[OSM-legal-talk] License question, user clicking on map

Jonathan Harley jon at spiffymap.net
Fri Mar 1 10:39:49 UTC 2013

On 28/02/13 14:58, Olov McKie wrote:
> Hello All!

Hi Olov, I'll give this a go. My answers are a long way down because I 
think cases 1-3 are all essentially the same:

> First off, thank you for the feedback I have gotten so far! I had an idea about what answers I would get on my questions, but some of your answers were not what I expected, so let me reason a bit about each case and I would love your feedback on my reasoning. Please also look on case 3 and 4 as no one has said anything about them yet.
> 1. If we present an OSM map to the user let them click on the map and use the coordinates they clicked on as part of the meta-data for a place in our application, will the resulting database be considered a derived database?  To clarify, we would not extract any information from the map, beside the coordinates that the user clicked on, they would by themselves navigate the map to for example London and then click somewhere in London.
> I was expecting this to be OK. If I were to use my old paperback world atlas to find the latitude and longitude of different places around the world, and then store those coordinates along with an awful lot of other information in a database, in no way would I expect whoever wrote that atlas to have copyright claims on my database. I see this as fair use of the atlas and I see the use of an application showing a map where the user clicks on the map as equivalent to an atlas and was therefor not expecting this to be an issue. As some of you see this as copying would I like to ask sub questions:
> 1a. What license would a coordinate extracted this way be under? As the application displaying the map keeps track of the coordinates and normally can display any map layer (OSM and others) and we are not extracting raw data from the database, but just using the rendered view (CC-BY-SA) to help us orient the applications coordinate position to a place we can find on the globe. Will the coordinates extracted from the application be CC-BY-SA or ODbL?
> 1b. If I move the map to a place that is not yet mapped, for instance a small village not at all represented on the map, but I know its location relative to surrounding places, roads etc., then I ask the application for its position, do you also see this as copying of map data?
> 1c. If I use another map layer (NOT OSM) to position my application to a specific place on the globe, then ask the application to change the mapping layer to an OSM representation, then ask the application for the coordinates by clicking on the map, would you consider this copying of OSM data?
> 1d. If I have a printed OSM map of the world and use that to find coordinates for places that I then put in our database, would you consider that copying of OSM data, if so, would the coordinates be CC-BY-SA or ODbL?
> 2. If we use the overpass API to find possible matches for a placename entered by a user, present the possible matches with markers on a map and let the user click on the map and use the coordinates the user clicks on, will the resulting database be considered a derived database?  Again, we would not extract any information from the map, beside the coordinates that the user clicked on. Presenting the markers would of course help the user find a place, such as London.
> I saw this as very similar to case 1, and using the same atlas reference as in case 1 (using the map registry to find the correct page), I would consider this OK. But I am in this case using raw map data to display positions on the map, so it is definitely getting closer to copying data, and I was expecting some people to find this as ok and some as not ok.
> 3. If we use the overpass API to find possible matches for a placename entered by a user, present the possible matches with markers on a map and if we have more then one result ask the user to fill in more details about the place such as, country, region, close to major city, local name, etc until overpass only returns on result, would the user entered data be considered a derived database? To clarify, in this case would we not extract the coordinates or any other data from the map.
> I can not see this as anything but OK, we are not storing any information from the map, just user entered data. But if someone has an idea how this could be considered copying in such a way that the ODbLs share alike clause would kick in, would I definitely like to hear it.

All these cases are of course "OK" in the sense that yes, you can do 
them. The only question here is whether you would you be required to 
release the resulting data for free.

It seems to me that all these points involve presenting your user with a 
map (a Produced Work in ODbL terms), and getting them to give you 
geodata relative to that.

Produced Works don't have to be licensed under ODbL - they don't have to 
have any specific license - so if you use map tiles whose license 
doesn't require any derived data to be shared back to OSM, you are fine 
to collect data and keep it to yourself - as long as you keep it 
separate from any other database (see further comments below). You can 
render the data you've collected as a layer on top of an OSM map or 
anyone else's map if you like. As long as it's derived from a produced 
work rather than a database, and remains separate from anything else, 
OSM's database share-alike doesn't reach it. OSM's produced work 
requirement to credit OSM does of course reach it and anything you use 
it for. (Because something based on a produced work is itself a produced 

> 4. If we present several places (all data about the place including coordinates originates from other sources than OSM) on an OSM map to help find duplicates, and then lets the user click on two places marked on the map, to merge them into one, would the resulting database be considered a derived database?
> I can not see this as anything but OK. A mapping application would solve this use-case with or without a map as a background layer, just visualizing the places on the coordinate grid with a scale present would immediately show duplicates so the OSM layer is only a nice visual touch and I can not see how it would be considered copying of OSM data. But if anyone has a different view please speak up. If you agree with my reasoning, on this or any of the other use-cases, please speak up as well.

Again, if your database remains independent of OSM, you don't have to 
share-alike. But this use case sounds more like you want to merge your 
user-derived data back into a geodatabase based on OSM, removing 
duplicates (and perhaps adding metadata in your earlier use cases?) If 
you modify OSM, or if you derive your own database which is partly OSM 
data and partly user-sourced data - for example with an OSM feature 
removed in favour of a user-sourced feature - then that's a Derivative 
Database and must be made available under ODbL.

So - *must* you make your database of user-sourced geodata available to 
the OSM community? I answer no, so long as it resulted from a produced 
work and you keep it separate afterwards. Anything you use it for is 
just another produced work.

*Should* you make your user-sourced data available to OSM? You don't say 
what it is, but if it's about real-world features (and it sounds like it 
is - particularly the identifying of duplicates) then yes you should. If 
OSM is useful to you, and you're collecting information which might be 
useful to other users of OSM, then why would you not share it with us?

HTH - Jonathan

Dr Jonathan Harley   :    Managing Director    :   SpiffyMap Ltd

md at spiffymap.com      Phone: 0845 313 8457     www.spiffymap.com
The Venture Centre, Sir William Lyons Road, Coventry CV4 7EZ, UK

More information about the legal-talk mailing list