Re: Web Semantics of Datasets (v0.2)

[snip]

I share these concerns. Time is important but it will be fiendishly hard to
formalise anything. Also note that these issues afflict many of our basic
building blocks. For eg, OWL InverseFunctionalProperty. We might imagine a
strictly static version or one that allowed the property to take different
values at different times, each truthful in its time. Or a half-dozen subtle
variations. Devilish details to fornalise and many existing Rdf/owl
properties have specs that don't give any clue what (if anything) their
authors thought on this point.

For an EU project --- fun stuff

For a Web standard --- we're not ready to go there

Dan