Announcement

Collapse
No announcement yet.

Linux is not ready for Steam

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Basically, I'm calling for an audio libfuse.

    Comment


    • Originally posted by darkphoenix22 View Post
      To the mixer. Without requiring that applications pass the audio through the wrapper. The mixer should just regulate the connections.
      But then the mixer is providing the same services as PulseAudio. See, you can't get around the need of that functionality.

      If you just want it provided by something that's not Pulse then that makes no sense. If it is required, it's required. No matter what you call the code.

      Comment


      • Something that manages the connections, but is transparent to the applications and drivers. And does not route audio, only autodetecting the audio devices and using the ALSA mixer to switch them.

        Comment


        • Originally posted by mugginz View Post
          But then the mixer is providing the same services as PulseAudio. See, you can't get around the need of that functionality.

          If you just want it provided by something that's not Pulse then that makes no sense. If it is required, it's required. No matter what you call the code.
          I have no problems with the ideals of PulseAudio. I just believe that per-app controls and audio mixing should be handled by the driver.

          And the auto-routing should just use the plain ALSA mixer to do the switching. Sound should not go through a daemon.

          Comment


          • Originally posted by darkphoenix22 View Post
            I have no problems with the ideals of PulseAudio. I just believe that per-app controls and audio mixing should be handled by the driver.

            And the auto-routing should just use the plain ALSA mixer to do the switching. Sound should not go through a daemon.
            But then you're advocating that the complexity of PulseAudio should go into the driver proper instead. That seems to be at odds with any goals of simplicity and low-latency at the driver level. You seem to be saying put Pulse directly into the driver.

            Comment


            • Obviously the mixer has the ability to switch audio devices on its own given the drop-down menu. Why can't the autodetection be added to it simlar to a Volume Manager? This way the applications can talk directly to the ALSA drivers, while providing the automated audio switching.

              Comment


              • Originally posted by mugginz View Post
                But then you're advocating that the complexity of PulseAudio should go into the driver proper instead. That seems to be at odds with any goals of simplicity and low-latency at the driver level. You seem to be saying put Pulse directly into the driver.
                Nah I saying the sound APIs and audio switching APIs should be complete and separate. The audio switching would be simlar to XRender, I suppose.

                Comment


                • Edit: XRandR

                  Comment


                  • Originally posted by darkphoenix22 View Post
                    Obviously the mixer has the ability to switch audio devices on its own given the drop-down menu. Why can't the autodetection be added to it simlar to a Volume Manager? This way the applications can talk directly to the ALSA drivers, while providing the automated audio switching.
                    But if an application has begun to play an audio stream, and it's going through the sound card, and then a bluetooth headset is activated, then the sound card driver needs to stop dmixing it through it's own hardware and strip out that stream and send it to the bluetooth driver until the bluetooth device is turned of and then the ALSA sound card driver then need to be signalled that this happened and then begin to re-mix that stream into the stream it's playing into its own hardware.

                    Do you see I problem with this type of architecture? I certainly do.

                    Comment


                    • Originally posted by darkphoenix22 View Post
                      Nah I saying the sound APIs and audio switching APIs should be complete and separate. The audio switching would be simlar to XRender, I suppose.
                      Xranr is augmented by another type of server. An x server.

                      Comment

                      Working...
                      X