[02:02] Eickmeyer: I have to go out for a bit... but I need to understand your very short bash script you added: autojacj-start [02:02] *autojack-start [02:03] Seeing as convert-studio-controls is gone, is it still needed? or does that part need to be removed? or? [02:04] OvenWerks: Hang on while I switch gears and refresh my own memory. [02:04] Eickmeyer: maybe I need to be using /usr/bin/systemctl --user restart pulseaudio || true rather than pulseaudio -k [02:05] OvenWerks: Removing the convert-studio-controls line is good, but the rest is probably still needed. [02:05] OvenWerks: Yes, see the comments I made. [02:05] do note that it will no longer be run by systemd [02:05] in the script. [02:05] but from /etc/xdg/autostart [02:06] Ubuntu is running pulseaudio via systemd now. [02:06] As is Fedora, and most likely Arch. [02:07] So, pulseaudio -k is actually deprecated. [02:07] OvenWerks: ^ [02:07] I know that bit. but when we start autojack in the same manner it seems many other system side users end up starting autojack as well. [02:08] Well, we do know that pulseaudio has to restart after autojack starts. [02:09] So, no matter which method, it's still going to have to kick pulseaudio via systemctl --user restart pulseaudio [16:49] Eickmeyer: I think they have fixed that.... at least pulseaudio -k does work. I think maybe pulseaudio doesn't respawn itself but systemd does [16:50] however the systemctl method is probably still better. [16:51] OvenWerks: Yeah, systemctl doesn't kill systemd, it basically tells systemd to restart pulseaudio at the user level. Killing/restarting systemd itself is impossible (oversimplified, but yeah). [16:52] Eickmeyer: So if I add/change to the systemctl method right in autojack we can remove autojack-start? [16:52] I think so, yes. [16:52] It was just a wrapper to be a workaround basicallly. [16:53] Initially, it was a patch. [16:53] Eickmeyer: yeah, I am not talking systemd, but the systemd pulseaudio.service already does a respawn if PA goes missing [16:53] That is what I was relying on [16:54] Right, but systemctl --user restart is still safer than systemctl --user stop. [16:54] however, it is possible that other distros don't [16:54] Better to be paranoid than rely on pulseaudio.service to do that when it might not. [16:55] I am still talking about pulseaudio -k, systemctl --user stop would tell systemd not to restart [16:55] pulseaudio -k just tells PA to exit [16:56] Right, systemctl --user stop would tell it not to restart, systemctl --user restart would tell it to restart. pulseaudio -k may or may not restart depending on how the system was set up, which might not have the desired effect. [16:56] Since we *want* it to restart, systemctl --user restart is the safer method. [16:57] So if I change to systemctl --user restart That would also remove the last thing that would interfere with cadence [16:57] Oh! Good point! [16:57] or allow cadence to interfere with contrtols [16:58] Still, having the two co-installed is still a bad idea. [16:59] That is true, but at least simply removing cadence would work. Or even simply not using the two at the same time. [16:59] Yeah, I just worry about stuff in ~/.pulse [16:59] user beware ;) [16:59] True. [17:00] That is totally a problem with cadence [17:08] Ok, I will switch things over. I think it is safe to assume anyone using controls is also using pulse and any system using pulse is using systemd [17:08] (except Fons?) [17:09] (actually Fons uses jackd not jackdbus or jackd2) [17:26] Yeah, I think one of our minimum requirements at this point is jackd2 for good reason.