The Birth of Linux: How Linux Got Started

2565

It’s hard to believe, but this year marks the 20th anniversary of Linux. If you’re a new Linux user, you might wonder how it all got started. As part of the Linux Foundation’s celebrations for the 20th anniversary of Linux, Linux.com is going to be running a series of stories that looks at the history of Linux — starting with the history of how it all began.

In a way, Linux got its start before Linus ever sat down to start work on the kernel. Specifically, the foundation that made the kernel possible was kicked off in September of 1983 when Richard M. Stallman announced the GNU Project with the words “Free Unix!” to a Unix “wizards” newsgroup. The announcement may also have included the first recursive acronym (GNU’s Not Unix) and set off a bunch of confusion about what Stallman meant by free.

Why did this guy from MIT start GNU? Stallman said that he believes “the golden rule requires that if I like a program I must share it with other people who like it. I cannot in good conscience sign a nondisclosure agreement or a software license agreement.”

Where did that come from? Stallman had a bad experience trying to get a printer to work. Specifically, Stallman couldn’t get source code for a printer being used in the AI Lab at MIT — at a time when it was more common for code to be available for hardware and for the systems being used. Even though the concept of free software or even open source was not around at the time.

Stallman’s frustration was our gain, though, because he set out to create a free replacement for Unix. This included pretty much everything Unix shipped with at the time, from compilers to text editors to documentation. Oh, and a kernel — which the Free Software Foundation (FSF) and GNU Project didn’t have when Torvalds came into the picture in 1991.

This is why you’ll hear some folks refer to Linux as “GNU/Linux,” because GNU utilities and applications make up an enormous part of most distributions that ship with the Linux kernel.

In Comes the Kernel

While the GNU folks were working on the rest of GNU, Torvalds was attending the University of Helsinki and took an interest in computers. Torvalds started development of Linux in 1991, and released the first “official” version in October with the 0.02 release. At that point, Linux was barely usable — it could run a shell, it could run a compiler, but most of the software that we take for granted was not available.

It probably wouldn’t have gotten very far, except for Torvalds’ decision to release Linux to the world and ask for help with development. If you’re feeling nostalgic or have a bit of a historian in you, check out the old versions of Linux on kernel.org, including the release notes for v0.12 which were the first to embrace GNU Copyleft. You might also want to read through the installation instructions — you will come away with a great appreciation for how far Linux has come in the last 20 years.

Note that Linux was not universally accepted as a good idea at the time. Torvalds created Linux, in part, as a response to the “education only” licensing of MINIX. Andrew Tanenbaum, who created MINUX, had quite a lot of disparaging things to say about Linux at the time. According to Tanenbaum, it was a bad idea to go with a monolithic kernel design. It was tied to “a weird [line] like the Intel” 386 line. Tanenbaum basically dismissed the Intel x86 line and said “5 years from now everyone will be running free GNU on their 200 MIPS, 64M SPARCstation-5.”

In short, Tanenbaum proved to be spectacularly wrong about pretty much everything in the famous MINUX vs. Linux debate — licensing, hardware, and operating system design. Thankfully, Torvalds was not easily dissuaded and stuck to his guns. If you read through the exchange, you can also see that Torvalds style of diplomatic debate has not changed a great deal in the last 20 years either.

If you’re using Linux today, you have a lot of things to thank for its availability. On one hand, you have the hard work of Stallman and the FSF, Torvalds and other kernel hackers, and many other folks. The other side of the coin is that Linux wouldn’t be around today if it weren’t for a few hurdles that offended Stallman and Torvalds — namely, licensing issues that prevented them from doing what they wanted to do with the hardware they owned or used.

There’s much more to the story, of course. We’ll be running more stories on the history of Linux — from the dawn of the distributions, to the businesses and projects that have shaped Linux into what it is today. In the mean time, check out the “Story of Linux” video.