From reading the article it's interesting to note that a 'pure-IL' solution was 'quite a bit slower by my performance requirements'. The workaround seemed to be to create a custom JIT compiler from IL to x86 specific for the requirements of the dynamic language:|
"A week more of work and I had resolved all the design issues for providing both my own jitter for a hi-performance .NET x86 version, and a pure-IL solution that would work [albeit quite a bit slower by my performance requirements] on any .NET implementation [like the PocketPC]."
And it seems the potential is there to be even faster:
Even better, my new .NET dispatchers were acceptably close to the performance of my own AOS.IL jitters in S#.AOS. Translated, this means that it is now faster than any interpreted Smalltalk, and within a couple of cycles [of AOS performance] on a per dispatch decisions point basis. The difference between S#.NET dynamic dispatch/call performance C#.NET static dispatch performance is very reasonable [roughly within 6-15 cycles] per call. It could be directly cycle-for-cycle competitive if my S#.AOS jitter mechanisms were incorporated into Microsoft’s .NET jitters.
For those unfamiliar with SmallScript, AOS is the virtual machine that David designed that SmallScript.AOS (or S#.AOS by its other name) sits upon. It's a virtual machine specifically designed for dynamic languages I believe.