NetNews Usenet Archive 1993 #3/NN_1993_3.iso/spool/comp/lang/ada/4086

Text File | 1993-01-27 | 6.7 KB | 130 lines

Newsgroups: comp.lang.ada
Path: sparky!uunet!charon.amdahl.com!pacbell.com!ames!elroy.jpl.nasa.gov!sdd.hp.com!zaphod.mps.ohio-state.edu!magnus.acs.ohio-state.edu!cis.ohio-state.edu!news.sei.cmu.edu!ajpo.sei.cmu.edu!wellerd
From: [e-mail] (David Weller)
Subject: Re: Leffler's Lies: A biased critique of a general
Message-ID: <[e-mail]>
Sender: [e-mail] (Netnews)
Organization: Ada Joint Program Office
References: <[e-mail]> <[e-mail]>
Date: Tue, 26 Jan 1993 10:20:36 EST
Lines: 118

In article <[e-mail]> [e-mail] (Gregory Aharonian) writes:
>>Saddest was his statement: "You don't see any financial
>>communities in the business world using (Ada).  It's not out in the
>>commercial venture any place that I'm aware of."
>[deletia]
>In the absolute sense, the general is obviously wrong.  In any
>application field you can think of, people are using Ada.  However
>in the relative sense, he is probably right.  Outside of the mandated
>world, Ada use is less than 5% of all total software development.
>I do not have exact figures, and unfortunately the Pentagon got ripped
>off by the Mosemann study consultants, who prepared a totally inaccurate
>survey of Ada and C/C++ use in all sectors of the economy.
>

Couldn't agree more.  Anybody who had a decent amount of knowledge
about the C++ and Ada communities knew this paper was a hoax, with
a capital "H".

I understand about the pitiful percentage you quited above, and it's
probably accurate (then again, C/C++ probably accounts for only about 15%,
maybe less).  For the reasons you cited below (and many more), the
Ada "community" needs to learn what "competition" really means.
That can begin by dropping the Ada mandate or making it an iron-fisted
(read that: HEAVILY enforced) law.  I'm leaning toward the former.

>The fact is that most statements about Ada outside the mandated world
>are impossible to prove true or false, as the data does not exist. And
>most economic statements made about Ada in the mandated world are also
>nearly impossible to prove true or false, because as the GAO pointed out
>many years ago and still unresolved, is that no one is collecting the
>data to make such assessments.
>

Statements like "We saved X millions of dollars in reuse" or "Because
we used C++ instead of Ada, we saved X thousand lines of code" are
indeed vacuous.  Your comments about poor economic models has always
intrigued me, and I look forward to more substantiative reports on
how to solve this problem.  We must also balance such comments with
the fact that every company that claims they're using C++ is actually
applying the extra features of C++ in a reasonable manner.  I don't
think we could make a fair economic comparision of modern languages
like C++ and Ada without finding ourselves hopelessly mired in a
tar baby.  C++ is winning the popularity war.  Plain and simple.
If the Ada community doesn't take agressive steps to fix this, no
economic model in the world will save it.

>>Why do people still accept 1985-based arguments for NOT using Ada?
>
>Wrong question.  It should be "not SUPPORTING Ada?".  Ever wonder why
>there are few entrepreneurs associated with Ada, why other than for a
>few defense contractors and Ada vendors, there are few companies able
>to sustain a marketing campaign for Ada tools and libraries?  Because
>starting an Ada business is a good way to loose money, while starting
>a C/C++ business is a good way to make money.  Who the hell is dumb
>enough to start an Ada business under these conditions?  I invested
>a lot of time and my own money building a large database of information
>on all of the reusable software available from all of the government
>agencies, in particular the DoD.  I went to a few Tri-Ada conferences,
>even exhibited, and all I got was a slap in the face.  No business from
>any contractors, no attention from any DoD offices.  I even published
>a column in a Defense journal monthly reviewing freely available defense
>software (talk about reuse advocacy) and nothing.  I learned my lesson
>and moved to other software arenas.  Multiply my experiences many times
>over, and you'll start to understand why few voluntarily support the
>Ada industry.  Between archiac DoD procurement regulations that discourage
>reuse (long ago identified but still on the books) and contractors too
>scared to take a chance in acquiring reusable software, the results are
>a stillborn Ada support business.
>

Good comments.  I think you said, in a nutshell, that Ada isn't succeeding
in a free-market society, the government only pays lip service to reuse,
and contractors who try to build or reuse reusable components are
rewarded by being punished.

I'll tell ya, the ONLY thing keeping me in the "Ada business" is
that I tried the "C++ business", and I'm 100% convinced that,
when done right, a well-put-together Ada-based organization
can deliver a higher quality product than an equally staffed
C++-organization.  Folks can challenge me on that one, but it
won't change my mind.

>About the General, hey at least he mentioned Ada in an interview.  It still
>is outrageous (but at least he's gone) that Strassman never mentioned Ada
>in any of the interviews in the mainstream computer industry.  When everyone
>else dropped the ball, it was his duty to set the example.  He didn't,
>no one doesn't, IBM still sabotages Ada by not including it in AD/CYCLE and
>AD/Platform, and Ada goes no where.
>

Yes.  Even funnier is watching IBM folks at OOPSLA work to undermine
their Ada efforts (by convincing you that C++ is a better choice),
but watching IBM people at the Tri-Ada love-in acquiesce and say
that C++ is entitled to a market just like Ada.  Sounds to me like
IBM's "official" policy is to agressively market C++ and let their
Ada business grow on it's own.   I'm also glad that Strassman is
out of the picture.  I never met a more two-faced bureaucrat.

>Start focussing you ire at inept DoD software initiatives and policies.
>Leave the poor general alone.  Instead write letters to your congressman
>asking why none of the truly inonvative software supposedly being
>developed by the STARS program is never exhibited at commercial CASE
>conferences and trade shows?

The general needs to know he's wrong and made totally false public
statements.  I intend to write my congressman too.  I don't expect
it to do much good, I'm sure AT&T would form a powerful C++ lobby
to counter my efforts :-)

>
>**************************************************************************
>Greg Aharonian
>Source Translation & Optimiztion
>P.O. Box 404, Belmont, MA 02178

P.S. -- Keep these comments coming Greg, I think you're doing a LOT
of good for us "Ada Bigots", even if we don't always like what you say :-)