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

Re: Apple Public Source License and more



Hello,

On Tue, Nov 05, 2002 at 08:22:05AM +0900, Jens Schmalzing wrote:

> 2. Apple Public Source License

>    A number of drivers are adapted from Darwin and were released under
>    the Apple Public Source License (APSL).  More specifically, version
>    1.2 applies, which to my knowledge was certified as free, but a few
>    posts on this list express doubts.  What is the current situation
>    on this?

I'm not sure why you have the impression that this license had been
"certified" as free.  If you mean "certified" in the sense of a formal
certification, the only body I know of which certifies licenses in such a
manner is the OSI, which uses the term "Open Source" -- not "free".  The
FSF (champions of the term "Free" :), for its part, states that the APSL
is not free.  I have read the license in the past in connection with my
employer's rollout of quicktime streaming server, also made available
under this license; and at the time my conclusion was also that the
license is non-free (and not DFSG-compliant).  I would have to look over
the license again to give you a specific citation; perhaps another
debian-legaler can be more specific.

> 3. no clear license

>    Some drivers are based on example code that Apple supplies as a
>    starting point for driver development.  This was before the APSL,
>    so Samuel believes it is possible to place them under a free
>    license.

If "based on example code" means that the drivers contain substantial
portions of the example code, either verbatim or in recognizable form,
the driver is a derivative work of the original code and we must have the
express consent of that code's copyright holder (Apple) before it can be
considered free.  Precisely how much of the original code must be
included before this is a problem is a question of paralegal mathemagic,
the answer to which may vary according to jurisdiction.  Note that if
there is no clear license on this code, upstream may also be violating
copyright law by distributing these drivers at all, and should consult
legal counsel if they have any doubts about this.

> | All the binaries above run within the MOL session and they are
> | coupled to MOL and the linux side through a syscall interface.
> | Running non-GPL code in this manner might slightly violate the mol
> | GPL license.

> Is this correct?  What modifications to the mol license would be
> necessary to keep mol DFSG-free while allowing for non-free drivers to
> run?

[As an aside, the DFSG-freeness of a license can always be determined per
se, without consideration for the nature of the licensed code or its
dependencies.  If a license is free, code distributed under that license
belongs in one of three places: the main archive, the contrib archive if
it depends on non-free code, or outside of the archive entirely if it
cannot be distributed because of a license conflict.  There is no
combination of circumstances that would require code distributed under a
DFSG-compliant license to be uploaded to the non-free archive.]

Run-time coupling of GPL and non-GPL code is never prohibited by the GPL;
it's only when you distribute such code together[1] that you run into
license issues.  You therefore cannot be held in violation of the GPL for
distribution that you are not a party to, and by definition, anything
that can legitimately be included in Debian's main archive does not
require the non-free code in question, so there should not be a problem
here.  As long as non-free code is not included, nor required for the
software to be useable, it should be ok to upload to main.

If there are still doubts, upstream may wish to clarify their license in
the same way that Linus has for the Linux kernel, making it clear that
running non-free applications on top of mol is considered acceptable.

Regards,
Steve Langasek
postmodern programmer

[1] Or separately, if the GPL binary depends on the GPL-incompatible code

Attachment: pgppHKwaZu5Cf.pgp
Description: PGP signature


Reply to: