> > > 1. Due to the path-copy semantics, the contention gets driven to the > root of the tree. > > > Out of curiosity, would reference counting (rather than or in addition to > "normal" GC) help with this? Or is reference counting problematic in a > highly concurrent environment? It seems like reference cycles will be less > of an issue with immutable data. >
Reference counting tends to be less efficient than a tracing collector in a concurrent environment. Reference counts must employ atomic updates that can be relatively expensive, e.g. similar to how AtomicLong.incrementAndGet() works. I've blogged on how atomic increments work. http://mechanical-sympathy.blogspot.co.uk/2011/09/adventures-with-atomiclong.html -- 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 --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
