[Telemetry] Disable forecast.io?tap as it is failing on reference builds.

[Telemetry] Disable forecast.io?tap as it is failing on reference builds.

PageActionNotSupported: Synthetic swipe not supported for this browser

TBR=rtenneti@chromium.org,dominikg@chromium.org
NOTRY=True
BUG=

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=245224

[tonyg, 2014-01-16 17:22:03]

Committed patchset #1 manually as r245224 (presubmit successful).

[Dominik Grewe, 2014-01-16 17:43:56]

On 2014/01/16 17:22:03, tonyg wrote:
> Committed patchset #1 manually as r245224 (presubmit successful).

I think even after the roll this won't work. We're trying to swipe and on
desktop systems this shouldn't do anything (at least that's what happening on my
Linux machine). And because we don't swipe, the wait_after condition will never
be true and the test times out.
So I'm wondering if we should run this test on desktop bots in the first place.
If we start running smoothness on this page set, many more pages will fail
because they use gestures that are meaningless on desktops without touch.

[tonyg, 2014-01-16 17:54:17]

On 2014/01/16 17:43:56, Dominik Grewe wrote:
> On 2014/01/16 17:22:03, tonyg wrote:
> > Committed patchset #1 manually as r245224 (presubmit successful).
> 
> I think even after the roll this won't work. We're trying to swipe and on
> desktop systems this shouldn't do anything (at least that's what happening on
my
> Linux machine). And because we don't swipe, the wait_after condition will
never
> be true and the test times out.
> So I'm wondering if we should run this test on desktop bots in the first
place.
> If we start running smoothness on this page set, many more pages will fail
> because they use gestures that are meaningless on desktops without touch.

I think there's a lot of value in running silk cases on desktop.

Perhaps we should have a key_silk_cases which runs everywhere and a
key_mobile_silk_cases which runs only on android.

[Dominik Grewe, 2014-01-16 18:32:00]

On 2014/01/16 17:54:17, tonyg wrote:
> On 2014/01/16 17:43:56, Dominik Grewe wrote:
> > On 2014/01/16 17:22:03, tonyg wrote:
> > > Committed patchset #1 manually as r245224 (presubmit successful).
> > 
> > I think even after the roll this won't work. We're trying to swipe and on
> > desktop systems this shouldn't do anything (at least that's what happening
on
> my
> > Linux machine). And because we don't swipe, the wait_after condition will
> never
> > be true and the test times out.
> > So I'm wondering if we should run this test on desktop bots in the first
> place.
> > If we start running smoothness on this page set, many more pages will fail
> > because they use gestures that are meaningless on desktops without touch.
> 
> I think there's a lot of value in running silk cases on desktop.
> 
> Perhaps we should have a key_silk_cases which runs everywhere and a
> key_mobile_silk_cases which runs only on android.

Yeah, maybe splitting it up would be more sensible if we care about desktop
performance of silk pages. Some pages, forecast.io or mobile-news, are clearly
mobile focused and make heavy use of touch-gestures that just don't work on
desktop.