Observational humor

[mark-OGhnF3Lt4opAfugRpC6u6w at public.gmane.org (Mark J. Dulcey)]

John Chambers wrote:
> markw wrote:
> | I'm compiling a custom Linux kernel.  Sort of walking down memory lane as
> | it were. Can someone please explain to me how doing this in 2009, with a
> | very fast dual core CPU and very fast SATA disks takes LONGER than it did
> | back in 1999 (10 years ago)? (rhetorical question)
> |
> | I mean come on? How much faster are systems today than 10 years ago? Has
> | the Linux kernel REALLY grown that much?

The limiting factor on kernel compilation is typically disk speed, and 
that very fast SATA disk drive isn't all that much faster than the one 
you had 10 years ago in RANDOM ACCESS SPEED -- likely less than a factor 
of two faster. The improvement in serial read speed is much larger, 
probably a factor of five, but compiling the kernel doesn't stress that. 
The growth of the kernel code has more than absorbed that improvement.