Subject: Re: "I've got more programmers than you"
From: "Stephen J. Turnbull" <turnbull@sk.tsukuba.ac.jp>
Date: Thu, 4 Oct 2001 07:15:50 +0900

>>>>> "Geoff" == Geoff Crawshaw <gcrawshaw@openair.com> writes:

    Geoff> I see no reason that over time the RDBM market does not
    Geoff> start to look like the OS market with the Linux/MS split.

I suspect Oracle could live with that.  As long as the OSS DBs don't
take more than a small share of Oracle's target market, which may or
may not include your company.

My point is not that FSBs can't do well.  Not even that they can't
develop world-beating technology.  It's that they have inherent
financial weaknesses that tend to restrict them to the "commodity
infrastructure" end of the market, and thus are going to suffer from
low margins---unless they target their market segment accurately.

I don't see any reason why successful small to medium-size businesses
cannot be built on FSB principles.  I do find the optimism of Tom,
Willem, and Peter worrying.  FSBs have few or no inherent advantages
over mixed-mode strategies, and a large potential financial
disadvantage.  This strongly suggests to me that FSBs are inherently
going to be niche businesses.  And that they are going to have adopt
strategies that don't expose them to large demands for cash flow.

I also think that many of the segments proposed for entry, such as the
desktop, office suites in particular, multipurpose databases, web
browsers etc have characteristics that make them ill-suited for
development by FSBs.  There are OSS strategies for these, probably
even FSB strategies, but _developing_ them is not going to pay
anyone's rent.  The problem is not that the "engines" won't be good
enough, it's the "layering around", the myriads of details that you
actually can deal with with an army of coders.  This incompleteness of
the whole solution makes it harder to get funded, which makes it
harder to get to market early with an engine, which makes the engine
less economically valuable.  (Thanks to Keith Bostic for the engine/
layering terminology.)

It worries me that people on this list talk about (ultimately) getting
rid of proprietary software, or developing office suites or RDBMSes,
in the context of business strategies.  I think that has to be based
on an unrealistic view of what FSBs are capable of achieving.  Not on
the technical side, but on the business side.  People regularly
explain why the technical logic is fine.  That's not what worries me.
I know there are a lot of fine OSS programmers, designers, architects
out there.

What worries me is the lack of _business_ logic.  They don't teach FSB
management at any biz school I know of.  It's got to be home grown.
(Yeah, there are a couple of Harvard cases now.  But Harvard cases are
like Pokemon---fad driven.  Or at least they're based on current
events, and can't really support mature analysis.  After a few years,
many of them just go into the archives, never to be seen again.)  I
wouldn't be surprised if I'm wrong about what can be done.  But I
don't think it'll be disproven by the kinds of logic the optimists
post.


-- 
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."