[01:36] <persia> ScottL, Um, huh.  You might need to run ubuntu-bug debian-installer from the installer environment, except that I don't think ubuntu-bug is available then.
[01:36] <persia> May as well try the procedure you describe: worst case, you'll end up with instructions on how to get the right data.
[02:09]  * ailo_at is learning Python very slowly and painfully
[07:21] <abogani> ailo_at: Good choice!
[07:28] <abogani> ailo_at: Test a realtime system it is very incredible hard thing but you are doing a really good job!
[07:55] <abogani> ScottL: Follow these steps:
[07:55] <abogani> 1) Install devscripts and dput
[07:55] <abogani> 2) Set-up a GPG environment
[07:55] <abogani> 3) Create a dedicated PPA on your launchpad page (In my case I called it "Broken").
[07:55] <abogani> 4) cd ubuntu-natty-meta-lowlatency/meta-source
[07:56] <abogani> 5) dch -i, add a comment, remove "ubuntu1", save and quit
[07:56] <abogani> 6) debuild -S -sa
[07:56] <abogani> 7) cd ..
[07:56] <abogani> 8) dput ppa:youruser/yourppa linux-meta-lowlatency_2.6.38.X.XX_source.changes
[07:56] <abogani> 9) Wait that PPA doing the rest
[07:56] <abogani> scott-upstairs: ^
[08:04] <abogani> ailo_at: The approach that I have found profitable is to "stress" the system as much as I can. I'm generaly use cyclictest (and not Ardour or other sound software simply because I'm not a musician) and put a lot of other works. For example: a linux kernel building (make -jx where X is a lot bigger of number of cores), execute hackbench, launch some "find /" and execute  a stupid dd if=/dev/zer of=/dev/null all these things launched in ba
[12:31] <ailo_at> paultag: Check out the code when you have a chance git://gitorious.org/ubuntustudio-controls/ubuntustudio-controls.git
[12:58] <ailo_at> abogani: Thanks. I will look into it.
[16:10] <ailo_at> abogani: What I found was that if a process is not running with realtime privilege (puredata can be run either with or without), it won't easily cause xruns on jack, even if the processes(I use one for each core) are flooding the cpu to a miximum. I've been thinking about what type of processes could cause xruns. I guess pulseaudio could, if trying to init it while jack is running (cause pulseaudio is using rtkit, right?).
[16:10] <ailo_at>  How could we find out what causes the xrun? What logs could I monitor to look for events that happen at the same time as the xrun occurs?