Re: parallel: bounds checker

[daef]

I've seen that already. But bounding the expression doesn't work that simple 
because it (more or less obviously) breaks the disjoint check.