Post AQ48Vhe6v9vRFxefAW by dflate@bitcoinhackers.org
 (DIR) More posts by dflate@bitcoinhackers.org
 (DIR) Post #AQ3E9kQaIpK052HZUu by lukedashjr@bitcoinhackers.org
       2022-11-28T04:20:08Z
       
       1 likes, 1 repeats
       
       PSA: #Bitcoin itself has _always_ been Full Replace-By-Fee (RBF), and if you implement anything else at the consensus layer, you will fork yourself off (likely within a few days). RBF is and has always been a hard _requirement_.1) No matter what your node policy is, you MUST replace any unconfirmed transaction with a conflicting transaction in a block.2) If a block gets reorg'd with another including a conflicting tx, you MUST accept the replacement.
       
 (DIR) Post #AQ3EBUlZVEHU5OvrCi by lukedashjr@bitcoinhackers.org
       2022-11-28T04:20:28Z
       
       0 likes, 1 repeats
       
       To remain on the #Bitcoin network, your node must reorg out one block in favour of another, if the replacement chain has a higher "fee" (proof-of-work).You can and have the right to implement whatever policy you want for unconfirmed transactions in your node, but at the end of the day, the RBF policy aligns with the consensus rules. FSS and other policies are a _deviation_ the protocol does not (and cannot) prevent.
       
 (DIR) Post #AQ3FJRgCxhxamgmJTE by kps@mastodon.online
       2022-11-28T04:33:05Z
       
       0 likes, 0 repeats
       
       @lukedashjr so for a Lightning node running Bitcoin core, what setting is recommended for full RBF. Use default or adjust Bitcoin.conf mempoolfullrbf option?
       
 (DIR) Post #AQ3UqBez2hjZy7HAbw by lukedashjr@bitcoinhackers.org
       2022-11-28T07:27:06Z
       
       0 likes, 0 repeats
       
       @kps Better to use Knots. ;)I prefer full RBF, but any policy should work.
       
 (DIR) Post #AQ48Vhe6v9vRFxefAW by dflate@bitcoinhackers.org
       2022-11-28T14:51:35Z
       
       0 likes, 0 repeats
       
       @lukedashjr not that it made much sense for the grown mature net and almost centralized mining we have now, but slowing mempool and the subsequent net spamming by forcing standard nodes to not relay, made some sense.Miners could had always mined blocks of any set of signed tx, but no one was then assumed default to relay in advance such replacements.Statistically interesting, as a means of defying possible double spent attempts, by damping tx propagating and indirect flagging bogus miners ip's