On Sun, Jan 30, 2011 at 12:06:31AM +0100, Landry Breuil wrote: > On Sat, Jan 29, 2011 at 10:23:51AM +0100, Landry Breuil wrote: > > On Thu, Jan 20, 2011 at 07:49:57PM +0100, David Coppa wrote: > > > On Thu, Jan 20, 2011 at 5:46 PM, Landry Breuil <lan...@rhaalovely.net> > > > wrote: > > > > On Thu, Jan 20, 2011 at 05:02:55PM +0100, David Coppa wrote: > > > >> On Thu, Jan 20, 2011 at 4:54 PM, David Coppa <dco...@gmail.com> wrote: > > > >> > On Thu, Jan 20, 2011 at 4:43 PM, Landry Breuil > > > >> > <lan...@rhaalovely.net> wrote: > > > >> > > > > >> >> spoke too fast, there's a regression for clients of http output. > > > >> >> Dunno what mpd > > > >> >> 0.16 outputs, but after the end of a track, instead of playing the > > > >> >> next > > > >> >> one mplayer goes into a loop : > > > >> >> > > > >> >> Cache not filling! > > > >> >> Cache not filling! > > > >> >> > > > >> >> 0.15.12p5 works fine. > > > >> > > > > >> > Is this with ogg vorbis? > > > >> > > > >> Or, more probably, it's related to this: > > > >> > > > >> http://musicpd.org/mantis/view.php?id=3149 > > > >> > > > >> http://git.musicpd.org/cgit/master/mpd.git/commit/?id=18b30b50197261a8f062f44b5e8b24b21fb2b280 > > > > > > > > I'm not sure.. mpd itself still runs fine, the libao output still plays, > > > > it's just the http client that fails. If i restart it, it resumes > > > > streaming fine. And yes, this is with ogg vorbis encoding, same setup: > > > > > > Can you please check if this happens with both: > > > > > > mplayer -demuxer lavf $url > > > > > > and > > > > > > mplayer -demuxer ogg $url > > > > > > If the latter works fine... maybe it's mplayer's fault > > > > The latter indeed works fine. But still, mplayer + default lavf demuxer > > _works_ with mpd 0.15.12, so that's a regression. > > oh, and another regression/change noted.. the mixer_type option seems to > have changed; now if i change the volume through a client, only the > libao output in my case gets changed, not the volume of other outputs > (like the http stream). i have mixer_type set to software for libao, and > no mixer_type set for http, and i'm pretty sure with the same conf mpd > changed both output volumes when changing mpd volume.
Now that the portstree is unlocked, that'd be good to get mpd 0.16 in, provided those two regressions are fixed... Landry
