replying to a point raised on Ben's weblog:
Another question for everyone I guess. In both RSS 1.0 and the Userland
proposal for RSS 2.0, you can have duplicate semantics. For example:
Which of these takes precedence? This is mostly an issue with RSS 2.0, as
there is a lot of overlap between the core elements and available modules.
(and RSS 1.0 has only one possible overlap, the title/dc:title one.) This
is one reason why I'd like to see the core elements become fewer in
Posted by Ben Hammersley at 05:04 PM
This is one of the reasons why _namespaces alone aren't enough_. (a
concern re RSS 0.94/2.0 btw)
We need rules for using and understanding XML namespaces, if they're going
to help us share, aggregate and exploit namespace'd data.
Here are some of the rules we wrote in the
RDF schema for http://purl.org/rss/1.0/
rdfs:comment="A short text description of the subject.">
rdfs:comment="A descriptive title for the channel.">
This means, roughly, that the relationship we call rss:title (between
something and its rss:title....) is a specialised case of the more general
relationship dc:title. Similarly with rss:description and dc:description.
This means that whenever we see someone claiming that "blah" is the
rss:title of something, and if we hold them to undersand the meaning of
rss:title under the rules of RDF, we can hold them to also be claiming
that the dc:title of that thing is also "blah". Same for dc:description.
If someone builds a tool for searching Web documents based on their
dc:title and/or dc:description, the rules associated with RSS 1.0 and RDF
(and RDF Schema) make it ok for them to use rss:title and rss:description
data instead of the more general Dublin Core versions. This is good; it
means the Dublin Core and RSS 1.0 communities can, for the most part, get
on with their lives without having to worry too much about interop. The fact
that both communities have bought into RDF means that a certain amount of
interop is acquired for free. Or for the cost of the syntactic burden of
RDF in the RSS 1.0 XML syntax, ie. the rdf:RDF and rdf:Seq bits that are
needed to allow RDF tools to understand an RSS 1.0 feed.
So to answer Ben's question. Neither as such takes precendence. But we
should realise that, since dc:title is defined as more general than
rss:title (the latter is an rdfs:subPropertyOf the former),
...in RSS 1.0 should be taken to imply:
since the rss title of something is always a DC title of it too. This
constraint helps fix the meaning of RSS title. For example, we
inherit the work done in the DC world for translating property defintions
into multiple languages. Unless we care to duplicate that work on RSS-DEV,
this sort of thing seems to be rather a good deal...