See also: IRC log
<HarryH> Scribe: bwm
<HarryH> Scribe next week?
<HarryH> brian suda is scribe
<HarryH> PROPOSED: to approve http://lists.w3.org/Archives/Public/public-grddl-wg/2006Dec/att-0011/06-grddl-wg-minutes.html__charset_us-ascii as a true record
<HarryH> RESOLVED: http://lists.w3.org/Archives/Public/public-grddl-wg/2006Dec/att-0011/06-grddl-wg-minutes.html__charset_us-ascii as a true record of Dec 6 2006 GRDDL Meeting
<FabienG> sends regrets from 20 Dec until wednesday 10 Jan included
<HarryH> bwm: Distinction between examples and test-case
<HarryH> bwm: DanC agrees distinction, but seems to want put everything in test-suite
<HarryH> bwm: Normative test-cases
<HarryH> bwm: Informative examples
<HarryH> Previous two lines were me
<HarryH> bwm: whether or not test-cases or normative is orthogonal.
<HarryH> DanC puts everything in test-cases
<HarryH> In charter, demonstrate how we work RDFa
harry: we must have relationship
... I would be happy to separate out examples that link to use case doc from the test suite
... we do stuff like the rdf/a as an example
... is that ok?
bwm: for me, yes, but we need DanC to decide.
Fabian: there are other issues
with this example
... the use of rel attribute in link elements in GRDDL collides with use in RDF/a
... we should look at this closely
harry: are you on rdf/a
Fabian: I'm on the mailing
... and submitted this example to them
... the group is not at full speed right now
hh: could you forward any responses ...
Fabian: no one noticed this
<HarryH> PROPOSAL: separate examples like RDFa that are covered by the charter from test suite and link examples to Use Case Document.
<FabienG> Brian's proposal: "The root cause of the problem here may be that RDF/a treats the value of
<FabienG> the rel attribute of a link element as a Curie and GRDDL doesn't. A
<FabienG> possible solution would be to mod GRDDL so that the rel attribute value
<FabienG> is also treated as a curie e.g. "grddl:transformation" with the grddl
<FabienG> namespace prefix defined."
Fabian: I'm in favour of the Brian's solution
hh: is everyone ok with these proposals
<HarryH> CURIEs might be controversial with DanC and the TAG.
bwm: good leave these for future discussion
<HarryH> We'll try to consensus on these with DanC and anyone else at next meeting
re action Fabien to make SAWSDL test case
fabian: my opinion is that we
... most of the SAWSDL annotation could be translated to RDF
... not sure what to do with xml schema part
... what is the base url
... I can't answer these right now
... we should wait
hh: I thought they were in last call
Fabien: they are currently at working draft stage
hh: is the wg aware of this issue
Fabian: I am not part of that group
hh: they have a comments list
Fabian: I could do post a comment
<HarryH> ACTION: Fabien to post to sawsdl list relevant questions about RDF mapping and relationship to GRDDL [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action01]
hh: I asked semweb cg to register turtle type
<HarryH> XInclude: What if document you're including as GRDDL links?
<HarryH> bwm: Both Ian and bwm thought both should be under control of author.
<HarryH> bwm: possibility of a transform-all link as opposed to transform?
hh: I have prejudice to keep
... i.e. define the 'atomic' operation of GRDDL applied to a single document
... and worry about more complicated structures later
<HarryH> By "single" document do you mean infoset pre-Xinclude or post-Xincllude?
<HarryH> Don't assume XInclulde is being done.
hh: that's the way I'm leaning to
<HarryH> Should we consensus?
<HarryH> bwm: 2 different issues
<HarryH> bwm: what about DTDs and Schemas validation?
<HarryH> bwm: what about XInclude?
<HarryH> bwm: Not assuming XInclude processing
<HarryH> bwm: No opinion about DTD Schema validation
hh: the order of processing of
xml documents is not in our charter
... murray is our liason the group responsible for that
... they aren't going to decide for a year
... it means that not every grddl implementatio will treat includes and validation the same
... but its in the author's remit to say what should be done
<HarryH> bwm: publisher of document should have control
<HarryH> bwm: of document.
<HarryH> bwm: They should know what they are publishing
<HarryH> bwm: No use to them to say XInclude may or may not get run.
<HarryH> bwm: Because they may not have control of transform they need answer.
murray: grddl can't answer this
... ultimately its up to the publisher or the transform author to answer this question
... a little snippet of xml transform could recognize an xinclude
... and puts a comment in the graph that something is missing
... doesn't belong in the spec
... but maybe in the tutorial
... we could publish a transform that did this
hh: you are suggesting an informative statement
... we could say in the spec there are dragons around xincludes
hh: I agree with that
bwm: are publishers responsible for triples that are produced by transforming an included document
murray: yes - xinclude specifies
... xinclude processing tends to happen on the server side
... here is a simple expedient
... we ask for a transform in the XI namespace
... as soon as you use xinclude that transform is available
hh: that is good thinking
murray: we don't control the namespace, we'd have to ask
hh: in whose domain is xinclude
murray: dunno - ask Liam Quinn
<HarryH> ACTION: Murray to e-mail Liam Quinn and DanC about possibility of GRDDL XInclude transform in XInclude namespace. [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action02]
<HarryH> PROPOSAL: That we do not mandante XInclude or XML Processing Order on the XML input document, and write a caveat in the spec.
<HarryH> Murray: XML transforms operates on an Infoset
<HarryH> Bwm: DTD or XML Schema transforms an Infoset by adding default values.
<HarryH> bwm: So if you run a Schema then GRDDL, you would get different RDF from the one that you ran pre-Schema
<HarryH> murray: So if you don't follow your nose you don't get the the full results, and we can't dictate following your nose.
<HarryH> bwm: what is the minimum amount of RDF published?
<HarryH> murray: The minimum is the output of the transformation run against the serialization of source document.
bwm: from the publishers point of view there are two questions - what must be in the GRDDL result of a transform and could be in the GRDDL result of a transform
murray: I think its wrong to define a minimum
fabien: DanC has written email
saying he is ok with content
... but is not happy with the word "scraping"
<FabienG> Scraping the web: Steffen wants to build a directory of the people he works with.
fabien: at the beginning there was discussion about ensuring that grddl was about transforms not on scraping
hh: do you have a transform attached to non xhtml compliant docuemnt
<HarryH> maybe just have a GRDDL transform attached to non-XML compliant HTML?
fabien: I tried to cover all the
... there is one where you have to tidy it
... there is an example of trying to extract as much as possible
... and that might be considered to be scraping
hh: could you phrase it as "you
are receiving a document and transforming it according to a
... could we just lose the word scrping from the title
fabien: DanC's message started with not sure wehther to use it and ended with maybe we should use it
murray: scraping is a bad word
stop using it
... there are multiple definitions
... its a pejorative term
... people have trouble with it
... so lets find a different word
<HarryH> Murray: "Extracting machine-friendly data from web-pages" :)
murray: try extracting machine friendly data from web pages
hh: fine with me
<FabienG> ACTION: fabien to remove the word scraping [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action03]
hh: I'll leave this in Fabien's hands
<HarryH> Propose: Given that a base URI parameter is a parameter whose value is
<HarryH> the base URI of the transform source document, the WG RESOLVES not to
<HarryH> define a base URI parameter for transforms, noting that triples
<HarryH> referring to transformed document, or documents named relative to the
<HarryH> transformed document can be created as illustrated in test case
test case files are in http://lists.w3.org/Archives/Public/public-grddl-wg/2006Dec/0011.html
correction should be http://lists.w3.org/Archives/Public/public-grddl-wg/2006Dec/0028.html
hh: why did we have a suggestion for a base param
murray: so that relative uri references work
hh: murray does brian proposal resolve t his
murray: I'd ask Dan
<HarryH> Would this cause problems for if the source doucment didn't end in http://www.ibiblio.org/hhalpin/links.html
<HarryH> Only use of base-param would be with would we want to don't to add a #?
hh: what do other people think
murray: I'm happy
<HarryH> PROPOSAL: Given that a base URI parameter is a parameter whose value is
<HarryH> the base URI of the transform source document, the WG RESOLVES not to
<HarryH> define a base URI parameter for transforms.
