Re: Network Layer Encryption History and Prior Art

["PALAMBER.US.ORACLE.COM" <[e-mail]>]

 
>Recall that Rick says the important feature is encrypting some 
>packets, based on their header information, while letting others go 
>through unencrypted.  I don't think any of the SDNS stuff that I saw 
>mentioned such ideas; they assumed you *did* want to encrypt and 
>explained the encrypted protocol. 
 
 
No, the "bypass" mode of operation that would allow selective encryption of 
packets was extensively discussed and documented.  Most of the documents were 
internal design memos and reports.  Selective encryption is critical to 
protect large existing systems ... you do not want to be forced to install all 
the encryption devices at the same time.   
 
 
Paul 
 
PS - I left all my patent files at Motorola when I changed jobs.  If someone 
wants to push on the UUNET patent I may be able to find someone at Motorola to 
help.  There is alot of documented prior art in their files... 
 
-------------------------------------------------------------- 
Paul Lambert                     Director of Security Products 
Oracle Corporation                       Phone: [phone] 
500 Oracle Parkway, Box 659410             Fax: [phone] 
Redwood Shores, CA  94065               [e-mail] 
-------------------------------------------------------------- 
  

--- Begin Message ---

> John (Gilmore), is this what you were looking for in terms of real Prior
> Art to take to Rick Adams, so he'll drop the patent claims? 

Nope, these are *pointers to* prior art.  Somebody (who understand
patentese) needs to read the patents, and then read the actual
published materials that Paul is mentioning.  If any of them cover
features that are specifically claimed by the Uunet patents, then
there's a chance that they count as "prior art".  (I'm not up on
exactly what qualifies as prior art.)  We would send the source
documents to Rick or his lawyer, pointing out the prior inventions.

If we get documents enough to mention ALL the claims from the Uunet
patents, then the patents will cease to be a problem.

Recall that Rick says the important feature is encrypting some
packets, based on their header information, while letting others go
through unencrypted.  I don't think any of the SDNS stuff that I saw
mentioned such ideas; they assumed you *did* want to encrypt and
explained the encrypted protocol.  Perhaps some of the implementations
of these NSA protocols supported mixed encrypted/unencrypted traffic,
though.  Their manuals might be valid prior art, if ordinary mortals 
could have obtained them by 1991.

	John

--- End Message ---