> Interestingly, [Erjang][1] (a port of Erlang to the JVM) apparently > performs TCO while claiming to stay "reasonably fast". The gimmick
I have never done extensive benchmarking of clojure, but given the frequent mentions of use of '-server' in order to achieve specific performance goals, I get the impression clojure (i.e. Rich) definitely wants to take advantage of all the optimizations the JIT can offer now and in the future. Trampoline-based TCO would, as far as I can tell, always defeat the JIT's notion of a call site - unless the JIT is made to understand the trampoline (but then are we getting close to full TCO support anyway? I dunno, I'm definitely not an expert on this topic...). So, I would expect that those cases which are aggressively optimized by JIT:ing, such as eliminating method lookups by inline caching in light loops, would suffer potentially very extreme performance impacts which aren't fixed by just avoiding allocation as is mentioned in the erjang wiki page. Or is this over-stating the problem? -- / Peter Schuller -- 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
