A brief history of open source - how free software fought with proprietary

    Today we’ll tell you how commercial solutions almost replaced the free code in due time. And which of the developers began to correct the situation. / photo Peter Hamer CC BY-SA

    Between Science and National Security

    The period of World War II was a time of technological breakthroughs for the United States. The collaboration of research institutes with military organizations has borne fruit in the fields of radio , cryptography and semiconductors .

    After the war, the research laid the foundation for inventions such as a transistor , and scientific ties turned into business contacts. The active development of computers began.

    The first commercial computer, the IBM 701, was unofficially called Defense Calculator . From 1952 to 1955, only 19 copies of this computer came off the assembly line. It was impossible to purchase them, but it was possible to rent for a month for a lot of money - about $ 12 thousand ($ 107 thousand by modern standards).

    The list of companies that have afforded such luxury, as expected, consists almost entirely of scientific and state organizations . Moreover, for that kind of money they received “bare” hardware without any software and operating system.

    To simplify the work, engineers passed programs to each other. A spirit of academic collaboration reigned in the industry. Research institutes, military customers, and big businesses created knowledge-sharing groups, and their labor products were in the public domain .

    Program Exchange Period

    The most famous groups whose members shared the source with each other were PACT, SHARE and DECUS. The first of these, PACT - Project For the Advancement Of Coding Techniques - consisted of representatives of military contractors like Lockheed and Douglas, as well as IBM.

    Together they developed a series of compilers of the same name for IBM 701 and 704, which used hashing methods. The leadership of the group emphasized the "value of cooperation" in the work on such projects and promised to maintain this spirit in the future.

    The successor to PACT was the SHARE group that appeared in 1955, which created the SOS operating system(Share Operating System). This primitive, by modern standards, solution for input and output of information has grown on the internal development of General Motors. It was SOS that laid the foundation for the first batch processing operating systems that performed several tasks prepared by one or different users. Similar systems dominated the computer market in the late 50s and early 60s.

    In 1961, another group appeared called DECUS (Digital Equipment Computer Users' Society). Its participants exchanged magnetic tape programs with each other. DECUS lasted quite a long time - in 1998 it still consisted of 50 thousand people.

    Thanks to scientific collaboration and software exchange , the Interlisp and UCI Lisp programming languages appeared , and this culture gave impetus to the development of the open Unix OS. But in the late 60s - early 70s there were several important events that suspended the development of open source. They made programs a product that can be monetized.

    Paid software and intellectual property

    If there was no software bundled with the IBM 701, then in subsequent mainframes it became more and more. All this software was included in the price of the system, and was supplied as a set. From the point of view of regulators, this was an attack on the monopoly. After a little pressure and the threat of litigation, IBM surrendered, and in 1969 began selling software separately. This led to the emergence of an IBM software market .

    Relatively high-level languages ​​- FORTRAN and COBOL - became widespread, and microcomputers came to the homes of enthusiasts. The line between users and programmers began to form. Complex software systems appeared, including those oriented for personal use. And their creators, quite logically, wanted to protect their work. This led to a public discussion: is it possible to patent software products? How do intellectual property laws apply to them?

    In 1974, a special commission was organized that secured the rights of programmers in American law. From that moment, software manufacturers stood on a par with the inventors. A little later, the U.S. Court of Appeal decided that copyrightapplies to computer programs. The developers were able to control who uses the results of their work. And in 1983, free software gave way to proprietary software. Then IBM stopped revealing the source code of its programs.

    So, the whole era of free software exchange has ended. But there were people who were not going to put up with this. People who make free software flourish now.

    Multics Lessons

    The software market in the 1960s resembled Cherkizovsky: there were many things, but they did not fit together well. For almost every new computer model, the OS was written from scratch. More and more programs were released with closed source, and the lack of standardization increased the cost of their development.
    The most vivid illustration of the situation in the IT industry at that time is possible with the example of the Multics operating system.

    Its creation was attended by MIT, Bell Labs and General Electrics, but they could not converge in a vision. The tasks assigned to the developers were ambitious, and the staff consisted of young, inexperienced programmers. The result of a five-year development was a system that was not destined to find success in the computer market. A detailed analysis of the miscalculations that contributed to the failure was subsequently devoted to a separate article .

    Shortly after the release of the Bell Labs system, they left the project, considering it a waste of time. But not all company employees shared this opinion. For Ken Thompson and Dennis Ritchie, this experience served as a lesson in OS architecture. They saw the potential inherent in the modular design of Multics - in it each task is performed by a separate utility - and decided to develop their own OS.

    In the summer of 1969, UNICS was born. Thompson wrote her first version on the PDP-7 home minicomputer during the holidays. After the project migrated to Bell Labs, where it changed its name to UNIX and received a small development team.

    / photo by Jason Scott CC BY

    UNIX and the academic market

    After the open source system was introduced to the public in 1973, it, unexpectedly for Bell Labs executives, began to rapidly capture the academic market.

