Re: [Core] new safeness condition

 From the new definition:

"A rule implication φ :- ψ is safe if ?V1, ..., ?Vn are the variables appearing 
in ψ or φ, ?V1, ..., and ?Vn are safe in ψ, in the context of ψ, and there is a 
mapping θ from the variables to {b, u} such that "

I'm having trouble understanding this context notion and to exemplify my 
misunderstanding I can't say whether "are safe in ψ, in the context of ψ," 
should be "are safe in φ, in the context of ψ," or "are safe in ψ, in the 
context of φ,".  Though I suspect the former, in other words the conclusion 
needs to be safe "in the context of" the condition, which I think is there to 
ensure equality doesn't sneak into the conclusion?


Jos de Bruijn wrote:
> I found a problem with the definition when considering equality atoms.
> I revised the definition; this should fix the problem.
> Best, Jos
> Jos de Bruijn wrote:
>> I completed ACTION-687: Write a proposed new definition of the safeness
>> restriction
>> Please criticize.
>> I suspect it can be a little more concisely.  When I find some time I
>> will go over it again.
>> I also invite anyone who is interested to go over the functions and
>> predicates in DTB and check whether the binding patterns defined are
>> appropriate.
>> Best, Jos

Dr. Christopher A. Welty                    IBM Watson Research Center
+1.914.784.7055                             19 Skyline Dr.                           Hawthorne, NY 10532

Received on Tuesday, 3 February 2009 15:54:54 UTC