[subwindow]

Not really excited. Sorry. To me this is an incremental improvement to a method of GC that is just not going to cut it.

Rubinius uses Generational GC. There are definitely some parts of Rubinius that are (currently) slower than MRI, but memory management stands to be one of places where Rubinius will demolish MRI in the long-run.

JRuby/JVM uses (from what I understand) a Generational-first hybrid GC. There's been so much work and analysis done on the JVM that there's no way in hell any mark and sweep GC is going to compete.

Honestly if MRI is going to remain competitive it needs to completely rethink the way it does GC.

[bluemoon]

Jruby seems like it has the most potential in the GC department. It seems like most of the interesting GC papers I've seen are work that's been done on the JVM.

[jacaetevha]

And (IIRC) much of the JVM's GC and VM work came directly from the Smalltalk world and the Strongtalk guys.

[Uchikoma]

No.JVM GCs (JRockit, Hotspot, IBM) have developed much further than what Smalltalk did. The G1 GC is especially nice.

I found

http://blog.dynatrace.com/2011/05/11/how-garbage-collection-...

a very interesting - and understandable - article.

[agumonkey]

I heard of SELF too.