Subject: Re: Is free software innovative ?
From: "Stephen J. Turnbull" <turnbull@sk.tsukuba.ac.jp>
Date: Tue, 12 Dec 2000 20:44:29 +0900

>>>>> "David" == David Fetter <shackle@fetter.org> writes:

    David> Bernard Lang wrote:

    >> Innovative means : containing stuff that could have been
    >> patented, according to a rather strict application of the novel
    >> and inventive rule.  (not the USPTO garbage criteria)

    David> OK.

    >> - gzip : compression utility, privately developed

    David> A slight improvement over compress.  Innovative?

(Aside: How about bzip2 then?  Also a compression utility, privately
developed.  With a rather different algorithm.  Don't "process
innovations" -- doing the same old thing but better or cheaper --
count?)

Etc, etc.  David goes on to propose Perl as innovative, but where's
the patent in it?  If the techniques that perl brings together in one
interpreter had been patented by their originators, isn't it highly
likely that perl would run a net deficit in royalty payments?  :-)

Some time back Peter Deutsch mentioned a potentially patentable
technique, I suppose developed in connection with Ghostscript.  If he
doesn't pop up here, you might ask him directly.  I suspect he could
probably give you a list of innovations in Ghostscript that were _not_
patented only because he wasn't looking for patentability.

Of course, I don't know that any of these meet Bernard's "strict
criterion".

    David> A very distinct pattern that's (omigosh) contrary to
    David> libertarian dogma is emerging here.

This would only be true if you could go down the list and demolish the
innovativeness of proprietary software, as well, I should say.

I'd like to see a (short) list of proprietary software that is
considered "innovative" (or not) for the sake of comparison.

Linux may not have any single "strict-criteria patentable" idea in it
(although surely it has hundreds by the USPTO's standards), but the
fact that it keeps up with heavily patent-based and well-funded OSes
in many applications surely indicates something.  Still I imagine
there must be patentable ideas in the various file system
implementations, and down in some of the device drivers and so on, and
in the several implementations of VM etc that have been tried.

Also, one thing we need to be careful about is the idea that a lot of
innovation in free software may not involve "patentable" software
techniques as such (by Bernard's strict standards), but may be
"emergent" from the ability to combine free software modules.  I
misdoubt that Apache's mod_* embedding of everything that begins with
"P" (perl, python, php, pgp, ...), and then some, would qualify on the
software level as "patentably" innovative, but that was one hell of a
great idea and has really paid off in interesting applications.

This business of rearranging common parts in perhaps not terribly
"innovative" ways, but producing great value, certainly doesn't
satisfy Bernard's request.  But maybe we have to attack the premise
that "creativity" and "innovation" can be measured by patentability,
or even a stricter criterion of "novelty and inventiveness."

In that connection, I'd like to point out that in economists'
discussions of "innovation", "inventiveness" hardly gets any play.
(Yes, yes, that's mostly because we can't get a grip on it.)
Innovation is defined instead as adoption of a technique before the
pack does, whether it's considered highly "inventive" or not.  So if
free software developers are busy picking up all the neat little
_obvious_ ideas that are just lying all around them, they can be
_creating_ a lot of value by _innovation_ without budging the ol'
patentability-meter.

-- 
University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences       Tel/fax: +81 (298) 53-5091
_________________  _________________  _________________  _________________
What are those straight lines for?  "XEmacs rules."