2009/10/7 Anthony Cartmell <[email protected]>: >> If I assign latitude and longitude to postcodes by using a map in >> which copyright subsists (but which does not have those postcodes) >> then, unless I am copying some element of the map in so doing (which >> is unlikely with users pointing and clicking) I very much doubt that >> the copyright could somehow attach to the postcode database I then >> generated. Not least because it is unlikely that copyright could >> attach to it. > > According to the OS Licensing team (who I now understand to be the Licence > Sales team) "tracing" information from a map creates derived data from > that map: I suppose the argument is that you are using OS's hard work to > locate your point. And OS-derived data, as we all know, is subject to > Crown Copyright and licence fees. Of course there must be a limit: if you > record a GPS tracklog for a bike ride that you navigated using an OS map, > that probably isn't derived data, although some paranoid people think it > might be. In my opinion you're deriving a line of points from the roads > that exist, not the OS map.
Clearly I am the world's worst communicator and should think several times before contributing anything about law to this list. My assumption (which was explained in what I said) was that there would be no tracing of features - that is no copying of the map. Having read Ed Parsons's blog post on the matter I can see that I am not the only one who has trouble explaining this distinction and I guess that a non-expert taking the very simple approach of "do not touch OS maps" is being suitably cautious. Just a pedantic point - but its worth getting these things right - there is no copyright in *data* per se, copyright subsists in works, be they maps, musical performances or books. Data may represent a work in which copyright does subsist so that copying the data is a copying of the work (because copies of a work can be made indirectly). The database right is really different from copyright and something else again, its important to keep the two separate in one's mind because the database right lasts only 15 years, has different rules on subsistence and infringement and so on. Both sets of rights may operate at any one time (and often do). There are difficulties with drawing lines here (in the legal metaphorical sense) and its expensive to trace them out by taking matters to court, hence confusion abounds. I had hoped to throw some light on the subject but fear I have failed to do so. > > But perhaps it _isn't_ derived data at all: no-one really knows. The > famous AA case was clearly mass-copying of mapping, and a breach of OS > copyright, but no-one has ever been taken to court for smaller instances > of "derived data" publication without a licence from OS as far as I'm > aware. Nor I. If anyone does come across something like that, I'll be interested to hear it. -- Francis Davey _______________________________________________ Mailing list [email protected] Archive, settings, or unsubscribe: https://secure.mysociety.org/admin/lists/mailman/listinfo/developers-public
