Only include PageLoad.AbortTiming histograms for histograms with data.

Only include PageLoad.AbortTiming histograms for histograms with data.

We currently have root PageLoad.AbortTiming histograms in histograms.xml,
along with custom suffixes. We only log values for the suffixes,
however, which means the root histograms show up in the UMA dash with
'No data' next to them. This change expands the histograms to only
include histograms we actually log for, to prevent 'No data' rows from
showing in the UMA dashboard.

This is similar to https://codereview.chromium.org/2355093002 which
did the same transformation for PageLoad.Clients histograms.

BUG=656119

[Bryan McQuade, 2016-10-14 20:22:19]

The CQ bit was checked by bmcquade@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-10-14 20:22:46]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Bryan McQuade, 2016-10-14 20:24:01]

bmcquade@chromium.org changed reviewers:
+ csharrison@chromium.org, isherman@chromium.org

[Bryan McQuade, 2016-10-14 20:24:45]

Description was changed from

==========
Only include PageLoad.AbortTiming histograms for histograms with data.

We currently have root PgaeLoad.AbortTiming histograms in histograms.xml,
along with custom suffixes. We only log values for the suffixes,
however, which means the root histograms show up in the UMA dash with
'No data' next to them. This change expands the histograms to only
include histograms we actually log for, to prevent 'No data' rows from
showing in the UMA dashboard.

BUG=656119
==========

to

==========
Only include PageLoad.AbortTiming histograms for histograms with data.

We currently have root PgaeLoad.AbortTiming histograms in histograms.xml,
along with custom suffixes. We only log values for the suffixes,
however, which means the root histograms show up in the UMA dash with
'No data' next to them. This change expands the histograms to only
include histograms we actually log for, to prevent 'No data' rows from
showing in the UMA dashboard.

This is similar to https://codereview.chromium.org/2355093002 which
did the same transformation for PageLoad.Clients histograms.

BUG=656119
==========

[Bryan McQuade, 2016-10-14 20:25:05]

Description was changed from

==========
Only include PageLoad.AbortTiming histograms for histograms with data.

We currently have root PgaeLoad.AbortTiming histograms in histograms.xml,
along with custom suffixes. We only log values for the suffixes,
however, which means the root histograms show up in the UMA dash with
'No data' next to them. This change expands the histograms to only
include histograms we actually log for, to prevent 'No data' rows from
showing in the UMA dashboard.

This is similar to https://codereview.chromium.org/2355093002 which
did the same transformation for PageLoad.Clients histograms.

BUG=656119
==========

to

==========
Only include PageLoad.AbortTiming histograms for histograms with data.

We currently have root PageLoad.AbortTiming histograms in histograms.xml,
along with custom suffixes. We only log values for the suffixes,
however, which means the root histograms show up in the UMA dash with
'No data' next to them. This change expands the histograms to only
include histograms we actually log for, to prevent 'No data' rows from
showing in the UMA dashboard.

BUG=656119
==========

[Bryan McQuade, 2016-10-14 20:25:33]

Description was changed from

==========
Only include PageLoad.AbortTiming histograms for histograms with data.

We currently have root PageLoad.AbortTiming histograms in histograms.xml,
along with custom suffixes. We only log values for the suffixes,
however, which means the root histograms show up in the UMA dash with
'No data' next to them. This change expands the histograms to only
include histograms we actually log for, to prevent 'No data' rows from
showing in the UMA dashboard.

BUG=656119
==========

to

==========
Only include PageLoad.AbortTiming histograms for histograms with data.

We currently have root PageLoad.AbortTiming histograms in histograms.xml,
along with custom suffixes. We only log values for the suffixes,
however, which means the root histograms show up in the UMA dash with
'No data' next to them. This change expands the histograms to only
include histograms we actually log for, to prevent 'No data' rows from
showing in the UMA dashboard.

This is similar to https://codereview.chromium.org/2355093002 which
did the same transformation for PageLoad.Clients histograms.

BUG=656119
==========

[Bryan McQuade, 2016-10-14 20:25:44]

PTAL, thanks!

[Bryan McQuade, 2016-10-14 20:33:26]

The CQ bit was checked by bmcquade@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-10-14 20:33:30]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-14 21:26:32]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-10-14 21:26:33]

Dry run: This issue passed the CQ dry run.

[Charlie Harrison, 2016-10-14 22:44:38]

LGTM

[Ilya Sherman, 2016-10-14 23:41:23]

Hmm, could I encourage you to instead implement some mechanism to indicate that
the base histogram is unused?  It's a shame to duplicate a bunch of text in the
histograms.xml file just to work around this minor issue.

