Re: [unrev-II]MS exec: Open-source approach is flawed - ZDNet Discussion Zaplet

From: Eric Armstrong (
Date: Mon May 07 2001 - 17:27:19 PDT

  • Next message: Eugene Eric Kim: "Re: [unrev-II]MS exec: Open-source approach is flawed - ZDNet Discussion Zaplet"

    Lee Iverson wrote:
    > And where does this claim come from? When has an open-source release
    > "killed an industry"?
    The difficulty with proving the proposition is the invisibility of the
    effect. Once, there were 20 different unix competitors. (That was not
    a good thing, mind you. The lack of standards was apalling.) Now, there
    is Linux and Solaris. Will others ever exist? It's hard to tell. It's
    *extremely* hard to identify how much work is not going on in a given
    area, due to the lack of ability to derive compensation from it.

    > A much better example is of an open source release "creating" an
    > industry cf. the Web. The Web exists today because Tim Berners-Lee
    > and the team at NCSA were committed to 1) simple, open standards, 2)
    > open source software, and 3) a *free* front end. An strong argument
    > can be made that these were the things that differentiated the 90's
    > WWW success from the failures of all previous hypertext efforts.
    That is a great example. It shows how great things can happen. But to
    whom does the browser industry now belong? To the only company that
    was actually making a profit at the time (and that from doing other

    How many browser competitors are there? How many tree-oriented browsers,
    XML-oriented browsers? How many of them are staffed to deliver a stable,
    professional-grade product, vs. a nice weekend hack that will never
    become much more than that?

    > There are definitely successful open source companies (Cygnus was one
    > and Codesourcery is another) that have very simple business models.
    > They develop and maintain open source software that other companies
    > depend on. The companies paying for the work get the advantage of
    > having a stable, centralized code base that they can direct the
    > development of with their dollars. Moreover, because other companies
    > are also paying to continue development, there is an economy of scale.
    > With Apache, it is slightly different, in that the vast majority of
    > development (as I understand it) gets done in-house by various
    > companies who then feed the changes back into the open source code
    > base. As Brian Behlendorf argues, everybody wins with this model.
    > Most bugs get fixed quickly and fed back into the baseline code.
    > There is a substantial incentive to not branch, since so much work is
    > being done by others that you wouldn't be able to take advantage of if
    > you were to branch or hold a large codebase private.
    > There *are* successful business models out there. Interestingly, they
    > are service-oriented and programmer-intensive. Seems just fine by me.
    Yes. The Cygnus example in particular points to the one case where
    open source *can* work. When you are selling a support to a company,
    who can easily justify the expense because it is making money, then
    the model can work. As I argued, it is consumer-grade products which
    really seem to fail with this model.

    I guess the lesson I would learn for the bootstrap project is to give
    client software away to promote the market, and charge good money to
    support the server side of the business!

    Community email addresses:
      Post message:
      List owner:

    Shortcut URL to this page:

    Your use of Yahoo! Groups is subject to

    This archive was generated by hypermail 2b29 : Tue May 08 2001 - 07:22:26 PDT