[01:31] len-dt: Do you have a personal page on wiki.ubuntu.com ? [01:35] ailo, A who? [01:35] Why would I do dat [01:36] ailo, ulatency doesn't work for me... [01:42] len-dt: Something like this https://wiki.ubuntu.com/ScottLavender [01:43] astraljava: Can't find your personal wiki page [01:43] I just got one myself. Very spartan https://wiki.ubuntu.com/ailo [01:44] len-dt: I'm about to start updating wiki.ubuntu.com/UbuntuStudio, and I will begin by determining who is doing what on this project [02:13] Ok, it's light enough to see where I'm running. I'm aiming at running 10km under 35 min next year, and have started extreme training for that. So, it's a fast lap almost every morning for me [03:27] ailo, it apears things swap no matter where swappiness is set... even with not very much memory use. unless my system started to hibernate itself? [03:38] len-dt: The first version of ubuntustudio-controls will only do one thing. Administer user realtime privilege. That is something you can't do graphically very easily on Ubuntu, and if you create a new user, there's nothing telling you you need to be apart of audio group to get realtime privilege [03:39] Gnome3, which Unity is based on, uses a much simpler user account manager, with no group editing [03:40] For people upgrading from any other Ubuntu derived distro, they will also not know about audio group [03:40] But, with -controls, they just need to use a toggle, and it will be set [03:43] ailo, Ok. [03:45] len-dt: My focus for the next couple of weeks will be the website and the wiki. So, I won't be helping out much with testing, which I think we need to set up a plan for. [03:46] We need to document what a tweak does and why we need it [03:46] Also, the testing must accurately prove it [03:46] Ya, my testing has been useful if not too practical. [03:47] but it does seem to show that swap does bad things for audio. [03:48] And it shows that swap happens all the time. It does not matter where swappiness is set, after enough time, things still swap out. [03:50] len-dt: Are you saying, the swappiness setting has no effect, and that swap is used no matter what? [03:51] ailo, the swappiness setting has a great effect in the short term. but after 26 hours swap has happened anyway. The system shows 22%memory use and 46% swap... [03:52] 22+46 is a lot less than 100% [03:52] Sounds mysterious. Perhaps there are kernel options that would help [03:53] Well, how many recording sessions would last 26 hours? would have lots of tasks sitting doing nothing? [03:54] My internet seems a bit more stable today. I will try to get my kernel PPA running [03:54] Jack was busy of course and never suffered dropout, but when I tried to do anything after that... almost everything had to be swapped in. [03:55] len-dt: Live recording, on festivals and such, though 26h is probably a bit unusual [03:55] Actually, there are other things too [03:55] Like live installations [03:55] Live recording would be different again because there would be disk use and disk cache buildup. [03:56] Live installations might, or might not use hard disk. Anyway, hard disk is a variable [03:56] Both read and write [03:57] ailo, I am going to try again with a swapoff command and see where that goes. [03:57] I should probably reboot first though :-) [03:57] just to start with zero swap use. [03:58] You could try formatting your swap partition, removing swap all together. But, what effects would that have on the system? [03:59] ailo, I am starting to think that no one has really done a thorough job of setting up a system just for audio. [03:59] swap off runs the system with no swap. [03:59] s/swap off/swapoff/ [03:59] len-dt meant: "swapoff runs the system with no swap." [03:59] Ah, right [04:00] Where did that guy come from? I thought he was turned off... [04:00] He's a lurker [04:00] thank you for the correction kubotu [04:02] Anyway, in the short term, swappiness set to 0 is helpful and probably safer than swapoff. [04:10] ailo, I am pretty sure any standalone disk recording box has no swap, just memory. The device is engineered to never run out of memory. [04:20] ailo: I don't think I have one. [04:26] ailo, There are things I could do to make a very good recording machine, but not such a good computer for anything else. [04:27] I don't want to see us ship US tweaked so it is no good for all the other things it is supposed to be good for. Or for a lots of desktop like stuff for that matter. [04:41] Great timing for this. Someone finished it 4 days ago http://python-gtk-3-tutorial.readthedocs.org/en/latest/index.html [04:41] astraljava: Don't know if you need reference, but the above should be somewhate usefule [05:00] ailo, I'm going to try this: [05:01] Mem: 1016584k total, 447240k used, 569344k free, 48092k buffers [05:01] Swap: 0k total, 0k used, 0k free, 252688k cached [05:03] To make sure no swap at all does get rid of strange stuff. Then I will try hard use on the system to see what happens as I get close to full memory use. === ubott2 is now known as ubottu [06:00] I wonder what the record is in replies to a thread on debian-devel list [06:01] They're discussing a possible change of their init system [06:27] ailo, init? are they using upstart? [06:31] ailo, anyway, playing around with no swap. It is interesting how much stuff must be swapped out normally. I can't run as many synths as I have with swap on. but qjackctl is always there when I need to change connections. I'll run it like this till tomorrow after noon and see where things are. [06:35] len-dt: They are still on the old system, but there has been discussion about alternatives, where upstart is mentioned. There are other ones too. Debian Wheezy is alpha now, so I guess the change will only happen for wheezy+1. [06:36] So they are using sys v then [06:36] Yep [06:36] Why change? [06:37] I wouldn't know, and I guess that is what they are discussing about, very passionately at times [06:37] Aside from faster startup [06:37] I think faster startup is probably one of the main arguments for [06:38] Yes, it is the configuration headache that is the downside [06:43] I guess it might be cool for laptops to be slighly faster. Other than that, who cares? [06:44] It's not like booting Windows Vista anyway, which is like submitting to torture [06:58] ailo: I believe I do. Thanks! [08:06] Been testing mixing with ardour2 [08:07] Routing seems very buggy [08:07] Hopefully A3 is a lot better when it comes to that [08:40] When's it coming out, anyway? [08:44] Should be anytime now. They haven't released another beta for a while [08:44] Also, one would think mixbus will release a new version after that, but who knows [08:46] Hi gang, please have a look at https://wiki.ubuntu.com/Xubuntu/Roadmap/Specifications/Quantal/MultipleMonitors and comment if needed. ochosi on #xubuntu-devel mentioned this, and (quite correctly) presumed it might be interesting for [us|US] as well. [08:46] Oh, and he's here, too. :D Oh well. [08:47] har har. [10:25] A3 is no doubt better at routing [13:54] astraljava, the one thing I would add is a default profile for when a second monitor shows up mid session. [13:57] In the case of a desktop user who always has two monitors, but forgets to turn on the second one at boot. The setup defaults to single/mirrored when it is turned on even though it was set fort side by side on last boot. [13:59] In other words yes profiles. Keep the default for only one monitor as is, but add a default for two monitors. [14:02] Also, should the second monitor vanish... the whole desk should fit on the monitor that is left... it doesn't right now. It leaves the screen size as if both monitors were there but puts the main monitor in the middle so there is no longer access to the menu even for getting settings to change things. [14:03] * len-dt is off to work...