[haiku-development] Re: Res: Scheduler algorithm improvements

  • From: Urias McCullough <umccullough@xxxxxxxxx>
  • To: haiku-development@xxxxxxxxxxxxx
  • Date: Fri, 6 Nov 2009 11:16:10 -0800

On Fri, Nov 6, 2009 at 11:05 AM, Axel Dörfler <axeld@xxxxxxxxxxxxxxxx> wrote:
> I have spend quite some time to reduce the latency with several system
> components during BeGeistert.

<snip>

> Even if the system itself works nicely (which I don't think it does
> yet), each driver alone could be responsible for such problems. I don't
> think it makes sense to start collecting data at this point, until we
> can rule out the problems the system itself has.

Would it make sense to develop latency-detection tests to test the
various suspect subsystems on different hardware configurations to
find where the biggest problems are over a variable set of
environments?

Or as you said, is it also too early for this type of data?

Other related posts: