Skip to Content.
Sympa Menu

coq-club - [Coq-Club]specification and Type

coq-club AT inria.fr

Subject: The Coq mailing list

List archive

[Coq-Club]specification and Type


chronological Thread 
  • From: "Jean.Duprat" <duprat AT ens-lyon.fr>
  • To: coq-club AT pauillac.inria.fr
  • Subject: [Coq-Club]specification and Type
  • Date: Thu, 02 Mar 2006 10:32:55 +0100
  • List-archive: <http://pauillac.inria.fr/pipermail/coq-club/>

Hi,

In plane geometry, where points are elements of a Set (Parameter Point : Set.)
and lines are defined as property on points (Definition Line := Point -> Prop.),
I would express one of the incidence axiom saying that I can construct three
points, say A, B and C, such that for every line, there exists an algorithm for
finding the point among these three points, that does not belong to the line.

So I try :
Axiom I4 : {A:Point &
           {B:Point &
               {C:Point &
                   (forall D : Line, {~(D A)}+{~(D B)}+{ ~(D C)})
               }
           }
       }.
but Coq refuses since (forall D : Line, {~(D A)}+{~(D B)}+{ ~(D C)}) has
type Type when Set is expected.

I would know if
- I have written a bad formule, and a correct one will be accepted by Coq,
- this is not possible for a fundamental reason,
- or it is not possible with the current version of Coq.

Thanks

   Jean Duprat





Archive powered by MhonArc 2.6.16.

Top of Page