[OSM-legal-talk] Viral can be nice

Frederik Ramm frederik at remote.org
Thu Apr 22 10:29:08 BST 2010


TimSC wrote:
>> The share-alike element in data is stronger with ODbL than it was with 
>> CC-BY-SA. 

> I assume you mean it is "stronger" as in "enforceable"?

Since CC-BY-SA is almost surely not applicable to a collection of facts 
like ours, it is not hard to make something that is more enforceable ;)

> Perhaps I am 
> missing another area of strengthening. The intent of CC-BY-SA is all 
> derived works are also SA. Otherwise, it seems ODbL is weaker - produced 
> works are not share alike?

Yes but we get something in return which is more valuable (if you define 
value in terms of how much something brings us closer to the stated goal 
of the project).

My standard example is that of a guy who makes a T-Shirt with an OSM map 
of his town on it. Say he finds the data for the town insufficient, and 
spends a few days surveying the bits that are important for his T-Shirt.

Under CC-BY-SA we have share-alike on the T-Shirt. This means everyone 
can buy the T-Shirt, put it under a scanner and hope to retrieve some 
meaningful data from it. (Plus people can simply upload the T-Shirt 
design to any T-Shirt making web site and undercut the original creator 
by one cent; a situation that reduces the likelihood that the T-Shirt 
will be made in the first place.) The improved data, however, need not 
be shared; since the improved data as such is never published, it does 
not fall under the CC-BY-SA share-alike clause. The data remains hidden 
to us; we only get a scanned T-Shirt.

Under ODbL, we don't necessarily get the T-Shirt design. We consider the 
T-Shirt design to be the creator's intellectual property and grant him 
every right to restrict access to that. But what we do get is his 
improved data - because we say that this data is a derived work of ours, 
he has to share his improvements. Which is, in my eyes, a stronger 
request - we're not happy with people releasing some processed, 
filtered, compiled version of their data, we want the data itself.


More information about the legal-talk mailing list