Alan Cox talks about cooperation with hardware vendors, patent law, microkernels, and GPLv3: “I think [the MS-Novell deal] is a bad idea and that Novell are going to get stung by the GPLv3, and rightfully so. The license is designed to keep the software free, if it fails to do this then it needs fixing, so GPLv3 hopefully will fix this flaw.”
Nice if quick ‘n’ brief interview. I like the way he mentions Intel and JMicron as being helpful towards Linux whereas AMD/ATI are acting strangely at the moment.
In particular his remarks throw into yet sharper relief the software patent question. Here we have Microsoft trying to start up a patent protection racket whereas you’d think they’d be at the forefront of attempts to get US patent law substantially changed so that software is a copyright and not a patent issue. It’s Microsoft who have the billions and who are therefore at greatest risk from patent trolls.
Also like the way he points out that working with people you get on with sounds great but isn’t always the best idea. We all need the different angles and other views that come from folks we probably don’t get on with.
Not really, it’s the little guy who is at the biggest risk simply becuase a company like MS can afford a few lawsuits from patent trolls, the little guys usually just go out of business.
The patent trolls have business models built around the biggest return for minimal investment; when you’re trolling a company like MS it’s easy to find a high-powered law firm willing to take the case on contingency fees for a piece of the pie. They’re less likely to if you’re targetting a little guy with little capital, and if you have to pay your own bills then legal fees involved can outweigh any potential gain. Particularly since it would be difficult to prove damages owing from infringement by a small, independent ISV versus a large, well financed or well funded one with a much bigger market base.
So by that measure, the little guys are only at threat from the commercial patent holders who don’t want their business threated. The patent trolls only target the low-hanging fruit in the hopes for a quick payout for minimal effort.
For everything you have contributed to Linux. yes the Microsoft/Novell and Microsoft/Xandros deal will die with the release of GPL 3. I cant wait for the MS FUD to stop.
I find his answer to question 10 to be a refreshing change from the usual black and white view people often take on the topic:
—
10) Do you think that licensing Solaris under the GPLv3 might weaken the position of Linux (which is about to continue using the GPLV2)? Are you worried about the possibility of Linux developers leaving to work on Solaris?
Not greatly. In fact I would hope we will see sensible discussions about dual licensing bits of code so that we can all benefit from them.
—
Edited 2007-06-11 00:12
I still font fully understand why Linux will stick with GPLV2. At first, Linus had some valid points but most of his concerns have been fixed. Now what’s the problem? Just stubbornness I guess.
“””
Just stubbornness I guess.
“””
No. Not stubbornness. Legal necessity.
http://osnews.com/permalink.php?news_id=18024&comment_id=245411
If the kernel devs could whimsically switch the license from GPLv2 to GPLv3 tomorrow, what is to keep wealthy and influential parties from gaining control and switching to a closed source license next year?
This topic actually came up on LKML today. Here’s an email from Linus on the topic:
On Sun, 10 Jun 2007, Tarkan Erimer wrote:
> >
> > Last heard, Linus was quite impressed with the toned down version of
> > the final draft of GPLv3.
I was impressed in the sense that it was a hell of a lot better than the
disaster that were the earlier drafts.
I still think GPLv2 is simply the better license.
I consider dual-licensing unlikely (and technically quite hard), but at
least _possible_ in theory. I have yet to see any actual *reasons* for
licensing under the GPLv3, though. All I’ve heard are shrill voices about
“tivoization” (which I expressly think is ok) and panicked worries about
Novell-MS (which seems way overblown, and quite frankly, the argument
seems to not so much be about the Novell deal, as about an excuse to push
the GPLv3).
Linus
So basically, instead of people asking why Linus doesn’t intend to push for v3, they should really ask why should he. Nobody has really given a single, valid reason yet that doesn’t wrap itself in FSF dogma.
Although interestingly, in another email, Linus did mention that if Sun goes through with v3, that would at least create a point of consideration for migrating the license so that there wouldn’t be two major kernels with incompatible licensing. Pragmatic to the end. Wonder how Sun would feel about that?
Since you plainly discard the idea of user freedom as “FSF dogma”, you will never hear a “valid reason”. You left nothing to chance here, well done.
Apart from the bit about the beer, most of that interview went way over my head
The thing that bugs me most about GPLv3 though, is whether it is possible that the FSF might have outsmarted themselves.
What TiVo did may have been in compliance with the letter of the law regarding GPLv2, but was also definitely against the spirit and intent of the GPLv2. Exactly the same may be said for the Novell and similar deals. I can understand the FSF’s frustration and anger.
Are we sure that the GPLv3 doesn’t have any loopholes at all to exploit? Is it now battened down so tight as to restrict “normal” use? If Novell do get “bitten” will they not just drop their support for Linux? Are they too committed to be able to change at this point? Could they adopt BSD instead? Will the GPLv3 cause a mass migration of business away from GPL’d software? What then?
I’m concerned. I would appreciate if someone could alleviate that concern.
Edited 2007-06-11 09:50
>Are we sure that the GPLv3 doesn’t have any loopholes at all to exploit?
You can never be sure. Even the really good GPLv2 has loopholes like you see today with DRM, DMCA, software-idea-patents,…
That’s why every project should keep the option to switch to a new GPL or another license when such a loophole appears.
> Is it now battened down so tight as to restrict “normal” use?
No, because you can do whatever you want with GPLv3 software (that’s freedom 0). GPL has only restrictions if you distribute software, the restriction is that you can only distribute if you give the recipients the same 4 freedoms you had.
>If Novell do get “bitten” will they not just drop their support for Linux?
Maybe, Maybe not. But who cares? GNU/Linux and Free Software exists before Novell, it exists with Novell and it will exists after Novell (almost no company exists forever).
>Will the GPLv3 cause a mass migration of business away from GPL’d software?
Time will tell. But Free Software and GPL especially isn’t about business it is about freedom.
Personally i think copyleft licenses like the GPL are in 99% of all cases the only valid option for business. Because than they can be sure that they will profit from improvements by other and they don’t have to worry that a competitor takes their work and uses it in non-free software to compete with them.
Edited 2007-06-11 10:03
>>If Novell do get “bitten” will they not just drop their support for Linux?
>Maybe, Maybe not. But who cares? GNU/Linux and Free Software exists before Novell, it exists with Novell and it will exists after Novell (almost no company exists forever).
Thank you for taking the time to address my concerns.
On this last point though, I believe companies such as IBM and Novell have done a great deal towards improving the quality and depth of free software. It would be churlish to so casually dismiss their contributions.
I for one do care about what happens to those who publically admit their support and friendship for free software and the community, even though it is certain that their motivation was profit.
>I for one do care about what happens to those who publically admit their support and friendship for free software and the community
If they commit them self to Free Software than it should be no problem, because GPLv3 is a Free Software license.
So every company which wants to do Free Software shouldn’t have any problem with GPLv3. Because GPLv3 just requests from them what they want to do anyway.
> On this last point though, I believe companies such as IBM and Novell have done a great deal towards improving the quality and depth of free software. It would be churlish to so casually dismiss their contributions.
Heh.. Please reference anything Novell has contributed to free software. You will find it lacking
Even if you check IBMs record, they have commited a small number of very interesting projects but nothing critical.
As far as complete projects go, you’re more right than wrong. However, they contribute to many projects by paying developers to work on them. I’d consider that contributing.
When did you or the FSF become the sole arbiters of judging user freedom? The four freedoms are simply one interpretation. My user freedom is choice in the tools I use, I support FLOSS because I believe it is important in assuring my choices continue to exist, but I reserve the right to use nasty, evil proprietary software when I’ve got to get something done that I can’t realistically achieve with my OSS alternatives.
RMS on the other hand, has stated time and time again that he would much rather use sub-par free software than better capable proprietary options. Fine for him, but not for me. For my personal requirements, binding myself in philosophical chains and restricting my options and ability to work effectively is not user freedom. But that’s just my opinion.
The point I was making is that people harping on Linus to adopt v3 make their arguments from their own personal points of view, as you did above by choosing to define the concept of user freedom to match your own opinion. Linus has rejected FSF philosophy, and has always stated right from the beginning that the GPL was selected for practical reasons. So with that in mind, if people want Linus to seriously consider v3, they need to quit posturing around the four freedoms and put it into terms relative to what *he* is trying to achieve, not the FSF. So in other words, what are the practical benefits for the linux kernel devs to adopt v3? Right now, nothing concrete that anybody has been able to articulate.
Well, given Linus & Co’s corporate funding (and wanting to stay in good with the corporate fundees), I think that the Solaris kernel going to GPL v3 would be the best thing for everyone.
I honestly see no reason to continue to use Linux @ GPL v2 if I can use Solaris at GPL v3 (especially since many would consider the Solaris kernel technically better in many areas than the Linux kernel). GPL v3 is my preferred license of the 2, because it stops corporate bastardisation. Read between the lines guys, Linus does NOT have a problem with what Tivo did, and that really tells me that he’s in it for the money and has no morals. I’d rather RMS any day to be honest.
I’m sick and tired of people bowing down before Linus, when in reality, he’s done one thing for the open source community. RMS has done an awful lot more, for a much longer timeframe imho. People praise Linus (and he is deserving of praise, don’t get me wrong), and they bag RMS.
All of these corporate donations have been nothing more than trojan horses imho, Linux did not need them, not for the average user. OK, Linux in the enterprise might have needed these improvements, but call me greedy, I don’t care one iota about Linux in the enterprise. Linux was always the people’s kernel, but not anymore. It worries me that others simply either cannot see this fact, or cast a blind eye to it.
I’ve always liked Alan [Cox], he understands the fundamentals of free software, and why it’s important. You’ll also probably find that Alan has contributed more lines of code to the Linux kernel than Linus has, and was Linus’ right hand man for many years until that other person [Andrew Morton] stepped into the role, and imho, ballsed the whole development cycle up.
Dave
Linus does NOT have a problem with what Tivo did, and that really tells me that he’s in it for the money and has no morals.
Do you think Linus got paid by Tivo? Linus has plenty of morals, they just don’t align perfectly with those of RMS. Not agreeing with RMS is not the same as being immoral. Linus wants to make the best possible kernel technology available to as many people as possible. What is so bad about that?
Out of curiosity to you consider Theo de Raadt an immoral person who’s only in it for the money? He would also have had no problem with what Tivo did if they’d used OpenBSD.
>Linus wants to make the best possible kernel technology available to as many people as possible. What is so bad about that?
There is nothing bad about it, but this is a technical decision/goal and not a moral decision/goal.
For example if he would achieve this goal better by writing a non-free kernel would he than write a non-free kernel instead of a free one?
There is nothing bad about it, but this is a technical decision/goal and not a moral decision/goal.
I’m not sure I agree. I’d say it is a moral decision as much as a pragmatic one. I’m not sure how you draw such clean lines between the two.
For example if he would achieve this goal better by writing a non-free kernel would he than write a non-free kernel instead of a free one?
Meaningless question. Would RMS still have founded FSF if he could have achieved his goals better by doing something opposite?
>I’m not sure I agree. I’d say it is a moral decision as much as a pragmatic one. I’m not sure how you draw such clean lines between the two.
building the best technology is clearly a technical goal.
>Meaningless question. Would RMS still have founded FSF if he could have achieved his goals better by doing something opposite?
The goal of RMS is that all software users should life in freedom. Sure if he could achieve this better without the FSF he wouln’t found it and if he could achieve this goal better without the GNU system he would never had developed the GNU system.
That’s the difference. It’s not about a foundation or a technical solution, it’s about a moral question and he would do whatever is necessary to achieve this moral goal but this could never end in something immoral.
On the other hand if you have only a technical goal: writing the best technology. Will you than also do everything to achieve it? Also writing non-free software if this is the best or even the only way to write the best technology? This is an important question to judge if a person with this goal does something moral only by accident (because at the moment it is the best way to achieve the technical goal) or because he belief that this is the right thing to do.
Edited 2007-06-11 09:46
For example if he would achieve this goal better by writing a non-free kernel would he than write a non-free kernel instead of a free one?
Of course he would. Linus has nothing against writing closed-source code (which he did when he worked for Transmeta, I believe) or using closed-source software. Linus has made it clear many times over that he’s only concerned about the technical aspects of software development. It just happens to be the case that the open source development model works quite well for the Linux kernel and that’s pretty much the only reason, as far as Linus is concerned, why the Linux kernel is open source software.
I’m sure that Linus is a very moral person but he just doesn’t think that morals have any place in software development. He has expressed this opinion quite clearly in many, many interviews.
It was Linus’ decision to use the proprietary BitKeeper revision control tool in the Linux development and Alan Cox was one of those kernel developers who refused to use BitKeeper due to its non-free license.
If Theo is “just in it for the money” then I would say he’s doing a pretty poor job of achieving his goal!
OpenBSD has done more stand up against the nosense of hardware vendors than anyone. OpenSSH is one of the greatest contributions to free software I know of. All distributions, BSD and Linux benefit from it. All of this is done without any big revenue stream from the big corporations.
It is a shame you got modded up for a statement like that. Really!
Way to miss the analogous point. Theo leads a project that makes a widely popular fres software product that is used by some of the biggest commercial organizations in the world, and rejects FSF philosophy while doing so. The argument is that if one feels justified for calling Linus a corporate whore for doing so, would the argument apply to Theo?
Probably, since the v3 fundamentalists seem to believe that they alone hold moral authority on free software philosophy and that any and all dissenting opinions are simply wrong.
Rereading, I do understand you’re analogy a little better, however, it was not worded well. Also, this following line still seems a little out there:
I’m not sure that is a factual statement.
I wasn’t the original poster, just pointing out his intent.
As for the Tivo issue though, I think his choice of the BSD license is factual enough. Tivo with linux is a drop in the bucket when you consider openSSH is co-opted by companies like Cisco as part of their multibillion dollar network infrastructure product portfolio; the only complaint that Theo has ever raised over that is the fact that deep-pocketed Cisco won’t kick back donated funding in return, he’s never claimed they don’t have a moral right to use the code the way that they do.
Read between the lines guys, Linus does NOT have a problem with what Tivo did, and that really tells me that he’s in it for the money and has no morals.
Well, there’s no need to read between the lines, he says it straight to your face. As for the last part of your sentence, try taking a look at how much Linus gets paid and ask yourself if that is really all he could get. Hint: it’s nowhere close.
Linus does have morals, they just apparently don’t match up with yours. Which is perfectly all right, but now you’re bagging on Linus just like you complained others bagged on RMS.
Ok, but it is clear that Linus is not educated nor interested, as himself often claims, on the social side of the equation. His oppinions on interrelation between technology and society are worthless from the standpoint of a general technology user – that is somebody not trying to sell it.
It probably doesn’t hurt if I tell it again: not everything you get for free can be used to hurt the good guy who gave that thing to you. GPL ensures that and people who don’t get the economy model behind it or don’t agree with the GPL in any other way can simply ignore it. But disrespect of other people rights will not be ignored.
All of these corporate donations have been nothing more than trojan horses imho, Linux did not need them, not for the average user. OK, Linux in the enterprise might have needed these improvements, but call me greedy, I don’t care one iota about Linux in the enterprise. Linux was always the people’s kernel, but not anymore. It worries me that others simply either cannot see this fact, or cast a blind eye to it.
Hey, it’s free software. Go fork the kernel and remove “corporate trojan horses”. I wouldn’t be in favor of corporate control of the kernel, but there is nothing wrong with corporate involvement and contributions. After all, many good improvements, including those that ‘the average user’ uses were developed with corporate funding. Be it journaling (ext3/reiserfs/XFS/JFS), LSM, many hardware drivers, better threading support, KVM, etc. One could probably name hundreds or thousands of other minor and major improvements. It’s a blend of community and commercial involvement. But in the end, everyone is an equal player, and Linus tends to pick the stuff that’s technically the most interesting.
No, I am no Linus groupie. But, it is free/open and it works well. Should I dislike it because Torvals is not Stallman? When was the last time Torvalds introduced proprietary drivers in the kernel?
I honestly see no reason to continue to use Linux @ GPL v2 if I can use Solaris at GPL v3 (especially since many would consider the Solaris kernel technically better in many areas than the Linux kernel).
Heh, speaking about kernels with corporate involvement
.
Edited 2007-06-11 12:11
Linux was never the “people’s” kernel, you’re re-writing history there and making it sound like Linus somehow cheated people. The pragmatic aspect of GPL combined with the lack of copyright assignment is specifically what encouraged corporate co-development.
The community already has a “people’s kernel” that existed before Linux. It’s called Hurd. Check it out if you want an idea of how the linux kernel would look today without commercial/community co-operation.
The community already has a “people’s kernel” that existed before Linux. It’s called Hurd. Check it out if you want an idea of how the linux kernel would look today without commercial/community co-operation.
You mean…it would have went nowhere?
You can be right about certain things, but could you elaborate a bit more on how or on what basis you think Solaris kernel is technically better? I think only GNU purists would switch to Solaris solely based on the GPL v2 vs v3 difference.
“””
I think only GNU purists would switch to Solaris solely based on the GPL v2 vs v3 difference.
“””
I would think that GNU purists would be wary of Sun Microsystems’ insistance upon being granted joint copyright on all code included in the official version of OpenSolaris, allowing them to do anything with it. Anything at all.
At least with Linux one can be sure that the official version can’t be taken proprietary by a single entity which happens to hold all the necessary copyrights and provide most of the developer resources.
If Sun were to decide to take development proprietary, for their own reasons, and withdraw devloper support, the Free project would be dead in the water, no matter what version of GPL the code was under.
Linus controls the process, so it is his decision. He does not have to state the reason why. Maybe he just don’t like FSF, he is not the only one.
If you don’t like it you can still gather developers and fork the kernel. Nothing keeps you from doing that, except that you can’t call it Linux, because it is a protected trademark.
What I realy don’t like about open source is that people who are using software for free are telling developers what to do. And developers are those who do the work. I’ve lived a part of my life under communism, I’ve seen enough of that.
as far as I know it is because soooo many people have dumped code into the kernel. its not really linus’s decision alone. unless the kernel devs want to rewrite all the code they can’t change to GPL3
Browser: Mozilla/4.0 (compatible; MSIE 6.0; Windows 95; PalmSource; Blazer 3.0) 16;160×160
>unless the kernel devs want to rewrite all the code they can’t change to GPL3
you can also change the license slowly:
– First step, make sure that you can. That means that all or at least most code should be licensed under “GPLv2 or later”. That’s quite easy (1) there is already a lot of code under “GPLv2 or later” (2) ask all active Kernel hacker to allow “GPLv2 or later” (3) accept from now on only code under “GPLv2 or later” and wait some years.
– Than when all or most code is under GPLv2 or later you can really easy upgrade to GPLv3.
– Hopefully learn from this process and make sure that next time it can be done easier.
Other projects/companies has already changed their license before so it is possible. You can try to do it as fast as possible than it is probably more work or do it slowly like described above than it should be rather easy.
Edited 2007-06-11 09:06
One of the problems with the kernel is that there’s so many people involved (and some who are actually dead) that getting all the copyright holders to relicense their code would be very difficult and time-consuming.
>One of the problems with the kernel is that there’s so many people involved (and some who are actually dead) that getting all the copyright holders to relicense their code would be very difficult and time-consuming.
As i described above you don’t have to get all copyright holder to relicense. Just have a consent on active kernel hacker (than you have already >50% of Linux under GPLv2 or later) and go on with only “gplv2 or later” commits for 2-3 years and you will end up with a kernel which is (almost) complete “GPLv2 or later”.
“””
As i described above you don’t have to get all copyright holder to relicense. Just have a consent on active kernel hacker (than you have already >50% of Linux under GPLv2 or later) and go on with only “gplv2 or later” commits for 2-3 years and you will end up with a kernel which is (almost) complete “GPLv2 or later”.
“””
You are overlooking the reason for having not used the “GPLv2 or later” language in the first place. It puts the entire project at the mercy of RMS, the terms of the GPLv4, GPLv5… whomever happens to get control of the rights to release future versions of the GPL, and perhaps… the MS-GPLv6. We, none of us, can predict the future. The original licensor, regardless of his opinions on the FSF today, really can’t know what that future might bring.
Eben is stepping down from the FSF and has hinted that RMS may also be ready to hand the reigns to… the next generation.
To anyone who wants to know that their project will remain under a Free *and* Sane license, the “or later” verbiage would be most unwise to include in their initial license… *or* to add later.
However, dual licensing of new code as GPLv2 and GPLv3… *after* the final version of the GPLv3 has been released, might make sense.
Edited 2007-06-11 17:28
The use of “the or later clause” does not “put the entire project at the mercy of RMS”. Firstly, the FSF (not just RMS) would have to approve and issue a new GPL revision. Next the distributors of the software have to choose to distribute the software under the new licence. This implies a willful choice to distribute under the new version. There is no language in the GPL that says “whenever RMS feels like releasing a new licence that all licensing terms are automatically update to the new terms”.
Furthermore, do you understand what we refer to when we mean GPL? It stands for the GNU General Public Licence. It is a legal document that has the FSF as the recognised authority that issues this particular document. If you read the GPL, you’d know that the “or later clause” only applies to further revisions of the GNU GPL released by the FSF. Microsoft could possibly create their own “General Public Licence” v5 and it would not mean anything to the GNU GPL.
As for the “or later” clause, it is there purely for convenience, nothing more. When the FSF releases a new revision, users of “or later” licensed program are free to distribute the program under the new licence as they already have the permission of the program’s copyright holder to do so. Without the clause, users need to spend time obtaining permission to distribute under the new licence. This may or may not be a good thing.
As for a possible “insane or unfree” GPL revision, I find this unlikely. In my experience, sanity is in the eye of the beholder. After considering the views of the FSF, I have concluded that their views are quite acceptable. Other people say that the FSF are extremists and would label me a “zealot” for having a viewpoint that parallels the FSF. The track record of the FSF shows reliable consistency throughout its lifetime. The names that run the foundation may change over time but I am willing to bet that these names will consistently maintain the viewpoint of today’s FSF.
I agree with you. I just want to add that even the fear about a “insane, unfree or complete different” GPL is handled by GPL:
9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns.…
So every new GPL will ever follow the same spirit (give all 4 freedoms and make sure that they will reach all users of the software) Otherwise the upgrade-clause would be void.
“””
The use of “the or later clause” does not “put the entire project at the mercy of RMS”. Firstly, the FSF (not just RMS) would have to approve and issue a new GPL revision.
“””
RMS has made it quite clear that he has full control, and the final say, upon what is in GPLv3. I believe he expressed that quite clearly during the Q&A after his initial presentation of GPLv3 Draft 2. Was that a deception?
(Furthermore, do you understand what we mean when we say GPLv3 Draft 2?)
“””
Next the distributors of the software have to choose to distribute the software under the new licence.
This implies a willful choice to distribute under the new version.
“””
False. They have already distributed copies with the “or later” verbiage and cannot take it back. They can only restrict newly added code to earlier licenses.
(You’re new at this, aren’t you?)
“””
There is no language in the GPL that says “whenever RMS feels like releasing a new licence that all licensing terms are automatically update to the new terms”.
“””
Yes, there is. When GPLv3 is released, all code ever released with the “or later” verbiage will automatically become dual licensed under both GPLv2 and GPLv3, with or without the approval of any of the copyright holders.
“””
Furthermore, do you understand what we refer to when we mean GPL? It stands for the GNU General Public Licence.
“””
Yes, sonny. I’ve been aware of that since about 1995.
“””
If you read the GPL, you’d know that the “or later clause” only applies to further revisions of the GNU GPL released by the FSF. Microsoft could possibly create their own “General Public Licence” v5 and it would not mean anything to the GNU GPL.
“””
Can you guarantee that you will always agree with what any future controllers of the FSF put into a future version of the GPL? Can you guarantee that no powereful interest will ever be able to influence the FSF or gain control of it by some clever means?
I maintain that if you want to know how your code, written today, is going to be licensed in the future, you should leave out the “or later” verbiage. If you don’t care… well… that’s your business, and potentially your problem, later.
Edited 2007-06-12 18:23
GPLv3 – companies such as tivo/linksys cannot use GPLv3 code, modify it and not submit source changes back to the author.
they cannot use it to prevent people from being able to fairly use their product – i.e. no drm/dmca, don’t prevent people getting mpegs off the disk, don’t sue people for patents etc.
i find this fine and fair. the thing that worries me is that this will simply stop companies using GPL software altogether and we’ll see a surge in bloody WinCE devices and longhorn servers.
i’ve even noticed it at a couple of places i’ve worked (network appliances mainly) they’ll shy away from GPLv2 even, in favour of either 1.proprietary – closed source by default, or 2.BSD – use opensource and do whatever we like and keep our source closed.
GPLv3 – companies such as tivo/linksys cannot use GPLv3 code, modify it and not submit source changes back to the author.
That’s true in GPLv2 as well. Tivo made all their kernel patches available for download.
they cannot use it to prevent people from being able to fairly use their product – i.e. no drm/dmca,
Last I checked there was nothing in GPLv3 that prevented you from implementing DRM, either playback of or encoding of. DMCA is something different and not relevant to GPLv3.
don’t sue people for patents etc
That’s already covered in GPLv2.
This begs the question, however: in what fundamental ways has the GPL changed that it would be a good thing for the Linux kernel to change?
I’m not being a jackass, I must admit that I’m not exactly sure what switching the kernel to GPLv3 would solve…
>in what fundamental ways has the GPL changed that it would be a good thing for the Linux kernel to change?
Nothing fundamental has changed. GPLv3 just make sure that the GPL will work in the future as good as GPL has worked in the past.
The environment has changed, both technically and legally. To keep the GPL as good as everyone would expect it the GPL has to take care about the new environment.
Edited 2007-06-11 14:12
That’s a little bit short on details…
Can you tell me specifically how the GPLv3 will improve things? I know about the Novell-MS deal, and I actually support efforts to make sure MS doesn’t extend its patent FUD campaign any further, but surely there are other aspects to GPLv3.
For example, what “loopholes” is v3 supposed to close?
>For example, what “loopholes” is v3 supposed to close?
In short:
– People can give you by copyright the 4 freedoms but than sue you because of patent infringement
– People can give you by copyright and patent law the 4 freedoms but than use DRM to prevent you from exercising them and use DMCA to make sure that you don’t circumvent the DRM.
That’s the two main areas. GPLv3 will also be compatible to more Free Software licenses than GPLv2, e.g. the Apache license.
Apart from that GPLv3 uses in many areas better wording to make the license more “international”. E.g. GPLv3 doesn’t talk about “distributing software” but about “propagate a work” this allows to define “propagate” in way you want to see it instead of relying on terms like “distribute” which have partially already a special meanings for lawyers and judges in different countries.
The details are far to much to write them down here. If you are interested look at http://www.fsfeurope.org/projects/gplv3/ Here you find articles about special topics of GPLv3 and links to all the talks of RMS and Eben Moglen which explain the changes in GPLv3.
Edited 2007-06-11 14:45
– People can give you by copyright the 4 freedoms but than sue you because of patent infringement
No they can’t. That’s already covered in GPLv2. Can you please describe a scenario where the above could happen without violation GPLv2.
>Can you please describe a scenario where the above could happen without violation GPLv2.
You can read about the changes of the patent language and the impact of it here: http://gplv3.fsf.org/gpl-rationale-2006-01-16.html#SECTION003120000…
Or look in one of the many GPLv3 talks by Richard Stallman or Eben Moglen. One example: http://www.fsfeurope.org/projects/gplv3/tokyo-rms-transcript.en.htm…
(a list of all texts: http://www.fsfeurope.org/projects/gplv3/)
(I’m curious how often i will have to link to FSFE’s GPLv3 information collection until everyone will be able to find answers to this basic questions by their own…)
Edited 2007-06-12 13:24
One of the big changes would be blocking what Tivo did. Bascially they released just enough code to satisfy the GPL, but not enough code so that you’d be able to compile and run that code on their hardware.
I apparently totally missed what tivo did/is doing that everyone is/was upset about. Can someone point me to some articles explaining what happened?
>I apparently totally missed what tivo did/is doing that everyone is/was upset about. Can someone point me to some articles explaining what happened?
I have already posted a link where you can find every information about GPLv3 you want: http://www.fsfeurope.org/projects/gplv3/
In every GPLv3 talk Richard and Eben explained “Tivoisation”.
Here two concrete links:
RMS: http://www.fsfeurope.org/projects/gplv3/barcelona-rms-transcript.en…
Eben Moglen: http://www.fsfeurope.org/projects/gplv3/barcelona-moglen-transcript… (here i don’t have a direct link to the passage in the text, just search for tivo and you will find Eben’s explanation (it starts at [0h35m20s]))
Edited 2007-06-11 19:07
oh, thanks. I have over looked your previous post.
Let me get this correct: You believe that as soon as the FSF releases the final version of the GNU GPLv3, all software that is licensed under “GPLv2 or later” will immediately be licensed under GPLv3. Your fears relating to the “or later” language are based on the idea of the possibility that an entity that could somehow exploit the premise of automatic retroactive changes to the distribution terms of a GPL licensed program.
A program licensed under the GPLv2 will remain licensed under GPLv2 regardless of the FSF’s actions. The licensee does not have the right to change the distribution terms of the GPLv2 without the licensor’s permission. The licensor permitting distribution under a different set of terms (as they become available) does not imply automatic retroactive changes to the distribution terms (when said different terms become available). ie, when the GNU GPLv3 is released by the FSF, the only programs that will migrate to the new licence are the ones that the author will specifically allow.
The point of the “GPL version 2 or later” clause is so a licensed distributor has the right to distribute the program under “GPLv2 exclusive”, or “GPLv2 or later”, or “GPLv3 exclusive”, or “GPLv3 or later”, …. “GPLv9 exclusive”, or “GPLv9 or later” and so on when new revisions of the licence becomes available. The clause is a matter of convenience because without the clause, the licensee needs to obtain permission before distributing the program under an updated set of terms.
Edited 2007-06-13 14:17