Subject: Re: From a mere reader
From: Ian Lance Taylor <>
Date: 27 May 1999 23:10:12 -0400

   Date: Thu, 27 May 1999 23:09:59 +0100
   From: "S.L. Kochanski" <>

   The collapse of Windows means market fragmentation, 

One lesson I take from Neal Stephenson's excellent essay
( is that as time passes
software becomes infrastructure (like many good ideas, it's obvious
once you hear it).  Therefore, I do not expect the collapse of Windows
will lead to market fragmentation.  The OS layer has become
infrastructure.  There are very few operating systems still standing;
the number is not going to go up significantly, and will probably only
go down.  I used more operating systems in the first four years of my
programming career than I have in all the time since.

Actually, I also rather doubt that Windows will collapse.  It may
collapse from Microsoft's point of view: it may become free, and it
may eventually become an emulation layer of other operating systems.
But I think it's too entrenched to actually disappear anytime soon.

   "Our team" = "people who have a stake in our product
   succeeding". Possibly they make a living from it, or just a certain
   amount of money, or maybe they get glory and fulfilment - enthusiastic
   word of mouth has always been our best marketing method.

   I see open software as a way of building such teams. But - forgive me
   - I don't want to give it all away. One can bask in the glow reflected
   glory of millions of users, but one can't eat that glow.

First, I suspect that most free software businesses start with a
commitment to free software, and continue by deciding what to do.  If
you skip that commitment step, then you probably aren't going to get
far.  Freeing software always leads to an obvious loss of potential
revenue; revenue gains are generally harder to see.

Looking at it that way, you could ask yourself the obvious questions.
What might you lose by making your software free?  What might you
gain?  For example, you might lose sales because people will simply
download the free version.  Or you might gain sales because the free
version will spread awareness of your software farther and wider.  The
fact that a free version is available does not mean that people will
not buy your version; if it did, Red Hat would be out of business

Anyhow, I think you are approaching it from another angle.  It seems
to me that you want something to happen: you want to build a larger
community of people to contribute to your software.  You are
considering freeing the software as a means to this end.

If I am right, then you should be aware that most software attracts
very few contributors.  The Mozilla project was unhappy because only
30 people outside Netscape contributed, but, frankly, based on my own
experience in free software, I thought that was quite a few.

You will attract contributors particularly if there are already people
who would like to contribute: for example, if there are programmers
using your product who are sending you suggestions for features, and
who might implement those features themselves if they could.

I think you will be hurt in this regard by being on the Windows
platform.  Windows, unlike Unix, does not have a tradition of tweaking
source code.  I'm not engaging in Unix snobbery here (I do that other
places); I merely mean that the Windows world tends to approach things
slightly differently.

   So what I found particularly interesting, among the shrapnel, were the
   odd business models that were emerging -- how firms manage to provide
   a public service but still, somehow, to make money 

The current spate of web based businesses seem to show that it is
possible to provide a public service and still make money.  That's
essentially what Yahoo is doing, and there are plenty of other
examples.  They won't all survive, of course.

I think most free software businesses which provide public services,
such as Red Hat's free downloads or Cygnus's, do
it essentially as a public relations move for their real money-making
strategy.  They don't lose any sales this way; the people who download
the free code would not have paid them money anyhow.

   -- what it is about
   an open-source project that attracts or repels people

I think the ideal case for attracting contributors is code that
basically works, that does something which programmers find useful
and/or interesting, and for which there is some reason to tweak it to
work better for a particular case.  Bad cases are code that doesn't
work, code which works perfectly (not many people hack /bin/true--
although I see that it does have a couple of options now so maybe I'm
wrong), and code which programmers don't use.

   -- stories of
   the transition from totally proprietary towards openness: can one do
   it and survive?

Most free software businesses started out free, and didn't make any
transition at all.  Netscape (now AOL, I guess) is trying to make the
transition from proprietary to free.  We'll see what happens.

   I do recognise that philosophical and constitutional issues have their
   place, but I would (if it is permitted) be interested to hear more
   history, more story, more anecdote: how it works in practice.

I'm afraid this list has had a philosophical side for a while, and I'm
afraid that people like me contribute to that.  But it does have a
practical side as well.

   How - I keep wondering - do the luminaries of the free software world *eat*??

I dunno, pick a luminary and we'll tell you, or at least make a good

I'm not sure I qualify as a luminary, but I used to work at Cygnus,
where I was paid a good salary to solve people's problems in the area
of software development tools.  Those people in turn paid Cygnus large
sums of money, because they had large problems.  Like the old joke
about the doctor who specialized in diseases of the rich, it's always
good to be in the problem solving business.  Then free software almost
becomes a tactical maneuver to gather up whatever changes get
contributed and to make life hard for proprietary competitors.