Fernando J. Corbato: the man who divided the time

Modern cloud technologies allow any user to get unlimited access to huge computing power from anywhere in the world. However, the core principle of this concept was developed in the late 1950s, decades before the emergence of concepts such as “cybersecurity” and “cloud”. Its pioneer was not a large technology company like IBM, but an ambitious young scientist with Spanish roots. Now it is customary to call him “the father of the computer password”. But Corbato’s merits do not end there.

Computation queue

Back in the 1950s, the young scientist Fernando “Corby” Corbato earned his Ph.D. in physics from the Massachusetts Institute of Technology. As a doctor of sciences, he was admitted to computing on one of the most powerful and promising computers of that time – Whirlwind… After some time, having figured out the principle of the computer, Corbato was delighted: this device could do much more interesting things than the “banal” solution of physical problems one after another. But the physicist did not have time for a detailed study of the capabilities of the computer. The fact is that the computer was shared by several MIT faculties at once. No one wanted to give up legal time to the young doctor’s dubious projects. At best, Corbato could be content with thirty minutes of computer work a day — and in the early hours of the morning. I had to reluctantly get up early and experiment for a few minutes, until colleagues with “serious” faces and calculations came.

Whirlwind Computer
Whirlwind Computer

Probably, this story left a serious imprint on the entire future career of Fernando Corbato. In the very early 1960s, he developed the first time-sharing operating system. It allowed several people to use the computer at the same time. In other words, the problem of queuing for machine time has disappeared. Users could work with the machine at any convenient time. The principles laid down by Corbato are still reflected in the operating system of modern electronic devices.

Corbato later recalled:

“It was a real fight! They saw me as nothing more than a skinny professor with a bow tie. Nothing more. At that time, computers were not distributed right and left to all Tom, Dicks and, sorry, Fernando. And the architects did not even want to hear about any breakthrough ideas “from the street”. Each had their own idea of ​​what a computer is. And a computer is certainly not a thing that can be shared with friends! “

Early years and early career

Fernando Jose Corbato was born on July 1, 1926 in Oakland, California. His father, Ermenegildo Corbato, was a migrant from Spain and worked as a teacher in a Roman Catholic school in China as a young man. He later moved to the United States to attend graduate school at the University of California, Berkeley. There he met his future wife Charlotte Jensen. With the birth of his son Ermenegildo Corbato, he got a job teaching Spanish literature at the University of California, Los Angeles.

Young Fernando did not follow in the footsteps of his parents: he was interested in humanitarian subjects insofar as he had a real passion for mathematics and physics. Fernando’s abilities showed up in high school. And with the outbreak of World War II, at the age of 17, he entered service in the Navy and received a specialized education as an electronics technician. This is probably what laid the foundation for Fernando’s future work. By his own admission, serving in the army taught him to solve trivial problems in a non-trivial way. This skill was critical to Fernando when he got involved in designing computer operating systems.

After the war, in 1950, he entered graduate school at MIT. Brilliant successes in the field of computer science, coupled with a host of extraordinary ideas and solutions, drew the attention of senior colleagues and professors to him. Fernando was appointed assistant management of the computer center on campus. As more students and teachers discovered computer computing, he saw firsthand the difficulties of efficiently exploiting computer resources.

In those years, working with a computer was structured as follows: at the beginning, programs were transferred to punched cards. The finished stacks of punched cards were handed over to the operators, and they ran the programs through the computer in batches for several days. It’s not hard to imagine how much trouble a single error in a program on a punch card could cause. Due to the smallest flaw or inaccuracy (provided that it could be detected immediately after the end of the calculations), it was necessary to re-fill the program and send it for execution.

“The equipment was very moody and complex,” Fernando recalled. “On average, computers crashed every 20 minutes.”

Corbato’s work with the Whirlwind project pushed him to a breakthrough idea: why not give multiple users the ability to simultaneously use a computer? There will be enough power for everyone, but due to the technical features of the first computers, it was simply impossible to implement the independent work of several programmers.

First time-sharing OS

In the early 1960s, Corbato, who had already received his doctorate, with the support of colleagues, studied ways to organize the sharing of computer resources.

