OSNews: http://www.osnews.com/story/22246/Linux_2_6_32-Rc1_Kernel_Released Exploring the Future of Computing en-us Copyright 2001-2012, David Adams adam+nospam@osnews.com Sat, 11 Feb 2012 10:30:18 GMT http://www.osnews.com/images/osnews.gif OSNews.com http://www.osnews.com Well, according to recent Linus' comments... http://www.osnews.com/thread?386638 http://www.osnews.com/thread?386638 ...won't that makes Linux even more bloated and ugly ? Shouldn't be a driver be... a driver, and not a kernel module ? Linus was bitching with Andrew about micro-kernel vs. monolithic kernel, it has then be proven the superiority of Andrew's solution : more stable, easier to upgrade and maintain, just more elegant. And all of this for what, 2% CPU slowdown ("the cost of the -ugly?- meat bags") Frankly, with current overpowered multi-cores CPU, I think current software can handle, even Microsoft is forwarding in that path !

Kochise Mon, 28 Sep 2009 19:11:00 GMT donotreply@osnews.com (Kochise) Comments RE: Well, according to recent Linus' comments... http://www.osnews.com/thread?386641 http://www.osnews.com/thread?386641 It is true, but Linux is modular now and the reason is to leave out all the bloat. However a monolithic kernel is harder to mantain compared to a microkernel.

In regards to performance, theoretically microkernel are slower, but L4 has demonstrated that it is possible to build a slim, fast, reactive microkernel with very low latencies.

Qnx is the same beast too. Mon, 28 Sep 2009 19:32:00 GMT donotreply@osnews.com (pabloski) Comments RE: Well, according to recent Linus' comments... http://www.osnews.com/thread?386672 http://www.osnews.com/thread?386672
...won't that makes Linux even more bloated and ugly ? Shouldn't be a driver be... a driver, and not a kernel module ? Linus was bitching with Andrew about micro-kernel vs. monolithic kernel, it has then be proven the superiority of Andrew's solution : more stable, easier to upgrade and maintain, just more elegant. And all of this for what, 2% CPU slowdown ("the cost of the -ugly?- meat bags") Frankly, with current overpowered multi-cores CPU, I think current software can handle, even Microsoft is forwarding in that path !


What on earth are you going on about? modules ARE drivers - and everything that can be a module is normally compiled as a module by the distributor. It has nothing to do with micro versus monolithic - the slow down is due to a natural increase in features required in a modern operating system - something that is unavoidable.

The question that should be asked is whether there these slow downs are unwarranted - that the features delivered cannot justify the slow down or increase in memory usage. So far I haven't seen that yet. Mon, 28 Sep 2009 21:40:00 GMT donotreply@osnews.com (kaiwai) Comments Main development phase of Linux 2.6.32 completed http://www.osnews.com/thread?386702 http://www.osnews.com/thread?386702 http://www.h-online.com/open/Kernel-Log-Main-development-phase-of-L...

With the first release candidate of Linux 2.6.32, last night, Linus Torvalds completed the main development phase of the next version of Linux on the main development branch. As the kernel hackers already integrate most of a new kernel version's major changes into the source code management system during this phase, called the merge window, 2.6.32-rc1 is already a good indicator of the most important new features due for release with Linux 2.6.32 in early December.


The linked article has some discussion of what is likely to be included, and what might be excluded, from the Linux kernel 2.6.32 release.

The release will likely not include ALL of the code which has been taken from the staging area and which is included in this release candidate. There is some "culling" to be done yet. Tue, 29 Sep 2009 00:15:00 GMT donotreply@osnews.com (lemur2) Comments