[said Wes responding to Sy]
> Are you hoping to generate code that will run in these target environments,
> or do you want a compiler that will run on the target?

I hope the former.
> There is no reason IMHO that a compiler cannot omit code that is not needed
> from the executable.  As you say, the fact that the compiler is able to
> generate the code may make it more expensive.  OTOH, the fact that a subset
> compiler implements a subset may make it less marketable and therefore MORE
> expensive.

Exactly. There is no technical reason why a GNU backend could not be
produced for Sy's targets. Assuming GNAT implements the Annex H
restrictions - that is, rejects programs that violate them - why
bother to try to strip down the front end?

In the long run, if this "restricted subset" is seen to have a market,
someone may wish to build a new compiler for it. In the short run,
since GNAT and GNU sources are widely available and stable, building
appropriate back ends, linkers, and runtimes is by far the more fruitful

Sy, did you say you had friends who could do this? If so, why
not get busy? If not, who should do it?

The Ada community is chock full of people with good ideas, but they
always seem to want someone else to implement these ideas for them.
If Sy were to say "OK, guys, help me define just what sort of Ada we
want, then _I_ and my friends will build will build it" I think Sy
would get a more enthusiastic response!

Instead, Sy, you persist in complaining and trying to work through
institutional processes that don;t seem to want to listen to you. Why
not try another, grass-roots, approach?

> ---------------------------------------------------------------------------
> W. Wesley Groleau (Wes)                                Office: 219-429-4923
> Hughes Defense Communications (MS 10-40)                 Home: 219-471-7206
> Fort Wayne,  IN   46808                  (Unix): [log in to unmask]
> ---------------------------------------------------------------------------
Mike Feldman