GNU bug report logs -
#11666
default-directory
Previous Next
Reported by: Dave Abrahams <dave <at> boostpro.com>
Date: Sun, 10 Jun 2012 17:33:02 UTC
Severity: normal
Tags: wontfix
Found in version 5.130006
Done: Lars Ingebrigtsen <larsi <at> gnus.org>
Bug is archived. No further changes may be made.
To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 11666 in the body.
You can then email your comments to 11666 AT debbugs.gnu.org in the normal way.
Toggle the display of automated, internal messages from the tracker.
Report forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Sun, 10 Jun 2012 17:33:02 GMT)
Full text and
rfc822 format available.
Acknowledgement sent
to
Dave Abrahams <dave <at> boostpro.com>
:
New bug report received and forwarded. Copy sent to
bugs <at> gnus.org
.
(Sun, 10 Jun 2012 17:33:02 GMT)
Full text and
rfc822 format available.
Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):
I find quite often that all of Gnu's many buffers are being created with
a default-directory set to something that's either accessed via TRAMP or
simply doesn't exist anymore. Either one will cause major problems,
with things either getting crazy slow, or just not working anymore. I
have no good ideas about how to track this down and fix it :-( but I
thought you should have a report.
Ma Gnus v0.6
GNU Emacs 24.0.94.1 (x86_64-apple-darwin11.4.0, Carbon Version 1.6.0 AppKit 1138.47)
of 2012-05-24 on pluto.luannocracy.com
200 Text.GigaNews.Com
500 syntax error or unknown command
--
Dave Abrahams
BoostPro Computing
http://www.boostpro.com
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Sun, 10 Jun 2012 22:02:02 GMT)
Full text and
rfc822 format available.
Message #8 received at 11666 <at> debbugs.gnu.org (full text, mbox):
Dave Abrahams <dave <at> boostpro.com> writes:
> I find quite often that all of Gnu's many buffers are being created with
> a default-directory set to something that's either accessed via TRAMP or
> simply doesn't exist anymore. Either one will cause major problems,
> with things either getting crazy slow, or just not working anymore. I
> have no good ideas about how to track this down and fix it :-( but I
> thought you should have a report.
Do you set `default-directory' or `gnus-default-directory' to something
explicitly?
I can't really imagine how Gnus would set the default directory in any
of its buffers to a TRAMP path by itself...
--
(domestic pets only, the antidote for overdose, milk.)
bloggy blog http://lars.ingebrigtsen.no/
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Mon, 11 Jun 2012 00:07:01 GMT)
Full text and
rfc822 format available.
Message #11 received at 11666 <at> debbugs.gnu.org (full text, mbox):
on Sun Jun 10 2012, Lars Magne Ingebrigtsen <larsi-AT-gnus.org> wrote:
> Dave Abrahams <dave <at> boostpro.com> writes:
>
>> I find quite often that all of Gnu's many buffers are being created with
>> a default-directory set to something that's either accessed via TRAMP or
>> simply doesn't exist anymore. Either one will cause major problems,
>> with things either getting crazy slow, or just not working anymore. I
>> have no good ideas about how to track this down and fix it :-( but I
>> thought you should have a report.
>
> Do you set `default-directory' or `gnus-default-directory' to something
> explicitly?
`M-x rgrep default-directory RET...'
Looks like... nope.
> I can't really imagine how Gnus would set the default directory in any
> of its buffers to a TRAMP path by itself...
I'm sure it doesn't do anything "by itself" :-). Maybe something
changes the global definition of default-directory, or dynamic scoping
is unexpectedly getting in the way.
--
Dave Abrahams
BoostPro Computing
http://www.boostpro.com
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Mon, 11 Jun 2012 00:15:02 GMT)
Full text and
rfc822 format available.
Message #14 received at 11666 <at> debbugs.gnu.org (full text, mbox):
Dave Abrahams <dave <at> boostpro.com> writes:
>> I can't really imagine how Gnus would set the default directory in any
>> of its buffers to a TRAMP path by itself...
>
> I'm sure it doesn't do anything "by itself" :-). Maybe something
> changes the global definition of default-directory, or dynamic scoping
> is unexpectedly getting in the way.
I see that all it takes is that you go to a tramp buffer and then say
`M-x gnus' for all the Gnus process buffers to get that as the default
directory.
That's probably not very nice. But would binding `default-directory' to
"/" (or something) for the Gnus process buffers be the right thing?
Would that work?
I have a feeling that this might be one of those "don't do that, then"
things.
--
(domestic pets only, the antidote for overdose, milk.)
bloggy blog http://lars.ingebrigtsen.no/
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Mon, 11 Jun 2012 01:52:01 GMT)
Full text and
rfc822 format available.
Message #17 received at 11666 <at> debbugs.gnu.org (full text, mbox):
on Sun Jun 10 2012, Lars Magne Ingebrigtsen <larsi-AT-gnus.org> wrote:
> Dave Abrahams <dave <at> boostpro.com> writes:
>
>>> I can't really imagine how Gnus would set the default directory in any
>>> of its buffers to a TRAMP path by itself...
>>
>> I'm sure it doesn't do anything "by itself" :-). Maybe something
>> changes the global definition of default-directory, or dynamic scoping
>> is unexpectedly getting in the way.
>
> I see that all it takes is that you go to a tramp buffer and then say
> `M-x gnus' for all the Gnus process buffers to get that as the default
> directory.
Sounds about right.
> That's probably not very nice.
Prolly not :-).
I think there are also circumstances where gnus *acquires* some
not-very-nice default-directory setting after it's already running.
It's possible that just happens via `M-x gnus' when the *Group* buffer
already exists --- I do that fairly often --- but I have the sense that
it has happened via subtler means.
> But would binding `default-directory' to "/" (or something) for the
> Gnus process buffers be the right thing?
Why not?
> Would that work?
I think so. How about binding it to something that is pretty much
guaranteed to exist as long as Emacs is running, say, the user's home
directory or user-emacs-directory?
> I have a feeling that this might be one of those "don't do that, then"
> things.
That'd be a shame; it bites me all the time. I can't see any reason not
to eliminate pitfalls for the unwary user.
--
Dave Abrahams
BoostPro Computing
http://www.boostpro.com
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Mon, 11 Jun 2012 02:12:02 GMT)
Full text and
rfc822 format available.
Message #20 received at 11666 <at> debbugs.gnu.org (full text, mbox):
Dave Abrahams <dave <at> boostpro.com> writes:
>> But would binding `default-directory' to "/" (or something) for the
>> Gnus process buffers be the right thing?
>
> Why not?
I was wondering whether that would bug out on some OS-es...
> I think so. How about binding it to something that is pretty much
> guaranteed to exist as long as Emacs is running, say, the user's home
> directory or user-emacs-directory?
I could imagine that "~" wouldn't be useful on some OS-es? But
`user-emacs-directory' would probably be safe?
I also wonder what buffers would need to have `default-directory' set
locally. The nntp and nnimap server buffers; sure. But what about the
smtpmail buffer, for instance?
Is this a problem that should be solved globally in Emacs, somehow?
Come to think of it, what is really the real problem? :-) What is it
that makes `default-directory' make any difference when opening a
network connection?
--
(domestic pets only, the antidote for overdose, milk.)
bloggy blog http://lars.ingebrigtsen.no/
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Tue, 12 Jun 2012 14:36:01 GMT)
Full text and
rfc822 format available.
Message #23 received at 11666 <at> debbugs.gnu.org (full text, mbox):
on Sun Jun 10 2012, Lars Magne Ingebrigtsen <larsi-AT-gnus.org> wrote:
> Dave Abrahams <dave <at> boostpro.com> writes:
>
>>> But would binding `default-directory' to "/" (or something) for the
>>> Gnus process buffers be the right thing?
>>
>> Why not?
>
> I was wondering whether that would bug out on some OS-es...
I suppose it's possible that on Windows the whole "current drive"
disappears, so "/" is just as incorrect as anything else...
>> I think so. How about binding it to something that is pretty much
>> guaranteed to exist as long as Emacs is running, say, the user's home
>> directory or user-emacs-directory?
>
> I could imagine that "~" wouldn't be useful on some OS-es?
Right, I meant ``the local equivalent of "~"''
> But `user-emacs-directory' would probably be safe?
Probably.
> I also wonder what buffers would need to have `default-directory' set
> locally. The nntp and nnimap server buffers; sure. But what about the
> smtpmail buffer, for instance?
I dunno.
> Is this a problem that should be solved globally in Emacs, somehow?
Dunno again. It's something that comes up for me with Gnus far, far
more often than with anything else I'm doing in Emacs... so solving it
in Emacs might be even more of an uphill battle than solving it in Gnus
is.
> Come to think of it, what is really the real problem? :-) What is it
> that makes `default-directory' make any difference when opening a
> network connection?
Great questions. Dunno #3. I just know it makes my life miserable :-)
--
Dave Abrahams
BoostPro Computing
http://www.boostpro.com
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Fri, 27 Jul 2012 15:50:02 GMT)
Full text and
rfc822 format available.
Message #26 received at 11666 <at> debbugs.gnu.org (full text, mbox):
For what it's worth, customizing gnus-default-directory to "~" makes
this problem go away for me.
--
Dave Abrahams
BoostPro Computing Software Development Training
http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Wed, 05 Sep 2012 18:37:01 GMT)
Full text and
rfc822 format available.
Message #29 received at 11666 <at> debbugs.gnu.org (full text, mbox):
Lars Magne Ingebrigtsen <larsi <at> gnus.org> writes:
> That's probably not very nice. But would binding `default-directory' to
> "/" (or something) for the Gnus process buffers be the right thing?
> Would that work?
>
> I have a feeling that this might be one of those "don't do that, then"
> things.
And I think that's the case. The expected thing for Emacs things is
that you get the current default directory when you start stuff. Like
if you say `M-x shell'. It would be odd if Gnus did something else
here.
So I'm closing this bug report.
--
(domestic pets only, the antidote for overdose, milk.)
http://lars.ingebrigtsen.no * Lars Magne Ingebrigtsen
Added tag(s) wontfix.
Request was from
Lars Ingebrigtsen <larsi <at> gnus.org>
to
control <at> debbugs.gnu.org
.
(Wed, 05 Sep 2012 18:37:01 GMT)
Full text and
rfc822 format available.
bug closed, send any further explanations to
11666 <at> debbugs.gnu.org and Dave Abrahams <dave <at> boostpro.com>
Request was from
Lars Ingebrigtsen <larsi <at> gnus.org>
to
control <at> debbugs.gnu.org
.
(Wed, 05 Sep 2012 18:37:02 GMT)
Full text and
rfc822 format available.
Information forwarded
to
bugs <at> gnus.org
:
bug#11666
; Package
gnus
.
(Wed, 05 Sep 2012 19:55:01 GMT)
Full text and
rfc822 format available.
Message #36 received at 11666 <at> debbugs.gnu.org (full text, mbox):
on Wed Sep 05 2012, Lars Ingebrigtsen <larsi-AT-gnus.org> wrote:
> Lars Magne Ingebrigtsen <larsi <at> gnus.org> writes:
>
>> That's probably not very nice. But would binding `default-directory' to
>> "/" (or something) for the Gnus process buffers be the right thing?
>> Would that work?
>>
>> I have a feeling that this might be one of those "don't do that, then"
>> things.
>
> And I think that's the case. The expected thing for Emacs things is
> that you get the current default directory when you start stuff. Like
> if you say `M-x shell'. It would be odd if Gnus did something else
> here.
>
> So I'm closing this bug report.
Well, that's your perogative of course. From my personal experience,
the pain caused by Gnus inadvertently started in a TRAMP directory (and
I think it can get switched somehow to using a TRAMP directory even if
it isn't started there) is massive, easy to achieve, and very hard to
understand, while customizing `gnus-default-directory' to "~" is
painless. I have never wished for my mail/news reader to hold onto some
other directory as its default, and since I've made the customization,
the problem hasn't returned. While this can't be called a bug, I think
it would be a service to your users if you'd protect them from this
problem by changing the default for `gnus-default-directory', even if it
causes some small apparent inconsistency with the way the rest of Emacs
works. BTW, if you get yourself in this situation, IIRC, you can't even
quit Gnus.
Anyway, more than 'nuff said.
--
Dave Abrahams
BoostPro Computing Software Development Training
http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost
bug archived.
Request was from
Debbugs Internal Request <help-debbugs <at> gnu.org>
to
internal_control <at> debbugs.gnu.org
.
(Thu, 04 Oct 2012 11:24:05 GMT)
Full text and
rfc822 format available.
This bug report was last modified 12 years and 267 days ago.
Previous Next
GNU bug tracking system
Copyright (C) 1999 Darren O. Benham,
1997,2003 nCipher Corporation Ltd,
1994-97 Ian Jackson.