Re: Eiffel for DoD development?

[emery@goldfinger.mitre.org (David Emery)]

Besides FRESCO, a group has been formed within the IEEE to standardize
Ada9X bindings to IEEE 1295 (IEEE Motif).  So the X Windows issue is
being solved.

But Dennis raises a more general question about Ada and bindings.  He
asserts that:
>[the costs of developing/maintaing an Ada binding] would
>overwhelm the savings provided by Ada's inherent software architecture
>advantages when compared to ANSI C.

The issue here is one of scale.  An Ada binding can (if done right)
provide substantial advantages to other parts of the system that use
the binding.  So, the issue is whether the fixed cost of maintaing the
binding is overtaken by the proportionate savings resulting from using
Ada and the associated Ada binding.  

We did a prototype client/server system, using a locally-developed Ada
binding to XVT.  The XVT/Ada binding was about 2k Ada statements, and
the application code was about 20k Ada statements.  In our estimation,
we 'broke even' with the Ada binding.  In particular, we isolated a
lot of C dependencies/warts and also XVT dependencies/warts within the
Ada binding, and hid them from the applications program, simplifying
the application.  Our prototype was about 15% of the total application
code.  Thus, we'd project a net win for the Ada binding, since there
would be savings/advantages in the other 85% of the code we didn't
write.

				dave
--
--The preceeding opinions do not necessarily reflect the opinions of
--The MITRE Corporation or its sponsors. 
-- "A good plan violently executed -NOW- is better than a perfect plan
--  next week"                                      George Patton
-- "Any damn fool can write a plan.  It's the execution that gets you
--  all screwed up"                              James Hollingsworth
-------------------------------------------------------------------------