Re: gerrit for release branches (bots ftw?)

Hi Petr,

On Tue, Jun 19, 2012 at 06:58:33PM +0200, Petr Mladek wrote:
> > submitter:
> > - find cgit link
> > - write to mailing list with cgit link asking for review
>
> This handles only people with commit access and review for stable
> branches. It does not handle newbies.

Right, see subject. To send a patch for review to master anyone can just push
to refs/for/master on the gerrit repo -- everyone means everyone with a OpenID
somewhere, no fd.o account needed. People with commit access on fd.o will
continue to have direct push access to master on fd.o, if they pinged Norbert
or me about it after the first login as asked to in the mail one month ago.

> > reviewer:
> > - check out the release branch
> > - cherrypick commit on release branch by exctracting the commit-id from the cgit link
> > - review the commit
> > - push commit on the release branch
> > - mail to list that the stuff is pushed
> >
> > using gerrit its a bit simpler for the reviewer:
> > - check out the release branch
> > - gerrit-cherry-pick the commit
> > - push the commit to the release branch (gerrit will take care of notifying the submitter
>
> Hmm, this is not much encouraging:

This is not how it would end up finally, but just the transitional until in
about one month. I expect everyone to finally have logged into gerrit by then
so we could give them them direct commit rights (bypassing review) on master,
whoever is _still_ missing by then might indeed run into a minor interruption
in direct commit access.

>       + if there are conflict, who and how would resolve them in
>           gerrit?

I dont assume too many conflicts happening on patches for release branches as
those are already rebased to the release branch by the submitter (or his
friendly helper bot, see below). For the few remaining conflicts, gerrit will
just ask you to rebase/resolve and resubmit.
Note again, that everyone who has commit access on fd.o right now will have
direct commit acccess on gerrit, so as a fallback you will be able to just do
stuff as you did before on the gerrit repo.

Note also that most of the annoying complexity of the current setup is
transitional an will be gone once we have everything on the gerrit repo.
Everything your could do on the fd.o, you will be able to do on the gerrit repo
(if you got your setup in time). In ADDITION, you can do the code review stuff
there.

>       + you need to write a comment when approving the change anyway;
>           some more words might motive the volunteer to come again, so
>           the gerrit tool might be rather antisocial

Once gerrit owns the repo, you just write a +1/+2 review with a nice "thanks"
note and gerrit pushes the patch itself. Push to fd.o is just a temporary
workaround as long as we have not given all (or at least most) people who had
direct commit access on fd.o the same on gerrit, so that we can switch over.

> This is even worse. We rather need to lower the barriers and make things
> easier.

Yes, that what the bot is for.

> >  If not, we might alternatively have either a email or an IRC bot, which
> >  you send a message like:
> >
> >  releasereview 213d5355d78a0a690e366645d6416f4a8fe5e666 libroffice-3-5
> >
> > and it will do all of the above on its own. Opinions on that? Is it better to
> > have an IRC or an Email bot?
>
> This looks better. It similar to the message from the openSUSE Build
> Service that I attached to the other mail.

Also note that this would cherry-pick _before_ the review - so if there are
conflict, it can complain directly and not leave the dirty work to the
reviewer.

> I still see it optimistic. I am just not happy with the style of pushing
> this solution. You direct style brought strong offense from others.

Unfortunately, I saw no other way to bring this forward. I saw the sentiment "I
have no idea what it does, will not give any feedback on what I need it to be
to be usable to me, but I will retain the right to complain viciously about it"
for far too long. I expect people -- without haste, but in due time -- to have
a look at gerrit and give qualified feedback on what exactly is wrong with it
and where we still need tweaking.

After some good initial flameage, today provided some good qualified feedback
in the end -- while before, mails to this topic fell on no response. To prevent
further flameage, it might be helpful for all to get involved with the gerrit
discussion _early_ on -- it prevents a lot of further bad blood.

Best,

Bjoern