECTED], emacs-pretest-bug@gnu.org, emacs-devel@gnu.org On 12/15/06, Eli Zaretskii [EMAIL PROTECTED] wrote: Actually, a cleaner way of fixing this would be to have a WINDOWSNT-only wrapper for execvp, called, say w32_execvp, that does TRT with quoting the arguments. You like this one better, then? Yes, thanks. However, please mention the Windows bug with execvp in the comment to w32_execvp, otherwise it is not clear to the uninitiated why it is needed, and why we quote the arguments we pass to the real execvp. ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: Emacsclient/server filename quoting error
On 12/15/06, Werner LEMBERG [EMAIL PROTECTED] wrote: Since those issues are full of subtleties I suggest that you have a look at the gnulib CVS (at savannah.gnu.org), inspecting the files `execute.c' and `pipe.c': Thanks, but I think I'll wait for people to test the current implementation. I don't think Windows users will have very sophisticate or uncommon alternate editor setups and if they do, they must be prepared for the possibility of it not working ;) /L/e/k/t/u ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: Emacsclient/server filename quoting error
On 12/15/06, Eli Zaretskii [EMAIL PROTECTED] wrote: However, please mention the Windows bug with execvp in the comment to w32_execvp I've shamelessly stolen your comment from a previous message. Thanks, /L/e/k/t/u ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: Emacsclient/server filename quoting error
Date: Fri, 15 Dec 2006 16:42:57 +0100 From: Juanma Barranquero [EMAIL PROTECTED] Cc: [EMAIL PROTECTED], emacs-devel@gnu.org, emacs-pretest-bug@gnu.org I've shamelessly stolen your comment from a previous message. That's okay, since I have a copyright assignment on file with the FSF. ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: Emacsclient/server filename quoting error
Date: Fri, 15 Dec 2006 16:14:38 +0100 (CET) Cc: [EMAIL PROTECTED], [EMAIL PROTECTED], emacs-devel@gnu.org, emacs-pretest-bug@gnu.org From: Werner LEMBERG [EMAIL PROTECTED] Since those issues are full of subtleties I suggest that you have a look at the gnulib CVS (at savannah.gnu.org), inspecting the files `execute.c' and `pipe.c': http://cvs.savannah.gnu.org/viewcvs/gnulib/lib/execute.c?rev=1.7root=gnulibview=log http://cvs.savannah.gnu.org/viewcvs/gnulib/lib/pipe.c?rev=1.6root=gnulibview=log Thanks, but I see nothing in these source files that could shed any light on the case in point. Similar code is used in groff too. Groff builds a pipe of several programs and then runs that pipe. By contrast, emacsclient needs to run a single program, and doesn't need to read/write its standard streams. So the complexity of the above URLs is not needed. ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: image.el doesn't associate image-mode with .JPG files
It would handle that one case, but it would still produce false matches. It would only produce false matches for cases where Emacs would have defaulted to Fundamental mode. Yes, and that is a mistaken outcome. There is no reason to make such mistakes happen. When an entry should match more than one case pattern, we write it to do so. That's a small amount of work, and it gives 100% correct results. Please drop this idea. ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: TRAMP password caching
*if* (!hwnd) *return* FALSE; *else* *return* TRUE; aka return !!hwnd; Stefan by eta-reduction ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: Mysterious fontification/C++ context issue
In the time Emacs 22 is the current release, a typical new PC will come to be around 20 GHz, and this slowness will not matter. Actually, recent trends indicate that this is not true. We'll probably see typical new PCs with 4-16 CPUs, each one running at 4-5GHz, but that won't help Emacs much. Hmmm. One processor for foreground editing, another for doing garbage collection, yet another four or five for background font-locking and syntactic cacheing, and so on. That could be quite a fast system. I'm anxiously awaiting your patch to Emacs's C code base to make such concurrency possible. Stefan ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: scroll-other-window scrolls wrong window for vc-revert-buffer
