File spoon-archives/list-proposals.archive/list-proposals_1998/list-proposals.9807, message 9


Date: Tue, 28 Jul 1998 00:58:21 -0400 (EDT)
From: malgosia askanas <ma-AT-panix.com>
Subject: Re: Wynship's proposals


I have actually worked on a number of anti-engineering projects.  My first
anti-engineering job was at Raytheon Data Systems, in the late 70s, where
they wanted to build a computer just like the IBM Series 1 with an operating
system just like Multics.  None of the people working on the OS had the 
slightest idea what an OS was, but we read many papers and each group --
there was a scheduler group and a memory management group and a file-system
group and a couple of others -- designed its piece using the latest research
papers.  The interpersonal dynamics were, for some reason, very passionate
and the groups all hated each other, so they purposefully kept from each
other the details of the interfaces.  So we had all these specs and all these
partly-written pieces and then they called in a very prestigious consulting
firm whose job it was to prove that the system would run correctly and efficiently.
And they proved it.  Two months later, the project was cancelled.

Then I worked on the world's most ambitious optimizing compiler.  We built this
unbelievably optimizing global optimizer which represented a materialization
of the most advanced research available at the time.  The project lasted 5 years
and towards the end involved 50 people.  The compiler produced excellent code
but ran 40 times too slowly.  The project was cancelled.

Then I worked on some factory-of-the future projects (this was the time of the
factory of the future).  Then I worked on some parallelizing compilers for
some state-of-the-art parallel computers.  Then I worked on Ada compilers
for the Brilliant Eyes racket and on rapid-prototyping systems for DARPA.
None of these yielded anything that was ever used, although all of them yielded
according to spec.

These were most definitely engineering projects, and in many cases the 
engineering was quite excellent.  They were intellectually challenging, 
rigorous and innovative.  Their ultimate uselessness was both sensed beforehand
and taken as a matter of course when it was confirmed.  Some were government
contracts, others were commercial and/or industrial.  I think that there are
thousands of people who make their living as anti-engineers.


-m

   

Driftline Main Page

 

Display software: ArchTracker © Malgosia Askanas, 2000-2005