Re: [Core] updated safeness condition

Jos de Bruijn wrote:
> I found that the definition is overly restrictive in the face of
> disjunction; it required all variables to be safe in every disjunct,
> even if the variable does not occur in the head. This is, however, a
> situation we do want to allow.
> I updated the definition yet again.
> 
> This disjunction in core is giving me a headache.
> Would the difficulties of Jos with the definition of safeness be
> sufficient grounds for getting rid of disjunction in core?  ;-)

Works for me :-)

Given that disjunction (and several other things like nested functions) 
are just syntactic sugar. One option might be to define the 
transformation to a stripped down Core without the sugar and define 
safeness in terms of that.

Dave