Who will prevail?

[MedivhX]

Linux Kernel or HURD?

P.S. Yea I know it may be the stupid question (like most of my questions)...

[denad]

Linux all the way.

[Sonrep]

Hurd is a better idea, it's just that there are more developers working on linux :(

[Lord Illidan]

Linux ftw!

[DC@DR]

Linux of course!!!

[MedivhX]

Why, why, why??? Tell me whyyyyyyyyyyyyyyyyyyyyyy??!

[agurk]

...and the winner will meet opensolaris in the finals. ;)

[MedivhX]

...and the winner will meet opensolaris in the finals. ;)

LOL! Maybe it will!!! OpenSolaris has big chances!!! And maybe the winner will meet BSD, but I doubt it...

[Omnios]

WTF is HURD?

[Lord Illidan]

WTF is HURD?

HURD = HIRD of Unix (http://en.wikipedia.org/wiki/Unix)-Replacing Daemons (http://en.wikipedia.org/wiki/Daemon_%28computer_software%29)
HIRD = HURD of Interfaces Representing Depth

http://en.wikipedia.org/wiki/Hurd

[BWF89]

The HURD is nothing, and is always going to be nothing. It's a good idea in theory but it had too many bugs to work in real life.

[IYY]

The HURD is nothing, and is always going to be nothing. It's a good idea in theory but it had too many bugs to work in real life.

Isn't this exactly what they used to say about Linux in the 90's?

Anyway, I do agree that Linux is the way to go, and not HURD. Linux is licensed under the GPL, after all.

[SunnyRabbiera]

HURD has potential, but for now linux is here to stay.

[Kimm]

I believe in the HURD. Since the HURD is written to be extended, it will survive longer when the software idustry moves to new standards.
However, the HURD will have a rough time ahead of it, but as it grows more Linux compatable (perhaps using code from Linux, to use Linux drivers, and maby Linux binary compatability) and the bugs are worked out, I think the HURD will get a larger userbase, mainly because distros will se the advantages of basing their system on that kernel while Linux may have to unergo a complete rewrite to keep up with new standards.
However, we never know what happens when Linux 3 hits the market.

[MedivhX]

Yeah... It would be good for Linux to transform into microkernel or hybrid kernel (if it is possible).

[zanglang]

Hurd... when Duke Nukem Forever gets released. :P

OK, crappy slashdot vaporware jokes aside I think there still needs to be more research done on microkernel development before the majority of developers will seriously consider it. Few more years maybe?

[Erik Trybom]

WTF is HURD?
...And that pretty much answers the question.

Many say that the Hurd is a better approach, but Linus denies this and says part of the Hurd's problems are not despite but BECAUSE the way it's designed. I don't think there's even a point in raising the question until the Hurd is finished enough to be of any use.

[MedivhX]

Yes I think so too... It can't run ony any architecture except on IA32. i don't think that we should listen to Linus' opinions about other kernels, becaouse no one would say that some other people's creation is better than his own...

p.s. Aren't microkernels only for some special use (i.e. medicine, astronomy...)???

[ssam]

if hurd catches up with linux then distros could switch to a hurd kernel with hardly and change from a user point of view.

see http://www.debian.org/ports/hurd/

[shining]

Yes I think so too... It can't run ony any architecture except on IA32. i don't think that we should listen to Linus' opinions about other kernels, becaouse no one would say that some other people's creation is better than his own...


I don't think I will ever say that. I wish I could though :)

[EdThaSlayer]

Hurd is interesting(easier to develop and one "server" process can't break down everything) but Linux is just so much better since it has far more developers.

[MedivhX]

WroNg! HURD is harder to develop! Microkernels (i.e. Mach, L4, Coyotos, EROS) are harder to develop, and monolithic kernels (i.e. Linux) are easier to develop.

[Omnios]

Found a Hurd info site.

http://www.gnu.org/software/hurd/hurd.html

[EdThaSlayer]

WroNg! HURD is harder to develop! Microkernels (i.e. Mach, L4, Coyotos, EROS) are harder to develop, and monolithic kernels (i.e. Linux) are easier to develop.

But...Wikipedia can't be wrong right?


The Hurd instead consists of multiple servers working together. Instead of a single large code base which includes everything from handling the clock to handling the networking, in Hurd each of these is handled by a separate server. This makes developing the Hurd much easier (at least in theory) as making changes to one is less likely to have side-effects in others.

Only in theory ](*,)
I guess your right:(

[jvc26]

Linux :) (A personal thought, but I think not uncommon ;))
Il

[deanlinkous]

uh...well...this isn't blood-thirsty money mongering corprate products so I will vote BOTH, each in their own time....

[Guardian_Bob]

Sorry I will be jumping ship as soon as a stable version comes out but then again I do not care for unix. Though I would like gnome to work with it.







100% generic driven.

[Hex_Mandos]

Well, they've only spent around 17 years developing it, and look how far they've come! It supports an architecture 99% of users will never hear of and which is ignored by most Linux distributions, but they'll surely port it to some other obscure hardware someday soon. I'm salivating in anticipation, thinking of the time I'll be able to use the HURD in my desktop computer. Maybe by 2040...

(I'm sorry, I just went through a phase of Linus vindication...)

[Artemis3]

Multiple kernel choice will indeed come. Debian is one of the few distros getting ready for non linux choices. With options like hurd; or the kernels used in both netbsd and freebsd; and maybe soon the one in OpenSolaris; the user will choose what it works best as much as today users choose their preferred file system or desktop enviroment when they use linux.

I don't think its a matter of "prevailing". Its much more about what works best for you; as usual with anything in free software. This is not a race, and it doesn't matter if few or many use certain kernel. It will be developed as long as someone wants to. Free software can never go bankrupt and dissappear from the face of earth like proprietary software can. It doesn't even matter if hurd remains as an ongoing experiment, it might inspire somebody to make a fully working derivative from it. Just because someone invented a wheel, doesn't mean you can't invent you own. It doesn't matter what others say, you will always have the freedom to do so as long as you want.

PS: You might be interested in reading this: http://en.wikipedia.org/wiki/L4_microkernel_family
Many hurd problems are being solved by moving it from its Mach roots...