Modules that integrate non-GPL PHP apps violate the GPL.

[Angela Byron-3]

On 8-Sep-07, at 5:51 AM, Thomas Barregren wrote:

> Angela Byron skrev:
>>
>> IMO, the only thing we can do is exactly what Joomla! did:
>>
>> - Do not fork the GPL by creating our own interpretation of it  
>> (adding exceptions, etc.).
>
> We should definitely *not* fork GPL. That would be committing hara-
> kiri.
>
> But adding a "FOSS Exception" or "Linked Under Controlled Interface  
> Exception" is *not* forking. On the contrary! It is a proper way to  
> solve situations like the one we are discussing. The technique is  
> proposed and fully described with template and everything on FSF/
> GNU's web site:
>
>    * http://www.gnu.org/licenses/old-licenses/gpl-2.0- 
> faq.html#GPLIncompatibleLibs
>    * http://www.gnu.org/licenses/old-licenses/gpl-2.0- 
> faq.html#LinkingOverControlledInterface
>
>
> In fact, it *might* be necessary to add a FOSS Exception for the  
> PHP-license. Why? The PHP-license is not compatible with the GPL.  
> That itself doesn't prevent PHP-programs to be distributed under  
> GPL. But since Drupal *make use* of GD and other libraries  
> distributed under the PHP-license, it is possible to argue that  
> Drupal in part is a derivative of GD and the other libraries. If  
> this cannot be deemed to fall under the platform exception in GPL,  
> it is not possible to distribute Drupal under GPL without adding a  
> notice saying that it is okay. Again, adding a such notice is not  
> forking GPL.

Fair enough. It still feels a bit "dirty" to me, though -- the fact  
that GPL doesn't have these exceptions written into "out of the box"  
it is definitely "by design," not an oversight.

>
>> Adding exceptions anyway is a physically impossibility; you'll  
>> never find all of the copyright holders of Drupal to sign off on  
>> it, and many of us would oppose such an action.
>
> I am not completely convinced that it is a "physically  
> impossibility". After all, it is a limited number of people who  
> have committed code to the core. I suppose most of them are still  
> members of Drupal.org, and hence possible to get in contact with.  
> Why not try? Likely, a vast majority of all core contributors will  
> accept a "FOSS Exception" and possible also a "Linked Under  
> Controlled Interface Exception" for a "Module Programming  
> Interface" (e.g. hooks and some utility functions).

Interesting. So the act of committing code transfers authorship? My  
offering up code and saying, "Please commit this to core" is  
synonymous with "I hereby abandon all rights I have as the author of  
this code, and trust that the core committers will not someday do  
something silly with it?"

That's something I didn't know before. It was my understanding that  
copyright was retained by each individual who has contributed code to  
the project, regardless of who actually pulled the "commit" trigger.

> BTW, something similar has been done before, and in much larger  
> scale, namely BSD.

True. I don't argue that it _can't_ be done. I do argue that it might  
take much more time/resources than we want to spend on it. ;)  
Especially when we have a workaround (just boot the code from contrib  
that attempts to create a derivative work).

However, your point about Drupal making use of GD and such is a good  
one that bears further investigation.

Thanks a lot for your helpful contributions, Thomas. I've learned  
quite a bit from your interactions in this thread.

-Angie

[Thomas Barregren]

Karoly Negyesi skrev:

>> That alone makes the module a derived work of Drupal.
>>
>> But if your program and the other program are linked, even if it is done
>> only at runtime, and make function calls to each other and share data
>> structures, your module is a derivative work of the other program as well.
>>    
>
> AAAAAAAAH! Finally somebody explained this whole situation clearly.
>
> So if I am writing a bridge to run with Drupal that's GPL because of Drupal and because it's considered a derivative of some non-GPL code that means I am breaking some clause in the *other* code licence and we know there is such a clause *because* it is non GPL compatible.

Correct.

> So the issue looked cloudy because we know the bridge breaks some rule but it can not be told what without studying the other licence. This makes sense to my obsessed-with-math-logic mind. For the first time, this debacle makes sense even if not a pleasant one.
>
> However, there seems to be an escape! http://forum.joomla.org/index.php?topic=190545.0 this ends with "the LGPL would allow someone else to derive code based on it without any restriction you can make a LGPL bridge to buffer your proprietary extention (such as SMF) to a GPL program."
>  

There is no "escape". Since the module is a derived work of Drupal which
is licensed to you under GPL, you must license your module under GPL as
well. It is not allowed to use LGPL.

The discussion in the J! thread is about integrating a third-party
software which is licensed under LGPL. That is perfectly legal since
LGPL is compatible with GPL.

