A Historical Look At The First Linux Kernel

LinuxFan writes "KernelTrap has a fascinating article about the first Linux kernel, version 0.01, complete with source code and photos of Linus Torvalds as a young man attending the University of Helsinki. Torvalds originally planned to call the kernel "Freax," and in his first announcement noted, "I'm doing a (free) operating system (just a hobby, won't be big and professional like gnu) for 386(486) AT clones." He also stressed that the kernel was very much tied to the i386 processor, "simply, I'd say that porting is impossible." Humble beginnings."

Comments

No sooner had I finished compiling...

[kaleco]

...than Gentoo using kernel 0.02 was made available.

Official kernel development strategy

[Anonymous Coward]

"When the first kernel pops, then toss in about 1/2 cup of open source developers and shake vigorously until the popping dies down. You don't want to leave it on until you hear nothing, because then it's sure to be burnt."

Too bad

[Anonymous Coward]

that that Torvalds guys project never amounted to anything useful.

That's it!

[Constantine XVI]

Instead of trying to convince Linus to go to GPL v3, we can just convince him to go back to the original liscence! :)

This kernel is (C) 1991 Linus Torvalds, but all or part of it may be redistributed provided you do the following: - Full source must be available (and free), if not with the distribution then at least on asking for it. - Copyright notices must be intact. (In fact, if you distribute only parts of it you may have to add copyrights, as there aren't (C)'s in all files.) Small partial excerpts may be copied without bothering with copyrights. - You may not distibute this for a fee, not even "handling" costs. Mail me at [email blocked] if you have any questions. Sadly, a kernel by itself gets you nowhere. To get a working system you need a shell, compilers, a library etc. These are separate parts and may be under a stricter (or even looser) copyright. Most of the tools used with linux are GNU software and are under the GNU copyleft. These tools aren't in the distribution - ask me (or GNU) for more info.

Linus even wrote a book about it

[bomanbot]

To get an even deeper look into the beginnings of the Linux Kernel, I like the book that Linus wrote which is called Just for Fun - The story of an accidental revolutionary (ISBN 1-58799-080-6, google the rest).

It contains the entire back history how Linux began as a side project and of course the famous spat with Andrew Tanenbaum over Minix and Linux and I found it to be a good (if very nerdy) read.

But the pictures in the article? Just sad, he reminds me so much of myself ;-)

It's Also a Great Story

[eldavojohn]

Anyone who cares already knows. Anyone who wants to find out, can find out. Well, that's not true, as it turns out, I care but I'm not afraid to admit I didn't know any of these things that Linus felt (having joined the game late, nor have I seen the documentation for the first release). And I don't go around looking at the history of software I use, I simply use the software. I use gcc but I'm not so sure its history would be that interesting to me.

Come on, this is like doing a history of the First World War again, complete with photos of the Arch-Duke. I don't think that's a very good analogy. This wasn't really a history of the Linux kernel, more so a short intro to the attitude of the single point in the beginning. In my opinion, reading through this and scanning the attached documents would be more like if you had given me the first three pages of the Bible. "In the beginning, there was this guy that wasn't very sure anything would come of a project ..." This is a good article because I kind of always thought that Torvalds started Linux knowing it was going to be big. I imagined him sitting down one day and saying, "F this noise, I'm going to write an operating system that works ... and I'll distribute it for free!" ... except in Swedish?

But that's not what happened and I think that's important for people to recognize. This was not unlike Frodo starting off on a quest thinking he wasn't going to get anywhere (though the motivation and implications are not so huge). It's the classic hero-by-accident story and since it's a true story, I love it all the more.

You know, I always thought about writing to Paul Harvey (if he's still alive) and asking him to do a "And that's the rest of the story" on Linus. That would be some classic stuff. Although most the listeners, probably not even sure what Linux is let alone know its creator Linus, would probably think he's suffering from some form of dementia set on by age ...

