Subject: Re: Market Forces and Free Software
From: Peter Deutsch <>
Date: Mon, 23 Sep 1996 18:51:01 -0400

   First, let's make it clear that I'm "J. Peter Deutsch"
   and he's "L. Peter Deutsch" and as far as we can tell,
   we're not related. He's the guy who did ghostscript,
   I'm the guy who did archie, and although we've never
   met face to face, judging from photos I've seen it
   seems I'm the better looking of the two!  ;-)

[ Another Peter Deutsch wrote: ]
.  .  .
} There are plenty of other models for successful businesses based around free
} software: I'd be interested in hearing some of them, particularly with
} respect to the marketing and channel issues that your message raised.  I'd
} also be happy to answer any other questions about my own experience.  I have
} no marketing background per se, so I'm sure the way I've described my
} experience doesn't quite use the usual vocabulary for discussing such
} things.

Okay, I'm normally a lurker on this list, but here's
another cut at how to make money with technology
development, without trying to sell only software or
relying solely on proprietary standards.

As the creators of Archie and founders of Bunyip
Information Systems we have focused on selling expertise
and services. We license some services, and develop
projects for others for a fee, often using some free
software components where appropriate. We specifically
tell people we are *not* a software company, but we do
have some services which require access to restricted code
we've produced, and often we'll produce some software to
accomplish the task at hand.

Although we don't support the free software model for all
our work, we do offer free access to some enabling
components we've developed, we participate heavily in open
standards development (e.g. with a relatively large
contingent at the IETF, given our size) and otherwise try
to help to build markets without actually giving away all
the value-added pieces we develop.

As one recent example, we've developed and released a
directory service technology called Digger, which is our
implementation of WHOIS++. In this case we have put out a
freely available version of this on the net and continue
to provide technical support, improvements, etc. We were
founding members of the associated IETF working group, we
operate a top level WHOIS-based index service, we supply
technical support to users of our server software and
otherwise continue to support this effort for the Internet

To accomplish this we've received some funded R&D money
from NSF and Industry Canada, as well as income from
commercial customers for whom a Digger server (or an archie
server, where appropriate) forms the basis of a service we
develop for them.

FYI, early versions of Digger were distributed with full
source, the current version is a binary-only release, but
we do make available specific components (including a
WHOIS++ component library, a UNICODE library, a WHOIS++
client and so on). For our source code offerings we use a
model similar to that mentioned earlier of granting free
access for non-commercial use, and requiring a license and
acknowledgement for commercial bundling.

How's it working? Fine so far from our perspective.  The
company was founded in 1992 with two employees, and today
we are 30 and still climbing, and this essentially without
a penny of outside investment (we did accept a couple of
small shareholders to cement relationships, but are
otherwise basically employee-owned).

Having said that, we do feel we're approaching a fork in
the road in which we will have to decide how much of our
growing company we will devote to the software side of
things and how much we continue to focus on the
higher-margin value-added expertise efforts. The 90-90
rule kicks in about now (wherein the first 90 percent of
the project takes the first 90 percent of the time and the
final 10 percent takes the other 90 percent).
Documentation, marketing, sales and support take huge
amounts of resources and the actual development costs
dwarf by comparison. We may even start considering either
investment, or a spin-off to handle software
productization, assuming we feel the markets are there and
need significant funding to exploit.

In effect, if we're to fund significant finishing,
packaging, marketing and sales of software, as opposed to
simply incorporating the components in our tool box and
using them (or allowing others to use them) in other
projects, we will need to up the revenue streams from this
end of things and we're tending towards a traditional
software packaging model to ensure that. Of course, we may
simply elect to continue doing what we're doing and let
others exploit any particular market, assuming there
really is a significant market for any particular
component we've developed. We consciously did that with
archie (We turned down several V.C.'s looking to take us
public a la Lycos and friends as it was simply not what we
did well and we felt we'd simply end up becoming someone
else's development division if we did that, which was not
in our plan).

On a different, but related topic and for what it's worth,
I feel the current IETF has lost much of its lead role in
steering and inspiring technical development for the
Internet through its failure to deliver on a promise to
promote open standards for applications. At this point,
the battle on that level is between Microsoft and
Netscape, and the IETF doesn't have much influence in
either camp (nor does the W3C, for that matter).

For our part, we will continue to remain active in
standards-setting and will continue to send people to the
IETF meetings, as we feel that we need to continue to push
the technical envelope and this is a good bully pulpit from
which to keep those spending the big bucks honest. I just
don't expect Bill Gates or Jim Clarke to pay too much
attention to either IETF working group mailing lists or
forums such as this one unless they demonstrate
significantly viable funding models that show them a
better path for exploiting their chosen markets once they
open up. All we've done so far is show them that this is a
good way to get that first proof of concept out quickly
and at little or no cost to those seeking to develop new

As one wag recently put it, Microsoft's line is
essentially "Thanks, and we'll take it from here". As an
explanation of what they're trying to do right now, this
rings pretty true.

Is the Microsoft tidalwave inevitable? Maybe not, but
let's realize that their goal is not to recover costs, it's
to make a significant profit. Only with that can you
reward risk-taking, cover the losers you've backed, put
aside something for a rainy day and still feed and cloth
your families (or investors). If you accept that these are
your goals, then you can analyze the free software model
as you would any other strategy for building your
enterprise. At times it works for us, but at times we feel
other, more traditional approaches are more appropriate.

Whatever works, is ethical and lets me feed my family and
sleep at night is what I'll choose to do...

					- peterd

     Peter Deutsch,                                  (514) 875-8611  (phone)
  Bunyip Information Systems Inc.                    (514) 875-8134  (fax)

  "This planet has - or rather had - a problem, which was this: most of
  the people living on it were unhappy for pretty much of the time. Many
  solutions were suggested for this problem, but most of these were
  largely concerned with the movements of small green pieces of paper,
  which is odd because on the whole it wasn't the small green pieces of
  paper that were unhappy."                      - Douglas Adams...