Skip to Content.
Sympa Menu

coq-club - RE: [Coq-Club] A question concerning certain variant of the Prod rule

coq-club AT inria.fr

Subject: The Coq mailing list

List archive

RE: [Coq-Club] A question concerning certain variant of the Prod rule


chronological Thread 
  • From: "Kouskoulas, Yanni A." <Yanni.Kouskoulas AT jhuapl.edu>
  • To: Matej Kosik <kosik AT fiit.stuba.sk>, "coq-club AT pauillac.inria.fr" <coq-club AT pauillac.inria.fr>
  • Subject: RE: [Coq-Club] A question concerning certain variant of the Prod rule
  • Date: Wed, 15 Apr 2009 07:55:53 -0400
  • Accept-language: en-US
  • Acceptlanguage: en-US
  • List-archive: <http://pauillac.inria.fr/pipermail/coq-club/>

> One of the trivial cases whose description was left out from the book is
> this:
> s = Prop
> s' = s'' = Set
> 
>        E,$B&#(B |- A:Prop      E,$B&#(B::(a:A) |- B:Set
>         ----------------------------------
>             E,$B&#(B |- $B"O(Ba:A,B:Set

Hi. The types of partial functions depend on this triplet, i.e. in

forall n:nat, [0<n -> nat]

the bracketed part uses it, see p 94 in the text.

Also, this triplet is very useful for making strongly specified functions 
that return values along with certificates or guarantees about the data they 
hold. The sig type is an example, see p. 252; its constructor has the type

forall (A:Set) (P:A->Prop) (x:A), [P x -> sig A P]

again the bracketed part depends on this rule.

-Yanni





Archive powered by MhonArc 2.6.16.

Top of Page