forcing tail/sibling call optimization

[Robert Dewar]

<<I happen to think that it would be good for ANSI/ISO language standards
to include requirements for warnings.  However, that is a separate battle,
and for this proposal I'm happy to amend the formal spec so that the
warnings are merely implementation advice rather than requirements.
>>

It is an unwinnable battle I would say, since a formal specification
of what constitutes a warning msg is impossible. Note that not even
the Ada standard *requires* error messages as such (it is just to
hard to define what an error message is). The only formal requirement
in Ada is that certain errors be "be detected prior to run time".

It is a common misconception that making things implementation advice
rather than implementation requirements weakens the "requirement", but
in fact IA is often much stronger, since you can say things in IA
that you could never get away with in formal definitions.