Slashdot Mirror
Should Developers Switch to GPLv3?
Isaac IANAL asks: "Victor Loh of ExtremeTech writes about the General Public License version 3's clause, which requires releasing digital signature keys — in other words, the software should be able to retain interoperability when modified. The article raises an objection, citing Linus Torvalds, that the so-called TiVoisation clause would inhibit open-source adoption in embedded devices among entities such as governments, health care providers, and finance firms. The issue has been discussed on Slashdot many times before. If you're a developer for a platform that needs to run signed code, could you use software under the GPLv3, or does the GPLv3 (at its current, unreleased state) truly inhibit your control as a developer over your device?"
Wait for the service pack.
If you are writing from scratch you lose no control as you can dual, triple or whatever license your own code as you see fit.
If I sit down and from scratch write a kernel I can release it under the GPL v2, v3, v8 and seventeen differrent closed licenses with no problems at all other than going mad from reading all of the legal junk that's required to define each one.
It would only impact on me if I decided to use someone else's work as the basis for mine, or as part of mine, and then I would either have to comply with their license or do the work myself. Doesn't seem that hard to me.
Think of the Children; Sleep with your Sister
If you are really interested in building a community around your project, choose a license that not only lets people contribute back to you (meaning that it has to be open to them in the first place) but also allows them to leave any time without having to forfeit their work (meaning that you have a cooperative relationship, not a dom/sub relationship).
GPLv3 is the worst of the series, IMO. Where it fails is in its insistence that if you want to be part of the community that you basically have to turn over every single thing to the whole community before you get the blessing to participate. Got a patent? Sorry, bud, check that at the door. Want to run specialized programs that require secrecy of code? Not on this platform, man. Want to mingle your closed code with our open widget? Give up all your source first.
It's not inviting at all except to anyone who has more to gain than lose from such a relationship. So what you get is a bunch of people who are actually leeches creating programs that no one else outside the community can even look at for fear of contamination.
If you want to share, then share. If you want to profit off of others and view everyone that looks at your code without contribution as suspicious, choose the new GPL. (The Artistic License for example, before it became GPL-compatible, was actually very cool and was able to gain a very large and loyal following for the Perl language. People contributed out of a sense of community, not out of coercion or because they were collecting a paycheck to do it.)
The GFDL is what you are looking for.
http://www.gnu.org/copyleft/fdl.html
Some people think it is antithetical to the purported aims of the GPL and the FSF.
Many of the Creative Commons licenses are more useful for non-software creative works. They have a wide variety of licenses, including "share-alike" (similar to the GPL), "attribution required" (similar to BSD), and so on. The GFDL is also meant for "documentation", but I personally don't like it (the bit about "invariant sections" is very crufty.)
See www.creativecommons.org for more info.
Okay, I may be a little tipsy, and legal loopholes may not be my strong point, but what exactly is wrong with v3? As I understand it, one of its main purposes is to prevent cases like Tivo from happening again, where the source is officially released (therefore GPL-compliant), but modified builds won't work anyway (not covered in GPLv2, therefore legally correct, but still against the actual spirit of the GPL)
Isn't it expected that licenses will evolve as technmology changes, and as loopholes are exploited? If v3 isn't adopted, what's to prevent everyone from locking down their software through keys?
Please clarify if I've misunderstood something, I greatly respect RMS and really can't see what he's doing wrong here. As I see it, without v3, the GPL will just end up just being a license where people can use the community's hard work and avoid giving something back in return.
I can explain it for you, but I can't understand it for you.
Spreading it is (either by ignorance or by malice) helping bad companies, like TiVo for instance.. Please read on the following to understand WHAT the GPL v3 draft says.
The draft version of the GPLv3 says that IF AND ONLY IF the software you want to run, needs some special digital signature, then and only then must the digital signatures acompany the source code.
So, what does all this blurb mean? Is Linus so obtuse he can't read english? No. So...?
I could understand it if he said that he felt he couldn't ignore the contributions of some hardware manufacturers, but what does he say? He says that GPLv3 "sucks" because it prevents legitimate businesses like those of TiVo. That if users don't like that hardware, they can use other hardware.
As usual, untrue pragmatism. The pragmatist doesn't idealize about perfect future conditions that may or may not happen. The true pragmatist solves the problem in a practical and definitive form: preventing the harm from happening.
For explanations of the changes in GPLv3, I highly recommend reading (or skimming) the transcripts of the GPLv3 conferences. Each transcript includes the subsequent Q&A session, and each begins with a list of links to the topics covered and the questions asked.
The freshest transcript is RMS in Bangalore in August. Here are the others:
Many also include links to audio and/or video recordings, and there's more general information about the timeline and how to participate on FSFE's GPLv3 page.
Also, if you want to help raise the quality of discussion, a useful and really easy thing to do is to pass these links on to others.
Please help publicise swpat.org - the software patents wiki
> If you're a developer for a platform that needs to run signed code, could you use software under the GPLv3
:)
:) Me--I consider myself a pragmatist too. I've used the BSD license, GPL, Apache, and many more, not to mention semi-free and proprietary licenses. I base my decisions on what I think is appropriate for the project I'm working on. Not on what a bunch of fanatics tell me. But the GPLv3 seems perfectly in line with the GPLv2 to me. It closes a couple of obvious loopholes, and little more. When I get some code released under the GPL, I expect to be able to fix it. TiVo showed us all that that wasn't necessarily true. If it were my code they were using, I'd be pissed as hell!
Yes!
> or does the GPLv3 (at its current, unreleased state) truly inhibit your control as a developer over your device?"
No! Any more questions?
(Ok, if you want to get picky: it doesn't inhibit your control over "your" device, but it may inhibit your ability to inhibit others. You know--the people who actually OWN "your" device! But that's the whole point!)
This whole "requires releasing digital keys" nonsense has to go! Whoever invented that meme should be shot. And I don't care how many of you like his fucking kernel!
Everyone's talking like this is going to have huge effects. The fact is that there is really, so far, only one company that would have been affected, and they won't be affected because the Linux kernel devs long ago decided to stick with v2. And now the devs want to justify that decision by pointing out all the supposed flaws with v3. I'm not impressed with their reasoning.
People talk about voting machines. The solution there is easy. The software needs to provide a signature of the results AND the software together. Then you can easily detect tampering while still providing all the freedom necessary to fix problems.
Going with GPLv2-only is the WORST possible solution, as far as I can tell. That will guarantee license-incompatibility in the future. Frankly, I see nothing in the GPLv3 draft that would justify the kind of headaches that going to GPLv2-only would cause. In fact, I see nothing in the GPLv3 worth bitching about. Yes, it's new, yes, there's some controversy, but my god, I was there when the original GPL was released, and this controversy ain't nothin' compared to the shitstorm of controversy back then! Well, Stallman turned out to be basically right about the GPL in the first place, and, by comparison, I see nothing but tiny, incremental improvements this time around.
The GPLv3 will be happening, and I, and probably tens of thousands of others, will be using it. Get used to it!
By the end of the next decade, I predict that people choosing GPLv2-only licenses will be being cursed as roundly and solidly as those who chose non-dual-licensed MPL or Artistic are today.
Governments, healthcare providers, and finance firms don't need to be able to make sure their software can be maintained? They want to be locked into a single source?
This is such a bullshit argument. Nothing about GPL3 prevents you from making your own machine tamper-proof. What they're really talking about, is distributing widgets to other people such that the other people cannot maintain or "tamper with" the widget. Governments, healthcare providers, and finance firms do not need that. Only media companies [think they] need that.
From submitter:
Before you finish that question, let's get something straight. When that platform is deployed, the end user will have the ability to install or choose what key(s) (perhaps even the end user's own key) the platform will accept, right? If so, then I really don't think you're going to have a problem with GPL3.
If the end user will not be able to sign code themselves, then fuck off. You sure as hell aren't talking about using DRM as a security feature, because users are the party who are ultimately responsible for their own security. Nobody cares if your project is GPL3 compatable or not, and nobody cares if your project uses Linux, because Linux is almost useless, like any other OS, if users cannot get maintenance whenever they want it. If your project can't get bugs fixed or features added (including features that you, the developer, think are bad ideas) then your project might as well run MS Windows. Maybe Torvalds doesn't care about users anymore, but Linux didn't get all the other developers working on it (including the ones who wrote those free drivers that you salivate over) by fucking the users over. Linux attracted people and became a successful project by not being user-hostile.
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
What part of the GPLv3 discussion draft 2 puts the restrictions on the end-user that you claim?
GPLv3 only puts restrictions on redistribution, not on use. (Just like v2.)
Please help publicise swpat.org - the software patents wiki
"Your" device? Once you've sold it to a customer, it's ceased to be "your" device. If a customer buys a device that runs GPLed software, they have the freedom to replace that software as they see fit. That's entire purpose of the GPL: to grant end users freedom. Complaining that the GPLv3 inhibits a developer's control over their device is like complaining that GPLv2 inhibits a developer's control over their software. Congratulations on identifying the core purpose of the GPL.
Next week on Ask Slashdot: "Can you use the Bill of Rights in your dictatorship, or does the it truly inhibit your control as a dictator over your citizens?"
Search 2010 Gen Con events
Your joke accidentally inspired a serious thought:
This shouldn't be named "GPLv3" when done and finalized. If they do that, there will be a big clusterfuck of confusion and uncertainty, coming from "GPL" softwares with crucially differing GPL versions -- v2 vs. v3 -- and this will harm business adoption of open-source software. Not completely clueful managers and officers get confused, they lose face, so they go elsewhere. (That is, stay with closed-source.)
"GPLv3" should be named "Stricter GPL -- SGPL" (or something like that), and "GPLv2" should be kept just "GPL" -- the familiar and famous thing that nobody has a problem with.
And anybody responding that we FSF hippies don't give a damn what the corporate world wants or needs... I understand the sentiment ("we do tools for ourselves and that's all"), but it would be good to have FOSS spread further, and in the biz domain any such ambiquity or other "perception problem" can be a bigger problem than anything related to quality or technology. Make the GPLv3 into what you want, but make it clearly separate from the current well-established GPL.
The freedom you're talking about is total freedom - which leads to fuedalism, which is not very free at all in practice.
This is total garbage, and can very easily be shown to be total garbage merely by pointing to those projects which *do* use MIT/BSD licenses and which work fine organisationally. Yes, forks happen, but forks happen with GPLed code too.
I've said this to a lot of the pro-FSF lemmings that I've seen on this site, and I'm going to say it to you too...Try using your own brain for a change, rather than constantly leaning on Stallman's. You might even find that you enjoy the experience.
First off, Linux is only a kernel. Did you somehow forget what else comes with a GNU/Linux distribution? The shells? The binary utilities? The network managers?
Last time I checked, Linux was best built with a GCC toolchain. That's right, a GNU C compiler is used to build Linux. Oh, and you should be using GNU make to configure it.
The FSF and its GNU project provide support utilities for virtually every Linux distro out there right now. Sadly, most of them, excepting Debian and its derivatives, have thrown away their acknoledgement of GNU and its importance in making Linux work. That is exactly how you talk -- as if GNU has done nothing for Linux.
What I hear from you is nothing more than fanboy's prattle. You honestly believe that Linux owes nothing to the FSF? NOTHING?
Without GNU, I would not have the following utilities:
Still feel that Linux doesn't need the GNU project or the FSF? Well, fine. Just don't call me an "uncompromising, radical, neo-Bolshevik extremist" anymore.
~ C.
Oh yes it does.
Support my political activism on Patreon.
I don't think that your analysis is correct here. The GPL is still relying purely on copyright law; it only applies if you engage in activity otherwise prohibited by copyright law (e.g. copying, distribution, preparing derivatives, etc.). The difference is that rather than limit what it requires in compensation for the copyright license it grants to the copyright field, it's now asking for more. What it asks for doesn't really weaken it. Lots of copyright licenses are made in exchange for money, which is also outside the ambit of copyright.
I don't think that it's getting into EULA territory, really, as the GPL still would not apply to end users. It only applies to people who want to engage in covered activity. Mere use isn't one of those.
And also, EULAs continue to grow in strength, though the main issue there has to do with how they're formed, not what they deal with. No one would have any argument at all against a EULA that was presented and agreed upon in a different fashion. In regard to formation, the GPL is on the same ground it always was.
-- This and all my posts are in the public domain. I am a lawyer. I am not your lawyer, and this is not legal advice.
The philosophy is the same. The primary motivating goal behind the GPL has always been to enable the end user the freedom to modify the software that they receive, in whatever fashion, and be able to share the software they're using (modified or not) with others. Patents inhibit that freedom. The restrictions on someone who uses patents as a weapon against Free Software are ONLY to the acts of modifying and distributing it, which is completely within copyright law.
It simply requires that you provide the user with the ability to use/create a functional key which provides identical functionality.
That way you can't end up with a situation where, say, Microsoft, uses their market clout to make hardware manufacturers release 'secure' boxes which only boot from Microsoft keys, and then they release a Linux kernel signed by Microsoft.... Now you have the source code to the Microsoft Linux kernel, but no 'comodity' box that will boot your recompiled kernel because they all require Microsoft's key.
Now, Microsoft (and Linus) can keep their private key private -- they just have to provide you with a key (any key) that will boot your box ... and They just can't punish you for using your own kernel (as long as it provides identical functionality).
Sometimes boldness is in fashion. Sometimes only the brave will be bold.