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

[Simon Poole]

Am 01.11.2019 um 19:03 schrieb Christoph Hormann:
> ...
>>  - the attribution guidelines should not require more than the law or
>> attempt to prohibit things the law allows because the ODbL itself
>> notes that certain rights are granted by law and the ODbL does not
>> interfere with them - (if I understand you correctly, Christoph, you
>> disagree with me on this point)
> Indeed.  My opinion is that the attribution guideline as a guideline
> about one specific requirement *of the ODbL* - the attribution - should
> not cover any explanation of the practical effects of the law that are
> generic, i.e. that would apply equally to ODbL data and non-ODbL data.
>
> More generally the attribution guideline should IMO not prohibit
> anything.  It should tell people under what circumstances attribution
> is required by the ODbL (essentially whenever a substantial amount of
> ODbL data is publicly used) and how people should design this
> attribution (i.e. explaining what "reasonably calculated to make any
> Person [...] aware that Content was obtained from [...]" practically
> means).  And as i said elsewhere these explanations should IMO be
> conservative and err on the side of caution, i.e. what is recommended
> should be clearly compliant with the ODbL.

We are not really doing anything else,  but that has to include giving
pragmatic guidance that doesn't create unsolvable catch 22 situations
for our users, for example in the case of using multiple sources which
as far as I can see remains the major bone of contention.

Consider for example using OSM data together with CC BY 2.0 sources in a
Collective Database, any requirement that the OSM attribution be more
prominent than the attribution of the other sources would create a
conflict and essentially not allow this to be done, the only attribution
solutions that will enable this are ones that display all sources equally.

>
> If the LWG and the board want to invest a significant amount of
> additional work to provide guidance to data users for some selection of
> jurisdictions on data use without the need to comply with the ODbL that
> would IMO be something for a separate "when you don't have to abide by
> the requirements of the ODbL" guideline.  This makes much more sense to
> be separate from the attribution guideline because it obviously also
> applies to any other constraints of the ODbL.  If for example some
> forms of showing an OSM map in a movie are fair use and therefore don't
> have to comply with the ODbL in the US that does not only mean there is
> no attribution requirement, it also means that such map could be made
> from a derivative database without share-alike.  In a way this would be
> an extension of what the "Substantial Guideline" currently covers.

There is no intent to do anything of the sort for the general case (we
are not a law firm). That doesn't mean that we can't take the legal
underpinnings of the ODbL and our own guidance in to account  (consider
the issue of insubstantial extracts and how that could effect
attribution and that we have thumbnails in use in the wild ).

Simon


-------------- 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/osmf-talk/attachments/20191102/53115673/attachment.sig>