Re: Air Force helping to undermine Ada

[kjmiller@mitre.org (Kevin Miller)]

In article <SRCTRAN.93Mar8175011@world.std.com>, srctran@world.std.com
(Gregory Aharonian) wrote:
> 
>      A recent article in Network World show one way in which the DoD is
> helping to undermine Ada acceptance.  I will quote from the article (without
> losing its essence) [Network World 3/8/93, 33]:
> 
>     XSoft, a division of Xerox, last week announced document-based work flow
> software called InConcert, which provides a graphical user interface that
> makes it possible for end users to construct a variety of complex work
> flow applications.  InConcert is client/server software that tracks and
> coordinates all tasks in a work flow process and automatically delivers to
> end users the business documents and applications they need to complete a
> work flow task.  Its server runs on Suns and RS6000s, as well as its clients
> along with MSWindows.
>   The problem with many high-end work flow products, according to consultants,
> is that they require programmers to build work flow applications using a high
> level scripting language.  This slows down the development process and takes
> end users out of the loop.  "InConcert's strength is its graphical work flow
> design tool, which makes it possible for end users rather than programmers
> to build work flow applications", says a consultant.
>     XSoft, which developed InConcert using the C++ programming language, has
> published more than 270 application program interfaces that enable end-users
> to integrate thiry part products with InConcert.  Developers can also use
> these APIs to embed software agents within InConcert that carry out processes
> in response to predefined events.
> 
>     Another beta user, the U.S. Air Force, is implementing a work flow
> system based on InConcert that specifies 1,400 tasks and 60 applications and
> will be used by 10,000 people.
> 
> ==============================================================================
> 
>    This is a good example of the beginning of the end of Ada inside the DoD
> for two reasons.  First, the Generals in charge are more and more going to
> see that the information processing needs can be meet by taking commercial
> products that are "open" and adapting them to defense needs, commercial
> products all being written in C/C++.  As 10,000 people and 60 applications
> is a serious integration effort, comparable in scope to most other DoD 
> software efforts, one success makes it easier to do this again and again,
> to the point where the Ada mandate becomes irrelevant.  It is impossible
> for the DoD RIGHT NOW to achieve a similar work flow system relying one
> any commercial or non-commercial system written in Ada.  As more and more
> systems inside the DoD are based on C/C++ systems, the Mandate also becomes
> uneconomical to retain.
>    The second reason that the Generals will desert Ada is that the success
> of these such efforts will get them to start asking what is going on with
> efforts like STARS to create Ada technology that increasingly is falling
> farther and farther behind the commercial world.  This questions the
> competence of either Ada or Ada contractors, neither of which does much
> good for retaining the Mandate.
> 
>     Thus my prediction: if you see more and more such stories about the
> DoD solving its problems using C/C++ commercial software products, then
> you know the handwriting will be on the wall for Ada.  Any given the
> technology I see at the commercial trade shows, I would bet MY money on
> seeing more and more such stories.
> 
> Greg Aharonian
> Source Translation & Optimization
>
 
Are you saying that the government should go out and develop an application
when there is a commercially available AND SUPPORTED product that meets the
need?  Doesn't sound very cost-effective to me.  In a situation like this
it would cost more for the government or Air Force to develop and maintain
a seperate package, irregardless of the language.  I would welcome the day
when the government can satisfy all it's requirements with commerically
developed and maintained packages.  Until that time comes (if it ever does)
the idea of having a single language for as much of the
government-developed software as possible makes sense.  I still maintain
that Ada is a reasonable choice for such a language.  Chasing after the
language de jour only componds the kind of problems that the use of Ada was
intended to address.

> 
> (In fact, maybe the DoD should change the name of the language.  When the
> disease AIDS first became well known, a diet product pronounced the same
> way changed its name because of confusion.  Given that most corporate
> executives think of the sound "ada" as standing for the Americans with
> Disabilites Act, are we not risking subconcious confusion for Ada as a
> disabled language :-)
> -- 
> **************************************************************************
> Greg Aharonian
> Source Translation & Optimiztion
> P.O. Box 404, Belmont, MA 02178

-----------------------------------------------------
Kevin Miller       | MITRE's lawyers can't moan,    |
MITRE Corporation  | 'Cause what's stated up there, |
Bedford, MA        | Is my opinion alone,           |
(617) 271-4520     | And not MITRE's to bear.       |
kjmiller@mitre.org |                                |
-----------------------------------------------------