Slashdot Mirror
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."
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.
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."
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
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.
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
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
Sadly, the Apache License v2 will probably remain incompatible with the GPL, even GPL v3. As desribed in the Rationale document, 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 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).
FSF have tried all along to achieve compatibility with the Apache licence. The current status is that the patent language is now similar enough for the two licences to be compatible in that respect, but a new problem has been found which would make them incompatible. I expect a solution is being looked for.
This is discussed in the transcript in the section Patent retalisation and the Apache licence (the transcript is split into sections and there is a menu for easy navigation and linking).
Please help publicise swpat.org - the software patents wiki
The Novell - Microsoft deal was the best thing that could have happened to encourage free software developers to switch to using GPL v3.
I'll probably be modded down for this...
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.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
RMS often says what he wants GPL to mean. That is very different from what the legal document means when interpreted as a legal document. This is the source of most of the confusion around most versions of the GPL.
If you read the GPL faq (http://www.gnu.org/licenses/gpl-faq.html) you get FSF's==RMS's interpretation of the GPL, but since that faq is not actually part of the GPL it is not binding on anyone's interpretation of the GPL.
A classic case in GPL2 is the handling of soft vs hard linking which is often debated. This is not mentioned in the GPL, though it is in the LGPL and faq - niether of which are binding on the GPL.
Engineering is the art of compromise.
Holy crap! That would mean your project goes from "First Commit to CVS" to "License Change"!
KUDOS!
"Anybody who tells me I can't use a program because it's not open source, go suck on rms. I'm not interested." (LT 2004)
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.
Bogtha Bogtha Bogtha
In that case, you might want to try complaining about copyright law. Though what copyright tries to achieve (chopping ideas into discreet units and assigning ownership[1]) is so different from how ideas otherwise exist I can't imagine the law every being simple.
In many cases, the GPL makes dealing with copyright less complicated - because it's a de facto standard, because you can focus on the four freedoms instead of the minutiae of the law, because you don't have to hire a lawyer every time you want to let someone use your software, because if you do let someone modify your software you don't have to worry about them turning around and suing you for using with the modified version, because there is a community to provide you with support and might even help you if your license is violated.
[1] I am aware that copyright is supposed to apply to the representation of an idea, not the idea itself. If you can figure out a way to reliabily differentiate the two, maybe you do know how to simplify copyright.
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.
Slashdot - where whining about luck is the new way to make the world you want.
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?
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
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.
I stole this Sig
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?
Live today, because you never know what tomorrow brings
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.
Please help publicise swpat.org - the software patents wiki
Am I the only one who sees "RMS" and thinks "Root mean square"?
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:
- 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.Science never settles, never rests.
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". 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.
DMCA, Hollings, Palladium. What might have sounded like paranoia is now common sense.
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.
I think we've pushed this "anyone can grow up to be president" thing too far.
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
"Democracy is two wolves and a lamb voting on what to have for lunch. Liberty is a well-armed lamb contesting the vote."
These days, using free software is easier and more productive than using non free.
Not if you want to play games, which you might recall was the subject of this thread.
Slashdot - where whining about luck is the new way to make the world you want.