--- Day changed Sun Sep 08 2019
01:14 -!- Giszmo [~leo@2407:7000:9d28:5100:5d26:c4d8:c4c1:6494] has joined #joinmarket     
01:32 -!- Sentineo [~Undefined@unaffiliated/sentineo] has joined #joinmarket     
01:55 -!- Sentineo [~Undefined@unaffiliated/sentineo] has quit [Ping timeout: 268 seconds]     
02:36 -!- MaxSan [~four@185.156.175.35] has joined #joinmarket     
03:14 -!- Giszmo [~leo@2407:7000:9d28:5100:5d26:c4d8:c4c1:6494] has quit [Quit: Leaving.]     
03:34 < fiatjaf> https://github.com/JoinMarket-Org/joinmarket-clientserver/blob/master/docs/USAGE.md#funding-wallet-and-displaying-balance
03:34 < fiatjaf> the listlabels call is returning {"code":-32601,"message":"Method not found"}
03:34 < fiatjaf> what can I do?
03:35 < fiatjaf> I've already updated my bitcoind
03:46 < takinbo> fiatjaf: what versions of bitcoind and joinmarket are you using?
03:49 -!- reallll [~belcher@unaffiliated/belcher] has joined #joinmarket     
03:49 < fiatjaf> bitcoind: 0.18.1
03:49 < fiatjaf> joinmarket I just downloaded from github, how can I see the version
03:50 < takinbo> can you confirm it wasn't built with the `--disable-wallet` flag?
03:52 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 258 seconds]     
03:52 < takinbo> if you cloned jm from github, that should be fine. the error though sounds like an issue from bitcoind and not joinmarket and my suspicion is that it was built without wallet support. 
03:53 < fiatjaf> I just downloaded the binaries
03:54 < fiatjaf> oops, actually I have disablewallet in my bitcoin.conf
03:54 < takinbo> ok. that might be it. 
04:17 < waxwing> "might" :)
04:37 < waxwing> as for that guy who wants to write about 'frontrunning' (or however you prefer to call it), i wanted to point him to that short discussion between belcher_ and gmaxwell from years ago, in case he comes back we can send him there.
04:45 < waxwing> https://github.com/JoinMarket-Org/joinmarket/issues/14#issuecomment-143509788 and he might find interesting my reference to it here: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2019-August/017208.html
05:20 < fiatjaf> I don't know how joinmarket works, what is tumbler or mixdepth. I'm about to fund the wallet. should I send a transaction to 3 addresses of mixdepth 0 and that will do it?
05:21 < fiatjaf> I'm reading the guide and the link about "sourcing" stuff, but I don't understand it
05:21 < waxwing> yes fiatjaf
05:21 < waxwing> you mean you followed the link to 'sourcing commitments for joins'. yes, it's a bit obscure, which is why that page prints in bold the only things you really need to know.
05:22 < waxwing> btw the 3 addresses thing (the 'sourcing commitments for joins') only matters if you are creating coinjoins yourself (being a "Taker", which is what tumbler.py and sendpayment.py do), not if you're being a "Maker" (yieldgenerator).
05:22 < waxwing> for the latter just sending to 1 address is fine.
05:23 < fiatjaf> I want to experiment being a taker and then become a maker
05:24 < fiatjaf> when I'm sending a payment it will use all my utxos?
05:26 < waxwing> fiatjaf, in that aspect it's like other wallets. it decides which utxos to use based on the amount.
05:27 < fiatjaf> what is mixdepth?
05:27 < waxwing> the main overarching concept to remember is, each mixdepth(=account) is separate.
05:27 < fiatjaf> sorry if I'm asking stuff that's already written
05:27 < fiatjaf> oh, it doesn't have a meaning besides account separation?
05:27 < waxwing> think of it like 5 pockets/accounts whatever. don't co-spend utxos from two mixdepths at once (no worry, the wallet won't allow it)
05:28 < waxwing> movements of your funds between different mixdepths happen via coinjoins, which obfuscates the linkage between them.
05:28 < waxwing> which is particularly important for running tumbler, and tumbler is the primary way that joinmarket tries to make your coin ownership private.
05:29 < waxwing> sendpayment is just something you can use opportunistically here and there (just do one coinjoin), see the comment about that at the top of the section on sendpayment.py
06:37 < fiatjaf> you mean I should do sendpayment to myself whenever fees are low?
06:37 < fiatjaf> coins at higher mixdepths are more anonymous?
06:47 -!- puddinpop [~puddinpop@unaffiliated/puddinpop] has joined #joinmarket     
07:42 < waxwing> read the tumbler guide and wiki page linked at the end of the usage guide for how to use tumbler.py to make sequences of coinjoins; use sendpayment.py to do a single coinjoin whenever you like - the destination address is up to you; it can be either another address in a *different* mixdepth (use from external branch), or a remote destination to a different wallet or a payment.
07:42 < waxwing> you can basically do anything you like, just don't do coinjoins where the destination address is in the same mixdepth.
07:43 < waxwing> think of your joinmarket as 5 different wallets (accounts). moving between them only via coinjoin, spending out of the wallet however you like (including coinjoin, if you like)
07:43 < waxwing> fiatjaf, ^
07:51 < kristapsk> fiatjaf, in case you run yield generator and deposit into mixdepth 0, yes, higher mixdepths likely will have bigger anonimity (because mixes usually go from mixdepth 0 to 1, from 1 to 2, etc)
07:53 < waxwing> over the long term there's no correlation between "depth" and number of coinjoins an output has gone through, though.
07:53 < waxwing> which is why belcher_ has often commented he regrets the term "mix*depth*" a bit; it's misleading, although it might temporarily be true.
07:55 < fiatjaf> I did a sendpayment.py with 6 counterparties
07:55 < fiatjaf> now where are the coins?
07:55 < fiatjaf> they don't seem to have appeared in my destination wallet
07:56 < fiatjaf> can I see the transactions involved?
07:58 < waxwing> did you use Qt or CLI?
07:58 < fiatjaf> cli     
07:58 < waxwing> you should have seen a txid at the end of the output. you can check that on a block explorer.
07:59 < waxwing> assuming it did actually go through; if it didn't you can check the joinmarket wallet still has the coins.
08:27 < fiatjaf> oh, sorry, it was asking me to confirm before actually sending the transactions and I didn't see it
08:27 < waxwing> heh, i guessed as much :)
08:28 < waxwing> the point there is that you want to know exactly how much  you're being charged (although you can strictly limit it with the setting it prompts you about at the start, to be fair).
08:28 < waxwing> there's also a --yes option if you don't want it, but i wouldn't recommend it really.
08:30 < fiatjaf> 2019-09-08 12:29:33,438 [INFO]  total cj fee = 138                                                                    
08:30 < fiatjaf> 2019-09-08 12:29:33,439 [INFO]  total coinjoin fee = 0.0445%
08:31 < fiatjaf> so I'll pay 138 sat in coinjoin fees, but how much in transaction fees?
08:31 < fiatjaf> total estimated amount spent = 350138
08:32 < fiatjaf> I tried to send 310000
08:32 < fiatjaf> Based on initial guess: 40000, we estimated a miner fee of: 827
08:32 < fiatjaf> fee breakdown for me totalin=369000 my_txfee=727 makers_txfee=100 cjfee_total=218 => ch
08:32 < fiatjaf> angevalue=58055
08:33 < fiatjaf> so I guess miner fees weren't shown to me before, but calculated on the go
08:36 < fiatjaf> so I got a transaction with 3 inputs from me and 3 from other people, 4 outputs equal to my desired destination output, 1 change output and 3 outputs with random amounts to somewhere else. 
08:37 < fiatjaf> isn't that good enough? in my complete naiveté I would consider everything here successfully coinjoined and hidden and unlinkable.
08:37 < fiatjaf> ok, now I'll read more
09:19 < kristapsk> great news, current Bitcoin Core master breaks JoinMarket - https://github.com/bitcoin/bitcoin/issues/16832
10:06 < waxwing> fiatjaf, yes calculated on the go because unlike a normal wallet, we don't know how many inputs the transaction has before we start. that's the nature of coinjoin.
10:06 < waxwing> kristapsk, oh dear. thanks for heads up.
10:10 < kristapsk> found it by accident, did testing on one Core PR, and had some testnet coins in JM wallet, which I wanted to transfer :)
10:49 -!- kristapsk [~KK@gateway/tor-sasl/kristapsk] has quit [Remote host closed the connection]     
10:50 < fiatjaf> how bad is the idea of writing a web UI for joinmarket?
10:56 -!- kristapsk [~KK@gateway/tor-sasl/kristapsk] has joined #joinmarket     
11:26 < kristapsk> waxwing, it isn't issue with new Bitcoin Core master, the same behaviour is with v0.18, it happens if multiple wallets are loaded in Bitcoin Core
11:27 < kristapsk> guess we should handle this somehow in JM code
11:27 < kristapsk> for example, assume default wallet (wallet.dat) and add some joinmarket.cfg option to specify different wallet, if user wants so
11:28 < waxwing> kristapsk, we have that
11:28 < waxwing> rpc_wallet_file in the config
11:28 < waxwing> i use it myself
11:29 < waxwing> i'm on 0.18 too, just checked
11:30 < kristapsk> ok, then just adding some try-block's at some places looks right thing to do, as currently one could use JM without problems, then creates and loads another wallet in core, and JM starts crashing with exceptions
11:33 < waxwing> kristapsk, oh, do you mean: if you create a new wallet while JM is still running, it crashes?
11:34 < kristapsk> no     
11:34 < waxwing> or is it like: if you don't set the rpc_wallet_file in joinnmarket.cfg, but you add a new Core wallet, then it crashes?
11:34 < kristapsk> JoinMarketQt crashes on JM wallet load in case there is more than one wallet loaded in Core 
11:35 < kristapsk> yes, the second option
11:35 < kristapsk> and that second wallet does need to be current, your selected one could be the same
11:36 < kristapsk> in any case, crashing with exception is bad, there should be either error message for humans or handling of such cases
11:38 < waxwing> right
11:38 < waxwing> thanks for noticing it
12:38 < waxwing> kristapsk, the issue is one of defaults i presume, see https://github.com/JoinMarket-Org/joinmarket-clientserver/blob/master/jmclient/jmclient/jsonrpc.py#L65-L68
12:38 < waxwing> i presume that if the Core instance hasn't had a wallet added then it's OK, but if it has, then there's an insistence that all calls use the /wallet/ path in the request.
12:38 < kristapsk> waxwing, yes, I kinda have fix already
12:39 < kristapsk> try listwallets and specify first returned in JsonRpc contructor
12:39 < kristapsk> assuming it is the default
12:39 < kristapsk> if rpc_wallet_file is not specified in joinmarket.cfg
12:40 < kristapsk> in v0.16.3 my default wallet is "wallet.dat", in v0.18 in testnet it's "", so can't be hardcoded
12:44 < waxwing> oh     
12:44 < waxwing> i see thanks
12:58 < kristapsk> waxwing, https://github.com/JoinMarket-Org/joinmarket-clientserver/pull/395
13:18 -!- Netsplit *.net <-> *.split quits: so
13:19 -!- Netsplit over, joins: so
13:34 < takinbo> i once had the issue kristapsk is referring to but mine was different, if you have the default wallet loaded with an empty name and others also loaded with different names, jm crashed if you didn't specify a wallet name in `rpc_wallet_file`
13:34 < takinbo> bascially, it wouldn't work with an empty file name if you intended to use the default wallet without a name
13:36 < takinbo> the solution i think i implemented was to use `self.url = '/wallet/'` without the `wallet_file` parameter.
13:38 < takinbo> kristapsk's solution will load the first wallet if none is provided and there are multiple wallets loaded. i suspect it has the same effect as loading the default wallet with no name.
13:38 < kristapsk> ahh, yes, there is len() > 0 check :)
13:40 < takinbo> assuming all the wallets have a name, this has the side effect of loading the first wallet which might not be what the user wants
13:40 < takinbo> maybe it should fail with a error rather than an exception and provide some guidance for the user to explicitly configure the wallet they want to use?
13:51 < kristapsk> I think advanced users actively using multiple wallets should know what they are doing and specify rpc_wallet_file in config
13:51 < kristapsk> this will not work as expected only if default wallet is closed
13:53 < kristapsk> and there isn't risk of losing funds, it will just import all JM addresses as watch-only in the wrong wallet and ask user for rescan
13:53 < kristapsk> but I think it's better to discuss this on github, not here
13:53 < takinbo> certainly
14:27 -!- reallll is now known as belcher     
15:02 -!- Zenton [~user@unaffiliated/vicenteh] has quit [Ping timeout: 264 seconds]     
17:57 -!- CgRelayBot [~CgRelayBo@p5DE4AE41.dip0.t-ipconnect.de] has quit [Ping timeout: 245 seconds]     
17:58 -!- AgoraRelay [~jmrelayfn@p5DE4AE41.dip0.t-ipconnect.de] has quit [Ping timeout: 258 seconds]     
18:09 -!- CgRelayBot [~CgRelayBo@p5DE4A2DC.dip0.t-ipconnect.de] has joined #joinmarket     
18:11 -!- AgoraRelay [~jmrelayfn@p5DE4A2DC.dip0.t-ipconnect.de] has joined #joinmarket     
19:51 -!- raedah [~x@192.30.89.51] has joined #joinmarket     
22:32 -!- Giszmo [~leo@2407:7000:9d28:5100:5d26:c4d8:c4c1:6494] has joined #joinmarket     
22:39 -!- Sentineo [~Undefined@unaffiliated/sentineo] has joined #joinmarket     
23:51 -!- Ulises68Trantow [~Ulises68T@ns334669.ip-5-196-64.eu] has joined #joinmarket