Subject: Re: a tool free software developers need
From: Tom Lord <lord@regexps.com>
Date: Fri, 15 Feb 2002 15:40:45 -0800 (PST)


        I must admit I was slightly confused by the misunderstandings
	between you and Michael Tiemann about what exactly business
	"B" is, (and for that matter, "A" as well). Could you
	*briefly* clarify?

This is getting silly but here's a short version: Business "A" is an
FSB in a world where the gap between public project distributions and
commercial quality, integrated releases is large, largely because of
immature processes and tools in fairly basic areas[1].  Much of what
business "A" implements is the process of bridging that gap, mostly
for the benefit of its specific customers.  Business "B" is an FSB in
a world where those problems are solved with the result that the FSB
is providing much higher-level and more aggressive innovation to its
customers.  The easier the public software is to work with, the more
valuable it is to customers, and the more that can be done with it;
there's a well documented wealth of basic improvements that can be
made to the public projects to make their software easier to work
with.  Recent business developments among the big server manufacturers
(several of which themselves contain elements of business "A") and
some adjacent industries suggest that they are a market for companies
that sell the service of helping to make the "A" -> "B" transition in
the public projects and that they want to be customers or sponsors of
legally independent type-"B" organizations.

And once again, businesses "A" and "B" are extreme points on a
spectrum and my take is that the real businesses are somewhere in the
middle, with a trajectory towards "B".

([1]: At the same time, the public projects achieve some
unprecendented successes in spite of the process and tool
shortcomings.).



	This [the test platform service] is a good idea. I would be
	able to utilize something like this quite often. But would I
	pay? not sure about that...

I think you would not.  I think it would be paid for mostly by
platform vendors of all types, with some augmentation of the network
by volunteers.  OSDL is a domain-specific implementation of the basic
idea and that's free (to selected projects).

	I don't see how much money could be made off this model
	though.

It's an expense for platform vendors and large consumers, offset by
lowered costs and/or hightened capabilities down the road.  It's
potentially a paid commercial service to help implement and deploy the
infrastructure.

-t