See also: IRC log
Previous: http://www.w3.org/2009/06/04-rdfa-minutes.html
<Steven> Good; I'm at a vFtF at the moment, so only on irc today
benadida: need to leave 15 minutes early
today
... try to get to target of RDFa processing at 1/2 past today
Manu: I talked with Robert Scoble last week
... he's talking about "2010 Web", launching "Building 43 Initiative"
... he'd like to learn more about Semantic Web and would like to do an
interview
... perhaps RDFa could be his first interview
... would you be interesting in being interviewed, Ben?
Ben: sure, nice to keep the current momentum going
<Ralph> Scobleizer
ACTION: Manu create a wiki page for discussion of issue-214 [recorded in http://www.w3.org/2009/06/04-rdfa-minutes.html#action12] [DONE]
http://rdfa.info/wiki/Rdfa-in-html-issues#Mixing_id_and_about_on_the_same_element
ACTION: Ralph find the statement on test suite copyright [recorded in http://www.w3.org/2009/06/04-rdfa-minutes.html#action13] [DONE]
<Ralph> Licenses for W3C Test Suites
ACTION: Ben to author wiki page with charter template for RDFa IG. Manu to provide support where needed. [recorded in http://www.w3.org/2009/05/28-rdfa-minutes.html#action10] [CONTINUES]
ACTION: Ben to put up information on "how to write RDFa" with screencast possibly and instructions on bookmarklet. [recorded in http://www.w3.org/2008/11/06-rdfa-minutes.html#action12] [WITHDRAWN]
ACTION: Ben to prepare "how to write RDFa" screencast with fragment parser [recorded in http://www.w3.org/2009/06/11-rdfa-minutes.html#action05]
markbirbeck: Do we need to keep track of these items? We do this sort of stuff as a natural part of what we do and it eats into the conversation time.
benadida: I will do this and we won't talk about it more until it's done.
ACTION: Manu to write summary for Semantic Web Use Cases for Ivan. [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action09] [CONTINUES]
ACTION: Mark create base wizard suitable for cloning [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action12] [CONTINUES]
markbirbeck: I'll try to work on this at SemTech.
ACTION: Mark to send Ben ubiquity related wizard stuff [recorded in http://www.w3.org/2008/11/20-rdfa-minutes.html#action11] [CONTINUES]
markbirbeck: next step in your fragment parser is to generate these wizards?
ACTION: Mark write foaf examples for wiki [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action13] [CONTINUES]
ACTION: Michael to create 'RDFa for uF users' on RDFa Wiki [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action14] [CONTINUES]
ACTION: Ralph make a request for an RDFa issue tracker instance [recorded in http://www.w3.org/2009/05/28-rdfa-minutes.html#action11] [CONTINUES]
ACTION: Ralph or Steven fix the .htaccess for the XHTML namespace [recorded in http://www.w3.org/2009/01/08-rdfa-minutes.html#action01] [CONTINUES]
benadida: Anybody else we can ping on the issue tracker thing?
Ralph: I'll just do it.
ACTION: Ralph think about RSS+RDFa [recorded in http://www.w3.org/2008/09/11-rdfa-minutes.html#action15] [CONTINUES]
Manu: let's defer discussion on tests 123, 127,
and 128; they're related to other design discussions
... 124, 125, and 126 are ready for discussion
... there's a bug in 124 in the test harness; missing a space
... the email is correct
<Ralph> Suggested test case 0124 [Shane 2009-05-26]
benadida: any thoughts on Philips test
cases?
... I want to be sure we don't overlook http://philip.html5.org/demos/rdfa/results.html
ShaneM1: Just want to go make sure they're thorough - looks good so far.
benadida: Can I run the test suite entirely in Javascript? Do we have a SPARQL implementation for the Javascript test suite?
markbirbeck: Had an issue creating the actual
parser itself.
... I construct JS objects that look like SPARQL and use those.
benadida: Let's talk about that offline.
benadida: Sam was on the call last week? That
was good.
... Appreciate all the work done to minimize the issue to the core
problem.
... Here's an idea...
<Ralph> The target of RDFa processing rules
benadida: There is an idea that Mark has
floated, which is to have an RDFa core processing rules document.
... It can be on an abstract document structure/tree.
... The way we make that happen in the long term, may be to not do that
first, but write an RDFa for HTML5 document that is what Henri suggested.
... Parse with html5lib and then work on the DOM - see if we can get a spec
written up doing that...
... We can have specific rules for DOMs that have namespace support and those
that don't.
... We can then see where the triples differ between html5lib and xhtml.
markbirbeck: Idea sounds okay, but the spec is
already general.
... It says it can work on SAX and "child nodes" - so I think we're talking
about namespace processing.
... Things like processing lower-case attributes.
... I think we can produce an HTML5 spec pretty quickly.
... This idea has come up before, before XHTML+RDFa
<Ralph> [[
<Ralph> Processing need not follow the DOM traversal technique outlined here, although the effect of following some other manner of processing must be the same as if the processing outlined here were followed. The processing model is explained using the idea of DOM traversal which makes it easier to describe (particularly in relation to the [evaluation context]).
<Ralph> ]]
<Ralph> -- http://www.w3.org/TR/rdfa-syntax/#s_model
<Ralph> (that's a pretty DOM-centric way of being "general" :) ]
markbirbeck: We should tightly control the
general processing rules because there is a deep understanding in this group
about what we're trying to achieve.
... If we can write the HTML5+RDFa spec, it's a quicker way of getting these
issues hammered out.
benadida: The rules don't have to be re-written
in any significant way, but we do have to focus on how we get to the DOM.
... First parse with html5lib - then use RDFa processing rules to process
that DOM.
<Zakim> ShaneM, you wanted to ask about processing rules
ShaneM1: It's difficult to divorce the
technical issues from the political/personal issues.
... I'm fundamentally opposed to repeating text in normative
specifications.
... It creates divergence.
... If processing rules are generic enough - and that was the intention with
the current set of rules.
... if the rules aren't sufficiently generic, we must fix it in the
underlying REC.
benadida: I can live with that - we should
revise XHTML+RDFa spec if the generic rules change.
... We should have one version of the rules.
<Zakim> Ralph, you wanted to prefer that the text match as closely as possible
Ralph: It's a shame if we have to have the text
replicated in several places.
... if the choice is to have HTML WG to go off on a divergent path because we
insist that they only reference an external document
... or if we can give them the text to place into their document.
... We should give them the text (which is not preferrable technically).
... as we were drafting our spec, early language didn't get some of this
stuff right and we tweaked the text.
... The fewer tweaks we make, the better off we will be to limit the
introduction of new bugs
... We should preserve as much of the text of the original spec as
possible.
... What if we discover more bugs?
... How much of this should we change?
ShaneM1: Isn't that what the errata process is for?
Ralph: Depends upon the nature of the
change.
... We shouldn't make design changes - only errata type changes are allowed.
We should start from that position.
ShaneM1: Errata is something we can publish
immediately - in order to keep the specs in steps... as we discover problems,
we should publish errata
... important to try and minimize changes, but we've already discovered one
issue that is critical - case sensitivity of CURIEs.
... That's not minor.
markbirbeck: Slightly different way of coming
at this - modularization idea is absolutely right.
... He says we should fix this in the core, but there is no RDFa core.
... Maybe we'll have a first, and second implementation.
... We might want to do an RDFa 1.1 - don't know how this fits with the
processes.
... We might be able to produce a new version of the spec that includes XHTML
and HTML family languages.
... If it's HTML5, run html5lib and skip to core processing rules.
... If it's XHTML, run X and skip to core processing rules.
... Because it's an update on our current spec, we don't introduce
fragmentation.
<ShaneM1> +1 on mark's suggestion assuming we have the ability to introduce material about HTML into the RDFa Syntax specification.
markbirbeck: In terms of the process, maybe this would be smoother approach.
