[Osmf-talk] Contributor Agreement is Dual Licensing

Gervase Markham gerv at gerv.net
Sun Dec 13 01:16:38 UTC 2009

On 12/12/09 14:42, Frederik Ramm wrote:
> Gervase Markham wrote:
>> So the official position of the OSMF is that the upgrade clause in the
>> ODbL is not sufficient insurance against future eventualities, and
>> that the OSMF must acquire (joint) full rights over the data of all
>> contributors
> No. "full rights" would mean that OSMF can do what they want with the
> data. Instead, they are just asking for the right to license it using
> any "free and open license agreed to by a majority of active mappers",
> which is much more limited that "full rights".

That's approximately the equivalent of what the FSF has. They also are 
only allowed to licence code under a free licence, using the same sort 
of "constitutional" restrictions. Probably stronger ones, in fact.

People still think it's a big deal to sign over copyrights to them.

>> There is a great deal about this which is strange. If the relicensing
>> happens, then all contributions at the time of the relicensing by
>> people who have agreed to the terms will be under ODbL 1.0 and
>> CC-BY-SA 2.0 anyway.
> No. If you look closely, the whole relicensing process as published
> today does not contain any element that puts the data under ODbL.
> Everyone just agrees, via the above statement, that OSMF may but the
> data under ODbL, and of course that's what they will do; but if the
> statement above would not contain the sentence about agreeing to let
> OSMF publish the data unter ODbL then there would be no legal basis for
> that.

OK, thank you, that's helpful. Although it seems an odd way to do it. 
What are we asking people in the relicensing process, if not "please put 
your contributions under the ODbL"?

> But now that you bring it up, the statement above has one funny bit and
> that's the mentioning of a version number ("ODbL 1.0"). ODbL has a
> built-in upgrade clause allowing anyone to take ODbL 1.0 licensed
> content and re-issue it, say, ODbL 1.1, but still, for clarity, the
> contributor agreement should say "ODbL 1.0 or later".

I agree it makes no legal difference, but this would be a reasonable 
change for clarity.

> You are right in saying that the definition is unclear but I think that
> certain licenses are surely *not* free and open. Moving to such a
> license would not render the vote invalid, but it would IMHO
> automatically terminate the contributor agreement. If I give you my data
> under the contributor agreement and you change your license to something
> not free and open, then you cannot automatically use my data - it is as
> easy as that! I'd say that ambiguity about what "free and open" is
> doesn't hurt here; if OSMF were to move to what some consider a not free
> and open license by majority vote and still attempt to automatically
> take with it all the data, then some contributors will probably sue (and
> that's why OSMF would not do it in the first place).

Hmm. Possibly. But that seems like the wrong way to have a corrective 
mechanism. Saying "oh, people can always go to court" isn't a great way 
of putting breach remedy terms in a contract.

> It would be wrong to simply drop the words "free and open" because that
> would of course remove any right for an individual to sue should OSMF
> suddenly decide to go proprietary.

I would much rather the OSMF dropped the whole "any other licence" bit.


More information about the osmf-talk mailing list