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."

No, it doesn't.

[Whiney Mac Fanboy]

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. [blogspot.com]

Anyway, thanks to the OpenBSD team for these great drivers. Thanks to the Linux team for including them (under the correct license).

Re:Strange

[Whiney Mac Fanboy]

The following was removed from the license:

* 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.

That sentence is pretty clear. Not all BSD code can be relicensed.

Let me remind you however, that this was the work of an individual who posted to a public mailing list. It hasn't been accepted into Linus's or Morton's tree.

Re:Strange

[fruey]

http://lkml.org/lkml/2007/8/29/69 [lkml.org]

Someone pointed out the problem and a patch is likely on its way.

Dual licensed

[xaxa]

Ignoring moral issues, is there a problem? The source was dual-licensed under GPL and BSD licenses ("Alternatively, this software may be distributed under the terms of the GNU General Public License ("GPL") version 2 as published by the Free Software Foundation."), so isn't it allowed to release it under just the GPL? IANAL.

I'll leave moral issues to another thread.

Re:Those in glass houses...

[moranar]

The response from the person involved is at least much more responsible and reasonable than in the earlier incident: http://lkml.org/lkml/2007/8/29/69 [lkml.org]

Re:No, it doesn't.

[phoebe]

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.

The contributor being the author of the wireless module makes this article a bit short on common sense.

First check the author of the patch, its Jiri Slaby.

Subject [PATCH 4/5] Net: ath5k, license is GPLv2
From Jiri Slaby <>
Date Tue, 28 Aug 2007 12:00:50 -0400
Digg This

ath5k, license is GPLv2

The files are available only under GPLv2 since now.

Signed-off-by: Jiri Slaby

Then check the copyright notice on top of the source files, there is a copyright to ... Jiri Slaby.

+++ b/drivers/net/wireless/ath5k_base.c
@@ -4,25 +4,9 @@
* Copyright (c) 2007 Jiri Slaby
* All rights reserved.

So an author changed the license of his own code, hit the presses!

Jury's Still Out

[Bob(TM)]

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 ...

Re:No, it doesn't.

[Lumpy]

Exactly,

I don't even know why this is news, Until Linus accepts it, it's some random patch submitted to the tree, tons of those are rejected daily.

The entire story and Slashdot submission is plain old FUD. if it was accepted and part of a new kernel tree I can see the story, but right now it's absolutely nothing but some random guy changed. Are we going to start getting stories submitted about what someone says on their blog now?

Legal Weirdness

[saterdaies]

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.

Re:Jury's Still Out

[dysprosia]

There's still a problem as not all the files are dual licensed (eg drivers/net/wireless/ath5k_regdom.c). They can't strip the entire license text from those files which are licensed BSD only and relicense as GPL.

Re:Mod parent up!

[Eivind Eklund]

Giving an exception seems within the spirit of the GPL. Reproducing the copyrights/licenses is actually onerous in many cases, but it doesn't usually interfere much with the end user's ability to change and redistribute the code as source, which is what the GPL seems to mostly be concerned about.

Personally, I would probably add another GPL poison pill to whatever I released after that, though - to require people to actually contact me and have the code relicensed if they want to hack around with the GPL. And convincing me to relicense would include convincing me that they knew exactly what they were doing and had sound reasons for needing/wanting to use the GPL instead of sticking with a more free license; I see routine licensing under the GPL as damaging, and want to do my little bit against it.

Eivind.

From the thread after TFA...

[Half-pint HAL]

An Anonymous Coward wrote this by the original article....

How much you will to bet this won't instantly appear on Slashdot

;-)

Re:No, it doesn't.

[kernelpanicked]

Wrong! But thanks for playing. I see you convinced the greater slashdot horde to give you a few mod points for your wrong answer, congrats. The correct answer is the code is copyrighted 2004-2007 Reyk Floeter 2006-2007 Nick Kossifidis. This is a clear case of someone grabbing BSD code, stripping and replacing the license with the GPL, and submitting it as a patch to the mainline kernel.

Re:Strange

[Ajehals]

The notion come from the fact that If I had taken the time in the first place I would have been able to post a more accurate and coherent post, rather than having to submit updates with corrections in, not good form.

Re:No, it doesn't.

[uglydog]

But what about the lines that say

- * Alternatively, this software may be distributed under the terms of the
- * GNU General Public License ("GPL") version 2 as published by the Free
- * Software Foundation.
- *
+ * This file is released under GPLv2

Doesn't that mean the the second person is opting to distribute under GPLv2? And the copyright notices are intact.

* Copyright (c) 2004-2007 Reyk Floeter
* Copyright (c) 2006-2007 Nick Kossifidis
*

Is BSD's "author reference" GPL's "restriction"?

[Anonymous Coward]

To begin with, I just contacted one of the authors and it appears
this is just miscommunication by some people who didn't bother to ask:
the code is really dual-licensed BSD and GPL, so that people
from all sides can get the benefit. The case that wasn't.

Let's remind that GPL and BSD are different licences.
You can turn a BSD code into GPL but not vice versa
and this has some serious implications, since GPL *does not*
enforce author back reference as long as the code remains GPL;
short of "the copyright holder is FSF" itself.

It means that this is a legally valid path:
BSD code with ref. -> GPL code with ref. -> GPL code without ref.

It is not clear to me though if BSD's request for author reference
should be considered "a further restriction" under GPL's regime.
A lawyer please?

We all agree that some back reference would be nice,
if not for credit at least for documentation reasons.

Re:No, it doesn't.

[Ed Avis]

So an author changed the license of his own code, hit the presses!

Nope... check the first patch that appears in the article:

@@ -2,17 +2,7 @@
    * Copyright (c) 2004-2007 Reyk Floeter
    * Copyright (c) 2006-2007 Nick Kossifidis
    *
- * 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.
+ * This file is released under GPLv2

See the two copyright holders? They would need to give permission.

N.B. 'This file is released under GPLv2' is not really the recommended notice to add to source files. See 'How to apply these terms to your programs' at the end of the GPL text.

Re:No, it doesn't.

[georgeb]

That's a load of crap. Being dual-licensed means that I can choose to abide whichever license suits me best. This _includes_ the right to redistribute the whole code under the GPLv2, ignoring the BSD license and all it's requirements altogether.

Did you even read the original patch?

[sethawoolley]

It appears that you can't read this particular patch style. The lines with + mean added, the lines with - mean removed.

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 .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.

Here's a link to the actual diff as provided in the original article:

http://lkml.org/lkml/2007/8/28/157 [lkml.org]

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 .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.

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.

Re:No, it doesn't.

[stevew]

Correct as far as you went - but did you bother to look at the followups???

Several people basically said Nope - can't do that. How about dual licensing?

The author replied - yes please, I'm away from my system right now - could someone do that.

(the above paraphrased..)

So in my mind - someone made a mistake, others pointed it out, and the original author asked for it to be corrected in the suggested manner.

-MODULE_LICENSE("Dual BSD/GPL");

[mwvdlee]

This line (taken from the diff in the link) says it all, really.

The original implementation was dual licensed BSD/GPL.

The submitter changed some bits and decided to pick the GPL license (both would have been allowed).

Now the submitted code is GPL-restricted.

It's a pretty pathetic thing to do, cutting off the source from any usefull changes, but perfectly legal nonetheless.

Re:Hmmmm

[Wannabe Code Monkey]

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.

Really? Because this is what I read from Theo [undeadly.org]: It boggles the mind. One writes legal text which says "You may not delete this", and their approach is to delete it, and splatter GPL-gizm all over it. "Screw the everyone and theirlaws, we are GNU...". He sounds like an ass to me regardless of who's right or wrong.

Re:Copyright is only good when it comes to the GPL

[teh_chrizzle]

it's a copy paste job:

http://www.kuro5hin.org/story/2003/2/13/8422/16656 [kuro5hin.org]

http://slashdot.org/comments.pl?sid=159323&thresho ld=1&commentsort=0&mode=thread&cid=13343214 [slashdot.org]

Stop this nonsense...

[Anonymous Coward]

Some points...

a) ath5k_hw.c /ah5k_hw.h / ath5kreg.h and ath5k.h do not come from openbsd tree, they have code from bsd (that's why Reyk's copyright is there) but are written from scratch. They first appeared on madwifi svn (http://madwifi.org/changeset/2232) and are part of madwifi-old-openhal project. They have a different layout (eg. code is not split per-chip as in openbsd cvs but it's common for all chips, lot more documentation on registers etc) and you can see that changes have been done since http://madwifi.org/log/branches/madwifi-old-openha l/openhal [madwifi.org] (initial register writes for example are done in a different way than original openbsd code). So it's a derivative work or a "fork", not a "copy" as the license says ("copyright notice and this permission notice appear in all copies."). So if those files had from start a GPLv2 license it wouldn't be a problem (since they are not "copies" of the original code and author's copyright -Reyk's- is still there so there is no copyright violation either).

b) Original author of those files (mickflemm) later uploaded them on madwifi svn repository again but now with a different license (http://madwifi.org/changeset/2670), GPLv2 as you see (Reyk's copyright is still there of course)...

