Subject: Re: How complexity in software helps (Was: Re: open source definition)
From: Craig Burley <burley@gnu.org>
Date: Mon, 27 Apr 1998 09:59:21 -0400 (EDT)

>>>>>> "CB" == Craig Burley <burley@gnu.org> writes:
>
>   CB>   -  Art is generally more appreciated the simpler it is, because
>   CB>      it's designed to meet the expectations of individual human
>   CB>      beings.  Software is generally more appreciated the more
>   CB>      complicated it is, because it's designed to convince as many
>   CB>      human beings as possible that they could not possibly have
>   CB>      participated constructively in its creation.
>
>I would give a different reason for why complex software can be appreciated.
>The complexity if architected properly greatly reduces the complexity of
>using the software.  People can then appreciate how they could not do
>the work efficiently without support from the software.

Oh, I agree; I hope you gathered that my point was somewhat tongue-
in-cheek above (see the sentence at the end of the list of points
in the original email).

I believe having a proper interface between the human end user and
the software means have an interface that includes a fair amount
of artistic consideration and, ideally (though it's not the case
for the interfaces I'm aware of today, but is for the ones I've
been designing in my head), that *supports* creative and artistic
expression *by* the end user in a fashion relating somewhat
usefully to the underlying productive reason for using the
software.

And, lo and behold, people who *really* care about that stuff have,
at least in the past, chosen the Mac OS over the Wintel offerings,
when making the relevant purchase decisions.  (But many more people
care more about "does it run all the software I'm likely to want
to use?  do lots of people support it?", and so on, so no surprise
where the money goes.)

>This also supports the notion that good software will continue to be
>expensive to produce relative to its reproduction and distribution costs.
>This leads me to believe that the needs for large infusions of up-front
>capital in the many software industries will continue to grow.  Drug
>companies are a good model from which trends could be derived.

I know so little about drug companies that I can't include the
last sentence, but otherwise I agree completely.  That's why
I still see the problem of making money by writing GPL'ed software
as a largely unsolved puzzle, even though I'm pretty hard to
convince that, somehow, in the roughly 20-year history of the
dominance of proprietary software, every possible avenue for
making money writing GPL'ed (or similarly free) software has
been tried and has failed.  (Specifically, since capital is
strongly attracted to writing proprietary software, as you
essentially point out, it's harder to find ways to get enough
to pay people from the same talent pool to help create GPL'ed
software, etc.)

        tq vm, (burley)