Johannes Sixt [off-list ref] writes:
The configured refspecs are printed almost verbatim, i.e. both the local
and the remote branch name separated by a colon are printed; only the
prefix 'refs/heads/' is removed, like this:
Local branch(es) pushed with 'git push'
master refs/tags/*:refs/tags/* next:next
Signed-off-by: Johannes Sixt <redacted>
---
[Resent; I don't see the patch applied nor were there any comments.]
The hunk that reads the 'Push'es from the .git/remotes file is taken
almost verbatim from the corresponding 'Pull' clause and is untested.
Maybe while waiting for application nor comments you could have
tested and removed that comment?
On Sunday 18 March 2007 21:37, Junio C Hamano wrote:
Johannes Sixt [off-list ref] writes:
quoted
The hunk that reads the 'Push'es from the .git/remotes file is taken
almost verbatim from the corresponding 'Pull' clause and is untested.
Maybe while waiting for application nor comments you could have
tested and removed that comment?
Fair enough ;)
While waiting for application or comments again, I did test now, and the patch
does what I expect even for .git/remotes files, that is, it prints the refs
to be pushed just like it does for remotes noted in the config file.
-- Hannes
Johannes Sixt [off-list ref] writes:
On Sunday 18 March 2007 21:37, Junio C Hamano wrote:
quoted
Johannes Sixt [off-list ref] writes:
quoted
The hunk that reads the 'Push'es from the .git/remotes file is taken
almost verbatim from the corresponding 'Pull' clause and is untested.
Maybe while waiting for application nor comments you could have
tested and removed that comment?
Fair enough ;)
While waiting for application or comments again, I did test
now, and the patch does what I expect even for .git/remotes
files, that is, it prints the refs to be pushed just like it
does for remotes noted in the config file.
Thanks, and sorry for that comment as I was in a grumpy mood.
Let's queue it in 'next', as I do not see anything obviously
wrong with it [*1*].
Now let's find other people who find this new feature
desirable. I am personally neutral right now.
[Footnote]
*1* Actually I think the refspec list on "Push: " lines should
be split out into separate array elements, because
Push: A B C
is equivalent to
Push: A
Push: B
Push: C
This does not apply to "Pull: " lines, as
Pull: A B C
and
Pull: A
Pull: B
Pull: C
mean quite different things.
This distinction does not matter to the current callers, but I
was hoping that somebody would build "interactive" mode to edit
Push and Pull list, and when that happens, having A B C as
separate entries in @{$hash->{$name}{'PUSH'}} would be handier
to code.
On Monday 19 March 2007 23:36, Junio C Hamano wrote:
*1* Actually I think the refspec list on "Push: " lines should
be split out into separate array elements, because
Push: A B C
is equivalent to
Push: A
Push: B
Push: C
This does not apply to "Pull: " lines, as
Pull: A B C
and
Pull: A
Pull: B
Pull: C
mean quite different things.
This distinction does not matter to the current callers, but I
was hoping that somebody would build "interactive" mode to edit
Push and Pull list, and when that happens, having A B C as
separate entries in @{$hash->{$name}{'PUSH'}} would be handier
to code.
The patch does just that, I think. It is only that everything is *printed* on
one line, just like the tracked branches.
BTW, the .git/remotes parser does not allow multiple refspecs on one 'Push:'
line, if I understand it correctly (I'm looking at get_remotes_uri in
builtin-push.c).
-- Hannes