RMS Explains GPLv3 Draft 3

H4x0r Jim Duggan writes "A transcript is now online of a talk Richard Stallman gave in Brussels earlier this week about the discussion draft 3 of GPLv3. Among other things, he explained how it will address the Novell-MS deal, from Novell's point of view and from Microsoft's, and he explained how the tivoisation clause was narrowed to make it more acceptable in the hope that it will be used by more people. After the talk he also gave an interview, and yesterday, draft 2 of LGPLv3 was released."

I wish for a ...

legal discussion of GPLv3, by lawyers for laymen, and how it impacts: developers, developers who use libraries, users, corporate users, etc...

Sorry, telling me that the GPL is obvious and that I'm stupid for not understanding it won't save me if I run into legal problems associated with the GPL. And I will have to spend money on an IP attorney to help me. It would be stupid to think that, without legal training and a license in my state, I know the law and can interpret a contract accordingly.

Yeah, it would be nice if it were so easy that I didn't need a lawyer. Then again, if laws and contracts were so obvious, why are there courts?

Incorrect answer (most of the time): Because everyone is dishonest and trying to screw the other guy.

Re:I wish for a ...

It's not a contract, it's a software licence, which isn't the same thing in law.

And it is stupid to have such things as EULA's which only a lawyer could possibly understand the full meaning and implications of - but haven't all the millions of computer users the world over "agreed" to them without such an understanding? Or has everyone else received their classes in understanding the licences for Microsoft Windows and Office, Adobe Acrobat Reader, Sun Java, Macromedia Flash and everything else that is loaded on their PCs as part of their basic education, on a day I was off school?

The question is whether the GPL (any version) is harder to understand than any other licence. If not, then you don't need a lawyer just for the GPL v3.

Re:Yes it must be by a lawyer.

"but since that faq is not actually part of the GPL it is not binding on anyone's interpretation of the GPL"

IANAL but as far as I understand it this is not actually true. Public statement from the author(s) of a license will affect the decision a judge makes if it ever comes to a trial. IBM lawyers used newletters published by AT&T where they clarified certain points in their license in the SCO case to show that SCO's interpretation of the license was not the same as AT&T's. So if it ever comes down to the question "what does this sentence here actually mean?" a judge will definitely look at any statements made by the FSF.

Where the law and the text of the license are clear there won't be much discussion, but both law and licenses are not science and open for many interpretations, that's why we need judges.

Re:Yes it must be by a lawyer.

Public statement from the author(s) of a license will affect the decision a judge makes if it ever comes to a trial.

This is a good point. When the intent of something is in doubt, it is common for judges to look to the context of things to determine what makes sense. Where things really get hairy is if I use the GPL to license my own work. In that case, am I bound by RMS's interpretation of the license? I may interpret the license differently (and provide my own FAQ about what I think it means). The reality is that in many cases, the FSF view of the license won't matter since the FSF is neither issuing or receiving the license.

Compatibility with patent-nullification licenses

My biggest question is whether GPL3 will be compatible with the many existing "GPL-esque" free licenses [1], which are currently GPL-incompatible, because they contain patent-nullification clauses?

There is a lot of software out there being developed under licenses which aren't compatible with the GPL, because the GPL doesn't allow patent nullification clauses -- this is dangerous, because the purpose of these clauses is to keep someone from slipping code into a major project that they have a patent for, and then torpedoing the whole thing later on when it's crept into wider use. I don't know if this issue just wasn't foreseen when GPL2 was written up, but I can't think of a more pressing issue at the moment.

Yeah, "Tivo-ization" and web services may keep some software out of the hands of the public, but they're not nearly as downright dangerous as submarine patents are.

[1] Examples: IBM Public License 1.0, Common Public License 1.0, Apache License v2, or any of the other licenses where the FSF cheerfully comments "We don't think those patent termination cases are inherently a bad idea, but nonetheless they are incompatible with the GNU GPL."

Re:Compatibility with patent-nullification license

My biggest question is whether GPL3 will be compatible with the many existing "GPL-esque" free licenses, which are currently GPL-incompatible, because they contain patent-nullification clauses?

As it is currently drafted, the GPL v3 has a patent-nullification clause itself: section 11. Furthermore, it allows under section 7 that derivative works have "Additional Terms" that are not in the GPL, but not incompatible with the idea of Free Software. The goal of section is to have less free software licenses that are GPL-incompatible; which directly addresses your question

Examples: (...) Apache License v2

Sadly, the Apache License v2 will probably remain incompatible with the GPL, even GPL v3. As desribed in the Rationale document [fsf.org], section 4.4, not because of the patent termination clause; but because Apache License v2's section 9 states that downstream redistributors must agree to indemnify upstream licensors under certain conditions.

In any case, if you have comments on the latest(L)GPL v3 draft, the FSF's comment page [fsf.org] is the best place to do it. The reason this whole GPL v3 thing takes so much time (the first draft for GPL v3 was published Monday, January 16, 2006!) is that the FSF takes serious comments seriously (and of course, because of certain vendors' deals as well).

Not patents, indemnity

It's not the Apache patent clause that now bars compatibility, it's an indemnity clause.

GPLv3 has been made compatible with Apache's patent clause.

Re:Why is Apache compatibility important?

People want compatibility with various licenses, e.g. GPL v3 and Apache, so that code used in one can also be used in the other by someone other than the original author.

Complexity

If it needs to be explained to intelligent people, it's too complicated.

Re:Complexity

If it needs to be explained to intelligent people, it's too complicated.

Unfortunately, the legal landscape is what's too complicated here. A certain level of complexity is necessary in anything intended to be functional on that landscape.

Re:Complexity

The GPLv2 is one of the simplest, straightforward software licenses I've ever seen. It uses plain English, virtually no legal jargon, and even includes a summary. And I always see people talking about GPL-this and GPL-that who don't appear to have even read it, much less understood it.

Now the GPLv3 is more complicated than the GPLv2, but the main reason for it having to be explained is because so many people already have misconceptions about it from the rumour mill and because of its novelty. I wouldn't say that the necessity for an explanation is inherently a cause for concern.

Re:Complexity

If it needs to be explained to intelligent people, it's too complicated.

Maybe. But legalese is what you get when you take English and try to make it into an exact language, like say a programming. Unfortunately general conversation is absolutely horrible at this, so it is a lot like putting a square peg in a round hole.

That leads to a very odd style with a lot of defintions, not just in the front but through-out the document. Further, unlike say source code where the code is the actual execution, you need to spell everything in detail so that a court can narrow its way down to a detail and say true or false, or hopefully be so clear it never gets there. It certainly tends to get wordy, and information density is very low. Most people fail to see the nuances and see repetition without meaning and react with "All that to say so little?"

Third, you have to assume bad faith on the other party. Some of the paragraphs are specificly there to prevent circumvention attempts, for example legally subdividing yourself deliver signed software with one hand and DRM hardware with the other. Unless you exaplin the whole scenario, that part is almost impossible to understand.

So well, even when you try to make it simple it gets complicated. For example, let me summarize chapter 15: "No warranty and absolute waiver of liability". It is simple, perfectly understandable and would probably get torn apart in court. So you get this blurb:

"THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MER- CHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/ OR CONVEYS THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.

If the disclaimer of warranty and limitation of liability provided above cannot be given local legal effect according to their terms, reviewing courts shall apply local law that most closely approximates an absolute waiver of all civil liability in connection with the Program, unless a warranty or assumption of liability accompanies a copy of the Program in return for a fee."

Now, did it honestly say more than the one line above?

Why GPL3?

With the wonderful growth of open source software and many using the GPL (v2) what is so wrong with it they need to make a new version. For the past couple months that I've been following everything has been pretty much bad concerning V3.

Re:Why GPL3?

I disagree completely.

Yes, the GPLv2 was/is a great license, but it isn't perfect. The FSF's main concern is that certain companies start to distribute software/devices that is in accordance with the letter of the GPLv2, but not with its spirit (specifically freedoms 2 and 3: the freedom to change the program to your own needs, and distribute changes). Example: Tivo.

Also, GPLv2 was written at a time where software patents weren't considered possible, even in the worst nightmares. Today, they are a reality in many parts of the world.

Besides Tivoisation and patents, there are some other good things in the GPL v3, as it is currently drafted. It will be no longer necessary to provide source code via snail mail if you distribute binaries without source. This is the 21st century, providing access to copy the Corresponding Source from a network server at no charge will be sufficient.

Yes, there is resistance to the GPL v3, especially to the earlier drafts (drafts 1 and 2). That's what this long comment period is about. The FSF is taking serious comments seriously. Also, there was opposition in 1991 about GPL v2 as well. Some resistance doesn't mean the whole thing sucks.

And if the GPL v2 continues to be fine by you, you are free to continue to use GPL v2 for your own software. Or dual license "GPL v2 or v3" if you want to stay compatible with GNU software. The FSF would argue that "GPL v2, or at the user's option, any later version" is even better.

Re:Why GPL3?

GPL is a Free Software license, it does not care about "open source" or "growth" at all. Tivoisation and patent deals make existing GPLv2 software effectively non-free software for end users. GPL is for people who cares about end users using their software. They want a license that would guarantee that no one could distribute their software without giving end users their Four Essential Freedoms they deserve. Use a different license if you don't belive end users deserves their Essential Freedoms.

CLEAR!

(RMS applies paddles to GPL3's chest. It twitches once, then lies still again)

RMS: Any response?

(EMS takes pulse of GPL3, shakes his head)

RMS: Alright, let's try watering down some of the shriller provisions. CLEAR! (RMS applies paddles to GPL3's chest again. It twitches, then lays still)

RMS: Anything?

EMS (peering into GPL's eyes) I'm sorry, I'm not seeing any Linus signs whatsoever.

Crow T. Trollbot

Novell - MS Deal Great for GPL v3

The Novell - Microsoft deal was the best thing that could have happened to encourage free software developers to switch to using GPL v3.

Ah, Richard, you're beautiful

From the Groklaw interview:

"there is no possible ethical way you could use [a game console]"

Fantastic. Absolutely fantastic. I haven't laughed so hard in my life.

backward persective.

heinousjay thinks this is funny:

"there is no possible ethical way you could use [a game console]"

What's not funny is how the console may use you. The point of software freedom is to avoid malicious use by the software's owner of the type seen in cell phone tapping. If the software is not free, you can't know what the device is doing. Giving your money to people who abuse you is a bad idea. The viewpoint is extreme, but consistent and sensible.

Now, something that is funny is Steve-o's iPod and Google "brainwash". [slashdot.org] The only reason he does not like either is because they represent another company's product. His language is just as disrespectful as his company and the picture drawn is simple bully.

Re:backward persective.

If the software is not free, you can't know what the device is doing.

I hate to be the devil's advocate on this argument, because I really like free software (though I can't bring myself to whole-kool-aid on the morality argument)...

The software being open-source doesn't give me the ability to know what the device is doing any more than the law being published and accessible gives me the ability to be my own lawyer. It merely allows lawyers (or independent software developers) to exist. Hundreds of thousands of lines of code go into modern working software, and a bit of trust on the part of the average end-consumer is strictly necessary regardless of the visibility nature. The average end-consumer simply doesn't have the time to learn enough computer science to eye-verify every line of code in every piece of software they run.

I trust open-source because many eyes have seen it, and my experience has been that those eyes are not in heads that are actively engaged in the business of doing evil. I trust much closed-source from big companies because the situations where actively lying to the customer is long-term profitable are more rare than many think they are. It's true that only one of these avenues has even the potential for exploitation, but if we always kept to the safe paths we'd miss out on half the fun, eh?

It's not a morality question. It's a risk-reward question.

Patent expansion?

GPLv3 is intended to keep a company from inserting covered by one or more of their patents from then suing downstream users of the modified code. As such it effectively lets them waive SELECTED patents.

But does it also prevent them from suing somebody downstream who inserts (or uses/distributes code where some middle-man inserted) additional code that infringes on OTHER patents than the ones covering what they themselves inserted?

IMHO it SHOULD do the former and not the latter. Otherwise distributing GPLv3ed code would effectively wipe out a company's entire patent portfolio - which would inhibit companies who have and value such a patent collection (if only for defense against others) from using GPLv3.

But IANAL - and haven't even studied the draft. Can someone who understands law AND has studied the draft tell us if this pitfall was avoided?

Tricky section

There was an interesting bit about how they're dealing with the Novell-Microsoft agreement.

The other paragraph, and these are both in section [11], is aimed at the Novell side in the deal, which is, it says that if you distribute the program under an arrangement you made with someone else, to gain promises of patent safety for your customers in a discriminatory way, then you're violating the licence and you lose your right to distribute.

This actually has a few more conditions because we were trying to avoid covering certain other things, for instance, consider a patent parasite, one of those companies that has only one business which is to go around threatening people with patent law suits and making them pay. When this happens, the businesses that are attacked often have no choice but to pay them off. We don't want to put them in a position of being GPL violators as a result. So we put in a condition: "this paragraph applies only if the patent holder makes a business of distributing software". Patent parasites don't. As a result, the victim of the patent parasites is not put in violation by this paragraph.

I'm a little nervous about "this paragraph applies only if the patent holder makes a business of distributing software", does that mean all the parasite has to do is put up a site that says "download a copy of ls for only $10" and they're a distributor? Ok, that's a bit extreme but a lot of patent parasites are dying software companies who would likely be considered distributors. As well it's possible for this clause to be exploited. What's to stop the next SCO from selling all their patents to a patent parasite who in return gives them a license? They've sidestepped the clause entirely and the parasite can threaten to sue whoever they want (except Novell & customers).

Really if one company is attacked for patent infringement on a piece of GPL'd software than every user is vulnerable and it's only a matter of time before the parasite makes the rounds through all the companies. I think the old form was better since it didn't allow a distributor to back down and pay them off, perhaps giving the community to gather and fight back. Of course the old version also allows an unscrupulous competitor to give their patents to a shell company with the understanding that the shell company will put the competition out of business...

Damn, patents suck.

RMS?

Am I the only one who sees "RMS" and thinks "Root mean square"?

Welcome to the World of Nuclear Regulation

Seriously, the situation surrounding the GPL has strong parallels with that of critical portions of the Code of Federal Regulations governing nuclear power (10CFR Part 50)... 10 years ago.

Even with a seemingly simple set of conditions (i.e., the license terms), someone will figure out a nuance or caveat that will challenge the conventional interpretation, and gum up the works for everyone else. Regulations (or licenses) are principally legal matters, and as such, precedents and case studies are key to understanding the bounds of their tenets.

Here's what we use in the nuclear industry to clear the air - as much as can be achieved, anyway. An industry group, like the Electric Power Research Institute or Nuclear Energy Institute, develops a guidance document that defines key terms, and most importantly, illustrates and amplifies the tenets of the regulation. These illustrations include case studies and examples that challenge the rule well beyond the obvious application. Ultimately, such a document may get accepted or endorsed by the Nuclear Regulatory Commission as a Regulatory Guide or other guidance document.

The parallels here are that:

• There is an industry group involved: the FSF (and perhaps others);

• A FAQ has been or (is being) developed for the new GPL. However, from what I've seen, the FAQ may not be comprehensive enough. There are examples, but I don't know that they are that challenging;

• As technology and understanding of the license (or regulation) evolves, there will eventually be new gray areas discovered. This is true in both fields;

Where the situational similarity fails is that there is no overarching authority to endorse (or deny) the final version of the GPL. So the best that could be accomplished really would be to have a rigorous FAQ or guidance document as possible - one that anticipates the future as best as its preparers can.

So tell me, RMS...

The licenses for most software and other practical works are designed to take away your freedom to share and change the works. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software - to make sure the software is free for all its users.

YET ...

we have designed this version of the GPL to prohibit the practice for those products.

So which way is it? Are you giving or just not giving for everyone?

Tivoisation

Caveat: I'm not an expert on exactly how the TiVo operates below the basic bash and web server add-ons that I use on my Series 1

With my semi-informed viewpoint, I can see why TiVo did what they did.

TiVo wasn't really ever in the hardware business, they provide subscription services. If I can modify the kernel I can intercept the low level calls that retrieve the unit number and steal the service by providing a different one.

Then the next step is to let you run a modified kernel, but not let you use the service any longer. I have a feeling that there would still be a tarring and feathering in the works should someone suggest that.

Building out the infrastructure to counter theft of service on the TiVo side of the connection would probably be cost-prohibitive.

TiVo was an awesome product... I still love my TiVo. I would hate to have squashed it before it began because they couldn't afford to counter theft of service.

Just my $.02
-Pyrroc

Overload!

• "...think of free speech, not free beer."
• Zero indexing the four freedoms
• GNU is a recursive acronym
• copyleft

I wish RMS would spend a lot less time working on grammatical cleverness. It's made worse by his compulsive need to explain his jokes (to let you know just how clever he is) every single time he speaks. Get to the point already!

Re:more acceptable

Holy crap! That would mean your project goes from "First Commit to CVS" to "License Change"!

KUDOS!

Re:we need freedom from Richard Stallman

nobody say you to use free software and when i say free software i say gnu general public license, if don't like you use not freesoftware, for example bsd. again about what is freesoftware or not lynksys for his own wifi wrt54g does not give copy of openwrt that use. the same for gnu/linux this is a violation of gnu general public license. the same dlink for openwrt, fon.... and they lock ssh and all the ports to don't use by terminal. they modify, change... but don't give copy of gnu/linux and openwrt and this is one of 4 points of gpl. if you modify, change ,distribute, you must give the changes [e-mail]