GNU bug report logs - #4243
Emacs 23 ido-mode vc-hg bug

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 4243 in the body.
You can then email your comments to 4243 AT debbugs.gnu.org in the normal way.

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

Message #5 received at submit <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Tracy Reed <treed <at> copilotco.com>
To: bug-gnu-emacs <at> gnu.org
Subject: Emacs 23 ido-mode vc-hg bug
Date: Sun, 23 Aug 2009 16:03:21 -0700

[Message part 1 (text/plain, inline)]

When I start emacs 23.1.1 which I compiled myself from code downloaded
from the official site:

GNU Emacs 23.1.1 (x86_64-unknown-linux-gnu) of 2009-08-05 on mail.copilotco.com

using emacs -Q (so my .emacs and site-lisp are not involved) I run
into an error when I activate ido-mode and try to exit emacs with
ctrl-x ctrl-c. Here is what appears in my Messages buffer:

("emacs" "-Q")
For information about GNU Emacs and the GNU system, type C-h C-a.
Ido mode enabled
Loading tramp...done
Quit [2 times]
(No files need saving)
Saving file /home/treed/.ido.last...
Loading vc-hg...done
Wrote /home/treed/.ido.last
vc-hg-registered: Args out of range: "", 0

I started emacs with -Q, then I did M-x ido-mode, then I did ctrl-x
ctrl-c. Then I get the:

vc-hg-registered: Args out of range: "", 0

error. I cannot exit emacs. I have to ctrl-z and kill it.  I suspect
vc-hg might have been activated because my home directory is under hg
version control. I'm not sure why it says "Loading tramp" there. I
did not do anything with tramp.

Please advise of any fixes/workarounds or if there is any more
information I can provide or anything I can do to resolve this.

Thanks!

-- 
Tracy Reed
http://copilotco.com

[Message part 2 (application/pgp-signature, inline)]

Message #8 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Dan Nicolaescu <dann <at> ics.uci.edu>
To: Tracy Reed <treed <at> copilotco.com>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Sun, 23 Aug 2009 23:43:20 -0700 (PDT)

Tracy Reed <treed <at> copilotco.com> writes:

  > When I start emacs 23.1.1 which I compiled myself from code downloaded
  > from the official site:
  > 
  > GNU Emacs 23.1.1 (x86_64-unknown-linux-gnu) of 2009-08-05 on mail.copilotco.com
  > 
  > using emacs -Q (so my .emacs and site-lisp are not involved) I run
  > into an error when I activate ido-mode and try to exit emacs with
  > ctrl-x ctrl-c. Here is what appears in my Messages buffer:
  > 
  > ("emacs" "-Q")
  > For information about GNU Emacs and the GNU system, type C-h C-a.
  > Ido mode enabled
  > Loading tramp...done
  > Quit [2 times]
  > (No files need saving)
  > Saving file /home/treed/.ido.last...
  > Loading vc-hg...done
  > Wrote /home/treed/.ido.last
  > vc-hg-registered: Args out of range: "", 0
  > 
  > I started emacs with -Q, then I did M-x ido-mode, then I did ctrl-x
  > ctrl-c. Then I get the:
  > 
  > vc-hg-registered: Args out of range: "", 0

We need to find out what calls vc-hg-registered with such arguments.

  > error. I cannot exit emacs. I have to ctrl-z and kill it.  I suspect
  > vc-hg might have been activated because my home directory is under hg
  > version control. I'm not sure why it says "Loading tramp" there. I
  > did not do anything with tramp.
  > 
  > Please advise of any fixes/workarounds or if there is any more
  > information I can provide or anything I can do to resolve this.

Can you please do:

M-x load-library RET vc.el RET
M-x load-library RET vc-hg.el RET
M-: (setq debug-on-error t) RET

this will produce a backtrace buffer, please email the contents of that
buffer.

Message #13 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Tracy Reed <treed <at> copilotco.com>
To: Dan Nicolaescu <dann <at> ics.uci.edu>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Sun, 23 Aug 2009 23:56:22 -0700

[Message part 1 (text/plain, inline)]

On Sun, Aug 23, 2009 at 11:43:20PM -0700, Dan Nicolaescu spake thusly:
> M-x load-library RET vc.el RET
> M-x load-library RET vc-hg.el RET
> M-: (setq debug-on-error t) RET

I did the following then reproduced the error by invoking ido-mode and
then trying to exit as I described previously. The traceback is
attached. Looks like it is looking at .ido.last or something. That
file is not managed by hg in my setup, if it matters.

-- 
Tracy Reed
http://copilotco.com

[emacs-error (text/plain, attachment)]

[Message part 3 (application/pgp-signature, inline)]

Message #16 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Dan Nicolaescu <dann <at> ics.uci.edu>
To: Tracy Reed <treed <at> copilotco.com>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 00:37:50 -0700 (PDT)

Tracy Reed <treed <at> copilotco.com> writes:

  > On Sun, Aug 23, 2009 at 11:43:20PM -0700, Dan Nicolaescu spake thusly:
  > > M-x load-library RET vc.el RET
  > > M-x load-library RET vc-hg.el RET
  > > M-: (setq debug-on-error t) RET
  > 
  > I did the following then reproduced the error by invoking ido-mode and
  > then trying to exit as I described previously. The traceback is
  > attached. Looks like it is looking at .ido.last or something. That
  > file is not managed by hg in my setup, if it matters.

What's the output of:

cd
hg status -A .ido.last

Message #21 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Tracy Reed <treed <at> copilotco.com>
To: Dan Nicolaescu <dann <at> ics.uci.edu>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 00:52:15 -0700

[Message part 1 (text/plain, inline)]

On Mon, Aug 24, 2009 at 12:37:50AM -0700, Dan Nicolaescu spake thusly:
> What's the output of:
> 
> cd
> hg status -A .ido.last

[treed <at> mail ~]$ cd 
[treed <at> mail ~]$ hg status -A .ido.last
[treed <at> mail ~]$ 

No output.

-- 
Tracy Reed
http://copilotco.com

[Message part 2 (application/pgp-signature, inline)]

Message #24 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Dan Nicolaescu <dann <at> ics.uci.edu>
To: Tracy Reed <treed <at> copilotco.com>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 01:14:40 -0700 (PDT)

Tracy Reed <treed <at> copilotco.com> writes:

  > On Mon, Aug 24, 2009 at 12:37:50AM -0700, Dan Nicolaescu spake thusly:
  > > What's the output of:
  > > 
  > > cd
  > > hg status -A .ido.last
  > 
  > [treed <at> mail ~]$ cd 
  > [treed <at> mail ~]$ hg status -A .ido.last
  > [treed <at> mail ~]$ 
  > 
  > No output.

And the file exists?
That looks strange because it should output something...
What version of hg?

Message #29 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Tracy Reed <treed <at> copilotco.com>
To: Dan Nicolaescu <dann <at> ics.uci.edu>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 01:49:46 -0700

[Message part 1 (text/plain, inline)]

On Mon, Aug 24, 2009 at 01:14:40AM -0700, Dan Nicolaescu spake thusly:
> And the file exists?

Yes, the file exists.

> That looks strange because it should output something...

hg status -A by itself produces a lot of output. hg status -A on a
non-existant file produces nothing just as it does on .ido.last

> What version of hg?

[treed <at> mail ~]$ hg --version
Mercurial Distributed SCM (version b81baf9e4dd6)

b81baf9e4dd6 is a recent stable version. I built it from source with a
pull from their stable repo:

http://selenic.com/repo/hg-stable/log?rev=b81baf9e4dd6

-- 
Tracy Reed
http://copilotco.com

[Message part 2 (application/pgp-signature, inline)]

Message #32 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Dan Nicolaescu <dann <at> ics.uci.edu>
To: Tracy Reed <treed <at> copilotco.com>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 06:12:06 -0700 (PDT)

Tracy Reed <treed <at> copilotco.com> writes:

  > On Mon, Aug 24, 2009 at 01:14:40AM -0700, Dan Nicolaescu spake thusly:
  > > And the file exists?
  > 
  > Yes, the file exists.
  > 
  > > That looks strange because it should output something...
  > 
  > hg status -A by itself produces a lot of output. hg status -A on a
  > non-existant file produces nothing just as it does on .ido.last

hg status -A non-existent-file
non-existent-file: No such file or directory

hg --version
Mercurial Distributed SCM (version 1.3)

And the behavior has been the same from before version 1.0.

  > > What version of hg?
  > 
  > [treed <at> mail ~]$ hg --version
  > Mercurial Distributed SCM (version b81baf9e4dd6)
  > 
  > b81baf9e4dd6 is a recent stable version. I built it from source with a
  > pull from their stable repo:

It looks like a bug in your hg version...

Message #37 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Tracy Reed <treed <at> ultraviolet.org>
To: Dan Nicolaescu <dann <at> ics.uci.edu>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 11:53:57 -0700

[Message part 1 (text/plain, inline)]

On Mon, Aug 24, 2009 at 06:12:06AM -0700, Dan Nicolaescu spake thusly:
> It looks like a bug in your hg version...

