Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site fortune.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxl!houxm!ihnp4!fortune!ojs
From: ojs@fortune.UUCP (Oliver Sharp)
Newsgroups: net.micro
Subject: Re: S1 & NCC
Message-ID: <3940@fortune.UUCP>
Date: Fri, 3-Aug-84 12:13:48 EDT
Article-I.D.: fortune.3940
Posted: Fri Aug  3 12:13:48 1984
Date-Received: Sat, 4-Aug-84 03:19:16 EDT
References: <642@sri-arpa.UUCP>
Organization: Fortune Systems, Redwood City, CA
Lines: 59

[]

  A couple of things about the article posted to the net recently:

I am currently learning UNIX (I ought to be, given that we sell a UNIX
box!) and have been learning some of the delights and pitfalls of this
OS.  Beginners are always the ones who complain about the system
because it is hard to learn, the documentation is not
beginner-friendly, etc, etc.  These complaints are, by and large,
true.  One thing that people seem to lose sight of quite often is that
UNIX was written by software developers FOR software development.  It
is a wonderful environment for just that.  I call the orientation of
UNIX expert-friendly, and I think it is great.  Once you know how to
use the system, a process that takes quite a while, you can be very
productive with it.  I have an IBM PC and I keep adding programs to it
trying to get up to the level of functionality a normal UNIX
system already has.  A couple of languages, make, communications
software, text editors, utilities galore, etc. are all standard.  Each
of these must be purchased separately for the PC and don't work
together nearly as well as the UNIX utilities do.  So, while it is
time consuming to understand the UNIX system, if you want to do
software development, it is worth the effort.

Another complaint mentioned in the article was that there are all
these versions of UNIX floating around and nothing will transfer from
one to the other without changes.  Again, this is true.  However, I
would like to see ANY other OS that runs on machines from an IBM PC up
to DEC VAXen and higher still that allows ANY transportability at all.
This is not really an argument in favor of UNIX over a new, even
better system, but it is a favorable note in the real world where
people are trying to get work done.  It is possible to port a UNIX
program between versions with RELATIVELY FEW changes most of the time.
If the program was written with portability in mind, it is usually
quite easy.  Porting across AT&T versions is generally not too bad,
and things generally don't get too tricky unless you start dealing
with Berkeley's new version, 4.2 BSD.  There have been a number of
changes in it which sacrificed compatibility for functionality.  A lot
of people don't like 4.2 for that reason.  Anyway, the point is that
there is a lot of compatibility and most programs port fairly easily
most of the time.  Part of this is also due to the fact that almost
EVERYTHING in UNIX is written in C (which I like a lot) and almost
nothing is written in Assembler (which I also like, but .....).  This
is why it is so easy (relatively speaking, of course) to bring up UNIX
on a new machine.  Once the assembly stuff is rewritten, the C that
sits on top of it doesn't need to be changed too much.

While I see some of the faults that others find in UNIX, there isn't
anything I've come across that does better.  BTW, the UNIX idea of
having a single file type is NOT bad - it is a cornerstone of the
whole system and is a GREAT idea.  I just thought that someone ought
to mention the other side.  I'm sure some of the UNIX gurus out there
could write a much more spirited rebuttal, so come on guys - out with
it!  I think we could get a good discussion going (in net.micro.pc???
- maybe we should switch groups; I can feel the heat already).  I love
mail - send flames, agreements, etc.

Oliver Sharp
.......!fortune!ojs