--- Day changed Mon Oct 09 2017
02:31 -!- Randall [~Randall@ns334669.ip-5-196-64.eu] has joined #joinmarket     
02:44 -!- coins123_ [~coins123@ip-244-225.sn1.clouditalia.com] has joined #joinmarket     
02:45 -!- coins123_ [~coins123@ip-244-225.sn1.clouditalia.com] has quit [Client Quit]     
02:47 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Ping timeout: 240 seconds]     
02:48 < GitHub68> [cjhunt] adlai pushed 1 new commit to master: https://git.io/vdgQt
02:48 < GitHub68> cjhunt/master 9806efb Adlai Chandrasekhar: how the fuck did this improve results
03:03 -!- coins123 [~coins123@ip-244-225.sn1.clouditalia.com] has joined #joinmarket     
03:03 -!- coins123 [~coins123@ip-244-225.sn1.clouditalia.com] has quit [Changing host]     
03:03 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket     
03:04 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Remote host closed the connection]     
03:13 < GitHub13> [cjhunt] adlai pushed 1 new commit to master: https://git.io/vdgda
03:13 < GitHub13> cjhunt/master ec4f49c Adlai Chandrasekhar: a more understandable improvement
03:14  * adlai has yet to find a sweep cj that doesn't produce a "credible" grouping, and yet to find a send cj that does
03:35 -!- coins123 [~coins123@195.32.46.14] has joined #joinmarket     
03:35 -!- coins123 [~coins123@195.32.46.14] has quit [Changing host]     
03:35 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket     
04:07 -!- arubi [~ese168@unaffiliated/ese168] has quit [Quit: bye]     
04:10 -!- arubi [~ese168@gateway/tor-sasl/ese168] has joined #joinmarket     
04:10 -!- wxxs [~chatzilla@212.84.170.109] has joined #joinmarket     
05:53 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 255 seconds]     
06:06 -!- belcher [~belcher@unaffiliated/belcher] has joined #joinmarket     
06:23 -!- puddinpop_u [~puddinpop@unaffiliated/puddinpop] has quit [Ping timeout: 258 seconds]     
06:26 < waxwing> adlai, weird (not that i'm reading the code but..) sends dont' result in plausible subset sum solutions? how's that?
06:54 -!- zxccxz [5db781f6@gateway/web/freenode/ip.93.183.129.246] has quit [Quit: Page closed]     
08:20 -!- beIcher [~user@unaffiliated/belcher] has quit [Ping timeout: 246 seconds]     
09:20 -!- beIcher [~user@unaffiliated/belcher] has joined #joinmarket     
09:25 -!- beIcher [~user@unaffiliated/belcher] has quit [Ping timeout: 255 seconds]     
09:29 -!- beIcher [~user@unaffiliated/belcher] has joined #joinmarket     
09:33 -!- beIcher [~user@unaffiliated/belcher] has quit [Ping timeout: 240 seconds]     
09:37 -!- beIcher [~user@unaffiliated/belcher] has joined #joinmarket     
09:50 -!- beIcher [~user@unaffiliated/belcher] has quit [Ping timeout: 240 seconds]     
09:56 -!- zxccxz [5db781f6@gateway/web/freenode/ip.93.183.129.246] has joined #joinmarket     
09:59 -!- beIcher [~user@unaffiliated/belcher] has joined #joinmarket     
10:16 -!- frank_ [5b566e51@gateway/web/freenode/ip.91.86.110.81] has joined #joinmarket     
10:17 < frank_> Hello is someone here ??? I need help fast
10:18 < arubi> what's the problem frank_ ?
10:19 < frank_> arubi I coinjoined some btc using tumbler but the script stopped unexpectedly, now when I use wallet-tool I see money in an internal address but there isn't any real money on it
10:19 < arubi> what does it mean that there isn't real money?
10:20 < frank_> in blockchain.info there's no transaction with the address
10:21 < frank_> it's been 5-6 days and I already rescanned
10:22 < arubi> and you're using bitcoin core as a source for the transactions correct
10:22 < arubi> ?     
