- From: Bijan Parsia <bparsia@isr.umd.edu>
- Date: Wed, 7 Nov 2007 09:22:00 +0000
- To: Bernard Vatant <bernard.vatant@mondeca.com>
- Cc: SW-forum Web <semantic-web@w3.org>
On Nov 7, 2007, at 9:04 AM, Bernard Vatant wrote: > > Tim >> Something funny about the XML: Line 2: doctype broken. Shoud not >> be a doctype. > Sounds like the standard output syntax of SWOOP, which is indeed > incorrect XML-wise in its DOCTYPE declaration. Oops. > This is something I've discussed with Bijan quite a while ago. And it disappeared into the bitbucket. Could you tell me what the correct output should be and I'll see what I can do. > That said, I've been using this syntax "as is" for about two years > now (see e.g. http://www.lingvoj.org/ontology.rdf) and it does not > seem to break any RDF tool. Only XML parsers are grumbling. > As a side note, apart of that, the SWOOP RDF-XML serialization is > the best I've found so far, given the messy output of other famous > ontology editors (no name :-) ). Bernard, two things...Protege4's output is moving to a cleaner form (partially inspired by the Swoop serialization) so you may want to check it out. Since this serialization strategy is built into the OWL API, it can affect more applications than just Protege4, so there's a good opporunity here :) I work with the author of the api (Matthew Horridge, my grad student) and he's quite responsive to this sort of feedback. Also, I've brought up the idea of canonical serializations in the OWLWG: <http://www.w3.org/mid/368C5358- B553-44F3-9585-3008E91E1F53@cs.man.ac.uk> So, please keep an eye on that. I hope we can get one that is *really* nice, bug free, and with a W3C spec behind it, that is widely implemented across editors and toolkits. Cheers, Bijan.
Received on Wednesday, 7 November 2007 09:22:21 UTC