Subject: Re: [FYI] Microsoft license spurns open source
From: "Stephen J. Turnbull" <turnbull@sk.tsukuba.ac.jp>
Date: Tue, 26 Jun 2001 00:34:54 +0900

>>>>> "Bill" == Bill White <bwhite@etnus.com> writes:

    Bill> On Mon, Jun 25, 2001 at 11:34:08PM +0900, Stephen
    Bill> J. Turnbull opined:

    >> You're confusing cost and price, I think.  Costs (to the
    >> vendor) surely don't change much; prices (to the customer) may
    >> go up or down.  Which do you mean?

    Bill> This seems wrong.  Different versions require multiple QA
    Bill> cycles.

Oops.  Yes, you are correct.  Vendor development costs will increase
substantially.  But these _fixed_ costs are (in economic theory,
anyway) irrelevant to the pricing question.  We want to look at
marginal costs.

But even in average terms, for the issue of versioning via "terms"
(which hardly seems to require much in the way of QA) I think they're
pretty close to negligible.  To take an example more favorable to your
case, consider the amortized unit cost of creating Windows NT
Workstation by editing 12 bytes or whatever in the registry, deleting
a slew of server binaries, and doing the QA, compared the the price
differential vs. the Server edition.  On that standard, isn't the cost
differential negligible, at least for a quick and dirty approximation?

A more important issue for marginal cost would be customer support.  I
can imagine that Server would be substantially more expensive per copy
to support.  These variable costs would surely demand a certain amount
of price differential.  But again, I find it far-fetched that they
would justify the price difference between Workstation and Server, at
least at the prices Gateway Japan quoted me last year.  For back of
the envelope purposes, isn't this cost differential also negligible?

On the other hand, those price differentials make a _huge_ difference
in the number of people willing to buy Windows NT, and to Microsoft's
revenues.  And the price differential to cost differential ratio
should be spectacularly higher for versions defined by "can't use on
Tuesday or for computing horse race odds" style terms.

-- 
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."