Omnibox: Create HistoryQuickProvider new scoring field trial.

Omnibox: Create HistoryQuickProvider new scoring field trial.

And add a beacon to the histograms to allow identification of users in it.

isherman: for field trial and histogram stuff (all files)
sky: for OWNERs approval history stuff

BUG=
TEST=by hand, changed probabilities of field trial, looked at omnibox behavior and histogram behavior


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=142451

[Ilya Sherman, 2012-06-15 01:24:58]

LGTM with some suggestions inline

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
File chrome/browser/autocomplete/autocomplete_field_trial.cc (right):

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
chrome/browser/autocomplete/autocomplete_field_trial.cc:159: if
(!base::FieldTrialList::TrialExists(kHQPNewScoringFieldTrialName))
nit: Perhaps if (!InHQPNewScoringFieldTrial())?

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
File chrome/browser/autocomplete/autocomplete_field_trial.h (right):

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
chrome/browser/autocomplete/autocomplete_field_trial.h:60: static bool
InHQPNewScoringFieldTrial();
nit: Is there any more descriptive name that we can give to this, rather than
just "new" scoring?  I'm imagining that you'll want to run additional field
trials tweaking the scoring after this one, and it'll get confusing if they're
all some variant of "new".

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
chrome/browser/autocomplete/autocomplete_field_trial.h:62: // Returns whether
the user should get the experiment setup or
nit: "experiment" -> "experimental"?

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
chrome/browser/autocomplete/autocomplete_field_trial.h:64: // group uses "new
scoring" (a complex multiplicate calculation that,
nit: Is "multiplicate" a typo or a fancy word I don't know?

[Mark P, 2012-06-15 12:26:12]

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
File chrome/browser/autocomplete/autocomplete_field_trial.cc (right):

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
chrome/browser/autocomplete/autocomplete_field_trial.cc:159: if
(!base::FieldTrialList::TrialExists(kHQPNewScoringFieldTrialName))
On 2012/06/15 01:24:58, Ilya Sherman wrote:
> nit: Perhaps if (!InHQPNewScoringFieldTrial())?

Done.

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
File chrome/browser/autocomplete/autocomplete_field_trial.h (right):

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
chrome/browser/autocomplete/autocomplete_field_trial.h:60: static bool
InHQPNewScoringFieldTrial();
On 2012/06/15 01:24:58, Ilya Sherman wrote:
> nit: Is there any more descriptive name that we can give to this, rather than
> just "new" scoring?  I'm imagining that you'll want to run additional field
> trials tweaking the scoring after this one, and it'll get confusing if they're
> all some variant of "new".

I thought about this.  The closest I got is Multiplicative, but I don't think
that explains much more.

If this experiment works, we'll switch over and get rid of the notion of "new".

If it doesn't, you're right; I'm going to have to run experiments like
NewScoringSquashPopularity and such, but having a different, more descriptive
name doesn't make it any easier to run a MoreDescriptiveNameSquashPopularity
experiment.

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
chrome/browser/autocomplete/autocomplete_field_trial.h:62: // Returns whether
the user should get the experiment setup or
On 2012/06/15 01:24:58, Ilya Sherman wrote:
> nit: "experiment" -> "experimental"?

Done.

https://chromiumcodereview.appspot.com/10532149/diff/1/chrome/browser/autocom...
chrome/browser/autocomplete/autocomplete_field_trial.h:64: // group uses "new
scoring" (a complex multiplicate calculation that,
On 2012/06/15 01:24:58, Ilya Sherman wrote:
> nit: Is "multiplicate" a typo or a fancy word I don't know?

Typo.  Fixed.

[sky, 2012-06-15 16:41:28]

LGTM

https://chromiumcodereview.appspot.com/10532149/diff/5001/chrome/browser/hist...
File chrome/browser/history/scored_history_match.h (right):

https://chromiumcodereview.appspot.com/10532149/diff/5001/chrome/browser/hist...
chrome/browser/history/scored_history_match.h:134: static bool initialized;
I missed this earlier, these should end with _

[Mark P, 2012-06-15 18:01:19]

https://chromiumcodereview.appspot.com/10532149/diff/5001/chrome/browser/hist...
File chrome/browser/history/scored_history_match.h (right):

https://chromiumcodereview.appspot.com/10532149/diff/5001/chrome/browser/hist...
chrome/browser/history/scored_history_match.h:134: static bool initialized;
On 2012/06/15 16:41:28, sky wrote:
> I missed this earlier, these should end with _

Fixed.

[commit-bot: I haz the power, 2012-06-15 18:04:27]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mpearson@chromium.org/10532149/8001

[commit-bot: I haz the power, 2012-06-15 19:44:35]

Change committed as 142451