Regardless, still thinking about getting into burst. It'd be nice if there were more poc coins though, if burst starts spiraling the drain it's nice to have more options.
I think cloning + maintaining + adding new feature for POC coin is not as easy as normal PoW , PoS coin. I suspect if there is ever a clone of burst it would be just copy and paste. No new feature and it would probably be a scam coin because it's not an easy coin to maintain.
Also , it's not a good time for alt coin in general for this period of time.
If burst were to be release last year, i bet it would reach 5,000 - 10,000 sat.
If alt coin pick p next year , we will probably see some cloning but by then burst could probably be the market leader in PoC coin.
exactly right.. meantime pools are in major disarray and SATA disks are expensive as heck 6T won't go down in price that easily.. so not just difficult to maintain but not that ultra cheap to mine, also it is very hard to hijack disk space of required magnitude so this is very anti-bot resilient coin
pool operators have to be cautious and provide steady as much as possible flow of deposits it is very important in general not just for individual miners it is quite unbelievable to me that pools established by core developers to which also assets are tied to are in such a weak state
java is not designed to scale quite well by simply adding memory to centralized hosts.
the solution is to port (start from scratch) the required burst protocol parts to a solid c++ backend and run the pools with that.
every single effort which is put into the java code base to stabilize the pool situation is in my eyes not worth it cause the next limit is already set.
what burst needs is a core functionality c++ api to host professional services.
this includes to act as full network peer (aquire and distribute the blockchain),
submit blocks to the network,
extract reward receipients from the chaindata,
send burst,
extract messages,
send messages
and to create addresses from private keys.
if this works the whole api functioniality for the asset exchange and marketplace may get added.
a intermediate solution may be to create a mysql backed partial blockchain with php access to take the pool load away from the java components on the pools. however, as far as i know is the core pool servercode still based on java.
another option to stay with java may be to add autonomous proxies to the pools which submit the round reward assignments to a hidden internal pool server which distributes the rewards. keeping them in sync may be much harder to code than starting to design the c based api from scratch.
a simple pool design may be this:
wallet (current or c based) --> distribute current block information and stats to the miner connect nodes
connect nodes --> provide block data to the miner, accept and verify shares, provide cached stats --> submit miner results to central hidden node
hidden node --> distribute rewards, provide stats
load balancer --> knows throughput of shares for each node and assigns miners to nodes
