See also: IRC log, previous 2006-09-25
ACTION: [DONE] Ben mail a full example of the need for bnode support [recorded in http://www.w3.org/2006/09/25-htmltf-minutes.html#action12]
-> better support for bnodes [Ben 2006-10-01]
ACTION: Ben prepare a sample XHTML2 document for Steven [recorded in http://www.w3.org/2006/09/25-htmltf-minutes.html#action05] [WITHDRAWN]
Steven: I have started such a sample myself
... the idea is to have an XHTML2 document with everything in it, right?
Ben: with XHTML2 headers and content
... mime type, http headers, and the html doc itself
... then we can plug-in RDFa examples
ACTION: [DONE] Ben to put together a work plan for the next 6-9 months of the TF [recorded in http://www.w3.org/2006/09/25-htmltf-minutes.html#action13]
Ben: I did a 3-month plan, propose that longer view be done after SWD WG is underway
<EliasT> yes
Ralph: I'm happy with the 3-month view for now
<Steven> me too
Elias: happy with 3-month plan
ACTION: [DONE] Mark contact schneegans.de folk about connecting with our work [recorded in http://www.w3.org/2006/09/19-htmltf-minutes.html#action10]
ACTION: Ben make sure RDFA bookmarklet runs locally [recorded in http://www.w3.org/2006/09/19-htmltf-minutes.html#action06] [CONTINUES]
ACTION: Ben start separate mail threads on remaining discussion topics [recorded in http://www.w3.org/2006/09/19-htmltf-minutes.html#action07] [CONTINUES]
ACTION: Ben update the issues list [recorded in http://www.w3.org/2006/09/19-htmltf-minutes.html#action08] [CONTINUES]
Steven: I gave a plenary talk this past weekend
at a European Information Architects conference
... got a very positive response
... but I could not show the bookmarklet example as there was no network
available
Ben: in my mail inbox I do have a proposed
patch from Creative Commons to add a '#'
... evidence that others are looking at the prototype and making
suggestions
Steven: at a recent Hypertext CG meeting there
was discussion about a DOM for RDF
... perhaps Ben's bookmarklet should be a part of this discussion
... as Ben has implemented a generic interface to accessing triples "in" a
page
Ben: what does HCG mean by RDF DOM?
Steven: I'll look for links to the discussion, may be Member-only
Elias: an RDF API will be more complicated that
what Ben has done
... statement-centric and resource-centric views
... we have good starting points in Jena and Sesame, for example
Ben: does rdflib qualify?
Elias: sure
... Ben's javascript API is a subset of what people have already been working
with
Ben: I've been following from an idea that I think Mark started; using an XML DOM and accessing "related" RDF triples at any point in that tree
Elias: rendering is different
Ben: yes, but it may be useful to be able to access triples assocated with a DOM node
Elias: the Document Object Model equivalence
for RDF is; given a document, how do you access the triples it contains?
... this is abstract; doesn't need to specify any particular data
structures
... a general (DOM) API for RDF
Steven: I'll track this HCG discussion
ACTION: Ben write a prototype hGRDDL profile for XHTML 1 [recorded in http://www.w3.org/2006/09/19-htmltf-minutes.html#action09] [CONTINUES]
Ben: the hGRDDL work is slightly delayed; it's not one of the primary use cases for the GRDDL WG
ACTION: Steven to put together sample XHTML2 doc with all mime type, etc.. [recorded in http://www.w3.org/2006/09/19-htmltf-minutes.html#action01] [IN PROGRESS]
move to next agendum
-> tentative schedule for next 3 monthss [Ben 2006-10-01]
Ben: it's an aggressive schedule but we've done
a lot of work on core issues
... good to refocus priority on getting RDFa to work in XHTML1
... my 5 target dates attempt to take holidays into account
... October 20th - updated Primer and Syntax documents for consistency
... there are a couple of issues we need to settle for this next version;
e.g. class and role
... October 27th - set of example documents with parsing code and output
RDF/XML or RDF/N3
... Steven's document will be a good start on this
... November 10th - new use case and requirements document for RDF-in-XHTML
1.1
... we need to think about how much of RDFa we can put in XHTML1.1 and what
changes we might want to XHTML1 to do this
... need to take the old requirements document(s) and update them
... December 1st - RDF-in-XHTML 1.1 details worked out
... this is an internal date; goal is to have all the basic details of RDFa
in XHTML1.1 worked out by 1 Dec
... December 15th - RDF-in-XHTML 1.1 module specification WD #1
