1 2011-11-17 00:15:15 <molecular> tcatm, people are asking about chart.glbse.com. They're really missing it ;). Do you maintain that?
  2 2011-11-17 00:33:36 <denisx> is the coder of phoenix here sometimes?
  3 2011-11-17 00:37:38 <luke-jr> denisx: at least one phoenix contributor hangs in #eligius sometimes
  4 2011-11-17 00:38:42 <denisx> I wonder what happens when I do not deliver the "X-Long-Polling" header
  5 2011-11-17 00:40:02 <luke-jr> then you don't enable LP&?
  6 2011-11-17 00:40:23 <denisx> luke-jr: and phoenix respects that?
  7 2011-11-17 00:40:42 <luke-jr> dunno
  8 2011-11-17 00:40:47 <luke-jr> phoenix is full of bugs
  9 2011-11-17 00:40:58 <denisx> ah fuck, learning by doing...
 10 2011-11-17 00:41:05 <denisx> I will just try it ;)
 11 2011-11-17 00:42:34 <markkio> What do I use to decrypt my Gpg again?
 12 2011-11-17 00:44:02 <luke-jr> you use GPG to decrypt PGP.
 13 2011-11-17 00:46:39 <denisx> luke-jr: I want to disable LP for the botnet, but instead it seems the traffic is increasing
 14 2011-11-17 00:46:43 <denisx> I don't know why
 15 2011-11-17 00:47:20 <luke-jr> denisx: of course. without LP, you get polls every 1-10 seconds.
 16 2011-11-17 00:47:35 <denisx> yeah, seems like 15
 17 2011-11-17 00:47:36 <graingert> gpg
 18 2011-11-17 00:47:40 <luke-jr> LP, like most extensions, was invented to REDUCE load
 19 2011-11-17 00:48:02 <denisx> luke-jr: but I can't handle LP calls from 500k clients
 20 2011-11-17 00:48:15 <luke-jr> denisx: then maybe you should join my project :p
 21 2011-11-17 00:48:28 <luke-jr> why the heck do you have 500k clients btw
 22 2011-11-17 00:48:37 <denisx> luke-jr: a botnet is hitting me
 23 2011-11-17 00:49:48 <luke-jr> denisx: anyhow, my Python pool server can probably handle 500k LP clients, when finished
 24 2011-11-17 01:09:45 <doublec> by the time you've sent 500,000 getwork requests to the LP clients, the next block would have come around
 25 2011-11-17 01:10:02 <denisx> doublec: no, its not that slow
 26 2011-11-17 01:11:36 <graingert> would it be possible to create a distributed pool miner
 27 2011-11-17 01:11:44 <doublec> I'd have thought 1,000 getworks/s was about normal? Then it's nearly 10 minutes to send out the longpoll results
 28 2011-11-17 01:12:40 <graingert> in which each miner shows the rest their block with the reward of the previous set in, if they only partially hit the target
 29 2011-11-17 01:13:19 <graingert> with their address in the extra nonce etc
 30 2011-11-17 01:13:43 <graingert> that way you could create a pool with no pool operator
 31 2011-11-17 01:14:00 <doublec> graingert: have you seen p2pool?
 32 2011-11-17 01:15:13 <graingert> cool
 33 2011-11-17 01:15:24 <graingert> 0.5% is kept as a fee ?
 34 2011-11-17 01:15:27 <graingert> a fee to whom?
 35 2011-11-17 01:15:43 <doublec> the p2pool creator I assume
 36 2011-11-17 01:16:08 <graingert> a hard coded bitcoin address
 37 2011-11-17 01:16:19 <graingert> would not be hard to make a free one
 38 2011-11-17 01:41:14 <luke-jr> denisx: so want to join me on my pool server? :P
 39 2011-11-17 01:42:14 <denisx> I need to discuss that with my partners in crime
 40 2011-11-17 01:43:50 <denisx> but did I understand that correctly. it is written in python?
 41 2011-11-17 01:56:03 <luke-jr> denisx: currently 100% python
 42 2011-11-17 03:08:51 <denisx> LP do not use keep-alive, right?
 43 2011-11-17 03:08:58 <denisx> LP calls that is
 44 2011-11-17 03:10:17 <luke-jr> LP calls are LP.
 45 2011-11-17 03:10:33 <luke-jr> they can use HTTP keep-alive or not, it's immaterial.
