We have a version of the consistency checkers, which is great. Some points from our discussion on possible improvements in the redesign of the system: - [ ] Better handling of compound constraints like the nogood propagator. - [ ] Better handling of (half-)reified constraints. - [ ] Allow bounds(R) in the implementation. I suppose we only have one potential candidate for this, multiplication?
We have a version of the consistency checkers, which is great. Some points from our discussion on possible improvements in the redesign of the system: