Hi, after reading various comments and OSM-wiki pages I'm still not sure how to credit sources correctly. The instructions etc. mainly state that in general the source should not be indicated with an object but with the changeset. This, however, in my view conflicts with licenses such as http://www.govdata.de/dl-de/by-2-0. They ask to indicate the source, which would often be for an outline (source:outline) of an object plus additional tagged information from another source (e.g. knowledge). A changeset might be derived from a mix of sources. So I would either need to a) add source:outline=* to every object (discouraged in the wiki) b) upload changesets derived from a single source only c) add all sources to a mixed changeset, in which case the credits would not be unambiguously linked to the objects, which might violate the licensing requirements. So how do I do this correctly? asked 09 Jan '20, 10:13 UliFR |
Hi UliFR, you have understood correctly, that sources cannot be attributed to individual objects, even less so if the data is being processed. That is one of the reasons why we must not use any data where the source requires such attribution. Information licensed under Datenlizenz Deutschland – Namensnennung must not be used by you to modify OSM data. We can only use information that can later be released under the terms of our OSM license. Have a look at the wiki for potential data sources. answered 09 Jan '20, 10:29 TZorn Hi TZorn, I looked up https://wiki.openstreetmap.org/wiki/DE:Potential_Datasources. There the statement is that the DOPs should not be used because of restrictions on commercial use (don't know if you know German, sorry). The sources of interest I'm thinking of, however, state the license I mentioned website, which explicitly allows the use for commercial purposes. Here's the relevant paragraph (1) from that site: "The data and meta-data provided may, for commercial and non-commercial use, in particular 1) be copied, printed, presented, altered, processed and transmitted to third parties; 2) be merged with own data and with the data of others and be combined to form new and independent datasets; 3) be integrated in internal and external business processes, products and applications in public and non-public electronic networks." Another site also explains how modified data should be credited under the same license (on a popup window opening on https://www.geoportal.rlp.de/) by addding "[Daten bearbeitet]" as a statement. So the argument on the wiki does not hold. If the license allows the use of the data even in processed form for commercial as well as non-commercial purposes, giving due credit, why should that not be allowed for OSM? In fact, for a case that is listed as permissible on said website (e.g.WMS Maps4BW), I see that numerous objects individually credit this source as source:outline=Maps4BW. For Thüringen the wiki has another (for me) ambiguous statement: on the one hand said license is stated as incompatible with OSM, on the other that aerial images can be used as background layers in JOSM. The latter would be my use case. I'm not talking about importing the data directly. I'd appreciate a clarification for this specific license.
(09 Jan '20, 11:16)
UliFR
1
@UliFR 1) pls do not use answers to refine or supplement your question , either amend the original question or add a comment. 2) the licence in question allows the licensor substantial leeway in their attribution requirements, in some cases we have received explicit permission for use, or the licensor has explicitly said our way of attributing is OK, in other cases permission has not been given.
(09 Jan '20, 11:31)
SimonPoole ♦
2
I think the argument on commercial use is indeed misleading or even wrong. The main reason why we cannot use the DOP is that the providers require attribution of the source (Namensnennung), which we and 3rd parties that use our data cannot provide for. The attribution might somehow be able to be fitted to an object in the database (and even that is going to be difficult once an object is changed, split, merged,...) but how do you want to accomplish that on a rendered map for example? On top of that OSM would have to change its own licencing terms would we want to use those data. Some Länder and other data providers gave explicit permission for OSM to use the data even if it would otherwise be not permissible due to general license. NRW will waive the attribution starting from March moving to Datenlizenz Deutschland Zero. For Maps4BW have a look at https://wiki.openstreetmap.org/wiki/Maps4BW#Genehmigung_zur_Nutzung where it is stated that BW is satisfied if the attribution happens once on the OSM main site and not on any derived work. btw: yes, I do speak German.
(09 Jan '20, 11:37)
TZorn
IMHO as a beginner (and in the wrong forum, as someone will probably say), and a scientist producing original material, I think it is absolutely good practice to credit material received from a particular original source (being a scientist myself). OSM is missing out on opportunities for great data sources, with information often not accessible otherwise, that these organizations are willing to give for free - just because OSM cannot credit them adequately. That's really hard to believe.
(09 Jan '20, 12:32)
UliFR
3
@UliFR you are misunderstanding the implications of the licence. While you can argue about if the source should be mentioned on the object or not, that is not the main issue. The problem is that the licence requires downstream attribution, that is on every map, geocoding result, route etc etc etc that is produced from OSM data containing data from the source, and that simply isn't possible both from a practical and from our distribution licence point of view.
(09 Jan '20, 12:59)
SimonPoole ♦
2
@UIIFR the issue is that any data in OSM can be substantially edited and changed by anyone so: a) we have no means of ensuring that a source attribution is adhered too; or b) that the data in OSM bears any relationship to the original source. Many official data sources contain inaccurate or outdated information so these are not trivial problems. As @TZorn says the solution is to ask for derogation for OSM for the attribution requirements. Equally OSM is not designed to provide a full audit trail of sources material which you might expect as a scientist: it would be far too onerous for ordinary individuals and vitiate the notion of anyone being able to contribute.
(09 Jan '20, 13:00)
SK53 ♦
showing 5 of 6
show 1 more comments
|