 46 2011-11-17 03:11:34 <denisx> luke-jr: yeah sure, but do I need to send an "Connection: closed" at the end?
 47 2011-11-17 03:11:52 <luke-jr> if you want to refuse the keep-alive&
 48 2011-11-17 03:11:55 <luke-jr> read the HTTP spec
 49 2011-11-17 03:15:30 <bd_> denisx: http://www.w3.org/Protocols/rfc2616/rfc2616-sec8.html section 8.1 :)
 50 2011-11-17 03:16:03 <bd_> note that the server sending a 'Connection: closed' is a SHOULD statement - it's not mandatory, but it's a good idea
 51 2011-11-17 03:16:10 <bd_> as otherwise you might end up wasting the client's time
 52 2011-11-17 03:16:41 <luke-jr> "close", not "closed" BTW
 53 2011-11-17 03:17:49 <denisx> luke-jr: thanks for catching that
 54 2011-11-17 03:18:06 <bd_> er, yes. oops :)
 55 2011-11-17 03:18:16 <denisx> I ask because I found that: http://comments.gmane.org/gmane.comp.lib.libevent.user/989
 56 2011-11-17 03:18:26 <denisx> (there he said "closed")
 57 2011-11-17 03:19:13 <denisx> and Iam not that Denis
 58 2011-11-17 03:19:25 <bd_> I haven't worked with evhttp
 59 2011-11-17 03:19:37 <denisx> pushpool does
 60 2011-11-17 03:19:42 <bd_> but a HTTP server can terminate the connection between requests whenever it wants. Connection: close is a courtesy.
 61 2011-11-17 03:23:31 <cocktopus> btw when discussing LP, be sure to take into account QoS for it - Slush has implemented it for his largest miners
 62 2011-11-17 03:23:38 <cocktopus> they get priority LPs
 63 2011-11-17 03:26:24 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 64 2011-11-17 03:26:25 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 65 2011-11-17 03:26:28 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 66 2011-11-17 03:26:29 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 67 2011-11-17 03:26:30 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 68 2011-11-17 03:26:32 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 69 2011-11-17 03:26:33 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 70 2011-11-17 03:26:35 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 71 2011-11-17 03:26:37 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 72 2011-11-17 03:26:38 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 73 2011-11-17 03:26:39 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 74 2011-11-17 03:26:40 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 75 2011-11-17 03:26:42 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 76 2011-11-17 03:26:44 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 77 2011-11-17 03:26:45 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 78 2011-11-17 03:26:47 <OneFixt> nanotube: theymos jgarzik
 79 2011-11-17 03:26:47 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 80 2011-11-17 03:26:48 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 81 2011-11-17 03:26:50 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 82 2011-11-17 03:26:52 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 83 2011-11-17 03:26:54 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 84 2011-11-17 03:26:56 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 85 2011-11-17 03:26:57 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 86 2011-11-17 03:26:59 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 87 2011-11-17 03:27:00 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 88 2011-11-17 03:27:02 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 89 2011-11-17 03:27:03 <p3N1S> nigger cracker beaner wetback penis slut fuck my asshole shit cunt jap whore vagina retard faggot @$$ ass nigga poof douche bag pussy beaver!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 90 2011-11-17 03:27:36 <OneFixt> jgarzik: thanks; could i get ops?
 91 2011-11-17 03:28:01 <OneFixt> are you able to set the flag?
 92 2011-11-17 03:28:07 <jgarzik> OneFixt: no
 93 2011-11-17 03:28:15 <OneFixt> ok; thanks
 94 2011-11-17 03:29:27 <cjdelisle> nice poetry, albeit redundant
 95 2011-11-17 03:30:07 <denisx> what is a wetback?
 96 2011-11-17 03:30:15 <denisx> the rest I know ;)
 97 2011-11-17 03:30:22 <cocktopus> derrogitory term for mexican person
 98 2011-11-17 03:31:12 <cjdelisle> @$$ <-- that's the only one that's new to me
 99 2011-11-17 03:31:26 <cjdelisle> must be a variable in perl
100 2011-11-17 03:31:43 <cjdelisle> hell, maybe it's a perl script
101 2011-11-17 03:31:58 <luke-jr> cjdelisle: it's invalid in perl
102 2011-11-17 03:32:03 <luke-jr> $$ is the PID
103 2011-11-17 03:32:18 <cjdelisle> ahh, so that's where bash got it from
104 2011-11-17 03:32:34 <luke-jr> you can't dereference a scalar (PID) to an array
105 2011-11-17 03:32:36 <cjdelisle> bad ideas die hard
106 2011-11-17 03:32:37 <luke-jr> list*
107 2011-11-17 03:39:55 <bitcoiner> luke-jr!!!
