Archives for: November 2008

Tuesday, November 25th 2008

Permalink 04:45:24 pm, Categories: Meeting summaries

Catching up on the blog!

Ah! I've finally got around to updating the blog. Apologies for being rather lax in this regard. The good news is that our 3 Rec Track documents, plus the Primer and POWDER-S vocabulary, have all been updated. See the WG homepage for links. That same page now also has links to running code - with more on the way in the coming couple of weeks. The request to register 'describedby' as a relationship type has been made to IANA (just need to sort out the MIME type request as well). We're only meeting every other week now, so next telecon is due for Monday 1st December. The group is open to Last Call comments through until Friday 5th December.
Phil ARCHER
Permalink 04:41:01 pm, Categories: Meeting summaries

Meeting Summary 17th November 2008

Almost all of the group's documents are being published today as a Second Last Call (with a signal that the group does not intend to make a separate call for implementations). Among other things this will be a trigger to contact IANA about registering the 'describedby' relationship type and the two POWDER MIME types. Meanwhile progress is being made on various code items:
  • The XSLT for POWDER to POWDER-BASE should be available any time now with the transformation from POWDER-BASE to POWDER-S to follow around the end of this week.
  • A POWDER-S processor that uses Jena and Pellet exists with code available at source forge but this needs additional documentation before we can point to it as an implementation. There may be one or two minor additions to make as well.
  • A POWDER-Processor written in PHP is under development and should be available by the end of the month.
  • A (Perl) validator is online now that also does the POWDER to POWDER-BASE transformation. This is very nearly complete will be moved to validator.w3.org before long.
  • The transformation function is also available as a separate script.
  • Much of the code is also used in a second POWDER Processor.
The important point to note here being that there is a growing body of running code that will be available within the next 2-3 weeks. We remain on target to complete our work and seek transition to Proposed Recommendation before Christmas. The group spent a little time discussing the future of POWDER. Areas such as building the trust model and using POWDER in profile-matching systems are those that seem most likely to be developed in future.
Phil ARCHER
Permalink 04:39:04 pm, Categories: Meeting summaries

Meeting Summary 3rd November 2008

The group took a look at recent work done on a test implementation of adding POWDER to a portal - it all looks solid and another strong case for HTTP Link! We also looked briefly at the work from Korea on a microformat based on POWDER. This is, of course, interesting and is a good indication of what is and is not clear about our work to date. The group will engage directly with Jonathan Jeong on this point. The Rec Track documents are all being finalised this week ready for publication next week. It was decided at the TPAC face to face that we will make a Second Last Call announcement but indicate that we do not intent to make a separate call for implementations (CR). The LC period should end Friday 28 November - giving us enough time to seek the transition to PR before Christmas. We are in touch with the RIF over perhaps including a new informative section in the formal document that would show how POWDER could be expressed in RIF. This will be added if there is time to do so, but this week is the deadline for making any further changes to our rec track documents if we are to hit the bigger deadlines.
Phil ARCHER
Permalink 04:37:45 pm, Categories: Meeting summaries

Summary of Face to Face Meeting 23 - 24 october

We got a lot of work done across the 2 days but it's hard to blog the detail. The time was taken up with line by line discussion of the Primer, POWDER-S vocabulary and our 3 Recommendation Track documents. The end results should be published soon as Second Last Call versions of those documents but with a note that the WG does not expect to make a separate call for test implementations (CR phase) and instead shoot straight for Proposed Recommendation in December.
Phil ARCHER
Permalink 04:29:03 pm, Categories: Meeting summaries

Meeting Summary 6 October 2008

Following the e-mail discussion, the group formally resolved to 'drop support for arbitrary RDF' in a POWDER document. This means that attribution information must be provided in a separate file, not within the POWDER doc itself; and that descriptor sets will only have terms for which the object is a literal or an RDF resource. This means that the feature at risk flagged in the formal document has been removed. POWDER-S is unaffected. Another issue highlighted in the current (Last Call) documents is that of the range of the issuedby property. After a lot of discussion it has been resolved that the range of the wdrs:issuedby property will be defined as being a class that includes, but is not limited to, foaf:Agent and dcterms:Agent. Resolving this also made clear that a POWDER-S document is an owl:Ontology, not an owl:Thing which has implications for what can and can't be said out it. Finally, we discussed the QA requirements as they relate to the various Rec Track documents. Conformance sections to be added! Activity in the next week should take us very close to being ready to seek transition to CR for the 3 Rec Track documents. A POWDER-S processor already exists (at NCSR as part of the Quatro Project) and at least one POWDER Processor is nearly complete. The group decided at its last face to face that the CR exit criteria would be two POWDER Processors and one POWDER-S processor - so we're well on target.
Phil ARCHER
Permalink 11:30:28 am, Categories: Meeting summaries

Meeting Summary 13 October 2008

The main substantive topic of conversation today concerned the issue of whether it was right to use foaf:depiction as an annotation property in a descriptor set class (in POWDER-S) as described by Andrea. The answer is no, it isn't! Hence, it has been resolved that we will define our own term wdrs:logo. The problem does not occur with dcterms:description since, unlike FOAF, dc does not define a domain of owl:Thing for its properties. The group also noted the substantial comments received today from Michael Schneider. These will be answered in the coming days. Finally, the group had a brief discussion about plans for the transition to CR. Preparation for this is due to be completed before the face to face meeting next week at TPAC and the end of CR should fall around the end of November.
Phil ARCHER 1 comment