10:22 < frank_> yes, I do
10:23 < arubi> can you run `bitcoin-cli listunspent 0 9999999 <your address>` ?
10:23 < frank_> the address with the wrong balance ?
10:23 < arubi> yes     
10:24 < frank_> Ok wait a sec
10:24 < arubi> I don't need the output, just if there is output at all, a txid should be there
10:26 < frank_> I get error parsing json
10:27 < arubi> don't paste the < >, if you did
10:27 < arubi> just the address
10:27 < frank_> I didn't
10:27 < frank_> but without the address it prints a json array
10:27 < arubi> oh sorry, probably `bitcoin-cli listunspent 0 9999999 '["your address"]'`
10:27 < arubi> without the address isn't interesting
10:27 < arubi> also with changing anything else turns it into not interesting
10:28 < frank_> Ok we have to pass an array
10:28 < frank_> still the same error
10:28 < frank_> I check the api
10:29 < arubi> bitcoin-cli help listunspent
10:32 < frank_> ok the problem was with powershell
10:32 < frank_> I get an array with one object
10:32 < frank_> txid, vout, address, ...
10:34 < arubi> okay, does it say how many confirmations on that txid?
10:34 < frank_> 0     
10:34 < arubi> cool, then you should be able to abandontransaction
10:34 < frank_> how do I do this
10:34 < arubi> `bitcoin-cli abandontransaction <txid>`
10:35 < frank_> Transaction not eligible for abandonment
10:36 < arubi> sigh.  okay, let's try restarting bitcoind with the -blocksonly=1 flag
10:36 < frank_> Ok I do it right now
10:36 < arubi> er wait lemme see if that's the actual flag
10:36 < arubi> maybe misremembering
10:37 < frank_> why everytime i ctrl-c bitcoind, it crashes?
10:37 < arubi> yes -blocksonly=1
10:37 < arubi> you need to run `bitcoin-cli stop`
10:37 < arubi> not ctrl+c the daemon window
10:37 < frank_> Ah didn't know :)
10:38 < frank_> Ok I started bitcoind
10:38 < arubi> can you retry the abandontransaction command?
10:39 < frank_> Ok it worked
10:39 < arubi> cool, now restart bitcoind without the blocksonly flag
10:39 < arubi> even better, before you do
10:39 < arubi> can you reach bitcoind datadir?
10:40 < waxwing> huh, this is interesting. so using blocksonly is a way to ditch un-abandonable-transactions?
10:40 < arubi> it's a weird thing, if the wallet rebroadcasted it and it's in mempool, it's not eligible
10:41 < arubi> afaict.  so now I want frank_ to move mempool.dat to mempool.dat.bak
10:41 < arubi> before restarting bitcoind
10:42 < frank_> ok done
10:42 < arubi> okay, so now you can check wallet-tool
10:42 < frank_> Ok my money is back
10:42 < arubi> yay     
10:42 < frank_> Thanks !
10:43 < arubi> np     
10:43 < frank_> But I still need some help :D
10:43 < arubi> hehe
10:43 < frank_> All of my addresses are in the second depth (1)
10:43 < frank_> Because I heard there was problems with the first one, is it possible?
10:44 < arubi> the problems?  never heard of it
10:44 < frank_> I don't know ^^
10:45 < frank_> So I have my money in depth 1
10:45 < frank_> what should I do for a perfect coinjoin ?
10:46 < frank_> I can not make the script work until the end, it always fails
10:46 < arubi> which one?
10:46 < frank_> the tumbler
10:47 < arubi> oh, are you passing any flags to it?  I'm not sure what happens if depth 0 isn't used at all and all funds are at depth 1
10:47 < arubi> you should be able to pass a depth flag tough
10:47 < arubi> *though
10:47 < waxwing> yes     
10:47 < frank_> python tumbler.py -m 1 wallet1.json
10:48 < frank_> That's all I do
10:48 < arubi> okay, so what's the failure there?
10:48 < frank_> I have 3 adresses with 0.0040665 btc each
10:49 < frank_> Different kind of error (unable to push tx, rebroadcasting, not enough money, ...)
10:49 < arubi> these look related to the false balance issue
10:50 < frank_> Maybe but i'm sure I'll get errors
10:50 < frank_> if I start the script
10:50 < arubi> well you could try now, worse that can happen you already know how to recover from
10:50 < arubi> you can't really lose money
10:51 < frank_> Yes but the best I can get is the script crashes at some time in the tumbling process and some (internal) addresses with coinjoined btc
10:52 < arubi> I really don't know what to say without an actual error.  also does tumbler pick inputs from internal addresses?
10:53 < frank_> I don't understand what you mean
10:53 < arubi> looks like you had one tumble done or more, but the tumbler run didn't manage to get all it's runs that it expected
10:54 < frank_> Should I specify output adresses ?
10:54 < arubi> no I don't think so
10:54 < frank_> Ok     
10:54 < frank_> So should I try with python tumbler.py -m 1 wallet1.json without other options ?
10:55 < arubi> yep     
10:56 < arubi> waxwing, does tumbler pick inputs from internal addresses?  iirc internal means funds that were used in a cj?
10:56 < frank_> Ok I started tumbling I come back to you if I get errors, thanks !
10:56 < arubi> cheers frank_ 
10:57 < waxwing> arubi, it picks from both (int. and ext.)
10:57 < arubi> ah cool
10:57 < frank_> First error :  ERROR outputs unconfirmed or already spent.
10:57 < waxwing> amounts 0.004 are far too small
10:57 < frank_> but the script continues
10:57 < waxwing> (although that error would be for a different reason)
10:58 < arubi> yea I thought the amount was too small.  didn't know for sure
10:58 < frank_> what is a reasonable amount for you ?
10:59 < arubi> see offers https://joinmarket.me/ob/
11:00 < waxwing> frank_, the tumbler page on the wiki gives some thoughts on fees, which have implications for amounts, but in any case i'd say ... less than 1 million satoshis is almost never going to work properly, for a tumbler run i think total amount about 0.1 is really bare minimum ... but that is kind of changing today, difficult to be sure.
11:01 < frank_> Ok I see, when you say the minimum, is it per address or for the total amount to tumble ?
11:02 < waxwing> i'm guessing you're setting bitcoin transaction fee rates very low, right? (i say that because of the problem you had earlier)
11:02 < frank_> Actually I'm not setting anything :p
11:02 < frank_> you mean max cj fee ?
11:03 < waxwing> there isn't really a minimum, arubi's right that running ob-watcher or looking at https://joinmarket.me/ob/ can give you some ideas
11:03 < waxwing> well that as well. but i meant bitcoin tx fee rates. see the tx_fees setting in the config file.
11:03 < waxwing> if you use defaults it's very likely you'll get fees that are very large chunk of your amount.
11:03 < waxwing> don't forget a coinjoin is not an ordinary bitcoin transaction. it is usually 4-7 times larger (varies a lot)
11:04 < waxwing> and so is multiples more expensive in fees.
11:04 < frank_> Yes I loose a lot
11:04 < waxwing> in tumbler you are doing a whole set of them.
11:04 < waxwing> again, this is explained in the tumbler page on the wiki. i put a big sort of "warning" paragraph at the start about it.
11:05 < frank_> Ok      
11:05 < frank_> Can I reach the minimum with multiple addresses ?
11:05 < waxwing> i would also mention that i made a new implementation of it in the joinmarket-clientserver repo which has some extra features for more robustness; but actually it may not be particularly relevant here.
11:05 < waxwing> and also yes do look at the non-default settings in `python tumbler.py --help` (again, some thoughts on the wiki page about it)
11:06 < frank_> non-default ?
11:08 < arubi> frank_, it would be very difficult to set find a set of settings and tweaks you can use to tumble ~0.004 amounts and not risk grinding the thing to dust, which will probably fail at some step during the tumble for being too small
11:10 < arubi> if you're not using segwit, these transactions can really explode.  that's part where the privacy comes from
11:13 < arubi> if you could first consolidate a few outputs for something close to 0.1, then you can expect things to work with no major issue.  the small amounts are not very common, and mixing with the same people over and over again might sound fun but not sure how helpful it is to you
11:14 < arubi> eventually the other participants are using up the small outputs, and if there aren't many to begin with, you can expect a very short run on tumbler which isn't the point of it
11:16 < frank_> So I should have a total of 4-5 addresses at least a total of 0.1 btc  ?
11:17 < waxwing> frank_, do you mean initial funding? well, at least 3 separate addresses with coins, in the starting mixdepth, is probably ideal. 4 or 5 is good too.
11:18 < frank_> Yes ok
11:18 < waxwing> i know on the new version (joinmarket-clientserver repo) it's not strictly necessary, because if your utxos are too new it just waits until they aren't. tumbler's supposed to take a long while anyway.
11:20 < frank_> Here I get coinjoining & <<pubmsg
11:20 < frank_> lot of times
11:22 < frank_> Also, How can I get an estimation of the output amount ?
11:22 < frank_> let's say I tumble 0.1 btc
11:23 < waxwing> yes, i get the question. i'm not sure we provide that; you can estimate it if you can get an estimate of sat/byte fee, but coinjoin fees are unpredictable. did you see the rough estimate on the wiki page? it's out of date but it gives you an idea.
11:24 < waxwing> if you set the bitcoin tx fees lower (use the N > 144 feature) you can change that a lot, but txs may take a long time to confirm of course.
11:24 < waxwing> (coinjoin fees are unpredictable, also, number of inputs in each tx is unpredictable, so it's going to be very much finger in the air)
11:25 < frank_> Okay but first I have to understand the cj fee
11:25 < frank_> in the transaction info at the beggining
11:25 < frank_> we have maker_count
11:25 < frank_> should a add every maker_count to get total makers ?
11:28 < arubi> frank_, some settings set maximums, some set minimums, some set even more elaborate stuff..  it's hard to give exact answers
11:29 < waxwing> also note the settings for maximum coinjoinfee absolute and relative, and read how it works in the help. although defaults are OK there.
11:29 < waxwing> generally for small amounts btc tx fees dominate, coinjoin fees are *mostly* rounding errors.
11:29 < waxwing> again, see the wiki page .... ;)
11:37 < frank_> Ok I understand the problems with tx fees and small amount to coinjoin
11:37 < frank_> But for cj fee let's say i set 1% and 100k satoshis
11:38 < frank_> What does it really mean
11:38 < frank_> I can lost up to 1% or 100k satoshis of my initial amout ?
11:39 < arubi> I think you're mixing up cj fees and transaction fees
11:39 < arubi> if I understand what you're asking
11:39 < frank_> why?
11:40 < arubi> there could be very nice participants to your cj that won't mind taking 0% cj fee, but still someone has to pay for the tx fees
11:40 < arubi> the more of these people participating, the higher the fees
11:41 < arubi> by setting the amount of participants, you set how much tx fees you expect.  by setting cj fees, you set how much you're willing to pay for each cj
11:41 < arubi> so the cj fees are multiplied times the transactions you're taking
11:43 < arubi> willing to pay each cj participant*  not the whole cj, I think
11:44 < frank_> So if there is x participant, max cj fee is x * max_cj_fee_per_participant
11:44 < frank_> Am I right?
11:45 < arubi> yea probably right, and the more participants, the higher the tx fees which are unrelated to the cj fee
11:46 < arubi> I don't know if that ^^ is a proper simplification though, don't take my word for it
11:46 < frank_> Yeah but I want to make sure I globally understand ^^
11:46 < waxwing> frank_, you understand what it says in --help for '-x', right?
11:48 < frank_> I do 
11:48 < waxwing> ok great; that's how you can set an upper limit on how much you pay in coinjoin fees.
11:48 < waxwing> and yes it's per participant
11:48 < waxwing> and bitcoin tx fees are quite separate.
11:49 < waxwing> btw the choice of what is paid is based on a random algo but quite heavily weighted towards cheaper (currently)
11:49 < waxwing> afk     
11:50 < frank_> yes but the cj fee is what I can lose per participant so I can lose much more ?
11:50 < frank_> if there are let's say 10 makers
11:53 < arubi> then as I said, even if they all just cj with you for free, you end up paying a huge amount in tx fees
11:53 < arubi> because 10 participants means that many inputs and even more outputs
11:54 < frank_> If I understand correctly cj fee is nothing compared to tx fees
11:55 < arubi> it can be, I don't know if it is really
11:55 < arubi> ob-watcher.py and a calculator for the answer :)
11:57 < frank_> Ok I get it but I have troubles to assemble all of this and make something under control :p
11:58 < frank_> without losing lots of money
11:58 < arubi> look at it this way, the more you tweak, the more you lose on privacy
11:59 < frank_> why?
11:59 < arubi> there's a reason for default values.  setting up some specific run (while also talking about it over irc no less) is making you stand out in a crowd
12:00 < frank_> Good to know :)
12:00 < frank_> For now I only used default values 
12:01 < arubi> yep, and if they fail, there's also a good reason
12:01 < arubi> probably means your expected gain is useless for you
12:01 -!- puddinpop_u [~puddinpop@unaffiliated/puddinpop] has joined #joinmarket     
12:02 < arubi> can you really use a very low amount output?  even if it's extremely tumbled and pretty much private
12:02 < arubi> you will end up consolidating it with some other input anyway
12:03 -!- zxccxz [5db781f6@gateway/web/freenode/ip.93.183.129.246] has quit [Quit: Page closed]     
12:03 < arubi> (which will completely destroy all privacy gained)
12:06 < arubi> the point of a coinjoin is to send to some destination that isn't yours, like paying someone.  the point of a tumble is to get some inputs consolidated into an output and gain privacy by cj'ing it a few times
12:07 < arubi> if you tumble 10 times with 5 participants each time, then it comes to a lot of tx fees, even if cj fees are low
12:07 < arubi> if you tumble only once, there's no point
12:09 < arubi> /dinner
12:20 -!- wxxs is now known as wxss     
14:22 -!- Randall [~Randall@ns334669.ip-5-196-64.eu] has quit [Disconnected by services]     
14:23 -!- Randall [~Randall@ns334669.ip-5-196-64.eu] has joined #joinmarket     
14:27 -!- takamatsu_ [~takamatsu@unaffiliated/takamatsu] has quit [Remote host closed the connection]     
14:27 -!- takamatsu_ [~takamatsu@unaffiliated/takamatsu] has joined #joinmarket     
14:28 -!- takamatsu_ [~takamatsu@unaffiliated/takamatsu] has quit [Remote host closed the connection]     
14:32 -!- zxccxz [5db781f6@gateway/web/freenode/ip.93.183.129.246] has joined #joinmarket     
15:47 -!- frank_ [5b566e51@gateway/web/freenode/ip.91.86.110.81] has quit [Ping timeout: 260 seconds]     
18:16 -!- Giszmo [~leo@pc-204-28-214-201.cm.vtr.net] has joined #joinmarket     
19:00 -!- wxss [~chatzilla@212.84.170.109] has quit [Remote host closed the connection]     
19:33 -!- Giszmo [~leo@pc-204-28-214-201.cm.vtr.net] has quit [Quit: Leaving.]     
21:39 -!- Randall [~Randall@ns334669.ip-5-196-64.eu] has quit [Disconnected by services]     
22:30 -!- Cody2 [~Cody@ns334669.ip-5-196-64.eu] has joined #joinmarket