Today wasn't a happy day for p...@. He did a seemingly innocuous roll
that broke the world: selenium, ui tests, layout tests. I am sure it
was stressful and probably added unnecessary gray to his hair.

Stuff like this happens to WebKit gardeners. We're used to breakages
upstream. That's the cost of being unforked, right?

The problem however, is that since we unforked, most of these
breakages and regressions are caused by fellow teammates. There are
two major issues:

1) writers of patches don't mention that the patch is two-sided and
will break Chromium if landed prematurely. I don't have to go far for
an example. Commit queue bot landed
http://trac.webkit.org/changeset/48659 a few minutes ago and broke the
canary. This means that the canary will be red all night and any
subsequent regressions will either not be noticed or create more
complications.

2) writers of patches don't test them properly. In Paul's case, it was
http://trac.webkit.org/changeset/48639, again by a teammate, and it
looks like the patch wasn't very thoroughly tested -- it showed a few
regressions in the canary, but because it had to do with V8 garbage
collection, the failures were intermittent and seemingly random.
However, landing it on trunk looked like a shrapnel blast.

This all means that we have to be a bit more diligent. We shouldn't be
paying these unnecessary costs. So, from now on, I propose a fairly
simple set of new rules:

1) if you write a Chromium patch for WebKit, you must provide URLs of
successful trybot runs with your submission. Chromium WebKit reviewers
will not r+ your patch otherwise. If you can't provide the trybot URLs
for some reason, please explain in detail why this patch could still
land.

2) if the two-sided patch you authored broke the canary and this
happened with no coordination with the WebKit gardener, you assume
WebKit gardening responsibility for the next 24 hours.

Hopefully, these amendments to our existing ways will bring a bit more
peace and stability to the Chromium land. What do you think?

:DG<

:DG<

--~--~---------~--~----~------------~-------~--~----~
Chromium Developers mailing list: [email protected] 
View archives, change email options, or unsubscribe: 
    http://groups.google.com/group/chromium-dev
-~----------~----~----~----~------~----~------~--~---

Reply via email to