From: Peter Dimov (pdimov_at_[hidden])
Date: 2005-03-21 14:16:32
Thorsten Ottosen wrote:
> 3. consider if the primary template should not take 1 range parameter
> and call hash_range;
> unordered containers can then be explicitly overloaded when they
> become available.
> this would also have the great effect that subranges like
> sub_range<string> "just works"
> by default
I don't remember this being discussed during the review, could you please
explain it in more detail? Do you mean that hash_value should have a primary
template that assumes that its argument is a range? If so, I strongly oppose
it. Ranges aren't special from the point of view of hash_value, they
shouldn't monopolize the primary template. The proper way to make hash_value
work with sub_range<T>, as with any other type, is for sub_range to define a
hash_value overload as an inline friend or in its namespace.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk