Re: [PATCH v2 0/9] Introduce publish tracking branch
- From
Felipe Contreras <felipe.contreras@gmail.com>
- Date
- Apr 11, 2014, 19:16 UTC
- Message-ID
- <53483f7984cb8_2c1f6e72ecc0@nysa.notmuch>
- In-Reply-To
- <vpqd2gng4nl.fsf@anie.imag.fr>
Matthieu Moy wrote:
Show 14 quoted lines
> Felipe Contreras <felipe.contreras@gmail.com> writes: > > > My patch series only affects push.default=simple, perhaps you have a > > different configuration. > > Good catch. I have push.default=upstream (essentially for compatibility > with old Git versions, I'd prefer simple actually). > > > Maybe we want the publish branch to override any push.default, so: > > Not sure actually. If a user says "push.default=upstream", it seems > weird to push to something other than upstream indeed. What's clear to > me is that your patch in its current form clearly makes "simple" a much > better default than "upstream" (good news, it it the default!).
As you said in another email; that's just the default. If the user explicitely told Git to use certain branch (git push -p), Git should use that branch.
Show 14 quoted lines
> That said, the advice given by "git status" is clearly wrong: > > > > $ git status > > > On branch master > > > Your branch is ahead of 'origin/new' by 4 commits. > > > (use "git push" to publish your local commits) > > It should say (use "git push origin new" to publish your local commits) > with push.default=upstream and the current behavior of the patch. > > Perhaps argumentless "git push" could warn when push.default=upstream > and branch.<name>.publish is configured, I'm not sure. > > Sorry, more questions and "I'm not sure" than actual suggestion :-(.
I believe in v3 of the patch series "git push" will actually do it correctly.
Cheers.
-- Felipe Contreras