Subject: Re: Computer Survey on perceived advantages of Open Source
From: "Stephen J. Turnbull" <>
Date: Sat, 14 May 2005 22:23:45 +0900

>>>>> "Cliff" == Cliff Schmidt <> writes:

    Cliff> On 5/13/05, Laurent GUERBY <> wrote:

    >> If a closed source software goes bad (abysmal support, no bug
    >> fix, no migration to new hardware or platform), you have no
    >> option, you're dead.  It happens just about continuously.

    Cliff> Having the source is clearly, at least marginally, better.
    Cliff> However, if a vendor-run open source project goes bad, for
    Cliff> most users the marginal benefit of having the source isn't
    Cliff> enough to help.  They are probably going to have to move to
    Cliff> another platform, unless they're able to start-up an
    Cliff> in-house staff to support and maintain the code base (very
    Cliff> costly) or take on the effort of starting a new open source
    Cliff> fork (also a costly venture with unpredictable results).

You can usually see a product going bad in advance.  Furthermore, with
an open source project, it is very probable that there are several (or
many) other customers with interest and experience---hackers do hack,
you know, even if their job description says they're not paid to do
it.  You'll probably have names and mail addresses of major
contributors in and out of the vendor from the ChangeLogs.  All this
implies that the cost of cobbling up a support consultancy is likely
to be substantially easier for open source.  And remember, many of the
staff at the former open source company are likely to be unemployed!

It's important to remember that if a closed-source vendor goes
belly-up, it is highly likely that the former employees do not have
the right to reveal sources or modify them, and the new owner may very
well be a bank or such in no rush to dispose of potentially valuable
assets.  All this suggests that Laurent is right on the money:

    >> Normal market forces, ie competition on support not destroyed
    >> by government granted monopolies thanks to the open source

[Competition on support is not destroyed by IP; it is simply made much
more expensive and clumsier.  That expense and clumsiness may go on to
destroy competition for support, but it's not a foregone conclusion.]

    >> setup, would also obviously respond to the incentive of getting
    >> some $ for restoring support and that mean that a company would
    >> already be there to continue support.

    Cliff> But would you bet a key piece of your business that these
    Cliff> market forces will take shape and come to your rescue when
    Cliff> you need it?

Don't kid yourself, you do every day, anyway.  The only things you're
not "betting on market forces" are the things you produce yourself and
have in stock in your warehouse.  Heck, you're betting your top
engineers on market forces!

Obviously, you do not want to count on market forces working in your
favor; you need to evaluate the risk, you need to prepare a cushion
against the larger risks, eg, by actively networking the other
customers and stroking the core people at the vendor.  Don't forget to
stroke your top engineer, too.  :-)

    Cliff> I think users should see software falling into one block of
    Cliff> a 2x2 matrix.  On one axis (excuse the re-use of "axis"
    Cliff> ;-), there is standards (proprietary APIs or industry
    Cliff> standards); on the other axis you have ownership or control
    Cliff> (vendor-controlled or community-controlled).
    Cliff> Vendor-controlled projects with proprietary APIs are
    Cliff> clearly the riskiest place to be.  Community-controlled
    Cliff> projects based on industry standard APIs are clearly the
    Cliff> safest.  My point is that the other two blocks are only
    Cliff> marginally different in the risk presented to the typical
    Cliff> business user.

Fine up to the last three sentences, which are all unwarranted
generalizations, but especially the last.  It will depend on the case,
because the kinds of risk in each corner are dramatically different.
In some cases one will be far preferable to the other.  Your business
needs to look at the options, not prejudge them.  It's possible that
the nature of your own business will strongly bias things one way or
the other, but in many cases it will be wholly up for grabs.

And don't forget ... if you last out the year, you're probably going
to be in the market for an upgrade.  The risk of lock-in and cost
escalation is much lower with open source.  100 piranha nibbles leaves
you just as bled out as great white shark chomp.

School of Systems and Information Engineering
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.