On Thu, Jul 26, 2012 at 01:08:31PM +0200, Michael Stapelberg wrote:
Hi,
in the past, we have briefly discussed using Code Review Software such
as Gerrit, Review Board and others for i3. Back then, we came to the
conclusion that none of the available tools is easy and compelling
enough to be used as-is and that the effort of starting something is too
high to be justified (see below).
I have serious doubts about the effort not being justified anymore. Code
Reviews as we do them currently (in a mixture of comments via IRC, trac
or our mailing list) pose a high cognitive load for me and a frustrating
experience for (potential) contributors.
Quite a while ago, I wrote about the features and non-features which I’d
like to see in an ideal code review system:
http://www.featurequest.org/FRQ$CodeReview
Now my question is: Who of you would actively (!) help developing such a
system?
Best regards,
Michael
PS: In case anyone wonders why we don’t like existing software:
• It’s often complex to use. The situation usually is that you have a
patch and you want to get comments on it. Most existing tools require
you to sign up first and then fill in a lot of fields. For reviewers,
getting to the actual side-by-side diff takes far too many clicks (it
should be precisely 1).
• Most don’t allow anonymous submissions, but we really need that (as we
do in our bugtracker).
• Some don’t even accept the git format-patch output (Rietveld for
example).
• Some require you to give up control over your git repository (Gerrit).
