From: iagoldbe@csclub.uwaterloo.ca (Ian Goldberg)
Date: Wed, 20 Sep 95 22:47:57 PDT
To: cypherpunks@toad.com
Subject: Patents and trade secrets was: Encryption algorithms used in PrivaSoft (fwd)
In-Reply-To: <3060D3B3@hamachi>
Message-ID: <43qu9q$nbd@calum.csclub.uwaterloo.ca>
MIME-Version: 1.0
Content-Type: text/plain


In article <3060D3B3@hamachi>, David Van Wie  <dvw@hamachi.epr.com> wrote:
>
>David Clavadetscher of PrivaSoft writes:
>> At this time our crypto engine is patented and proprietary.
>
>Ian Goldberg writes:
>> Waitasec...  I was under the impression that if you patented it, you had 
>to
>> reveal it.  That's why RC4 isn't patented (it used to be a trade secret).
>
>Many technologies have both patented parts and trade secret parts.  Often, 
>companies will maintain information that is in patent applications as trade 
>secret until they are granted.  I guess I should say _if_ they are granted! 

But don't they have to put something on the patent application?  Can they
claim trade secret status for something that was on a patent application,
but rejected?  That seems like they're getting it both ways.  They should
probably have to choose whther or not they want to show anyone their
"secret".  If not, it stays a trade secret.  If so, it's not a secret anymore,
and they hope it's "nonobvious, etc." enough to be granted a patent.

   - Ian "I heard that 'x*y=[(x+y)/2]^2 - [(x-y)/2]^2' is a patented way
		   to multiply numbers of the same parity.  Can anyone verify this
		   and/or produce a reference?"