1 2013-02-23 00:01:09 <dario1> lol, RealSolid is the maker of all evil in bitcoinland
2 2013-02-23 00:01:24 <amiller> it's tough to explain why it doesn't work because you can easily propose topologies that break it
3 2013-02-23 00:01:30 <amiller> but then they can say that those topologies aren't likely to happen
4 2013-02-23 00:01:44 <amiller> but to justify that requires describing what process a user should use to choose their trusted servers list
5 2013-02-23 00:02:05 <gmaxwell> Luke-Jr: Realsolid's decenteralized consensus was convergent... I'm not too confident that what they're proposing is...
6 2013-02-23 00:02:27 <Luke-Jr> gmaxwell: even if they keep control of the trusted nodes?
7 2013-02-23 00:02:35 <gmaxwell> amiller: right, I think they need to tell me what makes those topoligies unlikely.
8 2013-02-23 00:02:41 <amiller> in bitcoin there's a really simple decision - the winning chain is the valid one with the most work on it, but in ripple.com there's no guidance offered
9 2013-02-23 00:03:20 <amiller> consensus depends on a very subjective kind of judgment about whether peers are not only honest, but also trusted to behave in some "safe" way, which is then a circular definition
10 2013-02-23 00:03:35 <gmaxwell> amiller: it's not like the ones I described are the only ones that break itâ I just picked the simplest cases. I'm actually not smart enough to express what the requirement is, beyond "full mesh" being sufficient.
11 2013-02-23 00:05:10 <gmaxwell> amiller: and since safe also includes being not tricked by nodes they trust... or the not tricked by nodes N->â hops away.
12 2013-02-23 00:05:30 <HM> ripplepay.com still exists
13 2013-02-23 00:06:09 <gmaxwell> It seems clear to me thatâ assuming sutible implementation that with a stable and unchanging full mesh that this system would converge. But that could hardly be described as decenteralized or even strongly shut-down resistant.
14 2013-02-23 00:06:55 <gmaxwell> (though I think it would be a lot more fragile than bitcoin wrt implementation bugs, â it could easily flap forever, even with a full mesh, without a lot of care)
15 2013-02-23 00:07:29 ThomasV has joined
16 2013-02-23 00:08:38 test001 has joined
17 2013-02-23 00:08:56 <HM> A distributed ledger that details credit limits between friends
18 2013-02-23 00:09:05 <test001> hello all
19 2013-02-23 00:09:08 <HM> And allows debt to propagate
20 2013-02-23 00:09:15 <HM> makes sense ot me
21 2013-02-23 00:10:02 <test001> i need open de blk001.dat, what do you need? (IDE BerckleyDB, ODBC, ...???) Thanks
22 2013-02-23 00:10:26 <sipa> test001: it's just a concatenation of serialized blocks in network format
23 2013-02-23 00:10:40 <sipa> not a database
24 2013-02-23 00:10:49 <HM> ripple sounds like it turns friends in to scrooges and loan sharks
25 2013-02-23 00:11:07 <test001> where is ' the DB' ???
26 2013-02-23 00:11:13 <sipa> test001: blkindex.dat is a datbase
27 2013-02-23 00:11:16 <sipa> BDB
28 2013-02-23 00:11:40 <test001> ok, i need open de blkindex.dat, what do you need? (IDE BerckleyDB, ODBC, ...???) Thanks
29 2013-02-23 00:11:49 <sipa> berkeleydb
30 2013-02-23 00:12:06 bitafterbit has quit (Read error: Connection reset by peer)
31 2013-02-23 00:12:12 <sipa> also, since version 0.8, there is no more blkindex.dat, but two seperate LevelDB databases
32 2013-02-23 00:13:03 <test001> a IDE for berkeleyDB??? Configuration for Eclipse????
33 2013-02-23 00:13:22 <sipa> can you please use more question marks? it certainly helps
34 2013-02-23 00:13:33 * gmaxwell snickers (??)
35 2013-02-23 00:13:55 <sipa> and i have no idea about an IDE that would have BDB support built in
36 2013-02-23 00:14:07 <sipa> i doubt that exists, it's just a C library
37 2013-02-23 00:14:16 <CodeShark> you just have to set the linker flags
38 2013-02-23 00:14:20 <CodeShark> regardless of IDE
39 2013-02-23 00:14:38 TradeFortress has joined
40 2013-02-23 00:14:59 <HM> sipa: Oracle have something I'm sure
41 2013-02-23 00:15:01 <CodeShark> are you talking about tools for editing .dat files?
42 2013-02-23 00:15:22 <Luke-Jr> trying to decide if I should name split-up README files as *-README to README-*, anyone have any thoughts on this?
43 2013-02-23 00:15:32 <sipa> Luke-Jr: the second
44 2013-02-23 00:15:54 <test001> ok, thanks for all, bye
45 2013-02-23 00:16:00 <K1773R> rofl
46 2013-02-23 00:19:30 TradeFortress has quit (Ping timeout: 264 seconds)
47 2013-02-23 00:19:59 HiroWhite has quit (Quit: Leaving)
48 2013-02-23 00:20:14 HiroWhite has joined
49 2013-02-23 00:22:16 HiroWhite has quit (Client Quit)
50 2013-02-23 00:22:30 HiroWhite has joined
51 2013-02-23 00:26:18 denisx has quit (Quit: denisx)
52 2013-02-23 00:27:22 ppr is now known as peper
53 2013-02-23 00:30:39 agricocb has joined
54 2013-02-23 00:33:00 <Eliel_> gmaxwell: that one example that you provided in that thread about a consensus network that would allow you to be defrauded. There's one interesting factor in it, without which your node wouldn't be so easily fooled. Your node in it has less connections to other nodes than most other nodes have.
55 2013-02-23 00:34:11 <gmaxwell> Eliel_: I just drew the simplest network that allowed me to make the pointâ it doesn't really matter much, your own node in ripple is never a validator.
56 2013-02-23 00:34:34 <gmaxwell> he could be connected to every node on the graph and still get defrauded (though at least he'd know he was getting defrauded I guess)
57 2013-02-23 00:35:02 ThomasV has quit (Ping timeout: 252 seconds)
58 2013-02-23 00:37:42 <CodeShark> ripple isn't about zero-trust. people would have to choose to trust validators
59 2013-02-23 00:39:28 test001 has quit (Quit: Page closed)
60 2013-02-23 00:39:35 ThomasV has joined
61 2013-02-23 00:41:16 <CodeShark> in practice, though, people would most likely have their validators chosen for them by others
62 2013-02-23 00:41:32 <Eliel_> gmaxwell: I expect some kind of mechanisms to audit the trust network that forms from the UNLs of separate nodes probably needs to be built to evaluate the interconnectivity. With good enough interconnectivity, it'll be obvious when the validators are in disagreement and if that happens, none of the conflicting transactions should end up validated.
63 2013-02-23 00:41:57 <gmaxwell> CodeShark: you're failing to understand what I'm saying.
