Don Cameron wrote:

> As another regular user and sometimes OS developer, I don't wish to
> throw a wet-blanket over your vision of an OS driven utopia, but I do
> wonder if the vision fits reality...
>
> This would be wonderful, but I do wonder how the person located in this
> African village, with everything involved in terms of literacy, economy
> etc., would determine exactly what is needed from the latest version of
> Red Hat or Debian, let alone build it? - To begin with, who is compiling
> these systems into African village dialects? Also, it's perhaps a sad

If I can say so politely, Don, I think this is missing the point ;-)))
The advantage of OS is mainly in the Business-to-Business arena, not in
Business-to-Consumer. Software writing and adaptation will happen at a
far higher level than a user in a village - let's call it the super
regional level - , be implemented by more local companies - let's call
the national level - , and _used_ by people who quite possibly do not
even know what operating system they are running. Think of an analogy
with website building: when you visit a website, you don't know whether
the OS the web server is running on is Windows NT, Unix or Linux,
whether the web server is Apache, whether the HTML pages have been coded
by hand, or in Dreamweaver or Front Page. And so on. You certainly don't
have to write any software. You just _use_ the site. The image of the
person in the village trying to figure this out is therefore
inappropriate...

Paul Swider's point about the advantage of OS works at the B2B level.
Consider the opposite and current situation: software development is and
will remain impeded in a lot of LDC countries because they don't
represent enough of a market for anyone to develop software for them.
The situation is better in areas like South America, or francophone
Africa, which happen to use world languages with large footprints but in
other parts of Africa and places with smaller languages - particularly
using non-Latin alphabets - it's pretty dire. If you think, well it's OK
because all the people who need to use computers can speak or write one
of these world languages anyway, then you're predefining the debate by
saying only the top 1% in these countries will ever need computers.
Having said that, it seems to me a lot of the really small languages in
Africa and elsewhere will not be viable - 2,000 of the world's
approximately 6,000 languages have combined speakers of less than 50
million people - 33% of languages spoken by less than 1% of world
population . But OS could dramatically lower what might be a viable
'software language unit' down to say 10 million people in an LDC, which
would include a lot of African languages, certainly all major ones...
kiSwahili, Wollof, Zulu...

Here's the OS growth scenario: there are already regional software
companies who can undertake the software writing: Infosys in India has
6,000 programmers and there are strong companies in Latin America. In
Africa, too, South Africa, Ghana and Kenya to my knowledge have software
companies which currently write complex applications. They will grow
bigger and there will be more of them. The bigger gap is perhaps the
implementers: the smaller companies who take software written by others
and install it. As Don has pointed out, despite all their best efforts
Linux is still not an easy end user environment. But keep focused on the
fact that we're not talking about the home- or office- computing model
we've had in the North for the last decade - which is already subtly
changing month by month in any case. It's much more likely to be a
scenario where, for example, a government ministry, let's say health,
orders a drugs tracking system for it's X hundred hospitals and Y
thousand primary health care clinics. Super regional company writes it,
local companies install it. Many of these applications will be 'thin',
where a central system sits pulling in information from data processed
by thousand of client pieces of software - and that means the OS may run
on Windows systems at the user point of view (most major OS programs
have decent functioning versions for Windows). The hospital or client
does its work offline, then sends the data, either on disk or in an
online burst depending on connectivity.

The advantage of OS is then that the health ministry in the country next
door can adapt the system - either for free or buying it since OS
software does not _have_ to be free. Another ministry might want to
adapt it to track another commodity instead of drugs. And so on. As the
number of working applications grows, the value grows hugely because
we're into Metcalfe's law here which says the value of a network
increases exponentially with the growth in number of users. In this
case, the 'network' is the population of useable and comprehensible
software applications adapted to the LDC context.

> Paul also commented on the way by which OS can turn an old computer into
> something perhaps 'larger' than that available through commercial
> software - but of course modern versions of open-source systems are just

I actually think the 'adapt old computer' argument is marginal and
misconceived. The cost of computers is actually only a small part of the
cost of _computing_, as anyone who has run a network knows. All the
support and tuition and everything else are far higher costs. Analysts
reckon the Total Cost of Ownership of a $1,500 PC in a business
environment is currently anything from $12,000 to $20,000. The questions
of how to make computer interfaces accessible to people in an African
village, how to work with policy makers and create viable business
models are far bigger questions than whether more value can be salvaged
out of a few hundred thousand 486 computers. I think it also sets the
expectation level too low for the advantages OS can bring.

> Unfortunately it's the IT industry itself that is most aware of the
> benefits, and more importantly, of the shortcomings of OS... It does

OS's main shortcoming is it is not user friendly at the end-installer
level (which I have dealt with above). Do I detect a hint here of the
idea that OS is somehow not quite 'up to it', too fluffy and
well-intentioned to hack it in the big bad world? If so, tell it to
America's National Security Agency, which has just decided in the wake
of September 11 that Linux is the way to build an ultra secure national
defence software system. Or to Merrill Lynch, who are converting their
entire company operations, including trading however many billions they
trade, to Linux.


John



------------
***GKD is solely supported by EDC, an NGO that is a GKP member***
To post a message, send it to: <[EMAIL PROTECTED]>
To subscribe or unsubscribe, send a message to:
<[EMAIL PROTECTED]>. In the 1st line of the message type:
subscribe gkd OR type: unsubscribe gkd
Archives of previous GKD messages can be found at:
<http://www.edc.org/GLG/gkd/>

Reply via email to