On Thu, 29.10.09 15:43, Jeremy Nickurak (jeremy at nickurak.ca) wrote: > > That does rely on applications behaving well. > > What about a short-lived alsa audio program? There should be a way to > apply a match to these too... maybe a history of streams? Or simply a > minimum time that streams live in pavumeter before it disappears? This is admittedly a problem, but I kinda hope that it will fix itself by applications tagging event sounds properly. Even for legacy applications you can do that with minimal work most of the time: http://pulseaudio.org/wiki/ApplicationProperties I certainly would make sense to allow volume changes shortly after a stream died. The low-level technology in PA allows that, it's just that g-v-c and pavucontrol currently do not make use of this. But then again, I am wondering if the time you need to fix g-v-c/pavucontrol this way wouldn't be better spend in fixing the applcitions in question to tag their streams properly. I think generally only event sounds are short-lived anyway (at least noone was ever able to give me a convincing counter-example). Lennart -- Lennart Poettering Red Hat, Inc. lennart [at] poettering [dot] net http://0pointer.net/lennart/ GnuPG 0x1A015CC4
