The early days of Linux – join us as we celebrate 100 issues of Linux User & 20 years of Linux

He installed Linux on the Swansea Computer Society machine, and “started trying to make the networking code work properly, because it was buggy at the time,” and the university had a “nice, very noisy multi-protocol network which was just perfect for crashing your computer. I started fixing these things, which is how I ended up doing the networking code, building on the base networking code that other people had done.

“I was pretty much downloading standards and learning it as I read. The standards in TCP are small and clear, but there is an unwritten law behind it which does take time to learn, like queue theory. Some things weren’t really fixed until somebody became involved who really understood how the maths behind these things worked.” Cox is one of the better-known contributors to Linux, and is now employed by Intel.

Clueless
Another early contributor was Dave Miller. Miller was a less than successful student at Penn State University when he first encountered GNU/Linux, began hacking the code and sending the results to Torvalds. “At this point, I knew C for two days,” he told Glyn Moody, “and I had no idea what I was doing. I think the response was something like: ‘I think a better idea would be to try and…’
“So picture this. I’m clueless. I have no idea what I am doing, and yet he sends suggestions back as if I did have a clue.”

Miller stayed at Penn State for one semester, before moving onto Rutgers, where his academic career, as a computer science student, came to a rapid end. “I failed out of school after two semesters. My final semester was pure perfection; I failed every class,” he says. Miller is now employed by Red Hat, and has contributed more changes to the Linux kernel code over the last five years than any other single contributor.

Kleenex and ketchup
Just as important as the kernel developers were the packagers and maintainers, translators and the developers who worked on ‘userland’ applications. Almost as soon as the Linux kernel appeared, users saw the need to package the elements of Linux and GNU into a coherent ‘distribution’ with installers and partitioning tools that made the software accessible, and user tools that made the software friendly.

The first GNU/Linux distribution was Owen Le Blanc’s MCC Interim Linux, which appeared in February 1992, and was made available by FTP from the University of Manchester’s Computer Centre (MCC). The first popular distribution was SLS from SoftLanding Linux Systems, but SLS had its shortcomings. In the Debian Manifesto, Ian Murdock, the founder of Debian, was moved to say of SLS: “It is quite possibly the most bug-ridden and badly maintained Linux distribution available; unfortunately, it is also quite possibly the most popular.”

SLS fell out of favour and other distributions, such as Debian and Slackware, took up the slack and became the model for those that followed. The most famous of these, and the one most associated with the later commercial success of Linux and free software, was Red Hat.

Red Hat was founded in 1993 when Marc Ewing and his friends, Donnie Barnes and Erik Troan, decided that what the world really needed was a GNU/Linux distro that worked straight from the box. The inspired choice of name and logo was the stuff that marketing people dream of – and came about by chance, as Marc Ewing later confessed: “In college I used to wear my grandfather’s lacrosse hat, which was red-and-white striped. It was my favourite hat, and I lost it somewhere in Philadelphia in my last year. I named the company to memorialise the hat. Red and White Hat Software wasn’t very catchy, so I took a little liberty.”

But Red Hat didn’t really take off until Bob Young became involved. For the first months and years, Young worked “in my wife’s sewing closet in Connecticut, and Marc in a spare room in his apartment in Durham (North Carolina),” distributing the brand, free CDs, red fedoras and trademark T-shirts.

As Jon ‘maddog’ Hall tells it, Bob Young put Red Hat on the map, and would say “I’m building a brand”, but nobody listened. “By the time people recognised what he was doing it was too late. He had built the Red Hat brand to the point where a lot of people in the US would say ‘Red Hat is Linux, and Linux is Red Hat’. Like Kleenex and tissues, and Heinz and ketchup, Red Hat had become a generic term.

“Bob Young did an amazing job. He recognised that in building up the company, he needed to bring in some good executives with the right skills to take over when he stepped out of the way… He would say, ‘Hey, give away the software and sell the T-shirts and hats’, but what he really meant was: ‘Give away the software and sell the services.’ And still people didn’t understand.”

Under the radar
Free software was a world-changing idea, and the community that had grown around Linux managed to both capture and encapsulate the romance of something that was new and hopeful, and promote an atmosphere of do-it-yourself – ‘anyone can do that, I can do this, you can do that’ – which made it possible for everyone and anyone to become involved. And they did, and ported GNU/Linux to everything and anything, and if it didn’t work, there was every chance that it might later.

The ethos and community that grew around Linux gave everybody the chance to take part and make a difference. You could do anything with GNU/Linux. If you couldn’t write code, you could write and translate documentation, host websites, make friends or sell T-shirts. The GPL contributed to this effect and worked well for the long-term interests of the UNIX companies, but the success of GNU/Linux owed just as much to its growing ecosphere and popularity among a new breed of developers who brought something different to the mix.

Part of this effect was that there was a project to port Linux to the IBM mainframe, which was taken up by a skunkworks project inside IBM – from which other things followed. Linux became scalable from the wristwatch to the mainframe, because those who worked on it, and those who used it, said it would.

Against the odds, Linux flourished in the now-classic ‘under the radar’ scenario where sysadmins and programmers led management into conceding that Linux might be an acceptable alternative: ‘Okay. We’ll switch the web servers over to Linux. How long will it take?’ and the sysadmins would walk away and just make up a number because they had surreptitiously been running Linux on the web servers for months.

By the time of the dotcom boom and crash, LAMP (Linux, Apache, MySQL and PHP) had become the logical choice for startups to run their web servers, and Linux was well established in the enterprise – officially or otherwise – on file, print and web servers.

By the end of the decade, when Red Hat went public and was valued at upwards of US$5 billion, GNU/Linux was well on its way to becoming the standard offering on IBM, SGI and HP servers, and had already become the universal operating system that UNIX had always promised to be.