I absolutely love articles like this that humanize the creation of such a megalithic thing. It was just someone who was curious enough to tinker for a while on a project who was hanging out with friends. I love it.
I think it would be hard to overstate the impact of Linux and free software. I do climate science research and its pretty obvious that every single paper that contains an important finding was built on FOSS. Latex, Linux, python, all the gnu tools, and so much more. Imagine being a researcher who needed to write any of that, or pay for all of that, in order to work. It's incredible. Thanks to everyone who codes, builds, or contributes to such projects.
Also, it's laughable how little that is recognized in the science community.
>> One day, Linus accidentally attempted to use his hard drive to dial the university, resulting in his master boot sector starting with "ATDT" and the university modem-pool phone number. After recovering from this, he implemented file permissions in his kernel.
Now this nugget made my day. I love Finnish deadpan humor.
Reminds me of how I nuked my first Linux drive. I was using the cthugha[1] music visualizer app. You had to pass in a CD-ROM device as well as a DSP device (long before ALSA existed). Instead of the DSP I probably passed in my hard drive. Whoops.
I miss OSS and /dev/dsp, being able to write to the audio output like it was a file was great fun. I spent many hours back in the day just using bash scripting to make sound, I never got things good enough to make any worthwhile music but I did enjoy myself. Never accidentally specified /dev/hda instead of dsp but I did once do rm -r /usr instead of ./usr, switched to always using absolute paths for such things after that.
I've actually seen a rare case where a rebooting machine with messed up DMA tables managed to deliver incoming network packets to the hard drive, corrupting it. It took a bunch of really smart people a long time to figure that out.
Absolutely. But also, as a species, we're such a fan of these narratives that lionize people like Steve jobs. A brief article written by a friend about a project from their college years is much more humanizing than anything I've read about Steve Jobs. Sure, there's the "garage beginnings" part of the story but even that is very much lionized as "grind and hustle".
I think the way the myth of Steve Jobs and his impact on the world is a great example to juxtapose against Linus\Linux in so many ways. Linux runs nearly everything and was given away for free and the broader public is totally unaware. Steve Jobs sells hardware and constantly played up the lion narrative as a push to make money and be important. Im more interested in the former than the lot.
Inflammatory comment? Maybe. But I think we need to shift the values that we collectively encourage and this article is fantastic.
> But I think we need to shift the values that we collectively encourage and this article is fantastic.
Thank you for expressing that, I agree wholeheartedly. There's this increasing sentiment that everything is a product - something many people on HN are responsible for promoting. Where once was humility, there's now opportunism: Terminal emulator? Meet subscription service. In an industry defined by it's ability to generate hype cycles and profit off them, there's something brilliant and refreshing about FOSS and it's culture.
Steve Jobs got the private jet and the flashy keynotes, but Linus Torvalds is undoubtedly the bigger rockstar.
Woz had nothing to do with saving Apple in the 90s - and nothing for nearly 40 years - the only reason we talk about Apple today at all is because of Jobs and the small group he supported himself with in the late 90s. He was clearly smart enough to figure out how to save Apple. That’s more than “just” being a designer. 3 predecessors couldn’t figure it out and Gil Amelio wasn’t an idiot (Spindler wasn’t an idiot either - but oof talk about being out of your element)
The Apple of Wozniak is a historical footnote. Apple of today is NeXT - Jobs’ company.
A company doesn’t just succeed on technical prowess which is more where Woz’s skills lay. Trying to decide which one of them is more intelligent is pointless. Steve Jobs was smarter than the majority of the commenters here IMHO.
> the only reason we talk about Apple today at all is because of Jobs and the small group he supported himself with
It's fair to say that success can be attributed to the Apple II, which without Woz wouldn't exist. Yes, he did not make the iPod - but he bolstered Steve when his technical council was empty and was content doing so for next-to-nothing. There's no need to belittle his actions, obviously nothing he did competed with the work Jobs was doing. The two operated in their own lanes.
> Trying to decide which one of them is more intelligent is pointless.
I mean, I agree. Suffice to say that Steve Jobs could have never turned in his first Atari contract if Woz didn't do the work for him, though.
Yep, Apple needed them both. The Apple II was there early, but by the mid-80's it was very overpriced for its capabilities, compared to the Commodore 64 and Atari 8-bits. It needed Jobs' marketing abilities.
Late 80's and 90's Apple survived purely because of marketing. Both Commodore and Atari had them beat technically, with the Amiga and ST. Remember, both of those could emulate the Mac and were half the price.
I remember them, and I have to tell you the Commodore and Atari 68k machines were not obvious winners. Ugly software, ugly hardware. I know this isn’t what fans of those machines want to hear, but Apple’s design aesthetic blew them and the PC away; Apple was simply in another league.
Late 80’s Macs had the best user experience available. Again, this isn’t a popular take, but it is absolutely true. “Apple’s marketing prowess is what sold them” is bullshit; Apple was better at selling, but not so much better that it could overcome perceivably middle of the road products. Apple’s industrial and UI design led the industry. Everybody else followed.
You are, of course, right. Apple's UI was better. I was primarily an Amiga user, but also had a friend with an Atari ST around that time, and both their desktops were kinda ugly. Amiga got more professional looking with 2.0, but still no match for the Mac in terms of usability.
