In the six-and-a-half decades since SHARE began, mainframe computing has evolved and grown into the stable, secure platform that many enterprises rely on today. From the first meetings in the basement of the RAND Corp. headquarters building to its twice-yearly events today, SHARE has become the go-to place for mainframers and developers alike to share their knowledge and build their networks. All of this helps support the growth and improvement of the enterprise platform. We caught up with James Babcock, former SHARE president (1964-1965) and program committee member and former owner of Allen-Babcock Computing, to talk about SHARE's early days and his advice for the next generation of mainframers.
Early Computers Required Hands-on Learning
Babcock, who studied math and engineering at the University of Chicago and University of Michigan, started his career at RAND following his graduation in 1954 after an on-campus interview. "The idea of moving to Santa Monica, California, was particularly appealing," he said. After joining RAND, he worked as a "shavetail" computer programmer for a decade. RAND had government contracts, particularly with the U.S. Air Force, Atomic Energy, and Department of Defense agencies. One of the first machines he worked on was an IBM 704, which was the first mass-produced computer with floating-point arithmetic hardware. It used vacuum-tube logic circuitry and 36-bit binary words. "Because I didn't learn how to program at Michigan, I had to work out how to program the 704 on my own," explained Babcock. "Back then, there were no computer science courses."
During his time at RAND in Santa Monica, Babcock's team ran programs on the 704 to watch the winds and determine where the radiation from nuclear bomb tests in Nevada would go. The army would set off the nuclear bomb in the air, and RAND's computer department would run programs on the 704 to determine where the winds would carry the radiation. "Those machines would run all night long," he said. "I remember one night, we had to stop the program from running. The nuclear testing team in Nevada stopped their countdown to ignition because a sheep farmer had wandered into the testing zone," Babcock recalled. "I'll never forget that. That was a tense time."
The 704 machines, he described, ran on symbolic assembly language, which made it hard to quickly stop them from running programs. Unlike RAND's 704, IBM's 702 mainframe, he explained, only ran COBOL. Babcock said he never worked on that machine.
IBM's Introduction of the 360 Changed Everything
Babcock worked on a variety of mainframe machines, including most of the 7000 series, like the 7094. According to him, though, on April 7, 1964, everything in the mainframe world changed when IBM unveiled its 360, which aimed to solve the problem of mutually incompatible mainframe computer systems by allowing software written for one model to be run on any other model. Customers were now able to add or remove computing capacity without losing their software investments. Babcock said that moving from a machine that only programmed in FORTRAN over to the 360 with hexadecimal programming forced him to learn a whole new way of programming.
"Up until 1964, the computing industry was known as IBM and the seven dwarves," Babcock recollected. The dwarves included Burroughs, Sperry Rand, Control Data, Honeywell, General Electric, RCA, and NCR. Once IBM introduced the 360 and decided to be data dependent, it captured a majority of the computing market share. "IBM computers are known for their data management systems, and that's why they became popular with government agencies, financial companies, and insurers," he added.
"Birds of a Feather" Flock to SHARE
After about four years working at RAND, Babcock started to work on operating systems, which is about the same time he attended his first SHARE event in San Francisco in 1958. He says that it was during that time that he and his colleagues coined the term, "birds of a feather," for those who flocked together to discuss industry issues and systems. The SHARE “birds of a feather” sessions ranged in topic, and, in those days, the agendas were posted on simple bulletin boards with times and room numbers. "A lot of the sessions focused on operating systems and software," he said, "because at that time, SHARE members didn't have much sway in the hardware's configuration."
At the time of the IBM 360, some SHARE and IBM members started to work collaboratively on the SHARE operating system (SOS), which included FORTRAN, COBOL, and PL/1 languages for programming any IBM 360. "This collaborative work really demonstrated the importance of users and SHARE in the mainframe community," he said. (Note: This pre-dated the current MVS system.)
Babcock believes he last attended a SHARE event in Atlanta in the 1990s, but he still maintains many of the professional connections he's made over the years. "The connections you make at SHARE are lifelong," he pointed out. "When you're at SHARE, members often grab a beverage and talk business, but the evening SCIDS events were completely different." He recalled, "During my time, we would collect a small payment—usually $10—from those interested in going to SCIDS, and that money would be used to buy beverages, liquor, beer, whatever, (at local Liquor Stores) for everyone." He added, "SCIDS met four times during SHARE week, and more information was shared during those meetups than at any other time." Babcock also pointed out that, in the time between session events, a great deal of information was shared between colleagues. "When you join SHARE, you're joining a community, and you stick with them," he said.
When you join SHARE, you're joining a community, and you stick with them.
Babcock also shared a non-SHARE event in the 1970s in New York City where he was tasked with putting together an international panel on computing. Like at SHARE, computing experts from around the world wanted to share their expertise with others. One computer programmer in Russia was asked (by Babcock) to write a paper for the event, but when he was invited to New York for the event, Babcock got a call from the NSA and a visit from an army captain. "I was warned to be careful what I said in front of the Russian programmer, and to not talk at all about my work with nuclear testing," he said. "The captain added that if I got a phone call about being interviewed to be very careful. Sure enough, I did get a call from a photography studio, and I had to turn down their offer." Babcock chuckled, "I swear there's a file on me in some basement of a building somewhere."
Start Learning Early
For students today, Babcock recommended they start learning to code in elementary school because when they get to junior high or middle school, they'll need to know how to program. By the time today’s students get to high school, they should look into computer science courses if they plan to pursue computing as a career. He also recommended taking courses in computer engineering or anything to do with engineering or building computers. In college, students should be looking at computer companies, like Apple and IBM, for summer internships or jobs in order to learn additional skills not taught in the classroom.
Computer professionals are needed in every line of business, said Babcock, from insurance to computing. Students would be wise to start their journeys as early as they can. Computers have been at the forefront of history for the last 60 years, and as they evolve, so too have businesses. As business relies more on data, computers will continue to process more information faster and more efficiently