Explicit vs Implicit URIs (aka "blanket statements or not")

Hi,

This is to summarize a series of discussions we've been having recently about blanket statements. Please read this in preparation for our upcoming teleconference:


BACKGROUND:
There are two related but yet quite different use cases for implicit URI (e.g. "example.org" means all resources available under this domain name):
1. to facilitate more compact EARL reports
2. to enable blanket statements such as conformance claims

CarlosI took up an action item to work out a draft proposal which he sent to the list on Monday 10 April, 2006; and we had subsequent discussions on Wednesday 19 & 26 April, 2006:
  <http://lists.w3.org/Archives/Public/public-wai-ert/2006Apr/0010>
  <http://www.w3.org/2006/04/19-er-minutes#item02>
  <http://www.w3.org/2006/04/26-er-minutes#item02>


ISSUE:
It turns out that such implicit URIs also bring about ambiguity. For example, if a new page is added to "example.org" after an assertion was made, is then the assertion result still valid? Similarly, for blanket statements such as conformance claims, it is unlikely that all pages under "example.org" have been tested but probably only a sampling. It is therefore rather imprecise (but still useful) to give a blanket statement without further description of what has been tested, and which methodology has been used to test.


PROPOSAL:
The currently suggested proposal is that EARL should only focus on recording actual test results (ie. no implicit URIs, only explicit ones). In some cases, RDF features such as collections may be suitable to reduce verbosity (still, every tested URI will need to be recorded at least once per report). As to blanket conformance claims, other vocabularies (preferably RDF-CL) should be able to provide the required functionality of expressing these, and pointing back to the EARL report for more detail on what has been tested.


EXAMPLES:
* "example.org/page.html passes WCAG 1.0 CP 1.1"
 - the "usual" way of using EARL to record test results will still be available
* "example.org/page.html passes WCAG 1.0 Level A"
 - a contraction of individual checkpoints is possible due to the WCAG hierarchy
* "example.org conforms to WCAG 1.0 Level A"
 - not to be expressed by EARL but by a different set of vocabulary (RDF-CL)


ACTION:
We need to confirm that this is indeed the approach we want to take with EARL: it focuses only on recording tests, and relies on other vocabularies for blanket conformance claims. Reports would also become potentially more verbose but also more precise as a result of this decision.


Please send your thoughts and comments to the list for discussion, we need to vote on this in due course.

Regards,
  Shadi


-- 
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, 2 May 2006 12:11:47 UTC