WIP: I will probably never finish this. But it might be helpfull to some people out there. AirPlay audio for the rest of us. An (hopefully) easy to use implementation of the AirTunes v2 Protocol.
- Cleanup the code!
Send DAAP: Artwork, Trackinfo (use audiotools metadata)- PCM stream support (look at RAOP-Player for how to send a raw unencoded pcm stream)
Fix: Move the seq argument out of RTSP Client and work with the rtp time insteadFix: replace the status variable with locks (I don't know if this is required, because all rtsp request already wait until a response is in the queue)- Fix: http server lags
- Fix: Dmap classes have some problems with encoding certain ints
- Fix: rewrite the audio-sync class, so that it is actually working and thread safe
- Fix: Audio is sometimes only processed by some devices, but not all
- Fix: connection interrupted, because receiver went offline
- Fix: stop function (including when the end of the stream is reached)
- Fix: setup.py requirements
Fix: player hangs if the RTSP connection times out- Fix: seekable audiofile implementation (there is a bug .frames can't be written from python)
- Fix: throw error if no device is connected and you try to stream audio or something like this (this includes the connection is lost)
- use zeroconf information for encryption type and codecs (authentication and password as well ?)
- Test with Python2.7
Note: Sending RTSP is (is it ... ?) blocking, sending audio packets is not. This is expected behaviour
- Allow volume control of all and single devices
- Allow streaming muliple files after each other
- Add testcases
- Add examples
- Provide a convenient wrapper around the password / authentication api
- Fix: do not allow interaction while flushing (Is this correctly implemented [See: Note above] ?)
- Fix: insert low16 and low32 where required (are there any missing ?)
- Fix: ATV 4 not sending the timing_port (we don't actually need it, but it would be nice to have) (is this possible ?)
- Fix: replace hardcoded RSA-Key by a dynamic one (?)
- Fix: use the latency of each device from the header (is this a good idea after all ? the header seems to be incorrect in some cases and it would require restructuring the code)