[Phpmyadmin-devel] Double-checking a commit
Marc Delisle
marc at infomarc.info
Fri May 17 12:27:21 CEST 2013
Le 2013-05-16 23:11, Isaac Bennetch a écrit :
>
> On Thu, May 16, 2013 at 4:31 PM, Isaac Bennetch <bennetch at gmail.com
> <mailto:bennetch at gmail.com>> wrote:
>
>
> On Thu, May 16, 2013 at 4:08 PM, Marc Delisle <marc at infomarc.info
> <mailto:marc at infomarc.info>> wrote:
>
> Isaac Bennetch a écrit :
> > Hi, I just merged the pull request for AES_ENCRYPT [1], which
> required a
> > little bit of manual work to complete the merge. I thought
> everything went
> > well, except the commit history [2] looks strange to me.
> Would someone mind
> > double-checking that I didn't inadvertently do something
> wrong? I pretty
> > much followed the GitHub instructions, but this is the first
> time I've done
> > a merge where I committed back to the project repository
> rather than my
> > private one, so I'm a bit extra cautious.
> >
> > Thanks
> >
> > 1 - https://github.com/phpmyadmin/phpmyadmin/pull/290
> > 2 -
> >
> https://github.com/phpmyadmin/phpmyadmin/commit/8fbad56812f0b5b0730b07961c7db329bdef9dae
>
> Hi Isaac,
> indeed it looks strange. I would say that you merged master into his
> feature branch before merging back to master.
>
>
> Thanks. In the interest of not making this worse, I think the right
> thing to do is git revert 16febdb. That will take me back to the
> last commit before my mistake, "There should be message variable as
> well". Then I have to commit and push that revert. Then I can redo
> the ayusun merge correctly. I just want to make positively sure that
> I do the revert correctly.
>
>
> After a bit more testing, I think the even better response is to git
> reset --hard 16feb and push that change. One part of the git
> documentation suggests one, another part suggests the other. I'm not
> clear on what the differences are.
>
> The problem is that when I do a git diff and git status to make sure of
> what I'm about to push, it looks right...but then again, it also did
> when I messed things up.
>
> Any opinions?
I think it's too late to do a reset and push this, because there have
been other commits since (other than yours). If we lost nothing in the
source code, I would not care about the history.
--
Marc Delisle
http://infomarc.info
More information about the Developers
mailing list