'CustomSink' object has no attribute '__sink_listeners__'

[KuraiAI]

Summary

'CustomSink' object has no attribute 'sink_listeners'

Reproduction Steps

Ever since the DAVE encryption fix, I've been getting an error
'CustomSink' object has no attribute '__sink_listeners__'
when I send the custom sink into the start_recording function of the voice client. This was never a problem in the past, and unfortunately due to Discord's new requirements, I can't use the pre-DAVE versions of pycord. Other Discord Python libraries don't work at all (I actually used a different one before swapping to Pycord because sinks were too slow on that one), and non-custom sinks don't work for my use case.

I'll attach the init for my custom sink.

Minimal Reproducible Code

def __init__(
        self,
        queue: asyncio.Queue,
        loop: asyncio.BaseEventLoop,
        *,
        filters=None,
        data_length=50000,
        quiet_phrase_timeout=1.2,
        mid_sentence_multiplier=1.8,
        no_data_multiplier=0.75,
        max_phrase_timeout=30,
        min_phrase_length=3,
        max_speakers=-1,
        min_speaking_time = 1
    ):
        self.queue = queue
        self.loop = loop

        if filters is None:
            filters = default_filters
        self.filters = filters
        Filters.__init__(self, **self.filters)

        self.data_length = data_length
        self.quiet_phrase_timeout = quiet_phrase_timeout
        self.mid_sentence_multiplier = mid_sentence_multiplier
        self.no_data_multiplier = no_data_multiplier
        self.max_phrase_timeout = max_phrase_timeout
        self.min_phrase_length = min_phrase_length
        self.max_speakers = max_speakers
        self.min_speaking_time = min_speaking_time

        self.vc = None
        self.audio_data = {}

        self.running = True

        self.speakers = []

        self.temp_file = NamedTemporaryFile().name

        self.voice_queue = Queue()
        self.voice_thread = threading.Thread(target=self.insert_voice, args=())
        self.voice_thread.start()

Expected Results

Channel is found
Voice client is received from the connect function
start_recording uses the write function in my custom sink

Actual Results

I get an error when the start_recording function is called.

Intents

default, message_content, members, voice_states

System Information

- Python v3.11.9-final
- py-cord v2.8.None-final
- aiohttp v3.13.3
- system info: Windows 10 10.0.26200

Checklist

• I have searched the open issues for duplicates.
• I have shown the entire traceback, if possible.
• I have removed my token from display, if visible.

Additional Context

No response

[Paillat-dev]

pycord 2.8 does not support voice reception, please read the changelog. If you need it, use #3159.

[KuraiAI]

This changelog? It says that DAVE was added for Audio and Video, but I don't see anything saying that voice usage was turned off? Is there a different changelog somewhere that mentions that it's not usable or am I becoming blind?
I'll see if that PR works for me, I do need voice for my uses (it's the only reason I need pycord, the rest of the features I use are not necessities).

[Paillat-dev]

Added support for Discord DAVE (Audio & Video E2EE) for voice-sending related features. (#3143)

[Paillat-dev]

In any case, @KuraiAI , the PR I linked above for voice receive works fine at this point and it will likely be merged soon, and be released in 2.9

[KuraiAI]

Yep, thanks! I manually added some things from that PR when I was searching before making the issue, and was able to get my bot to stay connected to the voice chat in the same way that PR version is letting it stay connected, but the audio isn't going in anymore for voice recognition, but that might be something on my end (the DAVE encryption blocking happened at the exact same time I was setting up Gradio for GUI)
I look forward to seeing it in 2.9!