64 2013-02-23 00:42:27 <CodeShark> I'm sure I am :)
65 2013-02-23 00:43:51 <gmaxwell> I'm not complaining that you have to trust validators. (well, thats a negative for sureâ but at least its a simply understood one). I'm saying that even if the validators you trust are honest, you can still get ripped off (or the system can break entirely). And I don't have a way of expressingâ nor has anyone offered oneâ a necessary and sufficient set of requirements on the trust relationship for the system to be secure and ...
66 2013-02-23 00:43:57 <gmaxwell> ... available.
67 2013-02-23 00:45:00 <CodeShark> are you talking about defaults?
68 2013-02-23 00:45:06 <gmaxwell> No.
69 2013-02-23 00:45:16 <CodeShark> or doublespends?
70 2013-02-23 00:45:57 <CodeShark> default in the credit sense, not in the comp sci sense :)
71 2013-02-23 00:46:08 <gmaxwell> I'm not talking about the credit crud at all. I'm talking about the consensus system. (doublespends, as well as outright balance theft: it doesn't maintain a cryptographically secured transaction history, so a consensus failure allows balances to be rewritten)
72 2013-02-23 00:47:00 ThomasV has quit (Ping timeout: 252 seconds)
73 2013-02-23 00:47:03 <CodeShark> so you're saying that the system is nonconvergent
74 2013-02-23 00:47:23 <gmaxwell> Its non-convergent for some trust topologies. Correct.
75 2013-02-23 00:47:50 <gmaxwell> But not just non-convergent, it can be maliciously rewritten or double spent for some trust topologies under some conditions.
76 2013-02-23 00:48:17 <gmaxwell> In Bitcoin we obtain a single honest consensus in small finite time with very high probablity iff nodes can communicate (can't hide the best chain from them), and if no conspiring dishonest parties control a (near) majority hashpower. Thats pretty much complete and correct.
77 2013-02-23 00:48:25 <gmaxwell> I can't make a similar statement for the ripple consensus model.
78 2013-02-23 00:48:41 alexwaters has quit (Quit: Leaving.)
79 2013-02-23 00:48:47 <CodeShark> the question is whether such topologies would tend to arise in practice
80 2013-02-23 00:48:55 <gmaxwell> I believe I have a sufficient criteria: a completely regular full meshâ a single set of trusted validators who all trust each other with no missing edges. Thats sufficient, I think. At least abstractly.
81 2013-02-23 00:49:18 <gmaxwell> CodeShark: no, thats a question but perhaps not the most interesting one. Can attackers create one intentionally?
82 2013-02-23 00:49:20 <sipa> ... i hope it's sufficient but not necessary :)
83 2013-02-23 00:49:43 <gmaxwell> sipa: right, I would hope, but I don't know what the necessary criteria is.
84 2013-02-23 00:49:52 <CodeShark> arise in practice includes intentional as well as unintentional, gmaxwell
85 2013-02-23 00:50:08 <gmaxwell> CodeShark: in general, it sounds like small world networksâ which is what you get when you use _actual_ social graphs and trust, will not be stable in this model.
86 2013-02-23 00:51:20 <gmaxwell> CodeShark: okay, I was thrown by arise. A response to one of my graphs was "don't do that" but I don't see how it would be prevented, esp since there has been no mention of nodes disclosing the nodes they trust... so it sounds like it would be invisible.
87 2013-02-23 00:52:43 <amiller> i wonder if ripple.com will be susceptible to altcoins
88 2013-02-23 00:53:10 <amiller> if someone were upset about the xrp allocation then they could create a fork or something
89 2013-02-23 00:53:16 <CodeShark> I've had that same concern, amiller - since the XRPs are centrally distributed, there's nothing to stop someone else from doing that
90 2013-02-23 00:53:30 <gmaxwell> amiller: hm. interesting point, you could just reassign them.
91 2013-02-23 00:53:46 <amiller> then you could have confusing trust lists where some servers support both or prefer one or the other
92 2013-02-23 00:54:55 <amiller> there are lots of ways of being dishonest that aren't detectable like basically pretending that one of the servers is dropping packets, even if it isn't
93 2013-02-23 00:56:25 <CodeShark> that's actually more a concern of mine than the issue of fraudulent validators
94 2013-02-23 00:58:36 andytoshi has joined
95 2013-02-23 01:07:22 <CodeShark> gmaxwell: as for which nodes to trust, the vast majority of users are not going to be carefully researching and choosing their validators - they would most likely just use whatever default list is provided
96 2013-02-23 01:07:31 <CodeShark> (default here NOT used in the credit sense)
97 2013-02-23 01:07:35 rdymac has joined
98 2013-02-23 01:08:23 <CodeShark> so then it becomes an issue of who administers such lists, how carefully they are audited, etc...
99 2013-02-23 01:09:41 <CodeShark> with bitcoin, the only serious risk is that of cliques forming
100 2013-02-23 01:09:49 <CodeShark> as far as default lists being provided
101 2013-02-23 01:10:18 <CodeShark> but it only takes one node that links two cliques to form one graph
102 2013-02-23 01:12:53 J6So4W7Yb has quit (Ping timeout: 246 seconds)
103 2013-02-23 01:31:19 rdymac has quit (Read error: Connection reset by peer)
104 2013-02-23 01:31:36 rdymac has joined
105 2013-02-23 01:35:48 ralphtheninja has joined
106 2013-02-23 01:36:20 RazielZ has quit (Ping timeout: 246 seconds)
107 2013-02-23 01:37:55 <petertodd> is there any way to get bitcoin to rebroadcast a found block? I'm testing something on testnet and found a block with my node disconnected from the rest of the network
108 2013-02-23 01:39:08 Mrcheesenips has quit (Read error: Connection reset by peer)
109 2013-02-23 01:39:30 Mrcheesenips has joined
110 2013-02-23 01:39:30 Mrcheesenips has quit (Changing host)
111 2013-02-23 01:39:30 Mrcheesenips has joined
112 2013-02-23 01:40:07 <CodeShark> hah
113 2013-02-23 01:40:23 rbecker is now known as RBecker
114 2013-02-23 01:41:09 <CodeShark> I know of several ways to get that to happen - but none that will work quickly enough for your block to matter
115 2013-02-23 01:41:34 <petertodd> never mind, I found another block so I'm good
116 2013-02-23 01:41:43 <petertodd> (5 block re-org there)
117 2013-02-23 01:45:40 <CodeShark> bitcoind doesn't queue up found blocks to send when a connection is made?
118 2013-02-23 01:49:09 rdymac has quit (Ping timeout: 252 seconds)
119 2013-02-23 01:50:11 rdymac has joined
120 2013-02-23 01:53:18 <petertodd> doesn't seem too
121 2013-02-23 01:53:50 <petertodd> I restarted my node, and the block never got out until the next was found
122 2013-02-23 01:57:41 <CodeShark> hmm, sounds like a bug to me
123 2013-02-23 01:58:02 <CodeShark> albeit not one that probably arises very often in practice
124 2013-02-23 01:58:40 <petertodd> Ha, yeah not at all.
125 2013-02-23 02:03:42 TradeFortress has joined
126 2013-02-23 02:04:35 ralphtheninja has quit (Ping timeout: 248 seconds)
127 2013-02-23 02:09:54 rdymac has quit (Ping timeout: 264 seconds)
128 2013-02-23 02:10:00 RBecker is now known as rbecker
129 2013-02-23 02:10:21 <gmaxwell> CodeShark: most of my thinking has been around the topology of validators, â users are at the whim of their validators. And whatever opinion I have on such a model, at least I think I and anyone else would understand the basic consequences of it.
130 2013-02-23 02:21:45 LainZ has joined
131 2013-02-23 02:25:22 Zarutian has joined
132 2013-02-23 02:30:28 sgornick has joined
133 2013-02-23 02:34:26 swappermall has quit (Ping timeout: 246 seconds)
134 2013-02-23 02:42:30 swappermall has joined
135 2013-02-23 02:42:52 swappermall is now known as Guest37021
136 2013-02-23 02:47:09 coolsa has joined
137 2013-02-23 02:47:35 PhantomSpark has joined
138 2013-02-23 02:49:37 coolsa_ has quit (Ping timeout: 240 seconds)
139 2013-02-23 02:50:16 BitcoinG has quit (Quit: BitcoinG)
140 2013-02-23 02:51:24 LargoG has quit (Remote host closed the connection)
141 2013-02-23 02:51:35 <gmaxwell> Just a heads up: People are passing out wallet stealers and trojans in bitcoin channels again. I advise against following any links given here unless they are from people you know and trust. If you don't know and trust anyone here, don't follow any links. No matter who gives you a link here, don't run any executables.
142 2013-02-23 02:53:53 PhantomSpark has joined
143 2013-02-23 02:54:51 coolsa has quit (Ping timeout: 256 seconds)
144 2013-02-23 02:56:54 PhantomSpark has quit (2!~kvirc@pool-71-251-16-105.nycmny.fios.verizon.net|Ping timeout: 276 seconds)
145 2013-02-23 02:57:53 J6So4W7Yb has joined
146 2013-02-23 02:59:02 topace has quit (Read error: No route to host)
147 2013-02-23 03:01:58 none has joined
148 2013-02-23 03:15:48 topace has joined
149 2013-02-23 03:16:11 topace is now known as Guest32273
150 2013-02-23 03:21:54 ProfMac has quit (Ping timeout: 245 seconds)
151 2013-02-23 03:21:58 coolsa has joined
152 2013-02-23 03:31:02 rbecker is now known as RBecker
153 2013-02-23 03:38:18 none has quit (Quit: none)
154 2013-02-23 03:39:09 moarrr has quit (Ping timeout: 276 seconds)
155 2013-02-23 03:39:19 * muhoo reads links in w3m :-)
156 2013-02-23 03:42:54 <gmaxwell> I need to figure out how to get valgrind working in the selinux sandbox... so I can have kvm -> selinuxsandbox -> valgrind -> links for browsing.
157 2013-02-23 03:43:40 none has joined
158 2013-02-23 03:44:12 none is now known as BitcoinG
159 2013-02-23 03:48:24 <andytoshi> i can't get a second tty because pamd is missing some selinux rule, so i run without selinux
160 2013-02-23 03:48:31 <phantomcircuit> gmaxwell, you wouldn't happen to know the average number of transactions/block would you
161 2013-02-23 03:48:34 <andytoshi> my first tty auto-logins as root, so i can't use that one..
162 2013-02-23 03:48:39 <andytoshi> security!
163 2013-02-23 03:48:46 <phantomcircuit> i tried querying blockexplorer for it but it never returns heh
164 2013-02-23 03:49:11 <gmaxwell> phantomcircuit: see the shell oneline in scrollback.
165 2013-02-23 03:49:19 <gmaxwell> but remove the getrawtransaction from it.
166 2013-02-23 03:49:29 <phantomcircuit> wat
167 2013-02-23 03:49:33 <gmaxwell> and subtract one..
168 2013-02-23 03:49:54 <gmaxwell> lazy kids these days
169 2013-02-23 03:49:54 <gmaxwell> 07:23 < gmaxwell> (for i in {222552..222542} ; do ./bitcoind getblock `./bitcoind getblockhash $i` | grep ' "' | cut -d'"' -f2 ; done) | xargs -n1 -iblah ./bitcoind getrawtransaction blah | awk '{aa+=length($0)/2} END {print aa/NR}'
170 2013-02-23 03:50:21 fiesh_ has joined
171 2013-02-23 03:50:23 <gmaxwell> 07:23 < gmaxwell> (for i in {222552..222542} ; do ./bitcoind getblock `./bitcoind getblockhash $i` | grep ' "' | cut -d'"' -f2 ; done) | wc -l and then subtract 10. change the range as you wish.
172 2013-02-23 03:52:05 fiesh has quit (Ping timeout: 255 seconds)
173 2013-02-23 03:53:28 fiesh_ is now known as fiesh
174 2013-02-23 03:59:18 PhantomSpark has quit (Ping timeout: 276 seconds)
175 2013-02-23 04:05:31 FredEE has quit (Quit: FredEE)
176 2013-02-23 04:08:23 BurtyBB has joined
177 2013-02-23 04:11:28 BTCTrader has joined
178 2013-02-23 04:11:28 BTCTrader has quit (Changing host)
179 2013-02-23 04:11:28 BTCTrader has joined
180 2013-02-23 04:11:52 BurtyB has quit (Ping timeout: 245 seconds)
181 2013-02-23 04:14:52 Maged has joined
182 2013-02-23 04:15:07 BTCTrader has quit (Remote host closed the connection)
183 2013-02-23 04:21:46 ForceMajeure has joined
184 2013-02-23 04:38:11 freakazoid has joined
185 2013-02-23 04:44:26 swappermall_ has quit (Read error: Connection reset by peer)
186 2013-02-23 04:45:03 swappermall_ has joined
187 2013-02-23 04:45:27 swappermall_ is now known as Guest46294
188 2013-02-23 04:47:26 PhantomSpark has quit (3!~kvirc@pool-71-251-16-105.nycmny.fios.verizon.net|Read error: Connection reset by peer)
189 2013-02-23 04:54:42 Ramokk has quit (Read error: Connection reset by peer)
190 2013-02-23 04:56:09 Guest37021 has quit (Ping timeout: 252 seconds)
191 2013-02-23 05:00:15 TheSeven has quit (Disconnected by services)
192 2013-02-23 05:00:21 [7] has joined
193 2013-02-23 05:02:11 freakazoid has quit (Read error: Connection timed out)
194 2013-02-23 05:02:46 freakazoid has joined
195 2013-02-23 05:07:35 swappermall has joined
196 2013-02-23 05:07:51 PhantomSpark has joined
197 2013-02-23 05:07:57 swappermall is now known as Guest74266
198 2013-02-23 05:10:46 Ramokk has joined
199 2013-02-23 05:11:00 Guest46294 has quit (Remote host closed the connection)
200 2013-02-23 05:14:49 <Luke-Jr> http://bitcointroll.org/?topic=78192.msg1552015#msg1552015 <-- BFGMiner 3.0 alpha1, please help test
201 2013-02-23 05:19:48 Guest74266 has quit (Remote host closed the connection)
202 2013-02-23 05:23:18 axhlf has joined
203 2013-02-23 05:38:18 brwyatt is now known as Away!~brwyatt@brwyatt.net|brwyatt
204 2013-02-23 05:40:06 toffoo has joined
205 2013-02-23 05:45:17 <jgarzik> Luke-Jr: how do you support BFL? Did they publish a chip spec?
206 2013-02-23 05:45:37 <Luke-Jr> jgarzik: yes, they did
207 2013-02-23 05:48:09 <Luke-Jr> https://forums.butterflylabs.com/attachments/announcements/488d1360098512-bitforce-sc-communication-protocol-draft-revision-2-butterlylabs_bitforce_rev2.1.0_preliminary.pdf
208 2013-02-23 05:49:05 <Luke-Jr> jgarzik: I also work with them for clarifications; BFGMiner will be shipping on the MiniRig SC's embedded Nexus7 phone
209 2013-02-23 05:49:48 PhantomSpark has quit (Ping timeout: 276 seconds)
210 2013-02-23 05:52:17 J6So4W7Yb has quit (Ping timeout: 256 seconds)
211 2013-02-23 05:58:04 random_cat has quit (Remote host closed the connection)
212 2013-02-23 06:05:17 PhantomSpark has joined
213 2013-02-23 06:11:16 abrkn has joined
214 2013-02-23 06:12:02 Zarutian has quit (Quit: Zarutian)
215 2013-02-23 06:12:52 TradeFortress_ has joined
216 2013-02-23 06:15:47 <gmaxwell> OH MY GOD.
217 2013-02-23 06:15:57 <gmaxwell> BEST REASON FOR TESTNET EVER!!!!
218 2013-02-23 06:16:13 <gmaxwell> everyone spin up a testnet3 node on your *NIX desktops stat!
219 2013-02-23 06:16:39 TradeFortress has quit (Ping timeout: 256 seconds)
220 2013-02-23 06:16:42 <gmaxwell> This will not let you down! load up testnet now if you're not already running a node.
221 2013-02-23 06:17:58 <Luke-Jr> gmaxwell: does it have to be Qt?
222 2013-02-23 06:18:04 <gmaxwell> no.. cli.
223 2013-02-23 06:18:07 <gmaxwell> bitcoind is good.
224 2013-02-23 06:18:20 <Luke-Jr> TNIAB work? <.<
225 2013-02-23 06:18:31 <gmaxwell> no. needs to be on the network.
226 2013-02-23 06:18:41 <Luke-Jr> my TNIAB seems to have 9 connections, so I think it is
227 2013-02-23 06:18:42 <Luke-Jr> <.<
228 2013-02-23 06:19:30 <gmaxwell> "blocks" : 54528,
229 2013-02-23 06:19:31 <gmaxwell> ?
230 2013-02-23 06:19:32 none has joined
231 2013-02-23 06:19:41 none has quit (Remote host closed the connection)
232 2013-02-23 06:20:02 <Luke-Jr> yes
233 2013-02-23 06:20:55 <gmaxwell> plug in speakers
234 2013-02-23 06:20:56 <gmaxwell> bitcoind getrawtransaction 73e64e38faea386c88a578fd1919bcdba3d0b3af7b6302bf6ee1b423dc4e4333 | xxd -r -p | play -tul -
235 2013-02-23 06:21:45 <andytoshi> height 7885..
236 2013-02-23 06:22:00 <andytoshi> 14000..
237 2013-02-23 06:22:17 BitcoinG has quit (Ping timeout: 260 seconds)
238 2013-02-23 06:22:41 <Luke-Jr> gmaxwell: awesome
239 2013-02-23 06:22:44 <andytoshi> 2800..
240 2013-02-23 06:22:58 <petertodd> Luke-Jr: I'm working on the other half as we speak.
241 2013-02-23 06:23:13 <andytoshi> 35000..
242 2013-02-23 06:23:18 * Luke-Jr glares at petertodd if he makes a bitcoin-compatible audio uploader
243 2013-02-23 06:23:34 <gmaxwell> Luke-Jr: it's unrealistic for the main net, ... this is a 1mb transaction.
244 2013-02-23 06:24:17 <petertodd> I'm kinda annoyed I screwed up and was 17 bytes short for the whole block, but I also got another perfect 1MB block in a bit earlier tonight.
245 2013-02-23 06:24:32 <andytoshi> 54506..
246 2013-02-23 06:24:37 <Luke-Jr> gmaxwell: I'm tempted. Very tempted.
247 2013-02-23 06:25:04 <petertodd> Luke-Jr: So long as you put it all in one scriptPubKey, and you have > 10,000 bytes of audio, it is a prunable transation!
248 2013-02-23 06:25:19 <Luke-Jr> gmaxwell: think I could get away with it?
249 2013-02-23 06:25:36 <gmaxwell> Luke-Jr: no way. lol
250 2013-02-23 06:25:44 <Luke-Jr> <.<
251 2013-02-23 06:25:44 <andytoshi> hahaha, wonderful
252 2013-02-23 06:26:06 <Luke-Jr> what codec is this?
253 2013-02-23 06:26:16 <Luke-Jr> maybe I can make it fit in a normal txn and get some other miner to do it
254 2013-02-23 06:26:20 <andytoshi> "u-law"
255 2013-02-23 06:26:25 <Luke-Jr> wtf, uncompressed?
256 2013-02-23 06:26:37 * Luke-Jr ponders how to make it Windows-friendly
257 2013-02-23 06:26:52 <gmaxwell> using ulaw makes things a lot easier, because the samples are 1 byte.
258 2013-02-23 06:27:10 <Luke-Jr> I don't see how that makes it easier.. <.<
259 2013-02-23 06:27:13 <petertodd> Luke-Jr: I'd pay to see eligius mine that block
260 2013-02-23 06:27:22 <petertodd> (and give you my code)
261 2013-02-23 06:27:32 <Luke-Jr> petertodd: but gmaxwell says I couldn't get away with that
262 2013-02-23 06:27:36 <gmaxwell> opus or vorbis would have worked too, since libogg will capture anywhere in a stream of bytes, but .. because that would result in files small enough to put in bitcoin, uhh, I wouldn't have suggested it.
263 2013-02-23 06:27:38 <andytoshi> somebody should spend to a scripthash of it..
264 2013-02-23 06:27:51 <petertodd> andytoshi: you're terrible
265 2013-02-23 06:28:01 <Luke-Jr> lol
266 2013-02-23 06:28:03 <gmaxwell> you could never mine the spend.
267 2013-02-23 06:28:04 <petertodd> Unfortunately scripts > 10,000 bytes aren't spendable
268 2013-02-23 06:28:22 <Luke-Jr> just do a password
269 2013-02-23 06:28:37 <Luke-Jr> OP_HASH160 <hash> OP_EQUALS OP_NOP
270 2013-02-23 06:28:47 <Luke-Jr> but first let's figure out a way to make it decode on Windows
271 2013-02-23 06:28:48 <petertodd> AND PART 2!!! d85af546147ff78dfb06e9469ddfc84adc3ce00cda54db8d65b7617ff2b7661a
272 2013-02-23 06:29:09 <gmaxwell> Luke-Jr: if you have sox and cygwin you should be able to do it the same way.
