After the failure of the Multics project, Bell Labs programmers did not have a computer to use, and finally found the idle machine in the acoustics department in the corridor.

The Translation Bureau is a compilation team that focuses on technology, business, workplace, life and other fields, focusing on foreign new technologies, new ideas, and new trends.

Editor’s note: You may not know that the competing operating systems of Android and iOS are actually derived from the same system – Unix, and until now, they still provide support for these two operating systems.

This year is the 50th anniversary of the birth of Unix. At the time, this operating system, which almost drove all smartphones, was actually a product of a project failure. The programmers of the developer’s system did not receive support and could only use computers that others had left unused.

From a failed project to everywhere, in the early days of Unix, what happened? Ars Technica recently published a long article on the development of this project. The original title of the article is “Unix at 50: How the OS that powered smartphones started from failure” by RICHARD JENSEN. Compile, I hope to bring you inspiration.

Note: This article is longer and is published in three parts. This is the second part.

Unix 50 years: Android and iOS operating system, from the same failed project (2)

Ken Thompson (sitting) and Dennis Ritchie (standing) before the PDP-11 (Translator’s Note: PDP-11 is the most famous computer) One is a product manufactured by Digital Equipment Corporation (DEC) since the early 1960s.

Rickey joked that his hair was “more dense” than when he published his own photos in magazines such as Scientific American in March 1999.

Unix 50 years: Android and iOS operating systems, from the same failed project (1)

Comfortable environment

For Bell Labs programmers, MultIcs are not their full-time jobs, and the GE mainframe they use can also be used for other projects.

Thompson’s operating system is one of them. He had been doing this program that winter, and came up with the idea of ​​how to make each terminal connect to the machine and print “HELLO” at the same time.

But before he could further develop, the Multics project was cancelled and the mainframe was packaged and moved out of his computer science department.

Baker and Davis did not give McIlroy’s team a new job when they canceled the Multics project, which made them a bit uneasy. They worry that with the demise of Multics, their position at Bell Labs will not last long.

However, this thriving development team happens to be in an environment that is thriving for Unix. Unlike other places, Bell Labs is funded by a portion of the monthly income of almost all telephone lines in the United States.

Placing a small group of programmers on top of Murray Hill Complex will not let the company go bankrupt. Thompson’s department also has an ideal manager to support them in their curiosity.

Sam Morgan, who manages the computer science research department (composed of McIlroy’s programmers and a group of mathematicians), did not intend to rely on McIlroy’s team because they suddenly had nothing special to do.

Morgan is a professionally trained applied mathematician who was promoted to department head in 1967.

“I don’t think management is his favorite activity,” Brian Kernighan told Ars, his office was opposite Thompson. “But he worked very hard to get it done. He is very kind and tries to be kind to everyone.”

“The management principle here is to hire smart people and then provide them with an environment,” Morgan recalled in his Unix oral history project. “You give them general instructions, tell them what they need, give them a lot of freedom.”

So, Morgan did not provide a specific direction, but instead used what he called “selective enthusiasm” to encourage a particular research project, he pointed out, “If you mistakenly block or not respond, some later proved to be good. Something, if it’s really a good idea, it will come back.”

“He lets people do their own things and never tells anyone what they should do,” recalls Konyhan. At the time, Bell Labs also emphasized interdisciplinary cooperation. “Everyone has been opening the door, so if you have any questions, there is an expert nearby, you can walk in and ask for help,” recalls Cohenhan.

Nevertheless, for Thompson and his colleagues, there is still a small problem – no one has a computer. Although laboratory managersThe computer itself has no problems, but McIlroy’s programmers can’t convince their boss to buy a computer for them.

Because of the fiasco of Multics, they couldn’t convince anyone to give them a new computer so they could continue the research and development of the operating system.

But from the perspective of lab managers, Thompson and other team members seem to want to continue working on the Multics project.

Share computing wealth

The Computer Science Department shares a floor with the Acoustics and Behavior Research Department. This department is larger, led by Max Mathews, a pioneer in the field of electronic music, sound synthesis and voice recognition.

The Acoustics Division, which is clearly used in AT&T’s core business, is clearly more versatile than the computer science department.

In addition, there is a situation that seems to be intentionally irritating Rich and Thompson – the acoustics department is not lacking in computers. Rich and Thompson had some disdain for the bureaucracy of the company. In fact, the number of acoustic computers far exceeds their needs.

And, as long as the sound department’s programs become too complex to run efficiently on their computers, they only need to ask the lab manager for a new computer.

Unix 50 years: Android and iOS operating systems, from the same failed project (2)

Max Mathews plays an electronic violin in Bell Labs’ analog electronics lab.

Although they have some degree of embarrassment about the large amount of money available next door, there is some cooperation between the acoustics and computer science departments in the 1960s and 1970s. In fact, many of Bell Labs’ innovations in computer technology actually come from the Department of Acoustics.

For example, in the early 1960s, acoustic researcher Bill Ninke demonstrated a basic graphical user interface with the DEC PDP-7 microcomputer.

Although acoustics still keeps that computer, they have left it idle, didn’t use it, and placed it in a remote place on the sixth floor.

So, Thompson, after tirelessly smashing the corners and crevices of the lab, finally discovered the PDP-7. At this time, Davis and Baker just announced the cancellation of the Multics project shortly after.

In the teamWith the help of others, Thompson packed the various parts of the PDP-7, put it in a closet assigned to the audio department, and started running.

The size of the PDP-7 is equivalent to a refrigerator and is not a terminal. Moreover, they persuaded the acoustics department to provide space for the computer and let the department pay for the daily repairs of the machine from the budget.

McRoy’s programmers suddenly have a computer. Therefore, in the summer of 1969, Thompson, Richie and Canadi developed a basic file manager on the PDP-7.

This is not a simple task. Batch computing (running programs one after the other) rarely allows computers to store information permanently, and many large hosts do not have any permanent storage (whether disk or hard disk).

However, the time-sharing environment that these programmers like, you must add additional memory. When multiple users are connected to the same computer at the same time, the file manager must be written well enough to prevent one user’s files from being written to another user’s file.

Also, when reading a file, you must send the output of the file to the user who opened the file.

This is a challenge that McIlroy’s team is willing to accept.

They have seen the future of computers and want to explore it. They know that Multics is a dead end, but they have discovered the possibilities of shared development, shared access, and real-time computing.

20 years later, Rich said at Princeton University: “We want to retain more than just a good environment for programming, but a system that can work with teams.”

“In the late 1960s, I was still using the IBM 7094 from Princeton and the GE 635 from Bell Labs for batch calculations,” Brian Coneyhan told Ars.

“But I used CTSS at the Massachusetts Institute of Technology in the summer of 1966, which reveals how good interactive computing is.”

Once Unix started running, Conyhan moved from researching abstract topics to writing programs, and eventually collaborated with Rich to study C. When it came out, it immediately became the gold standard for programming manuals.

Translator: Scale.