Re: [boost] [Range] initial feedback on return values of algorithms

From: Thorsten Ottosen (tottosen_at_[hidden])
Date: 2006-02-04 08:00:20

David Abrahams wrote:
> Thorsten Ottosen <tottosen_at_[hidden]> writes: 
>>David Abrahams wrote:
>>>Thorsten Ottosen <tottosen_at_[hidden]> writes:
>>>>David Abrahams wrote:

>>>Sure you can; you just look at your implementation and if it is O(1)
>>>you write a specialization of the metafunction (or whatever).
>>
>>Well, I don't personally have access to all the implementations.
>>This seems like overkill.
> 
> 
> std::list is not the only sequence out there that might have O(1) size
> and no random access iterators!  Consider slist, hash_set, ...

right, so what do you propose exactly?

>>I was told that Duffing could actually hurt a modern optimizer.
> 
> 
> Loop unrolling helps unconditionally in this sort of case because you
> eliminate comparisons and branches. 

how much does it help and on which datastructures and for which
types in the container?

-Thorsten