Manu: one of the biggest issues I have is
generating more documents for people to review
... we have a list of issues that are pretty clear at this point
... Sam Ruby and Philip agree these are all the issues currently known
... so premature to decide on how to write an HTML+RDFa document until we've
addressed these issues
... Mark's idea on "RDFa in HTML Languages" could be a way to avoid
divergence
... separate test suites will be needed for HTML4+RDFa, HTML5+RDFa, ...; this
grows exponentially
... I don't think we're far from being able to generate an RDFa 1.1 spec that
addresses the current issue list
<ShaneM1> FWIW it is possible to have the same test suite exercise multiple languages.
benadida: Two quick things.
... ideally this RDFa 1.1 approach would take some time.
... We should provide some alternative to microdata
... I'm worried that doing another REC would take a long time, and it's not
the sort of thing that the HTML5 folks would include in their document.
Manu: why would we not be able to put out an
RDFa 1.1 draft spec in 2 weeks?
... we wouldn't expect this to be included immediately in HTML5
<ShaneM1> process issue - we don't have charter to say anything about HTML in a draft
benadida: How do we address text duplication between XHTML+RDFa and HTML5?
markbirbeck: You're effectively deprecating
RDFa 1.0 with RDFa 1.1.
... The major differences are going to be case sensitivity, maybe @prefix and
@token.
Ralph: You're going to have to work with the
HTML WG.
... You must work with HTML WG RDFa 1.1
benadida: I think we do agree on RDFa working on a model like this - first html5lib then generic RDFa processing rules.
Manu: I think we're saying that something produces a tree-like model, and then the RDFa processing rules operate on that tree-like model.
ShaneM1: Ralph - you said we have to work with HTML WG, I don't think we can say anything about HTML in an updated RDFa 1.1
Ralph: We are not going to get permission to publish RDFa 1.1 without working with HTML WG.
markbirbeck: I disagree with Manu's point that
we should have more discussion before publishing something.
... In the end, commit-then-review may be the way to go forward.
<ShaneM1> Remember that "proposals have momentum"
markbirbeck: It is a convenient way of discussing things.
Ralph: Feel free to do what you want with the editors draft, but they're not going to get WD status.
markbirbeck: I think we're mainly discussing
how to stop the fragmentation of the text.
... There is already some fragmentation - Shane's document and Philip's
document...
... Why can't we create a document and then use it as a discussion point.
... Perhaps we should focus on generating something to the HTML5 community to
show them that we're working on these issues.
... It also stops fragmentation.
Manu: What are we calling that document?
markbirbeck: We talked about RDFa Core.
Ralph: RDF Core means "the core RDF spec itself"
markbirbeck: RDFa Core would contain nothing
but the core processing rules.
... Maybe the other approach is to have one RDFa spec, but for that spec to
be continually updated.
... Maybe the document should contain XHTML, HTML, SVG Tiny.
... Maybe it should include Data+RSS from Yahoo?
... Everything being deferred to the core section of the document?
<Steven> I have to go now (not that I have done any more than lurk)
markbirbeck: Maybe this document is more than just RDFa+XHTML...
ShaneM1: My question is since the short name is rdfa-syntax - if we change the operational title, is it a process issue?
Ralph: It could be done, but there are other
issues.
... The longer we go trying to operate as an independent task force - the
worse it will look for us.
markbirbeck: We could talk about this on the
mailing lists - which we've done.
... We could also generate a document to talk about.
... We should make this document look like what we'd like it to be.
ShaneM1: Maybe Mark and I can work on a
structure of the document offline?
... Our target audience might not find it acceptable.
markbirbeck: We may want to focus on an "RDFa" document instead of several "RDFa in X" documents.
Ralph: The RDFa IG notion is on the table - and that group will be involved in decisions on RDFa moving forward.
markbirbeck: Where do LinkedData people come in?
Ralph: They don't have a formal group yet, if
pressed, they could be part of SWIG.
... SWIG has a practice of creating Task Forces - Healthcare/Life Science
came out of that.
markbirbeck: Wonder if there is a case for stating that RDFa has a place in Linked Data.
Ralph: Good thought, but no formal structure for Linked Data.