Re: Future of Ada?

["news.oxy.com" <Vladimir_Olensky@oxy.com>]

Jim wrote in message <369c6b78.7488219@news.nodak.edu>...
>I am just wondering what those who frequent this news group think
>about the future of Ada.  I've heard a lot of talk that Ada is kind of
>dying out and that not even the military is using it any more.  Any
>comments would be greatly appreciated.
>
> -Jim

Hi from Russia.


This reply is also  well suited for other dicussion thread  (Ada vs C++ vs
Java) .

As a matter of fact ADA recently has been adopted as  NATO programming
language standard (look at the end of the message ).
Believe me such kind of decisions are not taken easily and  just because
some top level people like ADA more than other languages but because it is
better suited for very complex real world and real time and embedded systems
(as ADA was designed for that and proved it).

I think that problem is that a lot of people do not try to look at the roots
(or essence) of things and just follow the easy way of living (just doing
what they have been taught and not trying to grasp new ideas, concepts
etc.). For many people (including programmers)  it is difficult to strain
themselves. It is easy to relax and do what is easier to do and just to
repeat what  other people say.

Now when good ADA95 compilers become available along with GUI tools (e.g.
CLAW from RR Software) and useful ADA libraries as well as ADA binding
(thick and thin)  to other language libraries ADA95 is at the starting point
for rapidly gaining popularity.
I think that  GNAT 3.11p  from ACT is such a starting point (it should be
soon publicly available).

My opinion that right now the best choices is ADA95 and combination of ADA95
and Java when appropriate (ADA95 being used for core systems and Java code
used for getting access to such core systems via Internet from any PC with
browser).  It proved to be that ADA95 and Java perfectly suits each other in
this area as now it is possible to generate Java class code directly from
ADA95 program and there exists bindings to Java for ADA95 (Averstar). This
option probably will be included in some future releases of ADA95 compiler
from ACT (according to ACT).

Pretty  soon there will be more ADA bindings and ADA libraries in addition
to existing ones.

My opinion that right now the best choices is ADA95 and combination of ADA95
and Java when appropriate (ADA95 being used for core system and Java code
used for getting access to such core system via Internet from any PC with
browser).  It proved to be that ADA95 and Java perfectly suits each other in
this area as now it is possible to generate Java class code directly from
ADA95 program and there exists bindings to Java for ADA95 (from Averstar).
This option should be included in some future releases of ADA95 compiler
from ACT (according to ACT).

Probably soon there will be more ADA bindings and ADA libraries in addition
to existing ones.

I wonder if anyone (professional  ADA programmers  and ADA software
Companies)   try to write ADA95  bindings (WindowsNT)   to ACE real time
communication subsystem  as well as TAO implemented in C and C++ by Douglas
C. Schmidt at Washington University in St. Louis
http://siesta.cs.wustl.edu/~schmidt/ ) or even implement it totally in
ADA95. This will be invaluable investment to ADA community. I know that this
is underway for Linux but I do not know anyone who is doing that for Windows
NT.

I consider this as a challenge to:
    Robert Dewar Ada Core Technologies,
    Tucker Taft   stt@averstar.com   http://www.averstar.com/~stt/
    Markus G. Kuhn, Computer Laboratory, University of Cambridge, UK,
    Marin David Condic, Real Time & Embedded Systems, Propulsion Systems
Analysis
    and other professionals in ADA.
------------------------------------

I am communication engineer that have been working in space industry for 13
years being involved in  Mir Space Orbit Station project ,Souz projects etc.
taking part in development of onboard  communication and control systems as
well as test equipment for that systems.
Since beginning of 1993  I've been  working as communication specialist in
American Company (Occidental Oil an Gas  www.oxy.com) in Russia (satellite
communications, data communications [frame relay, TCP/IP networks ], Nortel
Meridian1 switches, etc.).

I think that this prove that I am not a person which is interested in
promoting particular programming language. I am just a person who is
interested in having such kind of programming language and associated set of
tools that best suited for developing complex systems and my opinions stems
from my experience.

To say more though I am not a professional programmers but I have been doing
programming in many languages (DEC PDP-11 Assembler, x86 Assembler, Pascal,
Modula-2 (DEC PDP-11 implementation),  TopSpeed Modula-2,
Borland Delphi, Borland Paradox Object PAL (Accounting software to process
Nortel Meridian 1 PBX CDR data). I have been experimenting with C, C++ (to
understand what they are up to),Prolog, Modula-3.
I was studying OOP paradigms using C++ when ADA95 was not available.

One of the important things in software developing and maintenance is
readability of the code. ADA95 is the best in this matter. Anyone who knows
English can read and understand Ada code and this  is quite the opposite to
C and C++ no matter what C and C++ programmers claim.  This means that any
new person can easily be included in development/maintenance team  and can
easily understand what have been done so far. For long term projects this is
extremely important.

Also if it is necessary to deal with something that should be more efficient
(to programmer's view) the best choice is  to use small subprograms and
function written in assembler and call them from ADA code.
Assembler is not more difficult to read than C and C++  let alone code
efficiency.


Vladimir Olensky (Vladimir_Olensky@oxy.com)
Telecommunication specialist,
Occidental C.I.S. Services, Inc. ( www.oxy.com ).
Russia,
Moscow.



****************************************************************************
*********************

                                             STANAG 3912
                                             (Edition 3)

      NAVY/ARMY/AIR
                   NATO STANDARDIZATION AGREEMENT
                              (STANAG)
                Ada - REAL-TIME HIGH ORDER COMPUTER
                       PROGRAMMING LANGUAGE
Related Documents:  ANSI/MIL-STD-1815A                    ISO/IEC
8652:1995(E)

AIM1.  The aim of this agreement is to achieve easier
      development, improved maintenance and a decrease in life cycle
      costs of avionic systems, by using a standardized real-time,
      high order, computer programming language.AGREEMENT

2.  Participating nations agree:
    a.   To use the high order programming language Ada 95
         for programming avionics software in future avionics
         acquisitions.  Ada 95 is defined by ISO/IEC
         8652:1995(E).  Participants may continue to use
         Ada 83 as defined by ANSI/MIL-STD-1815A for systems
         already in production or in service where the cost
         to migrate to Ada 95 is prohibitive.
    b.   That future avionics systems will employ this
         standard when consistent with system requirements
         and constraints as determined through an acceptable
         systems engineering process.IMPLEMENTATION OF THE AGREEMENT

3.  This STANAG is implemented when a nation has issued
instructions that applications will be in accordance with thisagreement.

Implementing countries currently are:BE, CA, FR, GE, NL, NO, TU, UK

Promulgated on 11 September 1997