11 Mar 2002

  version 5.3

21 Mar 2002

  Fixed a bug in follow set computation.  If FIRST(c) is {epsilon,C}
  the follow set computed from the rule

    a ->.b c [C.d]

  was [C.b C.b].  The correct follow set is [C.b].  The follow set
  computation never ended because the follow sets continued to grow.

  No change to basilsim.

  version 5.3.1