Subject: Re: Licenses vs. public domain
From: kmself@ix.netcom.com
Date: Sat, 27 May 2000 19:44:42 -0700
Sat, 27 May 2000 19:44:42 -0700
On Sat, May 27, 2000 at 10:38:05AM -0600, Kevin S. Van Horn wrote:
> Here's a question for the group: What do you think of not having *any* license
> for a libre software project, and just putting it into the public domain?

The FSF covers this issue in their description of categories of free
software:

    http://www.fsf.org/philosophy/categories.html


The key distinction is this:

 - To put a work in the public domain is to relenquish all ownership
   claims to it.  You are then unable to state required procedures or
   terms of use, propretary, free, open source, or otherwise.
   
 - Licensing, under any of its various guises, reserves the author's (or
   copyright holder's) right to control use or terms of use of a work.

I like to employ the (reasonably accurate IMO) analog of real property
(real estate) and public parks.  A public park is open to all.  The
ideal land-use rules for a public park are limited to those which permit
enjoyment, use, or utilization of the land without denying the same
benefit to others.  The point being, it's possible to utilize real
property rights and land use regulation to provide a shared benefit --
you don't have to go and outlaw private property to have parks.


> I ask because several of the popular existing licenses are barely
> distinguishable from public domain.  

Which?  Artistic comes to mind.  None others off the top of my head.

> Yet I am aware of only one libre software
> project (the ISIP speech recognition system) that is actually public domain.

There are others, which have been listed.

> It seems to me that if your goal is to get the widest use of your code as
> possible -- especially if a large part of your code is organized as libraries
> -- that public domain has some advantages.  Users don't have to worry about
> license incompatibilities when mixing it with other libre software, nor do
> they have to puzzle out the legal implications of using your code.
> 
> One might worry that people will keep enhancements to the code proprietary,

This is, IIRC, Stallman's justification for writing the GPL.

> and so a public-domain project would not get the level of contributions that a
> GPL'ed project would get.  However, here are some considerations that argue
> against this being a problem:
> 
> - A larger user base may compensate for a smaller percentage of users
>   contributing back improvements.  A lot of businesses simply won't use your
>   code if they can't build something proprietary on top of it.

Eben Moglen talks and writes fairly persuasively about this line of
reasoning.  Essentially, his argument is that if you follow the
licensing line out far enough, you need something like copyleft.  He
promises me he's writing a book on the topic, so you can needle him
about this.

Note, too, that all license compatibility issues would go away if
everyone used the same license <g>.  In practical terms, limiting use to
the GPL, LGPL, BSD/MIT, and MozPL license (with a dual to [L]GPL under
Exhibit A) would provide for a range of licensing options, all of which
would then be compatible with the least-compatible license, the GPL.

> - Code drift.  A friend of mine works for a company that has made proprietary
>   extensions to one of the BSDs, for internal use.  This means that every time
>   there is a new release of the OS, they have to fold their own extensions
>   back in, again and again.  This is getting more and more difficult as time
>   goes on, and the code bases diverge.  They could have saved themselves a lot
>   of trouble by simply contributing their improvements.

This is called the Stupid Tax.  Attributions to origin vary, I've heard
it given to Cygnus, RedHat, and others.  Hmm...  An Alta Vista search
turns up....

  - This, which looks vaguely familiar:
    http://www.egroups.com/message/cni-copyright/11066 (I *hate*
    discovering I'm the authoritative reference on something).

...I *thought* the concept was covered either in  Open Sources  or
 Under the Radar , but it isn't indexed in the second and I'm
recompiling vgrep.  I've lent my copy of the first out....

> - Successful near-public-domain projects.  The Apache license isn't all that
>   far from simply being public domain, yet the Apache project hasn't suffered
>   from lack of contributions.

Apache is BSD.  It's an "attribution required" license.  Essentially
serves to keep people aware of who wrote the source.  PD needn't do
this.

> What do the rest of you think?

As I wrote here recently, I'm coming of the opinion that:

 - GPL is the One True License, but I'm content to wait for the rest fo
   the world to figure this out.  And might be willing to make minor
   exceptions.  <ob:  tounge-in-cheek mode>

 - The existence of strong copyleft licenses such as the GPL gives
   something for less strict licences to, essentially, "lean on".  The
   knowledge that there will be a body of free software, an operating
   system, development tools, and, importantly, developers, makes the
   sustainability of software licensed under free, but not copyleft,
   licenses such as MIT, BSD, and Artistic, easier.  It's a symbiotic
   effect, IMO.  I'd still favor strong copyleft where possible, but
   don't see immediate harm in other licenses, and quite possible some
   good.

 - Software licensing addresses only part of the concern of free
   software, and specifically doesn't address protocol standardization
   or other compliance certification, which should be handled through
   other means such as standards organizations and applicable trade or
   certification marks.

 - Chocolate is a perfectly wonderful ice cream flavor.  Unrelated, but
   it's what I think.

-- 
Karsten M. Self <kmself@ix.netcom.com>         http://www.netcom.com/~kmself
  Evangelist, Opensales, Inc.                       http://www.opensales.org
   What part of "Gestalt" don't you understand?      Debian GNU/Linux rocks!
     http://gestalt-system.sourceforge.net/      K5: http://www.kuro5hin.org
GPG fingerprint: F932 8B25 5FDD 2528 D595  DC61 3847 889F 55F2 B9B0


["application/pgp-signature" not shown]