Subject: Re: Successful FSBs
From: "Tim O'Reilly" <tim@oreilly.com>
Date: Mon, 04 Nov 2002 17:36:30 -0800

On 11/4/02 9:30 AM, "Benjamin J. Tilly" <ben_tilly@operamail.com> wrote:

> 
> If that was true, then you would have expected to see
> more successes with "almost open" licenses like Sun's
> Community Source License.  But they have uniformly
> failed abysmally.  My belief is that the blatant power
> grab kills any community spirit before it has a chance
> to get started.  And without that spirit you don't get
> the dynamics that can make free software work.

Yes, but isn't this social axis relatively independent of the license? I've
always liked to point to the success of the early Unix development
community, which worked under an "open but proprietary" model.  The social
dynamics were more important than the license, which became significant only
once ATT decided to enforce the rights they had reserved to themselves.  And
even then, what happened?  Proprietary unix stagnated in many ways, and the
community rebuilt itself around BSD and then Linux.  The open architecture
and the community's exposure to open code made it possible for the community
to eventually route around the broken license.

ATT's folly of course supports the GNU position that we need a protective
license because companies that have proprietary but open licenses can have a
change of heart.  But it also does underline the fact that the f/oss license
isn't essential.  It's a nice to have, not a must have.  Open code is a must
have.

As in my favorite Lao Tzu quote:

Losing the way of life, men rely on goodness.
Losing goodness, they rely on laws.

The "way of life" in this context is the law of information dissemination,
that openness has better survival characteristics for information than
closedness.  Goodness is the virtuous circle of a gift culture.  Laws are
the licenses that we rely on when we don't trust that people understand the
ways that open source works, or that they have our best interests at heart.

I'll also point out that the "failure" of software under the SCSL is hardly
proven, given that Java has a developer community whose size exceeds that of
any open source programming language.  And yes, they can't all contribute at
the level that they might in an OSS project, but there is a rich open source
layer on top of the Java core, just as there is with Perl or Python.

You also need to consider the relatively small number of projects under the
license.  There are thousands of projects under the GPL or BSD style
licenses; most of them are largely unknown and have very small development
communities.  It's easy to argue from the success of a few but you can't
draw too many conclusions.

-- 
Tim O'Reilly @ O'Reilly & Associates, Inc.
1005 Gravenstein Highway North, Sebastopol, CA 95472
1-707-829-0515 http://www.oreilly.com, http://tim.oreilly.com