[OpenAFS-devel] Gerrit reviews and the rate of acceptance

[Derrick Brashear]

On Wed, Apr 11, 2012 at 1:09 PM, Matt W. Benjamin <matt@linuxbox.com> wrote=
:
> Hi Jeff,
>
> I don't think number of reviewers is the limiting factor. =A0For example,=
 I submitted a change to enhance Simon's contributed red-black tree. =A0Sim=
on himself reviewed it, had no objections to committing the change. =A0I do=
n't care if the change is committed, but I don't think you can legitimately=
 fault non-present reviewers.
>
> That's not to say we don't need more reviewers, we surely do. =A0If you'd=
 like me to review work more frequently, a good way would be to ask me to d=
o so. =A0Gerrit has a feature to "add reviewer" which can be used for this =
purpose.
>
> More broadly, I have written a lot of code for OpenAFS over the years, mu=
ch of it not committed.
> So have others (e.g., Hartmut). =A0There are simply too many reasons to l=
ist why those contributions cannot be committed. =A0Most saliently, I've be=
en told not to send more XCB changesets, dirformat changesets, other random=
 YFS-funded changes, for many good reasons. =A0They are all to do with larg=
er change management policy, such as that governing afs3-stds.

I don't see any reason changes not yet standardized can't be sent, but
simply as long as the policy of OpenAFS is to be compliant with what
is standardized as AFS3, the stuff can't be merged until the standards
are ratified, and will sit in gerrit, reviewers or not.


--=20
Derrick