As promised, here is a summary of responses to my request for information about web users in developing countries. The response has been fantastic (see below for list of contributors) - thank you very much!

It would appear that the major obstacle in getting access does not necessarily lie in the computers and browsers being used - although there are exceptions, the equipment being used often seems to be just as capable and contemporary as equipment in industrialised countries.

The greater problem appears to lie in the connections to the Internet. Many places seem to have poor telecoms and network infrastructure, which translate into quite limited bandwidth, even where the user has access to technologies capable of high bandwidth. This constraint is often compounded by the relative high cost associated with Internet access, whether from home, work or Internet cafe, and frequent connection interruptions. In other words, downloading of pages and other files is a slow, expensive and frustrating process.

I realise that this sample will not necessarily be very representative but might I be right in assuming that older computers often will not be in use for Internet access anyway?

The findings do, however, fit in very well with observations from our own webserver's logfiles - although we appear to get a fair number of hits from developing countries barely any are via older or text-only browsers. This may, of course, be because users with older browsers are just avoiding our site!

In conclusion, people who create web pages should strive to cater for limited bandwidth by:

* keeping images to a minimum and ensuring they are compressed as much as possible.
* avoiding unnecessary presentational technologies, such as imagemaps, Java, Flash, etc.
* clearly signposting the size of downloads, such as PDFs, and also ensuring that these downloads can be 'previewed' as well as considering providing HTML alternatives.
* above everything else, keeping it simple!

More generally there is also a strong case for ensuring access through as many different types and versions of browsers, including text-only, as possible and avoiding frames. Although it will have resource implications it may also be worth considering providing versions in languages other than English and using CD-ROMs as an intermediate technology.

I have just realised this is pretty long for a summary and I am sure all this must seem to some like I have just been stating the obvious. In my defence I can only say that this exercise has been a real eye-opener for me personally and I hope to help ensure others facing the same tasks as me have better access to this type of information!

Best wishes - and thanks again,

Kenny

Thanks to the following for their contributions:

Douglas Ball, University of Zimbabwe, Zimbabwe
Peter Burgess, AfriFund Management Ltd, USA
Penny Cambell, KwaZulu-Natal Health Department, South Africa
Candy Day, Health Systems Trust, South Africa
David FitzSimons, World Health Organization, Switzerland
Zelda Jeffers, Nurse, Brazil
Miranda Lonsdale, BMJ Publishing Group, UK
Stixy Misheck Nyaluso, Malawi College of Medicine, Malawi
Helga Patrikios, University of Zimbabwe, Zimbabwe
Lenny Rhine, University of Florida, USA
Vasan Seshadri, Oxford University, UK
Sanjay Sood, National Telemedicine Project, India
Vasumathi Sriganesh, QMed Services Ltd, India
Chris Zielinski, Information Waystations and Staging Posts Network, UK

Several respondents (Irene Bertrand [WHO], Helga Patrikios [Univ Zimbabawe] as well as Tim Beanland and Ruth Brassington [Tropical Medicine Resource, Wellcome Trust, UK]) have provided personal contacts and other sources of information which, time permitting, I hope to follow up on and report back on.

['HIF-net at WHO' profile: Kenny Hermansen works at the London School of Hygiene & Tropical Medicine as Web Development Officer.]