how much right do I have on my project, if there are patches by others?

[Matthew Flaschen]

Rick Moen wrote:
> Quoting Matthew Flaschen (matthew.flaschen@...):
>
>> It would seem that any interpretation except a series of independent
>> (i.e. neither joint nor collective) derivative works...
>
> [USA law assumed for present discussion.]
>
> A third-party patch in _separate_ form is indeed logically treated as
> commentary; thus, as an independent work.

What if a significant amount of code is removed in the patch?  Could
that exceed the limits of fair use?

  Merging it into the original
> work strikes me as very likely to create a derivative work -- which then
> necessarily falls into either the joint- or collective-work category.

Is this really correct?  If a derivative work wasn't coordinated at all
with the original, how could it be a joint or collective work?  Also,
some derivative works are illegal.  How does that fit in?

> (not that copyright law has a mandate to bolster copyleft).

True.

>> If programs are collective works, the primary author can
>> relicense contributions (without explicit copyright assignment) to a
>> copyleft program under a proprietary license.
>
> If contributors prove that ye olde primary author has failed to
> safeguard their interests or has violated agreements with them, then
> nonetheless the primary author may find he/she lacks that option.
> Please note that courts tend to measure "interests" for civil-law
> purposes in economic terms:

So if a contributor uses copyleft for non-pecuniary reasons, they have
no recourse, right?

> But, anyhow, in that hypothetical, any party (obviously) can fork
> rev. n-1, taking over maintenance under copyleft.

But that's hardly enough, if the primary author has created a
proprietary fork against a major contributor's wishes.

Matt Flaschen

[Rick Moen]

Quoting Matthew Flaschen (matthew.flaschen@...):

> What if a significant amount of code is removed in the patch?  Could
> that exceed the limits of fair use?

These questions seem to indicate some serious confusion, since they
discuss an issue that has nothing to do with whether one work is
derivative of another in copyright law.  If you wish to better
understand how that concept is defined in USA copyright law, your best
starting point is the CAI v. Altai decision.  

> > Merging it into the original work strikes me as very likely to
> > create a derivative work -- which then necessarily falls into either
> > the joint- or collective-work category.
>
> Is this really correct?  

Attempting argumentum ad ignorantiam again, Matthew?  Your lack of
understanding of basic legal concepts really isn't my problem.

> If a derivative work wasn't coordinated at all with the original,
> how could it be a joint or collective work?

"Coordinated with" appears to have no meaning in copyright law,
so your question is not answerable as stated.  That aside, it is simple
logic that a collective (not "derivative") work _must_ be either joint
or collective, as those are mutually exclusive and exhaustive conceptual
subcategories.

> Also, some derivative works are illegal.  How does that fit in?

How can you _not_ understand that the creation and distribution of some
derivative works is tortious (not necessarily "illegal"), on account of
copyright violation?  And what on _earth_ does this question have to do
with the preceding thread?

Again, I don't know exactly what your problem is, here, but I'm unclear
on why it's suddently become my job to teach you.


> > If contributors prove that ye olde primary author has failed to
> > safeguard their interests or has violated agreements with them, then
> > nonetheless the primary author may find he/she lacks that option.
> > Please note that courts tend to measure "interests" for civil-law
> > purposes in economic terms:
>
> So if a contributor uses copyleft for non-pecuniary reasons, they have
> no recourse, right?

I said nothing like that.

Whether one party has committed a tort towards another depends on the
particular circumstances and the duties that person has _in_ those
circumstances.  I cited _one_ theory of law that could apply.  Many
others could, also.


> > But, anyhow, in that hypothetical, any party (obviously) can fork
> > rev. n-1, taking over maintenance under copyleft.
>
> But that's hardly enough, if the primary author has created a
> proprietary fork against a major contributor's wishes.

"Enough" for what?  Facts do not change just because you don't approve
of them, Matthew.  See .signature block, please.

--
Cheers,                                      "Reality is not optional."
Rick Moen                                             -- Thomas Sowell
rick@...

[Rick Moen]

I wrote:

> "Coordinated with" appears to have no meaning in copyright law,
> so your question is not answerable as stated.  That aside, it is simple
> logic that a collective (not "derivative") work _must_ be either joint
               ^^^^^^^^^^
> or collective, as those are mutually exclusive and exhaustive conceptual
> subcategories.

Should be "multi-author", not collective.

[Lawrence Rosen]

Perhaps Rick and Matthew are both a little confused? Take a look at
http://rosenlaw.com/oslbook.htm, chapter 2, page 32.

/Larry

Lawrence Rosen
Rosenlaw & Einschlag, a technology law firm (www.rosenlaw.com)
3001 King Ranch Road, Ukiah, CA 95482
707-485-1242 * cell: 707-478-8932 * fax: [phone]
Skype: LawrenceRosen
Author of "Open Source Licensing: Software Freedom and
                Intellectual Property Law" (Prentice Hall 2004)

> -----Original Message-----
> From: Rick Moen [mailto:rick@...]
> Sent: Sunday, July 08, 2007 6:00 PM
> To: license-discuss@...
> Subject: Re: how much right do I have on my project, if there are patches
> by others?
>
> Quoting Matthew Flaschen (matthew.flaschen@...):
>
> > What if a significant amount of code is removed in the patch?  Could
> > that exceed the limits of fair use?
>
> These questions seem to indicate some serious confusion, since they
> discuss an issue that has nothing to do with whether one work is
> derivative of another in copyright law.  If you wish to better
> understand how that concept is defined in USA copyright law, your best
> starting point is the CAI v. Altai decision.
>
> > > Merging it into the original work strikes me as very likely to
> > > create a derivative work -- which then necessarily falls into either
> > > the joint- or collective-work category.
> >
> > Is this really correct?
>
> Attempting argumentum ad ignorantiam again, Matthew?  Your lack of
> understanding of basic legal concepts really isn't my problem.
>
> > If a derivative work wasn't coordinated at all with the original,
> > how could it be a joint or collective work?
>
> "Coordinated with" appears to have no meaning in copyright law,
> so your question is not answerable as stated.  That aside, it is simple
> logic that a collective (not "derivative") work _must_ be either joint
> or collective, as those are mutually exclusive and exhaustive conceptual
> subcategories.
>
> > Also, some derivative works are illegal.  How does that fit in?
>
> How can you _not_ understand that the creation and distribution of some
> derivative works is tortious (not necessarily "illegal"), on account of
> copyright violation?  And what on _earth_ does this question have to do
> with the preceding thread?
>
> Again, I don't know exactly what your problem is, here, but I'm unclear
> on why it's suddently become my job to teach you.
>
>
> > > If contributors prove that ye olde primary author has failed to
> > > safeguard their interests or has violated agreements with them, then
> > > nonetheless the primary author may find he/she lacks that option.
> > > Please note that courts tend to measure "interests" for civil-law
> > > purposes in economic terms:
> >
> > So if a contributor uses copyleft for non-pecuniary reasons, they have
> > no recourse, right?
>
> I said nothing like that.
>
> Whether one party has committed a tort towards another depends on the
> particular circumstances and the duties that person has _in_ those
> circumstances.  I cited _one_ theory of law that could apply.  Many
> others could, also.
>
>
> > > But, anyhow, in that hypothetical, any party (obviously) can fork
> > > rev. n-1, taking over maintenance under copyleft.
> >
> > But that's hardly enough, if the primary author has created a
> > proprietary fork against a major contributor's wishes.
>
> "Enough" for what?  Facts do not change just because you don't approve
> of them, Matthew.  See .signature block, please.
>
> --
> Cheers,                                      "Reality is not optional."
> Rick Moen                                             -- Thomas Sowell
> rick@...

[Matthew Flaschen]

Rick Moen wrote:
> Quoting Matthew Flaschen (matthew.flaschen@...):
>
>> What if a significant amount of code is removed in the patch?  Could
>> that exceed the limits of fair use?
>
> These questions seem to indicate some serious confusion, since they
> discuss an issue that has nothing to do with whether one work is
> derivative of another in copyright law.

Perhaps you (or someone else) could explain.  If a patch removes code
from the original program, it must necessarily include some of the code
from that program.  Since the patch is a new work that contains part of
an earlier copyrighted work, I would think it to be derivative (and
possibly fair use if otherwise infringing).  Is that correct?

