Darren New escreveu:
> nemesis wrote:
>>> Firstly, you understand that by giving the FSF the copyright, you've 
>>> given them permission to prevent you from releasing your own code 
>>> with a different license, right?
>>
>> Yes, and I'm pretty sure all GPL contributors willing to do so, 
>> including commercial entities, are well aware of it.
>>
>>> Copyright is the right to restrict others from copying your work. 
>>> It's not the right to copy it, but to prevent others from doing so.
>>
>> [...]   The copyright holder
>> of the original GPL'd code is still the same,
> 
> Not if you've given the copyright to the FSF.

It was exactly about the FSF I was talking about, following from the gcc 
case.

>> but what does this mean when he explicitely grants others the right to 
>> use, copy and modify it?
> 
> That he can prevent you from copying it.

No, he can't because he explicitely stated otherwise by abiding by the 
GPL terms.

He can, as copyright owner, license it all under another license and 
begin anew, even closed source, from there on with worthy improvements 
to the code, just as some commercial entity can take away a MIT project 
and do the same.  The original GPLed code can't be closed by anyone, 
like MIT code.

>> The only power left to FSF is if they suddenly change from a 
>> foundation to a corporation and license the body of GNU software under 
>> a restrictive commercial license 
> 
> Yep. Not that I'm saying they will. I'm just pointing out that the 
> assertion that the original author can always dual-license the code is 
> incorrect if you've turned the copyright over to the FSF.

The majority of GPL'd code out there is not GNU and copyrights are 
retained by the original authors, which is what really fuels GPL 
projects, since everyone benefits from collaborative development and 
still retain their copyrights in face of leechers.

If you're willing to donate code patches to the FSF though, you are well 
aware of the consequences and are willing to take the risk perhaps 
because you think it'll be of benefit for everyone, including yourself 
(like, others can take the burden of maintaining the software, even your 
changes).  Even though Stallman may end up being nothing more than a 
evil leecher and a hypocrite.  I believe it's more likely though that 
he'll be assaulted by ninjas...

Post a reply to this message

nemesis wrote:
>> The brainwashing in this case is to make people think that if I 
>> improve your code, then it's still your code to control.
> 
> I don't think I get it what you mean.

You conflate "my code, with your code added" to be the same as "my code". 
The GPL isn't "more free" than the MIT license. It's just letting you 
exercise different control over the code you wrote yourself. You can prevent 
me from distributing your code in ways you don't like, but that's true of 
every license.

The thing that's "getting ridiculous" is when people go out of their way to 
apply the GPL to *your* code when it *isn't* a derivative of any GPLed code. 
See how that works?

-- 
   Darren New, San Diego CA, USA (PST)
   "Ouch ouch ouch!"
   "What's wrong? Noodles too hot?"
   "No, I have Chopstick Tunnel Syndrome."

Post a reply to this message

nemesis wrote:
> Darren New escreveu:
>>  The FSF is looking for technological mechanisms
>> to force you to GPL the code that you wrote entirely from scratch 
>> yourself without modifying or incorporating any other GPL code. 
>> They're not respecting my wishes about how I want to distribute *my* 
>> original code
> 
> Your gcc plugin code would only exist in the first place as an extension 
> for gcc, so without gcc, there would be no plugin of yours.

This is not true. Someone else may write a gcc-compatible compiler. Or it 
may be a large codebase that works for any compiler, and I want it to work 
for gcc also.

In any case, that's irrelevant. If gcc plug-ins fell under the GPL, they 
wouldn't need technological mechanisms to prevent them from working when 
they aren't GPLed, would they? The fact that it interoperates with gcc 
doesn't mean they're respecting my wishes for how I want to license my code, 
does it?

> So, why 
> would you even write a plugin for gcc in the first place?

And my Linux application wouldn't exist if I didn't have Linux to run it 
under. Does that mean my application is GPLed? My apache module wouldn't 
exist if apache wasn't around. Does that mean my module is GPLed? My PSQL 
stored procedures wouldn't exist if PSQL wasn't around. That doesn't make my 
stored procedures GPLed.

If every Linux application you wrote was required to use the GPL, would you 
find software companies like Adobe porting their software to it?

> If the GPL alone would be able to restrict it in this way, you wouldn't 
> enjoy it either, so what motivations would you have to submit any code 
> at all to help promote and spread a project purpoting the evil GPL license?

I'm not sure what you're trying to say.

You can certainly license code with a license that says "if you use this 
compiler, everything you compile with it belongs to us."  As you say, people 
don't do that, because people wouldn't use the compiler if they did.

I find it interesting that in the areas where FSF *is* the leader, they wind 
up acting like monopolists.

-- 
   Darren New, San Diego CA, USA (PST)
   "Ouch ouch ouch!"
   "What's wrong? Noodles too hot?"
   "No, I have Chopstick Tunnel Syndrome."

Post a reply to this message

nemesis wrote:
>> Not if you've given the copyright to the FSF.
> It was exactly about the FSF I was talking about, following from the gcc 
> case.

I was referring to the concept that "the original author can always 
dual-license it."  That isn't true if the copyright has been given to FSF.

> The original GPLed code can't be closed by anyone, like MIT code.

Original MIT code can't be closed by anyone either.

> The majority of GPL'd code out there is not GNU and copyrights are 
> retained by the original authors, which is what really fuels GPL 
> projects, since everyone benefits from collaborative development and 
> still retain their copyrights in face of leechers.
> 
> If you're willing to donate code patches to the FSF though, you are well 
> aware of the consequences and are willing to take the risk perhaps 
> because you think it'll be of benefit for everyone, including yourself 
> (like, others can take the burden of maintaining the software, even your 
> changes).  Even though Stallman may end up being nothing more than a 
> evil leecher and a hypocrite.  I believe it's more likely though that 
> he'll be assaulted by ninjas...

I agree with all of this. My disapproval is of the attempts to take code 
from people that doesn't fall under the GPL and doesn't contain any code of 
a GPLed project, and force them to release it under the GPL.

-- 
   Darren New, San Diego CA, USA (PST)
   "Ouch ouch ouch!"
   "What's wrong? Noodles too hot?"
   "No, I have Chopstick Tunnel Syndrome."

Post a reply to this message

On Wed, 28 Jan 2009 12:46:12 -0800, Darren New wrote:

> Jim Henderson wrote:
>> On Wed, 28 Jan 2009 10:22:27 -0800, Darren New wrote:
>> 
>>> Jim Henderson wrote:
>>>>>   This makes GPL incompatible with all other licenses. This is
>>>>>   rather
>>>>> restricting.
>>>> But protective of the rights of the original author, which is the
>>>> underlying goal.
>>> Which original author? The one giving away the code with more
>>> restrictions, or the one giving away the code with fewer restrictions?
>> 
>> The one who wrote the code that you are modifying and then
>> distributing.
> 
> What makes you think I'm modifying the GPL code?
> 
> If I include libjpg without modifying it at all, I can no longer use the
> MIT license on my code, even if my code is orders of magnitude larger
> and more valuable. This forces me, while building something large that I
> need to get paid for, to rewrite code that's already tested and works
> well and is bug-free and safe. Hence, GPL code leads to bugs in non-GPL
> code.

You can link to GPL libraries.  There are plenty of examples of this.

What you cannot do is incorporate GPL-licensed code into non-GPL licensed 
code.

There is, AFAIK, one single exception - if you own the code, you can dual-
license it.  There are also plenty of examples of this.

There's nothing that says you can't get paid for writing GPL code.  There 
are also also plenty of examples of this.

But if you're building something large and you want to leverage code 
others have written, you have to respect their license terms (whatever 
that license is, not just if it's GPL).

Jim

Post a reply to this message

On Wed, 28 Jan 2009 18:46:27 -0800, Darren New wrote:

> So you said "The GPL only affects those who modify the code." This is
> factually wrong, as evidenced by my example.

No, it's not factually wrong.

I just purchased X-Plane 9.  It uses openAL, openGL and a slew of other 
libraries.

X-Plane is a proprietary software package that links to these libraries 
which IIRC are licensed under the GPL.

If Austin (the author of X-Plane) had incorporated the libraries into his 
binary by compiling the code as part, then yes, he would've had to 
license his whole software package under the GPL.  He didn't and he's 
within the requirements of using GPL libraries.

The argument about "linking means you have to release your code under the 
GPL" has been disproven time and time again.

Jim

Post a reply to this message

Darren New escreveu:
> I was referring to the concept that "the original author can always 
> dual-license it."  That isn't true if the copyright has been given to FSF.

True.  Perhaps I should've written "the copyright owner can always 
dual-license it", but that would lead to further ambiguity, like in the 
case of someone submitting code to a large project only to find out it 
really wouldn't matter to dual code his lone little patch.  What sense 
does it make to release it in, say, a closed-source license, something 
that is but a little part of a far larger GPL'd whole and which wouldn't 
exist independently of it?

>> The original GPLed code can't be closed by anyone, like MIT code.
> 
> Original MIT code can't be closed by anyone either.

I thought I said that in that very sentence, but now reading it looks 
ambiguous, yes.  Sorry.

> I agree with all of this. My disapproval is of the attempts to take code 
> from people that doesn't fall under the GPL and doesn't contain any code 
> of a GPLed project, and force them to release it under the GPL.

It doesn't force anyone any more than the GPL forces everyone to use Linux.

Post a reply to this message

Jim Henderson escreveu:
> On Wed, 28 Jan 2009 18:46:27 -0800, Darren New wrote:
> 
>> So you said "The GPL only affects those who modify the code." This is
>> factually wrong, as evidenced by my example.
> 
> No, it's not factually wrong.
> 
> I just purchased X-Plane 9.  It uses openAL, openGL and a slew of other 
> libraries.
> 
> X-Plane is a proprietary software package that links to these libraries 
> which IIRC are licensed under the GPL.
> 
> If Austin (the author of X-Plane) had incorporated the libraries into his 
> binary by compiling the code as part, then yes, he would've had to 
> license his whole software package under the GPL.  He didn't and he's 
> within the requirements of using GPL libraries.
> 
> The argument about "linking means you have to release your code under the 
> GPL" has been disproven time and time again.

Hmm, this sounds weird.  You mean dynamic linking an app against a GPL 
lib doesn't require one to GPL the app code as well?  If so, what's the 
point of the LGPL?

And I didn't know OpenGL was GPL'd.  In fact, it's a specification only. 
  Are you talking about the mesa lib?

Post a reply to this message

Jim Henderson wrote:
> You can link to GPL libraries.  There are plenty of examples of this.

Then what's the LGPL for?

Note that OpenAL is apparently LGPL, so it specifically says you can link to 
it without invoking the GPL.

> What you cannot do is incorporate GPL-licensed code into non-GPL licensed 
> code.

Define "incorporate"? Look at the first paragraph of
http://www.fsf.org/licensing/licenses/why-not-lgpl.html

It doesn't sound like the FSF agrees with you.

> There is, AFAIK, one single exception - if you own the code, you can dual-
> license it. 

Unless it's a plug-in for GCC.

> But if you're building something large and you want to leverage code 
> others have written, you have to respect their license terms (whatever 
> that license is, not just if it's GPL).

Agreed. I'm not arguing the GPL is a bad thing, even. I'm simply pointing 
out that a lot of the slogans promoted by FSF supporters are wrong.

The gcc changes aren't aimed at making sure the plug-ins are "free 
software". They're aimed at making sure the plug-ins are "copyleft".

-- 
   Darren New, San Diego CA, USA (PST)
   "Ouch ouch ouch!"
   "What's wrong? Noodles too hot?"
   "No, I have Chopstick Tunnel Syndrome."

Post a reply to this message

nemesis wrote:
> True.  Perhaps I should've written "the copyright owner can always 
> dual-license it", but that would lead to further ambiguity, like in the 
> case of someone submitting code to a large project only to find out it 
> really wouldn't matter to dual code his lone little patch.  What sense 
> does it make to release it in, say, a closed-source license, something 
> that is but a little part of a far larger GPL'd whole and which wouldn't 
> exist independently of it?

It doesn't. I don't think anyone would complain that a patch to the 
internals of the Linux kernel won't get incorporated unless they too are GPLed.

It's major functionalities, like, say, plug-ins, that people might want to 
dual-license.

>>> The original GPLed code can't be closed by anyone, like MIT code.
>>
>> Original MIT code can't be closed by anyone either.
> 
> I thought I said that in that very sentence, but now reading it looks 
> ambiguous, yes.  Sorry.

I see.

>> I agree with all of this. My disapproval is of the attempts to take 
>> code from people that doesn't fall under the GPL and doesn't contain 
>> any code of a GPLed project, and force them to release it under the GPL.
> 
> It doesn't force anyone any more than the GPL forces everyone to use Linux.

No, the GPL doesn't. But that's exactly what the FSF is trying to do for gcc 
plug-ins. Remember where the thread started? Did you read the article that 
was linked? The article was specifically "how can the FSF force gcc plug-ins 
to be released as GPL even if they don't incorporate any GPLed code?"

-- 
   Darren New, San Diego CA, USA (PST)
   "Ouch ouch ouch!"
   "What's wrong? Noodles too hot?"
   "No, I have Chopstick Tunnel Syndrome."

Post a reply to this message