forwarded message from John H. Reppy

[Matthew Fluet]

> John Reppy sent a pointer to his new LCPS paper.  I haven't read it yet, but
> it's got a lot more than the previous one.

The new stuff includes a little more about the Moby compiler, another
application of LCPS, and some more implementation details and experiments.

The decontification optimization (the other application of LCPS) sounds a
little like the raise-to-jump optimization in MLton, although it is
expressed in terms of callcc/throw constructs (to which exceptions are
compiled in Moby).