UMA histrograms to track MW support

UMA histrograms to track MW support

BUG=329265
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=252121

[Miguel Garcia, 2014-02-17 19:13:30]

Follow up downstream in
https://chrome-internal-review.googlesource.com/#/c/155089

[Mark P, 2014-02-18 18:52:21]

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7735: +    Android: Whether Chrome is
resumed on a multiwindow capable device.
clarification for my sake: this happens only on resume, not on regular startup,
right?

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7735: +    Android: Whether Chrome is
resumed on a multiwindow capable device.
"Android:" is confusing here.  It's clear you're saying that only Android
devices are multiwindow capable.  But you should also be clear whether it's only
Android that logs this bool or whether all mobile platforms log it (and only
Android could ever set it to true).

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7740: +  <summary>Android: % of the
screen used during a multisession window.</summary>
nit: % -> percent

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7740: +  <summary>Android: % of the
screen used during a multisession window.</summary>
Please say explicitly when this is emitted.  Per page load?  Per window switch? 
Per session resume?  etc.

Please briefly add a definition "multisession window" to this description.

Please clarify what you mean by "used".  Used by the Chrome application?  Used
by the particular window with focus?  Does this count browser chrome (e.g.,
omnibox, if not hidden at the moment)?

[Miguel Garcia, 2014-02-18 19:37:33]

Thanks for the quick review!

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7735: +    Android: Whether Chrome is
resumed on a multiwindow capable device.
My apologies, this actually happens in onpause, every time the activity goes
away. We measure most of the session stats in this way.

On 2014/02/18 18:52:21, Mark P wrote:
> clarification for my sake: this happens only on resume, not on regular
startup,
> right?

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7735: +    Android: Whether Chrome is
resumed on a multiwindow capable device.
On 2014/02/18 18:52:21, Mark P wrote:
> "Android:" is confusing here.  It's clear you're saying that only Android
> devices are multiwindow capable.  But you should also be clear whether it's
only
> Android that logs this bool or whether all mobile platforms log it (and only
> Android could ever set it to true).

Done.

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7740: +  <summary>Android: % of the
screen used during a multisession window.</summary>
On 2014/02/18 18:52:21, Mark P wrote:
> nit: % -> percent

Done.

https://codereview.chromium.org/150573004/diff/120001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7740: +  <summary>Android: % of the
screen used during a multisession window.</summary>
Done, let me know if this explanation is enough. Happy to elaborate offline if
needed.

On 2014/02/18 18:52:21, Mark P wrote:
> Please say explicitly when this is emitted.  Per page load?  Per window
switch? 
> Per session resume?  etc.
> 
> Please briefly add a definition "multisession window" to this description.
> 
> Please clarify what you mean by "used".  Used by the Chrome application?  Used
> by the particular window with focus?  Does this count browser chrome (e.g.,
> omnibox, if not hidden at the moment)?

[Mark P, 2014-02-18 19:59:45]

https://codereview.chromium.org/150573004/diff/170001/tools/metrics/histogram...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/150573004/diff/170001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7737: +    devices.
Thanks for the expanded description.  It's still a little confusing though. 

1.  Combing the first sentence ("Whether Chrome is used on a multiwindow capable
device") with the third ("Only emitted on Android MultiWindow devices."),
implies the only value emitted is true.  I must be misunderstanding something.

2.  I read "the activity" in the second sentence as Chrome.  Is that correct? 
If so, say so.  (If not, then clarify; this clarification may also end up
clarifying my comment #1.)

https://codereview.chromium.org/150573004/diff/170001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7744: +    window. Emitted every time
the activity is paused. Only emitted on Android
Ditto about clarifying the meaning of "the activity"

https://codereview.chromium.org/150573004/diff/170001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7747: +    A multiwindow session is any
period of time that Chrome was not used in full
nit: Judging by this clarification, you probably meant "multiwindow session" in
the paragraph above, not "multisession window".  (This was partially what made
me ask for an explanation in my earlier comments on this changelist)

[Miguel Garcia, 2014-02-19 09:18:05]

https://codereview.chromium.org/150573004/diff/170001/tools/metrics/histogram...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/150573004/diff/170001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7737: +    devices.
It should have said only emmitted as "true" on Android MW devices. We would emit
false if it weren't a MW capable device.

In the downstream review we have decided to eliminate this metric since we are
not really interested in non MW capable devices so I have removed it now.

On 2014/02/18 19:59:45, Mark P wrote:
> Thanks for the expanded description.  It's still a little confusing though. 
> 
> 1.  Combing the first sentence ("Whether Chrome is used on a multiwindow
capable
> device") with the third ("Only emitted on Android MultiWindow devices."),
> implies the only value emitted is true.  I must be misunderstanding something.
> 
> 2.  I read "the activity" in the second sentence as Chrome.  Is that correct? 
> If so, say so.  (If not, then clarify; this clarification may also end up
> clarifying my comment #1.)

https://codereview.chromium.org/150573004/diff/170001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7744: +    window. Emitted every time
the activity is paused. Only emitted on Android
On 2014/02/18 19:59:45, Mark P wrote:
> Ditto about clarifying the meaning of "the activity"

Done.

https://codereview.chromium.org/150573004/diff/170001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:7747: +    A multiwindow session is any
period of time that Chrome was not used in full
On 2014/02/18 19:59:45, Mark P wrote:
> nit: Judging by this clarification, you probably meant "multiwindow session"
in
> the paragraph above, not "multisession window".  (This was partially what made
> me ask for an explanation in my earlier comments on this changelist)

Done.

[Mark P, 2014-02-19 17:56:52]

lgtm

[Miguel Garcia, 2014-02-19 17:58:28]

The CQ bit was checked by miguelg@chromium.org

[commit-bot: I haz the power, 2014-02-19 18:15:36]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/miguelg@chromium.org/150573004/220001

[commit-bot: I haz the power, 2014-02-19 22:47:37]

Change committed as 252121