Re: any reasons to recompile kernel?

My question comes from reading "Building XNU" in OpenDarwin HOWTOs, 
http://www.opendarwin.org/en/articles/building_xnu/. It barely provides 
the steps of obtaining the tool chain and the basic steps of compiling 
the kernel but nothing on kernel configurations. So it is hard to see in 
what circumstances where a compile is warranted.

I do have a different but related question here. This is also related to 
the thread on "building xnu for osx", 
http://lists.apple.com/archives/darwin-development/2001/Apr/msg00322.html, 
hence, the cross-posting. My apologies if anyone do take offense in 
this. From that thread and the HOWTO, it does seems that it is possible 
to build a Darwin 8 kernel (corresponding to OSX 10.4) and replace the 
Darwin 7 kernel in OSX 10.3. Besides masochism or in the name of 
experimentation, I cannot think of a reason why someone will want to 
build the same kernel as in their working system and install it at great 
risk.

Having said that, had anyone actually done that (building and replacing 
a new kernel into an existing OSX installation)? This is commonplace in 
Linux, but in Darwin/OpenDarwin/OSX landscape?

If so, how will the system then behave... as 10.3 or 10.4 in terms of 
compilation platform?

Is the version of OSX determined mainly by the kernel or other tools?

A very specific question will be this: if I do not care about user 
add-ons like widgets etc etc, but I want to install Java 1.5 into my 
"OSX 10.3", is kernel compilation and installation of Darwin 8 a way to 
get there?

Cheers
Maurice


Kevin Van Vechten wrote:

> If you do not know why you need to recompile the kernel, then you do 
> not need to recompile the kernel.
>
> - Kevin
>
> On Jul 20, 2005, at 5:33 PM, Maurice Ling wrote:
>
>> I've been reading the OD FAQ. There seems to be compelling reasons 
>> not to re-compile a kernel as almost all modules are loaded 
>> dynamically. So, is there a reason to call for a compilation?
>

begin:vcard
fn:Maurice Ling
n:Ling;Maurice
org:The University of Melbourne;Department of Zoology
adr:;;Gate 12, Genetics Lane;Parkville;Victoria;3010;Australia
email;internet:[email protected]
title:Ph.D. Candidate
tel;cell:+61 4 22781753
x-mozilla-html:FALSE
url:http://www.geocities.com/beldin79/
version:2.1
end:vcard