[FFmpeg-devel] Automated builds and some tests

[Mike Melanson]

Michael Niedermayer wrote:
> running benchmarks on a system which also runs other stuff like a webserver
> is not good, not in theory and in practice the FPS values look quite noisy
> no way anyone could notice a 1% change that way ...

It the autobuild solution I'm working on, everything is an execution 
module that is individually timed. For timing, I am using a Python 
facility that returns 5 time values, one of which is actual time that 
the child process has spent actually executing on the CPU. I hope that 
will be an effective benchmark.

-- 
	-Mike Melanson