On Thu, 4 Sep 2008, Ichthyostega wrote: > throughput, i.e. the bare speed of the system. So, generally > it's a tradeoff: do you need maximum speed, or do you need > the system to reliably react within a certain timespan?
Another point worth emphasizing is that RT kernels are designed for *predictable* latency, possibly even more so than just *low* latency. The average non-RT kernel may be very fast to respond to a given system call 99 times out of 100, but then takes much longer 1% of the time. You can't have that if you're feeding a buffer for audio (or CD burning, or a few other activities); so you need a kernel that will have a fixed upper limit on how long it takes to respond to the call. Even if the upper limit is not such a great response time, being able to guarantee no MORE delay than the limit is useful in planning how often you can afford to make the call. -- Matthew Skala [EMAIL PROTECTED] Embrace and defend. http://ansuz.sooke.bc.ca/ _______________________________________________ Cinelerra mailing list Cinelerra@skolelinux.no https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra
