* git's is_inside_git_dir too strict?
@ 2007-03-06 15:41 Nguyen Thai Ngoc Duy
2007-03-06 21:46 ` Johannes Schindelin
0 siblings, 1 reply; 8+ messages in thread
From: Nguyen Thai Ngoc Duy @ 2007-03-06 15:41 UTC (permalink / raw)
To: git
Hi,
I have a working directory separated from its git repository. When I
work on this directory, I specify --git-dir argument (actually all
happens in script so I don't have to type that much). It used to work
but no longer since git 1.5.0. Complaints include "cannot be used
without a working tree" and "This operation must be run in a work
tree". Can I have the former behaviour back in some ways?
--
Duy
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: git's is_inside_git_dir too strict?
2007-03-06 15:41 git's is_inside_git_dir too strict? Nguyen Thai Ngoc Duy
@ 2007-03-06 21:46 ` Johannes Schindelin
2007-03-07 12:17 ` Nguyen Thai Ngoc Duy
0 siblings, 1 reply; 8+ messages in thread
From: Johannes Schindelin @ 2007-03-06 21:46 UTC (permalink / raw)
To: Nguyen Thai Ngoc Duy; +Cc: git
Hi,
On Tue, 6 Mar 2007, Nguyen Thai Ngoc Duy wrote:
> I have a working directory separated from its git repository. When I
> work on this directory, I specify --git-dir argument (actually all
> happens in script so I don't have to type that much). It used to work
> but no longer since git 1.5.0.
Yes, we tried to make sure a few common errors were harder to make.
> Complaints include "cannot be used without a working tree" and "This
> operation must be run in a work tree". Can I have the former behaviour
> back in some ways?
Well, there is a real problem with "This operation must be run in a work
tree". What operation is that? We tried to make it hard to overwrite
things in $GIT_DIR by running Git operations meant for the working tree,
and it seems like you are running one of them.
However, I think that is_bare_repository() really should check if
$GIT_DIR is a prefix of $(pwd). Because if it is not, we are likely to run
in a detached work tree.
If you agree, then is_inside_git_dir() might actually be a better way to
ask if the repo is bare than is_bare_repository().
Comments?
Ciao,
Dscho
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: git's is_inside_git_dir too strict?
2007-03-06 21:46 ` Johannes Schindelin
@ 2007-03-07 12:17 ` Nguyen Thai Ngoc Duy
2007-03-07 15:53 ` Shawn O. Pearce
2007-03-09 12:00 ` Matthias Lederhofer
0 siblings, 2 replies; 8+ messages in thread
From: Nguyen Thai Ngoc Duy @ 2007-03-07 12:17 UTC (permalink / raw)
To: Johannes Schindelin; +Cc: git
On 3/7/07, Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:
> > Complaints include "cannot be used without a working tree" and "This
> > operation must be run in a work tree". Can I have the former behaviour
> > back in some ways?
>
> Well, there is a real problem with "This operation must be run in a work
> tree". What operation is that? We tried to make it hard to overwrite
> things in $GIT_DIR by running Git operations meant for the working tree,
> and it seems like you are running one of them.
>From this command:
git --git-dir=/home/pclouds/blog/data.git ls-files --others --modified
2007-03-07T18_52_41.txt
It was run in /home/pclouds/blog/data, which is the working directory
of data.git.
By the way, is it plausible to add --git-workdir option to specify
working directory? With that option, I won't need to chdir to the
working directory, run git commands and chdir back.
>
> However, I think that is_bare_repository() really should check if
> $GIT_DIR is a prefix of $(pwd). Because if it is not, we are likely to run
> in a detached work tree.
>
> If you agree, then is_inside_git_dir() might actually be a better way to
> ask if the repo is bare than is_bare_repository().
>
> Comments?
I didn't follow git development lately so I'm afraid I'm incapable to
comment on this :-)
>
> Ciao,
> Dscho
>
>
--
Duy
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: git's is_inside_git_dir too strict?
2007-03-07 12:17 ` Nguyen Thai Ngoc Duy
@ 2007-03-07 15:53 ` Shawn O. Pearce
2007-03-07 16:12 ` Nguyen Thai Ngoc Duy
2007-03-09 12:00 ` Matthias Lederhofer
1 sibling, 1 reply; 8+ messages in thread
From: Shawn O. Pearce @ 2007-03-07 15:53 UTC (permalink / raw)
To: Nguyen Thai Ngoc Duy; +Cc: Johannes Schindelin, git
Nguyen Thai Ngoc Duy <pclouds@gmail.com> wrote:
> On 3/7/07, Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:
> >Well, there is a real problem with "This operation must be run in a work
> >tree". What operation is that? We tried to make it hard to overwrite
> >things in $GIT_DIR by running Git operations meant for the working tree,
> >and it seems like you are running one of them.
>
> From this command:
>
> git --git-dir=/home/pclouds/blog/data.git ls-files --others --modified
> 2007-03-07T18_52_41.txt
>
> It was run in /home/pclouds/blog/data, which is the working directory
> of data.git.
Why not just symlink it?
cd /home/pclouds/blog/data
ln -s ../data.git .git
But really, why is your Git repository for the working directory
outside of the working directory? Why not just move data.git to
data/.git ?
--
Shawn.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: git's is_inside_git_dir too strict?
2007-03-07 15:53 ` Shawn O. Pearce
@ 2007-03-07 16:12 ` Nguyen Thai Ngoc Duy
0 siblings, 0 replies; 8+ messages in thread
From: Nguyen Thai Ngoc Duy @ 2007-03-07 16:12 UTC (permalink / raw)
To: Shawn O. Pearce; +Cc: Johannes Schindelin, git
On 3/7/07, Shawn O. Pearce <spearce@spearce.org> wrote:
> Nguyen Thai Ngoc Duy <pclouds@gmail.com> wrote:
> > It was run in /home/pclouds/blog/data, which is the working directory
> > of data.git.
>
> Why not just symlink it?
>
> cd /home/pclouds/blog/data
> ln -s ../data.git .git
It's currently my workaround.
> But really, why is your Git repository for the working directory
> outside of the working directory? Why not just move data.git to
> data/.git ?
The directory "data" is automatically manipulated by scripts. I feel
it safer for data.git being outside of data so if I make any mistake,
data.git _might_ remain intact. You see, if I link data/.git to
data.git, my scripts' bugs might affect data.git too.
> --
> Shawn.
>
--
Duy
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: git's is_inside_git_dir too strict?
2007-03-07 12:17 ` Nguyen Thai Ngoc Duy
2007-03-07 15:53 ` Shawn O. Pearce
@ 2007-03-09 12:00 ` Matthias Lederhofer
2007-03-10 1:11 ` Junio C Hamano
2007-03-10 11:49 ` Nguyen Thai Ngoc Duy
1 sibling, 2 replies; 8+ messages in thread
From: Matthias Lederhofer @ 2007-03-09 12:00 UTC (permalink / raw)
To: Nguyen Thai Ngoc Duy; +Cc: git
Nguyen Thai Ngoc Duy <pclouds@gmail.com> wrote:
> By the way, is it plausible to add --git-workdir option to specify
> working directory? With that option, I won't need to chdir to the
> working directory, run git commands and chdir back.
http://article.gmane.org/gmane.comp.version-control.git/38382
Since I did not need this feature that much and no one replied that
there is any interest I did not look any further into it.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: git's is_inside_git_dir too strict?
2007-03-09 12:00 ` Matthias Lederhofer
@ 2007-03-10 1:11 ` Junio C Hamano
2007-03-10 11:49 ` Nguyen Thai Ngoc Duy
1 sibling, 0 replies; 8+ messages in thread
From: Junio C Hamano @ 2007-03-10 1:11 UTC (permalink / raw)
To: Matthias Lederhofer; +Cc: git, Nguyen Thai Ngoc Duy
Matthias Lederhofer <matled@gmx.net> writes:
> Nguyen Thai Ngoc Duy <pclouds@gmail.com> wrote:
>> By the way, is it plausible to add --git-workdir option to specify
>> working directory? With that option, I won't need to chdir to the
>> working directory, run git commands and chdir back.
>
> http://article.gmane.org/gmane.comp.version-control.git/38382
> Since I did not need this feature that much and no one replied that
> there is any interest I did not look any further into it.
FWIW, I was once in favor of that, and currently I am neutral.
I vaguely recall that I suggested something like that back when
I was still a contributor. The proposal was shot down by people
on the list, but I do not recall the details of why.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: git's is_inside_git_dir too strict?
2007-03-09 12:00 ` Matthias Lederhofer
2007-03-10 1:11 ` Junio C Hamano
@ 2007-03-10 11:49 ` Nguyen Thai Ngoc Duy
1 sibling, 0 replies; 8+ messages in thread
From: Nguyen Thai Ngoc Duy @ 2007-03-10 11:49 UTC (permalink / raw)
To: Matthias Lederhofer, git
On 3/9/07, Matthias Lederhofer <matled@gmx.net> wrote:
> Nguyen Thai Ngoc Duy <pclouds@gmail.com> wrote:
> > By the way, is it plausible to add --git-workdir option to specify
> > working directory? With that option, I won't need to chdir to the
> > working directory, run git commands and chdir back.
>
> http://article.gmane.org/gmane.comp.version-control.git/38382
> Since I did not need this feature that much and no one replied that
> there is any interest I did not look any further into it.
>
You obviously have a fan now :) Are you going to rework the patch? I
think I could finish it if you don't have time for it. FYI, the patch
seems not working ("Not below specified working directory")
--
Duy
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2007-03-10 11:50 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-03-06 15:41 git's is_inside_git_dir too strict? Nguyen Thai Ngoc Duy
2007-03-06 21:46 ` Johannes Schindelin
2007-03-07 12:17 ` Nguyen Thai Ngoc Duy
2007-03-07 15:53 ` Shawn O. Pearce
2007-03-07 16:12 ` Nguyen Thai Ngoc Duy
2007-03-09 12:00 ` Matthias Lederhofer
2007-03-10 1:11 ` Junio C Hamano
2007-03-10 11:49 ` Nguyen Thai Ngoc Duy
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.