Re: Fwd: That whole "Linux stealing our code" thing

From: Constantine A. Murenin
Date: Sat Sep 01 2007 - 20:53:28 EST


On 01/09/07, Adrian Bunk <bunk@xxxxxxxxxx> wrote:
> On Sat, Sep 01, 2007 at 07:29:39PM -0400, Constantine A. Murenin wrote:
> > On 01/09/07, Adrian Bunk <bunk@xxxxxxxxxx> wrote:
> > > On Sat, Sep 01, 2007 at 05:27:03PM -0400, Constantine A. Murenin wrote:
> > > > On 01/09/07, Adrian Bunk <bunk@xxxxxxxxxx> wrote:
> > > > > On Sat, Sep 01, 2007 at 01:37:18PM -0400, Constantine A. Murenin wrote:
> > > > > > On 01/09/07, Jeff Garzik <jeff@xxxxxxxxxx> wrote:
> > > > > > > Constantine A. Murenin wrote:
> > > > > > > > This will hopefully help diminish certain myths about the code licensing.
> > > > > > >
> > > > > > > What myth? The myth that Theo understands dual licensing?
> > > > > >
> > > > > > Reyk's code was never dual licensed, so it's not like it even matters
> > > > > > to the original dispute.
> > > > >
> > > > > It's no longer dual licenced in the FreeBSD tree because the FreeBSD
> > > > > people removed the GPL choice of the dual licenced code 3 months ago.
> > > >
> > > > FreeBSD doesn't have Reyk's ath(4) HAL, which OpenHAL is based on.
> > > >
> > > > FreeBSD has a driver written by Sam, and a binary-only HAL, also written by Sam.
> > > >
> > > > > So all of Theo's accusations of people breaking the law by making this
> > > > > dual licenced code GPL-only apply as well to the FreeBSD people...
> > > >
> > > > How? FreeBSD doesn't have Reyk's ath(4) HAL from OpenBSD, so there are
> > > > no possible licensing accusations and violations.
> > >
> > > OK, I begin to understand this, there seem to be three different types
> > > of files changed by Jiri's patch:
> > > 1. dual licenced files planned to make GPL-only
> > > 2. previously dual licenced files with a too recent version used planned
> > > to make GPL-only
> > > 3. never dual licenced files planned to make GPL-only
> > >
> > > For files under 1. and 2. Reyk did contribute to dual licenced code
> > > without touching the licence, but I missed that there's also code unter 3.
> > >
> > > So there is a problem, but not with the code under 1. (unless you plan
> > > to change the semantics of the word "alternatively"), the problem is
> > > with some headers under 2. plus the code under 3.
> > >
> > > It's funny how Theo missed the part of Jiri's patch that actually is a
> > > copyright violation and instead complains about the part that is OK...
> >
> > I'm not sure how you conclude that Theo missed the relevant parts --
> > there were many messages posted to misc@xxxxxxxxxxx mailing list and
> > to The OpenBSD Journal in the last few days, and to me it appears as
> > all of the problems were discussed ad nauseam.
> >...
>
> Then it's your fault that you forwarded the wrong email - in the email
> you forwarded the only action for which Theo accused the Linux
> developers of breaking the law was for choosing one licence when using
> dual licenced code.

For the sake of the discussion, at the time I forwarded the message,
the obvious licensing problems (e.g. the original infamous patch by
Jiri) were already addressed.

Personally, these problems were so obvious -- entirely changing the
licence under Reyk's Copyright notice -- that I didn't even take them
for real when they first came across.

BTW, I've now once again re-read the original message that I've
forwarded, and it does contain Theo's reiteration of his response that
the original re-licensing patch had clear violations. E.g. re-read
this part of his message:

- If you receive ISC or BSD licensed code, you may not delete the
license. Same principle, since the notice says so. It's the law.
Really.

> > After the obvious copyright violations were addressed, I think the
> > problem started being an ethical one.
> >
> > As a free software user and developer, the question I have is how come
> > the Linux community feels that they can take the BSD code that was
> > reverse-engineered at OpenBSD, and put a more restrictive licence onto
> > it, such that there will be no possibility of the changes going back
> > to OpenBSD, given that the main work on the code has happened at
> > OpenBSD? (Obviously, such a scenario it is permitted by the licence,
> > but my question is an ethical one -- after all, most components of
> > OpenHAL were specifically based on the OpenBSD's ath(4) HAL code.)
> >
> > You can see that Christoph Hellwig agrees with this ethical problem,
> > as in the message below.
>
> Is it a legal problem or is it "only" an ethical problem?

I don't particularly like to repeat myself -- after the obvious
licensing issues were addressed, it has indeed become an ethical
problem: why do you think that you as the Linux community has to act
ruder to the *BSD community than the supposed corporations that we
always hear about in the BSD/GPL licensing arguments?

I really like the response that Bob Beck gave on this question:
http://lkml.org/lkml/2007/9/1/197

> If choosing one licence when using dual licenced code is not a legal
> problem then Theo repeatedly talking about it would "break the law" in
> the email you forwarded was very unethical and the worst he could do
> for his cause.

My understanding is that with dual-licensed code, you choose to comply
with all of the terms of either licence. However, you cannot simply
remove either of these licences from the code, unless you specifically
receive such right from the copyright holder (remember, with the
copyright law, unless the rights are specifically given, they are
retained). This is what Theo was trying to educate the community on. I
don't see anything unethical in explaining the legal issues.

Constantine.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/