[OSM-dev] Drawing ways with beziercurves rather than linesegments

Robert Hart Robert.Hart at BuroHappold.com
Thu Jan 25 17:05:13 GMT 2007

> You're missing the point.

> Ignorance is no defence... this is true.
> That isn't to say ignorance isn't an advantage.
> What you have to understand here is that the level of knowledge you
would > need to "understand all relevant patents" is way beyond anything
> going to manage if not a full-time army, that the work in coding
> every single possible infringement is way beyond all possible hope,
> the lawyers required to actually achieve any useful narrowing of this
> way, way too expensive for this project. 
> And ignorance of a patent is an extremely useful piece of ammunition
if you 
> stand any chance of fighting the validity of a patent. It helps imply
> patent may have been obvious. And from the other side of the coin, as
> as you start getting a full knowledge of the patent space, you cease
to be 
> an untainted mind, and are extremely unlikely to be able to prove you 
> didn't get your idea from a patent itself. 
> And you will miss things, which puts you in the same position as we 
> originally started with, just after spending thousands of hours coding

> round potential problems, which due to that lack of experts in the
> may have been unnecessary anyway.

> You can ask Yahoo, Google, NavTeq, TeleAtlas, whoever, to give a
> license of their portfolios to OSM if you like. They'll probably say
> or ask to negotiate with your lawyers. For each one that says yes
> less patents to not know about.

So basically your argument is that putting your hands over your eyes and
yelling "blah blah blah blah" at the top of your voice is going to make
the evil monsters go away?

My GPS came with a sticker on the front listing about 45 patent numbers,
and I assume everybody else's did too. I can hardly claim ignorance of
the possibility that some of the stuff it does could be covered by
somebody else's IP. I have used Google Earth, Tom Tom, Multimap, and so
on and so forth, and I'm smart enough to guess how a lot of it works.
Reading the actual patents isn't going to change anything.

The only reliable way to make a patent go away is to show *prior* art,
not independent art! I work with beziers and NURBS all the time, it
would take me about 10milli-seconds to "invent" a mapping system that
used them, but unless I can show I did it before the application date of
whatever patent, it ain't going to make much difference. Sure you *can*
appeal a patent on grounds of it being trivial, but it should have been
assessed on that regard already, so you will have a much harder time
than you think. (IANAL blah blah)

That said, there are definite advantages in not knowing about how other
people do things, because that does, in my opinion lead to innovation.
But OSM is well on the way to being an army - lots of people here are
already very knowledgeable about the way various aspects of IP involved
here. Nobody needs to know it all, and most of us don't need to know any
of it.

Spending 1000 man hours coding around a potential problem would be well
worth it, if it protects 100,000 man hours of mapping!

This message has been scanned for viruses by MailControl - www.mailcontrol.com

More information about the dev mailing list