NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c++-part2/12962

Internet Message Format | 1996-08-05 | 6.0 KB

Path: slwol1p47.ozemail.com.au!user
From: [e-mail] (John Maling)
Newsgroups: comp.lang.eiffel,comp.lang.c,comp.lang.c++,comp.object,comp.software-eng
Subject: Re: Beware of "C" Hackers -- A rebuttal to Bertrand Meyer
Date: Sat, 23 Mar 1996 04:13:01 +1000
Organization: Zip Nada and Zero, Attorneys at Law
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
NNTP-Posting-Host: slwol1p47.ozemail.com.au
X-Newsreader: Value-Added NewsWatcher 2.0b24.0+

In refeference to the Dr Meyer quote, as duplicated by Robert Martin in
message <[e-mail]> (I wont repeat it again), this
is my opinion.

I believe C is an old language (examples of it may have been discovered on
temples that predate the Roman Empire). Seriously though, my first
experiances with programming PC's was in 1982. Back then, when I looked at
what language I was going to initialy learn, there was C and Pascal. C was
undergoing the same 'push' that OO is undergoing now. It created
comparitivly faster and smaller executables than Pascal, it was a high
level language with lower level hooks and its libs were better optimised
(why you could save a whole 8K!). It became (in books and the press of the
time) The Way of The Future, precisly because of this- and *possibly*
because it was popularly believed UNIX was going to take over the world
(even down to PC's), and C was its anointed disciple.

Back then the most admirable thing was to squeeze the fastest code into
the smallest executable space, and 'Readability, Provability and
Extendibility' sat in the back seat, if they traveled together at all (at
least it worked this way for PC programming, perhaps those with the 'vast'
resources of mainframes could afford a more structured approach?). Does
anyone remember the bizarre 'one line program' competitions of the early
80's? Those where you had to create a program in 255 bytes?

I chose to learn Pascal, I thought the C language was too much like a
chainsaw, and I doubted UNIX's ability to take over the world (and did not
want to program in a world where it did). The potential to wield C in just
the way Dr Meyer describes was obvious. Of course that did not mean that
you *would* use it that way, you can write C with as much control as
Pascal, but since it *could* be done it was a natural choise for those who
wanted to go the path Dr Meyer illustrates.

This, I think, is where Dr Meyer puts the cart before the horse. C was
never inherantly evil, it was just the obvious language of choise for
those who wanted to use it in 'hack' ways, though I would argue that the
uses were not themselves bad, just necesary for the speed and space
demands of the day.

Perhaps it could be reasonably said that the nasty practices Dr Meyer
mentions were necesary practices in the early 80's, but now I find it very
difficult to believe that there are still so many of these so called 'C
Hackers' running around that their presence needs to be commented on.
Nasty practises can be achieved in any language, I seriously wonder what
Dr Meyer makes of bad OO designers who never had a C background. It seems
to me that hardware has way outpaced software since then. It needs to be
written by teams now, not by uncontroled genius mavericks working solo,
and OO programming is the anointed disciple of team programming (and
rightly so).

Up until a couple of months ago, OOD&P remained a curious mystery to me.
My logic centres were set way deep for structured programming. People
preaching OOD&P as 'THE WAY' left me stone cold- I thought it just another
language bigotry thing.

However, since then I have 'seen the light'. OOP is indeed The Way (at
least until something better is created). I was converted by <free plug
time> Robert Martins book, 'Designing Object-Oriented C++ Applications
using the Booch method.' I had periodically read books on OOD and always
missed the point, but Martins book lit my head up like a Christmas tree. I
had one of those 'Ahhhh!' experiences. Quite literally, I woke up sleepily
one early morning- sat on the edge of the bed, and suddenly the whole
concept of what makes OOD so good just hit me. Any parent of a newborn
whose baby cries in the night knows well the effect of being transformed
from sleep to hyper-alertness in a split second. It is not a pleasent
sensation. Nevertheless that was the effect on me when OOD principles
finaly clicked for me.

So I can understand what Meyers means by 'serious debriefing', it was only
after my notions of structured programming logic were deeply suppressed
that I was able to accept what OOD&P means (I believe in Zen this is known
as emptying the glass), and my appreciation of the value of OOD&P came all
at once. At the time I also had the insight that C++ is not an ideal OOP
environment. However, nodding my head in the direction of the QWERTY
keyboard and 90% of PC's I have chosen to study C++.

However, Dr Meyer setting this up as a religious issue appeals to the gut,
not to the brain. All he seems to do is cast a light into a pit of C
'hacks' and a spotlight on the pedestal of OOD&P, and say, in effect, that
those in the pit are in hell and do not know it, and those who are dragged
up, or drag themselves up, resent it and/or are unable to effectively
embrace OOD&P. I say kick this concept over to 'C++ beats the crap out of
Ada' or some such. Thats a thread from hell if ever I saw one.

In summary I think Dr Meyer needs to come back to the future.

Oh, and about this 'hack' thing....Grief! What is the point arguing about
the meaning of 'hack/er/s'? You can deduce what Dr Meyer means because he
defines what 'C hack' means to him. You don't have to agree or disagree
with the definition to argue whether his conceptualisation is flawed or
not. Dictionary meanings do change with common usage, even if you dont
like what they change into, much as 'hacker' has changed.

After three days without programming life becomes meaningless. <unknown>