Re: test timeout observation

I'll look at this. Feels like these ones should fail immediately with
NotSupportedError.

...Mark

On Tue, Aug 30, 2016 at 8:43 AM, David Dorwin <ddorwin@google.com> wrote:

> While looking for examples of timeouts not being reported as failures, the
> persistent session tests for Clear Key stood out. Only one of the six shows
> a result in the "all" page.
>
> http://www.w3c-test.org/encrypted-media/clearkey-mp4-
> playback-persistent-license.html times out in all three browsers, and
> when I run it, it is just blank for a long time until I see:
>
>> Harness status: Error
>> Uncaught ReferenceError: qualifier is not defined
>
>
> http://www.w3c-test.org/encrypted-media/clearkey-mp4-
> playback-persistent-usage-record.html, in contrast, times out immediately
>
>> Harness status: OK
>> ...
>>
> Timeout org.w3.clearkey, successful playback, persistent-usage-record,
>> mp4, set src before setMediaKeys
>
> NotSupportedError: None of the requested configurations were supported.
>
>
> Observations:
>
>    - Harness ERROR is reported as a timeout.
>    - A timeout in the test is not reported.
>    - It's strange that the second test times out immediately even though
>    it fails as we would expect.
>    - These two playback should in theory be very similar but fail
>    differently. Maybe that is because the first loads the session before
>    playback (I assume).
>
>