May Symbolics thrive...

[<@IU.AI.SRI.COM:slug-admin@iu.ai.sri.com>]

    Date: Tue, 5 Jan 1993 18:51 CST
    From: qobi@unagi.cis.upenn.edu

    Come on you guys....  The remainder [of Lispm capabilities]
    are more than made up for by the blazing speed.

An old debate.  Performance versus functionality, which matters most?

Who cares?  The relevant question is: why can't we have both?  Sometimes
speed matters most, sometimes specific capabilities matter most, and
sometimes they both matter a lot.  Is there any good reason that we
don't have both, together in one platform, already?  I said GOOD reason.

It's a sure thing that no Lispm can come close to a really spiffy RISC
workstation in performance any more.  It's just as clear (to me at least)
that no software environment has yet come close to the quality of Genera
for program development.  I put Lucid/GNU/whatever somewhere near halfway
between Borland on a DOS/Windows PC and a Lispm, with the moderately
large gaps on either side narrowing all the time.  Watch out, or before
long your argument in favor of the price/performance of your RISC machine
may turn against you and you'll end up using C on a 486/66, or something
similar.

I still use a 3650 regularly, and find it to be superior to anything for
certain development activities.  Not for 3D image processing, however!
What I'd like to see is the blending of the best of all possible worlds -
better user friendliness than a Mac, greater multiplicity of software and
hardware add-ons than a PC, even tighter integration, more powerful tools
and more open information access than Genera, and more speed than a RISC
workstation.  There are other things, but you get the picture.

Why should we waste time debating he relative merits (or rather, demerits)
of our various computer platforms?  To quote Scott Warren, "Computers don't
work."  If they did, we might be pretty much out of jobs ourselves.  So,
given that none of our computers are perfect, let's try to promote their
improvement, rather than indulging in divisive, unproductive argumentation.

    ....Symbolics made one fundamental mistake. They ignored performance....

This is silly.  Performance was not ignored, it was recognized as a battle
which Symbolics could not win, and therefore the performance battle against
conventional hardware was conciously avoided.  The mistake was not moving
their software to other, faster, non-proprietary hardware sooner.

    .... If they would have
    stuck to those three marketing goals: highest Lisp performance, highest
    performance/price ratio, and best LISP development environment, the areas of
    their unique strength, and did not attempt to reinvent the wheel of file
    systems, tape backup systems, network systems, window systems, etc. and
    instead leveraged off of other peoples work in those areas, they would probably
    still be in business today---and I would probably still proudly be using their
    products.

Not that it matters much, but I think your history here is a bit out of
sequence.  And if Symbolics is now out of business, it's news to me.  

    So what is the upshot of all this. I think that the Lisp community faces a
    much graver problem: not the demise of Symbolics but the demise of CommonLisp....

Possibly, but I doubt it.  Your crystal ball may be better than mine, but
Lisp's popularity has already outlived that of most of its contemporaries,
and from here it looks to me like the Lisp community is stable or perhaps
growing.  It isn't booming like it did for a brief period, but that's just
as well.  A slow, steady growth is safer and longer lasting.