| OvenWerks | Eickmeyer: if you have been monitoring #LAD, it seems RT is starting to roll into kernel upstream. | 15:49 |
|---|---|---|
| * OvenWerks doesn't know if he is excited or not | 15:50 | |
| Eickmeyer | Huh! interesting. People are still going to look for an RT kernel, though. :P | 15:50 |
| OvenWerks | Eickmeyer: from the sound of it, RT will be a switch or two in the build or even at boot. So an RT kernel would be mainline with RT turned on in the same way lowlatency is. | 18:25 |
| OvenWerks | in my experience switching on lowlatency or RT is not the whole picture anyway. People have gotten so used to plug and play they can't even think any other way. | 18:27 |
| OvenWerks | but RT/LL requires setup as well as just installing a new kernel. | 18:28 |
| OvenWerks | USB audio is not RT friendly and the way mother boards are built doesn't help. | 18:28 |
| OvenWerks | It is easy to end up with a mouse that has has higher priority than the USB audio | 18:29 |
| * OvenWerks has tripping fingers | 18:29 | |
| Eickmeyer | Agreed. | 18:39 |
| OvenWerks | https://github.com/jhernberg/udev-rtirq may help | 18:40 |
| Eickmeyer | For my USB audio interface, I just make sure it's plugged-in directly to the computer and not via any hub. | 18:42 |
| OvenWerks | It would seem to be able to prioritise an audio device above a mouse plugged into the same USB port... though I would imagine latency would still be higher. | 18:42 |
| OvenWerks | Eickmeyer: yes, but... there are too many USB ports on computers that are part of an internal hub. | 18:43 |
| OvenWerks | In fact pretty much all internal USB ports go through a "HUB" internally. Some only have one port to the outside world though. | 18:44 |
| Eickmeyer | Agreed. USB 3.0+ tends to be better with that since even 2.0 is given more throughput. | 18:44 |
| Eickmeyer | But older USB 2.0 systems tend to have more difficulty. | 18:44 |
| OvenWerks | a USB3 hub with SUB2 ports should be the same as separate ports | 18:45 |
| Eickmeyer | My system has all USB 3, which is probably part of why I don't have that much difficulty. | 18:45 |
| OvenWerks | I would love to have some script/daemon that when it detects USB audio devices plugged in, checks to see if any other device uses the same port or even IRQ and warn the user that is not a great port to use. | 18:46 |
| OvenWerks | but it seems it would need to detect USB2 over USB3 | 18:47 |
| Eickmeyer | My audio interface is USB 2, so yeah. | 18:48 |
| Eickmeyer | I know that if I have WiFi on that there's some interference, but other USB devices don't usually interfere. | 18:49 |
| OvenWerks | WIFI drivers are notorious for bad drivers | 18:50 |
| OvenWerks | (having large time consuming atomic code) | 18:50 |
| Eickmeyer | OvenWerks: Can I tell nilshi to go F himself? >:( | 19:22 |
| OvenWerks | Eickmeyer: there is always a new release every 6 months. nilshi is one of those advocating monthly release so he should just understand that. | 19:29 |
| Eickmeyer | Monthly release of what? Ubuntu? | 19:30 |
| OvenWerks | 12:19 < nilshi> That is beyond my motivation though | 19:30 |
| OvenWerks | plugins | 19:30 |
| Eickmeyer | Yeah. He wants us to have NSM so bad, but he's not motivated to do anything. That's very telling. Wants everyone to do the work for him. | 19:31 |
| OvenWerks | not really, more a lack of thought. He is willing to do things, but the reality linux lives with is that nothing happens without someone wanting to do things | 19:32 |
| OvenWerks | Even those wanting to do things have a time elecment that requires prioritizing | 19:32 |
| Eickmeyer | Well, he's trying his hardest to keep us from including RaySession. I can't include NSM because of NTK, and introducing an entire toolkit is beyond our scope/time. | 19:33 |
| OvenWerks | Eickmeyer: if one can truely just switch direct from NTK to FLTK by just changing a few lines of code... that would quiet everyone. | 22:11 |
| OvenWerks | Eickmeyer: I do see his point, if someone sends a session file along with everything else, it should be able to work on another system. RS does not and will never allow that. Having said that, Ardour/MixBus allows a session made in windows to be used Linux or macos. I don't see any linux sm ever being able to do that. | 22:23 |
| OvenWerks | ( I guess lash did) | 22:23 |
| OvenWerks | However, nsm (and rs) relies on the application storing it's last use exit point _somewhere_ so it ends up being a single system only deal anyway. | 22:25 |
| OvenWerks | the best a transfered NSM session could do would be to restart the same applications and create the jack connections. Each application would show up just started (unless one had really bad luck in which case the wrong file would be loaded) and the user would still have to load the project they wanted and make any settings changes needed to go with it. | 22:27 |
| OvenWerks | Eickmeyer: I did look at source for NSM... And it looks to use all FLtk stuff pretty much straight up. I don't see anything that must have NTK. I am not very good figuring out waf scripts... | 22:33 |
| Eickmeyer | OvenWerks: The other thing is that assumes that everyone has the same setup, which I guarantee isn't even possible. | 23:19 |
| OvenWerks | in my mind, that pretty much does away with any complaint about compatable. The only use case for being "more compatable" is to change SM part way through a project... | 23:38 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!