Slashdot Mirror
Linus on GPL3 In Forbes
musicon writes "In an interview via e-mail with Forbes, Torvalds discusses GPLv3, digital rights management and sharks with laser beams. From the article: 'I'm sure changes will be made [to GPLv3]. The fact that the FSF and I have some fundamentally different views of what the GPLv2 was all about makes me worry that we won't find a good agreement on the next version.'"
sudo apt-get install sharks-with-lasers_kernel_module
"Something's wrong with you...and I hope we never do meet again." - Deftones When Girls Telephone Boys
Welcome our laser beam-wearing shark overlords!
"Ein Volk, ein Reich, ein Führer." -Adolf Hitler
"We are one Nation, we are one People." -The One 'leader'
First off, please forgive my ignorance, but is it really *that* important for Linus to decide to move Linux from the GPLv2 to the GPLv3? Just because version 3 of the license becomes available does not automatically invalidate the version 2 license does it? Why is this such a hot button issue?
For the most part, I completely agree with Torvalds on his points--and I can't say I'm at all surprised to see Stallman and the FSF take this direction with version 3. Simply put: they are "zealots" for lack of a better term. For them, free software is less about open source and open development and more about a form of political agenda.
Now I'm not trying to bash Stallman or the FSF, they have made some wonderful contributions to the community. But let's call a spade a spade here and look at what GPLv3 is about: attempting to hide attempts to restrict developers under the guise of being an update to the world's most popular open source license. For all of the FSF's talk against bad copyright policy and software restrictions, this license introduces their own set as if to say, "we don't like their way; so you should definitely do it our way instead."
Too much politics and agenda and not enough open source development.
Torvalds Explains Dislike For GPLv3
Linus Says No GPLv3 for the Linux Kernel
I see a GPL that prevents companies from using DRM (which wasn't around for v2) to get around GPL requirements. Basically those same requirements that we liked from v2.
If Linus is fine with TiVo's method of coopting the kernel and making it for all practical purposes unmodifiable, that's his business. But lots of other people have contributed code to free software and are not.
PS: this is how I understand it so far. My opinion is subject to revision
My turnips listen for the soft cry of your love
I'm not sure....would maybe Firefox have more overall users? Seems that it's on 80-90% of Linux boxes, plus an ever growing number of Windows machines and other OS's as well.
\/\/oobie
Firefox is only on workstations -- headless servers typically won't have a web browser; my company's certainly don't. I was thinking gcc would be a better candidate: Not only is it installed on a strong majority of Linux-based systems, but also on a large number of traditional Unix systems elsewhere.
I don't think it's a matter of right and wrong, but a battle of ideas between purist "ivory tower" types and the real-world that has legitimate needs for OSS and the business community to work together. Like I said before, if you think that businesses like IBM have purely altruistic motives for supporting Linux and OSS then you are sadly, sadly mistaken. Businesses have a responsibility to their shareholders to make money. Linux/OSS is a means to an end. But in the meantime, the enemy of my enemy is my friend.
ConsultingFair.com
I wonder if Linus even has the real authority to unilaterly switch to an alternative license. I don't think so. By his own admission he is not a deep thinker about the philosophical (he says polical) part of the job. Many of his colleagues are. Any change would have to be accepted by the core kernel developers. If not a fork is all but inevitable (GNU/Linux anyone?). My guess is he will talk like this from time to time but will be under pressure to maintain the status quo.
an ill wind that blows no good
If I understand it right, and I prolly don't, you can install any modified version whatsoever on your sharks. Your obligations re: making keys available etc. do not kick in until you distribute the modified version. i.e. if you're a shark salesman rather than a mad scientist.
Is that right?
My turnips listen for the soft cry of your love
Linus isn't an activist. He's just a programmer. Sure, he made a wonderful kernel, but it's the GPL that made his kernel popular and freely-downloadable.
In any case, does it really matter if he redistributes his kernel under GPL2 or 3? It's not like it's the end of the world or anything. I think this is plainly media hype.
Due to the public outcry GPLv4 will no longer include sharks with laser beams. They have been replaced with Sea Bass. Extremely ill-tempered sea bass.
Looking at those disconcerting trends, I very much support the GPL v3's approach to software patents. But when it comes to DRM, I think the FSF goes too far and addresses an issue for philosophical reasons that isn't worth it. DRM is a lot more legitimate per se than software patents are. Categorically opposing DRM may be perceived as downright anti-commercial by a number of people, and it's a move that I fear will only hurt the FSF and the GPL without changing anything about the fact that DRM is here to stay.
Right at the end:
Are you participating in the GPLv3 process?
No, I'm not actively involved. And it's not so much because I couldn't be, it's more because I just can't find it in me to care too deeply. I'm the kind of person who hates office politics. I'm pretty happy with the GPLv2, and I just don't have the motivation or inclination to start talking to lawyers. I'm a programmer. I worry about kernel bugs.
For all of the FSF's talk against bad copyright policy and software restrictions, this license introduces their own set as if to say, "we don't like their way; so you should definitely do it our way instead."
RMS and the FSF aren't saying saying "All your old GPLv2's are invalid and now you must upgrade to our new GPLv3!!!"
They are giving developers the options to restrict what others from restricting the next guy down the line from doing something with their work.
You don't have to use GPLv3 if you don't want to.
If someone else releases their work with a GPLv3 license and it bothers you...
Then tough. The original author has the right to release it under any license he wants be it BSD, closed source, or GPL.
If Linus doesn't want to use GLPv3 then it is his right. He can keep v2 forever. The GPL license doesn't belong to RMS. He just made up the wording of the contract that others can use to release software with.
No one is being forced to anything they don't want to...
Well other than the people who are being restricted from adding DRM and various freedom restricting to other people's work released in GPLv3.
Well if you really want that DRM so bad... Then make your own program from scratch. Don't use someone elses open source code whose express wish is to not have his work used in ways he did not mean it to.
GPLv3 gives the original author this ability.
"I am the king of the Romans, and am superior to rules of grammar!"
-Sigismund, Holy Roman Emperor (1368-1437)
Simply put: they are "zealots" for lack of a better term. For them, free software is less about open source and open development and more about a form of political agenda.
Stallman repeatedly states that software freedom is his goal, and not its widespread adoption by "practical minded" corporations. He has nothing against corporations if they do not interfere with his primary goal. That make's him a zealot, I guess. I call it clear thinking. Time and again he has been proven correct in the face of criticism.
an ill wind that blows no good
Like I remember installing Ubuntu once and gcc wasn't on it. Some other nontechnical-user-oriented distros I've seen were like that. They assume you'll use package managers to do everything?
My turnips listen for the soft cry of your love
You cannot install it on your hardware (laser-equipped shark or otherwise) without also making sure that others can install another version. And that's my gripe.
Call me a fanatic, but open source isn't worth crap if it can't be redistributed. This is _THE_ principle of open source, that anyone can make AND RUN their own version. There are business-ready licenses out there, but the GPL was made to perpetuate the programmers' and users' freedom.
I think Linus needs a reality check. Perhaps a few months of working for Microsoft will make him realize his mistakes. There ARE evil people, evil corporations trying to take over the world, just look at the patent business.
I'm kinda disappointed after reading this, I always had seen Linus as a hero, and thought he was as enthusiastic about open source as many of us were. Sad to see he's just yet another programmer who went corporate, like Steve Jobs. He just happened to cooperate with the open source movement.
Oh well. We should be thankful he's still cooperating, and consider him an ally rather than a leader.
Yes! I'm confused as to how Linus gets this part wrong. Mad scientists (or the military, or Microsoft, etc.) can use modified GPLv3 code however they want. However, if they try to distribute the code to the public (they sell software, or hardware with software on it), then they have to make it possible for the recipients to:
1. See the code.
2. Modify the code.
3. Run the modified code.
Private shark zoos are not subject to any restriction. Shark salesmen, however, would be required to make the source available, so that you could modify the way your shark-laser system works.
The TiVo example is a better one. If I buy a TiVo, and it uses GPL code, then I should be able to modify that code and run the modified code on that hardware. Their modifications are functionally useless to me (to the world) if I cannot run the modified code on the hardware. Yes, in theory I could build my own TiVo from scratch... but that's not practical. So why should GPL programmers license their code that way?
GPLv3 is anti-evil-DRM in which the GPL would be circumvented by DRM methods. Such as providing you with the source code by unable to compile/run it because of DRM. Yet GPLv3 as I understand it does not say you can not include DRM in your software, you are free to do so.
I just find Linus too trusting of business. You would think he would have learned his lesson with BitKeeper but in the end I think he blames Andrew Tridgell instead of BitMover. Even RMS may be too distrusting of business, but isn't it better to be safer than sorry?
The best education consists in immunizing people against systematic attempts at education. - Paul Feyerabend
Why the facination with Mr. T's thoughts on the GPL? He really isn't that big of a believer in Open Source. It just happened to be the vehicle that propelled him to fame. IT IS HIS RIGHT TO NOT REALLY CARE ABOUT OPEN SOURCE!
/. seems to have about pretending there is this big NEW rift in the open source movement with Stalmans GPL v3? There is no rift, you have Stallman who is a believer, and Linus, who couldn't really give a crap as long as he can keep working unencumbered. That is why he chose bitkeeper, again, that is his right, as long as he does not pose as an Opensource poster boy(I don't think he usually does).
But what is this thing that
No, one of the fundamental GPL v3 changes is, by intentional design, antithetical to the continued proliferation of Linux in certain types of embedded devices, including TiVo-like devices, set top boxes, etc. Thus, there is and always will be a fundamental tension between RMS's notion of ideal freedom and the Linux community's goal of "Linux everywhere".
This isn't something that can be changed by a simple wording change. IMHO, GPL v3 is basically DOA as far as the kernel is concerned; you can pretty much be guaranteed that if Linus did try to push GPL v3 into the kernel, all the embedded Linux developers would fork, and that fork would result in some really ugly politics and a very dramatic decline in the number of Linux (v3) kernel developers.
Check out my sci-fi/humor trilogy at PatriotsBooks.
The FSF folks would be ecstatic to have busnesses actually embrace the open source model, be commercial, and sell lots and lots of support, installation, and maintenance for software that is still modifiable by the end customer.
People keep trying to paint the FSF folks as anti-commercial, or anti-business. They are most assuredly not. They are trying to educate companies and the public about a better way to do software, whether as a business or not.
And neither companies nor people should adopt FSF principals out of altruism. They should adopt them because they realize that once customers understand what the free software rights really do for them, they will begin to demand them by not doing business with companies that don't grant them. Just as you wouldn't buy a car from a dealer if you could only ever get it fixed at that dealership (for whatever rates they choose to charge), you will stop buying software that can only be modified by that software company. It doesn't mean you won't go to the dealer for some or all repairs, it just means you don't want to be forced to.
Of course, pushing the car analogy, this only really happens when you become aware of local car repair companies. And this is where companies like IBM can really help -- by offering the "Jiffy Lube" of free software -- a national, well known chain of software maintenance, configuration, and repair for open source.
- "History shows again and again how nature points out the folly of men" -- Blue Oyster Cult, 'Godzilla'
I see a GPL that prevents companies from using DRM (which wasn't around for v2) to get around GPL requirements. Basically those same requirements that we liked from v2.
The GPL does not require the licensor to allow the licensee to run modified code on particular hardware. It is not there, ergo it is not a requirement.
Arguments that it was an unintentional or unenvisioned loophole are wholly unpersuasive. RMS was fully aware that compiled code could be burned onto a PROM and incorporated into a hardware device to create a closed platform when he conceived of the GPL concept (whichever version you care to cite prior to the v3 draft). RMS was perfectly aware that the Xerox printer driver that so incensed him interfaced with Xerox printer software built into the printer. RMS, like most technically oriented computer users, also had to be aware that the PostScript software installed in various printers was more than simple firmware. RMS had ample opportunity to write the license that he intended, and he did not write the license that ought-to-have-been. Whatever his current stance, the license-that-is is the license that has gained such popularity and it now has a life independent of RMS and others of like mind.
You may not see a negative, but the business community certainly does. Considering that the business community is the one paying for support for GPLed software and purchasing GPL products, I fully expect to see GPL v2 versions of any commercializable collaborative apps exist far into the future. Whether you release particular code under GPL v3 or not, others can examine the code, document its interfaces, describe its processes, and rewrite it under GPL v2. Give RMS credit where credit is due, because he has created a software license that even he himself cannot defeat.
Suppose a vendor creates a distro, Blue Hat. It's designed for platform P but P is made to require binaries signed by Blue Hat, it won't run anything else. Now Blue Hat releases a body of source code and claims to have complied with GPL v.2.
Now has Blue Hat complied with GPL v.2? No one outside Blue Hat can know. The only way to verify that some source corresponds to the binary you're running is to compile it and run the result. If you can't do that without a key, and Blue Hat won't give you a suitable key, they could violate GPL with impunity.
It doesn't require that BH give up their ultimate private key, just one sufficient to sign source. This is all that GPL 3 requires in regard to DRM and keys.
Morally, Linus is on the wrong side of the DRM battle, since he supports it (and is willing to be used as a PR pawn by Forbes), however Pragmatically, he's on the right side of the battle, since DRM is ineveitable and perhaps by doing their bidding, the robber barons^W^W business world will allow him to continue living^w coding.
Stallman may be right morally, but so was John The Baptist; and look at what happened to him.
Just to throw my two cents in, that's exactly how I read it as well.
The thing is, if we really do get "Linux everywhere," enough people are going to want to start hacking that it'll create an economic incentive to cater to the hackers. Witness the Linux variant of the WRT54G.
IMHO, DRM will only die by collapsing under its own weight, and by heightening consumers' awareness of the issues. Fighting DRM head on by denying access to its underlying technologies (when, as Linus states, those technologies themselves aren't inherently evil) isn't going to work.
--JoeProgram Intellivision!
$ dmesg
sharks-with-lasers: module license 'mad scientist' taints kernel.
No, you can still sell the services of a machine that has GPL3 and not release your code. You can take a already take a machine that runs GNU/Linux modify it up one side and down the other, and not release a thing. Software copyrights also allow you to do this. That's why most companies like Microsoft won't sell you a copy of software, instead they "license" it to you.
..."Once the rockets are up, who cares where they come down. That's not my department", says Wernher Von Braun...
GPL kicks in only when you want to *publish*. If you want to publish a hybrid of your code and GPL code, like the Tivo kernel, you have to release source code. If you don't want to do that, you can release your product as closed source which is installed as a separate product from the Linux kernel.
That's basically what Tivo does with a twist. They've imbedded DRM into their kernel such that the kernel source is available, but you need their keys to unlock the content on the box. People have hacked around this, but it's a pain.
GPL3 is trying to fix this. Under it, Tivo wouldn't be allowed to use GNU/Linux unless the released code to everything necessary to make a Tivo work. Tivo, the company, is perfectly free to continue using the source they've already got, and to only release what they've already released, but they wouldn't be able to use new drivers and features that are GPL3'd.
Linus is using the Von Braun defense.
P.S. The other thing that GPL allows the end user to do, is to buy one copy and to run it on as many machine as they want to. Ordinary copyright and software licenses don't allow this.
All ideas^H^H^H^H^Hprocesses in this post are Patent Pending. (as well as the process of patenting all postings)
I was rather disappointed with the following quote:
It's not so much I generally care about being able to change my PVR's firmware to unofficial/open source releases it's just that doing so can sometimes be beneficial and add value (example: routers -> OpenWRT etc) and, well, being able to do this is just cool. I guess this disappointement is because I like hardware hacking as much as I do software.
RMS "Pragmatically speaking, thinking about greater long-term goals will strengthen your will to resist this pressure. If you focus your mind on the freedom and community that you can build by staying firm, you will find the strength to do it. ``Stand for something, or you will fall for nothing.''
Pretty much sums it up. I'm sure RMS doesn't like talking to lawyers either, he just has Beliefs and convictions that force him to. (No offense Mr MOGLEN)
I will adopt GPL v3 as soon as finalized. I have much more faith in the "Ramblings" of RMS than the casual "Who Cares" of Linus. Great as the kernel might be. Without the Convictions of Stallman and GPLv2, Most of us would still be Running Proprietary OS's and paying for several different Compilers and toolsuites and Graphical toolkits, ad infinitum. We have those choices now because Stallman sat down with a few lawyers despite his distaste for authority. There is something to be said about beating the big guys with their own stick, Software Licenses
Note to Article Author: I believe GCC is a little more popular (at least in terms of users) than Linux or Firefox. And that was written by?
Ramble on RMS
OSGGFG - Open Source Gamers Guide to Free Games
I watched some of the explanations on GPL3 given by RMS and Moglen here, and they seemed fairly adamant about the fact that GPL3 did not forbid the use of DRM-ish encryption/authentication in code that falls under it, just that any keys necesary for running it need to be given to the user.
Same deal for Trusted Computing-esque machines; it's perfectly fine to make a machine that uses keys to restrict usage of the machine maker's code that falls under GPL3, as long as they give the keys to the end-user. It's even fine to make a seperate key for each and every machine, only give it's unique key to one single end user tied to his or her machine. Just as long as the end-user gets those keys: Or as the GPL3 licence puts it: Reading through the DRM clause it mentions "no permission is given to distribute covered works that illegally invade users' privacy" which is already "illegal", so you wouldn't be allowed to do it anyway, and "[no permission is given] for modes of distribution that deny users that run covered works the full exercise of the legal rights granted by this License" which you wouldn't be allowed to do anyway, without breaking the licence.
Or is this dispute about some other part of the new licence?
If Torvalds declared he was simply walking away from it all, one really couldn't blame him considering the sometimes disgusting criticism levelled at him. And what has he done, apart from helping to give the world a free operating system, inspiring the open source movement and saying simply "Please treat me with the freedoms and respect I extend to you"?
Mr Torvalds is just one guy, an engineer with an optimistic, congenial outlook on life, not a professinal advocate or evangelist with a foundation or six behind him full of law professors. Have Lessig and Moglen condemned some of the personal criticism levelled at Torvalds or would that be too much like hard work?
Yet as a hard-working regular guy Torvalds has a better grasp of daily realities than many of his critics. Tightening up on software patents is a good idea, but rejecting DRM as the devil's work is a poor idea, perhaps simply immature. As the underdog in this affair he gets my support every time. If a better, revised GPLv3 emerges, one that two people or more can actually understand and agree upon, which is more than can be said for the present draft, then Torvalds will deserve our gratitude for sticking up for what he believes despite the hyena-like behaviour of some in the open source world.
Las qué passoun
tournoun pas maï
Let me rephrase your two categories, while trying to preserve the essential meaning.
First, DRM can be used to help stop people from breaking the law, e.g. by making copies beyond what is permitted by fair use.
Second, DRM can be used to restrict users from exercising rights that otherwise would be theirs by law, such as making fair use copies of copyrighted material or running a copy of a program that they have purchased.
Unfortunately, all existing DRM seems to fall into both categories or exclusively into the second category. Perhaps the wild-eyed idealists are the ones who are considering how DRM might in theory be used, rather than how it is used in practice.
He doesn't. The Linux kernel hasn't been under his exclusive copyright for most of its existence. Not even his fork is under his copyright alone because he doesn't collect copyright assignments from contributors to his fork. This seems to me to be right in line with the general lack of foresight and considerable confusion about software freedom I've come to associate with him (see his use of Bitkeeper and his objection to Andrew Tridgell's work on a Bitkeeper repo pulling program for other examples).
All the more reason why Forbes should have interviewed people who are deep thinkers about issues relevant to the GPL: RMS, Eben Moglen, or someone from the FSF who could have spoken with more insight and a clear understanding of what the license is meant to achieve. Interviewing Torvalds about licensing is usually fruitless because he gets another chance to demonstrate how much he doesn't understand the goals of the license and how much he doesn't agree with what he doesn't understand.
He claims that use of GPL-covered programs is restricted by the first draft of GPLv3: "You cannot install it on your hardware (laser-equipped shark or otherwise) without also making sure that others can install another version.". That is a good thing because that's critical to software freedom. His criticism is confusing: he professes to want to be allowed to fix things, yet he criticizes along the lines of preventing people from stopping users to be able to fix things. He also doesn't seem to understand that others might want to tinker things he doesn't want to tinker with (a dishwasher or a DVR). Heaven forbid anyone wants to change the length of a wash, rinse, or dry cycle or a DVR that only deletes recorded programs when the user says to do so.
He views the GPLv2 as a contract: "However, I don't think that's part of my GPLv2 contract." and Eben Moglen made it quite clear in his detailed discussion of GPLv3 that the GPL has not been and will not be a contract. There's even a section in the draft GPLv3 called "Not a Contract". I'd rather take Moglen's legal advice than Torvalds', particularly when it comes to interpreting the GPL.
It's also hard to take Torvalds' complaints seriously because he refuses to become a part of the year-long revision process, even by submitting comments to the GPLv3 FSF site.
GNU/Linux isn't a fork, it's the GNU operating system featuring the Linux kernel. This is distinct from the GNU operating system featuring a kernel from one of the BSD systems, or the official GNU operating system which runs with the HURD.
Digital Citizen
mad_scientist is completely compatible with the GPL, provided you load the wild_haircut and mismatched_socks modules first.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Why not? It's the data and binaries in the payroll system that need to be secure, not the source code.
That is all.
Linus is very confused. DRM is not necessary to protect your diary, and it is very easy to see why. I wrote down my thoughts here:
s _got_it_wrong_on_drm
http://www.fsfe.org/en/fellows/marcus/weblog/linu
"If the "Or later clause" is present, then the embedded developers are SOL, because their own fork must also include the "Or later" clause"
3 .html
No. You're missing the point of multiple licenses. If you distribute software to me that says GPL2 or later, it means that I can redistribute with any of GPL2, GPL3, or any future version. What I can't do is add "or later" to software that was licensed strictly GPL2.
It's the same as if you distributed software to me with a BSD or GPL clause. I can redistribute the software in something with a BSD license that is not GPLed.
Stallman talked about this at the bottom of http://www.onlamp.com/pub/a/onlamp/2005/09/22/gpl