Ben Lund wrote:
That depends entirely on the application you have in mind. For end- aggregators, it's mostly fine that they can ignore things they don't understand. But what about aggregating intermediaries? Unless there's a defined extensibility model, there's a large chance that the extra data in the exensions that the aggregator doesn't understand will be lost. RDF in RSS 1.0 make this a very simple problem to solve, whereas arbitrary XML namespaces makes it fiendishly difficult.
Could you show us an example? I'm having a hard time understanding why the intermediary could pass on blobs of (X)HTML but not blobs of XML.
I also don't understand why including rdf:RDF in atom:entry is insufficient. You'd have to consume it with an RDF parser and it would come in a container that isn't RDF itself, but I don't see a problem. It would eliminate the people pretending to use RDF that actually aren't--they remind me of American teenagers who get tatoos featuring chinese glyphs.
Robert Sayre
