--- Day changed Mon Sep 25 2017
01:51 -!- xcvvcx [53e42f33@gateway/web/freenode/ip.83.228.47.51] has quit [Quit: Page closed]     
02:56 -!- MaxSan [~one@213.152.161.85] has quit [Ping timeout: 240 seconds]     
06:30 -!- StopAndDecrypt [~StopAndDe@c-73-248-248-9.hsd1.nj.comcast.net] has joined #joinmarket     
06:48 -!- StopAndDecrypt [~StopAndDe@c-73-248-248-9.hsd1.nj.comcast.net] has quit [Read error: Connection reset by peer]     
07:13 -!- Giszmo [~leo@pc-204-28-214-201.cm.vtr.net] has joined #joinmarket     
07:14 -!- StopAndDecrypt [~StopAndDe@c-73-248-248-9.hsd1.nj.comcast.net] has joined #joinmarket     
07:38 -!- xcvvcx [53e42f33@gateway/web/freenode/ip.83.228.47.51] has joined #joinmarket     
07:43 < waxwing> that thread for #89 is becoming almost comical. might be better not to confuse the issue with virtualenvwrapper, but in any case, .. what the hell.
07:54 -!- delinquentme [~delinquen@142-254-104-244.dsl.dynamic.fusionbroadband.com] has joined #joinmarket     
07:55 < delinquentme> waxwing, belcher should we be concerned about access to the tx information available to the IRC channel that the bots use as an exchange?
07:56 < belcher> maybe
07:56 < belcher> there are measures taking to reduce any information being given out
08:01 < waxwing> delinquentme, did you see my answer to your Q the other day about that?
08:01 < delinquentme> I dont think so.
08:01 < waxwing> <waxwing> delinquentme, no
08:01 < waxwing> <waxwing> the 'server' in joinmarket's implementation is actually the "Taker", the person who starts the coinjoin
08:01 < waxwing> <waxwing> the messaging server is different; they don't learn the mapping because the messages being sent are end to end encrypted
08:02 < waxwing> there is of course significant metadata leakage, but that ^ was in response to your wondering about the server as per the original coinjoin post
08:02 < delinquentme> also is fivepence in here? I think i might have done a bad thing in telling the one guy in the bug reports to use virtualenvwrapper ... its probs best to standardize people running these nodes as much as possible
08:02 < delinquentme> cool. got it.
08:02 < waxwing> delinquentme, yes it's arubi here
08:02 < delinquentme> arubi = fivepence?
08:02 < waxwing> well, fivepiece but nearly :)
08:03 < delinquentme> Mind the global exchanges!
08:27 < delinquentme> Oh and I also missed the part if anyone answered my question about 'cash deposits' for localbitcoins =]
08:28 < delinquentme> Has anyone tried to do an atm 'cash deposit' and do you know if the final step asks you for confirmation of the users first and lastnames
08:33 -!- RainMan28 [~RainMan28@unaffiliated/rainman28] has joined #joinmarket     
08:42 < waxwing> delinquentme, maybe ask in #bitcoin ?
08:50 < delinquentme> On the install.sh or the setupall.py ... which is the preferred installation method for totally new nodes on joinmarket-clientserver?
08:51 < belcher> install.sh
08:53 < delinquentme> It seems thats in conflict w the explicit directions given in: https://github.com/JoinMarket-Org/joinmarket-clientserver/blob/master/docs/INSTALL.md
08:53 < delinquentme> the quickstart here: https://github.com/JoinMarket-Org/joinmarket-clientserver  says use install.sh 
08:54 < delinquentme> but the INSTALL.md uses setupall.py ( which is what worked for me )
08:55 < waxwing> delinquentme, the README says something like 'use install.sh but there is another alternative in INSTALL.md' which basically means, the latter is there for either expert users, or someone for whom the quick install doesn't work
08:55 < waxwing> that was the idea anyway
08:55 < waxwing> arubi just made a PR to allow the --develop flag to be used for install.sh, so prob even an expert user should use it, assuming everything works.
08:59 < RainMan28> does the joinmarket clientserver work with bitcoin core 0.15.0.1? just found out about joinmarket and going to set up a server and yield generator
08:59 < delinquentme> Cool. Oh and I have yet to get the tumbler test to complete.  Is this normal?
09:00 < delinquentme> OH! and my main cruncher box has finished downloading the blockchain and my generated wallets finally work <333 !
09:01 < waxwing> delinquentme, wait which tumbler test is this? i lost track
09:01 < waxwing> you can run ygrunner.py and then do a tumbler against it on regtest, i did that a bunch of times
09:01 < waxwing> to sy the least :)
09:01 < waxwing> RainMan28, sure, it should do
09:02 < delinquentme> the entire test for test/test_tumbler.py has yet to finish completely once.  Ive run it individually too
09:02 < waxwing> in which repo?
09:02 < RainMan28> thanks waxwing, installing it right now. are there any hardware requirements for running a clientserver or yield generator? as in RAM, CPU, etc
09:02 < delinquentme> RainMan28, yeah i've got the newest btc core running on a Raspi and it works just fine
09:02 < delinquentme> waxwing, joinmarket vanilla
09:02 < RainMan28> oh nice if it works on a raspi then i'm not worried about hardware :)
09:03 < delinquentme> ya haha.  Do mind the architecture though.
09:03 < waxwing> delinquentme, right. you'll notice it's not included in the build test. it's fiddly and takes a long time.
09:03 < delinquentme> waxwing, oh ok. So I shouldnt expect it to run completely.  Does it run better on clientserver?
09:04 < waxwing> well, it can take as little as 5-7 minutes i guess, but anyway. i'm not inclined to work on it.
09:04 < delinquentme> got it.  
09:04 < waxwing> yes one of the bigger things i did in the jmcs repo was to make the tumbler work more robustly, mostly via a "tweak" mechanism.
09:04 < delinquentme> OK so I want to start playing w coinjoins and the tumbler w real btc.  clientserver is the best for doing that right?
09:04 < RainMan28> waxwing: sorry if this has been answered on the wiki, i was looking around but didn't quite find it. when we run a yield generator, are our coins at risk? 
09:05 < waxwing> RainMan28, it's a hot wallet, so it's risky in that sense. it's basically similar to having a bitcoin core wallet on that machine.
09:06 < waxwing> in theory your coins are at risk, but only from that machine being hacked, or from software bugs; not from counterparties, due to how coinjoin works.
09:06 < waxwing> i don't think anyone's lost coins to a bug though. up to you.
09:06 < waxwing> delinquentme, as to what's best with real btc, no real answer. there's still more people using the non-segwit joinmarket version, for now.
09:07 < delinquentme> oh so joinmarket vanilla is non-segwit ... and clientserver is segwit?
09:07 < waxwing> delinquentme, yes
09:08 < delinquentme> and then bc # of users the market for joinmarket legacy is bigger?
09:08 < waxwing> yes     
09:09 < waxwing> im hardly seeing any transactions on the new version (there were a fair few the first week, people testing no doubt)
09:09 < RainMan28> waxwing: perfect explanation thank you
09:09 < waxwing> now you can use it without Core (electrum) i thought it might attract a *bit* more interest, but users have inertia, and there's also the matter of whether they want to send to '3' addresses.
09:11 < waxwing> quite happy for people to switch to it slowly, it's up to them. it'd be nice for it all to be in one group, but i fear sending to '1' addresses in a coinjoin where everything else is a '3' address is too much of a privacy downgrade.
09:12 < delinquentme> waxwing, ohhh are joinmarket wallets the only ones which are currently starting w '3'?
09:12 < delinquentme> IE were way ahead of the blockchain in its namespace?
09:13 < waxwing> delinquentme, no far from it. i'm talking about the difference between joinmarket and joinmarket-cs (latter is segwit p2sh wrap, uses 3 addresses, former uses 1 addresses)
09:14 < delinquentme> and the jm-cs wallets with 3 are downgraded in privacy bc what?
09:14 < waxwing> they're not, but they might be if your destination address is a '1' address. see description above ^
09:16 < delinquentme> yeah im a blockchain newb so I dont get why a 1 vrs a 3 addy downgrades privacy =]
09:16 < GitHub117> [joinmarket-clientserver] AdamISZ pushed 1 new commit to master: https://git.io/vdkQQ
09:16 < GitHub117> joinmarket-clientserver/master c591f90 AdamISZ: bugfix binascii ref
09:17 < waxwing> delinquentme, well it's a coinjoin thing i guess not a generic blockchain thing; the equal-sized outputs (that we call coinjoin outputs) are ideally indistinguishable (although there's a ton of things to say about that), so having 1 out of 5 of them be a different address type is suboptimal
09:18 < delinquentme> and the reason coinjoin outputs are going to 3 addresses is?
09:19 < waxwing> segwit addresses can't be '1' addresses, they're either '3' addresses (p2sh wrapped segwit scriptpubkeys) or bech32 (new addr format for segwit, not supported by most wallets yet)
09:25 < RainMan28> waxwing: if i am already running a core node for an electrum server, should i just run clientserver with that or are there any benefits to setting up a separate core node?
09:27 < belcher> using the same one should work perfectly fine
09:27 < belcher> it only needs the rpc username/password
09:27 < belcher> joinmarket will add addresses to wallet.dat as watch-only so watch out for that if you use the wallet.dat for other stuff
09:40 < RainMan28> belcher:  yeah i don't use the wallet.dat to actually store any funds currently, just use it for electrum users
09:43 < delinquentme> OHHH! So sagaman just found an interesting thing in the long install. https://github.com/JoinMarket-Org/joinmarket-clientserver/blob/master/docs/INSTALL.md
09:44 < delinquentme> It has the user create the venv before cloning joinmarket-clientserver ... therefore how could it be properly located within joinmarket-clientserver?
09:44 < delinquentme> additionally the python setupall.py --daemon will create another jmvenv... I can fix this if you guys like?
09:52 < delinquentme> belcher waxwing ^
09:52 < GitHub77> [joinmarket-clientserver] sangaman opened pull request #92: Update INSTALL.md (master...patch-1) https://git.io/vdkbv
10:14 < adlai> re: market segregation - couldn't we a hybrid system, where address types are randomized?
10:14  * adlai accidentally a verb...
10:16 < waxwing> yeah on reflection i was maybe too lazy about it adlai 
10:17 < delinquentme> I just dropped my phone. Can one of you guys pass it to me?
10:17 < adlai> well it's probably best to start off with a separate market; easier to reason about
10:17 < delinquentme> plez.
10:17 < waxwing> delinquentme, yes there's something wrong there
10:17 < waxwing> adlai, obligatory xkcd :)
10:17 < delinquentme> waxwing, in the install? Yeeah I think he fixed it.
10:18 < waxwing> he fixed it? \o/
10:18 < adlai> waxwing: which one? i might be partitioned here, haven't validated new xkcd blocks in looong time...
10:18 < waxwing> adlai, the "new standard" one
10:19 < adlai> heh. but i'm not thinking of a new protocol, per se, although it might be wisest to use a different offer type anyways to avoid incompatibilities
10:19 < adlai> depending on exactly how much validation is done on the transactions, it could work as just new behavior within the existing protocol
10:20 < adlai> but before we even think about that, there's the question of whether this even works
10:21 < adlai> trivial counterexample - perfect randomization could result in the taker's cjout being the only one of its kind... then again, there would also be false positives that look that way
10:21 < adlai> we could require that all cjouts be of the same type as the taker's cjout
10:22 < delinquentme> waxwing, yeah. He made a pull request too.  Shown above.  Im gonna push him through this on RasPi Stretch so we have an idea on what it needs
10:22 < adlai> but allow randomization for the others. the trouble is, you can't validate that your counterparties are randomizing...
10:23 < delinquentme> https://github.com/JoinMarket-Org/joinmarket/issues/634
10:23 < delinquentme> whos paying this bounty :D ?
10:25 < delinquentme> what does the command for tumbling a wallet look like?  $ python tumbler.py wallet.json ?
10:27 < arubi> phew that #89 sure did blow up
10:27 < adlai> delinquentme: https://github.com/JoinMarket-Org/joinmarket/wiki/Step-by-step-running-the-tumbler#example-tumbling-into-your-wallet-after-buying-from-an-exchange-to-improve-privacy
10:28 < delinquentme> arubi, yeah sorry I wanted to get to the bottom of it =/
10:28 < delinquentme> ( I see a big future for running raspi tumbler  nodes  =] )
10:28 < arubi> ah don't worry.  I'm not sure if I could help further without access to a pi device..
10:29 < arubi> so where does it stand now?  still the same error building secp256k1?
10:30 < delinquentme> The big issue was his packages.
10:30 < delinquentme> weve got that solved and now were seeing the simplest way to get the secp256k1
10:30 < delinquentme> also ... do we want people installing + compiling core for the default install?
10:31  * adlai has no problem with requiring this, although obviously it inhibits 'mass adoption'
10:31 < delinquentme> Seems there would be more version needed to get that to work ... instead of using more standardized release versions
10:31 < adlai> and, we do have an electrum interface now
10:31 < arubi> yea and tails folks aren't going to run a full node (mostly)
10:32 < arubi> with current jm-cs afaik you don't even need the full node on the same machine, so there's that
10:32 < delinquentme> IDK about electrum so i dont follow on the electrum node :D
10:32 < delinquentme> arubi, really?
10:32 < arubi> should be, yea.  you just set an ip and rpc user\pass
10:33 < delinquentme> oh i guess configuring the rpir interface to a local ip ?
10:33 < delinquentme> yah. hmm. Maybe its wise to give the users a single-install process as much as possible ... and leave configuration up for later?
10:34 < adlai> it's possible to use a remote node with an ssh reverse tunnel
10:34 < adlai> (for the walletnotify port)
10:34 < arubi> I'm not sure it uses the *notify stuff anymore really
10:35 < arubi> oh wait yea it does
10:35  * adlai means for the "vanilla" joinmarket
10:35 < arubi> ah, right
10:43 < delinquentme> arubi, do you know if secp256k1 depends on the cryptography pip package?
10:44 < arubi> no, the other way around
10:44 < arubi> err wait
10:45 < arubi> so if cryptography is the secp256k1 bindings, then what I said
10:45 < arubi> not sure if it's related.  libsecp256k1 has very few deps and non are python stuff
10:46 < arubi> delinquentme, sorry, deep in qemu here re-trying setting up a virtual pi :)
10:49 < delinquentme> a virtual pi?
10:50 < delinquentme> hahahah
10:50 < arubi> someone needs to run raspbian stretch right :)
10:54 < waxwing> pretty sure no deps/connections between cryptography and secp256k1-py
10:55 < waxwing> the former's only there for twisted
10:55 < waxwing> well, not sure, just think so
10:58 < delinquentme> arubi, watch youll finish it first
11:00 < arubi> :( https://i.imgur.com/G4jW9U6.png
11:03 < delinquentme> whew! bought me some time
11:04 < belcher> anyone tried to install jm-cs on tails ?
11:05 < belcher> jm on tails could tunnel to a full node on another computer to avoid having to sync on tails, but then traces of the wallet would be outside the tails environment which might defeat the privacy point
11:07 < arubi> it's running on debian 8 and 9, and I'm pretty sure I tested 7 at some point, so should run on tails
11:08 < delinquentme> is there something akin to updating repos for pip?
11:12 < arubi> pip --install --upgrade will upgrade a single package.  not sure if there's anything else
11:17 < delinquentme> oh and is this bounty real? https://github.com/JoinMarket-Org/joinmarket/issues/634
11:23 < belcher> joinmarket-clientserver already has that feature, with --restart
11:24 < delinquentme> should I link + close it?
11:28 < waxwing> yeah testing on tails is a good thing to do. testing an electrum connection would be interesting, i guess it *might* work with torsocks. belcher 's idea for Core is for sure interesting.
11:29 < waxwing> delinquentme, yeah it's real, was paid, and it's true that --restart on jmcs gives that effect, more or less (i'd argue it's a bit better, but not exactly the same).
11:30 < waxwing> belcher, re 'traces of the wallet', true. difficult to know how people'd think of that. at least for coin security it'd not be an issue, i guess they could wipe it between runs somehow.
11:30 < delinquentme> me gusta.
11:30 < belcher> yep     
11:30 < belcher> its better than nothing ofc
11:30 < belcher> off topic https://bitcointalk.org/index.php?topic=2199659.0
11:30 < belcher> armory SSS discovered to be broken
11:31 < waxwing> belcher, yes was just looking at it. very interesting.
11:32 < arubi> oh sweet, something to read while rebuilding qemu :P
11:35 < arubi> "where the first coefficient is deterministically derived from the secret, and the subsequent ones are derived successively from one another"
11:36 < delinquentme> ugh I might have borked something in my local dev env... whats a joinmarket command that will use secp256k
11:36 < delinquentme> ?     
11:37 < arubi> wallet-tool.py probably
11:37 < waxwing> yes, easiest, or bare-bones is `python; import secp256k1` in the venv
11:37 < arubi> oh right
11:38 < arubi> it's a python library eventually :)
11:38 < arubi> I feel like I could spot the first broken SSS but definitely not the second one
11:39 < belcher> waxwing you were saying that you connect to cyberguerilla via the onion and it always works?
11:39 < arubi> heh, easy to say that after reading the post :)
11:39 < waxwing> well i don't know about always, but it's been reliable for me last few weeks, yes
11:40 < waxwing> i had ob-watcher on clearnet and i noticed it went down for a while, a few days ago
11:40 < waxwing> while onion was ok
11:41 < belcher> cgan has been crappy for me for months, but yes the onion just worked perfectly
11:41 < belcher> interesting that joinmarket.cfg allows you to use the onion for one irc and clearnet for the other :)
11:42 < waxwing> arubi, what's f(x) in that write-up? i got stuck at that point
11:42 < waxwing> oh SSS f(x) i see
11:43 < belcher> oh btw who was asking about "unknown tx" in wallet-tool history
11:43 < belcher> i got the same issue because i didnt use a new wallet.dat when moving to segwit
11:43 < belcher> the way you fix it is `if len(our_inputs) == 0 and len(our_outputs) == 0: continue`
11:44 < belcher> i.e. dont print out the transaction if its relevant to our wallet
11:45 < waxwing> belcher, ok makes sense.
11:45 < waxwing> i assumed i got it on regtest due to coinbases
11:48 < belcher> i looked at the code and it turns out it basically obtains every single transaction in wallet.at
11:48 < belcher> which always worked for me because i only used wallet.dat for joinmarket, but in other situations it would give lots of "unknown tx"
11:48 < waxwing> right, got it
11:49 < waxwing> so this armory thing makes me wonder if there is an RFC or something for SSS(S)
11:59 < arubi> ah was just afk,  there this, https://tools.ietf.org/html/draft-mcgrew-tss-03 , not sure how related to sss it is :)
12:00 < arubi> "Shamir's original description treats the secret as a large integer
12:00 < arubi>    modulo a large prime number..."  maybe it is
12:04 < delinquentme> runnign my first tumbler: https://pastebin.com/raw/FTdG7Cpp
12:04 < delinquentme> hows that look?
12:04 < delinquentme> did i break it ?
12:06 < waxwing> delinquentme, is HAPPYHASH an address? you know you're supposed to use 3 addresses minimum?
12:06 < delinquentme> that was a single address
12:06 < delinquentme> 3 minumum?
12:07 < delinquentme> I wasnt sure. No examples of a run in the docs. 
12:07 < waxwing> ok, one thing that'll help you a bit: set console_log_level to DEBUG in joinmarket.cfg for more on-console information
12:07 < delinquentme> python tumbler.py wallet.json HASH1 HASH2 HASH3
12:07 < delinquentme> ?     
12:07 < waxwing> second thing is the main wiki page: https://github.com/JoinMarket-Org/joinmarket/wiki/Step-by-step-running-the-tumbler
12:08 < waxwing> third thing is some extra advice for jm-cs version: https://github.com/JoinMarket-Org/joinmarket-clientserver/blob/master/docs/tumblerguide.md
12:08 < waxwing> latter is a delta to the former; gives you some extra info about how it works in jm-cs vs original, but the same command line flags work
12:09 < waxwing> you can also do it in the GUI, it has a wizard, but, no need for that.
12:09 < delinquentme> cool ywah more stuff is happening
12:09 < delinquentme> will paste in a sec
12:10 < waxwing> i'd recommend `tail -f`-ing logs/TUMBLE.log to see the updates happening in real time. not necessary, of course.
12:10 < waxwing> it's kind of condensed, because what comes out on console is a bit wall-of-text, whereas TUMBLE.log gives a much better idea of what's happening/where you are in the process.
12:11 < delinquentme> Theres no reason I need to keep the external wallet addresses private is there?
12:11 < waxwing> in general not a good idea to try to do amounts < say, 0.2 btc. you'll too often hit edge conditions of too-small amounts.
12:12 < waxwing> well no reason to broadcast addresses, generally. 'external' can be given to people to pay to you.
12:12 < waxwing> is this mainnet? delinquentme 
12:12 < delinquentme> wait so I need to have > $780usd in this for it to run??
12:13 < waxwing> i think < $100 is basically not a good idea at all. even when fees are low, they add up for say 10-15 txs all of which are large. there is a paragraph about that on the wiki page, which is out of date re: numbers, a bit.
12:13 < delinquentme> waxwing, IDK what mainnet is =]
12:14 < waxwing> but to minimize fees, consider setting tx_fees = N where N is a number of satoshis per kB, if you set N < 144 it is interpreted as a block target for confirmation, if higher, then it's interpreted as satoshis per kilobyte
12:14 < waxwing> mainnet is "real" bitcoin as opposed to testnet or regtest
12:15 < waxwing> generally sunday is the best day to do testing on mainnet, if you feel the need to do that :)
12:15 < belcher> aww sadface "Is this supposed to work on windows? When I try to run install.sh I get the error, "./install.sh: line 61: virtualenv: command not found""
12:16 < delinquentme> waxwing, whats a typical way to test the functionality in increasing steps of 'seriousness'
12:17 < waxwing> delinquentme, regtest, testnet, mainnet i guess
12:17 < delinquentme> like am I jumping into the deepend trying to tumble?  should i be doing something else first w mainnet / real btc
12:17 < waxwing> there are still 3 of my lonely bots running on testnet :)
12:18 < waxwing> delinquentme, yes you hit a difficult point, i myself have only run tumbler tests on mainnet a couple of times, it's too expensive. if you know what you're doing you can do a test in ~ 1 day and spend maybe $20 in fees at the low end (like 0.15-0.2btc). i dunno, maybe less over last couple of weeks when we had very low fees.
12:19 < waxwing> but you can simply do 1 or 2 sendpayment runs to check it's working on mainnet, and set the fees on the low side.
12:19 < waxwing> i recently did coinjoins on mainnet for maybe 50c ... not sure about today, but recently it's been cheap, and segwit shaves off another 30% or so.
12:20 < waxwing> yeah come to think of it, i'm probably stuck in the old fees of earlier this year, maybe you can do a full tumbler run for < $10 nowadays without much trouble. not sure.
12:24 < delinquentme> ok noted. Ill be back in a bit. gotta get the girl
12:27 < waxwing> #noted
12:29 < adlai> belcher: did you catch my musings about how to unify the liquidity across address types?
12:29 < belcher> yeah i saw
12:30 < belcher> another idea was to have both types in one wallet and have them slowly move across to segwit as transactions happen
12:30 < belcher> but i think we decided to save developer time and did it this way
12:30 < adlai> i think the only hard requirement is that the cjouts all be the same format
12:30 -!- delinquentme [~delinquen@142-254-104-244.dsl.dynamic.fusionbroadband.com] has quit [Ping timeout: 240 seconds]     
12:31 < belcher> id say right now we should get as many users to move to segwit 
12:31 < belcher> if you're thinking of liquidity splitting across orderbooks
12:32 < waxwing> yes i was simply lazy about not doing the "change->p2sh address" thing. but i was also worried about assessing the privacy implications of mixed transactions. even if only changeouts, one could probably come up with arguments to say it's inferior privacy-wise.
12:32 < waxwing> but meh it was mostly laziness, too many things to do.
12:33 < adlai> the observation that destination format is really what matters, is a good one
12:33 < adlai> it doesn't matter how much users might want to use SW, if they're sending to a p2pkh address, they're currently still stuck using the p2pkh market
12:34 < waxwing> adlai, right, i didn't even think about that really, but it does seem a very important practical point.
12:35 < adlai> btw i've seen a rather steady trickle of SW CJs, although as you observed, far less than p2pkh
12:35 < waxwing> it's slightly amazing to me that not all exchanges, for example, are using p2sh addresses now. i've kind of miscalculated on that one.
12:36 < adlai> to me it seems too small an amount to characterize as having been higher at the start; then again, i wasn't providing liquidity there during the first days
12:36 < waxwing> yeah i guess first week i was thinking of. did see a bunch of txs that were clearly just tests. probably people here :)
12:38 < belcher> maybe time to turn on the spaminator again to increase segwit adoption
12:39 < adlai> https://jochen-hoenicke.de/queue/#8h - this looks reasonably "organic", although it could just be the spaminator starting up in slow gear
12:50 < waxwing> set phasers to spam
12:55 -!- delinquentme [~delinquen@184.250.69.92] has joined #joinmarket     
12:56 < belcher> tehehe
13:03 < delinquentme> this channel isnt logged is it?
13:04 < gmaxwell> adlai: just looks like monday to me, esp combined with bcash taking a bit of hashrate over the weekend.
13:05 < adlai> delinquentme: there are no official public logs, but some folks use IRC clients logs by default
13:06 < delinquentme> im jumping on and off of internet connections so I just wanna make sure im not missing any responses.
13:09 < adlai> gmaxwell: well yes, it's also completely dwarfed by the may and august floods
13:21 < delinquentme> adlai, you mean of users into joinmarket?
13:24 < adlai> delinquentme: https://jochen-hoenicke.de/queue/#all
13:25 < adlai> joinmarket usage seems relatively flat in 2017
13:53 -!- delinquentme [~delinquen@184.250.69.92] has quit [Ping timeout: 260 seconds]     
13:58 < belcher> flat even when blockr.io broke adlai ?
13:58 < belcher> interesting, maybe nobody actually used it
13:59 -!- arubi [~ese168@gateway/tor-sasl/ese168] has quit [Remote host closed the connection]     
13:59 -!- delinquentme [~delinquen@108-235-112-153.lightspeed.sntcca.sbcglobal.net] has joined #joinmarket     
14:00 -!- arubi [~ese168@gateway/tor-sasl/ese168] has joined #joinmarket     
14:18 < adlai> it seems to have degraded rather than vanishing at once, so i imagine people learned to avoid it
14:20 < belcher> im sure the high miner fees at many points did their job too
14:33 < delinquentme> cool. fixed the issue w secp256k1 on raspi stretch!
14:35 < delinquentme> $ sudo apt-get update 
14:35 < belcher> would you be able to write a short blog post about installing this all on a pi?
14:35 < belcher> any gotchas and unobvious things
14:35 < delinquentme> belcher for sure.
14:36 < delinquentme> I've got my install process =]
14:38 < arubi> what.. the fix was `sudo apt-get update` ?
14:42 < delinquentme> arubi, ;p
14:42 < delinquentme> yeap!
14:42 < delinquentme> it was an issue with wheel methinks
14:43 < arubi> with wheel?
14:43 < delinquentme> Building wheels for collected packages: secp256k1
14:43 < delinquentme>   Running setup.py bdist_wheel for secp256k1 ... error
14:44 < delinquentme> https://pastebin.com/raw/Qr4B1QVF
14:44 < delinquentme> it could also be python-pip though.
14:44 < arubi> the error is less important, is it that the person never ran `apt-get update` before?
14:45 < delinquentme> arubi, :P I dont often.  That update can take a long time
14:46 < arubi> what do you mean?  apt-get update is necessary to be able to use apt-get install at all
14:46 < arubi> apt-get update just fetches indexes and whatnot, it doesn't actually update packages
14:46 < arubi> apt-get upgrade, or dist-upgrade does
14:46 < delinquentme> its possible i've confused the two
14:46 < arubi> arg     
14:46 < delinquentme> yeah.  But I've been able to install packages without running update ...
14:47 < delinquentme> I think the most common situation is ec2 instances ... so maybe thats thanks to amazon.
14:48 < delinquentme> On tumbler.py ... this should only be run w 3 addresses that are of the same level ?
14:48 < arubi> I really don't follow, thanks for what to amazon?  I can't imagine running debian \ ubuntu without ever running apt-get update
14:48 < arubi> does not compute
14:50 < arubi> so the issue says `sudo apt-get upgrade`, now that makes some sense
14:51 < arubi> would be nice to somehow reproduce this
14:51 < delinquentme> I think we can get him to reproduce ... I think hes finishing the install process ... and then redoing it all again once he has the build.
14:52 < arubi> that would be very interesting
14:54 -!- takamatsu [~takamatsu@unaffiliated/takamatsu] has quit [Ping timeout: 248 seconds]     
14:56 < arubi> delinquentme, do you have the 'config.log' file from the successful secp256k1 build? 
14:57 < arubi> oh sorry I bet it's at some temp directory
15:00 < arubi> doh.
15:00 < arubi> https://bitcoin.stackexchange.com/questions/49937/cannot-build-secp256k1-c-library-on-debian-stretch
15:01 < delinquentme> damnit what were your search terms?
15:01 < gmaxwell> arubi: that is really old.
15:01 < gmaxwell> arubi: and doesn't apply anymore.
15:01 < delinquentme> maybe I should have just added "stack overflow"
15:02 < arubi> this is the exact error from config.log
15:02 < gmaxwell> arubi: then you're building a really old copy of libsecp256k1?
15:02 < arubi> not me, person on github issue with raspbian stretch
15:02 < arubi> apparently `apt-get upgrade` somehow solved it
15:02 < delinquentme> idk. its on a RasPI 
15:03 < delinquentme> but it runs now
15:03 < arubi> issue #89 on joinmarket-clientserver
15:10 < delinquentme> Building wheels for collected packages: secp256k1
15:10 < delinquentme>   Running setup.py bdist_wheel for secp256k1 ... done
15:11 < delinquentme> https://pastebin.com/raw/GbkBYUwu
15:11 < delinquentme> compared to the failing one
15:11 < delinquentme> cool! happy we fixed it <3
15:22 < arubi> so I'm guessing at some point `apt-get install libsecp256k1-0` or libsecp256k1-dev was run
15:23 < arubi> maybe even a while ago, and now with the upgrade it's at a newer version, system wide
15:24 < arubi> delinquentme, is this output from sangaman's machine?
15:24 < delinquentme> ya     
15:24 < arubi> alright, thanks
15:24 < delinquentme> yup. no problem.
15:26 < delinquentme> Now before I go dumping any real amounts of btc into this ... do we have anyone here whos more of a security person?
15:26 < arubi> delinquentme, could you also please run `apt-cache policy libsecp256k1-*` ? (no sudo)
15:26 < delinquentme> basically what kind of hardning should we do to a machine to make sure its tied down for holding btc?
15:27 < delinquentme> I think hes afk rn.. but ill ask
15:27 < arubi> okay
15:44 < delinquentme> ok so you guys are saying that ill be spending near $20 in costs .. to run a tumbler on real btc.... I understand that the tumbler is multiple levels of transactions happening ...but!  
15:51 < delinquentme> is the number of coinjoins the tumbler executes controllable?
15:53 < belcher> yes all of that is controllable
15:53 < belcher> you can even make tumbler pay lower miner fees
15:53 < belcher> for example pay low enough that you always confirm only in the middle of the night for max cheapness, then your entire tumbler run takes 2 weeks
15:53 < belcher> try tumbler.py --help
15:54 < belcher> you control miner fees in joinmarket.cfg
15:56 < delinquentme> cool so a balance of fees paid, time elapsed, and count of coinjoins can be tailored to roughly what I want to pay
15:58 < belcher> and to how much privacy you want
15:58 < delinquentme> more coinjoins = more privacy 
15:58 < delinquentme> got it.
15:58 < belcher> also you can configure number of coinjoin counterparties and range of coinjoin amounts
15:59 < belcher> and other stuff probably, read --help
16:00 < delinquentme> belcher do you want me to write up the post on the raspi in the wiki?
16:01 < belcher> yeah sure
16:01 < belcher> put somewhere "as of september 2017" since the instructions will probably change over time
16:02 < belcher> actually
16:02 < belcher> a reddit thread or medium blog post might be more appropriate, since the wiki is meant to be strictly for tutorials rather than experiances/feedback
16:04 < delinquentme> ya I think I can manage a reddit post
16:34 -!- delinquentme [~delinquen@108-235-112-153.lightspeed.sntcca.sbcglobal.net] has quit [Ping timeout: 260 seconds]     
16:53 -!- dan__ [b8bbb18f@gateway/web/freenode/ip.184.187.177.143] has joined #joinmarket     
16:56 < dan__> Question about 0.3.1.  I have had my maker bot up and running for over 24 hours.  No join yet but I know that is possibly within the norm.  In the past I would see messages going back and forth but now it just says JM daemon setup complete.  I do see my order at joinmarket.me, so I know it is there.  I am running ob-watcher.py but it just says started http server, visit http://localhost:62601/.  Nothing else.  Should I see more?
16:59 < belcher> dan__ you visit that url http://localhost:62601/
17:00 < belcher> that gives you a html page showing the offers
17:02 < dan__> ok, i see that.  It brings up the joinmarket.me/ob page.
17:03 < dan__> I recall in the past that that page did not always have the full list.
17:10 < belcher> joinmarket.me/ob is essentially just ob-watcher.py
17:10 < belcher> maybe it was broken or out of date at some points
17:13 < dan__> I assume you have seen some joins since 0.3.1 started?  I imagine it will take time for people to move over
17:16 < belcher> hmm     
17:17 < belcher> waxwing what did we say was the plan again? we talked about it a few days ago but i forgot what we concluded
20:19 -!- starsoccer [~starsocce@unaffiliated/starsoccer] has quit [Ping timeout: 255 seconds]     
20:25 -!- starsoccer [~starsocce@ns372404.ip-94-23-252.eu] has joined #joinmarket     
20:25 -!- starsoccer is now known as Guest93900     
21:15 -!- instagibbs [~instagibb@pool-72-83-36-237.washdc.fios.verizon.net] has quit [Ping timeout: 240 seconds]     
21:24 -!- instagibbs [~instagibb@pool-72-83-36-237.washdc.fios.verizon.net] has joined #joinmarket     
22:26 -!- instagibbs [~instagibb@pool-72-83-36-237.washdc.fios.verizon.net] has quit [Ping timeout: 248 seconds]     
22:29 -!- instagibbs [~instagibb@pool-72-83-36-237.washdc.fios.verizon.net] has joined #joinmarket     
22:59 -!- shinobimonkey [~truesdale@104.129.29.93] has joined #joinmarket     
22:59 -!- shinobimonkey [~truesdale@104.129.29.93] has quit [Remote host closed the connection]     
23:24 -!- arubi [~ese168@gateway/tor-sasl/ese168] has quit [Ping timeout: 248 seconds]     
23:41 -!- arubi [~ese168@gateway/tor-sasl/ese168] has joined #joinmarket