--- Day changed Wed Mar 13 2019
00:05 -!- vivarium [b916af9b@gateway/web/freenode/ip.185.22.175.155] has quit [Ping timeout: 256 seconds]     
00:08 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
00:17 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Excess Flood]     
00:37 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
00:37 -!- viasil [~viasil@185.107.94.165] has quit [Ping timeout: 252 seconds]     
00:42 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 252 seconds]     
00:46 -!- viasil [~viasil@185.107.94.165] has joined #joinmarket     
02:07 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
03:11 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Excess Flood]     
03:12 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
03:35 -!- undeath [~undeath@hashcat/team/undeath] has joined #joinmarket     
05:26 -!- felix34 [~muh@5.2.76.19] has joined #joinmarket     
05:27 < felix34> having to restart wallet-tool after each incoming transaction and re-enter the wallet password is phenomenally annoying
05:30 < belcher> is that for yieldgenerator felix34 ?
05:31 < felix34> no, basicically issue 328, it happens even if I fund the wallet from an outside source
05:32 < felix34> this is an old version of JM so could be fixed in the latest
05:41 < felix34> that said was impressed at the documentation/install process seemingly having gotten better since last I looked at the github
06:22 < waxwing> felix34, it's not fixed
06:22 < waxwing> also 'phenomenally annoying', really? :)
06:22 < waxwing> i mean, it's annoying.
06:23 < waxwing> but wait, now i'm not 100% sure what you mean. in which precise circumstances do you think you have to 'restart wallet-tool'?
06:23 < waxwing> wallet-tool is not long-running so i'm not sure what you mean.
06:24 < waxwing> also, it could be experienced as 'phenomenally annoying' if you have a bitcoin wallet.dat with tons of transactions, since it has to trawl through them and it can take ages. if you move your joinmarket wallet to a new/clean bitcoin wallet.dat (using multiwallet) it can speed it up a lot.
06:26 < waxwing> belcher, perhaps consider broadcasting the fact that the installation is basically one-step on the twitter account? i mean, i can't talk, i refuse to use the platform, so :)
06:27 < waxwing> i did actually make this point on mastodon about a month or two ago. but it's of limited use to try to tell people 'it's easy' because ultimately it is quite awkward to use, even if installation *can* be very simple nowadays.
06:28 < belcher> is it awkward to use with the GUI?
06:28 < waxwing> with the gui it can still be a bit awkward.
06:28 < waxwing> consider the problem of funding with multiple utxos for example, to source commitments.
06:29 < waxwing> if everything is 'just right' it's really quite cool, like you can get the tumbler running and it 'just works'. i've done like 3 real tumbler runs without encountering problems.
06:29 < waxwing> but that's kinda ridiculous because i know what settings have to be right.
06:29 < belcher> oh yes i keep forgetting about those commitments
06:29 < waxwing> yeah think about the config, too. and the amount ranges and the fee settings.
06:29 < belcher> if its not obvious i dont use joinmarket much myself, not for things like booking flights like you do
06:29 < waxwing> all of it *can* work very smoothly; but it doesn't take much bad luck for one thing to screw it up, and if you're not expert you may have a heck of a time figuring out the problem.
06:30 < waxwing> well i've used tumbler a few times but really for testing after some big upgrade. not that it's a bad thing, i hope us yield generators are willing to do it now and again.
06:30 < waxwing> great for privacy of system, i guess.
06:33 < belcher> i dont run an always-on yg either (only sometimes run it to slowly mix)
06:33 < belcher> similar to how people working on bittorrent never pirated hollywood films
06:33 < belcher> idk maybe im paranoid
06:40 < belcher> https://twitter.com/chris_belcher_/status/1105825707316441089 retweeted by @joinmarket
06:41 < belcher> funnily enough @chris_belcher_ has more followers than @joinmarket, i guess aiming to be a low-traffic announcement account doesnt make for high follower counts
06:42 < waxwing> many thanks, sorry for badgering you about it. may have been a mistake if a bunch of normies now try to install it and then start complaining :)
06:44 < belcher> mea culpa :p
06:44 < belcher> we'll just have to help them i guess
07:10 < waxwing> oh belcher i just realised, you linked to the windows section of the install doc; but casual readers will think that that doc is the install instructions, but on that doc it doesn't say "just ./install.sh and source the jmvenv", that's on the readme. that page has instructions for "manual" install on linux. ugh, it's so hard to get documentation to be clear without being stupidly detailed.
07:11 < waxwing> in fact, that document specifically says what you said on the podcast, i.e. that you have to first start by installing a bunch of dependencies.
07:14 < d3spwn> its hard to write good documentation if you dont know who your target audience is
07:18 < d3spwn> the current doc's could be structured better, but i dont think you can dumb it down much more. if someone can't figure it out using the current doc's i'm inclined to say maybe they shouldnt use jm
07:22 < waxwing> sounds about right. but the specific point i'm making above ^ wouldn't require the user to be dumb, they'd naturally assume INSTALL.md tells you how to install :)
07:31 < d3spwn> oh yes, thats unfortunate
08:14 < belcher> do you want me to delete my tweet and make a new one?
08:14 < belcher> in hindsight i shouldve shown you the tweet before i tweeted it
08:17 < belcher> waxwing ^
08:25 < waxwing> belcher, well if you reply to it directly i guess people will see it, sorry i feel bad being so fussy about it.
08:25 < waxwing> if people actually care they'll figure it out :)
08:25 < belcher> what should i reply
08:26 < belcher> dont feel bad this stuff is important, its public relations
08:26 < belcher> i bet in other places they spend hours wording tweets and speeches to not give the wrong idea
08:28 < belcher> "the correct link should be <url>" maybe
08:28 < belcher> going to this maybe https://github.com/JoinMarket-Org/joinmarket-clientserver#quickstart
08:37 < waxwing> belcher, yeah sure just 'to do the one-step installation go to' and then that link. sounds good to me.
08:38 < waxwing> you could also point out that you need bitcoin core to use joinmarket. just another thought, since people won't always assume that, and no need to make them annoyed by only realising it after installing.
08:39 < waxwing> i mean basically people don't read stuff, i guess that's the uhh underlying problem :) (not criticising, i'm lazy about reading too)
08:39 < belcher> its human nature i guess
08:40 < belcher> i read a nice book on communication about this, called Made to Stick, helped me write reddit threads that people actually read even if they're long
08:41 < belcher> not saying im any good at it but the book seems useful
08:42 < Sentineo> if something looks hard you stop reading :) 
08:42 < Sentineo> unless you are motivated to
08:42 < belcher> the trick is making people motivated
08:43 < belcher> https://en.wikipedia.org/wiki/Made_to_Stick SUCCESs, like this one about finding the core of an idea
08:44 < belcher> theres this idea of the curse of knowledge, where experts who have studied all the details usually end up writing all about those details and downplaying the big core idea because it seems obvious
08:44 < belcher> anyway, its probably not useful for writing install instructions
08:46 < waxwing> hmm, interesting point, yes.
08:47 < waxwing> weird thing is, i think i'm pretty good at explaining stuff, in person or classroom setting because i did that for years, i got the habit of interactive questioning type instruction. but on the page i tend to just drone on and on.
08:48 < waxwing> also the thing about explaining stuff is it's way harder than it seems at first, just understanding something is only very entry level. when you've taught something 10 times you see it a whole different way.
08:48 < belcher> you are def very good at explaining things in person IMO
08:50 < belcher> writing on a paper you're probably better than you think
08:52 < waxwing> it's definitely a skill you can learn; but otoh being a really good teacher is more than just knowing how to explain stuff, there's some 'special sauce' there that may not really be learnable.
08:53 < harding> 100% agree that explaining stuff is way harder than it seems at first.  :-)
08:55 < waxwing> need moar 'feed the unbanked' or something in optech newsletter to get the masses harding :)
08:56 < waxwing> i guess when it comes to on-the-page one must consider 'UI' features a bit, however primitive. diagrams, links etc. the main bitcoin developer docs were very helpful to me because they were navigatable
08:56 < waxwing> so, thanks for that
08:57 < harding> waxwing: oh, no, don't ruin the surprise on our exit scheme optech token.
08:59 < harding> The navigation on Bitcoin.org was all done by its old maintainer, saivann, who was amazing.
09:00 < belcher> this whole discussion got started because on that podcast i think i was asked why more people dont use joinmarket, i said installation problems.. so what do you guys think are other possible reasons?
09:01 < belcher> it requires a full node and sometimes those fiddly podle commitments, that might be why too
09:04 < harding> I don't think it's as heavily marketed as the other current major coinjoin implementation.
09:04 < waxwing> i honestly think full node is a write off for a huge proportion of people. but very limited UI (GUI but only very basic) and also, installation is not trivial. some people will just not use a terminal full stop.
09:05 < waxwing> and yeah marketing matters, but there'd be no point trying to mass market something that isn't designed to be used by mass market, anyway.
09:05 < belcher> true with the terminal, many people probably dont know how to turn it on, even "cd" takes a few minutes to learn id say
09:05 < waxwing> btw on the topic of wasabi, kevin loaec (sp) made an interesting point.
09:06 < waxwing> there's an incentive problem because the server earns, means it actually would make economic sense for them to fake volume.
09:06 < waxwing> chance of that actually happening is near zero, but it's an interesting example of how centralization affects things.
09:06 < belcher> can you explain that incentive ?
09:06 < belcher> is it higher volume = more anonymity if you use wasabi?
09:07 < waxwing> well, say only 5 people turn up to mix, but the round is set at 50.
09:07 < waxwing> wasabi only earns the fees if the coinjoins go through.
09:07 < waxwing> so they pay themselves the other 45 fees.
09:07 < belcher> so in some sense they're acting like joinmarket makers for those 45
09:08 < waxwing> yeah but the point is, they're the only ones in a position to sybil like that. any other sybiller has to pay.
09:08 < waxwing> i mean it's hardly likely to happen, but it's an interesting incentive skew.
09:08 < belcher> for sure
09:09 < belcher> i wondered if theres a way to change the incentives that the last few wasabi users pay less, to encourage the round to finish
09:09 < belcher> but that doesnt fix the thing you're saying
09:09 < waxwing> yeah you could suggest that, it's not a bad idea really.
09:10 < harding> Would that incentivize people who joined the round early to drop out and come back later?
09:10 < belcher> they'd need a new utxo i think
09:10 < belcher> but wasabi users have lots of utxos so it might be practical
09:10 < waxwing> hmm yeah maybe that screws it up. utxo blacklisting is messy. but i think they do that.
09:11 < belcher> we shouldnt forget that the irc servers used as joinmarket meeting points also have some power, they can hide all bots and make you only meet their own sybils.. but at least theres more than one network and presumably yieldgen operators will check their offers actually show up
09:12 < waxwing> yeah i think having 2, let alone 3, makes that problem very small indeed. but i agree it's not nothing.
09:12 < harding> Is SSL used for all the IRC networks?
09:13 < waxwing> almost everyone is connecting over Tor to HS
09:13 < harding> Oh, ok.
09:13 < belcher> yes for clearnet (i just checked, im connected to all 3)
09:13 < waxwing> i think the current ones all have ssl, i forget. so you can connect like that too.
09:14 < belcher> given that almost everyone uses tor HS, we have the option of moving over to only HS networks
09:14 < waxwing> it's kinda funny when you go over vpn to the HS and also are using e2e encryption for messaging. encryption to the max :)
09:14 < belcher> irc servers are very practical to run on home ASDL connections, since upload is approx equal download bandwidth
09:15 < belcher> ADSL* you know the common internet scheme everyone has for households
09:16 < belcher> but thats not true for http servers for example, where upload requirements are much more than download (or bitcoin full nodes but transfer speed isnt so important i think)
09:17 < belcher> but even so, clearnet irc networks might be useful for something like an eventual electrum plugin so that users dont need to also install tor
09:20 < belcher> another point: if we intend to add client-side block filtering then we need to make joinmarket seed phrases somehow include the wallet creation date, that makes it much easier to recover
09:20 < belcher> arubi_ i think once came up with an idea for grinding the seed like what electrum does, but to include an indication of the birthday
09:21 < waxwing> oh. clever.
09:21 < belcher> then its still backward compatible with bip39
09:21 < waxwing> the cryptography police have been alerted :)
09:22 < belcher> you dont get many bits, so maybe the birthday could be encoded as "number of quarters since january 2019" and then 9 bits is enough for 100 years or so
09:59 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Read error: Connection reset by peer]     
10:08 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
10:12 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Read error: Connection reset by peer]     
10:30 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
10:31 < belcher> theres also this other podcast i did https://twitter.com/ecurrencyhodler/status/1105861465397854208
10:37 < waxwing> oh right the whole walletexplorer thing, heh
12:54 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Read error: Connection reset by peer]     
12:54 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
13:21 -!- lnostdal [~lnostdal@77.70.119.51] has quit [Ping timeout: 245 seconds]     
13:24 -!- lnostdal [~lnostdal@77.70.119.51] has joined #joinmarket     
16:30 < takinbo> i've noticed that whenever the tumbler is moving all utxos from one mix level to the next (that is no change), it tends to pay 5000 sats per counterparty sometimes overpaying in fees; why is that?
16:59 -!- puddinpop [~puddinpop@unaffiliated/puddinpop] has joined #joinmarket     
17:52 < undeath> if there is no change it's very hard to make a good estimation for the cj amount because the fee is directly related to that in dependent on the makers
17:52 < undeath> so you have to apply some heuristics to make a good-enough guess for the right cj amount
17:52 -!- undeath [~undeath@hashcat/team/undeath] has quit [Quit: WeeChat 2.4]     
18:29 < qubenix> A lot of talk about the legitimacy/plausible deniability of participating in coinjoins recently. Is there any other way to earn passive income like this without giving up control of your bitcoin?
18:30 -!- d3spwn [~jimmy@a83-161-157-200.adsl.xs4all.nl] has quit [Ping timeout: 255 seconds]     
18:37 -!- d3spwn [~jimmy@a83-161-157-200.adsl.xs4all.nl] has joined #joinmarket     
18:45 < felix34> waxwing every time I get an incoming transaction (e.g to fund the wallet) and often after any outgoing one as well, when I then run wallet-tool after the transaction, I get the issue #328 error message and have to restart. it's merely annoying initially, but becomes phenomenally annoying after using JM for a while, as it really seem to be EVERY SINGLE FSCKING TRANSACTION
18:57 -!- AgoraRelay [~jmrelayfn@p5DE4A6BC.dip0.t-ipconnect.de] has quit [Ping timeout: 245 seconds]     
19:08 -!- felix34 [~muh@5.2.76.19] has quit [Ping timeout: 246 seconds]     
19:12 -!- AgoraRelay [~jmrelayfn@p5DE4A939.dip0.t-ipconnect.de] has joined #joinmarket     
19:29 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 252 seconds]     
19:34 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
20:17 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 252 seconds]     
20:55 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
22:04 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 252 seconds]     
22:06 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
22:10 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Excess Flood]     
22:10 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
22:47 -!- bbbbooo [a5e3060d@gateway/web/freenode/ip.165.227.6.13] has joined #joinmarket     
22:52 -!- bbbbooo [a5e3060d@gateway/web/freenode/ip.165.227.6.13] has quit [Ping timeout: 256 seconds]     
23:11 -!- felix34 [~muh@5.2.76.19] has joined #joinmarket