[OSM-talk] [Osmf-talk] Attribution guideline status update

Simon Poole simon at poole.ch
Mon Sep 9 12:58:09 UTC 2019


Look I'm sorry that the ODbL is not contact poison.

It is a, in some ways very permissive (so permissive that it isn't even
compatible with the CC BY licenses), open licence with some share-alike
aspects.

The ODbL allows creating extracts of all kinds, vertical, horizontal,
our guidelines ensure,

- that data extracted remains ODbL licenced,

- that third party data that is used in conjunction with OSM data in a
Collective Database remains cleanly separable,

- that OSM data is useful for the sector in general in that it is usable
in similar ways to other datasets if the other prerequisites are fulfilled,

- that the OSM specifics of the database (for example that we are
continuously producing new versions) can't be gamed to make the data
unusable,

- that the best way to fulfil SA obligations is to improve the original
dataset (strictly spoken this is actually a bit at odds with the ODbL),

and last, but not least, that the attribution guideline when it is
finalised is clearer than the current guidanceĀ  and helps us massively
reduce unattributed use of OSM data.

Simon

Am 09.09.2019 um 14:14 schrieb Christoph Hormann:
> On Monday 09 September 2019, Simon Poole wrote:
>> Am 09.09.2019 um 12:08 schrieb Christoph Hormann:
>>> Existing guidelines allow a lot of things that are clearly not
>>> allowed by the ODbL itself in terms of share-alike (like the
>>> regional cuts concept for example).
>> That statement is completely wrong. [...]
> I disagree.  And in any case - it does not matter, the regional cuts 
> were used just as an example.  I could likewise have used the 
> horizontal layers as an example in my argument.  And surely you could 
> for those equally present an interpretation of the ODbL that justifies 
> those.
>
> My point is not that you cannot interpret the ODbL in a way that allows 
> all this.  This is what corporate lawyers do and what they are good at. 
> My point is that within the spectrum of possible interpretations of the 
> ODbL all of this is on the far side of leniency and the OSMF - w.r.t. 
> share-alike - has already moved their frame of refrence of what is the 
> appropriate/neutral interpretation very far in that direction.
>
> And i strongly advise you not to do the same for attribution because you 
> are playing with fire here regarding the social cohesion of the 
> project.
>
>>> [...] If you disagree please list cases where commercial OSM data
>>> users have published derivative databases.
>> There is no requirement to publish derivative databases, only a
>> requirement to make them available to recipients of such databases
>> and Produced Works created from them.
> So your argument is that using derivative databases is common practice 
> and map producers routinely make them available to the users of their 
> maps.  But none of this is visible in public because the recipients do 
> not distribute them despite them being available under the ODbL as the 
> license requires?
>
> I am not convinced.
>
> For clarity i repeat and clarify my statement:  Share-alike is 
> functionally dead in the world of commercial OSM map rendering.  Map 
> producers universally route around it - or at least claim to route 
> around it and their claims are not challenged.  If you disagree then 
> show me the derivative databases.
>
>> It doesn't change the license at all, in general the guidance is more
>> -strict- than current practice, with the exception of the multiple
>> source case where there currently isn't any guidance at all.
> So you are essentially saying that commercial OSM data users with their 
> blatant ignorance of the requirements of the license have successfully 
> moved what is considered normal in the eyes of the OSMF so they have 
> adjusted their own frame of reference for what they may expect from 
> data users.
>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20190909/00045371/attachment-0001.sig>


More information about the talk mailing list