<briansuda> has to abstain - need to catch-up on back
<HarryH> I approve.
<HarryH> bwm: not quite quorum.
<HarryH> PROPOSAL should continue, HarryH as a proxy for bwm if he's not at next meeting.
<HarryH> Proxy to vote for "yes"
<HarryH> We've checked a new version of primer.
<HarryH> And all that remains is double-checking SPARQL and manufacturing more example data.
<briansuda> i have emailed HarryH some files
<HarryH> BrianSuda has manufactured most of the example data.
I nominate Harry as my proxy to vote on the baseparam proposal if I'm not at meeting where it is consdiered
<HarryH> ACTION: [DONE] HarryH to respond to Chime, clarifying song/album and discussing bNode vs URI. [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action04]
<HarryH> ACTION [DONE]: bwm to review testlist1#rdfa1.
<HarryH> ACTION [DONE]: Fabien to make a SAWSDL test sketch
<HarryH> ACTION:[DONE] bwm to review testlist1#rdfa1 [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action05]
<HarryH> ACTION:[DONE] Fabien to make a SAWSDL test sketch [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action06]
<HarryH> ACTION: [DONE] HarryH to poke semweb cg re N3 media type registration [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action07]
<HarryH> ACTION: DanC to add N3/turtle mime type to Atom/turtle test case. noting the unregistered status [CONTINUES] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action08]
<HarryH> ACTION:DanC to write rules about XSLT 1.0 processing context [CONTINUES] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action09]
<HarryH> ACTION: HarryH to integrate Murray's XInclude test case sketch into the test collection [CONTINUES] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action10]
<HarryH> ACTION:Fabien to add a tidy/tag-soup use case/paragraph, with caveats [DONE] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action11]
<HarryH> CTION: Ian to reconsider comments on cross-document introduction [CONTINUES]
<HarryH> ACTION: Ian to reconsider comments on cross-document introduction [CONTINUES] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action12]
<HarryH> * ACTION: iand to construct a content negotiation test case [CONTINUES]
<HarryH> ACTION:BWM to produce ~3 test cases for #issue-base-param [CONTINUES] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action13]
<HarryH> ACTION: Harry and Brian to rewrite second part of primer to use Brian and Dan's instance data [CONTINUES] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action14]
<HarryH> ACTION: DanC to add a sample implementation appendix to the GRDDL spec. [CONTINUES] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action15]
<HarryH> ACTION: iand to construct a content negotiation test case [CONTINUES] [recorded in http://www.w3.org/2006/12/13-grddl-wg-minutes.html#action16]
<FabienG> Chair: Harry
This is scribe.perl Revision: 1.127 of Date: 2005/08/16 15:12:03 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/that/ post a comment/ Succeeded: s/bwm/murray/ Succeeded: s/scaping/scraping/ Found Scribe: bwm Inferring ScribeNick: bwm WARNING: No "Present: ... " found! Possibly Present: Bwm CTION Fabian Fabien FabienG HarryH IPcaller Murray_Maloney P7 PROPOSAL PROPOSED Propose Sophia XInclude briansuda harry hh inserted murray murry You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy Got date from IRC log name: 13 Dec 2006 Guessing minutes URL: http://www.w3.org/2006/12/13-grddl-wg-minutes.html People with action items: brian bwm danc fabien harry harryh ian iand murray[End of scribe.perl diagnostic output]