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

Re: More test feedback

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Sun, 3 Oct 2010 14:18:30 -0700
Message-ID: <ef9b905e5f328e849df7b141b00ef897.squirrel@cp3.shieldhost.com>
To: "Simon Fraser" <smfr@me.com>
Cc: "Public CSS 2.1 test suite mailing list" <public-css-testsuite@w3.org>

> active-selector-002
> http://test.csswg.org/suites/css2.1/20101001/html4/active-selector-002.htm
> Test does not state whether visible red is a failure.


I agree with you that the red in that active-selector-002.htm testcase
does not help at all and will confuse unneedlessly.

> It's unclear what
> "these links" refers to.
> background-alpha-001
> http://test.csswg.org/suites/css2.1/20101001/html4/background-alpha-001.htm
> It's not possible to judge whether the alpha PNS is being rendered
> correctly, so the test is not self-documenting. Plus it makes me go
> blind.
> background-alpha-002
> http://test.csswg.org/suites/css2.1/20101001/html4/background-alpha-002.htm

That background-alpha-002.htm testcase definitely needs to be adjusted.

> background-attachment-004
> http://test.csswg.org/suites/css2.1/20101001/html4/background-attachment-004.htm
> Crazy test that is hard to judge. Not obvious what "flashes" should look
> like

may not be the most accurate testcase or obvious one (like a green
square testcase) but it is an OK testcase as far as I'm concerned.

> background-position-201
> http://test.csswg.org/suites/css2.1/20101001/html4/background-position-201.htm
> States "the diamond should be moving around the edge of the outer
> circle", but there is no outer circle visible

You must have javascript support enabled.
I have not tried this testcase locally.

> background-root-018
> http://test.csswg.org/suites/css2.1/20101001/html4/background-root-018.htm
> Text is virtually unreadable

Agreed. That background-root-018.htm requires adjustments. Here, the
needed code adjustments are doable and easy to do.

> border-001
> http://test.csswg.org/suites/css2.1/20101001/html4/border-001.htm
> Solid box or empty box?


> c525-font-wt-000
> http://test.csswg.org/suites/css2.1/20101001/html4/c525-font-wt-000.htm
> Does not state whether "This sentence should be very light" should be
> visibly lighter than normal.
> c527-font-001
> c527-font-002
> c527-font-004
> c527-font-005
> c527-font-007
> c527-font-008
> c527-font-009
> c527-font-10
> http://test.csswg.org/suites/css2.1/20101001/html4/c527-font-001.htm et
> al.
> These tests include words like "there should be 36px between lines". The
> tests need a reference marker of some kind to allow this to be judged.

I agree with you. I already said+explained so twice so far in the
mailing list:



> c5526c-display-000
> http://test.csswg.org/suites/css2.1/20101001/html4/c5526c-display-000.htm
> What am I supposed to do with this test? It needs a reference, if that's
> the point.


The reference ("from this reference rendering,") is mentioned in the
text of the testcase. This is acid1 test by the way..

> inline-block-003
> http://test.csswg.org/suites/css2.1/20101001/html4/inline-block-003.htm
> Like floats-126 etc, needs to say what should happen on window resize.
> Has a typo: "rustulting".

Agree that inline-block-003.htm needs to define what should happen on a
window resize.

> inlines-003
> http://test.csswg.org/suites/css2.1/20101001/html4/inlines-003.htm
> Not self-describing (what should happen on resize?).

I agree with your comment: that inlines-003.htm testcase requires IMO

> inlines-004
> inlines-005
> inlines-006
> http://test.csswg.org/suites/css2.1/20101001/html4/inlines-004.htm et
> al.
> "The paragraph of Xs below should remain perfectly aligned." Aligned
> with what?

Aligned with the baseline of letters, I would presumed.

> I think it means that the Xs should always lie on a grid.
> lang-selector-005
> http://test.csswg.org/suites/css2.1/20101001/html4/lang-selector-005.htm
> Missing the http flag?
> margin-collapse-160
> http://test.csswg.org/suites/css2.1/20101001/html4/margin-collapse-160.htm
> Should say that the first page is blank and 3 pages are expected (if,
> indeed, that is the intent).
> page-break-margins-000
> http://test.csswg.org/suites/css2.1/20101001/html4/page-break-margins-000.htm
> Test contains "[[UNFINISHED]]"
> text-transform-bicameral-009
> text-transform-bicameral-010
> http://test.csswg.org/suites/css2.1/20101001/html4/text-transform-bicameral-009.htm
> I think this test should attempt to specify a font that contains the
> glyphs on commons OSes. I have no idea what font name to enter to get
> the correct glyphs on Mac OS X.

That's also what I wondered myself too. What if I do not see the glyphs?
Then for which font should I be trying to load as a possible or suitable

One - bold, wild - idea would be for W3C to buy the rights to offer, use
code2000 font. Ideally, code2000 would be downloadable just like ahem
and code2000 would be a flag just like ahem .

"Code2000 is one of the larger fonts available and the latest build has
over 60000 glyphs."


and we can reach its author/creator.

That testcase is missing the dom flag:
<meta name="flags" content="">

> I have entered these notes in the wiki:
> <http://wiki.csswg.org/test/css2.1/issues>
> Simon

regards, Gérard
Contributions to the CSS 2.1 test suite:

CSS 2.1 test suite (RC2; October 1st 2010):

CSS 2.1 test suite contributors:
Received on Sunday, 3 October 2010 21:19:38 UTC

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