And they filed for Chapter 11 bankruptcy and were bought for a file sale price by SGI (I think Cray could have used some high quality market research in the 80s.)
> Steve Jobs was smarter than the majority of the commenters here IMHO.
Nah, you really think so?! Just because he was probably the most successful business leader in his generation you think he was smarter than us, HN commenters, despite our incredible feats like foreseeing the DropBox fiasco, predicting the LISP AI revival and so many others?
Kind of odd generalizing 'smart'. People have different skill sets and it's pointless comparing eg dev skills with marketing skills.
Imo, Jobs was good at making hype, that was his skill set and he was undoubtedly good at it. Woz, Torvalds, have very different skill sets from his but it's apples to oranges.
The brains behind the hardware and software of the Apple II - a remarkable accomplishment, to be sure, but even there, the marketing and packaging that Jobs contributed was not nothing. And Woz contributed nothing whatsoever to the Mac, let alone the iPod, the iPhone, or the products that came later.
Woz left pretty much no traces on contemporary Apple. Jobs is all over the place (and, if anything, it's Tim Cook who is the underrated contributor). And for all the narrative of Woz as the unsung hero, he parlayed his fame and wealth into an impressive string of failed products, up to and including a blockchain (which, I think, is still a going concern).
I think some of this is because people don't have the same freedom to tinker with things for fun or altruism without a profit motive, because life is so much more unaffordable now. Generations Y, Z and α are all struggling to just survive let alone potentially retire.
I think most of these opinion threads about Steve Jobs should be unhooked and connected to a root thread that points to Walter Isaacson's biography. It's painfully obvious most people haven't read it.
Steve Jobs was one of the greatest marketers in recent history, and from that perspective he really was "totally dependent on them for [his] life and well-being".
Others did the work going from zero to one, he did the work of going from one to a hundred.
While I do agree with you on the effort assessment, or certainly the "trailblazing" factor... (to me, writing a kernel and operating system that could be used by the entire world before there existed a notion of what that kernel and operating system might look do that is infinitely more impressive than being a salesman)
.. I do think it's a red herring. To me it's not so much about effort than it is impact. If every person to ever write code was trying to sell a product, progress would have halted as it started. It's both the effort, the philosophy, and the philanthropy of it. Which is also what has made it thankless-ish.
Jobs figured out how to collect and amplify the efforts of talented people into a industrial organization with products that have had a world-changing impact.
I don't know how much of that is his talents and efforts and how much of that is luck of being with teams that succeeded at a time when industries he chose were rising. I wouldn't discount either the possibility that he was distinctly good and industrious... or that dozens/hundreds/thousands of people of equal distinctiveness didn't make it for arbitrary reasons. And anybody who isn't sure the rewards are equitably distributed is probably correct.
But I do recognize that functioning in such a way that you can effectively collect and amplify the efforts of talented people is a non-trivial feat.
While I agree with everything said and celebrate the availability of open source tooling for just about anything, there's an indictment in here that the possibility of researchers paying for tools they need to work is relegated to the realms of imagination.
I think I get your sentiment, but what I think is the funny part is that when you PAY for apple HW/SW you are not allowed to use it as tools, instead it's locked down.
I gave up on iOS devices a few years ago when they forbid things like 'wifi scanner' on iOS. A tool I used regularly to install and measure the quality of WiFi network/coverage.
I just love the success Linus had with managing this gigantic project while not using means of SCRUM or any of the other fancy words and systems! Sure he might not always be the friendliest but I just can not imagine Linux being successful today if it followed any of today's modern software-dev-methodologies like scrum.
I just love my linux-workhorse-daily-driver :)
Thanks Linus and everyone else who had a hand in my awesome Linux desktop.
> After finishing the game, Linus started learning Intel assembly language. One day he showed me a program that did multitasking. One task or thread would write a stream of the letter "A" on the screen, the other "B"; the context switches were visually obvious when the stream of As became Bs. This was the first version of what would later become known as the Linux kernel.
Reminds me of: "Gall’s Law states that all complex systems that work evolved from simpler systems that worked. If you want to build a complex system that works, build a simpler system first, and then improve it over time."
I owe so much to Linux and the whole free software movement, as a self-taught programmer. I set up my first Linux computer in late 1995 - it was a laptop with 4 mb of memory (which wasn't much even then) running Slackware that I'd downloaded over a modem over the course of several days.
It was so cool to realize I could examine and modify the source code for... all of it! And interact with the people who had written all of it. I was hooked. I'm still using Linux as my primary development system these days.
I recall a friend installing Slackware about this timeframe (1995), where you could select which packages you wanted. We missed groff. As a result we didn't have man pages, had no idea how to fix it, and ended up doing a re-install. I miss those days.
This is a great read of a humble start of history. The A..B concurrency printing, helping Linus out implementing Sprintf, learning assembly. Building it all towards a coherent system.
What was truly profound is how Linux and other open source systems and libraries completely cracked open systems software for the masses. Back in the early 90s, everything around Unix, operating systems, compilers, tools, everything was crazy expensive to buy. And all of a sudden here came this thundering herd of code that was all completely free. Of course Gnu stuff had been around for awhile, but Linux is what propelled it into warp speed adoption. In thr 90s I was privileged to work at Bear Stearns and had a Sparc Station on my desk and regular interactions with SunOS, Solaris, and HPUX for the guys downstairs. Outside of University and business, it was hard for individuals to break into it.
