Lets assume 100% of miners can handle up 32Mb blocks and 80% can handle > 32Mb blocks with the block cap removed. Miner A for some reason decides its going to create a block of 64Mb (by filling it up with its own transactions as the mempool is too small) and mines and broadcasts this block.
Scenario 1:
No other miner accepts this block (all have a 32Mb soft cap) and miner A continue mining on his own chain or realize its stupid throwing money away and rejoin the 32Mb soft cap chain.
Scenario 2:
All miners that areable accept this 64Mb block which leaves 20% of miners unable to keep up which means they either upgrade or they leave and go mine BTC or whatever. The lost hashpower of those that left will probably equalize by new able miners joining because of the profit opportunity in the now lower difficulty.
Scenario 3:
50% accepts the 64Mb block leaving 2 competing chains without immediate clarity on what the real chain is. Yes this is a situation you want to avoid but miners are not incentivised to create situations like this. They don't need a block limit to tell them how to behave, they will use soft caps and other methods to be profitable as they possibly can be. Miners can create competing chains without block caps in any case but they wont as is evident.
Why is systematic block size increases preferred (which implies clients have to upgrade frequently) over just a single final upgrade where miners set the limits as appropriate?
[link] [comments]
source https://www.reddit.com/r/btc/comments/99c81o/please_explain_what_are_the_concerns_on_removing/
No comments:
Post a Comment