--- Day changed Fri Nov 25 2016
03:35 -!- DeathShadow--666 [~IDSE@S0106a84e3f595813.vc.shawcable.net] has joined #joinmarket     
04:22 -!- coins123_ [~coins123@ip-244-225.sn1.clouditalia.com] has joined #joinmarket     
04:22 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Read error: Connection reset by peer]     
05:25 -!- kakobrekla [~kako@unaffiliated/kakobrekla] has quit [Ping timeout: 256 seconds]     
05:26 -!- kakobrekla [~kako@unaffiliated/kakobrekla] has joined #joinmarket     
05:43 -!- coins123_ [~coins123@ip-244-225.sn1.clouditalia.com] has quit []     
05:43 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket     
06:31 -!- Empty2k12_ is now known as Empty2k12     
06:34 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Ping timeout: 260 seconds]     
06:35 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket     
07:03 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 256 seconds]     
07:03 -!- Giszmo [~leo@pc-40-227-45-190.cm.vtr.net] has joined #joinmarket     
07:18 -!- belcher [~belcher@unaffiliated/belcher] has joined #joinmarket     
07:19 <@waxwing> https://github.com/JoinMarket-Org/joinmarket/issues/669 <- hopefully saved this guy from paying a huge fee.
07:33 < tergi> hah wow
07:33 < tergi> good to have those kinda checks
07:34 < tergi> waxwing: got to the bottom of that password issue. not sure what can be done about it because im not much good in python really. but looks like a windows induced issue. It makes me concerned because you cannot paste a big password in from a password manager
07:34 < tergi> ie keepass or lastpass
08:03 <@waxwing> tergi: yeah i saw you mention it the other day. it's a good point, if someone could fix it that'd be great. i suppose it hasn't cropped up before because it's only windows. maybe.
08:07 <@waxwing> tergi: oh i see you added test results to the issue, appreciate the work, thanks
08:07 <@waxwing> i remember the code doesn't accept a null byte but does accept a null string for the password. that itself should have been fixed before.
08:07 <@waxwing> although all password issues are of course greatly ameliorated by recovery
08:07 <@waxwing> but still
08:29 < tergi> well the issue is security. if you are putting in a 100 character password on windows by pasting and it is getting reduced to null string... false security. 
08:30 < tergi> the getpass function in python is the issue it seems. and as far as i can tell the python people just push it into a windows bug
08:30 < tergi> Perhaps some basic password strength testing at least could be added to ensure non trivial passwords. 
08:31 < tergi> or perhaps a replacement for getpass
08:45 <@waxwing> i agree it should be fixed
09:16 -!- coins123 [~coins123@unaffiliated/coins123] has quit []     
09:20 < tergi> Might also be worth while to put a small note on the windows install guide to let ppl know pasting the password is not working in the current version
09:24 <@waxwing> tergi: yes, good thought. you can do that if  you like, the wiki is editable.
10:50 < Anduck> looks like lots of commit problems are around.. maybe due to increased fees
10:54 <@waxwing> Anduck: do you say that because you're seeing lots of "Authorisation failed"?
10:54 < Anduck> waxwing: yes
10:55 <@waxwing> ok, well, could be, very hard to say
10:55 <@waxwing> i guess you saw the github issue i linked above
10:55 < Anduck> the fee thing?=
10:55 < Anduck> yea     
10:55 < Anduck> its good to have the check in the joinmarket taker client :)
10:56 <@waxwing> indeed if people have tx_fees < 3 i would imagine it could partially cause that. or, could be other stuff. my maker has gone quiet last couple of days.
10:56 <@waxwing> !hp2s are pretty slow so i expect globally it's fairly quiet
10:56 < tergi> I have had no joins today so far
10:56 < tergi> only 5 yesterday
10:57 <@waxwing> yeah i think me neither. fee backlog *may* be involved, hard to say.
10:57 < Anduck> quietness started when fees went higher
10:57 < tergi> Yea     
10:57 <@waxwing> tergi: in the early days we would have been excited by that :)
10:57 < tergi> excited for 5? :)
10:57 <@waxwing> what's people's record per day i wonder? i think i had more than 50 only a couple of times.
10:57 <@waxwing> tergi: sure.
10:58 < nkuttler> 50 tx?
10:58 <@waxwing> yeah was wondering what people's peak throughput has been, per day.
10:58 < tergi> 44 on the 21st
10:59 <@waxwing> end of week is also slow usually, people have better things to do :)
10:59 < tergi> I was thinking it could be neat to have some type of notifcations when transactions happen.
11:00 < tergi> US holiday yesterday also
11:00 < Anduck> i had that
11:00 < nkuttler> 47 for me
11:00 <@waxwing> ah right, holiday.
11:00 < Anduck> when i did some tests months and months ago
11:00 <@waxwing> but .. bitcoin black friday!?!
11:00 <@waxwing> :)     
11:00  * nkuttler is actually surprised
11:00 < tergi> Yea. i expected a pretty busy day today with that but i dunno
11:01 <@waxwing> people buying consumer goods with bitcoin .. with coinjoin. huge market :)
11:01 < Anduck> tergi: i had emailer code attached
11:01 < Anduck> it's quite easy to add.
11:01 < tergi> yea     
11:01 < nkuttler> haven't had any tx today though
11:01 < tergi> i have been thinking about what to try and do
11:02 < Anduck> actually, my setup is currently quite kick-ass
11:02 <@waxwing> here's another curious thing: 5 days ago there were > 280 in the pit, now about 140
11:02 <@waxwing> cgan that is
11:02 < tergi> THinking about something to scrape the logs maybe and calculate some stuff on how much you have earned in fees each day, week, month etc
11:02 <@waxwing> maybe everyone's calculating their coinbase tax returns? :)
11:02 < Anduck> tergi: if you do that, please share in github
11:02 < Anduck> waxwing: ofc
11:03 < nkuttler> i have an ipython notebook with numpy and stuffs
11:03 < nkuttler> (probably a few bugs here and there)
11:03 < tergi> According to the coinbase tax filing report... i am in the negative for the last few years
11:03 < Anduck> did all coinbase customers get it
11:03 < tergi> So i shoudl be able to go back and claim those on my taxes hahga
11:03 < tergi> You have to request it
11:03 < tergi> its a beta report they are developing
11:04 < tergi> I dont know if they are really going to put up much of a fight against the IRS
11:04 < Anduck> no they wont
11:04 < Anduck> lol     
11:04 < tergi> its like and ant battling the bottom of your shoe
11:04 < Anduck> they will happily share EVERYTHING
11:04 < Anduck> coinbase is still very scared 
11:04 < tergi> I was under the impression they did already anyway
11:04 < Anduck> maybe even under control of gov
11:05 < tergi> well they should be scared... i mean they have that post they are getting the IRS on them... no ones going to be using them anymore that cares about their privacy
12:57 -!- lton [253074b8@gateway/web/freenode/ip.37.48.116.184] has joined #joinmarket     
12:58 < lton> Does anyone know if there are any wallet softwares or ways to create a watch only wallet with the wallet-tool pubkeys?
12:59 <@waxwing> lton: yes, joinmarket :) it creates a watch-only wallet in Bitcoin Core if you use it with that.
13:01 <@waxwing> another idea, you could use wallet-tool to display the addresses and then copy those addresses into other wallet software which has a watching-only feature
13:02 <@waxwing> not particularly elegant, at least not in general. not the same as using the master pubkey.
13:38 < lton> thank you, is there a way to list more of the addresses as well as the used addresses with wallet-tool.  btw, thank you so much for all your hard work.
13:39 <@waxwing> you're welcome
13:39 <@waxwing> there's two things: one, displayall shows all the addresses used thus far. two, the full solution is to import the master pubkey in a script and get the full list of addresses from that.
13:41 <@waxwing> that kind of script would be very simple, but it'd just be a way to get any and all addresses that joinmarket would use, you'd still need to figure out how to import it into whatever other software you'd be using to view the watch-only.
13:43 < lton> do you think a script like that will be included in future versions?  I would be willing to contribute a little btc.  ;)
13:44 <@waxwing> if you just want a script that prints out all addresses in joinmarket mixdepths up to N, i can knock that out in an hour (and so could other people). but i'm not sure if that's really what you're looking for. where do you want to import them? into Core? other wallet?
13:45 < lton> with all the keys I could import them into electrum for a watch only
13:46 < lton> I tried -displayall and got back no such option -d
13:46 <@waxwing> electrum, OK. well, it's late here, i guess i could do it tomorrow. probably like, 1 argument for how many mixdepths, 2nd argument how many addresses per mixdepth. then it just spits them out in a list.
13:46 <@waxwing> displayall is a method
13:46 <@waxwing> like python wallet-tool.py wallet.json displayall
13:47 < lton> aaaahhhhhh, that would be amazing!
13:47 <@waxwing> hmm on second thoughts probably better to add it as a method to wallet-tool. it could be a non-sync method, but it would be easier because the wallet decryption code is already there. i.e. easier than adding a new script.
13:48 <@waxwing> to be clear, i don't actually know that it'd work for electrum, since i never used it. if it imports *keys* (pubkeys) rather than addresses, might need to investigate that.
13:49 <@waxwing> i guess it just imports addresses though? that would make more sense
13:50 < lton> it can import address for a watch only and an xpub, but the xpubs from joinmarket aren't compatible.
13:50 <@waxwing> so either?
13:51 < lton> yeah, but it's a different type of xpub.  
13:51 <@waxwing> i'm not sure about that, but the wallet structure is different. the xpub format might be the same, i'd have to check, but either way it's academic, it won't work.
13:59 < adlai> can't you do this by increasing the gap limit?
13:59 -!- Empty2k12 [~Empty2k12@p579078EE.dip0.t-ipconnect.de] has quit [Remote host closed the connection]     
13:59 < adlai> although it's not a friendly interface
13:59 <@waxwing> yes, but i was thinking about a no-sync version
14:00 <@waxwing> hmm, would have to look at the code, but it's a good thought. and yeah it's bit hacky too :)
14:02 <@waxwing> yeah gap limit would involve sync of course :) prob just another wallet-tool method that's no-sync would be very easy and just dump to file.
14:03 <@waxwing> but still not as good as xpub ofc
14:04  * adlai found an obscure bug
14:08 <@waxwing> pls go on
14:22 <@waxwing> this: https://github.com/AdamISZ/joinmarket-clientserver now does sendpayment (including additional feature of scheduled sequences of transactions)  and electrum plugin is working although it'll need more refinements. i'm going to look into building binaries shortly. (the daemon and also the electrum plugin, after).
14:22 <@waxwing> i know people don't have much bandwidth for such things, but i'd appreciate any feedback :)
14:31 -!- mrkent [~textual@unaffiliated/mrkent] has joined #joinmarket     
14:33 -!- mrkent [~textual@unaffiliated/mrkent] has quit [Client Quit]     
14:57 < Anduck> why is joinmarket wallet not using bip44, btw?
14:57 < Anduck> bip44 would be good, so it could be used with electrum
14:57 < Anduck> to watch what's going on
14:57 < Anduck> (easier way to watch-9
14:58 < Anduck> what path does jm use?
14:59 <@waxwing> Anduck: the main issue is multiple mixdepths
14:59 < Anduck> ahhh, of course
15:00 <@waxwing> i don't think we could be compatible with electrum, i noticed in 2.7 they removed the accounts feature from the standard wallet unfortunately
15:00 < Anduck> yes, that's a tough one...
15:00 < Anduck> yeah
15:00 <@waxwing> there might be a way of making it closer to something that *might* be compatible, if only partially, but there'd be several tricky things to think about
15:00 <@waxwing> plus it wouldn't be backwards-compatible of course, not that that's necessarily a deal breaker
15:01 < Anduck> maybe use 1 wallet with tree-structure instead of different paths?
15:01 < Anduck> eg mod 20 or something
15:01 < Anduck> would give up to 20 depths
15:01 < Anduck> well, something wiser but that's the general idea
15:02 <@waxwing> you mean mod 20 on the index, to artificially split?
15:02 < Anduck> yes     
15:02 <@waxwing> (address index in a branch)
15:02 <@waxwing> funny, i was vaguely musing about some trick like that to do in the electrum plugin
15:02 <@waxwing> it's kind of a horrible hack to even think about though :)
15:02 < Anduck> well, yeah
15:03 < Anduck> maybe electrum could import addresses from paths?
15:03 < Anduck> jm wallet integrator plugin for electrum, which monitors the jm wallet (as it is today) and imports addresses as they're used
15:03 < Anduck> with big warning texts to only use it to watch
15:04 < Anduck> (or be *very* careful)
15:04 <@waxwing> for the plugin, one big motivation for me was to defer coin management to electrum itself.
15:04 <@waxwing> unfortunately i only have sending a single join working, which isn't exactly tremendously useful.
15:04 <@waxwing> i'd originally planned to try to use the accounts within the wallet, but that was removed (i didn't even get round to looking at it tho')
15:05 <@waxwing> you could do a poor-man's tumble through multiple wallets.
15:05 < Anduck> right
15:05 <@waxwing> which although is far from ease-of-use, at least on the other hand multiple wallets are ultra-easy to handle with electrum
15:06 <@waxwing> but so far, i don't have a particularly smart idea. the modulo thing crossed my mind, you could schedule multiple transactions like that. but means burrowing too much into coin selection. don't want to do that.
15:07 < Anduck> maybe it would best to be done with electrum accounts
15:08 <@waxwing> yeah, that fits the original bip32 design.
15:09 <@waxwing> that's what i wrote here: https://github.com/AdamISZ/electrum-joinmarket-plugin#future-developments - before it was removed :(
15:20 < Anduck> joinmarket/wallet.py", line 94, in select_utxos utxo_list = self.get_utxos_by_mixdepth()[mixdepth]
15:20 < Anduck> KeyError
15:21 < Anduck> any idea why this happens?
15:21 < Anduck> nvm, solved!
15:21 <@waxwing> there was a recent bugfix related to that, let me see .. but context? script?
15:21 <@waxwing> oh ok :)
15:21 < Anduck> btw -m should also do -M
15:21 <@waxwing> oh :)
15:21 < Anduck> wait, i mean -a
15:22 <@waxwing> yes. that was the bugfix
15:22 < Anduck> heh. alright, great!
15:22 <@waxwing> https://github.com/JoinMarket-Org/joinmarket/commit/c9aa523d2c4ee6ebdc0cc357a5addb57078de68f
15:25 < Anduck> right
15:25 < Anduck> btw, just thought of an attack vector...
15:25 < Anduck> what if someone starts proxying offers?
15:25 < belcher> <tergi> THinking about something to scrape the logs maybe and calculate some stuff on how much you have earned in fees each day, week, month etc <---- this would be best done by adding to the wallet-tool.py history tool IMO
15:25 < Anduck> e.g. say he proxyes each offer with 10 different connections
15:26 < belcher> i think with electrum you can give it a comma-separated list of addresses and it creates a watch only wallet from them
15:28 <@waxwing> Anduck: are you talking about someone repeating their own offers, or relaying others'? the latter (MITM) we have a defence against, the former is a problem we don't have a good solution for yet.
15:29 < Anduck> i meant the latter
15:29 <@waxwing> belcher: yeah sounds right, it'd be easy for us to add a nonsync method, but xpub (best way) i guess we can't do, although maybe it's possible.
15:30 <@waxwing> Anduck: ideas welcome. we've been discussing it for a while.
15:30 <@waxwing> it screws up privacy even when it works, could lead to bugs, but fundamentally it's an economic attack which we need a defence against.
15:31 <@waxwing> when i thought about it i hit a roadblock just because, how can someone privately prove that their offered utxos do not intersect with others'?
15:31 <@waxwing> adlai had some ideas, but schemes that were discussed just ended up getting a bit crazy as i recall.
15:32 < Anduck> ok     
15:34 < GithubBot5678> [joinmarket] adlai pushed 1 new commit to develop: https://git.io/v1T4p
15:34 < GithubBot5678> joinmarket/develop 70e8e78 Adlai Chandrasekhar: Fix array indexing underflow in coin selection
15:34 < tergi> isnt electrum a slightly sub optimal wallet to use for watching or anything related to JM? All your addresses go through a server that you dont control and as far as i know it always uses the same server till next start up
15:36 < belcher> yep     
15:36 < adlai> the [lack of] privacy is like that of using JM with blockr instead of your own node
15:36 <@waxwing> i wouldn't say it's the same as blockr, but it's nearer to that than a node
15:47 < lton> you could run your own private electrum server if you are already running a core node
15:50 < belcher> an electrum server is much much harder to ran than full node from what iv seen
15:51 < belcher> its incompatible with pruning, it uses much more disk space and takes a long time to sync since it must build a database of addresses -> tx history
16:14 < tergi> I used armory for a while when i started btc... 2x the block chain in disk usage
16:15 < tergi> but if you are running core anyway to get away from spys then you dont really need electrum anyway.
16:16 < tergi> perhaps wallet-tool.py will some day have a gui wrapper.
16:18 < lton> I'm looking for a way to monitor each mix depth in real time rather than the entire wallet.
16:19 < tergi> any particular reason? just curious whats going on?
16:20 < tergi> I think the mempool growth is due to bitcoin black friday... i think someones probably spamming the network to make it hard to buy anything
16:20 -!- windsok [~windsok@45.63.59.8] has quit [Ping timeout: 248 seconds]     
16:21 < lton> manages several yield generators, some for other people.  Being able some of the internal wallet activity would help me.  
16:22 < lton> I meant being able to monitor the internal wallet activity would help me
16:27 < belcher> if they all run off different bitcoin core instances could you look at the core wallets?
16:28 < lton> same core instance.  Either way, does core indicate which mix depth address is linked to?
16:31 < belcher> it doesnt, so that wont help then
16:31 < belcher> why do you need to know the mixdepth?
16:37 < lton> to have an idea of what is in each mixing depth in real time in case I want to sweep/withdraw/sendpayment without needing combine mixing depths.  Really a convenience thing.
16:38 < adlai> i'm not sure what to make of this travis-ci failure triggered by my commit: https://travis-ci.org/JoinMarket-Org/joinmarket/builds/178964625
16:38 < adlai> the error causing the failure is at this line: https://travis-ci.org/JoinMarket-Org/joinmarket/builds/178964625#L1337
16:39 < adlai> it seems to me like the bug is in the test, not the code
16:41 < lton> btw mr. belcher.  I would just like to say thank you for all your work.  I have been watching this project for some time and am very excited about it's progress.  so thank you.
16:41 < belcher> thank you!
16:43 < GithubBot5678> [joinmarket] adlai pushed 1 new commit to develop: https://git.io/v1T07
16:43 < GithubBot5678> joinmarket/develop 1a8da55 Adlai Chandrasekhar: fix type agreement in coin selection test
16:43 < belcher> yeah adlai it looks like it crashed after that assert failed
16:43  * adlai watches https://travis-ci.org/JoinMarket-Org/joinmarket/builds/178971937
16:43 < belcher> TypeError: cannot concatenate 'str' and 'function' objects <---- but the only string concatenation only happens after the assert results false
16:44 < belcher> probably "Selection algo: " + algo
16:44 < belcher> should be using algostr not algo, i think
16:44 < adlai> that's what i just changed :)
16:45 < adlai> it should probably report the specific amounts that failed
17:02 < GithubBot5678> [joinmarket] adlai pushed 1 new commit to develop: https://git.io/v1TE0
17:02 < GithubBot5678> joinmarket/develop 72537e2 Adlai Chandrasekhar: Proper fix to the fix
17:04 -!- windsok [~windsok@45.63.59.8] has joined #joinmarket     
17:30 -!- tergi [4a4aed8d@gateway/web/freenode/ip.74.74.237.141] has quit [Ping timeout: 260 seconds]     
18:28 -!- tergi [4a4aed8d@gateway/web/freenode/ip.74.74.237.141] has joined #joinmarket     
19:22 -!- megaddin [aladdin@gateway/shell/fnordserver.eu/x-reveowgibamtgisn] has joined #joinmarket     
20:10 -!- lton [253074b8@gateway/web/freenode/ip.37.48.116.184] has quit [Ping timeout: 260 seconds]     
21:38 -!- Giszmo [~leo@pc-40-227-45-190.cm.vtr.net] has quit [Quit: Leaving.]