Two countries, two systems for promoting free software

by Andy Oram

I had two fascinating interviews today about two initiatives, one in
the People's Republic of China and one in the United States, for
promoting public-sector computer applications--and thereby the use of
open-source software.

Beijing's rural, low-income, and educational initiative



The Beijing municipal government is getting ready to launch version
1.0 of a system aimed at bringing applications to rural areas. They
plan to use the same model for low- and middle-income residents of
cities, small and medium enterprises, and educational institutions. Linux (through the well-known

Red Flag

distribution) is a key part of the rural system. What's interesting for me
is the role of government.



Years ago, the Chinese government tried to promote computer
applications by simply funding ISVs to create applications. As
Americans might have guessed, these failed to produce significant
results and turned out to be wasteful. The government realized they
needed a different path.



The current approach is: "It is better to discover a market for ISVs
than just give funding support to ISVs," as summarized for me by
Liming Li of Intel China Ltd. Some details of this intriguing
philosophy came through in a hastily arranged interview Mr. Li set up
between me and Hu Qing Hua, director of an organization called the
Beijing Software Industry Productivity Center.



Mr. Li and I met at an O'Reilly party last night (he is a contributor
to an upcoming book, Open Sources 2.0) and realized after a brief talk
that China's new effort to promote rural and low-income development
would make a great subject for an article. He tried to set up a
meeting that evening with Mr. Hu, but could not. So I rushed out of
another meeting at 9 AM today and sped down to South San Francisco
with Mr. Li, guided by the car's navigational system, to hold the
interview in a spare hour before Mr. Hu had to take a plane.



Mr. Hu's English was fairly good, but we relied on Mr. Li for
translation. I still may have misinterpreted some statements, so I can
only report my best understanding of the topics.



The slogan for Mr.Hu's center is, "If you want to do software business
in China, go to Beijing through the Beijing Software Industry
Productivity Center." In the complex business/government environment
in China, the Center manages to hook up multinationals with local
companies--but now they're taking a significant next step and
representing the needs of ordinary Chinese end-users as well. The
project is named Strong Wind (Chang Feng).



In this new model, the Center provides funding and organization for
technical experts to meet with users and define detailed requirements
for some application to support farming, education, etc. These are
turned into applications based on commodity hardware from companies
such as Intel and local PC OEMs, running Red Flag Linux. Each project
has a coordinating committee that involves the various actors--local
and international, hardware and software--responsible for putting
together the solution.



The Center aims, therefore, not to request particular pieces of a
solution, but to put together have a comprehensive business model
covering the whole environment and to keep this in front of the
vendors. To save costs, Red Flag Linux is used and applications are
doled out to local ISVs. Large vendors such as Intel do integration
and support.



The Center keeps s firm hand on the project too, through an
organization of 80 people dedicated to testing and validation.
Considering that 20 packages may make up a single software solution,
this testing is very important. And after two months of testing in the
Center, some 500 units are sent to the end-users for field
testing. The vendors can then refine the system before its ultimate
deployment. Strong Wind is nearly ready for a 1.0 release, and they
plan to have 10,000 units deployed among users by the end of this
year.



Further funding can be provided for particular applications that the
Center recognizes a need for, based on the requirements gathering
mentioned earlier.



So far as they know, there is no one inside or outside China using
this particular model, but a lot of interest is being shown by
governments in South-East Asia, notably Malaysia, Thailand, and Viet
Nam.



The government is also working with the telephone company to bring
fairly higher bandwidth (such as ADSL) to as many areas as possible.



Because IT and Internet access are often associated with grass-roots
communications, I tried to formulate a polite question about whether
the system under development could contribute to rural and low-income
people communicating with each other. But the concept was too
abstract, so I couldn't get a conversation going around it. The system
is apparently designed around delivering content.

U.S. government representatives meet at LinuxWorld



The other news I got today was about "Linux in Government Day," held
at
LinuxWorld in San Francisco
for 50 representatives, aides, municipal employees, and others
interested in bringing free software into government applications. The
event was organized by Leon Shiman, who is an open
source developer, the secretary of
X.org,
and advisor to Massachusetts CIO Peter Quinn on bridging government IT
and community processes.




It was interesting to find out that government agencies develop
support for their software among their own staff, even if the software
is commercially provided and comes with support contracts (which often
go unused). This makes it much more feasible, in terms of staff
readiness, for them to adopt open source.



Government attendees were concerned with spreading code development
across government boundaries. They know they need to create and
sustain cooperation among agencies and sectors.



Leon's key point is that cost is not the main factor to consider in
choosing open source; rather; the license is fundamental. Free
software flourishes because it draws out the creativity of many
diverse contributors, but he says "You can't develop together if you
can't see into the software." He pointed to the license for the X
Window System as the most open (a subject of eternal debate, of
course), because it allows proprietary usage while protecting the
openness of code base. X is also uniquely "non-sectarian" regarding
operating systems and platforms--it has been ported just about
everywhere.



Highlights reported to me by Leon included:






  • An address by law professor and free software supporter Eben Moglen on
    the legal issues involved in governments using open source.







  • Reports by the CIOs of California and Massachusetts on state policies.
    Quinn closed the session with a detailed summary of the implementation
    issues that arose in the state when implementing open source policy:
    staff education, technical transition, etc.







  • A talk by an expert on Latin American government usage of open source,
    explaining its "dramatic adoption" and the roles played by in
    communities and government.







  • A talk by the CEO of
    Trolltech,
    about how that company fashioned a bridge between open-source and
    proprietary development, so their Qt libraries can be the basis for
    the highly popular KDE while still offering commercial licenses. This
    is important because governments, like many private institutions,
    worry about whether they can keep part of their code closed while
    making use of free software.







  • Two talks that demonstrated open source's strength through
    flexibility: one on Firefox functionality (central to government
    concerns because so many applications are delivered through the
    browser) and Linux Terminal Server Project.






In both China and the U.S., systems are bending in response to
changing environments. The supporters of open source and open
standards take widely varying steps to deal with the current state of
each system. I think everyone will be feeling their ways forward as
systems and users evolve.