[Webkit-unassigned] [Bug 74776] Commit queue building & running tests on Chromium shouldn't land JSC or WK2 changes

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=74776





--- Comment #20 from Ryosuke Niwa <rniwa at webkit.org>  2013-01-10 20:57:12 PST ---
(In reply to comment #19)
> ---8<---
> Like the style-queue, the EWS is purely advisory.  Contributors and
> reviewers are free to ignore the warnings if they believe the warnings
> are erroneous or they decide (for whatever reason) to break the build
> in question.
> --->8---
> 
> People shouldn't need to wait for bots.  If there's a JSC or a WK2 problem with a patch, then the EWS bot for that configuration should commit-queue- the patch automatically, stoping the patch from being landed.  If the EWS bot for that configuration doesn't run in time, that means you need to invest more resources in that configuration to make it run fast enough.  That doesn't mean we should slow down development to wait for the bot.

I'm sorry but I don't follow. On one hand, you say that we shouldn't wait for bots and EWS is only advisory. On the other hand, you say that bots should be able to cq- on time to recent commit-queue from landing patches. Either EWS is advisory, in which case, it shouldn't be required for commit-queue to the right thing, or it should be a required system for commit-queue to work properly.

If your position is that we can't fix this bug or this bug is invalid, then I'll make an announcement in webkit-dev instead saying that we shouldn't be using commit-queue to land JSC/WK2 specific patches and/or I'll propose to make changes so that cq+ is ignored on JSC/WK specific patches.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.