By the late 90s any old anybody could get a CD ROM with Red Hat or whatever.
I was born in 1989, so I missed the early days of Linux. However, I first heard of GNU and free software back in 2003 when I was looking for a free alternative to WinZip, a shareware utility for decompressing zip archives. That's when I discovered not only a free alternative (I believe it was 7zip), but also the GNU General Public License. Up until that point, I never heard of free, open-source software (this was before I heard of Mozilla as an alternative to Internet Explorer, which I found out about in early 2004, and later Firefox). I already knew how to program, but it was within the Microsoft ecosystem; I had a copy of Visual Basic and I longed for the day when I could save enough money for the entire Visual Studio suite. But once I read about free, open-source software, it was like a whole new world of software opened up to me. I was a bit of a "goody two-shoes" who didn't want to pirate software, but I couldn't afford licenses for many commercial software packages, so having access to Linux, OpenOffice, GIMP, GCC, and other applications was a very big deal to me.
One of the happiest moments of my teenage years came in 2004 when a teacher gave me his old 475MHz AMD K6-2 desktop with 64 MB RAM and Windows 98. Finally I had a computer of my own instead of using the family desktop! I downloaded and installed ZipSlack, a distribution of Slackware Linux that booted from DOS. Later that year a community college professor gave me FreeBSD installation disks. I credit these things for making me stick to a path of pursuing a career in computer science; at the time I was strongly considering majoring in linguistics, but I got entranced by operating systems and C thanks to Linux, FreeBSD, and gcc.
Nearly 20 years later it still amazes me how anyone with an Internet connection and a computer can not only download and install production-grade software that powers billion-dollar businesses absolutely for free, but also download and study the source code of these software tools for absolutely free as well. Sometimes I get pessimistic about the state of computing these days, but it's things like free, open source software that reminds me why I love computing so much and why I still pursue a career in this field.
> By the late 90s any old anybody could get a CD ROM with Red Hat or whatever.
That's exactly how I got into Linux. I bought a book about RedHat 5.2 with a CD ROM install disc in the sleeve :) Have been using Linux on all of my personal devices (and every work device that I'm allowed to) ever since.
Oddly enough, I also had a Sparc Station 5 on my desk at one time. I remember the filesystem seemed very slow but it was cool as hell to be able to use "real Unix." I remember it even came with Internet Explorer of all things!
Exactly. I was in university '95-99 and had access to amazing Solaris, IRIX, AIX, and other machines, grew to love UNIX, and wanted to run it on Intel hardware at home. RH Linux did just the trick.
RH 6.2 with Ximian Desktop and Crossover for Office support was peak Linux desktop. Then Microsoft released a new version of Office that purposely screwed with Crossover, and Nat and Miguel joined hands with the Great Satan, and desktop Linux spent a decade wandering in the wilderness.
i still remember reading when Carmack said (something like) anyone can learn to be any skill of developer, all they need is a used PC and linux CD.
this was an eye opening quote for me when i read it, for some reason. it meant that the craft of software was the most accessible in all of human crafting. really an amazing time to grow up.
Absolutely. I remember the articles in the trade press in the mid 90s asking "Is UN*X dead?" (Hindsight of course tells us that Betteridge's law applied)
But I was working for a small ISV making a windows based GIS and our customers were putting windows boxen on every desk in the office, and then ditching either intergraph based solutions or complex systems with Tektronix terminals. Our per seat license was a rounding error in their previous costs, and you could drag a selected rectangle from our map window and drop it into a word document. I occasionally helped with sales demos and you could see when the decision was made in the faces of the senior people when they saw this.
Are those some sort of standard teaching exercise? Because before this article I only saw them in the description of the “concurrency geek” archetype on OSDev Wiki: https://wiki.osdev.org/Eleanore_Semaphore.
If you're interested in Linux history, check out Just For Fun by Linus Torvalds himself. It's written in 1999 if I recall, which is so much closer to Linux's inception than it is to today, so it's so interesting to hear thoughts and predictions about Linux at the time.
Super cheap book. I think I got my copy for $4 on ebay.
I found this book in my local library when I was in college, it was a really fun read. I had recently finished the OS course which used PintOS[1] and wanted to learn more about the development process. Unfortunately my C is still fairly rudimentary, and at this point it's going to be a bit longer before I attempt to follow Linus's path.
To those that wondered what it was like to get Linux running in those early days, here is a post I made in April 1993 (30 years ago this month!) to the "comp.os.linux" Usenet group trying to dual boot Linux and DOD on my new 486 PC with I think no more than a 80MB harddrive. (BTW a2,a3,a3 refer to the first 3 floppy disks you need after the boot and I think root disk to start installing)
Help installing Linux.
5/4/93
Hi,
I,ve tried to install the SLS base system 99p6. I partitioned my 210 Maxtor drive to half DOS and then added a 15M /dev/hda2 (type 81 Linux,Minix) and 10M /dev/hda3 (type 83 Linux swap). After writing to the partition table and rebooting, I did a mkfs /dev/hda2 15504; mkswap /dev/hda3 4096; swapon
/dev/hda3.
After then running doinstall things seemed OK until it tried to start installing a2,a3 and a4. The system only looked at the disks for about a millisecond and then prompted me to insert the next one ( I was thinking, Boy! this installation is quick). Finally I got the prompt for formatted disk to make a boot disk. Thinking, I had finished, I attempted to boot this
disk to no avail. Of course during doinstall I got messages such as
"mv: command not found", "rm: command not found" and so forth.
I am guessing that my /dev/hda2 is not mounting and hence the file system is not being transferred. I have been trying to dissassemble "doinstall" as most errors seem to be redirected
to /dev/null.
One "error" I noted on booting is "WD8013 not found at 280 ". Can't the driver find my disk controller? I have a TMC 486/33 motherboard with a VESA IDE controller.
Linux seems to be able access the disk because the partitioning worked and mkfs -c works the disk out.
Please help me out , I would love to get Linux up and running.
I remember running X the first time on Linux just after it was reported working because I happened to have the exact chipset that Peter MacDonald supported in my 386 at the time. We ran it off of a boot and root floppy - both 3.5", 1.44MB disks.
We didn't do much with it at the time, but it was a big deal... we had only run X on workstations before that, and they were expensive and few and far between.
Aaah, yeah, the "good old days". I got started with Linux around 1995 or so. I don't actually remember which distribution I used first, as I was experimenting a lot back then. I remember using Turbo Linux, Yggdrasil, and Slackware at times. Sometime in 95 or 96 I was a student at UNC Wilmington and I remember a bunch of us were out in the courtyard by the C.S. building when somebody came up talking about this new company called "Red Hat" that was, get this, selling Linux! That led to some rather, erm, passionate, discussion that day.
I actually wound up gravitating to Red Hat linux myself in the end, and to this day I still use RH oriented distros (mostly Fedora) for most things. And I even wound up working for Red Hat for a spell.
So yeah, Linux and his "toy" kernel definitely had a huge impact on my life over the years. And so did Bob Young and his company that had the audacity to charge money for free software. :-)
I installed slackware as a 12 year old in the early 90s. I only have 1 PC in the house (other computer was amiga 500 with no internet). My ISP only supported PPP connection which didn't work very well at the time on Linux.
Since the PC was my only source to online help I literally had to reinstall windows and then slackware every time I made an attempt at getting it working. It took probably 15-20 rounds over a couple of months but when I got it felt like the biggest accomplishment of my life.
More or less the same (12 years in 1994), however we didn't have internet until some years later. I learned pretty much everything from a Dutch Linux book (which was on the market very early) and all the HOWTOs and guides from the Linux Documentation project.
I sort of remember the first one I had. It came on a couple of 3.5 floppies. I promptly ignored it for a year or so then came back and it was 20+ of them now for slackware with hundreds of utilities I had no idea what they did. That was a ton of fun digging thru.
This article is a goldmine of old, classic stuff and lots of information I have not yet read. Great read and nice of the author to share the experience, I hope people with these kind of experiences continue to share their subjective views on history.
Unrelated but fun highlight from the mentioned newsletter:
Not to be a GNU/Linux instigator but I remember when I first read the GPL how strange it was to have a legal document to give something away it seemed like a impish prank - will someone take it and insist you stop giving it away, or sue you when their production line breaks down.. And then the flamewars began, the pronunciation, the toolchain, freetards, 'open sores', etc. but it was the GPL making it all possible.
This just made me realize that even though I pronounce Linux as "Lee-nux" (the same way he does), I always read his name as "Lie-nus" (like the Charlie brown character).
> The first releases of Linux used a license that forbade commercial use. Some of the early contributors suggested a change to a free-software license. In the fall of 1991, Richard Stallman visited Finland and I took Linus to a talk given by Stallman. This, the pressure from contributors, and my nagging eventually convinced Linus to choose the GNU GPL license instead, in early 1992.
people often use 'git' for people they think are unpleasant, but that is in order to insult them — 'git' means 'stupid person', not 'unpleasant person'
> 'git' means 'stupid person', not 'unpleasant person'
Native English speaker (UK) and I've never heard 'git' used to mean 'stupid person'. I've only ever heard it used to mean 'unpleasant person', generally in the form of "complete git" or "utter git".
(Of course, language use can be different in different countries.)
The Linux kernel took off after Linus receive the MINIX book, a.k.a Operating System Design and Implementation by Tanenbaum, which include the source code of MINIX. Until the arrival of the book Linus was playing "Prince Of Persia" on his new 386AT desktop.
There were several microkernel-based research/hobby/free kernels in early 1990s, like L4, VSTa, or even HURD. Perhaps the reason why Linux succeeded as a general purpose OS (and not one of these) is because of its conservative design.
This has to be the funniest tech article I have ever read. I was literally laughing out loud. I don't even remember the last time I had such an emotional response to a tech article. Probably it was something by whytheluckystiff back in my teenage.
Lars has written about his interactions with Linus before in this article: https://liw.fi/linux20/
I fondly remember these lines:
That summer, I re-connected with my fellow students from the university,
including Linus. One of the students a few years older than Linus and me,
Patrik, had a sauna in the building he lived, and he booked it once a week
for the group of us. You perverts who think the word sauna is a synonym for
brothel can stop imagining orgies: in Finland, saunas are for bathing,
medication, even meditation, not sex.
