Talk:Community attribution advice

From OpenStreetMap Wiki
Jump to navigation Jump to search

Use of attribution in APIs

I think the expectation regarding attribution applies in a similar way to various APIs serving ODbL licensed data. An an example, the following services already explicitly hint towards the ODbL license:

If your API serves OSM data, you should probably include a similar hint as well. Mmd (talk) 16:11, 1 May 2020 (UTC)

Indeed we seem to have a well established practice and probably also the expectation in the community that OSM data delivered by APIs should have attribution included by the software. While such data extracts are not primarily meant for human consumption it is not uncommon for people to look at them and including human readable attribution there is definitely good practice. I am however somewhat reluctant to include that in the advice since OSM data - when of substantial volume - is also subject to section 4.2 of the license which has other requirements. 4.2b - the requirements to include the license text or an URL for example is only followed literally by the API and not Overpass or Nominatim. This has so far not been discussed that much (in contrast to attribution of produced works) so i am not sure if we have clear community expectations here is the details. --Imagico (talk) 16:54, 1 May 2020 (UTC)

Move to "Community_attribution_advice"

"advice" is the noun, "advise" is the verb form. --Jeisenbe (talk) 16:23, 1 May 2020 (UTC)

Ah, yes, thanks. --Imagico (talk) 16:54, 1 May 2020 (UTC)


I just want to register support for it Mateusz Konieczny (talk) 08:17, 5 December 2020 (UTC)

I also support this guidance. I hope a reasonable court in the USA would interpret the ODBL in this way, though I can't confirm if British courts use the same standards of reasonableness. --Jeisenbe (talk) 05:55, 6 December 2020 (UTC)
Glad to hear confirmation that the text seems to largely represent community consensus (which as indicated was my intention). I think the edits that have been made are largely beneficial in improving clarity - including the removal of the editorial remark about getting away with lesser attribution - which is a bit superfluous. --Imagico (talk) 15:44, 6 December 2020 (UTC)

Clarify relation to ODBL

(also posted to mailing list)

One thing that is missing to me is explicit mention that it is not overriding ODBL or related laws and is not adding any legal requirements.

If someone follows ODBL license or is in situation where following license is not needed for some reason, they can legally do this.

Maybe also mention that it is may be recommending more attribution than bare minimum that is required by ODBL, so it is a safe solution that should be also fine for any typical[1] project that is not hostile to OSM?

[1] "typical" - especially for very small objects things gets trickier, if you are making some special purpose map (tactile map for blind) then attribution also needs to be adapted, if map is going to be used in place where English is not understood in general you will definitely need to translate attribution etc etc. claim that smartphone has not enough space for attribution is clearly untrue. But if you show OSM map on screen of size 1cm x 1 cm or similarly tiny physical object then alternative attribution methods - that still comply with ODBL - may be preferable.

Mateusz Konieczny (talk) 08:18, 5 December 2020 (UTC)

I think one thing you have to keep in mind here is that the ODbL is ultimately a contract. The mappers are not formally party to that contract (the OSMF and the data users are) but they are directly connected to it through the contributor terms - they are licensing their work to the OSMF to be published under the ODbL and the OSMF can not change these terms of contract with the data user without consent from the mappers.
And the validity of a contract is grounded in there being an agreement between the contract parties. And what this document is meant to do is to formulate how this agreement should be interpreted from the side of the mappers as parties to the contract (directly in case of the contributor terms, indirectly for the ODbL itself). Therefore i think the idea that this text might be overriding the ODbL or adding legal requirements is pointing in the wrong direction (and stating that it does not implies the same), that direction being that the ODbL text has an abstract meaning independent of its concrete function in the contracts between mapper, OSMF and data users. But that is not the case. Any court that needs to make a decision on the ODbL would need to look at the meaning of the text in context of the contractual situation - which is what this page tries to formulate from the side of the mappers/the OSM community.
The attribution guideline drafts developed by the LWG and OSMF board are doing something different, they are collecting attribution recipes for data users to state that if a data user follows these rules to the letter they are guaranteed not to get into legal trouble with the OSMF. In that case a disclaimer that this guarantee is not part of the contracts between OSMF and data users or OSMF and mappers is prudent of course. But i would strongly advise against trying to turn this page into a similar kind of document. The better approach - and more like the OSM way so to speak as well - is to follow the idea of the ODbL itself and to formulate expectations in the results and this way clarify the intention behind section 4.3 of the ODbL and to allow data users to meet these expectations under their own responsibility.
--Imagico (talk) 16:21, 6 December 2020 (UTC)