[License-discuss] proposal to revise and slightly reorganize the OSI licensing pages

[Rick Moen]

Quoting Chris Travers (chris at metatrontech.com):

> Therefore, I think that lawsuits will more likely occur regarding the
> scope of requirements of the license than excusing one from the
> requirements directly....

So, basically, 'Yes, this licence articulates the licensor's
requirements, and he/she has a right to insist on them, but what do they
mean?'

Sure, anyone with standing and a bank account to suck dry can litigate
that.  Always.  About anything.


> So for example, suppose an OpenBSD developer sues a Linux developer
> for copyvio due to purporting to sublicense BSD code under the GPL
> with only de minimis modifications.

I keep hearing a limited group of people speaking of this alleged tort 
('purporting to sublicense'), but fail to find it in copyright law.
Reuse of 2-clause BSD code that satisfies its two requirements
(retention of copyright notice in source code, reproduction of copyright
notice and warranty disclaimer in documentation and other materials
provided with binaries) complies with the licence.

Your phrase 'purporting to sublicense BSD code under the GPL' is a bit
vague as to compliance or not, but the reuse scenario as I _think_ I
understand it does not appear to transgress either of the two
requirements.

I think people asking the question you ask, to be blunt, are a bit
unclear about what copyright law does and does not address.  In short:
'Mu.'  

Just because you don't like what someone else does, doen't 
make it unlawful.  FWIW, I think failing to very clearly state that
a reused upstream work is available under permissive licensing would be 
extremely tacky at best, and ungrateful.  However, come to think of it,
if the reuser creating and distributing a derivative work _does_ comply
with BSD clauses 1 and 2, aren't recipients fully enabled to seek out
the upstream instance and enjoy all permissive rights?  Then, what's the
alleged issue, again, please?[1]

> Similarly if Nusphere had persevered and prevailed against MySQL,
> would we count that as unsuccessful enforcement of the GPL?

More like complete and total failure of some court to understand
copyright law, since NuSphere's theory of law was hopelessly dumb.

Let's see:

MySQL AB brought a copyright infringement action against NuSphere,
alleging that NuSpheres's proprietary products NuSphere MySQL Advantage
and NuSphere MySQL Pro Advantage based on NuSphere's database component
Gemini Storage Engine were unauthorised derivative works of MySQL AB's
copyrighted property MySQL (which was statically compiled into it),
because MySQL AB asserted that NuSphere had not provided matching source
code as required by MySQL's terms of use for derivatives.  (There were a
bunch of other noisy causes of action I'll disregard, involving tortious
interference, trademark torts, tra la.)

NuSphere responded that (1) we released Gemini Storage Engine source
code the day before yesterday (I'm not kidding; they really said that),
and (2) we contend that NuSphere MySQL Advantage and NuSphere MySQL Pro
Advantage are aggregates of our code and MySQL's and are simply _not_
derivatives of MySQL AB's work, as that term is defined in copyright
law.

The parties settled on MySQL AB's terms a short while later, and so
NuSphere's novel and gutsy theory about the definition of 'derivative
work' never had a chance to make a judge laugh, but I'll just sum up to
say that I'll comment on what your hypothetical _is_ when and if it
happens.


[1] This dumb shibboleth about 'sublicensing BSD code' came up often
enough that Ernie Prabhakar at OSI nicely FAQed it, back in 2008.
http://ideas.opensource.org/wiki/help/license