273 2013-02-23 06:29:31 <Luke-Jr> most Windows users don't.
274 2013-02-23 06:29:53 <gmaxwell> (./bitcoind getrawtransaction 73e64e38faea386c88a578fd1919bcdba3d0b3af7b6302bf6ee1b423dc4e4333 ; ./bitcoind getrawtransaction d85af546147ff78dfb06e9469ddfc84adc3ce00cda54db8d65b7617ff2b7661a) | xxd -r -p | play -tul -
275 2013-02-23 06:29:57 <gmaxwell> shoudl work
276 2013-02-23 06:30:43 <Luke-Jr> gmaxwell: can we add a bitcoind -xxd option?
277 2013-02-23 06:30:45 devrandom has quit (Ping timeout: 276 seconds)
278 2013-02-23 06:31:11 <gmaxwell> Luke-Jr: make a python bitcoin rpc client?
279 2013-02-23 06:31:24 <Luke-Jr> most Windows users don't have Python either
280 2013-02-23 06:31:29 <petertodd> I should post in the forums about how the block size now needs to be exactly 1.8MiB...
281 2013-02-23 06:31:50 <gmaxwell> Luke-Jr: meh, then write a little C program.
282 2013-02-23 06:32:00 <Luke-Jr> gmaxwell: then I cna just embed the audio and it's no fun
283 2013-02-23 06:32:04 <gmaxwell> Yea my above commandline plays the whole thing with a little click.
284 2013-02-23 06:32:22 <gmaxwell> Luke-Jr: who says it only has one file in it? ... maybe we found a way to get people to use testnet. :P
285 2013-02-23 06:33:13 <petertodd> Yeah, checkout block 54512, I encrypted a super-secret file in the second transaction with a one-time-pad!
286 2013-02-23 06:33:32 <gmaxwell> petertodd: lol.
287 2013-02-23 06:33:40 <Luke-Jr> strings ~/.bitcoin/blk0001.dat | grep -i atheist
288 2013-02-23 06:33:43 <Luke-Jr> gmaxwell: ever see that? ^
289 2013-02-23 06:34:40 <gmaxwell> lol
290 2013-02-23 06:35:18 <andytoshi> in ~/.bitcoin/blocks/blk00003.dat for me
291 2013-02-23 06:36:13 <andytoshi> wow, there's a lot of stuff in that one..
292 2013-02-23 06:37:43 <Luke-Jr> including the famous "Stack is innocent"?
293 2013-02-23 06:38:24 <petertodd> I'm gonna regret this: https://bitcointalk.org/index.php?topic=146197.msg1552145#msg1552145
294 2013-02-23 06:38:37 <andytoshi> nope, but a lot of prayers and ascii bernankes
295 2013-02-23 06:38:55 <Luke-Jr> lol
296 2013-02-23 06:39:02 <Luke-Jr> hmm
297 2013-02-23 06:39:11 <Luke-Jr> I wonder if anyone here is too new to get "Stack is innocent" <.<
298 2013-02-23 06:39:18 * petertodd raises hand
299 2013-02-23 06:39:36 <andytoshi> Stack is innocent is in blk00006.dat
300 2013-02-23 06:40:04 <Luke-Jr> petertodd: http://www.imdb.com/title/tt2148561/
301 2013-02-23 06:40:18 <andytoshi> ohh, that's right
302 2013-02-23 06:40:35 <andytoshi> i got my IRC nick during that episode, was 'andyroo' before
303 2013-02-23 06:40:45 <Luke-Jr> lol
304 2013-02-23 06:40:59 <andytoshi> but somebody was suggesting we all be toshis in case people came on IRC looking for Mr. Bitcoin
305 2013-02-23 06:41:44 <petertodd> Luke-Jr: I wonder if I can compress that down to 1MiB...
306 2013-02-23 06:41:52 <Luke-Jr> petertodd: use vorbis
307 2013-02-23 06:41:59 <Luke-Jr> oggplay
308 2013-02-23 06:42:10 <Luke-Jr> hmm
309 2013-02-23 06:42:11 <petertodd> It's too bad we don't have an expert at, I dunno, codecs or something.
310 2013-02-23 06:42:11 <Luke-Jr> no oggplay
311 2013-02-23 06:42:21 * Luke-Jr keeps silent
312 2013-02-23 06:42:26 <petertodd> :P
313 2013-02-23 06:44:01 grau has joined
314 2013-02-23 06:53:32 <petertodd> cat ~/.bitcoin/testnet3/blocks/blk00000.dat | play -tul -
315 2013-02-23 06:53:33 <gmaxwell> yea, don't use vorbis. the result is too small. :P
316 2013-02-23 06:53:50 <petertodd> it's interesting how you can tell when people have been actually using testnet by the sound
317 2013-02-23 06:54:11 <petertodd> surprisingly soothing, a lot of low frequency components, probably from the zero'd fields in coinbases
318 2013-02-23 06:54:23 <petertodd> how small?
319 2013-02-23 06:54:28 <petertodd> mainnet small? :P
320 2013-02-23 06:54:32 <gmaxwell> hush
321 2013-02-23 06:54:44 <Diablo-D3> oh christ
322 2013-02-23 06:54:50 <Diablo-D3> what happens on the main net
323 2013-02-23 06:54:55 <Diablo-D3> when you get to SD
324 2013-02-23 06:55:00 <petertodd> (keeping data out of the blockchain is a serious, and I think good, argument for keeping block sizes small)
325 2013-02-23 06:55:30 <petertodd> Diablo-D3: lol, it'll be a few hours of listening though...
326 2013-02-23 06:55:58 <gmaxwell> opus would be betterâ 8kbit/sec .. get 16 minutes of speech into 1mb.
327 2013-02-23 06:56:05 <andytoshi> petertodd: i wouldn't say soothing, more like frightening
328 2013-02-23 06:56:11 <andytoshi> it sounds like it's trying to wake up
329 2013-02-23 06:56:15 ThomasV has joined
330 2013-02-23 06:56:28 <petertodd> andytoshi: depends what music you listen too; I like Autechre myself
331 2013-02-23 06:56:29 <gmaxwell> Obviously alice's restaurant would be the obvious thing to insert.
332 2013-02-23 06:59:17 <Diablo-D3> * several hours later *
333 2013-02-23 06:59:27 * Luke-Jr keeps playing Rickroll over and over
334 2013-02-23 07:00:44 <gmaxwell> Luke-Jr: lol.
335 2013-02-23 07:00:55 * gmaxwell too
336 2013-02-23 07:01:03 <andytoshi> i should setup a blocknotify to play these out loud
337 2013-02-23 07:01:04 robocoin has quit (Quit: ~_~;)
338 2013-02-23 07:01:09 <andytoshi> then if you guys ever need me, you'll know how to call
339 2013-02-23 07:01:11 <Diablo-D3> (sd starts saturating the chain, the fourth movement to beethoven's 9th symphony is heard)
340 2013-02-23 07:01:26 <gmaxwell> testnet phone
341 2013-02-23 07:01:33 <gmaxwell> world most inefficient phone system
342 2013-02-23 07:01:40 <Luke-Jr> not sure why people hate rickroll music so much
343 2013-02-23 07:01:48 <petertodd> I'm starting to like it
344 2013-02-23 07:01:50 <petertodd> catchy
345 2013-02-23 07:01:51 <gmaxwell> Luke-Jr: you'll know after you turn it off.
346 2013-02-23 07:02:03 <Luke-Jr> gmaxwell: ?
347 2013-02-23 07:02:07 <gmaxwell> 22:43 < petertodd> catchy
348 2013-02-23 07:02:11 <Luke-Jr> I didn't know the last 100 times I turned it off
349 2013-02-23 07:02:15 <gmaxwell> weird.
350 2013-02-23 07:02:32 <Luke-Jr> catchy = don't like? O.o
351 2013-02-23 07:03:02 <Luke-Jr> funny thing is, this music sounds nearly identical to that used in the opening theme for a crazy anime called Battle Programmer Shirase
352 2013-02-23 07:03:04 <Luke-Jr> <.<
353 2013-02-23 07:03:16 <petertodd> "Battle Programmer"?!
354 2013-02-23 07:03:26 <Luke-Jr> yes, it's about a computer programmer -.-
355 2013-02-23 07:03:39 <Luke-Jr> who's fighting crackers I think
356 2013-02-23 07:03:43 <Luke-Jr> I forget, it's been forever
357 2013-02-23 07:03:54 ThomasV has quit (Ping timeout: 264 seconds)
358 2013-02-23 07:04:22 <petertodd> sounds like quality tv
359 2013-02-23 07:04:25 <Luke-Jr> lol
360 2013-02-23 07:04:42 <Diablo-D3> its an old anime
361 2013-02-23 07:04:44 <Diablo-D3> like, late 80s
362 2013-02-23 07:04:44 <Luke-Jr> http://www.youtube.com/watch?v=L9IejLR8p7s
363 2013-02-23 07:05:09 <Luke-Jr> maybe not so identical, listening to this again
364 2013-02-23 07:05:26 <petertodd> Luke-Jr: see, rick is infecting your brain
365 2013-02-23 07:05:31 <Luke-Jr> lol
366 2013-02-23 07:05:37 <Luke-Jr> ah, this part is identical-sounding
367 2013-02-23 07:05:40 <Luke-Jr> about 50 seconds in
368 2013-02-23 07:05:47 <Diablo-D3> GODDAMNIT
369 2013-02-23 07:05:53 <Diablo-D3> IVE JUST BEEN RICKROLLED BY JAPAN
370 2013-02-23 07:06:24 <Luke-Jr> it would be funny if they intended it to sound rickrollish
371 2013-02-23 07:06:32 <petertodd> no, rick ashley BSP'd you
372 2013-02-23 07:07:01 <Luke-Jr> BSP = BS&P?
373 2013-02-23 07:07:18 <Diablo-D3> actually, thats not late 80s
374 2013-02-23 07:07:20 <Diablo-D3> thats more mid 90s
375 2013-02-23 07:07:34 <petertodd> s/BSP/BPS/
376 2013-02-23 07:07:37 <Diablo-D3> actually more late 90s
377 2013-02-23 07:08:15 <Diablo-D3> Luke-Jr: what year did this come out
378 2013-02-23 07:08:25 <Diablo-D3> the animation of the op is too shitty to be early or mid 00s
379 2013-02-23 07:08:38 <Diablo-D3> but the style they used is too late to be early 90s
380 2013-02-23 07:08:48 <Luke-Jr> Battle Programmer Shirase (BPS ããã«ããã°ã©ãã¼ã·ã©ã» BÄ«pÄ«esu Batoru PuroguramÄ Shirase?) is an anime TV series aired in 2003 and produced by AIC.
381 2013-02-23 07:08:55 <Diablo-D3> wow wtf
382 2013-02-23 07:09:00 <Diablo-D3> they should be ashamed of themselves
383 2013-02-23 07:09:04 <Diablo-D3> that is one low budget title
384 2013-02-23 07:09:08 <Luke-Jr> lol
385 2013-02-23 07:09:12 <Luke-Jr> I think the episodes were 5 minutes?
386 2013-02-23 07:09:19 <Diablo-D3> Ive never seen it
387 2013-02-23 07:09:30 <Luke-Jr> no, 11
388 2013-02-23 07:09:42 <Luke-Jr> 11 mins * 15 eps
389 2013-02-23 07:10:29 <petertodd> so... just over two star trek episodes
390 2013-02-23 07:10:40 <Luke-Jr> lol
391 2013-02-23 07:10:51 <Diablo-D3> thats like, first season of ah my goddess bad
392 2013-02-23 07:11:12 <Diablo-D3> "but diablo, whats so bad with the first season of ah my goddess?"
393 2013-02-23 07:11:38 <Diablo-D3> go watch the second one. one year apart really, 10 years apart in quality.
394 2013-02-23 07:12:20 <Luke-Jr> Diablo-D3: are you watching Psycho-Pass?
395 2013-02-23 07:12:28 <Diablo-D3> no
396 2013-02-23 07:12:36 <Diablo-D3> Luke-Jr: I like never watch new anime
397 2013-02-23 07:12:39 <Luke-Jr> o
398 2013-02-23 07:12:40 <Diablo-D3> unless its really really good
399 2013-02-23 07:12:48 <Luke-Jr> Psycho-Pass is the best airing now
400 2013-02-23 07:12:49 <Diablo-D3> or its kamen rider
401 2013-02-23 07:13:00 <Luke-Jr> which sadly isn't saying too much :/
402 2013-02-23 07:13:02 <Diablo-D3> I dont really care for /a/ animes really
403 2013-02-23 07:13:31 <Diablo-D3> now, if it had giant robots and shit blowing up and manly men saving the world? then Id watch it
404 2013-02-23 07:13:42 ThomasV has joined
405 2013-02-23 07:14:11 <andytoshi> about 19:15 into testnet, it's starting to vary more
406 2013-02-23 07:14:44 <Luke-Jr> Diablo-D3: instead, Psycho-Pass has a network of dismembered brains that judge society with brain scanners, in lieu of courts and laws
407 2013-02-23 07:15:14 <Luke-Jr> except for the worst criminals; instead, it has its human servants catch those so it can integrate them into itself
408 2013-02-23 07:15:20 nus- has joined
409 2013-02-23 07:15:25 <Diablo-D3> Luke-Jr: huh
410 2013-02-23 07:15:28 <Diablo-D3> I like the plot
411 2013-02-23 07:15:30 nus has quit (Read error: Connection reset by peer)
412 2013-02-23 07:15:36 <Diablo-D3> Ill have to remember that for later
413 2013-02-23 07:15:41 <Luke-Jr> Diablo-D3: well then you shouldn't have asked me to spoil it!
414 2013-02-23 07:15:42 <Luke-Jr> <.<
415 2013-02-23 07:15:43 <Diablo-D3> ALTHOUGH MY QUEUE IS LIKE OVER 9000 THINGS
416 2013-02-23 07:15:56 <Diablo-D3> seriously, who needs gigabit
417 2013-02-23 07:16:04 <Diablo-D3> Im already adding new shit to watch faster than I watch it
418 2013-02-23 07:16:18 <Luke-Jr> Diablo-D3: the hard part is finding *quality* stuff
419 2013-02-23 07:16:27 <petertodd> Diablo-D3: tvtropes.com <- way more efficient than primary sources
420 2013-02-23 07:16:42 <gmaxwell> yea, why would you watch stuff? reading is faster.
421 2013-02-23 07:16:50 <Diablo-D3> petertodd: I have that domain in /etc/hosts as 127.0.0.1 for safety reasons
422 2013-02-23 07:16:58 <Luke-Jr> gmaxwell: but I already read all the good books :<
423 2013-02-23 07:17:30 <gmaxwell> so read the bad ones next?
424 2013-02-23 07:17:43 <Luke-Jr> :/
425 2013-02-23 07:18:00 <petertodd> petertodd: Actually, so do I - I hire assistants to read it for me and give me summaries. (you're talking to one of my assistants right now)
426 2013-02-23 07:19:35 <Diablo-D3> btw
427 2013-02-23 07:19:39 <Diablo-D3> best android app
428 2013-02-23 07:19:41 <Diablo-D3> wifi analyzer
429 2013-02-23 07:21:25 * petertodd glares at iphone
430 2013-02-23 07:21:30 <Luke-Jr> petertodd: talking to yourself, eh?
431 2013-02-23 07:22:04 <Luke-Jr> [07:03:09] <osirisx11> does anyone know of any online btc unit converters (btc/mBTC/uBTC)?
432 2013-02-23 07:22:05 <Luke-Jr> wtf
433 2013-02-23 07:22:17 <petertodd> Luke-Jr: yeah, there have been mistakes made...
434 2013-02-23 07:22:29 <petertodd> wow...
435 2013-02-23 07:22:42 <gmaxwell> Must be american.
436 2013-02-23 07:22:57 <Luke-Jr> this is why we need tonal ?
437 2013-02-23 07:23:12 <petertodd> this is why we need elementary school
438 2013-02-23 07:23:37 <gmaxwell> petertodd: in america I think we cover that in grad school.
439 2013-02-23 07:23:58 <Luke-Jr> lol
440 2013-02-23 07:24:40 <gmaxwell> (they'd move it to highschool, but that would detract from time spent training for the nobody-gets-ahead standardized tests, which apparently don't test multiplying by 10)
441 2013-02-23 07:24:56 <petertodd> heh... we had interns this summer... our forth year electrical engineering student struggled to analyze a basic op-amp circuit
442 2013-02-23 07:26:09 <gmaxwell> I'm sure if you used exactly the formalism their class used they'd have done fine.
443 2013-02-23 07:26:12 <andytoshi> 30:15, a lot of chatter, some beeping noises, then back to static
444 2013-02-23 07:26:41 <petertodd> exactly... my boss said he wasn't half bad in the interview, and his grades were far better than mine
445 2013-02-23 07:26:50 <gmaxwell> aww, I didn't get that far, aborted it to play 73e64e38faea38 for my girlfriend.
446 2013-02-23 07:27:10 <andytoshi> gmaxwell: it's remarkable how much formal crap physics students can do without understanding any of it
447 2013-02-23 07:27:35 <Luke-Jr> andytoshi: are you streaming your office mic into testnet? -.-
448 2013-02-23 07:28:09 <andytoshi> i'm in an undergraduate physics course now, and the course material is incoherent nonsense -- but when i ask the other students how to make sense of it, they give me a bunch of symbol-shuffling rules
449 2013-02-23 07:28:15 <gmaxwell> petertodd: you've read the feynman books, right?
450 2013-02-23 07:28:19 <andytoshi> Luke-Jr: i'm streaming testnet out my laptop speakers
451 2013-02-23 07:28:33 <Luke-Jr> i c
452 2013-02-23 07:28:37 <petertodd> er... I've got them by my desk at work, but haven't read enough of them
453 2013-02-23 07:28:58 <andytoshi> ...then i ask where the rules come from, or even to give me some sort of actual axiomatic system to work with
454 2013-02-23 07:29:10 <gmaxwell> I was going to refer to this story: http://calteches.library.caltech.edu/46/2/LatinAmerica.htm but also going to insist you go get them now if you've not read them.
455 2013-02-23 07:29:27 <petertodd> the only physics courses I've actually done is the first year, although I had to do a bunch of basic electromagnetics crap to design some transformers
456 2013-02-23 07:29:47 <andytoshi> and they can't do it, they don't even know what they're writing
457 2013-02-23 07:29:49 <petertodd> ah, I've read that essay
458 2013-02-23 07:29:52 <petertodd> very good
459 2013-02-23 07:30:27 <gmaxwell> I suspect we've crept a little more like that in the US since the 60s, at least for undergrad education.
460 2013-02-23 07:30:48 <petertodd> I'd believe it
461 2013-02-23 07:31:34 <petertodd> first and second year math was a big eye opener for me, I was just barely passing the courses, but a big chunk of the students thought I was the smartest guy in the class soley because I could ask questions on the fly, frightening
462 2013-02-23 07:31:55 <Luke-Jr> lol
463 2013-02-23 07:32:16 <andytoshi> petertodd: i remember exactly that from a math analysis course
464 2013-02-23 07:32:33 <petertodd> nice, yeah, this was analysis for me
465 2013-02-23 07:32:35 <andytoshi> there were three of us who would ask questions and make jokes
466 2013-02-23 07:32:52 <petertodd> exactly!
467 2013-02-23 07:33:05 <gmaxwell> we can measure memorization and password guessing, but its harder to measure learning... so we measure what we can instead of what we should, then optimize for what we measure...
468 2013-02-23 07:33:34 <andytoshi> gmaxwell: i suppose that's what it is, since it doesn't happen outside of undergraduate courses
469 2013-02-23 07:33:39 <andytoshi> at least, IME
470 2013-02-23 07:34:12 <petertodd> for all the crap art schools get, the fact that it's irrelevant nonsense means the people who actually get it do really, really well and turn into solid self learners
471 2013-02-23 07:34:25 owowo has quit (Quit: sayonara)
472 2013-02-23 07:34:27 <gmaxwell> well it's less true the further you get from the cattle cars of people taking classes to get the grade and into groups of people with real interest in the subject matter.
473 2013-02-23 07:34:31 <petertodd> just too bad that leave 99% with burger-flipping jobs
474 2013-02-23 07:35:01 <petertodd> don't think most of the people in art schools have an interest in the subject matter :P
475 2013-02-23 07:35:18 <petertodd> it's just really easy to be someone who *does*
476 2013-02-23 07:35:20 <gmaxwell> plenty of people leave engineering school for burger-flipping jobsâ at least the art school people have lower debt on average (I hope?)
