XCSP Support

[Nick T.]

Hi,

At http://www.xcsp.org/tools, it's stated that OscaR supports the XCSP3 format.

Is that the case? If so, what's the best API entry point for this? Are there any relevant examples?

TIA,

Nick.

[Pierre Schaus]

Hi Nick,

Yes we are working on it.

It is not yet available though (should be the case for CP2016).

Cheers,

Pierre

--

---
You received this message because you are subscribed to the Google Groups "OscaR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to oscar-user+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Imre Kocsis]

Hi,

could you maybe share the status on this? If I'm correct CP2016 was in September :) and I see you have an "XCSP" branch on BitBucket with much activity lately. Should I check it out (for a project, I would need a CP solver with actual xcsp 3 support...) or wait some more?

Thanks in advance, best regards

Imre

[Pierre Schaus]

We are getting close but this is not yet fully functional (some constraints are still missing and it is not tested enough).

There will be a XCSP competition next year. That will be a big incentive for us to progress.

Out of curiosity, are you looking for any XCSP solver that could work, or is it OscaR specifically that you would need?

If this is only XCSP solver (on the JVM) probably the solver of Christophe Lecoutre (abscon) is the one that is the most ready.

Cheers,

Pierre

To unsubscribe from this group and stop receiving emails from it, send an email to oscar-user+unsubscribe@googlegroups.com.

[Imre Kocsis]

Hi,

at this point, it's pretty much any good solver that supports XCSP. Let's just say XCSP is an external project requirement (that I happen not to fully agree with). From my point of view problem description in MiniZinc would be equally fine. Running on a JVM is nice to have, but not an immediate requirement, either.

On the other hand, although I don't _need_ Oscar per se now (my current models are stock-standard finite step unrollings of composite automaton behaviors), I _am_ thinking about trying out a few things in it. Mixing various OR techniques in Scala seems to be a very neat possibility. I hope I can find the time in the near future to tinker with it.

Thanks for the pointer, earlier I dismissed abscon ("is planned to be published within the next few months", says its website); I will write to Prof. Lecoutre and ask about it. Hopefully I won't have to resort to using the XCSP Java parser directly.

Thanks, I appreciate the fast answer, best regards

Imre