>>> Merging it into the original work strikes me as very likely to
>>> create a derivative work -- which then necessarily falls into either
>>> the joint- or collective-work category.
>> Is this really correct?  
>
> Attempting argumentum ad ignorantiam again, Matthew?

I'm not making an argument at all.  You're probably completely right.
I'm simply confused, and seeking clarification.

> Your lack of understanding of basic legal concepts really isn't my problem.

I certainly never said or implied it was.  Feel free to ignore my
questions if you wish.  That said, I do appreciate what I learn on this
list.

>> If a derivative work wasn't coordinated at all with the original,
>> how could it be a joint or collective work?
>
> "Coordinated with" appears to have no meaning in copyright law,
> so your question is not answerable as stated.  That aside, it is simple
> logic that a multi-author (not "derivative") work _must_ be either joint
> or collective, as those are mutually exclusive and exhaustive conceptual
> subcategories.

I think I misunderstood you badly.  When you said:

>>>Merging it into the original work strikes me as very likely to create
>>>a derivative work -- which then necessarily falls into either the
>>>joint- or collective-work category."

I thought you were saying *any* derivative work was either collective or
joint (and thus a multi-author work).  I was then confused about why,
for instance, a fork of a program that is not actively developed (an
example of what I meant by a work that was not coordinated with the
original author) must be a multi-author work.

If I understand correctly now, you actually just meant that merging a
patch in this situation must create (or modify) a multi-author work.

>> Also, some derivative works are illegal.  How does that fit in?
>
> How can you _not_ understand that the creation and distribution of some
> derivative works is tortious (not necessarily "illegal"), on account of
> copyright violation?

I think I do understand that.  I should have phrased my question better.
 What I was wondering is how an infringing derivative work could be a
multi-author work (which I assumed implied licit cooperation).

That was based on my (mis?)understanding that you were saying *all*
derivative works were either collective or joint works.

>>> Please note that courts tend to measure "interests" for civil-law
>>> purposes in economic terms:
>> So if a contributor uses copyleft for non-pecuniary reasons, they have
>> no recourse, right?
>
> I said nothing like that.
>
> Whether one party has committed a tort towards another depends on the
> particular circumstances and the duties that person has _in_ those
> circumstances.  I cited _one_ theory of law that could apply.  Many
> others could, also.

Thanks for explaining.

>>> But, anyhow, in that hypothetical, any party (obviously) can fork
>>> rev. n-1, taking over maintenance under copyleft.
>> But that's hardly enough, if the primary author has created a
>> proprietary fork against a major contributor's wishes.
>
> "Enough" for what?  Facts do not change just because you don't approve
> of them, Matthew.  See .signature block, please.

Of course, I meant enough to satisfy the major contributor.  But I am
well aware that the major contributor may actually have no legal
recourse, however unhappy this makes them.

Matthew Flaschen

[Matthew Flaschen]

Lawrence Rosen wrote:
> Perhaps Rick and Matthew are both a little confused? Take a look at
> http://rosenlaw.com/oslbook.htm, chapter 2, page 32.

Thanks.  In general, I've found your book very valuable.  That section
says in part:

"Each contribution to a collective work is owned by its author, and that
author has the exclusive right to decide how that contribution is to be
licensed. A contribution to a joint work is owned by all of its
authors jointly."

Eric and Catherine's essay says:

"The difference is practically relevant because, according to 17 USC 201
the holder of the collective-work copyright is legally privileged to set
the distribution terms for the package as a whole (in the statute, this
expressed negatively as a statement that the collective-work copyright
holder acquires only those rights)."

It seems the essay mentions a single person (presumably the primary
author) holding a collective-work copyright, allowing them to relicense
the whole program without consulting the other authors.

But the book does not mention this, and seems to imply otherwise by saying:

"Relicensing a joint work is, in some ways, easier than reli-
censing a collective work because any one of the authors can
do it without consulting the others, but it may leave some
contributors angry with the results."

If the collective-work copyright holder can relicense without consulting
anyone, that isn't really harder (for them) than it would be for a joint
work.

