I've been following the CC-ODbL license discussions for quite long time and I have persistent question that I've been meaning to ask. The recent lively debates on osmf-talk that have spilled over here prompted me to ask now.
Why do people believe that there no creative copyright in OSM data (i.e., why is CC-BY-SA supposedly indefensible for OSM data)? I'm talking about the US-type of copyright that is based on sufficient creativity, and not on the sweat-of-the-brow copyright that is part of UK IP. The argument is that geodata, being factual, is not creative and therefore not afforded copyright protection. That is certainly true for individual pieces of geodata, like each and every node of a road. But I would argue that a selection of a finite set from an infinite possible nodes that can represent the centerline of a road is a sufficiently creative endeavor that is automatically afforded copyright according to the US copyright system. Therefore, the set of nodes that represent a particular road in OSM is a creative output and is copyrighted. By extension, the OSM database, that consists of such creative selections of road nodes is also copyrighted. As a comparative example, the fact that Obama is the 44th President of the United States and the fact that Obama is the first African-American to become U.S. President are two uncopyrightable facts, but those facts can be represented in many creative ways. And you can mix and match that with other facts. In the same way, while the factual data (e.g., position) attached to individual nodes is uncopyrightable, particular selections of such nodes, especially when the selection process is sufficiently creative, should be copyrightable. Sure, the copyright afforded might be "thin copyright" [1], but I don't think this matters because anyone who tries to derive a proprietary database from OSM data by relying on the underlying facts would be essentially be doing what Richard Fairhurst mentioned in his popular blog post [2] and is therefore allowed. Not to mention that they would have to expend a considerable amount of effort to avoid copyright infringement (by selecting a different set of nodes to represent a road) that they might be better off doing their geo database from scratch. [1] http://www.ivanhoffman.com/scenes.html [2] http://www.systemed.net/blog/?p=100 Eugene Villar (OSM: seav)
_______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk