On Tue, May 10, 2011 at 8:15 AM, Hannes Wallnoefer <[email protected]> wrote: > 2011/5/10 Charles Oliver Nutter <[email protected]>: >> It would work, but would rarely be applicable for production code. >> Better is to do what JVM does and mark the previous revision of the >> hidden class as "not entrant", and then new entries into that call >> path will use the newer code. >> >> I'm very excited to see Rhino getting updated. It is an embarrassment >> that the JVM can't run JavaScript as fast as V8. I'd love to help, if >> there's something I can do...I have looked at Rhino's codebase in the >> past and know the general lay of the land. > > Thanks! any help from you and other Indy mavens here is greatly > appreciated, of course. I'll happily consult this list whenever I > can't figure out something myself. > > As to matching the performance of V8, don't get your hopes too hight. > V8 uses some optimizations that probably can't be replicated on the > JVM. For example, regular expressions are compiled to machine code as > well. We have an experimental branch for using JRuby's Joni as regexp > engine, and it's still 5 or 6 x slower than V8.
Marcin(lopex) had originally planned on making Joni compile Regexps to Java bytecode. Maybe we can get him to pick that torch up again? -Tom -- blog: http://blog.enebo.com twitter: tom_enebo mail: [email protected] -- You received this message because you are subscribed to the Google Groups "JVM Languages" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/jvm-languages?hl=en.
