W3C home > Mailing lists > Public > public-css-testsuite@w3.org > February 2013

Re: Tests for CSS Regions + selection submitted for review

From: Rebecca Hauck <rhauck@adobe.com>
Date: Wed, 13 Feb 2013 15:44:14 -0800
To: Gérard Talbot <css21testsuite@gtalbot.org>
CC: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>, Mihnea-Vlad Ovidenie <mihnea@adobe.com>
Message-ID: <CD416154.287F4%rhauck@adobe.com>
Hi Gerard,

Thanks for taking the time to look at these. I probably should have
mentioned that Regions are not fully implemented or enabled. To run the
tests and see them pass, you can enable regions in Chrome in the
about://flags and Enabling 'Experimental Webkit features'  Also, since
these are prefixed properties, the -webkit- prefix will need to be added
to the flow-into and flow-from properties (2 references in the .css file,
one in the .js file).

As we'll be submitting more tests soon for Regions and Exclusions (also
requiring this additional step), how do you recommend we convey upon
submission to the W3C? Just in the email to this list or should there be a
comment in the test file?  If the latter, then the test would need to be
updated at some point when they're fully enabled - not sure that's

Either way, you should see failure if you run them in a UA where Regions
is not implemented or enabled. This is probably why you saw them fail
automatically. These tests are (I think) unique in that they only require
you the tester to move the mouse. You needn't verify anything (I.e., there
is no "the test passes ifŠ") because once the manual execution is done, JS
does the rest. As I mentioned, they're this way to use the same tests here
and in Webkit.  I considered adding a sentence along the lines of "All of
the tests below should pass." for additional clarity to the manual tester.
 Do you think that would help?

More of my comments are inline below.


On 2/13/13 10:35 AM, "Gérard Talbot" <css21testsuite@gtalbot.org> wrote:

>Le Mar 12 février 2013 18:46, Rebecca Hauck a écrit :
>> Hi,
>> I've submitted a suite of tests for CSS Regions that cover selection
>> behavior. Since they require mouse movement, they're manual, but the
>> verification is automated using testharness.js.
>As I load any of those 25 tests, the test results are already given
>(say, 2 Fail), are already in. And results don't change even after
>executing test instructions.

>> They're authored this
>> way in an attempt to import and submit the same files to Webkit, where
>> they run completely automated using WK internal stuff.
>And that's triggered by if(window.testRunner), right ?
Yes, that's right. I just added a couple extra comments denoting where
this happens. Again, I'm trying to have the same file(s) be used in both
places without having to maintain two separate copies of only slightly
different tests. I don't know if this is done anywhere else. I couldn¹t
find a good example to follow.

>> If anyone is interested and knowledgeable about Regions, I'd much
>> appreciate your time reviewing them.
>> Cheers,
>> -Rebecca
>line 21: <p>Click on the orange square, keep the mouse button down and
>drag it to the blue square, release the mouse over
>           the blue square.</p>
>This should mean that test results will be known once such selection
>task is performed.
>line 23:  <br />
>You can safely remove this <br /> from all 25 tests since <p> element
>have by default a bottom margin of 1em.


>line 42: <div id="log"></div>
>    </body>
>    <script>
>         runSelectionTest();
>    </script>
>v id="log"></div>
>    </body>
>    <script>
>         runSelectionTest([], true);
>    </script>
>line 56: </html>
>There must be some kind of copy-N-paste error or some search-N-replace
>error with this as I see approx. the same chunk of code in all 25 tests.
>I believe it should be
>line 42: <div id="log"></div>
>    <script>
>         runSelectionTest([], true);
>    </script>
>    </body>

Yep. Looks like this was a bad merge that happened in a mirrored github
repo I'm using for these tests. Oddly, running these in the Webkit harness
didn't show the issue. It's fixed now.

>+CC: Mihnea-Vlad Ovidenie
>Contributions to the CSS 2.1 test suite:
>CSS 2.1 Test suite RC6, March 23rd 2011:
>CSS 2.1 test suite harness:
>Contributing to to CSS 2.1 test suite:
Received on Wednesday, 13 February 2013 23:44:44 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:13:25 UTC