Mark: looks good but I wonder if the use cases
and requirements should be done first
... we have a lot of documents around that are fairly mature
... maybe we could work on the use cases and requirements in parallel
Ben: I agree that most of the core work has
bene done
... part of this task is public relations but we don't have as much latitude
to make changes to XHTML 1.1
... XHTML2 was wide open so we may need to work harder to justify changes in
XHTML 1.1
Mark: there are certain limitations that
current browsers impose
... so part of the effort is to see what we can make work
... for example, we may have difficulty putting href everywhere
... I'd like to document what we currently have and experiment with
implementations to see how far we can go
Ben: Nov 10 is not far off; if we can state
requirements by then that say something about validation, for example, that
would be good
... the problem we've been trying to solve is a big one; ideally we'd like to
do everything
... with our work over the past 2 years we now know better what is going to
be possible
Mark: I think we know pretty much what we want
RDFa to do by now
... it should be easy for us to write up the use cases now
... the requirements document may not match with what we can implement
Ben: I compare to the DAWG (SPARQL) group; a
lot of the SPARQL details were worked out before the requirements document
was finished
... my interpretation is that these requirements documents come after some
investigation of what is possible
Elias: it's not all one way or the other
... some of the SPARQL UCR document was added after experimentation
... but it is also true that the UCR document is compiled from requests
... for example, people asked for an XML output format so we added it
... UCR is based on both external and internal feedback
Mark: there's input, e.g. from IPTC, on the
RDFa spec; e.g. they wanted reification
... but for XHTML1.1 the requirements may be much narrower
... because we are concerned about what we can do with current browsers
Elias: that in itself is a requirement
... it might be a matter of identifying which things are addressed only in
XHTML2
... and other requirements, such as working with existing browsers, that only
apply to XHTML 1.1
Mark: we've a lot of implementation experience by now
<Zakim> RalphS, you wanted to support requirements work
Ralph: lots of information we hold in our heads
that we should put down on paper
... writing down what have been our goals is a good way to match "our" vs.
"their" expectations
Mark: RDFa in general or just the XHTML 1.1 module?
Ralph: both
Ben: some people will be coming to the table with interest only in XHTML 1
Mark: when we're talking specifically about RDF
in XHTML 1.1 that's a specific requirement
... RDFa is broader than that; it's not just XHTML 1.1
... getting RDFa into XHTML 1.1 is one requirement, figuring out what RDFa
should do is a larger set of requirements
Ralph: but we need one more level of detail than "RDF in XHTML"; we need to say whether, for example, reification is in or out, how much effort we put into making bnodes nice, ...
Mark: sounds like we're defining a new, subset language
Ben: not really; it's a matter of presentation. It's ok to say some features will not show up until XHTML2
Mark: I really don't want to spend my time on a requirements document
Ralph: I want feedback from the community on how hard we should work to fit every possible feature into the XHTML 1.1 module
Mark: we can look at (1) schema constraints and
(2) impact on rendering
... we're not asking fundamental questions about RDFa; it's mostly an
implementation thing
Ben: I propose that I take the lead on a UCR
document and Mark take the lead on the RDFa specification document
... these documents can feed each other
... some people come at this with no prior understanding and want to know
what problem we're trying to solve
Mark: we need this for all of RDFa, not just for the XHTML 1.1 module
Ben: yes, and we can point out which use cases we do not expect to be able to do in XHTML 1.1
Mark: a UCR document that only focuses on XHTML 1.1 is too narrow
ACTION: Elias start an FAQ [recorded in http://www.w3.org/2006/10/03-htmltf-minutes.html#action10]
Steven: I expect to be less available during December
Ralph: there's also a week of W3C meetings the end of November which affect my availability
next meeting: 9 Oct, even though it's a holiday for some in US
Steven: good news -- the HTML WG has just finished XHTML Print and gone to REC and there's more interest in RDFa now expressed