Subject: What does "Free Software Winter" mean and why should we worry?
From: "D. V. Henkel-Wallace" <>
Date: Fri, 02 Apr 1999 07:46:58 -0800

Apparently my "Free Software Winter" reference wasn't as clear as I
thought.  It _is_ a serious risk, so here's some clearer explanation.
I'm sorry this is long, but if you care about free software maintaining
even the level of popular support it has today, it's worth the read.

   Date: Thu, 1 Apr 1999 11:51:57 -0500 (EST)
   From: (Kragen Sitaker)

   I fear that some companies will start "open source" projects, and then
   not get all they'd hoped for, and loudly and publicly shut down the
   projects.  Think of Corel's Java suite.

   "Oh, I'll release this and then the Net will do all my work for me.
   Everyone will be happy to contribute to my code, and then I'll have a
   great piece of software I can license for huge sums to my competitors.
   And it will be bug-free, just like Linux."

   That's my fear.

And that's FS Winter.  The analogy is to AI winter (described by Terry
Winograd in 1984 as a warning, which sadly was not heeded).  If you
switch on the wayback machine you may remember the AI book of the
early '80s.  Salaries that make today's boom prices look small.  Lots
of companies touting their expert systems and whatnot as the answer to
all the world's problems.  And today?  "AI" is a dirty word in
business and worse, pretty much everybody has left the academic side
of the field.

I don't want this happening to the free software world either.

What causes this?  Arrogance, overselling, and getting caught up in
the froth.  These are human failings, not sinister acts of losers, so
it's easy to unthinkingly head down this path.

Running a good FS project is hard, but fortunately there are many good
examples.  The FSF started out well and pretty much has continued
well, except that it had a hard time sustaining the energy.  Linux is
run very well.  egcs started out well and has continued well, though
it is helped by the small size of the immediate community.  Apache,
bind...the examples are manifold.

But so are the counterexamples, though they are almost by definition
harder to remember.  Mozilla is a good example of one done (sadly, not
surprisingly) poorly.  I won't piss on people by singling out any others.

But there is a common theme.  Ian wrote:

   Date: 1 Apr 1999 14:27:27 -0500
   From: Ian Lance Taylor <>
			Once that has been done, it takes only a minor
   extra effort to release it on the net.

I don't think this is true.  For a free software project flourish
takes work.  Look at gilmore's work to bring gdb back to life when it
had splintered in the early '90s.  It took a long time and a lot of

Most people will just download the software, (perhaps) build it and
run it.  The "barrier to entry" has to be low.  All that takes an

One of the excellent things about libre software in the true sense is
that even if the author can't keep up that workload, as can naturally
happen for a bazillion reasons, someone else can step in and take

But the emphasis on "there's a gagillion zero-cost developers out
there just waiting to help you out" gets the emphasis backwards, and
is inherently dangerous for us.

   Date: 1 Apr 1999 22:51:29 -0000
   From: Russell Nelson <>

    > ...Yet they claim to act for good of the community, and
    > Eric Raymond, the president, refers to himself as ``public advocate
    > for the hacker tribe.''  And, in fact, people outside the community
    > appear to accept these claims to one degree or another.

   Advocate, yes.  Leader, no.  I don't see anybody following him; who
   could possibly mistake him for a leader?

Please don't be hyperliteral.  A businessman looking for an overview
of OS, or for "certification" that his OS plan is "acceptable" is
going to look for a single point of contact, self-appointed or not,
and will consider that contact to be the "leader" if it functions as
being legit.

Sure, you can stand on the words and claim that your trademark means
only what you say it does, but that isn't the way the world works.
MacDonalds thinks that their name stands for a hot, tasty nutritious
sandwich, but to many Indians it stands for imperialism, and to me it
stands for Frankenpotatoes.

So regardless of what the trademarks say, today the OSI is like the
federal government in Snow Crash: its legitimacy is recognised only
within its own grounds, and by a few faraway people.  And that's a
shame.  (I don't want L Bob Rife exploiting it!)

It would be great if an organization actually exhibiting true
leadership in the free software community.  The FSF did this for a
while but blew it.  Certainly no organization can please all the
people any of the time, but it can do an excellent job for most of us,
most of the time.  Perhaps the OSI is even the right one.  But
certainly not as it is constituted or operates today.

This message is long enough so I'll stop here.  If you agree that the
problem is a legitimate one I can talk more about what kind of
organization might function in such a role.  It's hard but not rocket
science because there are many such organizations in other fields.