Rumors of some commercial applications in Ada 
Author Message
 Rumors of some commercial applications in Ada


Quote:
(Gregory Aharonian) writes:
>     As long as IBM tells the commercial world that its choice for the
>object-oriented, enterprise programming language of the future is going
>to be either C++ or Smalltalk inside of AD/CYCLE,  Ada is going to be as
>relevant in the commerical world as Forth is in the engineering world.
>Sure there will be a few large systems done in Ada, but between IBM's
>public posturing about the advantages of either C++ or Smalltalk, its
>commercial silence about Ada, and the effects of corporate downsizing
>(i.e. adopting networks of workstations/parallel processes, all
>programmed in C/C++), Ada is not going to be widely accepted, not matter
>how much wishful thinking.

   [comments about Mosemann studies deleted]

Well, I've been waiting for you experts out there to respond. It's been
weeks and I've heard only silence. So I'll put my foot in my mouth:

I think a lot of the article makes sense. One quibble -- considering IBM's
troubles lately, I wonder how big their influence really is? C/C++ didn't
become popular because of IBM's blessings in the first place. IBM is only
reacting to what's happening.

But my major disagreement is this -- I think things are far from settled. A
wildcard in all this is Ada9X. An even bigger wildcard is the fortran
community. There is a HUGE investment in Fortran code out there, and there
are plenty of Fortran users who wish C/C++ would go away. Although Fortran
programmers have no great love for Ada either, a determined effort to woo
them might be worthwhile. I myself was a Fortran programmer who was won
over.

There are four things we should hammer on:

  (1) THE STEERING WHEEL'S IN THE RIGHT PLACE. Japanese resist buying
      American cars because the steering wheel is on the wrong side. C/C++
      has the same problem for scientific programmers. It does many things
      well but it wasn't designed for scientific programming -- consider
      the absence of an exponentiation operator. Fortran and Ada both have
      the steering wheel in the right place.

  (2) BIGNESS IS GOODNESS. The language of English is big. The language of
      mathematics is big. Why should the language of your computer be
      small? A large language makes it easier to express yourself. This is
      a point to which scientific programmers should be particularly
      receptive. Ask them if they would be happy using mathematics without
      vector and tensor notation? How easy is it to write a scientific
      paper using first-grade English? Are you SURE small is beautiful?

  (3) LOOK'N'FEEL. The look'n'feel of Ada is a lot closer to Fortran than
      C/C++ is to Fortran. Ada has no compound statements, for example. Nor
      does it have expression-level assignments. IMHO transitioning from
      Fortran to Ada is easier than Fortran to C/C++. It was for me,
      anyway.

  (4) RICHNESS IS BETTER THAN ORDERLINESS. C++ wins this one hands down. We
      can fight back with 9X. At the risk of being burned at the stake, I
      propose we also fight back with a variety of cheap, nonvalidated
      subset/superset (there! I said it!) compilers -- especially OO
      compilers. Validated compilers would still be used for serious stuff
      like flying airplanes. Nonvalidation hasn't hurt C/C++. The recent
      release of Ada/Ed was a step in the right direction. Reduced
      restrictions on the Ada name was another plus. Intel has sold over
      ONE HUNDRED MILLION 8051's. How about an 8051 punk ada cross
      compiler? Jeez, I'd buy one for the logo ALONE.

========================================================================
Frank Manning                                           I'm ZERO percent

========================================================================



Tue, 30 Aug 1994 19:58:01 GMT  
 Rumors of some commercial applications in Ada
        Do not pin your hopes on Ada 9X besting C/C++, as strong as your
arguments might be about its potential advantages.  The question of which
language will dominate is a question of market forces, which the DoD has
never understood and never cared to understand, much like their lack of
interest in the existence of a commercial software reuse industry.  The
DoD wanted and got one language to standardize everything in; all arguments
about technology transfer to the private sector is posturing, case in point
being everything associated with STARS, where a lot of money is going to
create tools that almost duplicate what's available on the market.  None
of that stuff is going anywhere, except to DoD contractors.  If that makes
the brass happy, fine, but anyone counting on Ada catching on in the hopes
of starting an Ada business should be prepared to go broke, as dozens of
companies have experienced with commerical Ada (beyond compilers).

        A good measure of the real short-sideness of Ada policies is the
complete lack of attention towards integrating Ada and VHDL (i.e. integrating
software and hardware design).  Given the similarity in the language
grammars, extending tools to do this is not hard.  Yet a big potential
benefit is ignored.

        And writing any of this stuff in ADA 9X won't change anything.

Greg Aharonian



Sat, 03 Sep 1994 11:58:12 GMT  
 
 [ 2 post ] 

 Relevant Pages 

1. Rumors of some commercial applications in Ada

2. Rumors of price change on Alsys First Ada

3. Lisp Fact/Rumors (Was Lisp,Ada,Flames)

4. Ada in Medical Devices: was RE: Inexpensive Developer Level Commercial ADA compile

5. Commercial CFW applications

6. using a functional language for a commercial application

7. language for commercial applications on Windows?

8. Large commercial application

9. Commercial application developed using Oberon??

10. Commercial Applications written in Eiffel ?

11. can anyone tell me about any commercial applications written in eiffel ?(for polytech report)

12. Commercial applications using an embedded lisp ?

 

 
Powered by phpBB® Forum Software