From OWL
Jump to: navigation, search

See also: IRC log

(Scribe changed to Michael Smith)

Agenda at

Michael Smith: introduces agenda, rqmts for in-line, external data types

Bijan Parsia:

Alan Ruttenberg: innteresting to hear what the motivation is for in-line datatypes. what would go wrong with just external?

Bijan Parsia: people don't like having definitions in external file, experience from Protege 3.x

Bijan Parsia: avoid overhead for always naming types

Alan Ruttenberg: is there an argument that necessitates a separate syntax for RDF/OWL?

Bijan Parsia: the argument is that its not clear tools could find XML Schema syntax in-line

Bijan Parsia: some argue its more familiar for people from RDF background

Jeremy Carroll: unnamed datatypes are advantageous, use-case for in-line user defined datatypes

Jeremy Carroll: issue with serialization of XMLLiterals in RDF/XML, particularly qnames in attributes

Bijan Parsia:

Uli Sattler: issue with XML schema types is closely related to using other types in OWL

Bijan Parsia: I can almost define rationals in xml schema
Bijan Parsia: Definiteately in XML schema + an additional constraint
Uli Sattler: bijan, ...i see - but will they look like the "true" xml schema types?
Bijan Parsia: yep

Alan Ruttenberg: concern over defiing, e.g. owl:rational

Bijan Parsia: owl extensibility requires URIs. owl:rational could be defined, doesn't seem to be a big deal

Bijan Parsia: clarifying Uli, Alan, are you saying reusing XML Schema syntax might put burden on future datatypes

Uli Sattler: concern was that other datatypes look like native types

Bijan Parsia: yes, I believe that they would

Bijan Parsia: we could use a relax ng type syntax as an inline syntax

Jeremy Carroll: take it that XML Schema types are a given, largely b/c OWL 1.0 supported XML Schema types

Jeremy Carroll: query about how the value space mapping for rational could work

discussion of representation of rational as pair of integers

discussion of XSCD, what they offer, are needed

Jeremy Carroll: the rqmt for us, is less than all of XSCD. we could take less of a solution if it provided re-usable URIs

Michael Smith: a query has been sent to the XML Schema WG chair on the status of XSCD

Alan Ruttenberg: can we say that if URI is present you can reference external

Alan Ruttenberg: suggest we move forward with that proposal, on Wed telecon

RRSAgent: I have made the request to generate MikeSmith