open.source: [ linux school computer lab project ]

<< mission >>
to promote universal access to computers and internet in schools,
to empower educators to create and refine open content,
to empower learners with a high-quality ICT learning environment,
and to emblazon our futures in an ICT-driven economy

<< advisory objective >>
we endeavour to advise appropriate organisations on proven
strategies for the mass deployment and maintenance of
high-quality low-cost opensource-based ICT infrastructure
(see article below)

<< industry objective >> will endeavour to mobilise south african
industries to empower every educator and learner in south africa
with computer and internet access to emblazon the future of
our people, our industries, and of our country

<< contact >>
please contact [ frequently asked questions ]

why should learners have access to computers?
learners are stimulated by access to information;
and the opportunity to participate in the global information society
and to gain skills in obtaining and disseminating information
will emblazon their futures in an ICT-driven economy
and imbue them with confidence and enthusiasm for information technology

why should educators have access to computers?
computers and internet communication can be harnessed by educators
to assist with school administration functions;
to empower educators with communication and access to content;
to deliver curriculum-related content and support

what is linux and opensource software?
linux is the kernel of the free computer operating system, gnu/linux,
which includes components from, and,
where "free" especially relates to our freedom to view the source code,
copy, modify and redistribute the software without limitation

what does .^|^. symbolise?
this is an "ascii art" bird, soaring...

<< empowering learners >> believes that learners will empower themselves through information,
and that every learner has the birthright to every opportunity we can afford them
including real access to computers and the internet

<< empowering educators >> believes that educators should be empowered
to create and customise electronic content; that open curriculum
content should be accessible to all educators; that
teacher training in ICTs should be available at all tertiary
education facilities in the country; and that distance e-learning
content should be delivered to all educators

<< opensource and free software >> believes that the culture of learning and the sharing of knowledge
in the scientific tradition, is furthered by the "freedom software movement"
which respects our learners as participants in the software community,
encourages an unhindered enthusiasm for information technology,
and fosters a culture of collaboration in our information society

linuxlab deployed at in London [ posted 24 april 2002 ] participated in the workshop in april in Surrey, England
providing additional technical support for the diskless linux demonstration;
Following the workshop, configured a diskless linux workbench for
the London-based charity for the purposes of recycling computers
as diskless linux terminals for schools in africa

zuza software foundation incorporates [ posted 3 april 2002 ]
zuza software foundation is registered as a section-21 non-profit organisation
to incorporate and projects, and to provide
a vehicle for future projects to promote opensource software for social development;
directors include dwayne bailey of, mike jensen of,
and Dr Sibusiso Sibisi (,

Alexander Sinton School deploys first linuxlab [ posted 15 february 2002 ] is deploying a pilot linuxlab at alexander sinton high school in athlone
( using linux server sponsored by the shuttleworth foundation,
and recycled 486 PCs as diskless X terminals, donated by dept of land affairs,
facilitated by and; with goal of achieving
a cost per seat of under R1000 including network and server goes diskless [ posted 1 November 2001 ], a sister project of, is employing diskless linux desktops
for their translation efforts, including project director dwayne.bailey
who now "even forgets he is using a diskless machine" at Africa Connects [ posted 14 August 2001 by evan.summers ] was presented and demonstrated at the "africa connects"
conference at UCT in July, using refurbished diskless clients
from and a linux server configured by;
and two diskless linux clients operating simultaneously off the linux server website created [ posted 16 June 2001 by evan.summers ]
Following the participation of in a meeting chaired by MEC Helen Zille,
in september 2000 the project was conceived and first approached,
and some eight months later, the website has been created
around June 16, this date choosen in honour of south african school children

This paper was presented at Africa Connects ( at UCT in July 2001

"Champagne at beer prices:
  Back to the future with Linux and Opensource"
  Evan Summers PhD,, 16 June 2001


Using thin-client Linux technology and opensource software,
schools are now able to create world-class computer facilities for
a fraction of the cost. For starters, opensource software does not
require license fees, and includes the robust Linux multi-user
operating system, the state-of-the-art KDE graphical desktop, and
other powerful applications such as StarOffice and the Forte Java
programming environment from Sun Microsystems. But what is really
interesting is that we need only low-spec diskless PCs
functioning as graphical terminals to access such applications on
a Linux server; and these days a powerful Pentium server with
512Mb RAM costs much the same as a Windows workstation.

