"Peter Dimov" <pdimov@mmltd.net> wrote in message news:016401c52e5e$0e08abe0$6501a8c0@pdimov2... | Thorsten Ottosen wrote: | > "Daniel James" <daniel@calamity.org.uk> wrote in message | > news:d1napl$8pu$1@sea.gmane.org... | > Peter Dimov wrote: | >> 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? | > | > yes. | > | >> 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. | > | > why should anything *special* monopolize the primary template? Why not | > something *general*? | | There can be at most one primary template (and for ADL customization points, | there should be none). why? In boost.range this is done and it is very convenient that the primary template works with all containers. | When you pick your favorite type category and put | that in a primary template, this makes this type category special with | respect to the primitive. speciel == works | It means that the primitive operates on this | specific category of types. this is your interpretation | hash_value is not a range primitive. It is a value primitive. is vector< sub_range<string> >::value_type a value or a range? -Thorsten