There’s a semi-famous picture of Linus, looking naked, drinking beer. The
picture was taken at Patrik’s, by a fellow student, Stina, and Linus is
wearing trousers, but it was a hot summer day, so he didn’t have his shirt
on. No orgies. Really.
There is almost a (probably misleading) glimmer of “i could have done that”, because the story is so human, he really did have to learn assembly by printing A and Bs, he didn’t have a grand design and just smash out an OS in vim and upload it! This is real agile programming.
I first downloaded Slackware in June '93. I needed 13 diskettes. Installed it on my Compaq 386/SX 4Mb laptop, and spent the rest of the year learning myself Linux.
In November '93 I founded one of the first internet providers in The Netherlands, and I'm pretty certain my family was the first Dutch family that had always-on public internet from the home.
Also I might have been one of the first companies in the NL to run Linux in a business-critical environment.
Linux always kept me from running Windows on my desktop. Although I switched to the Mac in 2005 after big problems get multimedia And graphics to work reliably.
Without Linux my life would have been completely different. It allowed me to create a new future for myself. So I'm hugely grateful for its existence.
I started using Linux in my early teen years. I'm now 39. When I started I had dial up internet and it took me a week to get my Win Modem to work. I would boot into Win95 join a IRC server like dalnet, jump on #Linuxhelp and ask for suggestions. Reboot the system into Linux and try it out. Then repeat.
I think the early community deserves a big thanks, especially those on IRC.
I was doing CS at a similar time. Used Ultrix and HP-UX at work. Until Linux and Minix it was very hard to get a Unix on a PC. Someone had SystemV running, not sure how.
Once Linux appeared, other home Unixes started evaporating, except for *BSDs.
I was at the very first Sydney Linux Users Group meeting (SLUG). Have pics! The only pics of that event. Have tried to get permission from others so I can share, but no response :-(. Might share them anyway.
Later in the 90s we had Linus over for a conference. Had a day out the harbour, good times. Got an excellent shot of Linus's head framed by the Opera House. Then found I had no film in the camera!
Linux came back to Australia more than a few times of course. On one of these he was infamously bitten* by a Fairy Penguin.
* Not sure if you could describe such a tiny animal as having a bite!
I'm pretty sure you should share them (statue of limitations and all that ;-) ). I probably went to my first SLUG meeting maybe '94 or '95. (I bought my 1st PC specially for dual booting DOS and Linux with Slackware early '93)
This is great. A few years later I would stumble upon this new operating system. I was hooked. I spent all my time tinkering with it, building with it, so much so that I ignored the outside world around me. Case in point, me, ignoring the world around me [1]. I was maybe 13 at the time. 1996 I think. Man, the wonder of the unknown and the potential of the future… :D
I can't find the article anymore (and I've looked for it several times in the past) but someone was making the point that open source was essentially an economic phenomenon. Insofar as distribution costs went to zero with the advent of mass-access to the internet then it was inevitable that people would start sharing software. Obviously this takes nothing away from all those early contributors, but it is food for thought.
Mind you while Linux was taking off the BSDs were apparently busy in lawsuits. So, while the zeroing of distribution costs should've benefited them, it seems Linux was at the right place at the right time, minus the baggage.
> Mind you while Linux was taking off the BSDs were apparently busy in lawsuits. So, while the zeroing of distribution costs should've benefited them, it seems Linux was at the right place at the right time, minus the baggage.
"With Linux, I just booted from a Linux boot floppy with my Linux install CD in the CD-ROM drive, and ran the installation. With BSD...it could not find the drive because I had an IDE CD-ROM and it only supported SCSI."
"It insisted on being given a disk upon which it could completely repartition. [...] Linux, on the other hand, was happy to come second after my existing DOS/Windows."
"By the time the BSD people realized they really should be supporting IDE CD-ROM and get along with prior DOS/Windows on the same disk, Linux was way ahead."
That was a failure of governance, which was a result of their failure to have an open development model. Linus was very liberal in accepting help from hobbyists and uncredentialed people (people who were hardware-poor, and hence needed support for stuff like dual-booting...), the BSD world has always been more opaque and closed.
In a way it was a victory of horizontal, open, "upstart" governance, versus aristocratic and elitist organization.
BSD's aren't really opaque. I understand what larger point you're trying to make about the difference between linux and bsd development. But the biggest BSD projects all have mailing lists where you can see the decisions being made, and the source of the core of each one is available to everyone.
BSD was Unix people used to workstations hardware so that's what they targeted on the PC. Linux was PC people who wanted it to run on whatever cheap hardware they had
The problem with BSDs was not much the lawsuits but the fact that they didn't figure out a governance model that could scale. Linux had open mailing lists and low barriers to entry, BSDs had "core" cliques. Linux begat git, a distributed VCS, when BSDs were happy with the likes of CVS, where control is rigidly centralized. Etc etc.
git didn't appear until much much later; and maintainers who can pull in code is not so different than a "core clique" with a commit bit: it's till a relatively small group of people who decide what does and doesn't get merged – it's more of a workflow thing than anything else. Of all the possible factors this seems like the least convincing one.
The "Net" in NetBSD (the first open source BSD community to coalesce after the lawsuits settled) refers to that, it's the network operating system. And not really in the sense of an OS for networks (though it certainly is that) but one from and of the network.
I don't know that you were necessarily thinking of this, but I did write a piece on exactly this in 2004.[0][1] I couldn't really find anything else at the time that was talking about this, so if you did/do find something else, I would love to read it!
Thanks to Lars Wirzenius for sharing his first Linux (and Linus) memories.
BTW, nowadays Lars teaches Rust using his own consultancy.
I well remember (the younger me, a school kid) having to install 50 3.5" floppy disks FTP'd from the University of Karlsruhe's servers, only to discover that some where faulty & having to do it all over again. And then the reward of having the X11 logo and xeyes on your own desktop - 1992 was a fantastic year!
What a great sense of humor this author has! I literally lol'ed at some of these jokes. It's hard to make humor shine through writing, so I'm double impressed.
Some of the best, but don't deprive yourself of the whole article (because it's a gem):
> More importantly for the future success of Linux was that the X11 system was ported to it, making 1992 the year of the Linux desktop.
> Thus, mine was the first PC where Linux was ever installed. While this was happening, I was taking a nap, and I recommend this method of installing Linux: napping, while Linus does the hard work.
> A couple of years later, he spent days playing Quake, ostensibly to stress-test kernel memory management, although that was with a newer PC. Much fun was had in that room, and there were no pranks whatsoever. None at all.
> Alas, early Linux networking code was occasionally a little rough, having been written from scratch. At one point, Linux would send some broken packets that took down all of the Sun machines on the network. As it was difficult to get the Sun kernel fixed, Linux was banned from the university network until its bug was fixed. Not having Usenet access from one's desk is a great motivator.
> In the spring of 1994 we felt that Linux was done. Finished. Nothing more to add. One could use Linux to compile itself, to read Usenet, and run many copies of the xeyes program at once. We decided to release version 1.0 and arranged a release event.
> I insisted that a version-control system be used. I had witnessed students in earlier courses do the shouting kind of version control: the students shared a source tree over NFS and shouted "I'm editing this file" when they were changing something. This did not seem like an effective method to me, so I insisted on CVS,
> In 1997 Linus graduated and moved to the US to take a job at Transmeta. I took a job at a different university in the Helsinki area. In the following years, many things happened. It turned out that there were still a few missing features from Linux, so people worked on those.
> The term "open source" was coined and IBM invested a ton of money in Linux development. Netscape published a version of its web browser as open source. Skipping a few details and many years, open source basically took over the world. LWN was started and covered much of this history on a week-by-week basis.
The 2nd part of the CVS section was funny too "[...] This experience is why Linus dislikes CVS and for years refused to use any version control beyond uploading tar balls to FTP sites."
I’ve had the great privilege of working with Lars. This deadpan humor is a constant feature of his discourse and even better in conversation than in this article.
That's funny! I'll still go ahead and attempt a somewhat serious answer. xD Arguably it's an addition to his original code that possibly modularizes his kernel in ways that he isn't familiar with, and that also might add completely new and unknown features to it. Thus it would probably behove him to read the Wiki either way. On the other hand, perhaps he's so 1337 that all he needs to do is to read the source code. My most major caveat here is that I don't myself know if there are any actual differences between his original kernel, and the code committed to the Arch Project. I thus think others are more qualified to say something intelligent about that, but these are my assumptions anyway. Hope you're having a great weekend! Skål!
> During this time, late spring of 1991, I wrote an implementation of the C sprintf() function for [Linus Torvalds], as he hadn't yet learned how to write functions with variable argument lists. I wanted to spare him the pain of having a different function for every type of value to write out.
That’s such a cool detail. Everyone starts somewhere.
(Shuffling off to commit change to Git and deploy to Linux…)
Circa 1996 I was getting my CS degree. At the same time, I got a job at a small local company running an ISP on the side. A dozen Hayes modems hooked up to a Livingston Portmaster. T1 uplink. And two PCs (named "hops" and "barley") running Slackware. That was my first experience with Linux. It was a barbaric OS compared to the Unix workstations we had in the CS department running HP/UX and SunOS. But it had promise...
I think my first open source contribution was while working there (I added the "-e" switch to chpasswd) and it's still part of Linux today. Let's see... found it:
Yes, absolutely. But it was solid Unix OS with excellent documentation. I also worked for the CS department. We installed software centrally to an NFS server which was then automounted. We used something called "depot" from CMU to setup a software depot:
The biggest PITA about this scheme at the time was that you wanted "make install" to put software in one location (/depot/...), but you wanted that software to have paths compiled in as if it were in a different location (/usr/local/...). For the most part, the Makefiles that came with open source software then just weren't designed that way. You could set PREFIX but it was used for both the compiled-in paths as well as where "make install" put things. I recall spending a lot of time wrestling things into the depot scheme.
Seems absurdly over-engineered looking at it now.
Later in my career, I spent a lot of time building RPMs which was similarly painful but got easier over time.
> In 1991, Linus wrote that Linux "won't be big and professional like gnu". In 2023. Linux is running on every continent, on every ocean, on billions of devices, in orbit, and on Mars.
