Uniqueness constraints on variables are broken in Clingo

I have a question into Potassco mailing list on this, will post a reply if I get one.  

I have a satisfiable Clingo program with this in it:



1{ does(robot,a,1); does(robot,b,1); does(robot,c,1) }1 :- time(1).

1{ does(robot,a,2); does(robot,b,2); does(robot,c,2) }1 :- time(2).

1{ does(robot,a,3); does(robot,b,3); does(robot,c,3) }1 :- time(3).

1{ does(robot,a,4); does(robot,b,4); does(robot,c,4) }1 :- time(4).

1{ does(robot,a,5); does(robot,b,5); does(robot,c,5) }1 :- time(5).

1{ does(robot,a,6); does(robot,b,6); does(robot,c,6) }1 :- time(6).



If I then add



1{ does(robot,a,T); does(robot,b,T); does(robot,c,T) }1 :- time(T).



the program is Unsatisfiable, even though my reading of the manual is that the latter summarizes and is consistent with the former.  (Also note I have to use ";" instead of "," to separate uniqueness set options, this is also a deviation from the Clingo manual, I have another question into them about that.)

It seems like Clingo is adding an extra quantifier in there so that I can only have one move for all times T if I use the latter construct.