[03:57] <OvenWerk1> zequence: running things for a while now (almost 24 hours).
[04:00] <OvenWerk1> The kernel oops just happens at boot. (login?) but I keep getting new bug windows unless I rm the file it creates in /var/crash/ And it creates a new one for each boot and they all just sit there. So when I first login I keep getting these bug screens... but there has only been the one.
[04:02] <OvenWerk1> Jack is actually pretty stable. It runs really well for a while and then gets a number of xruns all at once meaning there is something I can shut off that will fix that. (PA-jack bridge might help too)
[04:06] <OvenWerk1> Happens at login.
[04:34] <OvenWerk1> zequence: switched from nvidia to open gpu driver... no more kernel oops
[05:44] <OvenWerk1> latency seems more in line with what I have seen on this machine before too. nvidio driver seems to have stopped LMMS from being installed. This may have to do with the wine stuff having crossed dependancies. (32 bit ISO)
[16:38] <holstein> its too bad.. the pulse dbus is reallly nice for doing what it does.. but, its too complex
[16:38] <holstein> there was always a way to do it from alsa, and that was too complex as well..
[16:40] <zequence> The only complex part is setting pulse to use the jack output
[16:40] <holstein> nah.. its all too complex for a normal user
[16:40] <holstein> jack.. the whole thing.. really even installing an OS ;)
[16:41] <holstein> folks buy a mac, then install protools or whatever. these days they dont even touch installation media.. its a lot that we have to expect of them to get running
[16:41] <zequence> I think the biggest problem is we have three kinds of jack
[16:41] <holstein> im not saying i have an answer.. i think falk's KXstudio thing where jack is just running is an interesting way
[16:42] <zequence> If there was only one, it would be much easier
[16:42] <holstein> zequence: lol.. that too :)
[16:42] <zequence> I'll be looking into making that a reality for Ubuntu Studio too. I'll SRU ubuntustudio-controls with an option to have jack be started by default
[16:43] <zequence> there are tons of stuff that need to be worked out though
[16:43] <holstein> as long as i can find the disable button :)
[16:43] <zequence> It'll be a toggle - but having jack on all the time isn't going to bother anyone
[16:43] <holstein> i think its a nice solve, in most cases
[16:43] <holstein> zequence: let me know if there is something to test
[16:44] <holstein> having a way to have it fail, and break, and restart.. should be quite helpful
[16:44] <zequence> there are tons of things we could do to improve the situation. We haven't done much to do that so far
[16:45] <zequence> My long term goal for these changes will be 16.04. At least I know I will have time to complete things until then
[16:45] <holstein> personally, i lean toward adobe being able to provice jack support ;)
[16:45] <holstein> but, thats not constructive.. just honest
[16:45] <holstein> provide*
[16:45] <zequence> pulseaudio needs to be improved too. If jack grabs a card, it should disappear from pulseaudio settings
[16:46] <zequence> And, if pulseaudio was set to use that card when starting jack, it should automatically connect to jack
[16:47] <zequence> My idea is I start some discussions on LAD about this. I know there are some folks who do jack/PA integration code
[16:47] <zequence> the module needs to be improved. It's way too CPU hungry
[16:47] <holstein> agreed
[16:48] <holstein> and, any of it can be changed, and break any part of it, and the parts dont/cant care
[16:48] <holstein> its not like jack or pulse care if jack or pulse work with it
[17:12] <OvenWerk1> holstein: I think jack could benefit from having a "default" input and output pair.
[17:13] <OvenWerk1> The client that just wants to put sound out should be able to do that. Right now, if my audio out is 9 and 10 for left and right there would be no sound.
[17:13] <OvenWerk1> (9/10 is my s/pdif lines)
[17:15] <OvenWerk1> Jack should have some way of turning unused lines off and renumbering/naming it's hw ports so that all the clients see these names
[17:16] <OvenWerk1> In my case it would be nice to have inputs 1-12 be 9,10,1,2,3,4,11,12 with 5-8 not showing at all for example.
[17:17] <OvenWerk1> In fact 11-12 should be renamed monitor rather than system.
[17:21] <OvenWerk1> Having jack run all the time is not a bad idea, it would be nice if it could generate a set of alsa ports that look like a device and set it default the way pulse does. That would make things way simpler.
[17:24] <OvenWerk1> The thing is there are two use cases, audio work and desktop work. In audio work latency matters and it is at these lower latencies that the PA-jack bridge has problems with xruns and cpu use. A systray button that has a dropdown with latency settings (desktop, mixing, live sound or at least desktop and time values)would make sense.
[17:24] <OvenWerk1> Jack can change latencies on the fly.
[17:25] <OvenWerk1> The only problem with changing latencies on the fly is knowing what the lowest latency a card can handle. WE could make 64/3 minimum, but many people would like less (mine will work down to 16/2).
[17:27] <OvenWerk1> We could disable the PA-jack bridge below certain latencies (512/2).
[17:28] <OvenWerk1> PA running with no output does not take any cpu to worry about.
[17:29] <OvenWerk1> I have tried all of these things manually and know it is possible.
[17:29] <OvenWerk1> I do not know what ardour does if latency gets changed while it is running though :)
[20:09] <OvenWerk1> Ardour does not seem to care if latency changes while it is playing, there are no dropouts... maybe a slight audible delay when going to a higher latency.
[20:11] <OvenWerk1> I would not want to do this while recording or playing normally, but it is good to know it will not make things crash. I was using patchage to change latency on the fly.
[20:51] <OvenWerk1> I confirm that the xfce monitor unility works fine with two monitors. Unfortunately, even with the open video driver, two monitors gives the kernel oops it had gotten rid of :P
[20:51] <OvenWerk1> Anyway, we should be able to remove the ARandR application now.
[22:07] <zequence> OvenWerk1: inputs and outputs of the device are of course handled by the ALSA interface. Have you ever looked at those?
[22:07] <zequence> pulseaudio has it's own way of filtering those, according to a more desktop style logic
[22:08] <zequence> OvenWerk1: For instance, /usr/share/alsa/cards/ICE1712.conf
[22:10] <zequence> Seems lmms was just uploaded
[22:10] <zequence> We need to do a last minute fix concerning bluetooth, in our seeds, and update our metas. Finally, rebuild our ISO
[22:11] <zequence> Only after both lmms and the bluetooth stuff is fixed can we get our release candidate. I'll try to get that done tomorrow
[22:12] <zequence> OvenWerk1: You find any problems with the current ISO otherwise?
[22:12] <zequence> Starting to get late to make any more changes before release
[22:12] <zequence> I have a feeling that there may be a few problems with the final release.
[23:03] <OvenWerk1> zequence: I will try again with todays ISO. The only thing I am seeing is that it appears that lmms will not install if the closed nvidia drivers are being used. I will check for that.
[23:04] <OvenWerk1> This would only effect the 32bit ISO
[23:24] <OvenWerk1> Having two clocks in the top panel is a bit odd. The right one allows setting events. The left one only allows seeing a calendar.
[23:26] <OvenWerk1> I can get rid of the one I want to keep, but not the one I wish to be rid of.
[23:31] <OvenWerk1> dual monitors work a bit funny. The right one happens to be mon-0 and the left one mon-1. I have set mon-1 as primary, but somethings still end up showing on mon-0.
[23:31]  * OvenWerk1 is going to do yet another install.