> I had witnessed students in earlier courses do the shouting kind of version control: the students shared a source tree over NFS and shouted "I'm editing this file" when they were changing something
We did exactly this for a project my freshman year of college, in 2011, except it was Dropbox instead of NFS :)
Great read. The first distribution I used was TAMU (https://archiveos.org/tamu/) in 1992, when Linux wasn't yet at 1.0. I had to download dozens of floppy disk images. In those days, getting X to run involved real risk. Get some of the parameters wrong and you could destroy your monitor. My five kids (all in their 20s and 30s now) grew up running Linux. It's still the OS I choose when the option is available.
> X to run involved real risk. Get some of the parameters wrong and you could destroy your monitor.
The flexibility also allowed you to accomplish neat things.
I was spoiled by high res 21" workstation monitors at school, so my laptop with a 640 x 480 screen was very limiting. But, telling X that the LCD was a multiscan monitor made every other scan line go to /dev/null-- the effect was a pseudo-resolution of 640x960 with scrunched up, but very readable fonts. Ran that setup for years.
Same here for RH 5.1 in ‘98. I spent most of a summer messing around with trying to get the proper video modes working on my graphics card and learning the ins and outs of X configuration. Then came the sound card…
Some friends told me they ran into Andrew Tannenbaum at Embedded World a few years ago and asked him if he still believed that Minix was better than Linux. Apparently he said yes.
If I routinely had strangers coming up to me picking arguments about something argument I made decades before (always the same thing), I'd pretty quickly come up with the minimum necessary reply to end the interaction as well.
i started using linux in the mid-90s, after using commercial unices for the previous 10 years. i was amazed at how well-engineered it all was. i think i was using slack - too long ago to remember.
The only guy I knew who used commercial unixes at home back then, lived in Seattle and loved to play games.
Even got his girlfriend interested in them, which was unusual at the time.
Come to think of it, I remember one day some government goons took him away, right around that big nuclear scare. And no one has heard from him ever since.
If you want to take a closer look at the early days of the Linux kernel, I can recommend "A Heavily Commemted Linux Kernel Source Code", a listing of Linux kernel 0.12 with comments in the style of John Lions‘ commentary on 6th edition Unix (available in Chinese and English): http://www.oldlinux.org/
Wow, this brings me back to the crazy days of college in the fall of 1994. My roommates and I ran a mish-mosh of OSs, including Windows (3.11), OS/2, and Linux (Slackware distro). The last was certainly the biggest challenge, but once we had remote xterms to campus servers, there was no need to slog to a computer lab.
I also remember the era of free access to things. I had an account on some Stanford systems "just in case I wanted to run some stuff". It was a totally different time.
Yes, exactly. Around 1993-94 I was running through so many different OS's on my Micron Pentium. DOS, then OS/2 Warp, then Linux, then Windows 3.1, then back to OS/2, then to Linux. Drove my wife crazy.
I began tinkering with Linux about 6 months after it's inception. I think....think...it was what became Slackware in what I gathered up the things to make it run on my Micron Pentium.
This was in the day I was trying everything on that machine. I had DOS running on it at one point, then OS/2 Warp, then Windows 3.1, then Linux, then back to OS/2 etc etc. It was at a time where everything was up in the air, before Windows 95 took a major foothold.
This was in the day I was trying everything on that machine. I had DOS running on it at one point, then OS/2 Warp, then Windows 3.1, then Linux, then back to OS/2 etc etc.
Heh. I remember trying one time to get as many different OS's multi-booting on one physical box as I could. I think I had about 5 at one point, something like Windows 95, OS/2, Red Hat 6.2, FreeBSD (or maybe OpenBSD), Caldera Linux, and maybe something else. And then I tried to add one more (I think it was Solaris x86) and totally hosed everything and had to start over.
To me, one of the most remarkable aspects of this story and reading others' comments is how relatively quickly people seem to have picked up Linux despite the lack of modern high-speed internet and one-click away forums and networks we all enjoy today. I guess that's an inevitable outcome if the market is ripe for some change no matter the opposing forces, like missing features, strong alternatives or no commercial support, etc.
One of my earliest great Linux experiences was when I set up my 486 to download all of the floppy images (or maybe it was the Yggdrasil .iso), then went to Op Amp Bookstore in Los Angeles for lunch, finding myself an Yggdrasil distro, returning home, cancelling the download and getting on with it ..
That was really a watershed for me: the whole thing on a single CD .. wow! And bootable, also!
I first installed Linux in secret on a spare partition of a Windows 386 at University. It was 1994, and for a while it was like having my own workstation, normally the Suns and VAXstations were reserved for the Post Docs and researchers. Because PCs were in abundance and both Sun and VMS systems restricted and tightly controlled, many of us setup home on that PC. It was anarchy.
One big takeaway is how close the real Unix (in the form of 386BSD) came to being where Linux is today. If it had been available freely one year earlier, Linux wouldn't have gained momentum. Would it be better to have one true free Unix-like OS now, rather than two? Did the free-for-all development in Linux depend on it having no "pure" ancestry? Nobody knows.
> If it had been available freely one year earlier, Linux wouldn't have gained momentum.
I doubt that. Real Unix was a pretty horrible experience for a lot of hobbyists starting right from the disklabel, and many of those smug bastards where pretty horrible people to hobbyists.
The fact Linux worked on crap hardware meant people who could only afford crap hardware could learn it, and get good at it. Linux overtook Real Unix real fast in-terms of hardware support -- so fast I don't think *BSD ever stood a chance.
