Feeding audio from libdatachannel to gstreamer?

[wusspuss]

There is an example of feeding rtp to gstreamer via udp in examples, but I was wondering how I can do that via a function call in the same program?
So far I tried

source = Gst.ElementFactory.make("appsrc", "mysource")
source.props.caps=Gst.Caps.from_string('application/x-rtp')
sink = Gst.parse_bin_from_description('queue ! rtppcmudepay ! queue ! autoaudiosink', True)
pipeline = Gst.Pipeline()
source.link(sink)
pipeline.add(source)
pipeline.add(sink)
pipeline.set_state(Gst.State.PLAYING)

and feeding it via source.emit('push-buffer', Gst.Buffer.new_wrapped(rtp_bytes)) in rtcMessageCallbackFunc, but I can hear no sound coming out, and with

bus = pipeline.get_bus()
bus.add_signal_watch()
bus.connect('message', cb)

I am able to get these errors, consequently:
GstMessageError, gerror=(GError)NULL, debug=(string)"../gstreamer/subprojects/gstreamer/libs/gst/base/gstbasesrc.c\(3132\):\ gst_base_src_loop\ \(\):\ /GstPipeline:pipeline1/GstAppSrc:appsrc-140103022775504:\012streaming\ stopped\,\ reason\ not-negotiated\ \(-4\)", details=(structure)"details\,\ flow-return\=\(int\)-4\;";
and
GstMessageError, gerror=(GError)NULL, debug=(string)"../gstreamer/subprojects/gst-plugins-base/gst-libs/gst/audio/gstaudiobasesink.c\(1117\):\ gst_audio_base_sink_wait_event\ \(\):\ /GstPipeline:pipeline1/GstBin:bin1/GstAutoAudioSink:autoaudiosink0/GstPulseSink:autoaudiosink0-actual-sink-pulse:\012Sink\ not\ negotiated\ before\ eos\ event.";
What could I be doing wrong? I have been able to send audio successfully in a similar fashion, but got stuck really bad at receiving.

[wusspuss]

Also, if a track specifies several codecs, like

a=rtpmap:0 PCMU/8000
a=rtpmap:8 PCMA/8000
a=rtpmap:102 ILBC/8000
a=rtpmap:103 ISAC/16000
a=rtpmap:104 ISAC/32000
...

how do I tell which decoder gstreamer can use?