I'd like to point out that the proposal here for Architecture Review is
really not that different to what the ON (and some other consolidations)
already do for code review.
For integration of a bug fix we need a minimum of "3 sets of eyes":
1) the developer
2) the codereviewer who is not the developer
3) the CRT Advocate who approves the RTI who is not a
codereviewer
For an ARC fast-track under this new proposal we are actually only
asking for a minimum of "2 sets of eyes":
1) the project team (which could be one person)
2) the ARC case sponsor (which can be the same person)
3) a +1 from another ARC member which isn't the sponsor or
the project team.
Is that really too much to ask ?
I don't think so. Personally in the past I've read a lot of fast-tracks
where I've fully understood them been completely happy with them but
made no comment because our past policy was "silence is approval". The
problem with "silent approval" is you can't tell the difference between
"no problems but everyone read it", "no problems from the one person
that read it", and "nobody read it".
We already have escalation paths in place and this proposal really
doesn't change any of that. In fact it actually makes it easier in some
cases because it gives a heads up much earlier that there could be a
"skills gap" in review or "plain old apathy".
It is with this proposal and actually always has been the responsibility
of the ARC case sponsor (for a fast-track and "case owner" for a full
review) to ensure that the case is fully reviewed. All this proposal
does is formalise some language to make it easier for the case sponsor
to know when they need to step in and do something to keep the review
moving.
--
Darren J Moffat
