[Haskell-cafe] function result caching
tomasz.zielonka at gmail.com
Fri Oct 13 09:04:29 EDT 2006
On Fri, Oct 13, 2006 at 08:52:09AM -0400, Jan-Willem Maessen wrote:
> >I guess the reason is balancing. I can't think of any way of
> >balancing a
> >lazy tree that wouldn't break abstraction.
> Uh, Patricia trees aren't balanced in the usual sense.
> There is exactly one tree structure for a given set of keys,
> regardless of insertion order etc. (IntSet and IntMap workes
> approximately as Carl Witty described last I checked, though I won't
> swear to whether bits are taken low to high or vice versa.)
Ah, IntMaps! Forget what I said.
More information about the Haskell-Cafe