Fedora 9 'git' source kernel ??

Patrick O'Callaghan pocallaghan at gmail.com
Mon Jul 14 22:58:53 UTC 2008

On Mon, 2008-07-14 at 17:38 -0400, William Case wrote:
> Yes, I have been down that road with others.  It seems to be a bit of
> academic religious proselytizing.  You are a lowly student and
> therefore
> could never understand unless you devotedly sit at my feet and study
> for
> years.   

Then you misunderstand me. I have no desire to put you off, rather to
convey that this is exciting stuff.

> I have no expectation of starting at the top and working down.  But
> the
> kernel must have entry points and exit points.  Examining where the
> major services start and end must have some value.

Well they mostly start and *don't* end unless something bad happens.
That's one of the things that distinguishes kernels from the more
familiar "read input; process; produce output" pattern of conventional
software. Of course kernels also read input and produce output, but the
model is on the whole not all that helpful. One of the main things a
kernel does is juggle a zillion things happening at once so they don't
interfere with each other (but can still communicate when required).

To some extent a lot of application software is getting that way too
nowadays. The difference is that apps can rely on a well-defined and
semantically rich base of libraries to handle events and keep all the
balls in the air, but the kernel has to do the heavy lifting itself. 

> Besides most
> learning is iterative -- one starts somewhere, understands a bit, and
> keeps going around until they are back at the start ready add more.

Naturally. As Pratchett says "give a man a fire and you warm him for a
day; set him on fire and you warm him for the rest of his life". Sorry,
wandering again ...


More information about the fedora-list mailing list