These minutes have been approved by the Working Group
and are now protected from editing. (See IRC log of
approval discussion.)
(Scribe changed to Achille Fokoue)
Alan Ruttenberg: Agenda amendments?
Alan Ruttenberg: no agenda amendments
PROPOSED: accept previous minutes http://www.w3.org/2007/OWL/wiki/Teleconference.2008.03.05/Minutes
RESOLVED: previous minutes accepted
Registration fees and signup f2f
Peter Patel-Schneider: please register as soon as possible for the next f2f
Alan Ruttenberg: it is fine by me to have few observers at the next f2f meeting
Action item status
Alan Ruttenberg: Progress on Action 93
Alan Ruttenberg: Michael has started working on the OWL Full semantics, and commentts and feedback are welcomed
Michael Schneider: I don't want to force anyone to look at the OWL Full Semantics at this point
PROPOSED: Action 97 completed
Deborah McGuinness: Just joined - Deborah McGuinness - i tried to find the minutes from the last user facing documents meeting and failed. is that posted?
Due and overdue actions
Alan Ruttenberg: 3 out of the 5 due and overdue actions are Jeremy's actions, who is not here today, and 2 are Alan's
Alan Ruttenberg: So they will still be pending
raised issues
Alan Ruttenberg: Issue 96
Michael Schneider: there has already been some discussions on it. It is only an editorial issue
Zakim: pfps, you wanted to say that there is no inconsistency
Alan Ruttenberg: is Issue 96 an valid issue?
Peter Patel-Schneider: it is not an issue
Rinke Hoekstra: I don't think there is an issue, but if michael thinks there is, then that's an issue
Alan Ruttenberg: Should issue 96 be accepted?
Michael Smith: +0.1 on accept, it is a minor issue in RDF, but would change serialization for cosmetic reason (some tools already use the syntax)
Doug Lenat: (not a lot of strong sentiment on this one)
Alan Ruttenberg: I'll talk to Ian about it. I am inclined to accept it tough
Bijan Parsia: Is it subsumable in other issues with RDF mapping? I'd rather all such issues were clustered
Alan Ruttenberg: Issue 100
Alan Ruttenberg: it is not about last week discussion on RDF/XML serialization
Alan Ruttenberg: it came up on discussion related to punning with Boris
Alan Ruttenberg: In OWL 1.1 there seems to be ontologies that cannot be coverted to RDF
Bijan Parsia: does the issue come from the fact that some ontologies cannot be serialized or is it just an OWL DL serialization issue?
Bijan Parsia: let's not decide in the abstract but rather on a case by case basis
Michael Smith: -1 to accept
Issue 100 b/c drawing abstract lines in the sand doesn't seem necessary at this point
Boris Motik: -1 (i.e., I think we should *not* consider this issue; rather, we should reject it)
Michael Schneider: there *are* owl-1.0-dl ontologies, which are *not* expressible in rdf (I *don't* talk about rdf/xml!)
Peter Patel-Schneider: I will have to look carefully to find if it is really a problem
Bijan Parsia: separated vocabulary means e.g., no class is an instance
ACTION: Peter to look more carefully at Issue 100
Bijan Parsia: non-separated vocbualry means that there is some term that is used in more than one syntactic category
Sandro Hawke: achille, I don't think the chair has agreed to this action.
Michael Schneider: it should be accepted
Rinke Hoekstra: -1 think that case-by-case is probably the best way to deal with this "issue"
Alan Ruttenberg: I would accept it for discussion
Boris Motik: Four people voted against accepting the issue; why should we accept it? Nobody voted for it!
Alan Ruttenberg: there is no concensus. So we will accept it
Alan Ruttenberg: Ivan, michael and alan want to accept it
Alan Ruttenberg: Issue 101
Boris Motik: Michael, I was not for accepting the issue; I was for rejecting it.
Michael Smith: covered by statement in semantics doc, b/c such literals are "well formed constants from NV"
Michael Smith: These datatypes, as well as the well-formed constants from NV, are interpreted as specified in [XML Schema Datatypes].
Michael Schneider: meta - what exactly has to happen in order to reject a raised issue? I would say that it should be *strongly* rejected (many -1)
Peter Patel-Schneider: it is just a bug. I will fix it
Alan Ruttenberg: accept the issue as editorial
Alan Ruttenberg: Issue 102
Alan Ruttenberg: what happen to annotations on annotation property
Alan Ruttenberg: Should it be accepted?
