steve szmidt wrote:
> On Saturday 25 February 2006 19:44, Paul M Foster wrote:
>
>>michael hast wrote:
>
>
>>>is there any real advantage to 2.6 over 2.4? I know that some people
>>>have made ths kind of thing work on 2.6, but if there isn't a good
>>>reason to use it, I don't see the point of the effort. Thoughts?
>>
>>I think Eben best answered the overall question, but I wanted to make a
>>comment about the superiority of 2.6 over 2.4 kernels. The primary
>>benefit you'll notice is speed. The memory subsystem was considerably
>>reworked between the 2.4 and 2.6 series. That said, 2.4 kernels are more
>>stable than 2.6 kernels overall, because the kernel development method
>>was changed after 2.4. Not to say that 2.6 kernels are unstable. But
>>Linus changed the way things were done, and it has resulted in more
>>radical changes to the 2.6 kernels.
>
>
> Linus made changes which for the user meant your keyboard and mouse were more
> responsive even under heavy load. He greatly changed memory management from
> his originally flawed design by incorporating uClinux into the mainstream
> kernel. uClinux is Linux for microcontrollers project.
>
> The second big change was to support NUMA servers which makes Linux work well
> in a multiprocessor world, by running a lot more efficient over many CPU's.
>
> Linux started a new concept called subarchitecture which allows the separation
> of processor type and underlying hardware. This resulted in support for more
> platforms and a saner way of dealing with porting Linux to new systems.
>
> Supporting hyper threading is another big improvement. And a big change in
> threading which allows more stable threading with multiple CPU's. Preemtive
> processing is yet another. You can now interrupt the kernel mid-task, which
> makes all processes run a bit more evenly and give the user a faster overall
> experience.
>
> Support for more RAM and better scalability to newer and faster CPU's. Under
> 2.4 you could only have 255 major devices defined due to the limitation
> of /dev. With 2.6 it now supportss 4095 major device types.
>
> Linux I/O has had major improvements. Resulting in networking being even
> faster under 2.6. Faster disk I/O.
>
> By handling devices with the Unified Device Model it's internal understanding
> of hardware makes it run a lot more reliably. This also resulted in much
> better support for hardware, which is visible in the improved support on for
> example laptops since 2.6 was released.
>
> 2.6 was really a huge overhaul that changed the efficiency under which 2.6
> operates with internal and external devices. The improved support for scaling
> to many CPU's gave us the improved user experience as things need to run
> "tighter" to not stumble on each other.
>
> The 2.6 is not any less stable, quite the opposit. 2.4 has been considered
> stable as it's old and proven. What Linus said early on was that even though
> 2.6 had only been out for a short while, and had not undergone the same
> amount of testing, it was already more stable than 2.4.
>
> In effect you could say that the changes made were something like improving a
> vehicle which was only reliable if you stayed with the block, to make it able
> to reliably run cross country. Which of course made it even better at running
> around the block.
>
> The 2.6 kernel got Linux up to be able to compete with the big boys. Linux had
> finally grown up.
While all of this may be true, it wasn't the point I was making. I don't
closely follow the kernel lists, but my understanding is that the 2.4
kernel (and prior) was implemented under a much more plodding,
methodical paradigm. There was a very specific methodology to the way
kernel development was shepherded and administered. With the 2.6 kernel,
Linus changed this so that kernel development could be considered more
"haphazard" now. That's not really the right word, as it implies more
randomness than actually exists. But kernel development did change away
from the more "orthogonal" way in which it was previously administered.
-- Paul M. Foster ----------------------------------------------------------------------- This list is provided as an unmoderated internet service by Networked Knowledge Systems (NKS). Views and opinions expressed in messages posted are those of the author and do not necessarily reflect the official policy or position of NKS or any of its employees.
This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 18:53:02 EDT