TEAM-ADA Archives

Team Ada: Ada Programming Language Advocacy


Options: Use Classic View

Use Monospaced Font
Show Text Part by Default
Condense Mail Headers

Topic: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Content-Type: text/plain; charset="iso-8859-1"
Sender: "Team Ada: Ada Advocacy Issues (83 & 95)" <[log in to unmask]>
From: "Crispen, Bob" <[log in to unmask]>
Date: Wed, 22 Apr 1998 11:01:08 -0500
MIME-Version: 1.0
X-To: "Robert C. Leif, Ph.D." <[log in to unmask]>
Reply-To: "Crispen, Bob" <[log in to unmask]>
Parts/Attachments: text/plain (123 lines)
Bob Leif sez:

> You are absolutely correct! The purpose of MARKETING is to inform
> managers
> that the use of Ada and software engineering will result in "reliable,
> safe
> software that comes
> in on time and on budget" and that the software will also have the
> great
> added benefit of being maintainable.  In short, what a manager wants
> is NOT
> a software tool but risk reduction.
> For instance, at the Sun Meetings I have attended Java code was never
> shown! Most people buy products to solve problems. For instance, I
> have
> been telling the vendors that they must support the Windows CE
> platform. I
> suspect an Ada compiler for Windows CE might even help Boeing with its
> in
> flight entertainment software.
Let me mention another couple of corollaries,
inspired by your mention of Java.

A little while ago I said that every hour and dollar
spent on promoting the technology instead of the
solution was a precious resource wasted.

I was wrong.  Every hour and dollar spent on
promoting the technology is another knife in
Ada's back.

Here's why.  If you talk about technology, you force
the program manager to *look* at the technology.
And when he does, what does he find?

(a) Ada is hard.  It takes smart people to
program in Ada.  Now I can offer myself as a
counterexample, but in fact, even though a
dullard like me can program in Ada, I find it harder
than to program in other languages because
Ada makes you think about design and
architecture.  It also hasn't helped that we've
put some mighty bright people up there on
the platform selling Ada.

(b) Ada is expensive.  The days of the $20,000
compilers will follow us to our graves, despite
Ada compilers now being free, or at any rate
(when you add value to the compiler) cheaper
than equivalent C++ value-added compilers.
And while I routinely reuse Ada code, this is
invisible to program managers, who know that
even if compilers are cheap, smart people
(see (a)) are expensive.

(c) Ada is heavyweight.  Hell, we *sell* the
heavyweight features of Ada -- strong typing,
encapsulation, tasking, IPC.  If somebody gets
the idea that Ada is heavyweight, should we
blame them?

Yes, I know there are comebacks to every
item on this list.  Bad software is a hell of a
lot more expensive than good software, etc.,
etc., etc.

But by rubbing people's noses in the
technology, we show them Ada's worst side.

Sell them the results -- on time, on budget,
safe, reliable, maintainable, reusable -- on the
other hand, and you sell them not only Ada's
best side, but the very reasons why we love
Ada ourselves.  It ought to be natural.

In our case, the steak *is* the sizzle.  The
technology is nothing more than the butchering,
and you'll notice that very few restaurants kill
the cows on the premises.

Now let me name one example from the VRML
world that might start some people thinking.
Yes, I know the domains don't map one-for-one,
but bear with me.

A company called Intervista, founded by a
delightful man named Tony Parisi, a true
visionary in Virtual Reality Modeling Language
and one of its pioneers, has a couple of
products out there.  One of them allows you
to embed an animated 3D scene in a PowerPoint
slide.  Another lets you take an Excel
spreadsheet and turn it into a dynamic, interactive
3D model.

Both products rely on VRML, but the four letters
are *way* down in those products' sales pitches,
pretty much where you'll find "MOSFET" in the
brochure for an audio amplifier.

Tony has realized that real-world managers
and even time-pressed, harried engineers
don't want technology (VRML).  They want
solutions to their problems (complexity of data
and concepts requires 3D representation and
interaction) and they don't want to have to learn
something entirely new to take advantage of the
solution (Excel and PowerPoint).

Oracle may shortly putting 3D data interaction
into some of its database products, and you can
bet that if they mention VRML at all, it will
be because it's an international standard.

I don't know how to stretch these analogies to
Ada yet, but I think somebody had better figure
it out, and soon.

Bob Crispen
[log in to unmask]