Summary

  • Papua New Guinea’s 1997 “arrival” has several dates because the sources describe different layers: a carrier launch, the Tiare international gateway, provider activity, and a university website.
  • The strongest evidence supports a staged rollout. It does not establish one universally available service date, a single uncontested first ISP, or widespread access.

The date depends on what the word launch names.

A 1997 paper prepared for Papua New Guinea’s Waigani Seminar says the Internet was launched by PNG Post and Telikom on 12 March. It then places the launch of DataNets at an Internet seminar “a few weeks” later, followed by advertisements from two providers, and puts the first University of Papua New Guinea website on 19 August. A 1997 Pacific journalism review gives another threshold: it says the country came online in May. A later regulatory account dates the commencement of Internet services to April, when Telikom’s Tiare gateway was launched. Those dates should not be averaged into a false precision.

They describe different events, and the record does not contain a contemporaneous service log that proves when the first public customer sent a successful packet abroad.

The distinction matters because a national network does not become usable through one announcement. International carriage must reach a gateway. A service provider must be able to buy or lease access to that gateway. A customer needs a last-mile telephone line, modem, computer and money for both the connection and the service. A university can use email before a public Internet product exists; a newspaper can publish a website after the network is already running. Each is a real milestone, but none substitutes for the others.

A carrier date is not a user date

The Waigani paper by Nathan Kwasam is valuable because it is close to the events and names institutions, dates and public debates. It reports that PNG Post and Telikom launched the Internet on 12 March 1997. It also says that a first Internet seminar at Port Moresby’s Gateway Hotel came weeks later, and that DataNets was launched there as an ISP. After the seminar, two providers advertised in The National and Post-Courier. The account then marks UPNG’s first website in August.

That chronology can show that the carrier regarded March as a launch. By itself it cannot tell us which route was operational, who was eligible to use it, what service was offered, how often an international circuit was available, or whether an individual customer could complete a session. Kwasam’s paper is explicit about the limits of its own evidence: he says there had been no empirical study of Internet effects in the country and that providers did not give researchers usage statistics freely. His date is therefore a reported event, not a national access census.

A Pacific Journalism Review article published in 1997 places Papua New Guinea “on line” in May. The same account describes web publishing by local newspapers and the awkward work of building sites while users waited for connections. Its vantage point is media practice, not Telikom’s ceremonial calendar. It helps explain why an author watching a first public connection might choose May even if a carrier event happened earlier.

The gateway between trunk and provider

A separate set of records names April. In a 2008 University of Papua New Guinea paper on telecom regulation, lecturer and lawyer Lois Stanley dates the commencement of Internet services to April 1997. She distinguishes the retail providers from the gateway: Pacific Mobile Communications (PMC), a Telikom subsidiary, operated the Internet gateway; Telikom supplied wholesale bandwidth; PMC sold access under the Tiare name to ISPs; and those providers then sold services to users. She lists five licensed providers in the later market, including Datec and Online South Pacific, which used the Data Nets name.

That diagram makes the date differences easier to read. March can mark a carrier launch claim. April can mark a gateway service. The Internet seminar and provider advertisements can mark the appearance of a retail layer. May can mark an outside observer’s view that the service had become publicly reachable. August marks a visible institutional website. The 2008 account is a retrospective regulatory explanation, so it does not prove that every piece of this later-described arrangement was already identical on the first day. But it explains why “the Internet launched” and “an ISP started selling access” are separate propositions.

A record in the Network Startup Resource Center’s provider database says Datec’s entry was created on 10 May 1997 and describes Datec as “the first ISP in PNG.” The record was updated in 2005. Its creation date is a useful administrative trace: a provider account existed in the directory by May. The phrase “first ISP” is the provider’s own description preserved by the database, not an independently verified measure of first customer traffic. It also sits beside Kwasam’s account of DataNets as the first provider launched at the seminar.

The sources may be using different definitions—launch, registration, first offering, or first listed provider. They do not let us collapse Datec and DataNets into one company or select a winner without more evidence.

The network before the public story

The Department of Information and Communications Technology’s history records that the old Post and Telecommunication Corporation was split and corporatized in 1996, creating Telikom PNG, Post PNG and PANGTEL. An International Telecommunication Union document prepared for the 1998 World Telecommunication Development Conference says Telikom was formed in January 1997, was wholly government owned, and held a monopoly licence for telecommunications services through 2002. The document says Internet service had already “proven successful” by the time of its report.

The same ITU report gives the physical context. It describes international facilities at Gerehu in Port Moresby, links to Optus in Australia, satellite systems and a domestic network reaching 13 provinces. It also records that 85 per cent of the population lived in rural communities with access to only 4 per cent of the telephone lines. Those figures do not measure Internet access. They do warn against treating a gateway in the capital as national coverage. A network could connect internationally while the ability to reach it remained concentrated in a few urban institutions and households.

A 2002 response from Noel Mobiha, associated with the PNG Education and Research Network, gives a later snapshot of the commercial architecture: Telikom held the carrier monopoly; PMC leased wholesale capacity from Telikom; and five Internet Service Distributors sold service onward. Mobiha reported 6 Mbps of international capacity to Australia at that time, split between links to Telstra and Optus. That number belongs to 2002, not 1997. The value of the report is structural: it shows the trunk, gateway and retail provider remained different operating roles.

A launch could precede a public

The 1997 Waigani Seminar treated Internet service as a question about more than technology. The seminar’s report emphasized the cost of infrastructure and the need for information produced in Papua New Guinea, including content in local languages. Kwasam asked who owned the infrastructure, who could afford access, and who controlled the service. His paper noted that training could itself be expensive and that users needed skills to search effectively while telecommunications charges accumulated.

That concern prevents a familiar historical shortcut. A country can be “connected” in the sense that an international path exists, while most people cannot use it. A carrier may announce a service before there are competing providers. A provider may advertise while its addressable market is limited to Port Moresby. A website can make local publishing visible without making access cheap or geographically broad. The sources support all these distinctions; they do not offer a reliable user count for March, April or May.

Papua New Guinea’s 1997 story is thus not a contest over a single arrival date. It is a sequence of interfaces: the carrier’s announcement, an international gateway, companies packaging service for customers, researchers and journalists experimenting with the network, and institutions putting local material online. When later histories compress the sequence into one date, they hide the practical boundary that mattered most: which organization carried traffic, which one could sell a connection, and who had the means and line to use it.

Sources