[gecode-users] rtest_eq_dom doesnt lead to RT_TRUE !?!?!

[Martin Mann]

Greetings,

ok... i see, the costs for "normal" domain propagation is high. but if i 
m right you are using several special cases during propagation too.

also there exists already the function
rtest_eq_dom(View x, int n)

so why isnt it used for the special case that one of the variables is 
bound in the rtest_eq_dom(View x, View y) function? this would directly 
handle the problem using the fast bound or range handling.

Thanks,
Martin


> Hi.
> 
>> I found a strange behavior of the Gecode::rel function using IRT_NQ as
>> an reified constraint.
>>
>> I used 2 variables with the following domains:
>>
>> d1 = (1)
>> d2 = (0,2)
>>
>> also I ve got a reified constraint like that
>>
>> rel(home, d1, IRT_NQ, d2, isNQ, ICL_DOM)
>>
>> so isNQ should be true if d1 an d2 are unequal!
>>
>> BUT: isNQ is NOT assigned!
> 
> Giving ICL_DOM as a parameter does not always mean that domain consistency is 
> enforced. It only means that propagation will not be stronger than the level 
> given.
> For (dis)equality, of course domain propagation is easy - nevertheless, it is 
> significantly more expensive than bounds propagation. That's why it is 
> currently not implemented.
> 
> Cheers,
> 	Guido
>