Re: A reply on the RTLinux discussion.

[Jeff V. Merkey]

I've been reading all this discussion, and I think the solution
is pretty simple.  Patents are fairly easy to invalidate if you
can show prior art.  Novell and these big software companies 
do it all the time.  When they implement something that infringes
someone's patent, they wait until litigation is filed, then seek to
invalidate specific claims in the patent.  There are administrative
procedures in place wih the USPTO that take this into account.  It's
expensive and you have to be willing to risk litigation.

Patents describe "methods".  If you alter the methods, however slightly,
it makes it tougher for the patent holder to win an infringement case.
Based upon the whirlwind of discussion on this topic, it would seem 
that there is significant deviation from the patent claims to 
circumvent the probability that such claims would succeed.  

The bottom line is you can get sued anyway.  Patent cases are pretty
tough to defend, but the only test will be to implement it, then 
wait for the patent holder to bring claims in US District Court and 
attack the basic claims in the patent on the basis of prior art.  

This RTLinux patent appears, at least on the surface, to be another
software "trash patent".  Microsoft has thousands of such patents,
and it's questionable they will ever be able to win enforcement on 
many of them.  Ditto this case.

Jeff