On 11/26/2012 05:51 AM, Gregory Szorc wrote:
On 11/25/12 7:29 PM, L. David Baron wrote:
On Monday 2012-11-26 04:21 +0100, Robert Kaiser wrote:
Justin Dolske schrieb:
I think we should consider jettisoning/rewriting that part of the
policy. It doesn't match what we've been doing in reality(*)
Yes, that's why we almost f***ed up 17.0 and needed to do a
last-minute reversion of patches that changed IIDs while on beta,
for example.
If we wouldn't have caught that problem with applying proper sr, our
problem lies deeper, of course.
I don't think that's relevant, since I don't think anybody's
proposing that we have re-review and re-superreview of patches for
aurora and beta.
I concur. IMO we should be applying static analysis, auditing, etc to catch
these types of issues. [Super] reviewers shouldn't need to waste brain
cycles on style checking, IID change requirements, etc.
+1
superreview should be needed for those web/addon facing API changes which
aren't coming from Web specs
(or if the spec is still very unstable and/or controversial).
I concede we currently generally don't do a very good job at this.
(not sure what 'this' refers to)
As a reviewer and someone who cares about "quality," this annoys me because I
know
it is something that could largely be solved through decent automation and
tools.
Yes. We certainly should have at least coding style checker, and uuid update
checker.
-Olli
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform
