See also: IRC log
Note due to technical difficulties with Zakim, much of the automated support for meetings was down. The quality of these minutes may have been impacted.
PROPOSED: to approve GRDDL WG Weekly -- 21 March 2007 as a true record
These minutes do not have a correct list of who was there. At least BrianS, jjc, HarryH, DanC, simone, rreck and Chime were there.
<scribe> ACTION: jjc to correct the minutes from 21 March 2007 to get full list of attendees. [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action01]
RESOLUTION: to approve http://lists.w3.org/Archives/Public/public-grddl-wg/2007Mar/att-0139/21-grddl-wg-minutes.html pending jjc's corrections to attendee list.
* PROPOSED: to meet again Wed, 4 April 11:00-0500. John-l will scribe
Samsung and Citigroup hadn't filled out the web-form.
<scribe> ACTION:HarryH To communicate to Ian and Steve to figure out if Citigroup and Samsung not being in good standing. [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action02]
We may have more resources thanks to W3C Fellows Program.
<DanC> ACTION: [WITHDRAWN] Ian to clarify profileTransformation for JJC [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action03]
<jjc-scribe> > [[
<jjc-scribe> > The transformation property relates the XPath document nodes
<jjc-scribe> > to an RDF graph. These need not use RDF/XML as an intermediate
<jjc-scribe> > stage. To give an XSLT example, see testlist1#atomttl1, in which
<jjc-scribe> > the attribute-value media-type="text/rdf+n3" on the xsl:output
<jjc-scribe> > element indicates a different media type, from the default
<jjc-scribe> > value, within GRDDL, of "application/rdf+xml". GRDDL agents
<jjc-scribe> > that can process such a media type, can then produce an RDF graph
<jjc-scribe> > in accordance with the media type. Non-XSLT transforms may
<jjc-scribe> > indicate the RDF graph in some other, unspecified, fashion.
<jjc-scribe> > ]]
<scribe> ACTION: DanC to look at: http://lists.w3.org/Archives/Public/public-grddl-comments/2007JanMar/0079. [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action04]
Chime believes jjc's text addresses his concerns.
<DanC> 1.239 has a raw paste of the text in 0079, with a @@todo
Jeremy's previous non-normative suggestions on GRDDL Spec re validation and caching
I could see some informative text suggesting that you really try to minimize access to DTDs, namespace docs, etc.
DanC: Stick the namespace in your document explicitly and not rely on DTD
Chime: Does this address entity reference concern?
<DanC> I'm OK to to take an action around "stick the namespace in your document ; don't rely on getting it from the DTD"
jjc: Something will go wrong if you don't have DTD in there.
DanC: This would show up at GRDDL level.
<jjc-scribe> jjc: entity in html gets transformed into rdf content
I don't understand "who's on the phone"
<scribe> ACTION: jjc to to draft text to "don't write transforms that depend on whether or not its validated" [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action05]
Validation is not explicitly required by GRDDL.
My suggestion is that we add an explicit sentence in Faithful Infoset saying that "Validation is not explicitly required by GRDDL."
<jjc-scribe> text at end of section 6 in spec
<chimezie> I think commentary on entity issues should be placed after the faithful infoset section
<jjc-scribe> ACTION: jjc to draft test cases on validation [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action06]
<scribe> ACTION: Chime send out "We're still thinking about your comments, Stefano" [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action07]
jjc: I've added Jena tests to the pending list.
<DanC> "are you satisfied?" msg to ERH: http://lists.w3.org/Archives/Public/public-grddl-comments/2007JanMar/0078.html
<scribe> ACTION: jjc to update Elliotte with the fact that the tests that address his concerns are in pending. [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action08]
Discussion of Tony's concerns that RDF and RDF/XML are not adequately distinguished.
Tony did say: "Best close it out here. :)
Chime: thinks the informative rules help here.
<chimezie> ?RDFXML gspec:rdfParse ?G.
<chimezie> in particular
<DanC> "If an information resource IR is represented by a conforming RDF/XML document[RDFX], then the RDF graph represented by that document is a GRDDL result of IR."
<chimezie> " * Whenever the RDF/XML spec says that an RDF/XML document with root node ROOT represents a graph G, we have ?ROOT gspec:rdfParse ?G.
<chimezie> Some from XPath and XSLT:"
Jeremy suggested that above picture from Primer does not adequately differentiate the two concepts.
<scribe> ACTION: DanC to respond to Tony Hammond [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action09]
<scribe> ACTION: Chime to make a visual representation of RDF graph and RDF/XML difference, HarryH to check in when done for Primer [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action11]
<jjc-scribe> jeremy says that the above picture in primer seems to suffer the problem
<DanC> (the diagrams in the primer were made with MS ppt, I believe. I wonder if chime's diagram will look funky by comparison)
<jjc-scribe> discussion of oracle and product shipping
<scribe> ACTION: Ask Susie about GRDDL and Oracle deployment plans, and possible comments on spec if they actually read spec.. [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action12]
Scribe note: I'm not sure why this link is here, there was some discussion of these issues under item 3
<DanC> 1.240 <p class="ed">@@<a href="http://lists.w3.org/Archives/Public/public-grddl-comments/2007JanMar/0079.html">27 March from jjc</a></p>
<DanC> (if I haven't ack'd a comment by updating http://www.w3.org/2004/01/rdxh/spec in some way, please assume I forgot your comment and send it again next week)
<chimezie> The test case should call out to the spec about the right of the agent to follow a policy that doesn't allow GRDDL mechanism loops
<scribe> ACTION: jjc to modify #loopx to relate to discussion about SHOULD. [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action13]
<chimezie> where is harry's wand? (Scribe: Harry had just magicked Zakim into doing something, but not enough to make sense of these minutes).
<DanC> (Zakim has not regained all of his marbles)
Discussion of ACTION: Chime to fix the relative URL problem
Chime: A test-documet with internal anchors that are full URIs to test items.
<DanC> (I was trying to figure out who has the ball on test cases publication. Harry, I recommend you track the critical path much more clearly, at least in weekly agendas)
<jjc-scribe> chime - test doc has anchor to sections
<jjc-scribe> should have links to test material to test directory
<chimezie> 1) A test document
<jjc-scribe> also has URIs for test instances that will be in GRDDL results
(I thought it was pretty clear and the working draft had been published.)
<chimezie> with it's own uri and fragments to test section
<DanC> (eek. I advise against minting http://www.w3.org/TR/grddl-tests/#atomttl1 ; if it's been minted, I advise against further investment in it)
<DanC> (eek. I advise against http://www.w3.org/TR/grddl-tests/atom-grddl.xml too)
<DanC> just keep http://www.w3.org/2001/sw/grddl-wg/td/testlist1#atomttl1
<chimezie> so the test material should always be the uri in the test 'repository' - grddl-wg/td/...
<jjc-scribe> Danc suggests dropping fragments in munging of testlist1 into TR space
<jjc-scribe> in order to avoid having multiple uris for same resource
Chime: I feel like we have consensus on test material.
Chime: Feels strongly that we unify URIs for human and machine-readable purposes.
<john-l> <.w3.org/TR/grddl-tests/#atomttl1> a <http://xmlns.com/foaf/0.1/Document>; rdfs:sameAs <http://www.w3.org/2001/sw/grddl-wg/td/testlist1.html#atomttl1> ?
<jjc-scribe> Concerning how OWL did it: Description: (informative) <I4.6/Manifest003#test>
<jjc-scribe> links explicitly to http://www.w3.org/2002/03owlt/I4.6/Manifest003#test
<jjc-scribe> for the earl ouput chime is ambivalent about which uri to use?
<DanC> (I look forward to progress on: ACTION: jjc to produce EARL for his implementation, and then sort out multiple test passing in his EARL. )
We've published: http://www.w3.org/TR/grddl-tests/
Now we need to fix the relative URI in the next publication of test cases.
Discussion of whether we can/should fix problem with published test WD. Apparently we can, within 24hrs of publication in TR space. Discussion of what action to take, consensus builds around http redirects on offending content, in a .htaccess file
<DanC> options are 404 (NO), 410 (maybe), 303 ...
<DanC> 301 moved permanently works for me
<scribe> ACTION: jjc to manufacture .htaccess file for test cases using 301 "Moved Permanently" and e-mail Jean Gui. [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action15]
<DanC> the action is not just to email jean-gui; it's to get the redirect installed, with jean-gui's help
<scribe> ACTION: Chime to announce to email@example.com that GRDDL Test Cases is a WD. [recorded in http://www.w3.org/2007/03/28-grddl-wg-minutes.html#action16]