Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 3 Sep 2003 11:03:03 -0400
From:      Rahul Siddharthan <rsidd@online.fr>
To:        Terry Lambert <tlambert2@mindspring.com>
Cc:        chat@freebsd.org
Subject:   Re: Ugly Huge BSD Monster
Message-ID:  <20030903150303.GA2028@online.fr>
In-Reply-To: <3F55A19C.73A1614D@mindspring.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Terry Lambert wrote:
> Pretty much for any GPL'ed code example of a live company that
> you can come up with, I will be able to point out that they are
> singular in their market place because of the marginal costs of
> developement, and the inability to amortize costs of reaching the
> status quo with their competitors.

To get back to the original example: Qt?  It's been entirely developed
by Troll Tech, at certainly not "marginal" cost, and though they started
off with a not-quite-free licence, they're doing fine with a
GPL+commercial dual-licensing system now.  They have a competitor, Gtk+,
which is LGPL'd and thus can be used in commercial products without
"infection" or licensing costs, but many people find Qt superior and --
even though the GPL version is available -- are willing to pay to avoid
GPL "infection".  I'd argue that the GPL and the KDE project have helped
them by giving them and their toolkit wide exposure which they'd never
have got as a purely commercial product, and at the same time preventing
free use of their toolkit by commercial competitors.  Even so, they
switched to the GPL only when they were convinced it would not hurt
their business, and they have stated since that they are happy with that
move.

An older example, as I said, is ghostscript, whose cost isn't "marginal"
either; but the comparison is not quite the same, since they released
GPL versions a year after their commercial/AFPL versions.  Nevertheless,
it is a successful model using the GPL.

A third example is StarOffice.  Many people are buying it from Sun
rather than downloading OpenOffice, for support and additional features.
I'm not sure how much money it makes, but its problem is competition
from Microsoft Office, not OpenOffice; in fact it is doing better than
any other office suite in the last several years.

> The GCC is a good case study in this regard, as well: not as an
> "exception to prove the rule", but because the marginal costs of
> making contributions to the FSF's version of GCC were very high,
> compared to the costs of contributing to EGCS (assignments of
> rights to sign, paperwork, employer approvals).  It was a *lot*
> easier to contribute code to EGCS (via Cygnus), than it was to
> contribute code to GCC (via the FSF):

The EGCS team assigned all copyrights to the FSF, just as with all other
official GNU projects.  The reason it was easier to contribute was that
they used a more open development model and exposed their CVS tree to
the world, which the original GCC people did not do, and had open
mailing lists where user problems and patches could be discussed.  Eric
Raymond pointed out long ago that, in his terminology, most GNU projects
originally followed the "cathedral" model, rather than the "bazaar"
model of the linux kernel and the BSD's; the EGCS experiment was a
demonstration of the effectiveness of the "bazaar" model.  The GPL has
nothing to do with this.

- Rahul



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030903150303.GA2028>