Bug fixes for basic playback test.

Bug fixes for basic playback test.

Does the following:
- Removes race between autoplay and explicit .play().
- Removes race between seeked and .play().
- Adds logging for several other failure type events.
- Moves the events.join() code into Python so better errors can be
  reported.
- Instead of opening a new tab for each video in the matrix, reuses the
  same tab for each playback.  Ensures abort and emptied events fire on
  src switch.

Reusing the same tab was necessary to prevent the video from hanging on
the Linux bot, as a nice bonus it shaves seconds off the runtime.

BUG=111603
TEST=Ran on Linux bot.

[Ami GONE FROM CHROMIUM, 2012-01-31 20:26:16]

lgtm

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/data/med...
File chrome/test/data/media/html/media_basic_playback.html (right):

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/data/med...
chrome/test/data/media/html/media_basic_playback.html:17: var events = [];
Init unnecessary since startTest() does it anyway.

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/data/med...
chrome/test/data/media/html/media_basic_playback.html:19: // List of events to
log.
I worry this will be flaky b/c it's overly specific; if the timing of the impl
changes, you may get a stalled event, or not, and there's no reason for the test
to assert on it.
I think you want this to get better diags for failures.
Maybe the right thing to do is to have a logEventsForDiagnostics and
logEventsForCorrectness with two different sets of events.

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/data/med...
chrome/test/data/media/html/media_basic_playback.html:21: 'abort', 'emptied',
'error', 'playing', 'stalled', 'suspend',
2-space indent

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/function...
File chrome/test/functional/media/media_basic_playback.py (right):

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/function...
chrome/test/functional/media/media_basic_playback.py:28: # Correct events for
the first iteration and every iteration thereafter.
s/correct/expected/ here and below.

[DaleCurtis, 2012-01-31 22:31:54]

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/data/med...
File chrome/test/data/media/html/media_basic_playback.html (right):

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/data/med...
chrome/test/data/media/html/media_basic_playback.html:17: var events = [];
On 2012/01/31 20:26:16, Ami Fischman wrote:
> Init unnecessary since startTest() does it anyway.

Done.

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/data/med...
chrome/test/data/media/html/media_basic_playback.html:19: // List of events to
log.
On 2012/01/31 20:26:16, Ami Fischman wrote:
> I worry this will be flaky b/c it's overly specific; if the timing of the impl
> changes, you may get a stalled event, or not, and there's no reason for the
test
> to assert on it.
> I think you want this to get better diags for failures.
> Maybe the right thing to do is to have a logEventsForDiagnostics and
> logEventsForCorrectness with two different sets of events.

Discussed offline. Events in this list are those which are expected plus those
which are unexpected and will have a negative impact on playback.

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/data/med...
chrome/test/data/media/html/media_basic_playback.html:21: 'abort', 'emptied',
'error', 'playing', 'stalled', 'suspend',
On 2012/01/31 20:26:16, Ami Fischman wrote:
> 2-space indent

Done.

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/function...
File chrome/test/functional/media/media_basic_playback.py (right):

https://chromiumcodereview.appspot.com/9233046/diff/3001/chrome/test/function...
chrome/test/functional/media/media_basic_playback.py:28: # Correct events for
the first iteration and every iteration thereafter.
On 2012/01/31 20:26:16, Ami Fischman wrote:
> s/correct/expected/ here and below.

Done.