GPLv3 Second Discussion Draft Released

thppft! writes "The second discussion draft of the GNU General Public License version 3 was released, along with the first discussion draft of the GNU Lesser General Public License. Along with the text for the licenses , the GPLv3 website also includes an introduction by Eben Moglen along with markup changes to the rationale and the GPL itself."

Comments

Some more info

[H4x0r Jim Duggan]

After my submission was rejected, I figured another submission based on this story was in the queue, so I put the below links together:

• Video and audio recordings of both days of FSFE's recent international GPLv3 conference

Four transcripts which include the post-talk Q&A sessions from presentations by Richard Stallman and Eben Moglen:

• Transcript of Richard Stallman's talk from the 3rd international GPLv3 conference, Barcelona, June 22nd 2006
• Transcript of Eben Moglens's talk from the 3rd international GPLv3 conference, Barcelona, June 22nd 2006
• Transcript of Richard Stallman's talk from the 2nd international GPLv3 conference, April 21st 2006
• Transcript of Richard Stallman's GPLv3 presentation and Q and A from Torino, Italy, March 18th 2006

And two very useful docs:

• A diff from the first draft to the second
• A diff from the version two to this second draft of v3

Dear Slashdot

[Anonymous Coward]

In the future, please warn when linking to audio files.

Re:Dear Slashdot

[ameoba]

You must be new here - nobody ever actually reads the articles.

Re:more than diff

[foregather]

If you are looking for more that fsfeurope's plain text diff, FSF is providing a strikethrough version of this second Draft, that highlights all the changed text from Draft 1, in LaTex, Postscript, and PDF

Re:What Constitutes Distribution

[AuMatar]

People put code out under the GPL because they want changes to be redistributed. The fact that you could use the code as a webservice was never intended as a feature of the GPL, it was a loophole. It allowed a way for people to redistribute GPLed code as a service without releasing changes. If this change makes it into GPL3 it will be a good thing. IMO, its one of the two most important changes needed in the GPL (the other being patents).

Your option, if you don't like this, is not to use GPLv3 code in your webservice, just like everyone else using GPLed code. You no onger get to have your cake and eat it too.

What are the changes?

[phantomfive]

Looking through the new draft, they've made major improvements to the wording of things, which is good because in the first version the prose was thicker than molasses. Specifically, the part about releasing the keys necessary to run the source (the TIVO clause) is much clearer and easier to understand.
The other changes seem to be patching holes in the logic that might have allowed someone to get around the GPL.

Re:Microsoft doesn't need to do anything...

[AuMatar]

And gcc. And Gnome. And all the GNU utilities. ANd thats only the list of things I've used in the past 15 minutes or so.

Biggest Change

[Stalyn]

The word DRM and the phrase Digital Restrictions Mangement no longer appear in the document. Instead they define a clause called "No Denying Users' Rights through Technical Measures" which is basically the new anti-DRM clause.

Re:Biggest Change

[bcmm]

IMHO this is a very good move. It's a sad thing, but no one knows or cares what DRM is. However, everyone cares if his/her computer won't do what he or she wants. This makes it much easier for a normal person to understand things. Including a plain language explanation is an important strength of this license.

Re:What Constitutes Distribution

[kfg]

TNSTAAFL?

I may be an OSSer, but I have nothing against commercial code, per se. How about this?: choose the set of rules you wish to operate under. If you want to keep the code closed to make money, pay money for your code base. If you wish to use GPLed code, pay for it with GPLed code. And there's always BSD.

The GPL does not exist to promote the development of new and innovate web applications. It exists to promote the development of new and innovate code available to The People. Nor is the GPL the source of Microsoft's FUD. It is not its duty to ammeliorate it, but to oppose it.

Personally I don't really care whether you agree with it or not, but that is what the damned license is for.

KFG

Re:What Constitutes Distribution

[lordcorusa]

One of the things that was discussed regarding the GPL v3 was adding a provision that made web services considered distribution that would require eleasing [sic] the source as per the GPL...

The short story is that this definition of distribution (distribution is now called "conveying" in license language) has been rejected by the FSF and does not appear by default in this draft of GPL3.

