Who wrote 3.0 - from two points of view

Posted Jul 15, 2011 19:28 UTC (Fri) by Julie (guest, #66693)
In reply to: Who wrote 3.0 - from two points of view by mlankhorst
Parent article: Who wrote 3.0 - from two points of view

the thing that probably is lacking is the background knowledge

I'm a bit confused by your comment, I had taken Jon's remark about it becoming difficult for new volunteer developers to participate as meaning the same as the above (newcomers don't _have_ that knowledge). Along with a couple of other potential difficulties:

1. Volunteers necessarily have less time than full-timers and therefore anything they swot up on at the time is likely to be dated when they come to try to put it into practice due to the current fast rate of development

2. The kernel's pretty much saturated with regards to experienced (sometimes full-time, paid) devs (at least in the areas of core work; I know Greg K-H is always looking for more contributors to staging).

What did you mean?

Following lwn for a few years helps a lot to learn about that. ;)

Well, I can't agree with you here - I think that following LWN for a few years has become pretty _essential_ for learning about that ;-)

Who wrote 3.0 - from two points of view

Posted Jul 15, 2011 20:07 UTC (Fri) by mlankhorst (subscriber, #52260) [Link]

With background knowledge I meant all the things you have to worry about in kernel space that might not happen in userspace. Atomics, interrupt contexts, rcu, the fact that not the whole world is a VAX^Wx86.

When writing a bugfix for EFI I had to deal with early memory allocators and the transition to the full page allocator. For fun I looked up what allocators were used in the kernel. I came up with brk, e820, memblock, bootmem, xen's special brew, and the kernel's full page allocator. This list is probably incomplete. ;)