Development of Creative Commons licenses

[Terry Hancock <hancock AT anansispaceworks.com>]

drew Roberts wrote:
> I was referrring to unit of packaging and today unit of copyright. I find 
> you 
> talking about works and atoms. I think we may be talking concepts that are 
> close if not the same.

Finally decided on "assemblies" and "components" for my open hardware
paper. There might be a better word than "assembly" for general use, but
"component" is pretty clear.

I was trying to draw out the parallel. It's kind of vague with the
things we've been talking about, but it's extremely obvious when you're
talking about computer hardware: it's very clear that if you have a
copyleft license on chip design that is invoked when you distribute the
chips, that a copyleft which binds "any larger assembly" is going to be
ludicrous in the end (from my paper on the subject):

"""
Now, if copyleft did not respect design abstraction layer limitations,
then a copylefted IP Core could only be used in a copylefted chip (so
far so good), the copylefted chip can only be used in a copylefted PCB
(not so good), the copylefted PCB can only be used in a copylefted
entire computer (terrible --- no one would use it!), and the copylefted
computer can't be used in a train system, unless the plans to the entire
train system are released under the original license (ludicrous). And
what exactly are the limits of the "train system"? It is integrated into
the design of a city (so the city has to be copylefted?) which is in
turn connected to other cities, states, countries, and indeed, all human
creations.

Clearly, this is not what we want! Not that we don't want to encourage
free PCBs, computers, and even whole train systems. But to demand this
kind of control over other peoples' work seems decidedly non-free.

So, I propose that we construct a copyleft license that explicitly
recognizes the concept of an abstraction layer, and the resulting limits
on copyleft, extending both up and down.
"""

Intriguingly, I find that these "abstraction levels", even though they
are very clearly established, are somewhat arbitrary. You have to define
them for specific cases. OTOH, you have some freedom in setting the
boundaries. I think it's quite reasonable to have a license that
requires that free IP Cores only be used with other free IP Cores in a
single chip, but it gets pretty questionable when you make the leap to
the PCB level of abstraction.

I think it's fair to guess that if you did try to claim that your
licenses forces "all human creations" to be released under your license,
that the courts would be liable to rule against you. ;-)

Unless we explicitly define limitations to such abstraction levels, then
we are essentially asking for the courts to make such a ruling, and
until they do, we are leaving ambiguity and feeding FUD about free
licenses and their "infectious" "viral" behavior.  That seems like a
very bad idea to me.

Cheers,
Terry



-- 
Terry Hancock (hancock AT AnansiSpaceworks.com)
Anansi Spaceworks http://www.AnansiSpaceworks.com