Henry, please correct me if I'm wrong on your proposal A : 1. Atom core would remain essentially as-is*
2. Anyone wishing to use an RSS 2.0-style extension mechanism, can do (i.e. anything can go anywhere, leave it to vendors to figure individual interpretations) 3. Anyone wishing to use RDF-based extensions can do, by adding the attributes needed to make the syntax interpretable as RDF/XML, then re-using standard vocabularies in the same manner as RSS 1.0 * there may be minor readjustments needed to align the formats, but I can't see any yet (changing the case of Feed etc isn't essential - RSS 1.0 has lower-case classnames) This seems a low-friction route through the extensibility issue. I could live with it (to the extent that I'd be a lot more comfortable about Declaring Victory). James said "RSS 2.0 demonstrates the real-world fact that using namespaced modules works". I don't believe it does. There are very few extensions in common use. Last time I looked the usage was negligible. I would guess that it's grown now, but nothing like to the extent of extensions in RSS 1.0. If you can show me a few RSS 2.0 feeds carrying more than half-a-dozen extensions, I'll accept the demonstration. Dare, if I understand Henry correctly, I don't think this is really opening the debate as before as the effect on Atom Core is virtually nil. But you wanted a practical example - how about the query "Give me all the mp3 format audio files published by any of the people I know that work for Microsoft in the last week, that are less than 10 minutes long with a review rating better than 8/10, unless the artist name is "Happy Mondays" in which case just give them to me". That can be done today using off-the-shelf tools with virtually no coding (in particular no database structural building, the RDF provides the structure) and using mostly common extensions - MusicBrainz, FOAF, REV etc. Cheers, Danny. -- http://dannyayers.com
