fix rbf calculation - fixes #2010

[psq]

Fixes #2010. Current calculation overpays by a factor of 2048 (should use attempt-1 as multiplier, and fee is per kb, not per byte

[jcnelson]

For good measure, can you change attempt - 1 to attempt.saturating_sub(1) so it doesn't ever underflow?

[jcnelson]

Also, can you group the numerator explicitly? I.e. ((attempt.saturating_sub(1) * self.last_tx_len * self.min_relay_fee) / 1024);

[psq]

good idea. addressed in 00d7880

[jcnelson]

Thanks for catching this @psq! Are you able to mine blocks on the testnet with these settings?

Once my comment is addressed this will LGTM.

[psq]

Are you able to mine blocks on the testnet with these settings

yes, the miner is mining, but I'm no sure it is the second transaction is the one accepted when there is a second attempt based on how the UTXO seem to get updated, so I'll dig deeper tomorrow.

[psq]

seems to work fine, I changed the code to send 2 sats more per new attempt, and the replaced transaction is the one that gets committed, not the original one. So all good.

bitcoin-cli -regtest -conf=/Users/psq/src/perso/stacks-blockchain/bitcoin.conf -rpcport=19443 gettransaction f39adee67e803ee2faeec0d6733c8f6b905f34881e2d8ebdca7aeac8dc90d598 true
{
  "amount": -0.00022002,
  "fee": -0.00005852,
  "confirmations": 446,
  "blockhash": "291bf835aafebd56eb8e6529f821c82b1815fc61e16a407243520df9cb9c8e36",
  "blockheight": 3015,
  "blockindex": 7,
  "blocktime": 1604047618,
  "txid": "f39adee67e803ee2faeec0d6733c8f6b905f34881e2d8ebdca7aeac8dc90d598",
  "walletconflicts": [
  ],
  "time": 1604047618,
  "timereceived": 1604100263,
  "bip125-replaceable": "no",
  "details": [
    {
      "involvesWatchonly": true,
      "category": "send",
      "amount": 0.00000000,
      "vout": 0,
      "fee": -0.00005852,
      "abandoned": false
    },
    {
      "involvesWatchonly": true,
      "address": "mfWxJ45yp2SFn7UciZyNpvDKrzbhyfKrY8",
      "category": "send",
      "amount": -0.00011001,
      "vout": 1,
      "fee": -0.00005852,
      "abandoned": false
    },
    {
      "involvesWatchonly": true,
      "address": "mfWxJ45yp2SFn7UciZyNpvDKrzbhyfKrY8",
      "category": "send",
      "amount": -0.00011001,
      "vout": 2,
      "fee": -0.00005852,
      "abandoned": false
    },
    {
      "involvesWatchonly": true,
      "address": "mxtGS9DHV2TafucTmcDJhNipRZ7svFoi7C",
      "category": "send",
      "amount": -0.40695581,
      "label": "",
      "vout": 3,
      "fee": -0.00005852,
      "abandoned": false
    },
    {
      "involvesWatchonly": true,
      "address": "mxtGS9DHV2TafucTmcDJhNipRZ7svFoi7C",
      "category": "receive",
      "amount": 0.40695581,
      "label": "",
      "vout": 3
    }
  ],
  "hex": "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"
}

you can see 11_001 per address instead of 11_000, and the first transaction is not found.

[jcnelson]

LGTM. Thanks @psq!

[kantai]

LGTM!

[blockstack-devops]

This pull request has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.