Re: [TED] ACTION-306: suggestions for abstract syntax

Gary Hallmark wrote:
>>
>>> I like the ability to have ground facts not wrapped in a forall.
>>> I don't like the ability to have free variables (not scoped in a forall)
>>
>> This is specifically excluded in the deifinition I propose.
> 
> Well, it's not excluded in the syntax.  I'd rather have syntactic 
> enforcement rather than "footnote enforcement" (where possible using a 
> context-free grammar).

I'd like that too, but...

> I proposed a syntax that clearly forces wrapping 
> variables in a forall AND clearly avoids having to wrap ground facts in 
> a forall.

...your proposal, like the current abstract syntax and other 
alternatives I considered, forces an Implies to be wrapped in a Forall; 
it does not preclude variables to be left hanging free in the Implies, 
however.

Example:

Forall X, P(X,Y) :- Q(X,Y)

So, your proposal, like mine, the current abstract syntax and any other 
solution I have seen proposed or I could imagine, requires what you 
poetically call "footnote enforcement" :-(

I think Harold said during the discussion Tuesday that that requirement 
of having all the free variables within the scope of a forall could not 
be caught using a context-free grammar: Harold, is that what you said?

Just for the sake of completeness, because I am not sure that it has any 
practical impact: your proposed design requires that a ground Implies be 
wrapped in a Forall, too...

Cheers,

Christian

Received on Thursday, 14 June 2007 17:30:55 UTC