jd smith

I guess I've never really measured it. It seems the actual call to
CATCH + a test on the catch error variable take about a microsecond for
me, so really probably not a problem (i.e. it will always be a tiny part
of your motion event callback budget).

What I do always wonder is how setting up a set of nested catches bogs
down IDL, since it must always keep track of where it will jump to
when an error occurs. Probably just an over-defined sense of
optimization.

JD