--- Day changed Fri Oct 11 2019 01:36 -!- MaxSan [~four@195.206.105.227] has joined #joinmarket 01:48 -!- MaxSan [~four@195.206.105.227] has quit [Ping timeout: 265 seconds] 01:48 -!- takamatsu [~takamatsu@unaffiliated/takamatsu] has joined #joinmarket 01:51 -!- MaxSan [~four@195.206.105.227] has joined #joinmarket 02:05 -!- MaxSan [~four@195.206.105.227] has quit [Quit: Leaving.] 02:28 -!- stoner19 [~stoner19@unaffiliated/stoner19] has quit [Remote host closed the connection] 02:28 -!- stoner19 [~stoner19@vmi221374.contaboserver.net] has joined #joinmarket 02:28 -!- stoner19 [~stoner19@vmi221374.contaboserver.net] has quit [Changing host] 02:28 -!- stoner19 [~stoner19@unaffiliated/stoner19] has joined #joinmarket 02:30 -!- lnostdal [~lnostdal@77.70.119.51] has quit [Ping timeout: 268 seconds] 02:35 -!- lnostdal [~lnostdal@77.70.119.51] has joined #joinmarket 03:04 -!- undeath [~undeath@hashcat/team/undeath] has joined #joinmarket 04:31 -!- MaxSan [~four@195.206.105.227] has joined #joinmarket 04:37 -!- MaxSan [~four@195.206.105.227] has quit [Quit: Leaving.] 04:42 < belcher_> waxwing_: now that iv finished coding #387 im going to spend some time reviewing #359 04:45 < waxwing_> belcher_, ok thanks. it's basically done, going to merge #413 which is a small bugfix, been testing #359 a lot on main net and i think it's fine now except perhaps one minor aberrant log message, so i'd want to merge it probably in the next several days. 04:45 < waxwing_> once i've fixed that log message i'll look at #387. 04:46 < waxwing_> not sure of the status of #391 , if takinbo and arubi can report on that and if something needs doing. 04:46 < waxwing_> AlexCato did you have any other reports on running #359 ? 04:47 < undeath> i'm running #359 but my maker isn't seeing much activity lately, so nothing to report 04:48 < AgoraRelay> [agora-irc/AlexCato] waxwing_: running 359 also for quite some time on main net, noticed no problems yet. Will have a look at the logs when joins were happenning to see if anything is of interest 04:51 < belcher_> ok fine, ill read the code anyway 04:53 < AgoraRelay> [agora-irc/AlexCato] i think i found a problem with #359 waxwing 04:54 < AgoraRelay> [agora-irc/AlexCato] when a join happens, the offers for makers are created new. It seems to also consider 0-conf UTXOs to be available for joins 04:56 < AgoraRelay> [agora-irc/AlexCato] example: mixdepth0=1 btc currently, mixdepth1=2btc. If a join with 0.5 btc happens (using mixdepth0 to fill), it immediately offers a new max of 2.5 BTC (since those 0.5 moved from mix0 to mix1) when the transaction is seen on the network, but BEFORE confirmation 04:59 < AgoraRelay> [agora-irc/AlexCato] oh. Or it might be that the 0-conf TX amount it still considered for the mixdepth0 amount. The specific case in my logs would fit both scenarios. 04:59 < AgoraRelay> [agora-irc/AlexCato] in any case, it does consider 0-conf UTXOs for the amonut available in in either the mixdepth they were spent from (mixdepth0 in example above) or to which they are spent (mixdepth1 in example above) 05:01 < AgoraRelay> [agora-irc/AlexCato] checked another join: i can say with certainty that it is the case described in the line starting with "example:" 05:01 < AgoraRelay> [agora-irc/AlexCato] so unconfirmed coins are already considered for the mixdepth-amount available in the mixdepth they were spent to 05:28 -!- MaxSan [~four@195.206.105.227] has joined #joinmarket 05:29 -!- MaxSan [~four@195.206.105.227] has quit [Client Quit] 05:31 -!- MaxSan [~four@195.206.105.227] has joined #joinmarket 05:35 -!- MaxSan [~four@195.206.105.227] has quit [Ping timeout: 268 seconds] 05:54 -!- MaxSan [~four@185.156.174.155] has joined #joinmarket 05:57 < waxwing_> AlexCato thanks, checking it out now 06:07 -!- MaxSan [~four@185.156.174.155] has quit [Quit: Leaving.] 06:11 < waxwing_> so to explain the situation, it's like this: 06:12 < waxwing_> in the current master (and earlier), we remove utxos when the tx is seen on the network (so unconf), but don't add the new ones. when confirmation happens, we do wallet.sync_unspent. 06:13 < waxwing_> which gives rise to a temporary period where we're offering less than we "have", but we don't advertise unconfirmed coins for join. however, there is `listunspentargs` which means that when we select for a join, we can select unconfirmed if that's set to [0] 06:14 < waxwing_> no hang on i think that last statement is wrong, let me check. 06:17 < waxwing_> yeah it's wrong because of the manual changes to our wallet's utxos; we first manually remove the spent ones at broadcast time, so obviously not offering new unconf utxos then, and then we do sync_unspent *only* when the confirmation has happened (sync_unspent uses listunspent_args), which means whatever the setting is doesn't matter; those coins are now confirmed. 06:18 < waxwing_> so in the new version, we basically process every transaction notification and add and remove utxos when they're seen, and that's happening under the hood; and callbacks happen if a callback is attached to a particular transaction. 06:19 < waxwing_> reference: https://github.com/JoinMarket-Org/joinmarket-clientserver/pull/359/files#diff-963259b630dee1562e2bcaa5de98e960R211 06:20 < waxwing_> so what to do, hmm, well perhaps simplest is just have a flag to process_new_tx and when we reach that line ^ in the unconf case, we just pass the flag and it only removes, but does not add. 06:20 < waxwing_> however i thought about that and it does have a bit of a drawback; it means coins don't get seen when they happen on the network. 06:20 < waxwing_> at the moment, if i deposit into my JMQt wallet i see the coins within a couple of seconds (i mean, with #359 ofc) 06:21 < waxwing_> that was kind of the main point (and it'll help with other features, like snicker, or basically anything that wants to use the wallet without manually interacting with it, so to speak) 06:22 < waxwing_> i think the only full answer is to attach confirmation metadata to the utxos. 06:23 -!- lnostdal [~lnostdal@77.70.119.51] has quit [Quit: "Fascism, Nazism, Communism and Socialism are only superficial variations of the same monstrous theme—collectivism." -- Ayn Rand] 06:28 -!- AgoraRelay [~jmrelayfn@p5DE4A6CC.dip0.t-ipconnect.de] has quit [Ping timeout: 268 seconds] 07:38 -!- puddinpop [~puddinpop@unaffiliated/puddinpop] has quit [Ping timeout: 250 seconds] 07:59 -!- lnostdal [~lnostdal@77.70.119.51] has joined #joinmarket 08:01 -!- somethinglittle is now known as nothingmuch 08:30 < arubi> waxwing_, right, I'll look into it tomorrow (sorry, I forgot where we start with it) 08:50 -!- bonswouar [~quassel@static-176-158-108-152.ftth.abo.bbox.fr] has joined #joinmarket 08:52 -!- bonswouar [~quassel@static-176-158-108-152.ftth.abo.bbox.fr] has left #joinmarket [] 10:09 -!- qubenix_ [~qubenix@66.172.11.228] has joined #joinmarket 10:10 -!- qubenix [~qubenix@66.172.11.228] has quit [Read error: Connection reset by peer] 10:10 -!- qubenix_ is now known as qubenix 12:05 < waxwing_> undeath, so i'm about to add a commit that adds confirmations to the utxo data; (txid, index): (path, value, confirmations) .. reason is basically above. but obviously one has to think about this: my reasoning is that you always sync to find utxos when you startup so the new format won't hurt anything. 12:06 < waxwing_> does that make sense to you? is there a case where the fact that the format of the 'value' in the dict has changed, will cause something to break? 12:29 < kristapsk> waxwing_, was just thinking about Qt not showing number of confirmations anywhere, adding them to utxo data would be useful there too in the future 12:39 -!- puddinpop [~puddinpop@unaffiliated/puddinpop] has joined #joinmarket 14:34 < belcher_> waxwing_ undeath how about instead of confirmations you add block height 14:34 < belcher_> height never changes unless there's a reorg 14:44 < kristapsk> belcher_, that's good idea, but handling of reorgs needs to be added to the code 15:13 -!- dmar198 [3f61084e@63.97.8.78] has joined #joinmarket 15:14 < dmar198> if I want to add more money to my joinmarket wallet, should I stop my yield generator first? 15:19 < kristapsk> dmar198, no, that's not required 15:20 < dmar198> awesome, and if I do that, will the yield generator then "know" (without restarting it) that it now has more funds to make available for coinjoins? 15:24 < kristapsk> dmar198, currently no, not until next cj or restart happens (+ incoming tx is confirmed), https://github.com/JoinMarket-Org/joinmarket-clientserver/pull/359 changes that 15:25 < dmar198> excellent, thank you 15:29 -!- lnostdal [~lnostdal@77.70.119.51] has quit [Ping timeout: 265 seconds] 15:40 -!- undeath [~undeath@hashcat/team/undeath] has quit [Quit: WeeChat 2.6] 16:10 -!- dmar198 [3f61084e@63.97.8.78] has quit [Remote host closed the connection] 18:34 -!- achow101 [~achow101@unaffiliated/achow101] has quit [Ping timeout: 265 seconds] 18:39 -!- achow101 [~achow101@unaffiliated/achow101] has joined #joinmarket 19:44 -!- achow101 [~achow101@unaffiliated/achow101] has quit [Ping timeout: 268 seconds] 19:58 -!- achow101 [~achow101@unaffiliated/achow101] has joined #joinmarket 20:12 -!- technonerd [~techno@gateway/tor-sasl/technonerd] has quit [Remote host closed the connection] 20:13 -!- technonerd [~techno@gateway/tor-sasl/technonerd] has joined #joinmarket 22:48 -!- pog [8dc1bdc1@141.193.189.193] has joined #joinmarket 23:16 -!- pog [8dc1bdc1@141.193.189.193] has quit [Ping timeout: 260 seconds]