On 13/05/2011, at 2:21 PM, David Nolen wrote: > - instance? or custom predicate matches done w/ if
I did quite a bit of this in the context of a Smalltalk implementation. One of the key features is fast type testing. I note in your example case you are relying on an id generic to provide the type tags that are type-hierarchy aware. As long as your type tests are wrt a tree, rather than a graph, then maintaining a DFS number on the tree gives very cheap subclass testing. In a dynamic language you have to maintain that number in the face of changes to the tree, which can be optimised in various ways such as an initial sparse allocation, with an increase in sparseness as you note parts of the tree being updated dynamically. Also you need to allow your dispatch functions to either be updated in response to tree changes (not a great idea) or to use a level of indirection so that you reference the type tree in some way in your test and get the appropriate id range on the fly (and, of course, maintain the bounding pair of ids in the tree for each node). Antony Blakey ------------- CTO, Linkuistics Pty Ltd Ph: 0438 840 787 If you pick up a starving dog and make him prosperous, he will not bite you. This is the principal difference between a man and a dog. -- Mark Twain -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
