Re: css3 line-break tests

Le Mar 22 janvier 2013 11:40, Glenn Adams a écrit :
> On Tue, Jan 22, 2013 at 1:09 AM, "Gérard Talbot"
> <css21testsuite@gtalbot.org
>> wrote:
>
>>
>> Le Mar 22 janvier 2013 1:33, Glenn Adams a écrit :
>> > On Mon, Jan 21, 2013 at 9:36 PM, "GÊrard Talbot"
>> > <css21testsuite@gtalbot.org
>> >> wrote:
>> >
>> >>
>> >> Le Lun 21 janvier 2013 21:28, Glenn Adams a Êcrit :
>> >> > FYI, I intend to submit a number of tests related to the CSS3
>> >> line-break
>> >> > property, particularly those that relate to testing the
>> >> functionality or
>> >> > fixes deriving from:
>> >> >
>> >> > [1] https://bugs.webkit.org/show_bug.cgi?id=89235
>> >> > [2] https://bugs.webkit.org/show_bug.cgi?id=105692
>> >> >
>> >> > Some preliminary reftests have already been included in [3], but
>> >> need to
>> >> > be
>> >> > finalized and retargeted for the CSS test suite format.
>> >> >
>> >> > [3]
>> >> https://bugs.webkit.org/attachment.cgi?id=172397&action=prettypatch
>> >> >
>> >> > Regards,
>> >> > Glenn
>> >>
>> >> Glenn,
>> >>
>> >> I suggest to make single, unique tests and not a test made of 3
>> single
>> >> tests.
>> >>
>> >> Avoid using the word "box" in the description; for most people, a
>> box
>> >> is
>> >> a 3-dimension object.
>> >>
>> >> If you refer to identical rendering, then shape and size of
>> rectangle
>> >> should be identical and content of rectangle should be identical.
>> >>
>> >> From bug report https://bugs.webkit.org/show_bug.cgi?id=105692
>> >>
>> >> Test:
>> >> https://bug-105692-attachments.webkit.org/attachment.cgi?id=180625
>> >>
>> >> Expected result:
>> >> https://bug-105692-attachments.webkit.org/attachment.cgi?id=182513
>> >>
>> >> is not what I would have understood or expected by 3 identical
>> pairs
>> >> of
>> >> boxes.
>> >>
>> >
>> > Thanks for that input. Regarding the term 'box', I disagree with you
>> > that
>> > it should be avoided. The term has a well defined meaning in CSS [1]
>> > that
>> > is not a 3 dimensional object. We can't very well and go retrofit
>> the
>> > fundamental CSS concepts here, I'm sure you'll agree.
>> >
>> > [1] http://www.w3.org/TR/CSS2/box.html
>>
>>
>> It is a well defined term in CSS. But going through all the tests
>> should
>> be doable by anyone: your friend, your mailman, your pastry cook, your
>> butcher, your neighbour. Outside CSS, a box is definitely a
>> 3-dimension
>> object. Therefore, the pass/fail condition sentence of all tests
>> should
>> try to avoid all HTML and CSS related vocabulary and terminology.
>>
>> "
>> That self-describing test instructions are accurate, precise, simple,
>> and self-explanatory. Your mother/husband/roommate/brother/bus driver
>> should be able to say whether the test passed or failed within a few
>> seconds, and not need to spend several minutes thinking or asking
>> questions.
>> "
>> http://wiki.csswg.org/test/css2.1/review-checklist#test-design
>>
>> There are tests which are saying:
>> - "there should be a 100x100 green block": your mother would certainly
>> ask "what's a 100x100 block and how can I measure it"? Some people
>> will
>> even wonder if there is math multiplication involved here.
>> - http://test.csswg.org/suites/css2.1/20110323/html4/c527-font-000.htm
>> Your bread baker would certainly ask "What's 13px and how big is it?
>> What's Helvetica anyway?"
>> -
>> http://test.csswg.org/suites/css2.1/20110323/html4/c531-color-000.htm
>> and
>> http://test.csswg.org/suites/css2.1/20110323/html4/c532-bgcolor-000.htm
>> "what's the meaning of 'line' in your tests?" would probably be asking
>> a
>> few of your friends.
>> -
>>
>> http://test.csswg.org/suites/css2.1/20110323/html4/background-image-cover-001.htm
>> "What's a 'box' exactly in those tests" would probably be asking
>> others
>> - etc.
>>
>> I personally have neighbour-friends who are retired and never ever
>> used
>> a computer and they couldn't understand/figure out some tests.
>>
>> Ideally, you want the pass/fail conditions sentence to be short, clear
>> and understandable by non-web-people.
>>
>
> I don't agree with the last part of this assertion. CSS test suites are
> not
> designed for use by the public,

Right now, anyone (your mother, bread baker, mailman, butcher, bus
driver, etc) can use the test harness
http://test.csswg.org/harness/
and "do" a few thousands of tests; there is no exclusion and no
requirement.

Intentionally reducing the (audience) scope of test suites to web-savvy
people is not desirable if there are 16 thousands (and growing) tests.
I believe 90% (or more) of all 9735 tests of the CSS 2.1 test suite can
be taken by ordinary people without problems; the rest, 10%, are more
complex, involve a special font to install or to use a user stylesheet.

> and it would be a distortion of
> established
> terminology to adopt layman terms that, by definition, are imprecise.

Even if we were to recommend and adopt the "box" term in the pass/fail
conditions sentence of tests, we would quickly find that it's too
"broad", too general and not specific, not very descriptive. I have
recommended [1] [2] to avoid and to ban the "box" term in the pass/fail
conditions sentence of tests and to use exclusively instead
rectangle|square|stripe|bar|line|grid and also to use qualifiers when
needed, when required by the test. In the absolute, "a large filled
green rectangle" may still involve a bit of interpretation and
subjectivity but it's definitely a better description - in the
relativity - than speaking of "a box".

The terms rectangle|square|stripe|bar|line|grid have been already used
by Ian Hickson in several of his tests and, from a geometry point of
view, rectangle, square, line and grid are clear, understandable and
reliable terms we can use and that we learned in high school.

Again, in my opinion, the pass/fail conditions sentence should just be a
short, clear and relevant verbal description of the expected layout
preceded with "Test passes if there [is|are] (...)".


[1]: Filled box or just black-bordered box?; More relevant shape
description
http://lists.w3.org/Archives/Public/public-css-testsuite/2009Sep/0014.html

[2]: 9-Recommended shape descriptors
http://lists.w3.org/Archives/Public/public-css-testsuite/2011Dec/0004.html

Gérard
-- 
Contributions to the CSS 2.1 test suite:
http://www.gtalbot.org/BrowserBugsSection/css21testsuite/

CSS 2.1 Test suite RC6, March 23rd 2011:
http://test.csswg.org/suites/css2.1/20110323/html4/toc.html

CSS 2.1 test suite harness:
http://test.csswg.org/harness/

Contributing to to CSS 2.1 test suite:
http://www.gtalbot.org/BrowserBugsSection/css21testsuite/web-authors-contributions-css21-testsuite.html

Received on Tuesday, 22 January 2013 22:52:09 UTC