Re: ada and robots

[gwinn@res.ray.com (Joe Gwinn)]

In article <33B51B61.2E26@gsfc.nasa.gov>, Mike Stark
<michael.e.stark@gsfc.nasa.gov> wrote:

> Joe Gwinn wrote:
> <snip> 
> 
> > The attractions of COTS software will nonetheless draw many DoD entities
> > to C/C++.
> 
> Which makes sense for the MIS applications -- I would be wary of
> embedding COTS inside a weapons system though.

Well, don't look now, because that's exactly what's happened.  And, by the
way, C is far ahead of Ada in the Air Traffic field, with or without the
FAA.


> > In the aftermath of the AAS debacle, the FAA has turned against Ada, and
> > now gives us the choice of C or C++.
> > 
> God save us all!  The AAS debacle was due to a lot of factors, IMO
> primarily
> requirements instability and failure to do any useful risk management. 
> The 
> FAA would do far better to procure and adapt ATC systems already in use
> in
> Europe or Canada (and readers of Dr. Feldman's periodic postings on
> non-DoD
> usage know what language _those_ are implemented in ;) than to start
> over from
> scratch in C or C++.  As a taxpayer, I am apalled by the waste of money
> in
> starting over.  As an airline passenger, I don't want the controller in
> the
> tower at Dulles to have his system crash while _my_ flight is in the
> landing
> pattern.  Just about every common OO language, and especially Ada and
> Eiffel,
> are more appropriate for safety critical systems!

I generally agree.  I didn't claim that Ada was the reason that AAS
crashed.  I doubt that any language, however perfect, could have saved
AAS.  It's a long and ugly story.   But, fair or not, the FAA has in fact
turned away from Ada.


Joe Gwinn