Revert 84197 - Add one-time randomization support for FieldTrial, and the ability to

Revert 84197 - Add one-time randomization support for FieldTrial, and the ability to
disable field trials. I am going to have a need for both soon.

Cleaning up some comments about empty trial names, adding static
method TrialExists() and simplifying many call sites by using this
method.

While I'm in there and needing base/OWNERS approval, add an OWNERS
file for base/metrics that adds jar@chromium.org as an owner for that
directory.

BUG=none
TEST=base_unittests
TBR=jam@chromium.org
R=jar@chromium.org,phajdan.jr@chromium.org,mark@chromium.org,wtc@chromium.org

Reason for revert:  See http://crbug.com/81750

BUG=81750
TBR=joi@chromium.org
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=84373

[Jói, 2011-05-06 00:48:22]

Matt, thanks for letting me know about this problem (http://crbug.com/81750). 
It was introduced by the change because initialization of FieldTrialList is
deferred compared to before the change, and neither I didn't notice the field
trial you mentioned is initialized earlier in official builds.

BTW, to answer your remark regarding use of TBR for this change, it looks like
using the --tbr flag to [git cl dcommit] eliminated the actual list of
reviewers.  The change had several rounds of review by jar@chromium.org, plus
two reviews for OWNERS approval, and the --tbr was just for the last missing
one-line OWNERS approval that I hadn't realized I needed.  I'm also not sure why
the review link was missing, perhaps for the same reason (--tbr); the correct
review link is http://codereview.chromium.org/6883102/

Cheers,
Jói

[Jói, 2011-05-06 00:54:30]

I've committed the revert (r84373), will re-land the original change with fix
tomorrow once I have time to test (IIRC there is a way to build an offical-like
build locally, I will do that to test the case you were seeing).

Cheers,
Jói



On 2011/05/06 00:48:22, Jói wrote:
> Matt, thanks for letting me know about this problem (http://crbug.com/81750). 
> It was introduced by the change because initialization of FieldTrialList is
> deferred compared to before the change, and neither I didn't notice the field
> trial you mentioned is initialized earlier in official builds.
> 
> BTW, to answer your remark regarding use of TBR for this change, it looks like
> using the --tbr flag to [git cl dcommit] eliminated the actual list of
> reviewers.  The change had several rounds of review by
mailto:jar@chromium.org, plus
> two reviews for OWNERS approval, and the --tbr was just for the last missing
> one-line OWNERS approval that I hadn't realized I needed.  I'm also not sure
why
> the review link was missing, perhaps for the same reason (--tbr); the correct
> review link is http://codereview.chromium.org/6883102/
> 
> Cheers,
> Jói

[mattm, 2011-05-06 01:00:41]

I was using GYP_DEFINES='branding=Chrome buildtype=Official'.  You need
src-internal checked out too.

On 2011/05/06 00:54:30, Jói wrote:
> I've committed the revert (r84373), will re-land the original change with fix
> tomorrow once I have time to test (IIRC there is a way to build an
offical-like
> build locally, I will do that to test the case you were seeing).
> 
> Cheers,
> Jói
> 
> 
> 
> On 2011/05/06 00:48:22, Jói wrote:
> > Matt, thanks for letting me know about this problem
(http://crbug.com/81750). 
> > It was introduced by the change because initialization of FieldTrialList is
> > deferred compared to before the change, and neither I didn't notice the
field
> > trial you mentioned is initialized earlier in official builds.
> > 
> > BTW, to answer your remark regarding use of TBR for this change, it looks
like
> > using the --tbr flag to [git cl dcommit] eliminated the actual list of
> > reviewers.  The change had several rounds of review by
> mailto:jar@chromium.org, plus
> > two reviews for OWNERS approval, and the --tbr was just for the last missing
> > one-line OWNERS approval that I hadn't realized I needed.  I'm also not sure
> why
> > the review link was missing, perhaps for the same reason (--tbr); the
correct
> > review link is http://codereview.chromium.org/6883102/
> > 
> > Cheers,
> > Jói