Summary

  • A June 1996 account described a University of Guyana CUNET service: a DOS Waffle bulletin board with email exchanged over UUCP to the University of Puerto Rico.
  • The correspondent reported roughly 70 users and reliable operation since January 1995, but also a few interruptions, hardware needs, university-paid phone costs and volunteer troubleshooting.

“Internet service” in a 1996 Guyana update meant a particular, bounded thing: a bulletin-board system at the University, not a direct international circuit at every desk. Andrew Mancey described the University of Guyana’s CUNET service as DOS Waffle software providing a BBS and email over a UUCP link to the University of Puerto Rico. The route carried messages between systems; the report does not say subscribers had continuous, interactive IP access. [S1]

The account’s most revealing phrase was not a speed figure. Mancey said the service had worked reliably since January 1995, with a few interruptions attributed to hurricanes, exploding voltage stabilisers and other minor problems. “Reliable” here was a contemporary operator’s assessment of a service that still experienced outages. The archive records neither an uptime series nor the duration or frequency of the interruptions. [S1]

By June 1996, the system had about 70 users, according to Mancey, including about 20 outside the University. Those subscribers helped reduce phone costs, which the University paid. This arrangement made the university host more than a technical relay: it concentrated the local call expense and the equipment needed to reach the overseas UUCP peer. The source gives no tariff, total bill, call schedule or accounting formula, so it cannot show how much any subscriber paid or how the contribution was calculated. [S1]

People and hardware were also part of the route. Mancey hoped to upgrade the system—probably to Linux—when equipment became available. He said he was formally a physics lecturer and had taken on email support and troubleshooting voluntarily because no one else was doing it. The 1996 note therefore documents a service whose continuity rested partly on labour that was not described as a dedicated network post. It does not identify a staffing roster or establish how often that volunteer work was needed. [S1]

The same update says interested parties were waiting for the telephone company to offer overseas data lines at a reasonable price. A July 1996 note described work on a GT&T ISP design with dial-up and leased-line options, initially in 56K-to-256K bandwidth segments. That was a proposal, not proof that the service launched. In August 1997, Mancey reported that a UNDP-supported Sustainable Development Networking Programme operation ran only during working hours while it awaited a standby generator. These records show separate plans and operating constraints; they do not prove that SDNP inherited or replaced the University’s CUNET system.

[S1][S2][S4]

A UK observer’s October 1996 ping reached one gol.net.gy host: 10 replies from 14 packets, or 28% loss, with reply times between 811 and 876 milliseconds. It is a useful dated observation of one host from one vantage point, not a test of the CUNET mail path or a measurement of service across Guyana. A 1999 NSRC update later listed several ISPs using leased lines from GT&T and a GT&T satellite link upstream via MCI. That later snapshot cannot fill the missing transition history. [S3][S5]

The evidence supports a modest but important conclusion. International email reach preceded the record of a broad provider market, but it relied on a local institution to house the gateway, pay phone costs and supply a point of coordination. The June 1996 message is a first-hand operational account preserved in an archive, not a census. Its estimates and judgments belong to its author; they establish neither national adoption nor the exact start, cost or eventual fate of the service.

Sources