Re: Ada used in General Aviation (GA) applications?

[Peter Amey <peter.amey@praxis-cs.co.uk>]

Jeffrey Carter wrote:
> Marin David Condic wrote:
> 
>> Keep in mind that I have done a study that showed a 50% reduction in 
>> development time and a factor of 4 improvement in defects. You may in 
>> fact be thinking of that very study here. I've made the statement here 
>> before.
> 
> 
> One was a study in the domain of jet-engine control, which I think was 
> presented by some guy named Condic. Is there a link to this study? That 
> would be a nice thing to have. The other is the Rational/Verdix study in 
> the domain of Ada compilers, which is available on AdaIC.org. There are 
> other studies also available on the AdaIC site, such as John 
> McCormick's. They don't present their results in a comparable way, but 
> the general message is the same for all of them.
> 

On the subject of defect rates, I think data from the Lockheed C130K 
(Hercules II) project is relevant.  Independent V&V was carried out by 
the UK MoD (the lead customer).  The V&V covered a variety of systems 
produced by a variety of sub-contractors in a variety of languages.  All 
the software inspected had already been cleared to DO178B level A or B.

The results showed that:

1.  Significant errors remained despite the prior FAA clearance.
2.  Code written in C had, on average, 10 times as many errors as that 
written in Ada (and 100 times more than that written in SPARK).
3.  No statistically significant difference in error rate between level 
A versus level B systems could be found.

See: "Software Static Code Analysis Lessons Learned" by Andy German, 
QinetiQ Boscombe Down. DoD CrossTalk Journal, November 2003. 
http://www.stsc.hill.af.mil/crosstalk/2003/11/index.html

and

"Correctness by Construction: Better Can Also Be Cheaper"  (PDF 312kb) 
Peter Amey, Praxis Critical Systems Limited. CrossTalk Magazine, March 
2002.  http://www.praxis-cs.co.uk/pdfs/c_by_c_better_cheaper.pdf

Peter