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

[AI] conformance

From: Johannes Koch <johannes.koch@fit.fraunhofer.de>
Date: Thu, 11 May 2006 11:43:01 +0200
Message-ID: <44630725.40307@fit.fraunhofer.de>
To: public-wai-ert@w3.org

Hi group,

following our action item about reviewing the conformance/baseline 
thing, here are my comments:

<http://www.w3.org/TR/WCAG20/conformance.html>:

* Optional components of a conformance claim:
     "4. A list of user agents that the content has been tested on."
   If the scope contains more than one resource/web unit, is it necessary
   to test all resources/web units with all user agents/combinations?

* Examples of conformance claims -> Example 3:
   Does http://example.com/nav only mean
   1. "the web unit identified by the URL http://example.com/nav", or
   2. "all web units identifed by a URL starting with
     http://example.com/nav", or
   3. "all web units that are part of the web site section which the
     resource/web unit identified by the URL http://example.com/nav is
     the base of"?
   For me it's #1. If something else is meant, it must be made clear.

   "The technologies this content has been tested with"
   Should this be "The user agents this content has been tested with"?
   See "Optional components of a conformance claim" #4

* Conformance notes
   "Note: If multiple representations can be retrieved from a URI through
   content negotiation, then the conformance claim would be for the Web
   unit that is returned when no negotiation is conducted"
   This is really problematic as this "not negotiated" variant can be
   irrelevant, because most of the user agents may send information so
   that they won't get this variant. You could claim conformance for a
   website with an accessible default variant in basque while sending
   non-accessible variants in english, french, german, etc.
   Technique 2 in situation B for SC 4.2.1 (Understanding doc) talks
   about user agents' preferences indicating they can handle inaccessible
   content. How does a user agent indicate it can handle inaccessible tag
   soup in contrast to proper HTML?

* Aggregated content
   As mentioned in a nother mail it is not clear which referenced
   resources belong to a web unit. E.g. do all referenced style sheets
   for different media belong to one web unit identified by the URL for
   the HTML document where they are referenced from, although they are
   not "intended to be rendered together"?

* Scoping of conformance claims -> Example #2
   "The videos are located in one location (e.g.,
   example.com/movies.php). The conformance claim for the site or section
   of the site excludes the location that contains the videos."
   As the following sentence makes clearer, the videos seem to be
   _referenced from/embedded in_ the document example.com/movies.php,
   but not not _located_ there.


<http://www.w3.org/WAI/WCAG20/baseline/>:

Some things differ from the corresponding text in 
<http://www.w3.org/TR/WCAG20/conformance.html>, some things are missing.
-- 
Johannes Koch - Competence Center BIKA
Fraunhofer Institute for Applied Information Technology (FIT.LIFE)
Schloss Birlinghoven, D-53757 Sankt Augustin, Germany
Phone: +49-2241-142628
Received on Thursday, 11 May 2006 09:43:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:18:27 GMT