Microsoft's latest arguments against open source adoption are
feeble to say the least.
It is always entertaining to see Microsoft grappling with that most
shadowy of opponents, free software. Since it cannot attack
directly, Microsoft has been forced to take an oblique approach. A
few years ago it tried denying that anyone in business was using
free software - until it emerged that Apache ran more than 50% of
the entire public Web, and GNU/Linux was the number two server
operating system.
Today, Microsoft has adopted the more subtle approach of spreading
FUD (fear, uncertainty and doubt) hoping thereby to scare off users
that might otherwise be tempted to flirt with these alternatives.
The latest salvo in this stealth attack can be found in a speech
called The Commercial Software Model, given by Craig Mundie
(www.microsoft.com/presspass/ exec/craig/05-03sharedsource.asp). A
response (www.gnu.org/ philosophy/gpl-american-way. html) to some
of its technical inaccuracies has been put together by Richard
Stallman, who created the GNU General Public Licence that is
singled out for particular criticism, but here I want to examine
Mundie's arguments from an Internet perspective.
The opening sections play on concerns over the precarious state of
the new economy, and are based on a cunning, if manifestly false,
syllogism. Speaking of the dotcom disasters, Mundie claims that "a
common trait of many of the companies that failed is that they gave
away for free or at a loss the very thing they produced that was of
greatest value in the hope that somehow they would make money
selling something else." He then goes on to point out the
similarities of this with free software, which is also given away
for free, and draws the conclusion that open source is similarly
flawed and doomed.
Leaving aside the fact that Microsoft is the biggest proponent of
giving things away and then making money elsewhere - think of all
the bundled "extras" that come free with Windows - what really
unites the dotcom failures is the simple lack of any sensible
business plan.
The open source world as a whole does not need a business plan, and
so can happily give away the fruits of its labours without falling
foul of the iron laws of economics. Companies that wish to build
businesses on open source can do so drawing on alternative revenue
models - subscriptions, support, consultancy, etc - just as
Microsoft hopes to do with its .net initiative.
Mundie then contrasts the profligate dotcoms with virtuous
companies, such as Microsoft, that spend heavily on research and
development. Again, there is an implicit syllogism at the heart of
his arguments: Microsoft spends large sums on confidential R&D;
Microsoft is hugely successful; therefore, spending large amounts
on such R&D guarantees success.
But the truth is Microsoft's research seems to have produced very
little, save perhaps the talking paper-clip. Most of its new ideas
came from companies it has bought, and which usually build on the
insights of one or two gifted people, not teams of
researchers.
Secret R&D is instead a kind of tax that Microsoft would like
to see every software company impose on itself. By contrast, the
open source model allows anyone to benefit from the creativity of
everyone, and so reduces the barrier to entry - another reason why
Microsoft fears this alternative approach. One of the central
insights of the free software world is that if everyone contributes
a little to the common good, the resulting pooled benefits far
outweigh the nominal "loss" of placing that contribution at the
service of the public. The Internet is proof that this approach
works, and even Microsoft must admit that the Internet is effecting
one of the most profound transformations ever seen in business -
its entire .net strategy hinges on the fact.
The Internet also refutes Microsoft's tired allegation that the
open source model "leads to a strong possibility of unhealthy
'forking' of a code base". Not only has the Internet never forked,
but it stands in stark contrast to Microsoft's own perennial and
intentional forking: the deeply-incompatible Windows NT/2000,
Windows 95/98/ME, and Windows CE code-bases.