Interview with Sam Morgan, 9/12/89. Director of Computing Science Research at Bell Labs during development of UNIX.

 

Morgan was originally an applied mathematician in mathematical physics until he was put in charge of computing science research, not because he was ideal for the job, but because the previous director was transferred to Whippany, another comp center, and Morgan, though not in the market for a computer science job, was around and available. He knew the general rules of research management, "You get good people, you make sure they are aware of a wide class of problems and you let them do their own thing. But you give them feedback."

He became the director in 1967, after Bell Labs, jointly with MIT and General Electric, began research on MULTICS, and two years before the MULTICS project would be aborted in March, 1969. By the time he became director, doubts about MULTICS being the operating system of the future were already surfacing. He was unsure of whether there were doubts from the beginning about MULTICS or not since he was not the director when MULTICS research began, but it was clear by the time he became director that MULTICS research was going very slowly, wasting effort, and showing "no promise of turning into a user useful thing." However, when he became director, he was not told that there would be any changes in projects. Morganís understanding was that the group was to continue their ongoing work in language theory, numerical analysis, operating systems, switching theory, and MULTICS. MULTICS was only supposed to be a transient project, and his group was not committing to the software development business in any long term way, but MULTICS was a special case.

When the decision was made to end work on MULTICS, there were still people who continued working on it simply because they wanted to. To prevent MULTICS work from resurfacing, directors were discouraged from buying any machines big enough for MULTICSís shared environment work. Morgan stated that, "part of turning [MULTICS] off was not immediately buying hardware on which MULTICS could be continued." But on the side, Ken Thompson, a MTS under Morganís supervision, had been working on a new file system. Ken Thompson, Osanna, and McMahon, the three originally involved in the UNIX file system got together and asked Morgan to buy a PDP1120 for them to do their research on this new system. Morgan said no. Luckily, however, McMahon was working under another director, Max Matthews whom the three approached after Morgan rejected them. Matthews agreed to buy a machine for them because he was also interested in text processing and thought it a worth while expense. He ended up buying the 1145 for them.

By 1971, not long into the UNIX project, everyone around the lab realized the uniqueness of the UNIX project, and PDP11's were purchased so that eventually all the departments in AT&Tís Bell Labs, starting with the patent division, were using the UNIX system for their text and word processing. A few years after that, they began licensing UNIX to academia for a few hundred dollars. Morgan believes that "this piece of serendipity" was partly responsible for the world wide spread of UNIX. The top management of AT&T didnít foresee any possible profit from keeping UNIX to themselves. Bell Labs was not supposed to be in the computing business anyhow, so giving UNIX away and keeping in contact with the universities about UNIX was the only way it could have spread.

The public journal unveiling of UNIX was the 1978 special issue of BSTJ, of which Morgan was the guest editor. But that was only after the very informal spread of UNIX through academia. Before UNIX was even formally introduced to the public market, it had gained world wide attention through the involvement of the universities, especially UCBerkeley.

UNIX was certainly a very special development, one that Morgan believed could never have been forced by management. His management philosophy throughout his directorship was to "hire bright people, provide them with a stimulating environment," and provide what he called selective enthusiasm for their undertakings. UNIX was the work of artisans, not laborers. Management cannot give the specific task of writing a large scale operating system and expect to get one. If you order an operating system of a group of programmers, you get MULTICS. If you leave them to their creativity and support them when their work agrees with you, which is the selective enthusiasm idea, things like UNIX will grow out of their cultivation.