Patents and OpenMoko

Nils Faerber nils.faerber at kernelconcepts.de
Wed Feb 13 10:54:53 CET 2008


Jon Radel schrieb:
> Vasco Névoa wrote:
>> Hi. Sorry to barge in like this, but I don't quite understand the problem to begin with...
>> Isn't open source code by definition protected against subsequent patents?
>> It is part of the patenting process to search for conflicting publications; if they find any, then the candidate idea is not a novelty and cannot be patented. Publishing is the best weapon against (subsequent) patents: cheap and effective.
>> I think we should just add some way to automatically timestamp every code check-in in a legally binding way, like using some outside certification entity's digital signature (that carries a legally recognizable timestamp).
>> An open-source public repository is a valid publication of ideas, which are therefore not patentable.
>> What do you think?
> 
> Bogdan Bivolaru already pointed out some practical issues with your
> theoretical outline.  However, there are some additional issues:
> 
> The biggest in my view is that you seem to assume that open source
> developers somehow (magic?) manage to write only code which does not
> infringe (in somebody's eyes) on existing patents, or won't infringe on
> already filed patents that haven't been published yet.  There are even
> legal reasons to avoid doing a patent search before you start developing
> something, as you then avoid knowingly infringing, which makes a
> difference, at least in the US.

How do you want to proove that either you did or did not do that
research? This is quite pointless.
Of course it makes a difference if you knew about it or not but you can
perfectly well do an anonymous research yourself without leaving any
proof (and it is agood avise to do so in order not to start large
developments in areas that are already a patent mine field - been there
done that in the area of motion detection and picture recognition and
subesquently stopped certain developments).

> Then there's the practical matter that if you have a collection of
> patents you can frequently come to some cross-licensing agreement if
> someone else in your industry starts hassling you with their patents.
> If you have nothing, it's much easier for them to grind you into the
> dirt, if only with legal fees, if they so desire.

Here some more resonability would be good.

So let's think, who are the most likely players that could sue a company
like OpenMoko working in the mobile phone area? Other mobile phone
makers. Who are those? Nokia, NEC, Samsung, Qualcom, Lenovo, HTC,... All
those companies are magnitudes larger than a start-up like OpenMoko.

The cross licensing trick goes like this: You infringe on somebody
else's patent (knowingly or not aside). They come to you and threaten
you with sueing you - you shiver and fear they are taking you out of
business. Now you go down your cellar and browse your patent pool and
look at the other's products if they could possibly by coincidence (<-
this is the point here!) infringe on one of your patents.

So now step back and calculate the probably that the other party really
infringes one of your patents? If there is Qualcom vs. Nokia then there
is a good chance but those two do this business for decades and are huge
companies. For a small company even if you are able to find an infringed
patent they come up with at least a dozen more of your infringements.

As a small company you will loose, no way around it and the dream of
cross licensing is for small and new players just a dream. It is
unrealistic to assume that it could work.
For small businesses the only way to benefit from patents is to have a
real new invention and license it to some bigger player and make money
from it.

To the special case of OpenMoko as long as the GTA devices are concerned
they have no single piece (sorry, no offense!) of great invention in
them that would be a good candidate for a patent. And even if you can
probably construct, with a lot of extra "inspiration", say, 10 patents
on soft- and hardware? Maybe 20? But that's about it.
And now go back to those big players and count their's - IBM is well
known for being patent record holder, they file more than 10 patents - a
day!

So is cross licensing for a company like OpenMoko (and similar)
realistic? I think not.
Should someone like OpenMoko aim at doing so? I also think not. Do not
feed the troll.
The patent system is broken, especially regarding software. It could be
fixed but nobody has until now made any good proposal. Why? Because the
patent system in itself has become a huge business and doing a reform
would mean that many people loose a lot of profit - starting with lawyers ;)

So my advise is: Stick with the very old patent model. If you have a
really cool new invention, go for a patent! But this should be really
cool and really new technology, especially hardware. Be very careful
with software! I would advise not to touch software patents at all.
If you are behind patents just for the cross-licensing, forget about it.

This ultimatley leads to the bad situation we have today that everybody
tries to patent *something* only to have one more patent and the result
is devastating, just look at
	http://webshop.ffii.org/

(an example for rediculous patents that a typical webshop would infringe
- only a very primitive example)

> --Jon Radel
Cheers
  nils faerber

-- 
kernel concepts GbR        Tel: +49-271-771091-12
Sieghuetter Hauptweg 48    Fax: +49-271-771091-19
D-57072 Siegen             Mob: +49-176-21024535
--





More information about the community mailing list