Submitted by “I, Michael Buesch, am one of the maintainers of the GPL’d Linux
wireless LAN driver for the Broadcom chip (bcm43xx). The Copyright holders of bcm43xx (which includes me) want to talk to you, OpenBSD bcw developers, about possible GPL license and therefore copyright violations in your bcw driver. We believe that you might have directly copied code out of bcm43xx (licensed under GPL v2), without our explicit permission, into bcw (licensed under BSD license).” The entire thread can be found here.
Jesus, some over-reacting GPL boys, it’s opensource! Why not mail the guy and talk to him? Why “discuss” this mumbo jumbo first in public? Okay it isn’t right to take someones code, but it should be possible to talk in opensource and to avoid troll-bait in public.
Maybe it is sort of over reacting, but if the developers accept their (GPL’ed) code to be taken and relicensed as BSD, the next step is the code to be taken and possibly be used in proprietary projects.
Maybe it is sort of over reacting, but if the developers accept their (GPL’ed) code to be taken and relicensed as BSD, the next step is the code to be taken and possibly be used in proprietary projects.
Drivers should be public domain ONLY.
Naaaaaaah… no reason for that. LGPL would be better. Can be blended with proprietary licenses but still protects the original project.
I would agree with you if the LGPL was better, but the ‘dynamic linking’ stuff of LGPL is just stupid..
And I’m not the only one who think so: there are many variants of the LGPL which rephrase it without the ‘dynamic linking’ restriction.
Eehhh? You mean “static linking”
Drivers should be public domain ONLY.
I totally agree with you. But then again, all vendors should be the ones releasing the drivers, or at least the specs.
When one side plays dirty, the developers should take measures. the bcm43xx developers have chosen the GPL and other developers should respect this.
Edited 2007-04-06 13:49
You trust the vendors for drivers ? Strange, in a world where some multi thousand dollars devices has bogus drivers that make the whole thing almost unusable.
Specs should be public domain, at least to be able to rewrite a bad driver.
It is not your place to enforce your beliefs on the whole world. Whether the GPL is morally applicable to drivers is not the issue here.
Oh, but it’s ok for Stallman to use the GPL to enforce his beliefs about patents on the world with the GPLv3?
It is a two-way “street.”
No, Anyone is free to define their own license. The success of the GPL is a result of 3rd party adoption NOT someone enforcing their beliefs.
Ste.
The success of the GPL is a result of 3rd party adoption NOT someone enforcing their beliefs.
I’m sure Stallman would say it is successful because of the “freedoms” it guarantees (enforces in my view), not its 3rd party adoption…
Edited 2007-04-07 04:06
You miss the point. A license is only successful if it is used. The fact that it is adopted by a *lot* of projects is why the GPL is successful.
Developers choose the GPL. Who are you to say that they shouldn’t? It’s their code, they hold the copyright on it, it’s up to them to decide how it can or can’t be redistributed. There’s little else to say about this…
No, I don’t miss the point. You missed my point. My point is that you believe it is successful for different reasons than others believe it is successful.
Secondly, *some* developers choose the GPL. *Many* others do not. See the Apache, Perl, IBM Java, Eclipse, Apple, Cedega, OpenSolaris, Mozilla (originally), libsdl (LGPL, NOT GPL), *BSDs, and other projects.
In sheer terms of volume of code, there is *less* pure GPL software than software under other licenses. In fact, many GPL projects have *BSD* licensed code included in the project (that is still under the BSD license). Therefore, it is misleading to say flat out that “developers choose the GPL,” if they all did, we wouldn’t have license wars…
I’m sorry, but you’re not making much sense. Who are these others you are referring to? How do you know our opinions of the GPL’s success differ?
It seems like you’re trying to make a point here, but it’s really not clear what it is. Work on it a bit more and come back to explain it further when you actually know what you’re trying to say.
The GPL is still one of the most popular licenses around. But if you have some numbers you’d like to show everyone, please be my guest.
Of course. No one ever claimed that a majority of software projects used the GPL. However, it is still one of the most popular licenses out there, and I have yet to see you give us any kind of data that would prove this isn’t the case.
Way to put words into my mouth there. Did I say that *all* developers chose the GPL? No, I didn’t. I said that “developers chooose the GPL”. That means that an unspecified number of them do choose the GPL. The number is unimportant in that context. I’m not arguing about popularity here, only that, for those who choose the GPL, it is their right to do so, and not respecting that right is not respecting copyright law.
I don’t know, but it seems your anti-GPL agenda is clouding your judgement here…
Also, I never said they couldn’t. I just have a problem with them stipulating the terms of *my* code, which is a *non-derivative* work in *my view*. Why should *my* code be *forced* to be under *their* license. There’s plenty to say about it. If there wasn’t, everyone would be using the GPL…
It is irrelevant whether *you* think your code is non-derivative. That’s not for you to decide, but rather for a judge in a court of law.
What code are you referring to, anyway? Do you have an actual example, or are you just talking about vague, hypothetical scenarios that only serve your anti-GPL arguments?
If you don’t want to release your code under the GPL, then don’t modify and reuse GPLed code. It’s as simple as that. If you have problems understanding copyright law, then get a freakin’ lawyer and stop spreading FUD.
Actually, a lawyer will tell you that intent is a very important factor in cases. So it does matter what I believe, especially if I have sought legal counsel.
But that’s just it. What if I’m not modifying, redistributing, or using GPL code directly. What if a program is merely linking. Even the fsf seems to be unsure in some cases about whether or not specific cases are infringing. How then, as a developer, without seeking expensive legal counsel, am I to ensure I am in compliance?
Developers need simple, clear, easy to understand licensing that is not overly restrictive.
Ah, but is anything but simple. Stallman himself has said that things weren’t clear enough with the GPLv2, and that is one of the things they are working to correct in the GPLv3.
In fact, why not take the License quiz at fsf.org and tell me honestly how well you do first time around:
http://www.gnu.org/cgi-bin/license-quiz.cgi
You may be surprised.
Please tell me how exactly I am spreading FUD. Please see the WikiPedia definition:
http://en.wikipedia.org/wiki/Fear%2C_uncertainty_and_doubt
Note that I am not spreading “vague” information or “appealing to fear.” Instead, I am stating specific dislike for specific aspects of a specific license.
If you want a specific example, I’ll give you one:
Linux kernel drivers. nVidia somehow gets an nod from Linus for not being infringing (not being a derivative work), but other Linux developers say it is infringing.
If the original author/maintainer of the Linux kernel can’t agree with top developers of the Linux kernel, how can anyone expect the average developer to know what is and is not a derivative work?
You claim I must need a lawyer to understand what is going on, are you one? If not, how are you qualified to tell me what you tell me?
Actually, a lawyer won’t tell you that. Intent may lessen the penalty if you’re been found guilty, but it certainly won’t make you innocent of copyright infringement if that is indeed what you’ve done.
So, no, intent won’t do much for you in a case where you base your code on GPLed code and fail to comply with the terms of the license.
If you’re linking to GPL code, then you’re using it. If you don’t want to use the GPL license, then don’t write your program to use GPLed libraries. Fortunately, there are very few libraries that are linked under the GPL. Most use the LGPL, which allows linking, or other licenses which are equally permissive.
Do you have a specific example you want to share with us? I imagine that the problem for you is not purely hypothetical, and that this situation has come up for you before? I’d be curious to hear about it.
Really? Such as?
That’s very easy. Don’t base your code on GPLed code, or write it to use GPLed libraries, if you don’t want your own program to be licensed under the GPL.
Linking to libraries is *not* what he was referring to.
Uh, no. The modules are legal. It is not legal to redistribute kernels in which the modules are pre-compiled. Some developers didn’t like it, but found out they couldn’t legally prevent post-distribution linking, and since then this has pretty much been a non-issue.
Note that I’d rather have some examples of how this has affected you as a developer.
I’m not a lawyer, but I obviously understand the GPL better than you. (I got 8 out of 9 on that quiz, incidentally).
I’m not a lawyer, but I obviously understand the GPL better than you. (I got 8 out of 9 on that quiz, incidentally).
Would it suprise you to learn that I got the same score the first time? Then logically, I must understand it well as you do, and both of our understandings are not perfect. As such, I shall agree to disagree since I have sufficiently validated my argument.
Sure.
No one is forcing developers to license their code under the GPL. Developers who CHOOSE to use the GPL for their code know exactly what they are doing. They use the GPL because they agree with spirit of the license.
And people wonder why manufacturers won’t support the Linux platform with more drivers…
Edited 2007-04-07 04:06
And I wonder even more after reading this :
http://www.kroah.com/log/linux/free_drivers.html
BTW, Linux hardware support is really great, it’s one of the reasons behind opensolaris’ probable move to the GPLv3…
*bzzzt* wrong. That probable move is not even probable at the moment. Our community in general has not made any decision in regards to that, but I can tell you most of us are opposed at the moment, though we will see when the final version is released.
As people from Sun and others have already stated, even if Solaris was suddenly GPLv2 today, that would not really help our driver situation. Solaris drivers are *significantly* different than Linux drivers.
I fail to see the relevance of your response…
And *then* exactly then, you can bash him to death. But before that you can talk and like you see,they have deleted it at once.
As he wrote in the mail: He doesn’t want Broadcom, who aren’t cooperating with the OSS community at all, to profit from their work. If OpenBSD were allowed to re-release the code under a BSD license, Broadcom could use the work to improve their own drivers for Windows and OS X, while still not cooperating with the OSS community.
He doesn’t want Broadcom, who aren’t cooperating with the OSS community at all, to profit from their work. If OpenBSD were allowed to re-release the code under a BSD license, Broadcom could use the work to improve their own drivers for Windows and OS X, while still not cooperating with the OSS community.
Rewording this sentence as I see it:
We, GPL people, won’t allow Broadcom to make better drivers for Windows and OSX.
This is the nastiest philosophical difference between BSD and GPL style licenses. BSD code is written to be benfitting anyone (including users of propieritary products); GPL code is meant for GPL camp only.
I understand (and respect) GPL ideology and I’m not against using it for standalone applications, but for drivers it is not the best license. Like noted above by someone – drivers need to be in public domain.
We, GPL people, won’t allow Broadcom to make better drivers for Windows and OSX.
That’s exactly their intention. Broadcom is not a company that gives anything to anybody. For a public project to improve their drivers (allowing them to take and integrate at will) only helps to strengthen their position in the market, which hurts all of open source software. The biggest rip on Linux, and all the open kernels, is driver support. Why should any open project help a company that won’t help any open projects, and indeed furthers the fact that open kernels have poor driver support?
The ideal situation (outside of Broadcom actually cooperating with outsiders) would be for demand for Broadcom products to fall to such a degree that Linux’s poor support for their cards is a non-issue. GPL’d drivers aren’t going to accomplish that, but at least they’re not helping an entity that makes open kernels look bad.
We, GPL people, won’t allow Broadcom to make better drivers for Windows and OSX.
That would be true, if the GPL authors somehow obtained the authority to prevent broadcom and others from writing better drivers. The GPL nor the authors have that authority, nor are they attempting in this case to obtain or perpetrate such authority.
A better translation might be:
We, GPL people, won’t allow Broadcom to make better drivers based on our work.
