View Full Version : Linux Kernel Future

August 26th, 2009, 05:03 AM
How long will the Linux kernel continue to be up to date with current technology? What I mean is, how long until there is a breakthrough in OS design drastic enough that it can't be incorporated in another kernel update, but that offers enough benefits that it becomes worth switching to? Surely the Linux kernel will just simply become outdated eventually; will the whole free-software thing start over again then?

Sorry for all the questions, I'm simply curious :)

August 26th, 2009, 05:08 AM
Actually I think the Linux kernel has a brighter future then any other kernel, as it can be patched and modified to meet up with new standards.
The linux kernel is very diverse.

August 26th, 2009, 05:09 AM
I don't think the question is answerable. we could speculate a lot though :)

August 26th, 2009, 06:53 AM
If Linux Kernel can't, no one can...

August 26th, 2009, 07:14 AM
Was an interesting article on Ars Technical a while ago. I don't think Linux will be left behind any-time soon, I think its the priority kernels that are in trouble.

Linky (ArsTechnical) (http://arstechnica.com/open-source/news/2009/08/linux-dev-community-is-growing-5-patches-accepted-per-hour.ars)

August 26th, 2009, 07:20 AM
There are things in OS design that would be hard to do and preserve the notion we have of Linux today such as providing a fully managed kernel akin to Singularity. We will very likely though need to go this way if we want truly secure and reliable computing.

I think Linux will stick around for a long time but there are things we just can't do based on it's design choices. Not without a major rewrite and retooling at least.

August 26th, 2009, 08:36 AM
To add another cornerstone to this conversation:

What technologies could be implemented into such a kernel that makes it so overwhelming to the existing kernels?

August 26th, 2009, 10:31 AM
new type of CPU/architecture

August 26th, 2009, 11:09 AM
you can swap out the linux kernel in ubuntu/debian for other kernels.

if the windows kernel was open source i am sure someone would build debian GNU/NT

August 26th, 2009, 11:31 AM
new type of CPU/architecture

Linux runs on basically everything, a new arch shouldn't be a problem;)

August 27th, 2009, 12:22 AM
There are already some interesting technical innovations in computing that don't really fit with Linux's model. However, these things aren't found in PCs, and they won't be any time soon. It's hard to introduce them, because the software has to be ready for them. And the limiting factor there is Windows, which is far more brittle and far less adaptable than Linux. So, no worries.

Historically, disruptive changes have actually favored Linux. When Intel and AMD came up with 64-bit architectures, Linux was among the first OSes to take advantage of them (years before Windows). Linux found itself on netbooks before Windows, because it could adapt to their limited resources, and Linux now occupies many other devices where Windows can't go.

August 27th, 2009, 03:02 AM
@ O.P.: With respect, I think your question is a bit over-broad in its scope or you are making conscious or unconscious assumptions in it.

The Linux kernel is a community effort at this point, and has been for a very, very long time. People who do development work for Linux in general and the kernel in particular are very conscious of the changes and improvements in technology going on around us. Odds are they probably have good coping mechanisms at this point for various potential changes which could come down the pike, and given the brain power we're talking about, it's likely they could come up with a game plan pretty darned quick if something completely unexpected and unforeseen suddenly came out of left field.

As others have said, it's not just Linux but all the other operating systems which would also have to deal with the new technology. This isn't F/OSS's fight alone; rather, this would be the struggle of every OS maker period.

Frankly, I think you need to be a bit more specific in what you're asking, and then maybe we can get kernel-specific knowledgeable people to answer your question, whatever it may then be.

Don't sell the kernel devs short here -- heck, don't sell any of the major contributors and others short, either -- considering the conditions they work under and considering for many that we're talking about doing this part-time or in their spare time, Linux has faired rather well, considering. Without bragging, a lot of these folk represent some of the best, brightest and smartest people on the planet. I have a LOT of faith in them, and frankly you should, too.

The sun does not rise from, nor does it set on Redmond alone.