Re: Anybody in US using ADA ?

["John R. Strohm" <strohm@airmail.net>]

Uh, excuse me?

Cost per SLOC, in constant dollars, has declined SIGNIFICANTLY over the last
20 years.  This was one of the big findings in the SEI CMM stuff, that it
really is significantly cheaper to do things right the first time.

At one point, I was informally putting together a business case for Ada and
Rational R1000s at GD/FW, based on (a) Rational's productivity numbers and
"force multiplier" effects, (b) the known size of F-16C/D operational flight
software, (c) the expected size of the next airplane's software, (d) the
size of the parking lots at GD/FW (which were ALREADY overflowing).

"Mike Silva" <snarflemike@yahoo.com> wrote in message
news:20619edc.0301022244.2c16c563@posting.google.com...
> David Emery <demery@cox.net> wrote in message
news:<J_4R9.114758$pe.4406969@news2.east.cox.net>...
> > Mike Silva wrote:
> > > "Marin David Condic" <mcondic.auntie.spam@acm.org> wrote in message
news:<autmq3$4oe$1@slb6.atl.mindspring.net>...
> > >
> > >>Its a shame to see that many in the DoD arena are abandoning Ada
because
> > >>"Everybody else is using C++".
> > >
> > >
> > > Seems like we should be seeing some results of the "DoD arena" rush to
> > > C++ by now.  As the data comes in, we should either see that indeed
> > > C++ introduced problems and added costs that Ada would have avoided
> > > (my expectation), or that it didn't (in which case, time to re-examine
> > > assumptions).  So, where are the C++ disaster coverups, and/or the
> > > shining successes?
> >
> > In 1990 MITRE did an internal study based on then available data
> > saying that the cost to maintain Ada was -LINEAR- on SLOC, not
> > exponental as Barry Boehm says.  If you think about this, it's a
> > radical conclusion.  And it wasn't done by the Ada people at MITRE, but
> > by our cost center's statisticians and cost analysts
> >
> > Back when Emmett Paige was ASD-C3I, he held a series of Ada Dual-Use
> > summits.  At each, my position was:  DoD has the data available to show
> > if Ada has the value we claim.  Let's collect the data and let the facts
> > speak for themselves.
> >
> > Two results from this experience:
> > 1.  DoD didn't want to gather and then hear facts
> > 2.  Language decisions are generally not made on the
> >     basis of -any facts- but rather management perception of
> >     "acceptability", "training costs", etc.
> >
> > dave
> > (been there, done that, forgot my T-shirt)
>
> The only reason things *may* be different this time is that the
> amount, and cost of writing, software is so much greater today than
> 10-20 years ago that the risk of failure (or gross overrun) may be too
> great to ignore or sweep under the carpet anymore, or, inverting, the
> potential savings from using the right tools may be too great to
> ignore....
>
> Mike