So where is the problem ???

I see no violation, only people calling other people thieves (http://www.osnews.com/story.php/18528/Linux-Devel opers-Steal-OpenBSD-Code-for-Wireless-Driver) and this is really anoying !!!

Also have in mind that Madwifi team have provided patches on openbsd (you can see that on openbsd cvs http://www.openbsd.org/cgi-bin/cvsweb/src/sys/dev/ ic/ [openbsd.org]), so "bad linux developers against openbsd" scenario doesn't apply here...

To summarize the whole thing IMHO is nonsense, Theo just wanted to make a point against linux developers after a serious (even copyright was removed) violation commited on openbsd's cvs (http://lists.berlios.de/pipermail/bcm43xx-dev/200 7-April/004370.html), not a test branch like -mm, the core cvs. Also have in mind that Theo back then criticized Mike for doing this on a public mailing list etc and now he didn't say a thing about publicity.

Re:No, it doesn't.

[Macthorpe]

As far as I recall, BSD is basically "do what you want as long as the attribution remains" - and that attribution is still in the Microsoft code. However, the offenders stripped out the whole text and therefore infringed the license.

You can correct me if I'm wrong, of course.

Re:No... It's about something a little different..

[Anonymous Coward]

Except, that dual licenses means just that. In the case where a program is dual licensed, one must maintain compliance with both licenses, or do so to the fullest extent possible.

Um, what on earth gives you that impression? The whole point of dual licensing is that end users have the option of choosing one license to comply with, and disregarding the other license completely. This makes it possible for people other than the copyright holder to incorporate the work into a project that uses a license compatible with either of the dual licenses.

Re:No, it doesn't.

[Grax]

The patch file quoted removes only the BSD license notes, not any attribution to a creator. Since the file specifically states that it (the file) is available in either BSD or GPL license, it makes since to me that a GPL user would say "OK then. We are using it as GPL" and remove the BSD license notes.

Re:No, it doesn't.

[Soruk]

Some of the files changed are indeed dual licensed.

Unfortunately, some of them are only BSD-licensed. That's the big oops here.

Re:I wouldn't even go that far

[einhverfr]

The problem with BSD is that I can take a fork of a project today, throw a whack of talented resources at it and make a better product. Launch that into the wild and have everyone take my (non-BSD'ed) fork as the dominant project. Give it some time, and everyone has forgotten about FreeWhateverItWas and is now using PinkPanther_v5. Everyone is drinking from the PP-koolaid...then I start re$tricting acce$$.

And you are still competing with Free (both kinds). If you are extremely lucky the original development will be slow enough to actually let you be successful at this. In which case the project is more or less dead anyway (regardless of license).

Now, one of the BSDL projects I support is PostgreSQL. As your post suggests, there are a number of companies that either currently or in the past have offered proprietary versions of the software. These include Command Prompt, EnterpriseDB, Fujitsu, Green Plum, Pervasive, and SRA.

Of these, Command Prompt still has *one* proprietary add-on (but they no longer sell proprietary versions of the software), Fujitsu has dropped off the radar screen, Pervasive has given up competing with Free, and so has SRA. EnterpriseDB and Green Plum market niche products but they are hardly mainstream. In short, in a few years, pretty every proprietary version which even had a hope of being mainstream died.

In case you are wondering, EnterpriseDB offers a version of PostgreSQL with some extra Oracle compatibility. Nobody in the PostgreSQL community (myself included) wants this in our software. So we are happy to let them sell that. After all, they contribute a lot of code back to the main version. After all, they want to be competing against Oracle ($$$), not PostgreSQL (Free).

Similarly Green Plum makes a version of PostgreSQL aimed at buisness intelligence markets. They release a single-node version open source, and a version capable of parallelism under a proprietary license. The parallelism is what you pay for in BI space, so that is what they keep to themselves. Again, they want to be competing with Teradata, Oracle, and DB2 ($$$), not PostgreSQL (Free).

Pervasive tried to compete with Free and discovered it didn't work...

Where the BSDL has some drawbacks though is that it discourages businesses from being first movers in the development. The basic problem is this: You license your software, and your competitor can take that as you released it to get ahead. The GPL solves this problem, but in my view, but another option might be to approach some competitors and ask for contracts stating that for 1-2 years, they will contribute all the code thee write for it back. By then, you should have a larger community.