GNU bug report logs - #50976
28.0.50; Tramp read-file-name regression

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 50976 in the body.
You can then email your comments to 50976 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Juri Linkov <juri <at> linkov.net>
To: bug-gnu-emacs <at> gnu.org
Subject: 28.0.50; Tramp read-file-name regression
Date: Sat, 02 Oct 2021 22:19:25 +0300

This is a recent regression.  When visiting a remote Dired buffer,
marking one file with C-SPC (set-mark-command), then moving point
to another Dired file, then typing '=' (dired-diff) and 'RET'
fails with the error "Attempt to compare the file to itself".

This is because 'read-file-name' in the interactive spec of 'dired-diff'
now returns the remote directory name after typing RET,
while some time ago it correctly returned the default file name,
that is the marked file that was used as the arg 'file' of 'dired-diff'.

In local filesystems, RET still correctly returns the default file,
not a directory name as it does now for remote directories.

Message #8 received at 50976 <at> debbugs.gnu.org (full text, mbox):

From: Michael Albinus <michael.albinus <at> gmx.de>
To: Juri Linkov <juri <at> linkov.net>
Cc: 50976 <at> debbugs.gnu.org
Subject: Re: bug#50976: 28.0.50; Tramp read-file-name regression
Date: Tue, 05 Oct 2021 16:32:34 +0200

Juri Linkov <juri <at> linkov.net> writes:

Hi Juri,

> This is a recent regression.  When visiting a remote Dired buffer,
> marking one file with C-SPC (set-mark-command), then moving point
> to another Dired file, then typing '=' (dired-diff) and 'RET'
> fails with the error "Attempt to compare the file to itself".
>
> This is because 'read-file-name' in the interactive spec of 'dired-diff'
> now returns the remote directory name after typing RET,
> while some time ago it correctly returned the default file name,
> that is the marked file that was used as the arg 'file' of 'dired-diff'.
>
> In local filesystems, RET still correctly returns the default file,
> not a directory name as it does now for remote directories.

FTR, git bisect has identified the following guilty commit:

--8<---------------cut here---------------start------------->8---
commit 225ca617b70d3c70376c2d9bf38ced2f2323b26e (refs/bisect/bad)
Author: Michael Albinus <michael.albinus <at> gmx.de>
Date:   Fri Jul 2 14:51:23 2021 +0200

    Implement another fix for bug#49229

    * lisp/minibuffer.el (read-file-name-default): Respect remote files.
    (Bug#49229)
--8<---------------cut here---------------end--------------->8---

I will check how bug#49229 could be fixed differently.

Best regards, Michael.

Message #11 received at 50976 <at> debbugs.gnu.org (full text, mbox):

From: Michael Albinus <michael.albinus <at> gmx.de>
To: Juri Linkov <juri <at> linkov.net>
Cc: 50976 <at> debbugs.gnu.org
Subject: Re: bug#50976: 28.0.50; Tramp read-file-name regression
Date: Fri, 08 Oct 2021 16:36:02 +0200

Michael Albinus <michael.albinus <at> gmx.de> writes:

Hi Juri,

> I will check how bug#49229 could be fixed differently.

I must confess, w/o a recent build of Emacs 28.0.60 for Windows I cannot
do it :-(

Best regards, Michael.

Message #14 received at 50976 <at> debbugs.gnu.org (full text, mbox):

From: Michael Albinus <michael.albinus <at> gmx.de>
To: Juri Linkov <juri <at> linkov.net>
Cc: 50976 <at> debbugs.gnu.org
Subject: Re: bug#50976: 28.0.50; Tramp read-file-name regression
Date: Sat, 23 Oct 2021 15:29:55 +0200

Michael Albinus <michael.albinus <at> gmx.de> writes:

Hi Juri,

>> I will check how bug#49229 could be fixed differently.
>
> I must confess, w/o a recent build of Emacs 28.0.60 for Windows I cannot
> do it :-(

Finally, I've managed to install this beast on my MS Windows VM. I've
removed the incrimed patch, and implemented something different. Your
use case now works for me, on both MS Windows and GNU/Linux.

Could you please check?

Best regards, Michael.

Message #17 received at 50976 <at> debbugs.gnu.org (full text, mbox):

From: Juri Linkov <juri <at> linkov.net>
To: Michael Albinus <michael.albinus <at> gmx.de>
Cc: 50976 <at> debbugs.gnu.org
Subject: Re: bug#50976: 28.0.50; Tramp read-file-name regression
Date: Sat, 23 Oct 2021 21:09:56 +0300

close 50976 28.0.60
thanks

Hi Michael,

Thank you for fixing this.

>>> I will check how bug#49229 could be fixed differently.
>>
>> I must confess, w/o a recent build of Emacs 28.0.60 for Windows I cannot
>> do it :-(
>
> Finally, I've managed to install this beast on my MS Windows VM. I've
> removed the incrimed patch, and implemented something different. Your
> use case now works for me, on both MS Windows and GNU/Linux.
>
> Could you please check?

I confirm that now there is no problem anymore on GNU/Linux,
so I'm closing this report (hope that on MS Windows this works too).

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.