> So, if this holds up then we should allow LGPL code in our repo and ask the bridge module authors to LGPL their code. This is a ton easier than, say, changing the licence of Drupal core.
>  

Sorry, it doesn't hold up

However, if your module contains a large portion of code which is
independent of Drupal, you could release that part as a library with a
GPL compatible license, e.g. BSD, MIT or LGPL, and then use it with your
module. Your library can now be distributed on its own under any
license. But as part of your module, it will be distributed under GPL.

Best regards,
Thomas

[Thomas Barregren]

Chris Johnson skrev:

> On 9/7/07, Thomas Barregren <thomas@...> wrote:
>  
>> But if your program and the other program are
>> linked, even if it is done only at runtime, and make function calls to
>> each other and share data structures, your module is a derivative work
>> of the other program as well.
>>    
>
> If this is true and legally correct (and I understand it), then we are
> lucky the FSF doesn't have the money to sue thousands of cases in
> court.  Because every time a GPL program is compiled and run on a
> non-GPL operating system, it's going to be breaking this "rule."  No
> GPL software could be run on Mac OS or Windows, if it called the OS or
> any GUI API.

You comment only proves that you in fact have not read the license.  
Please do that before you comment.

Hint: § 3 of GPL v2.

Thomas

[Thomas Barregren]

ttw+drupal@... skrev:

> On 07.09-00:28, Thomas Barregren wrote:
> [ ... ]
>  
>> Assume you write a module for Drupal. Any meaningful module to Drupal
>> implements at least one hook. That alone makes the module a derived work
>> of Drupal.
>>    
>
> this is wrong.  it is an over zealous attempt to extend the legal
> coverage of copyright and licensing laws.  you may interface to
> software as you like, what you cannot do is copy and modify and
> distribute that software as your own.
>  

No, you are wrong.

If your code link to another code and call functions and share data
structures with that other code, your code becomes a derived work of the
other code.

It is however true that you don't have to worry about that if the other
code is licensed to you under an open source license and as long as you
don't distribute your code.

But i f you distribute your code, even if it is without the other code,
you must take the fact that your code is derived from the other code
into consideration. If the other code is under BSD, MIT or similar
license, it's easy: apply whichever license you like (almost). But if
the other code is under GPL, you must use GPL yourself.


> i believe this would be defeated in court as were proprietry software
> vendors when attempting the same trick.  i would also assert it is
> also an infrigement on the freedoms of free software and a direct
> contradiction of what free software is about.
>  

How can GPL infringe on the freedoms of free software? GPL was after all
developed by RMS and other to protect the freedom.

> i would re-state what an earlier poster said which is, it would be
> better to put together a list of copyright holders and correct
> current licensing discrepancies.

I agree that compiling a list of contributors, similar to the Linux
kernel, would be a very good thing to do (if feasible).

> then tackle how you wish to
> address this issue internally.  only then should you start
> propounding this sort of dictum.
>  

I have not propound any sort  of dictum. I have only tried to shed light
on something which great many people seems to be difficult.


Thomas

[Thomas Barregren]

Angela Byron skrev:

>
> On 8-Sep-07, at 5:51 AM, Thomas Barregren wrote:
>
>> Angela Byron skrev:
>>>
>>> IMO, the only thing we can do is exactly what Joomla! did:
>>>
>>> - Do not fork the GPL by creating our own interpretation of it
>>> (adding exceptions, etc.).
>>
>> We should definitely *not* fork GPL. That would be committing hara-kiri.
>>
>> But adding a "FOSS Exception" or "Linked Under Controlled Interface
>> Exception" is *not* forking. On the contrary! It is a proper way to
>> solve situations like the one we are discussing. The technique is
>> proposed and fully described with template and everything on
>> FSF/GNU's web site:
>>
>>    *
>> http://www.gnu.org/licenses/old-licenses/gpl-2.0-faq.html#GPLIncompatibleLibs 
>>
>>    *
>> http://www.gnu.org/licenses/old-licenses/gpl-2.0-faq.html#LinkingOverControlledInterface 
>>
>>
>>
>> In fact, it *might* be necessary to add a FOSS Exception for the
>> PHP-license. Why? The PHP-license is not compatible with the GPL.
>> That itself doesn't prevent PHP-programs to be distributed under GPL.
>> But since Drupal *make use* of GD and other libraries distributed
>> under the PHP-license, it is possible to argue that Drupal in part is
>> a derivative of GD and the other libraries. If this cannot be deemed
>> to fall under the platform exception in GPL, it is not possible to
>> distribute Drupal under GPL without adding a notice saying that it is
>> okay. Again, adding a such notice is not forking GPL.
>
> Fair enough. It still feels a bit "dirty" to me, though -- the fact
> that GPL doesn't have these exceptions written into "out of the box"
> it is definitely "by design," not an oversight.

You are definitely right. :-) RSM and FSF/GNU want everybody to use GPL.
But they are also realists, and therefore provide these means to use GPL
in conjunction with software not compatible with GPL.

>>> Adding exceptions anyway is a physically impossibility; you'll never
>>> find all of the copyright holders of Drupal to sign off on it, and
>>> many of us would oppose such an action.
>>
>> I am not completely convinced that it is a "physically
>> impossibility". After all, it is a limited number of people who have
>> committed code to the core. I suppose most of them are still members
>> of Drupal.org, and hence possible to get in contact with. Why not
>> try? Likely, a vast majority of all core contributors will accept a
>> "FOSS Exception" and possible also a "Linked Under Controlled
>> Interface Exception" for a "Module Programming Interface" (e.g. hooks
>> and some utility functions).
>
> Interesting. So the act of committing code transfers authorship?

No. I didn't said that.

> My offering up code and saying, "Please commit this to core" is
> synonymous with "I hereby abandon all rights I have as the author of
> this code, and trust that the core committers will not someday do
> something silly with it?"
>
> That's something I didn't know before. It was my understanding that
> copyright was retained by each individual who has contributed code to
> the project, regardless of who actually pulled the "commit" trigger.

It was not my intention to imply anything similar to what you say. Of
course the author of the code keep the copyright. I should have written
"contributed" instead of "committed" in my reply.

BTW, remember that not all contributions is "copyrightable". It must
have a certain level of creativity. For an example, a patch that just
correct a simple spelling error (or similar) is probably not protected
by the copyright laws. Furthermore, ideas are not protected by
copyright, only their expression as code.

>> BTW, something similar has been done before, and in much larger
>> scale, namely BSD.
>
> True. I don't argue that it _can't_ be done. I do argue that it might
> take much more time/resources than we want to spend on it. ;)
> Especially when we have a workaround (just boot the code from contrib
> that attempts to create a derivative work).

I agree that we first of all should boot code that links to code and
call functions and share data structures with another code which has a
licnse not compatible with GPL. But we should of course not boot code
that use another code using GPL compatible license (e.g. BSD, MIT and LGPL).

The second most important thing we can do is to use GPL v3 instead of
GPL v2. By doing so, many more licenses, most notable Apache, is
compatible with GPL.

But somewhere down the road I think we really should try to track
contributors very much like how Linux kernel does.


> However, your point about Drupal making use of GD and such is a good
> one that bears further investigation

That is one of the things I suggest Drupal Association bring up with a
lawyer.

> Thanks a lot for your helpful contributions, Thomas. I've learned
> quite a bit from your interactions in this thread.

I am glad to be any help. :-)


Best regards,
Thomas

[Earnie]

Quoting Angela Byron <drupal-devel@...>:

>
> Fair enough. It still feels a bit "dirty" to me, though -- the fact  
> that GPL doesn't have these exceptions written into "out of the box"  
> it is definitely "by design," not an oversight.
>

But even FSF adds exceptions to the GPL at times; e.g. libstdc++.

Earnie -- http://for-my-kids.com/
-- http://give-me-an-offer.com/

[Thomas Barregren]

Gerhard Killesreiter skrev:
>> Interesting. So the act of committing code transfers authorship? My
>>    
>
> No, no, no, and no. This does not happen. I think Thomas doesn't know
> how Drupal development works, which made him use the term "core committer".
>  

Well, I think I know how Drupal development works. :-)

As I answered Angie, I should have written "contributed" instead of
"committed" in my reply.

Best regards,
Thomas

[Earnie]

Quoting Gerhard Killesreiter <gerhard@...>:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Angela Byron schrieb:
>>
>> On 8-Sep-07, at 5:51 AM, Thomas Barregren wrote:
>>
>
>>> I am not completely convinced that it is a "physically impossibility".
>>> After all, it is a limited number of people who have committed code to
>>> the core. I suppose most of them are still members of Drupal.org, and
>>> hence possible to get in contact with. Why not try? Likely, a vast
>>> majority of all core contributors will accept a "FOSS Exception" and
>>> possible also a "Linked Under Controlled Interface Exception" for a
>>> "Module Programming Interface" (e.g. hooks and some utility functions).
>>
>> Interesting. So the act of committing code transfers authorship? My
>
> No, no, no, and no. This does not happen. I think Thomas doesn't know
> how Drupal development works, which made him use the term "core committer".
>

Where did Thomas you "core committer"?  I see "core contributer".  And
while authorship doesn't transfer copyright can but for Drupal thus far
copyright assignment hasn't been requested.  And yet while both
authorship and copyright have been retained by you; there is no
document to prove it within the source itself.  There is the ticket but
there is not even a ChangeLog entry to indicate the changes.  This is
in itself a violation of the GPL as clearly stated in article 2a.

