I'm experimenting with fusion and I have a question regarding the compatibiliy/interaction with MPL. in .. /libs/fusion/doc/html/fusion/introduction.html it states: "Fusion provides full round compatibility with MPL. Fusion sequences are fully conforming MPL sequences and MPL sequences are fully compatible with Fusion. You can work with Fusion sequences on MPL if you wish to work solely on types [1]. In MPL, Fusion sequences follow MPL's sequence-type preserving semantics" So when I compare the pages for the concept of AssociativeSequence for fusion: libs/fusion/doc/html/fusion/sequence/concepts/associative_sequence.html and that for MPL libs/mpl/doc/refmanual/associative-sequence.html the expression requirements are not the same. The metafunctions for fusion are: result_of::has_key<S, K>::type result_of::at_key<S, K>::type result_of::value_at_key<S, K>::type while for MPL they are has_key<s,k>::type count<s,k>::type order<s,k>::type at<s,k>::type at<s,k,def>::type key_type<s,x>::type value_type<s,x>::type So how can the statement in the fusion introduction be true? This question comes about as I'm making a new "thing" of my own invention where I want to inter-operate with purely type sequences created with MPL - just likt it says you can do in the documentation. Turns out at significant part of this "boost stew" depends on better compile time checking of the users's metafunction invocation. This ends up creating the need for concepts in both MPL which as far as I know haven't yet been coded. So when I do this, I run accross this question. I realize I can build a fusion sequence from an MPL one, but given the language in the introduction, I would have expected to be able to go both ways and this wouldn't seem to be true. Note that it's quite possible that I'm missing something here as I feel that I'm really starting to get this stuff only for the first time. Robert Ramey