[OpenAFS-devel] Rehashing gerritt auto-build security

Jeffrey Hutzelman jhutz@cmu.edu
Fri, 05 Nov 2010 21:45:44 -0400


--On Friday, November 05, 2010 08:51:05 PM -0400 Jeffrey Altman 
<jaltman@secure-endpoints.com> wrote:

>  1. the "Committer" is on a trusted list or a member of the
>     trusted list has given a positive "Code Review"

When last we discussed this, I seem to recall talk about having a separate 
review category for this.  Determining that a submitted patchset is safe 
for buildbot is often considerably simpler than deciding whether it is any 
good.  Thus, a reviewer may wish to do the former but not the latter, to 
give the buildbot slaves time to build the change while it is still being 
reviewed.

Of course, this gets more complex if/when the buildbot starts _testing_ 
changes as well as building them.  After all, today if you submit a change 
to rx.c, I can probably declare it "safe to build" without reading it.