Old School Linux Remembered, Parts 0.02 & 0.03

eldavojohn writes "Following our last history lesson of Linux 0.01, the Kernel Trap is talking about the following announcements that would lead to one of the greatest operating systems today. A great Linus quote on release 0.02 (just 19 days after 0.01): 'I can (well, almost) hear you asking yourselves "why?". Hurd will be out in a year (or two, or next month, who knows), and I've already got minix. This is a program for hackers by a hacker. I've enjoyed [sic] doing it, and somebody might enjoy looking at it and even modifying it for their own needs. It is still small enough to understand, use and modify, and I'm looking forward to any comments you might have.'"

Preservation

[Enderandrew]

I'm glad simply for the sake of history and preservation they're making these articles. I read the LKML frequently even if I don't fully understand the mechanics for the how and why the kernel operate, but I like to pretend that I do. I find this stuff rather fascinating. It is also interesting to wonder how Linux became what it is today considering its roots.

Linux today is a child of countless contributers, but it is still tied in name and perception very much to one man. I wonder if people think this is a good thing. I've often maintained that Linus is terse, but I've enjoyed that about him. If he rips into someone, I chuckle. But after this latest fiasco with Con and the schedulers, I'm wondering if this is a bad thing.

Still in development

[jesterzog]

At least, it looks as if the Change Log [gnu.org] is still being updated. (Click the link titled "ChangeLog in the main directory".)

terse?

[QuantumG]

that's an understatement [gmane.org] :)

In other words, I'm right. I'm always right, but sometimes I'm more right
than other times. And dammit, when I say "files don't matter", I'm really
really Right(tm).

Which is actually more funny than arrogant, so long as you know Linus' style.

Great quote by Linus

[schmiddy]

I really enjoyed seeing the quote by Linus (this is a program for hackers by a hacker). He clearly never, ever, expected his little hobby project to catch on the way it did. Hope this gives hope and inspiration to all the OSS developers out there, scratching their own itches. Just looking back on the history of the software industry, it seems like so many tremendous ideas and businesses got started around a small hobby project by one or two smart guys: Google, Perl, Python, Linux, GNU, and so on. Remember that one man change history.

Re:It's too bad about teh Lunix

[kcbanner]

I can tell your inexperienced because you said "program it right." No. The proper word is "code".

"...they never bothered to code it right the first time around."

And imho its coded very well indeed.

Re:The Linux alternate history game...

[AKAImBatman]

I think it's pretty clear that it's only the pace at which hardware has advanced in the last decade or so that has allowed Linux to continue monolithically.

What'choo talking 'bout Willis? Over the past couple of years, Linux has been slowly evolving toward a hybrid kernel design. Between the common use of FUSE [sourceforge.net] for powerful new file systems and the recent merging of user space driver support [slashdot.org] into the kernel, Linux is showing more and more Microkernel attributes every day.

In a sense, Tanenbaum wasn't really wrong. It's just that like most researchers, he was ahead of his time. Facets of Microkernel technology have made their way into nearly every major operating system on the market today. From Windows to Mac OS X to Linux, hybrid kernel design is proving to be a valuable feature that every moden operating system should have.

When it comes down to it, microkernels just make sense. It's in many ways simpler to develop than a monolithic kernel, and provides an easy-to-implement yet powerful firewall between the computer's subsystems. The catch is that early reseach ran into performance problems inherent in task switching on every system call. Hybrid kernels attempt to minimize that by designing around the monolithic "kernel space" vs. "user space" division already present in most OSes. Because the division already exists, the performance hit can be quite minimal for certain forms of application. (I haven't kept track to know if such performance has actually been achieved in any Linux hybrid code, so take a grain of salt with this.) Pure device drivers would still have performance problems due to the data bubbling up from the kernel rather than executing entirely in kernel space. Thus hybrid features are more useful for subsystems that already interact with userspace. (e.g. A new filesystem.)

Re:"Greatest operating systems today"?

[Urza9814]

Oh yea. 1.0, 3.1, 95, 98, 2000, ME, XP, Vista...real coherent scheme there. Goes from small decimal numbers to large two digit numbers to four digit numbers to two letters to words.

As for Linux being unstable...dude, have you ever even seriously used Linux? Hell, if you don't fuck with it, it'll run for YEARS. Funny story - Freenet, with default configuration, crashes Windows XP on my computer in under half an hour. Not even kidding. Just murders the bitch. Now take the exact same program (It's Java), put it on Linux, runs fine. Runs for weeks without a problem. Remove the bandwidth caps, change max allowed connections from 200 to 750, increase the data store size by 40 gigs, and remove the limits on allowed known routing nodes...and it still runs fine. for weeks. Windoze would die in under 10 minutes from the load my Linux box idles on.

Re:The Linux alternate history game...

[ozmanjusri]

It seems to me that had "real UNIX" been available for low-end systems in the early 90s

Amiga Unix was available in 1990, a time when Amigas were still selling well. Despite being one of the better Unixes of the time, it didn't set the marketplace alight.

Linux in the early 90's

[jbolden]

I was a Unix user in the early 90's and following the 386BSD saga closely. It was much more well known among the Unix community, but no one was getting it to work. The home unix of choice is one that almost never gets mentioned Coherent [wikipedia.org], and business Unix was Xenix and later SCO. The Linux community had a focus on working with existing hardware and a focus on being usable by non Unix people since the days of the Corsair project [wikipedia.org].

The idea that the 9 months made the difference is simple BS. Much as the FreeBSD people like to claim otherwise it was strategic choices made by the BSD camp all throughout the 90s (like focusing on reliability over functionality) that drove Linux's popularity.