[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Desert island test



On Fri, Feb 29, 2008 at 12:09:33PM -0800, Sean Kellogg wrote:
> The provision that I must post changes does not restrict ones ability to sell 
> or give away the software, it simply imposes a constraint. This constraint is 
> in no way different than the constrain imposed by the GPL that source code 
> must accompany the binary. Allow me to propose my own convenient test, which 
> I refer to as the "Bloody Murderer Test":

What are the definitions you're using here to distinguish between a
"restriction" and a "constraint"?

I'm inclined to regard the GPL's requirement for source redistribution as a
restriction, as well.  Qualitatively, the only line I can draw between the
GPL requirement, and other requirements that are considered non-free (all
the way down the spectrum), is that the GPL's requirement is a minimal,
necessary evil for encouraging the growth of Free Software.

Well, that and the fact that the GPL is explicitly named "free" in the DFSG.
I don't think we should start claiming that the GPL is non-free, but
unfortunately I don't see much of a bright line separating the GPL from
licenses with further restrictions.  Yet historically, folks in Debian were
agreed that the GPL was a free license, and postcardware licenses were not.

> While walking down the street, you are accosted by a a deranged lunatic 
> hell-bent on the destruction of the Free Software Foundation with particular 
> emphasis on undermining the GPL. He tells you that if you distribute any code 
> licensed under the GPL with the corresponding source code, he will hunt you 
> down and kill you in cold blood.

> If we follow the logic of the Desert Island test (or the even more fun 
> Dissident test), we plainly see that the GPL fails the Bloody Murderer Test. 
> Or, we can say, the license isn't what is dictating the distributability 
> (probably not an actual word...), but rather, it is the individuals situation 
> that is doing the dictating. I, for one, don't believe debian should be in 
> the business of ensuring every license covers every possible scenario a 
> debian user might possibly, some day, find themselves in.

The dissident and desert island tests are proxies for real circumstances
that affect millions or billions of people in the world.  Those of us who
work on Debian (predominantly affluent and living in North America, Europe,
or the Pacific Rim) may all have wireless antennas in our skulls, but
high-speed Internet access is far from ubiquitous.  There are plenty of
parts of Africa and Asia where LANs are possible, but Internet access is
non-existent.  Requirements to host changes on a public webserver, or to
send them to a central location, are insurmountable obstacles in these
circumstances; people should not have to choose between license compliance
and being able to modify the code in Debian for their needs as a consequence
of where they were born.

Anyway, applying the same reasoning as was used in the Dissident test, the
GPL still passes the Bloody Murderer test because the GPL doesn't require
that you disclose to Brett Glass^W^W the lunatic that you're distributing
code under that license.  I don't think Debian is responsible for making
sure that a license permits the user to exercise his rights in the face of
insane local "authorities"; but I do think we're responsible for ensuring
that licenses we bless as "free" aren't discriminatory based on geography,
income, status of diplomatic relations with the United States, etc.

-- 
Steve Langasek                   Give me a lever long enough and a Free OS
Debian Developer                   to set it on, and I can move the world.
Ubuntu Developer                                    http://www.debian.org/
slangasek@ubuntu.com                                     vorlon@debian.org


Reply to: