DragonFly kernel List (threaded) for 2004-04
Well, I think this is a little harsh isn't it?
Everyone has a time in which they must learn. I am sure the author of
the paper has gained insights and experience that will allow him to
write better articles in the future.
Just be thankful he did not choose white on bright blue, or some other
colour scheme that blinds the eyes ;-)
Benchmarks are kind of a grey area -- their relevance is questionable
because changes in hardware (among other things) often lead to new ways
of doing things that obsolete old algorithms. Unless you are developing
for a static system (ie: commodore 64) benchmarks can only be so useful
-- tomorrow we might not have hard drives or <insert your fav technology
here>, which would effectively change the rules of the game.
Randy Bush wrote:
this is not computer science, it must be a newspaper, and not
a very good one at that