Re: kernel-{image,headers} package bloat
On Mon, Apr 23, 2001 at 06:59:55AM +0800, zhaoway wrote:
> > this actually *helps* new users.
>
> Why?
>
> I can see your suggestion help mirrors (The trade-off is
> questionable.) But I can't see why it could help users.
because new users are better off when they learn how to control their
system.
ignorance is not bliss. ignorance is a crippling affliction, especially
when it comes to computers.
anything that promotes ignorance by hiding the "messy details" of
operating a computer is fundamentally wrong.
> I certainly don't think that make users (even make the task easier) to
> compile a kernel can do help to most of the users
nobody is making them do anything. it's only if they need the slightly
enhanced performance of a kernel optimised for their exact CPU that they
need to compile a kernel.
yes, it's better for every system to have a custom compiled kernel, but
a plain 386 kernel will suffice for most people. the CPU optimisations
for 486, k6, k7, p3, p4 and so on just aren't worth taking 5 times the
storage space and 5 times the number of packages per kernel...if a user
wants that negligible performance gain then they should compile the
kernel themselves.
the only excuse i can see for having all those kernel-image packages is
to cater to the idiots in the slashdot crowd who say "cool, distribution
X rocks because it has kernel x.y.z" and completely fail to understand
that any kernel can be run with any distribution as long as the support
tools (modutils, etc) are up to date.
is that worth hundreds of MB in the archive, gigabytes of bandwidth to
sync mirrors, and an extra CD in the release? no, not at all. no way.
> (That is *not* a task they should be bothered at all.)
if they don't want to do it, they don't have to.
if they can't be bothered, they're not even going to notice the difference
between a 386 kernel and a k7 kernel.
craig
--
craig sanders <cas@taz.net.au>
GnuPG Key: 1024D/CD5626F0
Key fingerprint: 9674 7EE2 4AC6 F5EF 3C57 52C3 EC32 6810 CD56 26F0
Reply to: