Subject: Re: the Free Software Movement in Industry
From: Tom Lord <lord@regexps.com>
Date: Thu, 23 Aug 2001 17:18:36 -0700 (PDT)



       There's no "de facto purchasing union".  Telco A puts out an RFP
       saying, "We need a billing system that supports prepaid phone cards."
       If the software company's product doesn't support prepaid phone cards,
       but the management expects that Telcos B, C, and D will soon be asking
       for the same thing in their RFPs, the software company will be willing
       to spend more on implementing prepaid phone cards than they actually
       get from Telco A for the job.

       If later it turns out that Telcos B, C, and D *don't* care about
       prepaid phone cards, or they're not offering enough for the feature to
       recoup the software company's investment, then the software company
       loses.  There's no collaboration between A, B, C, and D -- only a
       gamble by the software company that they have common interests.


I think we agree, other than on terminology.

Using my terminology, what you're saying is that there *is* a de facto
purchasing union, but that its success is probabilistic, not
deterministic, and that its membership is fuzzy, and subject to change
without penalty (to the members).  Investors in the software vendor
are accepting the financial risk of that non-determinism, which is one
reason why wind-falls as prizes make good sense.

That arrangement can still function perfectly well in a Free Software,
GPL world.

What does start to look financially suspect in a Free Software world
is the practice of creating software companies that are
one-hit-wonders rather than long term engineering facilities.  That's
one reason to make the vast all-encompassing transition to a Free
Software world in the replacement software, rather than trying to GPL
all the software that already exists.  The idea of bringing together
some programmers *only* to write a one-shot billing system, then
charge huge amounts of money selling that system, unmodified, over and
over again, without any significant reinvestment in the engineering
process -- is a little weird.  Fortunately, there are some really big
generalist players in the industry-as-a-whole (don't know much about
telco billing software, sorry) who are about as far away from being
on-hit-wonders as you can get.

-t