[webkit-dev] Breaking other ports

[James Robinson]

On Tue, Jan 29, 2013 at 6:29 PM, Ryosuke Niwa <rniwa at webkit.org> wrote:

> On Tue, Jan 29, 2013 at 5:46 PM, Adam Barth <abarth at webkit.org> wrote:
>
>> I understand that the new "rules of the road" for WebKit2 are that
>> contributors are allowed to break non-Apple ports.  However, those new
>> norms do not extend to WebCore.
>>
>> In <http://trac.webkit.org/changeset/138962>, Alexey broke form
>> resubmission confirmation in the Chromium port.  In
>> <https://bugs.webkit.org/show_bug.cgi?id=108214>, he is refusing to
>> fix the regression.  I don't view that as acceptable behavior from a
>> member of the WebKit community.
>>
>
> I understand your frustration but I don't think he intentionally broke the
> feature given that neither EWS nor commit queue detected the regression.
>

I don't see Adam or anyone else implying that he intentionally broke a
feature, but regardless when a patch to WebCore causes a regression in an
upstream port the options are to investigate and fix the regression or roll
the patch out.  If Alexey doesn't want to look into the regression, which
is valid, then the only clear option is to roll it out until someone who is
familiar with the breakage can look at it.


>  As far as I checked both of those bugs, there aren't even a Chromium test
> failing per his patch. So his patch broke something that had not been
> tested before.
>
> Also, given that the fix needs to Chromium port specific, it doesn't seem
> productive to ask Alexey, who presumably doesn't know much about Chromium
> port, to come up with a fix for it. And both of those bugs and
> https://code.google.com/p/chromium/issues/detail?id=172721 don't seem to
> contain any information as to why his patch broke the feature or what kind
> of changes he needs to make in order to fix it. Someone who knows how form
> resubmission confirmation works in Chromium needs to help him so that we
> can fix this regression.
>

If Alexey is interested in learning about this and fixing it, that's fine,
but in the meantime the regression can't be left in the tree.  As long as
I've been involved with the WebKit project there has been a strict
no-regression policy.

- James


> - R. Niwa
>
>
> _______________________________________________
> webkit-dev mailing list
> webkit-dev at lists.webkit.org
> https://lists.webkit.org/mailman/listinfo/webkit-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-dev/attachments/20130129/ac525dc2/attachment.html>