In 1978, members of the Montreal Amateur Radio Club began sending ASCII data over VHF frequencies with homebuilt equipment, after Canadian authorities authorized the experiments. It was a bold idea for its time: instead of speaking into a microphone, a ham would let a computer chop a message into small bursts, send each burst with an address and error checking, and let the station at the other end reassemble it. When the United States Federal Communications Commission authorized ASCII transmissions for American amateurs in 1980, the experiment was ready to spread.
Amateurs adapted the commercial X.25 data protocol into AX.25, adding callsigns to the address fields so a packet knew where it came from and where it was going. The missing piece was affordable hardware. Tucson Amateur Packet Radio, known as TAPR, answered that need with its TNC-1 kit in 1983, followed by the widely copied TNC-2. A terminal node controller sat between computer and radio, and suddenly bulletin boards bloomed across the bands. Digipeaters relayed packets from hilltop to hilltop, and a message could hop across a region while its sender slept.
Packet found a second life in the 1990s through APRS, the Automatic Packet Reporting System. Developed by Bob Bruninga, WB4APR, and presented to the amateur community in 1992, APRS used packet bursts to share positions, weather readings, and short messages in real time. Its roots reached back to Bruninga's position-mapping programs of the early 1980s, and paired with affordable GPS receivers, it turned a data mode into a living map of the amateur community on the move.
The next leap needed no special controller at all, just a computer sound card. In December 1998, English amateur Peter Martinez, G3PLX, introduced PSK31. Running at about 31 baud, close to a good typing speed, and using his Varicode scheme to send common letters with fewer bits, PSK31 squeezed a keyboard conversation into a sliver of spectrum only about as wide as its name suggests. Dozens of those narrow signals could share space occupied by one voice channel, and operators with modest antennas and low power worked the world by watching thin traces crawl down a waterfall display.
Then came modes designed to hear the nearly unhearable. Joe Taylor, K1JT, released WSJT in 2001, and JT65 followed in 2003, becoming a favorite for moonbounce and weak-signal work. On June 29, 2017, Taylor announced FT8, designed with Steve Franke, K9AN. The name honors its creators and its eight-tone modulation. FT8 completes an exchange in tidy 15-second cycles, decodes signals well below the noise floor, and trades a little sensitivity for far greater speed than JT65. Within two years it was the most popular digital mode on the bands.
Each generation of digital modes asked the same question in a new way: how much meaning can ride on how little signal? Packet answered with structure, PSK31 with narrowness, and FT8 with mathematics and patience measured in seconds. The shack sounds different now, full of chirps, tones, and whispers no ear could copy, but the spirit is the same one that animated those Montreal experimenters in 1978. Give hams a new tool, and they will find out how far it can carry a hello.