04:45 < CubicEarth> shesek: sipa: Doesn't it also write to disk if the db cache fills up?

13:05 < bitcoin-git> [bitcoin] MarcoFalke pushed 2 commits to master: https://github.com/bitcoin/bitcoin/compare/8ab4f282c06d...76e2cded477b

13:05 < bitcoin-git> bitcoin/master ae7faf2 Kristaps Kaupe: Exceptions should be caught by reference, not by value.

13:05 < bitcoin-git> bitcoin/master 76e2cde MarcoFalke: Merge #16095: Catch by reference not value in wallettool

13:05 < bitcoin-git> [bitcoin] MarcoFalke merged pull request #16095: Catch by reference not value in wallettool (master...catch-by-ref-wallettool) https://github.com/bitcoin/bitcoin/pull/16095

15:06 < bitcoin-git> [bitcoin] promag opened pull request #16102: Improve base_blob<BITS>::IsNull() (master...2010-05-baseblob-isnull) https://github.com/bitcoin/bitcoin/pull/16102

15:10 < phantomcircuit> shesek, the fee estimates are based on the mempool

15:10 < phantomcircuit> so why?

15:38 < bitcoin-git> [bitcoin] promag opened pull request #16103: Remove unnecessary call to CTransaction::IsCoinBase() (master...2015-05-avoid-iscoinbase) https://github.com/bitcoin/bitcoin/pull/16103

16:47 < bitcoin-git> [bitcoin] jacekdalecki84 opened pull request #16104: 0.14 (0.15...0.14) https://github.com/bitcoin/bitcoin/pull/16104

16:48 < bitcoin-git> [bitcoin] jacekdalecki84 opened pull request #16105: Add files via upload (master...master) https://github.com/bitcoin/bitcoin/pull/16105

16:51 < bitcoin-git> [bitcoin] sipa closed pull request #16105: Add files via upload (master...master) https://github.com/bitcoin/bitcoin/pull/16105

16:52 < bitcoin-git> [bitcoin] sipa closed pull request #16104: 0.14 (0.15...0.14) https://github.com/bitcoin/bitcoin/pull/16104

18:15 < bitcoin-git> [bitcoin] promag opened pull request #16106: gui: Sort wallets in open wallet menu (master...2019-05-followup-15957) https://github.com/bitcoin/bitcoin/pull/16106

19:03 < shesek> phantomcircuit, I was under the impression that fee_estimates.dat has additional data that improves the fee estimates accuracy. why keep it around between restarts otherwise?

19:04 < sipa> they are

19:05 < luke-jr> shesek: it's not the mempool alone that determines fee estimates

19:05 < luke-jr> shesek: more like studying the pattern of transactions that *leave* the mempool as new blocks come in

19:05 < luke-jr> fee_estimates.dat keeps track of that

19:06 < sipa> the estimates are based on statistics of the time between entering the mempool and being mined

20:26 < phantomcircuit> sipa, oh

20:27 < phantomcircuit> oops