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

Re: invalid tests about range of integer values

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Tue, 5 Oct 2010 09:02:22 -0700
Message-ID: <aab4a0439aa9ccac39f593fcddae674d.squirrel@cp3.shieldhost.com>
To: "L. David Baron" <dbaron@dbaron.org>
Cc: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>

> On Monday 2010-10-04 21:46 -0700, "Gérard Talbot" wrote:
>> > On Monday 2010-10-04 17:26 -0700, "Gérard Talbot" wrote:
>> >> > The following tests (in 20100917; I only spot-checked newer
>> releases
>> >> > to check that some tests were still problematic) are invalid
>> because
>> >> > they assume that CSS defines an allowed range of integer values,
>> >> > which it does not:
>> >> > counter-increment-013
>> >>
>> >> [snipped]
>> >>
>> >> {
>> >> (...)
>> >> > <meta name="assert" content="The property z-index set to a
>> minimum
>> >> value minus 1 is correctly truncated to the minimum value.">
>> >>
>> >> This assertion was discussed at the CSS WG F2F in Beijing. It was
>> >> decided that tests need to test some boundaries that were
>> reasonable.
>> >
>> > What was decided is, I believe, that it was reasonable to test that
>> > a reasonable range of values are supported.  It is not, however,
>> > reasonable to test that values outside that range are unsupported or
>> > truncated, based on the current spec.
>>
>> David,
>>
>> Isn't it what is supposed to happen with values outside a range?
>
> But there's no range in the spec.
>
> What we agreed to test is that values within a "reasonable" range
> are allowed.  In other words, we agreed that the test suite could
> test a *minimum* range for what's allowed.  Or, to put it another
> way, we agreed to test that browsers support *at least* a certain
> reasonable range.
>
>> many properties that allow an integer or real number as a value
>> actually
>> restrict the value to some range (...)
>
> There's no such restriction here.

David,

According to you then,

http://test.csswg.org/suites/css2.1/20101001/html4/z-index-001.htm

and

http://test.csswg.org/suites/css2.1/20101001/html4/z-index-012.htm

must also be considered invalid testcases because there is no range
defined for z-index.

Wouldn't it be best if CSS2.1 would define explicitly an allowed range
for integer values for properties like counter-reset, counter-increment,
z-index, widows, orphans, etc? For negative integer values when
applicable. And also define explicitly what should happen when minimum
(floor) and maximum (ceiling) values are exceeded?

Some tests will fail in some browsers only because their limit is
exceeded: over 255px for font-size, over 2^16 for z-index, etc..

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

CSS 2.1 test suite (RC2; October 1st 2010):
http://test.csswg.org/suites/css2.1/20101001/html4/toc.html

CSS 2.1 test suite contributors:
http://test.csswg.org/source/contributors/
Received on Tuesday, 5 October 2010 16:02:58 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 5 October 2010 16:03:04 GMT