Community Memory - History

History

The first terminal was a Teletype Model 33 ASR connected to the SDS 940 computer by telephone, using a 10 character per second acoustic coupled modem. It was located at the top of the stairs leading to Leopold's Records in Berkeley, right next to a busy conventional bulletin board. The Teletype machine was noisy, so it was encased in a cardboard box, with a transparent plastic top so what was being printed out could be seen, and with holes for one's hands while typing. This was the first time many individuals who were not studying a scientific subject had the opportunity to be able to use a computer.

Brief instructions were mounted above the modified keyboard showing how to send a message to the mainframe, how to attach keywords to it to make it searchable and how to search those keywords to find messages from others. To use a Community Memory terminal, the user would type the command ADD, followed by the text of the item, and then by any keywords under which he/she desired the item to be indexed. To search for an item, the user would type the command FIND followed by a logical structure of keywords connected with AND's, OR's and NOT's. By the side sat a CM assistant, attracting people's attention and encouraging them to add and find messages. In its approach, Community Memory adopted a creative method to funding the project. They provided users with coin-operated terminals which could be read without charge; however, in order to post an opinion, users were required to pay twenty-five cents or one dollar to start a new forum.

The record store and its bulletin board brought together drummers seeking fusion guitarists, bagel aficionados looking for sources, and the first poets of the medium, notably one who went by the nom de plume of Benway - the first net personality. Periodically directories of recently added items or of musician-related messages would be printed out and left there. In other terminal locations, users sought out complete strangers to assemble car pools, organize study groups, find chess partners, or even pass tips on good restaurants. According to Colstad and Lipkin, the rate of use of the system was fairly high and constant in relation to the environment of the terminals. About fifty searches and ten additions occurred each day at each location. Given the length of individual sessions with the system this was at least one-third the maximum capacity of a terminal.

Anonymity was possible with Community Memory because users were not required to share their names or register to use the system. All of the information on the system is community generated, which has two implications. Firstly, there was no central authority of any kind that establishes what information is available in the system. The second implication is that information is not imported from other sites.

When CRT based terminals became more cheaply available one was set up at the original Berkeley Whole Earth Access Store and another at the Mission Public Library in San Francisco. The character of the message base varied with location.

The Community Memory software was implemented as an extension of the ROGIRS keyword information retrieval system written by Bart Berger and John M. Cooney at Resource One, which in turn was derived from Robert Shapiro's MIRS (Meta Information Retrieval System). It was written in QSPL and ran on an SDS 940, an early timesharing system the size of eight refrigerators, originally used by Douglas Engelbart in The Mother of All Demos, which had been donated to Resource One for community use.

By 1974 it was apparent that Community Memory needed to move from its home on the XDS-940 (which was large, underpowered, and uneconomical) and be recast as a network of more modern minicomputers. It was shut down in January 1975; its staff left Resource One and began to explore funding for a new project which would develop the software for a replicable and networked version of Community Memory.