minix

Not to perpetuate the great flamewars of the early '90s, but if you read carefully, Tanenbaum [wa|i]s wrong about most everything (that "microkernels are better" nonsense, that linux shouldn't tie itself to the soon to be obsolete 80386 processor (which makes him misinformed, as well), or that GNU/Hurd is a sane example of good practice), and since he started a discussion about linux on a minix newsgroup, I think that makes him a troll as well. Albeit a softly-spoken, polite troll.
 
fronclynne said:
Not to perpetuate the great flamewars of the early '90s, but if you read carefully, Tanenbaum [wa|i]s wrong about most everything (that "microkernels are better" nonsense, that linux shouldn't tie itself to the soon to be obsolete 80386 processor (which makes him misinformed, as well), or that GNU/Hurd is a sane example of good practice), and since he started a discussion about linux on a minix newsgroup, I think that makes him a troll as well. Albeit a softly-spoken, polite troll.

Tanenbaum isn't wrong, but he didn't know back then that this creepy x86-architecture from the 70s would survive. Furthermore GNU/Hurd isn't an example for the failure of microkernels but for mismanagement in this community. Nowadays we see the failure of Linux kernel: it's fat and full of regressions. There are by the way examples for "hybrid-kernels" like NT, DragonFlyBSD, XNU in MacOS etc. or "pure microkernel" like QNX.

http://www.theregister.co.uk/2009/09/22/linus_torvalds_linux_bloated_huge/

Linus Torvalds, 2009

If the kernel _is_ slowly deteriorating then this won't become readily apparent until it has been happening for a number of years. By that stage there will be so much work to do to get us back to an acceptable level that it will take a huge effort. And it will take a long time after that for the kernel to get its reputation back.

http://lwn.net/Articles/257828/

Andrew Morton, 2007

So I don't think Tanenbaum is wrong, but he has to adjust some of his theories in praxis. Whereas Torvalds just did it for fun, today we call this fun Linux.
 
*takes the bait*

It's the difference between this and this. And you can go around all day about how the first is right and better and more standards compliant, but you'll still be talking about it while I'm riding the second back from having a couple of beers at the bar, twenty years from now.
 
Both micro and monolithic have their pros and cons.
That's probably why Apple and Microsoft are trying to get the best of both worlds by creating a hybrid, or bastard spawn of satan, depending on your view :e
 
fronclynne said:
*takes the bait*

It's the difference between this and this. And you can go around all day about how the first is right and better and more standards compliant, but you'll still be talking about it while I'm riding the second back from having a couple of beers at the bar, twenty years from now.

I didn't say microkernel is the best of the best of the best, it just has it points and I don't count the questionable "success" of Linux as evidence for the quality of a monolithic kernel. If I do so, taking Linux as example per se, then I proof rather the difference, as said above ;-)


http://www.mail-archive.com/freebsd-advocacy@freebsd.org/msg01530.html
 
I've used minix a few times and I like it. then again, I like about anything computer related that the normal person finds strange or complex.

Reading the old posts and mailing lists does no more for me then say,"Huh, I'll post a reply and see what happens."
 
I think that, above all, Tanenbaum's message is a lesson in how difficult it is to predict the future, no matter how smart and well informed you may be.

What he predicted was perfectly reasonable in 1992.
 
Carpetsmoker said:
I think that, above all, Tanenbaum's message is a lesson in how difficult it is to predict the future, no matter how smart and well informed you may be.

What he predicted was perfectly reasonable in 1992.

Tanebaum's message was based on purely scientific and technological considerations. He was right about everything. However, he was wrong that technology is the most important component of high-tech business. It as actually probably the least important. The most important components are initial price and PR bull-shit.
 
SirDice said:
Both micro and monolithic have their pros and cons.
That's probably why Apple and Microsoft are trying to get the best of both worlds by creating a hybrid, or bastard spawn of satan, depending on your view :e

Dont be so hard on microsoft, because of them there has been some great innovations.
For instance Windows for years could not truly multitask despite all their PR attempts (to redefine what true multitasking is). So anyways a few years back the boys from MS went to their friends at Intel and asked if they could design a multitaking CPU for them -- and viola, we got hyperthreading and then multicore cpu's.
MS wins the ability to claim they have multitasking (albeit with certain "conditions" - never "limitations", always "conditions" OIOW still not really true native mt), ...
... but the other winners were the 'nix-en (i.e. us), which rolled into multicore cpu's with about as much trouble as a new type of hard disk.

In a simpler examples because windows is such a hog it's been a great driver for faster machines, more onboard memory, bigger capacity and faster hard disks (and even death of the floppy).
 
Yes, let's praise the ingeniousness of Microsoft's marketing and PR departments.
<praise_start>Clap, clap, clap, whistles, bravo!!!</praise_end>


Intel designed hardware context switching (using the TSS) a long time ago. But no one really used it, and it became one of these many useless features.
Most operating systems only use(d) one 1 TSS and implement(ed) software context switching.

Preemptive multitasking existed under UNIX® when Bill Gates was still perfecting his HelloWorld! skills in BASIC.
Early versions of Windows used the less-than-perfect cooperative multitasking.

Hardware multithreading was first researched and developed around the time Intel was incorporated, and existed in its modern form in quite a few architectures other than Intel's.

Even most of the user interface was copied by Microsoft from other companies such as Xerox and Apple, e.g. the "Recycle Bin"/Trash, the desktop metaphor, icons, drop-down menus, drag-drop, etc.
 
>Even most of the user interface was copied by Microsoft from other companies such as Xerox and Apple

Different story: both of them copied by Xerox at the same time. If you don't listen to the MS hype, please don't listen to the Apple hype too ;-)
 
Back
Top