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

Re: Debian Legal summary of the X-Oz License

I'm going to try to be clear about where debian-legal is coming from.
We've gotten a lot more careful about licenses in recent years after being 
burned several times by surprising license interpretations.  And by people 
trying to do odd, non-free things with their copyright licenses (usually 
people trying to create pseudo-trademarks in a copyright license -- I don't 
know why that's so popular).

MJ Ray wrote:
>> Could you please look at the X.org and X-Oz licences again and notice
>> this difference? If you want to mimic the X.org licence, then will you
>> make that clause a notice in the footer instead, please? Copyright
>> licence conditions are the wrong way to police trademarks.
To be clear, what he means by this is that trademarks and copyrights are
quite separate and different things.  We are all for trademark enforcement.
However, putting conditions about trademarks in copyright licenses often
ends up restricting legitimate activity -- and it's never, ever necessary,
since a copyright license simply doesn't grant *any* trademark-related

> Except as contained in this notice, the name of X-Oz Technologies shall
> not be used in advertising or otherwise to promote the sale, use or other
> dealings in this Software without prior written authorization from X-Oz
> Technologies.
We're basically afraid of nasty interpretations of this -- we try to respect
the copyright holders' interpretation, and some have had unexpectedly nasty
interpretations of seemingly nice licenses (ask about the University of
Washington and Pine).

Basically, under the law of pretty much every country we've ever heard of,
these rights are guaranteed already -- nobody can use your name for
promotional purposes without prior written authorization regardless -- so
if it's just reasserting those rights, this clause is meaningless.

However, under a strict interpretation of the license, it could mean that if
someone (illegally) used the name X-Oz Technologies to promote their
variant of the software, they would be violating the copyright license, and
would then lose the right to create modified versions of the software
entirely -- and then the innocent people who got the modified versions
could suddenly have illegal, undistributable programs.  Or, indeed, under a
strict interpretation, if I happened to mention the name X-Oz Technologies
while I was promoting the software, even if not in a promotional manner,
that could be interpreted as violating the copyright license, etc.

Given that the clause is otherwise meaningless, we have to wonder whether
one of these stricter interpretations is intended.  So when a copyright
holder puts this in a license, we have to ask them, "Is this clause
meaningless, or do you mean something stronger than what's guaranteed by
law?"  If they say it's meaningless, that's great and we can stop worrying
about it -- though we ask them to please remove it because of its
meaninglessness.  :-)  If they say it means something stronger, then we
need to know what exactly they mean.

>The end-user documentation included with the redistribution, if any, must
>include the following acknowledgment: 
>"This product includes software developed by X-Oz Technologies
> Alternately, this acknowledgment may appear in the software itself, if and
>wherever such third-party acknowledgments normally appear.

Again, we're worried about nasty interpretations -- we have to ask the
copyright holders.

A strict interpretation might require all kinds of crazy things, like
including the acknowledgement line in documentation I wrote myself for
another program, simply becase it was "end-user documentation" and was
redistributed along with the X-Oz software.  We need reassurance from the
copyright holders.

There was a similar clause in the Apache 1.1 license; however, the Apache
copyright holders made it *very* clear that they intended this to be a
*very* weak requirement, so we concluded that it was free.

It may, however, be GPL-incompatible (i.e. you can't make a program which is
derived from both a GPL-licensed program and an Apache-1.1-licensed
program), because it imposes a very strict, specific condition which is not
in the GPL.  This has been argued about, but Debian had to be conservative
and assumed that it wasn't GPL-compatible.  For Apache that wasn't a
serious problem (although Apache has actually been revising their license
with the stated intent of making it definitely GPL-compatible).

For XFree86 this *is* a problem, because there's a very large number of
GPL-licensed programs linked with XFree86, using XFree86 header files,
etc., which makes them derivative works of XFree86 (at least in the view of
some of the copyright holders of the GPL-licensed programs, whose views we
try to respect).  So new XFree86 with this new clause might not be legally
usable with a lot of programs which are already part of Debian.

So we would want a definite statement from the copyright holder that it was
intended to be GPL-compatible.

We certainly hope that X-Oz intends its license to be Free in the strong sense
that Debian requires, and that it intends it to be GPL-compatible.  
Unfortunately, when a company comes up with a new license, if there is any 
significant ambiguity (as unfortunately there often is), we feel that we have 
to check that the company's interpretation does correspond to the favorable 
interpretation we would like to take -- since it doesn't always.  Hence 
Branden's rather detailed set of questions.

Thanks for your time, and I hope this helps.

Reply to: