NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c++-part2/11361

Internet Message Format | 1996-08-05 | 5.8 KB

Path: nnrp.info.ucla.edu!jmartin
From: [e-mail] (Jay Martin)
Newsgroups: comp.lang.c++,comp.lang.eiffel,comp.lang.c,comp.object,comp.software-eng
Subject: Re: Beware of "C" Hackers -- A rebuttal to Bertrand Meyer
Date: 14 Mar 1996 04:07:55 GMT
Organization: University of California, Los Angeles
Message-ID: <4i862r$[e-mail]>
References: <[e-mail]> <3taaha$[e-mail]>     <3taodp$[e-mail]> <3tap9h$[e-mail]>     <[e-mail]> <[e-mail]>
NNTP-Posting-Host: may.cs.ucla.edu
X-Newsreader: NN version 6.5.0.b3.0 #9 (NOV)


Geez, when was this thread, a year ago?  Well to recap ancient
history it started when Robert Martin (below) called Bertrand Meyer a
bigot for warning managers about C hacker types.  Rewind thread!


[e-mail] (Robert C. Martin) writes:

>   Jay Martin wrote:
>   > Rabid worship of C increases the probabilty the person is a hacker.

>Rabid worship of any language is inappropriate.  Languages are tools,
>no more, no less.  However, rabid worship of C in particular does not
>increase the probability that one is a hacker; since hacking is an
>attitude with regard to the way tools are used, not an attitude about
>*which* tools to use.  Hacking is "Programming for the moment", as
>opposed to "Programming for the Future".

No "hacking" is broader than that, it is writing poor code period. It
is writing low level or obscure code when it is unnecessary because
you think that its great.  You usually don't see such behavior in
then Eiffel, Ada, Modula*,etc communities.  You only compromise your
software (like writing low level code) when you have a "gun to your
head", then you must curse like hell!

Note that when I use the term "Software Engineering", I am talking
about optimizing the design and implementation of software which means
using the best tools and techniques.  I don't believe in defining SE
as something that only concentrates on processes.  Practical
limitations like "Johnny only knows COBOL" or "tool implementation
Y on target X sucks", are not software engineering technology
decisions that are at all satisfying.

>   > The basic tenents of C culture are anti-software engineering.

>There is no C culture.  C is a language, a tool, not a culture.
>Associating a tool with a culture is rather like associating a skin
>color with a culture.  There is a word for that.  The word is
>bigotry.

>Are there C programmers that are hackers?  Sure.  Are there Eiffel
>programmers who are hackers?  Sure.  What is the proportion?  I'd be
>surprised if a statistical analysis of the difference between the
>proportions of C and Eiffel hackers differed significantly from
>chance.  However, I doubt very much if any such study will be
>conducted.  Thus, any assertion regarding the whether the proportions
>are different is based upon opinion and not fact.

First see,"I Corporation"'s recent post under "Beware of "C"
Hackers..." in "comp.software-eng" which excelently conveys my basic
feelings on this issue.

You have no studies to backup that Eiffel and C programmers are
magically identical so this is just a red herring. (The reason for no
studies is that someone needs to shove a shotgun in the face academic
computer science and pull the trigger).  Competent software engineers
are not hackers.  Hackers do not like the restrictiveness of an
Eiffel or an Ada or any other software engineering language, thus
these languages are not likely to attract hackers.  C is
unrestrictive and attracts hackers in droves.  I have never worked
with an Eiffel programmer, but given an Eiffel programmer spouting
high level design and implementation concepts and a C programmer who
says "I use pointer arithmetic cuz it may be faster, I don't trust no
compilers" and "I write terse code all on my line cuz its cool like
mathematics and gee, I is a mathematical genius or something",
the Eiffel programmer is basically going to get the job.

In my book, to be a competent software engineer you must must also be
an very knowledgable at computer language design and also compiler
optimization.  Its common knowledge that C is poorly designed and
thus rabid worshipers of C are not too swift.

>   > (1) C makes low level code extremely easy to create and the C culture
>   >     says that low level code is cool.

>C also makes high level code easy to create, and there are many of us
>who say that "high level code is cool".

If C is so wonderful for making high level code, then what the bleep
was C++ invented for?  Most code written in C is low level crap. Is
it "easier" in C to write high level code than other "high level"
languages?  I don't have problems with high level C programmers who
understand the limitations of their tools.  I don't see "rabid C
fanatics" who mostly worship the low level in this catagory.

Culture is everywhere.  I have seen many FORTRAN programmers have
brain seizures when they first saw identifiers longer than 6
characters and of lower/mixed case.  I see correlations between
C/Unix fanatics and poor software engineering opinions and
attitudes.  Its politically correct nonsense for human beings not to
try to identify patterns of behavior.   So if a programmer starts
spouting nonsense and doesn't have what I label as a correct software
engineering opinions and attitude, I am not impressed.  I guess I am a
bigot, a bigot against incompetency.

To demonstrate the "C hacker culture" (filed under macho attitudes)
here's a random recent example:

[e-mail] (Robert C. Martin) writes:

>No, I disagree with your point.  It is not the responsibility of the
>language to make the engineer better.  It is the responsibility of the
>engineer to become adept at the language.

Clear demonstration of views that have left software engineering
and entered the "C Hacker Zone".

Jay