[whatwg] A Selector-based metadata proposal (was: Annotating structured data that HTML has no semantics for)

[Eduard Pascual]

Note: I wrote this yesterday. My internet connection wasn't working as
desirable, but GMail told me it had been sent and I believed it. Now I
have just noticed that it hadn't; and at least one person has been
confused by the changes in the document. Sorry for this issue, and
hope this time GMail does send it. What follows is the message as it
should have been sent yesterday:

Update: I have just put up a new version of the CRDF document. The
main changes are:
Section 0. Rationale: several corrections on the claimed limitations
of RDFa, which have been shown to be just limitations of my knowledge
about RDFa.
Section 2. Syntax: the syntax is now more formally defined (although
it still refers to CSS3's Syntax, Values, and Namespace modules for
some stuff). The content model for property values is now fully
defined: resource and "reversed" support has been added, and explicit
typing capabilities are now more prominent in the document. For
subject definitions, the "none" keyword has been redefined; "blank()"
now handles what "none" previously did, and a syntax has been added to
mimic EASE's nearest-ancestor construct. Finally, a subsection has
been added describing how to handle escenarios where a tool might have
to extract an "XML literal" from source in a non-XML language.
Section 3. The host language: expanded 3.3 (embedding inline CRDF) to
allow multiple brace-delimited blocks within the attribute value, to
enable stating properties for different subjects while reusing the
same element.
Section 4. The first examples don't make sense anymore after the
changes in section 0. They have been removed, waiting for further
feedback on that section before redoing them.



I'd like to reiterate what I said in the opening message: if someone
can suggest of a better place to discuss this document, please let me
know.