Re: UUNET Network Encryption Patents

[John Gilmore <[e-mail]>]

I mentioned the subject to Rick Adams, the patent holder.  He says
that if anyone has actual prior art, they'll drop the claim.  But it
has to be real prior art in the patent context, not just rumors or
speculation by people who don't know much about patents.  (Such rumors
aren't enough to be worth bothering Rick with, but might lead a
patent-aware person to info good enough to send to him.)

To invalidate the patent, the prior art needs to be prior to March
1992; what counts is the filing date of the original application, not
the grant date.  Uunet started work on this sometime in late 1991.

Per Unell said:
> I Just looked up the oldest version of the requirement specification
> for our KryptoLan system that I could find. It is dated May. 27 1991
> and includes IP and Ethernet encryption and is based on the IEEE
> 802.10 Draft 1 dated december 1989.

You're OK on the date.  Was this "reqirement spec" a public document?
Then the question is how many of the patent's claims it mentions.
Somebody should get the patent texts and figures and put them online.

Rick said:

> ... remember that a combination of publically available pieces is
> patentable as a whole, even if the individual components are not.
> 
> The key component is the decision to encrypt being made on a per
> packet basis based on the contents of the header of the packet. I'm
> fairly sure that Blacker did not allow unencrypted traffic to be
> mixed with encrypted traffic on the same network interface.
> ...
> I'll be happy to look at any well thought out argument, but I'm not going to
> do the digging. Comments like "I'm sure Blacker invalidates this" are too
> ambiguous to chase down.

We can feel free to chase such comments down and provide better info,
if we want.

	John