Subject: Re: open source definition
From: Phil Hughes <fyl@a42.com>
Date: Wed, 22 Apr 1998 03:10:46 -0700 (PDT)

Bob Young said:

> Proprietary is not my word either.  The current software funding 
> model is based on VC's investing in the "intellectual property" that 
> results from their investment.

Generally, this is what happens.  Or, more specifically, an individual
or a few individuals come up with a good idea, prototype a product and
then Mr. VC comes in and gives them a piece of their action.

In the case of Linux, that is the approach Linus could have taken.  I
think, however, we are all aware that it wouldn't have worked because:
  * if he had, there would not have been the level of cooperation
    from contributors to make it happen
  * no VC would have believed it could happen

> What is enabling us to succeed in competing with Microsoft (ok, ok, 
> in a very small way ;-) is that we are playing by a different set of 
> rules than those used by Microsoft, or IBM with OS/2, or SCO UNIX, 
> or any other proprietary OS vendor.

One thing missing here is that Linux is a better product.  In a free
market (which, of course, doesn't really exist) better products tend to
get market share.  What made Linux a better product was the thousands or
tens of thousands of people who freely contributed to that effort.  (By
freely, I mean they contributed their work--writing code, testing,
writing documentation, ...--with no strings attached.)

> > It's instructive to ask who can and cannot make money using this
> > model:
> > 
> > 	the publishers can make money
> > 	the packagers can make money because the packaging is a
> > 		difficult to reproduce tangible good
> > 	the marketers and distributors can make money
> > 
> > In fact, the ONLY party in this scheme who can NOT make money is the
> > party who makes the whole thing possible -- the original software
> > author.  This is because the terms prevent the software author from
> > getting royalties.
> 
> Again why the distinction?  Why can't the author be the publisher, 
> the packager, the marketer or distributor?

Let's look Linux again.  Linus didn't want to be a publisher.  He wanted
to contribute Linux to the good of the community.  His willingness to do
this was a large contributing factor to the success of the Linux
movememt.  If, on the other hand, he had said he wanted 10% of the gross
sales related to Linux ... (just a William dellaCroce phrase I picked up :-) )
he would have pissed off most everyone.

On the other hand, few people are having a problem with Red Hat,
S.u.S.E., Walnut Creek, Caldera, InfoMagic, ... selling packaged
products.  I think I lot of this gets back to people considering a CD
and possible a book as a tangible piece of property and "raw software"
as not being tangible.  We all know this isn't right but the number of
people that swipe software but don't swipe chewing gum tends to support
this conclusion.

> > > Publishing under a freely redistributable license does not
> > > necessarily restrict the commercial success of your product, 
> > > and can enhance that success.
> > 
> > This is a red herring, because you are talking about a different
> > product.  You guys don't sell software.  You sell packaging and
> > integration.
> 
> Again, why the distinction?  Money is money.

I see an important distinction.  The capital to develop a large software
product needs to come from somewhere.  In the case of Linux, it came
from a cooperative effort of a huge team.  If I go hang my shingle out
as a C programmer, I get paid by the hour or project to do something.
On the other hand, if I decide to rewrite UNIX I am electing to make a
huge investment before I start receiving anything back.  For example,
look at the HURD.  There is a large investment in it but is has never
prduced income for anyone.  

> > That's fine when someone else is giving the software away for you to
> > package, but kindly be honest enough to admit that your scheme only
> > works because you have thousands of unpaid laborers.
> 
> Now here is one of the dirty little secrets of our, and the free 
> software community's, success:  It is quite possible that the 
> majority of the non-Red Hat contributors to Red Hat Linux are quite 
> well paid for their labours.
> 
> From Linus Torvalds who went from being a starving Finnish undergrad 
> to getting a (presumably) well-paid job in Silicon Valley -because- 
> of his work with Linux, to Don Becker who earns a comfortable living 
> building technology for NASA that ends up as contributions to the 
> tools that he "borrows" from the free software community for his 
> work, to C and C++ compilers from Cygnus, to O'Reilly's free software 
> summit authors who make money off of the books they've published,
> most of (but definitely not all) the primary contributors I can think 
> of earn some or all of their living off of the work that the Red Hat 
> Linux OS benefits from.

Linus got the job because he is a great programmer.  Donald had a job
(at NSA) that didn't require him to write Ethernet drivers. ...  They
elected to give capital (their time) to make Linux happen.  The fact
that they were the lucky ones (as opposed to people like Orest Zabrowski
who got to work at Microsoft working on print spoolers because he ported
X to Linux) who get to work with what they love and get paid for it.
I expect this is true of Cygnus people and book authors.  It is
certainly true for me and a large number of the people at SSC.

In the case of Linux, there has been huge investment in making it what
it is today.  Red Hat has made investment, Linux Journal has made
investment, programmers have made investment, ...  This combination of
investment is what has made Linux both a good product and caused it to
get market recognition.  (I have been told over and over that being able
to take a copy of Linux Journal in to a supervisor has been what has
made it possible for a techie to get Linux to infiltrate another
commercial enterprise.)

I see Linus, the Free Software Foundation and a bunch of programmers as the
first-round capitalization of Linux.  The second round was the
Mark Ewing's, Patrick Volkerding's, Phil Hughes', ...  If there is to be
a third round, it certainly could be traditional VC firms.

To me, what we all want to do is figure out how to make this model--the
one where a bunch of people contribute for a long time without taking a
piece of the action--work for other software projects.  And I feel that
we have to look at how to start an effort--not just wait for the next
Linus Torvalds to come along and do it for us.

We think that the way to do this is to have "Open Software".  Whatever
that is as we continue to argue about it.  Maybe we need to back off and
see what models work and don't work and use that to help with the
definition of what we "are for".  For example, why did Linux succeed
commercially where FSF/GNU/the HURD did not? 

All that said, for Linux to continue to succeed, all the players need to
cooperate.  I think this will true in any situation where Open Software
tries to compete in a commercial marketplace.  For example,  one Linux
distribution vendor decides to try to distance themselves from the
others it just fragments the community.  My analogy in commercial
software was when PageMaker was the best brochure program, Ventura
Publisher was the best long document program and Quark was the best
magazine program.  Each product was clearly superior in its own market.
Then they decided to tell their marketing departments to "get all the
desktops".  It wasn't pretty, a lot of users got confused and wasted a
lot of time and we now see some seriously bloated products out there.

-- 
   Phil Hughes ++++ FYL ++++ fyl@a42.com ++++ Phone/FAX (360) 276-4232
              P.O. Box X, Pacific Beach, Washington 98571
  To find out about Pacific Beach, check out http://www.pacificbeachwa.com/