Ah-hah...I have tracked it down to this in my .hgrc file:

[defaults]
status = -q

I use hg mostly for sysadmin stuff and often have lots of files in a
directory that aren't checked in. But I use hg status often to see
what has been modified and do not want to see all the noise by the
files that are not checked in. So I use this default. Apparently it
confuses vc-hg. Not sure if emacs could handle this better somehow but
it isn't a big deal. I have commented it out of my .hgrc for now.

Thanks a lot for your help in troubleshooting this and sorry for
wasting your time with what turns out to be a non-emacs bug.

-- 
Tracy Reed
http://tracyreed.org

[Message part 2 (application/pgp-signature, inline)]

Message #40 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Dan Nicolaescu <dann <at> ics.uci.edu>
To: Tracy Reed <treed <at> ultraviolet.org>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 12:14:44 -0700 (PDT)

Tracy Reed <treed <at> ultraviolet.org> writes:

  > On Mon, Aug 24, 2009 at 06:12:06AM -0700, Dan Nicolaescu spake thusly:
  > > It looks like a bug in your hg version...
  > 
  > Ah-hah...I have tracked it down to this in my .hgrc file:
  > 
  > [defaults]
  > status = -q
  > I use hg mostly for sysadmin stuff and often have lots of files in a
  > directory that aren't checked in. But I use hg status often to see
  > what has been modified and do not want to see all the noise by the
  > files that are not checked in. So I use this default. Apparently it
  > confuses vc-hg. Not sure if emacs could handle this better somehow but
  > it isn't a big deal. 

Is there a command line flag to disable user's .hgrc? (like emacs has -q) 
If yes, the we can just pass that flag to "hg status", that way we
always get consistent results.
If not, not sure what we can do...

Message #45 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Tracy Reed <treed <at> copilotco.com>
To: Dan Nicolaescu <dann <at> ics.uci.edu>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 12:30:54 -0700

[Message part 1 (text/plain, inline)]

On Mon, Aug 24, 2009 at 12:14:44PM -0700, Dan Nicolaescu spake thusly:
> Is there a command line flag to disable user's .hgrc? (like emacs has -q) 

Not a command line flag but the command:

hg help env 

mentions the HGRCPATH environment variable. It is a typical :
separated path that tells hg where to look for its rc file. If it is
empty it reads only in the repository .hg/hgrc file and not in the
user's ~/.hgrc or anywhere else. So if vc-hg set HGRCPATH="" that
would fix the issue for me although someone who happens to put status
= -q in their repo/.hg/hgrc file would still encounter confusion. I
just have talked to the hg guys in #mercurial on freenode and they are
not aware of any way to completely disable all forms of hgrc reading.

-- 
Tracy Reed
http://copilotco.com

[Message part 2 (application/pgp-signature, inline)]

Message #50 received at 4243 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Tracy Reed <treed <at> copilotco.com>
To: Dan Nicolaescu <dann <at> ics.uci.edu>
Cc: 4243 <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 12:43:24 -0700

[Message part 1 (text/plain, inline)]

On Mon, Aug 24, 2009 at 12:30:54PM -0700, Tracy Reed spake thusly:
> Not a command line flag but the command:
> 
> hg help env 
> 
> mentions the HGRCPATH environment variable. It is a typical :

Hmm...the hg guys warn that disabling reading of the users .hgrc file
would "mess up the ability to utilize any extensions the user has
enabled, committing would be problematic as the username wouldn't be
set, and so on." So...

-- 
Tracy Reed
http://copilotco.com

[Message part 2 (application/pgp-signature, inline)]

Message #55 received at 4243-done <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Dan Nicolaescu <dann <at> ics.uci.edu>
To: Tracy Reed <treed <at> copilotco.com>
Cc: 4243-done <at> debbugs.gnu.org
Subject: Re: bug#4243: Emacs 23 ido-mode vc-hg bug
Date: Mon, 24 Aug 2009 12:47:31 -0700 (PDT)

Tracy Reed <treed <at> copilotco.com> writes:

  > On Mon, Aug 24, 2009 at 12:30:54PM -0700, Tracy Reed spake thusly:
  > > Not a command line flag but the command:
  > > 
  > > hg help env 
  > > 
  > > mentions the HGRCPATH environment variable. It is a typical :
  > 
  > Hmm...the hg guys warn that disabling reading of the users .hgrc file
  > would "mess up the ability to utilize any extensions the user has
  > enabled, committing would be problematic as the username wouldn't be
  > set, and so on." So...

I am closing this bug then, there's nothing we can do in emacs.
This sounds like a design bug in mercurial, tools (like emacs in this
case) cannot rely on getting consistent output from hg commands.

Previous Next

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