- From: Benja Fallenstein <b.fallenstein@gmx.de>
- Date: Mon, 29 Nov 2004 22:43:51 +0100
- To: Dan Brickley <danbri@w3.org>
- CC: "DuCharme, Bob (LNG-CHO)" <bob.ducharme@lexisnexis.com>, www-rdf-interest@w3.org
Dan Brickley wrote: > That's an oddity of CVSWeb. You're looking at a generated HTML page that > is _about_ some (yes, typically W3C RDF/XML) document that's in a CVS > repository. If you hit the 'revision 1.1' or whatever link, you'll get > an HTMLization of the content, plus navbar stuff. If you hit download, > you'll get the actual RDF. ... > In fact CVSWeb isn't doing anything wrong, assuming they're sending > correct content-type headers. People can choose to end their URIs in > whatever characters they prefer. But it is a little counter-intuitive. <garment webArchNitpickHat="on" speechAct:tongueInCheekFactor="quite"> In fact, it makes a lot of sense, since although the URI http://cvs2.oeone.com/index.cgi/penzilla3/apps/appliancemanagement/content/contents.rdf could represent anything, intuitively it represents "the file apps/appliancemanagement/content/contents.rdf in the repository penzilla3" (a digital resource), and giving a list of revisions of the file seems like a perfectly reasonable representation of such a resource. Even with digital resources, representations do not have to be instances (copies) of (a version of) the resource; they can be anything the authority (web space owner) likes, and this particular type of representation seems perfectly useful. </garment> (Leo Sauermann will get a fit if he reads it about people who solve theoretical problems instead of writing real code. But I think it's fun. ;-)) - Benja
Received on Monday, 29 November 2004 21:44:34 UTC