Free Software Foundation Begins Rewriting the GPL

Robert writes "The first update to the GNU General Public License in 15 years has begun. Details about the process and guidelines by which it will be updated by the Free Software Foundation, and the free/open source community at large, are now available. The FSF has announced plans to release the first draft of the new license for comment at a conference to be held at the Massachusetts Institute of Technology in mid-January 2006." From the article: "This is the first time the GPL has been open to a public development process. Stallman created version 1 himself in 1985 and introduced version 2 in 1991 after taking legal advice and collecting developer opinion. The rapid adoption of Linux and hundred of other software products licensed under the GPL makes the development of GPLv3 a significant event, and one that is now likely to involve some of the biggest vendors in the industry, with Hewlett-Packard, Novell, and Red Hat already having declared their intention to participate."

Comments

The largest amendment to this new draft...

[slughead]

... are the words "no, seriously."

Re:The largest amendment to this new draft...

[krgallagher]

"The largest amendment to this new draft..."

The requirement that all free software be called GNU/Software.

My First Question

[Ckwop]

Why? The GPL2 does everything I want it to.

Simon.

Re:My First Question

[zx75]

Then stick with using GPL v2. But just because it is adequate for your needs, doesn't mean that it neccisarily addresses the concerns of everyone who chooses to use it. Hence the rewrite.

Re:My First Question

[Ulrich Hobelmann]

One issue is that some companies use GPLed software and modify and extend it, but don't release it (the GPL only requires you to publish your modifications if you release the software). But these companies run the modified software on their webservers, so it is in use.

Now that more and more applications run simply over the web, with no publishing involved, some people (like RMS) are interested to extend the concept of Free Software to web apps.

Re:My First Question

[LiquidCoooled]

Doesn't this dilute the skills pot though.

If I see some GPL code, I cannot just use it. I will have to check if its the correct version of GPL before I can bring it in.

This will end in tears.

Re:My First Question

[SwiftOne]

There are a few issues that the GPLv2 doesn't cover, or is a bit too vague on. As I understand the desired improvements, some big points are:

* Language that is happier with different jurisdictions. (some legal terms have very different meanings in different countries)
* Patents. Patents Icky. Dealing with Patents Icky.
* Wrapping binaries. I think some parties want some more clear language here to prevent violations of the spirit of the GPL.
* with GPLv2, if you expose the service of the software but not the binary, you don't have to distribute changes. So I could take slash code (if it's GPL, which I don't recall), hack some changes, and sell access to the website using those changes, and never have to share my code, which violates the spirit of the GPL.

I don't think the idea is to toss the GPLv2, but instead to keep doing the same thing...only more so.

Re:My First Question

[Bogtha]

Now that more and more applications run simply over the web, with no publishing involved, some people (like RMS) are interested to extend the concept of Free Software to web apps.

The big problem is that this changes the GPL into a EULA. Right now, the GPL doesn't attempt to restrict anything, it merely grants privileges that would not usually be in effect. That's why it's such a strong license.

To change the GPL to include restrictions on how you use the software would seem to run counter to the ideals of Free Software; namely that you are free to use the software as you please. It's also vulnerable to the same criticisms of other EULAs. Basically, the only thing that allows copyright holders to bind you to terms is the fact that you are copying. But copying for the purpose of using the software (e.g. installation) is explicitly not copyright infringement under USA law. That means that if you are merely using the software, the copyright holder has no leverage to bind you to their terms.

Why does Linux make this important?

[N3Roaster]

I keep reading about this, and I can never figure why the rapid adoption of Linux makes GPLv3 important. Sure, there are a lot of projects that use the GPL that allow distribution under the current or any future version of the GPL, but Linux isn't one of them. From /usr/src/linux/COPYING

Also note that the only valid version of the GPL as far as the kernel is concerned is _this_ particular version of the license (ie v2, not v2.2 or v3.x or whatever), unless explicitly otherwise stated.

So what's the Linux connection here?

Re:Why does Linux make this important?

[cortana]

When the article meantioned "Linux" it meant "GNU/Linux" as opposed to Linux-the-kernel. A log of GNU/Linux software uses the GPL with the upgrade clause--hence it is important.

Stallman's GPLv3 mailing list

[ubiquitin]

Anyone subscribe to Stallman's new mailing list?
http://www.gplv3.fsf.org/index05

I hesitated because it didn't just say "subscribe".
The submit button says "I want to participate." which is hard to do without knowing exactly what you're participating in first.

Reasons for a rewrite ?

[johnhennessy]

Are there issues with the current GPL that need to be fixed ? Or even some parts that need to be clarified ?

If there are actual issues with the license, then a rewrite is a good thing - all I'm concerned about is that people don't waste time developing a new license when one isn't needed. In the end, its adoption will be decided by the various projects - on a case by case basis, so just because there is a version 3, doesn't guarantee adoption, unless it brings benefits. ... And hopefully it doesn't spawn pro- and anti- GPLv3 wars in every GPLv2 project ! While licensing is important, it shouldn't create huge overheads that distract developers from doing what they do best.

Re:Reasons for a rewrite ?

[rkcallaghan]

Are there issues with the current GPL that need to be fixed ?

The GPLv2 doesn't properly deal with patent issues, only copyright. Thus, with some legal smoke and mirrors, it is possible to comply with the letter of the GPLv2, gaining free use of other GPLv2 code in the process, while shipping your code/product under patent restrictions, preventing it from being redistrobutable or adjustable freely.

~Rebecca

Re:Reasons for a rewrite ?

[j00bar]

There are most definitely reasons for a rewrite, and most of them have to with developments that have taken our industry by storm since 1991 and which will continue to impact us in the near future. The GPLv2 does not successfully account for many of these.

Some questions which will likely be considered in the GPLv3 drafting process:

1) Back in 1991, the GPL was written centered on specifics to United States Copyright Law. With the diversification of international copyright law since the Berne convention -- some countries have implemented various manifestations of DMCA like laws, others have not -- how does a license that must govern international transactions of copyright account for these discrepancies?

2) How can software patents encumber free software? For example, let's say I write a word processor that is licensed under the GPLv2 and I submit and receive a patent for my word processor document format. If you write a derivative work of my word processor, are you infringing on my patent? Does that violate the principles of software freedom?

3) How does Trusted Computing encumber free software? For example, let's say I write the software for a DVR that uses GPL software and is licensed under the GPL. But let's further say that my DVR used TPM, and it won't run the DVR software if it is not signed with my private key. You can modify the source, and you may even be able to load a modified binary back onto the DVR, but without me signing your binary, it won't run. Does that violate the principles of software freedom?

I don't know the answers. They haven't been decided yet. These may not be all the questions -- they may not be among the questions. But that these questions are out there are symbolic of the need for a community-driven effort to reassess the future of software freedom.

The GPLv3 process will be a discussion of the free software community on how we can best ensure that the essential freedoms the GPL tries to protect are in fact protectable. And though rms is the final arbiter of what GPLv3 will actually be, these are questions that we the free software community as a whole need to discuss.

-jag, a.k.a. jag@fsf.org

Re:But that's not an option.

[mindstrm]

They are not forced into anything whatseover, ever.

When you download my code, licended under, say, GPLv2, and it says this.. it means that you may distribute deriviations of my code under the GPLv2 if you wish, or a later version, if you wish. How you want to apply this is up to you; I force you to do nothing.
If you wish to include the same clause, allowing future versions to be used, that was completely your choice.

Nobody is forced into anything at all.. the rightsholders deliberately gave you the right to pick a later version of the GPL becaues they trust the FSF.

Simple, right?

We don't need Microsoft to create "FUD"

[CyricZ]

Having talked to various developers, these recent licensing shenanigans have not been particularly good for the open source community.

Many developers wish to make use of open source software, but are getting to the point where they're not sure what exactly they're allowed to do with some particular piece of software.

These developers are not lawyers, and do not want to waste their time trying to figure out fairly complex licenses. Individual consultants and smaller development firms can't necessarily afford to hire a lawyer to verify that they're complying with the terms of all the licenses their project may be subjected to.

I know many professional developers who won't even touch LGPL'ed libraries. They stick with software released under the BSD license, for instance, because it has very clear and concise terms. They know what they can do with such software, and thus can focus on developing solutions, rather than getting bogged down in legal nonsense.

While the GPL v3 may offer some degree of protection with respect to patents, any such benefits may be mitigated by the fact that many developers out there are not interested in becoming lawyers. They don't want to get bogged down trying to interpret relatively complex licenses.

Re:We don't need Microsoft to create "FUD"

[real gumby]

That's cool. For some software I use a 3P license and pay a cash royalty; I generally use the compiler-included helper libraries (arithmetic conversions and startup code end the like) without worrying about it, and most of the time of use free code of various stripes. In all of those cases I check the license -- sometimes it's quite enlightening even in the case where you think you paid the developer $25K for a royalty-free buy-out! To me these are all just special cases of the same thing. The fact is that _regardless_ of where the code came from you have to do this checking. If you optimize by finding one or two that happen to work for you, great. The mere existence of someone who chose BSD as the one to standardize on doesn't especially distinguish it -- I can easily point to companies who only use code they pay for, for whatever reason.

By the way I wrote the original (?13 years ago?) draft that became the 1.0 LGPL you mention. John Gilmore suggested to me what turned out to be the most important clause: that you could dynamically link to such a library without having any licensing impact at all. Those terms are more liberal than any cash-royalty license I've ever seen (and don't even include the announcement clause of the BSD license), and made Linux's userland possible.

GPLv3 != EULA

[ObsessiveMathsFreak]

The big problem is that this changes the GPL into a EULA. Right now, the GPL doesn't attempt to restrict anything, it merely grants privileges that would not usually be in effect. That's why it's such a strong license.

To change the GPL to include restrictions on how you use the software would seem to run counter to the ideals of Free Software; namely that you are free to use the software as you please.

I would disagree with this interpretation.

The GPLv2 has never said anything nor placed restriction on how you use the software. In effect, the GPL only comes into play when you some to redistriute the software, and says that you must redistribute the source code when you distribute the binaries to others.

The GPLv3 again will not say anything or place restriction on how you, you use the software. You're still free to tinker as you please privately and keep the changes to yourself. Again on distribution you must include those changes.

However the GPLv3, in response to potential or actual shenannegans with web deployment, will specify that when you also come to offer your software as a service to users, you must also include the changes to your code, and make them and the original code visible to those users.

This makes a lot of sense. When I run a web app for users acting as thin clients, I'm effectively distributing my program to them. Albiet now the licence is for an extremely limited time, and the calculations are taking place on my machine. However, in effect, a binary of my program has been, however temporarily, placed at the disposal of that one user. They are a user after all. they are "using" the program.

In a way the GPLv3 is a lot better than GPLv2. The GPLv2 only covered the distribution of the binary of the program. GPLv3 covers the service of the program, or more succinctly, the program itself. If you offer the service of the program to someone else, directly, in whatever way, then you must show them your source code. You can see that binaries fall under this definition as well.

Of course companies will try to write wrapper programs to get around this, so that users are not directly using their app, etc, etc, etc. However, I think most won't go to the bother and will just publish their code. After all, how many trade secrets are going to be in your average php/asp page anyway.

It's all about making sure that users are empowered, and that software is both transparent and modifyable to everyone. Stallman originally argued on the grounds of modifyability, which of course is critical to the whole process. But the transparency conferred by this has benefits for the public at large which outweight even the benefits of modifyability. But you need modifyability in order for software to be free. Just having transparency would be like am autocracy having transparency in its government offices. You might be able to see what's going on, but you still can't do anything about it.