[OSM-talk] License plan

Ben Laenen benlaenen at gmail.com
Fri Feb 27 14:27:56 GMT 2009


On Friday 27 February 2009, Peter Miller wrote:
> Would it be appropriate to continue this conversation on legal-talk?
> Talk is very busy at the moment and we have a lovely list of our own
> :)

I think this discussion is important enough to take place on the talk 
mailing list. If it's held on the legal-talk mailing list this will 
just be obscured as there are not a lot of people subscribed to it.

> With regard to these questions, lets make a clear distinction between
> the question (which anyone is qualified to raise) and the answer (for
> which one is likely to need to be a lawyer)! Possibly Ben is a
> lawyer, I am not sure, but in the past we have tended to form a legal
> opinion without ever asking a lawyer.
>
> My company is taking legal advice on the licence and will publish the
> conclusions to the community. We will use the Open Issues page
> (http://wiki.openstreetmap.org/wiki/Open_Data_License/Open_Issues )
> as part of the input to our legal review. If you have questions then
> please ensure that they are mentioned on that page.

Well, IANAL. But the situation looks quite clear to me:

* person X enters data
* person Y uses that data to make his own data
* person X opposes the license change, person Y approves
* there's no way to tell the origin of all the data in the OSM database
* there could be no way to tell that some of person Y's data is derived 
from person X
* if the new license database keeps all of Y's data not based on the 
data from X as it can be found in the current OSM database, it will 
still miss heaps of the derived data.
* result: the new database will have improperly licensed data.

There's exactly one way to be sure this won't happen: get approval of 
*all* the people who've been editing OSM. And with a number of around 
100.000 mappers I'm very skeptical that you'll be able to manage that. 
If there's even one mapper not giving his approval, you're not able to 
move the database as you just can't tell if there's derived data or 
not.

Ben




More information about the talk mailing list