Summary

  • TeleYemen’s 1996 y.net interview described a dial-in number valid throughout Yemen but placed the network node in Sana’a. A common number did not mean that equipment, line quality or connection costs were equally local.
  • In 1998, the company’s Internet product controller reported that most customers were in Sana’a, that poor telephone lines led to modem disconnections, and that an Aden router was planned to avoid a 20-riyal local-call charge. The records do not prove that the router was completed.

A national number, one city’s node

The first public image of y.net was broad: “Next Week, You Can Access the Cyberspace Highway from Yemen!” the Yemen Times announced in an issue dated 26 August 1996. In an interview, TeleYemen manager Christopher Leather said the service was in its final testing stages. The company had missed an internal target of 17 August because system addresses had not yet been assigned. This is a useful distinction in the launch story: the interview is evidence of a planned service and a delayed test timetable, not a record of a particular customer’s first successful session.

The same interview described a universal access number valid from all parts of the country, then located the y.net node in Sana’a. The two statements were not contradictory. A national telephone number could direct calls to one service point. But the gap between the dialling instruction and the node mattered: a customer’s experience still depended on the local telephone network, the path to the service equipment and the charges attached to that path.

The tariff answer makes another distinction necessary. The interview called the access number “free” while listing a one-time sign-up charge of YR 5,750, a monthly subscription of YR 3,500 and YR 12 for each connected minute. It does not explain whether “free” referred to the number itself, a local access arrangement or something else. It cannot safely be read as free Internet service. Nor do these nominal charges, without household-income or comparison data, establish affordability for the country as a whole.

The service was marketed as dial-up Internet, but the physical access line was part of the product. TeleYemen said a modem could work at up to 28.8 Kbps, with actual speed depending on line quality. That qualification turned out to be more than fine print. In August 1998, the company’s Internet Product Controller, Khalid Mohammed Al-Dahbany, said the service had been increased to 28 Kbps, but some customers—particularly where telephone lines were poor—experienced disconnections. TeleYemen had reduced the setting to 14 Kbps. The interview also reported that the company had bought 48 new modems.

It does not include modem logs or a measurement of how many customers were affected, so the account shows an operating constraint without quantifying its reach.

The map behind the subscriber count

Al-Dahbany reported 1,545 customers and 32 home pages hosted on TeleYemen’s server. He said 1,189 customers were in Sana’a, 92 in Taiz, 26 in Mukalla and 102 in Hodeidah. The four city figures add up to 1,409, leaving a difference of 136 from the total. The interview offers no explanation. The gap could reflect omitted places or a different counting basis, but the source does not say; the numbers should remain unreconciled.

Even with that limitation, the interview gives the story a geography that the phrase “throughout the country” alone cannot supply. The company described most customers as being in Sana’a, identified three other cities in its breakdown and discussed a plan to install routers in Aden. The controller said the routers would let customers avoid a 20-riyal local telephone charge between Aden and Sana’a. That makes the routing plan legible as a customer-cost measure as well as a network change: moving a point of access closer could alter what a person paid to reach the service.

The wording is prospective. The interview says the company was in the process of installing routers; it does not confirm commissioning, traffic, new local access numbers or a completed Aden node. It is therefore evidence of the operator’s stated response to a geographic cost problem, not proof that the problem was solved.

A connection was also a service promise

The available contemporary record preserves both a provider’s pitch and complaints from outside it. A short Yemen Times item in January 1998 said customers had been unable to access email for a fourth time in less than three months. It supplies no outage duration, affected-customer count, technical log or independent verification. In September, Yemen Times web-site developer Walid Al-Saqqaf argued for competition and described his own problems with dial-up and Internet services. His article is first-person advocacy, not a network audit.

These accounts cannot establish a national uptime rate or prove that monopoly caused any particular failure. They do show that the design question was not only how to attach a country to the Internet. Customers also depended on the service provider’s modems, servers, telephone access and support. A planned router in Aden addressed one piece of that chain—the cost of reaching Sana’a—while the 1998 interview’s discussion of disconnections pointed to another: the quality of the local loop.

The 1996 and 1998 records thus describe different layers of the same early service. A common number widened the stated reach; a central node concentrated termination; a customer account opened a path to the network; and telephone quality and charges shaped whether that path worked in practice. “Internet access arrived” is not one event when these parts moved on separate schedules.

Sources