[css3-multicol] Re: Progress on multicol historical issues / updated DoC

Le 2017-08-20 15:53, Gérard Talbot a écrit :
> Le 2017-08-20 08:49, Rachel Andrew a écrit :
> 
> [snipped]
> 
>> *STATE OF THE TEST SUITE*
>> 
>> Repeatedly through the years is mention of the state of the test
>> suite, the fact that tests were submitted and not reviewed etc. I
>> think my next move involves having a look at where that all is. Again,
>> I don’t have historical context.
> 
> I have reviewed at least 107 tests of the Multi column test suite:
> 
> http://test.csswg.org/shepherd/search/testcase/spec/css-multicol-1/reviewer/gtalbot/load/t107/#t16
> 
> Each reviewed test by me has a
> 
> <link rel="reviewer" title="Gérard Talbot"
> href="http://www.gtalbot.org/BrowserBugsSection/css21testsuite/" />
> <!-- YYYY-MM-DD -->
> 
> with the exact, real date.
> 
> The reviews were done between june 29th 2013 and october 2013:
> 
> http://lists.w3.org/Archives/Public/public-css-testsuite/2013Jul/
> 
> http://lists.w3.org/Archives/Public/public-css-testsuite/2013Aug/
> 
> http://lists.w3.org/Archives/Public/public-css-testsuite/2013Sep/
> 
> 
> About 99% of all of my work is still available here:
> 
> http://www.gtalbot.org/BrowserBugsSection/CSS3Multi-Columns/
> 
> http://www.gtalbot.org/BrowserBugsSection/CSS3Multi-Columns/Opera/
> 
> http://www.gtalbot.org/BrowserBugsSection/CSS3Multi-Columns/Opera/NewTests/
> 
> http://www.gtalbot.org/BrowserBugsSection/CSS3Multi-Columns/howcome/
> 
>> Were all the tests that ended up on WPT reviewed by anything other
>> than the original submitter?
> 
> By definition, a test can not be reviewed by its author.
> 
> Gérard


Additional update
-----------------

I have entered 8 test results today; now coverage is 100% for Gecko and 
Blink:

Blink: Passed 75.29%, Coverage 100%

Gecko: Passed 26.47%, Coverage 99.41%

The very low passed percentage for Gecko is due to the fact that, before 
2016 (or before Firefox 52), Gecko was *_not_* supporting unprefixed 
property names. Today, Firefox 52+ supports unprefixed multi-column 
property names. So, the 170 tests would have to be re-taken with Firefox 
in order to get reliable, updated test results.

- - - - - - - -

3 quick test reviews:

1-
wrong reference file in

http://test.csswg.org/harness/test/css-multicol-1_dev/multicol-fill-auto/

I am 99% sure I have reported that before.


2-
Pass-fail-conditions sentence is not too clear, a bit obscur when 
printed (if the test mentions "even when printed", then I would think 
that "paged" flag is missing):

http://test.csswg.org/harness/test/css-multicol-1_dev/multicol-height-002/

     html, body { height: 100%; }
     div {
     ...
     height: 150%;
     }
implies that <strong>only and exactly 2 pages</strong> must be printed 
in print preview or in a page media view.


3-

http://test.csswg.org/harness/test/css-multicol-1_dev/break-before-always-001/

line 7:  <meta name="flags" content="page">

It should be "paged" and not "page".

Also, we usually start such print-related test with the preliminary:

"
PREREQUISITE: Switch to print preview or a page media view of the page.

Test passes if only and exactly N pages are printed and if ...
"

and then follow-up with other pass-fail-conditions if the test requires 
more.

+CC: public-css-testsuite mailing list

Gérard

Received on Sunday, 20 August 2017 22:18:19 UTC