Kernel testing
Ronan Jouchet
ronan at jouchet.fr
Thu Dec 9 03:14:47 GMT 2010
On 10-12-07 12:50 PM, ailo wrote:
> I noticed Alessio asked Tims opinion on whether or not the lowlatency
> kernel is needed.
> I don't know what has been suggested or talked about in the past, so
> forgive me if I repeat something here.
>
> [...]
>
> Maybe someone in this mail-list does tests to determine whether or not
> the generic kernel can do these things as well as the
> lowlatency/realtime kernels?
Hello everybody,
Alessio, Tim, I fail to understand this sudden position shift towards
-lowlatency. Half a dozen users tested the -lowlatency kernel this
summer, with the following results:
- Yes, lowlatency provides noticeably lower latency than generic
- Rt/realtime remains snappier and thus remain a requirement for
diehard audio users needing top notch performance.
- However, rt/realtime are very painful to maintain compared to
lowlatency that consists in "simple" build-time tweaks
This resulted in an apparently well-accepted *focus on lowlatency for
Natty*, summed up by Scott in end-of-September emails. We even nailed it
in https://wiki.ubuntu.com/RealTime , including a testing table
featuring two neat green "lowlatency" rows.
Alessio, don't get me wrong here: I respect this amazing work you've
been doing (and that I'd totally be unable to do), I know you are doing
it alone, and I know the absence of feedback can get discouraging.
I just want to understand why this question pops now. We should be
pushing for inclusion in Natty, and instead of focusing on testing and
bugfixing we are re-debating an already closed subject.
Anyway, *yes there is ongoing testing*: right now holstein, ailo, I (and
possibly other users) are testing it. Again:
Testers: please keep hammering the lowlatency kernel from Alessio's PPA,
and report your test results.
Alessio: please bear with us and continue your efforts, rest assured the
interest is there.
I hope this email doesn't sound harsh. No aggressiveness intended, I
only mean to be synthetic and to the point. Thank you,
Ronan Jouchet
More information about the Ubuntu-Studio-devel
mailing list