The statute (http://www.copyright.gov/title17/92chap2.html) says in part:

"the owner of copyright in the collective work is presumed to have
acquired only the privilege of reproducing and distributing the
contribution as part of that particular collective work, any revision of
that collective work, and any later collective work in the same series."

Is there a conflict here, or am I just misunderstanding again?

Matthew Flaschen

[Rick Moen]

Quoting Lawrence Rosen (lrosen@...):

> Perhaps Rick and Matthew are both a little confused? Take a look at
> http://rosenlaw.com/oslbook.htm, chapter 2, page 32.

I've read that chapter, and do appreciate your work (plus your pointer to
it).  

I'm not sure I was confused before, but definitely am now -- in that I
don't see anything there differing from what I was saying.  Care to
elaborate, please?

[Rick Moen]

Quoting Matthew Flaschen (matthew.flaschen@...):

> Perhaps you (or someone else) could explain.  If a patch removes code
> from the original program, it must necessarily include some of the code
> from that program.

It is, substantively, a set of instructions saying "delete these lines".
Therefore, I really doubt a judge is going to call that an unlicensed copy.
As has been said in this space before, judges are not Turing machines,
and with reasonable luck will leave their brains turned on while
hearing such a case.

> If I understand correctly now, you actually just meant that merging a
> patch in this situation must create (or modify) a multi-author work.

I'm saying that (in the general case) a patch by itself is an
independent creative work, and is likely to be treated in law the same
way as commentary on someone's work is.  By contrast, a patch being
_applied_ to the work is very likely to create a derivative work of the
latter.

That derivative work, being a multi-author work, must then necessarily
be _either_ a joint work or a collective work, by definition.  (This is
in reply to your suggestion that such a work with a third-party patch
applied to it _might be_ something else other than a joint or collective
work.  No.)

> I think I do understand that.  I should have phrased my question better.
>  What I was wondering is how an infringing derivative work could be a
> multi-author work (which I assumed implied licit cooperation).

Er, I'm having a difficult time parsing that.  Are you starting out by
positing something that is _both_ licit and infringing at the same time?
Neat trick, that.  Something one encounters on a noonday midnight,
perhaps?  ;->  Finding a way for it to then be multi-author, after you've
already found a way for it to be both licit and infringing, would
probably be simple by comparison.  

[Matthew Flaschen]

Rick Moen wrote:
> It is, substantively, a set of instructions saying "delete these lines".
> Therefore, I really doubt a judge is going to call that an unlicensed copy.
> As has been said in this space before, judges are not Turing machines,

Well, if they're not Turing machines, they might be non-deterministic,
and then it's even harder to predict what they'll do. ;)

>> If I understand correctly now, you actually just meant that merging a
>> patch in this situation must create (or modify) a multi-author work.
>
> I'm saying that (in the general case) a patch by itself is an
> independent creative work, and is likely to be treated in law the same
> way as commentary on someone's work is.  By contrast, a patch being
> _applied_ to the work is very likely to create a derivative work of the
> latter.
>
> That derivative work, being a multi-author work, must then necessarily
> be _either_ a joint work or a collective work, by definition.

I guess this is the part I don't understand.  By analogy, if you make a
film adaptation of Lord of the Rings, it's a derivative work of the
book, and there are two authors involved.  But I don't see how it could
be a multi-author work, since Tolkien is dead and thus there couldn't be
any collaboration.

Similarly, a fork that the original author ignores would seem to be just
as lacking in collaboration, and thus not a multi-author work.  Someone
could make this fork by writing then applying a patch, but I still don't
get how it would be multi-author (rather than a simple derivative work).

Now, if the original author applied the patch, I would understand it
being a multi-author work.  Are you talking about exclusively this
latter scenario?

  (This is

> in reply to your suggestion that such a work with a third-party patch
> applied to it _might be_ something else other than a joint or collective
> work.  No.)
>
>> I think I do understand that.  I should have phrased my question better.
>>  What I was wondering is how an infringing derivative work could be a
>> multi-author work (which I assumed implied licit cooperation).
>
> Er, I'm having a difficult time parsing that.  Are you starting out by
> positing something that is _both_ licit and infringing at the same time?

