Re: Minutes of the beer session

A while ago we had the discussion of the syntactic issues, with an informal 
discussion between me and Peter in Budapest.


http://lists.w3.org/Archives/Public/www-webont-wg/2003May/0349

Two of the issuettes remain incomplete, and I am still expecting resolution of 
these by the WG.


[[
B). B1, B2 in OWL DL Syntax

We agreed this was desirable:
ACTION: jjc Send proof sketch of extension to correspondence theorem with B1 
B2 under new intersectionOf semantics.
ACTION: pfps review proof sketch of correspondence with B1 B2 
(this replaces old action on pfps to work on this proof, withdrawn under (A) 
above).

C) rdf:LIst decision by RDF Core
We agreed in principle that:
C.1) we could handle this by simply making the rdf:LIst triple optional in the 
mapping rules
C.2) it is better to go further and make the type triple optional on all blank 
nodes in OWL DL

ACTION: jjc send proof sketch that nothing breaks without type triples for 
blank nodes
ACTION: pfps review proof sketch of blank nodes without type triples

Flow chart:
  Proof works  ==yes==> propose C.2)
                    ==no==> propose C.1)

We note that the desciption of OWL Lite and OWL DL  in english would become 
shorter with this change.
]]

The WG agreed with all these actions, and the e-mail log of them is as 
follows:

B)
initial proof jjc
http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0017
review comments pfps
http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0111
more detailed account
http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0294


C)
initial proof jjc
http://lists.w3.org/Archives/Public/www-webont-wg/2003Jun/0302

The result had to be weakened a little, and the semantics of owl:DataRange 
strengthened a little

(at some point I indicated that this proof was in doubt while we had semantic 
layering open, but it is fine now)

*****

I am particularly expecting change on the (B) one, since we discussed this in 
Boston and had consensus there to make the change subject to the technical 
details working out; we found those details hard, probably because there was 
a bug in the last call semantics that has been fixed. So the WG decided, 
mainly on the strength of Peter's technical concerns, to not make this change 
prior to last call, with the understanding that if the action could be 
completed the change would be made. This remained the case at the telecon 
discussing the beer session results.

At the moment I hear Jim ready to declare victory and I am feeling somewhat 
angry. I find the haste (i.e. between the draft to the WG and the msg on the 
comments list) with which a faulty message was sent off to Dave Reynolds 
disconcerting. The reply did not address the text
"A rationale for not permitting this in OWL DL
  should be given, preferably as a test case in OWL Full
  showing an OWL Full non-entailment that would hold in
  OWL DL if such triples were permitted."

The many group members who cheered that reply forward seemed to have failed to 
have read to the end to see that Jim had made a mistake (of course mistakes 
get made, but when checked off by three other people?) 

Jeremy

Received on Tuesday, 22 July 2003 16:17:25 UTC