Stability and quality is a double-edged sword: You think you gotta keep the crap out or it gets crap, but the reality is that crap is what makes it fun, and fun is why people used Linux and not Real Unix. If you want to make something great, you've got to do is focus on things that matter, and remember: fun matters more than scsi.
> Would it be better to have one true free Unix-like OS now, rather than two? Did the free-for-all development in Linux depend on it having no "pure" ancestry? Nobody knows.
Some people say "pure" and others say "inbred". I say this thing you're thinking about is a distraction: anything that prevents people from trying things out is slowing progress, there just ain't two ways around it.
Worse is not better: It just means you don't know what's important. Better is better.
Lovely article. My first install was Slackare 2.3, kernel version 1.2.13 on a 386sx. Thirteen floppies. I don't even remember why I installed it or where I'd heard of it. I don't think I even knew what Unix was, but I vividly remember the "darkstar $" prompt when it was all done and then my social life suffered for a few months...
Hah. Used SLS (and then Slackware) back then. I think the first version of the kernel I used was 0.98pl13.
I worked at DEC back then and did not have Internet at home, so I would schedule jobs at night to download the 1.3" floppy images and then copy them to actual floppies to bring home.
Yep I did the same (I worked at BHP the Aussie industrial multinational, and funnily ended up at DEC in '97 and still work there via HPE). Initially I was doing the regular downloads to a SGI workstation and writing the floppy images out on I think on a DOS PC. Eventually I was able to repurpose an unused 386 at work to become my Linux server at work.
For me the first one that sticks in my head is the MCC distrib on ~10 floppies, with (I think) a bunch more for X. Definitely used 0.9[89] before that (late '93) but no solid memory of details.
Kernel version 1.0.8 is burned into my brain though.
It's strange that Lars did not mention about Yggdrasil Linux, it's probably the earliest Linux distro even before Slackware existed [1][2].
According to the article, 1992 is the year of Linux desktop by the introduction of X11 system but every year since then is allegedly the year of Linux on the desktop.
Anyway this article has taken me back down memory lane, thanks Lars.
In fact my friend who studied in UMIST (now part of University of Manchester) introduced me to Linux. The fact that Manchester Computing Centre (MCC) is based inside UoM probably make it accessible to him because at the time Linux was installed from floppies.
> At one point, Linux would send some broken packets that took down all of the Sun machines on the network. As it was difficult to get the Sun kernel fixed, Linux was banned from the university network until its bug was fixed.
Well well well, who's running on the university network now?
I felt SOO late to the game when I got into Linux in the 90s. And now I'm that old motherfucker with a quarter century under my belt. Holy hell, time flies.
I found this article resplendent because Linux has given back to me what I put into it year after year since it launched the year I graduated college.
What I wonder is this: is there a story happening now that has similar good will for the future of AI without having commercial or social computing as its basis?
The reason I love Linux is that Linus started and finished before commercial and social concerns became the raison d'etre in software.
It's hella amazing to see how Linux has been used in so many different industries and applications, from chip design to web servers to personal computers and moree. It's a testament to the flexibility and power of the open-source model, and the ingenuity of the community that has built and maintained it over the years
As an earlyish (1995) Linux adopter I'd heard some of these tales, like the fact that it started as a simple task scheduler that wrote a's and b's to the screen. But it's cool to hear it from someone who was actually there to witness those early events as they unfolded.
I'm wondering, is there any tutorial to guide one from a bootloader to a very early version of Linux (but already have the barebone programs)? I guess any serious OS Dev course will do that, but would be interesting to find one specifically focused on Linux.
Get an old iso of Slackware and do it from scratch. For the authentic experience do it without internet access and using only the included install instructions. I once attempted this as an exercise on a very old thinkpad having only DOS+mscdex on the hard drive and a Slackware 2.1 CD-ROM. Bootstrapping that system is a feat that I’m still proud of. Though it was sort of a pointless exercise for my own gratification.
That's how I did it in 1995. I printed a few howto's, downloaded 27 or so Slackware floppy disk images, copied those onto actual disks, and biked home and booted from the first disk to start the installation process. At some point I had an issue with a corrupted disk so I had to bike back to the university to create a replacement. Following the howto, I got a booting system on my PC (dual booting with MS Dos/windows 3.x). And then using the bundled documentation, I was able to get X up and running, figured out how to rebuild the kernel with the sound drivers I needed, etc. Fun times.
I remember installing Linux on my friend’s PC. I think it was either 1997 or 1998 and he’d bought a retail copy of Mandrake. We took it back to his place and got it installed and running on his 200mhz Intel PC.
Such a great article. Much respect to Lars. I had the great privilege of working with him at Wikimedia and he's genuinely one of the nicest and funniest people you could ever hope to meet.
Posts like these are the reason why I'm checking HN. It's really nice to read about Linux's humble beginnings especially given how wide-spread it has become today.
I think it would be hard to overstate the impact of Linux and free software. I do climate science research and its pretty obvious that every single paper that contains an important finding was built on FOSS. Latex, Linux, python, all the gnu tools, and so much more. Imagine being a researcher who needed to write any of that, or pay for all of that, in order to work. It's incredible. Thanks to everyone who codes, builds, or contributes to such projects.
Also, it's laughable how little that is recognized in the science community.