b10c's avatar
b10c
b10c@b10c.me
npub1kyxq...c6vv
Bitcoin developer and observer
b10c's avatar
b10c 1 month ago
Hey @Nick - I wanted to have a look at your Bitcoin Node census data, but the site seems to be down. Is this related to cloudflare?
b10c's avatar
b10c 1 year ago
25 blocks to go, halving stream is live
b10c's avatar
b10c 1 year ago
Looking at the merkle branches that mining pools send to miners as part of stratum jobs, it's clear that the BTCcom pool, Binance pool, Poolin, EMCD, Rawpool, and possibly Braiins* have exactly the same template and custom transaction prioritization as AntPool. image Stratum v1 jobs include (parts of) the coinbase transaction and the merkle branches a miner needs to compute the merkle root. If all branches are equal across pools, the block template is exactly the same. This is a smaller merkle-tree with the stratum job branches highlighted. image Here, branch0 is the txid of the first transaction in the block after the coinbase. In the AntPool/Binance/BTCcom/.. template, it's a manually prioritized transaction paying 9.03 sat/vbyte Many of the other pools didn't prioritize a transaction and included the highest feerate transaction paying 301 sat/vbyte. ViaBTC prioritized another transaction and has a vastly different template. This indicates that F2Pool currently is not using the AntPool templates, even if they share the the same custody provider. We know from multiple occasions that F2Pool has been building their own templates (https://b10c.me/o11, https://b10c.me/o9, https://b10c.me/o8) Note that SigmaPool and SecPool both seem to share the same template (distinct from AntPools template). Two names, but probably one pool. Some overlap in the first branches of distinct pools is expected as these branches consist of few (usually high feerate) transactions. For example, Ocean's "default" and "core" templates overlap up to branch3. At least the first 15 tx (2⁴ - 1) in the template are equal. * similar to EMCD in the picture, Braiins sometimes differs from AntPool in the later branches of the template. This could indicate they're only helping with AntPools transaction accelerator and directly syncing with AntPools mempool, causing occasional differences in templates. I've heard speculation about this all being liked to an FPPS partnership with AntPool/Bitmain. Bitmain insures against bad pool luck but you have to use Bitmains templates/transaction prioritization and pay the mining reward to the insurer for later distribution. Essentially a pool of pools and a lot of mining centralization.