[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Pushing patches to staging

From: David Kastrup
Subject: Re: Pushing patches to staging
Date: Sat, 12 Nov 2011 20:43:54 +0100
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/24.0.90 (gnu/linux)

Graham Percival <address@hidden> writes:

> On Sat, Nov 12, 2011 at 07:58:55PM +0100, David Kastrup wrote:
>> "Phil Holmes" <address@hidden> writes:
>> > At some point in the near future, I'll need to push a CG patch to
>> > update the bug squad rota.  My understanding is that this should now
>> > be done to dev/staging rather than master.  I've read David's note
>> > about how to do this, and wonder whether that's all necessary for a
>> > simple patch that's in my master branch and which I'm not too
>> > concerned about.
>> To push, you need to rebase to your push target, regardless of where you
>> are pushing to.  And after that, you do
> What does "rebase to your push target" mean?

You can't push if the pushed branch is not a descendant of the branch
you push to.

It's not like this a surprising new thing.  It's just the same as

> Assume that we all agree to use "staging" as of midnight tonight,
> UK time.  EXACTLY what command-line should we use?
>> git push origin HEAD:the-branch-i-push-to
> Can I assume that the EXACT command-line will be
> git push origin HEAD:staging
> ?

If HEAD _is_ the branch you want to push, yes.

If your own local development branch is called "staging" as well and is
tracking the remote staging (perhaps because you created it with
git checkout -b staging origin/staging) then

    git push origin staging

will push it.  If you just write

    git push

then every remote-tracking branch named the same as an existing branch
on origin will get pushed, so likely not just staging, but also master
(if you made any changes to it).

When in doubt, do

git push --dry-run -v -v whateverIwanttotryout

and see whether it affects the branches you think it does.

> Done.  Midnight tonight, we switch to "staging".  Could you start
> a branch like that, so that nobody tries to push to a branch that
> doesn't exist yet and get confused?

I put it up as a copy of current dev/staging.  Let's hope that we don't
have to merge changes from five different locations in the next few

David Kastrup

reply via email to

[Prev in Thread] Current Thread [Next in Thread]