Re: An admittedly biased Ada/C++ comparison, by Ed Schonberg

Re: An admittedly biased Ada/C++ comparison, by Ed Schonberg

[Gregory Aharonian]

    Much like the Mosemann studies, what's the point of this message?
Schonberg at the outset invokes the old Turing-equivalence of languages,
and then spends time implicitly arguing against it with his Ada and C++
comparisons, ignoring, much like the Mosemann studies the socioeconomics
of software development in a free market, which in the United States favor
developing software in C/C++, despite many of the benefits that Ada does
have over C/C++.  It seems silly to spend billions to develop software to
defend the free markets of America if we refuse to participate in these
free markets, which to some extent the DoD is doing with software development
policies that completely ignore the free markets of software.  It's the
opposite of the old joke in the CASE world :  "Our product is so good that we
use it to develop the product" in the defense world becomes "Our free markets
are so worth defending that we don't use them for defense".

     I was at the CASE WORLD exposition in Boston this month.  THe CASE
world has software technology NOW that is a few years ahead of anything in
the STARS program and other DoD software initiatives, making me wonder why
the DoD it feels it has to spend money to develop such capabilities.  Two
of the STARS primes were at CASE WORLD - IBM and UNISYS - but their large
booths did not even mention Ada or STARS.  Are they embarassed to show the
crap they feed to the DoD to the more discerning general public.
    One of Ada's shining points - the Booch components - were at CASE WORLD,
but only a company advertising a C++ version of the components.  I asked why
not push the Ada version, and they said (heresy) there was no profit in it.

Ada versus C/C++ - has nothing to do about software and everything to do
about economics.

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

An admittedly biased Ada/C++ comparison, by Ed Schonberg

[pa.dec.com!nntpd2.cxo.dec.com!bonmot!wallace]

Gregory Aharonian (srctran@world.std.com) has put his finger on the real
issue behind the C++/Ada religious wars -- economics.

I can not agree with the slams that are put forward by Gregory.  The issue
behind the economic arguement is one that can be brought down to
implementation and education issues.  I list them here:

1) The "Law of first exposure" is the first criteria that sets an end user's
expectation of what to use as an implementation language.  The first exposure
that a potential programmer has is an inexpensive system (IBM/PC, Mac) in
high school (or at home) where the decision by the educational institution
(individual) is to maximize silicon over software.  So what is the cheapest
"real" language to use?  C of course (it used to be BASIC, remember?) because
a compiler is $100's per site or free (remember Apple's give-away?)  versus
$10,000's per site.  Next, if the student decides to attend university, the
next system seen is the Unix or Unix-style O/S with the "de facto" C
compiler, and if the University has sprung for a C++ compiler, that is
available too (don't forget the Gnu project G++...).

2) Now take that person to the job site and the "big machines" are the same
PC's and Mac's that were seen at high school and Sun/HP/Digital systems that
were seen at university; And of course, what language?  C/C++ of course.
Remember the silicon versus software cost issue?  It comes back in buckets
here because the software house is a)a VERY cost conscious operation, and
wants people in front of machines, and b) concerned with porting, so it will
choose a simple language that has alot of ports; and since C/C++ compilers are
dirt-simple to implement (not to mention de facto on Unix systems...) it is a
good business choice.

3) Gregory misses an important economic issue and that is the reliability of
the code that is produced under the C/C++ development paradigm.  I've seen
lots and lots of C code and now I'm seeing C++ code that has the same
trashed-out coding style I've seen from C "hot-shots."  I say to myself
"surely the compiler would have complained about that, if not then why not
the linter (we use several)?"  and the answer is that a) no the compiler
didn't, and won't, catch it, and b) the linter switches were turned off
because the programmer "didn't want to see that useless noise" that was
output.  These would have been fatal compilation errors out of *** ANY ***
Ada compiler I've used.  This is the issue that the U.S. Government wants.
Not to grow the individual (who is around for 20 years on a single program?)
but to grow the technology so that second-sourcing -- not a well understood
concept in the software industry -- can be done on the software of the
Government computer software.

Now if I could get an Ada given away to high schools and bundled with an O/S
then I don't think there would even be a C++...  but that is a topic for
another discussion.

Richard Wallace
Digital Equipment Corporation
301 Rockrimmon Blvd. South
CXO2-1/7A
Colorado Springs, CO 80919-2398
(719)548-2792
<wallace@cookie.enet.dec.com>

	"The opinions expressed are my own, Uncle Ken or Uncle Bob
	 may, or may not, agree with me."

