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

ERT WG feedback on "URI pattern matching"

From: Carlos Iglesias <carlos.iglesias@fundacionctic.org>
Date: Mon, 24 Jul 2006 19:30:08 +0200
Message-ID: <09700B613C4DD84FA9F2FEA521882819012DAAD1@ayalga.fundacionctic.org>
To: <public-xg-wcl@w3.org>
Cc: <public-wai-ert@w3.org>


Dear WCL group,

The ERT WG has looked at your work on "URI Pattern Matching for Groups of Resources" (Draft 0.1 17 June 2006) [1] with great interest. We have compared this work to our own (rough) requirements for identifying groups of resources in EARL [2]. During the ERT teleconference on 19 July 2006 [3], we have come to the conclusion that it may be too early for us to adopt this work as it currently stands but we will continue monitoring its development.

The following feedback for your consideration resulted from our discussion, we hope they will be of help to you in progressing this work:

#1: It may be useful to consider labeling for parts of Web resources (aka "snippets"). For example, making a statement about a specific element within an HTML document or CSS page that is part of a delivery unit. This could be interesting for syndicated and aggregated content. Note: ERT WG is currently working on "pointer" mechanisms to identify such snippets.

#2: We acknowledge a potential problem is that the XG has decided to adopt RDF-CL, in which sub-domains of given host are apparently always in scope by definition, which is in conflict with some of the Domain and Sub-Domain pattern matching requirements (the possibility of sub-Domain exclusions). We would like to emphasize the importance of resolving this issue for compatibility with EARL, since it is a requirement for us that any references to resources are unambiguous.

#3: It's not clear whether self-defined groups of pages (e.g. all the URLs that match an specific regular expression) are going to be covered or not. 

On one hand the "Non-Requirements" section explicitly says that "It is not required that patterns allow matching to any conceivable pattern..." but on the other hand the "Outline Design" section says "The matching a pattern consists of an optionally case sensitive leftmost, rightmost or exact [or regex ???] match of the normalised text of the relevant component from the candidate URI." 

Since this class of patterns are frequently used by automatic tools, and these are likely to be used as help to make assertions in several use cases (e.g. Web Accessibility or Mobile Web), it make sense to incorporate them into the labels.

Regards,
  Carlos Iglesias for ERT WG

[1] <http://www.w3.org/2005/Incubator/wcl/matching>
[2] <http://lists.w3.org/Archives/Public/public-wai-ert/2006Apr/0010>
[3] <http://www.w3.org/2006/07/19-er-minutes>
--
Carlos Iglesias

CTIC Foundation
Science and Technology Park of Gijón
33203 - Gijón, Asturias, Spain

phone: +34 984291212
fax: +34 984390612
email: carlos.iglesias@fundacionctic.org
URL: http://www.fundacionctic.org 
Received on Monday, 24 July 2006 17:30:19 GMT

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