--- Day changed Thu May 10 2018
02:09 -!- M1 [~Michail@michail.com] has quit [Ping timeout: 240 seconds]     
02:12 -!- M1 [~Michail@michail.com] has joined #joinmarket     
02:46 -!- asymptotically [~asymptoti@gateway/tor-sasl/asymptotically] has joined #joinmarket     
02:48 -!- windsok [~windsok@unaffiliated/windsok] has quit [Remote host closed the connection]     
03:06 -!- akrmn [~akrmn@79.red-83-54-170.dynamicip.rima-tde.net] has quit [Quit: Leaving.]     
03:11 -!- belcher [~belcher@unaffiliated/belcher] has joined #joinmarket     
04:53 -!- diogorsergio [~diogorser@81.92.202.8] has joined #joinmarket     
05:40 -!- diogorsergio [~diogorser@81.92.202.8] has quit [Quit: diogorsergio]     
06:53 -!- quitobro [~quitobro@pool-108-41-0-186.nycmny.fios.verizon.net] has joined #joinmarket     
06:57 -!- quitobro [~quitobro@pool-108-41-0-186.nycmny.fios.verizon.net] has quit [Ping timeout: 240 seconds]     
07:04 -!- quitobro [quitobro@gateway/vpn/privateinternetaccess/quitobro] has joined #joinmarket     
07:14 -!- Guest77756 is now known as Pilate     
07:39 -!- dx25 [~dx25@97.119.180.105] has joined #joinmarket     
08:00 -!- arubi [~ese168@gateway/tor-sasl/ese168] has quit [Remote host closed the connection]     
08:01 -!- arubi [~ese168@gateway/tor-sasl/ese168] has joined #joinmarket     
08:16 -!- raedah [~x@199.19.95.188] has quit [Quit: WeeChat 2.1]     
08:21 -!- GitHub185 [GitHub185@gateway/service/github.com/x-vhyciuecymdfbged] has joined #joinmarket     
08:21 < GitHub185> [joinmarket-clientserver] AdamISZ closed pull request #144: Don't calculate and display interest rate if profit is zero (master...zero-profit) https://git.io/vpXPM
08:21 -!- GitHub185 [GitHub185@gateway/service/github.com/x-vhyciuecymdfbged] has left #joinmarket []     
08:21 -!- GitHub178 [GitHub178@gateway/service/github.com/x-avlajmmivrklnuan] has joined #joinmarket     
08:21 < GitHub178> [joinmarket-clientserver] AdamISZ pushed 2 new commits to master: https://git.io/vp1PC
08:21 < GitHub178> joinmarket-clientserver/master c089063 Kristaps Kaupe: Don't display interest rate if profit is zero...
08:21 < GitHub178> joinmarket-clientserver/master 562b786 AdamISZ: Merge #144: Don't calculate and display interest rate if profit is zero...
08:21 -!- GitHub178 [GitHub178@gateway/service/github.com/x-avlajmmivrklnuan] has left #joinmarket []     
08:32 -!- GitHub4 [GitHub4@gateway/service/github.com/x-flbdamolbkpevhkb] has joined #joinmarket     
08:32 < GitHub4> [joinmarket-clientserver] AdamISZ pushed 2 new commits to master: https://git.io/vp1XB
08:32 < GitHub4> joinmarket-clientserver/master 76e4a7a James Hilliard: Alternative approach to connection reset.
08:32 < GitHub4> joinmarket-clientserver/master 622edaf AdamISZ: Merge #140: Alternative approach to connection reset....
08:32 -!- GitHub4 [GitHub4@gateway/service/github.com/x-flbdamolbkpevhkb] has left #joinmarket []     
08:32 -!- GitHub177 [GitHub177@gateway/service/github.com/x-tzdtqmhwuttitmsz] has joined #joinmarket     
08:32 < GitHub177> [joinmarket-clientserver] AdamISZ closed pull request #140: Alternative approach to connection reset. (master...conn-reset) https://git.io/vprgS
08:32 -!- GitHub177 [GitHub177@gateway/service/github.com/x-tzdtqmhwuttitmsz] has left #joinmarket []     
08:32 -!- Pilate [~pilate@infoforcefeed/pilate] has left #joinmarket []     
08:38 -!- d3spwn [53a19dc8@gateway/web/freenode/ip.83.161.157.200] has quit [Ping timeout: 260 seconds]     
08:42 < waxwing> oh balls it broke the orderbook watch test, heh
08:46 < waxwing> arubi, what's the difference between the two build jobs? As shown here: https://travis-ci.org/JoinMarket-Org/joinmarket-clientserver/builds/377331393?utm_source=github_status&utm_medium=notification
08:47 < waxwing> this is weird, i can't reproduce the error. either it's non-deterministic or perhaps it's due to being on a different enviro/OS?
08:49 < arubi> waxwing, one is a linux (some ubuntu flavor) that builds, installs the venv and then runs the run_tests.sh script.  second is osx that only builds and installs the venv
08:54 < waxwing> i'm a bit stumped, it seems to be reproducible on the linux build there, but i can't get it here; just get a connection refused on that one test: https://travis-ci.org/JoinMarket-Org/joinmarket-clientserver/jobs/377331394#L9256
08:56 < waxwing> oh i just remembered finding that PR mysterious because it didn't trigger travis. weird.
08:57 < arubi> did the failures start with this push?
08:57 < arubi> yea seems so
09:05 < arubi> seems that bitcoind is running and generating blocks.  weird
09:16 < arubi> gonna try here in a bit too
09:28 < waxwing> yeah it must be this one
09:30 < arubi> so waxwing, the command that tells bitcoind to generate blocks, does that go through rpc or does python actually executes a command in a shell \ subprocess?
09:31 < waxwing> hmm, i'll have to check, i think it's a shell
09:32 < waxwing> https://github.com/JoinMarket-Org/joinmarket-clientserver/blob/master/conftest.py#L100-L105
09:32 < waxwing> meh that code is old and a bit crappy, but not sure why it would be related here?
09:32 < arubi> yea so at least it makes sense that bitcoind generates
09:32 < arubi> well if it were by rpc then it would be weird that it worked
09:33 < arubi> I can see bitcoind generates blocks, so was wondering if it's either using a different rpc interfact, or a shell command
09:33 < arubi> interface*
09:38 < waxwing> arubi, oh sorry, i think i get your question now
09:38 < waxwing> yes there is a loop in regtest blockchain interface that generates blocks at intervals of a few seconds
09:39 < waxwing> tick_forward_chain etc. see the bottom of blockchaininterface.py
09:39 < arubi> ah cool.  I'm installing now and see if it fails locally for me
09:39 < waxwing> i can't remember, think it might be a thread actually. i made a point of removing all multithreading maybe apart from that.
09:39 < waxwing> it's so long ago i've got to read it now :)
09:40 < waxwing> ah no it's not a thread, it's a task.LoopingCall (twisted)
09:40 < waxwing> why are you looking at this though? are you seeing a connection?
09:40 < waxwing> cos i see the crash in load_program_config in the test_orderbookwatch test, no?
09:43 < arubi> reason I was asking is only because I was curious if 'generate' via rpc was working when the rest of the stuff was giving 'connection refused', but now I'm looking at RegtestBitcoinCoreInterface and seems that it does actually use rpc?
09:44 < waxwing> yes, sorry if i wasn't clear before, both are true
09:44 < waxwing> via shell at startup and in loop during run over rpc
09:45 < arubi> yep, anyway it's not refusing connection for me here, so must be travis specific..  trying to see if they offer updated versions of ubuntu, maybe python is old
09:45 < waxwing> btw (unrelated) i see this is the reason we don't get coverage: https://travis-ci.org/JoinMarket-Org/joinmarket-clientserver/jobs/377331394#L2093
09:45 < arubi> yea I just caught onto that too.  "incompatible".  weird
09:46 < waxwing> yeah i was just being dumb before, not reading the log from the start. of course it's crapping out all the way through, at least from where it's using the blockchain.
09:46 < waxwing> so maybe every rpc call is failing.
09:47 < arubi> seems like that's what's happening
09:47 < waxwing> ok i just realised my local tests are dumb too lol
09:47 < waxwing> because i'm not replicating your test script. how should i call the test script?
09:48 < arubi> from the jm root directory, `./tests/run_tests.sh`
09:48 < arubi> err, `./test/run_tests.sh`
09:48 < waxwing> ok. does it use the joinmarket.cfg in the root?
09:48 < arubi> it uses the regtest one.  copy yours somewhere else before you run the tests
09:49 < waxwing> k thx
09:49 < arubi> it does move it for you to ./joinmarket.cfg.bak if you have one, but better safe :)
09:50 < waxwing> arubi, is there a way to run it without the installation part?
09:51 < waxwing> i should just read it i guess. i wanted to replicate exactly how travis runs the test, but prob there's nothing interesting in this.
09:51 < arubi> the tests script won't install.  you just have to have venv active
09:52 < arubi> travis installs first with install.sh, then sources and runs run_tests.sh
09:57 < waxwing> arubi, but when i ran run_tests.sh i got a downloading thing and then "Could not open requirements file: [Errno 2] No such file or directory: './requirements-dev.txt'"
09:57 < waxwing> then 'packages could not be installed. exiting'
09:57 < waxwing> oh maybe it's this: i use a separate venv; i don't have it in jmvenv under that directory
09:57 < arubi> it's trying to pip install the packages required for tests.  are you running it from the jm git root directory?  should be run with ./test/run_tests.sh
09:58 < arubi> not ./run_tests.sh
09:58 < waxwing> yes i did that
09:58 < arubi> oh and that file is actually missing?
09:58 < waxwing> which file?
09:58 < arubi> './requirements-dev.txt'
09:58 < waxwing> hmm i'll check but: why is it trying to install anything?
09:59 < arubi> pexpect, coverage, pytest.. etc
09:59 < waxwing> yes the file is there.
09:59 < waxwing> yes but i don't want it to install anything :) but also, i have all of that in my virtualenv.
09:59 < waxwing> my virtualenv is not in ./jmvenv
10:00 < arubi> shouldn't matter, hm.
10:00 < arubi> well I can add a thing where it checks if these are already installed..  no way the tests will run without them
10:00 < waxwing> ok so then it is a pretty weird symptom: the file requirements-dev.txt is in that directory, but it complains it's not there.
10:01 < waxwing> ok i'll just try to make sure the pytest call in the script is the same one that i do anyway. this will all be irrelevant probably anyway.
10:01 < arubi> yea it shouldn't matter
10:05 < waxwing> Lightsword, ping; if you have any ideas that'd be great :) #140 breaks the travis tests, we get connectionrefused on either some or all of the rpc calls, consistently. i cannot replicate it in local tests. the travis test is on ubuntu.
10:05 < waxwing> i don't want to revert this merge since it seems to solve the mac os rpc problem, and i'm running it with no problems on linux.
10:07 < arubi> I'm reverting on my own branch and checking travis
10:10 < waxwing> what's the correct rpc port for regtest again?
10:10 < arubi> 18443 I think
10:10 < arubi> yea, "rpc_port = 18443"
10:12 < arubi> I'm almost positive it's travis' old python version acting up..  we'll see.  if this revert makes travis happy then I'll then check with the commit back and a newer python version.  travis has 2.7.5 which is really old
10:12 < arubi> yea tests are passing with the commit reverted
10:12 < waxwing> when i replace py.test with python -m py.test it craps out for some reason. i can't even remember why i originally preferred the former version.
10:17 -!- akrmn [~akrmn@79.red-83-54-170.dynamicip.rima-tde.net] has joined #joinmarket     
10:19 < waxwing> ah interesting. perhaps i can try with an old python, but i suppose it's not that easy if you don't have it already on the system.
10:21 < arubi> yea and if you wanna set it up, make sure it's not replacing the system python
10:25 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Excess Flood]     
10:25 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
10:49 < arubi> seems to fail the same with python2.7.12
11:02 < waxwing> ok, thanks for that.
11:02 < waxwing> i can't replicate it; so, you can replicate it locally arubi ?
11:04 < arubi> on travis https://travis-ci.org/fivepiece/joinmarket-clientserver/jobs/377383650
11:04 < waxwing> ah, got it
11:17 < waxwing> the delta isn't Core version right; i have 0.16.0 here and seems same on travis.
11:17 < waxwing> i also have 2.7.12 on this machine.
11:17 < arubi> yea seems the same..  really not sure what's going on
11:17 < waxwing> if anyone here is running JM on ubuntu, if you get a chance, please pull master and restart and see if you get any rpc connection errors either in wallet-tool or running a bot. i don't think you will; i'm not. but somehow travis does.
11:17 < waxwing> well ubuntu or anything really, doesn't for sure have to be connected with exact distro.
12:34 -!- quitobro_ [~quitobro@pool-108-41-0-186.nycmny.fios.verizon.net] has joined #joinmarket     
12:35 < belcher> waxwing what will you be speaking about at building on bitcoin?
12:37 < waxwing> i had a couple of different ideas: one was doing like an overview of the myriad different coinswap ideas. another was something similar to what i did in milan with coinjoin in contracts, tweaked perhaps. a third was doing a more didactic talk explaining schnorr's identity protocol as a zero knowledge proof and trying to make the signature algorithm intelligible. i actually kinda like the last one most, but it's not really appropriate for a 
12:37 < waxwing> conference like that i'm afraid.
12:37 -!- quitobro [quitobro@gateway/vpn/privateinternetaccess/quitobro] has quit [Ping timeout: 265 seconds]     
12:37 -!- quitobro_ is now known as quitobro     
12:37 < waxwing> actually i just sent the guy an email with a couple of paragraphs outlining, iirc those few ideas, i didn't realise they really wanted an "abstract", so it was a bit sloppy.
12:37 < waxwing> the shorter answer to your Q is "i don't know yet" :)
12:38 < belcher> ok     
12:38 < belcher> i didnt realize they wanted an abstract either, i sent a one-line description for each of three ideas (joinmarket, electrum personal server, lightning p2pool)
12:39 < waxwing> ah yeah. good idea. i think personally i'd find the last one the most interesting because i understand that stuff hardly at all. the middle one might appeal to people in terms of its practical applicability.
12:40 < belcher> if you do the coinswap one, make sure you mention how brilliant schnorr would be because then coinswaps would look exactly like a regular single-sig and be indistinguishable from non-mixing
12:40 < belcher> though i guess the same thing is possible with pallier in ecdsa
12:41 < belcher> getting into a situation where mixing happens without it even being visible is very good for fungibility
12:41 < waxwing> yeah. tbh i'm tending more in the direction of that one, out of the three. at some point i'll try to write an outline of it and will show you. if anything the problem will be keeping it from blowing up into too big of a topic.
12:43 < belcher> did you see the ML email about all the ideas for deploying schnorr/mast/taproot today
12:44 < belcher> i wonder if its worth actually emailing pointing out the fungibility benefits of having every feature in one kind of address/signature
12:44 < belcher> so people who want to use LN, people who want to aggregate signatures, people who want to use MAST, people who want to do coinswap, and people who want to do scriptless scripts all end up looking exactly the same on the blockchain
12:45 < belcher> the downside is then developers would probably have to invent and code everything at once and deploy it in one big SF
12:47 < waxwing> belcher, yeah i actually tweeted that ML post :) it was helpful
12:49 < waxwing> well hmm .. one kind of *address*, yeah - can this all end up p2sh though? i guess. as for one type of *signature*, hmm that seems like a more complicated question, def stuff to think about .. schnorr + taproot goes some way right?
12:50 < belcher> one kind of address and one kind of scriptSig/witness ideally
12:50 < belcher> yes schnorr + taproot + MAST goes a long way
12:51 < belcher> in that ML email he talks about possibly deploying MAST separately which might be bad from a fungibility point of view
12:51 < belcher> unless later deployments of schnorr also include taproot/MAST which then everyone adopts
12:52 < waxwing> this is why we off-chain :)
12:53 < belcher> can you rephrase?
12:57 < waxwing> well just stuff where it's really helpful for everyone to do the same thing, introduces this consensus problem, we've all got to agree to go the same way to get the benefit, you don't have that problem if the enhancement is something that doesn't affect the blockchain (base layer)
12:58 < belcher> oh yes
12:59 < belcher> i suppose you could say the great thing about all this schnorr stuff is it makes layer-2 stuff indistinguishable from just a single-sig on-chain transaction
13:00 < belcher> on the other hand, LN can lead to every UTXO of each channel being announce in public... but once channels are closed they wont be announced anymore and will likely be slowly forgotten
13:16 < waxwing> Here's a bit of fun: gmaxwell proposed a kind of "theorem" about contracts in the taproot post: "I believe that _any_ contract with a fixed finite participant set
13:16 < waxwing> upfront can be and should be represented as an OR between an N-of-N and whatever more complex contract you might want to represent." I have a theory that there's only 1 exception: coinjoin; and i think it's exactly because coinjoin does *not* involve transferring ownership of coins.
13:16 < waxwing> whereas notably, coinswap does, and taproot would help with coinswap in that way just like other contracts.
13:17 < belcher> it does in joinmarket
13:17 < belcher> the coinjoin fees get transferred from taker to makers
13:17 < waxwing> right i was *just* about to add that :)
13:17 < waxwing> but the interesting thing is, could you make a funky joinmarket that somehow offset the fees to different txs with some atomic secret passing.
13:18 < belcher> yes, and that would require multisig i think
13:18 < waxwing> for a minute, was thinking that might be cool, but removing fees from tx doesn't help, eh, because change is always trivially deducible.
13:18 < belcher> anything off-chain requires multisig
13:19 < belcher> coinswap/LN/atomic swaps are all off-chain you could say
13:19 < waxwing> yeah when i was doing that 'on-chain contracting' thing, i described it as "shared control point".
13:19 < belcher> well i suppose it depends how you define off-chain, because you could say coinjoin is off-chain because the information of which input corresponds to which output is off-chain
13:19 < waxwing> you always start from a shared control point, then you can "contract" from there.
13:22 < waxwing> with schnorr you could probably pass adaptor signatures as part of the coinjoin setup phase, in order to claim fees when the transaction is broadcast, but that doesn't sound like a practically useful setup compared to the other ways of using that technique
13:23 < waxwing> still it's interesting as a kind of "design space" i guess
13:24 < belcher> the reason for a shared control point is to stop doublespending
13:24 < waxwing> still, you never know. we could leverage simply the fact that joinmarket exists as a set of people doing txs together, to bootstrap people actually doing weird stuff like that.
13:24 < belcher> so for example in coinswap:
13:25 < belcher> you could imagine a protocol where Alice creates an signed tx to Carol and its left unbroadcast, while Carol does a Carol->Alice, but this scheme is trivially broken because Alice could doublespend her UTXO to invalidate the unbroadcast tx
13:25 < belcher> the 2of2 multisig shared control point stops that because the blockchain prevents doublespending like that
13:35 < arubi> one day we'll get confidential amounts in transactions, then our lives would be much simpler :)
13:36 -!- asymptotically [~asymptoti@gateway/tor-sasl/asymptotically] has quit [Quit: Leaving]     
15:40 -!- dx25 [~dx25@97.119.180.105] has quit [Read error: Connection reset by peer]     
15:41 -!- dx25 [~dx25@67-3-153-10.omah.qwest.net] has joined #joinmarket     
15:51 -!- gmaxwell [gmaxwell@wikimedia/KatWalsh/x-0001] has quit [Ping timeout: 240 seconds]     
15:52 -!- gmaxwell [gmaxwell@mf4-xiph.osuosl.org] has joined #joinmarket     
15:52 -!- gmaxwell is now known as Guest44828     
16:09 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Read error: Connection reset by peer]     
16:42 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
16:47 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Excess Flood]     
16:48 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket     
17:03 -!- Giszmo [~leo@pc-37-38-86-200.cm.vtr.net] has quit [Ping timeout: 256 seconds]     
17:10 -!- Guest44828 [gmaxwell@mf4-xiph.osuosl.org] has quit [Changing host]     
17:10 -!- Guest44828 [gmaxwell@wikimedia/KatWalsh/x-0001] has joined #joinmarket     
17:10 -!- Guest44828 is now known as gmaxwell     
17:16 -!- Giszmo [~leo@pc-37-38-86-200.cm.vtr.net] has joined #joinmarket     
17:26 -!- Giszmo [~leo@pc-37-38-86-200.cm.vtr.net] has quit [Ping timeout: 256 seconds]     
17:45 -!- belcher [~belcher@unaffiliated/belcher] has quit [Quit: Leaving]     
17:48 -!- Giszmo [~leo@45.232.92.190] has joined #joinmarket     
17:55 -!- Giszmo [~leo@45.232.92.190] has quit [Read error: Connection reset by peer]     
18:19 -!- Giszmo [~leo@pc-37-38-86-200.cm.vtr.net] has joined #joinmarket     
20:52 -!- AgoraRelay [~jmrelayfn@p5DE4A7D7.dip0.t-ipconnect.de] has quit [Ping timeout: 255 seconds]     
21:04 -!- AgoraRelay [~jmrelayfn@p5DE4A55C.dip0.t-ipconnect.de] has joined #joinmarket     
22:52 -!- quitobro [~quitobro@pool-108-41-0-186.nycmny.fios.verizon.net] has quit [Quit: quitobro]