Teleconference.2007.11.14/Minutes

From OWL
Revision as of 05:28, 29 November 2007 by IanHorrocks (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

See also: IRC log

(Scribe changed to Conrad Bock)

Action 9: Sandro will check with Josh about UFDTF zakim setup for tomorrow

{{{who}}}: RRSAgent, pointer?
RRSAgent: See http://www.w3.org/2007/11/14-owl-irc#T18-05-34
Bijan Parsia: I had an email about datetypes as well which might be helpful: http://lists.w3.org/Archives/Public/public-owl-wg/2007Nov/0052.html
Alan Ruttenberg: http://www.w3.org/2007/OWL/tracker/actions

Alan added items to the previoous minutes.

Bijan Parsia: Here- http://www.w3.org/2007/OWL/wiki/Teleconference.2007.11.07/Minutes#Summary_of_Action_Items
Peter Patel-Schneider: other business - the status of Issue 14

Discussion of how action items from previous action items didn't make it into previous minutes.

Action 10: Boris to send an email about Issue 11 fixed and how to fix Issue 28

Action 11: Boris to send an email on Issue 3

Action 12: peter to send an email about your proposal collocated with OWLED

Action 13: ian to send out email soliciting other proposals for the 2nd f2f

Uli Sattler: but this email has been sent out?
Peter Patel-Schneider: there are also lots of order issues in last week's minutes and lack of context

Discussion of minutes editing, remove "chit chat", general readability, 20-30 minutes in cleanup.

Bijan Parsia: I suggest we accept the uncleanuped minutes
Bijan Parsia: I mean, is there anything untrue in them?
Doug Lenat: +1

PROPOSED: accept previous minutes, as is

Peter Patel-Schneider: -1
Bijan Parsia: +1 to accept
Conrad Bock +1: {{{what}}}
Martin Dzbor: +1
Uli Sattler: +1
Michael Smith: +1 to accept previous minutes
Ratnesh Sahay: +1
Peter Patel-Schneider: status of Issue 13 and Issue 14

Vote on whether to accept uncleaned up minutes from previously.

Bijan Parsia: Weren't these resolved?
Alan Ruttenberg: Peter, is there something damaging about accepting these minutes if we agree future ones will be held to a higher standard? [Scribe assist by Sandro Hawke]
Jeremy Carroll: (I tend to feel that meeting mintues should be unanimous)
Alan Ruttenberg: +1
Jeremy Carroll: I'll side with peter if it makes this easier
Evan Wallace: agree with Peter

Action 14: Alan to clean up minutes

Jeremy Carroll: no
Sandro Hawke: proposal rejected

Vote and discussion result: not accept uncleaned up minutes.

Alan Ruttenberg: http://www.w3.org/2007/OWL/tracker/actions/4

ACTION Sandro: to explain how to fix red names on wiki pages.

Evan Wallace: Peter notes that the syntax document is still missing all but its first figure

Action 15: Alan to upload the rest of the inline images in the Syntax document

Alan Ruttenberg: http://lists.w3.org/Archives/Public/public-owl-wg/2007Nov/0283.html

April FTF

Bijan Parsia: Need a survey since most people are not on the call
Conrad Bock +1: {{{what}}}
Peter Patel-Schneider: Agree on the survey
Peter Patel-Schneider: +1
Sandro Hawke: +1 tthese days are fine for me
Evan Wallace: Alan asks if people can make first wk in April for FTF2
Evan Wallace: +1
Peter Patel-Schneider: wiki page looks good to me
Ratnesh Sahay: +1 days are fine
Michael Smith: + dates are fine
Peter Patel-Schneider: and send email, I guess

ACTION peter: post survey on April FTF dates.

Uli Sattler: 0 on the dates
Alan Ruttenberg: 0?
Uli Sattler: 0= i am not sure yet, but i am afraid that i won't be able to make them

Action 16: peter to make wiki page to get feedback on dates for F2F2, and send e-mail about it.

