Re: [GKD-DOTCOM] How Much Bandwidth is Necessary?

2003-11-06 Thread Sandra Roberts
My name is Sandra Roberts, I work with a project designed to support ICT
initiatives in the SADC (Southern African Development community) region.
We are represented currently in 12 of the 13 countries in SADC and have
nodal points in Tanzania, Zambia and South Africa.

Recently we conducted research on telecentres in SADC.  Here are my
answers to the questions.

KEY QUESTIONS:

 1. Are high-bandwidth connections necessary, or even important, to
 making a real impact on development? Or are the costs and problems
 inherent in establishing such connectivity too high -- and unsustainable
 -- for underserved areas?

Connectivity costs in Africa are too high, whether it is in urban or
underserved rural areas.  High bandwidth connections largely a dream in
many areas.  Also importantly there is not enough support for individual
telecentres who are often in very isolated areas.


 2. Are there cases that demonstrate the value of low-bandwidth (e.g.,
 store-and-forward email, packet radio) solutions to provide critical
 information access to under-served communities? How successful have they
 been?

 3. Can information distribution centers (e.g., public access
 telecenters) offer a viable economic solution to a community's
 information needs, by, in effect, sharing a single high-bandwidth
 connection among many users, and thus spreading the cost?

Telecentres and community multimedia centres have not fared very well in
Africa, this is due, in part to exorbitant connection costs, but also
because they need dynamic leadership. Management and technological
skills, yes, but leadership which is adaptive to the various conditions
which a telecentre/ CMC will face during its lifespan.  Unfortunately
practical barriers include high staff turn over - people with the skills
to run telecentres could get relatively high paying jobs elsewhere, and
have more security than telecentres can offer.

The practical reality is that many telecentres are donor dependent and
have no plans to become self sustaining, or possibly have plans and
haven't implemented them.

So, yes, they can, but practically they often don't.


 4. Are there new protocols that make more efficient use of the bandwidth
 that is available? For example, what role can the newer wireless
 technologies (e.g. Wi-Fi, MESH networks) play in bringing sufficient
 connectivity to underserved communities? Are the costs and maintenance
 demands of these technologies sustainable?

New technologies require new skills sets and new support mechanisms.
They should be adopted, but possibly not immediately as soon as the
technologies are available.  I think universities should be key in
experimenting with new technologies and slowly developing plans for
incorporation into their countries.


Please look at our site, it will be launched on the 17 November 2003.
www.cinsa.info


Sandra Roberts
Research and Information Coordinator
CINSA Project
SANGONeT
Tel: 27 11 838 6943/4
Fax: 27 11 492 1058
Email: [EMAIL PROTECTED]
Website: www.cinsa.info; www.sangonet.org.za





This DOT-COM Discussion is funded by the dot-ORG USAID Cooperative
Agreement, and hosted by GKD. http://www.dot-com-alliance.org provides
more information.
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
For the GKD database, with past messages:
http://www.GKDknowledge.org


Re: [GKD-DOTCOM] How Much Bandwidth is Necessary?

2003-11-06 Thread Mark Lediard
Here is a bandwidth sharing option I have been thinking about. I plan to
deploy this in Indonesia soon. The idea is to get a business, (perhaps a
bank?) that has some bandwidth in a district setting, to share its
bandwidth with a health center through a wireless access point placed
somewhere near the health facility. If a local business is willing to
share bandwidth with a health facility, the base costs of a router and
wireless access point to enable that are around 125 Euros. Security
software and routines exists to make sure that no one at the health
center can hack the host. Then, in pleasant and practical
public-private collaboration, the health entity that gets to connect
wirelessly via the on-all-the-time connection at no extra charge to the
host, can have free bandwidth to use for their ICT needs.

This gives a local business an easy and low-cost way to act in a way
that is socially responsible. There may also be a way for the health
unit to recover some costs by charging some fees for offering VOIP
(Voice over internet protocol) services such as the use of SKYPE or
www.net2phone for contact. Think about how you might apply such a
voluntary Robin Hood scheme. It's technically feasible. I have done it
already on a small scale. In fact, this note come to you via a wireless
setup...


Mark Lediard





This DOT-COM Discussion is funded by the dot-ORG USAID Cooperative
Agreement, and hosted by GKD. http://www.dot-com-alliance.org provides
more information.
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
For the GKD database, with past messages:
http://www.GKDknowledge.org


Re: [GKD-DOTCOM] Connectivity Is Not The Right Word

2003-11-06 Thread Guido Sohne
On Wed, 2003-11-05 at 09:26, Peter Burgess wrote:

 My vote is for narrowband EVERYWHERE connecting little local nodes.
 Improve the local infrastructure, and don't focus just on the
 international part of it. And my vote is for using technology to reduce
 the cost and price of basic communication rather than to maximise
 revenue for the technology producers by selling more and more complexity
 that adds a lot to the visual experience but not very much at all to the
 underlying messages being communicated.

This seems to assume that one size fits all. That narrowband will be
adequate because it serves the needs of more people, the vast majority
in fact. Or another way of saying it is they don't yet need bulk data.

Maybe I am biased, not being part of the vast majority in my identity
makeup, but I think that while moving the masses forward, you shouldn't
lose sight of the possibility that real change sometimes starts from the
ones who are few, so to speak. The best analogy I have is from Snow
Crash, where the infocrats are described as feeding off 'biomass' just
like whales feed off krill. Both parts are important for a successful,
functioning 'system' IMHO.

Putting narrowband everywhere and forgetting about broadband can stifle
the growth of a small number of different, more modern, more innovative
actors. Not everyone needs broadband, but don't forget those who do!

I also agree that connectivity is not the whole issue. More the tip of
the iceberg. Education, better health care and more capital (monetary,
HR etc) are much, much more important. Connectivity should be reduced to
the status of a tool that implements, or helps implement, a deeper, more
fundamental strategy. Without a clear workable deep strategy, I don't
think we should even start on solving connectivity.

Put in yet another way, like Simon alluded to, we should work on
connecting the people locally but without knowing or planning for what
they are going to do with that connectivity is another matter. You can
place as much broadband in a village as you please but when they don't
know how to leverage this bandwidth, it just 'lowers the barriers' as
the gentleman from Cisco mentioned - a roundabout means of saying that
there are still some problems ...






This DOT-COM Discussion is funded by the dot-ORG USAID Cooperative
Agreement, and hosted by GKD. http://www.dot-com-alliance.org provides
more information.
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
For the GKD database, with past messages:
http://www.GKDknowledge.org


Re: [GKD-DOTCOM] Bringing Connectivity to Under-Served Communities

2003-11-06 Thread Sudhakar Chandra
Hello Robert and others,

On 11/05/03 09:14, Robert Miller wrote:
 With regard to Ahmed's note and the great work he is doing by bringing
 Internet literacy to the students in his university in Nigeria, what if
 you could connect one Campus Content server to that Internet connection
 and locally store many times the content in the US Library of Congress?
 What if this provided simultaneous access for several hundred users on
 campus?

That is a great idea. When I connected a small college in Kenya to the
internet via a 64K VSAT connection, I installed a cacheing transparent
proxy server. The first time someone downloaded something, the content
would be fetched from the server and stored in the proxy server. For
all subsequent downloads, the content would be sent to the local
requestor's browser from the cache and not from the server. This vastly
improved performance and download speeds. Another advantage of using
proxy servers is that the administrator can set up access lists and
access times. So, for example, an administrator can configure the proxy
such that when a class is in progress, the students would only be able
to access the prescribed materials and nothing else.

A week after I connected my college, I discovered that the network usage
was inordinately high. Looking at the logs I saw too many connections
going to Brazil! It was a worm that had infected the lab computers. The
network usage was taking up precious bandwidth from legitimate packets.
I wrote a two line rule in the proxy server to drop all requests going
to the Brazilian site and the network utilization dropped dramatically.

Thaths
-- 
Slacker At Largehttp://openscroll.org/
Key fingerprint = 8A 84 2E 67 10 9A 64 03  24 38 B6 AB 1B 6E 8C E4






This DOT-COM Discussion is funded by the dot-ORG USAID Cooperative
Agreement, and hosted by GKD. http://www.dot-com-alliance.org provides
more information.
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
For the GKD database, with past messages:
http://www.GKDknowledge.org


Re: [GKD-DOTCOM] How Much Bandwidth is Necessary?

2003-11-06 Thread Njideka Ugwuegbu
Mark Lediard [EMAIL PROTECTED] wrote:

 Here is a bandwidth sharing option I have been thinking about. I plan to
 deploy this in Indonesia soon. The idea is to get a business, (perhaps a
 bank?) that has some bandwidth in a district setting, to share its
 bandwidth with a health center through a wireless access point placed
 somewhere near the health facility. If a local business is willing to
 share bandwidth with a health facility, the base costs of a router and
 wireless access point to enable that are around 125 Euros. Security
 software and routines exists to make sure that no one at the health
 center can hack the host. Then, in pleasant and practical
 public-private collaboration, the health entity that gets to connect
 wirelessly via the on-all-the-time connection at no extra charge to the
 host, can have free bandwidth to use for their ICT needs.
 
 This gives a local business an easy and low-cost way to act in a way
 that is socially responsible. There may also be a way for the health
 unit to recover some costs by charging some fees for offering VOIP
 (Voice over internet protocol) services such as the use of SKYPE or
 www.net2phone for contact. Think about how you might apply such a
 voluntary Robin Hood scheme. It's technically feasible. I have done it
 already on a small scale. In fact, this note comes to you via a wireless
 setup...


Mark, interesting scheme. The most challenging part, I think, will be
convincing the bank that sharing their bandwidth with a local health
center will be a socially responsible thing to do, especially if they
are aware that the health center will then turn around and use their new
connected state to make money.

I think even the banks will want to set up a monthly payment plan with
the health center where the center pays for their bandwidth usage - even
a nominal fee.



Njideka Ugwuegbu
Reuters Digital Vision Fellow
Stanford University
http://reuters.stanford.edu/
http://www.youthfortechnology.org







This DOT-COM Discussion is funded by the dot-ORG USAID Cooperative
Agreement, and hosted by GKD. http://www.dot-com-alliance.org provides
more information.
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
For the GKD database, with past messages:
http://www.GKDknowledge.org