[OSM-legal-talk] Houses of cards

[Richard Fairhurst]

Rob Myers wrote:

> All can fail at once is my point.
>
> If I am a US college student with a Tor client, an SSL connection  
> and a Hotmail account, I can grab the OSM tarball, remove the ODL,  
> and upload it to archive.org or a torrent host.
>
> - Copyright doesn't apply because it's data.
>
> - Database right doesn't apply because I'm in the US.
>
> - Trade secrets don't apply because the data is publicly available.
>
> - I have violated the contract but I am completely anonymous and  
> untraceable. Nobody can find out who I am to do anything about this.  
> And OSM would want copyright restored not damages anyway, which  
> isn't possible.

In which case a) we've done our best, b) we still have the  
"norms"/"peer pressure"/"negative PR" (call it what you want) to name  
and shame the offenders.

As you've noted, the most liberal regime is the US. Happily, the  
chance of any infringement there is much, much less because TIGER  
exists and is PD. A risk analysis on the part of the would-be  
infringer would probably say it'd be easier just to take/improve TIGER.

If not, you then have the issue that the contract "chain" has been  
broken - which is not ideal, but a) and b) above are still relevant,  
ODC-Database does go a small way towards addressing this in 4.7, and  
we could always organise a hit squad.


Incidentally, I'm not sure that "Copyright doesn't apply because it's  
data" is necessarily the full story in the US. The data may not be  
copyrightable _but_ the arrangement could be if it's sufficiently  
original. In Feist vs Rural an alphabetic listing of names and numbers  
was, unsurprisingly, deemed unoriginal. But OSM - well, how many other  
freeform key/value collaborative geodbs are there? :)

(see very roughly  
http://www.law.duke.edu/journals/dltr/articles/2001dltr0017.html)

cheers
Richard