[OSM-talk] It's all over! (patent fun)

David Earl david at frankieandshadow.com
Tue Nov 6 15:06:42 GMT 2007


On 06/11/2007 14:41, Frederik Ramm wrote:
> Oh dear,
> 
>     Navteq have a patent on driving around with a car and recording  
> the track:
> 
> http://patft.uspto.gov/netacgi/nph-Parser? 
> Sect2=PTO1&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch- 
> bool.html&r=1&f=G&l=50&d=PALL&RefSrch=yes&Query=PN%2F6674434
> 
> We're screwed. Oh well. Maybe we can at least patent the same for  
> roller blades, and if that's already taken, maybe for roller blades  
> using only one leg?

On the one hand, the claims aren't car specific (it's just the abstract 
that says "vehicle", the claims simply say "obtaining source data" 
without specifying how).

But on the other, the claims only cover reducing the amount of data by 
maximum chord distance from the track. The patent claims, which are what 
matter, aren't about collecting data but about reducing its complexity. 
But it doesn't say "automated" anywhere, so if you trace a track in JOSM 
omitting intermediate points where they are in a sufficiently straight 
line, maybe there's a potential infringement. And the plugin for 
converting traces to ways is particularly vulnerable.

Claim 3 is also about sampling roads while maintaining a desired level 
of accuracy.

Given it was filed in 1999, they've clearly been hammering away with the 
patent examiner, who has whittled down an original massively broad 
claim, to this relatively lesser one.

Nevertheless this is still so broad and obvious I'm surprised even by 
the lax standards of the US Patent Office that the examiner allowed this.

David




More information about the talk mailing list