GPLv3 Released

A GNU Dawn writes "The GPL v3 has just been released. Among other things, the released version grandfathers in the Novell deal so that Microsoft's SLES coupons will undermine their patent threats, replaces references to the Magnusson-Moss Warranty Act with more specific language, and clarifies that using BitTorrent to convey a GPLed work is not a breach of the license (it might be one, technically, in GPLv2). The GPL FAQ has been updated to cover the new changes." Commentary is available over at Linux.com (which is owned, along with Slashdot, by Sourceforge).

Comments

Yeah, but

[wiredog]

Does it run on an iPhone?

Re:Yeah, but

[shawnce]

...no but version 3 of the GPL will save us from the iPhone.

iPhone restricts users, GPLv3 frees them

Re:Yeah, but

[Anonymous Coward]

I'd love to see the GPLv3 on my iPhone! In fact, I'm reading this on my iPhone right now! The AT&T network is so gr39jf9,2c0mnklzpsp][/l/9 56(%*^

[CARRIER LOST]

Re:Yeah, but

[farmer11]

This part is interesting (from iPhone restricts users, GPLv3 frees them),

The GNU GPL is used by developers with various views, but it was written to serve the ethical goals of the free software movement. Says Stallman, "The GNU GPL makes sense in terms of its purpose: freedom and social solidarity. Trying to understand it in terms of the goals and values of open source is like trying understand a CD drive's retractable drawer as a cupholder. You can use it for that, but that is not what it was designed for."

Reminds me of the recent discussion on the linux kernel mailing list about Linus's interpretation of the spirit of the GPL.

Re:Yeah, but

[Anonymous Coward]

I have it on good authority, from a girl, that Stallman has a BIG ass dick. She held her hands apart like she was telling a fish story.

Seriously.

You brought it up.

Re:Yeah, but

[qualidafial]

In Utah, this is what we call a "faith promoting rumor."

Re:Yeah, but

[tkinnun0]

Perhaps it's time for BSD version 2, for example:

Redistribution and use in source and binary or other forms, with or without modification, are permitted provided that the following conditions are met:

* Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.

* Redistributions in binary or other forms must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.

* Neither the name of [original copyright holder] nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.

* This list of conditions is the only list of conditions imposed on redistributions in source and binary or other forms.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Re:Yeah, but

[jZnat]

So, uh, what's the point of this? It even has the obnoxious advertising-clause that makes it incompatible with several widely-used licenses out there (defeating the purpose of making a developers-are-free sort of license).

Re:Yeah, but

[tkinnun0]

Really, because the page I got it says "This is the BSD license without the obnoxious advertising clause". Are you sure you know what you're talking about?

iRonic

[vigmeister]

That it should be released on the same day as a device that teased otherwise, but was virtually closed to 3rd party development.
Cheers!

Re:iRonic

[fsmunoz]

Well, ironic, but perhaps not just a coincidence...

BOSTON, Massachusetts, USA -- Thursday, June 28, 2007 -- On Friday, June 29, not everyone in the continental U.S. will be waiting in line to purchase a $500 (370) iPhone. In fact, hundreds of thousands of digital aficionados around the globe won't be standing in line at all, for June 29 marks the release of version 3 of the GNU General Public License (GPL). Version 2 of the GPL governs the world's largest body of free software -- software that is radically reshaping the industry and threatening the proprietary technology model represented by the iPhone. (...) The iPhone is leaving people questioning: Does it contain GPLed software? What impact will the GPLv3 have on the long-term prospects for devices like the iPhone that are built to keep their owners frustrated? Peter Brown, executive director of the FSF said, "Tomorrow, Steve Jobs and Apple release a product crippled with proprietary software and digital restrictions: crippled, because a device that isn't under the control of its owner works against the interests of its owner. We know that Apple has built its operating system, OS X, and its web browser Safari, using GPL-covered work -- it will be interesting to see to what extent the iPhone uses GPLed software." The GNU GPL version 3 will be released at 12:00pm (EDT) -- six hours before the release of the iPhone -- bringing to a close eighteen months of public outreach and comment, in revision of the world's most popular free software license.

Irony - I love it

[$RANDOMLUSER]

The cute saying of the minute at the bottom of the page is:

An effective way to deal with predators is to taste terrible.

I think that says it all.

Re:Irony - I love it

[dpilot]

But it's only effective for your relatives. Maybe you've been spit out, but you're still dead.

At this point the analogy to OSS breaks down, because in this case the predator is after the kill, not the meal.

Re:Irony - I love it

[Mattintosh]

Only let them get a bite... It works better if you're poisonous. "Poison, poison, poison... Tastyfish!"

Re:Irony - I love it

[farmkid]

True, but that's the way evolution works: it tends to preserve the gene, not the individual. If your relatives survive as a result of lessons learned by your death: success.

Re:Irony - I love it

[Plutonite]

Good. One or two pieces of FOSS are brutalized, OSS in general lives on. I think I am missing something (no sarcasm).

Re:Irony - I love it

[dpilot]

OK 2 pieces of OSS to get brutalized.

#1 gcc
#2 glibc

But my other point was that in the predator analogy, part of the learning was by taste and part of the learning was by killing the predator. Genetic survival cuts both ways - predators who eat poisonous prey don't make offspring.

What's unique in this situation is that we have a predator that is only interested in killing, not in eating, not even in tasting. So maybe tasting bad simply doesn't matter, because one can kill without tasting. In that sense, it's not really a predator at all, it's a competitor out for the same food supply that competes by killing its competition.

Re:Irony - I love it

[Plutonite]

That's pretty interesting with avalanche-effect things like libc..etc, but isn't it possible to almost effortlessly license new, and equally free (offspring) versions given GPL3, which would also require to be killed?

Re:Irony - I love it

[dpilot]

Depends on how they "kill" it. Using patents, it's "dead" no matter what the derivation of the codebase, at least until worked around. If it's merely predatory corporate practices, obviously either or any could live on. Things like gcc and glibc have a pretty darned clean and open pedigree, from what I understand, so copyright attacks would likely not work.

Re:Irony - I love it

[Walter Carver]

The predator knows you taste terribly and that's why he doesn't even attempt to eat you. You taste is preventing him. This is not meant as a retaliation after he attempts to eat you.

Glad I sat in line

[Anonymous Coward]

I've been sitting at my computer eating nothing but hot pockets and red bull waiting for this. I got up only for bathroom breaks.
It's been worth it. Now I have to print it out so I can fall asleep with it.

Re:Glad I sat in line

[Antique Geekmeister]

Perhaps in honor if Mr. Stallman's hygiene habits, you could remember to bathe this year?

Re:Glad I sat in line

[Braino420]

I got up only for bathroom breaks. It's been worth it.

Pfft, if you were really hardcore, you would have worn diapers.

Re:Glad I sat in line

[sumdumass]

He is a geek, not an astronaut.

Yeah, but

[Captain Splendid]

Is Linus happy with it?

Oh My!

[Bohemoth2]

My head hurts now.

To all those complainers

[ILuvRamen]

For all those out there that are going to complain about certain parts of it, let me be the first to say that if you don't like it, DON'T USE IT! Write your own damn license.

Re:To all those complainers

[Archangel Michael]

"Write your own damn license."

Because the world needs yet another open source license. In the world of Idealists someone needs to be pragmatic. Find an existing license that gives you 99% of what you want. You'll find that you don't need the other 1%.

Re:To all those complainers

[BrianGKUAC]

Because the world needs yet another open source [browser, window manager, package management system, text editor, torrent client, ftp client, full distro, etc.] . In the world of Idealists someone needs to be pragmatic. Find an existing [browser, window manager, package management system, text editor, torrent client, ftp client, full distro, etc.] that gives you 99% of what you want. You'll find that you don't need the other 1%.

Sound familiar?

Re:To all those complainers

[Archangel Michael]

Yes.

How many [browser, window manager, package management system, text editor, torrent client, ftp client, full distro, etc.] do you need (choice wise) before you have 99% of what you need in what is available??

The answer is "one", unless the one that exists doesn't have what you need (security, stability, features etc).

There are, in the world of "open source" and "free software" already a myriad of licenses. I can name about 12 off the top of my head. Each, more or less, provides the use with free software that is community supported. Some licenses are really free (BSD), others not so (GPL3).

I can see the usefulness of most of these license and why someone would choose one or the other. Personally I'm a BSD style kind of guy, free is free is free. But I also understand the reasoning behind the likes of GPL3 and other restrictive licenses (ie control of the code and everything the code touches).

There is, however, a broad range of choices out there, all "free" or "open" licenses.

BTW, my opinion on GPL3 is the unintended consequences of GPL3 is going to be large fractures withing the Gnu/Linux/Gnome(KDE) environment, requiring seperated "distribution" channels. I can foresee the need for two or more install media to separate out what is GPL2, GPL3, and other licensed software, where one install media would do (RedHat GPL2 disk, RedHat GPL3 disk, RedHat Apache Disk etc). There is going to be a huge artificial wall of separation that is going to be problematic in the long term, mainly because it is not needed, or required by end users, but rather needed by licensing requirements.

Re:To all those complainers

[zsau]

Real question is, how many proprietary software licences are there? Just about every proprietary software producer has at least one...

Re:To all those complainers

[level_headed_midwest]

Yes, but they all say pretty much the same things:

1. You are only buying a license and cannot copy or redistribute.
2. You have no access to the source code.
3. The license is not transferrable, either from person to person (can't sell it), machine to machine (can't take from one computer, put on another) or from media to media (can't take from a CD and put on a DVD.)
4. We are not liable for any damage the software causes due to malfunctioning (this is shared with the GPL as well.)

Re:To all those complainers

[the_womble]

In what way is the GPL not free? It prevents you restricting the freedom of others. You can do exactly what you like with GPLed code: change it, sell it, use it as a platform for software using a different license, etc.

I can foresee the need for two or more install media to separate out what is GPL2, GPL3, and other licensed software

An interesting twist on anti-GPL FUD. As GPL software is already distributed on the same install media as BSD, MIT, MPL, QPL and proprietary licences, why do you think seperate install media are going to be needed for different versions of the GPL?

The only problem with GPL 3 is that it is difficult to move software with a large number of compyright holders from 2 to 3, if the "and later versions" wording was not used to start with. This will mean that that will be some GPL software that can not be statically linked with other GPL software. A lot of what people want to link will be LGPLed or under BSD style licences, which will mitigate the problem.

One project that might hae a problem is KDE, Does anyone know if Qt going to use GPL3?

Re:To all those complainers

[Tony Hoyle]

It's certainly going to murder the cross-development that happens at the moment. gplv2 and v3 are essentially incompatible in that they both have the viral aspect - you can't include gplv3 code without licensing your entire app that way, and since large chunks of code are v2 only would have to throw out or rewrite sections of it, which just means the new code won't go in.

I've got to face the annoying prospect of introducing submission rules to the projects I edit now, because if anyone copies in any v3 code we could be in trouble.. it's no longer enough to say 'this bit is gpl, this bit is lgpl, don't muck around with the licenses and keep the lgpl code independent of the gpl code'. Now we have to talk about license versions, the reason why gplv3 (and lgplv3 if it exists) are verbotten and probably end up in a half a dozen heated debates from rms fanboys.

Re:To all those complainers

[sumdumass]

BTW, my opinion on GPL3 is the unintended consequences of GPL3 is going to be large fractures withing the Gnu/Linux/Gnome(KDE) environment, requiring seperated "distribution" channels. I can foresee the need for two or more install media to separate out what is GPL2, GPL3, and other licensed software, where one install media would do (RedHat GPL2 disk, RedHat GPL3 disk, RedHat Apache Disk etc). There is going to be a huge artificial wall of separation that is going to be problematic in the long term, mainly because it is not needed, or required by end users, but rather needed by licensing requirements.

You cannot say they are unintended. They were known and talked about in great detail. If anything they are intended consequences.

Another intended consequence is that a third party software maker can craft their license into a discriminatory license that stops people from participating in the Four freedoms the GPL supposedly protects. Think about how many users that are tied to a piece of software that runs on windows. Now think about how many of them can only get support when it runs on windows. Now think about windows creating a little novel out of everyone who buys their software and clicks yes on the agreement after installing it. They would have effectively stopped the vast majority of people using windows from distributing any GPLv3 licensed software. They can use, but they cannot share their improvements, modifications or the software itself because of the new GPL software license. And all they have to do to make it a situation where "you make a payment based on", is to offer the software sans the little novell deal for ten time the normal price. And of course everyone will by the cheaper stuff.

A simple note saying "you cannot trigger this clause by the inclusion of a discriminator patent license inside the license of another software product" but they didn't. And they knew about it. So add that to your Chaos, I'm willing to bet some troll is sitting in wait until someone buy a license like this and then sues the FSF or some other project using the GPLv3 for lack for enforcement or something stupid like that when a contributer is part of a discriminatory deal because they bought a new computer with windows on it. I imagine that is the person cannot convey a covered work, they cannot contribute to a project or even be part of it if they are using the GPLv3 license.

And none of this is unintended, it is all intended because it was known and allowed to happen or become the case.

Re:To all those complainers

[sumdumass]

I was reading an article about this exact situation. The GPLed software on the BSDs are used in a way that people know they are GPLed. So in essence, they are separated but on different media. As far as I know though, the GPLed stuff isn't incompatible with the BSD licenses which is a little different case here. You would have to manage to keep the various versions separate for this reason. I don't know it is would require separate media though.

Re:To all those complainers

[zsau]

All free software licences say pretty much the same thing too, tho, by that standard. Different free software licences have different precise requirements about what to do if you're reused them, but if they're considered free or open source by the FSF or OSI, you don't have much to worry about. Different proprietry software licences have different precise requirements about e.g. whether you can transfer them from one computer to another or how many CPUs/Cores your computer can have, but as long as you put it on one computer and don't try to transfer/change it (and keep your licence/proof of purchase), you don't have much to worry about.

Re:To all those complainers

[Brotherred]

"There is going to be a huge artificial wall of separation that is going to be problematic in the long term, mainly because it is not needed..." Disagree. What is artificial is the invisible hardware issue in the Tivo and every other non Microsoft DVR that does not except modified code or custom devices (larger hard disks) added to the system. You say hardwere I say software (checksum). Firmware is still code no matter where it is put.

Let me be the first to say...

[niceone]

Let me be the first to say: GPLv4 - it's going to be terrible, it's unnecessary and unwanted, probably will destroy linux and maybe the world too. GPLv4 will eat babies!

Re:Let me be the first to say...

