feedXMPP with RDF Events
November 01, 2007 | 
co.mments
Danny Ayers on JEvent, an XMPP based EDA system:
"Standardized message format based on RDF compatible with OpenMetadir"
I'd suggest looking at this.
The adoption challenge for RDF is at least twofold:
1: an Atom Entry is also good starting basis for an event message, and there's a head of steam behind it. You need a few foreign elements (eg type), whihc would probably take a few ID drafts to thrash out, but quite a few existing Atom elements are useful for describing an event.
2: RDF is super flexible in away that's qualitatively different to XML or JSON dictionaries. this means managing mappings for the "mustHave" fields in the vent to your highly structured (ORM/RDBMS) backend. Or getting past all that and using RDF based storage. IMO, using RDF is more than a wire interchange/interop choice - it will impact implementation all the way through to persistence (not unlike a COBOL guy having to integrate XML with PICtures via xsd:maxlength restrictions, but this time it's happening at the domain model level). Yes, I know the flexibility provided by RDF's open world model is the much of the point.
One upside to using RDF is that it would integrate very well with the notion of applyting declarative rules for complex event processing (CEP). I have the impression that CEP is largely aspirational at this point (Esper notwithstanding), but I am happy to be wrong.
However, XMPP itelf is a no brainer as the backbone protocol. I can't imagine it not being used for everything that HTTP is unsuited for a few years from now. Although we might have to go through a wasteful EDA-* cycle first before everyone "gets it", a la what has happened with WS-*.
November 1, 2007 09:38 PM