Re: Question about interfacing C and Ada95

[Robert Dewar <robert_dewar@my-deja.com>]

In article <80hge1$upk$1@nnrp1.deja.com>,
  Ted Dennison <dennison@telepath.com> wrote:

> > Read the exact rules in the RM, and make sure that your
> > compiler documents that it follows them (this documentation
> > must be present in a conforming compiler, it is required,
> > see annex M)

> True. And (just to tweak Robert a bit :-) ) I'll also point
> out that it is meerly "implementation advice" that it work
> this way. But this is something you should be able to count
> on, and yell at your vendor if it
> doesn't work.

This really rates a huh? Ted, the *reason* that I said above
that you should check that your compiler follows these is
*precisely* that they are IA. Of course you do not have to
check documentation to see whether your compiler follows
required normative rules!

THe point is that in evaluating compilers, you must be able to
easily check if some given IA is followed, since compilers are
required by Annex M to provide this information in their
documentation (the relevant documentation in GNAT's case can
be found in the GNAT programmers reference manual for example,
but there must be equivalent documentation in any compiler
claiming conformance to the Ada 95 RM.

As Bob has pointed out, Ada 95 tends to go much further in
specifying such things than other languages. They are IA simply
because it is (quite obviously) impossible to formalize these
kind of interface rules.


Sent via Deja.com http://www.deja.com/
Before you buy.