tcamp@dukeac.UUCP (Ted A. Campbell) (03/26/90)
WHY THE PROMISE OF USER-SUPPORTED CODE
HAS FAILED (THUS FAR)
(from TCS Bits -- newsletter of the
Triangle Computer Society, Research Triangle Park, NC)
I am now coming to the realization that a generation has passed in the
history of microcomputing. It is a subtle change, perhaps, but my new
Atari Portfolio was supplied without BASIC (or any other programming
language). In the earlier years of microcomputing, this would have been
practically unheard of: almost every unit was supplied with a BASIC,
from the TI 99/4A to the original IBM PC, which booted up with ROM
BASIC if it failed to find a disk drive. (Yes, Virginia, there were IBM
PCs without disk drives.)
The idea, in those days, was that a computer's glory was its
programmability. Sure, you might use a "canned" program for this or
that, but "real" computer users would write their own programs.
Installing our new computers in Duke Divinity School these past few
weeks, by contrast, was more like installing fancy toasters -- you turn
it on, and presto, in a couple of minutes WordPerfect appears on the
screen. No DOS (that you can see), no BASIC, no bits, no bytes. The
user is perfectly "shielded from" the operating system and from any
need to use the computer as a programmable machine.
In the earlier, programmable-computer years, there seemed to be a sort
of unspoken goal or vision that we were all striving for: to have our
own, user-supported code for any purpose, so that we could string
segments together and create our own software worlds. The old FOG CP/M
library had a "Hacker's" section -- it was for code (almost always
MBASIC code) that, well, sort of worked and sort of didn't -- you know,
the kind of code that would scroll onto the screen, ask you a few
questions, and then probably error out. The idea was that it was
"definitely not approved," but might be a starter-point for other folks
wanting to develop similar code.
We seemed to have some actually working code for file-card programs,
telephone dialers, and the like. But what almost no one came up with
(at least, so far as I know) was user-supportable code for the
bread-and-butter microcomputer applications -- spreadsheets, word
processors, database managers, and the like. Hackers seemed to follow
one of two paths: either (a) they developed really crummy code for
really insignificant applications and then blessed us by putting it in
the public domain, or (b) they developed really good code for critical
applications, and then couldn't resist either going commercial or
distributing their product as shareware. The good code just isn't
publicly available.
Now, I already see a few of you Internet types out there ready to
proclaim the glories of Free Software Foundation and the GNU series of
programs. But (please correct me if I am wrong) it doesn't offer the
primary applications: GNU Emacs appears to be everything in the world
(including an asinine counselor) EXCEPT a word processor. I'm not aware
at all of a spreadsheet in the series. Moreover, you've probably
already heard my diatribe on how obfuscated the GNU code is.
Gripe gripe gripe. You can prove me wrong though. Here I offer a
challenge: send me your nominations for the best in user-supported,
publicly available microcomputer code (not canned applications), and
I'll summarize your suggestions for the net.
Ted A. Campbell, N4XCG Duke Divinity School
tcamp%numen@dukeac.ac.duke.edu Durham, NC 27706
dukeac!numen!tcamp (919) 684-6365