[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Proposed addition to Charter



On Mar 20, 2004, at 9:14 AM, Danny Ayers wrote:

The Charter is looking good!

One thing I'd like to propose is the addition of the third item here:

   * A conceptual model of a log entry.
   * A concrete syntax for this model.
   * An  RDF/OWL representation of this model
   * A syndication format using this syntax.
   * An archiving format using this syntax.
   * A weblog editing protocol using this syntax (the Atom API).

I'd suggest the deliverables would be an OWL ontology/RDF schema together with an XSLT stylesheet for an Atom => RDF/XML representation.


Tim Bray wrote:

I totally oppose adding this to the deliverables of the existing project. To start with, the experience here is overwhelmingly around blogging and aggregating, so we are poorly qualified to judge the quality, correctness, or usefulness of OWL/RDF deliverables.

On the contrary, there is a considerable amount of RDF experience around these parts, though many of the RDF-oriented developers on this list have gone quiet/left since the use of RDF/XML for serialisation was rejected. There has historically been considerable crossover, which might not be obvious to people arriving on the scene more recently and primarily encountering RSS 2.0 and associated promotion. Blogging, aggregating and RDF are not mutually exclusive, quite the opposite.Take for example SixApart's TypePad -out of the box it produces RSS 1.0 (RDF/XML), Trackback (RDF/XML) and FOAF (RDF/XML).

If you want to do this I can see no problem with doing it, but I don't think you're going to get a good result here, and I don't think Atom needs it, and I don't want it slowing us down, which it will. -Tim

I described some benefits to Atom and workload issues in my previous mail [1]. I see no reason it should slow things down, the draft schema is already mostly in place.

Ok, the RDF mapping can appear outside of Atom, but I suspect it would be poor politics to disenfranchise the RDF community even more, and further marginalise Atom. An informational schema/ontology and XSLT will appear whatever happens, hopefully in a way that will have most RDF developers working consistently when it comes to Atom data. A normative schema and XSLT would ensure that Atom had a consistent interface to RDF systems. On a practical, positive front, the interop that would provide from day one would be a huge benefit. I simply believe it would be beneficial for the Atom project to associate with the RDF community and take advantage of the technology, even if Atom's serialisation is plain XML and the primary description is the spec prose.

Cheers,
Danny.

[1] http://www.imc.org/atom-syntax/mail-archive/msg03218.html