[cc-licenses] If a middleman gives or sells a "by" licensed image...

wiki_tomos wiki_tomos at inter7.jp
Fri Sep 14 06:45:01 EDT 2007


Hi, Jen. I am sorry about the confusion. You are correct - 
the Miscellaneous section is indeed identical between by and by-sa. 
But the explanation of my interpretation remains basically the same. 
Remove the sentense "If you look at the -by-sa license, ... " 
from that email, and the rest can stand on its own. 

>If so, how is that different from the "by-sa" license?

The key difference between -by and -by-sa is that with a -by license the 
MiddlemanGuy is allowed to license his part of the derivative work 
under any license whatsoever, because the OriginalAuthor did not require 
the MiddlemanGuy to share alike, whereas with a -by-sa license, 
the MiddlemanGuy has to license his part under the same or other 
equivalent license. (Again, at least that's how I understand it). 

>Does the "by" license mean that every future generation of work that  
>includes an element from a previous generation of work requires an  
>attribution back to whichever artist introduced that element originally?

My answer would be yes, in many cases, with some footnotes.

<footnote 1:>
If, say, FifthGenerationWork includes the OriginalAuthor's creative 
expression, the Client who wants to use it has to attribute the OriginalAuthor, 
the creator of the FirstGenerationWork. CC-by is the only license offered to 
the Client. 

<footnote2:>
Not any "element from a previous generation of work" would result in the 
obligation to attribute the OriginalAuthor, though. Non-creative element does not 
has to do with a copyright license, so I think the element 
inhereted from the FirstGenerationWork should be minimally creative as 
to give it a copyright protection. Also, obviously, I think elements from 
people other than the OriginalAuthor, inhereted in whatever way, does not 
create the obligation to attribute the OriginalAuthor. 

<footnote3:>
If it is very small portion, there might be a good chance that 
the Client can go for a fair use. 

<footnote4:>
Another exception has to do with the case in which there is no 
attribution to the OriginalAuthor. As you know, -by licenses 
have the following restriction: 

"If You create a Derivative Work, upon notice from any Licensor 
You must, to the extent practicable, remove from the Derivative 
Work any credit as required by clause 4(b), as requested."

(taken from http://creativecommons.org/licenses/by/2.5/legalcode)

So there is a possibility that the Client of a FifthGenerationWork 
does not know, and cannot be expected to know that the OriginalAuthor 
created some elements in the work. In that case, I suppose it is 
somewhat likely that the Client cannot be blamed for not attributing 
the OriginalAuthor when he makes use of the FifthGenerationWork. But then, 
the Client might simply assume that he does not have license to make 
any use beyond fair use and some other statutorily defined use with 
regards to that unattributed elements. It is like some orphan work 
is in the FifthGenerationWork. 

----

Two related but different topics:


Is it surprising that, according to my interpretation, the 
CC-by-sa license works differently? In case of CC-by-sa, the 
whole derivative work is under whichever license the MiddlemanGuy 
chooses. OriginalAuthor seems to give permission the MiddlemanGuy 
to release the OriginalAuthor's creative expressions under the 
same or an equivalent license. And because of that, CC-by-sa 2.0'd 
work, a derivative of which can be licensed under CC-by-sa 3.0, 
and then if you create ThirdGenerationWork from it, it could be 
released under GFDL, when/if Creative Commons designates GFDL 
as a "Creative Commons Compatible License" as defined in the 
1.c. (http://creativecommons.org/licenses/by-sa/3.0/legalcode )

In other words, while all the creative expressions of the 
OriginalAuthor remains under CC-by license, and nothing else, 
in case of -by'd work, some portion of the expressions become 
available under some different licenses in case of -by-sa'd work. 

(Discussion about this topic is here: http://lists.ibiblio.org/pipermail/cc-licenses/2006-January/date.html )


Another one. Is there a possibility that if the FifthGenerationWork 
does not contain any creative expression of the OriginalAuthor, the 
author of the FirstGenerationWork, the Client who uses the fifth gen. work 
must avoid giving attribution to the OriginalAuthor? Could the attribution 
in that case constitutes "reverse passing-off"? Does this mean the Client 
has to investigate which elements are inhereted from which author, and 
determine if the inhereted elements are in any way creative enough to 
make it copyrighted? That sounds too much work to me, and I hope the 
answer is no. 

But I don't know. 

This is not a question unique to the -by licensed works, but the 
same can be asked about the -by-sa'd works. 

To make the matter even more complicated, in case of -by-sa, 
it seems possible to argue that the FourthMiddlemanGuy who 
created FifthGenerationWork and released under CC-by-sa-2.5 
is obligating the Client to keep the attribution, and ommission 
of the attribution to the OriginalAuthor constitutes a bleach 
of contract term, resulting in the termination.. 

I currently have no answer to this question. But I can say that 
this is not a hypothetical situation. I see cases like this, 
in which OriginalAuthor's contribution is zero in a FifthGenerationWork 
or FiveHundredthGenerationWork. You can find them on Wikipedia. 
Wikipedia is currently licensed under GFDL, but there has been 
discussions that CC-by-sa is a better license for them, and they 
might switch the license if the currently ongoing drafting of the 
next version of GFDL opens up such a migratory path. 


Best, 


Tomos



More information about the cc-licenses mailing list