The longer story: Some web services projects do want to include a link to allow users to download source, and they do want to limit server administrators from removing this capability. To appease this group, the FSF has added an optional license provision that forbids removal of such a feature. I repeat, this is an optional license feature that takes effect if and only if a given project explicitly activates it.

I suspect that you are right and that most web service providers will not want to use up resources with users downloading web service source. So, I suspect that the market will cause any such projects to diminish in popularity. The important thing to note is that the FSF is not forcing this notion of distribution on any project using the GPLv3.

On a related note, the GPLv3 drafts Section 7b contains a list of optional license restrictions (including the above mentioned restriction) that are permissible. All of these restrictions are things that the FSF does not believe are necessary to maintain a Free program, but that the FSF acknowledges won't seriously harm user freedom if individual projects choose to activate them. Mostly, this list is provided to improve the GPLv3's compatibility with other Free Software licenses which contain equivalent restrictions but are incompatible with GPLv2. This attempt at license compatibility with other Free Software licenses is a big improvement for the GPL.

Alan Cox on GPLv3 - one other thing I should link

[H4x0r Jim Duggan]

I had to forget something. Here's a transcript of comments by Alan Cox.

Re:Applicable?

[rm69990]

Yes, you're right that copyright law doesn't cover use. However, you DO have to have a license to modify the code as well, irregardless of whether you distribute it. As soon as the webmaster modifies the code for use in his web app, he must comply with the GPL. By removing that code that implements this feature, he is modifying the code outside of the terms of the GPL, and thus in violation.

Remember, copyright law places limitations not only on distribution, but also on modification and creation of derivitive works, even if there is no distribution of those modified works. So that is how this clause works. If you made a derivitive of Windows for internal use without Microsoft's permission, you would be infringing their copyright, unless you had a license to do so.

What about this...

[Wannabe Code Monkey]

So I haven't read through the entire draft just yet, but this section jumped out at me:

The Corresponding Source also includes any encryption or authorization keys necessary to install and/or execute modified versions from source code in the recommended or principal context of use, such that they can implement all the same functionality in the same range of circumstances. (For instance, if the work is a DVD player and can play certain DVDs, it must be possible for modified versions to play those DVDs. If the work communicates with an online service, it must be possible for modified versions to communicate with the same online service in the same way such that the service cannot distinguish.) (emphasis mine).

Should it really be fair to restrict some online service to have to treat all clients the same way just because one version was derived from another? Lets say I modify the source code of some browser that is covered by GPLv3. My version has some quirks that make it interpret css differently from the first browser. Would it then be illegal for a website to serve up different css based on my user agent string?

Re:What about this...

[Elladan]

The idea of this requirement is to stop the client from having to differentiate itself to the server. So in this case, the client could (at your - the user's - option) send the same user agent string, or not. It's up to you.

The point is to prevent people from putting out GPL systems that implement DRM client-executable authentication. An example would be if the client was required to transmit a SHA1 hash of its executable image to the server, and the server then decided whether to give you the page or not based on that.

For such a system to pass the GPLv3 requirements, then the correct SHA1 hash for the version the server wants to talk to has to be included along with the source code, and the source code, when you build it, has to send that hash instead of a real hash. In other words, it has to keep working when you build your own copy, even if you change it.

This requirement basically forbids that sort of thing entirely. You can give a hash, and the server can respond to that hash, but it can't be a secret hash. This is the sort of thing DRM systems sometimes do to make sure you can't access your data. For example, games like WoW and EQ do things like this to prevent you from using a GPLed game client.

Re:Microsoft doesn't need to do anything...

[rm69990]

Ummm, if you would learn to bloody read, you would know that the FSF TONED BACK the anti-drm provisions to state that DRM is completely fine, as long as the DRM doesn't impede on a licensees right to access to source code. In other words, if the DRM doesn't affect a licensees rights under the GPL, the DRM is fine, even if it restricts music files, video files, etc.

As for the anti-patent stuff, please explain how YOU would word the license to allow people to distribute works covered by others' patents to all third-parties, royalty free, while giving them the right to do the same. It's simply impossible if the patent holder required royalties, the patent license and GPL would conflict.

I love when people don't RTFA, and make themselves look like idiots in the process.

