GPL 3 Forking Risks Discussed

sebFlyte writes ""I fear a lot of unpleasant forking action when the GPLv3 comes out." The words of Debian maintainer Matthew Palmer. ZDNet has an interesting look at the possibility of forking when GPLv3 emerges, with lots of reassurance from Eben Moglen (the FSF's chief lawyer)."

Comments

Whew

[Captain Nick]

At least they aren't GPL spooning.

From the GPL v2 text:

[azmaveth]

TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION

9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns.

Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation.

Re:From the GPL v2 text:

[Daniel Vallstrom]

The problem is presumably that many choose not to add the "or any later version" option to their license.

The reason why people don't add the option is probably because if a later version of the license is weak or bad, they don't want to give users carte blanch to use the new bad license version.

However, since the new license version is, by the look of it, better and stronger, most should have no problem licensing their programs under the new GPL version.

On the other hand, people might quit licensing under the old versions since the new one is better.

Re:From the GPL v2 text:

[Anonymous Coward]

Ummmm... you are aware licenses can be revoked!

If GPLv2 is revoked in its entirety by GPLv3, then that clause is meaningless

Re:From the GPL v2 text:

[evilviper]

If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation.

GPL v5:
Sub-Section 12:
Agreeing to a License You Haven't Seen Yet

Any company may redistribute this software without releasing the source code of their modifications, so long as they have deposited $10 million USD in Swiss bank account: 111-222-333.

Is it just me...

[Xeth]

Or is this not the best time for the Open Source community to divide itself (admittedly, there may never be a *good* time for such an action...)? Is the GPL much of a problem in its current incarnation? Like they say, if it ain't broke...

Re:Is it just me...

[Aadain2001]

After reading the article, it seems like those who think there will be a "division" don't really understand the whole of the Linux world. There are MANY licenses in use by the FOSS world right now, and adding a new version (which addresses international copyright laws and patent issues) will not cause Linux to split into multiple camps. If GPLv3 turns out to be bad, then no one will use it and GPLv2 will remain the most used license. In the FOSS world, no one forces you to do anything like use a specific license when you want to use anyone instead (the developer that is, not the end user).

Re:Is it just me...

[natrius]

If GPLv3 turns out to be bad, then no one will use it and GPLv2 will remain the most used license.

This isn't the issue. Even if GPL v3 is good and lots of people want to move to it, projects that are licensed exclusively under GPL v2 (e.g. the Linux kernel) are going to have trouble moving forward to the new license. There are a lot of copyright holders to contact and agree to the change in order for it to happen. All that people are saying is that lots of forking could happen. Assuming everyone agrees to the changes presented in v3, it won't that much of a problem, but there's still work to do to change over to the new license.

There are MANY licenses in use by the FOSS world right now, and adding a new version (which addresses international copyright laws and patent issues) will not cause Linux to split into multiple camps.

If you haven't noticed, each license essentially creates its own community around it. You can't take code from the Linux kernel and put into into the BSDs. The problem with a new GPL version is that these incompatibilities will occur within an already established community, and break it apart unless everyone changes.

Re:Is it just me...

[Tony Hoyle]

I'm actually really worried by the GPLv3 because it stands a good chance of defeating the whole purpose of copyleft.

eg. Someone takes my GPLv2 code with the 'or later' clause in it, and makes a GPLv3 derivative. GPLv2 is very likely to be incompatible with GPLv3, so I can't then take that code and put it back in my GPLv2 work... I've lost control of that particular code unless I move wholesale to GPLv3 (which I'm not inclined to do based on the information available so far.. as a protection I've started removing the 'or later' clause already).

This is in fact slightly worse than the BSD scenario because it allows someone to fork a new opensource project with your code and you can't do anything but stand and watch.

Re:Is it just me...

[Aadain2001]

I think you are missunderstanding what GPLv2 allows end users to do. Could you take a current GPLv2 work that you download, make a small modification, and release it under the BSD license? No, you can't, because you don't have the right to change the license which the work is released under since it is based on a GPLv2 work. Same will go for the GPLv3 issue. If someone takes your code and makes a derivative of it they have to release it as a GPLv2. You can then put the code back into your project. The basic rule of thumb is that only the original developer, or copyright holder, can change the license a project and all derivatives are distributed under.

As for the issue about the kernel, I doubt very much that they will make GPLv3 completely different from GPLv2. If anything, they will just add clarification to international law and some patent issues like they have mentioned. And all this is YEARS away from happening, so there is a lot of time to make sure that GPLv3 is pretty much GPLv2 but with a few extra lines that doesn't make it incompatible with the previous version. Remember, this isn't some corporation making descision based on what's best for their bottom line. This is a a community driven and controled decision, so let them know what you want and what you don't want or stop whining.

I feel...

[Anonymous Coward]

there is a grave disturbance in The Force.

I'm always looking...

[Anonymous Coward]

for some forking action, but I never seem to get any.

Oh yeah, this is slashdot.

Re:I'm always looking...

[AvatarofVirgo]

Looking for forking action on slashdot... Reminds me of the beach scene on Eurotrip.

You only end up with naked over weight drunks chasing after teenage /. poongtang.;)(I think I spelled that right)

And if your wondering if I'm talking about the beach or slashdot than I'm talking about both. Well, except for the /. part.

Re:I'm always looking...

[Anonymous Coward]

You better not fork on the table you son of a bitch.

Re:I'm always looking...

[mysticwhiskey]

Practise safe forking - type with rubber gloves.

Pleasant forking action?

[Cruithne]

Has anyone ever encountered any pleasant forking action? This google image search is leading nowhere... perhaps safesearch needs to be off?

And from the Linux Kernel "COPYING" file

[Fruny]

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

Re:And from the Linux Kernel "COPYING" file

[Anonymous Coward]

That means there will be no GPL v3 as far as Linux is concerned. Linus et al won't get every contributor to the table to agree to a license change. Without that happening, the kernel will have to remain under GPL v2 only. The relicensing of the Mozilla codebase (to the MPL/LGPL/GPL trilicense) has been a tedious process for the Mozilla project, and that is a considerably younger and smaller codebase.

Re:And from the Linux Kernel "COPYING" file

[Lussarn]

If there is a will there is a way, we just have to wait and see.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

If there is a will there is a way,

I believe the correct quote is ``where there's a will, there are a hundred schemeing relatives''

Re:And from the Linux Kernel "COPYING" file

[azmaveth]

Exactly. It is highly unlikely that the Linux kernel will ever fully be licensed under the GPL v3. At least, not until all the old code under v2 is replaced by newer code.

The majority of the article, however, is talking about the forking of individual software projects. Some developers might prefer the new license and submit code that is only GPL v3. Some might prefer the old license and continue to use GPL v2. This is where the forking could occur.

The funny thing is that the reason to switch exclusively to the newer license would be to limit freedom. You can continue to license your code under "version 2 or later" of the GPL and that way people can choose whether or not they like the new changes. By using "version 3 or later" instead, you would be forcing people to accept the licensing changes.

I don't want to start a flame war, but this kinda makes the BSD-style license a bit more attractive, eh? ;)

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

Also note that the only valid version of the GPL as far as the kernel is concerned is _this_ particular version of the license

Does anyone know why Linus made this, at first glance, boneheaded decision?

What rights did he think the FSF were likely to give away in later versions of the GPL that he had to hold on to? It's hardly likely that RMS would have a deathbed convertion and make v20 of the GPL a BSD style licence.

Re:And from the Linux Kernel "COPYING" file

[Erik Hensema]

It's hardly likely that RMS would have a deathbed convertion and make v20 of the GPL a BSD style licence.

It is unlikely yet possible. To exclude any uncertainty Linus decided to 'lock' the Linux kernel to the GPL version 2 and version 2 alone.

The terms of the license work now, why won't they still work when GPLv3 is out? It's not as if all kernel developers will suddenly convert to it making their patches incompatible with the kernel.

Re:And from the Linux Kernel "COPYING" file

[Daniel Vallstrom]

Does anyone know why Linus made this, at first glance, boneheaded decision?

Probably because he didn't want to give a carte blanch *in case* a later version would be weak or bad in some way. (At least that was the reason I had once to not include the "or later version" option.)

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

It is unlikely yet possible [that a later version of the GPL would be substantially more free].

I think the possibility of the FSF giving up politics is far, far smaller than the possibility of some problem being found with a specific version of the GPL.

Re:And from the Linux Kernel "COPYING" file

[m50d]

He didn't want any risk at all. He thought v2 was good enough and wanted to make it completely impossible that linux could ever become propriety. That's also why he doesn't have contributors assign copyright - makes changing the license much harder.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

He thought v2 was good enough and wanted to make it completely impossible that linux could ever become propriety.

No, the `or later' clause can only loosen the conditions of use, so that can't be the reason.

The problem with tying it to a particular version comes if a bug is found in the licence.

Imagine if the US courts decided that some clause in the GPL v2 meant that commercial use was not allowed -- and they can if they want, common law system -- the linux kernel would be hobbyist only in the US, whereas emacs, having the or-later clause, could be `patched' to get commercial use back.

Re:And from the Linux Kernel "COPYING" file

[hankaholic]

But again if a hole were found, v2 code would have that hole, so fixing it with a v3 could only reduce the amount of leeway had by other parties.

Re:And from the Linux Kernel "COPYING" file

[Ingolfke]

The funny thing is that the reason to switch exclusively to the newer license would be to limit freedom.

All licenses that place any restrictions on the licensee restrict the freedom of the licensor. Even the BSD license requires that I not say the code was written by me. The intention of the GPL is to open up the code to any type of freedom imagineable. It is specifically designed to "Free software is a matter of the users' freedom to run, copy, distribute, study, change and improve the software." Patent law could limit the freedoms of the users of GPL code by requiring NDAs or licenses for the use of the technology. International copyright law could open the code up to any type of changes (including closing the source). Without restrictions or changes to the GPL, the purpose of the license could be undermined.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

But again if a hole were found, v2 code would have that hole, so fixing it with a v3 could only reduce the amount of leeway had by other parties.

How? You have certain options under v2. V3 may give you more or fewer options, but you still have the options you had under v2.

If v2 turned out to have a missplaced comma, meaning you could only use code from Emacs if you have red hair, then you never had the rights we all thought you had, so still v3 will give you mode rights.

There is no scenario where giving you the option of using the code under another licence, if you so desire, can reduce your options.

The only thing I can think of which could have motivated Linus to delete the or-later is a worry that the FSF might decide to give you more rights to use the linux kernel at some point in the future than he wants -- eg the FSF becomeing BSD converts (right after the Pope converts to Islam) -- but the sensible fix for that is to say `any version of the GPL >=2 approved by Linus Tovalds' or to create a new sequence of licences exactly the same as the GPLs except in name, and use those.

Perhaps the fact that he comes from a country with a civil law system made Linus underestimate the value of the or-later clause.

Re:And from the Linux Kernel "COPYING" file

[Erik Hensema]

If a problem were to be found in the GPLv2 then you're screwed anyway. It's at the users discretion what version to use: the current flawed version, or the new fixed version.

So it's never going to get any better. However, if the GPLv3 contains some legal loophole and the GPLv2 does not, then all GPLv2 programs suddenly contain the loophole.

In short, you have to trust the FSF to deliver perfect licenses from now until the distant future (when the copyrights on GPLv2 software start to expire).

So yes, IMHO the 'any later version' clause in GPLv2 is evil.

Re:And from the Linux Kernel "COPYING" file

[evilviper]

I think the possibility of the FSF giving up politics is far, far smaller than the possibility of some problem being found with a specific version of the GPL.

But by putting the "or later" clause in, you are opening yourself up to BOTH possibilities, for no good reason.

Re:And from the Linux Kernel "COPYING" file

[evilviper]

to? It's hardly likely that RMS would have a deathbed convertion and make v20 of the GPL a BSD style licence.

Not likely, but possible. Would you put a multi-billion dollar business in a position that it has to depend on one single person NEVER making a poor choice in their entire life?

Perhaps RMS/FSF will have a change of heart. Perhaps they will just put some unfortunately poor wording in the next version of the license.

Perhaps you shouldn't have given them a blank-check, and opened yourself up to an unlimited number of problems if the "unlikely" should happen.

Re:And from the Linux Kernel "COPYING" file

[m50d]

Sorry, if that wasn't clear what I meant was "make it completely impossible that there could ever be a propriety fork of linux". And yes, there's a problem if the license has a bug, but presumably linux weighed that against the risk of the FSF doing something stupid and decided that was more of a problem.

Re:And from the Linux Kernel "COPYING" file

[Anonymous Coward]

It's hardly likely that RMS would have a deathbed convertion and make v20 of the GPL a BSD style licence.

Depends... Mr. Balber, there are certain offers I can not refuse. Ny account number is 025 03296 3304, AIB Bank.

RMS.

Re:And from the Linux Kernel "COPYING" file

[Anonymous Coward]

That statement from the Linux kernel "COPYING" file seems to contradict the choice given in the GPL v2 that allows a user to select any version of the GPL as the license under which the use the code.

Does the statement in the kernel "COPYING" file override that choice in the GPL? If not, it should be removed to avoid confusion. If it does override the GPL, then can the kernel be considered to be licensed under the GPL at all? Can the GPL be modified or overridden and still be considered the GPL?

Re:And from the Linux Kernel "COPYING" file

[Taladar]

And perhaps the fact that the US is a country with a defective law system makes you overestimate the value of this clause. You can't make licenses immune to law-makers intentionally passing laws to break your license.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

But by putting the "or later" clause in, you are opening yourself up to BOTH possibilities, for no good reason.

No, you open the possibility of RMS suddenly BSDing your code as the price of allowing retrospective fixes should the GPL not turn out to allow what we all presume it allows.

Remember that the people drawing up the GPL has the normal problem that no one knows what the legal meaning of a contract really is until it hits the courts. There is no fundamental problem if the courts later found the GPL to be more free than they intended (they'd be itrritated, but at least the code would still be out there and usable by all the people who had assumed it was), but if the courts found that the GPL was more restrictive than expected, then lots of people could be in lots of trouble and the FSF would be on everyone's shitlist. The or-later is a reassurance that if that happens then the problem can be fixed.

Re:And from the Linux Kernel "COPYING" file

[Anonymous Coward]

The GPL explicitely includes later licenses only if the program does not reference a specific version of the license. Consequently there is no contradiction.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

Perhaps you shouldn't have given them a blank-check

But it's nota blank cheque, the only thing which can be done is to make the code more free. Given that RMS and the FSF are obsessive about not having the code be more free, that is hardly a great risk.

By not giving that possibility you are opening yourself to every politician and judge making the code significantly less free, and they certainly don't have the same dedication to openess that RMS has to enforced cooperation.

Re:And from the Linux Kernel "COPYING" file

[AssHatAnonymous]

...the only thing which can be done is to make the code more free.

No, the "or any later version" clause basically says, you can use this license or any license that the FSF chooses to release and call GPL v{n+1}. That v{n+1} license could be:

1. You can use this code for anything you like and do anything you like with it, without credit or compensation to any of the original copyright holders.

And then I could come along, being a complete and total asshole who wants nothing more than to fuck you over and shit all over your belief system, and choose to use v{n+1} of the license, take your code, slap it into a nice plastic box and stick a label on it: AHA's Goat Fuckingly Great Program, and sell it and make millions of dollars.

It is a blank check.

Re:And from the Linux Kernel "COPYING" file

[plague3106]

The majority of the article, however, is talking about the forking of individual software projects. Some developers might prefer the new license and submit code that is only GPL v3. Some might prefer the old license and continue to use GPL v2. This is where the forking could occur.

Likely though the project team will decide as a group which license to use. True one person might be so for or against one they attempt to fork..but it seems kind of unlikely, especially since the FSF is trying to gather so much feedback. Ideally, most developers would perfer the new license anyway. The minority that won't will be just that; a minority, and its unlikely that Samba GPL2 would be able to keep up with the much larger Samba GPL3...and since the GPL3 version is more active, users of samba will likely go with that...which in the end will likely end the V2 fork.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

That v{n+1} license could be: You can use this code for anything you like and do anything you like with it[...]

Exactly, a more free licence, hence making the code more free, which is what I said was the only thing which could be done.

Have you considered reading what you reply to?

Re:And from the Linux Kernel "COPYING" file

[bluGill]

Many rich people have left behind a foundation of money. Those who end up running the foundation often donate to political causes opposite of the ones the origional guy would have supported.

Sure today RMS is in charge and will keep the GPL pure. RMS will not live forever, in fact he could be dead now having been hit by a bus on the way to work, and in a few moment the story will make /. Unlikely, but it could happen. Once he is gone who is to say corporate interests won't take over and made a version 4 license that allows unlimited modification and distribution without making any source public. Linux is protected, while everything v2 or latter is not!

Of coruse Linux is taking the risk that something illegal will be found in the v2 license. That seems unlikely, but laws change all the time.

Re:And from the Linux Kernel "COPYING" file

[DustMagnet]

the normal problem that no one knows what the legal meaning of a contract

Another big problem is that no one seems to know that a license is not a contract.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

Once he is gone who is to say corporate interests won't take over and made a version 4 license that allows unlimited modification and distribution without making any source public.

A sane response to the above paranoia would be to add `or any later version approved by me'.

Re:And from the Linux Kernel "COPYING" file

[shaitand]

"Linus et al won't get every contributor to the table to agree to a license change."

He does not have to. The GPL technically does not really require forks to license under the GPL, just to provide compatible terms. That means Linus can download the code from kernel.org under the GPL and then relicense under the compatible GPL v.3 plus mods.

Re:And from the Linux Kernel "COPYING" file

[MrResistor]

I dispute that the BSDL is more free than the GPL. It is only more free at the point of origin. There is no incentive to keep derivatives open, and thus the code quickly becomes less free.

You seem to be taking a lot of things as absolutes that simply aren't. You should reread all these posts you've responded to, and consider them, as every one of them make a valid point.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

I dispute that the BSDL is more free than the GPL.

It gives the licencee more options, hence it is a more free licence.

and thus the code quickly becomes less free.

No, the code is always there licenced under the original terms, and so as free as it ever was. The worst that can happen is that someone else decides not to open up their work.

The point of the GPL is to sacrifice a little freedom in order to encourage the creation of further open source software.

The relevant point is that later licences can not restrict anyone's options, only open up more options. This means that all that is being `risked' is that encouragement to the creation of further open source derivatives, and that only if the FSF fundamentally changes it's nature.

On the other hand, congrtess could pass a bill, or a judge make a decision, tomorrow which changed the impact of the GPL, say making the linux kernel unusable in commercial shops. Wouldn't that give Bill a stiffie. Who do you think Bill would be more likely to manage to buy off, the FSF or some congresscritters?

The existance of a mechanism to patch a system is a security issue. The lack of any ability to do so is a limit to the useful lifetime of the system in a changing world.

Re:And from the Linux Kernel "COPYING" file

[Deven]

The GPL explicitely includes later licenses only if the program does not reference a specific version of the license.

I'm not convinced this would stand up in court. How can the language of one version of the GPL authorize the assertion that a different version of the GPL applies to the code? Unless the copyright holder explicitly offered the code under any version of the GPL, it seems necessary that one version of the GPL would have to be considered the "right" one, based on the author's intent. This would probably mean the latest version at the time of the license grant.

IANAL, but I'm not convinced that language in GPLv2 can really authorize the use of GPLv1 in its stead, as that would mean that the terms of GPLv2 wouldn't apply after all. (But someone could try to argue that GPLv1 was the intent all along.) It becomes a chicken-and-egg problem -- either version A controls or version B controls, but to pick and choose clauses from multiple versions seems wholly inappropriate.

It seems even more tenuous to read the license grant as authorizing GPLv3 or GPLv4 when GPLv2 was the latest version at the time of the grant. How can you license something under terms which don't even exist yet? And again, I don't know that you can leverage terms in GPLv2 to claim that GPLv3 retroactively applies instead.

It will be interesting to see what happens if this is ever tested in court. If GPLv2 is the most recent version at the time of the license grant, it seems probable that an unspecified GPL version would be interpreted as GPLv2 only. In the absence of any clear indication of intent from the author, I doubt the GPL's "any version" language would be held to be controlling...

Re:And from the Linux Kernel "COPYING" file

[AssHatAnonymous]

Exactly, a more free licence, hence making the code more free, which is what I said was the only thing which could be done.

The license could be anything. They could add invariant sections. They could add a clause that says all derivatives or anything the incorporates the software must have "GNU/" prepended. It's not a merge, it's a wholesale replacement.

Have you considered reading what you reply to?

Have you ever considered that you are just plain wrong?

Re:And from the Linux Kernel "COPYING" file

[Donny Smith]

> Even the BSD license requires that I not say the code was written by me.

Are you talking about this?
-------
Neither the name of the nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
-------
Source; http://www.opensource.org/licenses/bsd-license.php

If yes, then what you say is untrue - you can talk about it but not use your name or organization to **endorse** the products **without permission**.
That is to say, you can't hawk a BSD appliance shouting (for example) "Come on folks, only $999, here's a router appliance with AT&T routing programs!" unless you've got a permission.

>The intention of the GPL is to open up the code to any type of freedom imagineable.

Riiight.
I want to sell a mail server appliance based on a modified Postfix MTA and keep those Postfix modifications to myself.
Pray tell, how can I excercise such freedom?

The only truly free license would not have any duties, obligations or restrictions whatsoever.
GPL and suchlike licenses are full of rules and restrictions, for Christ's sake.

Re:And from the Linux Kernel "COPYING" file

[Donny Smith]

>There is no incentive to keep derivatives open, and thus the code quickly becomes less free.

Who cares if the code is free?
Freedom of people is the first thing to worry about, and BSDL gives more freedom.

Re:And from the Linux Kernel "COPYING" file

[CableModemSniper]

class License : public Contract {
...
}

class License extends Contract {
...
}

License subclass: #Contract ...

class License(Contract):
...

class License < Contract
...
end

package License;
push @ISA, 'Contract';
...

Re:And from the Linux Kernel "COPYING" file

[Ingolfke]

The intention of the GPL is to open up the code to any type of freedom imagineable.

Ahhh... that was a serious typo on my part. You're correct it should have been "is not to open up".

You're correct about the BSD license. My error.

Re:And from the Linux Kernel "COPYING" file

[HiThere]

I wouldn't call that paranoia. I'd call it "taking the long view". It's an experimentally determined fact that centralized positions of power tend to become occupied by people who will abuse that power to their own advantage.

Re:And from the Linux Kernel "COPYING" file

[kz45]

Once he is gone who is to say corporate interests won't take over and made a version 4 license that allows unlimited modification and distribution without making any source public.

you mean make the license free?

if this happened, someone would just create a new license, which duplicated the original GNU exactly, and programmers would start releasing software under the new license rather than the new "corporate" license.

Re:And from the Linux Kernel "COPYING" file

[DustMagnet]

You can write as much silly code as you want, but GPL is not a contract.

Re:And from the Linux Kernel "COPYING" file

[CableModemSniper]

From http://www.usdoj.gov/usao/eousa/kidspage/glossary. html
contract: An agreement between two or more persons that creates an obligation to do or not to do a particular thing.

Re:And from the Linux Kernel "COPYING" file

[DustMagnet]

contract: An agreement between two or more persons that creates an obligation to do or not to do a particular thing.

Right.

This isn't an agreement between two or more persons. The author of GPL code doesn't even know if I'm using the license. I never sign anything. If I violate the terms of a GPL license it isn't breach of contract. It's a copyright violation.

Re:And from the Linux Kernel "COPYING" file

[DustMagnet]

Please, before you reply again, follow some of these links.

Re:And from the Linux Kernel "COPYING" file

[wkitchen]

A sane response to the above paranoia would be to add `or any later version approved by me'.

For a project that has many contributors, some likely untraceable, who is "me"? And if your project does manage to resolve that issue, what happens when this "me" dies, or becomes antagonistic towards the project?

It seems to me that all you would have accomplished is to move the dependence from the FSF's future good behavior to someone else. With lots of projects, that's lots of someone else's. I suppose that could be good in the sense of not having the whole of GPL'd OSS susceptible to a single point of failure. But it increase the risks and complexity of developing with shared code. Risks and complexities that potentially could extend to users as well as developers. In which case, you wouldn't have to wait for the FSF to screw things up.

Re:And from the Linux Kernel "COPYING" file

[evilviper]

By not giving that possibility you are opening yourself to every politician and judge making the code significantly less free

The idea that the courts could or would change the GPL v2 to make it more restrictive is quite ridiculous. Courts don't change contracts to undermine the stated intention of the contract.

However, the GPL being selectively freer could be more insidious than you think. I suspect most people wouldn't be too angry in the case of GPL'd software going to a BSD-style license, but they'd be furious if RMS/FSF decided to give themselves the legal right to sell an unrestricted license for your software to the company willing to pay the most for it.

Re:And from the Linux Kernel "COPYING" file

[Minna Kirai]

The "ultimate freedom" you mention is a paradoxical impossibility.

I want to sell a mail server appliance based on a modified Postfix MTA and keep those Postfix modifications to myself.
Pray tell, how can I excercise such freedom?

I want to kill teenage girls and sew dresses from their skin.
Pray tell, how can I exercise such freedom?

The only truly free license would not have any duties, obligations or restrictions whatsoever.

The only truly free country would not have any obligations or restrictions whatsoever, including laws against murder or kidnapping.

But wait! If kidnapping is allowed, then it's not really free... error.

Re:And from the Linux Kernel "COPYING" file

[Minna Kirai]

Those who end up running the foundation often donate to political causes opposite of the ones the origional guy would have supported.

Often enough, that's what the rich guy wants. Many of the USA's billionaires earned their fortunes through borderline evil methods, but on their deathbeds decide that it's better to be remembered positively. What they do with the money can overshadow how they got it, securing ongoing praise for generations.

It's a "do as I pay, not as I do" situation.

Re:And from the Linux Kernel "COPYING" file

[Minna Kirai]

No, the `or later' clause can only loosen the conditions of use, so that can't be the reason.

Wrong, it can loosen or tighten. In fact, since there are two sides to every restriction, any change at all can be validly interpreted as both a loosening and a tightening.

For example, if GPLv3 turns out to be identical to BSD, and if Linux allowed "at your option, any later version", then Linux can immediately become proprietary, and all those network-hardware vendors can stop provide source code for their firmware.

Re:And from the Linux Kernel "COPYING" file

[Minna Kirai]

A sane response to the above paranoia would be to add `or any later version approved by me'.

That's a non-response. The effect is identical to having added nothing at all. (Prospective modifiers can ALWAYS ask the original author to re-release under another license)

Re:And from the Linux Kernel "COPYING" file

[Minna Kirai]

It gives the licencee more options, hence it is a more free licence.

The USA should legalize kidnapping! It gives the citzens more options, hence it is a more free country.

In real life, people are smart enough to recognize that every freedom has two sides, and giving more freedom to one party may produce a greater loss of freedoms from others.

GPL software is more free because the software itself is always guarranteed to remain free, never being enslaved like BSD code can be.

On the other hand, congrtess could pass a bill,

No they can't. That would be an unconstitutional "bill of attainder or ex post facto law". Why, if they had powers like that, they could also start giving people special rights to appeal state laws in federal court if and only if you have a child named Terry Schiavo... and that would just be CRAZY.

Re:And from the Linux Kernel "COPYING" file

[Minna Kirai]

How can the language of one version of the GPL authorize the assertion that a different version of the GPL applies to the code?

RTFL.

1. 9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will
2. 
   be similar in spirit to the present version, but may differ in detail to address new problems or concerns.
   
   Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any
   later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free
   Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation.
   

Re:And from the Linux Kernel "COPYING" file

[CableModemSniper]

You agree that if you modify and distribute a binary you have to distribute the modified sources. If you don't agree to that, you don't get to distribute your new binaries. Just cause the original author didn't shake hands or sign it with you there doesn't mean it wasn't an agreement.

Re:And from the Linux Kernel "COPYING" file

[MrResistor]

Freedom of people is the first thing to worry about, and BSDL gives more freedom.

I still disagree.

The GPL may grant less freedom to any given individual, but since derivative works created by others must also be free, that freedom has greater value as a whole, and thus the GPL is the more free license.

It's a matter of perspective. The BSDL is better only for people with an interest in creating proprietary software. For the rest of us, the GPL has far more value.

Re:And from the Linux Kernel "COPYING" file

[CableModemSniper]

links

Re:And from the Linux Kernel "COPYING" file

[eraserewind]

Seems like a perfectly sensible decision to me. He want's to see a legaly binding document before he distributes his code under it. It's irresponsible to take any other position.

Re:And from the Linux Kernel "COPYING" file

[MrResistor]

It gives the licencee more options, hence it is a more free licence.

You have failed to understand my arguement, and you will continue to do so as long as you insist on defining freedom only in the simplest, most immediate terms. Which provides the greatest amount of freedom: allowing the freedom to own slaves, or abolishing slavery?

I am simply applying the same principles to BSDL v. GPL.

On the other hand, congrtess could pass a bill, or a judge make a decision, tomorrow which changed the impact of the GPL, say making the linux kernel unusable in commercial shops.

Since that would result in a fundamental change in the way copyright works in this country, I think the GPL being invalid would be the least of our worries. For that reason, this arguement seems little more than paranoid/delusional ranting.

Who do you think Bill would be more likely to manage to buy off, the FSF or some congresscritters?

Today, congresscritters is the obvious answer, but who can say what tomorrow will bring? Just look at Caldera/SCOX, and see what a difference a few years can make.

Re:And from the Linux Kernel "COPYING" file

[DustMagnet]

Gads. Look at the first link. I can see this is hopeless.

Re:And from the Linux Kernel "COPYING" file

[Anonymous Coward]

You agree that if you modify and distribute a binary you have to distribute the modified sources.

No, I never agreed to such a thing. The license simply tells me, if I do ABC, I can legally do XYZ. If I do ABC and XYZ, it still doens't mean I agreed with the license.

Just cause the original author didn't shake hands or sign it with you there doesn't mean it wasn't an agreement.

My point was the the original author doesn't even know, pretty strange for a legal contract/agreement don't you think?

Re:And from the Linux Kernel "COPYING" file

[phats garage]

I dispute that the BSDL is more free than the GPL. It is only more free at the point of origin. There is no incentive to keep derivatives open, and thus the code quickly becomes less free.

No it doesn't. You download BSDL code. You cannot subsequently order the site where you downloaded that BSDL code to close. Therefore the origional BSDL code does not become "less free," no matter how many secret modifications you make to it.

Re:And from the Linux Kernel "COPYING" file

[phats garage]

Stupid argument. By your methodology, for software to be free it must cause grevious harm to innocent bystanders, and most civilized folks would disagree.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

The USA should legalize kidnapping! It gives the citzens more options, hence it is a more free country.

This is only true if you believe freedom from legal constraints is an overriding good.

GPL software is more free because the software itself is always guarranteed to remain free,

The software is guaranteed to remain free by having a copyright notice and a licence which says you can use it for whatever you like.

The GPL is intended to make it more likely that future software will be free, specifically derived works. The freedom of this bit of software is secure.

Eg, I have here a copy of the Apache source. I have quite wide rights to do things with it. There is bugger all you can do to `enslave' that software and reduce my rights. Well, without buying some congressbots.

Re:And from the Linux Kernel "COPYING" file

[R.Caley]

You have failed to understand my arguement

No, I simply pointed out that it was irrelevent, and indeed, just a restatement of something I said.

Yes, the GPL is intended to create more freedom for more people further down the line. But that is not what was being discussed. From the POV of what can happen to one specific piece of software, say a version of a file in the linux kernel, all that can happen by giving everyone the choice of another licence is that everyone has more freedom WRT that file.

This may or may not be bad for the FSF's political aims of there being more free software, but for this piece of code, nothing can happen which makes it less free.

[congress could undermine the GPL]

Since that would result in a fundamental change in the way copyright works[...]

Er, no, it wouldn't affect copyright at all. This is about the licence, not copyright. Imagine, for instance, a law putting a large insurance or auditing burdon on use of copyleft software. No, it wouldn't make sense, but we are talking about politicians here, the kind of people who pass laws outlawing cheese.

Re:And from the Linux Kernel "COPYING" file

[Deven]

I've read the license (many times), and you entirely missed my point.

A third-party document (the GPL) is claiming that you can use any version of the GPL if it's not specified. By what right can they make such a statement? Only the copyright holder may choose the licensing, not the author of the GPL.

Just because the GPL says "If the Program does not specify a version number of this License, you may choose any version ever published", doesn't mean much. That language has no authority unless that version of the GPL is the license intended by the author. If that is the case, then the version of the GPL is already fixed, despite that clause. If that version of the GPL isn't the license being used, then that clause in the license is just as irrelevant as any random clause in Microsoft's EULA.

The GPL isn't magic. It's only a legal tool. The clauses in any given version of the GPL can only apply if that version of the GPL license is used by the copyright holder. Otherwise those clauses are meaningless with respect to that particular copyrighted work. Even if every version of the GPL contains that clause, the one you (as the recipient) are relying on has to come from a particular version of the GPL, being the one which the author has actually licensed his code under! All other versions (despite identical language) are irrelevant because the author didn't release the code under those versions of the GPL.

So the question is which version of the GPL did the author intend to release the code under? Most likely, the most recent version of the GPL at the time of the license grant. At that point, the selection of which version of the GPL to use is locked -- the language of the GPL itself cannot usurp the copyright holder's privilege of determining the license to use.

If the author explicitly states "any version", then go ahead and use whichever version you want -- the author has granted multiple alternative licenses. But if the GPL version is unspecified, the presumption must be that one version of the GPL was intended, and I don't see any legal basis for the recipient to claim the freedom to arbitrarily select any version of the GPL, especially versions not released at the time of the license grant.

I'm not a lawyer, but this seems pretty obvious, once you think about where the authority of the GPL derives from in the first place -- the rights of the copyright holder.

Re:And from the Linux Kernel "COPYING" file

[MrResistor]

Imagine, for instance, a law putting a large insurance or auditing burdon on use of copyleft software.

If such a law applied only to a narrow subset of an industry, such as only copyleft software, I doubt it would be much more than a temporary inconvenience, as it wouldn't long survive the court challenge. I don't see the industry as a whole willingly taking on, let alone lobbying for, such a burden to their own profit margins just to hamper OSS.

No, it wouldn't make sense, but we are talking about politicians here, the kind of people who pass laws outlawing cheese.

Lucky for us the Founding Fathers took that into account when they wrote the Constitution. This is exactly why we have checks and balances.

Re:And from the Linux Kernel "COPYING" file

[MrResistor]

You're looking only at the short term, and not taking into account the long term consequences.

The BSDL only provides greater freedom to that narrow portion of the population which has an interest in producing proprietary software. There is no guarantee beyond that immediate payoff, since there is no incentive for derivatives to be free.

The GPL contains that incentive, resulting ultimately in more, and perhaps better, free software, as all derivatives must also be free. The freedom of the GPL thus has greater value to a greater number of people.

Which provides the greatest amount of freedom: allowing the freedom to own slaves, or banning slavery?

I'm simply applying that same reasoning the BSDL v GPL.

Re:And from the Linux Kernel "COPYING" file

[hankaholic]

Upon rereading your original post, I think we were coming at the same point from a different direction. I thought you were saying that later licenses could only loosen rights, when you were referring more specifically to the "or later" clause itself.

My post was unclear, and I apologize for the confusion.

Re:And from the Linux Kernel "COPYING" file

[phats garage]

First, let me dismiss a line of your argument:

Which provides the greatest amount of freedom: allowing the freedom to own slaves, or banning slavery?

While shooting slave owners on sight may seem like an honorable thing to you, I still refuse to kill another human being.

Now on to the actual discussion, that of open source software. BSDL allows an individual to redistribute derivative work in an open source form, just like GPL, but in addition, the BSDL allows an individual to redistribute derivative work in a closed source form. Now by definition the BSDL has resulted in additional choices for derivative works. If a closed source release of this code base subsequently reduces your freedom, this could be seen as a net loss of freedoms by naive folks such as yourself, however smarter folks like myself understand that the closed source releaser has himself initiated the freedom reducing act, and that you are partially to blame for depending on his actions for providing you a measure of freedom that you deem necessary.

Re:And from the Linux Kernel "COPYING" file

[MrResistor]

While shooting slave owners on sight may seem like an honorable thing to you, I still refuse to kill another human being.

If you're so smart, why do you have to resort to a straw man to refute my arguement? I asked which provides the greater amount of freedom, a question which you are apparently unable to answer.

I assert that banning slavery provides the greatest amount of freedom. Based on the same reasoning, I also assert that the GPL provides a greater amount of freedom than the BSDL, which allows the code to be "enslaved".

I said nothing about it being necessary, I simply stated that, if freedom can in fact be quantized, the GPL provides more of it than the BSDL.

If you want to insist that freedom only matters for the releaser of the code and those that recieve it directly from him, then you are a fool of limited vision. If you think you don't depend on others every day for everything you have, then you're even more foolish.

Re:And from the Linux Kernel "COPYING" file

[phats garage]

If you're so smart, why do you have to resort to a straw man to refute my arguement? I asked which provides the greater amount of freedom, a question which you are apparently unable to answer.

It wasn't even a strawman, it was entirely irrelevent to software freedom, as was your origional slavery diversion.

If you want to insist that freedom only matters for the releaser of the code and those that recieve it directly from him, then you are a fool of limited vision. If you think you don't depend on others every day for everything you have, then you're even more foolish.

It allows producers the freedom to provide proprietory software from open source software, but even then, consumers do not have to use that software so even in your brand of logic, consumers are not obliged to "enslave themselves" as you put it.

Even if I stoop to the level of using emotional, childish, but ultimately incorrect labels like "slavery", the logic still does not stand. Sorry.

Even if I totally descend to your level of debate, it still stands that BSD produced software in no way contractually binds subsequent derivative authors in a way that forces them to deprive others of freedom. These freedom depriving actions simply do not origionate with the BSD license users.

If I continue in your logical pursuit however, BSD license users could be seen as "enabling freedom inhibitors." Now a similar emotional allegory could be made that automobile manufacturers are enablers for drunk drivers, but when evaluating whats good for society, obviously automobile manufacturers produce goods that few want eliminated from the market, likewise with BSD licensed software, they simply do not provide a "sole source of evil" and thus fail your intellectually dishonest morality play. BSD licensed software only contributes to evil in the same way that morning cornflakes might have contributed to Hitlers regime, had these cornflakes been available to his staff officers for their breakfast. Sorry again but you fail to convince me of any causal contribution of BSD to a net loss of freedom, except in that it limits your particular freedom to offer up dumb arguments successfully.

PS, you can use my invokation of Hitler as an excuse to scurry away from this losing thread you happened to stumble into if you like. I'm generous that way.

Re:And from the Linux Kernel "COPYING" file

[MrResistor]

Even if I stoop to the level of using emotional, childish... Even if I totally descend to your level of debate

Please. Shooting slave owners on sight? Users enslaving themselves to software producers? BSDL producing a net loss of freedom? I never said any of those things. Every one is a strawman constructed entirely by you. If it's emotional, childish arguements you're looking for, I direct you to your own posts.

Clearly, both BSDL and GPL provide a net gain of freedom over the default state, which I will assume we can both agree is defined by US copyright law.

GPLed software, if it is worthwhile for anyone other than the origional author to do anything with it, always produces more free software. This is a gain over and above the origional free release, and it is precisely this cumulative gain that is the basis of my arguement. Indeed, this is the entire purpose of the GPL.

If the same software is instead released under the BSDL, it may produce more free software, but it is much more likely that any derivatives will be proprietary, thus the cumulative gain is greatly reduced in comparison to the GPLed version.

It's still a net gain when compared to the default state, but I supose you could say it's a net loss when compared to the GPL.

To be honest, I'm at a loss as to why the BSDL exists at all, at least in its current form. Without the advertising clause it's functionally no different from public domain. That's it's main weakness, in fact; it effectively becomes part of the commons, and is thus subject to the problems described in "The Tragedy of the Commons". (Clearly, the software itself isn't scarce, but the resources to create and improve it are.)

And don't think the irony of you declaring me the loser while invoking Godwin's Law on your own Hitler reference is lost on me. In keeping with Quirk's Exception, however, I will not be availing myself of your generousity. After all, for me to lose you would have had to address an arguement I actually made, and I'm curious to see if you will.

hmmm

[Anonymous Coward]

site's responsive
6 comments

oh no its the rapture

hello ...
can anybody hear me?

Mod parent up

[FidelCatsro]

I doubt we will see any unplesant forking ,The license must evolve just as the software does .
If we didnt patch the linux kernel and left security holes in it we would have alot of massive problems , the license is like any other code(all be it legal code) bugs will arise and it will need to grow to support new platforms and new inovations .
Thus the clause in the license that the parent states gives the backwards and forwards compatibility if you want it .
unplesant forking will rarely occur

Re:Mod parent up

[natrius]

Did you read the article? Linux itself is licensed exclusively under GPL v2. To change to v3, every copyright holder would have to be contacted and relicense their code. Every other project that does not allow the software to be licensed under future versions of the GPL will have the same problem. This isn't to say that the GPL should stay at v2 forever, it's just pointing out that there are going to be growing pains.

Re:Mod parent up

[FidelCatsro]

Yes , but i dont see where the pains will come from .
VS 2 and VS 3 will be fully compatible and i know the license allows for duel licensing (ala trolltech)
So the worst thing to hapen as i see it, will be the code having to be released as GPL2 and GPL3 code at the same time , and i know for a fact they will be compatible license
Perhaps im missunderstanding the situation , but i see this as a pure storm in the teacup

Re:Mod parent up

[KiloByte]

Some of the contributors are dead, too, in the non-figurative sense.
Try to get them to relicense the part they hold the copyright on.

Linux is the biggest GPLed project, with many thousands of separate copyright holders. The mere code audit would take years, not to mention trying to actually contact them. They are often unreachable, their mail address may be no longer valid, they may use the name Anonymous Chinese Dissident #75483, they may be in a persistent vegetative state, etc etc.

And the Berne convention forbids you to ignore even a single copyright holder.

Re:Mod parent up

[Baricom]

Dual-licensing only works if every person with a copyright interest in the code agrees. In practice, this is extremely difficult unless either the copyright holder doesn't accept outside patches, or requires a transfer of copyright to do so.

Re:Mod parent up

[natrius]

VS 2 and VS 3 will be fully compatible

I'm not an expert on the GPL, but I don't think this is true. Version 3 of the GPL will add additional restrictions on top of what v2 does. GPL v2 explicitly states that you can't add more restrictions. The only way GPL v3 would be compatible with v2 is if it took away restrictions, which I don't think is the case.

Re:Mod parent up

[FidelCatsro]

True enough , and witht he kernel that is alot of people ,However I think there will be little problem here .
As i understand it , the GPL v3 address major issues with international laws which is a benifit to us all , and im sure the kernel developers will see it this way (anyone out there who is a kernel dev please correct me if im wrong)
The license is problemeatic if people in say Belarus dont have to abide by it and can just take the code for their own .The only problem will be caused if someone is trying to cause trouble , and if its just one person there code can be removed and replaced

Re:Mod parent up

[R.Caley]

Yes , but i dont see where the pains will come from .

Every compyright holder must agree to a change in licence, no matter how compatable the licences.

Do you want the job of goiong back through a the history of a major open source project and identifying everyone who made a non trivial contribution to the code, then finding them (all you have is an email address from 10 years ago), confirming it is the right person, getting in touch, getting them to sign a bit of paper, chasing them when they have better things to do, dealing with the heirs of the ones who have died... and doing that under many different legal systems.

Yes, there will come a point where the remainig small contributions can be deleted and reimplemented, but they you have to do all the testing required to reassure everyone that the new function is at least as stable and secure as the previous one.

No fun at all.

Re:Mod parent up

[Welsh Dwarf]

The big problem though is that not all kernel devs are still contactable (or even still alive), so there would be a fair amount of rewriting involved in anycase.

Re:Mod parent up

[FidelCatsro]

Hm i had not though of that , However will the vs 3 add restrictions as such or will they be safegaurds ,or explicit wording to ensure compatibility with international laws as i understand it
well it would be odd if they made it incompatible
((Joke)) unless ofcourse they are trying to crush linux for the glorius rise of HURD((/joke))

Re:Mod parent up

[QuantumG]

Ohhh.. I have a doomsday scenario. Let's say one of those dead dudes' brother - who cares nothing about computers and software - gets a phone call from a lawyer saying they need him to grant permission to change the license on his code in the linux kernel. He quickly calls his lawyer who gets in touch with Microsoft and sells the rights to that little bit of the linux kernel to Bill. Now Bill can say "nope, you can't upgrade to GPL v3." Linus has to get someone to strip out the code that Bill now has a claim to and anything that is a derivative work. They have to be really careful cause if Bill's lawyers do a code audit and find out that just a little bit of the code is still in the GPL v3 linux kernel Bill can use that little known clause of the GPL which allows to you prohibit people from further distributing your source code. Bill could actually revoke Linus' right to distribute the linux kernel!

Of course, then the linux hackers would just remove the last traces of the code Bill owns and release a legit GPL v3, but Bill could probably sue the shit out of them all anyways.

Re:Mod parent up

[FidelCatsro]

Would this be so much of a problem , if so then i see that as a major flaw in the GPL v2 , if they are not avaliable to contact then i would move ahead witht he change and if they later contacted me and disagreed it could be removed , if they are dead then IANAL but would the copywright not either drop or pass to a family member

Re:Mod parent up

[FidelCatsro]

Correct me if im wrong , but is there not a list of kernel developers with contact info .It is good honest Office administration work ;) it would bore the hell out of me but it may be needed.
Lets just hope we learn from this atleast ,as i would hate to see the same hapen with GPL 4.
Though the point being , GPL3 only needs to work with GPL 2 , GPL 2 does not have to work with GPL 3 as such , the kernel and other projects could continue on with GPL 2 and new projects could stay with GPL3 and keep the GPL 2 licensed code GPL 2 licensed

Re:Mod parent up

[natrius]

Would this be so much of a problem , if so then i see that as a major flaw in the GPL v2

No, it's a flaw in the Linux kernel's use of the GPL v2. It'd be easy to move forward if they had included the "v2 or later" clause, but they didn't. I'm sure there was a reason behind it, but this issue isn't the fault of the GPL itself.

Re:Mod parent up

[Anonymous Coward]

Things not specificly stated in a will general revert to the government

Re:Mod parent up

[FidelCatsro]

Too true ,The reason it was licensed as GPL 2 explicitly is something i think that requires some investigation , as it may shed some light if this will be a major pain or not.
If anyone has any background please don't keep it to yourself
I would fathom a guess that it may have had to do witha flaw in GPL vs1 or fears about the changing nature of the GPL license but those are just guess

Re:Mod parent up

[R.Caley]

Is all that information up to date for everyone who has been a developer?

Is every contribution from a non-developer (eg emailed in pateches) legally signed over to one of the developers?

Lets just hope we learn from this

The normal way of licencing under the GPL would be `versionN or later', so any project caught by this now either has a reason, and so will be in the same position next time, or was at some point headed up by an idiot, and we can't avoid that happening again.

Re:Mod parent up

[sumdumass]

Yes there is a list. I read about it when the SCO BS started going down. It is how they checked for any violating ip initialy.

This problem may have been made less of an issue then it would seem. I think they have already done some legwork in tracking down alot if not all of the contributors and have current information because of SCO claiming rights to ip in the kernel. I may be wrong on the amount of work done but a simple website with a "send email with your willingness" to area and instructions to to incluide the contact information that was in use when the code was submited should verify most of the intentions. It would at the minimum narrow the scope of the project down. This site could also be "hidden" from the public and a mass email to all the "known" contributers could direct them to the site.

While this wouldn't be complete, it would give somewhat of an idea as to if it is possible to even change and what kind of timeline would be neccesary if it is. You might also have holdouts liek companies that have gone belly up or are in litigation and the ip contributed is seen as an asset in a court. You also might have problems with persons in certain areas that see the GPLv.3 as a limitation to freedom because the copyright laws it sanctions or limits goes beyond what thier local government permits-enacted.

I can see all kinds of reason why one wouldn't want to change. I'm also wondering why it would be neccesary to change? The clause of "future versions" makes me worry about someone placing "for non-comercial use only" or somethign simular into the GPLv.20 and having it auto-enacted. I doubt that would be the case but does concern me.

Re:Mod parent up

[FidelCatsro]

What is perhaps needed , is a GPL license license , Something to insure that the GPL must assure some basic freedoms and restrictions with the code (IE: copyleft).
This is perhaps needed to stop the GNU licenses ever going sour with these clauses
Infact i would hope that this already exists

Re:Mod parent up

[R.Caley]

The clause of "future versions" makes me worry about someone placing "for non-comercial use only" or somethign simular into the GPLv.20 and having it auto-enacted.

Can't happen. Since you have the right to pick the version of the GPL you licence the code under, you must always have at least the rights in the named GPL version. All a later one can do is give you more rights. Eg, if v6 GPL had a no commercial use clause, you could continue to use it commercially under the v5 GPL.

The reason for the `or later' clause, AIUI, is that it is always possible that the current GPL will be found too restrictive for it's intended effect (eg in some jurisdiction no one considered when drawing it up, or because of some legal decision in the US or EU). If that happens the FSF can release a new version, and everyone now has the choice to take emacs under those terms instead of the broken original ones.

Re:Mod parent up

[m50d]

The recommended way to put your stuff under GPL is "version 2 or any later version published by the free software foundation". It's not a flaw in the FSF setup, it's a flaw in linus' deliberate decision to say v2 only.

Re:Mod parent up

[m50d]

I've thought of something like "this software may be licensed under any license meeting the DFSG provided such a license would not allow the subsequent licensing of the software or any derivative under a license which did not meet the DFSG". Not sure how good that is legally though.

Re:Mod parent up

[Simon]

Version 3 of the GPL will add additional restrictions on top of what v2 does. GPL v2 explicitly states that you can't add more restrictions

Clause 9 from the GPL 2 text says that you *can* use v2 with a later version of the GPL.

It may appear that the "no additional restrictions" clause in v2 might come into conflict with clause 9 and a GPL v3 (or later). But in this case I think that clause 9 clearly takes precedence. That appears to be spirit of the license and the intention of clause 9. You would have a hard time arguing that "no additional restrictions takes precedence" is how the GPL v2 should be read. IMHO.

--
Simon

Re:Mod parent up

[mattyrobinson69]

The KDE team owns the copyright of the code if you send a patch (iirc). This could be worked round in the kernel by some legal bollocks like

"This code is released under the GPL3 although may be relicensed by the maintainer of the Linux kernel to any OSI compatible license)"

Re:Mod parent up

[Anonymous Coward]

Clause 9 from the GPL 2 text says that you *can* use v2 with a later version of the GPL.

If the notice with the code specifies "or later".

Clause 9 also states "Such new versions will be similar in spirit to the present version". If I'd licenced my code under version 2 or later but I disagreed with the new version then clearly I wouldn't be of the opinion that the new version is similar in spirit to the present version. Thus I'd probably have a good shot at arguing (in court if need be) that one of the conditions on my "or later" grant had been violated and therefore the grant is void.

Re:Mod parent up

[Wudbaer]

I find the limitation to V2 quite reasonable from Linus' point of view.

Look at it like that: There is no GPL V3 at the moment, at no one definetely knows what will finally be in there. By putting your software under a license (among others) that just does not exist at the moment you give a blank cheque to the one creating the new version of the license. Ok, the FFS seems to be quite trustworthy at the moment, but many organizations have totally changed over the years and many good intentions didn't turn out good results but goddamned disasters. So I think Linus' decision to limit Linux to the known to be functional GPL V2 seems to me quite sane.

Re:Mod parent up

[FidelCatsro]

Possibly , but that does leave yourself open to problems with the current license and no possible get out claus , As i understand it the GPL can only be replaced with a license in the same vent as the origional , the only aditions allowed are to clear legal issues up such as the current international law issues.
This leaves the kernel vunerable to these problems , However i can see why it was done as the GNU FSF needs to clear this point far more and perhaps give stronger legal asurances that the license will and must remain in its current spirit

Re:Mod parent up

[AssHatAnonymous]

Ok, the FFS seems to be quite trustworthy at the moment...

FFS is more "trustworthy" (I personally would use stabile, robust, etc.) than EXT2FS. It's been more rigorously tested and designed. (Hint: It's funny if you are a nerd, you retarded moderators.)

Re:Mod parent up

[AssHatAnonymous]

Wrong. The clause doesn't say "this license and any future addendums that may be necessary to tweak the spirit of this license." It says "this license, or any future version." It's a wholesale swap of licenses. Not a merge.

Re:Mod parent up

[R.Caley]

It says "this license, or any future version

The important word is `or'. It's not a swap it's an alternative. There is nothing the FSF can do in GPLv3 to reduce what I can currently do to the Emacs code. They could give me more rights under GPLv3, but if they tried to give me fewer, I'd just say `personally, I have chosen to take my copy of this code under GPLv2' and have the same freedoms I do now.

Re:Mod parent up

[Anonymous Coward]

sure lee you want Ed to type "albeit" wrath err than you sing sing gull sill a bull words?

bitches

Re:Mod parent up

[Tony Hoyle]

There's a list of kernel developers but that isn't anything like the number of submitters.

Hell, there are a few lines of code in there that I wrote and I'm not listed.

To change the license you would have to get permission from *everyone* including people like me that once sent a patch via email 2 years ago.

Re:Mod parent up

[AssHatAnonymous]

I understand now, you're an idiot. The issue isn't with the recipients of the code and license. It's with the authors of the code, those who are granting the license.

All of you ranting is based on the fact that you just don't get it. You have no idea of what the problem is.

I'll spell it out for you. A developer choosing to release software under an unmodified GPL v2 license effectively gives the FSF an unrestricted, perpetual right to sublicense in any fashion the FSF desires. At a simple level, there's no problem, developers trusted the FSF. Then the FSF came up with the GFDL, which instead of guaranteeing freedom, guarantees that you will join their cause and propagate their message regardless of whether you agree with it or not.

Re:Mod parent up

[HiThere]

There's a fairly simple way around this, though admitedly it's a slow way. Just start licensing the kernel code under GPL version 2.0 or 3.0. Continue on in this vein as each piece of software is upgraded until everything remaining can be used with 3.0 (code not being developed tends to become unused). Eventually you'll be able to make the shift, and then start the 3.0 to 4.0 conversion.

I see definite advantages to having a relatively stable license. Modular sections could be created where the code linking the modules could be less restricted as to which licenses it could link to, e.g. the RPM module, say, could link to any code under a GPL license approved by Linus, or, after his death, approved by the FSF before 4/1/2012. (Putting a time limit on lessens the potential for future corruption such that the FSF could approve a GPL descendent license saying, e.g., "may be linked with any previous GPL license code, or any code under a license issued by Sun Microsystems". (Or insert some other company's name.)

It seems unlikely NOW, but if you look at history, things change, and not always for the better. (Also, admittedly, not always for the worse, but the BSD code is around to handle that case.)

The other thing to remember is "Linux is a temporary phenomenon!" It's only been around a bit over a decade. It will be lucky to last another decade. Something else WILL appear that renders it irrelevant, though who knows what...AI based operating systems? Something! Possibly even the Hurd. (It's small and ineffective now, but THINGS CHANGE!)

Re:Mod parent up

[Minna Kirai]

Just start licensing the kernel code under GPL version 2.0 or 3.0.

And then compile that code into an executable kernel, which makes you liable for copyright infringment lawsuits from Linus Torvalds, or any of the other 400+ Linux contributors.

Right.

Re:Mod parent up

[HiThere]

No, because it's licensed under GPL 2.0 if that's what's required.

Re:Mod parent up

[tricorn]

However, let's say that GPLv3 has additional restrictions AND additional freedoms - you can't get the additional freedoms unless you agree to the additional restrictions - you either get v2 or v3, not a mixture. In addition, let's say that v3 has only additional restrictions. You're right that it won't affect a v2-or-later-licensed program, UNTIL someone adds in v3-or-later code. Then the project as a whole can only be distributed under the v3 terms. You could strip out the v2 code and release that under v2 terms, or continue using only the original version, but if you want those particular changes you'd have to accept v3 terms.

An impractical question

[Anonymous Coward]

Since the majority of GPL-ed code (with Linux kernel being the most notable exception) is licensed under "GPL v2.0 or (at your option) any later version", wouldn't it be theoretically possible for FSF to publish, for example, GPL v10 saying "You can use and modify this software without any restriction whatsoever, but only if you are Microsoft (Sun, Google, whatever)"? In other words, they can effectively sell the rights to close the sources off all GPL-ed Free Software. Of course, Richard M. Stallman wouldn't do this, but, sadly, he will not live forever, and anyway, whole system shouldn't base its existence and safety on one man.

Re:An impractical question

[FidelCatsro]

http://linux.slashdot.org/comments.pl?sid=143673&c id=12044418
As this poster states , the clause in the license allows us to avoid any situation such as this and secures the copyleft .
However if said did hapen (which i strongly doubt) the only thing to occur would be the license would become more BSD license like allowing people to use the code with your supposed vs 10 for closed source projects

Re:An impractical question

[azmaveth]

Yes, the GPL could be changed in the future this way. No, it wouldn't effectively "sell the rights to close the sources" as you can still choose to use the earlier version of the license. You are not forced into using a later version - it is left as an option.

Re:An impractical question

[latroM]

Quoting GPL: "The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns".

Re:An impractical question

[Anonymous Coward]

Copyright expires after 28 years (*), after which the work passes into the public domain. Microsoft will have the right to use Linux 1.0 eventually, even is the GPL never changes.

(*) the length of the 1710 statute, which I hope someday returns into force.

Re:An impractical question

[Anonymous Coward]

Microsoft will have the right to use Linux 1.0 eventually, even is the GPL never changes.

But not before Bill Gates expires!

Re:An impractical question

[natrius]

No, it wouldn't effectively "sell the rights to close the sources" as you can still choose to use the earlier version of the license.

You can never sell the rights to close the sources without reassigning copyright.

If the GPL was changed like the grandparent poster suggests, as long as someone has a copy of the code out there that says "GPL v2 or later", they can use the code under the terms of the GPL v10, which can say whatever the FSF wants it to say. That's the whole reason behind why some people restrict the license to v2.

With that said, it doesn't sell your rights to close the source, because you have the copyright over the software. The version you have released will always be under the GPL as long as someone has a copy of it, but you can add new stuff to it and only release binaries.

Re:An impractical question

[jeroendekkers]

No, the FSF can't modify the GPL whatever way it wants. Read the GPL:

9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns.

Re:An impractical question

[Anonymous Coward]

The theoretical version 10 would only affect code already extant at the time, or any code that those mentioned in this theoretical version chose to release under this licence. Thus people would be free to take code released under version 9 and develop it as before.

Re:An impractical question

[Anonymous Coward]

But if Microsoft (google, whatever) chooses to fork linux kernel into v10, they could right? And from there develope on it themselfs as closed source...

Re:An impractical question

[m50d]

No, because the GPL says future versions will be "similar in spirit". The FSF publishing a license like that would invalidate the GPL so no-one but the copyright holder would be able to distribute GPLed code at all. It would cause a mess, but it wouldn't let you close the source.

Re:An impractical question

[m50d]

No, but if program foo says "version 2 or later at your option", then Steve Ballmer licenses a copy under v10 to Bill Gates who then can make a derivative of foo called foobar and sell it without giving users the sources.

Re:An impractical question

[Ingolfke]

International copyright law shold be observed (see here). 28 years is a pipe dream. Try life+25 years for most countries as a minimum. Now if a 15 year old kid added code the kernel and lived to be 60. That would mean the code wouldn't be out from under the copyright for 80 years from the point of release.

Re:An impractical question

[Ingolfke]

But not before Bill Gates expires!

Bill Gates' cyborg body and vast nueral network will never expire.

Re:An impractical question

[anthony_dipierro]

wouldn't it be theoretically possible for FSF to publish, for example, GPL v10 saying "You can use and modify this software without any restriction whatsoever, but only if you are Microsoft (Sun, Google, whatever)"?

No. The FSF is a non-profit charity, and isn't allowed to inure to the benefit of Microsoft (Sun, Google, or whatever).

That said, they can make an awful lot of changes legally, without the permission of the copyright holder, and that's why I personally wouldn't license anything I write under "GPL 2 (or later)".

Re:An impractical question

[Urchlay]

So you're saying that in this worst-case scenario, the GPL degrades into the BSD license? Interesting...

Re:An impractical question

[stratjakt]

Only the code he adds.

Kernel 2.6.(whatever today's is) will be free for everyone 25 years after the last contributing author dies.

That is, the source tree that exists right now, this very second.

And if the guy who wrote some obscure little MIPS compatable just won't kick off, you can take his chunks out.

I wouldn't trust RMS

[iamacat]

He has a particular purpose in mind, which is forcing everyone to release source to their software. Personally, I have a different goal - releasing my pet projects for free while making sure any commercial users will talk to me and negotiate attribution, compensation and so on. GPL V2 seams Ok for that, but I will never put an "... or later" clause. Maybe eventually FSF will prevent me from using my own code in commercial products or something. I am not sure intellectual property laws are beneficial (at all or beyond say 5 year duration), but even if people are allowed to copy binaries, I sure shouldn't be forced to give up my source.

Given that Linus/many Linux developers seem to have somewhat different goals than RMS as well, it would indeed make sense for Linux developers to fork the license. It's time for something that follows pragmatic wishes of most free software developers rather than one person's political agenda.

Re:I wouldn't trust RMS

[Anonymous Coward]

I'd just love it if Linux "forked" to a new license and RMS would STFU about it being some kind of subcomponent of the fantasy "GNU" system that he likes to call "GNU/Linux". WTF would he call it if the "GNU" utils and the Linux kernel had a different license? The GNU/kinda-sorta-GNU-Linux OS? blah.

Re:I wouldn't trust RMS

[latroM]

...Maybe eventually FSF will prevent me from using my own code in commercial products or something.

They can't do that. If you have written the code you can do whatever you wish with it. GPL allows you to use the software under it commercially.

It's time for something that follows pragmatic wishes of most free software developers rather than one person's political agenda.

I don't know about you but I value my freedom with free software. The pragmatical POW is too narrow.

Re:I wouldn't trust RMS

[latroM]

It isn't about the license, it is about the building blocks of the operating system and about the principal author of the system whose development begun as GNU.

Re:I wouldn't trust RMS

[MathFox]

For slightly different reasons than you state I would not add a "or later" clause to the copyright notices in my programs. As copyright owner I can allways change the conditions for my open source distribution of my programs. I do trust that the FSF has the well being of Open Source in mind.

I will certainly read the GPL V3 and consider using it on new projects. (Or use an "this project can be distributed under the terms of the GPL V2 or GPL V3" clause.) Only when I find the text acceptable.

Re:I wouldn't trust RMS

[KiloByte]

Note that RMS still haven't realized his error where it comes to GFDL. In nearly every reviewer's but the FSF point of view, GFDL is flawed to the point of uselessness, and yet RMS sticks to his standpoint.

From what I see, people are escaping the leaking GFDL ship like crazy, with FSF manuals and Wikipedia being the biggest bastions.
The former is governed by RMS, the latter is simply too big to allow relicensing.

Re:I wouldn't trust RMS

[femto]

The GPL states that you can use either the version at the time of release OR a later version. Consequently a user is free to ignore any new versions and continue to use code based on the license it was released under.

As for authors, are they going to sue themselves for copyright infringement? An author can do whatever they want with their code.

Re:I wouldn't trust RMS

[anthony_dipierro]

The problem is when you have projects with large numbers of contributors. It's very hard (and probably impossible if you didn't think hard and set things up right in the beginning), to get agreement among all the contributors to switch. Perhaps one possibility is to change the clause to "or any later version published by the Free Software Foundation and approved by [insert your name here]". Then in order to switch the project, you'd have to have a new version which is both published by the FSF and approved by you.

Personally I don't like using the GPL in the first place, and I'll only use it where I have to. I like to have the option of keeping my source code hidden, and if I used the GPL, and someone made modifications to my software, I couldn't make further modifications to that without distributing the source.

Re:I wouldn't trust RMS

[Anonymous Coward]

Then in order to switch the project, you'd have to have a new version which is both published by the FSF and approved by you.

Then the license can be changed until such time as you get hit by a bus, at which point the license frozen as a sort of memorial to you. I like it.

Re:I wouldn't trust RMS

[squiggleslash]

s/commercial/proprietary/g, and even after that it's still wrong.

RMS can't force you to do anything as long as you're the copyright holder, as has been pointed out many times before. If you write some code, and add it to a GPL'd work, you make a work that's derived from that GPL'd work and the combination (but not your original) is licensed under the GPL, which will restrict usage of the combination in proprietary products. (But not commercial projects, people are still free to sell copies, sell support, sell programming time to add custom features, use within a business as an engine for a critical process that makes the company money, etc.)

But your original code is still copyrighted by you, and you can still use it as part of proprietary products when you don't combine it with GPL'd works.

You can even release it on its own under the GPL, and offer companies that want to buy rights to release proprietary versions custom licenses, in exchange for hard cash, of course.

Nothing RMS will ever do will change that. It's not really possible to modify the GPL to remove rights from the copyright holder. A license is something that grants rights to someone who didn't previously have them. The best RMS can do is be elected President, have the rest of the FSF take over Congress, and abolish copyright, and if that happens, "trusting RMS" as far as new versions of the GPL goes will not be your problem.

Re:I wouldn't trust RMS

[runderwo]

He has a particular purpose in mind, which is forcing everyone to release source to their software.

Clueless flamebait. If this were the goal, why would he choose a copyright license as the vessel of his agenda? A license that is optional for a developer to use? Why not just lobby for legislation that would remove the developer's choice in the matter?

Not that mandatory escrow of source code to the Library of Congress would be a bad idea. But you're attributing a false characteristic to Stallman.

Yes, but... is it workmanlike

[Anonymous Cowdog]

All this talk about forking is well and good. But what inquiring minds want to know, is will the GPL3 be written in a workmanlike writing style.

Can't see the problem

[rescendent]

Presumably this is to allow more protection from/for patents and copyrights.

Most distributions include demo versions of commercial software, software in the public domain and software under other licences.

Apache is included in most distributions and apache is under the Apache License and apache aren't entirely convinced their licence is compatible with the GPL...

If Apache can be included, where's the problem?

Just some early morning thoughts...

Re:Can't see the problem

[chadruva]

I don't see a problem to include Apache on a Linux distro, even if apache license is not GPL compatible.

The GPL license dosen't cover materials distributed on the same medium as GPL Software.

Altough it would be nice having Apache to be GPL conpatible, soo we can use GPL plugins with it.

Re:Can't see the problem

[kgp]

But your comments aren't actually relevant to what's being dicussed as the GPL says in the last paragraph of Section 2 "In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License."

Stop being afraid of Change

[ites]

Remember that the version 2 of the GPL dates from June 1991. It is an incredible document, and I agree with Moglen's assertion that it's the basis of a multi-billion dollar industry. Stallman will go down in history as a visionary.

But after 14 years, GPL/2 is starting to age. Yes, it addresses current problems, but remember that software written and licensed today must still be protected and viable in 15 years' time.

There is absolutely no point in postponing the introduction of GPL/3. There must be a migration, and there will be a period of overlap.

But change is not something to fear in itself. It's something to plan and to manage, and in this case, it's essential.

Last thing: if you followed the FSF's recommendations as to how to use the license, your code would contain this text:

# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License as
# published by the Free Software Foundation; either version 2 of
# the License, or (at your option) any later version.

Which all my company's GPL software contains.

Thanks to Moglen, and the FSF for their fantastic work.

Re:Stop being afraid of Change

[Anonymous Coward]

There is absolutely no point in postponing the introduction of GPL/3. There must be a migration, and there will be a period of overlap.

Personally I think they shoul come up with an interim licence. One which is a "beta" of version 3 that also allows for code to be used under version 2 or the future version3. I think it's important to test the waters with new code and get some feedback. Any licence that is going to retroactively apply to some code should have a period for input. Without peer review, who knows what will end up in it. (Search for "enconomy", I doubt the World Bank is going to call anything that...)

Re:Stop being afraid of Change

[Masker]

The problem is that not all Linux kernel sources are licensed in the way that you have (with the any later version addendum). In fact, not all kernel sources are even under the GPL; here is an example of a file that's under the MPL (Mozilla Public License) and GPLv2 (only). Then there are the files (example provided) that simply state the license as GPLv2 without the "or later".

In other words, the Linux kernel has many, many copyright holders, and many, many different variations of the license language in the source files. If Linus had only accepted files with a certain license language, that would be one thing, but I think he was more worried about the technical aspects than the religious war aspects of the kernel.

In short, you can't just promote the kernel to GPLv3, and I, for one, don't really see what the outcome is really going to be...

Re:Stop being afraid of Change

[Anonymous Coward]

I was just thinking: "International Monetary Fund" would be a great name for an indie/emo band.

Misunderstandings

[bonniot]

I will never put an "... or later" clause. Maybe eventually FSF will prevent me from using my own code in commercial products or something.

I think you have two misunderstandings:

1. As the copyright holder, you can always decide to release your code under another license.
2. The "or later" clause is at your option, "you" being the licensee. This means that that clause can only grant new rights, not remove rights, since anybody can always decide to chose to see the software as licensed under GPL v2. This is similar to dual licensing.

Both points mean that your fears are not founded.

Re:Misunderstandings

[Minna Kirai]

The "or later" clause is at your option, "you" being the licensee. This means that that clause can only grant new rights, not remove rights, since anybody can always decide to chose to see the software as licensed under GPL v2. This is similar to dual licensing.

Yes, but anyone who exercises that option may possibly find himself in the sad position of seeing commericial projects modify and re-release her code, without giving source OR compensation.

The "or later" clause gives RMS (or whoever takes over the FSF someday) the option to do ANYTHING with any code released with GPL version X "or later". It sounds a little insane, but we should remember that Microsoft has enough dollars to buy almost anything, included the FSF!

Re:Misunderstandings

[mattyrobinson69]

We should also remember the part of the GPL that says something like "future versions of the gpl will be of the same spirit but with different bits to address new concerns"

Re:Misunderstandings

[argent]

# As the copyright holder, you can always decide to release your code under another license.

But the code under GPLv2, GPLv3, etcetera will still be out there.

# The "or later" clause is at your option, "you" being the licensee. This means that that clause can only grant new rights, not remove rights, since anybody can always decide to chose to see the software as licensed under GPL v2. This is similar to dual licensing.

The issue isn't whether anyone can see the software as licensed under GPLv2, but whether they can see it as licensed under GPLv3/4/5/2010/..., which is what the "... or later" clause allows.

The concern the OP seems to have is that some later version of the GPL may contain clauses that are less restrictive to some proprietary use, and thus reduce the value of his own dual-license strategy. I personally don't think it likely that something like the LGPL will be included in the GPL, but I sure wouldn't want to say it would never happen.

Re:Misunderstandings

[argent]

We should also remember the part of the GPL that says something like "future versions of the gpl will be of the same spirit but with different bits to address new concerns"

That still leaves plenty of room for licenses that reduce the value of dual-license strategies. Some future GPL may include LGPL-like terms, for example, and the OP's commercial customers will find they can switch to the GPL version of his library and quit paying him. It's not likely, but it's not impossible, and it certainly fits in the "same spirit" as the GPL.

Re:Misunderstandings

[anthony_dipierro]

Microsoft has enough dollars to buy almost anything, included the FSF!

The FSF is a non-profit charity. No one "owns" it so you can't "buy" it. Microsoft is a company, so it couldn't even be on the board. Bill Gates could theoretically be on the board, but only if the current board of directors voted for him to be on the board.

Re:Misunderstandings

[mattyrobinson69]

My point was that it cant be too bad. Plus its arguable that an LGPL type license is not in the spirit of the GPL, that would be the spirit of the LGPL.

Re:Misunderstandings

[argent]

My point was that it cant be too bad.

It doesn't have to be "too bad", it just has to be "bad enough".

Plus its arguable that an LGPL type license is not in the spirit of the GPL, that would be the spirit of the LGPL.

The fact that it's arguable means that if the OP's customers pull the ripcord and bale out, if he wants to get them back he'll have to go to court. And don't forget "the courts are always open, but so's the Hilton"... the OP may not be able to afford that. It's *much* easier and more practical to leave the "... or later" bit out and then pick up on later versions of the GPL when appropriate.

Re:Misunderstandings

[orkysoft]

... which would be sometime after the FSF had a flying pig and an ice skating Satan on the board?

Re:Misunderstandings

[xbsd]

...I will never put an "... or later" clause. Maybe eventually FSF will prevent me from using my own code in commercial products or something.

I think you have two misunderstandings:

1. As the copyright holder, you can always decide to release your code under another license.
2. The "or later" clause is at your option, "you" being the licensee. This means that that clause can only grant new rights, not remove rights, since anybody can always decide to chose to see the software as licensed under GPL v2. This is similar to dual licensing.

You are both PARTIALLY right. His fears are well-founded because, in practice, the Free Software Foundation request copyright assignments in order to enforce the GPL "more effectively" , as stated by Prof. Eben Moglen himself:

"In order to make sure that all of our copyrights can meet the recordkeeping and other requirements of registration, and in order to be able to enforce the GPL most effectively, FSF requires that each author of code incorporated in FSF projects provide a copyright assignment" Source: http://www.gnu.org/licenses/why-assign.html

Since most people use GPL'ed software while using Linux, and since the average Linux distro borrows heavily from the GNU Project, most of the code copyrighted under the GPL do not belong to their authors.

My personal advice for him/her is to copyright his code under a specific GPL version while keeping the copyright under his/her name. The rest of what you said is true.

Re:Misunderstandings

[bonniot]

Your advice does not match your analysis. Your are stating that the problem is that FSF asks for assignments. Then you advise to keep the copyright under his name and to license under a specific GPL version. You gave no motivation for the latter.

Re:Misunderstandings

[Minna Kirai]

The FSF is a non-profit charity.

It's as charity today. It could turn into a publically-traded corporation tommorrow with a board vote. "Free Software Foundation" is just a trademark, after all, which can be passed on or even sold to any other group.

No one "owns" it so you can't "buy" it.

I should point out that nobody "owns" democraticly elected officials, but they are bought and sold every day.

Microsoft is a company, so it couldn't even be on the board.

On one side, add up the dollar dollar value it would take to buy closed-source licenses for all the code released under GPL "any later version". On the other side, add up the money it would take to bribe each FSF board member, none of whom are fantastically rich.

As the ratio between those two numbers rises, it becomes more and more cost-effective for a large corp to "buy the FSF", through one of several possible mechanisms.

Re:Misunderstandings

[anthony_dipierro]

It's as charity today. It could turn into a publically-traded corporation tommorrow with a board vote.

No it can't. You can't turn a charity into a publically traded corporation.

"Free Software Foundation" is just a trademark, after all, which can be passed on or even sold to any other group.

Nonsense. The license isn't referring to the name, it's referring to the actual foundation. That's like saying that I can cash a check made out to "Bill Gates" by changing my name to Bill Gates.

On the other side, add up the money it would take to bribe each FSF board member, none of whom are fantastically rich.

Yeah right. RMS is going to just roll over for the millions. C'mon, this guy is an idealist. He'd love to be offered the bribe just so he could brag about how he didn't take it.

Besides, even if the board was bribed, they still couldn't inure to the benefit of Bill Gates or Microsoft. They're a charity, and they can only do things for charitable purposes.

Re:Misunderstandings

[Minna Kirai]

You can't turn a charity into a publically traded corporation.

I've seen examples of charities becoming private corporations (typically because they want to move into political endorsements that support their cause but violate their charitable legal status), and of private companies going public. Logically, those two transitions can happen sequentially.

They're a charity, and they can only do things for charitable purposes.

They can call it "charitable economic development of Pacific-Coast computer science institutes".

