1 2013-10-12 00:00:07 <skinnkavaj> there is a boss on google
  2 2013-10-12 00:01:31 <skinnkavaj> gmaxwell: do you hear anything from dan kaminsky?
  3 2013-10-12 00:02:02 <skinnkavaj> i have just seen him on a bitcoin panel in a conference @ youtube
  4 2013-10-12 00:02:08 <skinnkavaj> never on forum, irc, maillist
  5 2013-10-12 00:02:10 <skinnkavaj> or anyting
  6 2013-10-12 00:02:21 <gmaxwell> skinnkavaj: I have mail from him from last week. He's posted on the forum before.
  7 2013-10-12 00:02:21 <skinnkavaj> and yet he makes bet with jgarzik on twitter about bitcoin
  8 2013-10-12 00:02:29 <gmaxwell> I thought he declined that bet? :P
  9 2013-10-12 00:02:30 <skinnkavaj> gmaxwell: do you know his forum nick?
 10 2013-10-12 00:02:40 <skinnkavaj> haha, who?
 11 2013-10-12 00:02:46 <skinnkavaj> jgarzik or dan?
 12 2013-10-12 00:02:52 <sipa> dakami, right?
 13 2013-10-12 00:02:57 <gmaxwell> dan declined the bet jgarzik offered.
 14 2013-10-12 00:03:32 <skinnkavaj> i like how dan kamisky talks, he is really good at talking and making everything sound cool
 15 2013-10-12 00:03:55 <skinnkavaj> gmaxwell: haha
 16 2013-10-12 00:03:57 <sipa> he does seem to like exaggerating, though :)
 17 2013-10-12 00:04:01 <skinnkavaj> wasnt it his idea
 18 2013-10-12 00:04:05 <MC1984> he is rather bombastic
 19 2013-10-12 00:04:13 <skinnkavaj> to have the bet
 20 2013-10-12 00:04:52 <gmaxwell> skinnkavaj: he was making some bombastic claims that us techies thought were crazy. I would have rushed to make the same bet with him, if I had a twitter account.
 21 2013-10-12 00:05:20 <skinnkavaj> and jgarzik replied haha
 22 2013-10-12 00:05:24 <gmaxwell> He's a bum for not taking it. :P
 23 2013-10-12 00:05:49 <skinnkavaj> i still like dan
 24 2013-10-12 00:05:52 <skinnkavaj> he can admit he is wrong
 25 2013-10-12 00:06:01 <skinnkavaj> like the famous bashing bitcoin video
 26 2013-10-12 00:06:04 <skinnkavaj> from defqon
 27 2013-10-12 00:06:13 <skinnkavaj> and now he is in bitcoin panels
 28 2013-10-12 00:06:20 <skinnkavaj> takes a lot of balls to admit that
 29 2013-10-12 00:06:48 <sipa> he bashed bitcoin?
 30 2013-10-12 00:06:56 <skinnkavaj> lol
 31 2013-10-12 00:06:58 <skinnkavaj> yeah
 32 2013-10-12 00:06:59 <skinnkavaj> YEAH
 33 2013-10-12 00:07:04 <skinnkavaj> i mean really much
 34 2013-10-12 00:07:06 <skinnkavaj> its like 1 hour
 35 2013-10-12 00:07:09 <skinnkavaj> bashing video
 36 2013-10-12 00:07:11 <skinnkavaj> on youtube
 37 2013-10-12 00:07:14 <skinnkavaj> talking on how bad bitcoin is
 38 2013-10-12 00:07:18 <skinnkavaj> on a defqon conference
 39 2013-10-12 00:07:36 <sipa> k
 40 2013-10-12 00:07:46 <sipa> (defcon)
 41 2013-10-12 00:08:11 <MC1984> i think he tried to blind the wrong audience with science
 42 2013-10-12 00:08:26 <skinnkavaj> haha
 43 2013-10-12 00:08:58 <skinnkavaj> talk about that, there was a guy buying bitcoin from me on localbitcoins
 44 2013-10-12 00:09:05 <skinnkavaj> i met him, he make his first buy
 45 2013-10-12 00:09:11 <skinnkavaj> i ask how did you get into bitcoin?
 46 2013-10-12 00:09:23 <skinnkavaj> he say "my girlfriend is a programmer and she explained it to me"
 47 2013-10-12 00:10:05 <skinnkavaj> wow you have a smart girl i replied, he smiled and lived happy ever after rich on bitcoin
 48 2013-10-12 00:21:51 <skinnkavaj> gmaxwell: care to give a short explanation on coloured coins? i have watched youtube videoes, read about it etc.. i still don't understand, why is it good?
 49 2013-10-12 00:22:07 <skinnkavaj> and also many in bitcoin development seems to be against it, why?
 50 2013-10-12 00:23:57 <MC1984> fucks with fungibility i expect
 51 2013-10-12 00:24:03 <gmaxwell> It's unclear to me that it actually does anything useful. Many of the applications people cite can simply be done without color coins at all. Sometimes colored coins exemplify the same kind of tunnel vision that plagues people with DHTs.  "Bitcoin is the only kind of distributed system I know" or "Bitcoin is the only way I know to transfer ownership"  "thus I should use bitcoin for this applicatoin".
 52 2013-10-12 00:25:28 <gmaxwell> By itself that stuff is harmless if somewhat time wasteful for people who get caught up in it.  ... but colored coin applications can potentially be abusive of the blockchain as a limited scarce resource, resulting in inefficient transaction patterns and bloat which makes bitcoin less decenteralized and more costly to run for everyone.
 53 2013-10-12 00:25:58 <nanotube> anyone care to give a brief summary or link to the failed bet between jgarzik and dankaminsky/
 54 2013-10-12 00:25:59 <gmaxwell> I like neat applications, but if they're ones that can be done without burdening third parties with additional costs, they ought to be done that way.
 55 2013-10-12 00:26:00 <nanotube> ?
 56 2013-10-12 00:26:38 <jgarzik> nanotube, google "jgarzik dakami bet"
 57 2013-10-12 00:26:46 <gmaxwell> nanotube: dan said that within the year (?) mining would need to use another pow than SHA256 because omg asics. Jgarzik tried to get him in to a modest bet on that (10 BTC?) but failed.
 58 2013-10-12 00:26:57 <MC1984> kaminsky said the hash algo would change by this year or something, someone bet him it wouldnt
 59 2013-10-12 00:27:16 <sipa> gmaxwell: not "would need to use", "would use"
 60 2013-10-12 00:28:32 <jgarzik> https://twitter.com/jgarzik/status/336210942717214720
 61 2013-10-12 00:29:27 <skinnkavaj> The Internets
 62 2013-10-12 00:29:27 <skinnkavaj> We can fix it. We have the technology. OK. We need to create the technology. Alright. The policy guys are mucking with the technology. Relax. WE'RE ON IT.
 63 2013-10-12 00:29:32 <skinnkavaj> ^^ Kaminsky
 64 2013-10-12 00:29:34 <skinnkavaj> Legend
 65 2013-10-12 00:30:31 <Luke-Jr> skinnkavaj: the thing with coloured coins is that they get their value outside the blockchain. So long as that outside party exists, the coloured coins are centralised even in the blockchain. centralised systems can work far more efficient and private without a blockchain, so might as well do that instead.
 66 2013-10-12 00:30:56 <gmaxwell> Luke-Jr: at least for 99% of the things people talk about.
 67 2013-10-12 00:30:59 <MC1984> why not merge mine a colourcoin chain
 68 2013-10-12 00:31:08 <MC1984> if people are set on doing it
 69 2013-10-12 00:31:33 <gmaxwell> MC1984: becuase you never needed the colored coins to begin with for those applications, and the merged mined chain runs into the "doing cross chain trades is tricky" problem.
 70 2013-10-12 00:31:43 <Luke-Jr> gmaxwell: is there another use case?
 71 2013-10-12 00:32:00 <MC1984> people want to use a blockchain so shit cant be seized
 72 2013-10-12 00:32:13 <Luke-Jr> gmaxwell: merged mining can be done in a way that makes the slave dependent on the master (like what Mastercoin should be doing)
 73 2013-10-12 00:32:24 <gmaxwell> Luke-Jr: I'm not willing to say there is none. But all the smart property / stock ones don't really need it.