108 2011-11-17 03:42:14 <cocktopus> help help, #bitcoin-court is being flooded too
109 2011-11-17 03:44:58 <theymos> Maybe someone should complain to the Freenode staff.
110 2011-11-17 03:46:09 <Graet> they sigh, "bitcoin again"
111 2011-11-17 03:47:42 <cocktopus> it seems to have stopped for now
112 2011-11-17 04:06:41 <neofutur> if needed we have freenode staff on #bitcoin-ops fyi
113 2011-11-17 04:08:41 <Graet> cant join :/
114 2011-11-17 04:08:43 <Graet> :P
115 2011-11-17 04:20:07 <neofutur> its for bitcoin channel ops Graet
116 2011-11-17 04:20:26 <neofutur> ask to nanotube to be added on the access list is you re op here
117 2011-11-17 04:20:41 <neofutur> SomeoneWeird: you should be on #nitcoin-ops ;)
118 2011-11-17 04:21:10 <Graet> lol
119 2011-11-17 04:21:16 <SomeoneWeird> invite only neofutur
120 2011-11-17 04:24:09 <Graet> i have ops in a few bitcoin related chanels neofutur , but i'm guessing thats not what you mean ;)
121 2011-11-17 04:25:05 <neofutur> yes it is
122 2011-11-17 04:25:48 <SomeoneWeird> neofutur, /invite me?
123 2011-11-17 04:25:49 <neofutur> any #bitcoin-* channel can ask to be on #bitcoin-ops
124 2011-11-17 04:26:26 <neofutur> SomeoneWeird: I /invite you but you still need to ask nanotube to be in the access liste and be able to invite yourself
125 2011-11-17 04:26:52 <SomeoneWeird> i know, but i got a bounce so I wont be leaving anytime soon
126 2011-11-17 04:28:36 <Graet> hmm several bitcoin related, but #bitcoin-police is only public #bitcoin- one
127 2011-11-17 04:29:20 <SomeoneWeird> :)
128 2011-11-17 06:29:16 <Prattler> don't ban pls :D
129 2011-11-17 06:30:04 <lolcat> Prattler: ?
130 2011-11-17 06:30:35 <Prattler> check out channel logs, hehe
131 2011-11-17 06:34:20 <lolcat> Prattler: ahh
132 2011-11-17 06:34:40 <Prattler> that was funny :D
133 2011-11-17 08:25:16 <tcatm> molecular: what are they missing on charts.glbse.com?
134 2011-11-17 08:25:47 <tcatm> oh it's broken again
135 2011-11-17 09:29:51 <molecular> tcatm, it's been broken for some days ;)
136 2011-11-17 09:30:39 <molecular> tcatm, I actually thought "out of coins" meant: I need some donations, if I don't receive any, this will be "offline" ;)
137 2011-11-17 09:35:17 <tcatm> molecular: no, it's the old 500 error page
138 2011-11-17 09:36:28 <molecular> yeah, people told me
139 2011-11-17 09:36:38 <molecular> tcatm, thanks for fixing
140 2011-11-17 09:36:41 <molecular> you're great!
141 2011-11-17 09:52:39 <erus`> wow setting up a windows server is as eassy as arch linux :)
142 2011-11-17 09:53:05 <erus`> i havnt set up any services yet tho
143 2011-11-17 11:10:53 <coderrr> damn, the find a subset that sums to x is actually solvable really fast
144 2011-11-17 11:13:27 <Mqrius> knapsack problem?
145 2011-11-17 11:14:04 <coderrr> Mqrius, yea
146 2011-11-17 11:14:27 <coderrr> http://www.diku.dk/hjemmesider/ansatte/pisinger/subsum.c taht shit solves it really quickly
147 2011-11-17 11:14:31 <Mqrius> Isn't that supposed to be difficult? :P
148 2011-11-17 11:14:42 <coderrr> i was looking into how feasible it would be to identify multiple btc transactions that sum to a given amt
149 2011-11-17 11:14:50 <coderrr> yea i thot so, but not since '99 i guess
150 2011-11-17 11:15:10 <coderrr> wikipedia only mentions thsi doods paper in passing
151 2011-11-17 11:15:20 <coderrr> http://en.wikipedia.org/wiki/Subset_sum_problem#cite_note-Pisinger09-2
152 2011-11-17 11:17:45 <coderrr> Mqrius, i create a random array of 1mill numbers and it a subset that matches any # i throw at it in less than a second
153 2011-11-17 11:18:35 <coderrr> it *finds
154 2011-11-17 11:18:53 <coderrr> s/matches/sums/
155 2011-11-17 11:20:32 <coderrr> hrm, its probly that the hard part is finding the smallest set
156 2011-11-17 11:21:38 <Mqrius> Heh, yeah
157 2011-11-17 11:23:29 <Mqrius> the bitcoin problem is different though, you want any set without transaction fees :)
158 2011-11-17 11:34:50 <epscy> coderrr: why would you want to do that?
159 2011-11-17 11:35:39 <coderrr> epscy, was just thinking about anon stuff, like if you want to pay someone from multiple addresses in multipel tx's as to not link the addresses
160 2011-11-17 11:35:53 <coderrr> if someone knew what the payment amount would be, how hard it would be for them to identify the transactions
161 2011-11-17 11:41:11 <CIA-89> libbitcoin: Kamil Domanski * r184d9813d529 /README: updated README
162 2011-11-17 12:05:49 <epscy> oh i see
163 2011-11-17 12:57:58 <MaxSan> Hey everyone
164 2011-11-17 12:58:18 <MaxSan> anyone point me to some information about communicting with the bitcoin network via a web service?
165 2011-11-17 13:01:48 <[eval]> MaxSan: you should check out etotheipi_'s posts on bitcointalk.org
166 2011-11-17 13:02:13 <MaxSan> any link [eval]
167 2011-11-17 13:02:26 <[eval]> https://bitcointalk.org/index.php?topic=29416.0
168 2011-11-17 13:05:21 <MaxSan> cheers
169 2011-11-17 13:05:37 <[eval]> oh wait... via web service
170 2011-11-17 13:05:43 <[eval]> i missed that part
171 2011-11-17 13:05:57 <[eval]> are you talking about the JSON-RPC interface of bitcoind?
172 2011-11-17 13:06:36 <MaxSan> yeh think thats whats needed
173 2011-11-17 13:07:11 <MaxSan> wanna setup a WCF service between bitcoind and vanitygen to make produce address's on the fly
174 2011-11-17 13:08:48 <[eval]> you're going to have your web service talk to bitcoind and get it to import the addresses generated by vanitygen?
175 2011-11-17 13:09:05 <[eval]> (and their corresponding keys, i assume)
176 2011-11-17 13:09:19 <MaxSan> naaaw idea is request a new address w/e
177 2011-11-17 13:09:26 <MaxSan> via asp.. which talks to legacy C vanitygen
178 2011-11-17 13:09:41 <MaxSan> but only begins it once bitcoin transaction is logged in the blockchain
179 2011-11-17 13:09:51 <[eval]> oh i see!
180 2011-11-17 13:10:03 <[eval]> you're trying to sell vanitygen addresses
181 2011-11-17 13:10:40 <[eval]> https://en.bitcoin.it/wiki/API_reference_(JSON-RPC)
182 2011-11-17 13:10:56 <[eval]> i don't know how to use JSON-RCP in WCF; i haven't used .Net in years (since before WCF existed)
183 2011-11-17 13:12:09 <MaxSan> ahh cheers
184 2011-11-17 13:12:18 <MaxSan> yeh thats the idea.. just a random wee project i came up with on the fly
185 2011-11-17 13:12:31 <MaxSan> not exactly a moneymaker but a service nontheless
186 2011-11-17 13:13:49 <[eval]> check this out: https://bitcointalk.org/index.php?topic=25804.msg615923#msg615923
187 2011-11-17 13:14:27 <[eval]> i don't know enough about ECC to verify it (and don't have time to look into it at the moment; i'm at work) but you might find it interesting
188 2011-11-17 13:14:59 <MaxSan> ahh yeh i was thinking about implementing that
189 2011-11-17 13:15:29 <MaxSan> well, unreadable priv keys by me
190 2011-11-17 14:30:30 <roconnor> [eval]: that link doesn't make sense to me
191 2011-11-17 14:30:42 <roconnor> a private and public keys aren't even the same type; they cannot be added
192 2011-11-17 14:30:53 <roconnor> public keys are points on the elliptic curve
193 2011-11-17 14:31:10 <roconnor> private keys are a number modulo N
194 2011-11-17 14:31:16 <roconnor> (where N is the order of the curve)
195 2011-11-17 14:32:00 <roconnor> [eval]: maybe when he says adds, he means multiply
196 2011-11-17 14:32:34 <roconnor> even still, I don't think the system works
197 2011-11-17 14:32:47 <roconnor> granted I'm not an elliptic curve expert either
198 2011-11-17 14:33:38 <roconnor> Oh, BTCurious's later posts are more clear
199 2011-11-17 14:34:30 <[eval]> roconnor: that's where it started... i've worked with RSA before (on the down-and-dirty level), not ECC, and i haven't had time to learn how ECC works yet.
200 2011-11-17 14:34:56 <[eval]> so i can't tell whether it makes sense or not, but i figured if someone's gonna be generating vanity addresses for someone else...
201 2011-11-17 14:35:10 <[eval]> it looks almost like a version of blinding
202 2011-11-17 14:35:22 <roconnor> [eval]: post #367 looks more readable
203 2011-11-17 14:35:31 <roconnor> I mean looks more plausible
204 2011-11-17 14:36:17 <roconnor> okay I believe it now
205 2011-11-17 14:36:37 <gmaxwell> roconnor: it works fine, it's not being descrbed clearly there.
206 2011-11-17 14:36:39 <roconnor> heh, this looks like a nice service
207 2011-11-17 14:36:40 <[eval]> i'm reading through the post now
208 2011-11-17 14:36:49 <gmaxwell> roconnor: bytecoin and I have both described doing this before.
209 2011-11-17 14:37:12 <roconnor> gmaxwell: ya, I get what he is saying now
210 2011-11-17 14:38:03 <[eval]> i see how it works now
211 2011-11-17 14:38:11 <[eval]> neato
212 2011-11-17 14:39:36 <gmaxwell> It uses the same property I describe here, https://bitcointalk.org/index.php?topic=19137.msg239768#msg239768
213 2011-11-17 14:40:40 <gmaxwell> ah, here is where we discussed it for vanity address generation: https://bitcointalk.org/index.php?topic=1387.msg196529#msg196529
214 2011-11-17 14:42:16 <[eval]> have you guys read... whoever's... analysis of bitcoin's resistance to quantum computing?
215 2011-11-17 14:42:35 <[eval]> vanity addresses will be one of the first things to go once it's practical
216 2011-11-17 14:44:00 <copumpkin> huh
217 2011-11-17 14:45:30 <gmaxwell> [eval]: The amount of quantum logic required to do a vanity key search would be quite enormous.
218 2011-11-17 14:45:55 <gmaxwell> [eval]: people greatly inflate the capabilities of quantum computing.
219 2011-11-17 14:46:11 <[eval]> not vanity key search
220 2011-11-17 14:46:43 <[eval]> the analysis basically said that because we're double-hashing the public key to make an address, the address should be safe from quantum computers finding the private key
221 2011-11-17 14:47:00 <copumpkin> o.O
222 2011-11-17 14:47:06 <[eval]> but as soon as the public key is broadcast to the network, redeeming the transaction to the address, a quantum computer could find the private key
223 2011-11-17 14:47:14 <copumpkin> [eval]: how would it do that?
224 2011-11-17 14:47:19 <[eval]> so you can only use an address once (or at least redeem the address once)
225 2011-11-17 14:47:21 <gmaxwell> Until you spend.
226 2011-11-17 14:47:25 <[eval]> right
227 2011-11-17 14:47:35 <gmaxwell> Yes, I posted about that too. Were you talking about my post?
228 2011-11-17 14:47:40 <gmaxwell> :)
229 2011-11-17 14:47:42 <[eval]> i guess so!
230 2011-11-17 14:47:43 <[eval]> :D
231 2011-11-17 14:47:45 <gmaxwell> But I see what you're saying.
232 2011-11-17 14:47:52 <[eval]> my first thought was "vanity addresses go bye-bye"
233 2011-11-17 14:47:56 <[eval]> single-use addresses only
234 2011-11-17 14:48:08 <gmaxwell> Yea, you wouldn't want persistant addresses if there were attackers with QCs.
