ERT WG

Check-in on open action items

<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2007Oct/0014

SAZ: check for open action items

http://pruebas.fundacionctic.org/uaw/Pointers_DRAFT.rdf

http://lists.w3.org/Archives/Public/public-wai-ert/2007Sep/0026.html

SAZ: what about RDF schema for pointer methods?

CI: have an early draft
... based on the last model we discussed about
... some comments about open issues in the schema

SAZ: CV what about content draft?

CV: work in progress...

SAZ: TP next week, try to get some advance in the meanwhile if BTW allows
... DR would you be able to read the last HTTP in RDF?

<shadi> ACTION: Shadi finalize the updated "EARL 1.0 Schema Editors Draft" [recorded in http://www.w3.org/2007/10/31-er-minutes.html#action01]

DR: OK

<shadi> ACTION: CarlosV and Johannes prepare first "Content" Editors Draft [recorded in http://www.w3.org/2007/10/31-er-minutes.html#action02]

<shadi> ACTION: David peer-review HTTP Vocabulary in RDF Editors Draft and send coments to the list [recorded in http://www.w3.org/2007/10/31-er-minutes.html#action03]

SAZ: interesting to know relation with MWBP checker moki vocabulary

Review HTTP Vocabulary in RDF

<shadi> http://lists.w3.org/Archives/Public/public-wai-ert/2007Oct/0015

DR: can't help, not in this TF

<shadi> http://www.w3.org/WAI/ER/HTTP/WD-HTTP-in-RDF-20070914

<shadi> http://www.w3.org/WAI/ER/HTTP/WD-HTTP-in-RDF-20070914#bodyProperty

SAZ: some open comments in HTTP in RDF

JK: 2.2.1 body property
... you can have various representations of the same content

<shadi> http://www.w3.org/TR/rdf-schema/#ch_alt

JK: byte, text, xml...
... how to model?
... could use rdf:Alt?

<shadi> "The first member of the container, i.e. the value of the rdf:_1 property, is the default choice."

SAZ: what should be the default representation may be another discussion

JK: maybe Bag too generally

SAZ: certainly not a rdf:Seq
... options are rdf:Alt rdf:Bag or no list at all

CV: prefer not list at all

SAZ: what is your reasoning to avoid using the semantic of a list

CI: the key question is they are equivalent not alternative
... you can choose one or several
... musn´t be forced to choose just one

JK: can create a new object, a sort of equivalent container

SAZ: not sure about the value
... let's keep it for now and put an editorial's note asking for feedback

JK: is your proposal havin multiple bodies properties?

SAZ: yes
... looks good, just a little bit of clean up and wait for the Content

<shadi> http://www.w3.org/2006/http#

<shadi> http://www.w3.org/2006/http-headers#

<shadi> s/http://www.w3.org/2006/http-headers#/http://www.w3.org/2006/http-header#

JK: references for files, not namespaces
... rdf, not rdfs

Draft RDFS schema for Pointer Methods

<shadi> http://pruebas.fundacionctic.org/uaw/Pointers_DRAFT.rdf

CI: subclass rdf:Bag and reuse it for specific purpose
... to use the rdf:li items

JK: think that should be legal, yes

CI: easier to understand this approach, and also more logic
... have containers with groups of pointers
... we can think about Sequences or Alternative lists too

JK: problem with RDF lists is that you can't describe the types of object in the lists
... it would be legal to add *anything* into the group

CI: can't subclass the rdf:li property
... to change the range of the property?

SAZ: did you try running this code through some of the RDF validators and parsers?

CI: it is valid, haven't tried subclassing the rdf:li property

<JohannesK> <http://www.w3.org/TR/rdf-schema/#ch_container>

JK: there is no real rdf:li, it is a variable property

<JohannesK> <http://www.w3.org/TR/rdf-schema/#ch_containermembershipproperty>

<JohannesK> <http://www.w3.org/TR/rdf-schema/#ch_member>

SAZ: the RDF validator focuses primarily on the schema and validity, others try to explore the semantics
... mainly OWL validators, they spot important issues such as if certain restrictions can not be met
... there is a mail about some of these parsers somewhere on the mailing list

JK: no range for the reference property?

SAZ: no range, it could be anything

[discussion if should have a range, and what it would be]

<scribe> ACTION: CarlosI to finalize draft RDFS (check with validators/parsers), and send comments/questions to the mailing loist [recorded in http://www.w3.org/2007/10/31-er-minutes.html#action04]

<scribe> ACTION: all to review draft RDFS proposal by CarlosI and send comments to the list [recorded in http://www.w3.org/2007/10/31-er-minutes.html#action05]

CI: should we make equivalent pointer class a subclass of single pointer?
... all the equivalents point to the same place

SAZ: sound good, would be easier to see in the hierarchy model (see thread on the mailing list)

JK: reference properties could be different between the container and the pointers

CI: we have the same problem in many other instances

JK: we removed it from the equivalent pointer class to avoid this problem, we are reintroducing it by using single pointer

CI: we have the same problem in compound pointer and elsewhere

JK: agree, we also agreed that it may sometimes make sense to have different references (although the pointers are equivalent)
... however, what is the logic in this case?

CI: we have to think more about the reference and see how we can provde more guidance

Next Meeting

next meeting tentatively 14 Novemeber (regrets from Shadi)

meeting after on 21 November