 74 2013-10-12 00:32:31 <Luke-Jr> MC1984: there's no evidence blockchain ownership can't be seized, ESPECIALLY of coloured coins
 75 2013-10-12 00:32:46 <warren> Luke-Jr: you mean a merge coin is possible that can't be merged with <master that isn't Bitcoin>?
 76 2013-10-12 00:32:54 <gmaxwell> Luke-Jr: I know, I've pointed this out to people, I'm just saying that for the 99% of things which there is no need for a blockchain consensus at all, you also don't need a merged mined thing.
 77 2013-10-12 00:32:57 <Luke-Jr> warren: yes
 78 2013-10-12 00:33:00 <warren> Luke-Jr: how?
 79 2013-10-12 00:33:13 <Luke-Jr> warren: the child block is only valid in the context of the master chain
 80 2013-10-12 00:33:13 <nanotube> jgarzik: gmaxwell thanks :)
 81 2013-10-12 00:33:22 <gmaxwell> warren: trivially. The nodes in that other thing are also bitcoin nodes (or at least bitcoin SPV nodes)
 82 2013-10-12 00:33:23 <Luke-Jr> warren: a forked bitcoin master chain would result in a forked child chain
 83 2013-10-12 00:33:29 <skinnkavaj> mastercoin seems to be good but i hate that it was premined, but wait.. satoshi just have 1 million coins for himself xD
 84 2013-10-12 00:33:42 <gmaxwell> skinnkavaj: I don't think there is a single good thing about it.
 85 2013-10-12 00:34:05 <nanotube> skinnkavaj: i'm not spanish, though i speak it thanks to some years of schooling.
 86 2013-10-12 00:34:35 <skinnkavaj> nanotube: cool, perhaps you know if there has been some attempts to start a spanish bitcoin exchange?
 87 2013-10-12 00:34:40 <skinnkavaj> thats what i wanted to ask you.
 88 2013-10-12 00:34:52 <gmaxwell> MC1984: in any case, what asset are you talking about?  E.g. if its ownership in some company... if $foo can seize the company then some chain records of the stock are worthless.
 89 2013-10-12 00:35:08 <warren> Luke-Jr: is that criticism of colored coins written down anywhere?
 90 2013-10-12 00:35:31 <Luke-Jr> warren: dunno, someone should write up about it on the wiki..
 91 2013-10-12 00:35:42 <gmaxwell> Ripper would just delete it. :P
 92 2013-10-12 00:35:57 <Luke-Jr> gmaxwell: lock the page and/or ban him?
 93 2013-10-12 00:36:00 <nanotube> skinnkavaj: not aware of anything like that.
 94 2013-10-12 00:36:21 <Luke-Jr> gmaxwell: though he hasn't been too much trouble lately
 95 2013-10-12 00:36:29 <gmaxwell> warren: it's a really common one, I've seen lots of people independantly produce it.
 96 2013-10-12 00:36:39 <warren> that's pretty much the same problem with Ripple
 97 2013-10-12 00:36:49 <skinnkavaj> while talking about exchange, does anyone know any good open source exchange software, or can i buy it from someone?
 98 2013-10-12 00:36:50 <MC1984> ok i will stop attempting to rationalise coloured coins
 99 2013-10-12 00:37:22 <warren> skinnkavaj: exchanges are things that really must be understood by the people running it...
100 2013-10-12 00:37:37 <sipa> and the hardest part of an exchange is certainly not the software
101 2013-10-12 00:37:37 <skinnkavaj> warren: yeah but i could hire a guy.
102 2013-10-12 00:37:39 <warren> skinnkavaj: far simpler things (like pools) aren't understood by the people running it and hilarity ensues.
103 2013-10-12 00:37:58 <skinnkavaj> sipa: i know. coding is not my thing either.
104 2013-10-12 00:38:03 <warren> skinnkavaj: you have source code and legal code to deal with, so you need to hire more than one guy...
105 2013-10-12 00:38:08 <gmaxwell> MC1984: in any case, I'm not saying there aren't any uses at all, just that virtually everything people can suggest can be done basically as well without sticking the limited scalablity blockchain in the middle of it.  Plus you get the fun extra concerns like we have no utxo expiration, and all your colored coins just become worthless min-value outputs if the company in question dies.
106 2013-10-12 00:38:15 <skinnkavaj> for me the most fun is to debate bitcoin and all the laws applied to it.
107 2013-10-12 00:38:35 <gmaxwell> (and crap about colored coins caused people to try to oppose really important anti-dos measures. :(  ... mixing applications really stinks)
108 2013-10-12 00:38:57 <warren> gmaxwell: what anti-dos measures in particular?  I missed this.
109 2013-10-12 00:39:15 <sipa> the dust output
110 2013-10-12 00:39:45 <gmaxwell> warren: people were opposing having the dust output threshold because they were thinking that being able to make 1 satoshi outputs was essential to their colored coin plans.
111 2013-10-12 00:40:10 <MC1984> MUST USE SMALLEST UNIT
112 2013-10-12 00:40:23 <skinnkavaj> anyone in this channel must have an exchange software they can run if i pay them? i really want to hire.
113 2013-10-12 00:40:43 <MC1984> fuck colour coins simply due to fungibility imo. dont need any more abstract justification than that
114 2013-10-12 00:41:06 <MC1984> skinnkavaj you will enter a universe of pain
115 2013-10-12 00:41:14 <skinnkavaj> sipa: whats hard about running an exchange? isn't it enough to keep bitcoin software up to date and it runs by itself?
116 2013-10-12 00:41:23 <sipa> skinnkavaj: legal threats
117 2013-10-12 00:41:24 <warren> 
ACTION facepalm

118 2013-10-12 00:41:33 <gmaxwell> interfacing with things that aren't bitcoin.
119 2013-10-12 00:41:36 <sipa> interacting with other businesses
120 2013-10-12 00:41:39 <sipa> banks
121 2013-10-12 00:41:39 <skinnkavaj> sipa: the legal is what i want to handle.
122 2013-10-12 00:41:49 <skinnkavaj> i dont want anything to do with exchange software
123 2013-10-12 00:41:52 <skinnkavaj> its not my thing.
124 2013-10-12 00:42:00 <warren> yet you want to run an exchange
125 2013-10-12 00:42:31 <skinnkavaj> warren: yes
126 2013-10-12 00:43:24 <sipa> if you're able to run an exchange business, i doubt you'd have much trouble hiring someone to do the software for you
127 2013-10-12 00:43:46 <skinnkavaj> sipa: no probably, but always when i hire a coder they are like. yes this will be done in 2 weeks, and 6 months later
128 2013-10-12 00:43:49 <skinnkavaj> its not finished
129 2013-10-12 00:43:53 <MC1984> is it ever very profitable
130 2013-10-12 00:43:55 <skinnkavaj> so hard to find anyone i know can do a good job.
131 2013-10-12 00:44:27 <MC1984> stop hiring programmers off craigslist
132 2013-10-12 00:44:36 <skinnkavaj> MC1984: i have hired people from this chat room.
133 2013-10-12 00:44:50 <skinnkavaj> it has failed.
134 2013-10-12 00:45:03 <warren> did they ask to be paid up front?
135 2013-10-12 00:45:07 <skinnkavaj> no
136 2013-10-12 00:45:14 <skinnkavaj> i did not apy them upfront either
137 2013-10-12 00:45:16 <skinnkavaj> that would be stupid
138 2013-10-12 00:45:17 <gmaxwell> Time estimation in serious development is very difficult.
139 2013-10-12 00:45:54 <skinnkavaj> i think i need to find someone that is already running an exchange
140 2013-10-12 00:46:08 <skinnkavaj> dont tell me to go to mtgox
141 2013-10-12 00:46:12 <skinnkavaj> i will kill you.
142 2013-10-12 00:47:03 <sipa> you could try hiring MagicalTux
143 2013-10-12 00:47:09 <gmaxwell> haha
144 2013-10-12 00:47:10 <skinnkavaj> hahaha
145 2013-10-12 00:47:43 <skinnkavaj> i still think MagicalTux will be back bigtime and be like "hey im the king again"
146 2013-10-12 00:47:54 <sipa> k
147 2013-10-12 00:47:56 <gmaxwell> Considering the administrative challenges mtgox has had, selling regional exchanges to other people who want to try their hand at the logistics locally might be a viable business.
148 2013-10-12 00:47:57 <skinnkavaj> so hard to lose status
149 2013-10-12 00:47:59 <skinnkavaj> it takes a lot of time
150 2013-10-12 00:48:19 <warren> gmaxwell: that approach seems to have not gone well ...
151 2013-10-12 00:48:35 <gmaxwell> well thats not what that was.
152 2013-10-12 00:48:48 <skinnkavaj> gmaxwell: have they tried doing that?
153 2013-10-12 00:48:50 <skinnkavaj> oh
154 2013-10-12 00:48:53 <skinnkavaj> yeah u mean coinlab
155 2013-10-12 00:48:54 <skinnkavaj> lol
156 2013-10-12 00:49:01 <MC1984> gox sold?
157 2013-10-12 00:49:27 <gmaxwell> skinnkavaj: Are you on a phone?
158 2013-10-12 00:49:31 <skinnkavaj> kraken.com i love that site!
159 2013-10-12 00:49:43 <skinnkavaj> but too technical for the average joe.
160 2013-10-12 00:49:54 <skinnkavaj> gmaxwell: no, why?
161 2013-10-12 00:50:45 <gmaxwell> skinnkavaj: Just wondering why you're using "u" for you? Usually I expect that to go along with "loldongs!". :)
162 2013-10-12 00:51:02 <sipa> roflcopter!
163 2013-10-12 00:52:09 <skinnkavaj> =)
164 2013-10-12 00:52:14 <skinnkavaj> maybe i should write a mail to bitstamp
165 2013-10-12 00:52:24 <skinnkavaj> but they are probably too big to care about more money.
166 2013-10-12 00:52:29 <skinnkavaj> and have their own shit to deal with
167 2013-10-12 00:52:37 <skinnkavaj> but its a nice looking exchange
168 2013-10-12 00:52:43 <skinnkavaj> perfect for the average joe buys.
169 2013-10-12 00:55:01 <MC1984> do people even really straight up start an exchange any more these days
170 2013-10-12 00:55:18 <MC1984> i thought those carefree days were gone
171 2013-10-12 00:55:19 <sipa> i'm sure some try :)
172 2013-10-12 00:55:31 <skinnkavaj> MC1984: i see new exchanges popping up all the time
173 2013-10-12 00:55:33 <skinnkavaj> and its good
174 2013-10-12 00:55:38 <skinnkavaj> because there will always be a need
175 2013-10-12 00:55:59 <MC1984> arnt people afraid of getting steamrolled by tax inspector scary or agent frightening now
176 2013-10-12 00:56:15 <skinnkavaj> MC1984: you can pax taxes right
177 2013-10-12 00:56:21 <skinnkavaj> if you handle a bitcoin exchange
178 2013-10-12 00:56:27 <skinnkavaj> god, why does everyone think its soo hard?
179 2013-10-12 00:56:35 <skinnkavaj> its not that hard to follow all the laws and pay all the taxes
180 2013-10-12 00:56:44 <sipa> sounds like a good start
181 2013-10-12 00:56:54 <skinnkavaj> yeah but hilarious is that
182 2013-10-12 00:56:56 <skinnkavaj> all exchanges
183 2013-10-12 00:57:00 <sipa> but judging by how many have tried and failed, i doubt it's that easy :)
184 2013-10-12 00:57:00 <skinnkavaj> did not follow any laws
185 2013-10-12 00:57:02 <skinnkavaj> like all big ones
186 2013-10-12 00:57:07 <skinnkavaj> did not follow AML, KYC etc.
187 2013-10-12 00:57:12 <MC1984> home si doubt the taxman knows what taxes you should be paying on that shit
188 2013-10-12 00:57:16 <Luke-Jr> skinnkavaj: yes they did
189 2013-10-12 00:57:22 <skinnkavaj> Luke-Jr: not from start
190 2013-10-12 00:57:23 <MC1984> but when he figures it out you dont want to be the first
191 2013-10-12 00:57:24 <Luke-Jr> skinnkavaj: and yes it is hard
192 2013-10-12 00:57:40 <skinnkavaj> Luke-Jr: it's not hard to follow laws
193 2013-10-12 00:57:50 <warren> skinnkavaj: the way accounting works, there's no way to calculate what each customer should owe in taxes with all the random deposits and withdrawals.  there's also the problem where how exactly to classify it on taxes isn't defined in most countries.
194 2013-10-12 00:57:51 <Luke-Jr> skinnkavaj: as I understand it from my very limited experience, it's basically impossible for a foreign company to comply with US laws
195 2013-10-12 00:58:11 <skinnkavaj> warren: is that the exchange owners problem? No. totally wrong.
196 2013-10-12 00:58:15 <MC1984> not hard to follow laws? Cool now we can kill all the lawyers
197 2013-10-12 00:58:35 <skinnkavaj> people should pay their own taxes on profit
198 2013-10-12 00:58:47 <skinnkavaj> on each tax year.
199 2013-10-12 00:58:49 <skinnkavaj> they can report it.
200 2013-10-12 00:58:53 <skinnkavaj> if they want to
201 2013-10-12 00:59:01 <skinnkavaj> its not up to the exchange to do it
202 2013-10-12 00:59:19 <skinnkavaj> the exchange should pay taxes on their profits as well
203 2013-10-12 00:59:59 <MC1984> try it then
204 2013-10-12 01:00:08 <MC1984> make sure you have legally limited liability though
205 2013-10-12 01:00:25 <skinnkavaj> MC1984: you just have to have balls
206 2013-10-12 01:00:35 <warren> skinnkavaj: other "exchanges" in the U.S. report capital gains directly to the IRS every year.
207 2013-10-12 01:00:36 <MC1984> instead of brains maybe
208 2013-10-12 01:01:08 <skinnkavaj> warren: you cannot do that with bitcoin, there is no way.
209 2013-10-12 01:01:13 <MC1984> if i ran an exchange i would ip block the whole of US at least
210 2013-10-12 01:01:14 <skinnkavaj> its technically impossible
211 2013-10-12 01:01:38 <MC1984> its just not worth fucking around with seeing as they can extradite from anywhere that speaks english for anything
212 2013-10-12 01:02:33 <skinnkavaj> look, the government will have to find other ways to collect taxes and they will find. they just have to adapt.
213 2013-10-12 01:02:38 <skinnkavaj> just like the file sharing industry has adapted
214 2013-10-12 01:02:41 <skinnkavaj> with spotify.
215 2013-10-12 01:02:51 <MC1984> omh you are so misinformed
216 2013-10-12 01:03:22 <MC1984> fucking with tax lines will bring a violent reaction that makes the copyright shenanigans look like gentle pleading
217 2013-10-12 01:04:00 <skinnkavaj> MC1984: ok so forbid bitcoin and drive it underground and make no taxes at all? is that a better alternative?
218 2013-10-12 01:04:12 <MC1984> DRUG WAR
219 2013-10-12 01:04:35 <skinnkavaj> MC1984: bitcoin wont be forbidden
220 2013-10-12 01:04:40 <skinnkavaj> only tin foil hats believe that
221 2013-10-12 01:04:51 <MC1984> i didnt say they were rational, i said they were violent and powerful
222 2013-10-12 01:06:37 <skinnkavaj> (02:57:48) (Luke-Jr) skinnkavaj: as I understand it from my very limited experience, it's basically impossible for a foreign company to comply with US laws
223 2013-10-12 01:06:44 <skinnkavaj> you are probably right
224 2013-10-12 01:07:18 <skinnkavaj> you don't have to serve other users then your own country
225 2013-10-12 01:07:46 <skinnkavaj> bitcoin works even if you have an local exchange for every country
226 2013-10-12 01:08:09 <MC1984> im british so im really getting a kick out of that
227 2013-10-12 01:16:31 <MC1984> aww shit speaking of drug war
228 2013-10-12 01:16:42 <MC1984> krok has made it to america
229 2013-10-12 01:17:37 <MC1984> good lord that shit makes crytal meth looks like vitamin pills
230 2013-10-12 01:18:44 <gmaxwell> getting a little too far off topic here.
231 2013-10-12 01:20:46 <MC1984> ok
232 2013-10-12 05:17:45 <warren> gmaxwell: is your position on centralized issuers on Ripple -> might as well do trading at the centralized issuer written somewhere in public?
233 2013-10-12 05:19:47 <gmaxwell> Yep.  There may be some "value" in basically obfuscating what they're really doing, to try to call it a decenteralized system when it really isn't in order to evade regulatory attention. I don't assign much value to that kind of argument, legal and leglislative systems have neigh infinite power to see past obfuscation (to the extent that they can see 'past' obfuscation that isn't).
234 2013-10-12 05:21:23 <warren> gmaxwell: can I use that as a quote?  I'm pointing to various opinions on colored coins and ripple.
235 2013-10-12 05:29:13 <gmaxwell> if you think it would be helpful. I'm not sure it will actually be informative to people other than you. I suspect it might need more background, e.g. the idea that ripple isn't really a decenteralized system is very surprising to many people, most attention on ripple has been on the weird gigantic premine, and all the astroturf support thats been bought with it. :)
236 2013-10-12 05:29:35 <gmaxwell> and the few people that get past that get hung up on the economics of IOUs. :P
237 2013-10-12 05:30:02 <gmaxwell> Whats the ripple premine worth at current exchange rates? has it topped a trillion dollars yet?
238 2013-10-12 05:31:03 <warren> If it really were about a network of IOU's, with everyone actually using the system that way, it might be different.
239 2013-10-12 05:31:45 <warren> Then there's also the issue of the doubtful security of their consensus method.
240 2013-10-12 05:32:00 <gmaxwell> warren: see https://bitcointalk.org/index.php?topic=144471.0 (and the subsiquent thread with some sparring betwen me and joel katz)
241 2013-10-12 05:32:32 <gmaxwell> warren: well I think their consensus method is "secure" as long as its controlled by a single administrative domain.
242 2013-10-12 05:33:01 <gmaxwell> if it's not I still have no clue what the necessary criteria are for convergence. I was never able to extract a theory about what one might be from the ripplers.
243 2013-10-12 05:33:14 <warren> So their consensus system works as long as you agree with yourself.
244 2013-10-12 05:34:51 <gmaxwell> It seemed a little stronger than that, e.g. it worked so long as it was 100% honest parties with fully meshed trust graph. But I don't actually know how you can achieve that at any scale unless there really is just one administrative domain controlling it. Maybe there is a way, but I've not seen it described.
245 2013-10-12 05:35:42 <gmaxwell> (also, the system would have to go offline if 51% percent of the parties in the fully meshed trust graph were offline for any reason, otherwise you risk unresolvable partitioning.)
246 2013-10-12 05:36:04 <warren> " unless there really is just one administrative domain controlling it" could be analogous to broadcast checkpoints
247 2013-10-12 05:37:22 <gmaxwell> yes, and you could imagine a slightly different version of this that didn't require the trust graph to be fully meshed but unwedged stuck states with broadcast checkpoints.
248 2013-10-12 12:09:16 <darsie> hi
249 2013-10-12 12:11:17 <darsie> I'd like 2 buttons in bitcoin-qt. 'Send all' which puts all your balance in the amount field and 'subtract fee - yes/no' which subtracts your fee from the send amount  if there is not enough balance left for the fee.
250 2013-10-12 12:11:40 <swulf--> send all would be terrible
251 2013-10-12 12:11:45 <swulf--> someone would accidentally hit it
252 2013-10-12 12:11:57 <darsie> It just fills out the amount field.
253 2013-10-12 12:12:04 <swulf--> with your entire balance
254 2013-10-12 12:12:07 <darsie> ye
255 2013-10-12 12:12:09 <darsie> s
256 2013-10-12 12:12:16 <swulf--> that's just asking for mistakes
257 2013-10-12 12:12:44 <darsie> I've seen it on at least two bitcoin exchanges.
258 2013-10-12 12:12:48 <darsie> it's good.
259 2013-10-12 12:12:48 <sipa> you can't know the fee before constructing the transaction, unfortunately
260 2013-10-12 12:13:24 <wumpus> a way to send the entire balance would be useful nevertheless
261 2013-10-12 12:13:31 <sipa> agree there
262 2013-10-12 12:13:39 <wumpus> but indeed, it's not trivial
263 2013-10-12 12:13:54 <sipa> just a checkbox "receiver pays fee" for example
264 2013-10-12 12:14:07 <swulf--> oooh neat
265 2013-10-12 12:14:16 <wumpus> yes that'd be nice
266 2013-10-12 12:14:26 <wumpus> would also make it more generic
267 2013-10-12 12:14:27 <swulf--> receiver pays fee, set fee equal to 200btc for hilariousness
268 2013-10-12 12:14:48 <wumpus> you're still the person that pays in the end
269 2013-10-12 12:15:03 <swulf--> yeah, receiver would just raise the price of whatever he's selling
270 2013-10-12 12:15:42 <darsie> I usually let the buyes of my btc pay the tx fee and deduct it from their btc balance.
271 2013-10-12 13:40:45 <_Sam---> hi, can someone explain to me the dangers of what can occur if one party controls more than 51% of the found blocks?
272 2013-10-12 13:45:50 <MC1984_> https://www.virustotal.com/en/file/64f6087c248b471be4ee8909dd5e605d6cef7f7e9f87c1330c282245c87b096e/analysis/1381545651/ interesting
273 2013-10-12 13:46:08 <pigeons> _Sam---: I don't know the answer to your question as asked "more than 51% of the found blocks" usually the danger of 51% of the hashing pwer is referred to as 51% attack
274 2013-10-12 13:46:21 <MC1984_> thats a report for the secp256k1 build warren made
275 2013-10-12 13:46:45 <MC1984_> ofc its a false positive because it actually does contain a bitcoin cpu miner
276 2013-10-12 13:46:45 <pigeons> _Sam---: https://en.bitcoin.it/wiki/Weaknesses#Attacker_has_a_lot_of_computing_power
277 2013-10-12 13:49:32 <_Sam---> thank you, reading up.
278 2013-10-12 13:52:34 <jouke> MC1984_: >_<
279 2013-10-12 13:58:21 <_Sam---> can anyone comment on the legality of running a bitcoin pool in the USA?
280 2013-10-12 13:58:23 <_Sam---> http://www.fincen.gov/statutes_regs/guidance/html/FIN-2013-G001.html
281 2013-10-12 13:59:01 <MC1984_> so dont
282 2013-10-12 13:59:21 <_Sam---> it was a question, you didnt provide an answer, but thanks
283 2013-10-12 13:59:22 <MC1984_> run it anywhere except the us
284 2013-10-12 13:59:35 <_Sam---> i see.  so its not OK in the usa?
285 2013-10-12 13:59:51 <MC1984_> i dont really know, but why take the risk?
286 2013-10-12 14:00:07 <MC1984_> bitcoin is global
287 2013-10-12 14:00:11 <Belxjander> _Sam---: right now the US situation is too risky...
288 2013-10-12 14:00:12 <_Sam---> the point isnt where the pool is located ; if i lvie in the USA then its illegal for me to run it.
289 2013-10-12 14:00:48 <MC1984_> well...........enjoy your citizenship i suppose
290 2013-10-12 14:01:14 <_Sam---> i wasn't trying to be a jerk ; i just was not aware what a touchy subject it was.
291 2013-10-12 14:01:22 <_Sam---> but now i am.  thank you.
292 2013-10-12 14:01:23 <Belxjander> _Sam---: you need to find out from fincen and all the rest of the administrators financial and jump through their hoops for the clowns going to town if you really want to be 100% legit... and it seems a whole lot of crap for very little in return
293 2013-10-12 14:01:56 <_Sam---> by the same standards, it would be impossble for someone in the USA to mine legally and turn that BTC into USD
294 2013-10-12 14:02:09 <MC1984_> since when has pool opping been in serious legal question anyway
295 2013-10-12 14:02:24 <Belxjander> _Sam---: they can register as a state-specific MSB and that seems to cover them
296 2013-10-12 14:02:24 <MC1984_> _Sam--- that may be the case
297 2013-10-12 14:03:23 <Belxjander> But I am not an american and have very little knowledge beyond what I hear on #bitcoin#? channels which is mostly 2nd hand at best
298 2013-10-12 14:03:55 <_Sam---> thank you for the information.  disappointing for me, but the law is the law.
299 2013-10-12 14:04:30 <MC1984_> this is not definitive advice
300 2013-10-12 14:04:41 <_Sam---> i understand -- i can read and ersearch further.
301 2013-10-12 14:05:00 <_Sam---> but the law seems to say, i live in the USA, if i a run a pool, im a money exchanging service and need to be registered.
302 2013-10-12 14:05:07 <_Sam---> to have to worry about being hassled over .0001 btc
303 2013-10-12 14:05:09 <_Sam---> not worth it
304 2013-10-12 14:05:26 <MC1984_> land of opportunity
305 2013-10-12 14:06:29 <_Sam---> im willing to at least see what it would take to become registered and legitimate ;  i have nothing to hide if its not ridiculous to do.
306 2013-10-12 14:07:03 <MC1984_> didnt i read being registered as a msb is like 5m dollar per state
307 2013-10-12 14:09:02 <_Sam---> dont know, but i will let you know.
308 2013-10-12 14:09:04 <_Sam---> "Registering with FinCEN is very easy, takes about 20 minutes and can be done through the BSA E-Filing System, described below."
309 2013-10-12 14:09:43 <MC1984_> so what then. what does fincen really want from you
310 2013-10-12 14:10:00 <MC1984_> you must log who found every block or something?
311 2013-10-12 14:10:17 <_Sam---> you know as much (if not more) than i do already.
312 2013-10-12 14:10:38 <_Sam---> it was just a surprise to find that 1.99 worth of BTC that i sell now and then, or the pools that i just finished and wanted to go live -- are not legal.
313 2013-10-12 14:10:59 <MC1984_> or do they want you registered the same way people want there whereabouts of each gun registerd....you know just in case
314 2013-10-12 14:12:14 <MC1984_> i dont know if anyone definitively knows theyre not legal
315 2013-10-12 14:12:45 <MC1984_> its probably one of those things thats fine and dandy until the day the man gets grumpy about it and decides to make an example of someone
316 2013-10-12 14:13:33 <_Sam---> i will find someone at my state level to speak with ; im friends with the attorney general.
317 2013-10-12 14:14:05 <MC1984_> well now thats a handy contact to have for sure
318 2013-10-12 14:14:42 <_Sam---> the other option is to work with someone who is already an established money service business.
319 2013-10-12 14:15:14 <MC1984_> i think thts what the likes of bitpay did?
320 2013-10-12 15:10:12 <super3> just put up a pull request. mainly just doc cleanup stuff: https://github.com/bitcoin/bitcoin/pull/3085
321 2013-10-12 15:35:02 <cjrouge> hi
322 2013-10-12 16:21:15 <sipa> https://gist.github.com/sipa/6951837
323 2013-10-12 16:21:22 <sipa> jgarzik: maybe interesting ^
324 2013-10-12 16:23:45 <sipa> petertodd, gmaxwell: ^
325 2013-10-12 16:38:37 <nanotube> seen in my bitcoind log: 2013-10-12 16:37:09 ERROR: Non-canonical signature: R value excessively padded <- some misbehaving third-party client?
326 2013-10-12 16:41:25 <sipa> nanotube: it's a malleability in bitcoin signatures we've been trying to kill for over a year now, i think
327 2013-10-12 16:42:32 <MC1984_> warren should probably take down that win32 libsecp256k1 build
328 2013-10-12 16:43:14 <sipa> nanotube: since 0.8.0, transactions with non-standard signatures (or public keys) are not relayed anymore
329 2013-10-12 16:43:19 <sipa> nanotube: but there are several old codebases out there that still create such transactions
330 2013-10-12 16:43:48 <nanotube> ah
331 2013-10-12 16:43:57 <sipa> they're valid in blocks, though
332 2013-10-12 16:44:01 <sipa> (for now...)
333 2013-10-12 16:44:05 <nanotube> i see thanks for the info :)
334 2013-10-12 16:44:38 <sipa> nanotube: one problem is that currently, which signatures are valid is entirely determined by openssl (since the reference client uses openssl to parse them)
335 2013-10-12 16:44:47 <sipa> and openssl accepts many not-really-correctly encoded ones
336 2013-10-12 16:45:08 <sipa> so any alternate implementation must really be bug-by-bug compatible with openssl
337 2013-10-12 16:45:09 <nanotube> aye, i've seen talk about this earlier. hence your secp256 custom code. :)
338 2013-10-12 16:45:31 <sipa> well it's a prerequisite for switching reference validation code to libsecp256k1
339 2013-10-12 16:45:50 <sipa> but also good for other full node implementations
340 2013-10-12 16:45:55 <nanotube> mm
341 2013-10-12 16:46:58 <sipa> not clear?
342 2013-10-12 16:49:05 <nanotube> i buy it. so once libsecp is well tested and used, i presume we'll set a sunset date for non-canonical encodings, and that'll be that?
343 2013-10-12 16:49:57 <sipa> well there are several reasons for wanting canonical encodings
344 2013-10-12 16:50:13 <sipa> one is being able to use non-openssl based code for validating nodes
345 2013-10-12 16:50:37 <sipa> another is malleability (the fact that you can change a valid transaction into another valid transaction, with a different txid)
346 2013-10-12 16:50:58 <sipa> both are very good reasons individually to kill non-canonical signatures
347 2013-10-12 16:51:36 <sipa> but we just can't do a softfork to enforce them until there are near-zero transactions being created anymore that use them
348 2013-10-12 16:51:58 <sipa> as that could kill some client/site/... without anyone even knowing about its existance
349 2013-10-12 16:52:17 <sipa> so for now, it's just a policy rules, preventing relay of such transactions
350 2013-10-12 16:54:09 <MC1984_> how far is libsecp from actual deployment really
351 2013-10-12 16:54:55 <MC1984_> i was under the impression it was one of those things that wanted like atleast a year of testing after you think its done just to make sure
352 2013-10-12 16:54:58 <sipa> MC1984_: i don't want the responsability of stolen coins or hard forked chains
353 2013-10-12 16:55:17 <sipa> and i'm not even sure what could happen to make me trust my own code enough for deployment
354 2013-10-12 16:55:27 <sipa> some smart people looking at it, i suppose
355 2013-10-12 16:56:03 <sipa> and i don't think that if there are bugs in it, we'll find them through randomized tests
356 2013-10-12 16:56:22 <MC1984_> i wonder if/why cant the foundation have a fund to hire some well known clever clogs to audit things like that
357 2013-10-12 16:57:55 <MC1984_> i mean how does the openssl project do it? Same sort of risk exposure it seems
358 2013-10-12 16:58:01 <MC1984_> more actually
359 2013-10-12 16:58:53 <sipa> more risk, but certainly many many more eyes as well
360 2013-10-12 17:00:17 <MC1984_> i wonder why we havent seen more papers from research departments about it
361 2013-10-12 17:00:34 <MC1984_> there are plenty that like researching p2p stuff
362 2013-10-12 17:01:09 <MC1984_> seems like distributed consensus would be a very interesting brand new area of research
363 2013-10-12 17:01:32 <MC1984_> maybe they are as we speak
364 2013-10-12 17:12:00 <nanotube> after restarting a node, i see in getinfo: "errors" : "Warning: Please check that your computer's date and time are correct! If your clock is wrong Bitcoin will not work properly." timeoffset is -1. box time is correct within a few milliseconds (ntpdate). is there a way to clear the message, or will it go away after some time?
365 2013-10-12 17:52:43 <michagogo> cloud|sipa: Have you spread it to crypto communities, forums, mailing lists, etc to look at?
366 2013-10-12 17:53:33 <michagogo> cloud|There are plenty of places you could put it for assorted people to look at and comment on
367 2013-10-12 17:54:11 <michagogo> cloud|And there are sure to be at least some smart cryptographers who can evaluate it
368 2013-10-12 17:54:22 <michagogo> cloud|(Btw, what's the license?)
369 2013-10-12 18:08:50 <MC1984> hmm my mail program is threading the dev list nicely
370 2013-10-12 18:09:09 <MC1984> downside: 400> unread. time for a catchup
371 2013-10-12 18:56:34 <MC1984> "Bitcoin-Qt 0.9 will (probably) have Pieter's work in this area to be
372 2013-10-12 18:56:34 <MC1984> usable very quickly, and download/verify the history in the background "
373 2013-10-12 18:56:38 <MC1984> really?
374 2013-10-12 18:57:03 <MC1984> i didnt know anyone was even seriously working on that
375 2013-10-12 18:59:38 <sipa> you nean headersfirst?
376 2013-10-12 18:59:56 <sipa> or where do you read that?
377 2013-10-12 19:00:16 <MC1984> devlist
378 2013-10-12 19:00:41 <MC1984> bout a month ago
379 2013-10-12 19:00:57 <sipa> that sounds like an spv node upgrading itself to full
380 2013-10-12 19:01:06 <sipa> not actually what i'm working on
381 2013-10-12 19:01:16 <MC1984> oh
382 2013-10-12 19:01:19 <sipa> but headersfirst is certainly a step in that direction
383 2013-10-12 19:02:05 <sipa> we don't actually have an spv wallet implementation
384 2013-10-12 19:02:13 <MC1984> i think we kond of need ot to stop haemorrhaging nodes......
385 2013-10-12 19:02:21 <MC1984> but no one has any hard data on it
386 2013-10-12 19:05:10 <warren> MC1984: you sure that isn't a false positive?
387 2013-10-12 19:05:25 <MC1984> ?
388 2013-10-12 19:05:49 <warren> virustotal
389 2013-10-12 19:06:03 <MC1984> oh, i know its a fasle positive
390 2013-10-12 19:06:10 <MC1984> bitcoin still has a cpu miner in it
391 2013-10-12 19:06:32 <MC1984> just found it strange/amusing some AV picked it up. I wonder if they flag the other miners too
392 2013-10-12 19:06:50 <maaku> MC1984: where was that postd?
393 2013-10-12 19:07:03 <MC1984> ???
394 2013-10-12 19:07:11 <maaku> "Bitcoin-Qt 0.9 will (probably) have Pieter's work in this area to be
395 2013-10-12 19:07:11 <maaku> usable very quickly, and download/verify the history in the background "
396 2013-10-12 19:07:19 <MC1984> devlist
397 2013-10-12 19:07:25 <maaku> ok
398 2013-10-12 19:07:46 <MC1984> seems it was somewhat mistaken though
399 2013-10-12 19:08:05 <Luke-Jr> Pieter is of course the ultimate authority for what his work does and doesn't do :P
400 2013-10-12 19:08:15 <sipa> haha :)
401 2013-10-12 19:10:40 <sipa> MC1984: what was the subject/sender of that mail?
402 2013-10-12 19:11:09 <MC1984> luke
403 2013-10-12 19:11:23 <Luke-Jr> 
ACTION hides

404 2013-10-12 19:11:46 <sipa> MC1984: regarding that virus warnig, i suspect many botnets distribute a bitcoind binary to mine against?
405 2013-10-12 19:11:54 <sipa> MC1984: or have done so in the past
406 2013-10-12 19:12:42 <MC1984> id have thought they would just mine to a pool
407 2013-10-12 19:12:55 <MC1984> but int he past when cpu mining was stilla thing, could have been viable
408 2013-10-12 19:13:28 <MC1984> i think i heard of cgminer etc getting put into payloads, i wonder if AVs flag that
409 2013-10-12 19:14:23 <MC1984> i tend to run every executable/file-that-does-stuff thru VT automatically when downloading
410 2013-10-12 19:14:40 <sipa> Luke-Jr, MC1984, warren: headerstfirst is just a different way of synchronizing the blockchain (which builds a headers-only representation withing seconds/minutes, and then fetches the blocks along the best path). I guess once we have that, implementing an SPV mode in the wallet becomes easy (disable the block fetching, and replace it by the wallet itself fetching filtered blocks). once all that is done, you could have a hybrid where you start in...
411 2013-10-12 19:14:46 <sipa> spv, and upgrade to full in the background
412 2013-10-12 19:15:29 <sipa> anyway, pulltester seems to break on headersfirst, and perhaps it's due to an actual problem in how it works
413 2013-10-12 19:15:33 <MC1984> so its not usable straight away with just headersfirst?
414 2013-10-12 19:15:52 <sipa> headersfirst is not different in user experience than what we have now
415 2013-10-12 19:16:01 <sipa> except it downloads faster :)
416 2013-10-12 19:16:09 <maaku> MC1984: SPV is about wallets. headers-first has nothing to do with wallets
417 2013-10-12 19:16:50 <sipa> indeed
418 2013-10-12 19:16:54 <warren> "(disable the block fetching, and replace it by the wallet itself fetching filtered blocks)" uses bloom?
419 2013-10-12 19:16:59 <sipa> yes
420 2013-10-12 19:17:03 <MC1984> oh its just a more efficient way to bootstrap a new node than the shitty way it does it now. well thats good
421 2013-10-12 19:17:25 <maaku> MC1984: the connection (i think) is when you consider committed index hashes
422 2013-10-12 19:17:51 <MC1984> wut
423 2013-10-12 19:17:56 <maaku> then you *could* use headers-first to get the hash of the most-work validation index, download that, and then validate backwards
424 2013-10-12 19:18:18 <maaku> gradually transitioning from SPV to full node security
425 2013-10-12 19:18:25 <maaku> but that's a lot further off than 0.9
426 2013-10-12 19:19:26 <sipa> i'm reimplementing all changes i made in my frst headersfirst implementation now as smaller pullreqs
427 2013-10-12 19:19:52 <MC1984> well, we need that
428 2013-10-12 19:19:57 <sipa> so it's not suxh a large patch, and the actual change to implement the behaviour change is smaller
429 2013-10-12 19:20:28 <MC1984> along with making the chain svelte enough going forward so that most people dont even notice or care its being upgraded to full node
430 2013-10-12 19:20:48 <sipa> yeah, it's nice prospect to have that
431 2013-10-12 19:21:35 <MC1984> id put a stronger point on it that nice myself. but thats just my beliefs
432 2013-10-12 19:29:30 <MC1984> background and use battery+radio can easily be seen as "abusive" by end
433 2013-10-12 19:29:30 <MC1984> "Gavin believes the future of computing
434 2013-10-12 19:29:30 <MC1984> have much longer battery life than laptops. Apps that spin up in the
435 2013-10-12 19:29:30 <MC1984> is mobile and tablets. I don't know about that, but let's assume for the
436 2013-10-12 19:29:30 <MC1984> sake of argument he turns out to be right. These devices are expected to
437 2013-10-12 19:29:31 <MC1984> users."
438 2013-10-12 19:29:39 <MC1984> tssk so many line breaks
439 2013-10-12 19:30:19 <MC1984> but oh god hes right, were heading into some sort of thin terminal nightmare becuase batteries havent improved in 2 decades fffffffffff
440 2013-10-12 19:30:45 <MC1984> graphene cells when??
441 2013-10-12 19:31:03 <petertodd> MC1984: welcome to the reality that IC moores law scaling is completely unprecidented in human history...
442 2013-10-12 19:31:29 <sipa> petertodd: seen https://gist.github.com/sipa/6951837 ?
443 2013-10-12 19:31:35 <petertodd> MC1984: it's normal for technology to move sloooowly, which is my my analog electronics designs use parts that are 10 years old
444 2013-10-12 19:31:37 <MC1984> were all living though many things are that unprecedented
445 2013-10-12 19:31:52 <petertodd> sipa: oh cool!
446 2013-10-12 19:31:57 <MC1984> THE SINGULARITY IS NEAR
447 2013-10-12 19:32:04 <MC1984> woops sorry that just slipped out
448 2013-10-12 19:32:08 <petertodd> MC1984: don't count on it :)
449 2013-10-12 19:32:34 <sipa> the singularity has been reached long ago
450 2013-10-12 19:32:44 <petertodd> sipa: you had a chance to use the script in anger yet?
451 2013-10-12 19:32:47 <sipa> they're just smart enough to not let us realize it
452 2013-10-12 19:33:00 <petertodd> lol
453 2013-10-12 19:33:02 <sipa> petertodd: i've tested all parts except the push at the end
454 2013-10-12 19:33:27 <petertodd> sipa: I'd try it out, but I have even less of a reason to use it :P
455 2013-10-12 19:34:08 <sipa> can you send some tiny pullreq to bitcoin-seeder or something? :p
456 2013-10-12 19:34:51 <petertodd> sipa: ha, I gotta look into TD's report about testnet acting up, but I've got two deadlines this weekend :/
457 2013-10-12 19:35:20 <sipa> weekend deadline is worst deadline
458 2013-10-12 19:35:56 <petertodd> sipa: heh, yup, I was supposed to be down south caving, but someone missed their deadline for getting a car :P
459 2013-10-12 19:36:29 <MC1984> caving is great
460 2013-10-12 19:36:54 <petertodd> MC1984: meh, the hipster in me says it was better when it was still underground
461 2013-10-12 19:37:04 <MC1984> hueehehhe
462 2013-10-12 19:38:38 <michagogo> sipa: question about https://github.com/bitcoin/bitcoin/pull/3085#issuecomment-26202486
463 2013-10-12 19:38:45 <michagogo> why is 0.8.5 in parentheses?
464 2013-10-12 19:38:52 <sipa> because it doesn't exist
465 2013-10-12 19:38:59 <sipa> oops
466 2013-10-12 19:39:05 <sipa> it does!
467 2013-10-12 19:39:54 <sipa> fixed
468 2013-10-12 19:40:09 <michagogo> k :-)
469 2013-10-12 19:40:37 <petertodd> sipa: thre's your pull-req
470 2013-10-12 19:42:07 <sipa> petertodd: and it's merged :)
471 2013-10-12 19:42:34 <petertodd> ha, nice
472 2013-10-12 19:42:46 <petertodd> oh, and reminds me, sign my PGP key :P
473 2013-10-12 19:42:58 <sipa> i can't remember that i verified it!
474 2013-10-12 19:43:21 <petertodd> lol, was awhile ago... and you probably don't have an email client that can verify my signatures...
475 2013-10-12 19:43:42 <petertodd> though you can do git log --show-signature on the bitcoin github and notice all my sigs
476 2013-10-12 19:44:00 <petertodd> 
ACTION has a looser definition of verify than the pedants

477 2013-10-12 19:44:34 <sipa> 
ACTION -pedantic

478 2013-10-12 19:45:16 <petertodd> The real irony with web-of-trust verification, is that for all the effort they go to to make a nice decentralized system, signing parties depend on government issued photo-id...
479 2013-10-12 19:46:32 <sipa> i'm sure i've mentioned this before, but that's why i think pgp identities should list what authority they claim to link with
480 2013-10-12 19:46:48 <petertodd> Verification based on "I have a lot of evidence that you're the guy commonly thought of as Peter Todd within my social community" makes far more sense.
481 2013-10-12 19:46:59 <michagogo> petertodd: Hmm, what kind of verification process would you use for a key that represents a user that's not their government-recognized name?
482 2013-10-12 19:47:02 <petertodd> Yup, and WoT authority should generally be thought of as social group.
483 2013-10-12 19:47:13 <sipa> and that could be either "government-issued identity of $COUNTRY", or "github committer $X to project $Y"
484 2013-10-12 19:47:22 <michagogo> (re: "[22:45:08] <petertodd> The real irony with web-of-trust verification...")
485 2013-10-12 19:47:43 <petertodd> michagogo: Nothing more than "Is Foo Bar the name people in this social group associate with that rough idea of an identity?"
486 2013-10-12 19:48:06 <petertodd> michagogo: which means in the case of a name conflict I might very well ask someone to disambiguate themselves in some way in theory
487 2013-10-12 19:49:05 <michagogo> petertodd: Isn't signing a key supposed to be a statement of "I have verified that this person is who he claims to be"?
488 2013-10-12 19:49:18 <petertodd> michagogo: define "claims" :P
489 2013-10-12 19:49:29 <michagogo> "who the key says he is"
490 2013-10-12 19:49:52 <petertodd> "Hi, my name also happens to be Michagogo, and I don't want to give out my email"
491 2013-10-12 19:50:06 <michagogo> My point exactly :-P
492 2013-10-12 19:50:25 <michagogo> hence "[22:46:49] <michagogo> petertodd: Hmm, what kind of verification process would you use for a key that represents a user that's not their government-recognized name?"
493 2013-10-12 19:50:37 <petertodd> michagogo: common sense
494 2013-10-12 19:50:56 <petertodd> See, if IRC was PGP signed, I'd happily sign your PGP key at some point.
495 2013-10-12 19:51:00 <sipa> michagogo: it's all about what identity you claim to have
496 2013-10-12 19:51:25 <sipa> if you claim your name, as recognized by country X, is Y, then i'll want to see a government-issued id before signing your key
497 2013-10-12 19:51:56 <sipa> if you claim your registered nick on IRC is X, i'll just ask to verify that you indeed control that nick
498 2013-10-12 19:52:42 <michagogo> So, would you sign 628ECF0C ?
499 2013-10-12 19:53:13 <Luke-Jr> name/nick are meaningless
500 2013-10-12 19:53:21 <Luke-Jr> multiple people have the same name/country
501 2013-10-12 19:53:21 <sipa> eh, no
502 2013-10-12 19:53:35 <Luke-Jr> and it really says nothing about who someone IS
503 2013-10-12 19:53:35 <sipa> michagogo: no, i won't sign that
504 2013-10-12 19:54:27 <Luke-Jr> really, signatures should have comments associated for what they mean
505 2013-10-12 19:54:58 <Luke-Jr> who sipa is to me: bitcoin expert, bitcoind developer responsible for ultraprune etc
506 2013-10-12 19:55:11 <Luke-Jr> "Pieter Wuille" is meaningless in itself
507 2013-10-12 19:55:13 <michagogo> sipa: What would make you change your mind? (note that this is hypothetical -- just using my key as an example, not actually asking you to sign)
508 2013-10-12 19:56:11 <sipa> michagogo: if the identity (the comment in it), would explicitly say something like "registered nickname michagogo on freenode IRC", i might
509 2013-10-12 19:56:39 <sipa> as then it's clear that i'm only signing the claim that this key belongs to the one controlling that nickname
510 2013-10-12 19:56:43 <petertodd> speaking of, this is part of what I used with warren to verify my identity: https://s3.amazonaws.com/peter.todd/pgp-statement.tar.bz2
511 2013-10-12 19:57:03 <michagogo> Ah, Interesting
512 2013-10-12 19:57:14 <michagogo> s/I/i/
513 2013-10-12 19:57:26 <skinnkavaj> Please explain, why do people always say that you should sign a message with an adress? Why isn't enough to send like 0.001 BTC from that adress to prove that you own it?
514 2013-10-12 19:57:47 <Luke-Jr> skinnkavaj: you can't send coins from addresses at all
515 2013-10-12 19:57:48 <petertodd> michagogo: I would argue that, in conjunction with other verification, should be enough for you to sign my PGP key.
516 2013-10-12 19:57:54 <Luke-Jr> skinnkavaj: addresses are only used to recieve
517 2013-10-12 19:58:10 <sipa> skinnkavaj: why 1) pay  2) burden the network  3) reveal your identity to the entire world
518 2013-10-12 19:58:26 <sipa> skinnkavaj: signing a message and sending it to the one who cares has none of these disadvantages
519 2013-10-12 19:58:30 <skinnkavaj> sipa: Don't you have to pay anyway to sign a message? Doesnt it get broadcasted to the network?
520 2013-10-12 19:58:35 <sipa> no
521 2013-10-12 19:58:49 <sipa> it's purely local
522 2013-10-12 19:58:55 <sipa> you do with the signature what you want
523 2013-10-12 20:00:07 <Luke-Jr> just because a coin had previously been sent to said address previously, also does not prove the sender of the transaction owns that address
524 2013-10-12 20:01:27 <petertodd> ah, and in my pgp-statement.txt I said I timestamped the statement; this is the tx I did it in: ad3d32ae11947ae3f07322dfadcc1c418b529091bffd89c901682ebdb8bd5404
525 2013-10-12 20:02:05 <petertodd> sipa: there, that should be enough for you to sign my key, lol
526 2013-10-12 20:04:42 <michagogo> sipa: So something like that?
527 2013-10-12 20:08:27 <petertodd> michagogo: see pm
528 2013-10-12 20:23:43 <gribble> slush was last seen in #bitcoin-dev 28 weeks, 2 days, 11 hours, 25 minutes, and 28 seconds ago: <slush> not really :(
529 2013-10-12 20:23:43 <ThomasV> !seen slush
530 2013-10-12 20:39:34 <MC1984_> http://goldbug.sourceforge.net/ anyone know anything about this? Not sure i trust and encrypted comms thing with a cartoon dog..
531 2013-10-12 20:40:34 <pigeons> i trust cartoon dogs more than cartoon paperclips
532 2013-10-12 20:44:19 <MC1984_> IT LOOKS LIKE YOURE TRYING TO MAKE A JOKE
533 2013-10-12 20:47:59 <sipa> http://www.deanliou.com/winrg/
534 2013-10-12 20:48:21 <skinnkavaj> gmaxwell: whats your opinion on bitshares?
535 2013-10-12 20:51:24 <MC1984_> sipa shit i havent seen that for years
536 2013-10-12 20:53:05 <MC1984_> its a perfect simulation of windows ME. id know.
537 2013-10-12 20:54:22 <sipa> MC1984_: i switched to linux shortly after winme :
538 2013-10-12 20:54:24 <sipa> :p
539 2013-10-12 20:55:05 <MC1984_> i tried to opensuse a few years ago but never got wifi working
540 2013-10-12 20:55:08 <MC1984_> so i gave up
541 2013-10-12 20:55:24 <sipa> i think it was debian 2.2
542 2013-10-12 21:14:26 <michagogo> sipa: So are the 2 new UIDs I added to 628ECF0C the kind of thing you were talking about?
543 2013-10-12 22:17:45 <sipa> wumpus: there?
544 2013-10-12 22:26:14 <sipa> ;;calc [diff]*2**48/65535
545 2013-10-12 22:26:15 <gribble> Error: invalid syntax (<string>, line 1)
546 2013-10-12 22:26:26 <sipa> ;;diff
547 2013-10-12 22:26:27 <gribble> 1.8928124928103292E8
548 2013-10-12 22:26:49 <sipa> ;;calc log([diff])/log(2)+32
549 2013-10-12 22:26:50 <gribble> Error: invalid syntax (<string>, line 1)
550 2013-10-12 22:28:39 <gribble> 59.4959687518
551 2013-10-12 22:28:39 <sipa> ;;calc log(1.8928*10**8*2**48/65535)/log(2)
552 2013-10-12 22:29:57 <gribble> 3511950090279321600000
553 2013-10-12 22:29:57 <sipa> ;;calc 1.8928*10**8*2**32/600*86400*30
554 2013-10-12 22:30:31 <gribble> 71.5727623349
555 2013-10-12 22:30:31 <sipa> ;;calc log(1.8928*10**8*2**32/600*86400*30)/log(2)
556 2013-10-12 22:32:32 <sipa> oh, we passed 2**72 hashes a month ago
557 2013-10-12 22:33:02 <sipa> wait, what?
558 2013-10-12 22:34:08 <gmaxwell> We still haven't beat the lowest hash of 0000000000000000004bb6e7e2661661ba9809062d90c3121933d6d02c8bd763.
559 2013-10-12 22:35:09 <sipa> ;;calc 2**(72.6968-71.5727)
560 2013-10-12 22:35:10 <gribble> 2.17965530138
561 2013-10-12 22:35:30 <sipa> tje entire chain is 2.18 PoW-months now? :o
562 2013-10-12 22:35:33 <sipa> *the
563 2013-10-12 22:36:26 <MC1984_> wat
564 2013-10-12 22:36:30 <gmaxwell> sipa: it's been fightening like that for a while now.
565 2013-10-12 22:39:16 <sipa> ;;nethash
566 2013-10-12 22:39:17 <gribble> 1834299.06352
567 2013-10-12 22:40:36 <sipa> ;;calc 2**(72.6989-log([nethash]*10**9*86400)/log(2))
568 2013-10-12 22:40:37 <gribble> 48.3688392051
569 2013-10-12 22:40:47 <sipa> even worse... 48 days
570 2013-10-12 22:40:57 <gmaxwell> hm? I get (2^72.697147/(2^72.697147-2^72.696993))/144. = 65.06
571 2013-10-12 22:41:58 <sipa> that's assuming there are 144 blocks per day
572 2013-10-12 22:42:08 <sipa> there are more now :)
573 2013-10-12 22:43:47 <gmaxwell> Hm. I think your way is double-counting part of the increase, and my way is not counting part of it. (your way double counts the part thats included int the last difficulty update, any my way doesn't include the increase over that)
574 2013-10-12 22:44:13 <gmaxwell> oh well actually I dunno what [nethash] comes from.
575 2013-10-12 22:44:26 <sipa> from my site :)
576 2013-10-12 22:44:35 <gmaxwell> I retract my complaint then. 48 it is.
577 2013-10-12 22:45:12 <sipa> that's pretty awful
578 2013-10-12 22:45:46 <sipa> i should make a graph of pow-days in the chain over time...
579 2013-10-12 22:45:55 <gmaxwell> yup, but the awfulness was really there all along it was just easy to pretend that someone couldn't buy a million dollars in chips and rewrite the network when no one had a million dollars in chips yet.
580 2013-10-12 22:46:06 <sipa> true
581 2013-10-12 23:32:50 <sipa> gmaxwell: http://bitcoin.sipa.be/powdays-50k.png
582 2013-10-12 23:34:21 <sipa> it's been worse, though: http://bitcoin.sipa.be/powdays-ever.png
583 2013-10-12 23:39:40 <sipa> not really the shape i would have expected, really
584 2013-10-12 23:42:02 <gmaxwell> I don't quite get the initial part.
585 2013-10-12 23:42:17 <gmaxwell> like where it hits 1200.. how is that possible?
586 2013-10-12 23:43:33 <sipa> the 7-day hashrate estimate dropped to ~1 MH/s around august/september 2009
587 2013-10-12 23:44:20 <gmaxwell> oh right, this is hashrate estimate not difficulty.
588 2013-10-12 23:44:38 <sipa> perhaps i should use / max(x.hashrate for x in past)
589 2013-10-12 23:44:42 <gmaxwell> how hard would ... lol
590 2013-10-12 23:44:58 <gmaxwell> I was going to suggest looking at max hashrate ever.
591 2013-10-12 23:45:19 <gmaxwell> (er, by ever I mean in the past from each point)
592 2013-10-12 23:45:56 <gmaxwell> yea, means that it won't improve during times when the hashrate fell... which I think its a reasonable measure of security, since that other rate could come online.
593 2013-10-12 23:46:53 <sipa> ... that looks interesting
594 2013-10-12 23:54:24 <warren> how do you define POW-equivalent-days?
595 2013-10-12 23:57:24 <gmaxwell> These are graphs of  "at day X how long would it take for day X's hashrate to replace the chain (fork from genesis)"
596 2013-10-12 23:57:28 <sipa> warren: the number of days you would need at the current hashrate, to do as much work as the entire chain
597 2013-10-12 23:58:07 <sipa> in theoretical conditions, it should just be a straight line with a slope of one day per day
598 2013-10-12 23:58:14 <sipa> (=constant hashrate)