    The reason for this was four factors:

    Availability. Antitrust laws specifically regulated Bell Labs, a subsidiary of telecommunications giants AT&T and Western Electric.

    In 1956, firms entered into a deal with the government, the terms of which limited their business outside the core business. In particular, the sale of products not directly related to telecommunications was to be carried out at a “nominal price”. This condition also extended to UNIX. More information about the legal aspect of the situation and its consequences can be found in the book “ Quarter Century UNIX ” by Peter Salus and this research work .

    The result of the regulation of the activities of Bell Labs was the fact that academic institutions were able to purchase the OS at a price slightly higher than the cost of a physical copy.

    Focus on minicomputers. Until 1975, there was nothe concept of "personal computer" as such. Public attention was mainly focused on mainframes. But the market for minicomputers - low-power alternatives to mainframes - has grown rapidly. Largely due to the limited project budget, UNIX was originally developed specifically for such machines.

    The third version of the operating system was sharpened for DEC PDP-11, more than 170 thousand copies of which came off the assembly line in the 70s. For comparison, sales of the popular IBM 1401 mainframe did not exceed 20 thousand units for the entire model. As a result, the OS has gained a large number of users in the amateur and educational segments.

    Simplicity.Large monolithic programs are hard to write and even harder to modify. UNIX consisted of small modules, each of which was responsible for a separate function and easily changed. Moreover, starting with version 4, the operating system was almost entirely written in the high-level language C (and not in low-level assembler). As a result, it was easier for university students to figure out how the OS works.

    Portability. To run UNIX on a different architecture, it was enough to port the C compiler. For this reason, working versions of UNIX for new processors appeared quickly. This became especially important in the era of microcomputers, which began in the late 70s. UNIX ports for home machines based on the Intel 8086 and Motorola 68000 were very popular.

    UNIX users around the world, mostly academics, actively exchanged software for the system. Communities of enthusiasts existed in Britain, Australia and Japan. In 1975, more than forty American educational institutions used this operating system.

    The birth of BSD

    One of the first academic institutions that began to actively use UNIX was a branch of the University of California at Berkeley. Professor Robert Fabry received a copy of the system personally from Ken Thompson in 1973. Employees of the organization quickly fell in love with UNIX and took up their own developments to improve the OS.

    University additions to the system were distributed under the name Berkley Software Distribution (BSD). The first version of BSD included a modified Pascal language compiler, the code of which was obtained directly from Thompson, and the text editor Ex - now known as Vi . Many of the BSD innovations were implemented in the “official” versions of UNIX. But AT&T were unhappy that control over the fate of the system was left out of the hands of the company.

    The cooperation of the University of Berkeley with the military agency DARPA, which brought advanced networking technologies to the BSD, has aggravated the situation. As is typical of an academic project, BSD commercial licenses were available at a low market price of one thousand dollars - and were sold in thousands. For AT&T, it was a lost business.

    While Berkeley employees used the UNIX code primarily for scientific purposes, other companies viewed the popularity of the system as a chance to make money. In 1978, the first commercial C compiler appeared. AT&T decided to put an end to this and at the end of the 70s restricted the distribution of the OS source code. The company's lawyers resumed the struggle for the right to fully monetize the product and protect it from third-party modifications.

    In 1983, Bell Labs formally separated from AT&T and Western Electric by court order. UNIX has become a completely commercial product, and its cost has increased to tens of thousands of dollars - often more expensive than the iron on which the operating system was running.

    / photo Sam Williams CC BY-SA / Richard Stallman

    Richard Stallman and the Birth of GNU

    In the 1970s, Richard Stallman worked as a programmer in the MIT artificial intelligence systems lab. He was known as one of the authors of EMACS, a text editor for minicomputers of the PDP family. Like many other software of the time, the source code of EMACS was freely distributed, and the final product was the sum of the efforts of users.

    All changes in the culture of academic development occurred in front of Richard in front of his eyes. The university has ceased to be a place for an open exchange of ideas and software tools. Even the creator of the EMACS version for UNIX, which initially allowed the free distribution of its program, sold it to a commercial distributor in 1983. Stallman considered such an act a sabotage.

    He knew that something needed to be done, but to rectify this situation was not easy. Free distribution of programs did not have solid legal ground. Most of these products were simply considered “public domain”, and nothing prevented the commercialization of their modified versions.

    Moreover, the development for proprietary systems limited the distribution of the product, requiring the user to purchase commercial and, from the point of view of Stallman, immoral software.

    Stallman faced two tasks - to create a free operating system and a legal base for its distribution. Therefore, in 1983, the GNU project (GNU's Not Unix) was born, designed to become an open and improved copy of the already proprietary platform at that time.

    The decision to use UNIX as a basis was not accidental - Richard was attracted by the portability of the operating system. We will describe how this project developed and what it led to in one of the following materials.

    Posts from the First Corporate IaaS Blog:

    Fresh materials from our Telegram channel:

    Also popular now: