See also: IRC log, previous 2006-05-01 record
<Ralph> ? I'm getting 404 on information-resource-debate-and-rda.html
Mark: it doesn't always seem easy to tell when the subject is a car or a Web page. Current usage infers implicitly that a document that has both a dc.author and a geo.location property is likely talking about the author's location
ACTION: Ben, Mark be ready on 15 May to discuss WWW2006 presentations [recorded in http://www.w3.org/2006/05/01-htmltf-minutes.html#action04] [CONTINUES]
Ben: the 3-minute talk will be the more challenging one
ACTION: Ben, Mark send outlines of their XTech and WWW2006 presentations this week [recorded in http://www.w3.org/2006/05/01-htmltf-minutes.html#action05] [CONTINUES]
Ben: I hope to send mine today
Mark: me too
ACTION: Ben start separate mail threads on remaining discussion topics [recorded in http://www.w3.org/2005/12/06-swbp-minutes#action04] [CONTINUES]
ACTION: Ben to draft full response to Bjoern's 2004 email [recorded in http://www.w3.org/2006/01/24-swbp-minutes.html#action03] [CONTINUES]
Ben: it's been more work than I'd planned to integrate comments from Mark
ACTION: once Steven sends editors' draft of XHTML2, all TF members take a look and comment on showstopper issues only [recorded in http://www.w3.org/2006/02/06-htmltf-minutes.html#action01] [CONTINUES]
Ben: (summarizing previous discussion) -- concern about using class attribute to create triples due to legacy usage in many existing documents. The hGRDDL approach allows an author of a new document to say explicitly in the profile attribute that this document is intended to have explicit semantics
Mark: sounds like a nice idea. In which namespace are the profile URIs declared?
Ben: we would declare the specific profile URIs for XHTML1 and XHML2 documents. GRDDL Working Group would need to collaborate with us on this
-> Proposal: hGRDDL, an extraction from Microformats to RDFa
-> draft GRDDL WG charter [Member-only link]
Ben: the RDFA.GRDDL.addProfile examples in my mail of 29 April should be considered as examples of adding _additional_ profiles above some XHTML 1.1 pre-declared ones. At the time I wrote that proposal message I had not considered that XHTML 1.1 could declare its own default profiles as well
Mark: I think this is a much nicer way of getting around people writing "class='x'" and lots of other local intranet-only stuff that people do. This could give authors a way to opt-out of QNames usage as well
ACTION: Ben write a prototype hGRDDL profile for XHTML 1 [recorded in http://www.w3.org/2006/05/08-htmltf-minutes.html#action06]
Mark: this would be a transformation step? Perhaps we should define this in terms of triples; e.g. if rel='stylesheet' really means rel='html:stylesheet' then it could be defined in transformation terms .or in triple terms
Ben: I don't intend to specify the profile in terms of javascript but rather in DOM terms. The point is to do in-place substitution of the DOM tree
Ralph: I expect we'll attract more developers if we adopt a more operational specification; that is, hGRDDL is specified as a GRDDL transform of the DOM tree. A question, and possible requirement, for GRDDL WG is whether explicit pipelines of transforms can be declared; e.g. hGRDDL followed by RDFa to get triples
Mark: some of Jeremy's transformations have to be reworked; e.g. how rel='transformation' is handled. It would be nice to define an hGRDDL transform at a high-level rather than in a specific implementation language
Ralph: yes, the GRDDL Note supports this; it is neutral w.r.t. the language in which any given transform is implemented and it should not be a problem to provide multiple implementations of a given transform so we could define hGRDDL in an abstract way and give javascript, XSLT, and other implementations
ACTION: Ben update hGRDDL issue to note a possible requirement for explicit pipelining in GRDDL [recorded in http://www.w3.org/2006/05/08-htmltf-minutes.html#action07]
Ben: hGRDDL does give us a way to resolve our backwards compatibility constraints
Mark: after working out details of <head about="">, I realized that <body about="#"> would be a way to say the triples in the body are about what the body is about
next meeting: 15 May, 1300 UTC
adjourned