← history
Read stories →
October 29, 1969

ARPANET

First documented ARPANET host-to-host transmission

Technology
ARPANET View full image ↗
Illustration generated by Ideogram 4 (fp8) (48 steps, 1216x832) on an NVIDIA RTX 5090 · 2026-08-27
How was this image made?

The creative brief sent to the image model. These are instructions, not a record of what actually happened.

Original prompt & settings (JSON) ↗

The first famous “message” was really a failed login: L arrived, O arrived, and then the distant computer crashed.

29 October 1969: two letters, then a crash

At UCLA on 29 October 1969, student programmer Charley Kline attempted to log in from an SDS Sigma 7 host to SRI's SDS 940 host in Menlo Park, while SRI programmer Bill Duvall monitored the receiving system.

Kline typed LOGIN, but only L and O reached SRI before the receiving system crashed. The original handwritten UCLA log records the event at 10:30 p.m., noting that the host crashed after the O.

Working within the hour

Duvall corrected the problem, and within about an hour the teams were successfully transmitting and completing the login between the two hosts. Kline and Duvall kept a telephone line open throughout, verbally confirming each character the computers sent.

The data passed through BBN-built Interface Message Processors at UCLA and SRI, making this the first documented host-to-host connection on ARPANET.

What “LO” was not

“LO” was not a deliberately composed message: it was the first two letters of the LOGIN command before the receiving host crashed. Nor was this the first occasion on which any two remote computers had communicated — it was the first documented ARPANET host-to-host transmission.

Only UCLA and SRI were connected that evening. The four-node diagram commonly shown for 1969 represents the network at year's end, and the permanent UCLA–SRI link is separately dated 21 November 1969.

Sources

Researched 23 Aug 2026 7 sources 3 audit passes

How this was checked

Researched from the web into a fact sheet, rewritten from that sheet, then audited against it by a different model. How the pipeline works →

What the sources leave uncertain

  • Institutional retellings differ over the clock time: the Computer History Museum says Kline began typing at around 9 p.m. and that the connection was fully working by 10:30, while UCLA, IEEE, DARPA’s log caption, and many later accounts assign 10:30 p.m. to the documented first transmission. The date and sequence—failed “LO,” repair, successful login—are well supported.

Checked against