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.
No orthography among ACs, eh?
Im suddenly reminded of this, where linux gpl'd code found its way into BSD via a wireless driver.
Those in glass houses shouldnt throw stones
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.
There are shills on slashdot. Apparently, I'm one of them.
http://lkml.org/lkml/2007/8/29/69
Someone pointed out the problem and a patch is likely on its way.
Conversion Rate Optimisation French / English consultant
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.
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 move is clearly against the BSD license. (Also, combining GPLv2ed code and BSDed code is subtly against the GPL, as the requirement to reproduce the license - as shown and violated here - is an extra requirement compared to the GPL, violating the "no additional restrictions" clause of the GPL.)
Eivind.
Doubting the existence of evolution is like doubting the existence of China: It just shows that you're uninformed.
Absolutely. This situation is unbelievable.
A few months ago, a GPL'd Linux driver was incorporated into the mainline OpenBSD kernel tree (albeit some months before release, and largely non-working.) The Linux developers concerned contacted the OpenBSD team via the regular mailing list, Ccing some relevant Linux and legal people. The email was polite, more or less friendly, and constructive, offering help to the OpenBSD people to ensure the situation was resolved with both projects having a working driver at the end of the day.
The OpenBSD team's response was to go nuclear. Theo called the Linux developers "inhuman". Many argued that the copyright violation was legitimate performing coding acrobatics to pretend that real, copyrighted, code was never being distributed under the BSD license; others argued this proved the superiority of the BSD license because if it had been the other way around, the OpenBSD team would never have objected, given the BSD license allows you to do (apparently) anything, whereas the GPL prevents use in closed systems.
Well, what a bunch of, frankly, hypocritical two-faced liars. The OpenBSD team's response to an apparent BSD license violation (which we were assured would never happen, because the BSD license is so liberal) is to directly accuse the Linux developers of copyright infringement. Rather than involve appropriate mailing lists and relevant people, the complaint is made on the public Undeadly.org website. Rather than offer help, the OpenBSD developers just spit blood. And none acknowledge that the copyright infringement hasn't even happened yet, that is to say, the proposed code is a patch that has yet to be accepted into the mainline kernels.
This is the second time the OpenBSD team have owed Linux developers an apology, and I bet it's the second time we're not going to hear one, instead hearing the same self-righteous fraudulent apologetics we hear one.
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.
You are not alone. This is not normal. None of this is normal.
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
Theo and Company happen to be VERY obnoxious at times. This would be one of them.
They were guilty of the very thing they're accusing the Linux crowd of back a while
back and the Linux crowd handled it rather nicely and helpfully, but Theo went ballistic
and basically got all bent out of shape indicating that they weren't really violating
the GPL licensing on a kernel driver (they were, but...) and so forth.
Now, we see a percieved violation being "observed" by Theo and Company
and in reality, the people in the discussion thread all bent out of shape over it weren't
paying close attention. The original author did the change- which is legit all the way
around.
This isn't about dividing BSD and Linux. This is about Theo and some of the OpenBSD
crowd being a little more mouthy than usual and simply going off half-cocked.
Nothing new here- move along.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
No it isn't. You can reply to your own posts all you like. Where do people get this weird notion that it's bad form?
I agree with you, nothing wrong with that.
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.
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.
Slashdot social media options: AIM, ICQ, Yahoo, Jabber and Mobile Text. Why no MySpace?
And that's what many of us BSDers are against. We want our software to keep freedom. Freedom isn't binary, nor can it easily be discussed without qualifiers. In this case, BSD is a case of the developer saying "I want freedom, I want it for myself, and who cares what everyone else does with it" whileas GPL is a case of the developer saying "I want freedom, I want it to rest with the end user of my software and never mind if that causes me some inconvenience in the process". They are different kinds of freedom, resting in different parts of the software ecosystem.
sigs are hazardous to your health
Really? Because this is what I read from Theo: 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.
We always knew Comcast was corrupt, here's the proof: http://tech.slashdot.org/comments.pl?sid=1909890&cid=34545432
Some points...
/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 (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).
l opers-Steal-OpenBSD-Code-for-Wireless-Driver) and this is really anoying !!!
/ ic/), so "bad linux developers against openbsd" scenario doesn't apply here...
0 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.
a) ath5k_hw.c
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-Deve
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
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/20
"Except that it's the author and copyright holder who did it, and he's allowed to do what he likes with his copyrighted works."
The submitter wrote some of the driver, but there are a few other names in the Copyright list. There is no information in the article indicating their (dis)approval.
Doing the same thing in the other direction (including GPL-licensed code in a BSD-licensed project) is a no-no, by the way. The GPL has additional requirements that the BSD license doesn't satisfy.
$x='S24;r)>63/* h@<5+oZ)32"5cz';$me='phroggy'x$];
$x=~y+ -xz+\0-Tx+;print$_^chop$me for split'',$x;
I think the difference is how freedom is best preserved. In a BSDL community, you encourage everyone to contribute because it benefits them and everyone else by doing so, and it hurts them not to contribute. This works becasue if one doesn't contribute back then it becomes prohibitive to upgrade to the latest version fairly quickly, but if one does contribute then everyone else can supply patches and improvements. Furtermore, if two people create a different fix to the same problem and only one contributes the patch, the person who didn't gets screwed, especially if their version is better since now they have to maintain the difference or lose functionality.
:-)
BSD uses economics to protect freedom. GPL tries to use the force of law.
Generally I prefer the BSD approach but tend to feel safer with the GPL
LedgerSMB: Open source Accounting/ERP
Please cite me any source of information that says dual license means you are bound by two separate and distinct licenses. All definitions of Dual License I have seen are an either license context, not BOTH licenses. A good example is MySQL. Dual licensed under a commercial license where anyone can do what they want for a fee. They in no way are required to redistribute the code if they choose a commercial license.
-- I'm the root of all that's evil, but you can call me cookie..