RE: Proposed response to Golbeck regarding imports issue

Massimo:
[[
As an aside, note the related counter-can-of-worms: leaving this unspecified, 
there is the potential risk that there is no minimal official "conformance 
level" an OWL reasoner could rely on as far as imports are concerned. Which 
leaves the way open to abuses (suppose you put some info, and then you import 
your license statements in another file... some apps could consider them, 
some other ignore them).
]]

I note that the only relevant OWL processors that we have defined are OWL 
Syntax Checkers and OWL Consisteny Checkers.

The former MUST process the imports closure, unambiguously.

The latter SHOULD process the imports closure (unless an inconsistency is 
detected), since an OWL Consistency Checker SHOULD NOT return Unknown.
Complete consistency checkers MUST process the imports closure.


Personally I feel confortable with this level of normativity for imports 
(SHOULD rather than MUST)

Jeremy

Received on Monday, 26 May 2003 10:28:21 UTC