Yes, you need to fail as well. Did you try switching recomputation off?
Cheers Christian -- Christian Schulte, Professor of Computer Science, KTH, www.ict.kth.se/~cschulte/ -----Original Message----- From: users-boun...@gecode.org [mailto:users-boun...@gecode.org] On Behalf Of kel...@cretin.net Sent: Monday, March 31, 2014 1:00 PM To: users@gecode.org Subject: Re: [gecode-users] commit on already-assigned variable > This is most likely okay and can happen due to recomputation during search: > it might be the case that the variable has already been assigned by > propagation during recomputation before the commit operation is executed. Is the correct behavior in this case to fail if commit() attempts to commit a value which is already outside the possible range? It seems like with IntVars this would be the case. e.g. variable x assigned to 13, i.e. the range [13,13] commit with n = 10 choice 0: x <= 10 --> empty domain, FAILED choice 1: x > 10 --> no change My case is similar: variable x assigned to 1010, i.e. the domain [1010,1010] commit with n = 0001 (i.e. choose this bit to branch upon) choice 0: set bit specified by n to 0 --> no change choice 0: set bit specified by n to 1 --> fail? Cheers, Kellen _______________________________________________ Gecode users mailing list users@gecode.org https://www.gecode.org/mailman/listinfo/gecode-users _______________________________________________ Gecode users mailing list users@gecode.org https://www.gecode.org/mailman/listinfo/gecode-users
