Subject: Re: What does "Free Software Winter" mean and why should we worry?
From: craig@jcb-sc.com
Date: 2 Apr 1999 16:31:02 -0000

>But the emphasis on "there's a gagillion zero-cost developers out
>there just waiting to help you out" gets the emphasis backwards, and
>is inherently dangerous for us.

Exactly.  The fact that forces similar to Darwinian evolution seem to
act more quickly (and perhaps more effectively) on free software does
not change two other facts:

  -  It is better (by most of mankind's standards) to use cognition to
     avoid being part of the evolutionary process (that is, to
     avoid going extinct, or nearly so).

  -  Most people willing to use technology will try and avoid getting
     involved in areas clearly dominated by evolutionary processes
     (which they'll define as "too many unpredictable failures")
     just as they try to avoid unpredictable environments in other
     areas of their life.

If we emphasize the importance of up-front, solid engineering,
and other cognitive elements of R&D (&M, maintenance, &S, support),
we put free software in a market shared by most other prominent
forms of software, where it can compete for the attentions of
those who prize cognitive analysis.

We can *also* point out the evolutionary processes at work, and
the fact that, to the extent those cognitive efforts *fail*,
free software is more rapidly "judged" by the clammy hands of
those evolutionary processes than proprietary software, which
is usually better for end users (since they're a big part of
what drives those evolutionary processes), sometimes better
for developers.

However, to focus *primarily* on the evolutionary processes puts
free-software development at a substantial apparent disadvantage to
proprietary software, in that this approach makes it seem like,
no matter *how* well-thought-out and well-engineered a product
is, the free-software market might decide it prefers dreck, and
force that product to *become* dreck...when that's not necessarily
the case (mainly because the people in charge of accepting pathes
ultimately prefer cognitive analysis over evolutionary processes --
Linux and EGCS are examples of this preference).

(Focusing on evolutionary processes also, as explained by others already,
makes it seem like no matter how poorly-thought-out and/or shoddily
implemented a product is, the free-software market will magically
transform it into a good one.  In nature, what really happens is that
scavengers eat the carcasses of failed experiments.  On the Internet,
what really happens is that the free-software market doesn't even
glance at the carcass as it drives by at 100kph, so it just sits
there and rots, smelling worse to the people further back in The Wave.)

        tq vm, (burley)

P.S. I know there's some slipperiness in distinguishing "cognitive
processes" from "evolutionary processes" as per the above.

Basically, the distinction is this: when *you* think you know how
something should work, or agree that you might have been wrong earlier
and have Seen The Light, then any activity that reflects agreement
with how you currently think represents, to *you*, a cognitive
process.

Whereas, activities that reflect *disagreement* with how you currently
think represent, to you, evolutionary processes.

Yes, those *other* people are being cognitive, to some degree or another,
and we're all supposedly the result of evolutionary processes, including
our cognition.  The salient point is that if enough people disagree
with you, or, more precisely, if too few people agree with you to help
you realize your goals, you're being thwarted by evolutionary processes
in the sense I use the phrase above.

That doesn't explain what cognition truly is, or what evolution truly
is, of course.  Generally, the more people that employ cognition,
the more likely a given market will produce results that better
optimize the combination of its inhabitants' desires and the resources
of the environment.

In this light, a key advantage of free software is that, even without
the existence of sufficient numbers of truly cognitive thinkers to
give properly designed products *market supremacy*, the few cognitive
thinkers who *do* exist (or, those who agree about how things could
be so much better if only...) have an added channel of communication
available to them -- source code.

And, source code is one of the most precious things in the universe,
for it, as well as books, teachers, and, sometimes, art, are the only
objects known to be capable of changing individual beings from mere
participants in evolution into cognitive thinkers.

(There is substantial evidence that politics and GUIs achieve the
opposite effect.)