Re: what does a patent restrict?

On Thu, 7 Oct 1999, Lynn Winebarger <owinebar[_at_]free-expression.org> wrote:
>
> The truth is exactly what I wrote. Software is a written
> description of an algorithm(s), generally designed so that a machine
> will be able to follow it. (Don't know of any cases where descriptions
> that machines can't follow are considered software, but I wouldn't rule
> it out).

But such would be much less likely patentable, though I suspect that their protection under copyright would not suffer that much.

> The description is not the device. The computer simulates the
> process (or implements it if you prefer) by following the description
> provided by the software. The description is expression (which is why
> it's copyrightable in the first place), protected by the 1st amendment.
> There is no process that is infringing _until_ the computer starts
> executing the expression.

Not exactly correct from a couple of points of view. First, as noted elsewhere by both Andy and me, software may be claimed (and indeed I suggest should be claimed if done competently) as encoded or whatever in an article of manufacture (floppy disk, etc.). Also, claims are often written to claim a system including the software loaded in some sort of memory (disk, RAM, etc.). But technically, you are right, when software is claimed as a "method" or "process", then it typically cannot be directly infringed without being executed.

Also, it is not clear exactly how much software in textual form is protected by the 1st Amdt. For example, it clearly is subject to both copyright and patent protection. Thus, you cannot use the 1st Amdt. as justification in ignoring a patent or copyright.

> > You directly infringe a patent by making, using, offering for sale
> > or selling (collectively, "practicing") THE CLAIMED INVENTION.
>
> Last time I checked, algorithms alone cannot be patented, though
> specific processes they are used for can be. Is my understanding in
> error? As far as I know, you still can't patent expression alone,
> only the process. As I said in another post, contributory infringement
> is a different matter.

Your understanding is a bit simplistic. It really depends on how you define "algorithm". An algorithm is not unpatentable (any more) just because it is an algorithm. Rather, a more complex investigation is required. A fairly accurate (though still a bit too limited) view of the current state of the law (at least as the PTO sees it) can be found at:

http://www.uspto.gov/web/offices/com/hearings/software/analysis/computer.html

-- 
--------------------------------------------------------------------
The preceding was not a legal opinion, and is not my employer's.
Original portions Copyright 1999 Bruce E. Hayden,all rights reserved
My work may be copied in whole or part, with proper attribution,
as long as the copying is not for commercial gain.
--------------------------------------------------------------------
Bruce E. Hayden                      bhayden[_at_]acm.org
Phoenix, Arizona                     bhayden[_at_]ieee.org
                                     bhayden[_at_]copatlaw.com