Hi!
While experimenting with the FD structure i found, that there is a discrepance between
the implementation and the docu. According to the website there should be a
"power" function in the structure, which is not there. However there are some
functions in the structure, not mentioned in the docu.
Is there a reason why "power" has gone?
The power propagator has been removed from Gecode because it was buggy,
and the code was not in a state that could be easily repaired.
Therefore, it has also vanished from the Alice interface.
Is a current documentation of the structure available?
Not at the moment. I'll prepare one and make it available.
And i have another remark/request about the usability of the Modeling structure:
I find the alice script much harder to read than the corresponding mozart
scripts.
At least for the examples i came across so far, all constraints are posted to
the same space and typically with the same conlevel.
So the scripts are bloated with unnecessary information.
It would be nice to have a Structure that can be parametrized with a Space and
perhaps a conlevel, such that you can directly write something like:
post'( constraint, conlevel )
or even
post''(constraint)
instead of
post( space, constraint, conlevel );
This might indeed be more convenient for modelling. However, there are
two things that would become a little clumsy: 1. the structures with
implicit spaces cannot be used in models with branch&bound search, as
you have to post constraints in different spaces. 2. for accessing the
solutions, you have to use the procedures from the Reflect
substructures, which again require different spaces. One could of course
just instantiate the functor again each time you need to use a new space.
If you send us a version of the functorized modules, and it really eases
modelling, we'll be happy to include it!
Cheers,
Guido
_______________________________________________
alice-users mailing list
[email protected]
http://www.ps.uni-sb.de/mailman/listinfo/alice-users