> I hereby state that for none of the patches not written by me that I've
> committed to core or contrib I accept transfer of copyright or
> authorship (the latter is actually impossible under German law) of the code.
>

That is colorfully stated but no one has asked for transfer of ownership.

>> offering up code and saying, "Please commit this to core" is synonymous
>> with "I hereby abandon all rights I have as the author of this code, and
>> trust that the core committers will not someday do something silly with
>> it?"
>>
>> That's something I didn't know before. It was my understanding that
>> copyright was retained by each individual who has contributed code to
>> the project, regardless of who actually pulled the "commit" trigger.
>
> That is my understanding too.
>

The only means by which you can transfer your copyright is by legal
terms spelled out in writing and then signed by the original owners of
the code.  There can be stipulations that state that copyright is
dually maintained so that both parties own a copyright.

Earnie -- http://for-my-kids.com/
-- http://give-me-an-offer.com/

[David Metzler]

On Sep 8, 2007, at 9:11 AM, Thomas Barregren wrote:

> ttw+drupal@... skrev:
>> On 07.09-00:28, Thomas Barregren wrote:
>> [ ... ]
>>
>>> Assume you write a module for Drupal. Any meaningful module to  
>>> Drupal implements at least one hook. That alone makes the module  
>>> a derived work of Drupal.
>>>
>>
>> this is wrong.  it is an over zealous attempt to extend the legal
>> coverage of copyright and licensing laws.  you may interface to
>> software as you like, what you cannot do is copy and modify and
>> distribute that software as your own.
>>
>
> No, you are wrong.
>
> If your code link to another code and call functions and share data  
> structures with that other code, your code becomes a derived work  
> of the other code.

You keep insisting this, but I don't agree its that cut and dry.  
Current literature on the subject suggests otherwise when calling  
code that is "designed as a library".  Regardless as to whether it is  
statically linked or otherwise.  Creating a derivative work is not a  
technical definition, it's a legal one.  Most of the reading I've  
done outside of the GPL FAQ agrees with this.

Drupal's strength IMHO relies on its community, and defining the  
risks we're willing to take in terms of technical semantics feels  
like a step in the wrong direction.  I agree, that the SMF bridge  
case takes it too far.  Distributing non-gpl code using the drupal  
infrastructure would be a step backward for us also.  I think we  
should continue to feel free to distribute code that says it's GPL,  
no matter what, but not try and second guess lawyers on what is  
considered a derived work or not.

If the drupal foundation decides that it want's to go after people  
who might bundle and distribute drupal core with other non-gpl code,  
I'd probably be supportive of that.  I would not be supportive of the  
foundation going after contributers who develop modules to integrate  
drupal with proprietary systems and release those modules as GPL,  
whether we think they're wrong or not.

And with that, I think I'm done with this thread.....

[Peter Wolanin]

Can I again suggest that the most pragmatic change would be to license
core under "GPL 2.0 or any later version".    Moving to only 3.0 would
be a mistake, since there are contrib modules and other open source
projects that may stay as GPL 2.0 only.

-Peter

On 9/8/07, Thomas Barregren <thomas@...> wrote:
...

> I agree that we first of all should boot code that links to code and
> call functions and share data structures with another code which has a
> licnse not compatible with GPL. But we should of course not boot code
> that use another code using GPL compatible license (e.g. BSD, MIT and LGPL).
>
> The second most important thing we can do is to use GPL v3 instead of
> GPL v2. By doing so, many more licenses, most notable Apache, is
> compatible with GPL.
>

...

> Best regards,
> Thomas
>

[Karoly Negyesi]

> The second most important thing we can do is to use GPL v3 instead of
> GPL v2.

There is a slight problem with this. I do not like GPL v3. So, you do not need to chase the hundreds of core contributors, because I won't sign off such a change.

[Thomas Barregren]

Karoly Negyesi skrev:
>> The second most important thing we can do is to use GPL v3 instead of
>> GPL v2.
>>    
>
> There is a slight problem with this. I do not like GPL v3. So, you do not need to chase the hundreds of core contributors, because I won't sign off such a change.

Karoly,

I would be very interesting to here your reasons for not liking GPL v3.

Best regards,
Thomas

[ttw+drupal]

On 07.09-15:34, Earnie Boyd wrote:
[ ... ]

> >>Assume you write a module for Drupal. Any meaningful module to Drupal
> >>implements at least one hook. That alone makes the module a derived work
> >>of Drupal.
> >
> >this is wrong.  it is an over zealous attempt to extend the legal
> >coverage of copyright and licensing laws.  you may interface to
> >software as you like, what you cannot do is copy and modify and
> >distribute that software as your own.
> >
>
> I don't see how Thomas is wrong.  What he says is the heart of the GPL.

because the GPL is a copyright license and this statement is in
direct conflict with copyright law and, as such, is a zealous
attempt to extend the coverage of copyright.  for example, if your
write a book and i quote your work (for any wide range of reasons)
you cannot sue me under copyright.  this is the comparitive legal
situation.

i would much rather assert a notion of freedom than discuss legal
aspects of law.  what is relevant on this list is what this
community choses to enforce - let the legal aspects sort themselves
out after that.

[ttw+drupal]

On 07.09-15:34, Earnie Boyd wrote:
[ ... ]
> The copyright is owned by the producer of the code and the license
> allows the copyright holder to give you the right to use it. [ ... ]

giving someone freedoms to use your code does not allow you to
dictate basic freedoms to them .. i.e. you cannot restrict the basic
rights given to them in copyright law just because you are giving
them more access than copyright law dictates.

[ttw+drupal]

On 07.09-15:32, Larry Garfield wrote:
[ ... ]
> (This email, by the way, is copyright 2007 Larry Garfield.  Forwarding
> it, including quoting it in a reply, counts as redistribution and is
> illegal under modern US copyright law without license.  License to
> redistribute this email and its contents are hereby granted to all
> readers under the GNU GPL.  That does not in any way shape or form
> diminish the copyright claim of Larry Garfield upon the work herein.
> Welcome to modern copyright law.)

:-D
appreciate the humor but, actually, it's a good example.  you can't
do that.  under copyright i may use your material, reply to it and
comment on it even use it in satire or parody and there's not a
thing you can do about it .. well you can sue me but you'd lose (i
hope, or i just got myself in a world of #@$% ;-)

[Earnie]

Quoting Karoly Negyesi <karoly@...>:

>> The second most important thing we can do is to use GPL v3 instead of
>> GPL v2.
>
> There is a slight problem with this. I do not like GPL v3. So, you do
> not need to chase the hundreds of core contributors, because I won't
> sign off such a change.
>
>

IMO, you already have because of the "any future version" clause.  So
if Drupal.org decides to move to GPL v3 they do not need to contact
anyone for permission as permission has already been given.

Earnie -- http://for-my-kids.com/
-- http://give-me-an-offer.com/

[Earnie]

Quoting Peter Wolanin <pwolanin@...>:

> Can I again suggest that the most pragmatic change would be to license
> core under "GPL 2.0 or any later version".    Moving to only 3.0 would
> be a mistake, since there are contrib modules and other open source
> projects that may stay as GPL 2.0 only.
>

This is resolved by changing to the GPL v3 version with a major
release.  Then modules must update with along with support for the
newest version of Drupal.

Earnie -- http://for-my-kids.com/
-- http://give-me-an-offer.com/

[Karoly Negyesi]

> IMO, you already have because of the "any future version" clause.  

Which can be found exactly where..? Drupal tarball is distributed with the GPL v2, the contributions FAQ link to the GPL v2 and however hard I seek I can't find a single mention of that.

[Angela Byron-3]

On 9-Sep-07, at 10:42 AM, Karoly Negyesi wrote:

>> IMO, you already have because of the "any future version" clause.
>
> Which can be found exactly where..? Drupal tarball is distributed  
> with the GPL v2, the contributions FAQ link to the GPL v2 and  
> however hard I seek I can't find a single mention of that.

It's written in LICENSE.txt:

"    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***.
"

[Darren Oh-2]

Further simplification of the claims being made now appears possible.  
There are two:

1. You cannot use a GPL-incompatible license for an app if it calls  
functions in Drupal or in a Drupal contributed module. This claim is  
not disputed.

2. You cannot use the GPL for code that calls functions in a third-
party app if you cannot distribute your code and the third-party app  
together as a single work using the GPL. This claim is disputed.

On Sep 7, 2007, at 6:28 AM, Thomas Barregren wrote:

> Since you already have accepted to distribute your module under GPL  
> and nothing by GPL, the license of the other program must allow its  
> derivative works to be licensed under GPL. A license which allows  
> that is said to be compatible with GPL. Examples of compatible  
> licenses include the revised BSD and the MIT licenses. If the other  
> program's license is not compatible with GPL, then you cannot  
> distribute your module. It is as simple as that.