Example: Sync NumPy¶

A broadcast variant of the NumPy stream where every viewer shares the same frame feed. Demonstrates coordinated fan-out with background tasks.

Location¶

• Repository: https://github.com/cetic/pyplet_examples
• Path: apps/pyplet_examples/examples/sync_numpy_*
• Client entry point: sync_numpy_client.py
• Server loop: sync_numpy_server.py

Fetch the examples submodule:

git submodule update --init --recursive

Highlights¶

• sync_numpy_client.py:1 reuses the standard NumPy client loop to stay aligned with the shared feed.

from .numpy_client import websocket_client_loop

• sync_numpy_server.py:17 spawns a background task that keeps generating and broadcasting frames.

async def send_to_all():
    while True:
        if sockets:
            bytes = io.BytesIO()
            array = np.random.randn(10, 10)
            np.save(bytes, array)
            value = bytes.getvalue()
            for ws in list(sockets):
                try:
                    await ws.send(value)
                except tornado.websocket.WebSocketClosedError:
                    sockets.remove(ws)
        await asyncio.sleep(1)

asyncio.create_task(send_to_all())

• sync_numpy_server.py:24 prunes disconnected sockets while iterating the subscriber set.

for ws in list(sockets):
    try:
        await ws.send(value)
    except tornado.websocket.WebSocketClosedError:
        sockets.remove(ws)

Start here when you need a baseline for multi-client synchronized streams.