Michael Smith: i.e., you can't create an annotation property without using it in an annotation
Boris Motik: the source of the problem is that annotation properties are not entities
Boris Motik: I don't think we need a seperate issue for this
Boris Motik: reject this issue, but take in the comment
Zakim: pfps, you wanted to argue that this is also a bug (but Boris may not agree) and the fix is easy
Peter Patel-Schneider: disagree with Boris
Alan Ruttenberg: to be decided on email
Deborah McGuinness: +1 to any reasonable solution that allows annotations on annotations - it sounds like peters current solution does this
Alan Ruttenberg: peter's solution seems reasonable to me
General discussions
Alan Ruttenberg: formal vote to publish documents is needed
Carsten Lutz: +1 for some more reviewing time for fragments document
Alan Ruttenberg: on the fragments, it is not clear that it should be published by the next f2f
Alan Ruttenberg: reviewers needed for the documents to be published
Alan Ruttenberg: let's start with the fragment document
Alan Ruttenberg: I want to make a distinction between must publish documents as opposed to documents that would be nice to publish
Bijan Parsia: The fragment spec should be published as is
Bijan Parsia: i don't see any showstopper
Alan Ruttenberg: the naming of the fragments is still an open issue
Ivan Herman: there is nothing wrong to ask explicit questions in the published documents
Bijan Parsia: I have no problem with including such questions, I just don't think they are showstoppers
Ivan Herman: it's ok to openly request feedback on open issues
PROPOSED: publish fragments document as is, asap
Achille Fokoue: I feel like it has to be made more accessible. One point raised by Jim, there is an assumption that you are already familiar with the full vocabulary of OWL DL [Scribe assist by
Rinke Hoekstra]
Achille Fokoue: although I think the document is well-written, I think it needs to be more accessible. The production rules assume you're already familiar with the full details of OWL DL. It will not really help get feedback from the people who are only interested in their one fragment. [Scribe assist by
Sandro Hawke]
Bijan Parsia: How about an appendix with a complete grammar for each fragment
Achille Fokoue: My assumption is that people who might usefully review parts of the fragments document, might not be ready to understand all our documents. [Scribe assist by
Sandro Hawke]
Boris Motik: Achille, this document is a specification for three fragments. IMHO it should not be a *guide* for three fragments.
Bijan Parsia: Does having a complete grammar for each fragment move things at all to address this issue?
Michael Smith: we can publish want we want now, and make them self contained later
Ivan Herman: we can publish want we want now, and make them self contained later
Sandro Hawke: Not tutorial, Boris -- Just written in the jargon of the dialect, instead of your jargon.
Deborah McGuinness: at one point, we discussed having a statement about a description of a class of users that each fragment is aimed at
Bijan Parsia: how about an appendix?
ACTION: ppatelsc to check whether all OWL 1.0 ontologies are representable in RDF
trackbot-ng: Created
Action 103 - Check whether all OWL 1.0 ontologies are representable in RDF [on Peter Patel-Schneider - due 2008-03-19].
Bijan Parsia: some example ontologies might also be useful
Bijan Parsia: maybe a little more about the design
Michael Schneider: The fragment document has not been published since the recent important changes
Alan Ruttenberg: There are a number of reasons why we make fragments. I don't think we can say from here what people should be looking for in fragments. I think the sentiment is toward publishing it. [Scribe assist by
Sandro Hawke]
Bijan Parsia: I think boris's overstates...there are many possible fragments with good comptuational properties (e.g., FL family) which are expressively horrid and thus aren't proposed
ACTION: achille start editing the document by the end of next week
trackbot-ng: Created
Action 104 - Start editing the document by the end of next week [on Achille Fokoue - due 2008-03-19].
Deborah McGuinness: While some fragments are computation-oriented, some are not. Some populations which are not DL-literate could benefit from the fragments document. It would be nice if, reading the document, you could see which audience each fragment is aimed at -- and Bijan's idea of identifying which KBs are in which fragments. [Scribe assist by
Sandro Hawke]
Bijan Parsia: I'll note that there is also a fragment intro, then a feature overview...
Achille Fokoue: I want people to be able to read the section on a fragment and understand what it's about, without understanding the rest of OWL. [Scribe assist by
Sandro Hawke]
Alan Ruttenberg: Achille, I'm comfortable with you editing the document in that direction, with the edits being reviewed in 10 days. Alternatively, we can publish more like what we have now, and let your edits wait. [Scribe assist by
Sandro Hawke]
Achille Fokoue: In that case, I will try. But my schedule is very tight, so I might not make it. So I'm going to try the edits on a copied wiki page, in case they do not get done. [Scribe assist by
Sandro Hawke]
Boris Motik: Sometimes you want to make a speculative change to a document. I am in favor of making such copy-and-modify documents often. [Scribe assist by
Sandro Hawke]
Achille Fokoue: +1 for boris
Bijan Parsia: we agree that the current design is ok for publication
Bijan Parsia: It seems like we agree that the current design is okay for publishing; the concerns have been editorial. [Scribe assist by
Sandro Hawke]
Bijan Parsia: I suggest we publish the current draft, pretty much as-is, and note that we are working on the editorial reason. [Scribe assist by
Sandro Hawke]
Alan Ruttenberg: next week, we will have a formal vote on publication
Alan Ruttenberg: reviewrs will be needed for the documents to publish
Michael Schneider: Question - is it possible to compare the differences between two *different* pages in the wiki? This would help when creating speculative copies to see the differences
Alan Ruttenberg: volunteers?