That’s a far more accurate assessment, and something within their rights as copyright holders who are free to license the code in whatever manner they feel is appropriate.
This is the nastiest philosophical difference between BSD and GPL style licenses. BSD code is written to be benfitting anyone (including users of propieritary products); GPL code is meant for GPL camp only. I understand (and respect) GPL ideology and I’m not against using it for standalone applications, but for drivers it is not the best license. Like noted above by someone – drivers need to be in public domain.
The way I see it, I’m glad there are both GPL and BSD licenses. That is the true freedom. Freedom to license the code and have it used in the manner you think best is appropriate.
We, GPL people, won’t allow Broadcom to make better drivers for Windows and OSX.
This is the nastiest philosophical difference between BSD and GPL style licenses.
No, not at all. If Broadcom were to make GPLed drivers for Windows and OS X and share the code back, then no problems. That’s how things move forwards, and how the Linux kernel works as well as it does. That’s the primary concern of the developers in this – not making better drivers for Windows or OS X.
We, GPL people, won’t allow Broadcom to make better drivers for Windows and OSX.
This is the nastiest philosophical difference between BSD and GPL style licenses.
No, not at all. If Broadcom were to make GPLed drivers for Windows and OS X and share the code back, then no problems. That’s how things move forwards, and how the Linux kernel works as well as it does. That’s the primary concern of the developers in this – not making better drivers for Windows or OS X.
Yes and no.
From (GPL) developers standpoint I can agree – they did the dirty work, but hardware maker won’t give anything back – why then allow it to use this code?
But from users standpoint this is nonsense. User wants well working hardware and that’s all. In current case it’s clear that GPL actually goes against the normal users, it doesn’t make them any good.
Or if Broadcom propieritary drivers are good enough, then initially cited claim (we don’t want Broadcom to take our code) is unreasoned, because Broadcom wouldn’t want their code anyway.
Somehow seems that GPL people are afraid that their code may not be so good as they think after all…
Of course this is speculative opinion, I don’t know about neither (GPL nor propieritary) drivers quality.
BTW, differently from Linux kernel development, in current case GPL does not move driver development forward.
Neither Broadcom nor OpenBSD people will give anything back to GPL team, so are there no other commercial parties, who would be interested in developing these drivers. Single developers – yes, but they would contribute back to BSDL code either.
Please explain why this is the case…the GPL has *nothing* to do with who uses what, it doesn’t affect users at all. It does affect what you can do with the code if you modify and redistribute it, though.
This is completely irrelevant. The developers *chose* to release their code under the GPL, therefore the BSD devs *cannot* use that code and re-release it under the BSD license. Doing so would be *against the law*. Are you guys advocating copyright violation? Because from where I stand, it certainly looks like it.
We, GPL people, won’t allow Broadcom to steal our work to make better closed/proprietary drivers for Windows and OSX.
I never saw such arrogant words.
Just as broadcom need them to make the driver for THEIR OWN hardware.
That’s just egotistical to think that ONLY open source developers develop quality drivers.
That’s not what they’re saying at all. This is about copyright violation, people!
THey have no problem with Broadcom making better drivers. What they have a problem with is Broadcom appropriating THEIR work without appropriate compensation to use for whatever purpose, whether that be better drivers or not.
The key phrase, “appropriate compensation,” is what matters. In their case, that compensation might simply be the release of the new driver source code back to the community, it does not have to be monetary in nature.
I don’t follow this particular vendor, but some hardware manufacturer’s are notoriously closed about their hardware and the specs needed to write good drivers. Assuming Broadcom is one of those companies, why should they be permitted to take the hard work of another group that wants their code open, use it for Broadcom’s benefit without complying with the terms of the original authors and closing the code?
Suppose the reverse engineered driver doesn’t take advantage of something in the chipset because Broadcom didn’t release the specs. Broadcom takes the bcm43xx code, adds support for that special chipset, making the driver work a little better. Only now, instead of releasing that change back to the community under the terms of the GPL, they keep it closed. Broadcom benefits at the expense of the original developer’s work.
To me, it doesn’t matter what license you work with, GPL, LGPL, BSD, Artistic, whatever. If you are using code, respect the license that code was created under. If the license won’t work for what you want to do with it, contact the license holders and ask to license it differently.
Broadcom takes the bcm43xx code, adds support for that special chipset, making the driver work a little better. Only now, instead of releasing that change back to the community under the terms of the GPL, they keep it closed.
Do they(Broadcom) care at all if it is GPL or BSD licensed code? How can you tell if they actually used/embraced FREE code or not?
The can steal “ideas” from the code, whenever they want to. So it’s nonsense to talk about protecting the code. It’s one mistake on the OpenBSD side and hate on the other side. Man to man would be the right way, not this childish behaviour!
The phrase offset by commas, though not gramatically required for correctness, is essential to the meaning of this sentence. It means the difference between denying someone help out of being petty, and denying someone help because they are being belligerent.
It’s the difference between:
We don’t wish to provide aid to Iran.
And:
We don’t wish to provide aid to Iran, which is threatening to blow us up.
If Broadcom doesn’t want to be nice to the OSS community, we don’t need to be nice to them. Simple, pure playground rules, unsullied by nonsense about “helping everyone”.
Broadcom can use the GPL’ed driver, but they can’t close it, and they will have to share improvements they make to the GPL’ed code. That would all be great. Under a BSD license, there would be little incentive for companies to share their improvements.
They can steal it and nobody will know it – that’s the problem with closed-source
…so, you think that Broadcom should release their drivers in the public domain, then?
I don’t understand how you can both say that GPL is bad because it prevents Broadcom from using the GPLed drivers and make closed-source versions of them afterwards, while saying that drivers should be public domain. You don’t seem to fully understand this issue, at least that’s the impression your contradictory arguments leave me.
No, this isn’t about keeping out OS X and Windows, it’s about not letting Broadcom take your work. Anyone can take the drivers and port them to Windows, as long as they are distributed according to the GPL. In essence, the GPL enforces the spirit of public domain through copyright.
So let us analyze it:
GPL believes people are dishonest and there is a need of a license to legally force people to accept GPL terms.
Commercial license believes people are dishonest and there is a need of a license to legally force people to accpet their terms.
hmm which one is better..none…ignore the license and select the better product.
Which one is best? The one that promotes survival of the fittest i.e. BSD.
The motto of BSD is:
We love programming and it is our passion. We write code and we make it public domain. If you can do better then feel free to take it and do what you want to do with it. In the end better will survive.
Ahh BSD world is so much more pure and honest.
PS: I wish GPL never existed. It is really a cancer to the spirit of open source.
Edited 2007-04-07 20:18
PS: I wish GPL never existed. It is really a cancer to the spirit of open source.
Really?! well i guess you are using Firefox as browser. Firefox is unter MIT, LGPL and GPL.
And what do the firefox foundation did some monaths ago? They donate 10.000$ to the OpenSSH (and OpenBSD) project.
So next time think befor you’re writing.
Thx.
Two reasons:
1) The IP theft was done publicly.
2) The devs for the linux-driver don’t want Broadcom to create a proprietary driver based on their code.
However, the linux-driver-devs could have handled it less mean, but Theo de Raadt is just as mean.
The IP theft should not have happened in the beginning, the linux-bcw developers could have been nicer and so could Theo de Raadt.
Both parts agree that the IP theft was wrong and it will be resolved. And that’s the important thing here.
I’m with Theo on this one. The best way to get licence compliance is to talk to someone privately first. Telling the world is the last step.
But I don’t agree that “IP theft” happened. What was involved was copyright. Talking about copyright by using a blanket term is just adding confusion, and using a senseless blanket term such as “IP” is just being silly. The OpenBSD guy did not “thieve” any geographic designations, or plant varieties, or design schematics – which are all encompased by “IP” – he “thieved” copyright. Good article: http://www.gnu.org/philosophy/not-ipr.html
But I wouldn’t use the word “theft” either. If I steel your lunch, then you will be hungry in the afternoon and I could understand you being unhappy. If I copied your lunch while your back was turned, you would be none the wiser or thinner.
“Theft” and violation of copyright are two very different things. Each can be wrong in various circumstances, but the two should not be confused.
In all fairness, as I read the thread, it was Theo who started to state, that Micheal Buesch et al make the implication :
GPL license violation == thief
Examples for this behavior:
which I quoted from this post [1], other examples of Theo accusing everybody else from implicitly equating license violation with theft are for example [2]
which is from [2]. I’m pretty sure you can find other instances too.
I try to do not take sides in this issue, as I think,
that the license violation should be out of discussion (the GPL kicks in at distribution, whether the code works, is functional or not, period) and both sides have not acted very luckily.
But I can’t help to think, that while I would have prefered too to be addressed in private first, the fact that the code was already out in the wild under a incompatible and more permissive license
without the attribution of the rightful copyright holders makes it understandable, that they tried to inform all involved parties (note the explanation/motivation for the CC list here [3]), so yeah, going public is a valid option.
Furthermore, note that if they “really” wanted to go full public, they could have sent out a statement to the tech press instead or blogged about it (or even submitted it to /. ). THAT would have been clearly over the top.
The step to take the matters in a first step to mailing list instead of starting with a private notification may also have something to do with previous bad blood between the wireless teams at the OpenBSD and Linux camps, but that is just my interpretation. (see the sub-thread in reply to [4])
[1]http://article.gmane.org/gmane.linux.kernel.wireless.general/1588/m…
[2]http://article.gmane.org/gmane.linux.kernel.wireless.general/1600/m…
[3]http://article.gmane.org/gmane.linux.drivers.bcm54xx.devel/4002/mat…
[4]http://article.gmane.org/gmane.linux.kernel.wireless.general/1615/m…
I’m with Theo on this one. The best way to get licence compliance is to talk to someone privately first. Telling the world is the last step.
While I do agree that this could have (should have) been handled much more maturely, it did need to be publicized that the re-licensed code was taken from a GPL’d project and re-licensed illegally. People freely “borrow” BSD-licensed code whenever they see it; the Linux bcm authors didn’t want that happening to their code, since that’s not the license they chose to use. It would have been nice if they had given the BSD contributor a chance to correct and publicize his own action though…
I’m with Theo on this one. The best way to get licence compliance is to talk to someone privately first. Telling the world is the last step.
Why? The bcw code was public, so it was the bcm43xx code. The violation was therefore already public -they just noticed the community of that. People must be aware of public consequences of public actions.
I often admire Theo de Raadt, but this time he acted as a completely hysterical 14-y.o. teen. I read the whole thread. On the latest half of it, Theo just ravages about “he did a mistake and you imply he was doing that on purpose” -How can you copy code by “mistake”? Marcus even admitted that he did it on purpose! He was surely in good faith, but it was not a “mistake”. He was fully aware of copyright issues. He just thought that fixing these later was OK.
If I was Marcus, 1)I’d have answered as in his first mail, 2)I’d have *temporarily* deleted the code from the CVS tree 3)I’d have talked with bcm43xx people about relicensing etc.
The fact he just bitterly dropped the ball after a perfectly reasonable first mail in which he apologized etc. makes me think there’s the shadow of Theo behind.
