--- Day changed Wed Jan 18 2017 01:41 -!- cbits [~cbits@47.148.176.74] has quit [Ping timeout: 255 seconds] 03:36 -!- waxwing [~waxwing@14.174.32.23] has quit [Ping timeout: 240 seconds] 03:49 -!- waxwing [waxwing@gateway/vpn/mullvad/x-tvnejrunuzkxiydr] has joined #joinmarket 04:03 -!- Trinity [~Trinity@unaffiliated/trinity] has joined #joinmarket 04:30 -!- zxccxz_ [4f7c12d7@gateway/web/freenode/ip.79.124.18.215] has quit [Ping timeout: 260 seconds] 05:08 -!- waxwing [waxwing@gateway/vpn/mullvad/x-tvnejrunuzkxiydr] has quit [Ping timeout: 264 seconds] 05:20 -!- waxwing [waxwing@gateway/vpn/mullvad/x-jfnyamketxfmfkgd] has joined #joinmarket 05:45 < waxwing> i've been away on and off last week or two, has there been any discussion further about either the possible alternative blockchain interfaces or addressing the duplicate utxo issue? 05:56 < waxwing> re: the latter it looks like the numbers of duplicates is increasing if anything, so at some point even 'minimum_makers' tricks are not going to be enough, well, depending on details. 06:34 < lebeev> belcher_: I think reddit operates under the assumption that there are moderators actively approving posts 06:34 < lebeev> just FWIW 06:36 < lebeev> although maybe you can just turn off all filtering within the subreddit, or set automoderator to auto-approve 07:02 -!- Yohkii [~Yohkii@unaffiliated/yohkii] has joined #joinmarket 07:18 -!- juscamarena [~justin@47.148.176.74] has quit [Remote host closed the connection] 07:18 -!- berndj [~berndj@mail.azna.co.za] has quit [Excess Flood] --- Log closed Wed Jan 18 07:18:33 2017 --- Log opened Wed Jan 18 07:29:03 2017 07:29 -!- kanzure [~kanzure@unaffiliated/kanzure] has joined #joinmarket 07:29 -!- Irssi: #joinmarket: Total of 68 nicks [1 ops, 0 halfops, 0 voices, 67 normal] 07:31 -!- zmachine [uid67411@gateway/web/irccloud.com/x-rfhhalndyiuchhhz] has joined #joinmarket 07:32 -!- Yohkii [~Yohkii@unaffiliated/yohkii] has quit [Quit: Mutter: www.mutterirc.com] 07:39 < waxwing> https://www.reddit.com/r/joinmarket/comments/5oavdt/has_jm_been_abandoned_tor_bug_not_fixed_questions/dcl1rwx/ this is just reminding me that we don't even have a solution for non-responsive for sweep. 07:40 -!- Irssi: Join to #joinmarket was synced in 691 secs 07:40 < grubles> hey wb waxwing :) 07:41 < waxwing> huh see what you're referring to lebeev .. uhh interesting :) 07:42 < waxwing> grubles: hi, wb? 07:42 < grubles> welcome back :p 07:42 < waxwing> ah thanks :) i'm sort of off and on, i should really get a bouncer, too lazy 07:42 < waxwing> and yeah this should be logged but meh 07:43 < waxwing> or maybe it shouldn't, could never really decide 07:43 < grubles> hm yeah not sure about that either 07:44 < lebeev> waxwing: for a small subreddit like that https://www.reddit.com/r/AutoModerator/comments/42km6z/auto_approve_new_posts_from_any_users/ seems like a reasonable setup unless there are actual spam issues 07:47 < grubles> does insight.bitpay.com have the tor captcha? 07:48 < waxwing> i was wondering something similar today, does bc.i work over Tor? 07:48 < grubles> insight-api has this: https://github.com/bitpay/insight-api#transaction-broadcasting 07:48 < grubles> umm i think bc.i actually does have a hidden service 07:49 < waxwing> i wonder that specifically, not because i think bc.i is great, but because we have an open PR which i believe works (or did, after i spent many hours finding a pernicious bug in it) 07:49 < waxwing> yes the HS never seems to work though 07:49 < waxwing> i mean, your Q is good too, i can't imagine it's terrifyingly hard to create a blockchainterface class for another api 07:50 < lebeev> doesnt bc.i have a hidden service waxwing 07:50 < waxwing> i mean it's fiddly and slow to get it right and tested, but it's not hard 07:50 < waxwing> lebeev: yeah i just said, it never seems to work 07:50 < waxwing> for me, anyway 07:50 < lebeev> ah 07:50 < lebeev> let me try 07:51 < lebeev> worked great in the past but not 100% sure if it provides the functionality needed here 07:51 < lebeev> looks fine to me 07:51 < lebeev> https://blockchainbdgpzk.onion 07:51 < lebeev> api is a different deal 07:51 < grubles> with insight, people could run their own instances of it too (if they wanted) 07:51 < waxwing> needs an indexed node presumably? 