Development of Creative Commons licenses

[Terry Hancock <hancock AT anansispaceworks.com>]

On 4 Jan 2006 13:22:30 +0900
wiki_tomos <wiki_tomos AT inter7.jp> wrote:
> In this example, Greg's poem is under CC-BY-SA 2.0, and 
> Helen's lyric based on the poem is under CC-BY-SA 3.0. 
> 
> And the central question is when CC-BY-SA 2.0 says 
> Ian "may distribute, ... a Derivative Work only under 
> ... a later version of this License with the same License 
> Elements as this License..." does that mean that Ian 
> can now switch the license of Greg's poem from 2.0 to 3.0?
> 
> Or, as I suggested, is it simply requiring Ian to release 
> his contribution to the book under one of the licenses,
> while  keeping the Greg's part still under 2.0? 

Greg's license already gives you the right to use his
work under the CC-BY-SA 2.0. Ian isn't the one
granting that.

But you can *also* use Greg's work under the terms of the
CC-BY-SA 3.0 because of the text allowing the license to be
converted.

You're always free to use the earlier license on the work
that is covered by it.

> I believe you support the former interpretation. I see one
> problem  with that interpretation nowadays.

> (3) If this were the case, one consequence is 
> that when Bob *minimally* change Alice's CC-BY'd short
> story  to make it a derivative short story, Bob can
> release the whole  piece under, say, a license which does
> not require any attribution.  By that move by Bob, most of
> Alice's piece is now usable without  attribution to her.
> Is this what CC-BY allows? I doubt. And this  is the
> problem with that interpretation.

ISTM there are two problems with this:

1) If the changes are indeed sufficiently minimal, then the
work may not even be legally a "derivative" -- it may be the
same work, thus Alice's license still applies.

2) The legal precedent for preserving attribution is both
much less ambiguous and much older than that for preserving
copyleft. So I don't necessarily think that conclusions
based on the legal status of attribution are necessarily
transferrable to copyleft redistribution.

> The fact that term "restriction" is used for the SA
> licese's  provision regarding derivative work's license
> choices means,  instead, this: CC-BY does not impose
> restriction regarding  how Bob releases "Bob's creative
> contribution in the derivative work  Bob created using
> Alice's work" (i.e. Bob's part). It is up to Bob  how Bob
> licenses Bob's part. But whatever license Bob chooses, 
> Alice's work remains under the same license, CC-BY. In
> contrast, SA  liceses restricts Bob's release of Bob's
> part. It has to be under a  SA license (and/or GFDL). But
> again, Alice's part is still under the  same, initial SA
> license. So if Alice's work was under SA 2.0, and  Bob
> chooses SA 3.0, Alice's part embedded in Bob's derivative
> is  usable only under SA 2.0, not 3.0. 

Except that the license makes none of these distinctions
about "original part" and "derivative part" -- it just talks
about combined or derivative *works*.  So I don't think this
interpretation really follows.

I do think that, in as much as Bob's derived work includes
in full (or essentially full) part Alice's work, that it is
a mere *re-distribution* of Alice's work (not "modified
re-distribution"), which is (IIRC) covered by a previous
clause which doesn't provide for the upgraded license. That
is to say, it is covered by CC-BY simply because it *IS*
Alice's work, not because it is "based on" Alice's work.

> That's how I came to interpret what SA licenses says. 
> 
> >Whether this is desireable or not is an issue for CC (and
> >this list) to discuss -- has Greg been unfairly taken
> >advantage of by this sequence of events?
> 
> I agree with you on this.

That we should discuss it, or that Greg loses in this
arrangement? ;-)

CC-BY is, AFAIK, very similar to the BSD/MIT type license
for software, and it is actually quite common for that
software to be distributed without attribution or source
code, so ISTM that indeed there isn't a whole lot of
security that the attribution can't be stripped as you
describe.  Only when the source files are included -- the
verbatim distribution of the author's original work -- do
you see the attributions.

Of course, there's an element of ethics as well as law here
-- people usually don't try to strip authorship out like
this, because the consequences of being deemed unethical
when found out are a lot worse than any diminution of value
in not being the original author.  And, since the original
distribution is there to be seen, such exposure is almost
guaranteed.

That ethical factor strikes me as a major difference between
"BY" and "SA" terms -- the ethics of the former are
clearcut, whereas the latter's are very much a matter of
opinion.

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