On Fri, Dec 18, 2009 at 6:23 PM, John Cowan <[email protected]> wrote: > On Fri, Dec 18, 2009 at 7:19 PM, Charles Oliver Nutter > <[email protected]> wrote: >> I figured I'd do a feasibility study on whether JDI could be used to >> implement ObjectSpace (Ruby's feature for walking all objects on the >> heap of a given class). > > Wouldn't maintaining a weak set initialized by the constructor be preferable?
That's what we do now, but I'd like to just rip that code completely out so we don't have to check a flag "should I add myself or not". And the weakref method means you have n * <ruby object count> extra objects in memory, along with the tremendous perf hit of constructing a weakref for every ruby object construction. In the end, this is too slow, so it's only an interesting JDI hack. But it's cute :) I may look into a JVMTI version instead, since it would be a lot closer to the metal and we may be able to bind the JNI methods dynamically using JRuby's FFI layer. - Charlie -- 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.
