Over the last year, several government organizations such as Brazil,
China and the City of Munich have stated their intentions to build a
version of desktop Linux. While their motivations are understandable, I
think an examination of history illustrates that it's an unwise
and there's a better path to achieve their goals.
Government officials I speak to routinely express exasperation with
Microsoft's pricing inflexibility. Even in emerging markets with
reduced capacity to pay for software, Microsoft prices software at
similar rates to the United States. Governments want to create
environments where intellectual property is respected (read: where
their citizens pay
for software, music, and movies). But when incomes are a fraction of
the United States, the undeniable realities are that businesses are
forced to pirate as a simple matter of economic survival. Additionally,
Microsoft uses proprietary formats, switching to other software becomes
difficult, if not impossible, so endless upgrades and maintenance are
forced upon them.
At the same time, the non-stop pounding of hacker attacks on Microsoft
software creates considerable support costs and worrisome national
security issues. Weekly patches, virus services and periodic
reformatting/cleansing of machines add significant costs to government
organizations. Of more concern is the security risk that critical
government data will be leaked out. Whether from sloppy programming
that hackers exploit or from intentional or even unintentional back
doors inserted into the software, there's a real concern that national
security is at risk by depending on Microsoft software.
With this backdrop it's understandable that progressive countries are
looking for alternatives and several have announced they intend to
build their own Linux distribution. They absolutely should insist on
from all software vendors so they can avoid vendor lock-in. Having the
ability to switch improves bargaining leverage and ensures competitive
pricing, but creating their own nationally branded Linux product will
actually cost them more and leave them with an inferior product in the
I think there's an expensive lesson to be learned from airlines. There
was a time last century where nearly
every country thought they needed their own national airline.
Politicians spoke of airlines as strategically important and a national
security imperative (much like they are talking about operating systems
and software now).
Approximately 190 state-owned airlines were created. The vast majority
of those airlines have been economic disasters with total losses well
into the hundreds of billions of dollars. Over time, these countries
commercial companies could provide a much higher level of service at a
lower price. By 1985, 130 states recognized their mistake and had
plans to privatize their airlines, and more countries such as Hungary's
Alitalia, Brazil's Varig, Greece's Olympic, and Poland's LOT are
struggling trying to keep their national airlines afloat. Even
sophisticated and rich
governments like the U.S. and Switzerland are routinely bailing out
airline companies with tax dollars costing billions of dollars.
Meanwhile, low-cost carriers with no government backing are performing
well, reinforcing the point that the free market can deliver services
much more cost effectively than a state-sponsored institution.
The same national furor that created massive money losing airlines is
pushing a few countries like China and Brazil to explore their own
software entities. Countries are desperate for an
alternative, but Microsoft's successful business
tactics, legal or illegal, have vanquished other options. However, the
best way to ensure
quality products at a fair price is to reinvigorate competition by
supporting some of the commercial companies like Novell, Mandriva and
Building your own country-specific Linux operating is an illogical
decision in an
era when privatization of government services is almost universally
recognized to deliver superior results. If the goal is to end up with
affordable software options, commercial companies, which can spread
costs across many countries, will produce a much higher quality product
at a significantly lower price. It's important to note that building a
capable desktop software product requires an army of people - not just
a few programmers. It's imperative to have quality assurance personnel
to test products; field engineers who can partner with companies like
AMD, Intel, Via, ATI, Nvidia to insure compatibility; certification
teams to interface with OEMs (computer manufacturers) to
enable them to ship fully functioning computers; and support personnel
will need to be trained on the product and offer end user deployment,
support and maintenance, and constant, ongoing development will be
required to support the wave of never ending products and technologies.
Each country having their own personnel will make software more
expensive than it needs be.
who feel the need to create their own
operating system, should, using this same
logic, want to have their own automobile factories, apparel
manufacturing plants, telecom companies, microprocessor fabricators,
on. Governments tried to prop up many of these industries in the past.
History has shown time and again that these well-intentioned efforts
become economic disasters. Ultimately, they have turned to commercial
vendors who bring the benefit of economies of scale and competition.
goal is to recapture ultimate control of their data and
information technology systems, which I would concur is critically
important, then that is best accomplished by insisting that all vendors
provide source code and use open formats for all their data.
Fortunately, all the Linux based companies are doing precisely that
today, which is why they deserve support from progressive governments.
It's exciting and unusual to see states taking the lead in technology.
I hope their actions stimulate the marketplace, which will create long
term solutions in the most cost effective manner.