235 2011-11-17 14:48:30 <copumpkin> link to this analysis?
236 2011-11-17 14:48:54 <[eval]> i get confused as to who posts what... i remember content more than attribution, unfortunately... apologies to you, gmaxwell, for not remembering
237 2011-11-17 14:48:55 <gmaxwell> copumpkin: there isn't much analysis required.
238 2011-11-17 14:49:15 <gmaxwell> copumpkin: it's just the observation that no ecc public key is disclosed until you spend from an address.
239 2011-11-17 14:49:57 <gmaxwell> copumpkin: so bitcoin has some resistance to ecc compromises, because if you use addresses just once an attacker would have to compromise your key in the time between when you spend from it and when that spend is mined.
240 2011-11-17 14:50:12 <copumpkin> ah, okay
241 2011-11-17 14:50:18 <copumpkin> you could thus receive multiple times with an address
242 2011-11-17 14:50:22 <copumpkin> as long as you don't spend multiple times
243 2011-11-17 14:50:27 <gmaxwell> Yes.
244 2011-11-17 14:50:52 <etotheipi_> this is what might also save BTC if no changes are made to it before quantum computing becomes reality
245 2011-11-17 14:51:12 <etotheipi_> quantum computing can compute your private key from your public key... but not if it doesn't have you public key
246 2011-11-17 14:51:13 <copumpkin> I doubt that's happening any time soon, anyway :)
247 2011-11-17 14:51:29 <copumpkin> regular computing can compute your private key from your public key, too :D
248 2011-11-17 14:51:41 <etotheipi_> copumpkin, not before the sun runs out of energy
249 2011-11-17 14:51:42 <gmaxwell> yea, to break ecc you don't just need a real quantum computer, you need a rather large one.
250 2011-11-17 14:51:45 <copumpkin> anyway, we should switch to lamport signatures
251 2011-11-17 14:52:03 <gmaxwell> copumpkin: I proposed that.. but the keys are really big. :(
252 2011-11-17 14:52:04 <copumpkin> clearly the easy answer ;)
253 2011-11-17 14:52:49 <copumpkin> yeah
254 2011-11-17 14:52:53 <gmaxwell> though with pruning and the fact that you use a hash of the key... it's not completely terrible.
255 2011-11-17 14:52:59 <gmaxwell> but not good either.
256 2011-11-17 14:53:01 <etotheipi_> interesting... I'm not actually familiar with Lamport sigs... I guess I'll be reading about it tonight :)
257 2011-11-17 14:53:03 <gmaxwell> <3 lamport signatures.
258 2011-11-17 14:53:05 <copumpkin> clearly the answer there is to run it through lzma ;)
259 2011-11-17 14:53:48 <gmaxwell> They're the only kind of signature that you can explain to a moron and have them understand it and trust it for the correct reasons. They're also strong against quantum computers.
260 2011-11-17 14:54:03 <Eliel> hmm? I thought I read somewhere that ECC cryptography is not vulnerable to quantum computing
261 2011-11-17 14:54:24 <gmaxwell> Eliel: dunno where you read that, but it's not true.
262 2011-11-17 14:54:30 <etotheipi_> *most* asymmetric encryptions are threatened by QC
263 2011-11-17 14:54:46 <etotheipi_> unfortunately, it's all the ones that are in user everywhere in the world right now
264 2011-11-17 14:55:06 <gmaxwell> Though we use fairly large ecc... and you're really need an insanely large quantum computer to attack it, there may turn out to be physical limits that make such machines impossible.
265 2011-11-17 14:55:26 <etotheipi_> yeah, we'll have *plenty* of time to prepare...
266 2011-11-17 14:55:46 <etotheipi_> if anyone's interested, I gave an hour-long seminar on QCs... but it's very math-y
267 2011-11-17 14:57:42 <etotheipi_> actually, it's more like 2 hours:  http://dl.dropbox.com/u/1139081/QC_Lecture_Slides.pdf
268 2011-11-17 14:58:13 <etotheipi_> either way, if anyone wanted to learn about QCs, that presentation gives you all the intro math, and then starts explaining some of the basic QC algorithms
269 2011-11-17 14:58:52 <nanotube> what was that about ops and stuff?
