2010/3/10 Steve Purcell <st...@sanityinc.com>: > On 9 Mar 2010, at 23:22, Michał Marczyk wrote: > >> In the way of early feedback -- that's looks super neat! I've got this >> instant feeling that this would be a great clojure.contrib.memoize. > > > +1 > > That would be wonderful.
Well, in the way of early feedback too (alas not much time to argument in length), there are some points which annoy me : * usage of refs : I had a bad feeling, and cgrand confirmed this to me by pointing an even more interesting counter-argument. Me: using refs is not mandatory since you do not need to synchronize this change with anything else. Christophe: And by using refs, you synchronize the change with a potential uber STM transaction, and if this uber STM transaction retries, you will not benefit from the memoization, since the memoization itself will be discarded by the retry. * lookup function responsibilities: I cannot (yet) offer a better way to approach the problem, but I have a bad feeling with the lookup function changing things "behind the scene". My 0.02 euros, -- Laurent -- 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
