Re: constraints order

[Daniel Diaz]

A FD solver is not complete. This means that you can "trust" the value 
of a variable only when you enumerate it (using for instance 
fd_labeling). In the case of your constraint and since you have not 
defined initial domains for X and Y the solver will will take a long 
time to try each value for X and to see there is no good value for Y.


Michel Levy wrote:
> I do not understand why these two requests have different answers :
> | ?- X #\=Y, X #=Y.
> X = _#2(0..268435455)
> Y = _#21(0..268435455)
> | ?-  X #=Y, X #\= Y.
> no.
> This contradicts the assumption that the order of the constraints is 
> unimportant.
>
> Furthemore, in the official documentation, I read :
> fd_all_different(List) constrains all variables in List to take distinct 
> values. This is equivalent to posting an inequality constraint for each 
> pair of variables.
>
> But this equivalence is no true :
> | ?- X #=Y, fd_all_different([X,Y]).
> X = _#0(0..268435455)
> Y = _#0(0..268435455)
> yes
>
> Can you explain me the difference ?