See also: IRC log
<HarryH> Convene GRDDL WG Meeting of 2007-03-07
<HarryH> Scribe: BrianS
<HarryH> PROPOSED: to approve GRDDL WG Weekly -- 28 Feb 2007 as a true record
<HarryH> RESOLVED: GRDDL WG Weekly -- 28 Feb 2007 as a true record
<HarryH> Chime to scribe next meeting.
<HarryH> http://lists.w3.org/Archives/Public/public-grddl-wg/2007Mar/0021.html
<HarryH> everyone needs to fillout the form from Ian
<Murray>, the form is for AC not members?
<HarryH> BrianS?
<brianS>i emailed the microformats list,
<brianS> i also got Operator to support some GRDDL
<brianS>nothing more at the moment
<Harry>, other advocacy ideas? popular XML dialects
<Danny>, working to get Profile URIs out there
<Danny>indifference in getting them set-up from the MF list
<Danny>the XMDPs need to mint a good profile URI
<john-l> Technical question: what happens when microformats and GRDDL interpret a document's content in conflicting ways?
<Danny>lack of profile URI doesn't seem to be a problem
<Danny>, create some bad formats
<Danny>, XFN has profile, but probably not GRDDLed
<HarryH> ACTION: Danja to update the ESW wiki to have list of profile URIs and a sort of check-box for which ones are GRDDL-enabled or not. [recorded in http://www.w3.org/2007/03/07-grddl-wg-minutes.html#action01]
<danja> http://dannyayers.com/misc/microformats/soupdragon
<Danny>, created a test case to find issues
<Harry>, primer document take-up
<Murray>, can give it a solid review
<Harry>, is looking for a co-editor on the Primer
<Jeremy>, can review it as well
<Harry>, needs to push the changes to the primer
<Harry>, people think the primer should go rec track? needs reviewer and an editorial changes
<HarryH> ACTION: Jeremy and Murray to review Primer after HarryH's adds in changs [recorded in http://www.w3.org/2007/03/07-grddl-wg-minutes.html#action02]
<Harry>, should the test cases go to rec?
<HarryH> Chime to edit the Test Cases.
<Harry>there is questions about normative, versus others
<Harry>help from jeremy?
<Jeremy>, i can help, but not co-edit
<John>, i can help test cases documents
<HarryH> Chime is definitely editor of doc.
<Chime>, order test cases, by normitive
<Harry>, base-param is normative
<Zakim> jjc, you wanted to mention OWL test approval process
<Chime>, can order them himself,. no need to do this now
<Jeremy>, on the OWL list got approved if two implementations passed it
<Jeremy>approval by software, only objections we discussed
<Chime>, good coverage by implementation
<Jeremy>, pending tests for a long time, until implementations caught-up. Then tests got approved quickly
<Chime>, do test cases cover all the issues
<Jeremy>, needs more work on the schema profile
<HarryH> Using EARL to generate HTMl from implementations.
<Harry>, this helps to show which tests pass, etc
<Harry>, for next week, which tests exist and which don't
<HarryH> ACTION: Chime to take pending tests for approval from agenda and see if at least 2 implementations pass them. [recorded in http://www.w3.org/2007/03/07-grddl-wg-minutes.html#action03]
<HarryH> http://lists.w3.org/Archives/Public/public-grddl-wg/2007Feb/0212.html
<Harry>, discussion about rel="transformation" issues when multiple values
<Murray>, sees the text as just fine... others mis-read it
<Murray>, sent some words to fix this normative issues
<HarryH> http://lists.w3.org/Archives/Public/public-grddl-wg/2007Mar/0015.html
<Harry>, security issuses from multiple people
<HarryH> ACTION: Murray to resolve potential ambiguity with multiple values in transformations. [recorded in http://www.w3.org/2007/03/07-grddl-wg-minutes.html#action04]
<Jeremy>, nothing new from Last Call....
<Jeremy>adding texts to specs, so implementations can reference the spec in dangerous places
<Jeremy>end users take the risks, but only after we give/list those risks
<Harry>, hard to document zero results as ok?
<Harry>, GRDDL agent can not produce no results based on local policy
<jjc> http://lists.w3.org/Archives/Public/public-grddl-wg/2007Mar/0037
<Harry>and still pass the tests
<Murray>, can't pass the tests if there was no result
<Jeremy>, a conforming GRDDL agent that produces nothing, is worthless
<Jeremy>non-issue
<Chime>, conformance label requires GRDDL agents to produce a reult
<Chime>, if we are pushing conformance, then we should remove the ambiguity
<Harry>, GRDDL agent can alter local policy to produce results, or test cases go normative and then you need results
<chimezie> I prefer we have a rec track test document without scenarios where there are no results
<Jeremy>, doesn't think this should be a worry. Reality, people may or may not use the label ....
<Jeremy>mis-use of the label doesn't get them sued, so what is the benefit of labeling....
Jeremy, requiring a GRDDL agent to set secuirty policy to none
Harry, add that to the test case, so all agents must turn off secuirty policy before running tests.
Murray, useful to have a paragraph which describes how to set the secuirty to match output
<HarryH> ACTION: Chime adding sentence to Test Case Doc specifying that local security policy must be set to none before running tests [recorded in http://www.w3.org/2007/03/07-grddl-wg-minutes.html#action05]
<Harry>, emails and comments about Javascript developers
<HarryH> http://lists.w3.org/Archives/Public/public-grddl-comments/2007JanMar/0068.html
<Harry> for javascript developers
<john-l> ... and basically any other language besides XSLT
<Harry> he wanted more guidance
<Chime>, his concerns were about non-normative texts
<Chime> be more specific about
which languages supported
... there is reply to his comments in the archives
<Chime>impossible to implement a transform with JUST javascript
<Chime>, the output format issue language gave explicit examples with XSLTs
<Danny>, thinks it is impossible to implement directly in Javascript
<Danny> seperate document which
describes how it should be done
... redirection, similar to the profiles
<chimezie> 1) Sections preceeding normative sections seemed to suggest the spec covered non-XSLT scenarios, but they don't 2) The output of transforms should either be only RDF/XML or at least associated with a mime-type
<Danny>, just leave it as it is
<Harry>, continue dialog with stefano, and get feedback
<danja> my latest blog comment on the matter - http://dannyayers.com/2007/03/03/grddl-and-language-neutrality
<Harry>, it was underspecifed by RDFa, you can output as HTML/RDFa, JSON, etc
<Harry> need more experience with non-XSLT
<chimezie> we would be faced with the same issue that XProc would if they attempted to accomodate non-workflow-oriented components in an XML Pipeline
<Jeremy>, XSLT2 implementations as well
<danja> chimezie, couldn't an XML pipeline include a Javascript component?
<Chime>, for a time we tried to specific a mimeType comming out
<Danny>., N3 output example, spec was silent on the final representation
<Chime>, greenboxes only refer to RDF-XML as output
<Harry>, i was to be an RDF graph not just RDF-XML
<Harry>as output
<Harry> rules in the spec that go
against the WG decisions
... informative text to say the Graph can be RDF-XML, N3,
etc
<chimezie> http://lists.w3.org/Archives/Public/public-grddl-wg/2007Mar/0022.html
<chimezie> is the suggested change to accomodate mime-type-based parsing
<HarryH> ACTION: To deal with suggested changes for mime-type based parsing and RDF/XML issues in spec. [recorded in http://www.w3.org/2007/03/07-grddl-wg-minutes.html#action06]
<Jeremy>, go through a couple quick tests?
<Jeremy> content-negotiation on language and mime-type
<Jeremy> is the merge of those two results correct?
<chimezie> I agree
<Jeremy> get all the representations and merge ALL the results? the spec is more about get one rep based on your preferences
<HarryH> ACTION: HarryH to update #langconneg test-case to deal with merge from different representations [recorded in http://www.w3.org/2007/03/07-grddl-wg-minutes.html#action07]
<Jeremy>, SVG test has a similar issue
<HarryH> Meeting adjourned
AJOURNED