[Anonymous Coward]

Sorry, you aren't the first to say. Rumoured clause to be included in GPLv4 will mandate the usage of GNU/Emacs for all GPL software development.. Eating babies is just about the least evil thing it will do!

Re:Let me be the first to say...

[niceone]

Wow, not by a long way:

gplv4.org
Registrar: Go Daddy Software, Inc. (R91-LROR)
Creation Date: 2005-12-14 16:52:22

OK, the first on this thread then.
I'm all for the emacs thing though.

tivoisation

[Kuciwalker]

I don't understand the furor over Tivoisation, and think it's a really bad move on the FSF's part to ban it. In all the other articles I've seen tons of comments along the lines of "well if Tivo doesn't want to give back, then they can't use my code!" The thing is, Tivo does give back - they contribute any source code they add. And there's nothing preventing you from using that code on another system (with similar hardware, of course). I don't see the benefit in forcing them to open up their hardware just because they want to use GPLv3 software on it. For most devices like this, it's important to the proper functioning of the network for the servers to be able to trust the clients, and so there have to be limitations to the software you can run on that device. GPLv3 won't convince Tivo (or others) to open their hardware, it will just force all of them to stick with GPLv2 code. That hurts adoption of Free Software.

The irritating part is that the FSF has the business products exception, where Tivoisation is okay for hardware sold for business use. Stallman et. al. recognize that in some cases it's ultimately beneficial to the user to be unable to run modified software (e.g. a business that has to have accountability, or a console gamer who wants to know that no one is running a hacked game in multiplayer), but they think they can somehow figure out where that line is for everyone.

Re:tivoisation

[zellyn]

That's fine if it were just Tivos. However, what would happen if every piece of hardware you bought was Tivoised? Only properly signed binaries would run at all. I can think of several companies that would love that situation.

Re:tivoisation

[gad_zuki!]

I dont like it either and I think in the long run its going to hurt oss adoption, but at the same time its almost trivial for manufacturers to 'tivoise' all their equipment. If this happens en masse it would hurt oss quite a bit. I'm not sure who wins in this scenario, but the neckbeards have decided to shoot back and declare hardware as theirs. This is something to consider before releasing anything under v3.

Re:tivoisation

[Kuciwalker]

That's fine if it were just Tivos. However, what would happen if every piece of hardware you bought was Tivoised?

Then I wouldn't be able to run the FSF's software on any of it, by their design!

But, let's be realistic. It would be absolutely impossible to Tivoise all general-purpose computers. And it's not like you couldn't build your own.

Re:tivoisation

[UtucXul]

I don't see the benefit in forcing them to open up their hardware just because they want to use GPLv3 software on it.

No one is asking them to open up their hardware. As far as I'm concerned they can do anything they want with their hardware. The problem is that they want to lock MY hardware and won't give me the key. I purchased a series 2 TiVo. I understand that software has licenses so I don't own the software when I buy it in a way. But the TiVo box is a physical item that I gave a store money for. The only owner of that box sitting under my TV is me (and I guess my wife).

Re:tivoisation

[Anonymous Coward]

The spirit of the GPL is that of freedom for the user. Tivoisation, the restriction of actually modifying the code on the hardware, was an unforeseen issue when GPLv2 (which guarantees that right) was written and thusly had to be corrected in GPLv3.

If you don't understand the significance of that then I invite you to a thought-experiment of what it would be like in a world where you can see the source, but not modify it, even though that right is guaranteed by the license of the software. Oh, and in regard to your comment of using the code on similar hardware. What if said hardware was proprietary and only available through the systems provider? What then?

As an aside, where does it state that GPL licensed software must "give back?" As I read it, it merely states that you must pass on the code and the same rights to your clients. If the clients are public, "at large" then I can understand the concept that the source is released publicly and considered to be giving back. But, this concept of upstream patches that is constantly heralded on Slashdot I just don't understand. Who is espousing this misinformation? Heck, even the concept of forking would be moot if upstream patches were mandatory.

Re:tivoisation

[R2.0]

Don't kid yourself - your wife is the only owner - she just pretends otherwise.

Re:tivoisation

[Kuciwalker]

The spirit of the GPL is that of freedom for the user. Tivoisation, the restriction of actually modifying the code on the hardware, was an unforeseen issue when GPLv2 (which guarantees that right) was written and thusly had to be corrected in GPLv3.

And yet the FSF went to great lengths to permit it in some cases (the business use exception), recognizing that Tivoisation isn't a restriction on fundamental freedom, and in fact in many cases is beneficial to the user. That's why they have the convoluted definition of "consumer device" so that they can distinguish between consumer and business products - because lots of businesses have an interest in devices that will only run signed code. And I gave an example of where the consumer wants the same thing - games, where the user wants to know that everyone is running the same version in multiplayer. In general, it's often good for networked devices (other than general-purpose computers) to only run signed code, because it makes it significantly easier to guarantee network stability. So even if we accept that there are cases where Tivoisation is bad, and that the FSF ought to try to prevent them, we're left with the fact that it can only do so with a broad brush, eliminating a lot of good uses along with a lot of bad ones. They're taking away as much freedom for the user (freedom to use GPLv3 on their trusted platform while maintaining that trust) as they're giving.

Re:tivoisation

[Evets]

It's pretty simple. If I buy a product, let's say it's a PDA and the PDA interface uses software I developed. I find a major security hole in my software, so I want to patch it.

Tivoisation prevents me from patching it. It prevents me from adding features. It prevents me from fixing bugs.

It's my software. I want to update it, and the only thing preventing that is licensing. That and a hardware based security system enforcing the licensing. When I wrote the software originally, I did not intend for products to be developed using my software that I could not update. In fact, I licensed my software in order to prevent that kind of a thing. Unfortunately, the PDA manufacturer found a loophole that stuck to the letter of my license, but not the spirit of it.

This is pretty much what Tivo did. The anti-Tivoisation language in the new GPL effectively closes that loophole.

Now, if Tivo wants to do the same thing in the future, they can either utilize software who's authors don't mind (and it's widely available for what Tivo wants to do), or they can contact the original author team for alternative licensing. In the case of large scale community projects where no such licensing option exists, they can either stick to the license or they can develop their own similar project in-house.

All the FSF is doing is ensuring that software licensees abide by the spirit of what the original authors intended. They recognized that business enterprises were taking advantage by using loopholes and they've attacked that problem - pretty effectively.

---
I can't think of a single instance where Tivo actually contributed code back to the community as you state.

Regardless, they are not without free options. All they need to do is shift to BSD licensed products - which they should have done in the first place given their long term strategy.

Re:tivoisation

[bill_mcgonigle]

For most devices like this, it's important to the proper functioning of the network for the servers to be able to trust the clients, and so there have to be limitations to the software you can run on that device.

Then they should pick a software product based on a license that lets you do that. To oversimplify, but in a way that most people can grok, the BSD line maximizes the rights of the individual user(developer-sense) of software. The GPL has always been about maximizing the rights of the community, at the expense of what the user wants (to keep his changes secret, e.g.). GPL3 just clarifies some of the loopholes found in GPL2 that minimize community rights.

TiVo clearly takes rights away from the end-users of the hardware they've purchased, and this is contrary to the spirit of the GPL.

Re:tivoisation

[TrueJim]

It's like a restaurant chef refusing to tell you what's in your dinner. Sure, it's the chef's recipe...but it's your dinner. You're the one about to put the mystery food into your belly. You have a right to know what's going to go into your stomach. That doesn't mean you intend to steal the chef's recipe, but daggumit...it's your stomach! You have a right to know what's going in'it!

Likewise, my Tivo is my computer. It's running in my house. It's connected to my network. I have a right to know what it's doing. Would you be as forgiving of a builder who wouldn't let you keep a copy of the plans to your house? Would you tolerate a car dealer who's welded your hood shut? It's your house. It's your car. You have a right to know where the plumbing goes. You have a right to look at the engine.

I'm not saying I entirely agree with this philosophy...but I fully understand it's impetus. GPL aside, how do you balance the chef's right to keep his recipe a trade secret, against the diner's right to know what's a'goin' into his own belly? It's a thorny question. Anybody who claims the answer is "obvious" isn't giving due consideration to the counter-arguments. Both sides have a valid point.

In the case of Tivo, the chef is saying, "You won't eat without knowing my recipe? Well, I guess you'll have to go to a different restaurant!" The diner is saying, "But we had a deal...you built your recipe on mine." In this case, I have to side with the diner. We had a deal. Tivoisation is wrong.

Re:tivoisation

[PeterBrett]

But, let's be realistic. It would be absolutely impossible to Tivoise all general-purpose computers.

That is exactly what the likes of Microsoft and Sony want to do for anything consumer-available.

And it's not like you couldn't build your own.

They want to stop you from doing that too (or at least make it economically infeasible).

Re:tivoisation

[PeterBrett]

In general, it's often good for networked devices (other than general-purpose computers) to only run signed code, because it makes it significantly easier to guarantee network stability. So even if we accept that there are cases where Tivoisation is bad, and that the FSF ought to try to prevent them, we're left with the fact that it can only do so with a broad brush, eliminating a lot of good uses along with a lot of bad ones. They're taking away as much freedom for the user (freedom to use GPLv3 on their trusted platform while maintaining that trust) as they're giving.

Surely having code signed by the owner achieves the same results in this case as having code signed by the vendor, assuming that the owner is not stupid enough to keep the private key for the signing process on the box which they are requiring to run signed code?

I challenge you to provide me with one example of a case where using the user's key to sign the code running on their box is inferior to the same box running the same software signed by a vendor key which the user does not have access to.

Re:tivoisation

[PeterBrett]

Likewise, my Tivo is my computer. It's running in my house. It's connected to my network. I have a right to know what it's doing. Would you be as forgiving of a builder who wouldn't let you keep a copy of the plans to your house? Would you tolerate a car dealer who's welded your hood shut? It's your house. It's your car. You have a right to know where the plumbing goes. You have a right to look at the engine.

Actually, this is a poor analogy. In defense of Tivo, they are very good about complying with the letter of GPL v2 in that they distribute the sources to all of the GPL software that they use.

This explanation is much better, in my opinion.

Another way of looking at it is this (I don't have a Tivo, but hopefully my argument will make sense). Suppose you have a great idea for a way to improve the way the Tivo interface works for you, and you see that the modifications you need to make to the code which Tivo distribute to you are quite trivial. However, when you send the patch to Tivo they tell you they won't implement it because the business case doesn't support it -- you're the only one of their customers likely to use such a change. Then you're left hanging: as you mentioned, you own a more or less generic computing device and now you also have some code to run on said computing device, but you're at the whim of a third party (non-enforceable EULAs notwithstanding) as to whether or not you can put those two components together.

As many people have pointed out, the right to make modifications and improvements and to use said modifications and improvements is pretty fundamental to the whole point of Free Software, and that's exactly what Tivo have been trying to get around.

I agree that they need to be stopped, and I support the GPL v3 in its strategy to deal with this.

Re:tivoisation

[Kuciwalker]

Jesus Christ, I've mentioned it twice already: networked multiplayer games. Because I can't run a modified executable on my machine, I can't cheat, and neither can anyone else.

Re:tivoisation

[Timothy Brownawell]

The irritating part is that the FSF has the business products exception, where Tivoisation is okay for hardware sold for business use. Stallman et. al. recognize that in some cases it's ultimately beneficial to the user to be unable to run modified software (e.g. a business that has to have accountability, or a console gamer who wants to know that no one is running a hacked game in multiplayer), but they think they can somehow figure out where that line is for everyone.

The other irritating part is that this is discriminatory licensing, and I therefore don't see how it can properly be considered "Free".

Re:tivoisation

[PeterBrett]

Allow me to make some introductions: Peter, meet BSD. BSD, meet Peter.

Oh yes, GPL3 anti-Tivoisation clause, meet irrelevance and backlash.

I suspect that you don't understand what I'm talking about.

Quite a lot of open-source software (or at least user-written software) has been created in response to the encroachments of Digital Restrictions Management on the abilities of users to view the media which they have purchased in the false belief that such purchase entitles them to view where and when they like.

Requiring general-purpose consumer computer equipment to prevent code unsigned by the vendor from being run provides a number of benefits from the point of view of both the media cartel and large software companies:

1. Lock-in of consumers to "approved" software is easier, increasing sales and thus profits for a small number of "certified" vendors.
2. Pernicious anti-DRM tools can be stopped at source.
3. Market barriers to entry are greatly increased, reducing profit-limiting competition.

Sony, in their position as a vendor of "trendy" hardware but also as a major stakeholder in the media market, is in the ideal position to introduce such a device, and I would expect them to do so within the next two years or so. Assuming Microsoft survives the damp squib that is Vista, I would expect the device to be announced jointly by the two players as a new "secure mobile computing platform" (not marketed as a "traditional" PC).

Software vendors would be able to acquire signing keys (relatively) easily, but they would not be available for operating systems and would likely require small per-user fees in order to squash out open source software. In addition, a Sword of Damocles contract would be required, whereby distributing software later found to be non-compliant with the terms of the signing key contract would result in all software signed with that vendor's key to be remotely deactivated.

Your mention of BSD is entirely irrelevant in the context of my GP comment, so (at the risk of being accused hypocritical) I respectfully suggest that you shut up and stop being a condescending moron.

Re:tivoisation

[2short]

"I challenge you to provide me with one example of a case where using the user's key to sign the code running on their box is inferior to the same box running the same software signed by a vendor key which the user does not have access to."

Not that I actually want to jump into this debate, but I love a challenge, so:

The user is a casino, the box is a slot machine. If someone accuses them of cheating (which will happen), it is great for them if they never had any ability to modify the code in the first place.

More boringly, I've worked on software that a company uses to help decide what franchises to open and, more significantly, close. They want to actively prevent their own people from influencing this process once it's designed, so that when they decide to close Bob's franchise, Bob can't claim they did it because someone doesn't like him rather than for legitimate business reasons. They don't go as far as tivoised hardware, just hard-coded random number seeds so their results are reproduceable. In any case, there are perfectly good reasons why a user might find it preferable that they cannot modify software, or particularly, that they can show others that they can't, and thus didn't.

The FSF gets this, and hence the business exception. But a console gamer might want to be able to show he's not cheating, and might prefer this ability to the ability to modify his software.

Re:tivoisation

[rockhome]

What if said hardware was proprietary and only available through the systems provider? What then?

There should be no expectation that a manufacturer make its hardware available for third party development.
Proprietary systems have there drawbacks, but they are a fact of life. If one has an issue with the fact that a given consumer
device is a closed platform, one ought not purchase the device. This ought to be how the issue of so-called Tivoization ought
to be decided, by making the business model non-viable for the company, not some poorly constructed, arbitrarily defined licensing
scheme that restricts only the freedom of some, and not all. The fact that there is a "commercial device" exception indicates
that the case for so-called Tivoization has a valid application and should therefore be allowed. Banning the practice for
"consumer devices" is simply an arbitrary confinement of freedom that is a slippery slope.

Again, the quandry should not arise because those who even care about being able to run aribtrary code on the device would not
buy it in the first place, and would therefore not be harmed.

Re:tivoisation

[PeterBrett]

Jesus Christ, I've mentioned it twice already: networked multiplayer games. Because I can't run a modified executable on my machine, I can't cheat, and neither can anyone else.

Fair point. And although you might (probably should) agree that this is a pretty pathological example, I did only ask for one!

My response would be to ask whether a community so rife with anti-social behaviour is one which you want to be part of, but I suppose you've already answered that question too.

Re:tivoisation

[PeterBrett]

The user is a casino, the box is a slot machine. If someone accuses them of cheating (which will happen), it is great for them if they never had any ability to modify the code in the first place.

There are such a thing as ROM chips, and there are such a thing as tamper-proof stickers. This is a good example of a situation in which both would be appropriate, and compatible with both the relevant legislation and the GPL v3 (neither vendor nor user has ability to change code after delivery).

Secondly, as you mention, both of the examples you give do not fall within the category of "consumer product" as defined by the GPL v3, and therefore the Tivoisation sections don't apply. I should probably have mentioned this when posing my challenge!

As I mentioned in another posting, the gaming problem is pathological, and I repeat my response to the other poster: do you you really want to be a member of a community where the problem of cheating is such an issue that you need to surrender your freedom to cope with it?

Re:tivoisation

[PeterBrett]

There should be no expectation that a manufacturer make its hardware available for third party development. Proprietary systems have there drawbacks, but they are a fact of life. If one has an issue with the fact that a given consumer device is a closed platform, one ought not purchase the device. This ought to be how the issue of so-called Tivoization ought to be decided, by making the business model non-viable for the company, not some poorly constructed, arbitrarily defined licensing scheme that restricts only the freedom of some, and not all. The fact that there is a "commercial device" exception indicates that the case for so-called Tivoization has a valid application and should therefore be allowed. Banning the practice for "consumer devices" is simply an arbitrary confinement of freedom that is a slippery slope.

The clauses in the GPL v3 simply say: "If you use my software, for free, to create a device which you then sell back to me at a steep price, I'd better bloody well be able to put any further modifications I make to that software onto said device. Furthermore, because I think that's pretty neat, I'd like you to extend that ability to your other customers too. If you don't like it, write your own damn software."

I paraphrase and over-simplify, but the gist is there. It's not legislation, banning the practice generally. It's very specific.

You have an objection to this?

Re:tivoisation

[PeterBrett]

But closed hardware and code does not prevent cheating. The PS3 and Motorstorm are pretty closed, yet online game play is rife with people cheating with infinite boosts.

Which to my mind makes whether or not you're allowed to restrict the user to running vendor-signed game binaries a rather moot point...

Slashdot's told me to "Slow down, cowboy!" so while waiting for the timer to run out I'll write this rather pointless line at the end of my post.

Re:tivoisation

[mrchaotica]

I don't understand the furor over Tivoisation, and think it's a really bad move on the FSF's part to ban it. In all the other articles I've seen tons of comments along the lines of "well if Tivo doesn't want to give back, then they can't use my code!" The thing is, Tivo does give back - they contribute any source code they add.

Her's what you're missing: the GPL exists to benefit USERS , not developers! As it stands, the users of TiVos are being screwed because they can't actually modify the code that runs on their own devices. The GPLv3 is designed to fix that.

If this still doesn't make sense, remember that the incident that incited RMS to start all this was that some company refused to give him the source code for the driver for his printer (it had a bug that he wanted to fix). The GPL exists to fix that problem. Sure, RMS could have just gotten a different printer, but that wasn't the point.

Re:tivoisation

[aoliva]

> And yet the FSF went to great lengths to permit it in some cases (the business use exception)

I don't see that the language of GPLv3 actually permits it. Since a number of people (myself included) believe that GPLv2 didn't permit tivoization of any devices whatsoever, GPLv3 is no different as far as products that don't qualify as "user products" are concerned.

Re:tivoisation

[maxume]

And that would work just fine, just only play against people that are known to be signed by the vendor, rather than their own personal key. A little more work than a complete lock-down, but nothing onerous.

Re:tivoisation

[2short]

"do you you really want to be a member of a community where the problem of cheating is such an issue that you need to surrender your freedom to cope with it?"

No, I don't, as I'm not much of a gamer in any case. But some people apparently do want to be part of such a community. Do "consumer" users get the freedom to make that decision, like businesses do, or does the FSF make that decision for them? Business and gaming are two communities where it's nice to be able to give people more reason to trust you than they would have inherently. There are many potential communities like this.

  I emphasize I'm mostly playing devils-advocate here: I'm not sure what the right answer is, I'm just sure it's not cut-and-dried.

It's the creation of two different classes of users that strikes me as the red flag: why do businesses vs. consumers, who may well be the same people in different contexts, deserve different freedoms?

Re:tivoisation

[init100]

I can't think of a single instance where Tivo actually contributed code back to the community as you state.

I think that he meant that TiVo publishes the modifications it makes to the GPL-covered software, as required by the license.

Re:tivoisation

[AnyoneEB]

Why don't you just write a modified client that claims to be signed by the vendor? Or have your cheating program run somewhere between the client and server? Signed game executables for trusted multiplayer seems like a rather silly way to prevent cheating.

Re:tivoisation

[BRSloth]

IMHO, if you want to prevent hacking, put more checks on the side you have control: the server. You have no control over the user machine (it is HIS/HER machine) and no control over the network. There are too many pieces "in between" that could go wrong (and you may punish the user for something that may not be his fault).

It is just a little bit hard, but no impossible.

Re:tivoisation

[jythie]

The problem with this type of argument is you are having to force your technical design based off idealogical legal requirements.

'Rolling the dice' on the server for a large number of connections is more secure and correct, but can be sufficiently inefficient as to make games unplayable. When possible, technical decisions should be made on technical merits.

Re:tivoisation

[pilot1]

There's no need to prevent you from running a modified executable. You only need to be prevented from connecting to the game's server with a modified executable.

Re:tivoisation

[babbling]

I never want see a program I wrote on a device that does not allow me to make changes to my own program running on that device.

Re:tivoisation

[lucifron]

If you don't have a problem tivoization, why would you use GPL in the first place?

Pick BSD/MIT then..

Re:tivoisation

[rockhome]

First, the relevant text :

A "User Product" is either (1) a "consumer product", which means any tangible personal property which is normally used for personal, family, or household purposes, or (2) anything designed or sold for incorporation into a dwelling. In determining whether a product is a consumer product, doubtful cases shall be resolved in favor of coverage. For a particular product received by a particular user, "normally used" refers to a typical or common use of that class of product, regardless of the status of the particular user or of the way in which the particular user actually uses, or expects or is expected to use, the product. A product is a consumer product regardless of whether the product has substantial commercial, industrial or non-consumer uses, unless such uses represent the only significant mode of use of the product.

"Installation Information" for a User Product means any methods, procedures, authorization keys, or other information required to install and execute modified versions of a covered work in that User Product from a modified version of its Corresponding Source. The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or interfered with solely because modification has been made.

If you convey an object code work under this section in, or with, or specifically for use in, a User Product, and the conveying occurs as part of a transaction in which the right of possession and use of the User Product is transferred to the recipient in perpetuity or for a fixed term (regardless of how the transaction is characterized), the Corresponding Source conveyed under this section must be accompanied by the Installation Information. But this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Product (for example, the work has been installed in ROM).

The requirement to provide Installation Information does not include a requirement to continue to provide support service, warranty, or updates for a work that has been modified or installed by the recipient, or for the User Product in which it has been modified or installed. Access to a network may be denied when the modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network.

Corresponding Source conveyed, and Installation Information provided, in accord with this section must be in a format that is publicly documented (and with an implementation available to the public in source code form), and must require no special password or key for unpacking, reading or copying.

I have a problem with this for multiple reasons. First, the license makes a clear distinction between products for a home and a business. If freedom is the essence of the GPL, this clause is the antithesis of that. By making the distinction, since the specification of a user product provides a way to classify a non-user product, the license admits that practice is a valid one. If the practice has some validity, then the license should err on the side of more freedom by allowing the practice generally, or the license should not allow for the practice at all. Again, the GPL is aimed at freedom, and this freedom ought to extend to all users, as has been its application. The cost of "tivoization" needs to be considered as a by product of freedom. By not applying the freedom equally and targetting one behavior while simultaneous admitting that a similar behavior is within the bounds of the license, the FSF is no longer supporting true freedom.

My second issue regarding the anti-tivoization clause is simply that there has been no sufficient argument made that the practice actually harms anyone. With the source available, users are free to build there own versions that are not closed and will run any arbitrary code.

With no proof of real harm, and a license clause that contradicts the aims of RMS and

Re:tivoisation

[VGPowerlord]

That's fine if it were just Tivos. However, what would happen if every piece of hardware you bought was Tivoised? Only properly signed binaries would run at all. I can think of several companies that would love that situation.

No, I think Microsoft would hate to have to show all their source code, because that means we could just install it on a non-tivoized machine.

Microsoft loves Palladium, though.

Re:tivoisation

[VGPowerlord]

Her's what you're missing: the GPL exists to benefit USERS , not developers! As it stands, the users of TiVos are being screwed because they can't actually modify the code that runs on their own devices.

The two bolded sections together make an oxymoron.

Re:tivoisation

[Ragingguppy]

I don't know I think this version of the GPL is a little too heavy handed. I don't think its right to restrict people from using the software to create DRM tools. If a commercial vendor wants to stupidly restrict their customers from being able to use their product then thats their problem. I don't think its necessary to prevent that in the GPL. There are already allot of people fleeing the GPL because its considered too restrictive already. You see it when it comes to libraries etc. I mean firefox doesn't use it. And there are plenty of apps out there that don't use it. They make it too restrictive and they will scare away all commercial support for open source.

Re:tivoisation

[fferreres]

Linus comment is that that is a great goal, but goes beyond the scope of a software license. I mean, If we are unable to obtain uncloked hardware, Linux can't stop that from happening. We will only make sure that ALL efforts for software-specific hardware become closed NOT open source. We lose all the efforts and insights, and the propietary vendors win. Would you contribute to a project that instead of the effort put into enhancing the code, wanted to limit your legitimate ability from doing usefull solutions? Would you prefer TIVO used closed software? What's the benefit from a developers perspective? Because from an end user perspective, TIVO will go propietary and you'll lose that right.

And open hardware solutions will be harder or impossible to obtain, and even worst, most technological advances in the field of PRV will go closed platforms. With TIVOs allowed to lock down HARDWARE, you get to enable competitors to use TIVOs code to make open SOLUTIONS more likely to success (after all, they have all the software they need, and only need to make it work on non locked hardware).

Now the community is divided. You have better patent coverage + nasty limitations that involve which hardware can be used with GPL software, or you have a more fair and intelligent license like GPL v2 but that is not great for dealing with patent bombs.

For the record, I predict that GPL v4 will get rid of the stupid anti-tivoization elements, and keep the patent protections in place, and that Linux will migrate to GPL v4 by that time.

Federico

Re:tivoisation

[Kjella]

The part about network stability is not very important, if you ask me. You should always assume that the other part in a network connection can send corrupt/invalid/flooding data, since there are many hacks that change data in transit etc. and it hardly seems a big issue for non-signed code. As for Tivos, they're using it to enforce DRM which I don't approve of anyway, because it's restricting contents in many ways that are wrong. In the general case, I don't think you shouldn't care how something happens, for example whether you send me a static HTML file or generate it from PHP or pull it from a database, none of my business. Nor should you as reciever care how I choose to parse or render it. You gave it to me, my business.

I agree networked games are a valid point - because you make up special rules that only have meaning within the context of the game. You agree to use your human aiming skills instead of an aimbot, and even though the game needs more data to render it, I limit myself to only seeing what I'm "supposed to" see and not use a texture hack. Signed code helps enforce this agreement. I don't see how you can possibly separate between the two though, even within that same game there could be an annoying bug which it would be legitimate for you to fix. So they're left with either two choices - either don't enforce the rules of make-believe, or allow a lot of real user restrictions.

I guess we disagree on the "equally much" bit. Lastly, the FSF can't take anyone's freedom away. Using GPL code of any version is entirely optional. Tivo has no more right to use newer versions of software under the same terms than you have the right to use Windows XP because you once paid for Windows 98. If you don't like the new terms (or price tag), don't use it. Ultimately, while the FSF owns the copyright to a few central pieces of code they're not in control of the greater GPL community. If developers choose to use GPLv3 and think it serves their purpose better, Tivo has no grounds to claim it's not fair. Let's wait for the users then, to form a grassroot protest against the rights they've lost in the move to GPLv3, mmmkey? Something tells me you'll be waiting for a loooooong time.

Re:tivoisation

[SEE]

Am I supposed to actually believe converting the whole world of GNU/Linux-on-Intel servers so they can't run modified versions of software would be no big deal, while the TiVo is a major threat to freedom?

I mean, c'mon. Under this definition, a standard x86 Dell sold for "Small Business", "Medium & Large Business", or "Government, Education and Healthcare" can be locked down. Ship consumer PCs with Windows, and sell non-Windows machines only through the business/organization sales channels, and a vendor can perfectly legally lock down every single computer such that it can only run vendor-authorized versions of software, regardless of the GPL 3. Microsoft can even pay the motherboard chipset makers to do the lockdown, and consequently permanently lock GPL3 OSes out of the entire consumer channel while allowing business demand for GNU/Linux-on-Intel servers to be filled.

If opposing TiVoization had genuinely been a matter of principle to the FSF, an actual matter of defending freedom, the license wouldn't only prevent it on "tangible personal property which is normally used for personal, family, or household purposes, or (2) anything designed or sold for incorporation into a dwelling".

Re:tivoisation

[NemosomeN]

Actually, with Tivoisation, you know exactly what's in the dinner. But when you try to add salt, the chef swats it out of the air. If you want to change the recipe, do it at your own home with your own equipment.

Re:tivoisation

[level_headed_midwest]

Four letters: T-C-P-A. You could probably still build your own, but it would have to be some kind of custom FPGA job because all "usual" hardware would require all of the hardware DRM nonsense. About the only part of the PC that doesn't have DRM applicable to it today is the power supply. Even your monitor has DRM in the form of HDCP. You also would not be able to connect to the Internet because "untrusted" machines would not be allowed to connect. So unless you want to build a cobbled-together FPGA unit and run it on an underground POTS network like in the old days, 'Tivoization' is really really bad news and Stallmann et. al. are right to see how bad it could get and try to prevent it.

Now WILL the doomsday scenario happen? That's up for debate. My guess is that it won't completely as there are both a bunch of legacy devices and programs that won't work and that DRM implementations are rarely well-implemented and commonly piss off Joe User. There are many more Joe Users than RMSes out there, so Joe User is much more influential in what manufacturers will push on people. But the mechanism of action is known and the hardware will certainly allow for it, so I'd be at least a bit antsy.

Re:tivoisation

[AusIV]

Likewise, my Tivo is my computer. It's running in my house. It's connected to my network. I have a right to know what it's doing. Would you be as forgiving of a builder who wouldn't let you keep a copy of the plans to your house? Would you tolerate a car dealer who's welded your hood shut? It's your house. It's your car. You have a right to know where the plumbing goes. You have a right to look at the engine.

Really bad analogy. First, Tivo does release their source as required by the GPL version 2, what they don't allow is putting modified (unsigned) versions of the software on the TiVo. While I understand this is an unpopular viewpoint with the open source community, I think it makes sense economically. TiVo sells their units below cost, so that they will be able to sell their service along with it. If people could put whatever software they wanted on it, it would be a below-cost computer, and there would be people who would sign up for the shortest TiVo subscription so that they could get a cheap computer. Most Tivo users don't care about modifying their firmware, and if it makes their Tivo less expensive, they're not going to complain.

So following the chef analogy - the chef is telling you what's in your food. You could go home and make your own derivative of the recipe in your own kitchen, but when you're at this chef's restaurant, you get the chef's version. Likewise, Tivo is giving you the Linux kernel that they use, and on your own hardware, you can do whatever you want with it, but on Tivo hardware, you only get to use the Tivo kernel. Some may say this is a circumvention of the intent of the GPL. I think it's perfectly valid. Tivo is making contributions to Linux - everything they've done can be adopted back in to the main kernel - you just can't use their hardware with derivative works.

Re:tivoisation

[Treffster]

Hell, i don't need a modified executable to cheat! Nothing stops me changing my memory live while the game is running!

Whats that, i just bumped my local ammo/cash/lumber/HP counter back to full? that was easy!

While i agree that its a much more difficult way to cheat than just making a executable mod and building in an auto-aiming bot, the simple fact is that it doesn't matter WHAT you do to lock down a client, at the end its just a flawed approach in the first place. The only place you have any real chance of having this system work is a proprietary console running proprietary software -> and mod chips get around that anyway.

The only way to properly protect an online game is to validate inputs (are the actions they are telling the server valid according to the game rules) and behaviors (is this a human action or a computer-assisted action?).

Its orders of magnitude harder to detect aim-bots at the server, anybody with any programming experience knows that. But that doesn't mean its impossible. Its not simply a case of saying "is this person "too-good", because you can't filter that out. But you can cross-check the movement instructions with the mouse velocities and build a signed key from the pair, and on the server validate that the aiming instructions correspond with the hardware signals, rather than just coming from a software command. There are other methods in this vein, like introducing noise into the inputs sent to clients and detecting if the return signals have accurately returned the noise (if i oscillate the player 5 pixels to the left and right, does the shoot-aim return signal accurately follow the oscillations? bot!)

Couple this with some basic run-time application hashing (hash the contents of memory and the executable 2 minutes in, once the game has started, and test against THAT signature) rather than just the up-front vendor keys, and you have a much more hardened and reliable system.

You don't even have to do these tests all the time, but if they exist in your system you have ways of validating client behaviour regardless of whether you trust the client.

Re:tivoisation

[shaitand]

'recognizing that Tivoisation isn't a restriction on fundamental freedom, and in fact in many cases is beneficial to the user'

Tivoisation is never beneficial to the user. It is sometimes beneficial to the OWNER. In a business they are typically different people. The exception is to allow businesses to intentionally order the equipment that way for security purposes, not to allow vendors to prevent businesses from modifying the GPL'd code and running it on their hardware.

'In general, it's often good for networked devices (other than general-purpose computers) to only run signed code, because it makes it significantly easier to guarantee network stability.'

It is never good when someone else holds the keys to YOUR hardware. There is nothing wrong with a user controlable ability to restrict hardware to only run software signed with a key belonging to the user. As for network stability, it is my network, if I feel the vendor software is more stable than I can run that to guarantee network stability. At no point is it in my interest to block me from intentionally loading whatever software I want on my own device.

Re:tivoisation

[shaitand]

But this isn't needed for networked multiplayer games at all. The key can be verified by the server when the client connects. There is no reason to give control of your individual device over to a third party when everyone has to connect to a device that is under the control of that third party.

Re:tivoisation

[shaitand]

'Under this definition, a standard x86 Dell sold for "Small Business", "Medium & Large Business", or "Government, Education and Healthcare" can be locked down.'

If the owner requests it. The idea is to allow the guy who recieves the software to be able to modify it and use the modification. There is no violation of the spirit of the GPL if your boss buys a signed system to prevent YOU from running modified software on it.

Re:tivoisation

[shaitand]

Thousands of DD-WRT users who couldn't code to save their souls and all intentionally run modified code on their routers beg to differ.

Re:tivoisation

[mrchaotica]

They didn't back in the 80s, when RMS came up with the whole thing.

Besides, it's still relevant today: even if a particular user doesn't have the skill to make the changes he wants himself, he can still hire somebody to do it for him. The GPL is necessary to guarantee that he be allowed to do so.

Re:tivoisation

[PeterBrett]

I have a problem with this for multiple reasons. First, the license makes a clear distinction between products for a home and a business. If freedom is the essence of the GPL, this clause is the antithesis of that. By making the distinction, since the specification of a user product provides a way to classify a non-user product, the license admits that practice is a valid one.

No, it doesn't.

I believe that the practise of locking down software from user modification is a really bad thing, in general, and in the vast majority of cases where this is done it's for bad reasons. Tivo, XBox, PS3: all these are sold to me, and theoretically they are my property to do what I wish with, under consumer protection law. The locking down is a deliberate attempt to stop me.

In business, however, there are a few cases where this is desirable. For instance, e-mail is beginning to be much more important in court cases involving unfair dismissal, anti-trust, tortuous business interference, etc. If a company has an e-mail logging server which they have read-only access to, and is completely locked down by the vendor against interference from the company or its agents, it makes the e-mail records much more reliable as evidence in a court of law. Note that in this case, the restriction adds utility to user, rather than removing utility from the user. This is an important distinction.

There is absolutely no reason to stop GPL software from being used to implement such a business model, and according to the drafters of the GPL v3, the vast majority of instances which they consider safe fall outside the realm of "User Products". However, note that the definition of "User Product" used in the GPL v3 is deliberately painted with a very broad brush.

I hope that clears things up for you. As I said, I'd personally rather a blanket prohibition, but I can see why the drafters put the language in.

Finally, for your own products you can quite freely add an "Additional Permission" under the meaning of section 7 paragraph 1 which removes or weakens the requirements of distributing "Installation Instructions".

(I am not a lawyer and this is not legal advice)

Re:tivoisation

[Tony Hoyle]

You do know that DD-WRT along with all the wrt54g firmwares run on an unmodifiable proprietary broadcom version of Linux? (about 50% of it is a binary hard linked into the linux kernel so you can't even change the compile options without breaking it).

Broadcom wouldn't know what the GPL was if it came up and bit them, but it hasn't stopped people creating new software for it.

Likewise with Tivo - my Tivo is hacked to hell and back, and runs loads of third party code. I don't see the problem.

Re:tivoisation

[VGPowerlord]

I was under the impression that they had to compile it first, e.g. it's a binary, not code.

I guess it really depends on your definition of code. I interpret code (noun) as a synonym for source code and code (adj) to mean to write source code.

Re:tivoisation

[EsbenMoseHansen]

[...] fleeing the GPL because its considered too restrictive already. You see it when it comes to libraries etc. I mean firefox doesn't use it. And there are plenty of apps out there that don't use it. [...]

Firefox is licensed under GPL 2.0 or later and LGPL 2.1 or later, as well as a couple of other licenses.

They make it too restrictive and they will scare away all commercial support for open source.

They would be scared of their own shadow, then. There is a plenty of space in the OS based market as the proprietary, legally.

Re:tivoisation

[A.K.A_Magnet]

Thanks, that's the first time I read an anti-GPLv3 comment on Slashdot with a good point and a clue :).

Would I prefer TiVo to run closed software ? Yeah, definitely. Or BSD code (still Free). The thing is, different developers have different objectives when they release their code under a F/OSS license. My own goals are that I want the code to remain Free (GPLv2 does that very well), that it won't get encumbered with software patents from a 3rd party developer (GPLv3 does this), and that you can actually *use* the freedoms granted once you get the software. You can't use those freedoms with TiVo, so you're suggesting someone (ie a company) will clone a TiVo without the TCPM chip, and use the code TiVo contributed to.

Except that TiVo's changes to GPL software are virtually nothing. They don't need to release their user-space applications as Free Software (and fortunately they don't, everyone is free to choose a license to his interests/liking), so the few changes (on GNU/Linux itself) should be hardware dependent: ie the ONLY changes TiVo clones won't need (as they should have their own hardware).

My point here being: with low level/infrastructure software such as GNU and Linux, I don't care for contributions from TiVo. They use GNU/Linux precisely for two reasons: it's free as in beer, and they don't have much to contribute back. And I can't even hack the hardware (whatever the reasons are)? So yeah, in this case, I prefer they run closed-source software or a system under a non-copyleft license (most likely BSD).

What if the GPLv2 code TiVo used was a F/OSS presentation suite that they'd heavily contributed to (now I don't believe they would do so but let's imagine anyway). TiVo-clones would have something useful from TiVo, but you still couldn't change anything to the application on your own TiVo. With GPLv3, if TiVo chooses to use this code, they can't do that, so they may not use it from the start and go with something else or a solution from scratch ; here we lose if we consider that we give more value to contributions, software advancement and "market share" than to the freedom itself to be able to modify the software on your OWN product. Here we are on the good old rms problem "I want to hack drivers for my printer". Precisely the reason why Free Software was defined. So, yeah, I think it's a Good Thing(tm) from a "freedom" point of view.

Now, as gp already told, there will be tons of TiVo-like appliances (from MP3 players to iPhone clones, tablets, GPS sytems, watches!, ...); actually I think most of the appliances are going to be signed. With GPLv3, developers say "use my code as long as you allow unsigned-code to run with the same privileges". This is gonna be needed to avoid a world where all devices ask for signed code: using GNU/Linux cuts down development cost for many devices (no need to code much of low-level code, no need to get a license from MS, ...) and it will remain advantageous for many small companies trying to get into the market, and these won't be the one who won't force digitally signed code. I feel we're securing a bit of freedom for the future.

Re:tivoisation

[Alef]

"If you use my software, for free, to create a device which you then sell back to me at a steep price, I'd better bloody well be able to put any further modifications I make to that software onto said device.[...]"

To meta-paraphrase your paraphrase, that is essentially to say: "If you use my software, and put it on a device which you sell, that device must not be of kind X". A similar requirement would be: "You may not use my software in weapons", or "You may not use it in polluting industries" and so on.

While I find the cause to prevent vendors from locking down consumer hardware with cryptographic keys sound, I'm not yet convinced that the fight really belongs in a software license.

Re:tivoisation

[PeterBrett]

To meta-paraphrase your paraphrase, that is essentially to say: "If you use my software, and put it on a device which you sell, that device must not be of kind X". A similar requirement would be: "You may not use my software in weapons", or "You may not use it in polluting industries" and so on.

I see what you did there! You drastically abstracted what the FSF have written, then used the fact that other restrictions may be abstracted similarly to say that what they have written is the exactly the same as something totally different! Clever!

That isn't particularly meaningful in any way whatsoever. The restriction in no way stops a vendor from giving a user capabilities (like the ability to kill things, or the ability to watch movies, or the ability to hide several boxes behind a single IP address), it merely stops them from removing capabilities which would otherwise be implicit.

While I find the cause to prevent vendors from locking down consumer hardware with cryptographic keys sound, I'm not yet convinced that the fight really belongs in a software license.

Given that they're using my software in a device which they are locking down with cryptographic keys, and given that I don't want them to put my software in such a locked-down device, logically the correct thing to do is to license my software to them in a way that stops them from using it in such a device. This is a means appropriate to the end, is it not?

Anyway, where if anywhere do you think it does belong?

Re:tivoisation

[Alef]

Given that they're using my software in a device which they are locking down with cryptographic keys, and given that I don't want them to put my software in such a locked-down device, logically the correct thing to do is to license my software to them in a way that stops them from using it in such a device. This is a means appropriate to the end, is it not?

And you are certainly allowed to do that. However, unless your software is superb and unique, I doubt it will affect the course of events on a greater scale very much. To me, it is more of a moral standpoint than a pragmatic one, which is OK too of course.

Anyway, where if anywhere do you think [the fight] does belong?

Well, for starters, I'd like to see it in public debate and in the development of hardware standards.

Personally, I think that the impact the GPLv3 will have is primarily to attract attention to the issue.

Re:tivoisation

[shutdown -p now]

GPLv3 won't convince Tivo (or others) to open their hardware, it will just force all of them to stick with GPLv2 code. that hurts adoption of Free Software.

Well, if the software adopted in such a manner ceases being "Free" (per FSF definition), why would they want it?

Re:tivoisation

[mgiuca]

Furthermore, I believe GPLv2 effectively said that as well. It just wasn't specific enough about hardware signing so they got away with it - once again breaking the "spirit" of the GPL.

As far as I can tell, GPLv3 is exactly the same intention as v2; it has the same spirit; it just embodies that spirit in a more legally-binding way.

Re:tivoisation

[shaitand]

'I interpret code (noun) as a synonym for source code and code (adj) to mean to write source code.'

Yes, and running modified code (obviously) is running the binary that is produced by compiling said modified code. The point is that you don't need to be a programmer to run modifications to the software a vendor ships for your device.

