Re: Full nonentailment and consistency tests

>> We are not expecting complete Full reasoners, so we are not expecting 
>> to prove the Full nonentailments or consistency tests.

>This is not correct. It is perfectly possible to have an incomplete
>reasoner that can prove some or all of the Full nonentailments or
>consistency tests.

I wasn't trying to talk about what is possible, more about what we are 
expecting ...

I am intending to enumerate the (approved) Full Nonentailments and Consistency 
tests and suggest we demote them to extra credit, which does not prohibit 
implementations from proving them but merely marks it as not within *our* 

Given the logical possibility of automatically constructing the proofs would 
you oppose such reclassification?


Received on Monday, 15 September 2003 09:20:20 UTC