Agility

Perry E. Metzger pmetzger at macports.org
Wed Apr 25 20:13:24 UTC 2018


On Wed, 25 Apr 2018 21:40:46 +0200 Rainer Müller
<raimue at macports.org> wrote:
> This seems a problem that only occurs in the context of pull
> requests.
> 
> I think the openmaintainer policy was intended for changes coming
> from other project members and not necessarily for patches coming
> from external contributors. openmaintainer allows other project
> members to push changes directly to master, without explicit
> approval by the maintainer.
> 
> Pull requests are like the original workflow in which patches are
> attached to a ticket on Trac. That ticket would either be handled
> by the maintainer or run into maintainer timeout. It does not seem
> common that another project member takes a patch from a ticket and
> pushes the change under the openmaintainer policy before the
> maintainer timeout is over.
> 
> Of course, in the original workflow that also involves more than
> just the press of a button...

And the original workflow didn't have a CI system involved. Ours is
less than perfect of course, and doesn't run any tests.

So the question is, now that GitHub sort of encourages a more rapid
paced development cycle, what are the minuses of going along with
that? The most obvious one is, I think, the possibility of breaking
things that are messy to fix. (The positives are obvious and needn't
be belabored.)

Perry
-- 
Perry E. Metzger		pmetzger at macports.org


More information about the macports-dev mailing list