Subject: Re: patent trolls and X-licensors
From: "Ben Tilly" <btilly@gmail.com>
Date: Mon, 5 Jun 2006 14:50:05 -0700

On 6/5/06, simo <s@ssimo.org> wrote:
> On Mon, 2006-06-05 at 13:56 -0400, Forrest J. Cavalier III wrote:
> > Ben Tilly wrote:
[...]
> The problem with patents is that economically they make sense only in a
> market where you need huge investments for any kind of development.

I'm not sure that I believe this.  My understanding is that the patent
system makes the most sense when dealing with ideas that are difficult
to come up with but cheap to copy.  An ideal example would be the
zipper - developing a good zipper took one person close to 20 years,
but once you see one it is easy to duplicate the design.

Even though the actual research investment was not particularly huge
(well it was big for one person, but not for a company), I'd cite this
as an example where the patent system did what it was supposed to do.

Ironically by this statement of what patents are supposed to protect,
the Amazon One Click patent is actually a *good* patent.  After all
given the state of the art that existed at that point of time, it was
not obvious.  (Useless trivia, the programming team that was supposed
to build the one-click system assumed that the lack of a confirmation
screen was an oversight in the specification and built one anyways.
They were quite surprised when they were told that it was not a
mistake, and the extra screen should be removed.)  But as soon as
you've seen it, it is immediately obvious why it is a good idea, and
it is trivial to copy.

Whether patents like this accomplish a social good, in particular
whether theyl "promote the progress of science and useful arts" is
another question entirely.  (That question is particularly relevant in
the USA, or would be if people paid more attention to the
Constitution.)

> In software that's simply not true, it is very easy and cheap (compared
> to any other sector) to find new ideas and to code them up.
> The hard part is QA and making your software robust, a thing that
> patents can only worsen, because when you have a monopoly you have no
> incentives at getting better, only direct competition can achieve that.

The economic shelflife of software strongly suggests  that, if nothing
else, the duration of a traditional patent is not limited in any
useful sense.  Which strongly   I also believe that this pace will not
let up until some time after Moore's Law breaks.

Cheers,
Ben