Subject: Re: commercial / proprietary
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: Fri, 18 Feb 2005 13:52:12 +0900

>>>>> "David" == David N Welton <davidw@dedasys.com> writes:

    David> "Stephen J. Turnbull" <stephen@xemacs.org> writes:
    >> >>>>> "David" == David N Welton <davidw@dedasys.com> writes:
    >> 
    David> Are there examples that come to mind of developers, whether
    David> individuals or corporations, who have released software as
    David> open source in order to maximize their monetary returns on
    David> the time invested in creating the software?

    >> Define "maximize monetary returns."

    David> I'll put it in my terms then:

    David> "Make as much money as possible."  In other words, not
    David> doing it out of the good of their hearts, or to thumb their
    David> nose at some competitor, but because they think it's the
    David> best way to make the most money.

Sorry, I misspoke.  What I meant was "under what constraints?"

Ie, the problem here is "as possible" and "best way".  Is knocking
over a bank acceptable?  Not on this list.  How about lying about
known flaws in the product?  Not on this list.  How about dual
licensing?  Not to everybody on this list, and it's not obvious to me
that you have that in mind.

There's also the problem of "nearly open" source, in all its variety.
Aladdin License, gated communities, etc.  Why do open source companies
not use "nearly open" licenses when they use mutually incompatible
open source licenses?  In theory if people just buy the best product
under the best terms, it seems obvious that licenses tuned to the
product would make more money.  So you could argue that social
pressure is the issue here; we're just catering to the "pinko" market
by going beyond the profit-maximizing price discrimination strategy
(charge commercial redistributors who can afford to pay, let others
distribute at _their_ expense to those who can't afford to pay anyway,
thus saving on lawyer and PR expense :-)[1] to full OSS.

Finally, there's the issue of whether "making money" means "earning
revenue" or "weakening a competitor in his strong market, thus making
it harder for him to compete for your cash cow."  The latter is
arguably what IBM and Sun are up to.  Do _you_ want to include a
company that earns its revenue on proprietary software, while dumping
OSS code in a competitor's market?

I chose X because the companies apparently felt releasing sample
implementations as OSS was the cheapest way to get the standard
developed.

    >> The example that I would say best fits is the X Window System,
    >> and even that is really a dual-license setup.

    David> How so?  I don't know enough about the history of X.  The
    David> notion I have is that it was a big-company consortium kind
    David> of affair.

Exactly, but not only did the companies (and their major contractor,
the Athena project at MIT) cross-license to each other (in theory, it
_could_ have been limited to a sort of copyright pool), they did so to
the public.  Why they chose to do this, I don't know; it may have been
the easiest way to go with NSF etc money mixed in or something like
that (that's pure guesswork).

However, in most cases the example code in the X Consortium
distribution, although it works fine, was substantially augmented by
proprietary code (in particular, in the X servers and in Motif) in the
commercial products.  Thus, dual license.


Footnotes: 
[1]  Of course the price discrimination occurs anyway, but it'