With GNU Prolog 1.5.0 (64 bits) on Win 10 with 16GB RAM, and the somewhat maximized settings (i.e. anything I try and increment further just makes gprolog fail at launch):
The following test code simply fails (no warnings or errors, just the call fails), up until I uncomment F #= 4 (I happen to know the solution), then it works:
pu4(Vs) :-
Vs = [A,B,C,D,E,F,G,H,L],
fd_domain(Vs, 1, 9),
fd_all_different(Vs),
A #= 9, B #= 1, C #= 2,
D #= 5, E #= 3, % F #= 4,
% G #= 7, H #= 6, L #= 8,
100*A*E*H + 10*A*E*L + 10*A*F*H + A*F*L +
100*B*D*H + 10*B*D*L + 10*C*D*H + C*D*L +
100*B*E*G + 10*B*F*G + 10*C*E*G + C*F*G #=
1000*B*E*H + 100*B*E*L + 100*B*F*H + 10*B*F*L +
100*C*E*H + 10*C*E*L + 10*C*F*H + C*F*L,
fd_labelingff(Vs).
Maybe it is indeed overflowing memory-wise, but I suppose it should issue an error in that case.
With GNU Prolog 1.5.0 (64 bits) on Win 10 with 16GB RAM, and the somewhat maximized settings (i.e. anything I try and increment further just makes gprolog fail at launch):
The following test code simply fails (no warnings or errors, just the call fails), up until I uncomment
F #= 4
(I happen to know the solution), then it works:Maybe it is indeed overflowing memory-wise, but I suppose it should issue an error in that case.