From: John Unekis <etn-rad!jru>
   Newsgroups: comp.lang.ada,comp.lang.c,sci.space,sci.space.shuttle
   Subject: Re: "C" vrs ADA

   Date: 18 Aug 87 18:09:26 GMT
   Keywords: Any suggestions?

Glen asks various questions, to which John replies:

   The ada language is far more than just a language.  Ada includes
   standards for editors,compilers, and run-time symbolic de{*filter*}s.

Ada is the name for a language, only.  There are other standards
which were developed to go with it, which have their own names.
The other standards are not being very successful, so far.

   The C language evolved at AT&T in the process of developing the UNIX
   operating system. There were, beleive it or not, an A language and a B
   language that preceded it.

C is based upon B is based upon BCPL is based upon CPL.
A is also based upon BCPL, but is not an ancestor of C.

   Finally with the C language the original developer of the UNIX
   operating system (which was done on a PDP-7 microcomputer) felt that
   he had what he wanted.

The C language was developed by Dennis Ritchie, who with Ken Thompson,
the inventor of B, jointly developed Unix.  Unix was originally
developed on the DEC PDP-7 and PDP-9 minicomputers, but C was
developed for the port to the PDP-11.

   It was a moderately structured language , with a syntax that was
   similar to the UNIX c shell (or vice versa).

This is a wild notion!  The C Shell was developed by Bill Joy
at UC Berkeley about ten years after Unix and C were invented.

   As UNIX gained wide acceptance the C language became more popular. It
   has the advantage over fortran of having structured variable types,
   but without the overly elaborate type checking done by a language like
   Pascal.

Pascal's type checking is not elaborate, and is actually simpler than
C's.  However, its more restrictive.  Pascal checks for type
compatibility by name, C by structure.  C also does more automatic
conversions - except for function arguments, which makes function
calls fragile and dangerous (ANSI is fixing this).  Each style of type
checking has its advantages.  Some languages provide both.

   It does share with Pascal the fact that each was developed by
   a single individual and thus represents that individuals prejudices in
   computer languages.

Ada was also developed by a single individual, Jean Ichbiah.  However,
the specifications were developed by many people, so this distinction
has some truth.

   C is now widely available outside the UNIX community and is a defacto
   standard with many companies.  It is often the case in
   military/aerospace procurements that in order to generalize the
   request for computer hardware so as not to sole source a single vendor
   the government will ask for UNIX because it is the only operating
   system that can be had on a non-proprietary basis on a variety of
   computer hardware. UNIX of course brings C right along with it.

C's popularity with the DOD and their vendors is no longer dependent
on the popularity of Unix.  In my experience with these people, most
are using C with PC-DOS.

   Because UNIX does not handle real-time applications (such as
   interrupt handling) very well, and because there was no non-
   proprietary standard for real-time operating systems, the
   government(DOD) wanted to develop such a standard.

Here, operating systems are confused with languages.  Ada is often
used under Unix; Unix has not been changed.  Also, there are many
realtime versions of Unix, from AT&T and others.  Ada is suitable
for programming realtime applications, provided that the underlying
O/S, Unix or other, can support such.

   Also, the DOD had a problem with each branch of the service having
   its own non-compatible languages for software
   development(COBOL,JOVIAL, CMS-II,FORTRAN,C,etc.). It has been decided
   that the DOD will develop a standard computer language that will
   include standards for editors, compilers, run-time de{*filter*}s, and even
   operating system functions for real-time processing. This standard was
   named ADA, (the name of the {*filter*} of Charles Babbage, who
   invented a punched card driven loom, considered to be the first
   computer, she was rumored to be the first person to ever write
   a program on punched cards- why her name is appropriate for a
   real-time language is a mystery).

Again, the name Ada refers only to the language.  This history of Ada
and Charles Babbage is so off the mark it sounds like intentional
satire.  Here's the real story:

Ada Agusta (Countess of Lovelace and daughter of Lord Byron, the poet)
and Charles Babbage were lovers.  Babbage did not invent the punch
card controlled loom - Jacquard did that.  Babbage used the idea in
his design for his Analytical Engine, an early design (perhaps the
first) for a mechanical computer.  The construction of the Analytical
Engine was never completed.  It was too ambitious for the machining
capabilities of the time.

Ada was credited with a number of insights about the programming of
such a machine, including the notion of using it for symbolic
computation and reasoning, instead of just for arithmetic.  (A recent
biography of her suggests these ideas were actually Babbage's.)

   If you are interested in the military/aerospace field then ADA is
   definitely the language to learn. Be aware that it is a very complex
   language (Carnegie Mellon University is rumored to have required it
   for all sophomores- which resulted in flunking out half their
   sophomore class) and that to learn it properly you must find a DOD
   certified implementation which includes the editors, compilers, and
   de{*filter*}s as well.

Ada is indeed complex relative to C or Pascal (although much simpler
than PL/1).  My opinion: Ada is not a good language for teaching
beginners.  Its constructs require lots of knowledge and experience to
use correctly.  I've never heard this rumor about CMU, and intend to
disregard it unless I hear it from a more reliable source.  If you are
an experienced programmer and want to learn Ada, find a book that is
well recommended, and/or a good teacher (shop carefully for these).
You will want to have access to a validated Ada compiler, but special
editors and de{*filter*}s are not needed.

   The DOD plans eventually to require ALL software to be done in ADA,
   but they realize that there is enormous inertia against it.  Thousands
   of programmers must be retrained, and millions of lines of code
   converted. Don't expect to see ADA used very widely outside of the DOD
   environment. It will fail for the same reason that Pascal, Modula2, C,
   PL1, and others have failed - IBM is the {*filter*} force in the
   commercial market(~75 percent of all commercial installations) and
   COBOL dominates the IBM installed base (~90 percent of IBM
   applications are in COBOL).

Ada was designed for writing systems programs, primarily for embedded
systems.  It is not in competition with COBOL, nor is C.  Far from
failing, C has become the most popular systems programming language
ever invented, and its popularity is still rising fast.  Pascal was
designed for teaching programming, and is widely used for this purpose
at Universities.  It is being pushed out by C for many programming
tasks, but it has not failed in its primary purpose.  My personal
opinion: Pascal is much better suited for learning programming than
either C or Ada.  Time spent learning Pascal well is time well spent.
My students who know Pascal do far better with C than the others.

   As long as computers remain basically Von Neuman processors, no
   language is going to offer any advantages in the real world to a
   language like COBOL. No business is going to go through the 3 to 5
   years effort of retraining and converting of existing code just to
   satisfy the dogmatic prejudices of computer-science weenies.

Note the strong prejudice expressed here.  This kind of demeaning
characterisation has no place on the network.

   The DOD is perfectly capable, however, of making contractors like
   Boeing, Lockheed, TRW, Eaton,etc. jump through the ADA hoop just
   by refusing to accept bids which do not include ADA. Therefore if
   you want a career in military/aerospace, go for ADA.

   ---------------------------------------------------------------
   ihnp4!wlbr!etn-rad!jru   - The opinions above were mine when I
   thought of them, by tomorrow they may belong to someone else.

No service is provided to the community by this kind of article.  My
sense from the style of the article is of little concern for accuracy.
This is well supported by the remarkable number of errors.  The
mixture of facts with opinions hurts both, and the virulent prejudices
serve no one.  We all make these errors from time to time and to a
certain extent.  This particularly bad example may inspire us all to
be more careful!

_Greg

J. Greg Davidson                          Virtual Infinity Systems
+1 (619) 452-8059        6231 Branting St; San Diego, CA 92122 USA

By the way, Babbage not only understood the machine technology of his time,
he may have made an indirect major contribution to it.  His early Difference
Engine used interchangeable parts many years before Eli Whitney "invented"
this concept.  One of the people involved in that project was Whitworth [?]
[this is from memory of a seminar some years ago], who later devised the
world's first standardized{*filter*}thread.
--

Next time, we should open it.        | {allegra,ihnp4,decvax,utai}!utzoo!henry

If you want to learn Ada, by all means, learn it.  Just don't expect
it to magically solve all your problems.

-Dave Sill                     "It is the tragic fate of most individuals

The opinions expressed above are those of the author and do not
necessarily reflect those of the Department of Defense or the U.S.
Navy.

I am sorry Greg, but I could not disagree with you more.

The right to hold and express an opinion is VITAL to our basic freedom, both
as engineers/programmers/users, and as citizens.  Mr. Unekis's posting gave
the world a chance to hear what he thought, which is the whole purpose of this
net - the free and open exchange of opinions, ideas, answers, questions, facts,
rumours, gibberish, and just plain information.  I DO NOT think that John's
posting was "virulent", and I think that he did indeed provide a service to
the community, by providing topic for discussion.  This makes people think, and
thinking is a good thing.

I am now going to make a bold statement, which I will stress is my opinion.

ADA SUCKS !!!!

I am of the opinion that there are only three good high level computer
languages for most
situations.  I have seen and programmed in all the major languages ( I have
probably not seen some estoteric ones that are quite nice, I am sure, but
limited in scope or usage).  

For doing serious programming, the ONLY language is "C".  I have NEVER seen
an application written in another language that would not have been better/
faster/more flexible/more useful if re-written in "C".  This includes ADA and
COBOL (retch!!).  I proved this to a previous employers by redoing a DBMS
update program in "C", where before it had been in COBOL.

For hacking device drivers, writing critical code segments, or doing very
time critical applications, write it in "C", and then hand-optimize the
critical segemnts in Assembly Language.

For quick & dirty problem solving type programs, do it in FORTH.  FORTH is
also wonderful to play around in.  I have written some great games in FORTH.

For teaching programming to new CS'ers, use PASCAL.  PASCAL is a brute becuase
of its strong limitations, but it teaches good programming etiquette and style.
I also like PASCAL, particularly TURBO PASCAL for the home hacker.

The above are my opinions, and reflect only my face in mirror.

Randy Martens - !rochester!kodak!elmgate!ram
"Reality - What a Concept !!" - R.Williams

  It's niece to be back from vacation.....now back to some bashing :-)

  This stuff really doesn't belong in comp.lang.c it should be comp.lang.misc
or something like that but ......

   I think you've missed a very important aspect about programming in
different languages.  The languages we choose to program in, shape the
way we think about a solution to a problem.  They also affect the style
and methodology we use.

  An example...  Many people abbuse pointers in C.  The classic fast
4 (sizeof(int)) byte copy code.

                * (int *) x = * (int *) y;

This is not portable. It should be coded as:

                x[0] = y[0];
                x[1] = y[1];
                x[2] = y[2];
                x[3] = y[3];

IF portability is an issue - it might not be.

  A language like Pascal will not allow the first construct.  There are many
other examples where C allows one to write code that is machine specific,
that one could not write in a language like Pascal.  Does this make Pascal
a better language then C?  Some people would say yes, I would say no but then
again thats another story. ;-)

  Is this a good enough example of how a language can effect the way think
about a solution to a problem?

--

                                        Eddie Wyatt

Babbage, Whitworth and two of the other team members co-authored *the*
foundation text in the field of precision machining, a weighty tome with a
title that escapes me at the moment. Ask any historian of technology about it.

The article I got all this from claimed that the book is *still* occasionally
cited by the older generation of instrument makers and tool-and-die men, who
remember and revere Babbage's name without having retained as part of the craft
lore the original purpose of his work.

Sorry, I can't give hard sources for this. It might have been one of James
Burke's coffee-table books on the history of technology. --
      Eric S. Raymond
      UUCP:  {{seismo,ihnp4,rutgers}!cbmvax,sdcrdcf!burdvax,vu-vlsi}!snark!eric
      Post:  22 South Warren Avenue, Malvern, PA 19355    Phone: (215)-296-5718

Dale
--

UUCP: ...!seismo!harvard!mit-eddie!cullvax!drw
Apollo was the doorway to the stars - next time we should open it.
Disclaimer: Don't sue me, sue my company - they have more money.

    The right to hold and express an opinion is VITAL to our basic freedom,
    both as engineers/programmers/users, and as citizens.  Mr. Unekis's
    posting gave the world a chance to hear what he thought, which is the
    whole purpose of this net - the free and open exchange of opinions,
    ideas, answers, questions, facts, rumours, gibberish, and just plain
    information.  I DO NOT think that John's posting was "virulent", and I
    think that he did indeed provide a service to the community, by
    providing topic for discussion.  This makes people think, and thinking
    is a good thing.

The right to express one's opinions are not in question--it's the proporting
them as fact which is being called into question here.  There is a spectrum
of experience in the readership of this newsgroup.  Many of us are able to
discern fact from opinion, but not all of us.  Expressing misconceptions and
heresay as fact, without any disclaimers on its accuracy, is a disservice to
this community.

--

- Stephen

Oh, excuse me, there is a another subject I wanted to bring up.
What has all this Ada/COBOL/Babbage/370 stuff got to do with C,
anyway?  Not much.  Let's try to keep our postings a little more
germane to the discussion of the C language.  Before I get a string of
"pot calling the kettle black" messages,  I'll admit I'm guilty of
this myself.  Sometimes it's awfully hard to resist the urge :-)

-Dave