Comment 15 for bug 1325752

pulseaudio releasing a device because it is asked to do so, is not a bug. That is, it seems pulseaudio is acting correctly, if an application wants direct access to the audio device, Pulseaudio must give it up. There can be only one one application attached to an ALSA device at a time. The idea of pulseaudio (and other audio backends) was to get around that by adding a mixer with psuedo alsa ports. This allows more than one application to think it has access to a sound card. However, there are some kinds of applications that need direct access to an audio device to be useful. In order to accomodate these applications pulseaudio has a command to release the device or pause itself so that this application can run. When this happens all applications that are using pulseaudio to play sound will stop (in various ways). This is not a bug, but a nessessary feature needed for pulse to exist in the (semi)pro audio world.

My reason for attaching the bug to mixxx is that this is the only place I can reproduce it. And it tells me that Mixxx is using the pause pulseaudio command incorrectly. It should not pause pulse when it is not using an alsa device or starting jackd on it's own and it does. This is a bug because mixxx suggests that it can be used for Internet radio and one of the most common things done on radio is to take phone calls to air... on a computer that generally means Skype for a lot of people. Skype needs pulse to work and can be bridged to jackd(bus) but not with pulse paused.

I will look more closely at musescore and see if it does the same thing without using jack as my audio device.