[Bryan McQuade, 2016-10-15 03:05:26]

On 2016/10/14 at 23:41:23, isherman wrote:
> Hmm, could I encourage you to instead implement some mechanism to indicate
that the base histogram is unused?  It's a shame to duplicate a bunch of text in
the histograms.xml file just to work around this minor issue.

Totally agree - I'd really like to have a mechanism to indicate the base
histogram is unused. Do you have a sense for how much work that would entail? If
it's straightforward I'm happy to do it. If it's more involved I think I'd
prefer to land this and file a bug, then circle back and revert this change once
support has been added.

Feel free to start an email thread to discuss what adding support for this might
look like, thanks!

[Ilya Sherman, 2016-10-17 23:46:24]

On 2016/10/15 03:05:26, Bryan McQuade wrote:
> On 2016/10/14 at 23:41:23, isherman wrote:
> > Hmm, could I encourage you to instead implement some mechanism to indicate
> that the base histogram is unused?  It's a shame to duplicate a bunch of text
in
> the histograms.xml file just to work around this minor issue.
> 
> Totally agree - I'd really like to have a mechanism to indicate the base
> histogram is unused. Do you have a sense for how much work that would entail?
If
> it's straightforward I'm happy to do it. If it's more involved I think I'd
> prefer to land this and file a bug, then circle back and revert this change
once
> support has been added.
> 
> Feel free to start an email thread to discuss what adding support for this
might
> look like, thanks!

I think it should be as easy as updating [1] to allow an empty suffix marked as
obsolete.  In fact, this might already be possible -- I don't know whether
anyone has ever tried it =)

[1]
https://cs.chromium.org/chromium/src/tools/metrics/histograms/extract_histogr...

[Bryan McQuade, 2016-10-18 09:58:16]

On 2016/10/17 at 23:46:24, isherman wrote:
> On 2016/10/15 03:05:26, Bryan McQuade wrote:
> > On 2016/10/14 at 23:41:23, isherman wrote:
> > > Hmm, could I encourage you to instead implement some mechanism to indicate
> > that the base histogram is unused?  It's a shame to duplicate a bunch of
text in
> > the histograms.xml file just to work around this minor issue.
> > 
> > Totally agree - I'd really like to have a mechanism to indicate the base
> > histogram is unused. Do you have a sense for how much work that would
entail? If
> > it's straightforward I'm happy to do it. If it's more involved I think I'd
> > prefer to land this and file a bug, then circle back and revert this change
once
> > support has been added.
> > 
> > Feel free to start an email thread to discuss what adding support for this
might
> > look like, thanks!
> 
> I think it should be as easy as updating [1] to allow an empty suffix marked
as obsolete.  In fact, this might already be possible -- I don't know whether
anyone has ever tried it =)
> 
> [1]
https://cs.chromium.org/chromium/src/tools/metrics/histograms/extract_histogr...

Excellent, thank you! I'll put together a patch to add support for this and send
it your way.

[Bryan McQuade, 2016-10-18 14:51:23]

On 2016/10/18 at 09:58:16, Bryan McQuade wrote:
> On 2016/10/17 at 23:46:24, isherman wrote:
> > On 2016/10/15 03:05:26, Bryan McQuade wrote:
> > > On 2016/10/14 at 23:41:23, isherman wrote:
> > > > Hmm, could I encourage you to instead implement some mechanism to
indicate
> > > that the base histogram is unused?  It's a shame to duplicate a bunch of
text in
> > > the histograms.xml file just to work around this minor issue.
> > > 
> > > Totally agree - I'd really like to have a mechanism to indicate the base
> > > histogram is unused. Do you have a sense for how much work that would
entail? If
> > > it's straightforward I'm happy to do it. If it's more involved I think I'd
> > > prefer to land this and file a bug, then circle back and revert this
change once
> > > support has been added.
> > > 
> > > Feel free to start an email thread to discuss what adding support for this
might
> > > look like, thanks!
> > 
> > I think it should be as easy as updating [1] to allow an empty suffix marked
as obsolete.  In fact, this might already be possible -- I don't know whether
anyone has ever tried it =)
> > 
> > [1]
https://cs.chromium.org/chromium/src/tools/metrics/histograms/extract_histogr...
> 
> Excellent, thank you! I'll put together a patch to add support for this and
send it your way.

For others who may be interested, the proposed change to add support for 'base
histograms' is here: https://codereview.chromium.org/2426473006