Re:Misunderstandings

[anthony_dipierro]

I've seen examples of charities becoming private corporations (typically because they want to move into political endorsements that support their cause but violate their charitable legal status)

They're still a non-profit organization, though. Just not a 403(c) charity.

and of private companies going public. Logically, those two transitions can happen sequentially.

Except that a non-profit organization can't go public, since it doesn't have any shareholders.

Re:Misunderstandings

[xbsd]

Your advice does not match your analysis. Your are stating that the problem is that FSF asks for assignments. Then you advise to keep the copyright under his name and to license under a specific GPL version. You gave no motivation for the latter.

I think it is quite obvious that if he or she keeps the copyright under his or her name and release the code under a specific GPL version (say, v.2 instead of "v.2 or any later version") s/he will always be in control. No matter what RMS may do, his decisions wont affect the copyrighted code. The code is released under v.2 only, not the latest GPL.

Re:Misunderstandings

[bonniot]

I think it is quite obvious that if he or she keeps the copyright under his or her name and release the code under a specific GPL version (say, v.2 instead of "v.2 or any later version") s/he will always be in control.

Let me recall the fears of the OP: Maybe eventually FSF will prevent me from using my own code in commercial products or something. This fear is unfounded as long as the code is release under at least GPL v2, so advising to release under a specific GPL version does not seem appropriate. The only "risk" I see in having "or any later version" is that the later versions will be more permissive. The risk of not having "or any later version" is that if/when the copyright holder goes AWOL/dies and GPL v2 becomes unacceptable for some reason (new laws, incompatible with some other license) nothing can be done about it. I suppose this is especially problematic when there are so many contributors that contacting all of them becomes virtually impossible.

Wow,

[QuantumG]

if you don't trust RMS then why the fuck are you using his license? Seriously. RMS is the pope of Free Software. When he says X is Free Software and Y isn't, that's it, end of discussion. So if you don't trust RMS, bugger off and use a BSD license or some other license.

Re:Wow,

[Anonymous Coward]

RMS is the pope of Free Software.

Does that make Bill Gates the antichrist?

Re:Wow,

[Minna Kirai]

if you don't trust RMS then why the fuck are you using his license?

People change. They can turn greedy and resentful. They may grow old and senile. And inevitably, they will die, and the organizations they lead will be repopulated with other people whose ideas are non-identical.

Someone's past actions are infinitely more trustable than his future.

If you trust RMS today, then use the GPL. If you trust RMS and all his succssors in the future forever, certain they can never be bought, bribed, or bludgeoned, then use GPL plus "at your option, any later version"

Re:Wow,

[QuantumG]

Well firstly, RMS has never changed. He's been the same since the 70s. Even physically, he's like the unaging man. As for the Free Software Foundation, there is absolutely no possibility that it could suddenly change policy, and even if it were to do so, some future version of the GPL which wasn't in the spirit of the current version of the GPL would simply not be valid. There's checks built into the GPL specifically to prevent that.

Re:Wow,

[SA Stevens]

RMS is gonna die. Don't let anybody ever tell you otherwise. Hopefully it won't be for a long, long time, but since he's about 10 years older than I am, I suspect I'll be around to see it happen.

As to there being 'absolutely no possibility' the FSF could change policy, keep the faith, brother.

Re:Wow,

[QuantumFTL]

If you trust RMS today, then use the GPL. If you trust RMS and all his succssors in the future forever, certain they can never be bought, bribed, or bludgeoned, then use GPL plus "at your option, any later version"

GPLv2 states clearly:

Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation.

This means unless you specifically do not allow future versions (by claiming a specific version number), implictly you do.

Re:Pro-fork!!!

[Anonymous Coward]

Story of a Software Project Aplha:

Step 1:
Project Alpha released under GPL2

Step 2:
Developers work on 'Alpha' for many years. Peace and contentment rules the land.

Step 3:
GPLv3 is released, includes software patent language.
Developers revolt! Some want GPLv2 Some want GPLv3..

Step 4:
Alpha.fork1 is GPLv2, and Alpha.forkb is GPLv3.
developers work on their respective forks, mutual resentment burning.

Step 5:
All the improvements from .fork1 get incorporated into .forkb.. because ALL GPL VERSIONS ARE AUTOMATICLY COMPATABLE WITH NEWER VERSIONS.

Step 6:
Developers in .fork1 collectively go 'DOH!' and decide to stop being buttsticks and use GPLv3 like everybody else is. The forks combine.

Step 7:
Developers work on 'Alpha' for many years. Peace and contentment rules the land.

Linux: GPL2 *and* GPL3

[Stephen Samuel]

I would expect that the GPL version 3 will be backwards compatible with GPL Version 2. As such some parts of Linux may remain under the GPL2 while others are updated to GPL 3 as they are maintained.

Making GPL2 and GPL3 incompatible with each other is the kind of thing I'd expect Microsoft to do.

Re:Linux: GPL2 *and* GPL3

[Minna Kirai]

I would expect that the GPL version 3 will be backwards compatible with GPL Version 2.

Although there will be top legal minds writing GPLv3, that compatibility is likely to be impossible. I haven't seen a draft of GPLv3 yet, but I know that one focus is to enable users of software to get access to the source, even if they don't have access to the binaries.

Such a restriction (something like "you must offer the users of the software a copy of the source, or the same offer you got") would be constitute a "further restriction", violating section 6 of the GPLv2: "You may not impose any further restrictions on the recipients' exercise of the rights granted herein."

(On the other hand, if there is no further restriction in GPLv3, then it won't vary substantially from GPLv2, and there will be little motivation for anyone to change)

The only way GPLv3 and v2 may be compatible is by falling back to clause 10, the brute force way: "If you wish to incorporate parts of the Program into other free programs whose distribution conditions are different, write to the author to ask for permission"

Re:Linux: GPL2 *and* GPL3

[ShieldW0lf]

I haven't seen a draft of GPLv3 yet, but I know that one focus is to enable users of software to get access to the source, even if they don't have access to the binaries.

That would be pretty stupid. Being forced to distribute source if you elect to distribute the binary is one thing. Being forced to distribute ANYTHING when you are just USING the software, however, is too ornerous to be tolerated. I would actively look for a GPLV2 fork of the code or use a closed source alternative before I would accept a license like that. And I sure as hell wouldn't recommend it to any of my clients.

Re:Linux: GPL2 *and* GPL3

[anthony_dipierro]

Being forced to distribute ANYTHING when you are just USING the software, however, is too ornerous to be tolerated.

I don't think that's the idea. I think the idea is to require that you give source to anyone if you distribute binaries to anyone. Right now you only have to give source to the person you distribute the binaries to.

Re:Linux: GPL2 *and* GPL3

[dtfinch]

Microsoft would bribe the FSF to add the clause "ALL YOUR CODEBASE ARE BELONG TO MICROSOFT"

Re:Linux: GPL2 *and* GPL3

[plague3106]

Such a restriction (something like "you must offer the users of the software a copy of the source, or the same offer you got") would be constitute a "further restriction", violating section 6 of the GPLv2: "You may not impose any further restrictions on the recipients' exercise of the rights granted herein."

Why would you change the license on an existing version? If you do want to move to the new license, you'll do so during your next release...

Besides your example is not a restriction put on the users or developers...its forcing them to do something, not forcing them NOT to do something.

Re:Linux: GPL2 *and* GPL3

[plague3106]

Not using it...but if you modify and sell something, then people can get your modified code without having to 'buy' your binary. At least thast how I interprate it. But ordinary users wouldn't have this apply to them because they never distribute binaries.

Re:Linux: GPL2 *and* GPL3

[geoffspear]

So if you give a copy of a GPL'ed program to your friend, you should be required to run a server from which anyone on the planet can download the source? That's ridiculous.

Re:Linux: GPL2 *and* GPL3

[Anonymous Coward]

"if you modify and sell something, then people can get your modified code without having to 'buy' your binary."

Maybe, but not from you. By means of the GPL you only HAVE TO deliver source to whoever you deliver binaries AND asks for that source. That's the bare minimum.

Of course, I can do more than the bare minimum and leave the sources to whoever wants them (at sourceforge, for instance), or you can get the sources from anyone else that wants to give them to you, but the bare minimum still aplies.

"At least thast how I interprate it"

Then, I'd suggest you to reread the license to look for where it is said you can get sources from me if I didn't give binaries to you.

Re:Linux: GPL2 *and* GPL3

[Tony Hoyle]

How?

Any substantial project is going to use a mixture of libraries from different sources. It probably has multiple contributors - many of whom you have no contact with any more.

Moving to GPLv3 is limited in that:

1. *All* the libraries you link with must be v3 compatible.
2. *All* the users that use your software must also be prepared to jump to v3 (otherwise you lose a large part of your userbase).
3. You must contact everyone who has ever contributed to your source code and ask their permission. If a single person objects you cannot change.

Re:Linux: GPL2 *and* GPL3

[Tony Hoyle]

I would not be prepared to allow my code to be relicensed with such a clause.

The BSD advertising clause was far less onerous - it only involved a line of copyright.

*forcing* source distribution even to those that have no access to binaries means someone could post a slashdot story like 'foosoft has a GPL light saber in development' and they would be forced to send their incomplete development code to 5 million slashdotters who asked for it. That could easily kill the project.

Re:Linux: GPL2 *and* GPL3

[plague3106]

Any substantial project is going to use a mixture of libraries from different sources. It probably has multiple contributors - many of whom you have no contact with any more.

I'll address your library points below.

Usually if you are contributing to a project you transfer your copyright to the project, so the current project maintainer would decide to change the license. Any contributions you make to Moz. belong to Mozilla, not you.

1. *All* the libraries you link with must be v3 compatible.

Since we don't know what the GPL3 will say, how can you assert this?

2. *All* the users that use your software must also be prepared to jump to v3 (otherwise you lose a large part of your userbase).

The GPL only matters if you redistribute the software; the GPL means little to an end user, unless the user cares about seeing the source.

3. You must contact everyone who has ever contributed to your source code and ask their permission. If a single person objects you cannot change.

If you're working together on a project, the project owns the copyright, not the indvidual authors.

Re:Linux: GPL2 *and* GPL3

[Donny Smith]

>*forcing* source distribution even to those that have no access to binaries

Heh, heh - of course, aparently GPL "hawks" didn't like the fact that thousands of companies run modified code without distributing it (and therefore without having to distribute the modified source).

xSPs, service providers and others will be scared shitless. I wonder if Google's MTAs are modified GPL or commercial.. They'd have to disclose that fact so that people can ask (or not) for access to source code such as Gmail's MTA and perhaps even more than that.

Woo hoo! Great times for Solaris and xBSD...

Re:Linux: GPL2 *and* GPL3

[Anonymous Coward]

You have to either deliver the sources with the binary right away, without being asked for them, or you have to provide them to *any third party*. That includes people who did not buy your product.

The simplest and least risky way of fullfilling the GPL obligations is to put the source right next to the binaries, on a CD or otherwise customary medium. It puts the whole redistribution burden on the client. If you do not deliver the source right away, the chances of someone getting the source who will actually know how to do something with it is much higher AND you have to ensure source availability for THREE YEARS.

Re:Linux: GPL2 *and* GPL3

[anthony_dipierro]

Personally I think it's ridiculous that if you give a copy of a GPLed program to your friend, you are required to give her the source. But I'm not a big fan of the GPL. I only use it when I have to.

Re:Linux: GPL2 *and* GPL3

[geoffspear]

You're not. For noncommercial distribution, you can provide a copy of the offer to get the source that you received with your copy of the binary.

Re:Linux: GPL2 *and* GPL3

[brlancer]

If you're working together on a project, the project owns the copyright, not the indvidual authors.

As an absolute statement, that is patently false.

Usually if you are contributing to a project you transfer your copyright to the project....Any contributions you make to Moz. belong to Mozilla, not you.

Now we're in usually. Better.

One of the major concerns of this thread is the linux kernel, which is bound to a specific version (2.2?). I know that all contributors have not transfered their rights--it only takes one instance to disprove.

Unless you can get every contributor to a project to agree, you cannot change the licensing because each contributor much re-license their part.

Re:Linux: GPL2 *and* GPL3

[anthony_dipierro]

For noncommercial distribution, you can provide a copy of the offer to get the source that you received with your copy of the binary.

Only if I received a copy of an offer to get the source with my binary. Which rarely happens. And as you said, only for noncommercial distribution.

Re:Linux: GPL2 *and* GPL3

[Minna Kirai]

Being forced to distribute ANYTHING when you are just USING the software, however, is too ornerous to be tolerated.

Some people tolerate it already (scroll down to section 2d).

The Affero GPL is a modification of GPLv2 that requires all users to be provided the source, even if they only execute the program remotely. RMS has stated an interest in this provision, and was at least considering putting something along those lines into GPLv3. Whether that is still under discussion or not, I don't know (GPLv3 drafts don't seem to be publically released).

The concept obviously introduces all kinds of confusing and far-reaching implications: the line between users and non-users of a certain piece of software is difficult to draw. If my ISP's router runs Linux, am I "using" Linux when I view this webpage on Windows XP??

Re:Linux: GPL2 *and* GPL3

[Minna Kirai]

someone could post a slashdot story like 'foosoft has a GPL light saber in development' and they would be forced to send their incomplete development code to 5 million slashdotters who asked for it

(a) The hypothetical clause didn't do that. It only forced distribution to "users", and none of those slashdot readers are using the software yet.

(b) Even if a different hypothetical clause did require you to release the code to anyone who asks, even non-users, that wouldn't kill the project. That's because you are allowed to charge whatever money you need to cover the time & effort you expend providing that copy.

Re:Linux: GPL2 *and* GPL3

[Minna Kirai]

Heh, heh - of course, aparently GPL "hawks" didn't like the fact that thousands of companies run modified code without distributing it (and therefore without having to distribute the modified source).

This is a confusing point, because even a FAQ on the GNU website gets it wrong- but it's almost impossible for a company to make productive use of modified code without distributing it.

That's because the majority of companies have more than 1 employee, and when one employee transfers a copy to another, that is distribution, an action which is illegal under copyright law, unless you have permission from the original author. It's no excuse that both copies are still controlled by one legal entity- they have been spread out geographically and electronically, and that is (by definition) "distribution".

If you have trouble understanding this, just imagine Wal-Mart buying a single copy of Windows XP for their 20,000 managers' PCs, and making the excuse "copies for internal use aren't really copyright infringement"

The conclusion is that if a manager directs a programmer to alter a GPL program, and then provide that program to a different employee to use, he has given permission for either employee to post that source code on sourceforge. If he tries to stop them, then he is in violation of GPL section 10 ("You may not impose any further
restrictions on the recipients' exercise of the rights granted herein"), meaning the company can't make copies of the altered program.

Re:Linux: GPL2 *and* GPL3

[plague3106]

One of the major concerns of this thread is the linux kernel, which is bound to a specific version (2.2?). I know that all contributors have not transfered their rights--it only takes one instance to disprove.

I'll take your word for it then; however, it still isn't a big deal. The ones that disagree can have their code removed, and someone else can recode the functionality and put it under the new license. As long as it does the same thing, it doesn't matter too much how its done.

Why this would be allowed in such a large project is beyond me, it seems like it would cause more problems than its worth.

Re:Linux: GPL2 *and* GPL3

[ShieldW0lf]

That's nice. Very interesting. Never heard of it before despite it's having been released in 2002, and it sure doesn't seem to have taken the world by storm.

So is that co-incidence, or are you trying to support my argument?

Re:Linux: GPL2 *and* GPL3

[Anonymous Coward]

Getting >400 people, some of whom are dead, to agree on a license change is no big deal? Have you ever actually looked at the Linux kernel sources? Do you have any idea how huge it is (and, frankly, something of a mess)?

Also note that outside of official FSF projects and a few others, individual contributors owning their own copyrights is the rule, not the exception.

However, with this in mind, the GPL was written to allow any later version to be used; the Linux kernel is the exception, because Linus doesn't have sufficient trust in the FSF.

Re:Linux: GPL2 *and* GPL3

[plague3106]

I find it hard to believe that each contributor to the kernel retained their copyright. Not every 10 line piece of code is copywritable by the way.

Otherwise we'd be in alot of trouble for everyone who wrote a simple for( i = 0 ; i 10 ; i += 1) { printf( "%d", i ); }

Re:Linux: GPL2 *and* GPL3

[geoffspear]

Sure, if by "rarely" you mean "in every single case of GPLed software you can name".

Re:Linux: GPL2 *and* GPL3

[anthony_dipierro]

Really? I guess I've just never noticed it. For instance, where's the written offer that comes with Firefox? I figured most people were using the fact that "offering equivalent access to copy the source code from the same place counts as distribution of the source code".

Re:Linux: GPL2 *and* GPL3

[geoffspear]

You figured wrong. Offering access to download the source code is not the same as including the source code with the binary. I suggest you read the GPL a few times before making pronouncements about what it says.

Re:Linux: GPL2 *and* GPL3

[anthony_dipierro]

Offering access to download the source code is not the same as including the source code with the binary.

The GPL says "If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code."

I suggest you read the GPL a few times before making pronouncements about what it says.

I wasn't making a pronouncement, I was making a direct quote. That's why I used those quotation marks.

Whats all the hubbub? Bub?

[mnmn]

So what is in GPL3 thats causing all the commotion? All I hear is people saying they'll do this and that and them saying no we wont, and when its out you wont worry. Does anyone even know what differences will be in GPL3?

The only change I'd like to see is " this code cannot be used by Microsoft or SCO or its subsidiaries, or employees in any fasion ". Or better "this code cannot be used by GW Bush to kill innocent people in any country under any circumstances whatsoever" or something to that effect.

Theyre using WindowsNT to drive the battleships anyway.

Re:Whats all the hubbub? Bub?

[latroM]

That wouldn't be neither free software nor open source then

Re:Whats all the hubbub? Bub?

[Cardinal Biggles]

The only change I'd like to see is " this code cannot be used by Microsoft or SCO or its subsidiaries, or employees in any fasion ". Or better "this code cannot be used by GW Bush to kill innocent people in any country under any circumstances whatsoever" or something to that effect.

Then the GPL would no longer be an Open Source license, or even a Free Software license.

See items 5 and 6 of the Open Source Definition ("No Discrimination Against Persons or Groups", "No Discrimination Against Fields of Endeavor"), or the Free Software Definition ("you should be free to redistribute copies, either with or without modifications, either gratis or charging a fee for distribution, to anyone anywhere .")

Re:Whats all the hubbub? Bub?

[CDarklock]

> what is in GPL3 thats causing
> all the commotion

Possibility.

Knowing that there will be a GPL v3, there are many people screaming of doom and gloom, because it could be TERRIBLE! It's POSSIBLE that RMS could do HORRIBLE things, just look at this protracted convoluted example! WE'RE ALL GOING TO DIE! DIE OR GO TO JAIL! OR BOTH! And then the philosophers will starve and you'll have to reboot.

Of course, if you like the GPL v2, chances are you're going to like the GPL v3. It's probably not going to be that different. The GPL exists and goes through versions primarily to keep the lawyers happy, not to push us into a trap.

On the other hand, if you think the GPL v2 *is* a trap, v3 will be pretty much the same trap. Meet the new boss, and all that.

Re:Whats all the hubbub? Bub?

[SA Stevens]

The only change I'd like to see is " this code cannot be used by Microsoft or SCO or its subsidiaries, or employees in any fasion ". Or better "this code cannot be used by GW Bush to kill innocent people in any country under any circumstances whatsoever" or something to that effect.

One of the biggest threats facing the Free Software movement is people like you.

Microsoft actually SELLS certain packages (well, they used to. Services for UNIX are free now) that contain GPL'd software, you know. When I've run the install.exe to install it, there are explicit click-through popups that notify the installer of the fact.

Not that easy

[Anonymous Coward]

If GPL3 mentions anything that is not in GPL2 (ie it places restrictions relating to patent litigation etc) then it cannot be compatible with GPL2.

The only thing the GPL3 can do and still be compatible with GPL2 is to have fewer restrictions. In which case, what's the point, we already have BSD.

Re:Not that easy

[anthony_dipierro]

The only thing the GPL3 can do and still be compatible with GPL2 is to have fewer restrictions. In which case, what's the point, we already have BSD.

BSD is not copylefted.

Re:Not that easy

[shutdown -p now]

Yes, but isn't GPL the most free copyleft license one can possibly write? In other words, if you remove a single restriction from it, would it still be copyleft?

Re:Not that easy

[mrwiggly]

If GPL3 mentions anything that is not in GPL2 (ie it places restrictions relating to patent litigation etc) then it cannot be compatible with GPL2.

And who says that v3 has to be backward compatible?

Backward compatibility is very useful for API's, but not desireable for a FOSS license that has a loophole that needs to be closed (ie: the application server loophole)

Re:Not that easy

[Anonymous Coward]

BSD? You mean the steal this code license?

Re:Not that easy

[SA Stevens]

And who says that v3 has to be backward compatible?

Anybody who has released any code under the old license that the maintainers have any hope of relicensing under the new license. It's really that simple.

Re:Not that easy

[Dwonis]

I think you could drop stuff like section 2c, or rewrite the GPL to be the same as the LGPL, and still be copyleft, but I can't think of any reason why that would be done.

Re:Not that easy

[temojen]

Or just less rant.

Re:Not that easy

[Anonymous Coward]

If GPL3 mentions anything that is not in GPL2 (ie it places restrictions relating to patent litigation etc) then it cannot be compatible with GPL2.

Not true. Depends on what your project claims to be licensed under. See section 9 of the GPL:

Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation.

In other words, the GPL v3 could be the MS EULA, and if your product claims to be licensed under v2 or any later version people would be allowed to consider it licensed to them via the MS EULA.

The big irony is that GPL upgrades are the only way to introduce more restrictions on how GPL'd code can be licensed. Any other license applied to the same code must have the same or fewer restrictions as the GPL, but the new versions of the GPL itself get a free pass.

Re:Not that easy

[ArekRashan]

But why would you choose the more restrictive latter license if the older version is still an option?

Re:Not that easy

[agbinfo]

• Step 1, get a copy of s/w released under GPLv2 or later
• Step 2, add new functionalities and bug fixes
• Step 3, release under GPL v3

So contributor released under v3 because he believed that it was a safer version and better suited to protect his code and you choose the v3 license because it contains the additional functionality, bug fixes.

Re:Not that easy

[aichpvee]

I always thought it was more the "free, you haul" license.

Re:Not that easy

[anthony_dipierro]

Yes, but isn't GPL the most free copyleft license one can possibly write?

Nope, the QingPL is (ShareAlike 1.0 is a close second though).

In other words, if you remove a single restriction from it, would it still be copyleft?

No, if you removed the restriction that you can't distribute binaries without distributing source, it'd still be a copyleft.

Re:Not that easy

[anthony_dipierro]

I think you could drop stuff like section 2c

Sure, 2c, 2a, 3a, 3b, 3c, 4, parts of 1...

but I can't think of any reason why that would be done

Because you want to make the software more free. "[Carrying] prominent notices stating that you changed the files and the date of any change" is a real pain in the ass for a program which has undergone hundreds of thousands of changes by thousands of contributors. Forcing people to distribute the source code of any change they make is even worse.

Re:Not that easy

[Stephen Samuel]

The FSF promises that any new version of the GPL would be in the same sprit as the earlier versions. If the FSF were to try and make the MS EULA be GPLv3(or 4 or 5) someone would almost surely sue them for breach of contract (and/or trust) and get it rescinded.

Re:Not that easy

[Per Abrahamsen]

GPLv2> Such new versions will be similar in spirit to
GPLv2> the present version, but may differ in detail
GPLv2> to address new problems or concerns.

I believe a convincing case can be made that the MS EULA is not "similar in spirit" to the GPLv2...

Concisely put:

[mas5353]

while(1) fork();

Re:Concisely put:

[Zeebs]

Cease and Desist!

You are unlawfully distributing the source code to one of my projects, rabbit.
-God

I don't trust YOU

[Jonti]

(RMS) has a particular purpose in mind, which is forcing everyone to release source to their software.

Please could you cite your source for this? I mean, I agree that there is a case to be made that any software offered for sale should "display its ingredients" as it were. One could argue that people and corporations should have a right to see what the software they need to run actually does.

But I'm not sure that RMS has specifically called for legislation on this. Can you help out with a reference to back up your claim, please?

Shocky !

[ReidMaynard]

Shocky is, quite simply, for people who have trouble waking up. When the alarm clock goes off and the snooze button is pressed, Shocky will give a mild 500 volt shock. Each successive Snooze press increases the voltage by a factor of 10 !

Great gift for Xmass .. Stock up Today.

Re:Shocky !

[mysticwhiskey]

Wake up - you've posted in the wrong forking topic! BZZZZZZZZZZZZZZZZT!!!!

Re:Shocky !

[koreaman]

LOL but...
posted in the wrong topic

So called doomsday scenario

[nickco3]

Let's say one of those dead dudes' brother ... [snipped] ... sells the rights to that little bit of the linux kernel to Bill.

OK. And then what?

Linus has to get someone to strip out the code that Bill now has a claim to

OK. Shouldn't take long.

So what else? You mean that's it? That's all? That's your doomsday scenario?

If you pick a piece of the kernel code at random it's probably a driver, hardly a doomsday scenario. Even if we lucked out and lost a critical component like the scheduler, there are any number of competing implementations that have either been discarded or never made it into the official kernel. It's not like we'd be starting from scratch.

All that so-called "wasted effort" on competing implementations that Slashdotters piss-and-moan about is actually genetic diversity that helps guarantee Open Source software's long term survival.

Re:So called doomsday scenario

[QuantumG]

Uhh.. say it is a driver. Say it's self contained. What about those people who wrote a similar driver by taking the existing one, stripping bits out and adding bits in? I mean every driver looks kinda similar.. Bill's lawyers could easily claim that every driver that "looks like" the one Bill has a claim on is derived from it. If you argue he'll drag your ass into court (not IBM, not some legal fund, YOU) and ask you questions like "so you're tell me you've never read the source code to another driver before you wrote yours?" "Oh, you did? Then how can you be so sure you never read the source code to our driver and accidentally copied it into yours?" And that's the kind of shit lawyers will ask you. They'll make it look to a judge like you've never put any effort into making sure you had "clean hands" because the license said you were free to copy. Then once Bill's lawyer has shown you can't possibily have clean hands he'll say that you have no authority to change the license on your own code without his permission. Your driver won't be going into the GPL v3 linux kernel either.. and when some significant portion of the linux kernel has been put on dubious ground it'll be really easy to say that the whole kernel is dubious.

Now, with all that pain fresh in your mind, can you think of a really good reason why we'd ever need to move the linux kernel to GPL v3? Is it really worth it?

Re:So called doomsday scenario

[SA Stevens]

You are forgetting all the 'cease and desist' orders sent out to any entity that continues to distribute the older kernel source tarballs, and/or any ISO containing the 'offending' binaries.

There are vulnerabilities there that make the 'opening' that SCO exploited look minor.

Re:So called doomsday scenario

[nickco3]

No, no, no. You can't retrospectively withdraw GPL licensing. If Bill inherits the rights to some code and then declares it's no longer licensed by the GPL everybody who received a GPL'd copy right up until that point still has a valid license. Only the new recipients (if any) get the new license. Everyone who ever had a GPL'd copy is still legit. And so are all its derived works.

There is no real mechanism for an author to withdraw GPL'd code from the public once it's out there. We are in no danger when authors change their mind (it does happen), and we are in no danger if somebody awkward inherits the rights.

Re:So called doomsday scenario

[Anonymous Coward]

there are any number of competing implementations that have either been discarded or never made it into the official kernel... competing implementations... is actually genetic diversity that helps guarantee Open Source software's long term survival.

You're neglecting the fact that these "competing implementations" didn't make it into the official kernel for a reason. Most of them are unfinished, buggy, incorrect, or poor performers. So while, yes, you can simply grab some other code, you're taking a step backwards in functionality and reliability.

If you're married to some Platonic notion of "open source software" that merely "survives", you may not care about the distinction. If, however, you want open source software to be actually widely adopted and useful, rather than confined to a niche of people that use it only because of ideological preference or simple lack of cash, then mere survival is not enough.

Re:So called doomsday scenario

[QuantumG]

Read the GPL. It specifically says that if you violate the license you lose any rights you have under that license. That means you have no rights to copy and distribute the software, forever. No-one else loses any rights, just the person who violated the license.

Re:Pro-fork!!!

[dadjaka]

Actually, I read on /. that several of the clients just copy work that others have done. While I doubt that they will ever join up, there may not actually be that much duplication.

Version 3: to good to use

[Anonymous Coward]

Forking seems like less of a problem than the simple that Linux simply cannot practically be released under a license other than the GPL v2. Linus established Linux as being licensed specifically under version 2 of the GPL. All contributors to Linux hold the original copyright to the code the contributed. The contributions -- as modifications to code explicitly licensed under only version 2 of the GPL -- are themselves explicitly licensed under only version 2 of the GPL. Even if there were compelling reasons to migrate Linux to GPL v3 (patent provisions, etc.), the only way to do so legally would be to contact every single contributor to Linux ever and get them to explicitly agree to re-license their contributed code under the new version of the GPL.

Re:Version 3: to good to use

[Anonymous Coward]

Obviously, contacting everyone and getting their consent to re-license would be too much. But you could make a transition over a period of 5-10 years if you had every new contribution dual-licensed under GPL2 and GPL3. Once the GPL3-licensed code has critical mass, someone could decide that it's worth their while to go back and rewrite any remaining GPL2-only portions, so the whole thing could be GPL3. At that point, you could take GPL3-only contributions, and the transition would be complete. The first GPL3-only contribution would ensure that the future versions of Linux would be available as GPL3 only, since the newest features would only be available under that license.

Obviously, this is a slow undertaking, but it's a viable way to make the transition, if that's what seems best. At minimum, asking for new contributions to be dual-licensed gives people options down the road.

Wrong

[Anonymous Coward]

Yes, the GPL could be changed in the future this way.

And at that point a lawyer would be involved, and it would undoubtably be ruled that the proviso:

Such new versions will be similar in spirit to the present version

had been violated and the "upgradable" nature of the old code would be void.

This is the main danger with instituting a newer version. The more it does over the current version the less obviously it will be in the "spirit" of the current version.

acces to souce.

[leuk_he]

That would be pretty stupid.

Imagine a embedded device running linux. You can use it but the distibuter gives you no updated "firmware" and thus no binaries. The only way to find out it runs linux is to "hack" it.

This might a way builders of embedded hardware try to circomvent the GPL since they give you no access to the binaries. (This is the way the embeded hardware builder would explain it, this is open to discussion. )

Now comes the strange part: give out firmware updates would violate the GPL. now lets talk about stupid.

Re:acces to souce.

[codeguy007]

Imagine a embedded device running linux. You can use it but the distibuter gives you no updated "firmware" and thus no binaries. The only way to find out it runs linux is to "hack" it.

Come on this is just getting really stupid.

If they have box running the firmwire, they have the binaries. Because the firmware is the binary. You can't run the software without the binaries.

Re:acces to souce.

[Mycroft_VIII]

It seems to me if it's using gpl'd code anywhere in the machine they are indeed distributing binaries when they distribute the machine.
I fail to see any difference of a meaningfull nature between code on a (flash)rom on a cd.
Of course I'm not a lawyer so there could be some braindamaged distiction written in some law or rulling somewhere. But such a thing would be false to fact imho.

Mycroft

Re:acces to souce.

[SA Stevens]

A lot of embedded controllers have the firmware directly burned into an area of ROM in the CPU that is not aceessable to the outside world. Some have a provision to extract and/or verify the contents of this ROM. Others have hardware 'protection' schemes, such as a bit that can be set, that permanmently disables the ability to access the ROM code. With some parts, you'd have to have access to the firmware used to produce the silicon mask pattern to access the binary.

Re:acces to souce.

[shaitand]

It doesn't matter. Access to manipulate the binary is not a requirement of the GPL. The binary changing hands from vendor to end user is enough, it doesn't matter what media/hardware it is distributed on or whether that media/hardware is open.

If a binary passes from you to me, you are obligated to provide me with the code and your modifications to that code.

Re:acces to souce.

[sjames]

Imagine a embedded device running linux. You can use it but the distibuter gives you no updated "firmware" and thus no binaries. The only way to find out it runs linux is to "hack" it.

Nevertheless, the device DOES contain binary code which was sold (distributed) to you, so you DO have the right to the source code that resulted in that binary.

Hiding a license violation doesn't make it go away.

Re:acces to souce.

[jbolden]

I don't see how this makes any difference. Lets say I create a program. An embedded software manufacturer uses this program in his product. His product is a derived work of my program. He then distributes the product, he's distributing a derived work of mine.

It makes no difference how he goes about distribution. Its still a copyright violation unless I gave him permission. The GPL only gives him permission under limited circumstances. I can't see a hardware manufacturer winning a lawsuit if Linus were to sue here.

Re:acces to souce.

[Stephen Samuel]

This might a way builders of embedded hardware try to circomvent the GPL since they give you no access to the binaries. (This is the way the embeded hardware builder would explain it, this is open to discussion. )

The GPL kicks in when you distribute the code (in any form), not when a user sees that code as a separate entity. If I give you an embeded product containing GPL software, I am required to give you a copy of the source code.

If this 'hack' worked for the GPL, companies would also be able to distribute machines with embeded versions of Windows CE without paying royalties to MS. Releasing the source code is, effectively, the royalty for distributing GPL code.

The only reason why some companies think that they can distribute GPL code in embeded products without following the GPL requirements is that they don't think that the authors will be able to identify the code as theirs. (and many of these companies have been caught with their pants down).

Re:acces to souce.

[benb]

> Imagine a embedded device running linux.
> You can use it but the distibuter gives
> you no updated "firmware" and thus no binaries.

I got the binaries as part of the device. ROM is also a medium.

Re:acces to souce.

[Binestar]

The fact that the software is on the device means they have given you a binary. (does the device recompile the software everytime it boots or something?) GPL doesn't say you get source when you get "access" to the binaries, you get the source when you get the binaries.

dead copyright holder.

[leuk_he]

Bill could actually revoke Linus' right to distribute the linux kernel!

It might not be possible to distibute it under gplv3, but that part was licences under gpl v2. and ssince you licenced it under gplv2 it may still be distibuted under point 4:

4. ......However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.

That makes the gpl inrrevokeable.

Nice thought expriment, but i would liked to have seen the part of the licence that would be violated.

but Bill could probably sue the shit out of them all anyways.
As always this is true. He could be "not right" but still sue, as the sco tries. MS could sue many many small competitors to death.

Re:dead copyright holder.

[QuantumG]

Yes, but say if someone was to distribute a GPL v3 linux kernel that contained code which had not been released for GPL v3 distribution, the owner of that code can revoke the right of that person to distribute their code. The rest of the controversial section 4 of the license:

4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.

To me that means everyone who distributed the tainted GPL v3 linux kernel would lose the right to distribute the unauthorised code, even if they were to subsequently aquire any copy of the code under compliant terms. A lot of people distribute the linux kernel. That's a lot of people who our antagonist could prohibit from distributing the linux kernel until such time as the offending code (and all derivative works) was removed from the linux kernel.

Even if I'm wrong it still sounds like a legal mine field. The only real safe thing to do is to distribute the code under the exact license under which it was written for release to the public. For the linux kernel that's the GPL Version 2, June 1991.

Re:dead copyright holder.

[danielk1982]

That makes the gpl inrrevokeable.

No it doesn't. From what I understand the copyright holder (e.g. code contributor) can always relicense his contribution. It just hasn't been done yet and tested in court.

Re:dead copyright holder.

[Tim C]

No, he can licence it to other people under a different licence, and he can stop distributing it under the GPL, but he can't revoke licences that have already been granted.

If I write some code and give it to you under the GPL, then sell it to IBM/Apple/MS, you still have your GPLed copy with all the rights that affords you.

Re:dead copyright holder.

[danielk1982]

No, he can licence it to other people under a different licence, and he can stop distributing it under the GPL, but he can't revoke licences that have already been granted.

If I write some code and give it to you under the GPL, then sell it to IBM/Apple/MS, you still have your GPLed copy with all the rights that affords you.

The original copyright holder remains the owner and can revoke the permission that he grants you through the GPL. A license is just that, a license to use or make use of a property not belonging to you. Its not a contract which states that as long as you follow your end, the owner has to follow his.

There is nothing in the GPL that makes in irrevocable.

Re:dead copyright holder.

[Physics Dude]

Yes, but say if someone was to distribute a GPL v3 linux kernel that contained code which had not been released for GPL v3 distribution

That's exactly why people here have been saying that it'll likely be impossible to release the kernel under the GPL v3. It's just not going to happen.

Re:dead copyright holder.

[jbolden]

I agree this is standard but I don't see how this would hold up. You get another copy of the program you have a new license. The old license is terminated. The author would have to show you didn't get another copy from a clean source since then you would have acquired a new license even though your old license had been terminated.

The law doesn't recognize the platonic ideal of "the source code" seperate distinct from specific instance of the source code.

Re:dead copyright holder.

[Minna Kirai]

To me that means everyone who distributed the tainted GPL v3 linux kernel would lose the right to distribute the unauthorised code, even if they were to subsequently aquire any copy of the code under compliant terms.

No. As it says, any attempt to violate that license will "terminate your rights". But, your rights having been terminated once doesn't prohibit you from re-acquiring them. Just look at, say, an ISP lease or any other kind of service contract. Repeated nonpayment may terminate your usage rights for the service, but it certainly doesn't block you from signing up again in the future. People terminate and then re-establish licenses all the time.

There is no kind of "Free Software Blacklist", where people once convicted of attempting to welch on the GPL are never allowed to use GPL code again.

Re:dead copyright holder.

[QuantumG]

Uhh no. There is only one license, issued by the owner of the copyright. The GPL specifically says the owner can revoke that license for people who violate it. It would be a nonsense clause if the user could just go and aquire another license for the software from someone other than the owner of the copyright. The court is more likely to uphold something that makes sense than something that is nonsense.

Re:dead copyright holder.

[QuantumG]

There is if the copyright holder wants there to be. If you violate my license and I feel like I'll prevent you from ever distributing my software again by using that clause of the GPL. Any judge would read that clause and agree with my interpretation of it. Especially seeing software licenses are more often used to restrict people's freedom than protect them.

Re:dead copyright holder.

[Minna Kirai]

If you violate my license and I feel like I'll prevent you from ever distributing my software again by using that clause of the GPL.

Nope. Once you've released something under the GPL, you can never pull it back in.

Re:dead copyright holder.

[QuantumG]

Look, read the freakin' license, it specifically says that if you violate the license your rights under the license are terminated. That's it, there's no regaining the rights from someone else, the copyright holder has terminated your rights. Even more so, countries like Australia (which is where I'm at BTW) specifically state that a copyright holder can terminate a license at any time and for any reason. The Berne Convention would permit me to exercise that right in any country that has signed the WIPO treaty, so frankly it doesn't even matter what the GPL says.

Re:dead copyright holder.

[jbolden]

The courts have alread upheld this "nonsense" idea multiple times. For example the in the Cleanflix case the courts were very clear that each copy purchased carried a seperate license for that copy. That is a copyright holder must stop selling you copies (even indirectly) if he wants to stop giving you additional licenses. In this case all Clearflix has to do to get an additional license to a commercial movie was to purchase a valid VHS copy from any legal source, in particular any distributor.

Programmers like to think that the license applies to a platonic version of the code and not a specific copy but that's simply not the case. Copyright law covers exactly what it sounds like literally who has the right under what conditions to make a copy of a physical work.

For a GPL word anyone is empowered to grant a license who has the code and hasn't violated the GPL. The author doesn't have rights under the GPL that are different than other people's (of course the author has the copyright and thus has the right to create copies under a totally different license which others don't have).

Re:dead copyright holder.

[Minna Kirai]

Look, read the freakin' license, it specifically says that if you violate the license your rights under the license are terminated.

Yes, I've read it. You violate it, your rights under the license are terminated. So then download one of the thousands of other copies with identical licenses from all over the internet, and proceed using it.

You might be confusing "this license" with "other licenses that are identical copies of this one". Those things are not the same, for copyright purposes at least. Each license file stands on its own.

specifically state that a copyright holder can terminate a license at any time and for any reason. The Berne Convention would permit me to exercise that right in any country that has signed the WIPO treaty

The Berne Convention does not allow for license termination in any way (without affirmative consent from licensee). Only if the work was copyrighted before your country became a Berne signatory might you have that ability, and only towards fellow citizens. (And Austrialia probably joined the treaty too long ago to be relevant to modern programming)

Hypothetically though, if you had modified a GPL program back before attaching to Berne, and if you then revoked your license, you would have violated the GPL, and no longer allowed to distribution modifications of the original yourself. (Until, of course, you re-acquired a new GPL license, which you don't believe is possible)

Re:dead copyright holder.

[QuantumG]

The Berne Convention is that pesky treaty that allows any country to enforce its specific blend of copyright law anywhere in the world. So, for example, if I reverse engineer software owned by someone in the US and they've got a clickthrough license that says I can't do that, they can sue me, even though here in Australia we have laws that state that reverse engineering is legal under specific circumstances. That's one of the reasons why Australia was reluctant to sign the onto the Berne Convention, but they did it anyway. So, it's not the Berne Convention that says a copyright owner is free to terminate a license at any time, it's Australian law, and the Berne Convention permits Australian copyright owners to enforce that law in countries that don't have that law. If you think that's stupid, then I agree with you, it's fucked up, but no matter how stupid and far reaching copyright laws are, copyright owners seem to get whatever they want.

Re:dead copyright holder.

[Minna Kirai]

Look, read the freakin' license, it specifically says that if you violate the license your rights under the license are terminated.

That is true, but irrelevant, so I will explain slowly:

1. 1. I download Linux.
2. 
   2. I modify and restribute Linux. This would be a copyright violation, except that the license allows me to do so.
   3. I violate the GPL license in some way. My permission to distribute modified copies goes away.
   4. I stop violating the GPL license.
   5. I download Linux again. It comes with a new license.
   6. Since the new license I got hasn't been terminated, I proceed to distributed modified copies again.

Maybe you still think that because the first and second licenses I got are identical to each other (and to the widely-used GPL), that they are the same license. They aren't. Many kinds of licenses and contracts have identical boilerplate text, yet when they say "This License", they mean that license, not all other licenses sharing those terms. If the publisher sells you a second copy of her work with an identical license, you now have two licenses, and you can invalidate either of them by breaking its rules, while retaining the other.

That's it, there's no regaining the rights from someone else, the copyright holder has terminated your rights

True, if the copyright holder doesn't want me to provide me her software under the GPL anymore, she is free to not give me anymore copies. But as long as she (or anyone else) is providing new copies under the GPL, I can grab any of them and use it. (And obviously, its virtually impossible for the original author to stop 3rd parties from redistributing, since she gave them explicit permission to do so by released as GPL)

Even more so, countries like Australia (which is where I'm at BTW) specifically state that a copyright holder can terminate a license at any time and for any reason

If that were true, then Linus Torvalds would absolutely not accept patches from Australians. That would mean that an Australian who'd contributed a 5-line fix to Linux could point to any random person using that OS and demand him to power down and erase his bootloader. There would be no commercial use of Linux (or any software with an Australian contributor), because the risk that the programmer will spontaneously decide to revoke your license in the middle of operations is too large.

Re:dead copyright holder.

[Minna Kirai]

The Berne Convention is that pesky treaty that allows any country to enforce its specific blend of copyright law anywhere in the world.

No. The USA, at least, was forced* to re-write some of its copyright laws to "follow the international standard" before it could join Berne. At the same time, some European countries were made to change their laws to match the USA, with the net effect being that both regions adopted the most restrictive parts of both their laws.

* I shouldn't quite say "forced". Rather, lobbyists for the USA's publishing corporatations used "We've got to comply with the Berne standard" as an excuse to get stricter laws passed.

Re: I didn't find this comment in the COPYING file

[Ingolfke]

This clause was added in the 2.4.18 version of the file It did not exist prior to that. This seems that this would create a problem. If I submitted my work into the Linux kernel prior to 2.4.18 I would have submitted my work under the generic GPL and applied the license allowing for future versions of the GPL to be used. For Linus to make a blanket change to the licensing would violate the copyright holders intention and would be essentialy a violation of the GPL. Linus may have avoided this by securing certain rights from the copyright holder when the code was submitted. I doubt that happened. Also, any derivitive works that changed the licensing of the code to be more restrictive would seem to violate the copyright holders license.

The problem gets worse, I think, because if I submitted code after the 2.4.18 release I would have submitted my code under a license that restricted my code to be license under GPL v2 only. Which would mean there is code in the kernel that is licensed under two different sets of restrictions.

GPL v2.0, not any later version

[inc_x]

There are two reasons I only use GPL v2.0 and not any later version:
1) I don't want to license my software under terms I have never seen or read and over which I do not have any control.
2) I strongly suspect that the "or any later version" part is not legally enforcable towards the copyright holder because the copyright holder (in this case that's me) had no opportunity to review the terms of a later version when he put that line in and does not have any control over such later version. The clause would be void in most jurisdictions IMHO. (but IANAL)

Re:GPL v2.0, not any later version

[SLi]

I strongly suspect that the "or any later version" part is not legally enforcable towards the copyright holder because the copyright holder (in this case that's me) had no opportunity to review the terms of a later version when he put that line in and does not have any control over such later version. The clause would be void in most jurisdictions IMHO.

Well, it's not much different from giving it to the FSF under a BSD type license. Can't see what would be unenforceable about that (esp. since it's a license in question, not a contract, which would make a big difference)

Re:GPL v2.0, not any later version

[inc_x]

Under german law it could be unenforcable due to 31 Abs. 4 UrhG. And I recall a lawsuit in the US about whether publishers could publish copyrighted material in digital form that had been licensed to them previously (for publication in paper form). I think that was http://www.edwardsamuels.com/copyright/beyond/case s/tasini.html

I hope that ppl realize how wrong this is

[WindBourne]

First off, how many ppl code on projects that they think are wrong? Few if any. The only time that occurs is when there is a paycheck involved. Otherwise, why do it? So when ppl think that a single project will get done, well no. It just does not work that way.

How many projects have been converted into different languages just for the sake of doing a perl project in C? Far fewer than would be guessed by most. However, there are plenty of projects which started in perl-Tk/Python-TK/etc. and somebody saw it, and decided that they could do a better job with it in their language. Hey, that is the way ideas get shared.

As to re-inventing the wheel, Well Yes they are. And I say, keep it up. A good example is the automobile wheel. Automobiles used solid tires for eons, and then went to air tires (better ride). Now, they are about to go back to airless (new technology and ways of thinking about it). So, it is advantagous to do that. In fact in the OSS world, there are many great examples of re-invention helping out. Right now, Linux is a re-invention of Unix. Even as a long-time Unix hacker, I would say that Linux has caught up (and surpassed in many ways).

Another good example is NCSA web server vs. Apache. NCSA was resting on their laurels and did not really wish to allow others to play. So Apache forked and has taken over the web.

Then there is XFree vs. X.org. X.org IS run by old style unix folks who did not wish to play with the OSS world. So XFree was created and took on X.org. But along the way, XFree became just like a number of world leaders; We know what is best for you and we will dictate what you can and can not have and how you do it. SO X.org recently spun it off and is now attacking it in a truely open fashion. Somewhere down the road, the X.org will most likely try to pull a fast one and will try to move to a closed arch. like they had for ages (SCO, Sun, MS and others are in this group). When they do, then the current stuff will be forked again and it will continue all over.

Plain and simple, the ability to fork is good overall, and most forks only help the community.

How to make Linux not GNU/Linux

[dosius]

How to make Linux not GNU/Linux: For crying out loud, use the BSD userland!

Moll.

Re:How to make Linux not GNU/Linux

[SA Stevens]

Lots of us just use one of the BSD Unixes. However, all three of the main BSD 'forks' still use the GNU C Compiler.

Re:How to make Linux not GNU/Linux

[dosius]

Isn't OpenBSD trying to move to Tendra?

Moll.

correction

[Anonymous Coward]

That should say "then clearly I wouldn't be of the opinion that the new version is not similar in spirit to the present version"

Re:Pro-fork!!!

[lilo_booter]

I know that the parent troll is going to disappear into oblivion eventually, but I would just like to point out that forking have little to do with the fact that there are lots of implementations of a particular thing.

Ultimately, anyone wishing to contribute to the open source community can do it anyway they wish - they can join an established team, create a new project, fork an existing one or just start doing something from the ground up. Others pay for project development, others take the DIY approach.

But so what? How does this in anyway affect the usefulness of Open Source? The best will float into the public consciousness, many good ones will get overlooked and the worst will disappear into oblivion (similarly to the parent post I guess).

It's similar to saying to an amateur dramatic team that there's no point in making their own version of {insert name of play} when there are plenty of {films or tv productions} of the same work.

Or like saying to the professionals: Why remake Battlestar Galactica? Why make new episodes of Dr Who? Why remake the The Ladykillers?

The answer is always the same - the individuals involved believe they have something new to give. Don't like it? Ignore it and stick with what you like. Like it? Excellent.

So what if there are 1000 messengers or chat clients - no one is forcing you to use them.

Re:acces to source.

[leuk_he]

If they have box running the firmwire, they have the binaries.

GPL v2 only speaks about:"object code or executable" . Do you have the executable if you use a washing machine running linux? do you have the executable if you access a device of the LAn, do you have a executable if you run an application on a gpl webserver? Where exactly is the limit? (Note: shis is a real problem with gpl libraries.) The gpl is not clear about this, and a vendor will explain it his way. Do you want to sue a chinese vendor over something that is unclear to get the sources of your dvd player?

The gpl should be clear about this.

Stallman's master plan revealed

[Ingolfke]

With the Linux kernel locked into v2 of the GPL and exposed to international copyright and software patent attacks only code secured within the high walls and protective defenses of GPL v3 will be safe for FOSS developers. With the kernel sufficiently undermined developers and users alike will need to find a new, more protected, codebase from which to operate. That codebase is The Hurd.

Re:Stallman's master plan revealed

[Anonymous Coward]

Damn you, Stallman!

Re:acces to source.

[Anonymous Coward]

The medium is a little awkward, but yes, you have the object code. It's on a flash-chip inside the device. In principle there is no difference to object code on a CD or harddisk.

I trust Eben

[sl4shd0rk]

He's a smart shit. People don't like change - fear it rather. There are some valid concerns about license "portability" but I can't think of anyone else I would rather have sorting it all out.

codegraves on sourceforge

[N3wsByt3]

Well...the parent poster is being ironic, but at least partially right.

I mean, no one can deny, when even having a superficial look at the different projects that are on sourceforge, that an enormous amount of them are just plain dead, or whithering away. Exept for the really big projects - which have like, a treshold of minimum 3 developers (or people that at least keep busy themselves a bit with the code) and half a dozen 'helpers' - almost all the smaller projects really just sizzle out.

And then, some day, a new lonely coder gets up with the same idea, and he begins from scratch again, even though there are already myriads of dead projects that do the same. So, indeed, small projects keep being replicated, and, contrary to what one might exept, rarely is it working on top of an already existing (dead) project. Mostly they invent the wheel all over again, then they whizzle out (if they can't muster enough critical interest), and the whole process repeats itself.

The result is what you see on sourceforge: some big thriving projects, a lot of smaller almost-one-man projects that usually go completely dead real soon (you always have exeptions, ofcourse), and already massive amounts of complete stone-cold-graves of forgotten small projects. Which anyone hardly seem to notice even when they decide to do similar things.

It is rather mysterious how this is possible, seen the fact that FOSS projects are open to all. Why does there have to be 8 little projects that do in essence the same (but starve to death), instead that they all pull together and make one viable project? why do people reinvent the wheel, when there are so many basic (yet dead) projects they could use to build upon? Something is missing here...

I think, the answer has partly to do with ego's: ppl want it to be "their" project, and even if others are welcome to contribute, those that started with the project (especially if it are one-man-projects) like to feel it is and remains 'theirs'. So, *even* if they know there are other, similar projects, they will rather steal (well, in case of OSS it's just allowed use ;-) code from other projects and incorporate it in theirs, then just to join an already existing one.

But that doesn't explain it all, because not all coders are like that, and even those don't seem to be able to make efficient use of other works. The plain fact is, some do not really bother, or think it's to dificult to get to learn an already existing codebase (and simply prefer to start with one, so they know it well), and - more importantly - sourceforge sucks in finding projects that are similar to others, based on their internal code. Yes, sure you can search for generic terms on the application-level, but it's real hard to actually know what code could be useful or similar to some project you envisage.

In any case, it's very clear which curve the projects on sourceforge follow: a very large part of dead or near-dead small projects at one end, a certain amount of medium projects that never seem to amass the critical level but still keep hanging on, and then a few big projects that have 3 or more active developers, a buch of 'helpers' and a large userbase, which will thrive.

I'm not sure if all this is good or bad or 'normal', but I do think a system should be found to pull together all the working forces and/or code of (similar) small projects, so the chance of survival rises, there is less redundancy and reinventing the wheel and a critical mass can be more easily abtained. For that to happen, I fear sourceforge (and the likes) will have to become more efficient and just plain capable of letting people more easily recognise and bundle together similar projects in the first place.

Re:codegraves on sourceforge

[Spaceman40]

"I think, the answer has partly to do with ego's ... some do not really bother, or think it's to dificult to get to learn an already existing codebase ... sourceforge sucks in finding projects that are similar to others"

Another reason, especially here on Slashdot, is that some programmers just want to do something because they can. They host it on Sourceforge because it's a good, free hosting solution, and later lose interest in the project. I've done this several times, but I try to always notify Sourceforge that they can remove the project (save them some storage space, and allow someone else to use the project name).

I'm actually about to do this next week - I'm writing an assembler, about 70% of the code is done (it scans, it parses, it just doesn't "assemble" yet), and when I get to a working assembler, I'll start a project and put it up.

A lot of those dead projects are just people that start a project whenever they think of a cool idea, and then later lose interest or give up. (Or never get out of the planning stage...)

It's an interesting system, I'll give you that. :)

Re:codegraves on sourceforge

[eraserewind]

Because writing software to do something is a fun hobby you can do in your own time ragardless of whether a dozen people have done the exact same thing before. Why don't people just wait for the crossword answers tomorrow instead of doing it themselves today?

Project managing 8 developers to pull together to make something "viable" is a job, not a hobby. If you were doing that (or indeed working for such a person) you would usually have to be paid to do it -> not compatable with open source software, unless it falls into the narrow niche of being useful to, and fundable by, some business.

I agree that there are many projects on sourceforce that are dead or dying, but except for the fact that sourceforge has a crappy search engine, they are largely not hurting anybody, and "somebody" may find them useful at some stage. Not every piece of software has to be linux or apache.

a problem either way

[pgilman]

one of two possible scenarios exists here:

a) this new version of the gpl is going to cause a real problem, and there will be all sorts of forking and license incompatibility issues; or

b) this is a non-issue, but nevertheless the community now has to deal with clearing up a bunch of misconceptions, muddy water, and FUD.

either way, it's a problem. why does there even have to *be* a new version of the gpl? didn't they write it properly the first two times? personally, i admire the ethical intent of the gpl, but it's this kind of aggravation that makes the BSD license the only way to go.

Re:a problem either way

[Zachary Kessin]

because there are a number of issues that have surfaced over the last 13 years that were not covered in version 2. For example use in embeded systems of various forms. What are the rules if you are using linux to drive a Tivo or an elavator or something?

I suspect that it will end up being a big non issue.

Re:a problem either way

[SA Stevens]

What are the rules if you are using linux to drive a Tivo or an elavator or something?

The rules are that the litigants WILL be able to dig into the source and compile a list of people to sue, when the elevator crashes and kills a loved relation.

Believe me, in the current tort environment, the 'NO WARRANTY' section of the GPL might not suffice. If you've contributed to the kernel source tree, better keep your long-term savings in the form of Kruegerrands in a steel chest down in the celler next to your gun safe.

(only halfway tongue-in-cheek here)

Re:a problem either way

[Anonymous Coward]

You say that like the BSD licence has never been forked. By the way, to which BSD licence are we referring?

Re:a problem either way

[eraserewind]

I don't really agree. If linux is used in a controller for something, then at some point, some individual or company will have had to certify it's suitability for use in an elevator or whatever. It's that party (or the owner of the building using uncertified elevator parts) that carries responsibility.

GPL v3 and patents

[Danathar]

I'm curious as to how the GPL could help with the problem of software patents. Does anybody have information on what might go into the license that will address this problem? And how it will work?

Re:GPL v3 and patents

[ctr2sprt]

You could put in a clause stating that you grant an irrevocable, non-transferrable, royalty-free license to use any patents owned by the copyright holder in the source code. Probably "irrevocable" would need to be lawyered up, since the GPL permits you to use (but not distribute) a program even if you don't agree to the license. (So you would want to revoke the patent license in that case.)

Anyway, that would let open source developers use code contributed by IBM and other big IP holders free of worry about being sued some day. Even if IBM (for example) contributed code using patents it doesn't own (such as SCO), developers would have an extremely good defense. "IBM and its horde of patent lawyers told me I could use it." Continuing with my hypothetical situation, it would also apply to people who use the software, so long as they have agreed to the GPL, and so it would protect Linux users and not just developers.

That's my thinking. But neither I nor anyone else who has posted to this story has any clue what will really be in GPLv3, so... you know, don't be surprised if I turn out to be totally wrong.

Transitioning linux to GPL 3

[sicking]

I don't see that big of a risk that projects fork. The only thing that would require forking is if some devs wanted to go with GPL 3 and some wanted to stick with GPL 2. However I don't think the differences between them will be big enough that anyone would go to such drastic measures as forking.

What I am worried about though is how will large projects like the linux kernel transition to the new version? You'd have to hunt down all developers and request permission from each and every one of them. This seems like a next to impossible task for a project that's comming up on its 14th year birthday.

This isn't a problem for projects that either put "or later" in the license, or that transfer the copyright of all contributions to a central body. But neither seem to be the case with the kernel, or did I miss something?

Re:acces to source.

[leuk_he]

I agree.. But under this licence you may offer a washing maschine with the sourcecode in firmware for a fee. (It is a hell of a job to transfer the washing maschine, so the fee might be a little more than you think you would pay. ).

It is all in the wording.

Re:acces to source.

[penix1]

"Do you have the executable if you use a washing machine running linux?"

By definition, YES! Let me give you a hint...YOU CAN'T RUN A PROGRAM FROM SOURCE! Besides, you are using it not distributing it (see below)...

"do you have the executable if you access a device of the LAn,"

You are using the program not distributing it (see below)...

"do you have a executable if you run an application on a gpl webserver?"

You are using it not distributing it (see below)...

Using a program is not covered by the GPL and shouldn't be covered by it. Distributing is however. If you distribute binaries then according to the GPL you also have to distribute source. It is that simple. So unless you were the manufacturer of the washing machine, LAN device, and web application then the GPL only applies to you if you modify and DISTRIBUTE.

Hope that clears it up for you..

B.

Vaporware issues..

[iamsure]

If they'd stop talking about what it "will be" and "wont be", and put out a draft, they'd have *constructive* discussions instead of guessing!

Look, seriously, yes, there are grave concerns, and its a hideously important document. However, there is no reason why they cant put v2 into a wiki, add some proposed changes, and start working with the community on modifications.

This is at least the 10th story that has discussed A DISCUSSION OF WHAT WILL BE IN THE NEW VERSION!

Its not even 5 pages long. They've already mentioned the high points of areas they want to improve/change (patents, webservices), and everyone is well informed!

So just get on with it, and stop playing the vaporware game.

In the meantime, the only GPL-like license that actually closes the web services loophole (the Affero GPL), which is mentioned as a template for the GPLv3, ISNT GPL compatible!

It would be nice to have a GPLv2 compatible license that closes that loophole, so I'm waiting anxiously for a look at a license that will do it.

Enough talk - WRITE!

Re:Vaporware issues..

[slapout]

So why don't we just have GPL 2 and allow extensions to be written for it like Firefox? :-)

Re:Vaporware issues..

[Anonymous Coward]

It would be nice to have a GPLv2 compatible license that closes that loophole, so I'm waiting anxiously for a look at a license that will do it.

The GPLv2 explicitly prohibits imposing any additional restrictions. Adding a clause about using GPL software as a web application would be an additional restriction. Any modifiction that adds such a change must be incompatible GPLv2.

- krafty

Re:Vaporware issues..

[Rich0]

only GPL-like license that actually closes the web services loophole (the Affero GPL), which is mentioned as a template for the GPLv3, ISNT GPL compatible!

That really isn't a problem. If software is licensed under GPLv2 you just can't use the new license. If it is licensed as GPLv2 or later, then you can even though it is not compatible. Future versions of the software will be a mix of parts which are GPLv2 or later and GPLv3 or later.

The web services loophole won't be closed until all the GPLv2 code is replaced - otherwise you can just use GPLv2 to make your web service application...

Re:Vaporware issues..

[asuffield]

In the meantime, the only GPL-like license that actually closes the web services loophole (the Affero GPL), which is mentioned as a template for the GPLv3, ISNT GPL compatible!

As far as we know, a license that "closes the web services loophole" can neither be GPL compatible, nor a free software license. This "loophole" is necessary for a work to be free software; any license clause that attempts to "close" it will cause somebody to lose the ability to modify the work (exactly who depends on the clause). For example, the Affero GPL prohibits you from modifying and using the work if you do not have an HTTP server available (and many, many other cases).

Re:acces to source.

[Mycroft_VIII]

It seems simple in all but the remote execution case to me.
If you have the machine that runs the code, you have the binary, and thus the gpl requires the source be made available to you.

Mycroft

Ha.

[Anonymous Coward]

What I'm wondering is: How much closer is this going to get us to a common installer? Oh, and one that's easy to use.

That's not important though.

And you wonder why OSS hasn't taken off more.

Re:Ha.

[Luke-Jr]

I wonder if the lack of a common installer is related to the fact that none of the competition has even an attempt at an installer...
Windows? The most it does is display a list of programs that go to the effort to add themselves to that list.
Mac OS X seems to have a standard installer, but most programs are designed so as to not use it.

No matter how crappy the package manager is, be glad you have one at all.

Hey!

[The Spoonman]

What's wrong with spooning!? :)

Re:Hey!

[Drooling Iguana]

There's no such thing.

There is no spoon.

Re:Hey!

[Brandybuck]

Imagine RMS and AC spooning, and after you've scrubbed that image out of your brain with lye soap, you'll understand the gravity of the problem.

Mod parent up

[MemeRot]

It seems that Linux may already have an explicit conflict between version 2 and version 2.4.18. Any comments on the copying file overriding the submitters' intent?

Do not fear unpleasant forking action

[Oyvind Eik]

Instead, try to realize the truth -- there is no fork.

Re:acces to source.

[Anonymous Coward]

YOU CAN'T RUN A PROGRAM FROM SOURCE!

Sure you can. Interpreters have been around for ages. Some firmwares run Forth interpreters. Interpreters even exist for C.

I know that doesn't weaken your argument, but I hate to see a totally false statement in all caps.

Make it compatible with the CPL and the Apache 2.0

[srmq]

It would be wonderful (especially in the Java camp) if the new GPL was compatible with the CPL and the Apache 2.0 licenses. Apparently the FSF even agrees somewhat to the additional restrictions that these licenses make (see http://www.gnu.org/licenses/license-list.html ), so it would be nice to address the problem.

Being able to use Apache code and Eclipse with GPL projects would give a great boost to GPL Java projects.

Constant revising of the GPL is stupid and lame.

[Anonymous Coward]

Nevermind the fact these revisions happen every couple of years... Can we show a little more stable minded thinking rather then trying to prove how smart and fancy we are? It makes OpenSource look like it's run by a bunch of high school heavy metal leather wearing anarchists...

Re:acces to source.

[shaitand]

"Do you have the executable if you use a washing machine running linux? do you have the executable if you access a device of the LAn, do you have a executable if you run an application on a gpl webserver?"

YES. Just like you have the MS windows binaries if you bought a pc with them preloaded. Just because the software is preinstalled on a piece of hardware doesn't mean you do not have the software!

The binaries are not the installer and the GPL does not require you to be able to manipulate the binaries. If they sent you the device and that device has a binary image of gpl software installed on it, then that counts. Period. There is no limit. The GPL is very clear on this.

Re: I didn't find this comment in the COPYING file

[Tony Hoyle]

You can submit your code under GPLv2 or any later version... the project maintainer chooses to use GPLv2 and *only* GPLv2. That doesn't change the status of your code, only of the project that it is a part of.

If as seems likely GPLv3 is incompatible with GPLv2 (thereby forcing projects that use it to use it exclusively) the only issue would be submitting GPLv3 code to a GPLv2 only project, which would not be allowed.

Fork that

[Anonymous Coward]

Everyone should use the BSD license anyway. Unlike the GPL, BSD is not about controlling the world.

But wait GPL represents freedom.
Freedom, ha!! Free as in what.
Yes, more strings attached.

Now dead Linux contributors

[Anonymous Coward]

Yes, it would be hard to get Linux contributors who are now dead to agree to a change of license. On the other hand, they aren't likely to sue for copyright infringement if we just start treating their contribution as if it falls under the new license, are they?

You misstate RMS' goal

[swillden]

He has a particular purpose in mind, which is forcing everyone to release source to their software.

This statement is the cornerstone of your argument, and it's not true. RMS thinks it would be nice if everyone released source to their software, but his goal/purpose is explicitly *not* to force everyone to release source. Rather, his goal is to encourage the construction of a whole computing environment that is Free, so that people can choose to use it rather than being forced to use opaque, unmodifiable software.

That said, you are correct that your goal, which is, effectively, to advertise your skills and your code with the goal of being able to "monetize" it later, is not compatible with his. And also not compatible with the goals of most other Free Software developers, either. Most of them (I think it would be presumptuous to say "us", even though I've contributed to a few projects) just want to build the software we need and see it grow so that we can have the kind of software we want.

I am not sure intellectual property laws are beneficial (at all or beyond say 5 year duration),

I am. I think they're very important, and that it makes sense to protect IP for longer than five years, for some sorts of IP, anyway. There are many different things that fall under the IP umbrella, and what makes sense for each is different.

but even if people are allowed to copy binaries, I sure shouldn't be forced to give up my source.

As I've explained elsewhere, I think if you want copyright protection for your binaries, you should have to disclose source (which should also have copyright protection). If you choose not to disclose your source, you should not have copyright protection for your binaries, though you would still have it for your source and you would still be able to use contract and trade secret law to protect your binaries.

FSF needs a Constitution

[Anonymous Coward]

The Free Software Foundation needs a clear plan of succession and controls that make this scenario very unlikely.

It can include a permanent ban on any changes that are not "in the spirit" of GPL version 2. Then there can be disagreements and challenges over "controversial" changes (e.g., requiring public servers that run on GPL code to make their modifications to the code available), but obviously unreasonable changes (such as your scenario that MS and only MS is allowed to fork a proprietary version) would be found by a judge to not be in the spirit of GPL2 and therefore not allowed.

*sigh*

[Anonymous Coward]

You need remedial GPL 101. The GPL does not purport to cover software that is merely bundled on the same distribution medium. The other software has to "link" to or "derive" from the GPL'd software in some (controversial to describe exactly) way.

Not that big of a deal

[xenocide2]

Debian represents the group of developers most beholden to the FSF. Thats not entirely a bad thing, although there's always a set of people loudly pining for the affections of their chosen God, RMS.

Fortunately, this isn't that big of a deal. Most all of the developers who care so deeply about following RMS's every whim and order have already jumped on the Hurd bandwagon. In theory, the GPL should be compatible with itself, meaning that the kernel.org source could opt to distribute under GPL v3 without asking every developer. Either way, the Linux kernel team will probably spend a good deal of time examining and move very slowly on this, even after the public discussions are over and the final draft is made... final. v2 works well enough at persuading infringers to settle, so there's no hurry to change.

don't worry

[idlake]

You may not have much opportunity to practice it yourself, but, trust me, forking can be quite pleasant and good for you, actually.

Re:acces to source.

[Dwonis]

GPL v2 only speaks about:"object code or executable" . Do you have the executable if you use a washing machine running linux?

Do you think it is legal to take MS Windows, install it on the same washing machine, and sell a few million of these washing machines to the public without paying for Windows licenses?

If the GPLv2 only speaks about "object code or executables", and for some reason, the courts don't consider the programs on the washing machine to be object code or executables, then a distributor has no legal basis for distributing that washing machine at all.

As has been mentioned before, the GPL grants specific rights to people, provided that they meet certain conditions. Any distribution rights not granted by the GPL are implicitly reserved, and it would be a violation of most countries' copyright laws to exercise these ungranted rights.

Rights Assignment

[adavies42]

This is why rights assignment is a good idea in a large project. If there had been a linux foundation to which all kernel contributers had to assign their rights, this would be trivial to do. The FSF requires all rights in HURD code to be assigned back to the FSF, so they will never have this problem.

Hmm.... x.y?

[Anonymous Coward]

Think we'll have 3.0.0.1? :) (a little 2.6 humor)

Re:Pro-fork!!!

[MntlChaos]

How are all GPL versions automatically compatible with newer versions. What does this statement even mean?

If file 1 is licensed only with GPL v2 and file 2 is licensed only with GPL v3, unless GPL v3 says that you can redistribute under either the terms you received or a previous version, then you must distribute one file with one license and the other with the other license. Adding in "or any newer version" is dangerous. What if the FSF somehow gets taken over by Microsoft? Then MS can do whatever they want with your code.

Re:acces to source.

[arkanes]

Do you have the executable if you use a washing machine running linux?
Yes

do you have the executable if you access a device of the LAn
Yes

do you have a executable if you run an application on a gpl webserver?
No

The "web application" is considered by some to be a loophole in the GPL, personally I dont'.

The limits of the GPL are exactly as clear as copyright law is. Unfortunately, thats not very clear, but there's nothing the GPL can do to clear that up. The GPL applies if you're creating a derived work. If you work isn't sufficently derived so as to be covered under copyright law, then you can safely ignore the GPL. Knowing where that boundy is is hard, because (unfortunately) copyright law provides only the most minimal guidelines. The FSF posts what it thinks the limits are, and sticking to those is a pretty safe legal bet.

GPL3 SUX

[rice_burners_suck]

I don't see a problem here, as long as the GPL 3 solves enough of the current legal issues. (Hey, maybe there should be a Bugzilla database for the GPL license itself.)

Switching licenses shouldn't be a problem because the GPL notice in every source file states that you can use this (at your option) under version 2 or any later version of the GPL. Therefore, all GPLed software is already licensed under version 3, in a sense.

Besides, the GPL license should state that although an infinite number of developers might contribute to a project, the person who originally released that project under a specific name is the owner, and has the right to change license schemes at any time, and the new license takes effect upon the release of the first copy with that license. This would mean several things:

First, that, e.g., Linus could say, "Ok, Linux is going to be GPL 3 from now on." And no contributers or others who have their hand in Linux could complain about it, because Linus was the first to name and released Linux.

Second, that if Linus takes crack one day and decides to put Linux under, say, a Microsoft EULA, then all versions up to that version are still GPL (remember, you can't stop others from having the same rights as you had), so a fork could form at that time, where Linus can continue his work under a EULA version, and everybody else in the world can continue working on the GPL version.

By the way, this would require someone to change the name and release a version. Probably, there will be hundreds of versions released right away, with hundreds of names, but one or two of them will survive and become the new "Linux"...

This goes for all other projects using the GPL, not just Linux.

What a minute....

[deanj]

Why is everyone worried about this? We've been assured time after time when the "Let's GPL Java" topic comes up that forking the project won't hurt a darn thing.

Now, we have people up in arms about forks caused by a new version of the GPL itself..... Why is this a problem for GPL, but not for Java when we're talking about that?

SCO?

[Anonymous Coward]

Good luck getting SCO to sign off on that!

Meta-modding

[Anonymous Coward]

Meta-modded 'troll' mod down. Next time, think for more than two seconds before modding. Thanks.

Re:acces to source.

[jbolden]

You don't have to sue the chinese vendor. The American distributor is the one breaking US law. You sue him and let him handle the issue with the vendor.

How they plan to address Trusted Computing?

[Alsee]

They have stated that they are planning to address the Trusted Computing issue somehow in GPL3. I really really want to know what they have in mind. I've put a fair bit of thought into it and I can't see any way to deal with it.

In case anyone isn't aware of the problem, I'll explain. Trusted Computing defeats the GPL. It makes the source code usless. You can modify the software and it will run, but it will no longer work. If the code makes a call to a Trust chip then the chip will forbid any modified software from accessing any existing data files for that software. The software will run, but fail on file access. Something similar will end up happening with any attempt to interoperate with other software or to communicate over the internet. If you change so much as a single line then the Trust chip spying on your system will report that the software is incompatible. It will be rejected by other software, and it will be denied a network connection by other software.

The GPL may ensure you get a copy of the source, but the source does you no good.

I don't see how the GPL can be revised to fix this. Even forbiding the code to make a call to a Trust chip wouldn't do it. The operating system EXE loader can scan each program as it loads and make the Trust chip call. It's a little roundabout, but it can enforce the exact same problems. I'm stumped.

-

No

[Kjella]

I think you are missunderstanding what GPLv2 allows end users to do. (...) Same will go for the GPLv3 issue. If someone takes your code and makes a derivative of it they have to release it as a GPLv2. You can then put the code back into your project.

You've misunderstood the concept of multiple licenses. Take the Qt code for example, it is both QPL and GPL. It means that you can choose either QPL or GPL. If I make a GPL derivate, they can not put that code back in their QPL project.

GPLv2 and GPLv3 work as separate licenses. You can create a derivative GPLv2-only project, or a derivate GPLv3-only project. If you want to keep it dual-licensed, you would have to explicitly state so. However, this would be an additional restriction and in violation of the GPLv2.

Note that the way all such software accept contributions is by making you either sign over copyright or making you agree to both licenses. That way, they force you to fork your own project if you want to make a single-license version, but you are free to do so.

The FSF could try to resolve that by making a "GPLv2.5", which is code that can be used in both GPLv2 and GPLv3 projects, where the code itself must remain GPLv2 compatible (LGPL-style clause, but only for GPLv2) but where GPLv2.5 code does not "taint" a GPLv3 project with the conditions of the GPLv2. It would still require authors to relicense their code, but this allows for a smoother transition from GPLv2 to GPLv3.

Kjella

Re:No

[tricorn]

No. If something is currently v2-or-later, you can release a version that is v2-only (by adding code that is v2-only), v3-only (by adding code that is v3-only - the v2-or-later code is compatible), v3-or-later, or keep it all v2-or-later. The original code that was v2-or-later doesn't change, but the modified program you're distributing can be any of those. If you don't want that to happen, then you should release your code as v2-only from the start (which doesn't stop anyone from adding v2-or-later code, but does prevent any v3 code from being combined with your code).

GPL v2 and GPL v3 have to be incompatible, one way or the other or both ways. If they weren't, they'd be the same license. If v3 removes restrictions, then v2-only code is incompatible with it (since the essence of any version of the GPL is that additional restrictions aren't allowed - if it were to include an exception for v2-only code, for example, then it is no better than if you just continue to license under v2-or-later). If v3 adds restrictions, then it is incompatible with v2-only code. If it adds some restrictions and removes others, then they are mutually incompatible. Only v2-or-later code is compatible with v3-only or v3-or-later. Two licenses that both require there be "no other restrictions" must be mutually incompatible unless they have exactly the same restrictions. Only dual-licensing of at least one of them (which "or-later" is) can make them compatible.

may as well be PD

[DM9290]

Last thing: if you followed the FSF's recommendations as to how to use the license, your code would contain this text:

# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License as
# published by the Free Software Foundation; either version 2 of
# the License, or (at your option) any later version.

If you are going to allow undefined licensing of your work. Why not just make the work Public Domain and be done with it.

You dont know what text future GPL licenses may contain.