Peter Patel-Schneider: +1 to accept :-)
Bijan Parsia: +1 to accept
Michael Smith: +1 to accept editorial Issue 6, Issue 7. Issue 27. Issue 33. Issue 37.
Ratnesh Sahay: +1 to accept

Alan surveys on accepting five issues in agenda

Jeremy Carroll: +1
Martin Dzbor: +1
Peter Patel-Schneider: s/accept/close/
Doug Lenat: +1 to accept editorial Issue 6, Issue 7. Issue 27. Issue 33. Issue 37
Jeremy Carroll: (I have fixed scribe conventions page wrt :)
Uli Sattler: +1
Jeremy Carroll: wrt colon)

RESOLVED: Close Issue 6, Issue 7. Issue 27. Issue 33. Issue 37 based on Peter's changes.

Datatype issues not in n-ary discussion

Sandro Hawke: http://www.w3.org/2007/OWL/tracker/issues/29
Jeremy Carroll: I do
Uli Sattler: ...jeremy, i can barely understand you, you sound "muffled"?
Bijan Parsia: Really?!?!?
Bijan Parsia: What part of the rdf document forbids that?
Jeremy Carroll: ACTION jeremy to show why rdfs:datatype and owl:DataRange are different

Action 27: jeremy to show why rdfs:datatype and owl:DataRange are different

Bijan Parsia: Using XML Schema's namespace requires doing in a way they can accept. In particular, need approval for using their namespace.

Bijan Parsia: A URI is different from a qualified name. Only true in RDF land.

Bijan Parsia: A qname is *not* an abbreviation for a URI. It is a pair. [Scribe assist by Sandro Hawke]
Bijan Parsia: Concatenation is sanction by the RDF specs, for use in RDF -- but not widely accepted in XML. [Scribe assist by Sandro Hawke]
Jeremy Carroll: +1 to giving XML Schema WG some say
Bijan Parsia: If we're introducing an alternative serialization for XSD definitions, then XML Schema WG should be involved. [Scribe assist by Sandro Hawke]
Alan Ruttenberg: One issue is polutiong (squatting) a namespace. Another is using a term the namespace publisher has defined. [Scribe assist by Sandro Hawke]
Peter Patel-Schneider: Yes, but XSD-WG did not define minInclusive, etc, as being in this namespace. [Scribe assist by Sandro Hawke]
Alan Ruttenberg: http://lists.w3.org/Archives/Public/www-archive/2007Sep/0043.html

Bijan Parsia: Can't use existing element as attribute or vice-versa.

Bijan Parsia: I think there's a syntax issue here, too, eg using minInclusive as an attribute when it's defined as an element, or some such. [Scribe assist by Sandro Hawke]

Bijan; You can derive a URI from a qname, but the XML WG might not approve of the URI.

Alan Ruttenberg: even if they have defined the qname and used it
Peter Patel-Schneider: there is a tag issue on this http://www.w3.org/2001/tag/issues.html#rdfmsQnameUriMapping-6
Alan Ruttenberg: Mike Smith is Liason to the XML Schema working group
Carsten Lutz: After the discussion we had on of n-ary datatypes, I actually believe that using XML Schema for defining
Carsten Lutz: datatypes is a very bad idea in general, i.e., *also for unary datatypes*. I am not sure, though, whether it
Carsten Lutz: makes sense to bring this up now. I will write a mail.
Bijan Parsia: I can't find the translation to RDF of the datatype facets

Action 17: Smith to take the issue of URI/qname approival from XML WG

Jeremy Carroll: fine
Bijan Parsia: Ah, table 5

Action 28: Jeremy to edit Issue 29 to deal with RDFS data type

Alan Ruttenberg: mike will spawn new issue, jeremy will add note to Issue 29 saying it it narrower

Issue 31

Sandro Hawke: trackbot-ng, list users
Sandro Hawke: trackbot-ng, help
trackbot-ng: See http://www.w3.org/2005/06/tracker/ for help (use the IRC bot link)
Bijan Parsia: Did XML schema ever finish component designators?
Sandro Hawke: trackbot-ng, status
Uli Sattler: jeremy, i still can't understand you
Bijan Parsia: q+ to say we should pick a winner

Alan Ruttenberg: We can define datatypes in XML, and if we lefft everything alone, could import documents with XML datatypes. Would be annoying because of one file per datatype.