And it isn't something the is deserving of a place on this website! Are you kidding me? This actually makes me want to start some open source project even though I recognize it will probably go nowhere. Of course this belongs on this site.

Oh No!

[MyLongNickName]

He looks just like Bill Gates at that age!

At least two people agree on 640kB

[Centurix]

From memory.c

* NOTE 2!! When from==0 we are copying kernel space for the first
  * fork(). Then we DONT want to copy a full page-directory entry, as
  * that would lead to some serious memory waste - we just copy the
  * first 160 pages - 640kB. Even that is more than we need, but it
  * doesn't take any more memory - we don't copy-on-write in the low
  * 1 Mb-range, so the pages can be shared with the kernel. Thus the
  * special case for nr=xxxx.

uh huh

[thatskinnyguy]

I have a poster in my office with all the lines of code for kernel 0.0.1. It's uber-geek. A must have for most slashdotters!

Note the mention of GNU

[DrXym]

Yup, GNU Mach was well into development BEFORE Linux was even written. This is an example why open source projects are more effective when they're driven by pragmatism and not politics.

Re:Note the mention of GNU

[sayfawa]

Maybe if the GNU folks had only been working on a kernel instead of also doing the hundreds of other programs as well, they would have made more headway with HURD. And if Linus had been trying to do a whole OS and not just the kernel, Linux the kernel would still be early in development.

The mention of GNU should merely point out how important the GNU is in GNU/Linux. As Linus said in the post: Sadly, a kernel by itself gets you nowhere. To get a working system you need a shell, compilers, a library etc. These are separate parts and may be under a stricter (or even looser) copyright. Most of the tools used with linux are GNU software and are under the GNU copyleft.

Re:Note the mention of GNU

[Just Some Guy]

This is an example why open source projects are more effective when they're driven by pragmatism and not politics.

The problem is that politics is interested in you even if you're not interested in it. The pragmatic approach involves taking politics into account even if you're personally bored to tears by the subject.

Well

[akkarin]

"Oh, Linux will *never* have a fan base. Never. Ever. Period."
- Bill Gates

Re:History of GCC

[gr8_phk]

I use gcc but I'm not so sure its history would be that interesting to me.

I'm no historian, but I think the origin of GCC is very interesting. When RMS started the GNU project, he was not a compiler guy. He had written system tools, but never a compiler. He recognized that any free operating system (GNU in particular) would require a free compiler and that this was a requirement, not an option. So he sat down and wrote the first GCC himself. It wasn't something he wanted to do, but it was fundamental to his vision so he did it anyway. That speaks to his insight, ability, and dedication. Most important was the insight. Imagine where GNU/Linux would be today if it had to rely on commercial compilers. Yep, that's right - it wouldn't exist.

Also, there is much discussion about GCC transitioning to GPLv3 license. Apparently once the 4.3 branch is released, 4.2 will no longer be maintained under GPLv2. I believe this is because the FSF knows the compiler is fundamental and the license change is so important they don't really want patches contributing to the version under the old license.

More than just a kernel

[sepluv]

just as a preemptive strike against all the Linus-entirely-made-the-OS-himself trolls before they come out of the woodwork, here is the last paragraph of section 2 from the announcement in the TFA. Torvalds says it better than I could:

Sadly, a kernel by itself gets you nowhere. To get a working system you need a shell, compilers, a library etc. These are separate parts and may be under a stricter (or even looser) copyright. Most of the tools used with linux are GNU software and are under the GNU copyleft. These tools aren't in the distribution - ask me (or GNU) for more info.

Re:How would one build this?

[squiggleslash]

It got quiet a bit of interest, but was hampered by two major issues initially.

First, it wasn't free software. It was shared source, you could obtain the source after buying a copy and share that source with others who had also bought a copy, but you couldn't just modify it and pass on your modifications to anyone who wanted. That extra step of "Receiver must already have a license" was an issue, and reduced the number of experimenters and tinkerers drastically. GNU/Linux has achieved much of its popularity through the ability of virtually anyone who has a copy to pass on that copy to others, with freely downloadable LiveCDs and other ways to be exposed to it with little commitment on your part.

