> Either Ada must be perceived as showing short-term benefits
   > (power, features, etc.) or
   > language decision makers must be promoted based on long-term
   > criteria (life cycle cost, reliability, and all that)

Although there are reform movements, the current metrics are not
based on life cycle cost, reliability, etc. So Ada will have to
market primarily on Short Term Benefits (power, features, etc.).

It is easy to compare things based on speed, number of resuses,
time per fix, number of hanging pointers, number of memory leakages,
etc. Since Ada already gives the capability of not using pointers,
and is already fairly safe, the PRIMARY CRITERIA of language selection
are:
  (a) how fast can I get my sequential one-computer program converted
      to run in a client-server gui environment on the Web?

and

  (b) how must faster does it run in Ada versus C++ and Java?

Mike Brenner