--- Day changed Fri Mar 15 2019
00:06 -!- cohibaa [~CohibAA@gateway/tor-sasl/cohibaa] has quit [Quit: cohibaa]     
00:08 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 250 seconds]     
00:30 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
02:11 -!- undeath [~undeath@hashcat/team/undeath] has joined #joinmarket     
03:10 < takinbo> is it possible to do a payjoin and a coinjoin in the same transaction?
03:17 < undeath> technically yes, but not implemented afaik
06:30 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 252 seconds]     
07:12 -!- puddinpop [~puddinpop@unaffiliated/puddinpop] has joined #joinmarket     
07:15 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
07:43 < takinbo> great. thanks undeath 
07:45 < waxwing> takinbo, yeah good idea really, and precisely because it doesn't make sense :) (i.e. the idea was to have payjoin be indistinguishable from ordinary non-coinjoin payments)
07:45 < takinbo> ah, yes. definitely.
07:54 < undeath> it gives a whole new level to coinjoin's subset sum problem
09:06 < waxwing> undeath, i'm a bit confused why the storage has an entry for 'utxo' currently (inserted by the utxomanager), is it always empty (as it seems to be in my active wallet)?
09:12 < undeath> the utxomanager is initialized on wallet creation but not actually stored
09:19 < undeath> utxomanager.save() is implemented but not called. (this was removed, there was a discussion in the pr iirc)
09:20 < waxwing> thanks, i'll keep reading
09:20 < undeath> or maybe it was just discussed here
09:26 < waxwing> undeath, heh, maybe, i just checked #104 it wasn't there anyway.
09:26 < waxwing> can you outline what you do remember, in terms of reasoning around this?
09:27 < waxwing> the reason it comes up, in case it isn't obvious, is i've been writing some code to allow a user to disable a utxo, but obviously it'll need to be persisted.
09:27 < undeath> it was removed because we always sync and syncing is bound to be more accurate
09:27 < waxwing> rather ignorantly i was adding it as feature to the utxo manager before realising that it isn't persisted at all, heh.
09:27 < waxwing> right
09:28 < undeath> technically it wouldn't break anything to save the utxomanager and simply clearing the data upon syncing
09:29 < waxwing> yes, it's that now we have a different use-case for it. i'm calling it 'metadata' on the utxos with the idea that we might have other stuff to associate with utxos later (obvious one being labels)
09:30 < waxwing> but anyway, i have to keep reading, i often find there's some aspect of the logic of the program that completely escapes me if i haven't read it for a few months :)
09:32 < waxwing> like specifically right now, why is the maker/yieldgenerator code not calling add_new_utxo(_) when a transaction arrives?
09:33 < undeath> you mean add_utxo()?
09:34 < waxwing> wallet.add_new_utxo(_)
09:34 < waxwing> sorry with an 's', utxos
09:35 < undeath> ah     
09:35 < waxwing> cf https://github.com/JoinMarket-Org/joinmarket-clientserver/blob/master/jmclient/jmclient/taker_utils.py#L256-L257
09:36 < waxwing> this is what the tumbler does between transactions to update its utxo-set
09:36 < undeath> yeah, I see
09:36 < undeath> good question
09:36 < undeath> but things seem to work, so somehow it has to work
09:37 < waxwing> remove_old_utxos is called in the jmclient.client_protocol.JMMakerClientProtocol.(some function about unconfirm update)
09:37 < waxwing> unconfirm_callback
09:40 < waxwing> undeath, ah ok i remembered/saw it
09:41 < waxwing> we don't re-apply the new utxos until they're confirmed.
09:41 < undeath> how is it done?
09:41 < waxwing> in the confirm_callback it calls sync_unspent
09:41 < waxwing> which does call add_utxo (s)
09:41 < undeath> heh     
09:41 < waxwing> so the thing i can't remember is why we do it differently in tumbler. i guess because it's non-parallel it's safe to do it "in ram" so to speak.
09:42 < undeath> sync_unspend sounds like a pretty expensive call (not having looked at the code)
09:43 < waxwing> yes, there's probably some interesting questions there. if volume were very high it could be an issue. obv it's the safest, though.
09:44 < waxwing> so any thoughts on persisting 'utxo_metadata', basically, in the utxomanager. we could have it save to disk but then as currently, wipe the actual utxo list and get from sync.
09:47 < belcher> sync_unspent() probably calls the RPC listunspent which is quite cheap
09:48 < belcher> waxwing maybe save information that says "if this utxo exists, then mark is unspendable"
09:48 -!- devachandra [~devachand@vmi179228.contaboserver.net] has quit [Quit: devachandra]     
09:48 < waxwing> belcher, yup i think that code is unchanged from when you wrote it :) still has that timing check which of course is funny.
09:49 < waxwing> belcher, yes, i agree (if your point is: we can mark it as unspendable and don't really need to track if it's since been spent)
09:49 < belcher> the timing check was part of like 20 things i tried to make yieldgenerators more stable back then
09:50 < waxwing> hmm i just noticed we have that literal_eval in that function for the arguments. can you crash bitcoind with rpc arguments? :)
09:50 < waxwing> i mean it's just crashing your own node, but still :)
09:51 < belcher> i wonder why literal eval was chosen rather than just minconfirm, since i think the point of that code is to allow unconfirmed utxos to be seen/spent
09:52 < waxwing> well but adlai gave in the comments 3 or 4 possible settings i think. it's not just 0/not 0 eh.
09:53 < belcher> ok     
09:55 < waxwing> but, prob i think you're right. a minimum single number is probably enough (i guess almost nobody pays attention to it except it being 0 or not)
10:00 < belcher> its not great safety issue since its only exposed to the user and most people will leave the default
10:02 < waxwing> no indeed, not really an issue, agreed.
10:14 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 245 seconds]     
10:28 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
12:23 -!- kristapsk [~neonz@62.63.157.78] has quit [Quit: Leaving]     
14:35 -!- undeath [~undeath@hashcat/team/undeath] has quit [Quit: WeeChat 2.4]     
15:32 -!- belcher_ [~user@unaffiliated/belcher] has quit [Ping timeout: 245 seconds]     
15:42 -!- lnostdal [~lnostdal@77.70.119.51] has quit [Quit: https://www.Quanto.ga/ -- Immutable state trading system in Clojure]     
15:43 -!- belcher_ [~user@unaffiliated/belcher] has joined #joinmarket     
15:51 -!- lnostdal [~lnostdal@77.70.119.51] has joined #joinmarket     
18:56 -!- AgoraRelay [~jmrelayfn@p5DE4ABEA.dip0.t-ipconnect.de] has quit [Ping timeout: 252 seconds]     
19:09 -!- AgoraRelay [~jmrelayfn@p54866F4E.dip0.t-ipconnect.de] has joined #joinmarket