Re: Questions

[Brian Behlendorf <brian collab net>]

On Wed, 27 Nov 2002, Linas Vepstas wrote:
> cc'ing the foundation-list

Well, you snipped out a lot of context to my words, so I'm not sure how
useful that was; I also replied privately because I did *not* want to get
into a debate about licensing, I am not a Gnome contributor (though a
daily user) and I have no interest seeing Gnome change its licensing
policy.  I took it private when Linas asked how many new projects were
non-GPL, and I thought I'd give a private perspective.

> Yikes, I did not mean to slander ASF.  I guess I was thinking "small
> projects" rather than big ones.   Certainly, the big projects each have
> a different "taste" to them; its hard to distinguish the end-user from
> the license.  Gnome "tastes" like lots of small projects with
> predominantly individual users.  ASF "tastes" like a few mega-projects,
> whose users are corporations running server farms.  Are these differences
> due to the type of the project, or the type of the license?  ASF
> developers "taste" older, more experienced, more imbued with corporate
> culture w/standards-body interaction.  Whereas a gnome developer
> wouldn't recognize a standards body even after sitting through
> a 3-day long meeting, (assuming incorrectly that they could sit that long).

I think both members of the ASF and Gnome would debate your
characterizations of them.  We have a different area of focus, but I don't
know right now if we've got a difference in people-paid-to-hack vs.
pure-volunteers, or not.  Ultimately it shouldn't really matter - both
projects still value the quality of the code and the willingness of the
contributors to work in a community.

> I run a small, public-domain 3D tube-drawing library.  In ten years, it
> has had zero major contributors (aside myself).  I beleive its in a
> number of commercial products, including not a few games; I beleive
> its also inside of MS WinNT.  It has (I beleive) zero acknowledgements
> in the commerical products (as there is no license to even demand
> such acknowledgment), and at best a smattering of thank-yous.
>
> I've never felt angry about this, so "rip-off" may be too strong a word.
> But it is a bit disappointing, as I crave recognition.  Is it the
> license/lack thereof that give GLE its flavour? Or is it the user
> community?
>
> By contrast, in all of my GPL'ed projects, I sense that zero of them
> have been incorporated into a commercial product (aside from linux
> distro's).   But the user community seems to be more active, more
> appreciative, and send in more patches.   Is it the license, or
> the types of users?

Or the task to solve?  Perhaps your tube-drawing program is perfect.  :)
Lots of one-person projects start out and remain that way.  The license
does matter when it comes to building communities, of course.  But I
really don't think you can apply your example too broadly.

> This is what I mean by the "gut feel" of the application of a particular
> license to a particular project:  what reward does the developer
> /contributor expect to get?  Recognition? a salary?  What are the
> developers future plans for the code?  Not uncommon seems to be the
> syndrome: "People are going to contribute to my BSD-licensed project,
> and I'm going to be CEO of my own company that sells the proprietary
> version.  Gosh I'm so smart I'm gonna be rich."  And even more common
> seems to be the syndrome: "I am the smartest coder alive.  People
> will bow in awe before my great code.  Its GPL'ed for all the world
> to see, and even great corporatinos will be humbled."  (these are
> extreme examples; but they do shade the picture).

They're extreme, but I'd hardly say they are representative.  Most coders
on Apache, I believe, are there because they use those tools to solve
their own problems, and reuse is enhanced by an active development
community. I don't feel qualified to generalize about developers on GPL
code, but many of the ones I know have similar motives.
 Egos and community-killing moves by anyone looking to make a
profit tend to get called to the mat pretty quickly - we actually just had
our first committer suspension recently when it was clear a particular
developer was not interested in working with the community.

> In reality, few get rich, and the license doesn't seem to be a
> correlator.  The sleepycat & reiserfs & kompany license arguments
> are more subtle, and are driven strongly by thier expected end-user/
> customer expectations.
>
> In reality, the I beleive the decisions that developers make about
> licenses are very subconscious, subliminal.  I don't think they
> really think it through; I beleive that most users of the GPL
> could not withstand cross-examination about why they picked that
> license, other than  a vague pledge of allegience to gnu principles.

I don't know if I'd go with "most", but I'd use the word "many".

	Brian