Are slots first class objects in Stk3.0?? (was Re: STk plans...)

From: J.Marie Kubek <kubek_at_gerii.insa-tlse.fr>
Date: Fri, 6 Oct 95 17:55:36 +0100

>>>>> "Erick" == Erick Gallesio <eg_at_kaolin.unice.fr> writes:

    Erick> I hope to be able to put the beta1 version of STk3.0 this
    Erick> evening (tommorow morning, if not). I think it is
    Erick> relatively stable and it integrate Tk 4.0.

Hi,
        I would like to know if there will be some changes in the
Object model for Stk3.0. More precisely, will slots become
object instances, as already are methods, generics and classes ?

        I think such a modification will be an improvement. It will be
easier:
        + to define new allocation's scheme for virtual slots
        + to define new facets for a slot, such as demons activated on
        a value change.

        I'm specially interested in the first problem as I'm trying to
extend Stk towards Constraint (logic) programming. In my case some slots
are actually allocated as prolog variables or as prolog relations
governing values for the "variable's slots". Relations and variables
are then handled by some prolog process.

        As I'm investigating different solutions I'm getting some
troubles : trying to extend an already existing code involving slot
manipulation is really a pain.

        In case such modifications are not done in Stk3.0, I'm ready
to devote time to 'em. I ran into the code of the "stklos.stk" file for
stk2.2 and I think required changes may be done without too much
nuisance. But surely I'm wrong :-) Any thought is welcome...

Cordialement,
        Jean-Marie.
Received on Fri Oct 06 1995 - 17:46:08 CET