270 2011-11-17 15:23:04 <cocktopus> nanotube, SomeoneWeird wants access to #bitcoin-ops
271 2011-11-17 15:23:43 <Graet> mm i think he was offered and i wanted :P
272 2011-11-17 15:24:19 <cocktopus> oh yeah (it was last night nad i forgot)
273 2011-11-17 15:26:35 <Graet> :)
274 2011-11-17 16:13:50 <sipa> gavinandresen: regarding rc6... you mean restoring the wallet backup while bitcoin is running?
275 2011-11-17 16:14:17 <gavinandresen> sipa:  no.  Use case is:
276 2011-11-17 16:14:25 <gavinandresen> User decides to encrypt their wallet.  So they do.
277 2011-11-17 16:14:48 <gavinandresen> Bitcoin encrypts and then shuts down.  They then backup the encrypted wallet (good time to do that, right?).
278 2011-11-17 16:15:07 <gavinandresen> ... and are then extremely unhappy later when they restore from that backup.
279 2011-11-17 16:16:09 <sipa> gavinandresen: haven't tested it, but what's the problem with that?
280 2011-11-17 16:16:35 <gavinandresen> The keypool isn't regenerated until you request a new key from the encrypted wallet.
281 2011-11-17 16:16:48 <sipa> ow, i believed it was
282 2011-11-17 16:18:50 <gavinandresen> There is still an issue with upgrading previously-encrypted wallets, though-- that is done at startup, before we have a wallet passphrase
283 2011-11-17 16:19:03 <gavinandresen> ... and without a wallet passphrase we can't generate new keypool keys
284 2011-11-17 16:20:54 <luke-jr> gavinandresen: won't your solution mean the NEW keys get left UNENCRYPTED on disk?
285 2011-11-17 16:21:46 <luke-jr> perhaps the best solution is to require, or at least prompt for, opening the wallet after the restart
286 2011-11-17 16:21:52 <gavinandresen> luke-jr: No; the steps are:  Invalidate old keypool in old wallet.  Write new keyppol unencypted to old wallet.  Rewrite entire wallet, encrypted.
287 2011-11-17 16:22:12 <luke-jr> gavinandresen: that puts new keypool on disk unencrypted
288 2011-11-17 16:22:41 <gavinandresen> sorry, mis-typed--  write new keypool ENCRYPTED to old wallet
289 2011-11-17 16:25:04 <sipa> when doing rewriting on upgrade, you'd only want the flush the keys from the keypool that were available before the wallet for encrypted
290 2011-11-17 16:25:18 <sipa> and if there are no such keys left, the harm is already done anyway
291 2011-11-17 16:25:27 <sipa> s/for/got/
292 2011-11-17 16:25:37 <gmaxwell> Can you really tell?
293 2011-11-17 16:25:39 <sipa> unfortunately, you don't know that
294 2011-11-17 16:25:56 <gmaxwell> yea.. it's not like flushing the keypool is a costly operation in any case.
295 2011-11-17 16:29:44 <sipa> and in the average case, you'd still want most of the keypool to be flushed
296 2011-11-17 16:29:48 <luke-jr> gavinandresen: oh ok, then just have the "omg your wallet is unsafe" step require the passphrase, and just quit unchanged until they give it? :p
297 2011-11-17 16:30:08 <sipa> so i guess the only real solution is indeed to ask for opening the wallet upon upgrade
298 2011-11-17 16:30:19 <sipa> but what to do in bitcoind? just not rewrite?
299 2011-11-17 16:30:28 <luke-jr> ew
300 2011-11-17 16:30:43 <sipa> you cannot "ask" for a passphrase there
301 2011-11-17 16:30:52 <luke-jr> how about not fork, and do stdio?
302 2011-11-17 16:31:02 <luke-jr> ie, until it's fixed, bitcoind becomes a CLI app
303 2011-11-17 16:31:11 <sipa> and have all automated scripts that depend on it fail?
304 2011-11-17 16:31:15 <luke-jr> yes
305 2011-11-17 16:31:23 <luke-jr> that's already the case
306 2011-11-17 16:31:30 <luke-jr> since it would otherwise just quit after upgrading
307 2011-11-17 16:31:38 <sipa> what about making backupwallet require opening the wallet
308 2011-11-17 16:31:40 <luke-jr> actually
309 2011-11-17 16:31:45 <sipa> and do the rewriting there
310 2011-11-17 16:31:52 <luke-jr> bitcoind users aren't likely to do the "backup immediately after rewrite" thing anyway
311 2011-11-17 16:31:57 <sipa> indeed
312 2011-11-17 16:32:09 <luke-jr> backupwallet idea is good too
313 2011-11-17 16:32:30 <luke-jr> if it hasn't got a valid keypool, refuse to backup until it's opened?
