W3C home > Mailing lists > Public > public-css-testsuite@w3.org > August 2007

Re: Implementation report template for 20060307 css3-selectors test suite

From: L. David Baron <dbaron@dbaron.org>
Date: Fri, 3 Aug 2007 00:50:35 -0700
To: public-css-testsuite@w3.org, w3c-css-wg@w3.org
Message-ID: <20070803075035.GA17359@ridley.dbaron.org>
On Friday 2007-08-03 01:39 -0400, fantasai wrote:
> >Should I be listing 27a and 27b as failure tests?
> 
> I'm not sure. It would depend on whether
>   foo:not(foo)
> would be considered a failure test. There's nothing invalid about those 
> tests,
> they're just using selectors that will never match.

I looked at a bunch of other tests categorized as failure tests, and
many of them did seem to be tests that selectors didn't match (e.g.,
182, 155a, 155b, 155c, and 175b).  So I think this is ok.

(Perhaps 155d and 175c are incorrectly marked as a failure tests,
though.)

> 
> >And does anybody see a better solution for how to list 27a?
> 
> List it under :root? 27b might also fit better there.

I think 27b is pretty clearly testing the descendant combinator --
that it doesn't combine to anything above the root element.

And I think 27a is testing something that's a clear conformance
statement for each of the selectors I put it under, whereas it's not
testing any conformance statements for :root.  I think it really
does belong in multiple places.  My main question was whether it's
better to list it in all those multiple places, or once, with a
special heading that indicates it belongs in all of them.

-David

-- 
L. David Baron                                 http://dbaron.org/
Mozilla Corporation                       http://www.mozilla.com/

Received on Friday, 3 August 2007 07:50:45 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 20 September 2010 17:51:55 GMT