I'm surprised Greg didn't mention the AFC (Ada Foundation
Classes) which is their binding to the MFC.  I would think
with that ability it would be relatively simply to interface
between the two.  Assuming the bindings contain the class
in question.  Even if it were a user defined class, you
could at least follow their examples to accelerate the
interface process.

Of course, you have to buy the right level of their tool
to get it.

Frank

-----Original Message-----
From: Ada Marketing [mailto:[log in to unmask]]
Sent: Monday, September 11, 2000 11:19 AM
To: [log in to unmask]
Subject: Re: Integrating Ada95 and C++?


Crispen, Bob wrote:
>
> The fire has burned under the open-source HLA RTI
> pot for long enough that it's starting to bubble.
>
> One of the issues that's come up is that there are two
> codebases which might be used: one in Ada95 and
> one in C++.
>
> Has anyone here ever integrated C++ and Ada?  I'm
> talking about C++ on the class level, not some little
> library functions with a C-like interface, but medium-
> sized reusable elements in C++ as well as Ada95.

We have some customers doing this.  I believe the main trick is to declare
the
C++ methods as static at the interface point.  This principally turns off
the
C++ name mangling, so the Ada side knows what the function name is.

>
> It seems that it might be theoretically possible given
> the codebase behind gnat and g++, though I gather
> egcs has complicated things?  The problem is, I've
> fallen too far behind in understanding some of the
> underlying stuff to even make an educated guess.
>
> I don't want to recommend something to an effort I
> care very much about if there isn't a solid background
> of success doing this.  There are simply not enough
> hands on the open RTI project to do research into
> making Ada95 play with C++.
>
> Bob Crispen
> [log in to unmask]

--

Greg Gicca
Aonix Ada Marketing account
[log in to unmask]