In 1961, Corby, along with his colleagues (Bob Daly and Marjorie Mervyn-Daggett) at MIT demonstrated the first version of the time-sharing operating system – CTSS. The system was implemented on the basis of the IBM 709. Users were asked to connect to the machine using modified Flexowriter terminals. Four tape drives were used to save programs that were suspended.

In 1962-1963. development of the CTSS continued. The tube IBM 709 was replaced by a new machine – the IBM 7090. The computer was specially modified by IBM to work in time-sharing mode. The connection to it was still implemented through the terminals installed in the offices of MIT and even at some employees of the institute at home.

CTSS presentation
CTSS presentation

By the way, the MIT Press CTSS manual is still available today. here

The CTSS was the spark that convinced MIT to open Project MAC, the forerunner of the Computer Science Laboratory (LCS), which would later become the Computer Science and Artificial Intelligence Laboratory (CSAIL). At the first major Project MAC event, the Summer Study, computer scientists from various research institutions, government agencies and industry were able to get to know CTSS in person. This hands-on experience not only convinced the participants of the effectiveness of timesharing, but also helped to form a community of early users around CTSS.

It became clear that CTSS had gone beyond the pilot project. In the fall of 1963, Project MAC acquired its own IBM 7094 with ARPA funding, and CTSS made the science machine resources available to all MIT research projects.

In addition to making computing more efficient, CTSS has played an important role in shaping the concept of data privacy. Since different users wanted to restrict the access of “car neighbors” to their files, Corbato came up with the idea of ​​creating individual accounts with personal passwords.

In 1963-1964. Dr. Corbato and his team began work on a much more ambitious project. The new operating system was named Multiplexed Information and Computing Service, or Multics. Such a major project required the appropriate hardware support, and Corbato began to knock the doorsteps of large technology companies. To his surprise, he found that IBM, the undisputed leader in the computer industry, was not interested in timesharing at all. I had to turn to General Electric Co. Unlike IBM, GE caters to software developers in every possible way. In this way, she was trying to gain a competitive advantage over IBM, which, according to Corbato, “made every effort to force customers to buy another new computer instead of somehow upgrading the old one.”

OS Multics

The Multics project is the first operating system to implement a flat storage model. It had a clear separation of the concept of files (segments) and memory of computational processes. Moreover, each segment of which the memory consisted had its own address space.

Multics logo
Multics logo

With some reservations, it can be recognized that Multics was the first fully centralized file system with support for a tree-like hierarchical structure.

In addition, it was in Multics that the concept of dynamic linking was first implemented between the executable program and the code libraries stored in the computer’s memory.

We recommend everyone interested to watch a short A video explaining how virtual memory works in Multics:

The project lasted until the mid-1970s. The OS did not gain commercial success: the relatively small market, IBM’s ironclad policy towards its own mainframes, and a host of other factors prevented Multics from gaining popularity.

However, years later, Dr. Corbato said more than once that the concept of Multics itself had a positive impact on the development of computers in general. It was Multics that inspired the Bell Labs researchers to create UNIX and became the training ground for many famous programmers – Dennis Ritchie, Ken Thompson, Dan Bricklin and Bob Frankston.

“It was an excellent school, through which an incredible number of professional programmers from all over the world went through,” said Fernando Corbato in a 2006 interview with the Museum of Computer History.

Corbato on Multics:

In 1990, Fernando Jose Corbato received the Computing Machinery Association’s Turing Award for his work on time-sharing computer systems. Died on July 12, 2019 at the 94th year of life.

“It’s no exaggeration to say that Corby’s time-sharing work has fundamentally affected the way we know computers today,” says CSAIL Director Daniela Rus. “The digital revolution is rooted in the projects he led at MIT nearly 60 years ago.”

“Corby was a key figure in the research that made computers available to many people and purposes at once,” recalled his longtime colleague Tom Van Vleck. “He saw that these concepts not only make computing more efficient, but that they are fundamentally changing the way people use information.”

Corbato’s vision of making high-performance machines available to everyone was anticipated back in the 1960s. Who knows how technology would have evolved today, had it not been for his work and enthusiasm more than half a century ago?

Similar Posts

Leave a Reply