retroelectro
cabinets that offered improved cooling. The rebuild process was plagued with issues. The air conditioning would fail often, and both rotating drum memory units got sand in them, causing them to crash. It is Hawaii after all. Apparently, sand was such an issue that they turned that section of the building into a clean room with positive air pressure, air filtration and cleaning systems, etc. The BCC 500 system first became operational as a complete working system in February 1973. It was initially only helpful for system programming due to poor reliability. Still, by March 1973, the system had a schedule for four user hours per day, with the remaining time dedicated to hardware and system development activities. This guaranteed user schedule was modified and expanded several times. By May 1974, the system operated 24 hours a day, seven days a week, and was continuously available to users except for Saturdays, reserved for hardware and software maintenance. System reliability was better than 95% uptime during hours when Task personnel were present to assist with operations. Despite being often
mini-computers, the University purchased an IBM 360/65 to use with a time-sharing system, like many other Universities of the day. The central computer would be on one island, and students would have to ‘dial in’ to the system using antiquated inter- island phone lines that AT&T laid in the early twentieth century. The lines were unreliable, and the age and quality of the cables limited speeds. In 1968, members from the then-new ethernet consortium were meeting in Honolulu for a conference, and members of the University of Hawaii’s engineering department had the opportunity to pick the minds of some prominent figures in computing history. Soon after, in September, the University of Hawaii began researching radio communications for computer- to-computer networking. Early on, they were able to get funding
stopped funding BCC, deciding to stick with IBM equipment. With this they lost a contract for mainframe leasing with DPF&G along with additional funding. When the Berkeley Computer Corporation closed in January 1971, the members split up. Many of the top talents went on to Xerox PARC to invent the first modern computer, the Xerox Alto, while some others followed the BCC-500 on its fated Hawaiian trip. 1968: beginnings of the ALOHA System at the University of Hawaii In the 1960s, the University of Hawaii consisted of half a dozen campuses spread across many islands and several research institutes with researchers all around Hawaii, within a 200-mile radius of Honolulu. With the increasing availability of mainframes and
from the Office of Aerospace Research (SRMA) and they started developing the system. In 1970, while at a conference in Washington, DC, the Principal Investigator of the ALOHA system, Norman Abramson and Frank Kuo met with Larry Roberts and Bob Taylor at the Pentagon. During this meeting, they proposed an easily deployable, resilient, and repairable wireless system. This, of course, caught the attention of many people who saw the potential of such a system in the wars of the day. In June 1971, the central UHF station of the ALOHA System had been tested by the first remote terminal. By the end of 1971, four remote terminals had been connected to the University’s central computer through the ALOHA System.
‘Converting to this kind of communication could save a lot of money. For example, ARPA spends in excess of $1 million per year for line charges. With this much money, it could buy a transponder on a domestic satellite.’ - Norm Abramson
awarded contract number NAS2- 6700 for the ALOHA System. The contract consisted of two primary tasks. Task I is the one that gets all the attention, The ALOHA System network. This is where all the blood and glory were made, interconnecting network systems wirelessly with radio and eventually with satellites. A lot of groundbreaking things came out of Task I, including the first use of packet switching, addressing, wireless collision detection, and ended up creating the first commercial use of wireless network computing in 1975. This story is not about Task I, it is about Task II. Task II was titled ‘Research in Multi-Processor Computing Systems’ for the University of Hawaii to collect the prototype BCC-500 and move it to Hawaii to finish its development. The University of Hawaii built new facilities to house the computer, Holmes Hall, which was still under construction while the
machine was being moved. In February 1972, a team from Hawaii and previous members of BCC started to disassemble the machine to move it to Honolulu. Some sections could only be disassembled by cutting cables, but the final report submitted to ARPA notes that less than $100 of damage occurred throughout the move. In 2024 dollars, this is equal to $767 when accounting for inflation. The equipment was crated and flown to Hawaii on a 707 Jet Freighter. When the equipment arrived, it was all intact but far from operational. According to the report, they spent a lot of time deeply examining each piece for design flaws and inefficiencies. Several design weaknesses were identified and rectified. Often, this required adding wires to the existing printed circuit boards. However, in some instances, completely new boards were necessary. The revised items were then thoroughly tested and installed into newly designed
1971: ARPA grant and moving from Berkeley
It is unknown how or why ARPA gained possession of the BCC- 500, but this writer speculates that the University of California Berkeley took it from the Berkeley Computer Corporation as a return for their very recent one-million-dollar investment, then it was given or sold to ARPA. In November 1971, ARPA
Figure 5: April 1970 Proposal for The Aloha System (The Aloha System – Another Alternative For Computer Communications)
we get technical
36
37
Powered by FlippingBook