Re: An approach to xsd:dateTime

Ian Horrocks wrote:
 >> -    In the current proposal, every dateTime specifies a “moment” 
 >>represented by a point on the timeline.  No time interval entities 
 >>would be supported.  The lexical values for dateTime must contain 
 >>values for Time of Day components, so this wouldn’t be  misleading. 
  >>Note that this would be in stark contrast to OWL  Time’s Date-Time 
 >>support, where the granularity of the time entity  indicated is 
 >>determined by the finest attribute of  DateTimeDescription component 
 >>values provided.
 >
 >
 >I didn't quite understand this -- OWL didn't have any Date-Time 
 >support did it?

Sorry for not being more explicit.  OWL Time is the name of the Time
ontology created by a DAML project.  The name is confusing, I agree.
In the text above, I was referring to that ontology, not any OWL
language support for time.

 >All seems reasonable. I presume, however, that we wouldn't actually 
 >*spec* this? -- we would leave it up to applications to implement 
 >whatever repair strategies seem most appropriate (although Issue-56 
 >does talk about specifying "standard" repairs).
 >
 >Ian

No, we don't need to spec. these things.  I am just trying to understand
the ramifications of what we do put in the spec. (i.e. what it will and
won't support, and how we can coerce it into doing what users will need).

-Evan