[License-discuss] a Free Island Public License?

[Chad Perrin]

On Mon, Dec 12, 2011 at 01:45:28PM -0500, Clark C. Evans wrote:
> I'd love your high-level thoughts on a "Free Island" 
> license or anything that might be similar in nature.

I'll see what I can offer.  I speak for myself, only, in this.

Note that I am not a lawyer, and my comments should not be taken as legal
advice.  What follows is my interpretations alone.


> 
> FREE ISLAND PUBLIC LICENSE (v0.2 on 12 DEC 2011)
> 
> This software is licensed for any purpose excepting
> the right to make publicly available derived works 
> which depend exclusively upon non-free components.

Based on this statement of intent, it seems your intent is to create a
license that disallows distribution of any materials covered by its terms
as part of a larger proprietary work.  This is similar to some part of
the intent of the GPL, and much of the reason that many GPL users choose
that license, but with the added benefit that this statement of intent
suggests you want to allow the licensed materials to be used with other
open source licenses without overriding their license terms.  Is this a
fair assessment?

Of course, I wonder if strong copyleft works would still not be able to
use such a license, because of the fact that they would legally demand
that the work be distributed under the terms of that umbrella license.
It might be pretty hairy legal territory, and I am probably not qualified
to judge this aspect of the license.


> 
> So long as this copyright and license are included 
> in all substantial copies of this work you may:

The phrase "substantial copies" here seems imperfect for what I believe
to be your wishes.  Instead of "substantial copies", you may wish to say
something like "copies in whole or significant part".  I say "something
like" because I'm sure that phrasing could be better refined.


> 
> 1. Publicly copy and use verbatim copies of this
>    work including public distribution and performance.

There are some phrasing details here that make me wonder how it might be
interpreted in court, though I think it is generally clear to the layman.
A copyright lawyer could of course offer more substantial analysis, but I
unfortunately do not know of any copyright lawyers prepared to freely
offer such advice.


> 
> 2. Privately deal with this work in any way you wish,
>    including internal usage, copying, and modification
>    of this work.
> 
> You may also make publicly available via distribution 
> or public performance any Derived Work only if the
> following conditions are met:
> 
> 1. the preferred source code for the Derived Work must
>    be made freely available under this license;

The use of the term "source code" here makes this a software-specific
license which, while fairly typical of licenses suitable to use in a
software context (including notable copyfree licenses, despite their
typically simple language), strikes me as an unnecessary limitation on
the license whose effects when applied to non-software works has not to
my knowledge been tested in court.

This condition is essentially a copyleft licensing condition, of course.
I am sure you are aware of this, but I figured I would point it out, just
in case.


> 
> 2. the Derived Work must pass the Free Island test.
> 
> By "Derived Work" we mean a modified copy or adaptation
> of this work or a separate work such as a plug-in,
> protocol adapter, or wrapper which is designed to have
> intimate interactions with this work's operational
> details, or interfaces.

This seems to make the work strongly copyleft because it implies (or, at
least, I infer) that use of the work as a library dependency in any way
would also qualify something as a "derived work".  This would make this
more-copyleft than the LGPL at least.


> 
> A Derived Work passes the "Free Island" test if it could
> be prepared, modified, compiled, tested, installed, and
> operated in a manner advertised or expected using only
> Commodity Hardware, Free Software, this software, and the
> Derived Work itself.  In particular, the Derived Work
> fails this test if it depends upon proprietary software,
> remote services or hardware to provide features that do
> not have a corresponding Free Software implementation.
> 
> By Free Software we mean any software which is readily
> available to the public without fee and with this
> license, any license approved by the Open Source
> Initiative or any license considered free by the Free
> Software Foundation.

This places substantial power to determine what does or does not pass
your "free island" test in the hands of third parties that are not in any
way answerable to you or the intent of this license.  A better approach,
for purposes of ensuring the independence and satisfaction of intent of
this license, might be to include a list of brief conditions that
encompass the qualities of "Free Software" and open source software you
wish to promote, probably inspired by the FSF/GNU Four Freedoms and the
Open Source Definition.


> 
> By Commodity Hardware we mean a computing device which
> has substitutes in a marketplace and is priced under
> fair, reasonable and non-discriminatory terms (FRAND).

This is very hand-wavy and, I think, might be judged effectively
meaningless in court.  Again, I am not a lawyer, so this is speculation
on my part.


> 
> A safe harbor for passing the Free Island test is if the
> Derived Work is fully usable as intended when compiled &
> installed on a set of networked Debian virtual machines
> using software only from its 'free' distribution, the KVM
> emulator, and no outside network connectivity.  If the
> work is created for the purposes of interfacing with
> proprietary hardware or services, then a sufficiently
> complete emulation of those components must be made
> available as Free Software.

This is even more dependent on third parties who are (presumably) not
accountable to you or the intent of the license, and thus probably a bad
idea for this license.


> 
> THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS "AS IS" AND ANY
> EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
> IMPLIED WARRANTIES OF MERCHANTABILITY, AGAINST INFRINGEMENT, TITLE AND
> FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
> AUTHOR OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
> TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
> PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
> LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
> NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
> SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

This disclaimer text looks fairly standard to me.  No particular comment.


TL;DR Summary:

    My take would be that this satisfies the conditions of the Open
    Source Definition, though I may have overlooked something in my first
    reading.  It appears to also satisfy the conditions of the FSF/GNU
    Four Freedoms and the Debian Free Software Guidelines, but fail to
    satisfy the conditions of the Copyfree Standard Definition.  It
    appears to qualify as a copyleft license, but a somewhat atypical
    example of a copyleft license, in that its proliferation mechanism is
    not tied directly to proliferation of itself.

I hope that helps.