Re: CSS Test suite format

Right now I'm the designated liaison since I've hired the student Bert 
is referring too (Nadia), specifically to work on that kind of stuff :-)
But I think Karl has been tracking more her work that I did since she
arrived because, and Charles McCathieNevile from WAI (where EARL
originated).

Anyway, it's covered and you'll hear about it as it moves forward.

You can look at the thread starting at
 http://lists.w3.org/Archives/Public/w3c-wai-er-ig/2002Feb/0028.html
for details.

> Lynne forwarded this to me (thanks, Lynne).  This seems like something that 
> we should establish liaison to, or track carefully, etc.  EARL is called 
> out specifically in QAWG deliverables.  Should QAWG designate a WG member 
> to be the focus/responsible person for starting to move forward on the 
> deliverable?
> 
> -Lofton.
> 
> >>Resent-Date: Fri, 1 Feb 2002 10:53:53 -0500 (EST)
> >>From: Bert Bos <bert@w3.org>
> >>Date: Fri, 1 Feb 2002 16:53:22 +0100
> >>To: w3c-css-wg@w3.org
> >>X-Mailer: VM 7.00 under 21.4 (patch 6) "Common Lisp" XEmacs Lucid
> >>Subject: Test suite format
> >>Resent-From: w3c-css-wg@w3.org
> >>X-Mailing-List: <w3c-css-wg@w3.org> archive/latest/6128
> >>X-Loop: w3c-css-wg@w3.org
> >>Sender: w3c-css-wg-request@w3.org
> >>Resent-Sender: w3c-css-wg-request@w3.org
> >>List-Id: <w3c-css-wg.w3.org>
> >>List-Help: <http://www.w3.org/Mail/>
> >>List-Unsubscribe: <mailto:w3c-css-wg-request@w3.org?subject=unsubscribe>
> >>
> >>I was looking a bit at our test suites, because we have a student who
> >>just started a project here for our QA activity. She will see how test
> >>suite formats can be harmonized across W3C and the test results
> >>processed semi-automatically, using the EARL (Evaluation And Report
> >>Language[1]) format that WAI developed.
> >>
> >>One idea is that she makes a common test harness, which adds a form
> >>with three or so buttons to every test page (Succeeded, Failed,
> >>N/A...) and a system that stores the results in EARL. A system that
> >>works off-line would be useful as well, but is probably harder.
> >>
> >>Since EARL is RDF, those results could then be input to an inference
> >>engine that combined the conformance requirements (also expressed in
> >>RDF) with the test facts, and conclude whether a certain test
> >>environment conformed or not. This latter part depends on further
> >>developments in RDF, of course, but we expect that they will happen
> >>eventually.
> >>
> >>Our CSS tests would fit that system very well. Thanks to Ian's
> >>scripts, such a test harness would be just another generated
> >>framework.
> >>
> >>I'll probably come back to the project later, when the planning is a
> >>little more advanced.
> >>
> >>But meanwhile I came up with two comments:
> >>
> >>   - Among the (excellent) metadata in the XML-based tests one thing we
> >>     used to have in CSS1 seems to have disappeared: a link back to the
> >>     specification.
> >>
> >>   - Ian has produced a lot of variants, do we need all of them? They
> >>     are cheap to make, and W3C can afford the disk space, but I'm
> >>     worried that people will get confused by the sheer number of
> >>     choices, most of which don't work in their browser.
> >>
> >>
> >>[1] http://www.w3.org/2001/03/earl
> >>
> >>
> >>Bert
> >>--
> >>   Bert Bos                                ( W 3 C ) http://www.w3.org/
> >>   http://www.w3.org/people/bos/                             W3C/INRIA
> >>   bert@w3.org                             2004 Rt des Lucioles / BP 93
> >>   +33 (0)4 92 38 76 92            06902 Sophia Antipolis Cedex, France

Received on Friday, 15 February 2002 12:34:23 UTC