Re: RES: What makes software copyrightable anyway?
On 5/19/05, Adam McKenna <adam@flounder.net> wrote:
> > This "absolute protection" did not seem to protect Napster, nor did
> > the home recording act.
>
> Despite their claims to the contrary, Napster's *primary function* was to
> facilitate the illegal distribution of copyrighted materials. That is
> clearly not the case with Debian.
Ok, here's what I'm thinking:
First off, is Quagga + libsnmp + libssl a work which is subject
to copyright protection in and of itself? If not, there's no
problems under copyright law in distributing this work.
I think that this is subject to copyright protection -- this
work has unique characteristics which mean it isn't just some
random collection of bits. Someone had to carefully select
these pieces to make routing software which performs properly.
The work which had to go into that selection process means
that this is a collective work, and is subject to copyright
protection.
Once again, if I'm wrong on this point, there is no need to
worry about the GPL.
Next question: why do we need to worry about the GPL if this
is the case?
The answer is that copyright law (which is the basis for all
non-free software) says that you need a license to distribute
a copyrighted work. In the case of a collective work, you
also need permission on each of the components to distribute
the work as a whole.
So this question can be rephrased: does the GPL grant permission
to distribute this collective work? I think the answer is:
yes it does, but it puts some conditions on that distribution.
M.K.Edwards has argued that these conditions do not apply.
His argument seems to be that because no one has every done any
modifications to this work as a whole, section 0 doesn't apply
(which seems to me to mean that we wouldn't have permission
at all from the GPL) and that the "mere aggregation" clause
does apply (though how it applies when section 0 does not is
beyond me).
M.K.Edwards has also indicated that the Progress v. MySQL case
is precedent which shows that the above don't matter, but since
the judge didn't specifically address any of the conditions
which I'm thinking are important, I'm thinking that MySQL did
not contend that those were violated. More specifically, you'd
need to establish that each of the terms 0 through 6 apply to
the case, and assert that 4 is violated. You might need to
show more than that (for example, being a contract case, you'd
have to show how valuable contributions of the community are,
and how much value you're losing on a section 4 violation).
I'm not a lawyer, and I'm not preparing a legal case here --
I'm only trying to rather broadly characterize the issues.
I'm going to ignore that rat hole for now and continue...
The next question is: "Are we distributing this copyrighted
work?"
I think we are. We're distributing:
The unmodified sources
The modified sources
The binaries without libssl
The binaries with libssl
These are packed efficiently, but by issuing slightly different
commands you can get each of these. The distribution of
any one of these does not mean that we're not distributing
the others. The rather clever mechanisms we're using to
efficiently distribute these don't really matter, either.
The fact that a lot of people worked together on this doesn't
really make a difference. The fact that the first three use
a different set of programs to unpack than the last doesn't
really make a difference -- not when we've been so careful to
make sure that the last works.
The "dynamic linking" issue doesn't seem to matter at all --
the collective work still exists. We've still modified that
collective work so that we can distribute both binaries which
include libssl and binaries which do not. Dynamic linking is
just a way of putting different parts of a work in different
files, which allows some editorial changes on one file without
having to re-create the others.
So, anyways, that's what I think we're doing which is somewhat
analogous to what Napster was doing.
Napster, by the way, was enabling people to make recordings
of music at home. There was law on the books (the home audio
recording act) which seemed to say that these activities were
perfectly legal. Court ruled that they were not.
--
Raul
Reply to: