[haiku-development] Re: To become haiku to real micro-kernel ?
- From: Adrien Destugues <pulkomandy@xxxxxxxxxxxxx>
- To: haiku-development@xxxxxxxxxxxxx
- Date: Sat, 17 Mar 2018 05:26:30 +0100
On Fri, Mar 16, 2018 at 09:21:24PM +0000, Alexander von Gluck IV wrote:
March 16, 2018 3:41 PM, "aikizen" <aikizen@xxxxxx> wrote:
Hi,
I would like know if you thinking to become haiku kernel to a real
micro kernel like minix3 ?
I like very the principles of the u-kernel, for the little that i know.
Any else, i learn C++ by the book thinking C++, you think is the good
way to learn C++ for help haiku ? By what can i help for know better
the code and the structure of haiku, by UML schema or else.
Every once and a while people bring stuff like this up.
Haiku is a hybrid, modular kernel.
Haiku's kernel is stable, has a large selection of drivers,
and redesigning it would only delay R1 further without any
real benefits.
To word it differently: a microkernel would be nice, but also something
that would take a LOT of time to get done (look at the Hurd project if
you don't agree). And we already have a working kernel.
So, if someone sends us patches that convert our kernel to a
microkernel, we would probably consider them (as long as they don't
break any features). However:
- Working on the kernel side is a level of difficulty above working in
userland
- We have more important things to work on first
- We don't need ideas, we need people implementing them
If you want to start contributing to Haiku, the best thing to do is to
solve existing bugs, not create new ones :) It's too easy to dream about
"what would happen if..." and get distracted. Focus on fixing the bugs,
this is what matters.
--
Adrien.
Other related posts:
