W3C home > Mailing lists > Public > public-css-testsuite@w3.org > October 2010

Re: Time for tests with script to settle?

From: fantasai <fantasai.lists@inkedblade.net>
Date: Sat, 02 Oct 2010 21:01:31 -0700
Message-ID: <4CA8001B.5010801@inkedblade.net>
To: Simon Fraser <smfr@me.com>
CC: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>
On 10/01/2010 10:24 AM, Simon Fraser wrote:
> On Oct 1, 2010, at 10:10 AM, fantasai wrote:
>
>> On 09/30/2010 11:30 AM, L. David Baron wrote:
>>> Is there a known time that I can use for when to screenshot a test
>>> in the CSS 2.1 test suite after its onload fires?
>>>
>>> For example, if a testcase makes DOM manipulations 100ms after
>>> onload and requires those manipulations to have happened for the
>>> test to be valid, I can't screenshot it until after those 100ms have
>>> passed.
>>>
>>> Currently I'm running screenshots after a 2 second pause, but that
>>> makes doing the test runs really slow; I'd like to reduce it.
>>
>> I have no idea. Most of the ones you and bzbarsky submitted use
>> other means to force reflow. The others I've come across were
>> by hixie and Gabreiele; I'm not sure about those.
>>
>> The dynamic tests should all be flagged with the 'dom' flag,
>> which should help you with this... Can you grep for setTimeout
>> or something?
>
> I think the ideal solution would be for the test harness to inject some JS
> that the script can call, if present, when the test is complete.

If we're modifying the tests, I think Mozilla's reftest-wait
method makes more sense. A test with <html class="reftest-wait">
needs waiting, and once that class is removed it is ready for
a screenshot.

~fantasai
Received on Sunday, 3 October 2010 04:02:07 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 3 October 2010 04:02:18 GMT