I came across notes that indicate Neko has tail recursion optimizations, but does it also support general tail call optimizations? By this, I mean passing and calling closures in tail position frees stack space before allocating a new frame.
Sandro -- Neko : One VM to run them all (http://nekovm.org)
