taverna-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gale Naylor <Ga...@noventussolutions.com>
Subject Re: To pull or not to pull
Date Fri, 04 Nov 2016 18:52:35 GMT
Thanks for the explanation. :)

On Fri, Nov 4, 2016, 7:59 AM Stian Soiland-Reyes <stain@apache.org> wrote:

> It's very low tech, subscribe to commits (or look at archive), just look at
> quickly at every email and reply (back to dev@) if something is unclear.
> Revert git commit if something breaks.
>
> (But no-one are necessarily "on watch" so commits might go unnoticed.)
>
> Apache Commons work like that, which is a bit important there as they have
> granted all ASF committers write access and the Commons libraries have
> hundreds of thousands of users.
>
> You could also look at the git log locally or at GitHub, but we have many
> repositories and a couple of branches. This might be good before a release.
>
> On 4 Nov 2016 1:39 pm, "Gale Naylor" <GaleN@noventussolutions.com> wrote:
>
> > RTC for large changes and CTR for maintenance sounds good. I'm curious -
> > What is the process for CTR with the @commits list?
> >
> > On Fri, Nov 4, 2016, 5:31 AM Stian Soiland-Reyes <stain@apache.org>
> wrote:
> >
> > > CTR can be done with the commits@ list, but with git it can be way too
> > > noisy to follow or understand. Pull Request have very good UI for code
> > > review. I think we also have an ASF Gerrit instance we can use.
> > >
> > > How about we do RTC for large things or where a committer is not quite
> > > sure, but CTR for maintenance things?
> > >
> > > Also I would put an informal 1w deadline on any pull requests after
> which
> > > the committer just merges themselves.
> > >
> > > On 4 Nov 2016 11:48 am, "Andy Seaborne" <andy@apache.org> wrote:
> > >
> > > > There are two styles
> > > >
> > > > CTR - "Commit then review" -- its still up for review
> > > > RTC - "Review then commit"
> > > >
> > > > and hybrid forms such as committers doing CTR for small, "obvious"
> > things
> > > > (e.g. "Doh!" bug fixes; emergency repair) and RTC via PR when larger
> or
> > > the
> > > > committer is seeking review.
> > > >
> > > >         Andy
> > > >
> > > > On 04/11/16 04:11, Thilina Manamgoda wrote:
> > > >
> > > >> HI,
> > > >>
> > > >> I think this is a good idea. There may be mistakes in my code
> because
> > > >> still  i am not a expert thus code review is a good approach.
> > > >>
> > > >> Regards,
> > > >> Thilina
> > > >>
> > > >> On Thu, Nov 3, 2016 at 10:05 PM, Ian Dunlop <ianwdunlop@gmail.com>
> > > wrote:
> > > >>
> > > >> Hello,
> > > >>>
> > > >>> I think we need a policy decision on how to add new code to
> existing
> > > >>> projects. Apache Taverna commiters can just merge straight into
> > master
> > > >>> but perhaps we should have a policy of using pull requests so
that
> we
> > > >>> can review the code first. It might mean there is a slight overhead
> > but
> > > >>> maybe long term it means we get better code out of it. Myself
and
> > Sagar
> > > >>> were just having a chat about this with respect to the TavMob
> project
> > > so
> > > >>> it might not be appropriate for every repo.
> > > >>>
> > > >>> Discuss.
> > > >>>
> > > >>> Cheers,
> > > >>>
> > > >>> Ian
> > > >>>
> > > >>>
> > > >>>
> > > >>>
> > > >>
> > >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message