[PLUG] Article in Linux Today about M$ kill UNIX at Universities..

[A Braunsdorf]

I think most of you are missing the point.  That's because you've
come in too late.  Let me lay a little historical perspective on
you:

It used to be that Purdue turned out some damn fine programmers
and computer administrators.  Do you know where they were taught?
CS?  By degree, maybe, but that's not where they made their bones.
EE?  Nope.

PUCC!

It may be hard to fathom in these days when they're considered a
Great Satan, but PUCC used to be one of the top training grounds
for computer people.  I know- I was there.

We used to take in students who wanted to learn, trained them as
programmers and admins in a real environment, cranked up the duties
and responsibilities until they were pretty darn good.  Then they'd
graduate and either work for us or go get a "real job".

The system worked great through the 1980s and into the early '90s
until PUCC management made us stop.  Their reason of note was that
the part-timers churned too fast- just when we got them up to speed,
they'd leave.  We were training people and then not getting to reap
the benefits.

Duh, that's what a University is supposed to do!

My brother (ksb) tried to restart the program as a volunteer effort,
figuring that if the guys weren't on the payroll, there'd be nothing
to stop it.  Well, it got harder when we couldn't provide office
space or workstations, but that evolved into the Purdue Daemons,
which wasn't as productive, but was still fun.

PLUG is a similar organization, though you don't have a ksb to
pontificate at you. :-)


This and other (mis)management decisions led to a massive exodus
from PUCC.  How massive?  I stopped counting at about fifteen
resignations (I was number eight, I think).  We didn't leave for
more money (though most of us got it).  We didn't leave for fame
(you wouldn't know any of these guys from Adam).  We left because
the culture died.

And >that< was the culture that got stuff done.  And, yes, it was
the UNIX culture.  More importantly, it was about learning, exploring,
sharing, and teaching.

The current situation is a desire to run stuff "off the shelf".
That'd be fine if it worked, but it doesn't.  It doesn't work
because it doesn't scale and most of the software's not that good.

And guess what?  It's not going to get any better.

If something is wrong on a UNIX machine, and you've got source and
some good people, you can fix it.  It gets better all the time.
If you share with people at other sites, it gets better faster.
That's the open source idea, and it's been intrinsic to University
computing throughout the UNIX era.

If you have an NT system, it doesn't matter how many people you
have.  Unless you're at Microsoft HQ, you >can't< fix it.  It won't
get better.  No user serviceable parts inside.


Off the shelf stuff will improve when computers hit market saturation.
That's still a ways off.  When that happens, people will get fed
up with applications and systems that crash.  Now they don't know
any better.

Microsoft is following in IBMs shoes, and they'll meet the same
"end".  Microsoft convinced people more than IBM did that you could
use microcomputers in place of mainframes.  That was mostly because
IBM was trying to play both sides.  Microsoft is trying to push
compatibility and proprietary technology at the same time.  We
already got them to accept the Internet, other things will unravel
as we go along.

Other companies will step up and change the way microcomputers (and
other computing devices) are used.  Microsoft will continue to be
big (do you have any idea how big IBM is?!), but their profile will
change.  It's got to.  They've been riding the wave, but there's
another one coming.


But what's that got to do with education and culture.  I'll tell
you what.  Five years ago no one used Windows 95.  No one.  Look
at the name.

Ten years ago, no one used the web.

Twenty years ago, no one used an IBM PC.

Big big changes happen all the time.  That scares big companies
that can't move that fast.  It doesn't scare me.

What does scare me is the conflicting forces on teaching our new
generations.

On the one hand are the reactionaries who teach the flavor of the
month.  I see a lot of that in CS here.  They taught FORTRAN for
many many years in CS 180.  Since they quit that they've been
through how many languages?  Are they teaching programming or a
particular language?  They should be teaching the former.  My guess
is they are really teaching neither.

On the other hand are people who refuse to improve things.  That
counts most UNIX people I know.  Me too when my heart's not in it.

They're both wrong and they're both headed for trouble, both in
practice and as teachers.


If you coddle computer people in training and make them think that
what they're doing now is what they'll actually be doing later,
you do them a great disservice.  I assure you I am doing nothing
with computers that has much resemblance to what I was taught in
high school and earlier- and much of it uses only theory from what
I learned in school later.

If you teach people that everything in the field is new and that
there's nothing to learn from the past, you're perpetuating The
Myth of The New that keeps people thinking that everything was just
invented yesterday and so of course it's thrilling and buggy.

What's most important to learn and to teach is that you need to be
able to use what you've learned in new situations.  That requires
you to learn things, learn why they're important, and to learn how
to use that knowledge in different ways.

In these days when little kids are made "computer literate" by
learning to operate a web browser (that seems to be the new standard),
I'm sure we're losing not only our knowledge, but the ability to
understand that knowledge and put it to good use.  That's not
education.  It's sheep farming.


So, do I think it's a tragedy that people are giving up UNIX culture
for NT... whatever?  Yes, but for several reasons:

First, the stuff doesn't work.  NT isn't reliable enough to use
for mission critical stuff and it can't be repaired.  This will
teach people that computers are hopelessly unreliable and to believe
in voodoo fixes instead of systems programming.  ("Let's reinstall"
"Maybe a reboot will help"  "If I smack the monitor just right it
gets better")

Second, it doesn't scale.  For just one example: can we really make
as many accounts on an NT machine as The PUCC/ECN Axis has?  How?
One at a time with a mouse?!  We used to developer a new accounting
system every two years at PUCC until they decided to stick with
acmaint, which was developed with the awareness of what'd come
before.  Has Microsoft done that much work on the problem?  Why
not?  Who will?

Third, we already know that single-sourcing is bad.  Microsoft
helped teach us not to bet all our money on IBM.  What does that
say about Microsoft?

Finally, all the usual Microsoft criticism- the justified stuff at
least.  Microsoft makes a living by stealing from other companies.
As the sources of stuff to steal dry up, they won't have much left.

The University brain drain is another example of this.  They won't
make NT-ready systems people- they won't make any systems people
at all.

Either that or they'll convince more people to come back to UNIX.

ab