- From: Chris Lilley <chris@w3.org>
- Date: Tue, 28 Feb 2017 15:57:16 -0500
- To: Gérard Talbot <css21testsuite@gtalbot.org>, Richard Ishida <ishida@w3.org>
- Cc: Public CSS Test suite mailing list <public-css-testsuite@w3.org>
On 2017-02-28 15:32, Gérard Talbot wrote:
> Le 2017-02-17 13:16, Chris Lilley a écrit :
>> On 2017-02-16 17:22, Gérard Talbot wrote:
>
>> If the test uses @font-face then it doesn't need the font flag.
>
> Record set of files declaring the font flag and using a .woff resource:
>
> http://test.csswg.org/shepherd/search/flag/font/content/.woff/
>
> Right now, 890 tests use a .woff or .woff2 resource and declare the
> font flag. I think this isn't a big error ...
From http://web-platform-tests.org/writing-tests/css-metadata.html
"font
Requires a specific font to be installed. (Details must be provided
and/or the font linked to in the test description)"
Unless the test requires a font to be installed (for example, as visible
fallback for test failure or success when the webfont does not load)
then it does not need, and should not use, the font flag.
>
> There is also 466 reference files that do the same.
>
> I am adding Richard Ishida as recipiendaries because Khaled Hosny
> (Nantonos)
Khaled and Nantonos are different individuals.
The woff2 testsuite is correctly using the font flag, because it *does*
require locally installed files. Some tests pass when the woff2 is
loaded (so the local font, if used, produces a fail); others pass if the
(malformed) woff2 is used, so pass if the local font is used.
--
Chris Lilley
@svgeesus
Technical Director @ W3C
W3C Strategy Team, Core Web Design
W3C Architecture & Technology Team, Core Web & Media
Received on Tuesday, 28 February 2017 20:57:30 UTC