Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: henry@utzoo.UUCP (Henry Spencer)
Newsgroups: net.legal
Subject: Re: RSA cryptographic algorithm patented?
Message-ID: <5799@utzoo.UUCP>
Date: Thu, 18-Jul-85 14:55:32 EDT
Article-I.D.: utzoo.5799
Posted: Thu Jul 18 14:55:32 1985
Date-Received: Thu, 18-Jul-85 14:55:32 EDT
References: <9028@ucbvax.ARPA>, <468@stc-b.stc.UUCP>
Organization: U of Toronto Zoology
Lines: 47

Doubtless the legal eagles will correct me if I blow this, but into
the fray anyway...

> 	1 algorithms cannot be patented, hence the use of copyright
> 	  and/or trade secret law to protect software.

The theoretical intent of patent law is to protect inventions, i.e. new
and useful ways of doing things.  At least some algorithms clearly ought
to qualify.  In reality, the situation is rather unclear.  Scientific
discoveries, i.e. laws of nature, definitely are not supposed to be
patentable, and there is a problem of definition here:  just what exactly
are algorithms?

In practice, you can probably patent an algorithm which is expressed
as a hardware implementation (it's been done), and you might be able to
use that patent to sue people who implement the same thing in software.
I would guess that this is the situation for the RSA encryption scheme.

One reason why copyright and/or trade secret law are the primary means
for protection of software is that most pieces of software are not
innovative enough to qualify as inventions.  They are routine exercises
of technical skill, rather than striking new inspirations.  Another
reason is that the murky situation regarding patents on algorithms has
scared people off; trade secrets don't have that kind of legal confusion
surrounding them.  A third reason is that our patent system has been
allowed to deteriorate to the point where it no longer provides much
protection to inventors, and trade secrets are once again on the rise
as the only effective protection method (patents were originally devised
to make it possible to protect inventions *without* keeping them secret).

>	2 software independently developed to perform a given function
>	  is the property (copyright etc.) of the developer to use as
>	  he sees fit, and it is only that based on someone else's 
>	  *CODE* which is open to legal attack.

"But I did it myself, borrowing only his idea" is insufficient; it is the
idea itself that is protected by patents, not the details of its expression.
(Copyrights, on the other hand, protect form of expression rather than
underlying concepts.)  In fact, completely independent invention of the
basic idea is not a defence against patent infringement either.  Once the
inventor has a patent on his invention, it is *his* *property* until the
patent expires.  Exclusively his.  You are thinking of things like the
Unix licence, which are based on trade-secret law; independent invention
*is* a defence there.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry