Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: [News] Linux Kernel Approaching Perfection

Roy Schestowitz wrote:

> Is There Perfection in The Linux Kernel?
> ,----[ Quote ]
> | In a perfect world, you could compile a brand-new Linux kernel
> | without the need for much configuration and without error.
> |
> | According to Linus Torvalds, the new 2.6.19 Linux kernel is such
> | an entity.
> |
> | "It's one of those rare "perfect" kernels," Torvalds wrote in a
> | Linux kernel mailing list posting announcing the new kernel.
> `----
> http://www.internetnews.com/article.php/3646456

You know I think he is probably right. I'm surrounded by 2.6.16's round here
but actually I couldn't tell you the last time I had a kernel crash with
previous kernels.

Of cause applications have crashed, but never the kernel and I am not just
talking about a couple of years, I mean many many years. In fact almost at
the beginning when we first used Linux as nothing more than a test bed for
code/scripts before using them on a mainframe or live server, even then
kernel crashes were very rare.

Then when Linux was being used as a live server, once set up it could sit
there for years, there would be drive changes, some application problems,
but still the kernel would plod along like marvin the paranoid robot, as if
everything you threw at it was too meneal to build up a sweat.

It is also many years since I compiled a kernel, except one experimental for
sf, but for my own or company server/clients I haven't compiled one for

There was that one release wasn't there, was at the beginning of the 2.x's
when there was a bad release. But that wasn't a problem for long, we just
booted the previous and waited for the one after.

I know that some groups are wanting to make major changes to the Linux
kernel, and they can do that provided they make it clear that it is a new
alternate thread in the kernel development, but the main kernel I believe
is sacred ground, it needs to be carefully controlled.

[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index