Re: kernel firmwares: GR proposal
On Sat, Sep 09, 2006 at 01:57:31AM -0400, Nathanael Nerode wrote:
> > The arsenic case was more problematic, since the copyright seems to have
> > landed at broadcom too, but they don't care since they don't sell it
> > anymore,
> Given this, we actually should have a decent chance of getting them to
> license it under a free software license (after all, what do they care
> about it?) provided we can talk to the right people.
Well, the thing is they can't be bothered to make sure they are actually the
right people or something. Notice that even in a friendly case, like the ocaml
library which ended up at HP, and i asked bdale over it, it was not possible
to find the papers back, but then it was easier to reimplement that code.
May guess is that the source code is lost in some obscure case and nobody
cares, but right, if you find the right person ...
> > and they probably are not even aware of the fact that they are
> > actually copyright holders.
> FYI, there is a way to deal with a copyright of unknown ownership.
> Get a license from everyone who *might* have the copyright. Of the
> form "*If* Broadcom holds any copyrights in this code, and we don't know
> whether we do, *then* Broadcom licenses its copyrights under license X.
> Broadcom does not license any copyrights which it does not hold as of
Yes, i was thinking something such.
> Repeat with all other companies which might have some of the rights. Once
> you've gone through all of them, you have a proper license, even though you
> don't know who holds the copyright.
> This is essentially similar to getting "quitclaims" for contested land.
> > I had a similar problem with some ocaml library, which was developed
> > together byt the ocaml team and the digital labs, which ended up at HP,
> > and even asking bdale about it, did not help free that code, which is now
> > lost forever and upstream reimplemented it.
> At least we actually have the source for the acenic code, even though we
> don't have a free license for it.
Euh, no, it is a binary-only firmware blob last i checked, but i may be wrong.
> > I think the quote from bdale
> > was "i think i know in which set of boxes it may possibly be".
> >> I think that throwing Debian's name around with 'offical' status may
> >> be helpful to get responses from some of these companies; I didn't do
> >> that, since I couldn't!
> > Well, assuredly, but i think that another difference may have been the
> > more reasonable and well though-out mail with some legal analysis, and the
> > fact that what we demanded was quite easy for them to do. Also, the
> > timeline was maybe one of more maturity and sensibility on this subject,
> > and we had a rather huge thread on LKML when it happened. From your past
> > posts on the subject, i believe that maybe the wordings you chose where
> > not the best ones, but as i have not seen said mails ...
> Probably not the best words. I was very polite to them, since I really
> figured it was just an honest mistake (which it was), but I probably came
> across as an "unimportant nobody" and they figured it wasn't worth wasting
> their time.
Yep, which is why it is important to find out the list of licensors/copyright
holder of the other problematic cases, and someone, preferably imbued with an
official and announced delegation from our DPL, will contact them on Debian's
behalf. Ideally, we would do it in such a way that it is precedeed with a
slashdot article, and/or other widespread press campaign, or something.