[m-dev.] Reconsidering initialisation of solver types

[Peter Stuckey]

On 27/07/2006, at 3:37 PM, Ralph Becket wrote:

> Peter Stuckey, Thursday, 27 July 2006:
>> It makes Mercury unusable as a constraint programming language!
>
> Currently automatic initialisation seems to make Mercury a bit of a  
> pain
> as a constraint programming language :-)

True.


>> It will add a burden to Cadmium to initialize all variables  
>> explicitly!
>
> I thought Cadmium was a Zinc to Zinc translator.  The interpreter
> actually runs the model and it *has* to initialise all the solver
> variables.

OK it leaves the burden to someone.

I agree that for G12 automatic initialization is not required.

I suppose what I dont have is the feeling on how many errors are caused
for G12 because of automatic initialization.

Presumably this is what you are discovering now?

PEter
--------------------------------------------------------------------------
mercury-developers mailing list
Post messages to:       mercury-developers at csse.unimelb.edu.au
Administrative Queries: owner-mercury-developers at csse.unimelb.edu.au
Subscriptions:          mercury-developers-request at csse.unimelb.edu.au
--------------------------------------------------------------------------