Re: [PATCH] push: error out when the "upstream" semantics does not make sense

From: Matthieu Moy <matthieu.moy@grenoble-inp.fr>
Date: Apr 6, 2012, 07:16 UTC
Message-ID: <vpqiphd49fx.fsf@bauges.imag.fr>
In-Reply-To: <7vy5q95354.fsf@alter.siamese.dyndns.org>

Junio C Hamano <gitster@pobox.com> writes:

Show 13 quoted lines

>> "To push the current branch to this remote, run:
>>
>>   git push <remote> <branch>
>>
>> "
>
> I am afraid that the above advice is a lot worse than leaving it unsaid.
>
> We are in no position to assume that the user wanted the "current"
> semantics when we issue this message.  Otherwise, we would be better off
> switching the default semantics to "current", not "upstream".  But the
> working assumption in this series is that "upstream" is an improvement
> over "current", no?

I'm not sure. I do support "upstream", but I also think that in many (most?) cases, the user will want to set "upstream" to point to the branch with the same name. If there is any confusion between "current" and "upstream", then avoiding situations where they do something different is not stupid. This is already what we do when there is no upstream configured at all:

  $ git push origin
  fatal: The current branch master has no upstream branch.
  To push the current branch and set the remote as upstream, use
  
      git push --set-upstream origin master

OTOH, if the user sees this message, he already has several remotes configured, and we can probably expect him not to be a total newbie. Then suggesting

  "git push --set-upstream %s %s:<remote-branch-name>", remote->name, branch->name

would make sense.

-- 
Matthieu Moy
http://www-verimag.imag.fr/~moy/