Re: [boost] [contract] importance ordering (was "diff n1962")

Subject: Re: [boost] [contract] importance ordering (was "diff n1962")
From: Thorsten Ottosen (thorsten.ottosen_at_[hidden])
Date: 2010-04-14 16:42:42

Lorenzo Caminiti skrev:

> I would personally recommend for Boost.Contract to support importance
> ordering -- what do you think?
> 

I think I would only need two orderings. One for preconditions that are 
equally or more expensive than the actual function and one for those 
that are faster (big-O wise).

But it is not something I have found a need for very often.
is_sorted() on binary_search() is the only immediate one I can find, but 
it seems too expensive even for debug builds.

-Thorsten