19:00:39 <achow101> #startmeeting 
19:00:39 <core-meetingbot> Meeting started Fri Jun 17 19:00:39 2022 UTC.  The chair is achow101. Information about MeetBot at https://bitcoin.jonasschnelli.ch/ircmeetings.
19:00:39 <core-meetingbot> Available commands: action commands idea info link nick
19:00:43 <achow101> #bitcoin -core-dev Wallet Meeting: achow101 _aj_ amiti ariard BlueMatt cfields Chris_Stewart_5 darosior digi_james dongcarl elichai2 emilengler fanquake fjahr furszy gleb glozow gmaxwell gwillen hebasto instagibbs jamesob jarolrod jb55 jeremyrubin jl2012 jnewbery jonasschnelli jonatack josibake jtimon kallewoof kanzure kvaciral laanwj larryruane lightlike luke-jr maaku marcofalke meshcollider michagogo moneyball morcos Murch nehan NicolasDorier
19:00:43 <achow101> paveljanik petertodd phantomcircuit promag provoostenator ryanofsky sdaftuar S3RK sipa vasild
19:00:50 <instagibbs> hi
19:00:53 <Murch> Hi
19:01:02 <provoostenator> hi
19:01:03 <furszy> hi
19:01:07 <achow101> There are no pre-proposed wallet meeting topics. Does anyone have anything to discuss?
19:01:23 <provoostenator> Not really
19:01:30 <kanzure> hi
19:02:20 <achow101> any prs we should be prioritizing?
19:04:00 <achow101> or does anyone want to talk about what they're working on?
19:04:11 <fjahr> hi
19:05:59 <Murch> I have found a coin selection approach that looks very promising in that it outperforms anything with Knapsack except in keeping the UTXO pool tiny.
19:06:06 <fjahr> maybe #25351 could get a 24.0 milestone tag? The predecessor had 21/22 tags.
19:06:08 <gribble> https://github.com/bitcoin/bitcoin/issues/25351 | rpc, wallet: Scan mempool after import* - Second attempt by fjahr · Pull Request #25351 · bitcoin/bitcoin · GitHub
19:06:18 <Murch> However, Knapsack is notorious for being too aggressive in consolidating.
19:07:12 <achow101> fjahr: added to the milestone
19:07:30 <instagibbs> Murch, still spends negative effective value, yeah?
19:07:32 <fjahr> achow101: thanks!
19:07:33 <achow101> Murch: Is the change in utxo pool explainable by not spending negative ev?
19:08:01 <Murch> Working on writing it up so that I hopefully can convince it does better or equally well in other regards while removing a bunch of worst case-outcomes we've seen in the past.
19:08:46 <Murch> instagibbs: My idea is to allow that only at feerates < 2 ṩ/vB, while adding a FIFO selector.
19:08:58 <Murch> That way it does happen, but not when it hurts by paying 10× as much than necessary
19:10:01 <Murch> Knapsack would often just tack it on with the next best transaction (since it reduces the change amount, and knapsack optimizes for smallest change)
19:10:33 <Murch> So, be it 10 ṩ/vB or be it 200 ṩ/vB, knapsack would grab it.
19:11:05 <achow101> have you simulated at low and high fee rates?
19:11:36 <Murch> Yeah, I have simulated with the peak-and-tail and the 2019–2020 scenario
19:12:26 <achow101> what about different scenaiors?
19:12:53 <Murch> I used the tiny bustabit scenario, though, so it's from the perspective of a wallet that gets twice as many deposits as it makes payments. I still need to run on other scenarios, although they're generally easier, because they are less imbalanced towards building a huge UTXO pool
19:14:19 <Murch> Anyway, if someone wants to seed some concerns that I should address, I'd be happy to hear you what your bar for convincing a change in the coin selection strategies would be ;)
19:15:35 <achow101> cool, any other topics?
19:16:58 <furszy> from my side, have been implementing your suggestion in 25297
19:17:30 <furszy> and.. jumping across the sources a bit
19:18:21 <furszy> found few things doing it
19:19:15 <achow101> furszy: that's great, looking forward to those changes
19:19:59 <furszy> yeah :), I still owe you finish 24914 review. It's at the top of my list.
19:21:44 <achow101> #endmeeting