On 6/12/07, Charlie Clark <charlie@xxxxxxxxxxxxxx> wrote:
With comments like that I can understand why the poster wished to remain anonymous!
Hey, hey. Nobody asked me for anonymity. *I* guessed that it would be better not to mention names because *I* didn't ask for permission to quote anyone. If anyone here is at fault, it's *ME*.
Linux's scheduling and desire to be all things to all people is one of its bigger problems.
Hey, don't get me wrong. I think the energy in the Linux development world is wonderful! And slowly but surely it will approach a very balanced compromise between all things people want Linux to do. Even if such compromise is simply extreme tailoring (way beyond what is currently possible) at compile time with the right config options. I'm only wary of people believing Linux has "Solved Every Problem Right Now", and that everything that they come up with is a "Fresh New Look at the Issues that Plague the World Since the Big Bang". ;)
Thank you André for a very informative paper. We don't have a heap of coders out there ready, willing and desperate to reinvent the wheel at every passing programming fad so we need research based development.
?
+1 for Minix +1 for Tanenbaum +1 for microkernel
You DO realise however that Tanenbaum only implemented a very simple and linear scheduling algorithm for Linux, right? :) He did however cite Waldspurger's paper. And while I ended up finding it again independently, if I read the damn bib references I'd save myself quite some time, too. The difference is that Ingo Molnár is a *key* Linux developer for quite some time and he's a RedHat employee; I'm just unemployed Joe Nobody finishing college!
Charlie
See ya, A.