Even a 486 with 16Mb is sufficient for the function of a graphical
"X terminal", since all users' applications run on the server.
This provides schools with the opportunity to make effective use
of existing low-spec computers, and inexpensive refurbished PCs
which companies are willing to donate because they are not
upgradable to the minimum 64Mb of RAM required to run the latest
Windows systems.

Using recycled computers as low-cost diskless terminals enables
schools to deploy terminals in staff rooms and classrooms,
to facilitate the effective use of ICTs by educators.

Linux infrastructure and opensource software provides the highest
quality learning environment from an IT vocational perspective.
What is exciting is that this higher quality can be achieved at
a lower cost than the purely end-user oriented software environment.

This means that for the first time since the advent of computers
in education, universal access is truely achievable since
diskless Linux technology and opensource software can provide
high-quality computer education facilities at a small fraction
of the cost previously possible -- "Champagne at beer prices" :)


Using thin-client Linux technology and opensource software,
schools are now able to create state-of-the-art computer facilities 
for a fraction of the cost.

For starters, opensource software does not require license fees.
The opensource desktop under Linux includes the KDE and GNOME
graphical desktop environments, KOffice and StarOffice office
suites, Forte Java programming environment and Netscape web browser
and email client.

But what is really interesting is that we need only minimal
"low-spec" diskless PCs functioning as graphical terminals to
access applications on a Linux server. Using a Linux application
server for all users allows us to achieve large economies of
scale, and to reduce the total cost of the required hardware 
for a computer lab.

The cost of a new minimal entry-level PC with no hard drive and
minimal RAM -- basically just a motherboard with a display card
and network card -- and no software licenses, is half the cost of
a typical new PC. (In the case of a minimal diskless PC we are
however omitting multimedia capabilities, ie. audio and video,
where these can be demonstrated on the server only, and CD-ROM
and floppy disks need to be inserted in the server to be accessed
by a user.)

However what is interesting is that even a 486 with 16Mb is
sufficient for the function of a graphical "X terminal", since
all users' applications run on the server. This provides schools
with the opportunity to make use of refurbished PCs, which
companies are willing to donate because they are not upgradable
to the 64Mb of RAM required for Windows 2000. It is important to
note that the most unreliable component in computers in general
and more so in old computers, is the hard drive (since it is a
sensitive mechanical device) and this is precisely the component
we omit when using a PC as an X terminal using Linux diskless

An interesting trend is that organisations in Europe and the US
are looking to offload large quantities of old PCs which are not
suitably upgradable for Windows 2000, since dumping them would
contravene US environmental legislation owing to their lead
content. It is predicted that the number of PCs that will become
obsolete in the US alone over the next ten years, is in the 
hundreds of millions. This could give further impetus to the PC
refurbishing industry in South Africa, creating welcome
employment opportunities, and increasing the availability of
inexpensive low-spec refurbished computers, which are suitable as
diskless Linux X terminals for school labs.

Diskless terminals boot via the network off the application
server. This is achieved by configuring the server appropriately,
and inserting a "boot-ROM" in the network card of the diskless
PC. This boot-ROM then fetches the Linux kernel off the
application server and boots the diskless terminal into Linux. 
It is also possible to use a boot floppy disk, and this 
can be useful for using an existing PC on the network to boot 
into Linux (and access applications on the Linux server) 
without interfering with the existing setup of that PC.

The only software that needs to run on the diskless Linux PC is
the X Windowing System, which provides the user with a graphical
interface to the server. This software requires approximately 8Mb
of RAM, and the Linux kernel a further 4Mb RAM, so 16Mb of RAM is
sufficient for using a PC as an "X Terminal." Moreover, the
function of the X Windowing System is drawing graphics on the
screen and sending user keyboard and mouse input to the server,
and for this purpose even a 486 processor is sufficient. 
(An accelerated graphics card does make changes to the screen 
appear smoother and faster, but unaccelerated graphics does 
not degrade the performance of the applications themselves, 
as they are running on the server.)

All users in the lab are really working on the server, which
should be a fast Pentium PC with 512Mb of RAM (or more). Linux
transparently makes optimal use of the RAM resources of the
server so that only a single copy of each application -- such as
the KDE applications, Netscape and StarOffice -- is loaded into 
RAM and shared by any number of users wishing to execute that 

The response to the user is immediate even if they are using a
486 with 16Mb to access the application, since it is typically
already loaded into RAM on the server, where it is  executed for
all users. On average only a fraction of the resources of a
Pentium 4 CPU are required for a single user -- workstations are 
on average upwards of 95% under-utilised, waiting on user input
(assuming that the RAM resources are sufficient, since swapping
to hard drive is the chief performance killer).

Looking at it from the users' perspective, this architecture
makes a 486 with 16Mb RAM -- when used as an X Terminal in
conjunction with a powerful Linux server -- appear more
responsive, more powerful and more useful than a Pentium
PC with 32Mb operating as a standalone workstation. This
is because modern applications such as StarOffice and the
Forte development environment for Java, require at least
64Mb of RAM to operate effectively, otherwise their use leads to
excessive memory swapping to the hard drive, making the system
unusable. However in case of the X Terminal, the users'
applications execute on the server, which might have 768Mb RAM,
and which shares memory for multiple instances of the same 

An interesting option for schools with older computers which
are unsuitable for running modern applications such as JBuilder
and StarOffice (because they are "slow" and have only 16Mb or 
32Mb of RAM) is to "side-grade" these computers into X Terminals.
These same PCs can then be used most satisfactorily to access
those very same applications on a Linux application server. This
means that a computer lab with "obsolete" PCs can be "modernised"
without having to replace or upgrade its computers. We rather
just add a new Pentium Linux application server with sufficient
RAM, which should cost less than two new entry-level PCs. 

What is also very attractive about this architecture of a single
host computer with multiple diskless and stateless clients, is
that such a setup is simpler to administer and maintain. In fact
UNIX was designed for exactly such multi-user environments in US
universities. Moreover remote system administration and
maintenance via modem is totally feasible, except in the event of
modem or hard drive failure, and these failures can be remotely

Linux has RAID-1 capability in the kernel for transparent disk
mirroring, and so for the cost of a second hard drive in the
server, a hard drive failure would not cause any disruption.
Hard drive failures are the cause of most maintenance problems in
computer labs (since hard drives are the most sensitive devices
in computers, and yet the most critical), and these problems can
be eliminated by using diskless clients and a server with two
redundantly mirrored hard drives.

UNIX servers and X Terminals have been the technology and
architecture of choice for computer labs in US universities in
the 80s and 90s, but it is the recent advent of inexpensive PCs,
diskless Linux technology, and opensource software that makes
this now also the most affordable architecture of all.

This means that for the first time since the advent of computers
in education, schools with more limited financial support are no
longer at a major disadvantage -- since diskless Linux technology
and opensource software can provide computer education facilities
comparable to the best available facilities today at a small
fraction of the cost.

"Champagne at beer prices..." :)

* Evan Summers is a director of 
  and co-founder of the and

The following comment submitted to Business Day
by Dr Sibisi, Deputy Vice-Chancellor (Research and Innovation)
at the University of Cape Town.
"A Haven of Open Choice?"
  Sibusiso Sibisi, 17 May 2001

Robert Davies and Alexsis De Raadt St James make a timely
contribution to the discussion on the digital divide between
countries "Technological garbage of the rich is a poor man's
high-tech haven," (Business Day, 8 May 2001). They propose
cross-border partnerships to transfer decommissioned computers
from developed to developing countries. They can then be
refurbished and deployed at schools, where e-learning centres can
be created for enterprise, training, health and social projects. 

They also note that the use of IT as a platform for development
requires the integration of telecommunication links, access to
computers and software, and relevant content and training.
However, like most discussions on the fundamental challenge of
the digital divide, the rest of the article dwells on a strategy
for acquisition of computers, with passing reference to software.
Yet basic software can involve prohibitive licensing costs, which
pose a significant barrier to narrowing the divide, quite apart
from the costs of connectivity, training and higher level
specialised content and applications. 

