Subject: Re: Here's a-what I'm a-gonna do.
From: Magnus Redin <redin@signum.se>
Date: Thu, 13 Jun 96 12:30:56 MET DST


From: Michael Tiemann <tiemann@cygnus.com>

    As far as I know, Cygnus, John's company, is the only substantial
    business that doesn't fall into one of the above cases.  I still
    can't understand how Cygnus, and only Cygnus, can generate enough
    excess revenue to support the overhead of a good-size business
    solely by doing support and contract development of GNU-Licensed
    software.  Or does Cygnus also generate revenue by developing
    non-free software?

> Well, Cygnus has yet to offer a single product that derives any $$
> on licensing fees.  We have substantially tightened up our support
> contract wording that has the effect of a strong licensing
> agremeent, but only if you want support.  While the relationship may
> be terminated, the freedom to use the software remains.

At Signum we have done this "sin" and afterwards let the product free.
We dident realy dare to develop it as free software from the beginning
when we only had a handfull of customers that understood that Internet
were a good thing and that we could deliver something useful, we were
afraid of scaring that tiny and critical group with a new concept. And
we did anyway not want to release something unfinished as free
software.

I am curious about your support contracts. I have heard that there is
another contract culture USA then in Sweden. Here they usually are a
couple of pages and the rest is covered by a "gentlemans agreement"
that we will deliver what is needed to fulfill our promises and not
only excactly what is specified and the customer will pay and of
course pay for extras. Its fuzzy, practical, might be dangerous but it
is very convenient. Why did you tighten up your contracts?

> I think that the reason for Cygnus's success is because we have
> chosen to use the free software model largely as a means rather than
> an end.

Wise, customers rarely want to make an ideological statement. :-)

> This has meant a frustrating wait on the sidelines of Linux while
> the market sorts out what it wants (I CAN'T WAIT!), but it has meant
> a really successful ride on GNU, due to the nature of the market and
> our ability to drive solutions into the heart of the problem.

This is intresting, care to elaborate some? We did choose to sell
support on linux installations and on others distributions. What
frustrated us was that Ydddrasil lagged in its development and
shipment and the focus moved to Slackware. (The movement of the
"focus" was most probably not to due to the late cdrom releases but
due to Slackware being a good and modular distribution that could be
distributed with ease with floppies and over the net. )  As it is now
Redhat and Debian looks promising. I like Redhats ambitions with
getting commercial software to ship for Linux. Perhaps we should have
worked together with someone with developing a good distribution?
Perhaps it is not to late?

> Now, given what I said about ends and means, one might conclude that
> if the right chance came along, we'd do something proprietary.
> There's no easy answer to that question, just as there's no easy
> answer to a sound policy for the use of military force.  In the mean
> time, our fundamental strategy continues to win favor in the market,
> the measure of success we set forth in the beginning.

The company(board) has goals, the owners has goals and those working
inside it has goals.  It is great if they are the same ones and all
want to use the same tools and ways to reach them. If one is forced to
compromise this greatly I suspect that the company will become
something else, perhaps loosing people, perhaps getting other people.
For us it is nice to after a few years be able to get a little closer
to the initial goals that were very inspired by Cygnus and the free
software movement. And most of what we do, even the proprietary stuff
has something to do with Internet in ways that help create the
infrastructure for the working and cooperation model that has given us
free software. 

For instance we will help release a binary-only free version of a 3-D
CAD system for linux in a few days. Its a unique system where one
describes the objetcs in a programming language or draws the objects
to create the descriptions. If done right one can build flexible
libraries of parts and set up systems where one can change for
instance the height of a stair and the calculations ripple down thru
the objects creating a descripton for all objects needed to be
manufactured for such a stair. Its used in niches, mostly in Sweden
with for example designing houses and parts in SAAB:s aeroplanes.  If
this experiment goes right there will be free libraries of parts
created brining the free software principle to manufacturing companies
and those who has written the program hopes that spreading it via the
Linux community will help them getting good support contracts.  If
this works out well and people start writing and sharing routine
libraries it could start competing with CATIA and AUTOCAD in a few
years. 

What do you think about this reasoning? Is it likely it will work?

Regards,

--Free software--I  Magnus Redin        I  WWW   http://www.signum.se/
Ibland får man   I  Signum Support AB   I  email redin@signum.se  
mer än vad man   I  Teknikringen 8      I  fax   +46 (0)13 214700 
betalar för.     I  S-583 30 LINKOPING  I  tel   +46 (0)13 214600