Air force and Ada

Air force and Ada

[Kenneth Mays]

Greetings,

Kill the thought of Ada95 going away in the Air Force.
Give me an official paper, like the one in 1990 on how 
the Air Force is pushing Ada, and I might take a peek.

The problem is the "general purpose" of Ada95. If you look at
GNAT V3.03 under ACVC 2.0.1 (the latest validation module),
it is the only (I think) Ada95 compiler that supports the
majority of the annexes. The other Ada95 compilers focus on
the first two. So, this means that GNAT is the standard (oh,oh)
that others have to follow or beat - also GNAT is FREEWARE. It
is also available for all (almost) popular systems (not my C64).
This is a major benefit, since it is cheap, efficient, cost-effective,
easy to maintain, easy to manage, and cross-platform capable.

Basically, you can write Ada95 code on your SGI or AS/400 and I can
take your code and compile it on my IBM PC or VAX with the SAME
compiler!! This is not posible with Rational, Thomson, or many others.
This should leave a thought in many people's head. Your embedded system
controller doesn't need (usually) high powered graphics or a super
duper sound library. It just needs a program that controls it and tells it
what to do (and when not to). Send this whatever, track this whomever,
or blow up this something. There are more complex things out there, but
I'm talking about the basic things Ada95 will become involved in.

Now, you can look at the ACVC V2.0.1 and see it covers distributed 
systems. The ACVC also covers information systems as well as other 
areas of interest. So, if maybe there can be a switch for different 
things. Like a Ada83 switch for compiling Ada83 code, and val95 
switch for compiling Ada95 code and checking that it doesn't break 
the rules of validation or hardware
dependency.  The market is big enough to let some proprietary commercial
and noncommercial compilers fill in for hardware specific compilers 
(Visual C++ and Visual Basic works off this - you don't see MFC/OWL 
on a SGI
workstation). But, can't we stick with ONE compiler (GNAT) that is 
the motherly true VALIDATED ACVC V2.0.1 compiler of all Ada95 
compilers? Maybe.

Like I have mentioned before, if the Air Force is sticking with Ada - 
give Ada to all their engineers and however else works for them. Give 
them the
tools they need to work with and train them in the way they should go. I
preach a lot on GNAT since I like the darn thing (its good), but make sure
you're compiler is available on a MSDOS machine - or it will get left 
on someone's desk for "advanced training only" in the corporate 
offices and not on the desktops of every hacker's (hobbyist) home. 
(think Windows95/NT, Linux, or just plain coffee)
-Ken

"Leave it at work and its a 9 to 5 thing, leave it at home and its
a part of your life."

Re: Air force and Ada

[Robert Dewar]

Kenneth said

"I preach a lot on GNAT since I like the darn thing (its good), but make sure
you're compiler is available on a MSDOS machine"

Well some good news here, Doug Rupp of ACT just completed the port of
tasking to DOS (he was the original author of the GNAT port), and checked
the DOS tasking sources into our CM space. We expect a release of GNAT
3.04 in the reasonably near future which will have tasking support.