--- Day changed Sat Feb 25 2017 00:17 -!- cluelessperson [~cluelessp@unaffiliated/cluelessperson] has joined #joinmarket 00:25 -!- coins123 [~coins123@93.68.153.203] has joined #joinmarket 00:25 -!- coins123 [~coins123@93.68.153.203] has quit [Changing host] 00:25 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket 01:49 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Remote host closed the connection] 02:00 -!- coins123 [~coins123@93.68.153.203] has joined #joinmarket 02:00 -!- coins123 [~coins123@93.68.153.203] has quit [Changing host] 02:00 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket 03:08 -!- stachrom [d4338865@gateway/web/freenode/ip.212.51.136.101] has joined #joinmarket 05:11 < belcher> i thought of a way to deal with the "bad maker sends a bloom filter with all 1s", that is to have some algorithm that rejects the bloom filter that matches the most others 05:12 < belcher> i.e. if theres filters A, B, C, D and E and E matches A and B but A,B,C,D dont match each other, then reject E, not A and B 05:12 < belcher> not sure if such an algorithm will be scalable, hope its not O(2^N) or something 05:13 < belcher> but you might be able to do a binary search thing, note that you can batch verify many bloom filters by ANDing them all together, then if theres a match split them into two groups and AND all those, etc 05:14 < waxwing> belcher: back now, i need to read up a bit, it may take me a while :) i only briefly read about bloom filters ages ago, so i'll have to read again. 05:14 < belcher> ok 05:14 < waxwing> what's the best place to get a good summary specifically as it relates to bitcoin proposals? 05:16 < belcher> the bloom filter wikipedia page and the bitcoin.org developer guide (searching for the one link im thinking of) 05:16 < waxwing> thx 05:16 < belcher> a bloom filter is a bit array, essentially they work by hashing your input and using that to set certain bits, then to test if an element is in the bloom filter you hash it and check if all the bits are 1s 05:17 < belcher> if they are all 1s its either in the filter, or the 1s were all set by chance (i.e. its a false positive) 05:17 < belcher> if some of the bits are 0, then that element is for sure not in the bloom filter 05:17 < belcher> and you can tune certain parameters to get the false positive rate you want, the wikipedia page has the maths for that 05:18 < belcher> also you can test if two bloom filters contain elements in common by ANDing them 05:18 < belcher> heres the bitcoin.org page, contains some code which can help with understanding https://bitcoin.org/en/developer-examples#creating-a-bloom-filter 05:26 < waxwing> ok, think i understand the basic idea. so when i add a new item to an already existing filter, i bitwise-OR its bits into it right? (not XOR) - i.e. previous 1s stay 1, previous 0s flip to 1. 05:29 < belcher> yep 05:29 < belcher> so the fp rate is also determined by how many elements you will add 05:31 < waxwing> yeah. i think at first glance, losing the property of 'you can't predict others' commitents' is very hard to overcome. i get your 'too many matches' idea but that seems very chewy to analyze. 05:31 -!- HostFat_ [~HostFat@host25-242-dynamic.32-79-r.retail.telecomitalia.it] has joined #joinmarket 05:31 < waxwing> just a first impression though 05:33 -!- HostFat [~HostFat@host141-180-dynamic.249-95-r.retail.telecomitalia.it] has quit [Ping timeout: 240 seconds] 05:39 < belcher> thought of an attack, i think it works for your podle method too 05:41 < belcher> it works by rejecting the second podle hash it sees, the irc server (or any message channel we use) could delay the sending of the genuine podle hash and replay it with its own bot, the taker would see the genuine hash come second and then reject it 05:41 < belcher> its still a help though, it still doesnt allow makers to use their wallets twice 05:41 < waxwing> ah right. because initial phase is not e2e. 05:42 < waxwing> but the false maker cannot do a valid reveal on that 05:42 < belcher> yep, so it would be a DOS 05:42 < belcher> well actually... the irc server can do a DOS just by switching itself off 05:42 < waxwing> hehe 05:42 < belcher> so yes not worse than the status quo 05:43 < waxwing> well i guess the more to-the-point counter is that a current IRC server can block any makers from participating (selectively) 05:44 < belcher> " i.e. if theres filters A, B, C, D and E and E matches A and B but A,B,C,D dont match each other, then reject E, not A and B" <--- the DOSer could still DOS by connecting with multiple maker bots and each one sends a matching bloom filter, AA matches with A, BB matches with B and so on 05:44 < belcher> then there'd never be a situation of "this bloom filter matches with everyone so we'll reject it" 05:44 < belcher> however... we're thinking of maker's replying to !orderbook by PM, so other makers wont actually see what bloom filter they send 05:45 < waxwing> hmm i forgot the point of multi-MC being DOS resistance to one server operator 05:48 < waxwing> good lord, signal messaging server uses sha1 cert :) 05:48 < waxwing> https://twitter.com/veorq/status/834832713584668672 05:48 < waxwing> but muh moxie? 05:49 < belcher> hey i had a suggestion for reducing the size of podle commitments, what if we used sha1 instead of sha256 because sha1 is smaller 05:49 < waxwing> lulz 05:49 < waxwing> #yolocrypto. ACK. 05:49 < belcher> actually to save more space we could use crc32 :D 05:50 < waxwing> even with length extension trickery, sha1 probably wouldn't be unsafe. i think :) 05:50 < waxwing> but i did envisage just using 16 bytes, "ought to be enough for anyone" 05:55 < belcher> 16 bytes instead of 32 means its 2^16 easier to brute force right? 05:55 < waxwing> no 128 bits 05:55 < belcher> and you need to bruteforce it in the time between sending the commitment and the taker doing a coinjoin 05:55 < waxwing> i always used to make that mistake :) 05:56 < belcher> please explain? 05:56 < waxwing> 16 bytes is 16*8 bits 05:58 < belcher> ok 05:59 < belcher> ah ok so 128 times easier 05:59 < belcher> well even if this bloom filter thing doesnt work it probably wont be too bad to just use the podle hash thingy 06:00 < waxwing> yes i could imagine it working, realistically, for fending off the economic "attack". don't see it stopping determined DOS of various flavours. 06:01 < waxwing> as for the data transfer issue, i don't know. 06:04 < belcher> we cant get much worse than bitcoin itself, requiring as it does the blockchain download :p 06:04 < belcher> although, only at the start, after that it uses very little resources from what i see 06:04 < waxwing> well yeah. but i remember in the early days, you were quite concerned about how slow a transaction is, it'd be nice if it got faster or at least not a lot slower. 06:05 -!- stachrom [d4338865@gateway/web/freenode/ip.212.51.136.101] has quit [Ping timeout: 260 seconds] 06:05 < waxwing> yeah it's pleasantly not resource intensive, but then i guess that's not a surprise. not much it has to do really. 06:05 < belcher> yeah.. 06:05 < belcher> i think the rise of the fee market made me realize fast broadcasting isnt that important 06:06 < belcher> it will take 10 minutes on average at minimum before you can trust the transaction anyway 06:06 < belcher> but it would be annoying if makers were running in ultra slow raspberry pis that took 30 seconds to reply to anything 07:34 -!- Giszmo [~leo@pc-240-13-215-201.cm.vtr.net] has joined #joinmarket 09:10 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Ping timeout: 240 seconds] 09:29 -!- MaxSan [~one@213.152.161.35] has joined #joinmarket 09:31 -!- MaxSan [~one@213.152.161.35] has quit [Client Quit] 09:31 -!- stachrom [d4338865@gateway/web/freenode/ip.212.51.136.101] has joined #joinmarket 09:41 -!- shinobimonkey [~vagabond@162.219.176.19] has quit [Ping timeout: 260 seconds] 09:44 -!- puddinpop [~puddinpop@unaffiliated/puddinpop] has quit [Remote host closed the connection] 10:03 -!- solomun [5999b126@gateway/web/freenode/ip.89.153.177.38] has joined #joinmarket 10:07 -!- himmelhund [c276de76@gateway/web/freenode/ip.194.118.222.118] has joined #joinmarket 10:51 -!- coins123 [~coins123@93.68.178.64] has joined #joinmarket 10:51 -!- coins123 [~coins123@93.68.178.64] has quit [Changing host] 10:51 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket 11:29 -!- zxccxz [d4159dee@gateway/web/freenode/ip.212.21.157.238] has quit [Quit: Page closed] 11:34 -!- zxccxz [d4159dee@gateway/web/freenode/ip.212.21.157.238] has joined #joinmarket 12:29 -!- puddinpop [~puddinpop@unaffiliated/puddinpop] has joined #joinmarket 12:35 -!- stachrom [d4338865@gateway/web/freenode/ip.212.51.136.101] has quit [Ping timeout: 260 seconds] 12:50 -!- megaddin [aladdin@gateway/shell/fnordserver.eu/x-iqcdhyawoskcjsdq] has quit [Quit: https://fnordserver.eu] 13:22 -!- megaddin [aladdin@gateway/shell/fnordserver.eu/x-xdfmuxkkqsdznwgt] has joined #joinmarket 13:44 -!- owowo [~ovovo@unaffiliated/ovovo] has quit [Ping timeout: 268 seconds] 13:54 -!- owowo [~ovovo@unaffiliated/ovovo] has joined #joinmarket 14:09 -!- wumpus [~quassel@pdpc/supporter/professional/wumpus] has quit [Read error: Connection reset by peer] 14:11 -!- wumpus [~quassel@pdpc/supporter/professional/wumpus] has joined #joinmarket 14:19 -!- owowo [~ovovo@unaffiliated/ovovo] has quit [Ping timeout: 268 seconds] 14:38 < himmelhund> hello, i have a problem with the yield-generator and hope that someone can help me. wenn i start the generator i get the following messages: 14:40 < himmelhund> Date Time [Mainthread] [INFO] Downloading Wallet History 14:40 < himmelhund> Date Time [Mainthread] [INFO] Starting Yield Generator 14:40 < himmelhund> Date Time [Mainthread] [INFO] Connecting to Message Channels 14:42 < himmelhund> Date Time [MChread] [INFO] Connecting to Host irc.cybergorilla.org6697 14:43 < himmelhund> Date Time [MChread] [INFO] Connecting to Host agora.anarplex.net14716 14:44 < himmelhund> Date Time [MChread] [INFO] Failed Authentication irc.cybergorilla.org 6697 14:44 < himmelhund> Date Time [MChread] [INFO] Ending irc 14:45 < himmelhund> And this repeats every few minute and nothing else happens 14:45 < himmelhund> ho can help me with what i am doing wrong? 14:56 -!- owowo [~ovovo@31.7.59.226] has joined #joinmarket 14:56 -!- owowo [~ovovo@31.7.59.226] has quit [Changing host] 14:56 -!- owowo [~ovovo@unaffiliated/ovovo] has joined #joinmarket 15:32 -!- himmelhund [c276de76@gateway/web/freenode/ip.194.118.222.118] has quit [Quit: Page closed] 18:07 -!- HostFat__ [~HostFat@host178-172-dynamic.250-95-r.retail.telecomitalia.it] has joined #joinmarket 18:10 -!- HostFat_ [~HostFat@host25-242-dynamic.32-79-r.retail.telecomitalia.it] has quit [Ping timeout: 240 seconds] 18:15 -!- HostFat__ [~HostFat@host178-172-dynamic.250-95-r.retail.telecomitalia.it] has quit [Ping timeout: 260 seconds] 19:02 -!- MaxSan [~one@213.152.162.89] has joined #joinmarket 19:29 -!- Giszmo [~leo@pc-240-13-215-201.cm.vtr.net] has quit [Quit: Leaving.] 19:45 -!- MaxSan [~one@213.152.162.89] has quit [Ping timeout: 260 seconds] 20:04 -!- MaxSan [~one@213.152.162.94] has joined #joinmarket 20:08 -!- MaxSan [~one@213.152.162.94] has quit [Client Quit]