Kenya's technology experts are developing a community Web site to monitor quality of service by ISPs.

According to Michuki Mwangi, one of the leaders in the project, it will be Kenya's largest technology-related online community-based initiative, where all members are expected to come up with a solution that will make the project work better.

"It is a joint community venture. We will develop applications and all the information will be available on the Web site," Mwangi said.

The idea of a community Web site was brought up at one of the monthly meetings of Kenyan techies, known as "Skunkworks." During the meeting, hosted at Google East Africa offices, members of the IT community underscored the need to monitor Internet services.

Joseph Mucheru, Google's East Africa office lead, promised to support the team in developing and maintaining the initiative.

"We will develop the applications that will run, put them online and set the criteria of what we want to achieve," Mwangi said.

Noting that the Web site is still in the early stages, Mwangi said that the exercise is designed to depend on volunteers. These volunteers will host applications to monitor service, via ISPs, to the Kenya Internet Exchange Point (KIXP), and also monitor international traffic.

As the applications provide information, it will be collated in a Web site that anyone can access.

Asked about involving the ISPs, Mwangi said that the providers could support or reject the initiative, depending on how they view it.

There has been an argument that Kenyan providers thrive on users' ignorance of technology matters. Mwangi thinks the problem lies in users' failure to understand or question the contention ratios provided by ISPs.

"People do not understand the contention ratios. People pay dedicated costs, yet they are given contended bandwidth. It could be 1:4 or 1:12, yet the customer pays for dedicated capacity," Mwangi notes.

To determine the capacity provided, Mwangi noted that one can measure the throughput at peak time and off-peak. Usually, the capacity is faster at off-peak hours (after 5 p.m.) compared to peak hours (8 a.m.-5 p.m.).

If the capacity is dedicated, then there should be no difference in performance during day and night. Organizations provide graphs, but Mwangi says that the graphs only show an average of what one is using, and not the actual bandwidth at peak time.

The emergence of video uploads and downloads has affected the bandwidth provision. Downloads take a shorter period at night compared to daytime. According to Mwangi, the downloads provide another way of monitoring the bandwidth contention ratio. This is because the speeds or performance during the day and at night should be the same if the bandwidth provided is the same.