Re: bug#7854: emacs-23 (regression) - fix breaks `dired'

Re: bug#7854: emacs-23 (regression) - fix breaks `dired'

[Chong Yidong]

Could someone with access to Windows please test the regression reported
at Bug#7854?  I could not reproduce it, replacing "u:/tmp" with "/tmp/".

> $ emacs -Q -l files-emacs23-git.el --eval '(dired "u:/tmp/")'
> (files-emacs23-git.el =
> http://git.savannah.gnu.org/cgit/emacs.git/plain/lisp/files.el?h=emacs-23&id=8588a5a723f4e5cfa115341a242126a9efc71808)
>
> I get:
>
> | if: Reading directory: "ls -al -- u:/tmp/" exited with status 1

If someone could come up with a smaller testcase, that would be good
too.

Re: bug#7854: emacs-23 (regression) - fix breaks `dired'

[Eli Zaretskii]

> From: Chong Yidong <cyd@stupidchicken.com>
> Date: Sat, 12 Feb 2011 15:20:06 -0500
> Cc: 7854@debbugs.gnu.org, Reiner Steib <reinersteib+gmane@imap.cc>
> 
> Could someone with access to Windows please test the regression reported
> at Bug#7854?  I could not reproduce it, replacing "u:/tmp" with "/tmp/".
> 
> > $ emacs -Q -l files-emacs23-git.el --eval '(dired "u:/tmp/")'
> > (files-emacs23-git.el =
> > http://git.savannah.gnu.org/cgit/emacs.git/plain/lisp/files.el?h=emacs-23&id=8588a5a723f4e5cfa115341a242126a9efc71808)
> >
> > I get:
> >
> > | if: Reading directory: "ls -al -- u:/tmp/" exited with status 1
> 
> If someone could come up with a smaller testcase, that would be good
> too.

I cannot reproduce this with the current trunk (replacing u:/tmp with
a directory I actually have on my system).  Up comes Dired with the
listing of that directory.  The only way I can cause it to fail is by
removing ls.exe from my PATH -- but that produces a different error
message:

  Debugger entered--Lisp error: (file-error "Searching for program" "no such file or directory" "ls")
    call-process("ls" nil t nil "-al" "--" "h:/test/.")
    apply(call-process "ls" nil t nil ("-al" "--" "h:/test/."))

This error is expected: loading files.el overrides the replacement
functions installed by ls-lisp.el (which is preloaded on Windows), so
now Dired on Windows must have a working ls.exe executable.  (I
actually don't understand what is the purpose of loading files.el from
Emacs 23 into Emacs 24, but that's me.)

Reiner, can you still see this problem in today's trunk?

Re: bug#7854: emacs-23 (regression) - fix breaks `dired'

[Reiner Steib]

On Sat, 12 Feb 2011 23:11 +0200, "Eli Zaretskii" <eliz@gnu.org> wrote:
> > From: Chong Yidong <cyd@stupidchicken.com>
> > Date: Sat, 12 Feb 2011 15:20:06 -0500
> > Cc: 7854@debbugs.gnu.org, Reiner Steib <reinersteib+gmane@imap.cc>
> > 
> > Could someone with access to Windows please test the regression reported
> > at Bug#7854?  I could not reproduce it, replacing "u:/tmp" with "/tmp/".
> > 
> > > $ emacs -Q -l files-emacs23-git.el --eval '(dired "u:/tmp/")'
> > > (files-emacs23-git.el =
> > > http://git.savannah.gnu.org/cgit/emacs.git/plain/lisp/files.el?h=emacs-23&id=8588a5a723f4e5cfa115341a242126a9efc71808)
> > >
> > > I get:
> > >
> > > | if: Reading directory: "ls -al -- u:/tmp/" exited with status 1
> > 
> > If someone could come up with a smaller testcase, that would be good
> > too.
> 
> I cannot reproduce this with the current trunk (replacing u:/tmp with
> a directory I actually have on my system).  Up comes Dired with the
> listing of that directory.  The only way I can cause it to fail is by
> removing ls.exe from my PATH -- but that produces a different error
> message:
> 
>   Debugger entered--Lisp error: (file-error "Searching for program" "no
>   such file or directory" "ls")
>     call-process("ls" nil t nil "-al" "--" "h:/test/.")
>     apply(call-process "ls" nil t nil ("-al" "--" "h:/test/."))
> 
> This error is expected: loading files.el overrides the replacement
> functions installed by ls-lisp.el (which is preloaded on Windows), so
> now Dired on Windows must have a working ls.exe executable.  

Thanks for the explanation.

> (I actually don't understand what is the purpose of loading files.el
> from Emacs 23 into Emacs 24, but that's me.)

As I use a pre-compiled Emacs (when using Windows), I loaded the most
recent version of files.el for the emacs-23 branch from git-web (with
Stefans's fix for the original problem) on the command line (using emacs
23.2) because the old files.elc is dumped (if I understand correctly).
The original problem was reported for the trunk, but the (potential)
regression was found in the emacs-23 branch.  Sorry for the confusion.
 
> Reiner, can you still see this problem in today's trunk?

Now I downloaded the most recent binaries: 23.3 pretest
(emacs-23.2.93-bin-i386.zip) and trunk (emacs-20110207-bin-i386.zip).

I cannot reproduce the problem with these versions.  So please close
this bug again - sorry for the noise.

Bye, Reiner

-- 
http://www.fastmail.fm - mmm... Fastmail...