See also: IRC log
<scribe> Scribe: Steven
Absent: Ralph, DanBri
Ben: I talked with DanC this week
about GRDDL
... we would like to add a GRDDLY transform somewhere
<MarkB_> Ben asked if the XHTML schema would live at the namespace URL address.
<MarkB_> Steven said no decision had been taken yet.
Steven: Maybe
... depends on content negotiation
Ben: we could specify a GRDDL transform in the head of the HTML doc that lives at the end of the NS URI
Steven: Sounds fine
Ben: Is my usage of RDF:bag OK?
Mark: We changed the inheritance
rules, and haven't yet redrafted the RDF/A document
... the example you gave had nested elements in an about, and
I'm not sure if that was the final version
Steven: I am sure we agreed to inherit the about
Mark: I need to check
... especially check the use cases in the rdf/a document
... IPTC are thinking of adopting rdf/a
... one of the examples that someone from IPTC showed asked how
we were going to do lists
... this was one of the reasons I wanted to put the rdf:bag
stuff on hold
... A new requirement from IPTC is the ability to use qnames in
the object as well as the properties
... and I think we do need to think about the bag/list
stuff
... to avoid repetition
... One possibility is to rephrase the example that uses
bags
Ben: can we give you an action?
Mark: Yes
<benadida> ACTION: Mark to check edge cases of inheritance in RDF/A [recorded in http://www.w3.org/2005/06/21-swbp-minutes.html#action01]
Mark: If those are OK, we can tweak the example
Ben: I would like to resolve it before we reply, if possible
Mark: Part of the requirement
from IPTC is to keep the markup short
... and consistency between how you refer to object and
subject
... We could use a special character in hrefs, such as "[" to
mark qnames in href, or different attributes for the two cases,
such as href and qref
Ben: This is for brevity?
Mark: Yes, because their use
cases adds MASSES of metadata
... and apparently in the WAI area this is also an issue
... One way would be to do something special with meta and
link, such as adding extra attributes, but not generalise those
attributes, allowing them only on those elements.
Ben: I think it would be good if IPTC adopted this, so we should consider their requirements
<scribe> ACTION: Mark to create an rdf/a wiki [recorded in http://www.w3.org/2005/06/21-swbp-minutes.html#action02]
Ben: I am convinced, I appreciate your summary
http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2005Jun/0010.html
Ben: But would like more feedback from other members of the TF
<benadida> ACTION: Ben to put together the "ACID" test for XHTML2 RDF/A [recorded in http://www.w3.org/2005/06/21-swbp-minutes.html#action03]
Mark: Another IPTC
requirement
... was that they would have many attributes from different
namespaces on one element
http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2005Jun/0024.html
Mark: we could make any 'foreign'
attribute a predicate
... this would then work with other attributes within
XHTML2
... such as media type
Ben: This sounds a little worrying
Steven: It might break backwards
and forwards compatibility
... since attributes might mean different things between
different versions of XHTML2
Mark: Well, it's only metadata;
it shouldn't matter if it gets added to the triple store, even
if it means something else as well
... the problem is working out what the content of the foreign
attribute is
Ben: I won't be able to come to
the IPTC coordination meeting in July
... Eric Miller will be there
Mark: I have already talked with Eric about it
Steven: So have I; don't know if I can/will be there
Mark: Not clear for me either
Ben: We should coordinate with Eric
<scribe> ACTION: Ben chat to Eric about IPTC [recorded in http://www.w3.org/2005/06/21-swbp-minutes.html#action04]
Mark: Their issues are soluble;
their higher level issue is whether to use XHTML2
... instead of their in-house language
http://www.w3.org/TR/xhtml2/mod-metaAttributes.html#sec_23.2.
Ben: The question is does the wording imply that a rel= implies an rdf statement
?
Mark: Well, the idea is that you are making RDF statements whether you like it or not
Steven: Would you like to replace
"The metadata attributes can be used to generate RDF
statements." with
... The metadata attributes represent RDF statements.
Ben: Yes, that would be fine. I
thought it might just be a wording issue
... It is basically to do with author intent
... whether they really meant to say what we will extract in
rdf triples
... Are there examples where the author wouldn't be generating
rdf?
Steven: Well rel="next" can be
used by the user agent to automatically go to the next
document
... and the author may not have been thinking rdf
... but I see no harm in this also representing rdf
Ben: I want to avoid the situation where there could be any ambiguity in whether the author really meant what is in the markup
<danbri> (vmtf overrunning...)
<benadida> please join us!
Mark: XHTML2 reserves all unprefixed values
Ben: So any prefixed values mean nothing in XHTML2, and can only be interpreted as RDF
Steven: I think we need to say
that there is an equivalence regardless of prefix
... I don't see why it matters
Mark: I'm trying to think of how to word it without being too proscriptive (for instance to allow flexibility in the use of rel="next")
Ben: That brings us to the question of DanBri's schema, and our action item
Steven: Well, most of the questions just bring up questions in my mind, so I would rather discuss it
Ben: We will discuss it next week
then
... Steven/Mark, would you please send suggested wording for
23.2 to the mailing list?
Steven: OK
<scribe> ACTION: Steven send wording for 23.2 to mailing list [recorded in http://www.w3.org/2005/06/21-swbp-minutes.html#action05]
<benadida> Steven, can you generate the minutes?
This is scribe.perl Revision: 1.126 of Date: 2005/05/16 16:49:48 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/;/:/ Succeeded: s/tis/this/ Succeeded: s/woule/would/ Succeeded: s/THis/This/ Succeeded: s/Mark/Ben/ Succeeded: s/WOuld/Would/ Succeeded: s/liek/like/ Succeeded: s/authomat/automat/ Succeeded: s/23.3/23.2/ Found Scribe: Steven Inferring ScribeNick: Steven Default Present: Ben_Adida, Steven, MarkB_ Present: Steven Ben Mark Got date from IRC log name: 21 Jun 2005 Guessing minutes URL: http://www.w3.org/2005/06/21-swbp-minutes.html People with action items: 23.2 ben chat for mark send steven wording[End of scribe.perl diagnostic output]