No.  My understanding is that joint and collective works are legally
created (by definition).  But derivative works are not necessarily
legal.  So my point was that an infringing derivative work couldn't be
joint or collective (/because/ that would be a conflict)

Matthew Flaschen

[Rick Moen]

Quoting Matthew Flaschen (matthew.flaschen@...):

> I guess this is the part I don't understand.  By analogy, if you make a
> film adaptation of Lord of the Rings, it's a derivative work of the
> book, and there are two authors involved.  But I don't see how it could
> be a multi-author work, since Tolkien is dead and thus there couldn't be
> any collaboration.

Er, there's nothing about multi-author works that necessitates collaboration.

The 2004 recording of Rhapsody in Blue and An American in Paris, where
Michael Tilson Thomas and the Columbia Jazz Band play the
small-orchestra part along with George Gershwin's 1925 reproducing-piano
roll, most assuredly was _not_ collaborative -- unless Tilson Thomas has
developed previously unknown acumen with Ouija Boards.  And yet it
provably had multiple creators.

> Similarly, a fork that the original author ignores would seem to be just
> as lacking in collaboration, and thus not a multi-author work.

Again, I deny your premise about "collaboration".  I have no idea where
you got that.


> No.  My understanding is that joint and collective works are legally
> created (by definition).  But derivative works are not necessarily
> legal.

Nowhere did I say that derivative works are necessarily legal, so I'm
utterly unclear on what you're driving at.

> So my point was that an infringing derivative work couldn't be
> joint or collective (/because/ that would be a conflict)

First, I have not a clue why you are suddenly talking about
"_infringing_ derivative works" in this context.  (I sure hope that, by
"infringing", you mean infringing copyright law.)  A copyright owner who
heard of such a work would deal with it simply by demanding that the
unauthorised party cease and desist, and if necessary sue to enjoin.

The question of whether it's joint or collective is pretty silly because,
rather more to the point, it's pretty much going to get killed off, upon
being noticed.

I realise computerists tend to have a rabid fixation with odd edge cases,
but c'mon, now.

[Matthew Flaschen]

Rick Moen wrote:
> Quoting Matthew Flaschen (matthew.flaschen@...):
>
>> I guess this is the part I don't understand.  By analogy, if you make a
>> film adaptation of Lord of the Rings, it's a derivative work of the
>> book, and there are two authors involved.  But I don't see how it could
>> be a multi-author work, since Tolkien is dead and thus there couldn't be
>> any collaboration.
>
> Er, there's nothing about multi-author works that necessitates collaboration.

See below.  The key issue I'm still wondering about is:

Must every derivative work (made by a different author than the
original) involve either a joint or collective work?

If not, which ones must be?

> Again, I deny your premise about "collaboration".  I have no idea where
> you got that.

From the Raymond essay, "Campbell vs. Lavery is appellate case law
indicating that collaboratively-written software is a collective rather
than joint work when programmers function in identifiable
author/contributor roles."  This (and probably other parts) gave me the
understanding that collective and joint works must involve collaboration.
>
>> No.  My understanding is that joint and collective works are legally
>> created (by definition).  But derivative works are not necessarily
>> legal.
>
> Nowhere did I say that derivative works are necessarily legal, so I'm
> utterly unclear on what you're driving at.

Okay, this is becoming a tangent.  I was trying to make a
counter-example, but it really isn't that relevant to the core issue.

Matthew Flaschen

[Rick Moen]

Quoting Matthew Flaschen (matthew.flaschen@...):

> Must every derivative work (made by a different author than the
> original) involve either a joint or collective work?

The law sets that up as a logical dichotomy concerning multi-author
works, based on the authors' intent.

If you can, nonetheless, think of some bizarre way for a multi-author
work to be (in your view) something else, please do let us know.

> From the Raymond essay, "Campbell vs. Lavery is appellate case law
> indicating that collaboratively-written software is a collective rather
> than joint work when programmers function in identifiable
> author/contributor roles."  This (and probably other parts) gave me the
> understanding that collective and joint works must involve collaboration.

Some multi-author works are collaborative.  _Those_ (specifically, those
involving software) were addressed by the cited court decision.  Not all
multi-author works are collaborative (e.g., MTT, Columbia Jazz Band, and
Gershwin).