* possible regression (or me abusing git) in v1.5.5
@ 2008-04-11 19:21 tarmigan+git
2008-04-11 21:06 ` Junio C Hamano
0 siblings, 1 reply; 2+ messages in thread
From: tarmigan+git @ 2008-04-11 19:21 UTC (permalink / raw)
To: Junio C Hamano, git
Hello,
After upgrading to v1.5.5, I noticed a change in the way that git
treats symbolic links.
Git will follow symbolic links to directories when they were added as follows
$ git init
$ mkdir foo
$ touch foo/bar
$ ln -s foo foo1
$ git add foo1/
# Note the trailing slash above.
In v1.5.5 (but not v1.5.4.5 and earlier),
$ git add -u
will remove foo1/bar from the index.
I have bisected this down to f58dbf (diff-files: careful when
inspecting work tree items).
The ability for git to follow links (instead of recording them) was a
useful feature for me, but I may have been abusing this "feature". If
that is the consensus, it might make sense to have
$ git add foo1
and
$ git add foo1/
behave the same way?
Thanks,
Tarmigan
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: possible regression (or me abusing git) in v1.5.5
2008-04-11 19:21 possible regression (or me abusing git) in v1.5.5 tarmigan+git
@ 2008-04-11 21:06 ` Junio C Hamano
0 siblings, 0 replies; 2+ messages in thread
From: Junio C Hamano @ 2008-04-11 21:06 UTC (permalink / raw)
To: tarmigan+git; +Cc: git
tarmigan+git@gmail.com writes:
> After upgrading to v1.5.5, I noticed a change in the way that git
> treats symbolic links.
>
> Git will follow symbolic links to directories when they were added as follows
>
> $ git init
> $ mkdir foo
> $ touch foo/bar
> $ ln -s foo foo1
> $ git add foo1/
In this step "git add" adds "foo1/bar", which is a bug that is shared
between 1.5.5 and previous versions.
Recent "git add -u" after that was taught to notice this bogosity (it
actually is git-diff-* family that was fixed), and fixes it.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2008-04-11 21:07 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-04-11 19:21 possible regression (or me abusing git) in v1.5.5 tarmigan+git
2008-04-11 21:06 ` Junio C Hamano
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).