Patented algorithms in kernel

[David Woodhouse]

[e-mail] said:
> [e-mail] wrote:
> > If you have a WinModem, then you implicitly have a license to use that
> > patent on that modem.  Otoh, if you want to use it on, say, a spare
> > ISDN channel, you do not.

> Good point.

... and one which brings me to an important question I would like to ask 
Linus.

I have code which implements patented algorithms. As with WinModems, if you 
have the hardware, you have permission to use that code. 

As far as I see it, it's fine to license that under GPL. The GPL explicitly 
states that it only affects "copying, distribution and modification":

	"Activities other than copying, distribution and modification
	are not covered by this License; they are outside its scope."

Thus, there's no problem which distributing GPL'd code that people may not
have permission to _use_ in certain jurisdictions and situations, be that
because of patent restrictions, restrictions on the use of crypto, pure
tinpot dictator insanity, or other reasons. You can GPL the code for 
detonating warheads, you're not prevented from doing so by that fact that 
many uses of it would be illegal.

I have kernel code for stuff which is patented, which I would like to 
contribute for inclusion in the 2.5 kernel - or even 2.4, as it's only new 
drivers. 

The question is: Linus, will you accept that code, or would you reject it
simply because of the patent question. Obviously I assume that it would also
have to meet all the normal criteria for inclusion - but would you even
consider it?

Please note that this is copied to l-k for information only. I am not
particularly interested in anyone else's opinion on this question - _only_
whether Linus is likely to accept the code. Please do not start another
licence flamewar or theoretical discussion. 

--
dwmw2