From: David Van Wie <dvw@hamachi.epr.com>
Date: Wed, 20 Sep 95 23:43:44 PDT
To: "'cypherpunks'" <cypherpunks@toad.com>
Subject: Re: Patents and trade secrets
Message-ID: <3061091A@hamachi>
MIME-Version: 1.0
Content-Type: text/plain



>> After a patent is granted, it is usually a good idea to also maintain 
some
>>trade secrets in your products -- since trade secrets never "expire," 
unlike
>>patents.  If the patent isn't granted, you still have the option of 
treating
>>the contents as an intellectual property under trade secret protection.
>
>Rubish, disclosure is required for a grant of a patent. Unless someone
>skilled in the art can duplicate the invention from the patent claim
>you don't get a patent issued.

Sure disclosure is required.  There is no requirement; however, that an 
invention be your _whole_ product.  For example, most automobiles have 
thousands of patents involved in their creation.  It is entirely possible, 
even commonplace, as I said in my mail, that one or more portions of a 
product represent practice of patented inventions, and one or more _other_ 
portions represent trade secrets.

>Trade secret protection is very tricky in any case. Its practically
>useless if you want to protect a product rather than a procedure.

That depends.  Obviously, trade secret protection can be very effective for 
processes involved in manufacturing physical goods.  In software, it depends 
on whether what you are treating as a secret becomes widely known (after 
which, self-evidently, it is no longer a secret!)  For example, if one where 
to keep the mathematics behind MD5 a trade secret, it is plausible that it 
would never be "figured out" just from examining object code that implements 
the algorithm.  It doesn't matter if it theoretically could be done, just 
that it hasn't actually happened.

dvw