On Fri, Jul 18, 2008 at 11:52:33AM +0200, CJ van den Berg wrote: > If opening a device causes pulseaudio to block indefinitely, then the > device needs to be fixed (as it was in your case). I don’t see what > pulseaudio can (or should) do about this.
It could timeout, so that the rest of the system can finish its startup. In fact, it might already timeout: I waited some time but maybe not enough, and didn't check the source code. > > In understand that the origin of the bug doesn't lie in pulseaudio, and that > > is has been fixed already (see #491114). But, whatever causes pulseaudio to > > fail its startup, I think it shouldn't freeze the whole boot process, but > > rather die gracefully. > > I don’t see how bad permissions could cause pulseaudio block > indefinitely on a device. What makes you think that is what happened? Because everything under /dev was 660 instead of 666 (belonging to root). But you made me realize that it might as well have been the case that the devices did not even exist (#491114 messed up *many* things). I don't see why a device that do not exist at all should be treated differently from a device that you can't open because of bad permissions. Then again, I might be wrong on the real origin of the problem. If this the case, please forgive me and close this bug report. But seeing pulseaudio hanging on startup and blocking the whole system (for whatever reason) didn't seem the "right thing" to me — unless it times out and I wasn't patient enough. Thank you for your quick reply anyway. Regards, -- Gabriel Kerneis _______________________________________________ Pkg-pulseaudio-devel mailing list Pkg-pulseaudio-devel@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/pkg-pulseaudio-devel