Re:tivoisation

[VGPowerlord]

Yes, and running modified code (obviously) is running the binary that is produced by compiling said modified code. The point is that you don't need to be a programmer to run modifications to the software a vendor ships for your device.

I never said you needed to be a programmer to run modifications to the software. I was saying that "not developers" and "modify the code" (which I was interpreting as source code) together made an oxymoron. I said nothing about running modifications.

Re:tivoisation

[shaitand]

Splitting hairs aren't we? ;)

Re:tivoisation

[sumdumass]

However, what would happen if every piece of hardware you bought was Tivoised?

Their fix didn't stop this, it just stopped this with GPLv3 software. But more importantly, it sets the bar for licenses to cover concepts outside their realm. With a software license now covering hardware, it isn't unimaginable that hardware could contain a license that covers software. So imagine all computer with windows pre-installed violating the license if any operating system other then windows is installed on it ever?

You cannot say they can't do that, the GPLv3 just did, so why cannot someone else? Placing restrictions on hardware is the same concept no matter what the restrictions are. Running a stop sign is the same offense as running a red light, you get the same results unless the act causes a wreck.

The problem is that the Tivo wasn't a general purpose computer yet every one thinks it is. Not once ever has Tivo claimed their device to be anything other then an appliance computer. And the difference between an appliance and a general purpose computer is only that the appliance is designed, marketed and intended to be used in a specific way. A router is an appliance, some can run a desktop OS but it is still an appliance computer and nothing more. Same with the Tivo. And it is their discretion if someone could use it on ways other then they intended it to be used. I don't see a problem with that.

I would also expect to see hardware now with limitations on what software can be used because of the firmware's license inside it.

Re:tivoisation

[sumdumass]

I find this a funny argument all the time. Your saying that you purchased an appliance device and want to use it in a way the manufacturer never intended it to be used and you are pissed that they built in restrictions to keep it the way they intended. Circumvent the restrictions, but you knew full well that you weren't buying a general purpose computer when you made your decision. It doesn't matter who owns the device, the manufacturer shouldn't have to assist you in making it work in any way other then how they intended for it to work. It is that simple.

Re:tivoisation

[sumdumass]

The GPLv2 says is considers anything outside the software outside it's scope. The GPLv3 specifically includes the hardware in it's scope. There is no similarity in this spirit. You have all the freedoms the GPLv2 attempts to protect even with a Tivo'ed system, you just don't have them on that system unless you get around their protections. That isn't impossible so you could even have it there too.

The GPLv3 isn't in the same spirit, it is in the newly defined spirit. And it doesn't really stop tivo or anyone from locking consumers out either. It just stops one way of them doing it. I can think of at least two ways right now that would allow Tivo to lock you out and still use GPLed software including the GPLv3.

MS SLES Coupons

[Kadin2048]

Is it really true that they're "distributing" Linux by selling the SLES coupons?

The way I read this, it's a pretty high-risk maneuver. They've essentially given Novell a free pass, hoping that they can undermine Microsoft with the SLES coupons thing. But if Microsoft can argue that they're not really distributing Linux by selling the coupons, then the whole thing, I think, unravels -- Novell gets a free pass for nothing and can continue on its merry way, remaining in the death-pact with Microsoft and using GPL3 code.

Given that Microsoft isn't screaming bloody murder about this, I think they must not see it as a big risk, and that to me isn't a good thing. It means they think they can avoid having their patents undermined, and if that's true, then excepting Novell will be for nothing.

Re:MS SLES Coupons

[Anonymous Coward]

"Given that Microsoft isn't screaming bloody murder about this, I think they must not see it as a big risk, and that to me isn't a good thing. It means they think they can avoid having their patents undermined, and if that's true, then excepting Novell will be for nothing."

The reason that Microsoft isn't screaming bloody murder about this is because if they gave any impression that this would be a major impact to their IP Licensing and bottom line, share holders will start to dump stock.

I'm not saying that GPL3 is a huge hole is MS core business, just that IF IT WAS, they'd sure as hell wouldn't scream about anything if they were in a tight spot, drawing attention to themselves during a bad time.

Re:MS SLES Coupons

[Todd Knarr]

Not quite a free pass. Notice that the grandfather clause only includes the bit about distributing GPLv3'd software while a party to such an agreement. It doesn't extend to failing to pass along any patent license (which includes things like covenants not to sue) or rights to pass along such a patent license in turn. So Novell can distribute GPLv3'd software because of the grandfather clause, but they can't distribute any of it subject to the agreement with MS without violating either the agreement or the software's license (license requires that they pass that coverage to all indirect recipients, agreement prohibits doing so).

"Getting in is easy. Getting out, that really isn't hard either. Getting out alive, that's the tricky part."

Re:MS SLES Coupons

[kripkenstein]

Is it really true that they're "distributing" Linux by selling the SLES coupons?
[...]
Given that Microsoft isn't screaming bloody murder about this, I think they must not see it as a big risk, and that to me isn't a good thing. Well, to be honest, the whole Microsoft-saying-Linux-infringes-patents thing is FUD, and vice versa, the FSF-says-Microsoft-are-distributing-Linux thing is FUD as well. Both are dubious and the results of testing them in court would be very uncertain, as no precedent exists for either. Their purpose is basically to deter people from doing things (from using Linux or from suing Linux, respectively).

So, this basically seems like a 'fight fire with fire' tactic on the FSF's part. If it makes Microsoft even more wary about suing Linux vendors, then why not, I guess (although I don't think Microsoft would have sued them anyhow).

Re:MS SLES Coupons

[shaitand]

'Novell can continue to distribute GPLv3 programs and stay in the deal with Microsoft'

If Microsoft owns patents on the software then Novell doesn't have the legal right to grant the permissions the GPL requires. In that case they lose their right to distribute under the GPL. If they distribute anyway they are guilty of copyright infringment regardless of what Microsoft does or does not do.

Re:MS SLES Coupons

[jimicus]

So Novell can distribute GPLv3'd software because of the grandfather clause, but they can't distribute any of it subject to the agreement with MS without violating either the agreement or the software's license (license requires that they pass that coverage to all indirect recipients, agreement prohibits doing so).

We'll find out exactly what Novell and Microsoft think of how that part of the GPL v3 plays with their agreement when the next version of SuSE comes out and we see whether or not it's been updated with all the FSF's "GPL3 or later" versioned utilities.

Until such time, the exact text of their agreement remains under wraps (yes, I know we've been given the gist of it, but that's not what would be argued in a court case) and we can't be certain what's going to happen. For all we know, there's a line which says "if Novell are forced to choose between this agreement and distributing software, then the agreement becomes null and void".

Let the forking begin!

[shawnce]

...and ain't talking about no salad fork.

Why is this under "Linux"

[Anonymous Coward]

Linux is under the GPL v2, not v3.

Re:Why is this under "Linux"

[wbren]

Linux is under the GPL v2, not v3.

Probably because Linus is at least considering using GPLv3 for at least parts of Linux. Here's a quote from the man himself:

Btw, if Sun really _is_ going to release OpenSolaris under GPLv3, that _may_ be a good reason. I don't think the GPLv3 is as good a license as v2, but on the other hand, I'm pragmatic, and if we can avoid having two kernels with two different licenses and the friction that causes, I at least see the _reason_ for GPLv3.

I know it's a bit of a loose connection to Linux, but there has definitely been quite a lot of news lately about Linux, Linux, GPLv3, Sun, etc.

Re:Why is this under "Linux"

[eMbry00s]

Because the Linux section is not just about the kernel, but about the entire GNU/Linux system. Everything GNU is going to become GPL3, since the GNU project own all the code copyrights.

This includes things extremely vital to running a Linux system, such as the gcc program.

Re:is this the license that bans DRM?

[Spy der Mann]

Well, that's too bad for you. See, what you view as "the real world" will soon become "the corporation-dominated world". But we'll live in whatever's left of the free world.

Farewell.

Apache Licence

[smartr]

So is the GPLv3 compatible with Apache License v2 style licenses, or has that idea been abandoned?

Re:Apache Licence

[pinky0x51]

Yes, GPLv3 is compatible to the Apache License v2.

Re:Apache Licence

[fsmunoz]

Yes,it's compatible. I don't see it explicitly in the final documentation, but it was mentioned in a previous comment (GPLv3 Final Discussion Draft Rationale:

We are pleased to report that the Final Draft makes the Apache License, version 2.0, fully compatible with GPLv3. We are grateful to the Apache Software Foundation for working with us to achieve this long-sought goal. The concerns we stated in the Draft 3 Rationale were based on varying literal readings of section 9 of the Apache license that diered from the interpretation of section 9 held by the ASF itself. During the course of productive discussions with the ASF following the release of Draft 3, we ascertained that, to the ASF, the words \by reason of" in the section 9 upstream indemnication clause meant nothing broader or vaguer than \directly as a result of." Read in this light, section 9 seems to us a reasonable and fair approach to protecting upstream developers, even though we do not wish to adopt such a provision in our own license. The Final Draft makes the Apache indemnication clause compatible with GPLv3 by adding a new category of additional conditions in section 7 that may be applied, with appropriate copyright authorization, to material added to a covered work. Subsection 7f allows terms that require indemni- cation of upstream licensors and authors of the material by a downstream distributor who conveys with contractual assumptions of liability to the recipient, for any liability that such assumptions directly impose on those upstream parties.

Also, from the Why Upgrade to GPL Version 3 document:

Further advantages of GPLv3 include better internationalization, gentler termination, support for BitTorrent, and compatibility with the Apache license. (For full information, see gplv3.fsf.org.) All in all, plenty of reason to upgrade.

Re:Apache Licence

[VGPowerlord]

Further advantages of GPLv3 include better internationalization, gentler termination, support for BitTorrent, and compatibility with the Apache license. (For full information, see gplv3.fsf.org.) All in all, plenty of reason to upgrade.

I'm still trying to figure out how "better internationalization" applies to a software distribution license. Did GPL3 get released with some hot new i18n library, and if so, where does one get it?

Re:Apache Licence

[fsmunoz]

Ehehe. It's in the context of international copyright law, I'm assuming they made it in a way the is less directly attached to the US copyright system and its legal framework and more applicable and easily interpreted worldwide, e.g. "FSF board member Moglen has stressed the need for GPL 3 to become more legally cosmopolitan so the license becomes more accessible to lawyers outside of the U.S."

OpenSolaris

[starseeker]

Now the interesting part - will this license prove to be one Sun feels they can use for OpenSolaris?

If so, and the copyright holders for the parts of the Linux kernel of use to Sun are willing to license their code under GPLv3 as well, we may begin to see some major impacts on the open source OS landscape.

Fingers crossed.

Re:OpenSolaris

[gowen]

Linux relicensing won't happen. It pretty much can't happen. Any contributor to any file that does not have "or later version" (and there are many) in the license would have to sign off on the change. Some of them are dead, so you'd probably have to track down their descendants / executors.

Of course, OpenSolaris may use GPL3 for exactly that reason.

Re:OpenSolaris

[pinky0x51]

>Linux relicensing won't happen. It pretty much can't happen. Any contributor to any file that does not have "or later version" (and there are many) in the license would have to sign off on the change. Some of them are dead, so you'd probably have to track down their descendants / executors.

That's not quite correct. Sure it's would be hard to relicense Linux. Especially if you want to do it at once. But it can happen like Mozilla managed to relicense their code.

Probably the easiest way would be to ask any known (GPLv2 only) developer for permission, accept from now on only "GPLv2 or later" and wait some years. In 3-5 years almost 100% of Linux would be "GPLv2 or later" so that you could relicense relatively easy.

Re:OpenSolaris

[starseeker]

Not as a whole, true. But Solaris doesn't NEED the whole kernel - they just need some key parts they don't have. Drivers are an obvious one, and the usual list of ZFS and dtrace - I wonder what dtrace would be able to do as far as improving drivers imported from the Linux kernel. There are probably other parts, but I'm not up on the details enough to know what they would be - anybody have a list?

Of course, incorporating any of Linux into Solaris would probably mean instrumenting it and/or other non-trivial work, but I'm guessing it would still be easier than working from scratch.

Re:OpenSolaris

[joe_bruin]

Consider this possibility:
If Solaris moves to GPLv3 while Linux is stuck with v2, Debian might adopt the Solaris kernel as their main kernel (after a long transition process), since it is more in line with their principles of creating a pure GNU system. And then, what will happen to the Debian derived systems (such as Ubuntu)? If Sun plays their cards right, they can effectively shift a big chunk of the Linux world to Solaris.

Re:OpenSolaris

[notamisfit]

It may be more in line with GNU principles (although Debian has broken ranks with the FSF in the past; the GFDL comes to mind), but is it in line with Debian's goal of being "The Universal Operating System?". Debian can offer GNU/Linux on something like 13 different processor architectures; Solaris supports something like 3. They may pick it up as a sideline if anyone's interested (like GNU/Hurd, or GNU/FreeBSD, or GNU/NetBSD) but I don't see it becoming the mainline anytime soon, nor do I see Ubuntu switching over.

Re:OpenSolaris

[Bruce Perens]

Linux relicensing won't happen. It pretty much can't happen.

This is an urban myth. Linux can be relicensed at any time, with a simple legal process. It is not necessary to find all of the developers to get their permission.

How can the Linux kernel project, with its thousands of developers, ever change its license? We can't even reach them all, and some of those developers are dead and their estates don't know software licenses from driver's licenses. But changing the license is easier than most people think.

First, it's not a fundamental change: the intent of GPL 3 is that of GPL 2, the change is in the implementation. Given that, what would be required for such a change would be for Torvalds (or someone else) to publish his intent to start making releases with the new license, as a legal notice. A certain number of people would object, and they would have the right to require that their contributions be removed from the new release.

The kernel team has never been loath to replace code when necessary, and never slow to handle the job, no matter how large the item to be replaced. Just look at the replacement of Bitkeeper with "git", a big job that took a ground-up rewrite and yet was working in five weeks. So, code belonging to GPL3-objectors would be swiftly dealt with.

After some time passed, the release would happen under the new license, and life would go on. There is precedent for this, as Torvalds has already made two significant changes to the prelude to GPL2 on the kernel, publishing his intent and then making a release.

Bruce

Re:OpenSolaris

[Kjella]

First, it's not a fundamental change: the intent of GPL 3 is that of GPL 2, the change is in the implementation. Given that, what would be required for such a change would be for Torvalds (or someone else) to publish his intent to start making releases with the new license, as a legal notice. A certain number of people would object, and they would have the right to require that their contributions be removed from the new release.

