In an industry where systems routinely become obsolete in 5 years or less, the 30-year life span of MTS — the Michigan Terminal System — is almost unprecedented. This article, written primarily by Susan Topol of the University of Michigan Information Technology Division, looks back at how MTS has influenced computing at the University of Michigan and elsewhere.
Developed at the University of Michigan, and then supported and enhanced by a consortium of universities, MTS has made enormous contributions to computing and to computer users. It was a ground-breaking system, one that introduced concepts and technologies that were unprecedented in their time.
Perhaps most importantly, MTS provided a venue for sharing information and ideas in such a way that it helped students, faculty, and researchers to achieve their own goals. In this way, MTS played a key role in helping the University achieve its mission.
MTS programmers at the U-M Computing Center enjoyed a development environment that was second to none. Gary Pirkola, architect of the MTS filesharing system, stressed, "We had the opportunity to work on problems that weren't being solved anywhere else, and Dr. Bartels gave us the flexibility and freedom to just go ahead and do it."
He and other early MTS staff members credit Robert Bartels, the first director of the U-M Computing Center, with creating an ideal working environment. Said Pirkola, "His managerial and administrative style of letting his technical people just do their jobs was in large part responsible for not only attracting, but also retaining many of the people who made MTS the success that it was.
Mike Alexander, one of the MTS architects, said of Bartels, "He, more than any other single person, was responsible for creating the kind of environment that let MTS happen."
Pirkola added, "We were a small staff working in close proximity, and we had informal meetings only when necessary. Individuals had a lot of responsibility and were able to work autonomously on different parts of the system. Mike Alexander had the overall picture of how everything fit together and how it needed to work."
This development environment benefited users also in that developers had the ability to respond to users' requests directly. Bernard Galler, who was an associate director of the Computing Center during the early years of MTS, stated, "I remember a time when I made an off-hand remark to Don Boettner [an MTS architect] about how it might be nice to have a command that allowed you to append some additional characters to the end of a line. The next day there was an APPEND command in the Editor."
Over and over, when asked what MTS meant to those who used and developed it, the response was "a sense of community." Because MTS enabled electronic communications, and networks were linking together major institutions for the first time, an electronic community was formed. This community fostered friendships and professional collaborations of many kinds.
"MTS was not just a system, it was a community," said Elizabeth Barraclough, director emeritus of Computing Services at Newcastle University. "Those of us who saw the benefits that the community brought to its members will never think of it as just a system."
Whereas other systems made users feel like it was just them one-on-one with a computer, MTS was designed with many features that enabled sharing and collaboration. Users were able to collaborate with MTS developers, and vice versa. According to Bob Parnes, architect of the Confer system, "MTS was our system; it belonged to the University, not to a corporation."
Al Anderson, assistant research scientist at the Population Studies Center, concurred. "MTS supported a computing environment and computing community for many years with characteristics that I suspect we will not see or experience again," he said.
This electronic community also became a vehicle for the development of MTS itself. As other universities began to run MTS, they began to participate in its development also. Said Jeff Ogden, then an MTS programmer and associate director of the Computing Center, "MTS became a cooperative development effort of eight universities in three countries."
Alexander added, "It was one of the first 'networked' development efforts, and it was coordinated across international sites before the advent of the Internet." Indeed, developers had to rely on commercial networks, direct dial to other machines, and periodic visits to other MTS sites.
An annual workshop was held to share information about MTS development and support. A different MTS university hosted the event each year, and MTS developers and support personnel gathered to share their knowledge and experiences.
In fact, the sense of collaboration between these universities became so strong that the annual workshops continued even after MTS development work tapered off. Called simply the "community workshop," these gatherings continued until recently as people from the eight universities met to share their expertise and common experiences in providing computing services to their users, even though most of them no longer run MTS.
In the mid-1970s, the next great computing revolution on campus further expanded the U-M MTS community. Bob Parnes, then a graduate student studying experimental psychology, was attending a seminar in which Professor Merrill Flood was discussing the new concepts of e-mail and electronic conferencing and their use in decision making. Flood had a magnetic tape of a prototype system and approached Parnes about getting it to run on MTS. Parnes declined, but offered instead to attempt writing a similar program for MTS.
Because of a graduate teaching assistant strike, Parnes was temporarily relieved of his teaching duties and had some extra time to devote to his experimental system, which he called "Confer." MTS served as an excellent development environment for Confer, which was built on top of the MTS file structure and exploited its filesharing features. According to Parnes, "I don't think I could have written Confer anywhere but on MTS."
