Adds set noparent to webkit/OWNERS

Adds set noparent to webkit/OWNERS

Without this owners are inherited from third_party/OWNERS, which isn't what we
want here.

R=dglazkov@chromium.org, jochen@chromium.org
BUG=none

[jochen (gone - plz use gerrit), 2016-12-15 18:36:01]

jochen@chromium.org changed reviewers:
+ thakis@chromium.org

[jochen (gone - plz use gerrit), 2016-12-15 18:36:02]

we used to have that but Nico didn't like it for some reason I forgot

[Nico, 2016-12-15 18:42:22]

On 2016/12/15 18:36:02, jochen wrote:
> we used to have that but Nico didn't like it for some reason I forgot

I don't think we should use noparent more than absolutely necessary, and
third_party owners should all know that they shouldn't approve changes in
WebKit. Did the absence of this cause any issues? (Eventually this will move to
blink/, rendering the issue moot)

[sky, 2016-12-15 18:50:11]

On 2016/12/15 18:42:22, Nico (out Thu Dec 15) wrote:
> On 2016/12/15 18:36:02, jochen wrote:
> > we used to have that but Nico didn't like it for some reason I forgot
> 
> I don't think we should use noparent more than absolutely necessary, and
> third_party owners should all know that they shouldn't approve changes in
> WebKit. Did the absence of this cause any issues? (Eventually this will move
to
> blink/, rendering the issue moot)

John echoed that 'noparent' is an anit-pattern, so I'll close this. His
reasoning is that with 'noparent' you can't get a top-level owner to review
mechanical changes that happen to touch a lot of files. This is certainly true,
but to counter that in large mechanical changes more often than not a top-level
owner is going to rubber stamp, so that you could just as well TBR the change.

As to the specific case. It's not that I approved a change to webkit (I wouldn't
do that), it's that I was put on a review with a bunch of reviewers that
happened to touch webkit and didn't realize the author of the patch didn't have
a webkit reviewer. So, my LGTM for chrome also covered the webkit change. If we
had a 'noparent' that would not have happened.
https://codereview.chromium.org/2545693002/ is the change.