pull down to refresh
300 sats \ 2 replies \ @petertodd 30 Jul 2023 \ on: policy: Enable full-rbf by default by petertodd bitcoin
I'm glad to see this is getting no discussion. Because frankly this ridiculous issue should have been resolved 10 years ago.
Full-RBF by default settles it. We can remove the remaining BIP125 junk after that.
just when you will look here:
"The next Bitcoin halving cycle is going to put the entire Bitcoin mining sector at risk. The 2024 Bitcoin halving will double the sector average total business cost per Bitcoin or cut Bitcoin mining revenue by half. The sector average total business cost per Bitcoin ranges from $30k-$60k while the currently known Bitcoin all-time high is only $69,000. Many Bitcoin mining companies might not survive a 50% cut in Bitcoin mining revenue."
https://seekingalpha.com/article/4619872-riot-platforms-ambitious-expansion-to-be-overshadowed-by-2024-bitcoin-halving
It looks like we can have only short 4 years for tail emission or demurrage fix against network difficulty/security regression.
And I could even risk a statement, that the acceptance for such solution - is (slowly, but still) rising:
"There is currently no solution to this problem. Maybe a tail emission may be a necessary evil in the future to secure the network."
https://old.reddit.com/r/Bitcoin/comments/15fkcmv/profitability_of_validation_after_all_btc_is_mined/jue204r/
You will have my full-rbf support here... (i.e. on bitcoin-dev ;)
reply
very well done!
P.S. and now is good time for the longterm security budget challenge ;)
as soon as Bitcoin is only at half of gold inflation level - in worst case scenario we may not be able to see global network difficulty/hashrate not lower in 2028 than it was in 2024...
reply