Development of Creative Commons licenses

["Greg London" <teloscorbin AT gmail.com>]

On 6/26/06, drew Roberts <zotz AT 100jamz.com> wrote:
> On Monday 26 June 2006 10:17 am, Greg London wrote:
> > The AGPL is copyleft. Phil's project isn't.
>
> My point was that you can come up with new copyleft
> licenses and perhaps new copyleft like schemes.

And my point is that this is a non-sequitor.
Phil's project isn't Free, Libre, or Shared.

> look at the mods I suggested. How is it that if a project adopted
> the ideas I suggested that contributors would not end up with pretty
> much copyleft protection?

This may create a copyleft scenario, but it isn't Free because
it creates way more overhead to contribute than something like
the BSD advertising clause. The BSD license is nearly public
domain, but people protested the overhead associated with
the advertising clause, so it was removed.

So, you could just as easily craft some sort of copyleft license,
a license where the work and any derivative of teh work all
must be put under the same license, but you could add things
to the license that create extra overhead that hinders contributions,
that hinders people from creating their own derivatives, that hinders
people from forking the project if they wish.

There is overhead associated with Attribution in early CC licenses,
and it was fixed by allowing people to assign attribution to a single
site.

So, it isn't enough to be copyleft, like some hypothetical scheme
that you come up with to replace Phil's scheme, and it isn't enough
to be almost Public Domain like BSD or CC-BY. Small demands
placed in a copyleft of Public Domain license will make the license
be not Free, even though the license it copyleft or Public Domain.

Specific licenses, such as the latest versions of GNU-GPL and
CC-BY-SA are written to be copyleft and to also avoid any
bolt-ons (advertising clauses, accumulating attribution) that add too
much overhead to weigh down any Free project.

> How would someone in Phil's position have any advantages?

Maintaining someone at an advantage is only one problem.
Even with a license such as the old BSD license with its
advertising clause, no one is at an advantage, but the
overhead can become problematic to the point that
contributions become too much work.

Phil's approach keeps his contributers at a serious
disadvantage compared to his position.
Your approach may solve the problem so that no one
is at an advantage over anyone else, but it still has
overhead problems of having to report all mods back
to Phil.

So when Alice starts her fork, she has to notify Phil.
And when Bob wants to fork off of Alice's version,
he has to notify Alice and Phil. And when Charlie
wants to fork off of Bob's version, he has to notify
Bob, Alice, and Phil.

The effect is cumulative, exactly like the BSD advertising
clause, exactly like requiring individual attribution in the
early CC-BY licenses. Every fork'ed generation becomes
more and more work to create, which in a sense, keeps
Phil's version at an advantage, because Phil doesn't have
to notify anyone.

This is the effect of any cumulative license requirement:
The first version is always easiest to modify because
it has the least cumulative requirements. Which means
that the first version is always at an advantage over
everyone else.

The only way to keep the field completely level so that
no one is at an advantage and so no overhead accumulates
no matter what version you fork off of, you have to allow
anyone to create a derivative off of any version they wish,
and you can't require that they notify you of those modifications.

> Look, if I am mistaken and the mods I suggested would
> behave other than I imagine, please point out the problem.
> I am all ears. If they would behave like I imagine, but three
> are still major problems, point them out, I have tried to.

They would behave like you imagine, but there are
still problems with the approach that make them
not Free.


> I don't think so. From what I grasp, he is cool with anyone
> forking and making their own version "on wiki" and then
> taking them "off wiki" in a BY-ND like fashion. He just
> doesn't want forked versions existing in the wild that are
> not also "on-wiki." (That is what I understand his desires to be.)

Exactly. He has created a licensing and contract scenario
that maintains his version of the work at an advantage.
What if Alice wants to create a Fork of the project?
And then Bob wants to take Alice's version and fork that?
and Charlie wants to fork Bob's version?

Charlie has to notify Bob, Alice, and Phil.
Therefore it is easier to modify Phil's version,
therefore Phil has created a legal requirement
that demands his version maintain an advantage
over all other versions.



> So, if he commits as a part of the copyright grant or
> assignment process to also not go off wiki as anything
> other than BY-ND, everyone will be on equal footing.

except Phil legally establishes himself as gate keeper.
the work is not Free, you have to get Phil's permission.
If Alice makes a mod that Phil doesn't like, Phil can revert it.
If Alice wants the modification in the work, she cannot create
her own fork without Phil's PERMISSION. If you need someone's
permission, you are not on equal footing with them.

> Anyone can fork "on-wiki"

As long as Phil permits it. Will Phil also agree to allow forks
of the work "on-wiki"? Or will he be able to reject modifications
if he doesn't like them?

And again, see all the problems with "overhead".
Actually, with Phil's approach, you can't have multiple
generation forking because only Phil can fork.
It prevents cumulative overhead from forking, but
only because it doesn't allow forking at all.

> What am I missing/overlooking? Would this not work?

All Rights Reserved "works", It gets people to create books
and software and all sorts of things.
So the question isn't whether it "works".

The question is whether it's Free/Shared or not.

And every hoop you've tried to jump through,
every hypothetical suggestion you've come up with,
every piece of legal maneuvaring you've suggested,
ends up with a scenario that maintains Phil at an
advantage over all other contributers, because everyone
must go through Phil but Phil doesn't have to go through
anyone. There's also the question of what happens
when Phil decides he doesn't like someone's modification
and won't allow it to remain on the wiki. It becomes
awfully hard for Alice to create a forked version of
the work, if Phil keeps reverting it.

All these hypotheticals are pointless, in my opinion.
The point is a much more practical matter:

Phil can do whatever he wants to do to the extent
allowed by copyright law. He can license his stuff
however he wants to license it. He can require
contributers sign whatever contract he wants them
to sign.

AND his approach isn't Free, nor is he Sharing.

There is no way to mandate that all modifications must
flow back to Phil without mandating that Phil is kept
at an advantage over all other contributers.

To create a scenario that is (1) Free and that (2) requires
Phil be notified of all mods is to satisfy two mutually
exclusive requirements. One necessarily excludes
the other.

And the practical answer is that GNU-GPL and CC-SA
already exist, so if Freedom is what you want, then use
them. If you don't want Freedom, then use something else.

Phil's choice of license has broadcasted what he wants.
Freedom is not his highest priority. If it was, he would
have chosen a different license. Phil's priority is to maintain
himself at an advantage over the community because he
believes he knows what's best for the community.
He can do that since copyright allows it, but I'm not going
to spend time discussing hypothetical scenarios trying
to solve an impossible problem.


Greg