Michael Smith: I will be a reviewer for any of the 3 documents. At present, I am most familiar with fragments.
ACTION: dm to review primer
trackbot-ng: Created
Action 105 - to review primer [on Deborah McGuinness - due 2008-03-19].
ACTION: achille to review the XML doc and the fragment
trackbot-ng: Created
Action 106 - Review the XML doc and the fragment [on Achille Fokoue - due 2008-03-19].
Doug Lenat: Deb, coordinate with me if you want so we can produce a joint review.
Deborah McGuinness: i do not have enough time to review more than one but on the fragment document, can we include a request to describe a class of users for whom each fragment is targetted?
Jeff Pan: (I am only available on IRC today)
ACTION: bijan to review the XML document
trackbot-ng: Created
Action 107 - Review the XML document [on Bijan Parsia - due 2008-03-19].
ACTION: Sandro to review XML document
trackbot-ng: Created
Action 108 - Review XML document [on Sandro Hawke - due 2008-03-19].
ACTION: boajie to review the fragment
ACTION: bijan to review just the fragment document
trackbot-ng: Created
Action 109 - Review just the fragment [on Bijan Parsia - due 2008-03-19].
Jie Bao: my user name is "baojie" on the wiki
ACTION: DougL to review the primer
ACTION: dlm to review the primer
ACTION: msmith to review the fragments document
Sandro Hawke: jiebao, I don't suppose you'd be willing to pick one form of your name and stick to it, for us? (using the name you want to be called first.)
Jie Bao: i will stick to baojie to be consistent to my ids everywhere
ACTION: smith to review the fragments document
trackbot-ng: Created
Action 110 - Review the fragments document [on Michael Smith - due 2008-03-19].
Alan Ruttenberg: which syntax should be used for the primer?
Alan Ruttenberg: should it be addressed before the submit the draft?
Bijan Parsia: "We also recognize Bijan's email about the syntax issues and his very good suggestion of having an automated hookup to the owl api (or similar translator) so for the moment will not address the syntax issue."
Ivan Herman: we have to be careful that any language we use should have a clear and stable description
Alan Ruttenberg: for the short term, can we simply voice Ivan's concern in the status of the document
Peter Patel-Schneider: q+ to say that the Manchester syntax could be turned into a WG note in short order
Peter Patel-Schneider: A Logical Interpretation of RDF, Conen, W., Klapsing, R..Circulated to RDF Interest Group, August 2000.
Ivan Herman: not sure. the issue is for the long term (once the primer becomes a standard)
Peter Patel-Schneider: let's publish it in its current version
Zakim: pfps, you wanted to ask why an informative document cannot reference non-W3 stuff and to say that the Manchester syntax could be turned into a WG note in short order
Ivan Herman: turtle addition will be very valuable for people from RDF background
Zakim: pfps, you wanted to say turtle only if there was a spec that showed how turtle maps to RDF
Deborah McGuinness: could we include an action to include the minutes from the task force where this document was discussed
Bijan Parsia: Manchester syntax was there first partly as a historical reason...we found it easier to compose in that
Deborah McGuinness: *action on alan - include minutes from user facing group discussion of documents on the wiki page for the task force
Bijan Parsia: Translation and incorporation of more syntax is at the moment tedious so we've not done any more than the three we started with
Michael Schneider: but isn't there a "natural" mapping from turtle to ntriples? and a mapping from ntriples to RDF graphs?
Issue Discussions
Alan Ruttenberg: I did not undestand Issue 16
Boris Motik: it has to do with asymmetry
Boris Motik: suggestion: annotations should be external to what is annotated
Alan Ruttenberg: why do we need a new desgin for annotation?
Michael Schneider: if someone would propose a solution to the problem, then I would perhaps understand the problem :)
Bijan Parsia: There was a semantic change...entailments are different
Boris Motik: annotation should be some kind of comments (without any semantics)
Jeremy Carroll: From the rdf point of view, annotations are the substance of their concerns (e.g. Dublin Core)
Alan Ruttenberg: we should continue this discussion on email
Jeremy Carroll: (only some rdf subcommunities - some rdf-ers have same view as hcls)
Boris Motik: An annotation from its very name is something unimportant
Michael Schneider: I understand an "annotation" to *not* inherently belong to the annotated thing. For the case of ontologies: Annotations for parts of an ontology should not add to the semantics of the ontology itself
Bijan Parsia: both parts (annotation and axioms) are important
Boris Motik: +1 to bijan (even though he disagrees with me :-)
Alan Ruttenberg: the current design of annotations breaks some use of ontologies
Alan Ruttenberg: three issues with annotations
Alan Ruttenberg: Clarifying 3 aspects of annotation properties 1) Not being in the same domain of discourse 2) Able to have both individuals and literals as values 3) being able to create associations to classes and properties [Scribe assist by
Alan Ruttenberg]
Bijan Parsia: I think fixing the current backward compatibility point does not affect any o ther design consideration, IMHO
Bijan Parsia: AnnotationProperties puns data and object properties :)