I’m new here but I already have an ambitious idea,
There was a Google Earth KML file published some time ago with a vast wealth of data covering North Korea called “North Korea Uncovered”. From Google searches and forum searches I haven’t seen any indication that anyone’s thought about including it yet, nor does OSM data show any evidence of Containing NK Uncovered.
I contacted the author, Curtis Melvin asking for permission to use the data and he said that he already spoke to someone about it and determined that it isn’t public domain by virtue of the fact that it was created in Google Earth, i.e. using their software and satellite images.
To me it makes no sense, since I’ve used Windows to work on OSM and yet it doesn’t mean Microsoft owns my work, or is part of Google’s EULA?
Any input would be greatly appreciated.
There are two aspects here: Purely legal considerations on the one hand, and project policy on the other.
The legal questions that are usually discussed in similar contexts are whether map data traced from copyrighted aerial imagery is a derivative work limited by the copyright associated with the original work, and whether Google’s Terms of Use impose relevant restrictions on using their services for tracing. OSM contributors who know much more about legal matters than I do have discussed that question repeatedly. The result is usually that tracing and publishing the result might actually be legal in many cases. For example, I’ve found this blog post in the depths of my bookmark archives, if you are interested in these aspects.
However, OSM tends to take a much simpler position: We trace from images only if we are given permission to do so, avoiding the decision whether we would actually need that permission. We have permission from Bing, Yahoo, and various local imagery providers. We don’t have permission from Google, so we don’t trace from Google’s images. The objections against importing data traced from Google would probably be the same as if you directly traced data from Google. This would be an issue even if you could convince the author that he can and should provide his data under a sufficiently liberal license.
Maybe the people on the legal-talk mailing could provide you with more detailed replies (if they aren’t busy with heated discussions about the license change…).
It looks like someone already read into Google’s license on the wiki and as for the blog, my understanding is that if Google filed a lawsuit it would be up to the judge to figure out what’s derivative work or not, which I’m sure is the last thing OSM needs.
The unclear thing is this: theoretically, if one decided to start a North Korea mapping project using Yahoo maps instead, would simply looking at the “Google Earth derived map” for a hint as to where to look constitute personal knowledge or copyright infringement? A better question is as follows: what level data indirection is accepted by the OSM community? E.g., is personal geography knowledge derived from using Google Earth ok to use? Is it only acceptable to use knowledge if one forgot where it came from? Where does OSM draw the line?
If no one knows I can always take it to the legal mailing list but it would certainly help to know!
And thanks I just hope mapping isn’t too addicting…
Hello Cheburashka, did you manage to get replies from someone else yet?
Looking at copyrighted content to find out where you need to look for something interesting, but retrieving or at least independently verifying all data that ultimately ends up in OSM from a legally clean source, is widely accepted by the community. For example, one of the frequently cited uses of popular map comparisons like http://tools.geofabrik.de/mc/ or http://sautter.com/map/ is to find out where data is incorrect or incomplete in OSM, usually in preparation for a survey on the ground.
Personally, I also assume that adding facts from memory should be fine, regardless of original source, simply because it most likely doesn’t constitute a substantial extract. Of course, that’s only true as long as you do not intentionally and repeatedly memorize numerous facts from a protected map database simply in order to add them to OSM on a large scale. Distinctions like this tend to be rather fuzzy, perhaps inevitably so.
