Slashdot Mirror
Linux Wireless Driver Violates BSD License?
bsdphx writes "After years of encouragement from the OpenBSD community for others to use Reyk Floeter's free Atheros wireless driver, it seems that the Linux world is finally listening. Unfortunately, they seem to think that they can strip the BSD license right out of it."
It appears that someone's submitted a patch to the LKML that wrongly strips the BSD atheros driver of its license - a clear violation of copyright.
However, until it's in Linus's tree (or even the MM tree), the violation is not by "linux", but the contributor, Jiri Slaby.
Anyway, thanks to the OpenBSD team for these great drivers. Thanks to the Linux team for including them (under the correct license).
There are shills on slashdot. Apparently, I'm one of them.
The fact is the original patch post was on Tue, 28 Aug 2007 12:00:50. Since then, the discussions are ongoing as how best to proceed. Recently, this was posted:
...
Date Wed, 29 Aug 2007 08:35:05 -0200
From "Jiri Slaby"
Subject Re: [PATCH 4/5] Net: ath5k, license is GPLv2
On 8/29/07, Johannes Berg wrote:
> On Tue, 2007-08-28 at 12:00 -0400, Jiri Slaby wrote:
>
> > The files are available only under GPLv2 since now.
>
> Since the BSD people are already getting upset about (for various
> reasons among which seem to be a clear non-understanding) I'd suggest
> changing it to:
yes, please. Can somebody do it, I'm away from my box.
> + * Parts of this file were originally licenced under the BSD licence:
> + *
> > * Permission to use, copy, modify, and distribute this software for any
> > * purpose with or without fee is hereby granted, provided that the above
> > * copyright notice and this permission notice appear in all copies.
> > *
> > * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL
> WARRANTIES
> > * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
> > * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
> > * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
> > * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
> > * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
> > * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
> + *
> + * Further changes to this file since the moment this notice was extended
> + * are now distributed under the terms of the GPL version two as published
> + * by the Free Software Foundation
>
> johannes
>
As mentioned before, it is the LKML, not the Rosetta stone. Things change
The little guy just ain't getting it, is he?
Basically, you are allowed to link GPL'd code to BSD code. So if I wrote "The boy hit the baseball" under the BSD license and you alter it to "The large boy hit the baseball well" under the GPL, the original statement is still available for use under the BSD license - even in your second statement. As long as they remove your GPL'd addition (the intertwined words "large" and "well"), they are free to use it under the BSD's terms.
The practical point is that the BSD code, when linked with GPL code, must adhere to the restrictions of both licenses. Most people just say that it has been relicensed under the GPL. That isn't exactly true. From most practical standpoints, the BSD license has so few restrictions that it doesn't matter, but technically that BSD code is still under the BSD license and it's requirements must be met.
So, that BSD code can easily be linked and intertwined with GPL code, but those few requirements of the BSD license must be met so long as there is any BSD code in the GPL'd derivative work.
The original BSD vaersion is still out there.
Just like it would if the code was taken properitary.It's a bit rich to deny people to keep their own changes proprietary, wouldn't you say?
Except that's what the GPL tries to do. It's removing freedom.
And that's what many of us BSDers are against. We want our software to keep freedom. Including the freedom of future developers to keep their own changes private, or get paid for them. Thereby, we also allow the end users the freedom to buy those changes - a freedom they wouldn't have if the code was GPLed, because the incentive to make the changes wouldn't be there.
As an example, we have Apples operating system, partially made on code I wrote. And I'm a very happy user of it, even though I (or rather, my employer) had to pay for the extra stuff Apple has added. The ability to do so is a freedom I have partially gotten from having released my software under the BSD license.
Eivind.
Doubting the existence of evolution is like doubting the existence of China: It just shows that you're uninformed.
The BSD and GPL communities are two groups with much in common and some differences. That's especially true of the OpenBSD people who are very much committed to freedom for their software (and just disagree with the GPL people about how that is best defined). This kind of article is about trying to set up atificial disagreements between those communities so that they don't cooperate well. I write copyleft software and content (GPL/GFDL/CC-SA) but I would mostly relicence it if an important project like OpenBSD or X.org asked for it. I would make that decision based on the value of the project. If I feel that a project is harmful overall then I probably wouldn't.
The trick is that we have to not be divided and work together sensibly.
OpenBSD developers have time and time again claimed "moral superiority" over GNU and Linux due to their adoption of a license that allows code to be used in closed projects. It always was a specious argument, but it's looking all the more absurd today.
Claiming any particular licence (BSD/GPL) to be superior is like asserting that cars are superior to helicopters. In far too many cases, the licences are dragged in to try to justify a bad argument, and the fault lies with both camps. GNU and BSD zealots alike adopt Talibanesque positions that do nothing but harm to the community.
This story should have been a simple clear-cut case it weren't for a small rabble-rousing group. Funnily enough, Theo posted a fairly decent and non-inflamatory respones in the early discussions. This is in stark contrast to the earlier GPL case (mentioned in your post) where his reaction was indefensible.
Incidentally, the BSD licence infringement has already taken place. That happened as soon as the author distributed the code with the licence stripped from it. Doesn't matter whether or not it hit a main-stream kernel. As soon as he made it available to others, distribution kicked-in. That said, the author has a case to answer for but certainly not the entire Linux community the "OMG LINUX STOLE OUR CODE!" crowd would have us think.
-- Using the preview button since 2005
It appears that you can't read this particular patch style. The lines with + mean added, the lines with - mean removed.
.c files were handled appropriately and it was merely the .h files that had the license completely ripped out. The .c files were dual-licensed and said you could choose either. They just removed the BSD license as that was "choosing" GPLv2. The .h files are just some interfaces and don't change often anyways, so the BSD license is good enough for them (they should have left those). The .c files are the actual implementation, which would change between operating systems.
.h files since they didn't have the dual-license notice in them. If they aren't dual-licensed under both, you can relicense as GPLv2, but you have to include the BSD notice under its own terms. The GPL itself even says not just attribution, but the original notices themselves must be preserved. One additionally might say that since the GPL says to preserve the original notice, that even in the dual-license case you must preserve the BSD license in order to initially comply with the GPL, although that's a requirement of the GPL and not a dual-licensing/BSD provision. A dual-licensing (as you can see in this case) clearly says you can pick either, since the word "Alternatively" (e.g. the ath5k_reg.h license) implies if you chose the following path, you can ignore the provisions of the previous path.
The lines without either mean that's context for the differences.
If you look at the original patch, no attribution was removed. The attribution was in the context lines.
It looks like the
Here's a link to the actual diff as provided in the original article:
http://lkml.org/lkml/2007/8/28/157
You'll also note that the dual-licensed code had the committer's copyright notice on it. In some cases it was only his notice, originally. With the data immediately available, maybe he stripped it out in a commit before this one, but they don't seem to be accusing him of that. They are mainly accusing him of ripping out the BSD license from a couple
In summary, it looks like a lot of this was nit-picking over how to actually do the license notice preservation, rather than preserving somebody's attribution. I imagine it'll be fixed up in very little time and few people will care about this in more than a day or two.