Re:What Constitutes Distribution

[ObsessiveMathsFreak]

So, if I used a GPLed programming language to make my website, must I offer the full source to that language? What if I use GPLed software to generate a movie, then put it on the web? Now, what if I use GPLed software to generate a random file, then put it on the web? What if that file is html? Where's the line?

Mmmmmm!!! Delicious FUD cakes! Straight from Bullshit Lane Bakeries.

The GPL has never, will never, and can never cover the generated output of any GPL'ed program. This can only occur in the mind of a poor deluded fool, such as yourself. The GPL covers only the source code, and binaries generated from the source code. Not, I repeat, not binaries generated from the binaries, or anything else they might produce.

As for a "GPL'ed programming language", I don't even know what the hell that's supposed to mean. Languages exist independantly of the programs that interpret them, in theory at least. Language symantics cannot be copyrighted anymore than mathematical relationships can.

Re:What Constitutes Distribution

[Just Some Guy]

I may be an OSSer, but I have nothing against commercial code, per se. How about this?: choose the set of rules you wish to operate under. If you want to keep the code closed to make money, pay money for your code base. If you wish to use GPLed code, pay for it with GPLed code.

I have to disagree with you on this one. I'm not convinced that transmitting the output of, say, phpBB2 is the same as distributing the source of phpBB2. If I install it and modify a page, should I be obligated to make my patch available to anyone who views that page? That's not a rhetorical question, by the way, but a real issue that the GPLv3 introduces.

The same could be said of any other GPLv3ed software that uses some kind of templating system to generate output. Word processors, code generators, database frontends, drawing programs: those all incorporate parts of themselves into their end products. Should graphics drawn with The GIMP be GPL because they contain circles made by The GIMP's copyrighted code? If not, then what's the fundamental difference between The GIMP and phpBB that should restrict the output of one and not the other?

I apologize if this comes off like a troll, but I'm really curious. I don't understand this viewpoint but would be interested in seeing it logically supported.

Re:free as in "free to shoot yourself in the foot"

[rm69990]

Ummm, the old GPL allowed you to as well. In-fact, I was talking about the GPL v2, and it was the GPL v2 that I had open in-front of me when I wrote that comment. The GPL v3 increases the restrictions, but against the patent holding companies. Could you please actually read it (both 2 and 3)?

Commercial software companies are unable to filter every single patent in the world to make sure they are not infringing any of them, what makes you think a 2 developer team would be able to? For the GPL to require anyone to do so would be impractical...hell, to require Microsoft to do so would be impractical. The patent system is broken, and it is impossible to do what you think the GPL should require developers to do. The FSF found a balance, only requiring a ceasing of distribution once the patent holder requires it.

Do you honestly think every distributor of software, both proprietary and open source, knows every single patent in the world and whether or not they infringe it? If you do, you are incredibly naive.

Re:What Constitutes Distribution

[lordcorusa]

Oh, joy. Now, when trying to use multiple open source projects, we can't even assume that two GPLv3 projects have compatible licenses. "libAardvark is GPLv3 with restrictions 4, 7, and 19, and gLlamaBoy is GPLv3 with restrictions 1, 8, and 21-36. We'll have to rewrite one of them."

Remember, any GPLvX code can automatically be linked to any other GPLvX code (although not necessarily to GPLvX-1 code). The allowable optional restrictions of section 7b do not impose contradictory burdens (that is, option 2 does not contradict option 3, etc) nor do they really add significant burdens to the vanilla GPLv3 (with the exception of 7.b.4 which is the web services option and the one everyone is still up in arms about). They are just minor differences in effect and the purpose of Section 7 is to allow modules with such minor licensing differences to be linked.

In fact, not only is the purpose of this section to allow you to link code under all permutations of the GPLv3, it also allows you to link GPLv3 code to various Free Software licenses that were not previously linkable due to minor wording differences or patent retaliation clauses. In fact, the controversial 7.b.4 section was to allow linking with the Affero Free Public License. The big debate should not just be whether 7.b.4 is okay, but also whether the Affero Free Public License is really a Free license.

Treacherous

[wiredlogic]

*2 DRM becomes nastier when based on Treacherous Computing and other changes
in computer hardware which deny users the possibility of running modified or alternate
programs.

It looks like they had RMS personally writing the footnotes for this one.