07:51 < grubles> yeah 07:52 < grubles> very resource intensive 07:52 < grubles> but, maybe people could connect to others' insight instances 07:52 < grubles> so we aren't all connecting to blockr 07:52 < lebeev> api is indeed disabled over tor 07:52 < grubles> i dunno. i'm just brainstorming. 07:53 < lebeev> https://blockchainbdgpzk.onion/q/addressbalance/1FeexV6bAHb8ybZjqQMjJrcCrHGW9sb6uF 07:53 < lebeev> vs 07:53 < lebeev> https://blockchain.info/1FeexV6bAHb8ybZjqQMjJrcCrHGW9sb6uF 07:53 < lebeev> er 07:53 < lebeev> https://blockchain.info/q/addressbalance/1FeexV6bAHb8ybZjqQMjJrcCrHGW9sb6uF 07:54 < lebeev> now I seem to be able to hit the api just fine over tor though 07:54 < lebeev> the clearnet api that is 07:55 < waxwing> I get 'Disabled' 07:55 < waxwing> for https://blockchainbdgpzk.onion/q/addressbalance/1FeexV6bAHb8ybZjqQMjJrcCrHGW9sb6uF 07:56 < waxwing> hmm, but the main page of https://blockchainbdgpzk.onion loads now. it never did for me, before 07:56 < waxwing> i think maybe i had the wrong onion link, always got 'Quota exceeded' 07:57 < waxwing> lebeev: oh, maybe i didn't understand your report. you can, or can't, access the api over Tor? 07:57 < lebeev> 15:52:55 lebeev | api is indeed disabled over tor 07:57 < lebeev> clearnet api looks fine over Tor right now 07:57 < lebeev> i.e. https://blockchain.info/q/addressbalance/1FeexV6bAHb8ybZjqQMjJrcCrHGW9sb6uF 07:58 < waxwing> yeah but then you said "now seem to be able .. fine over Tor" 07:58 < waxwing> i see, so not HS 07:58 < waxwing> that makes sense 07:58 < lebeev> yeah I thought they had cloudflare before on it 07:58 < lebeev> but looks like they've turned that down/off 07:58 < waxwing> yup works for me too 07:58 < lebeev> not cloudflare, but the cloudflare thing that harasses Tor users 07:59 < waxwing> it's painful though, someone spends a week testing it, release it, then they turn on cloudflare again... 07:59 < waxwing> would be great if cloudflare started using some hashcash or something 07:59 < waxwing> although i guess the painful slowness would be even worse 07:59 < lebeev> the 'Quota exceeded' afaik is per path ratelimits (as in /, or /q/addressbalance/address) most of the time 08:00 < lebeev> I'd imagine they do same per-ip, and that probably affects the .onion too 08:01 < lebeev> rather than using hashcash it'd be better if they gave a damn, but from friends that work there they don't seem to 08:01 < waxwing> i mentioned it not purely whimsically, i had a vague memory that there were some ideas along those lines. but probably won't happen :) 08:02 < waxwing> it's a bit "out there" 08:02 < waxwing> some people would probably say it's never going to work as a general strategy 08:02 < lebeev> yep 08:03 < lebeev> but they only ever came up with it to deflect 08:03 < lebeev> they're not blocking Tor users because they need to 08:03 < lebeev> no sorry 08:04 < lebeev> that was terribly worded 08:04 < lebeev> they're not blocking Tor users because their customers demand it 08:04 < lebeev> rather because of their current lacking anti-abuse solutions they do need to 08:05 < waxwing> what's the right way to do it then? personally i don't have any too-smart ideas on how to rate limit apart from hashcash. 08:06 < lebeev> well the single biggest issue I guess is comment spam 08:06 < lebeev> serving captchas at that point makes sense 08:06 < lebeev> that's how akamai does it too 08:08 < lebeev> after that you get the random hacking attempts and wp-admin bruteforce 08:08 < lebeev> this is mostly relatively easy to filter because you know how it looks 08:08 < lebeev> and that's how they mostly filter it 08:09 < lebeev> just not very well 08:09 < lebeev> maybe they'll figure out some clever ML for it at some point 08:10 < waxwing> bitcoin style things are basically *the* solution, otherwise it's always just bots vs bots :) 08:10 < lebeev> hm 08:10 < lebeev> well 08:11 < lebeev> I guess you could really easily detect likely attacks in GET requests and serve captchas for those 08:11 -!- Anduck_ is now known as Anduck 08:11 < lebeev> honestly don't see why you'd ever serve a captcha for a GET request on / or other normal path like that, unless the request has some malicious looking headers 08:13 < waxwing> does captcha not help against DDOS? i've never really thought about this much. wouldn't it stop requests reaching the endpoint? 08:14 < lebeev> it helps against DDoS when it's happening 08:14 < lebeev> that's when cloudflare serves it for all IPs, not just Tor 08:14 < lebeev> it doesn't help against DDoS when there isn't one going on, and when one starts you usually know within a second or so 08:15 < lebeev> you can turn on the "I'm under attack" mode in CF control panels and it should keep showing captchas to everyone that visits the site 08:15 < waxwing> hmm yeah think i see your point, you mainly need to detect the attack and then take countermeasures. not really needed for that. 08:15 < lebeev> right now it's primarily a really elaborate solution to comment spam 08:16 < lebeev> I use luminati.io regularly for things, and it makes cloudflares goal seem even more futile 08:16 < lebeev> I've got 19 million relatively clean IP addresses I can use, what's cloudflare gonna do to stop me? 08:17 < lebeev> "relatively clean", as in completely clean unless I'm really unlucky and get one that someone else used for noisy bad things before 08:18 < lebeev> hard to see what showing captchas to Tor users prevents :) 08:18 < grubles> looks like limunati.io is $500/mo o_o 08:20 < lebeev> well some other people gladly pay 30k+ a month for it to track down ad fraud or to scrapeflight/hotel prices for example 08:22 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Remote host closed the connection] 08:22 < lebeev> the guys running it are really shady (http://adios-hola.org/), but the service itself is pretty neat 08:22 < lebeev> but that's the legal way to defeat cloudflare & co., if you want to do it illegally there's certainly much easier approaches available 08:23 < waxwing> mirai ? 08:23 < waxwing> i do wonder about ipv6 , seems like it could be a problem here 08:23 < lebeev> yea I was gonna suggest slapping proxy code onto that but I'm not sure how well it spreads after the code got released 08:24 < weex> i think it's better to just say, you want jm, you run a bitcoin core node 08:24 < weex> it can be pruned and we can make it easy to setup 08:24 < waxwing> weex: yes, the problem is TAILS 08:24 < lebeev> I guess anyone can get like 100k or so bots with it 08:24 < lebeev> honestly I'd just tell people not to use TAILS but that's bad manners 08:24 < weex> vagrant or docker you can bring the # of steps way down, for example I made this https://github.com/ReinProject/devsetup 08:25 < weex> what baggage does TAILS bring? 08:25 < lebeev> lack of persistent storage I guess 08:25 < lebeev> nothing besides that 08:25 < waxwing> well people run it off USB and the issue is where do they put 100GB for initial sync, i guess? 08:26 < waxwing> maybe you can do it if you already have your blockchain. not sure. 08:26 < waxwing> tailsjoin did a full write up on how to run it with full node + TAILS, but that was 1-2 years back 08:26 < waxwing> recently arubi updated the install process, but nobody's tested it i think. or, we have no reports. 08:27 < waxwing> it seems a common usage pattern is, have TAILS (probably using built-in electrum), run TAILS, install JM, use blockr default. 08:27 < waxwing> we might think it's a bad way to use it, but it seems a common one. 08:27 < weex> pruning works though so only need more like 10gb 08:28 < waxwing> weex: yeah but need initial sync. unless they already pruned outside TAILS and can somehow import? 08:28 < waxwing> weex: for things like docker, how distributabe is it and what's the trust model? 08:28 < lebeev> I feel that Tails should be avoided given how deanonymization attacks using RCE sploits are everyday things now, but that's not particularly relevant to this conversation 08:28 < lebeev> but I'd direct people to set up whonix instead 08:29 < lebeev> and provide bci interface for those that still want to use tails 08:29 < lebeev> in case bci actually works :) 08:29 < weex> i like to reduce dependencies and tails seems pretty specific, do we have an idea of how common it is for people to use tails, or just setup their own box? 08:29 < waxwing> lebeev: those RCEs though, isn't it mostly about the browser? 08:30 < lebeev> waxwing: yeah, but tails by design does a poor job defending against those 08:30 < waxwing> weex: don't think we have any statistics, but i know a lot of people asked about tails setups 08:30 < lebeev> weex tails is extremely common 08:30 < lebeev> basically all the guides recommend it because it's easy 08:30 < lebeev> whonix defends against the RCEs by not being able to leak your IP 08:31 < lebeev> tails defends against the RCEs with some iptables rules 08:31 < lebeev> and in the past they've had particularly silly LPE issues 08:31 < waxwing> an important feature of TAILS i think, is that it's a poor man's cold wallet, being completely disconnected from your "live" machine, but not needing a second piece of hardware (other than usb stick) 08:31 < weex> those would be RCEs in Tor or just in general? 08:31 < lebeev> just in general 08:32 < weex> then in the tails use case, these are folks who just boot up for an hour of several to do their jm thing then, put the machine back to their regular OS?? 08:32 < lebeev> these days people get deanonymized when FBI takes over a site and puts a 1day or 0day firefox exploit on there 08:32 < lebeev> waxwing: is that how people normally use tails? 08:32 -!- bsm117532 [~mcelrath@38.121.165.30] has joined #joinmarket 08:32 < lebeev> it seems hard to use it as a cold wallet for bitcoin storage 08:33 < lebeev> since AFAIK you can't actually store things 08:33 < waxwing> conjecture. i think that's a common thing though. it's a lot easier because it's shipped with electrum 08:33 < lebeev> which would presumably include your private key 08:33 < waxwing> you don't need to store data, just use the electrum seed (or joinmarket seed) 08:33 < lebeev> I'd assume people just use it because all the Tor market vendors use it 08:34 < weex> waxwing i pm'd a rewrite of the first page of the wiki, did you get that or should i send it again, or just post it? 08:34 < waxwing> the main point is you run it live off the USB, with no persistence, nothing is stored anywhere 08:34 < weex> i feel like sprucing up the docs could be helpful as there's a lot of wordiness i detect 08:35 < weex> that and making the case for privacy broadly so people are willing to go a little further for the project 08:35 < waxwing> weex: thanks for reminder, i still have it. could you PM it to belcher_ i think it's somehow more his thing (i *think* he wrote most of it originally) 08:35 < weex> for example i feel like jm shouldn't lack for funding and paid devs since it's got to be important for hodlers 08:36 < weex> belcher_: pm'd, i'm willing to tackle more pages too if you like that 08:36 < weex> I've been working on a system to make that kind of work more efficient, Rein, it's basically a multisig escrow manager but has some decentralization in the design 08:37 < weex> i've been using it to get Rein moving a bit faster and would love to see the same thing happen for jm 08:37 -!- zxccxz [4f7c12d7@gateway/web/freenode/ip.79.124.18.215] has joined #joinmarket 08:37 < weex> once there's a multisig escrow setup for a job, anyone can fund which i think is a neat trick for bitcoin in general 08:40 < waxwing> it's not inconceivable to use a bounty approach. it wouldn't help me much though, when i do things like (a) test PRs (b) work on a new refactoring. those are things that don't have a direct perception of value to a user. 08:41 < weex> doesn't always have to be a bounty, it just means that when people donate to the project they can be assured there's value there 08:42 < waxwing> multisig is a nice extra, i've done that before, but it just spreads the decision a bit. doesn't change things *too* much (not like the whole lighthouse style thing, which is probably a bit more trouble than it's worth) 08:42 < weex> test these X PRs, refactor these N modules, etc 08:43 < weex> the description can be as general or specific or even private if people like 08:44 < weex> most of the work i've needed to do was in finding devs and the multisigs just helped me know i'm being efficient 08:44 < weex> if the work doesn't get done, the money can come back, assuming the mediator is honest 08:45 < weex> not the pinnacle of a giving spirit but bitcoin's special territory 08:46 < waxwing> yes returnable with honesty is nice (you could do that anyway), but multisig is clearly better. 08:47 < weex> what i'm realizing though as well is that jm is kind of important if payments are made somewhat publicly 08:47 < weex> jm seems to defend best against casual observers and your mediator, worker, or client become easy casual observers in REin 08:48 < weex> it would be kind of neat if payments at the close of a job were routed through jm by default 08:49 < weex> not sure if that's a net win for privacy though, just a thought 08:49 < waxwing> yes using JM as an adjunct to other privacy measures is good 08:49 < waxwing> best of all is to have wallets that mix roles; run as ygs for some time, then send payments. this can also be done with patientsendpayment 08:50 < waxwing> the latter works by offering coins to join for some period, then finally sending the remainder/all after a (long) timeout 08:50 < weex> long being a day or days? 08:50 < waxwing> entirely up to you 08:51 < weex> Rein is getting a bip32 wallet on next release, anyway gotta go, keep it up! 08:51 < waxwing> ok good to talk, thanks for suggestions 08:53 -!- sturles_ is now known as sturles 09:11 -!- waxwing [waxwing@gateway/vpn/mullvad/x-jfnyamketxfmfkgd] has quit [Ping timeout: 245 seconds] 09:12 -!- waxwing [waxwing@gateway/vpn/mullvad/x-zyxuqxqfzxbagpla] has joined #joinmarket 09:34 -!- waxwing [waxwing@gateway/vpn/mullvad/x-zyxuqxqfzxbagpla] has quit [Ping timeout: 245 seconds] 09:35 -!- Trinity [~Trinity@unaffiliated/trinity] has quit [Remote host closed the connection] 09:50 -!- waxwing [~waxwing@14.174.32.23] has joined #joinmarket 10:19 -!- waxwing [~waxwing@14.174.32.23] has quit [Ping timeout: 240 seconds] 10:36 < grubles> hm is agora down? 11:02 -!- zxccxz [4f7c12d7@gateway/web/freenode/ip.79.124.18.215] has quit [Ping timeout: 260 seconds] 11:38 < grubles> hm nevermind :) 12:12 -!- zxccxz [4f7c12d7@gateway/web/freenode/ip.79.124.18.215] has joined #joinmarket 12:20 -!- deafboy [quasselcor@cicolina.org] has quit [Quit: No Ping reply in 180 seconds.] 12:22 -!- deafboy [quasselcor@cicolina.org] has joined #joinmarket 12:24 -!- windsok [~windsok@45.63.59.8] has quit [Ping timeout: 260 seconds] 12:29 -!- juscamarena [~justin@47.148.176.74] has joined #joinmarket 12:29 -!- juscamarena_ [~justin@47.148.176.74] has joined #joinmarket 12:30 -!- juscamarena_ [~justin@47.148.176.74] has quit [Client Quit] 12:51 -!- windsok [~windsok@45.63.59.8] has joined #joinmarket 13:13 -!- waxwing [~waxwing@14.174.32.23] has joined #joinmarket 13:53 -!- cbits [~cbits@2607:f380:a61:650:1410:ce00:3445:4a8a] has joined #joinmarket 14:01 -!- cbits [~cbits@2607:f380:a61:650:1410:ce00:3445:4a8a] has quit [Ping timeout: 255 seconds] 14:19 -!- cbits [~cbits@139.182.205.184] has joined #joinmarket 14:27 -!- bsm117532 [~mcelrath@38.121.165.30] has quit [Ping timeout: 240 seconds] 14:32 -!- bsm117532 [~mcelrath@38.121.165.60] has joined #joinmarket 14:34 -!- bsm1175322 [~mcelrath@38.121.165.30] has joined #joinmarket 14:35 -!- bsm1175322 [~mcelrath@38.121.165.30] has quit [Remote host closed the connection] 14:36 -!- bsm117532 [~mcelrath@38.121.165.60] has quit [Ping timeout: 240 seconds] 14:37 -!- bsm117532 [~mcelrath@38.121.165.30] has joined #joinmarket 14:50 -!- nkuttler [~nkuttler@unaffiliated/nkuttler] has joined #joinmarket 14:50 -!- nkuttler [~nkuttler@unaffiliated/nkuttler] has quit [Client Quit] 14:51 -!- nkuttler [~nkuttler@unaffiliated/nkuttler] has joined #joinmarket 15:19 -!- cbits [~cbits@139.182.205.184] has quit [Ping timeout: 252 seconds] 15:21 -!- cbits [~cbits@2607:f380:a61:650:78a2:2ea:d1c4:c2bf] has joined #joinmarket 15:28 -!- IanKelling [~ian@c-73-53-70-213.hsd1.wa.comcast.net] has quit [Remote host closed the connection] 15:35 -!- cbits [~cbits@2607:f380:a61:650:78a2:2ea:d1c4:c2bf] has quit [Ping timeout: 255 seconds] 17:09 -!- cbits [~cbits@2607:f380:a61:650:48d:dec2:683a:dc04] has joined #joinmarket 17:09 -!- cbits [~cbits@2607:f380:a61:650:48d:dec2:683a:dc04] has quit [Remote host closed the connection] 17:09 -!- cbits [~cbits@2607:f380:a61:650:48d:dec2:683a:dc04] has joined #joinmarket 17:33 -!- cbits_ [~cbits@2607:f380:a61:650:48d:dec2:683a:dc04] has joined #joinmarket 17:33 -!- zxccxz [4f7c12d7@gateway/web/freenode/ip.79.124.18.215] has quit [Ping timeout: 260 seconds] 17:34 -!- cbits [~cbits@2607:f380:a61:650:48d:dec2:683a:dc04] has quit [Ping timeout: 255 seconds] 17:56 -!- cbits_ [~cbits@2607:f380:a61:650:48d:dec2:683a:dc04] has quit [Ping timeout: 256 seconds] 17:59 -!- cbits [~cbits@2607:f380:a61:650:1c62:599f:8331:fe6] has joined #joinmarket 18:17 -!- cbits_ [~cbits@139.182.205.184] has joined #joinmarket 18:20 -!- cbits [~cbits@2607:f380:a61:650:1c62:599f:8331:fe6] has quit [Ping timeout: 255 seconds] 18:33 -!- cbits [~cbits@139.182.205.184] has joined #joinmarket 18:35 -!- berndj-blackout [~berndj@mail.azna.co.za] has joined #joinmarket 18:36 -!- pigeons_ [~pigeons@94.242.209.214] has joined #joinmarket 18:37 -!- wump [~quassel@pdpc/supporter/professional/wumpus] has joined #joinmarket 18:39 -!- cbits [~cbits@139.182.205.184] has quit [Quit: Leaving] 18:40 -!- Lightsword_ [~Lightswor@2604:a880:1:20::1d3:9001] has joined #joinmarket 18:40 -!- rotarydialer_ [~rotarydia@unaffiliated/rotarydialer] has joined #joinmarket 18:42 -!- Netsplit *.net <-> *.split quits: wumpus, arubi, pigeons, rotarydialer, Lightsword, cbits_, puddinpop, nanotube, berndj 18:42 -!- Lightsword_ is now known as Lightsword 18:48 -!- nanotube [~nanotube@unaffiliated/nanotube] has joined #joinmarket 18:51 -!- berndj-blackout is now known as berndj 19:05 -!- cbits [~cbits@139.182.205.184] has joined #joinmarket 19:06 -!- puddinpop [~puddinpop@unaffiliated/puddinpop] has joined #joinmarket 19:15 -!- arubi [~ese168@unaffiliated/arubi] has joined #joinmarket 19:31 -!- pigeons_ is now known as pigeons 19:51 -!- cbits [~cbits@139.182.205.184] has quit [Ping timeout: 255 seconds] 20:10 -!- cbits [~cbits@2607:f380:a61:650:29be:5663:2e1f:b3e3] has joined #joinmarket 20:55 -!- cbits [~cbits@2607:f380:a61:650:29be:5663:2e1f:b3e3] has quit [Ping timeout: 256 seconds] 22:03 -!- wump is now known as wumpus 22:40 -!- xpub [054f49f6@gateway/web/freenode/ip.5.79.73.246] has quit [Ping timeout: 260 seconds] 23:02 -!- zxccxz [4f7c12d7@gateway/web/freenode/ip.79.124.18.215] has joined #joinmarket 23:16 -!- cbits [~cbits@47.148.176.74] has joined #joinmarket 23:18 -!- cbits [~cbits@47.148.176.74] has quit [Client Quit]