May 29, 2018

@neillm a quick question, how do you mitigate congestion out outbound messaging in a P2P channel in LB? so i see here https://github.com/libbitcoin/libbitcoin-network/blob/version3/include/bitcoin/network/proxy.hpp#L62 "dispatch_.lock" here, that must be a way to serialize individual messages' submissions... what does the dispatch lock do here actually, will it process that code (that is "&proxy::do_send") at first after unlock? and also, say i'm the full synced node logics and i just got a valid block and now i like to broadcast it to all the peers that are connected to me, and there's a tooon of them of course, and some are slow or just tanked or random malfunction, right. when the LB fully synced node wants to relay transactions and blocks, what are the logics it applies to discern if it should? also say a P2P channel was jammed right now - the TCP connection was slow and a previously committed 1MB block was taking say 25 more seconds to go through - at the point it has gone through, is there any pushback feedback notification where the fully synced node would learn to know that a particular channel C was busy sending stuff when I wanted to forward to it before, but, now 25 secs later you can send? where is that logics :))