--- Day changed Sun Feb 12 2017
00:15 < waxwing> didn't i put absurd at 150?
00:16 < waxwing> anyway i will try to get to reviewing it soon alexcato thanks
00:57 -!- juscamarena [~justin@47.148.176.74] has quit [Quit: Lost terminal]     
01:04 < JM-IRCRelay> [AlexCato1] yeah, absurd is at 150. Still have an old joinmarket.cfg, local config on my side
03:40 -!- himmelhund [d5f05661@gateway/web/freenode/ip.213.240.86.97] has joined #joinmarket     
04:21 -!- FiveBroDeepBook [~gk.1wm.su@46.148.182.82] has joined #joinmarket     
04:21 -!- FiveBroDeepBook [~gk.1wm.su@46.148.182.82] has left #joinmarket []     
05:19 < MaxSan> can i see the requests that have been made to my client
05:19 < MaxSan> to the bitcoind client
05:20 < waxwing> rpc calls are in the <nick>.log file
05:20 < waxwing> search "rpc:"
05:26 < MaxSan> cheer
05:26 < MaxSan> s     
05:27 < MaxSan> anyone going to d10e?
05:27 < waxwing> no, where is it?
05:30 < MaxSan> bucharest
05:31 < MaxSan> usually in SF but this ones over here
05:31 < MaxSan> "Could not spend from a mixdepth which is not max"
05:31 < MaxSan> what does that mean?
05:31 < waxwing> you can ignore that
05:32 < waxwing> it's yg-pe trying to avoid re-announcing an offer, but it can't
05:32 < waxwing> nothing wrong
05:32 < waxwing> or *randomizer, i forget whether that also uses that algo
05:32 < waxwing> MaxSan: is there a website?
05:33 < waxwing> ok got it d10e.biz i think
05:33 < MaxSan> https://d10e.biz
05:33 < MaxSan> heh sorry bit slow there
05:34 < waxwing> i'd go maybe but i wouldn't pay for it :)
05:34 < MaxSan> where you based?
05:34 < waxwing> latvia
05:34 < MaxSan> im just getting a ticket now
05:34 < MaxSan> oh cool
05:34 < MaxSan> im UK
05:35 < waxwing> keynote mcafee oh dear, heh
05:35 < waxwing> hmm not sure it looks that good really, but sometimes fun to meet people, there's that
05:36 < MaxSan> yeah thats it exactly
05:36 < MaxSan> bucharest is awesome place too
05:39 < waxwing> several of the "thought leaders" there i consider borderline scammers :)
05:44 < MaxSan> many successful peoople are sadly. im just going for a bit of a crack anyway
05:44 < waxwing> sure, i can understand that
05:45 < waxwing> and yes, sadly the first half of your sentence is also true ;)
06:01 < MaxSan> i am very happy
06:01 < MaxSan> the place i stayed at in bucharest last time is right beside the conference
06:01 < MaxSan> so i know my way about already
06:01 < MaxSan> :D     
06:34 < waxwing> alexcato ping, i have a Q, see the PR
06:40 < JM-IRCRelay> [AlexCato1] interesting, havent thought about that. The algo indeed does allow this
06:40 < JM-IRCRelay> [AlexCato1] first thought: an economically motivated maker would probably want to offer the whole range so he's not missing out on some potential joins
06:41 < waxwing> yeah that was my 'are we saying' bit, but then i thought of the counter-example
06:42 < waxwing> tbh i was never as bothered by the overlapping ranges as by the volume issue, but ... it's kind of all connected. if people even think they get advantage like this, even if they're wrong, it leads to this undesirable spamming up the orderbook situation.
06:42 < waxwing> stamping out even the slightest possibility of advantage (and giving a massive negative advantage) should do the trick
06:43 < waxwing> i mean the PR as is, is a guaranteed improvement anyway.
06:43 < waxwing> if it were me, i'd tend to worry less about the scaling and more just killing it, even if it involves a sort step. but, not sure exactly.
06:44 < JM-IRCRelay> [AlexCato1] currently working on the 2nd PR to disregard makers with > X offers (configurable). Then offerbook-spam by individually identifiable makers should be history.
06:45 < waxwing> right, doing it separately makes sense. thanks.
06:45 < JM-IRCRelay> [AlexCato1] but about that overlapping thing: i could of course change the algo to check each range for overlap, but I dont know how to do it without O(n*n)
06:45 < waxwing> forgive me being a computer science idiot, but what's the scaling of sort?
06:46 < JM-IRCRelay> [AlexCato1] i'm sure i've learned that 20 years ago, but cant remember
06:46 < belcher> so is the idea of this filtering to stop the orderbook database getting too big? since right now it is stored in RAM
06:46 < waxwing> O(nlogn) i'm seeing here, that's vaguely what i remembered
06:47 < waxwing> http://bigocheatsheet.com/
06:47 < waxwing> well, there's best, average and worst just to be annoying
06:47 < waxwing> also what i was proposing earlier, i'm pretty sure it's not a complete sort in general
06:48 < JM-IRCRelay> [AlexCato1] @belcher: my main goals are to have orderbook that humans can actually look at and reduce the load on the IRC servers (after all, we're just guests here)
06:49 < waxwing> yeah seconded, that's what i'm after here. but i'm also slightly worried that we may have missed an edge case that actually *does* give an advantage to overlapping orders.
06:49 < belcher> just because takers filter the offers doesnt stop makers send them, but yes i agree with you
06:49 < waxwing> if their only intent is DOS, they could spam up IRC with anything; so we can only be attempting to improve by stopping an either hypothetical, or real, economic advantage
06:50 < JM-IRCRelay> [AlexCato1] well, makers will learn quickly that what they're doing gets them ignored
07:06 < belcher> you'd hope so, a few of them are doing the spam thing today even though it doesnt help them
07:07 < JM-IRCRelay> [AlexCato1] they might not know or realize that. But the filtering will make it dead obvious: no more coinjoins for them, as takes just ignore them
07:07 < JM-IRCRelay> [AlexCato1] currently they still get CJs, but theres no way for them to know if they get more than they would get otherwise
07:09 < waxwing> right, that's the situation, except i repeat that we should not assume that our filtering 100% works to remove the advantage; quite a few times i tried to see if there was some flaw in our reasoning, but i haven't found it. but doesn't mean there isn't one.
07:10 < waxwing> actually it's conceivable to set a test up to try to prove it, but i haven't got round to it. it would involve collecting a ton of data.
07:10 < waxwing> and if the edge case is something network-y (yeah i know it sounds crazy, but who knows), then might not catch it.
07:14 < JM-IRCRelay> [AlexCato1] yes, combined with the 2nd PR which is done soonish, i can change the first one to check for actual overlaps, as the number of offers to cross-check will be small anyways. So nlogn or even n^2 wouldnt matter at that point
07:14 < waxwing> yeah that sounds ideal, or at least, a solid practical approach. just kill the spam then we don't have to wonder about weird edge case stuff.
07:15 < JM-IRCRelay> [AlexCato1] my next thing to do soonish would be to finally start writing tests, at least for the stuff i added. But that might still be a bit in the future, doubtful i'll have the time soon
07:16 < waxwing> yes i think getting it done is a bit of a priority; as long as we review and we have the regtest sanity tests. but since it's db stuff we have to be fairly thorough in checking it. the db stuff is the part of the app i've spent least time thinking about (hardly at all).
07:42 < GithubBot5678> [joinmarket] AlexCato opened pull request #711: Takers: filter out makers with many offers (develop...limit_maker_max_offers) https://git.io/vDrf6
12:20 < MaxSan> can send payment take multiple addresses
12:22 < waxwing> no. tumbler can though.
12:23 < MaxSan> oh ok
12:24 < waxwing> sendpayment does 1 coinjoin (or just an ordinary bitcoin tx with -N 0), we don't currently support a coinjoin with multiple output values.
12:24 < waxwing> the idea is to have 1 equal sized output for each participant, and 1 change output for each participant. it'd be possible to have multiple of each, but we haven't done that.
12:25 -!- MaxSan [~one@91.214.169.69] has quit [Read error: Connection reset by peer]     
12:25 < waxwing> you can also have no change output, for 'sweep'. with the idea of cleaning out that mixdepth
12:28 -!- MaxSan [~one@91.214.169.69] has joined #joinmarket     
12:45 < MaxSan> Traceback (most recent call last):
12:45 < MaxSan>   File "patientsendpayment.py", line 255, in <module>
12:45 < MaxSan>     main()
12:45 < MaxSan>   File "patientsendpayment.py", line 238, in main
12:45 < MaxSan>     mcs = [IRCMessageChannel(c) for c in get_irc_mchannels()]
12:45 < MaxSan> NameError: global name 'get_irc_mchannels' is not defined
12:45 < waxwing> are you running off the develop branch?
12:45 < MaxSan> im not sure good question
12:45 < waxwing> only in develop branch is patientsendpayment fixed up to work
12:45 < MaxSan> ohhh
12:46 < waxwing> it was broken for ages waiting on a protocol update, belcher fixed it up a few weeks back
12:46 < waxwing> that one does have multiple address destinations, i forgot :)
12:46 < waxwing> (because can be multiple coinjoins)
12:47 < belcher> MaxSan theres a short tutorial about it on the wiki on github in case you didnt see
12:47 < MaxSan> yeah ive been reading over the wiki
12:48 < belcher> wikis are better for howtos because they're easier to edit, editing a --help page requires a PR
12:49 < MaxSan> true aye
13:09 < waxwing> i wonder how quickly estimatefee updates; couldn't hurt to tweak/randomize fee rate to avoid watermarking if it's slow.
13:11 < belcher> was thinking of making an issue about this blog post https://medium.com/@bramcohen/how-wallets-can-handle-transaction-fees-ff5d020d14fb#.2icg4zob6
13:11 < belcher> where its recommended there is some randomness partly for the reason you said
13:12 < MaxSan> how many confirmations does it need
13:12 < MaxSan> before signing a new transaction with coins?
13:14 < waxwing> JM only waits for 1 conf before considering coins usable.
13:14 < waxwing> actually that can be edited in cfg but wouldn't do that
13:15 < waxwing> well,not downwards in any case
13:16 < MaxSan> hmmm
13:16 < MaxSan> just seen in the logs it said 5 oO
13:16 < MaxSan> it has 1 confirm
13:16 < waxwing> you need 5 confirmations for a PoDLE commitment, that's separate from the coins you're spending
13:17 < waxwing> it's important to read this: https://github.com/JoinMarket-Org/joinmarket/wiki/Sourcing-commitments-for-joins
13:23 < MaxSan> belcher: you coming to d10e? :D
13:23 < MaxSan> i dont know where your based either actually not even sure if europe lol
13:24 < belcher> i dont know what that is.. googling now
13:24 < belcher> okay a conference on decentralization
13:24 < MaxSan> https://d10e.biz
13:24 < waxwing> thanks for the bramc article, he does some good stuff
13:25 < MaxSan> you in europe?
13:40 < waxwing> heh, fluffypony will approve: https://twitter.com/Ethan_Heilman/status/830834575161044993
13:40 < fluffypony> *clicks*
13:40  * fluffypony giggles
13:41 < waxwing> send 2c to the developers ethan, don't forget :)
13:41 < belcher> https://coinmarketcap.com/currencies/zcash/#charts
13:42 < belcher> remember the "one zcoin worth more than one bitcoin" headlines?
13:42 < waxwing> chart is misleading, first trades were at 3000BTC :)
13:43 < belcher> interesting
13:43 < waxwing> it was actually a beautiful example of "painting the tape"/ "anchoring bias"
13:43 < belcher> yep     
13:43 < waxwing> some smart people (i guess zcash miners) figured out they could post offers for 5000 btc before any coins reached the exchange
13:43 < waxwing> thus giving the impression that that price meant anything at all
13:43 < belcher> so those offers were never actually filled? price was
13:44 < belcher> s/price was//
13:44 < waxwing> well the initial ones for 5000 weren't. i was watching it on poloniex, it was fascinating
13:44 < fluffypony> there were some filled, but small amounts
13:44 < waxwing> they kept coming and going, about an hour or two before the maturity made any available
13:44 < waxwing> the first were ~ 3000, i'm talking about before that
13:44 < belcher> ah     
13:48 < fluffypony> ah yeah
13:48 < fluffypony> also there was some exchange that offered leverage trading before the token existed
13:48 < waxwing> oh right, but iirc they placed a hard upper limit
13:48 < fluffypony> so it was trading based on pure speculation 
13:48 < waxwing> (bitmex)
13:48 < fluffypony> 40x or something
14:17 < Aleph0> is anyone aware of any micropayments in the space?
14:20 < belcher> theres that experimental video streaming website that uses a payment channel
14:20 < Aleph0> xotica lol?
14:20 < belcher> streamium
16:26 -!- himmelhund [d5f05661@gateway/web/freenode/ip.213.240.86.97] has quit [Quit: Page closed]     
16:50 -!- Giszmo [~leo@pc-165-227-45-190.cm.vtr.net] has quit [Quit: Leaving.]     
16:50 -!- Giszmo [~leo@ip-253-233.219.201.nextelmovil.cl] has joined #joinmarket     
17:03 -!- HostFat [~HostFat@host219-95-dynamic.3-87-r.retail.telecomitalia.it] has joined #joinmarket     
17:40 -!- Giszmo [~leo@ip-253-233.219.201.nextelmovil.cl] has quit [Quit: Leaving.]     
17:40 -!- Giszmo [~leo@pc-165-227-45-190.cm.vtr.net] has joined #joinmarket     
17:50 -!- Giszmo [~leo@pc-165-227-45-190.cm.vtr.net] has quit [Remote host closed the connection]     
18:04 < Aleph0> thx     
18:16 -!- HostFat_ [~HostFat@host220-91-dynamic.245-95-r.retail.telecomitalia.it] has joined #joinmarket     
18:19 -!- HostFat [~HostFat@host219-95-dynamic.3-87-r.retail.telecomitalia.it] has quit [Ping timeout: 245 seconds]     
22:45 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Remote host closed the connection]