Uli Sattler: it's an option, but do we need to take their semantics?
Evan Wallace: datatypes could be defined in one document

Jeremy Carroll: Can put multiple datatypes on same file with different ID's.

Uli Sattler: i.e., do we need to take the XSD semantics as well if we use their datatypes?
Peter Patel-Schneider: yes, that would assume that we take the XSD semantics - of course, we could restrict the allowable syntax (I guess)
Bijan Parsia: WE already do :)
Bijan Parsia: perhaps the liaison should ask
Bijan Parsia: I'm on the queue
Zakim: bijan, you wanted to say we should pick a winner
Evan Wallace: -1 on not allowing use of outside defined xml datatypes if we are using the syntax
Jeremy Carroll: yes, I agree with Bijan too
Peter Patel-Schneider: I don't see much need to either point to external XSD documents or reuse the XSD syntax

Bijan Parsia: Need both inlining and referencing datatypes.

Evan Wallace: if we use the XML syntax people are going to expect to be able to reuse XML datatypes
Peter Patel-Schneider: If we do use the XSD syntax, then there is no reason not to allow pointing to external XSD documents
Zakim: jeremy, you wanted to suggest way forward

Bijan Parsia: WOuld rather have one way of referencing datatypes.

Jeremy Carroll: Final document does have one way (picks a winner).

Bijan Parsia: -1 to not solving it even if XML Schema fails

Alan Ruttenberg: Would be good to settle ourrequirements on this.

Peter Patel-Schneider: it's not that we don't need the datatypes, it's that I don't think that we *need* the syntax

Alan Ruttenberg: Heard four opinions on requirements so far.

Bijan Parsia: Submission doesn't make a choice
Peter Patel-Schneider: no mechanism for datatype import in OWL 1.1
Evan Wallace: No.

Peter Patel-Schneider: No mechanism for datatype import in the member submission.

Jeremy Carroll: I think the BP WG Note should be on the table

Alan; Any objection to discussion on email on requirements?

Jeremy Carroll: jena uses that too
Jeremy Carroll: (that = DAML+OIL soln as documented in SWBP WG Note)

Bijan Parsia: No bias in sub mission against importing.

Evan Wallace: MIght as well use best practices, otherwise fall bak to inline.

Bijan Parsia: Some is better than none
Peter Patel-Schneider: where is the DAML+OIL solution?

Alan Ruttenberg: Agreement to discuss on mailing list?

Jeremy Carroll: In a version of BP note, maybe one before last
Bijan Parsia: http://www.w3.org/TR/2005/WD-swbp-xsch-datatypes-20050427/#sec-daml-soln
Jeremy Carroll: q+ to note status
Zakim: jeremy, you wanted to note status

Action 29: Mike to lead discussion on datatype import.

Bijan Parsia: I would rather datatype identification be in a spec, near the inline syntax for user defined datatype, but that's fine
Bijan Parsia: +1 to getting info from XML Schema WG
Evan Wallace: some bits of the issues Jeremy has brought up with n-ary datatype are also in the BP note
Evan Wallace: so this needs to be reviewed before this were to be published by us
Evan Wallace: I would like to be included in the datatypes discussion group
Bijan Parsia: What about XML SCheam syntax?
Jeremy Carroll: (I think I am done, but I am happy for others to continue)
Jeremy Carroll: no opinion

Clarification of Mike's last action: to convene half-hour telecon on data types issue

Uli Sattler: I would like to be included in the datatype sub-group discussion, too

Alan scribing

(Scribe changed to Alan Ruttenberg)

Carsten Lutz: I would also like to be in the datatype group

Ontology elements

Alan Ruttenberg: suggested in the xml syntax have a mandatory uri on the ontology
Michael Smith: for xml schema discussion, I will use the minutes to co-ordinate info, if anyone else wants to be include, be sure to chime in here
Peter Patel-Schneider: what happens with local names if you don't have an ontology name?
Alan Ruttenberg: relax that proposition
Peter Patel-Schneider: ah, right, currently in OWL 1.1 there are no relative IRIs
Sandro Hawke: (sorry, do we have a scribe...?)
Bijan Parsia: No
Alan Ruttenberg: alan raises imports processing issues in relations

(Scribe changed to Sandro Hawke)

Bijan Parsia: Tools should not use the Ontology Name to check for inclusion cycles.

