- From: Simon Cox via GitHub <sysbot+gh@w3.org>
- Date: Wed, 22 May 2019 05:43:34 +0000
- To: public-sdwig@w3.org
In [Allen's original formulation](https://www.sciencedirect.com/science/article/pii/0004370284900080?via%3Dihub) all temporal entities are considered to be **intervals**. Time instants are merely intervals for which the beginning and end cannot be separated using the available level of precision. So in principle the topological relations should also be used on instants - as the degenerate form However, in OWL-Time we followed the [2006 draft by Hobbs and Pan](http://www.w3.org/TR/2006/WD-owl-time-20060927/) - see https://www.w3.org/TR/2006/WD-owl-time-20060927/#summary which specifies that both domain and range of the interval relations are `ProperInterval` - which is the subclass of `Interval` in which beginning and end can be distinguished. So under the definitions in OWL-Time you can't use the topological relations for instants :-( Was this a mistake? Perhaps. Would correcting it (make the domain and range of the interval relations `time:Interval`, or even just `time:TemporalEntity) be disruptive? Not much. It would change entailments, but wouldn't 'invalidate' existing data. Its a pretty straightforward change though it would affect all the text definitions and global constraints in [these property definitions](https://www.w3.org/TR/owl-time/#time:intervalAfter). Should we consider correcting it ... ? Not sure. -- GitHub Notification of comment by dr-shorthair Please view or discuss this issue at https://github.com/w3c/sdw/issues/1126#issuecomment-494660402 using your GitHub account
Received on Wednesday, 22 May 2019 05:43:36 UTC