Kim F. Storm [EMAIL PROTECTED] writes: emacs -Q C-x C-f some vc-controlled file with many uncommitted change RET C-x v u emacs frame spits into two with the diffs in the lower window Prompt is Discard changes? (yes or no) I would like to review the changes, so I try to scroll the diffs C-M-v But this scrolls the source window rather than the diffs window. Not very useful... Did anybody look at this? But it gets worse: I don't like the yes or no prompts, so I have this in my .emacs: (defalias 'yes-or-no-p 'y-or-n-p) Now, C-M-v doesn't work at all --- it just beeps. I looked into this ... and it is because yes-or-no-p uses read-from-minibuffer, and thus obeys all key bindings. In contrast, y-or-n-p does the reading itself, and looks up the bindings ONLY in query-replace-map (which obviously don't tell how to scroll-other-window. I don't see an easy fix for this... -- 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: try-completion crashes on non-obarray vector
On 12/8/06, Johan Bockgård [EMAIL PROTECTED] wrote: (try-completion []) = Program received signal SIGSEGV, Segmentation fault. 0x004d7e67 in Ftry_completion (string=10456851, alist=9781604, predicate=9337233) at minibuf.c:1315 1315 if (XSYMBOL (bucket)-next) Fixed. Thanks, /L/e/k/t/u ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: ido.el reports possible completions wrongly
Chris Moore [EMAIL PROTECTED] writes: Chris Moore [EMAIL PROTECTED] writes: [EMAIL PROTECTED] (Kim F. Storm) writes: Look at ido-max-directory-size. In this case C-a is your friend. The bug is that the message says: There are no possible completions of what you have typed when in fact there are possible completions of what I have typed. I just fixed this, so it will actually load the directory (despite it is [too big] and show you the possible completions. I just noticed also, that if I immediately type a 2nd '?': C-x C-f / u s r / b i n / g n u p l o ? ? then I see a giant list of 'possible completions': This is what you would normally get once you have scrolled through the list of possible matches the first time (i.e. all directory entries). So since the first ? (and subsequent ?'es to scroll the completion buffer) now works correctly, this also works ok now. -- 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: image.el doesn't associate image-mode with .JPG files
On 12/14/06, Richard Stallman [EMAIL PROTECTED] wrote: That seems like a valid reason. Would someone please install that change? It would be useful to recognize PNG and GIF as well this way. image.el already defines `image-type-header-regexps' and `image-type-from-buffer', so it seems wasteful to recreate that. It would be better IMHO to simply use `image-type-from-buffer' from `set-auto-mode'. Other than changes to docstrings and info files, the path is as simple as this, and works quite well. /L/e/k/t/u Index: lisp/files.el === RCS file: /cvsroot/emacs/emacs/lisp/files.el,v retrieving revision 1.868 diff -c -r1.868 files.el *** lisp/files.el 11 Dec 2006 04:40:41 - 1.868 --- lisp/files.el 16 Dec 2006 02:51:21 - *** *** ,2230 (narrow-to-region (point-min) (min (point-max) (+ (point-min) magic-mode-regexp-match-limit))) !(assoc-default nil magic-mode-alist ! (lambda (re dummy) ! (looking-at re)) (set-auto-mode-0 done keep-mode-if-same) ;; Compare the filename against the entries in auto-mode-alist. (if buffer-file-name --- ,2232 (narrow-to-region (point-min) (min (point-max) (+ (point-min) magic-mode-regexp-match-limit))) !(if (image-type-from-buffer) !'image-mode ! (assoc-default nil magic-mode-alist ! (lambda (re dummy) ! (looking-at re))) (set-auto-mode-0 done keep-mode-if-same) ;; Compare the filename against the entries in auto-mode-alist. (if buffer-file-name ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
Re: image.el doesn't associate image-mode with .JPG files
On 12/16/06, Juanma Barranquero [EMAIL PROTECTED] wrote: path is as simple as this, and works quite well. s/path/patch/ sleep /L/e/k/t/u ___ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
