Re: Is Ada likely to survive ?

[safetran <safetran@kaiwan.com>]

My ISP does not seem to be getting all the messages from the news server
at sw-eng.falls-church so I can't reply to each of your messages
individually - I can only post my reply to my news server.  I've got a
complaint in with the ISP and hope he will fix it soon.   Sorry about
that.  Also as a result of that this email is pretty long.

>Stanley Allen writes:
>Worries #1 and #3 would also be facing you if you choose C++ or Java.

I would not be worried about C++ that much. Considering the volume of
C++ code produced and the number of programmers of C++  its going to be
around for quite a while.  C came out in the early seventies and its
still around.   Java - I agree.  Its new and one does not know how long
it will be around for.  I also do not know how well Java will turn out
in the hard real time world.  I know that a lot of work is going on in
trying to make it deterministic and there are a couple of vendors who
have succeeded (somewhat).

>Worry #2 is intersting ... are you accepting resumes?  I think a
>lot of Ada program

Yes we are hiring - have about 4 slots open.  Though you may not find
the work very interesting :).   One important bit of info I left out was
that we do **safety critical** systems.  As a lot of you know this can
be a rigorous environment to work in.

Robert Dewar writes:
>I am a little puzzled by (2), hard to find = high market value. But perhaps
>what you should be looking for anyway is *good* programmers. 

What I meant was that (sometimes) people do not want to program in Ada. 
The languages in vogue today are C/C++/Java.  And everyone wants these
to be on their resume.  So sometimes programmers/engineers do not want
to work in Ada as it has "lower market value" -- whatever that means !!

>As to (3), whether your code is maintainable or not depends on whether it
>is maintainable code.

I was not very clear here : yes Ada is very maintainable but if (in 10
years) you can't find programmers who know the language then it does not
really matter how maintainable the language is.   Like a lot of you have
said a good programmer can pick up any language -  I agree with that.  
Its just that if you hire someone and he has to learn a new language -
well that is additional cost vs hiring someone who already knows it.

>If you are trying to look ahead ten
>years to see what will be the popular language-du-jour in the year 2007,
>I think that is an idle excercise. I doubt it will be any of Java, C++
>or Ada, but that really does not matter.

Don't know about that.  C has survived a long time.  If someone was to
ask me to bet the odds of Ada vs C++ I would go for C++.    Again I am
not saying that C++ is better; like I mentioned in my original post I
have used Ada for over 7 years and also use C/C++.

Paul Van Bellinghan writes:
>I think the man is sold on Ada and does not have a problem training "good"
>programmers in using Ada for the company's RT embedded apps. 

Correct.

>Many of the candidates we
>interview are concerned about the future of the Ada language. Even the DOD
>may be moving away from it (there are advocates for both C++ and Ada in the
>DOD). 

Amen.

>What matters, I told them, is having
>experience in applying good programming techniques to RTE designs. Of
>course, it fell on deaf ears.

You are right about this.  I also find that those people who have worked
with Ada generally make better programmers - their code is more
readable.  I have a few colleagues who used to work with Ada but now
work with C/C++ and their code is infinitely more readable that just
straight C programmers.  This is not to say that there aren't good C/C++
programmers out there.

Brian Rogoff writes:
>How many languages *haven't* survived? Cobol, Fortran, C, PL/1, and REXX
>are still thriving. There is even a company selling Algol 68 compilers.

My question was not survival as just survival - but more will Ada become
a bit more mainstream or at least stay as "mainstream" as it is today. 
I agree with you that all these languages have survived but how
mainstream are they  compared to (say) C/C++.

>Unless you are writing code for one of those embedded microprocessors with 
>no Ada compiler, why do you want to switch? I assume you are using Ada 95, 
>right?

I think I have explained why I am thinking of switching.  We were using
Ada 83.  But on the new projects we would try to use Ada 95.

>I believe so, but there is a good free (as in free source) compiler, so 
>you'll never be "orphaned". 

We work in the embedded world and the "Free" compiler is a host
compiler. While GNAT can be ported to be cross we also use emulators and
need a debugger to work with this etc etc..  So GNAT by itself is only
part of the answer.

Jeff Carter writes:

>Ada is the language of choice for very large systems and for
>safety-critical systems.

Yes you are correct.  However a lot of safety critical systems are being
done in C/C++.  A lot of medical electronics company's,  the car
industry,  rail-road industry etc etc do their safety critical systems
in C - I know these to be correct as I have spoken with people from
these industries and work in one of them myself.  Also, even when Ada
was mandated,  I understand that DOD contractors would get waivers and
do their systems in C.  Don't know to what extent this is true but I
have friends who work in DOD related companies and this is what I heard
from them.    So atleast some military and aerospace related software is
in C.    There is a book on writing safety critical software in C and in
the book the author acknowledges interviews he has done with
DOD/aerospace firms and they were using C.

Please do not mis-understand - I am not endorsing any of the above or
saying it is good practice. Just that this is how things are. 

--
Rakesh
Rakesh.Malhotra@Safetran.com