Re: An admittedly biased Ada/C++ comparison, by Ed Schonberg

[Val Kartchner]

wallace@bonmot (Richard Wallace) writes:
: 1) The "Law of first exposure" is the first criteria that sets an end user's
: expectation of what to use as an implementation language.  The first exposure
: that a potential programmer has is an inexpensive system (IBM/PC, Mac) in
: high school (or at home) where the decision by the educational institution
: (individual) is to maximize silicon over software.  So what is the cheapest
: "real" language to use?  C of course (it used to be BASIC, remember?) because
: a compiler is $100's per site or free (remember Apple's give-away?)  versus
: $10,000's per site.  Next, if the student decides to attend university, the
: next system seen is the Unix or Unix-style O/S with the "de facto" C
: compiler, and if the University has sprung for a C++ compiler, that is
: available too (don't forget the Gnu project G++...).

Actually Pascal is used in more high schools as a teaching language than C.
My high school currently teaches Pascal (though they used to teach BASIC,
FORTRAN, and COBOL), and (to my knowlege) has never taught C (I have visited
a few times since graduating).  "The Law of First Exposure" would then
dictate that students would migrate to Pascal or a Pascal-like language.

At Weber State University, Ada is the required language.  However, it is not
the most popular.

"The Law of First Exposure" can also be generalized to mean that the order in
which computer programming languages are learned is the order of preference.
If this were applied to me specifically, then my most prefered languages
would be BASIC, FORTRAN, COBOL.  This is definitely not the case.

What should actually be applied is "The Law of Only Exposure".  Many
programmers (unfortunately) only learn one language.  This makes discussions
about languages on technological merits difficult if not impossible.
Additionally, since most "programmers" don't know an assembly language,
they generally lack an understanding of what a computer actually does with/to
their HLL program.  Lacking this understanding makes it difficult if not
impossible to discuss efficiency issues.

Programmers need an understanding of language merits and efficiency to be
able to pick the right tool for the job.  So, programmers should know
several languages well.

--
|===== The previous was my opinion -- val@csulx.weber.edu =====///============|
| "AMIGA: The computer for the creative mind." (tm) Commodore /// Weber State |
| "Macintosh: The computer for the rest of us."(tm) Apple \\\///  University  |
|============== "I think, therefore I AMiga." =============\///=== Ogden, UT =|
 

An admittedly biased Ada/C++ comparison, by Ed Schonberg

[pa.dec.com!nntpd2.cxo.dec.com!shodha!wallace]

Val Kartchner clarifies what I wrote.  I, indeed, was refering to what he 
writes as "The Law of Only Exposure" thus:

: What should actually be applied is "The Law of Only Exposure".  Many
: programmers (unfortunately) only learn one language.  This makes discussions
: about languages on technological merits difficult if not impossible.
: Additionally, since most "programmers" don't know an assembly language,
: they generally lack an understanding of what a computer actually does with/to
: their HLL program.  Lacking this understanding makes it difficult if not
: impossible to discuss efficiency issues.

Val does seem to miss the point I was making regarding economics though.
Pascal is still a more complex compiler than is any BCPL-based (C, C++
according to Horowitz) language to port; and as such would not be
available to the Pascal/FORTRAN/COBOL writer at either unversity or 
job site.  Having just said that I know that there are a great deal of
universities and businesses that have these -- as well as a myriad of
other languages -- available.  What I want to focus on here in this
thread is the cost of compilers.  I haven't done any scientific survey
so what I'm writing about is knowledge of personal experience at buying
compilers.  To bend a phrase from the movie/book "The Right Stuff"

	"It isn't technology that dictates what compiler is used,
	 its funding."

Val continues to write:

: At Weber State University, Ada is the required language.  However, it is not
: the most popular.

I'm glad to hear this.  If a survey were taken with the following type of
questions I wonder what language would come out the majority response?
The first question is a control to make sure the questions are a)
understood and b) an indicator to understanding of reserved words in a
computer language.

1) What is your native language (ex. English, French, Spanish, et.al.)?

2) What was the first computer language that you wrote programs in?

3) What assembly/macro assembly language did you first learn?

4) To your personal knowledge, what compiler would you buy for your
personal computer if cost is not a problem (you have unlimited dollars
to use)?

5) To your personal knowledge, what compiler would you buy for your
personal computer if cost is a problem (you have to buy the cheapest)?

Other questions in this vein could be asked.  Give it a try Val.
University is a great place to try this.  You could even design a survey
program to collect the data.  I, for one, would like to see the results!

Aloha,
	Richard

Richard Wallace
Digital Equipment Corporation
301 Rockrimmon Blvd. South
CXO2-1/7A
Colorado Springs, CO 80919-2398
(719)548-2792
<wallace@cookie.enet.dec.com>

	"The opinions expressed are my own, Uncle Ken or Uncle Bob
	 may, or may not, agree with me."