Re: Ada Core Technologies and Ada95 Standards

[dewar@cs.nyu.edu (Robert Dewar)]

John McCabe says

"You seem to be defending a "make it fast, THEN make it work"
philosophy here (which I completely disagree with) and confusing
quality with run-time performance.

I have seen the effect of this kind of philosophy - my current
compiler supports the ATAC co-processor, yet has had trouble compiling
some basic Ada constructs.

I would choose 2. over 1. as I believe a _quality_ compiler is one
which compiles the language completely, not one which does some parts
of the language quickly."

John, I can for SURE conclude from your writing here that you have never
looked at the ACVC tests closely, and also that you have limited experience
in the definition of language definitions.

The fact of the matter is that for *any* language of any complexity, you can
find extremely marginal cases which are not worth testing. Study for
example the Rosen tasking anomoly as an example (I don't want to waste
space describing this case, since it is well known to anyone who has
followed the ACVC process).

It is a common naive viewpoint that (a) the RM precisely defines the Ada
language -- of course it does not, since it is not a formal document,
and that (b) it is therefore absolutely clear what conformance means,
and that (c) it is therefore valuable to test any aspect of this conformance.

I guess that neither you nor Ken have paid any attention to what is going
on with ACVC 2.1, but the whole idea of this effort is to make sure that
the ACVC suite better reflects actual user usage. I would be interested
in comments from either of you on this reformalation, and in your reaction
to the new tests, but to make useful comments you will have to sepdn more
time actually studying the tests and the ACVC process.