But I wouldn’t use the word “theft” either.
This I fully agree. Using the word “theft” is always wrong when talking about “IP”, and in this case is plain ridicolous.
>However, the linux-driver-devs could have handled it less mean, but Theo de Raadt is just as mean.
Theo isn’t equal to Marcus, they have no dictatorship in OpenBSD. If you’re thinking of that, you have to handle Linus the same, he too has *leadership* in the kernel development, but it isn’ dictatorship! And Linus too uses some harsh words (look at Gnome e.g. or FreeBSD, interface-nazis and idiots) ….
I haven’t written anything about the way the projects are being lead. This is not about the “leadership” (if any) but about the behaviour of individuals. ***** the collective.
And yes, Linus can be mean too
I love it when you are talking about IP theft. You are among the same people who will say Microsoft should not charge licensing fee to open their protocols which is their IP.
You will be the same person who will be against patents:)
It is funny to see double standards all around in GPL world.
BSD world is where i live and it is heaven there…
So you think it’s okay to violate copyright when the GPL is involved?
Never mind that you’re completely changing the subject (and going into ad hominem attacks) by bringing patents and Microsoft in.
I, for one, respect *all* copyrights, whether they are closed-source, GPL or BSDL.
Nothing better than some OSS drama. It’s been months since the last Linux kernel “tainted” upheaval. We’re long over due here.
Well, ronaldst, copyright is just as important to the OSS community as it is to Microsoft, SCO, and the rest of the proprietary software industry. These seemingly minor issues are a really big deal. It’s a shame that OSS developers have to worry about intellectual property, but the law says that the code that we write is valuable, and that by default, nobody but the original author can distribute it.
For many developers, this isn’t how we want it to work, so we have to carefully waive many of our rights without giving up what remains important to us. What’s important to one developer may not be important to another. But just because we waive some of our rights in the interest of collaboration doesn’t mean that we have given up all of our rights. If anything, the fact that we waive so many of our rights might even make us more protective of the rights we reserve. We didn’t just reserve these rights because the law granted them to us, we reserved them because they’re important to us.
In this case, the crux of the issue is that an OpenBSD developer copied portions of a driver that had been licensed by some Linux developers under terms that that are important to them and committed the code to a project that is distributed under different terms that don’t reflect the values that are important to the original developers. They did so without asking permission and without warning. He didn’t want to “taint” the OpenBSD kernel with GPL code, so he decided to take the liberty of relicensing it to reflect the preferences of the OpenBSD developers.
Did the Linux developers respond in the kindest of all possible ways? No. But I don’t think they should have felt compelled to handle this privately with the developer in question either. They were right to post to the OpenBSD mailing list, but the itemized list of evidence showing the widespread copying was unnecessary in the initial disclosure. They should have simply posted saying that they have identified possible copyright issues in the OpenBSD bcw driver and would like to work with their community to resolve them.
If the OpenBSD folks would prefer to handle the details of this situation in private (for whatever reason), then they should be allowed to arrange for such correspondence. Unless a potential security exposure is involved, the development mailing list is almost always the proper forum for initiating a dialog with a community project. Otherwise, too frequently you never get a response from the individuals who are supposedly the right ones to contact.
They could have handled this in private. After the first post Theo could have sent mail:
“Thank you for bringing this to our attention. We should talk about this in private to resolve the issues. We’ll let everybody know how that turns out.”
and there’d be no bruhaha. No need to admit fault even, just a simple statement and it could have gone to a small private group of people.
Instead, we have Drama! Feh! None of that needed to happen.
Probably but the “damage” had already been done by the miles long CC list. It should have been the GPL guys talking to that developer off list, in private, period. It could have all been over and done with and no harm done. Two wrongs don’t make a right, and both sides became culpable but it didn’t need to happen if the GPL folks would have done the right thing from the start.
The whole freedom thing irritates me. The real freedom is that the author of the code gets to decide how the code is licensed. The power is with the author, which is where it should be.
They Should of contacted Marcus and worked out out that way, Development could of continued on “both” drivers.
BUT Michael Buesch’s initial email was obviously designed to publicly attack Marcus.
The CC’s:
John Linville
Greg Kroah-Hartman
bcm43xx <Mailing list>
linux-wireless <Mailing list>
license-violation <Mailing list>
It was also sent to other OpenBSD developers including Theo de Raadt..
This entire thing could of been handled directly & privately between Michael, Marcus and the bcm43xx team instead of posting it on multiple mailing lists to publicly humiliate him, inevitably forcing him to delete the code from CVS.
Again, It wasn’t handled well.. And now.. Apparent by this damn OSAlert topic and other articles it’s now way out of control.
Edited 2007-04-06 20:27
They Should of contacted Marcus and worked out out that way, Development could of continued on “both” drivers.
Marcus put GPLed code into a public repository and relicensed it without discussing it with anyone first. It was discussed in the way it should have been, with all the right people who should know.
BUT Michael Buesch’s initial email was obviously designed to publicly attack Marcus.
No. That’s Theo’s version of the whole thing in order to try and have something to say on the matter. If the whole thing had stayed on subject and Marcus and everyone else could have discussed what to do, either removing the code or rewriting it, then this could have been solved with minimum fuss.
This entire thing could of been handled directly & privately between Michael, Marcus and the bcm43xx team instead of posting it on multiple mailing lists to publicly humiliate him
Marcus didn’t discuss this first before using GPLed driver code, so this was the correct way to handle it. As Michael stated in his e-mail I referenced above, if Marcus had discussed this privately before using the code then he would have discussed it in private e-mails if he so wished.
As it is, there’s nothing wrong with discussing this on a mailing list and with others, which is where it should be discussed.
inevitably forcing him to delete the code from CVS.
No one forced him to do anything, but the code in question is GPLed code relicensed – which is a clear violation. If the discussion had stayed on topic, without rants from Theo, then Marcus could have discussed more rationally what to do – whether that be deleting the code or rewriting it.
Michael even gave him the option of relicensing:
http://article.gmane.org/gmane.linux.kernel.wireless.general/1587
“I don’t understand your reaction, really. If you were really interrested in doing a Broadcom wireless driver for openbsd, you would have chosen the option to relicense some code (and therefore drop only that code which I refuse to relicense)”
He didn’t want Marcus to drop the code, and offered a sensible resolution.
Again, It wasn’t handled well.. And now.. Apparent by this damn OSAlert topic and other articles it’s now way out of control.
That’s funny. The only reason why this has blown up is because Theo insists on prolonging the whole thing by implying that everyone is calling Marcus a thief. Marcus was never called a thief, and everything he came up with in response was totally irrelevant.
If anyone has prolonged this and embarrassed and scared off Marcus Glocker, it’s Theo de Radtt. As usual. There was no problem here at all – but Theo’s reacted and gone off on one.
That’s very nice of you to repeat and summarize what I read and know. I was surprised to discover that I have a fan here on OSAlert.
IMO Theo should be added to “Everybody loves Eric Raymond” comic. Those are hilarious.
http://geekz.co.uk/lovesraymond/
Jesus, some over-reacting GPL boys, it’s opensource!
No. Taking GPLed code and putting it under another incompatible license is just plain wrong – and obviously wrong at that.
Why not mail the guy and talk to him? Why “discuss” this mumbo jumbo first in public?
The code is in a public repository, mailing lists are the place to talk about this and the issue is quite serious. Surely, Marcus Glocker knew what the implications were?
It seems as though Theo de Raddt just didn’t grok the seriousness of that in his e-mail exchange. He starts talking about how he thinks OpenBSD has the cleanest codebase around (right….), it’s OK because the driver doesn’t run yet (OK……) and it will be resolved in the tree……somehow. He then goes on to talk about it as infighting between two groups, which it isn’t. Relicensing GPLed code under another, incompatible license, is just a huge violation.
There’s then a bizarre comment he puts on the end in response to Stefano Brivio at then end of another comment to trivialise copyright and license issues, in response to how the code had been copied verbatim with whitespace and variable names intact:
“Copyrighted whitespaces and variable names, you mean, right?”
Well yes, Theo. That’s why we have these things called licenses.
I officially own the copyright to the variables i, j, k. As well as any and all white space.
Everyone is in violation.
Now then, shall we procede with that ridiculous arguement?
To clarify though, it appears as if there were actual copied functions, apparently as place holders while he was writting something for himself, according to
http://article.gmane.org/gmane.linux.kernel.wireless.general/1647
It could have EASILY been commited to CVS by mistake. Now then back to the real issue of why was there no private email saying “could you please remove the gpl’d code” first. No, not even close. They went straight for public humiliation.
Edited 2007-04-06 15:39
I’m sorry, but while it is definitely important to discuss the social and moral implications of the “how do I weight the wish of others to be informed in private about their wrongdoings against my wish to protect my own contributions” question, the real issue here is imho a very hand – waiving approach towards incompatible licenses from (at least) the author of the bcm BSD driver.
It is perfectly OK to use foreign code snipplets in my own, private development branch as placeholders, as long as I conform to the licenses of said snipplets before I merge it with the public branch. But if I read the example [1] correctly, some of the bcm43xx snipplets were for instance used to replace already existing but nonfunctional code (quote from [1] :
At least this code segment was not a place keeper. This one replaced a (non-functional, but fixable) existing part and went – if I interpret the rest and the note correctly – into code intended for public testing.
I’m currently not involved with kernel development stuff, but I have to work with code from my colleagues and even if they most of the time release their code into the public domain or use a very liberal MIT-style license (academia), I would *NEVER* commit samples of their code into my repository without
– Placing comments before and after the snipplets, so that I can identify these contributions easier in diffs
– Write down file and linenumber in the commit note, so that all other folks that reuse my code know of all sources.
I always thought, that developers of projects with a far wider audience would take additional steps to prevent a tainting of their code base, but if this all was really just a mistake (although I somehow doubt this if [2] is correct and the the violations prevail in several commits and not in just one) then some mechanisms are needed to prevent such nuisances in the future.
[1]http://thread.gmane.org/gmane.linux.kernel.wireless.general/1558/fo…
[2]http://thread.gmane.org/gmane.linux.kernel.wireless.general/1558/fo…
Well… so much for everyone working towards a common goal. Lets blast the guy in public. Like the other comment above, talk with him in private. I guess the open source community really doesn’t work together all that well.
Open Source is primarily about doing things in the open. Why hide it? Openness is the core of Open Source.
Too bad the moron stole the code. He can only blame himself.
This should have been dealt with in a private forum. It’s sad. If it had happened in private everything could have been worked out nicely for everyone.
But the code was released to the public. Therefore there should be public disclosure to ensure that there is legal enforceability at a later date.
Both parts agree that the IP theft was wrong and it will be resolved. And that’s the important thing here.
Subject: Re: OpenBSD bcw: Possible GPL license violation issues
From: Theo de Raadt <deraadt () cvs ! openbsd ! org>
Yes, this driver has other problems though. To begin with, it does not even run yet, in any sense. Since it is not actual using code, there will be those who argue that the full impact of the GPL does not come to bear yet — noone is “using” the code yet. But beyond that, these types of problem should not exist in our tree. It will be resolved.
So that’s OK. Just steal all the Windows Vista source code, and remove the bootloader stuff. Now you can legally distribute it to anyone because it doesn’t work. Nice.
<sigh>And people still say GPL is the most “free” license</sigh>
How come is it “free” if you can’t use it the way you want, not even in proprietary software?
let’s not start a GPL vs BSD license war here…
they both have different goals in mind, and the bsd people should respect the goals of the gpl.
So Americans are free (aparrently), how come they can’t go out and commit legal mass-homicide?
It seems that the law impinges on their freedoms in an attempt to maintain a coherent community.
Chuck: The GPL is the most free license as it guarantees that you won’t make the code un-free. You want proprietary, write it from scratch. Stop being a troll.
“The GPL is the most free license as it guarantees that you won’t make the code un-free.”
Prohibiting this activity could be seen as un-free, too. But in general I agree. The GPL does not offer the right to take something developed and / or driven by the community and make it proprietary. The BSDL allows this, as long as the original sources are quoted and their license are offered. So one may not cut out the developers’ names and give an application a new name and then sell it.
“You want proprietary, write it from scratch.”
The BSDL allows making proprietary products from BSDL’ed code. Seen morally, this may be wrong. The BSDL has often be called a “rape me license”. On the other hand, the GPL does not provide this… incorporation…
I feel a bit strange about the fact that GPL code may walk 1:1 into BSDL code, allowing anyone to “steal” it. If OpenBSD’s developers say: “Here, it is a working driver, do what you want, sell it, but please, include our license.”, it’s okay, because it’s their choice. They may not do this with code they got from GPL fragments or drivers. But as it has been mentioned before, inspecting and analyzing GPL code in order to write own (BSDL) code is no problem at all.
But as it has been mentioned before, inspecting and analyzing GPL code in order to write own (BSDL) code is no problem at all.
I wonder. Section 2b of the GPL says:
You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.
I would say that reading the source code of a GPL program to get ideas, or to see how something is done, in order to incorporate into another program, would constitute “derived”.
But maybe not. Let’s look at the GPL FAQ:
Q. Do I have “fair use” rights in using the source code of a GPL-covered program?
A. Yes, you do. “Fair use” is use that is allowed without any special permission. Since you don’t need the developers’ permission for such use, you can do it regardless of what the developers said about it–in the license or elsewhere, whether that license be the GNU GPL or any other free software license.
Note, however, that there is no world-wide principle of fair use; what kinds of use are considered “fair” varies from country to country.
A bit ambiguous. Here’s what Wikipedia says about fair use:
the fair use of a copyrighted work … for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright.
OK, that seems to mean that looking at GPL code as part of your research would be OK. But then it goes on:
In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include—
1. the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;
2. the nature of the copyrighted work;
3. the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
4. the effect of the use upon the potential market for or value of the copyrighted work.
Briefly,
– if your purpose is not to supercede the original work for your own profit;
– the nature of the orignal was facts and ideas;
– you didn’t use too much of the original;
– and you didn’t destroy the markey for the original
then it should be OK. So I guess it’s OK for BSD people to look at GPL code a bit. But you know a cleanroom would be a whole lot simpler.
I’m sorry this thread has developed into another annoying “GPL is bad” vs. “BSDL is stupid” thread…
“I would say that reading the source code of a GPL program to get ideas, or to see how something is done, in order to incorporate into another program, would constitute “derived”.”
I’d interpret “is derived from” if the derived program contains parts and / or functional structures of the original program, such as “DragonflyBSD is derived from FreeBSD”; so forks may be seen as derived projects, too. But that’s my own point of view, which is discutable.
“Briefly,
– if your purpose is not to supercede the original work for your own profit;
– the nature of the orignal was facts and ideas;
– you didn’t use too much of the original;
– and you didn’t destroy the markey for the original
then it should be OK. So I guess it’s OK for BSD people to look at GPL code a bit.”
I think so, too. (The other way round is no problem at all.) This is really complicated, because intention has to be known first. Imagine a very early stage of development where it’s not clear if the final product will be a commercial or a noncommercial one. Complicated too, how much is “not too much”? Hmmm…
You misunderstand.
The freedom is not for you. The freedom is for the code.
GPL is free and will remain free.
This gurantees that users and developers will have access to free code.
GNU GPL is the most “free’ license for code, which is what a license is actually applied to.
– Jesse McNelis
The “Freedom” is for users with regards to the rights bestowed upon them by Free Software.
Almost no license is free in the ordinary sense of the word. They all have restrictions and limitations. Otherwise there would be no need for a license. And if your project has no license, be prepared to be sued for damages that may be caused by your software.
The purpose of “Free Software” licenses like the GPL is to grant and protect “Freedom” of users, not software. As such the GPL grants users, not software, the following rights:
– “Freedom” to use
– “Freedom” to study
– “Freedom” to modify
– “Freedom” to distribute
In fact, all Free Software including the BSD grant users these rights. However, the GPL goes a step further. It protects users. It ensures that software and its derivatives always grant users the aforementioned rights.
This is the reason the GPL has a reputation for being contagious or viral. “It infects everything it touches.” This is also the reason it is the most widely used, and in my opinion the most successful, Free Software license.
A lot of people think the GPL is about software. This is only true at a superficial level. More profoundly, the GPL is primarily about users. In particular, the rights they have to use software. It’s a political weapon first, before it’s a technical documentation.
Edited 2007-04-06 16:51
The GPL does not prevent usage with proprietary software. It is perfectly legal to link proprietary software to GPL’ed software and opposite (depending on the country).
What you cannot do is distributing the resulting binaries under a proprietary license. But you can use it as you want to.
Distribution != usage.
You can use it in any way you want, including in proprietary software.
You just can’t DISTRIBUTE the binaries of GPL’d code without adhering to the license.
I not an expert in these religious matters, but I would recommend:
12 Hail Richards
3 Months working with Visual Basic developers
That should do it!
That’s okay. A bit harsh but reasonable.
Now now, behave! That’s SOOOOO evil and uncalled for. Nobody deserves that kind of punishment
For those of you wondering whether (a) the Linux device developers were being unfair and (b) whether the GPL applies if the code supposedly doesn’t run, you need to reconsider whether “borrowed” code was being distributed under a different licence. If so, it’s a public matter *and* a violation of the GPL. If it was being done in public, any notification of the matter merits widespread publicity.
Of course, De Raadt plays the political games for which he is infamous, but on this matter he should tone down his response: it’s far better to work with people who want a reasonable solution than to provoke them into legal action. Or perhaps we’ll be hearing that OpenBSD’s legal fund needs topping up in the near future.
Or perhaps we’ll be hearing that OpenBSD’s legal fund needs topping up in the near future
Not nearly as much as Linux vendors when Microsoft finaly nails them for stolen intelectual property.
Excuse me, how is De Raadt playing political games? all he says is “give us the damn specs!” where as the GPL camp seem to have announced a fatwa and getting ready a mardi-army together to launch a jihad.
There is a big difference between pushing a vendor into the corner to demand the specifications for hardware to be disclosed vs. the move by the roudy bunch over in the other camp.
If it were me, and I wrote a driver, I’d be bloody excited if heaps of people were using my code, and a vendor decided because my code is so good, that they would use it for their own driver development.
If it were me, and I wrote a driver, I’d be bloody excited if heaps of people were using my code, and a vendor decided because my code is so good, that they would use it for their own driver development.
But then you’d perumably release your code under a license that allows it. The orignal developer doesn’t share your goals and that should be respected.
Usually I might argue against Theo’s somewhat exasperated email replies however in this particular case if you bother to read the entire thread you’ll find that Theo makes a good point… eventually. I don’t think anyone would be any less discouraged from such a public ousting. Then again broadcom has to be one of the worst wireless chips/drivers I’ve ever had the unlucky chance to use.
If the ‘development’ you are referring to includes copyright theft, then a bit of discouragement is a good thing.
I read the original email and didn’t see anything wrong with it, it was a public announcement that a public piece of code was a copyright violation and a request that something be done about it. The only antagonistic statements that I read came from Theo.
I don’t think anyone would be any less discouraged from such a public ousting.
I’m afraid he should have considered carefully beforehand when mixing two codebases – of differing licenses. It should have been pretty obvious. Whether he is discouraged from working on it is neither here nor there.
There is and was no need for a public debate over this.
The only reason I can see for the GPL people posting this in public right away instead of starting off privately to resolve it is their usual anti-BSD and/or anti-everything-that-is-not-GPL opinions and jumped on the opportunity to put something BSD related in bad light.
If, after initial private communication, the issue does not look like it will be resolved, THEN it might be time to bring it to public.
The code was public, therefore there should have been public disclosure of a violation. Simple.
They could still have discussed AND resolved it in private before anything was said in public.
Why?
What’s the difference?
Marcus wrote public code. He did a public act this way. He should have expected public consequences, both good and bad.
If he wanted private mails, he could have wrote his code privately and sent it to the bcm43xx guys for review.
What do you possibly gain by starting a complaint like this in public? Nothing at all. Except being able to spit on *BSD in general.
Talking it over in private does NOT exclude a public statement AFTER it’s been resolved.
Also the code was in the public cvs and had been committed/recommitted quite a few times.
I believe the public disclosure was more for the parties who incorporate BSD code into their private projects.
Michael did offer to relicense some parts of the driver under the BSD license so that the BSD projects could implement it, but the developer just decided to stop working on it and Theo blames the linux devs rather than find an amicable solution between each other.
I believe Michael could have been a bit more diplomatic, but then again Theo is almost never diplomatic as is evidenced here.
I’m not a fan of the GPL, but I’ve been following the development of the bcm43xx for a long time (I wrote a HOWTO for 2.6.15), and aside from a few isolated incidents, I’ve always been impressed with the development process of Michael & crew. Reading the email list and the responses, I cannot help but be shocked and appalled by the behaviour and pettiness of Theo. I knew that he had a reputation for high-handed drama, but I had always retained a grat deal of respect for him for running the OpenBSD project. No attempt is made by Theo or the others to work with the bcm43xx developers, after it has been clearly shown that they are in the wrong. GPL code has been copied verbatim and placed under the BSD liscence. If the code was being ‘borrowed’, to be used as a placeholder while development on bcw went ahead, ok, but Marcus should have contacted the bcm43xx crew first before comitting it to the OpenBSD tree. When Michael brought this to light, Theo complained that it was done too publicly, (seemingly intentionally) taking much of Michael’s message out of context, resorting to ad hominem attacks, and eventually shutting down bcw development, blaming bcm43xx for the whole issue when it was his own developer that had created the problem in the first place!
Let’s look at this from another angle – what if Sun took some bcm43xx and placed it under their CDDL liscence? What would have been the community reaction then? Are we just whining because we see OpenBSD as ‘the underdog’? Even underdogs have to play by the rules.
If you bother to read the list, as I have, you’d see that someone made a mistake of copying. No doubt thats somewhat of a big blunder to make, copying GPL code, I won’t even begin to argue that one. Theo tends to exasperate on the mailing lists no doubt.
The big point he made though, which I agree with, the public ousting should’ve gone private first then public. It appears the developers of the BCM linux driver had no intention of firing off a curteous warning shot. No, they as Theo put it, went right for the jugular of public humiliation.
Should all be forgiven for what appears to be blatent copying and relicensing? No, of course not.
Should this have been in the public view without a curteous warning? I think no, others may differ on that. It could’ve easily been resolved privately before any of this hoobla ever hit the street
I find the “copied by mistake” arguement to be total bullshit, that simply shouldnt happen. That is blatant lie first of all…
“No, they as Theo put it, went right for the jugular of public humiliation.” i find this bit humorous, the pot calling the kettle black. Doesnt make it any more right, but lets not pretend that theo would have been level headed about it, as another poster said, fire and brim stone would have been involved and i believe it.
Actually I’d say it’s pretty plausible the “copied in to CVS by mistake” reason is legitimate. Working on code like a driver, Marcus may have simply used some of the working functions in his personal home development to test out his own functions. Mistakenly he may have added the wrong file or commited the file without much thought simply thinking “someone else should test out what I have.” To repeat what Theo and Marcus (the developer) in the list already agreed, they did in fact make a mistake and copy the code and would remove it. The real issue is more so the public way in which this was done.
lets think how Theo would react if it were the other way around.
I imagine brimstone and hellfire, myself.
What’s your point? GPL using developers take BSD code and relicense it all the time. Hell, if you actually read the email thread you would see the jackass that started this mess flaunting that fact at one point. There is certainly no hellfire and brimstone.
“GPL using developers take BSD code and relicense it all the time.”
The BSDL allows this procedure explicitely (“Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.”), including proprtietarisation and forking, while the GPL does not allow it without contributing back. So it’s okay what GPL developers do.
What is BSDL can get something else (GPL, proprietary).
What is GPL will stay GPL.
What is BSDL can get something else (GPL, proprietary).
BULLSHIT
http://www.opensource.org/licenses/bsd-license.php
Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
It is still proprietary, since you are not allowed to modify the derivative work when licensed under a proprietary license.
BSD and MIT allows for sublicensing essentially resulting in some derivative works being GPL’ed and some derivative works being proprietary. The latter part is controversial for some, and the first part is controversial for others.
What we often forget is that there’s nothing in the BSD license that guarantees the availability of corresponding source code. You can distribute a modification in binary form only regardless of whether you want to retain the BSD license or would rather relicense it altogether. The BSD license is just as suitable a license for proprietary “freeware” as it is for open source software. Often folks claim that GPL/BSD is an issue of free software vs. open source software. But BSD is not even an open source license by my interpretation (although the OSI might disagree). It’s just a really permissive license that is often used by open source projects.
It’s not even a very good license for what it’s intending to accomplish. Many developers and vendors that seriously consider the available ultra-permissive non-copyleft, non-share-alike licenses seem to favor the Apache version 2 license.
Maybe I need to explain more verbose. Because you’re getting offensive, it seems you did not understand my statement, which may be caused by the fact that english is not my native language. So let me give an example.
Let’s have a distribution A consisting of the parts Q, W and E. All of them are under BSD license. A developer uses them as a base for his work and derives his own Q’ and W’ from Q and W, he adds T and Z which he has developed on his own. His final product is called U and is released under a proprietary license. The developer ensures the BSD license to be included in U, as well as he mentions that his Q’ and W’ are based on Q and W, originally programmed by the respective authors (names follow). As far as I understood, the developer now may release U under the terms he wants, such as T and Z where he holds the definite copyright.
A = (Q, W, E) -> U = (Q’, W’, T, Z)
BSDL(Q), BSDL(W), BSDL(E) -> BSDL(A)
BSDL(Q’), BSDL(W’), PL(T), PL(Z) -> PL(U)
but also:
PL(Q’), PL(W’), PL(T), PL(Z) -> PL(U)
Please note: Q, W, E and A won’t be affected at all. They stay BSDL.
Let’s now assume A not being BSDL, but GPL. The rest of the conditions are assumed to stay the same. Now the developer may not release Q’ and W’ under any other license than the GPL. Because his final product U contains GPL parts Q’ and W’ next to his own work T and Z, the whole product U is GPL now. Furthermore, he needs to contribute back Q’ and W’ to the original Q and W.
A = (Q, W, E) -> U = (Q’, W’, T, Z)
GPL(Q), GPL(W), GPL(E) -> GPL(A)
GPL(Q’), GPL(W’), PL(T), PL(Z) -> GPL(U)
Here, Q, W, E and A won’t be affected at all. They stay GPL, of course.
Please correct me if I’m wrong. Please explain clearly, backed up by facts and examples (if needed or useful), don’t scream around nothing else but “BULLSHIT”, that’s plain impolite.
Personally, I like the idea of the GPL. The community and the developers benefit from the works other developers do on basis of existing code. So everyone benefits from the improvement made by one. No one may take GPL code, cut out the license and the names of the developers and sell it with a new sticker on the box. But the BSDL allows this under special circumstances, as tried to explain above.
What are blathering about?
I read the email thread.
Read my comment again, and see if you can figure out the big words.
Marcus Glocker used “simple” functions as place holders in his development, Maybe accidentally importing these into the CVS..
But the driver in question DOES NOT EVEN WORK YET, Michael Buesch should of contacted Marcus privately and asked him to either credit him or re-write the functions of similar design.
It is neither illegal or immoral to study GPL code when working on something under BSD/ISC etc, Copying might be wrong.. but the bcm43xx crew should not of made this such a public humiliation of Marcus Clocker, He is a good developer and DID write a majority of the OpenBSD bcw driver.. investing hundreds of hours of his personal time.. (as a hobby..)
Now because of this public attack the driver was deleted from CVS.
Seriously, You GPL zealots really piss me off.. Theo is definitely right!!
Because of a few similar functions.. Apparently copyrighted whitespace and variables names, The code and any hope of bcw on OpenBSD is lost.
Michael could of contacted Marcus privately and the driver could of been cleaned up, and BOTH projects could of continued.. But Michael was obviously out to personally attack this developer, and embarrass him publicly.
Feel proud everyone.. You totally made someones many months of work on a “free” bcw driver useless.
Edited 2007-04-06 15:16
Surely the code can be re-entered after its been cleaned up? They should have been a bit more subtle than what they did. But dont blame the GPL or people who support it. A**holes are part of every group. We have ours, you have yours. But I am sure they dont represent the majority of people in either group. This is really unfortuate if this was all over a non-intentional infringement.
But the driver in question DOES NOT EVEN WORK YET
You can’t look at things in terms of whether they work or not. We’re talking about code distribution, and it’s incredibly stupid that Theo thinks that the GPL doesn’t apply if something doesn’t work.
It is neither illegal or immoral to study GPL code when working on something under BSD/ISC etc
It is not illegal at all to do things with GPLed software and code if you don’t distribute it. However, Marcus did distribute it via a public repository, and you just can’t do that.
He is a good developer and DID write a majority of the OpenBSD bcw driver.. investing hundreds of hours of his personal time.. (as a hobby..)
Then he should have invested that time wisely by making absolutely sure where his code was coming from, and that he could legally integrate it – as hundreds of open source developers do on mailing lists everywhere.
Seriously, You GPL zealots really piss me off.. Theo is definitely right!!
There’s nothing zealous about it. You have to look very carefully into the licenses you use, and where your code comes from. Theo de Radtt is most certainly not right, but it’s difficult to see what he’s ranting on about because very little of what he’s wrote is relevant to the discussion.
But Michael was obviously out to personally attack this developer, and embarrass him publicly.
Nope. Marcus knowingly (or unknowningly) lytook GPLed code, put it into his driver – and then committed it to a public repository, which is the crucial part. He may have known about it, he may not have, but when Marcus used GPLed code it should have been immediately obvious to him that he needed to get clarification, or take it out and write something new.
Feel proud everyone.. You totally made someones many months of work on a “free” bcw driver useless.
That’s Marcus Glocker’s problem I’m afraid. Developers in open source projects everywhere get license and code issues sorted out first (dual licensing was even discussed, and nothing happened on the BSD side), and then they put the effort in when they know where they stand.
Edited 2007-04-06 15:52
As both a Slackware Linux and OpenBSD user I am not to take sides with any of the parties.
I would like my hardware to be both supported on GNU/Linux and on *BSD. The fact that the minority OpenBSD development community cannot reuse GPL source code while it is possible the other way puts them at a natural disadvantage.
Shouldn’t these kinds of license issues be prevented from happening in the future by having common specification documents from where each kernel community creates its own drivers?
There is nothing wrong with licenses that have to be complied with. But at least specs should be created that don’t compel driver developers to peeking at another OS’ source code.
I have found that there is a certain arrogant part of the Linux kernel community that doesn’t even want to share documentation for creating such drivers.
They forget that it is not about furthering GNU/Linux at the cost of other free operating systems but to gain market share from proprietary OSes.
Hi, I’m using Slackware (my private OS of choice) and some OpenBSD boxes mostly at work myself
If I interpret this message in the thread correctly [1], then the specifications from the reverse engineers are open for everybody to work with[2][3] and the reverse engineering team has even stated, that they would be happy to discuss /clarify the specifications with developers from the BSD camp.
As others have said before, the effort that the bcm43xx
developers went through (complete clean room reimplementation, separate reverse engineering team)[4] is impressive and should act as a model for other projects too. And better sharing of specifications is certainly something the teams should work on, but this was hardly the problem here, where the bcm43xx reverse eng. team seems to have been very open minded.
[1]http://thread.gmane.org/gmane.linux.kernel.wireless.general/1558/fo…
[2]http://bcm-specs.sipsolutions.net/
[3]http://bcm-v4.sipsolutions.net/
[4]http://bcm-specs.sipsolutions.net/ReverseEngineeringProcess
Ok, the reply button obviously works for some people, but not for me. Damnit.
It would have been brought into the public light soon enough. Bring the issue up in private and try to get it fixed. If it can’t be fixed, then out him. If it is being worked on it could easily be announced as a win for both sides.
I was reading through all this and a thought came to me. Free Speech isn’t really free and open source isn’t really open. What I mean by is that as long as you have licenses, patents, and copyrights even free (as in freedom) and open software or hardware is going to have restrictions.
In a Utopian society BSD could use any GPL’d code without restriction because their would be no need for licenses that direct how the code can be used. All would share and all would benefit. It would be true software communism. (Remember that communism in it’s purest form isn’t a bad thing.)
Unfortunately we don’t live in that world. We live in the dirty and often shady real world were patents, copyrights, and licenses are a real necessity. With that said I have to agree with other that say BSD should not take code and relicense it without permission (if they in fact did) and the original developers shouldn’t take to the public forums until they have exhausted all attempts at a resolution with the offending party.
That’s life though and like electricity we all tend to take what we believe is the path of least resistance.
Anyone who wants this whole situation summed up in two mails, read this one from Michael Buesch:
http://thread.gmane.org/gmane.linux.kernel.wireless.general/1558/fo…
If Marcus (and Theo) were going to use existing GPLed code, he needed to get relicensing of it sorted out – something that developers think about all the time in various open source projects. He didn’t do that, and distributed GPLed code under a BSD license.
Theo, responded with this:
http://thread.gmane.org/gmane.linux.kernel.wireless.general/1558/fo…
Personally, I think it is far better this was made known and sorted out, rather than Marcus wasting even more time and it becoming an even bigger issue in the future.
Anyone who wants this whole situation summed up in two mails, read this one from Michael Buesch:
http://thread.gmane.org/gmane.linux.kernel.wireless.general/1558/fo…..
Drat. Use of frames on web sites………..
First e-mail from Michael Busch:
http://article.gmane.org/gmane.linux.kernel.wireless.general/1602
Second e-mail from Theo:
http://article.gmane.org/gmane.linux.kernel.wireless.general/1604
….is to have as little as possible in common with GPL, FSF and Linux community.
Really? I would think that the best thing to do is not commit code into cvs that isn’t yours.
Again someone who obviously didn’t read the thread. All the accuser could come up with were variable names and whitespace. That’s stretching the meaning of copyright IMO. If I write a copyrighted perl script and use $_ and @ARGV, should I then humiliate publicly anyone who dares use these again? Like the GPL zealots love to say “Show me the code!”
I read part of the thread, until Theo’s ranting and whining got the better of me and I just tuned out.
I did read far enough to see where there were not only variables copied, but it also appear very specific functions were copied that were not part of the specification, but part of the linux driver itself.
Further, I believe there were even comments copied verbatim.
Unlike the SCO incident, the linux bcm43xx driver guys have actually shown what they think is direct copying. Yet, in their message, they made it clear they SUSPECTED it was literal copied.
The original message looked very well thought out and attempted to be neutral and non-confrontational as much as possible. There weren’t any threats, veiled or otherwise. They left the door WIDE OPEN for a sane response to come back and either confirm yes it was copied, or no it wasn’t literal copying because of a. b. and c. In fact, either type of reply would probably have ended the issue or laid the framework for a solution agreeable to both sides.
Did Theo and the OpenBSD people react appropriately, or dare I say it, sanely? No. The OpenBSD guys got caught and called out, and they didn’t like it. Instead, they do what many guilty people when caught: distract from the real issue and redirect blame. By changing the argument from copying to making much ado about the public communication, they have distracted from the real issue and attempted to point the fingers back at the bcm43xx people. In my opinion, this level of distraction seems to imply that _someone_, and I don’t know who, knew there was something fishy about this code.
As I read it, many of those included on the CC list had a reasonable expectation to be there. They worked on some aspect of the project on one side or the other. It looked like reasonable lists were copied, so of course the message went to a wider audience.
Assume that the GPL violation was true. There has been a public commit of GPL code under a BSD license in a completely different project, and short of the driver it self, an unrelated project. This was done in full view of the world as part of the open process.
At this point outside observers can see new code under a BSD license, without knowing the origins. Perhaps even Broadcom themselves. If this matter were handled completely in private, then those outside observers that saw the code enter in public view might not gain the benefit of knowing the disputed origins of the code involved. Furthermore, less scrupulous might attempt to check out previous revisions (before this came to light) in an attempt to use the “BSD” version of this code and claim it was an honest mistake.
Making this public in the manner it was done seems perfectly reasonable given the circumstances, and it makes it much more difficult for someone to later claim they didn’t know the code was not supposed to be released under the BSD license at the time. Nevermind that these are open projects where issues generally will be ironed out in open forums.
One other thing that I’d like to bring up. How exactly do you define release? This driver has never been a part of an OpenBSD release, doesn’t even work, and is under development. Granted it was put into a public CVS repository, but in my mind that does not constitute a release. The CVS repo is where the devs do development. It just so happens to be publicly accessible as a nice benefit to users.
It may not be an official release, but according to my layman knowledge (inserting IANAL and all the other disclaimers here … done) incorporating code samples into code that goes to a public accessible CVS server should qualify as distribtion.
And the GPL is crystal clear about distributing derived works.
In the light of [1]
and to a lesser extend [2] and [3], I would guess, that an internal revision control system for the (Open)BSD project without access to the general public would have prevented this situation, if it really has arrived by accident.
EDIT: I don’t suggest, that using an NDA to exploit the secnario in [3] should be a viable path for a FOSS project. But it is one method, if sealing of the CVS /SVN server from the public is not feasable / practicable either.
But such a construction would be very awkward for an open source project, as in the FOSS community development and distribution go hand in hand.
Feel free to correct me, if I’m wrong! Thanks in advance
[1]http://www.fsf.org/licensing/licenses/gpl-faq.html#InternalDistribu…
[2]http://www.fsf.org/licensing/licenses/gpl-faq.html#TOCDoesTheGPLAll…
[3]http://www.fsf.org/licensing/licenses/gpl-faq.html#DevelopChangesUn…
Edited 2007-04-06 18:42
Mr Michael Buesch you are a thief by breaking and circumventing locks and reverse engineering the GPL driver. YOU f–kING HAVE THE GALL TO BITCH AT BSD developers?
Why can’t you run BCM drivers under ndiswrappers and just wait for Broadcom to provide official support? Your open source drivers don’t help anybody maintain it because you are the ONLY person who knows how the device works because of reverse engineering.
LISTEN YOU DIPSHIT! Open source is about people understanding code so that IF YOU GOT HIT BY A BUS, someone else could maintain the code.
Here and it’s the lesson. Frankly, wrong or not wrong here is obviously a question of debate. But the way handling stuff is just outraging.
Going public instead of private is really messed up. I agree with Theo all the way, and I have to say that removing the code from CVS straight away shows how committed Theo and others are to upholding their believes.
This is why I’m happy to regularly buy the OpenBSD CDs and would never ever consider buying or supporting anything GPLed financially at any occassion ever.
That said, I wouldn’t mind a clause added to the OpenBSD license saying code may be used to anything except for GPLed projects.
Going public instead of private is really messed up.
So why did he publicly write code?
You write code in your room. Fine, it’s your private affair. You PUBLISH it. Now it’s a public affair, and it’s perfectly fine discussing it publicly. Maybe I wouldn’t have done it, but I think it’s not disrespectful to do it.
I’m 100% sure that Marcus was in good faith, but the bcm43xx guys were behaving perfectly fine, and I wouldn’t ever had been offended by them, if I was in Marcus’ clothes. I would just have pulled off the offending driver from CVS and collaborated with bcm43xx for relicensing-rewriting. Instead of that Theo went amok, and Marcus was now forced to take sides.
So much for Theo’s “empathy”.
I hold Theo in great esteem, but here he behaved nothing but hysterically.
This is why I’m happy to regularly buy the OpenBSD CDs and would never ever consider buying or supporting anything GPLed financially at any occassion ever.
Call me when you can compile OpenBSD with a BSD-licensed C compiler.
Seriously: GPL and BSD should learn to cooperate, that’s true, on either side. I think GPL guys should allow more nicely for their code to merge with BSD code, and viceversa. No need for blood on either side.
(EDIT: Sorry, This is supposed to be Re: Before we chastise this guy…)
What is BSDL can get something else (GPL, proprietary).
What planet do you live on? You can’t convert BSD licenced code to GPL.
You can add BSD licenced code to a GPL project if you include it’s original licence, Extracting that code and using it an another BSD licenced or commercial project IS permitted as long as the person follows the licence.
Edited 2007-04-06 16:15
“You can’t convert BSD licenced code to GPL.”
BSD licensed code can appear in non-BSD licensed projects (such as GPL or proprietary contexts), as long as the original source is quoted and the BSD license is included. Of course you’re right, this won’t affect the original code which still is under BSD license.
“You can add BSD licenced code to a GPL project if you include it’s original licence, Extracting that code and using it an another BSD licenced or commercial project IS permitted as long as the person follows the licence.”
That’s correct, but please refer to the statement I replied to:
RE: Before we chastise this guy… By fsckit (2.74) on 2007-04-06 16:56:59 CET: “What’s your point? GPL using developers take BSD code and relicense it all the time. Hell, if you actually read the email thread you would see the jackass that started this mess flaunting that fact at one point. There is certainly no hellfire and brimstone.”
Feel free to comment the highlighted part.
One of the issues that the BSD crowd had with the early Linux crowd was the tendacy to take BSD code, file off the license, slap on a GPL and call it good. People did it all the time, and as far as I know all the folks that have done it have been hunted down and told the error of their ways and to restore the original license (and, if they want, add the GPL for their parts).
This was more complicated back in the days when the BSD license was 4 clause and not 2 or 3. The peksy advertising clause wasn’t GPL compatible, per Stallman.
Edited 2007-04-06 16:47
To see GPL guys afraid because they don’t want their code to be ‘steal’ when they just have stolen the code from the BSD guys …
http://undeadly.org/cgi?action=article&sid=20070406104008
BTW, I think I know the topic of the song of OpenBSD 4.1 …
Edited 2007-04-06 16:47
They haven’t stolen any code from the BSD guys. The BSD license explicitly allows for sublicensing as long as you follow the BSD license.
Microsoft also uses BSD-licensed code including code from ESR and the GNU project.
“Stealing” is something you do when you violate the copyright. The “GPL guys” (whoever they are???) haven’t violated the BSD license.
SCO/Caldera however… that company has violated the BSD license (by removing copyright notices).
How is using the license exactly as the original developers intended others to “stealing”
I thought the BSD was designed to to allow as many people to use their code as they want with a few minor restriction. But if they expect others to give back because they will consider it stealing if they dont, well isnt that exactly what the GPL already is?
The GPL is designed for one purpose, the BSD is for another purpose. GPL utilizing BSD code is exactly why the BSD is created, to allow others to use it. If they feel slighted for people doing that, maybe they dont want to develop under the BSD license (which is just fine, a person should pick a license for what they intend others to do with it)
In the past, some people have suggested I am a bit lacking in empathy
but my god, compared to you I am a pinacle of human understanding.
Theo de Raadt
The guy makes a mistake (and I believe it a mistake) in the commits. Instead of being contacted, he is blasted (and I read all the threads). The driver is deleted (so there is no infringement). Both sides act poorly (still) and now both sides don’t like each other. Nobody wins.
All because a simple thing wasn’t done. A simple email contacting the developer “Hey, did you know?” and it would have been taken care of, no hard feelings, no poor public displays by anyone.
After reading the threads, I have to side with the BSD side on this.
Yes, they should’ve emailed the guy publically and seen if he could clean things up. But Theo’s ranting is just so over the top. He is the classic over-reactor. After reading the thread, all I can say about Theo is “what an A-hole!”
It seems to me like the code was DIRECTLY copied and then they sent warning along with invitation to help fix it. *Maybe* they sent it too publically. But either way, Theo is the one who should be embarassed by his reactions.
Edited 2007-04-06 18:27 UTC
Like Linus and his interface-nazis? Get serious …
Richard Stallman: “Software should not be owned”
Michael Buesch: “Obey MY rules or suffer!”
HARDLY! Michael Buesh is saying that they took GPL’ed software and relicensed it under a more liberal license. The point Theo is missing is that for all this nonsense, he’s actually making it MUCH less attractive to license something under the GPL (or any open source license) for fear that someone will do something like this (intentionally or not).
If this were MY code and someone copied and pasted it into a BSD’ed project AND COMMITTED IT, I’d be pissed too. Theo is wrong and he knows it, and that’s why he’s doing this dog and pony show about how it was “inhumane” to expose the infraction on a public list.
Whatever. The BSD guys ripped off the code, and rather than fix it peacefully without anyone noticing, Theo has inspired headlines like this: “Open Source coders caught stealing Open Source code”
http://www.theinquirer.net/default.aspx?article=38746
Way to go, de Raadt. Shoulda kept your fat mouth shut and let the comitter fess up and move on.
That’s quite a misquoting.
RMS’ view is that it is okay to own code. That’s not the issue. The problem is not ownership but lack of freedom to run, study & adapt, redistribute the program – and improve the program and distribute it.
As a number of people have already pointed out, the GPL is as draconian as they come. BSD licenses give you real freedom, truly open source style.
Theo’s argument seems a little flawed though if he thinks because it doesn’t work yet the GPL doesn’t apply.
This raises the value of all drivers being truly open source BSD style so all can benefit. Why waste countless hours re-inventing a (square) wheel for each OS when the whole bunch could work together for rounder wheels?
Dan Farrell makes an interesting post here.. Read it.
http://article.gmane.org/gmane.os.openbsd.misc/121406
He is right, Theo is right, Denying it is idiotic.
Edited 2007-04-06 20:51
Dan Farrell makes an interesting post here.. Read it.
Theo makes the point that Buesch and Co. are treating Marcus like a
thief.
There’s that word thief again. It was never stated and never even implied. People chose to read that into the mailposting.
That clearly exposes that Buesch thinks Marcus didn’t make a ‘mistake’, but did something on purpose… like a
No. He just finds it a bit daft that people chose to use GPLed code and license it under a BSD license without having a long hard think and talk about it first. It’s damn obvious.
Buesch cc’d a large part of the community on his initial email, which itself was rude. Why?
Because it’s a clear public violation, which wasn’t discussed or hinted at by anyone on the OpenBSD side before the code was publicly distributed.
Because Buesch clearly thinks he and his driver are more important than treating Marcus like a human being
Theo and the rest of them are just clutching at straws by trying to turn this into a personal thing. That’s not what’s at issue. Neither Marcus or anyone else thought at all about the code they were using, and it needs to be resolved.
This is Buesch’s big moment! His 15 minutes of fame! And unfortunately, it worked.
No. This has become a big thing because of Theo’s reaction. It could have been solved in less than fifteen minutes otherwise.
That, to me, means Buesch is an asshole. He made more of a name for himself in this bullshit fight….He did it at the expense of Marcus, he did it
on purpose
Bugger. Did you say he had a good point to make?
Theo gains nothing from this except the understanding amongst OBSD devs that anyone that develops for OBSD won’t be left in the cold when this shit happens.
On the other hand, he could just resolve this whole thing in no time by getting Marcus to discuss it and get the code relicensed – as Michael Busch has offered. In future, instead of mouthing off he could also encourage OpenBSD developers to be more careful about relicensing other peoples’ code rather than being stupid and taking people miles away from the point.
In future, instead of mouthing off he could also encourage OpenBSD developers to be more careful about relicensing other peoples’ code rather than being stupid and taking people miles away from the point.
How many times we have to tell people like you that only code that is “stolen” was variables and whitespace. Leaving original(non-working) code for reference is normal practice.
GPL is viral, unusable, restrictive and should be banned from any self-respecting free-open-source project.
Sure. If you don’t mind Microsoft walking away with your idea and with no responsibility to the original project. Sorry but I’d rather have people give back to the original project rather than just takes the jewels and run. Viral or not this is the GPL’s purpose to stop people from closing the code and not giving back. Its freedom for the code, if not the author.
It’s “freedom” for the simple minded. Real freedom is freedom without any trade-offs.
It’s “freedom” for the simple minded. Real freedom is freedom without any trade-offs.
If you feel like that. However, Linux is used as much as it is because everyone knows where they are with it. Everyone, including competitors, know where they are, everyone contributes on a level playing field with no hiding places, there is no keeping certain code hidden for perceived competitive advantage and there are no binary blobs that people can attach and distribute (directly) that cause problems.
If Linux was BSD licensed then you’d see companies everywhere sticking closed extensions on to it, most of which would duplicate each other and cause problems when run together, and all we’d be left with would be an open sourced nucleus of code that would do absolutely nothing by itself.
The reason Linux has kept its integrity is in large part thanks to the GPL.
If Linux was BSD licensed then you’d see companies everywhere sticking closed extensions on to it, most of which would duplicate each other and cause problems when run together, and all we’d be left with would be an open sourced nucleus of code that would do absolutely nothing by itself.
sorry to break it to you, but leaving out the “closed” part in your sentence I would say that you’ve explained exactly how Linux and BSD differs. BSDs in having ONE working solution and Linux having zillions (some working, some not), making it hopeless to have something universal working across all distros etc. Closed or not, the hazzle is there.
While at it.
The reason Linux has kept its integrity is in large part thanks to the GPL.
I wouldn’t consider Linux having kept it’s integrity. It’s scattered in philosophy, code base, distros you name it!
How many times we have to tell people like you that only code that is “stolen”…
Please stop inserting words into the argument. Nowhere have the words ‘stolen’ or ‘thief’ been used anywhere, so just stop using them as some sort of silly comeback on the terms of the GPL.
…was variables and whitespace. Leaving original(non-working) code for reference is normal practice.
Variable and whitespace? What on Earth does that mean? You’ll find that a huge amount of the code is a 1:1 match. How many times are you going to be told that that angle that Theo came up with is straw clutching, and irrelevant? Leaving original code in, whether working or not, relicensing and distributing it in a public repository is a violation, and a slightly silly one. End.
If you merely want to test out the code for research purposes, then don’t distribute it. Problem solved.
GPL is viral, unusable, restrictive and should be banned from any self-respecting free-open-source project.
What a load of rubbish. Again, this is a totally irrelevant rant. The GPL is the GPL, a license is a license, people know what it means and if you take code from anywhere else and distribute it then you need to make sure of what you’re doing – GPL or not.
“On the other hand, he could just resolve this whole thing in no time by getting Marcus to discuss it and get the code relicensed – as Michael Busch has offered. In future, instead of mouthing off he could also encourage OpenBSD developers to be more careful about relicensing other peoples’ code rather than being stupid and taking people miles away from the point.”
You guys still don’t quite understand it. Nobody on the BSD side wanted the damn code for distribution purposes. What code was used from the Linux driver was only there for placeholder purposes. The code was also not distributed in any release. A CVS repository is where developers keep the f–king code they’re working on. It is not and has never been considered an official distribution channel.
It most certainly is a distribution channel if others can download the code from it.
Adam
CVS is a distribution channel.
Marc Balmer clears more things up, Definitely a good read.
http://article.gmane.org/gmane.os.openbsd.misc/121329
Posted Below:
Marcus Glocker, mglocker <at> openbsd.org, knows a big deal about wireless LANs. He has been involved in many of our wirelesss driver, he has also written applications for wireless applications like rtunes. He wrote the nostromo webserver. He is certainly the person who knows how to write original code.
When it comes to bcw, a piece of hardware for that no documentation exists, he decided to use the docs the linux folks have.
He began a rewrite of a bcw driver, inspired by the work of the linux folks. His driver was not working yet, to give him a headstart, he used some code of the linux folks with the clear intent to replace it with
his own. Just to make sure this shit works.
To ease his work, and to let others in our group to step in in his efforts, he committed it to our work area which we call cvs.
The linux folks tooks this as the grounds to ride attacks agains Marcus, claiming license violations.
Marcus, devoting his spare time to OpenBSD decided that this is kindergarten and best left to the Linux amateurs and deleted his driver from the OpenBSD cvs tree.
Now everyone has won, the Linux people, Broadcom and the OpenBSD users.
Thank you, Linux BCW developers!
Edited 2007-04-06 21:24
Um… Am I the only one who sees the obvious here? Its clearly stated in this quote that he took the code from the linux driver, ofcourse that is not in contention, but he knowingly distributed the code and it says so here. I think its the openbsd team thats acting childish and I’ve read the whole thread that started this mess and this Theo character is the type of person people use as examples of how bad the OSS community is. At one point Theo gloats about that whole issue with the Atheros drivers, and now he sobbing about how inhumane the linux bcw developers are. Give me a break!
Lol, this is classic ‘look what you made me do’-behaviour. As I child, on a few occasions I would destroy a toy or rip apart a comic book of mine to ‘punish’ my mom. Of course, my mom didn’t care, and I soon realized I was only punishing myself.
Linux kiddies, they don’t know anything about pro development. Real developers are using cvs and you do need a base for development – this is the usual way in opensource. This so-called violation is a flamewar toward OpenBSD, it was no release stuff, it was a working base only.
BSD should be using the original BSD license, because it’s incompatible with this “closed-source”-GPL-nonsense. The opensource-community would be better without the GPL-zealots! Fanatics!
First off, when you distribute code it doesn’t matter to who, you are still publicly posting code you have no right to. It doesn’t matter if it was working or not, the fact is the code was taken copied and redistributed under another license, regardless of who it was intended for or if the code was just a placeholder. The PL is pretty explicit about these things. Like you said it was put on cvs. What is cvs? Hmmm, I don’t know a way for others to gt access to your code. What does that mean? It means that the code was distributed. The issue is not that it was distributed but that the code was redistributed under another license that the original author didn’t use for specific reasons. I really don’t see you argument here. The code was distributed and neither Marcus, Theo, or any of the OpenBSD team have denied this. Was it a mistake, I happen to think it wasn’t but who knows.
They already said it was a “mistake”. Did you not read all the threads? Your whole post is not the point of the issue at all. The point is “how” it was brought to the attention of the developer. Theo has said he believes it was a mistake but the correct way to fix it wasn’t to blast the whole universe but to talk to the developer. That is common courtesy and it was severely lacking in this case.
He has been involved in many of our wirelesss driver, he has also written applications for wireless applications like rtunes. He wrote the nostromo webserver. He is certainly the person who knows how to write original code.
Then he should know to be careful about relicensing code published under the GPL, and get that sorted out first.
When it comes to bcw, a piece of hardware for that no documentation exists, he decided to use the docs the linux folks have.
Using docs to write code is absolutely fine.
His driver was not working yet, to give him a headstart, he used some code of the linux folks with the clear intent to replace it with his own.
You simply can’t do that, and he should know, if you’re going to put it in a public repository. You need to seek clarification, get the code properly relicensed and carry on working.
I also don’t understand why he needed to put GPLed code into a repository under a different license to check that it worked. It’s not going to help him, and there’s also the risk that some of it could have been left behind. Not a great way to work. It doesn’t make a lot of sense.
Marcus, devoting his spare time to OpenBSD decided that this is kindergarten and best left to the Linux amateurs and deleted his driver from the OpenBSD cvs tree.
Boo, hoo, hoo. That’s his problem. He didn’t need to do that at all, given that he had the option of getting the code properly relicensed. I also love the cheap ‘amateur’ shot.
100% agree. That’s the best explanation yet of how things went down. Like it or not the Linux devs were dead wrong in this deal.
Couldn’t disagree more. I don’t see how it’s a “mistake” that someone commits GPL’ed code they didn’t write to a BSD project. That shows lack of intelligence at the very best. It SHOULD have been reported as it was. The only legitimate complaint was that it COULD have been reported more privately, but I don’t see how the way it was done was wrong.
The code was improperly committed. If you think the way he reported it was crappy, it doesn’t change that fact.
BTW, Theo behaved like such a 4 year old that I can’t image any company ever taking him seriously or ever caring to work with him ever again.
while the “mistake” might be in bsd side, is there any need to make it public?
http://www.theinquirer.net/default.aspx?article=38746
what could be more offending than this when the matter can be solved privately?
and BTW, I guess that’s how Theo always behave…
I would say it wasn’t really “public.” It was on a mailing list, and no one would’ve given it a second thought until Theo opened up his big fat mouth and made it into something bigger.
After the initial email, the coder could’ve written back and said “I’m sorry, I mistakenly committed code I shouldn’t have, and I will fix it.”
Instead, Theo chimed in with his usual tact and said “You’re calling him a thief!” and started a massive flamewar, one that has left me with one feeling only: Theo is an immature sod. Now it’s a headline, and Theo is the only one to blame.
Edited 2007-04-07 03:09 UTC
This is actually the most offensive and inflammatory comment I’ve seen in this whole fiasco, and I’m pretty sure I’ve read nearly all of them, all across the net.
I would have voted it down, but was surprised to find that I can’t, because it’s coming from an OSN Staff member.
Oh, right. I’m sure. Because any comment that disagrees with your stance is “offensive and inflammatory.”
Actually, users CAN mod OSAlert Staff members. You could vote on my comment if you had earned enough trust on the site, except your peers haven’t judged *your* comments as worthy enough. So let that speak for itself.
The more I read this, the more it infuriates me. I want to understand the mentality of the open source zealot: if you steal my code – knowingly or not – and commit it into a public CVS tree that anyone else knows is BSD licensed and therefore could – at this very moment – be using it in a proprietary driver or application – am *I* wrong for calling it to your attention? Or am I just wrong if I “cc” too many people in my reply? What is the acceptable amount of people for the first notice?
Am I less right for calling you out if I’ve embarrassed you for your massive screw up?
This whole thing is absolutely RIDICULOUS, and the people who are ignoring blatant copyright infringement should not only be ashamed of themselves, they should not be trusted to ever touch, write, or release Linux (GPL), OpenOffice (LGPL), and most GNU apps. – they obviously have no respect for these licenses.
Edited 2007-04-07 13:46 UTC
I don’t think this is a matter of someone trying to steal someone else’s work and pass it off as their own.
This is a matter of someone using someone else’s code temporarily and pushing it off onto a CVS repository without thinking about the consequences (allow us all to re-read this statement multiple times, less we should forget it again).
am *I* wrong for calling it to your attention? Or am I just wrong if I “cc” too many people in my reply? What is the acceptable amount of people for the first notice?
Right and wrong can be argued to no end (as philosophers have been doing for eons), and there are so many varying shades of grey, however, I myself would have chosen personal (private) contact first, so that I could find out what was going on before potentially making an ass out of anyone (myself included) — and as far as I can see, Michael Buesch’s approach, only served to show the worst side of everyone, both those involved and uninvolved, himself included — although I doubt that was his original intention — and I wonder if any of us can look at this objectively enough to learn from it?
And in the end what have we really gained? Michael does not have his recognition, the OpenBSD driver was deleted, Marcus is publicly humiliated, and everyone has gotten needlessly worked up over spilt milk. Good lord — we’re talking about a wireless network driver here, in fact, only a portion of it, one that does not even work, yet we treat it as the oncoming of the Third Reich. Some fights are worth fighting, I am doubtful that this is one of them.
It seems to me the original issue could have been remedied, and the subsequent lynching prevented, had personal contact been made first. To accuse someone of stealing without first investigating and talking to them is a demonstration of a true black and white mentality — the very same mentality we see taking extremely polar view points in these comments. It is for these reasons that our respective countries have chosen judicial systems that separate the victim from the process, and presume innocence before guilt.
Let us not forget that this was a work in progress, and the fact that it didn’t work certainly provides evidence to support this claim. Marcus was not actively going around distributing the code pretending it was his — yes — it was on a CVS server, but as mentioned before, CVS servers require you to go out of your way to get anything off them. They are a passive system of distribution (you have to want to get something off them, know how to do it, and know what you want), so I think it’s easy to see how one can mistakenly commit something without thinking they’ve done anything wrong.
YES, proper recognition should have been given, even if the code was only going to be used temporarily, however, I don’t think that anyone here would want to be in Marcus’s position — and to be publicly labelled a thief, with no real investigation, trial, or conviction beforehand.
Well said, but I maintain that had Theo not initially over-reacted, it wouldn’t have garnered the attention it did. But I hear ya.
I agree. Certainly he should have quit after the first 30 or so posts in to the thread (as should have everyone else). Oh well, I suppose it’s over with now, kind of a shame how it turned out, but maybe someone will take something positive away from it — a good lesson of what not to do in the future.
The more I read this, the more it infuriates me. I want to understand the mentality of the open source zealot: if you steal my code – knowingly or not – and commit it into a public CVS tree that anyone else knows is BSD licensed and therefore could – at this very moment – be using it in a proprietary driver or application – am *I* wrong for calling it to your attention? Or am I just wrong if I “cc” too many people in my reply? What is the acceptable amount of people for the first notice?
Am I less right for calling you out if I’ve embarrassed you for your massive screw up?
For one thing, the code wasn’t “stolen”, it wasn’t attributed properly. And the idea that someone could close source it and use it is ridiculous. The driver isn’t even close to functioning yet.
The point is that yes, there was a screw up, but the way which the issue was raised was entirely inappropriate. This was an incomplete driver. It has not shipped with any oBSD releases. It was well known that the developers were going to use the linux driver as a reference (http://undeadly.org/cgi?action=article&sid=20061121194620)
A simple email directed to the developer and Theo would have fixed the issue rather quickly. Michael Buesch *KNEW* that cross posting that email would cause a fire storm. He is an absolute idiot if he thought otherwise.
I’m sure that’s true. But it was published on a public CVS. And just because it was incomplete doesn’t mean it was OK to publish it.
Maybe. A simple email certainly would have sufficed, but he chose another way. That might make him a dick, but it doesn’t make him wrong.
– On the subject of public disclosure
HAHAHA. Yah, I’m sure if it was discovered that microsoft had copied and pasted GPL’d code and accidently released it in Vista, we’d be up in arms if someone had the audacity to point it out in public.
– On the subject of being polite vs ethical.
Yah, when that guy tried to make off with the ladies purse in NY, then punched her when she tried to fight him off – it was really rude of them to put him on the news. After it all, it could have been a accident. (with about the same probability of this code being committed being an accident).
– On cutting to one’s soul
Being called a thief for copying and pasting some code from one open-source project to another, shouldn’t cut to ones soul. Especially if no lawyers are involved. (lawsuits can be scary). Watching people die, watching innocent people get really, truly hurt. That cuts to one’s soul.
Watching Theo making an ass out of himself as usual, and being a complete hypocrite to boot.
That’s entertainment.
*Hmm. But I can’t help wondering if there is a non-corporate edition of ssh that has been reviewed by people outside the theo bubble?
I read the whole thread, and about half of the useless comments on here — the fact that I’m writing this is probably a waste of my time as some hot head is only going to mod it down, but I’m going to write it anyway.
Let’s start with some questions here:
1. How many of you have written code before? I’m talking real code, not hello world.
2. How many of you have tried to piss around with CVS on a semi large project?
3. How many of you have tried to reverse engineer a driver before?
Okay, now given my feeble experience, I can tell you that it’s pretty easy to: a) temporarily use someone else’s code for something you plan to rewrite later on (especially while you’re still learning how something works); and b) start committing half-finished stuff to a CVS repository not thinking anyone’s going to see it.
Which leads to another question: 4. How many of you regularly sift through CVS repositories? — yeah, probably not many, you know why? because the vast majority of people that checkout from public CVS repositories (with the intention to look at code) are the very same guys who are committing to them. So yes, technically they are public, but effectively they’re not so public. Don’t believe me? Go find a public CVS server and see if you can figure out how to get anything useful out of it. Unless you already know what you’re doing you’re going to find it to be a fruitless task.
The fact of the matter is that some dude made the mistake of committing a bunch of work-in-progress code not thinking that anyone was going to see it and freak out about it. I highly doubt there was any intention to ‘steal’ code, as the OpenBSD guys are pretty anal about eliminating as much GPL code as possible (and I can only imagine this fiasco will make them all the more jaded toward it).
Should the code have been in CVS and not properly attributed as GPL code? No, probably not, and I bet had they just sent the guy an email asking for proper recognition he would have given it — and why wouldn’t he?
Why was it committed? Seems to make sense to me that it allows other people working on that particular chunk of code to see it and change it — MAYBE — even for the purpose of encouraging other developers to help replace the temporarily borrowed code.
Certainly, had I been the guy who had written the borrowed code, I would have at least made some kind of attempt to find out what was going on before jumping to crazy conclusions and creating a stupid soap opera in a public forum.
I don’t particularly agree with Theo’s response in it’s entirety, but it’s his developer, and it’s his prerogative (as project leader) to stand up for the guy.
Michael Buesch should have apologized after the initial misstep of making a public mockery of the guy (which was unnecessary), and it would have ended the whole preposterous thing before it had a chance to escalate to such a ridiculous level.
I too read the entire thread (I wonder how many of my brain cells died in that excercise in absurdity?) and I come away amazed by several things:
* That someone entrusted with leadership of a project as significant as BSD would engage in such infantile behavior on a public mailing list (or should my amazement be that someone given to such infantile behavior would be given leadership over such a significant project?). Theo had the demeanor of a 7-year old on a school playground, not an adult in a position of responsibility. Granted, there were infantile responses on both sides, but Theo kept surpassing all expectations in that regard.
* That others kept responding as though continued efforts would change his position or demeanor — or as though it really mattered.
* That I kept reading…
Not a glorious day for either camp, unfortunately.
To sum up all this nonsense- code is gone, GPL nazis are happy, OpenBSD devels got spanked, baby seal died unexpectedly.
MAKE LOVE NOT WAR.
I have read the majority of the comments on this “issue” both here and on Slashdot and I have come to the conclusion that companies looking to potentially work with F/OSS developers could be reluctant to cooperate or work with F/OSS developers based on their inability to handle simple problems without unnecessary publicity and flame wars, in other words act like adults.
I am sure that this has not escaped various people at Broadcom, who could use this as further justification not to open up their hardware and software to further F/OSS development.
We Linux people can’t stand to have other free operating system like FreeBSD or Open Solaris move in on our turf. We do all the reverse engineering and those tea-baggers make off with our hard earned work.
Edited 2007-04-07 20:21
If Openbsd indeed used some of the GPL code I’m sure it was an honest mistake and not done purposely to cause any kind of problems. I hope Openbsd and linux devs can work this out to make the open source community stronger.
Edited 2007-04-07 20:26
Again, the great thing about these licenses is that, as the author, you get to choose which license you want to use.
Who wants someone telling them: “No, you can’t use X license. You must use Y license. It has been decreed!”