On Wed, 15 Dec 2004 18:48:57 -0500, Robert Sayre <[EMAIL PROTECTED]> wrote: > Danny Ayers wrote: > > > > "Said of a system (e.g., program, file format, programming language, > > protocol, etc.) designed to easily allow the addition of new features > > at a later date, e.g. through the use of hooks, an API or plug-ins." > > So you can't do that with RSS2 or Atom?
Not really and not yet. See my previous post. XML Namespaces can enable pluggability, but don't automatically provide it. The core language (or a common underlying language) has to provide some form of hooks into its constructs. RSS 2.0 is generally lacking, Atom is getting there (IMHO). > > Does RSS 1.0 allow this? Yes, without a doubt, there is a complete > > supporting framework/language in which it is straightforward to define > > extensions on the same level as RSS (RDF/RDFS) or by building layers > > on top (OWL, rules etc). > > Yes, but you can't use your standard RDF serializer to write the RSS1. True, but that doesn't mean it can't be done. And most of the time in this context it really isn't an issue. My own RSS 1.0 feed is produced using WordPress with the FOAF Output plugin, and last time I looked used terms from 9 different namespaces. I'm fairly certain there's no generalised RDF modelling going on, but the output is (usually) spec-valid RSS 1.0 & also spec-valid RDF/XML. If I wanted to apply RDF processing, or use RDF queries, I'd read the stuff into a store, along with any other data I wanted to query in the same model (Before long I probably will move to a primarily RDF backend, but probably still make use of an RDBMS for the underlying store and templating for presentation). Whatever, currently the serialized data is still expressed in the rich language, even though as far as my blog templates are concerned it's dog-HTML. > > Does Atom allow this? That remains to be seen. Personally I'm hopeful > > that we've got past the assumption that namespaces alone will save us, > > and that we (will) have usable hooks which will make the format & > > protocol genuinely extensible, rather than just inheriting an 'X' in > > the title. > > OK, fair enough. How about > > <entry > xmlns="http://purl.org/atom/ns#draft-ietf-atompub-format-03"; > xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#";> > <title>Atom-Powered Robots Run Amok</title> > <link href="http://example.org/2003/12/13/atom03"/> > <id>vemmi://example.org/2003/32397</id> > <updated>2003-12-13T18:30:02Z</updated> > <rdf:RDF>...</rdf:RDF> > </entry> > > This is just like RSS1, except there's a URI identifying the entry, and > the syntax defaults to XML literals. Also, you can use your standard RDF > serializer to write your statements. Of course, you'll have to use an > actual RDF parser to get them back out. The URI on the entry is a very good hook into the core constructs, for sure. Something like this could quite easily provide the same kind of extensibility RSS 1.0 has - exactly the same kind really, assuming the obvious triple interpretations of atom:title etc. I'm not really trying to suggest that it wouldn't be fairly straightforward to give Atom strong extensibility (or that RDF is needed somewhere, though it does seem the easy way). Just that I think the simple/extensible split in the Nature folk's diagram is (or maybe was, until relatively recently) a fairly reasonable observation. Cheers, Danny. -- http://dannyayers.com
