- From: Richard Cyganiak <richard@cyganiak.de>
- Date: Sat, 12 May 2012 18:04:32 +0100
- To: Eric Prud'hommeaux <eric@w3.org>
- Cc: Sandro Hawke <sandro@w3.org>, RDF Working Group WG <public-rdf-wg@w3.org>
Hi Eric, On 12 May 2012, at 13:57, Eric Prud'hommeaux wrote: > I'm not sure all of this defines "conformance". In fact, it was my > impression that specs which define a language which does something > (e.g. SPARQL maps RDF graphs and queries to solution sets) have > conformance criteria while languages whichs simply are (XML, RDF, > Turtle) don't benefit from expressing a notion of conformance. I find this distinction rather strange. Languages don't “do something”. Agents do. And surely, the XML and Turtle specs need to explain how to “do something”: mapping certain kinds of strings to, respectively, XML infosets or RDF graphs. My understanding is that a recommendation-track W3C specification without conformance clause violates W3C's normative QA Guidelines: http://www.w3.org/TR/qaframe-spec/#specifying-conformance Personally I find the W3C QA Guidelines very well thought out and compelling. Best, Richard
Received on Saturday, 12 May 2012 17:05:03 UTC