right now, all the price data is sent over the broadcast channel so if one function using that channel is replicated over many threads, might start using too much memory compared to one thread(the price sender thread) keeping all the prices in memory and only sending the ones requested