W3C home > Mailing lists > Public > public-wai-ert@w3.org > August 2006

Re: Action Item: Testable Statement class

From: Shadi Abou-Zahra <shadi@w3.org>
Date: Tue, 01 Aug 2006 21:52:47 +0200
Message-ID: <44CFB10F.3000503@w3.org>
To: Carlos Iglesias <carlos.iglesias@fundacionctic.org>
Cc: ERT group <public-wai-ert@w3.org>


Carlos Iglesias wrote:
> - No Testable class
> - At least one required Requirement per Assertion (an Assertion could be related to several requirements e.g. WCAG, 508 and mobileOK)
> - An optional (just one) Test Case per Assertion

The question that was raised from this proposal was "what was the outcome of the individual tests and which logic was used to combine these partial results?". We concluded that we needed to have a mechanism to record the "lower level" test results as well as the assertions on the requirements level.

Another option could be to require that if earl:TestCase is used as an earl:Testable, then it *must* contain a dc:isPartOf property that points directly or indirectly (via in-between earl:TestCase instances with further levels of dc:isPartOf pointers) to a published (not necessarily in the same report) earl:Requirement.

This is a simple way to resolve which requirement(s) a test case relates to, and vice-versa (but not really *how* they relate). It is a fairly flexible model and should be compatible with most test description languages to "sit on top". However, it still doesn't cover the "which logic was used" aspect but that certainly belongs into a test description language.

While we're at it, we could also add that every earl:Requirement *must* contain a dc:hasPart property that points back to the relevant earl:TestCase instances if they are published. We could also just soften that into a *should* or something.

Is this a viable solution?


>> Johannes Koch wrote:
>>> Charles McCathieNevile wrote:
>>> [TestRequirement and TestCase subclasses of Testable]
>>>> I still think this is a bad idea, since I don't see the value in 
>>>> having  the two kinds of subClass. If we adopt this, the range of 
>>>> earl:requirement  needs to be made earl:Testable too. (Since we are 
>>>> shifting the range to a  superclass, I think we can get away with that.
>>> If the object for the earl:testable is either an earl:TestRequirement or 
>>> an earl:TestCase we can create both assertions about subjects 
>>> passing/failing/... a certain test case or meeting/not meeting certain 
>>> requirements.

Shadi Abou-Zahra     Web Accessibility Specialist for Europe | 
Chair & Staff Contact for the Evaluation and Repair Tools WG | 
World Wide Web Consortium (W3C)           http://www.w3.org/ | 
Web Accessibility Initiative (WAI),   http://www.w3.org/WAI/ | 
WAI-TIES Project,                http://www.w3.org/WAI/TIES/ | 
Evaluation and Repair Tools WG,    http://www.w3.org/WAI/ER/ | 
2004, Route des Lucioles - 06560,  Sophia-Antipolis - France | 
Voice: +33(0)4 92 38 50 64          Fax: +33(0)4 92 38 78 22 | 
Received on Tuesday, 1 August 2006 19:52:47 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:55:54 UTC