Summary

  • A 1 August 1995 Dawn report said users could reach online email and other Internet services through Digicom’s head office in Karachi. It described store-and-forward email elsewhere as a process that could leave messages waiting for hours.
  • The same report named Lahore, Islamabad, Peshawar and Faisalabad as places where service was operating, but said online services in those cities were still a future effort. The wording does not establish four live Internet gateways.
  • A 1996 column blamed a shortage of Pakistan Telecommunication Corporation telephone lines for constraining Digicom’s subscriber growth; a 1998 field review still located its online service in Karachi and described email-only networks in three other cities. These sources show a gap between a national claim and a geographically uneven service, not universal access.

The launch had two geographies

The first sentence of a country’s Internet history often compresses several different events: a private computer reaches a foreign network, an institution begins exchanging email, an Internet gateway opens to customers, or a service becomes available in multiple cities. Digicom’s 1995 launch is clearer when those events are kept apart.

A Dawn Wire Service report published on 3 August 1995 carried a Karachi dateline of 1 August. It called the launch the first online Internet link by a company in Pakistan. Arif Iftikhar, identified as Digicom Pakistan’s country manager, said customers could connect to the Internet through the company’s head office in Karachi. The service included online email and other Internet services. A later academic history by Peter Wolcott likewise dates Digicom’s first online service to Karachi in 1995 and says the international connection ran at 64 kilobits per second.

That is the first geography: the point from which an online service was being offered. The second appears in the next lines of the Dawn report. It says service was operational in Lahore, Islamabad, Peshawar and Faisalabad, then adds that efforts were under way to bring online services to those cities in the near future. The article does not spell out the product carried there. Its preceding paragraph explains the existing alternative as off-line email, which stored a user’s message on a provider’s computer and forwarded it later.

The cautious reading is that Digicom had service activity in those four cities, but the online form announced from Karachi had not yet reached them.

The distinction matters. A city name in a launch report is not the same thing as an Internet point of presence. Nor does “online” prove that every household could subscribe, that a line was free when a user called, or that every application was usable. The 1995 report documents an offer and a planned expansion; it does not provide a coverage map, a subscriber count or a household-access measure.

A message could wait for the next connection

The service transition was not simply from “no email” to email. The newspaper described an existing store-and-forward arrangement. A user sent a message to a local provider; that provider kept it and later sent it onward to an Internet user. Iftikhar said the wait could last hours. The account does not identify every network involved or the schedule of the mail runs, so it cannot establish one national system. It does capture the user-facing difference between deferred exchange and an online session.

In a live Internet session, a customer could interact with remote services while connected. Email was one application among others, rather than a package of messages waiting for the provider’s next forwarding opportunity. But the wider Internet did not remove the local access path. A customer still needed a computer, a working telephone connection, a provider’s modem or access line, and enough capacity at the service point. The international route could be operational while these local conditions remained uneven.

This is why a 64-kilobit international link, cited in the 1999 study, must not be treated as the speed each subscriber received. It describes the provider’s connection to the global Internet, not an individual user’s dial-up rate, the number of simultaneous calls, or the quality of a session. The sources do not give those measurements for the 1995 launch.

The telephone line became the contested control point

In June 1996, Dawn columnist M. Ziauddin argued that Pakistan Telecommunication Corporation was restricting online access by limiting dedicated telephone lines. He wrote that Digicom could not accept additional subscribers because it was not being given more lines, and that the few dedicated lines were mostly unavailable. The column quoted a charge of roughly 100 rupees per hour. Those statements are evidence of a contemporary complaint and of the way users experienced access; they are not an independent audit of PTC’s allocation decisions or a complete tariff schedule.

The complaint nonetheless identifies a practical dependency. Digicom could supply an international route and sell a service, but reaching customers required local telephone capacity as well. If the dial-in number or dedicated line pool was constrained, the usable network was smaller than the overseas circuit suggested. The Internet’s distributed architecture did not make the last connection administratively or physically distributed. In this case, an incumbent telephone organization and the provider controlled complementary parts of the route.

A 1998 review by Tariq Mustafa, then submitted to the National Science Research Center’s Connectivity Providers Database, provides a later snapshot. Mustafa described Digicom’s online service in Karachi as serving about 4,000 users. He listed a 64-Kbps satellite connection to Singnet and a separate 128-Kbps connection from Comsats. He also described Digicom’s networks in Islamabad, Lahore and Peshawar as email-only, with hourly rates varying by peak period. Mustafa said his information had been personally collected and warned that it could become outdated. These are reported estimates, not audited accounts.

They nevertheless show why the 1995 phrase “service operational” should not be silently converted into “online Internet access in every named city.”

“First” is a bounded claim, not a national switch

The evidence supports a specific milestone: Digicom’s commercial online Internet service in Karachi in 1995. It does not establish that the country had no earlier Internet-related activity. Pakistan already had store-and-forward email, including services described in a 1999 study as using UUCP, and institutions could have distinct forms of connectivity. “First online service” is therefore narrower than “first email,” “first computer network,” or “the day every Pakistani went online.”

It also does not mean that Digicom’s announcement instantly created a national retail market. Wolcott’s study says 16 companies received licenses for email and Internet services in February 1996, and describes the field as growing after Digicom’s 1995 service. That sequence places the commercial launch before the later licensing round; it does not by itself prove that every licensee operated or that all access barriers disappeared.

The 1995 report is valuable precisely because it preserves an intermediate state. Karachi had a route into online Internet service. In other cities, the same company’s service footprint and its online reach were not identical. Users and institutions encountered different delays, telephone constraints and service modes under one national label. The launch was an opening, not a countrywide switch.