Alan Ruttenberg: but protege-4 does that.

Bijan Parsia: I think that's a bug

Alan Ruttenberg: Matthew believes it's right, I think....

Alan Ruttenberg: And it's not that he's confused. He's sure his approach is right.

Alan Ruttenberg: If this issue can proceed independently, something something something.

Bijan Parsia: Definition of ontology closure can be separated from ontology naming.

Alan Ruttenberg: the basic issue is what happens when you say owl:imports "foo"

Bijan Parsia: We could just use XML Include, and things would come out pretty much the same.

Bijan Parsia: If nothing else is different other than the ontology name.... I don't know what one would make of that.

Bijan Parsia: Import closure reaches a fixpoint when doing more imports doesn't add anything.

Alan Ruttenberg: Your defining an import-processing-model, calling in 'include', in which the ontology name doesn't matter.

Peter Patel-Schneider: in OWL 1.0 the ontology name doesn't matter
Peter Patel-Schneider: in OWL 1.1 the ontology name does matter

Peter Patel-Schneider: in 1.0 import is defined completely independently of ontology names.

Bijan Parsia: REALLY!?!??!!

Peter Patel-Schneider: in 1.1, it's completely different. [ some O....O' definition ]

Bijan Parsia: Omg, OWL 1.1 is broken there

Alan Ruttenberg: If one wants to sign the ontology, ....

Bijan Parsia: """An ontology O directly imports an ontology O' if O contains an import declaration whose value is the ontology URI of O'."""

Peter Patel-Schneider: I'm not saying what it should be, I'm saying what 1.1 says.

Bijan Parsia: I see why matthew was confused
Bijan Parsia: Because OWL 1.1 got it wrong :)

Let's deprecate owl:imports!!!!!!!!!!

Bijan Parsia: +1 to xml:include
Bijan Parsia: http://www.w3.org/TR/owl-semantics/direct.html#3.4
Bijan Parsia: Aside from this local meaning, an owl:imports annotation also imports the contents of another OWL ontology into the current ontology. The imported ontology is the one, if any, that has as name the argument of the imports construct. (This treatment of imports is divorced from Web issues. The intended use of names for OWL ontologies is to make the name be the location of the ontology on the Web, but this is outside of this formal treatment.)

Alan Ruttenberg: there is a requirement from users that people be able to sign their ontologies, include metadata, include versioning, ....

Evan Wallace: Alan notes a common requirement to annotate ontologies with metadata such as author

Bijan Parsia: The 1.1 defn is a bug because it's less-liberal than 1.0, so it's not backward compatible.

MikeSmith, when you use "/me" like that, it's left out of the IRC log.

Peter Patel-Schneider: an email on problems with OWL 1.0 imports would be useful

Bijan Parsia: I think we should align the current specs with OWL 1.1

Bijan Parsia: ... and then think about a deeper redesign, eg using XML Include.

Evan Wallace: Alan proposes a breakout session at F2F1 where we can meet with Alan Rector, etc and talk about reqs

Alan Ruttenberg: I don't know anyone who like owl:import as is.

Evan Wallace: and possible designs to address them
Michael Smith: I note (to be logged) that xml:include may be a path to datatype import as well. Something to revisit.
Uli Sattler: sounds good

Action 18: Alan to brain dump on owl imports issue to mailing list

Peter Patel-Schneider: should rich annotations be on the agenda at F2F1?

Bijan Parsia: I believe alan rector is only available on the first day, but I shall check
Bijan Parsia: Sean and Matthew are available both days

Action 19: Alan to include imports, possibility of annotations as subject for f2f

ADJOURN