Physical infrastructure and enabling software are so intimately
coupled that a conversation on basic infrastructure ought to
speak to both in the same breath. The minimum level of software
needed to add "intelligence" to physical pipes and boxes
comprises operating systems as well as Internet access and basic
desktop productivity tools.

It is thus equally timely to complement the discussion on the
digital divide with a hard look at the availability and
affordability of software. After all, the license cost of a given
software package does not diminish with the age or cost of the PC
on which it resides.

Given the importance of its role in the digital age, the Internet
is the natural starting point.  The strength of the Net derives
from its lack of barriers. It is founded on non-proprietary, open
standards developed by bodies like the Internet Engineering Task
Force, a body open to any interested party.  

Internet services like email and the World Wide Web are based on
"open source" software. The free service enabled by such software
is in itself a sufficient boon to most users.  However, the open
source model goes beyond this - the source code revealing the
inner workings of the software is also made freely available.
This affords any interested party the freedom to probe, modify
and learn from such code without compromising the integrity of
the tried and tested working version.

Hence schools can benefit not only from refurbished computers,
but also from open source software for connectivity and desktop
use. They need encouragement and support to build their own Web
sites on products such as Apache and, notably, the flagship of
the open source movement: the Linux operating system.  

The primary purpose of IT at schools is presumably to empower
learners to become fluent in the fundamental principles of IT
through guidance as well as self-initiated probing and
exploration. This suggests an environment that allows routine
productivity as well as the freedom to experiment.

The networked open source desktop computer equipped with Linux
and, say, StarOffice (for documents and spreadsheets) is a sound
base configuration for such purposes. Indeed, this would suffice
for many people's regular needs. 

Beyond this, learners would soon discover the Free Software
Foundation site and similar online repositories of the wealth of
open source offerings that are available to be freely downloaded.  

Even more significantly, the eager learner has the freedom to
"hack the source" as part of the learning experience. Indeed,
such a learner may ultimately wish to join the distributed
community of online volunteers that form the typical mode of open
source software development; entry requirements are merely
interest and constructive ideas. 

Peer scrutiny ensures the generation of robust software within
this vast community of collaborating volunteers. Akin to the
culture that has faithfully served the advance of scientific
research over generations, the virtues of this model have begun
to capture the attention of the commercial world. The raft of
open source articles in The Economist in recent months is ample
demonstration of this recognition. 

So, what is the downside? A corporate need to capture business
processes online will more than likely be met by a proprietary
software package at this stage. Equally, delivering some
specialised educational or administrative solution to an
education or health care institution may well point to a specific
proprietary product. 

The intention is not to embark on an evangelical crusade against
proprietary software. It is to highlight the choice where it
exists so that properly informed decisions can be made. IT
procurement policy for government institutions in general might
do well to consider open source solutions on an equal footing to
proprietary alternatives, as has already been suggested in
several other countries, both developed and developing.

This choice also speaks to another important matter. Respect for
intellectual property dictates that one should neither use nor
encourage the use of pirated proprietary software. However,
unlike the pharmaceutical dilemma, there is no law against
developing (superior) alternatives - at least in the absence of
patents on software, or even more sinister broadly scoped patents
on algorithms. Piracy is simply inapplicable to open source
alternatives, for there are no proprietary restrictions in the
first place.  With that recedes the prospect of the dreaded
piracy-busting early dawn raid. 

It would be a profound irony if an earnest attempt to bridge the
digital divide unravelled because of prohibitive software license
costs. Even with educational discounts and so forth, the
proprietary model does not offer the unfettered choice to
participate in the development or modification of the very
technology that can only increasingly become an intimate part of
any developing society as it ventures into a digital future.

The poor man's high-tech haven will have been truly attained when
there is open choice.

Poverty might then turn out to be short-lived.

* Dr Sibisi is Deputy Vice-Chancellor (Research and Innovation)
  at the University of Cape Town

please contact

this site best viewed through an opensource desktop