Just because the license text is produced by the same organization, and some other code is using some other license which permits relicensing to it, doesn't make GPLv2-only code related to GPLv3 any more than any other license. Having the same "intent" is a concept with no legal weight. Nor can Torvalds make any kind of legally binding "objection period". When someone specificly removed "and later versions" from their license text, you can't assume that silence now is permission to relicense.

A certain number of people (many?) would not object or do anything at all, but they would have the right to sue Linus and every Linux distro for copyright infringement any time in the next decades, not only cease and desist letters but also for damages, and they'd be right. Why? Because they might want money, because they're asshats, because they got bought out by MS, because they're peeveed' that you relicensed without permission, because they don't like the license, take your pick. Also copyright doesn't expire if not enforced for a while, so if you thought this was anything like not enforcing a patent when someone makes public claims to use it, you're wrong.

In short, to me it sounds like a very dangerous plan to me, but IANAL so I wouldn't know. Then again I think neither are you, but maybe you've gotten better legal advice than my understanding...

Re:OpenSolaris

[Per Wigren]

Nexenta OS: Ubuntu userland with OpenSolaris kernel. Has been in development since 2005 and it's quite usable already (especially as a server) if your hardware is supported. The latest beta allows you to boot straight into a ZFS root filesystem.

Re:OpenSolaris

[kripkenstein]

Consider this possibility:
If Solaris moves to GPLv3 while Linux is stuck with v2, Debian might adopt the Solaris kernel as their main kernel (after a long transition process), since it is more in line with their principles of creating a pure GNU system. And then, what will happen to the Debian derived systems (such as Ubuntu)? If Sun plays their cards right, they can effectively shift a big chunk of the Linux world to Solaris. First, Debian already has multiple kernels (Linux and BSD). Second, sure, if OpenSolaris were as functional as Linux, and GPL3 to boot, it might be a nice replacement. However, at this point OpenSolaris is far behind Linux with regards to device drivers and other things. So, this won't happen in the near future. In fact, Nexenta are already basically doing what you suggested, with little popularity thus far (although it is an interesting project to watch).

Re:OpenSolaris

[petrus4]

This is an urban myth. Linux can be relicensed at any time, with a simple legal process. It is not necessary to find all of the developers to get their permission.

Yep, and like I've said before, it's more or less imminent that it will happen, regardless of the delay, because the people who want it can just keep chipping away.

Linus can have whatever perspective on the license he wants, whether good, bad, or indifferent. Although it will probably be spun to look that way, the ability to ultimately make this decision has been taken out of his hands. The reason why is because no matter how many times he or anyone else might say no to the FSF, they can ask one more time after that. It's a simple battle of wills, and by virtue of the number of people the FSF has, Linus will inevitably be the first to blink. As I said about this earlier, in the end, he will do it if for no other reason than purely to get them to shut up and leave him alone about it.

If anyone wanted the kernel to remain its' own project, they never should have got involved with the FSF in the first place. That was the first, and only, real opportunity Linus ever had to say no...and he didn't.

Re:OpenSolaris

[trifish]

Given that, what would be required for such a change would be for Torvalds (or someone else) to publish his intent to start making releases with the new license, as a legal notice.

Torvalds is not the copyright holder of the software released by other copyright holders under GPL2 without the "v2 or any later" clause (except portions of the software he wrote himself). Hence, if he ever attempts to release it under a different license (v3 or whatever) without the consent of the copyright holders, he immediately violates copyright law (remember, he's not the copyright holder and only the copyright holder can change the license terms).

Re:OpenSolaris

[EsbenMoseHansen]

Most people are far more interested in seeing parts of Solaris (ZFS and dtrace) make it into Linux.

Really? I'll take oprofile over dtrace, and for filesystems I prefer something simple that just works. And keeps working. Like ext3. I fail to see what ZFS offers me that LVM+ext3 doesn't give me, and I dislike the idea of mixing LVM and filesystem parts.

Another point is that dtrace and zfs would be hard to port, from what I hear on the kernel lists. While device drivers would be easier.

A third point: Without dtrace and zfs, linux might be sligtly less attractive for a while. Without more device drivers, Solaris can't boot on most developer's machines. So I'd say that Solaris would gain most.

Re:OpenSolaris

[sgtrock]

Please re-read this paragraph:

The kernel team has never been loath to replace code when necessary, and never slow to handle the job, no matter how large the item to be replaced. Just look at the replacement of Bitkeeper with "git", a big job that took a ground-up rewrite and yet was working in five weeks. So, code belonging to GPL3-objectors would be swiftly dealt with.

Get it now? If Linus made the kind of commitment that Bruce is talking about, the kernel team would simply write out all the code from the people who can't be found or object to release under the GPLv2.

Re:OpenSolaris

[gowen]

Given the unrecorded provenance of most of the code prior to 2.4.x, that would pretty much require a rewriting of all the major subsystems that haven't been clean-room implemented since that time (even heavily modified code would be a derived work and dependent on the original copyright holders). Sure, that might happen, but to call the resulting massively-forked work "Linux" as we understand it would be merely an exercise in trademark semantics.

Re:OpenSolaris

[Bruce Perens]

A certain number of people (many?) would not object or do anything at all, but they would have the right to sue Linus and every Linux distro for copyright infringement any time in the next decades

Folks who ignore a public notice in a combined work like the kernel and then choose to sue as their first action would not have much chance in court. I think it would be sufficient to just write them out when they are heard from.

you can't assume that silence now is permission to relicense.

But this is how a large collaborative work like the kernel is different from a stand-alone work. If the majority of copyright holders go along with a change in license and some small minority does nothing, the best assumption may be that the minority has abandoned their copyright. Indeed, in a work like the kernel individual work tends to be overwritten to the point that an older copyright can cease to exist.

And yes, I have better legal advice than you on this. Or I should say that I have legal advice.

Bruce

Re:OpenSolaris

[Bruce Perens]

I don't really think it is necessary to write out unrecorded 2.4 code contributors. Let them respond to a public notice, or not. Their contribution is in general being eroded to the point of unrecognizability as a copyrighted work, anyway.

Bruce

Re:OpenSolaris

[Bruce Perens]

Given the unrecorded provenance of most of the code prior to 2.4.x

Works of unrecorded provenance have probably already been overwritten to the point that the survival of the original copyrighted work in 2.6.22 would be difficult to show out to a court. The person with the most standing in pre-2.4 works is Linus himself, he can make a pretty good case that the work continues to be a derivative of his original creation. Folks who feel they have a surviving claim should not ignore a public notice. Folks with early works who ignore public notices and whose work has been massively overwritten probably don't have much standing to file any sort of complaint.

In other words, I don't think it's a problem.

Bruce

Re:OpenSolaris

[trifish]

You guys seem to be treating copyright law and works protected by copyright like public domain. Also, it is also a very difficult, time-consuming and complex task to decide who holds copyright to which part of a derived work when hundreds of people collaborate. That's why it's nearly impossible to decide which portions of the kernel would need to be rewritten from scratch. Without the consent of all the hundreds of kernel developers, it's practically infeasible to do it legally.

Re:OpenSolaris

[hxdmp]

Another point is that dtrace and zfs would be hard to port, from what I hear on the kernel lists.

Really? Take a look at the FreeBSD ports of ZFS and Dtrace. It only took one person to port ZFS and one person to port Dtrace. Both in a reasonable amount of time.

Really? I'll take oprofile over dtrace

Feh. Are you've really sure you've used both Oprofile and Dtrace??

Re:OpenSolaris

[gowen]

The person with the most standing in pre-2.4 works is Linus himself, he can make a pretty good case that the work continues to be a derivative of his original creation.

Undeniably. But that doesn't make it his sole copyright. Copyright belongs jointly to the creator of the original work

Folks who feel they have a surviving claim should not ignore a public notice.

Morally, you're quite right (except for the dead ones, obviously), and I agree with you. But being morally right and legally right (especially where copyright law is concerned) are basically orthogonal.

Folks with early works who ignore public notices and whose work has been massively overwritten probably don't have much standing to file any sort of complaint.

The notice of intent has absolutely no legal weight. All this would do is open the kernel up to further speculation and whispering concerning its legal status. Do we really need other OS vendors to say "Linux infringes unspecified patents and was recently relicensed in a legally questionable manner, raising the risk of copyright infringemnt suits from former developers."

What you say is lovely, but its not the law.

Re:OpenSolaris

[runderwo]

How in the world do you guarantee that such a notice would be considered public by a court? A notice included with an operating system kernel is hardly public, in the usual legal sense where the notice is published in a brick and mortar newspaper. How do you notify the estates of deceased developers, who probably don't even know what Linux is, even if they are aware they have an interest in it.

Do what now?

[Rob T Firefly]

over at Linux.com (which is owned, along with Slashdot, by Sourceforge). Does Slashdot (or Linux.com or Sourceforge for that matter) no longer welcome their OSTG overlords?

Re:Do what now?

[Archangel Michael]

Because you missed it the first time .....

http://web.sourceforge.com/news_archive/2007/1799. php

VA Software Corporation (Nasdaq: LNUX), the online media and e-commerce leader in community-driven open source innovation, today announced it has changed its name to SourceForge, Inc. The change reflects the company's strategic focus on its network of Web properties following the disposition of its enterprise software business. The company's Nasdaq ticker symbol will remain the same.

Re:Do what now?

[Rob T Firefly]

Thanks! Don't know how I missed that.. might have something to do with my complete inablility to pay any attention whatsoever to the world around me or anything in it.

Re:Do what now?

[Archangel Michael]

Hey look, pretty shiny geek toy over ---------------> there

Pretty good

[katterjohn]

But I generally place my works under the MIT license, or possibly some other (really!) free one.

But, then again, I contribute to Nmap regularly (and am now an SoC student for it this summer), and I have no problem that my contributions are placed under the GPL, because it's not a bad license...

And least not that I can see yet anyway :)

note nice!

[Spazmania]

From http://www.gnu.org/licenses/gpl-faq.html

What if the work is not much longer than the license itself?

If a single program is that short, you may as well use a simple all-permissive license for it, rather than the GNU GPL.

Hey, that's not nice! Some pretty remarkable software has been written on T-Shirts that are much shorted than the GPL.

Re:note nice!

[Brandybuck]

Then release them under a simple all-permissive license! Think about it. Don't put it under the GPLv3 if you're not planning on suing the shit out of people who "exploit" the code on the tee-shirt.

I would say the same about many larger pieces of software as well. GPLv2 and v3 are legal clubs you use to whack people over the head. They may not have rusty nails in them like a Microsoft EULA, but they are still clubs. If you have no intention of suing people over possible misuse of your code, then use an unrestricted license.

Re:note nice!

[Rakshasa Taisab]

Those small yet pretty remarkable programs could easily be rewritten under a new license. At that length, it's in the thinking that all the work lies.

Re:note nice!

[gangien]

If you have no intention of suing people over possible misuse of your code, then use an unrestricted license.

Reallly? i mean if I release some code under the GPL, I would probably not sue someone over it. But does that mean that I want people to ignore the license i gave it? no I wouldn't. I guess it's similar to saying someone is on the 'honor' system.

Re:note nice!

[Brandybuck]

The GPL is not the honor system. When you go to a dance, and there's a can by the door saying "$10, on your honor", you will note that it does NOT add "or we will sue".

If you want the honor system, then make it a true honor system. Use an unrestricted license (just to keep your warranty disclaimer attached), and add a note that says "please". No need to threaten people with attorneys.

Re:note nice!

[gangien]

you missed my point. To me it _is_ the honor system because I have no clue if they don't follow the GPL. ANd even if I knew they didn't chances are I wouldn't sue.

Re:Yea-eah!

[Anonymous Coward]

That's not chocolate.

Why is this under Linux?

[JeanBaptiste]

Correct me if I'm wrong, but doesn't the GPL cover a lot more things than just Linux?

Make it an ideology

[athloi]

"The educational aspect of GPLv3 has, in my opinion, been the greatest success," he says.

I agree. The open source movement has always wanted a focal point, a figure like Mao or Roosevelt, who can champion its ideal and point out its obvious implications. It's slightly anarchistic. It's anti-ownership. It's Darwinistic (let the best code win). All that is now clear, and many people agree.

The real question to me, as an observer of "current history," is whether people will take this to logical conclusions outside the world of the net. Will the Wal-Marts burn, and Paris Hilton be forced to do her own laundry? Will the morons of the world be forced to dwell in antartica? Or will this be a pithy statement like those of forgotten rock stars, "changing the world" inside a few minds separated by vast spaces of time, distance, economic instability and doubt?

Holiness Unto the Prophet

[WED Fan]

There is no God but FOSS and Stallman is its Prophet. Peace be upon him. The writ of holy scripture GPL shall burn into the souls of the infidel. If they do not submit, as the Prophet, holiness is his name, to the will of God, then we will force governments to ban the infidels. We will use the holy war of antitrust on the infidel. God wills it, we are the followers of God and its Prophet, paradise awaits him.

There is not God but FOSS.

Stallman is its Prophet.

The Holy Writ of God is GPL.

Re:Holiness Unto the Prophet

[fsmunoz]

Heathen!

There is no system but GNU, and Linux is one of its kernels.

Sainthood in the Church of Emacs requires living a life of purity--but in the Church of Emacs, this does not require celibacy (a sigh of relief is heard). Being holy in our church means installing a wholly free operating system--GNU/Linux is a good choice--and not putting any non-free software on your computer. Join the Church of Emacs, and you too can be a saint!

Re:Holiness Unto the Prophet

[WED Fan]

Please, allow me to correct you, you apostate sectarian:

GNU/Sainthood in the GNU/Church of Emacs requires living a GNU/life of GNU/purity--but in the GNU/Church of Emacs, this does not require GNU/celibacy (a sigh of relief is heard). Being GNU/holy in our GNU/church means installing a wholly GNU/free GNU/operating GNU/system--GNU/Linux is a good choice--and not putting any non-free software on your computer. Join the GNU/Church of Emacs, and you too can be a GNU/saint!

The GNU/Holy Writ proclaims that you are GNU/Free to accept the Prophet's, Holiness Issues From His Orifices, version of GNU/Freedom. That is, you become GNU/Free from GNU/Choice.

Jihad on apostates and infidels.

Re:Holiness Unto the Prophet

[fsmunoz]

Pfff. That GNU prefix is for the non-initiated. The ones that are in the Inner College know, from the grade of Adeptus Exemptus onward, that they don't need to use it, it's implied in everything per secula seculorum.

Re:Holiness Unto the Prophet

[WED Fan]

HERESY!!!

Re:Holiness Unto the Prophet

[psyburn]

The internet is serious business.
Check the link and read the line above the hard rule.
Get the joke already... XP

Re:Holiness Unto the Prophet

[glwtta]

Stop and say to yourself people. It is only a license. There are more important things in life.

By definition, there is only one thing for which there are no more important things in life. This may not be that thing (whatever it is), but it's still important - for many of us it has a significant impact on several facets of our professional lives.

(also, I'm pretty sure Stallman hasn't actually killed anyone)

Re:Holiness Unto the Prophet

[Eli Gottlieb]

Liar and terrorist!

Truly Richie was the greatest of all n'viyim that HaShem ever sent among men, though the noble men of Berkeley did follow him with their Software Distribution, and the Open Source Software movement did receive and do interpret that Torah.

And truly did Richie reward dominion over his Eretz Internet to his Chosen People, the Hackers.

Sh'ma Hackers! C, lashoneinu, C echad!

Re:Holiness Unto the Prophet

[Antarius]

but in the Church of Emacs, this does not require celibacy (a sigh of relief is heard)

I think that a requirement of Celibacy from Geeks, especially those that are Unix, is rather redundant...

Time for Sun to Shine

[bill_mcgonigle]

OK, now it's time for Sun to grab the bull by the horns. They've been waiting for GPL3 for a year and a half - and just recommitted to it a couple weeks ago, pending final language - if Java and OpenSolaris get released with GPL3 things are going to get *very* interesting.

Everybody please join me in exhorting Jonathan to take the bungee jump.

Re:Time for Sun to Shine

[ForumTroll]

if Java and OpenSolaris get released with GPL3 things are going to get *very* interesting.

I'm not trolling but I'm genuinely curious at to what exactly you think would get interesting? Unless Linux is able to move to the GPLv3 as well, which is doubtful, I doubt much will change. Java and Solaris are already open source and it hasn't made much of a difference. Java on the desktop is still shunned and it will probably never be adopted significantly outside of enterprise data entry applications. In terms of web applications, there are a plethora of web frameworks that are written in other languages that people seem to be adopting and migrating to. Java, to some degree, is also still tied down to the archaic processes that exist at the JCP. From what I've seen, people seem to be moving away from Java if anything.

In terms of Solaris, while it might be a great operating system, for most use cases it doesn't really offer much incentive to switch away from Linux or one of the various BSDs. The main reason people are excited about it possibly being released under the GPLv3, is because Linus has publicly stated that if that happens he might attempt to push the Linux kernel to the GPLv3. This is simply because people want to see ZFS and dtrace in Linux. From my experience, most people don't seem to care what happens on the Solaris side of the fence.

The main reason I write this is because I know a lot of Sun supporters that seem to think that if Solaris and Java are released under GPLv3 that there will be a significant adoption of the two technologies. I just don't see it happening. It didn't happen when they were released under other open source licenses and I can't see it happening if they're released under the GPLv3. What exactly does the GPLv3 offer that makes people think it will be any different?

Re:Time for Sun to Shine

[notamisfit]

The people pushing for a GPLv3 Solaris really don't care about the system one way or another. They just want to rip the kernel out to put together a completely free GPLv3 "GNU" operating system and settle an 11-year-old chicken-or-egg pissing contest. Notice the silence from the GNU crowd over software patents (an unspeakable evil when used by companies like Microsoft) being used to strong-arm Linux out of the "open" ZFS and dtrace. If the end is "free software" (i.e. free as in strong copyleft with restrictions for people we hate), any means are justified.

Re:Time for Sun to Shine

[fsmunoz]

Notice the silence from the GNU crowd over software patents (an unspeakable evil when used by companies like Microsoft)

Oh, the silence...

In Patent Absurdity, Stallman discusses the impact of software patents by using literary patents as an example. Next month, the European Parliament will vote on the issue of whether to allow patents covering software, so now is an especially important time to share information about their dangers.

The Danger of Software Patents: This speech will be accessible to all audiences and the public is encouraged to attend. Richard Stallman will explain how software patents obstruct software development.

It is difficult to find organizations that officially speak up against software patents. There are many that should do so but are under the control of large corporations.: The Free Software Foundation and its European arm are clearly against software patents

Saving Europe from Software Patents: Imagine that each time you made a software design decision, and especially whenever you used an algorithm that you read in a journal or implemented a feature that users ask for, you took a risk of being sued.

FSF Europe: No software patents in Europe

Richard M Stallman: Software Patents- Dangers to development

GPL: Finally, every program is threatened constantly by software patents. States should not allow patents to restrict development and use of software on general-purpose computers, but in those that do, we wish to avoid the special danger that patents applied to a free program could make it effectively proprietary. To prevent this, the GPL assures that patents cannot be used to render the program non-free.

Does GPLv3 have a "patent retaliation clause"?In effect, yes. Section 10 prohibits people who convey the software from filing patent suits against other licensees. If someone did so anyway, section 8 explains how they would lose their license and any patent licenses that accompanied it.

FSF annual members meeting: "As we expected, 2006 is proving to be a pivotal year for the Free Software Movement. With the release of the GPLv3, we have brought to focus the debate on the threats posed by Digital Restrictions Management, Software Patents and Treacherous Computing.

FSF releases the GNU General Public License, version 3: But even more importantly, these different groups have had an opportunity to find common ground on important issues facing the free software community today, such as patents, tivoization, and Treacherous Computing

Boycott Amazon!: The boycott can also indirectly help change patent law--by calling attention to the issue and spreading demand for change

Fighting Software Patents - Singly and Together: Software patents are the software project equivalent of land mines: each design decision carries a risk of stepping on a patent, which can destroy your project.

Software Patents and Literary Patents: On July 6, 2005, the European Parliament will vote on the vital question of whether to allow patents covering software--a policy that would restrict every computer user, and tie software developers

Re:Time for Sun to Shine

[bill_mcgonigle]

I'm not trolling but I'm genuinely curious at to what exactly you think would get interesting? Unless Linux is able to move to the GPLv3 as well, which is doubtful, I doubt much will change.

Right, I don't think Linux can ever leave GPL2 - too many authors, known and unknown, insufficient licensing. So, now if we assume that GPL3 is a better GPL than GPL2, and will lead to more "Free" (adjective sense) software, what does a community that values freedom above other concerns do? One could argue that section 3 of the DFSG is biased towards a GPL3 kernel over a GPL2 kernel. NexentaOS has already shown that Debian (Ubuntu) on OpenSolaris can be done technically.

Java and Solaris are already open source and it hasn't made much of a difference.

But not with a popular license. I'm not interested in hacking on a CDDL-licensed Java. A GPL3'ed Java, yes.

Java on the desktop is still shunned and it will probably never be adopted significantly outside of enterprise data entry applications.

Because of bigotry or technical failings? Having a developer-friendly license can help with at least the latter.

In terms of web applications, there are a plethora of web frameworks that are written in other languages that people seem to be adopting and migrating to. Java, to some degree, is also still tied down to the archaic processes that exist at the JCP. From what I've seen, people seem to be moving away from Java if anything.

Anecdotally, I've seen the opposite, but regardless it's not going away in the web space. It's also really picking up momentum in the embedded space. I'm using it for a project I'm doing, and I've seen more new J2ME apps for my Palm (Google, Opera, etc.) in the past 6 months than in the past 6 years.

In terms of Solaris, while it might be a great operating system, for most use cases it doesn't really offer much incentive to switch away from Linux or one of the various BSDs. The main reason people are excited about it possibly being released under the GPLv3, is because Linus has publicly stated that if that happens he might attempt to push the Linux kernel to the GPLv3. This is simply because people want to see ZFS and dtrace in Linux. From my experience, most people don't seem to care what happens on the Solaris side of the fence.

So, there's the issue of if GPL2 and GPL3 can be made compatible. RMS thinks they can't. I think that stands on his arbitrary line for linking, and that there's room for debate. Sun, at least, thinks drivers can move from linux to Solaris.

I agree, the OpenSolaris userland isn't all that interesting, but the kernel is pretty good, better than Linux in some cases.

The main reason I write this is because I know a lot of Sun supporters that seem to think that if Solaris and Java are released under GPLv3 that there will be a significant adoption of the two technologies.

Here's one Sun-hater who thinks that could happen.

I just don't see it happening. It didn't happen when they were released under other open source licenses and I can't see it happening if they're released under the GPLv3. What exactly does the GPLv3 offer that makes people think it will be any different?

Freedom. That may not be significant enough to motivate you, but don't discount its power in the hacker community. The GPL has its detractors, but there's real evidence that many coders are inspired by it.

Re:Time for Sun to Shine

[ForumTroll]

Thanks for the great response.

Re:Time for Sun to Shine

[eviltypeguy]

But not with a popular license. I'm not interested in hacking on a CDDL-licensed Java. A GPL3'ed Java, yes.

Newsflash! Java is GPLv2; not CDDL. Research before posting. Thanks.

Re:Time for Sun to Shine

[WebMink]

You may be interested in the commentary I have written explaining Sun's position on GPLv3. Despite having been involved in the discussion process from the beginning, we still only saw the final text at the end of this week. Whereas some people (who ought to know better) think we can just change the licenses over-night, in fact we have to take into account:

• The meaning of the final text and what its implications are. The team has started the discussion but with July 4 week next week many of them will be taking vacation so I don't expect a firm statement for several weeks.
• The fact (which I explain at length in my blog) that Sun is only the steward of the copyrights and the decision to change licenses takes more agreement than just Sun's.
• The fact that we've 6+ million lines of code in each of OpenSolaris (which is under CDDL) and OpenJDK (which is under GPLv2) and it will be heavy lifting to make a change.

So please do let Jonathan know how you feel, but be aware he's not the only guy who has to decide.

Re:Time for Sun to Shine

[bill_mcgonigle]

Thanks for the response and the link to your blog - much appreciated.

Despite having been involved in the discussion process from the beginning, we still only saw the final text at the end of this week. Whereas some people [zdnet.com] (who ought to know better) think we can just change the licenses over-night

Yeah, he seems to take a rather naive position. I do think that by piecing together the investors' mail he published and the snip from your blog one could get the wrong impression, but thinking people would agree it's absurd for him to think that Sun could have made a decision on the final version within hours, hence we make a diagnosis of premature interpretation. If the letter had instead read, "We congratulate the FSF on the release of a very good license. You'll be hearing more from us about it after we spend some time considering the final version of the license," he might have come away less confused. Arguments about whether journalists ought to possess superior reading comprehension skills would make good dinner conversation.

* The meaning of the final text and what its implications are. The team has started the discussion but with July 4 week next week many of them will be taking vacation so I don't expect a firm statement for several weeks.

And don't feel that anybody serious is asking you to rush this. I have a truckload of fireworks to send off, so I won't be thinking about GPL3 on the 4th, except in the general freedom framework.

* The fact (which I explain at length in my blog [sun.com]) that Sun is only the steward of the copyrights and the decision to change licenses takes more agreement than just Sun's.
        * The fact that we've 6+ million lines of code in each of OpenSolaris (which is under CDDL) and OpenJDK (which is under GPLv2) and it will be heavy lifting to make a change.

These two I'd expect, but I'd also have expected this work to be well under-way by now, and hopefully nearing completion. I read about Sun being interested in GPL3 almost 18 months ago, and I would have thought a review and documentation process would have begun then, if not earlier. Even if GPL3 winds up being wrong for Sun (I hope not for the sake of Sun and the sake of the community) at some point it's just good corporate governance to have a handle on who owns your code. I appreciate that you may be restricted on what you can say about this (though open-ness is a good goal) but if Sun is seen to be waiting for GPL3 final text before beginning such a process it would be seen as much less serious.

Re:Time for Sun to Shine

[WebMink]

These two I'd expect, but I'd also have expected this work to be well under-way by now, and hopefully nearing completion. I read about Sun being interested in GPL3 almost 18 months ago, and I would have thought a review and documentation process would have begun then, if not earlier.

Right - I've got a good handle on the license overall. However, there were some very significant changes in the last draft, final call and final version which may have significant consequences in our business (not least with how we are able to defend ourselves against patent aggressors) and we'll need to do some deep thinking about them.

Even if GPL3 winds up being wrong for Sun (I hope not for the sake of Sun and the sake of the community) at some point it's just good corporate governance to have a handle on who owns your code.

Oh, we know where every line of code that's been Freed came from, have no fear! The engineering teams for both OpenSolaris and OpenJDK did a phenomenal job of source code archaeology. No, my point was that changing the file headers is actually a non-trivial task that triggers a wave of regression testing. Not to be undertaken lightly.

I appreciate that you may be restricted on what you can say about this (though open-ness is a good goal) but if Sun is seen to be waiting for GPL3 final text before beginning such a process it would be seen as much less serious.

Yep, we were waiting for the final text. We engaged fully, contributed throughout but the final text was still news to us at the end of the week. And even when due diligence is done, we've still the need to get approval from the communities around the code in each case. I remain positive though, and I'll be trying hard to stay transparent (though some things just have to be kept internal of course).

Re:Time for Sun to Shine

[bill_mcgonigle]

The engineering teams for both OpenSolaris and OpenJDK did a phenomenal job of source code archaeology.

Cool, that must've been quite an effort. Make sure the guys who got the grunt work know they're appreciated!

No, my point was that changing the file headers is actually a non-trivial task that triggers a wave of regression testing.

Gotcha. Assuming you're talking about license comments, maybe the regression tests can implement a test cache such that, say an md5, of the resultant binary can be checked for pass/fail criteria? That is, making comment changes ought not produce a different binary, and if binaries A and B have identical md5sums, and binary A has passed tests, binary B gets a pass. I say this having absolutely no idea what the build/qa systems look like, but I probably could have got that past the FDA in a prior life. :)

Thanks again.

Yeah, but

[bioglaze]

does it run GNU/Linux?

Botched Link

[Odin_Tiger]

The link on 'A Gnu Dawn' points to xkdc.com; I'm going to take a not-so-wild guess here and say it was supposed to be the geeky comic xkcd.

I updated to version 3 today

[MarkWatson]

This was an easy decision for me - I agree with the new license terms.

I did not originally use the "or later version" verbage, and I decided not to this time, not that it matters: I write what could best be called "small market" OS projects :-)

A bit off topic, but it continues to frustrate me that my customers don't take more advantage of the GPL. I have been an independent consultant for a decade, and I almost never get customers to support open source development. I went so far as to offer a 30% discount for work on GPLed projects - no bites, but lots of offers to work on proprietary systems. My take is that there is too much emphasis on protecting intellectual property and not enough on reducing costs and improving quality by building on top of existing GPLed projects. From my experience, and a bit of opinion thrown in: most value in intellectual property is in unique data sources and human knowledge. I would bet that most companies would do better on financial and quality metrics by having a few proprietary systems for specific data processing, application of unique algorithms, etc. - and use GPL (or Apache, BSD, etc.) for as much infrastructure software as possible.

Re:I updated to version 3 today

[orra]

I did not originally use the "or later version" verbage, and I decided not to this time, not that it matters Ignoring that you said it didn't matter, there's new language about in the versioning section. This means you could release your code under "The GNU General Public License version 3, with Mark Watson as a proxy to decide which future versions of the license could be used instead." I'm not sure the best way to word it, but there you go: in one public announcement, people using your code could, for example, use it under the GPL 5, even if you didn't like the GPL 4.

Re:I updated to version 3 today

[Perky_Goth]

You could do it with GPLv2, and the FSF has adviced it for a while. It just isn't common practice and widely publicized.

My head hurts too

[mbstone]

Thank you GNU for writing a GPL that is sufficiently ineffable that it will create millions of billable-as-in-expensive-imported-beer hours for attorneys.

To go with your analogy...

[everphilski]

It's like a restaurant chef refusing to tell you what's in your dinner. Sure, it's the chef's recipe...but it's your dinner. You're the one about to put the mystery food into your belly. You have a right to know what's going to go into your stomach. That doesn't mean you intend to steal the chef's recipe, but daggumit...it's your stomach! You have a right to know what's going in'it!

And if you don't like the fact that the good chef won't tell you what's in your dinner, then you have every right to leave and not purchase dinner.

In my opinion, that's where tivoization should be fought. with the pocketbook, not a license.

I cannot help but feel some sympathy towards them

[thesymbolicfrog]

I mean, their condition was pitiable before, but on this day, it seems moreso than usual.
http://hoodyhoo.com/

In related news..

[jshriverWVU]

Tivo announces their new top of the line Tivo unit now running VxWorks.

No, but...

[SuperKendall]

It's the perfect thing to read while waiting in line FOR an iPhone!

Microsoft has nothing to fear

[paulpach]

They did not change the language from the last draft.

If Microsoft wants to engage other companies into Microsoft-Novel kind of deals in the future, all they have to do is give the patents to a proxy company that is not "in the business of distributing software", and the proxy company can do the deals for them. I don't think it would be hard for them to find the proxy patent troll for a reasonable fee.

The license leaves the door open not only for Microsoft proxies, it also leaves the door open to any patent troll to do the same, so long they are not "in the business of distributing software"

Re:Microsoft has nothing to fear

[init100]

I don't think it would be hard for them to find the proxy patent troll for a reasonable fee.

They already have one. It's called Intellectual Ventures, and is run by the former Microsoft manager Nathan Myhrvold.

the future of GNU/Linux distros?

[FudRucker]

what is to become of future releases of GNU/Linux distros? is GPLv2 compatible with GPLv3? will this cause problems with distributors of GNU/Linux? will it now be GNU -_- Linux forever separated by incompatible licenses unless the Linux kernel and other GPLv2 products are upgraded to GPLv3?

i surely hope RMS did not make a mess of things...

i am no uber geek capable of writing my own OS & apps, but i do manage to get Slackware installed & configured in to a pretty sweet desktop. if this turns out to be a fly in the ointment i guess there always is FreeBSD, PCBSD or DesktopBSD, i tried these BSDs mentioned they are ok & usable, but i really like Slackware and would hate to see this turn Linux in to an un-usable litigious conundrum leaving me with only the BSDs to use, i refuse to use MS-Windows as i don't trust for anything (especially connected to the intertubes)...

Re:the future of GNU/Linux distros?

[notamisfit]

Incompatible licenses really don't matter; the kernel doesn't depend on anything else, and the kernel license doesn't infect user-space programs. And shipping the two together is mere aggregation and is perfectly acceptable (the "GPL" Linux distributions you use are in fact a hodgepodge of licenses, some GPL-compatible, some not).

Re:bah, lawyerese, td;lr

[PeterBrett]

The nice thing about GPLv2 was that it was just about readable by the average human, written by people with fairly clear ideals wanting to get across a message. Now it's the culmination of special interests, reading like something designed by committee, dealing with such contemporary minutiae as to confuse anyone coming to Free software development after a few months from now, while completely turning off any of the "Just click Next until the damn software installs" crowd. GPLv3 will have nothing like the staying power of GPLv2.

Firstly, the Free Software movement has ballooned massively since 1991. It wasn't even a dot on the landscape back then. Now it's a movement whose ideals are sweeping the world, and have diversified into the arts, law, and legislation too. According to Eben Moglen at a recent talk in Edinburgh, over ten thousand people were directly involved in the process of designing the license. It was designed by a committee -- and a very large one!

Secondly, there are many parties who are desperately intent on finding chinks in the armour of the Free Software movement. As we've seen with Microsoft and their recent unsubstantiated patent blustering.

Unfortunately, the English language is an imprecise tool, and there are about 192 countries where the license has to be legal, and uphold the idea of Free Software, and have as few exploitable holes as possible. It's a bit like writing a tool that has to run on x86, x86-64, MIPS, SPARC, ARM and Power when you have to write in assembly. It's going to be long, and it's not going to be as neat as you'd like.

Re:Yea-eah!

[larry bagina]

That's not a beard.

Our turn to Embrace, Extend and Extinguish(?)

[realdodgeman]

What if any commercial company released some GPLv3 code, would any program containing this code be safe from that company's patents? That would have some really interesting effects...

Re:Our turn to Embrace, Extend and Extinguish(?)

[david_thornley]

Pretty much the same as GPLv2 code, in this respect. GPLv3 says it more clearly.

Mod parrent up!

[realdodgeman]

What the title says

Looks like they got the MS Lawyer

[tjstork]

I have to admit that I am more than a bit put off by the excessive wording and length of this license.

I think if I put out free software, I'm going to put it into the public domain, and I would encourage people to do the same. The fixation on limiting other people's profits from your work because you choose to do so is absurd, and in the days where it only costs $20 a month to give away a fairly large amount of source code to a fairly large amount of people, the whole point of linking the publication of the source with the distribution of a derived work is largely obsolete.

The GPL is starting to shape up as something that redistributes rights more than actually creates new ones. Software developers publishing under the GPL essentially have to waive their patent rights, waive their copyrights, so that consumers will benefit from software. Seriously, why do you care if someone gets rich off of what you wrote for free? Honestly, I'd almost rather have that, knowing that what I did made someone something, rather than none at all. Somebody else's profits do not block you from anything, and trying to make other people not earn money and not close software seems more jealousy than any practical social concern.

The absolute freest you can make your software is public domain, because public domain doesn't take away someone else's right to do something with it, and the GPL does.

GPLv3 Doesn't Have to be the Only Licence

[tjstork]

I guarantee that if I built a product, and wanted it closed source using GPL materials, and sold it for a ton of money, I could negotiate separate license deals with all of the copyright holders and hire some Indians to replace the code from those holders who wouldn't sign.

So, in a sense, this license doesn't really STOP someone from making money on software, as it does, discourages it.

I bet I could even buy the right from all of GNU for less than 10 billion dollars.

Re:GPLv3 Doesn't Have to be the Only Licence

[realdodgeman]

No you wouldn't. Some GNU'ers are not interested in money, but are true FOSS fighters. They wouldn't give you rights to re-licence their code for any reason (maybe except shutting down MS or something like that.)

I hereby

[ericrost]

License this post under the GPLv3. To see your rights, go to http://www.fsf.org/

"Version 2 or later" vx "Version 2 only"

[Dan Ost]

Does anyone have a good feel for how many GPL'd projects are "version 2 or later" vs "version 2 only"?

It seems to me that if the ratio is something like 90:10, then the transition to version 3 should be relatively painless. On the other hand, if the ratio is 50:50 or even 10:90, there's going to have to be an effort to migrate/replace code with version 3 friendly code.

thanks Linus

[QuantumG]

I've seen tons of comments along the lines of "well if Tivo doesn't want to give back, then they can't use my code!" The thing is, Tivo does give back - they contribute any source code they add. I love the "don't contribute back" argument. Cause it's just such a straw man, propped up by none less than Linus Torvalds. That's how Linus uses the GPL, it's not what the GPL is about. If Linus was asked to summarize what the Linux kernel license should today, I think he'd say:


You can distribute this code however you like but, if you make any changes, and you give them to others, you should make sure those changes make their way back to me.


Which, ya know, is a very permissive license which I think has gotten him a lot of code. But if you were to ask all the other people out there who use the GPL the same question, you'd get a lot of different answers. If you were to ask the FSF I think you'd get this:


Uses of software have 4 freedoms that you must ensure are not quashed when you distribute this software:

* The freedom to run the program, for any purpose (freedom 0).
* The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this.
* The freedom to redistribute copies so you can help your neighbor (freedom 2).
* The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). Access to the source code is a precondition for this.

If you can't distribute the software without quashing these freedoms, don't distribute the software.


And that's the difference. As the FSF wrote the GPL, I think we should look to them for its spirit.. not Linus.

Re:bah, lawyerese, td;lr

[VGPowerlord]

reading like something designed by committee

Yesterday, PJ from Groklaw inadvertantly addressed this point (emphasis mine):

Tomorrow at noon GPLv3 will be released. I was on one of the committees, so it's a relief it's finally done.

So, apparently it was designed by multiple committees.

You underestimate the geek appeal of spaceships

[tjstork]

If you put a spaceship in front of any true geek, they'll sell out. It doesn't matter what the cause is. You give a geek enough money to build their own spaceship, or at least launch an unmanned probe to a place of their choosing, and get pictures, scientific and commercial data back, and they will sell out. Sure, advancing some quasi socialistic ideal is one thing, but, putting mankind into the stars is quite another. You just can't be a self respecting geek and not want a spaceship, or at least a cool submarine. I mean, come on!

Re:bah, lawyerese, td;lr

[Sesostris III]

There seem to be four committees; A, B, C, and D. There is also something called International GPLv3, although I'm not sure if this is another committee, as I couldn't find membership. Details at:

http://gplv3.fsf.org/discussion-committees

PJ was on A, along with Bruce Perens and Larry Rosen.

(Whether you think that they have managed to come up with a sleek thoroughbred horse, or a four-humped camel, is a matter for speculation!)

Sesostris III

Good Point

[Evets]

I see your point. I'm not entirely excited about this particular clause myself, but strangely I didn't comment on the draft. I meant to.

I see the reason the GPL moves in this direction, but I don't agree with it myself. I think it can be dangerous.

I will say, however, that the GPL has done a great job of fostering free software development and the FSF is doing a good job of migrating the license towards the future. I believe that future versions of the GPL will change on this issue if a need arises that pushes it out to the forefront. We'll just have to wait and see.

FSF Chief and iPhone FUD

[gig]

I saw the iPhone FUD coming out of FSF chief, that was really low class. He's like we're very interested to see if there is any GPL on there, if there is there will be hell to pay. I was like, is that FSF or BSA? Two sides of the same fucking coin is all. There is no excuse for that kind of comment.

Apple has been using BSD since 1988, Apache since 1998, OS X since 1999, WebKit since 2003 ... how did porting them to the iPhone suddenly get some GPL on there?

Releasing today was also low-class. If there were 365+ nerdy events this year that would be one thing, but there is no doubt Apple claimed this week six months ago. Even if it is not petty it has the appearance of petty. It's like Bill Gates showing up at D with a really low-tech multi-touch screen, oh that is pathetic. If the hottest thing at MS really was their enormous multi-touch screen, that would be a reason NOT to bring it to D, when people are asking Steve Jobs to show them his personal iPhone with all the same features but in a ready-to-ship consumer product that goes in your pocket and runs all day on batteries.

Finally, there is something disturbing about knocking the iPhone for not being open source enough. It is clearly the market leader in phones when it comes to both open source and interoperability standards. At a time when there are no real Web browsers on phones, Apple puts in an open source, standards-compliant Web 2.0 browser and says "make Ajax if you want to be on the user's phone." The iPhone is compatible with the entire Internet instead of a menace to it because of BSD. It is a remarkable open source success story. To the extent that it has anything to do with FSF they should be proud of their contribution and to the extent that it doesn't, they should recognize that also.

Re:FSF Chief and iPhone FUD

[PeterBrett]

Apple has been using BSD since 1988, Apache since 1998, OS X since 1999, WebKit since 2003 ... how did porting them to the iPhone suddenly get some GPL on there?

Webkit is derived from KHTML, which is GPL software. Webkit is, therefore, also GPL (note that it may be the default HTML rendering engine for KDE 4).

Next!

Re:Yea-eah!

[Tony Hoyle]

That's not a beard... it's a space station!

What the %&@* does it mean?

[james_gnz]

From the linked commentary:

"In a sense it's only from today that we have enough text. So this is where the work begins. There's a lot of work to be done beyond the actual words themselves to get people to understand the license."

Not to worry, I've already got that sorted: http://gplv3.fsf.org/comments/rt/readsay.html?file name=gplv3-draft-4&id=3445

PLEASE MOD UP

[fritsd]

(hey, I have to... because it's commandment nr. 8 on the draft) ;-)

Re:Slow news day

[qualidafial]

Attention moderators: parent post was a joke. Loosen up

Re:Yup! My bad.

[Odin_Tiger]

Ah, didn't see. Mods to you, if anybody has 'em. :-/

I see no changes

[theologu]

To all the people who want to hack TiVo or submit patches... as Linus said, there are other companies that produce tivo-like equipment, and allow modifications, support them buying their products, and let TiVo with the merchendise in the stores. If you buy a car...let`s say a Mercedes, you think they will cover the warranty if you modify the car? You bought it because you liked it as it is, if you want to chip tune it, you are on your own! Many of us turns to be programmers and hackers of the tivo box, but if it was to allow mdifications I guess nobody cared. They respected the GPLv2, in the manual they even printed the licence. You buy a TiVo box because you like it as it is, otherwise you buy something else! You want acces to the hardware? Maybe make a clone? There must be a business part of all this, and from this perspective TiVo didn`t do anything wrong! If we want to change something, maybe it is time to support those companies that sell devices that we like, hack, improve, and maybe, TiVo will be forced by business to comply to market.

Re:I see no changes

[spitzak]

If you reprogram your Tivo, you *will* violate the warranty. That is not what the GPL3 is trying to do. Trying to confuse this with warranties is just more FUD.