The other was that it was (usually, at least in x86) 16 bit. Applications generally ran in 64k memory spaces (albeit different spaces for code and data.) This severely limited the available functionality.

Linus, in part, wrote the first Linux kernel to try to overcome the second issue. By using Linux + the GNU toolchain instead of Minix, you had a full blown 32 bit operating system. Things like the X11 Windowing System suddenly became possible. His eventual adoption of the GPL also gave Linux users the freedom needed to ensure they could build a much bigger community around that kernel than Minix was able to achieve.

Today, Minix version 3 is available as free software, and in 32 bit form, but it happened too late to stop the GNU/Linux juggernaut from rolling right over it.

Embarassing change of context

[Anonymous Coward]

Am I the only one twisted enough to have fixed for a moment on the once perfectly innocent sentence : "Could someone please try to finger me from overseas?"... :)

Ready?

[bollucks]

...and it was almost ready for the desktop.

Ah, the good old days...accounting

[wandazulu]

Because I was a computer-room assistant back in college I got a couple of Unix accounts (that's what they were called) to learn and possibly help the grad students who were doing all the "cool" stuff on them (as opposed to showing a freshman how to print from WordPerfect 5.1 for DOS (F7 baby!)). The problem with the Unix machines (a SunOS and an Ultrix box) was that they both had accounting turned on and so I had $5000 of computer time to use until I had to go back and ask for more, which they actually gave only begrudgingly. I guess some departments really paid some $$$ for access.

Anyway, along comes Linux (not .01 but some very pre-1.0 version) and somebody else put it on a Gateway2000 486 machine) and all of a sudden I had, along with all the other assistants, a Unix-like machine we could call our own, do whatever we want and not worry about screwing up the "real" work being done. So when it came to learning how the Unix-world worked, I learned far more on that early Linux box than either SunOS or Ultrix if only because I didn't have to deal all the accounting stuff.

The funny thing is that I remember feeling that the Linux box responded better than the Sun machine or the VAX in that it seemed to handle more users better (though I suppose on the Linux box we were just mucking about with standard commands instead of doing heavy-duty work).

Re:How would one build this?

[LWATCDR]

Well I think that is a bit unfair.
Minix and Linux where different proposes in mind from the start. I would consider them be both be highly successful.
Minix was included with a text book. Minix was written to teach students how a Unix like OS worked so they could learn how write operating system code! Minix was very portable, clearly written, and would even run on an 8088 and 68000. It's technical limitations where just a logical trade off.
Requiring people to own the book to have the OS was probably a mistake but My guess is that the author wanted to prevent people from reselling Minix. Plus he really wanted people to buy his book.
Linux was some guy that wanted to write a free Unix Kernel for his 386 and he didn't care if it worked on anything but a 386 or frankly anything but his own computer.
Frankly at that time I and everybody else was waiting for the real free UNIX that the GNU project was going to write. The future was going to be GNU Unix and it was going to be a state of the art micro kernel based UNIX like OS. Of course the future doesn't really feel obligated to follow our plans.
Minix was a brilliant success. How many of the early Linux Kernel developers read Operating Systems: Design and Implementation by Andrew S. Tanenbaum?
I would say that Minix it did it's job very well.

Now Minix3 is a very new project. Frankly I find it very interesting. It is micro kernel and it runs drivers in user space. The goal is to create self healing OS. AKA a driver crashing will not take out the OS.
It uses BSD instead of the GPL which I am beginning to favor because of what I consider the bad spirited anti-Tivo clauses in GPLV3.
I really hope that Minix3 does get the attention that it deserves. Just as I hope the OpenBSD, FreeBSD, NetBSD, DragonflyBSD, and Linux continue to grow and thrive.