Confer played a tremendous role in enlarging the electronic community at the University and in removing the traditional geographic borders of the classroom and campus. Said Parnes, "Confer enabled a lot of people to talk together who wouldn't have otherwise."
The U-M Center for Research on Learning and Teaching was an early sponsor and proponent of Confer and saw great promise in it for expanding learning environments. Those working on the Merit Network were also excited by the potential for Confer, and they created the MNET:Caucus conference to help users get quick answers to their questions and take some of the load off their consulting staff. It turned out that the participants — both consultants and users — learned a lot from each other through the conference. MNET:Caucus, a statewide conference, later became the first campuswide computer conference.
Not only did Confer offer the opportunity for various forms of group discussion, it also served as the first e-mail system on campus. The MTS message system (or "$MESSAGE") was introduced in 1981. Written by Jim Sterken, $MESSAGE allowed MTS users to send and receive e-mail. Gavin Eadie and Jim Sterken then enhanced the message system to include remote mail — the ability to exchange e-mail with users on other systems. The early e-mail exchange was done over Mailnet. Mailnet was eventually replaced by BITNET and the Internet.
Although $MESSAGE eventually surpassed Confer as the e-mail facility of choice on campus, the computer conferencing portion of Confer continued to thrive. Parnes went on to form his own company — Advertel Communication Systems, Inc. — which markets and supports Confer.
MTS was a system for everyone, not just those who were traditionally thought to need computing. Alexander stated it more simply: "MTS let ordinary people do routine things easily."
Paul Whaley, former MTS developer at the University of British Columbia, said, "MTS provided a service to the UBC campus community at a time when other computing resources were phenomenally expensive and were only available to smaller groups."
When Alexander and the other MTS developers began working on MTS in 1966, one of their goals was to make the system easy to use for non-technical people, back during a time when systems were used primarily by mathematicians and scientists.
Alexander said, "I always thought it was a sign of success that some computer scientists hated MTS."
According to Barraclough, "MTS allowed those of us providing a computing service to protect our users from the horrors of raw computers and manufacturers' operating systems."
"MTS didn't require you to be very technical in order to use it; the user interface shielded us from the techno-wizardry behind the system," said Anderson. "This is one of the reasons it grew to have so many users."
MTS was a democratic system in that all users were treated equally and had equal access to services. Ogden added, "The system didn't really know or care if you were a student or a faculty member."
One of the developments that was instrumental in truly bringing computing to anyone on campus who wanted it was the addition of a second MTS system in 1985. Called UB ("U-Blue") to distinguish it from the first system (UM — retroactively dubbed "U-Maize"), this system provided accounts by request to anyone who wanted to use them.
Users no longer needed a specific reason, such as a research project or a class assignment, in order to get these "request accounts" to gain access to the computer. A request account could be used for any purpose. This allowed many users to experiment with computers for the first time. The UB system and request accounts allowed U-M to offer a campuswide e-mail service much earlier than many other universities.
Confer became a common means of communication as students organized their own conferences and CRLT staff members convinced instructors to set up course-related conferences.
According to Ogden, "It was the Computing Center's ability to create that second system very quickly, and at very low cost, that made it possible for the U-M to extend computing access to all students, faculty, and staff." With a campus network already in place, and printing and other resources that could be shared, the second system was easy to implement and operate (on a used, but still powerful, 470V/8 mainframe donated to the University by Amdahl).
One of the things that MTS provided to the U-M community was a virtual place where people could come together to share data, tools, information, and ideas. They could share their programs and data in disk files or on magnetic tapes. They could use MTS as a home base from which to send e-mail or from which to participate in group discussions via Confer.
MTS allowed sharing to occur between people in different departments and in different schools and colleges — and at other universities. It was one of the resources that made it possible for the University of Michigan to be more than a collection of individuals, departments, schools, and colleges.
Today that role is being taken over by the Internet and the World Wide Web, which allows sharing to occur on a much more widespread basis. But that kind of electronic sharing was already occurring at U-M more than 15 years ago.
"MTS and Merit/UMnet allowed many people to communicate electronically for the first time both one-to-one — using e-mail — or one-to-many — using e-mail, newsgroups, and conferencing," said Christine Wendt, then computer systems consultant for Merit. "So many people today are impressed by the Internet and the World Wide Web, but after more than 15 years of conferencing and e-mail using MTS, I have a feeling of 'been there, done that.'"
"MTS was a tool used by everyone at the university," said Alexander. "It got tied into the life and work of the university and became as fundamental as the phone system."