Subject: Re: Tim's paradigm shift
From: "Stephen J. Turnbull" <>
Date: Mon, 12 Jul 2004 13:49:31 +0900

>>>>> "Tim" == Tim O'Reilly <> writes:

    Tim> I completely agree.  That's been the rallying cry I've tried
    Tim> to use in the talks I've given on the topic.  But I also
    Tim> believe strongly in the lesson that I try to bring out in the
    Tim> article, that it's not in appeals to volunteerism or the idea
    Tim> that information *ought* to be free that the solution is to
    Tim> be found, but in architecting free and open source systems in
    Tim> which the architecture promotes the creation of open
    Tim> collective works on the data level as well.

But how do we get people to use them?  Jamie Lokier gives a salient
example: access to travel data.  I have a friend who runs a B&B, and
he says that the last thing he wants to do is share his customers'
experience with other B&Bs, and especially not with the chain motels.
That's what puts bread on his own dinner table, not just the meals and
the quaint old farmhouse.  He provides an Internet kiosk, and it works
just like IE: it's in your face as soon as you log on, and it has a
reasonably convenient interface to the things he wants you to see---
especially the form where you can contribute advice/reviews of local
attractions.  For his customers to contribute to a global database
would require _effort_.

Of course, a third party (say Lonely Planet or even Triple-A) could
start up a wiki-style site, but this would depend on volunteerism/ego
on the part of the travelers.  Amazon depends on the same thing to
collect its reviews.  But just like my hotelier friend, they will be
pretty fierce about locking up the the data.

Is this a problem?  The static monopolies we can usually live with, as
you point out for CDDB.  The problem is that the reason why the
_world_ (as opposed to a rival business) wants a fork is that some
rival has (a) a better way to present the data or (b) a useful way to
combine this database with another or (c) some brilliant idea that I
as yet lack the imagination to present here.  Will our database
monopolist implement this idea?

A lot of papers in the literature on incentives and R&D say _no_.  One
thing that comes up over and over again is that IP gives monopolists
disincentive to innovate at the same time as it provides resources
with which to innovate.  (Don't get me wrong, the jury is still out,
even in the economic theory community, but that's where we would place
bets for being the received wisdom in undergrad texts 25 years hence.)

And sometimes the database does matter all by itself.  Cf. the WestLaw
mess, where they claimed property in the only set of electronic index
numbers that U.S. courts accept as references.  (Of course there are
other ways to fix this than to open-source the database, I'm just
pointing out that the social effects can be significantly large.)

    Tim> That's another whole way to leverage the paradigm shift --
    Tim> get ahead of the monopolists, and get large critical pieces
    Tim> of data infrastructure into the public domain BEFORE the
    Tim> monopolists get into play.

But they have a big advantage here; they can provide incentives that
open source cannot.  And what do you do if the only way to get at it
is through Hailstorm/Passport?  Scary.

    Tim> I think where I disagree with you is in the idea that we need
    Tim> to figure this out for every database.

I don't think that we need to figure it out for every database.  My
point is that we can't know a priori which databases are going to need
"liberating" because they are blocking development of a complementary
technology owned by a rival (or some 12-year-old genius somewhere)
"Give me the monopoly, you'll never know what you're missing!" is a
_serious_ threat here.

Institute of Policy and Planning Sciences
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.