After watching most of Rich's Clojure presentations over the weekend, I found myself playing with ants.clj again on my netbook. The ant simulation runs brilliantly on my quad-core machine at work. Not so much on my netbook. The problem seems to be that with only a single (hyperthreaded) core the render agent is almost constantly interrupted by some pesky ant while attempting to snapshot the world, forcing the render agent to automatically retry. And so, the ants run merrily around the world, only I can't see it.
This raises a question. Clojure's refs and dosync sure are neat, but this experience would seem to indicate that there are potential scalability problems when combining long-running and short-running transactions. Under load (or on a slow machine) a long-running transaction may never get a chance to complete and may be stuck forever retrying, burning CPU but producing no useful output. This makes me uneasy. I was able to get ants.clj to work reliably on my netbook by playing around with the sleep times in such a way as to increase the probability of the renderer actually completing a snapshot, but this process felt hacky and ad-hoc. What works on my netbook might well be sub-optimal on another system. How could one change the design of ants.clj to work reliably (i.e. update the screen periodically) even on slower systems? // Ben --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
