Subject: Re: FW: Why would I pay for Ximian software?
From: "Perry E. Metzger" <>
Date: 03 Jan 2002 01:05:34 -0500

Mike Linksvayer <> writes:
> > I'd like to think there is good reason to believe that the public
> > goods/market failure theories that would traditionally be applied here
> > aren't true. Note that I say "would like to think". However, there is
> > some strong indicative evidence in this direction.
> OSS overconsumed?  Evidence: MSWindoze 90%+ market share.
> OSS underproduced? Evidence: How many free general purpose OSes,
>                              desktops, productivity applications,
>                              etc. can you use in a day?

That isn't really any evidence in this instance.

How about the following example instead, from my own world. It turns
out there are a very large number of companies using NetBSD in the
embedded market -- it has become quite popular with many
vendors. There are very few of them, however, who are contributing
back to development in any way. If they all did in fact contribute
back, a lot more resources could be put into OS development, but of
course, most of them don't.

> Conclusion: OSS is underconsumed and overproduced.  "Market failure"
> of a different sort.
> This counter-intuitive market failure is caused by copyright rather
> than any problem with free software.  Proprietary software has
> crowded out free software in most markets (underconsumption).

We would expect that since proprietary software is much more expensive
to buy and use than open source software, that it would lose market
share if it had the same level of functionality. In an environment
where proprietary and non-proprietary software compete under similar
levels of functionality, we would expect that with time the
non-proprietary software would win. In fact, I think this is (rather
rapidly by many standards) happening. The fact is that at this point,
GCC has destroyed most of the market for C compilers on non-Windows
platforms, ISS is a second to Apache, and Linux based servers are
probably the single biggest threat to Microsoft's dominance of the
server market. Since there isn't yet a reasonable OSS desktop for
ordinary users (I'm going to get flamed to death for that one but I'm
not interested in debating it) it is not surprising that we don't yet
see market penetration there, but I would expect that a truly friendly
OSS desktop (with a really good set of personal productivity apps)
would win very easily.

> Market leaders are nearly all proprietary,

In the areas where there are good pieces of free software available,
the proprietary alternatives have been severely damaged if not
crushed. The problem is that developing good alternatives requires a
lot of resources. The purpose of exploring ways to make money in the
free software arena is to assure that those resources are available
and directed where the market has the greatest pent up demand. We
would expect that if we could find a way for the market to properly
deal with the situation, Ximian could make a lot of money competing
with Microsoft, and use a good chunk of that cash in producing better
and better product, ultimately crushing Microsoft. However, if they
cannot make money at it, the problem of supporting the programmers who
work on such software gets harder. The fact that software, unlike
building bridges, can be done by a large group of people working part
time certainly helps matters, but I think you'll find many of the most
important free software projects do in fact have (and need) full time
help, and that does have to be paid for.

I have no doubt that whether or not free software businesses succeed
free software will continue, but if we can bridge the "market failure"
issue and make free software businesses able to pay for lots of
programmers, it will make it much easier for the world to shift models.

Figuring out how such companies can truly succeed in the sorts of ways
that attract enough money for the programmers involved that they don't
have to worry about what will fund their work is a matter of key

> A facile argument no doubt.  I post because the argument seems
> obvious yet I haven't heard it, and market failure bugs me. :)

Well, obviously I believe that you can make money in the free software
business or I wouldn't be in it, and that the "market failure"
question can be addressed, but I must admit that the problem is a hard
one and that we will not get anywhere by pretending it doesn't exist.

There are deeper issues here for me. As for one example, in the
future, nanotechnology will make everything in the world essentially
software. If we can make free software businesses succeed, we can
eliminate the need for "closed source" designs for real world goods in
a nanotech world, and thus make everyone on the planet measurably
richer in a material sense. However, if we cannot find a way for
engineers to make a good living in that sort of world, we won't have
nearly the levels of interesting and virtually free (as in beer)
physical constructs available, and the world will be poorer for it.

In a similar vein, it is often said that we need government in order
to deal with situations in which there are externalities and market
failure, such as the market for defense. A systematic understanding of
ways to avoid needing government to deal with such externalities might
well radically reshape human liberty for the better.

It is thus quite important that we face the "public goods/market
failure" issue square on and find ways to get past it, rather than
ignoring it. The free software business is only one part of a much
larger picture to me.

Perry E. Metzger
NetBSD Development, Support & CDs.