On Mon, 08.06.09 19:11, pl bossart (bossart.nospam at gmail.com) wrote: Heya, > While I am at it, what was the rationale for the new PA_SINK_IDLE > state introduced in 0.9.15. That state existed much longer than that. > From what I understand this is a running > sink with no incoming data (no non-corked inputs). That is correct. > But the state transitions in alsa-sink essentially consider this > state as a PA_SINK_RUNNING one. So if you have a suspended sink and > add an input in corked mode, the sink will go to IDLE mode, which > will start transmitting silence and after a timeout the sink will go > to suspend again. D'oh! Why not suspend if you don't have anything > to transmit? mdule-suspend-on-idle needs a bit of fixing I guess so that it doesn't resume a device prematurely if the stream that is create is actually created in corked mode. I have now added this to my todo list. Should be a simple fix. Cheers Pierre, -- Lennart Poettering Red Hat, Inc. lennart [at] poettering [dot] net http://0pointer.net/lennart/ GnuPG 0x1A015CC4
