Linux — Linus Torvalds' 21-Year-Old Announcement

On 25 August 1991, Linus Torvalds, a student in the computer-science department at the University of Helsinki, posted to the Usenet newsgroup comp.os.minix:

Hello everybody out there using minix -

I'm doing a (free) operating system (just a hobby, won't be big and professional like gnu) for 386(486) AT clones. This has been brewing since april, and is starting to get ready.

Subject line: "What would you like to see most in minix?" Summary line: "small poll for my new operating system". Dated 25 Aug 91 20:57:08 GMT. "Just a hobby, won't be big and professional" — he really wrote that. The postscript adds: "Yes - it's free of any minix code".

Thirty-five years later, the same author surely knows it became larger, and considerably more professional, than he had any intention of making it.

The Constraint Called Minix

Minix was Andrew Tanenbaum's teaching operating system, bundled with his 1987 textbook Operating Systems: Design and Implementation. Its source could be read and modified. It was not free, and the reason was ownership: the copyright belonged to the publisher. When the licence finally changed, in April 2000, Tanenbaum announced it himself on the same newsgroup — "I finally got permission from Prentice Hall to change the MINIX license to the BSD license. The lawyers sort of sat on this for two years." The new licence header still read Copyright (c) 1987,1997, Prentice Hall.

For nine years before that, a student could study Minix and patch it but could not simply hand a friend an improved system. The practical shape of that constraint appears in Torvalds's own 5 October 1991 announcement of Linux 0.02:

ALERT! WARNING! NOTE! These sources still need minix-386 to be compiled (and gcc-1.40, possibly 1.37.1, haven't tested), and you need minix to set it up if you want to run it, so it is not yet a standalone system for those of you without minix.

A contemporary reader put the same point in a single clause in the August thread: "Mach needs a MMU and Minix is not free."

The technical complaint was separate, and specific. Torvalds wanted his own Intel 80386 to be used as a 386. In the follow-up post of 26 August 1991 he is blunt about how far he had gone:

Yes, it needs a MMU (sorry everybody), and it specifically needs a 386/486 MMU (see later).

Simply, I'd say that porting is impossible. It's mostly in C, but most people wouldn't call what I write C. It uses every conceivable feature of the 386 I could find, as it was also a project to teach me about the 386.

Segmentation was the part that bound it hardest to the chip: "every task has a 64Mb segment for code & data - max 64 tasks in 4Gb. Anybody who needs more than 64Mb/task - tough cookies." The ambition, at that moment, was modest and explicitly stated: "first I'll make it a simple minix-lookalike, keyword SIMPLE." So was the licensing intent, a full five months before the GPL: "It will be free though (probably under gnu-license or similar)."

The Early Development Model

Torvalds had begun with a terminal emulator that exercised 386 task switching. A file system came next, then I/O. By 3 July 1991, when he posted to the same group asking where to find a machine-readable copy of the POSIX standard, "some of the device drivers were ready, and the harddisk actually worked."

Version 0.01 went up on nic.funet.fi in mid-September 1991, and was not announced. By Torvalds's own account a year later: "0.01 sources weren't actually runnable: they were just a token gesture to arl who had probably started to despair about ever getting anything" — arl being the correspondent whose exchange with Torvalds that July had produced the Linux subdirectory on nic.funet.fi in the first place. "Note the lack of announcement for 0.01: I wasn't too proud of it, so I think I only sent a note to everybody who had shown interest."

The public announcement came with 0.02, on 5 October 1991, and it was addressed to a specific kind of reader:

Do you pine for the nice days of minix-1.1, when men were men and wrote their own device drivers? … This is a program for hackers by a hacker.

Then the cadence that would define the project for three decades: "max 2-3 weeks was the time between releases even back then." 0.03 followed within weeks; the next was numbered 0.10 because things had started working; 0.11 was current by 19 December 1991, with a 0.11+VM build circulated to a handful of testers just before Christmas because people were failing to compile the kernel in 2MB. Torvalds's own .plan file at the time read: "Free UNIX for the 386 - coming 4QR 91 or 1QR 92." His own estimate of the ceiling: "It will never be the kind of professional OS that Hurd will be (in the next century or so :), but it's a nice learning tool."

Choosing the GPL

The change was announced in the release notes for version 0.12, which came out in January 1992. The licence until then carried a "you may not distribute it for money" clause — or, as Torvalds glossed it later, "The early copyright was in fact much more restrictive than the GNU copyleft: I didn't allow any money at all to change hands due to linux."

The Linux copyright will change: I've had a couple of requests to make it compatible with the GNU copyleft, removing the "you may not distribute it for money" condition. I agree. (…) Otherwise The GNU copyleft takes effect as of the first of February.

So the GPL took effect on 1 February 1992. That choice was decisive. Copyleft requires that anyone redistributing the software, modified or not, do so under the same licence — so Linux could not be enclosed by a commercial vendor and turned into a proprietary product. Patches began arriving from developers around the world.

The Objection, Stated Properly

The most serious criticism of Linux arrived within weeks of the GPL, from the author of Minix. On 29 January 1992, Tanenbaum opened a thread on comp.os.minix under the subject "LINUX is obsolete".

His argument, in the opening post, was that Linux was a monolithic kernel in a decade that had moved on: "among the people who actually design operating systems, the debate is essentially over. Microkernels have won." Of the design itself: "LINUX is a monolithic style system. This is a giant step back into the 1970s. That is like taking an existing, working C program and rewriting it in BASIC."

The line the argument is now remembered by came later in the same exchange, after Torvalds had replied: "I still maintain the point that designing a monolithic kernel in 1991 is a fundamental error. Be thankful you are not my student. You would not get a high grade for such a design :-)"

He was not wrong about the academic consensus; he was wrong about which property would decide the outcome. The debate is worth keeping in the record precisely because the losing side had the textbook on its side.

Big, and Professional

• 1994: Version 1.0.
• 1996: Tux the penguin became the mascot.
• 1998: Major vendors — IBM, Compaq, Oracle — announced Linux support.
• 2003: SCO sued, alleging IP infringement; SCO ultimately lost.
• November 2007: Google's Open Handset Alliance announced Android, whose kernel is Linux.

The scale is measurable. Version 0.01 was 10,239 lines across 88 files. Running cloc over the Linux 7.2 git tree on 28 June 2026, Phoronix counted 108,158 files and 43,898,743 lines in total — 33,653,681 of them lines of actual code, the rest blanks and comments — up from 42,924,382 for Linux 7.1 one release earlier. That is roughly 1,200 times the files and 4,300 times the lines of the first tarball.

The release machinery is now equally unlike a hobby. On 9 August 2026, kernel.org listed mainline at 7.2-rc6, stable at 7.1.7, and six longterm branches still receiving fixes, the oldest of them 5.10.

W3Techs, surveying on 8 August 2026, puts Linux at 61.7% of all websites whose operating system is known, and the Unix family at 91.9%. Every system on the TOP500 supercomputer list has run Linux since November 2017. Android devices number in the billions. And in February 2021 Tim Canham, Mars Helicopter Operations Lead at JPL, told IEEE Spectrum: "This the first time we'll be flying Linux on Mars. We're actually running on a Linux operating system."

The 21-year-old who wrote "just a hobby" is now a fellow of the Linux Foundation, and still personally signs off on each release of the Linux kernel.