"Fredrik Blomqvist" <fredrik_blomqvist@home.se> wrote in message news:cvfhbf$n4i$1@sea.gmane.org... |I haven't studied your proposal in detail yet, but - I think it would be | valuable | for further discussion if you could compare your tree implementation to | Kasper Peeter's version: http://www.damtp.cam.ac.uk/user/kp229/tree/ looks nice. | Kasper did propose his tree to Boost in late 2002: | http://lists.boost.org/MailArchives/boost/msg36876.php | .. altough it didn't seem to gain momentum at the time. well, it does show there is great interest in tree containers. | I've been using Kasper's tree in a couple of smaller personal projects and | would | very much like to see a good generic n-ary tree container in Boost. yeah, I guess a reasonble tree container library would start of with namepace boost { template< class T, class Container = std::vector<T> > class tree; template< int N, class T, class Container = std::vector<T> > class n_ary_tree; } the possibility that we can specify the underlying container is important. Given that, we can easily make a tree of heap-allocated objects by typedef boost::tree<T*, boost::ptr_vector<T> > heap_tree; br Thorsten