314 2011-11-17 16:32:37 <sipa> indeed
315 2011-11-17 16:32:41 <luke-jr> actually
316 2011-11-17 16:32:49 <luke-jr> breakign automated backups is the LAST thing we want to do IMO
317 2011-11-17 16:33:21 <sipa> there are three concerns: breaking automated backups, breaking functioning wallet, breaking security
318 2011-11-17 16:33:44 <sipa> you will need to break one
319 2011-11-17 16:34:32 <gavinandresen> If people read the release notes, and backup AFTER requesting their first key after upgrading a previously-encrypted wallet, then they'll be OK.
320 2011-11-17 16:34:57 <sipa> so for bitcoind, you'd go for just flushing without refilling?
321 2011-11-17 16:35:24 <sipa> or only rewrite at the point where the first new key is requested?
322 2011-11-17 16:37:04 <gavinandresen> rewriting means a shutdown, so rewriting at first key request would be a nightmare to test
323 2011-11-17 16:37:34 <sipa> do we really need to shutdown bitcoind?
324 2011-11-17 16:38:05 <sipa> can't you do something like make the shutdown function call init/main/... again?
325 2011-11-17 16:38:08 <gavinandresen> More testing is welcome, but in my testing if a shutdown wasn't done immediately after rewriting then old keys could end up unencrypted in the new wallet.dat
326 2011-11-17 16:39:05 <gavinandresen> sipa:  be my guest.  The shutdown code is a tangled nightmare, I doubt a soft restart can be done without a very large refactoring
327 2011-11-17 16:40:12 <jgarzik> indeed... that would be a pain
328 2011-11-17 16:40:17 <luke-jr> sanest approach to soft restart right now is exec()
329 2011-11-17 16:40:23 <luke-jr> which seems to have other psosible issues
330 2011-11-17 16:52:49 <sipa> gavinandresen: let's try :)
331 2011-11-17 16:53:52 <gavinandresen> let's try what?  restarting?  go for it, I'm busy debugging ugly plan B....
332 2011-11-17 17:06:30 <JFK911> rain rain :(
333 2011-11-17 17:10:41 <sipa> there is a CNetCleanup that is not used anywhere? :S
334 2011-11-17 17:11:17 <sipa> oh, just a single instance that is destroyed at shutdown
335 2011-11-17 17:26:24 <gavinandresen> sipa, I'm about to throw my keyboard out the window-- a simple little routine to recreate the keypool after encryption is locking up deep in the bowels of CWallet/CEncryptedKeyStore/CKeystore, and I can't figure out why. Could you give it a try?
336 2011-11-17 17:35:27 <sipa> gavinandresen: i hope to have restart working soon, and then i'm off for a bit
337 2011-11-17 18:02:13 <gavinandresen> ... figured it out...
338 2011-11-17 18:22:12 <CIA-89> bitcoin: Gavin Andresen master * r1c15f88 / (src/db.cpp src/db.h src/init.cpp src/wallet.cpp): Only remove database log files on shutdown after wallet encryption/rewrite - http://git.io/2p5rfA
339 2011-11-17 18:22:13 <CIA-89> bitcoin: Gavin Andresen master * r37971fc / (src/wallet.cpp src/wallet.h): Create new keypool for newly encrypted wallets. - http://git.io/WmNeRg
340 2011-11-17 18:36:11 <CIA-89> bitcoin: Gavin Andresen 0.4.x * r90de05e88ecc bitcoind-stable/src/ (wallet.cpp wallet.h): Create new keypool for newly encrypted wallets.
341 2011-11-17 18:36:13 <CIA-89> bitcoin: Gavin Andresen 0.4.x * rc4a3bf9e552f bitcoind-stable/src/ (db.cpp db.h init.cpp wallet.cpp): Only remove database log files on shutdown after wallet encryption/rewrite
342 2011-11-17 23:30:23 <CIA-89> bitcoin: Kamil Domanski * rbb6295..400cf9 gentoo/ (7 files in 3 dirs): (6 commits)