> Has any one had any luck (or lack therof) getting the linux port of
> Mathematica 3.0 running under NetBSD 1.2?  (or 1.2.1 for that
> matter)

For whatever reason, under NetBSD 1.2, the Linux MathKernel binary is
treated as an SVR4 binary and dies very quickly with a SIGSYS.  I
don't know if the situation is improved under -current.  I can try to
collect some more data, since we don't yet know if the binary works
after you get over this obstacle.

(I know that it's difficult to distinguish SVR4 and Linux ELF
binaries, but most Linux ELF binaries don't have this problem, and I
never looked into the presumably gross hacks NetBSD uses to
distinguish them.)