477 2013-02-23 07:35:45 <petertodd> in canada tuitions are both fairly cheap, and fairly consistent for whatever field
478 2013-02-23 07:36:12 <petertodd> we have maybe 2x differences at most, because for some fields the regulation is lifted a bit
479 2013-02-23 07:36:24 <gmaxwell> and if they raise themâ out with the pots and pans? :)
480 2013-02-23 07:36:29 <Luke-Jr> groups of people with real interest in the subject matter <-- does this exist in unis anymore? O.o
481 2013-02-23 07:36:30 <petertodd> yup
482 2013-02-23 07:36:46 <andytoshi> Luke-Jr: gmaxwell: yes to both :P
483 2013-02-23 07:36:51 <petertodd> for all the flack the quebecers got, the reality is those students did get what they wanted
484 2013-02-23 07:37:13 <andytoshi> Luke-Jr: there's really not many, and they tend not to participate in cattle-car courses
485 2013-02-23 07:37:32 <gmaxwell> Yep! though its still pretty scarry, I'm surprised at how US-style conservative the canadian government has become.
486 2013-02-23 07:37:48 <Luke-Jr> US-style conservative = neocon?
487 2013-02-23 07:37:53 <andytoshi> yeah, i was hoping it would be conservative-style conservative :(
488 2013-02-23 07:37:59 <petertodd> not a hope
489 2013-02-23 07:38:09 <petertodd> I don't think conservative-style conservative really exists anymore
490 2013-02-23 07:38:13 <gmaxwell> Luke-Jr: yes, after a fashion.
491 2013-02-23 07:38:19 <Luke-Jr> petertodd: it doesn't, that'd be a real monarchy
492 2013-02-23 07:38:30 <petertodd> heh, well there is that
493 2013-02-23 07:38:42 <Luke-Jr> afaik the only "monarchies" left are facades
494 2013-02-23 07:39:11 <jgarzik> think Middle East
495 2013-02-23 07:39:16 <petertodd> the UK is a really interesting example, because of the issues with their monarchy is the money they get from the state really is owed to them
496 2013-02-23 07:39:53 <petertodd> jgarzik: I'm waiting for their oil to run out...
497 2013-02-23 07:40:10 <andytoshi> 45:20 a lot of cutting out, then silence..
498 2013-02-23 07:40:41 <petertodd> andytoshi: ooh, maybe my megabytes of \xff's?
499 2013-02-23 07:40:47 DiabloD3 has joined
500 2013-02-23 07:41:06 <andytoshi> not sure, but i'm just over 22Mb in
501 2013-02-23 07:41:26 <petertodd> hmm... maybe a bit early
502 2013-02-23 07:41:30 <petertodd> or someone else had the same idea
503 2013-02-23 07:41:55 ThomasV has quit (Ping timeout: 252 seconds)
504 2013-02-23 07:41:58 nus- has quit (Ping timeout: 276 seconds)
505 2013-02-23 07:42:47 <andytoshi> 48:00, 23Mb, still silent, some blips here and there
506 2013-02-23 07:43:01 Diablo-D3 has quit (Ping timeout: 252 seconds)
507 2013-02-23 07:43:28 <gmaxwell> you're done
508 2013-02-23 07:43:36 <gmaxwell> perhaps?
509 2013-02-23 07:43:47 <petertodd> testnet3 is up to 32mb
510 2013-02-23 07:43:50 <gmaxwell> the file is preallocated.. so there is a wad of zeros at the end.
511 2013-02-23 07:44:04 <petertodd> oh, true, how much prealloc?
512 2013-02-23 07:44:16 <petertodd> andytoshi: you heard rick yet>?
513 2013-02-23 07:44:19 Diablo_D3 has joined
514 2013-02-23 07:44:20 <andytoshi> oh, but i didn't hear rick
515 2013-02-23 07:44:31 <gmaxwell> scratch that!
516 2013-02-23 07:44:38 <petertodd> probably my few mb of non-rick testing
517 2013-02-23 07:44:58 <andytoshi> i was hoping he'd show up around 30 minutes in, so i could distribute the file along with the "cover your eyes with ping pong balls and play static for half an hour and you'll trip" thing
518 2013-02-23 07:45:08 <andytoshi> there's just been a couple blips, so i'm not done
519 2013-02-23 07:45:34 PhantomSpark has joined
520 2013-02-23 07:45:45 <petertodd> (I meant to not let those blocks get onto the network, but forgot to firewall the ipv6 side of things and found that information is hard to stifle...)
521 2013-02-23 07:45:55 <andytoshi> haha
522 2013-02-23 07:45:57 <gmaxwell> lol
523 2013-02-23 07:46:12 <gmaxwell> you bloat blocks want to be free!
524 2013-02-23 07:46:19 <petertodd> yeah...
525 2013-02-23 07:46:27 <andytoshi> in a way, you're the real victim then
526 2013-02-23 07:46:45 <petertodd> don't give me any ideas...
527 2013-02-23 07:46:53 <petertodd> (though I'd be real sad if anything happened to rick)
528 2013-02-23 07:47:03 DiabloD3 has quit (Ping timeout: 252 seconds)
529 2013-02-23 07:47:25 <petertodd> hmm... pullreq for a testnet checkpoint
530 2013-02-23 07:48:33 <gmaxwell> petertodd: using a coinbase was a mistake
531 2013-02-23 07:48:47 <gmaxwell> if you used a regular txn and he got reorged outâ nodes would put him back in
532 2013-02-23 07:48:52 <petertodd> yeah, I know, they won't get put back in
533 2013-02-23 07:48:58 <petertodd> but I still have the code :P
534 2013-02-23 07:48:59 <andytoshi> 54:15 rick starts
535 2013-02-23 07:49:07 <petertodd> lol
536 2013-02-23 07:49:16 <andytoshi> it was all silence, then a short burst, then rick
537 2013-02-23 07:49:43 * gmaxwell wonders if the alt-ideas pages needs "serializations which are musically pleasing"
538 2013-02-23 07:49:54 phungus has quit (Quit: leaving)
539 2013-02-23 07:50:06 Diablo-D3 has joined
540 2013-02-23 07:51:15 Diablo_D3 has quit (Ping timeout: 248 seconds)
541 2013-02-23 07:51:21 <petertodd> as an artist, I second your idea
542 2013-02-23 07:51:52 <andytoshi> as a stereo-bicyle enthusiast, i third it
543 2013-02-23 07:52:03 <gmaxwell> I low cows.
544 2013-02-23 07:52:06 <gmaxwell> er like
545 2013-02-23 07:53:34 <andytoshi> all silence after rick, i think it's really over now
546 2013-02-23 07:53:48 <petertodd> that was a valient effort
547 2013-02-23 07:53:50 <andytoshi> i'm 28.3mb in, can't imagine the remaining 4mb actually has content
548 2013-02-23 07:53:54 <andytoshi> thx
549 2013-02-23 07:53:55 random_cat has joined
550 2013-02-23 07:54:00 <andytoshi> :P
551 2013-02-23 07:54:41 * andytoshi hits ctrl-c after exactly 60 minutes of blockchain techno
552 2013-02-23 07:55:19 profmac has joined
553 2013-02-23 07:55:48 * andytoshi turns on the bitcoin chain and goes to bed
554 2013-02-23 07:56:02 RBecker is now known as rbecker
555 2013-02-23 07:58:42 X-Scale has left ()
556 2013-02-23 08:09:29 freakazoid has quit (Ping timeout: 255 seconds)
557 2013-02-23 08:17:10 X-Scale has joined
558 2013-02-23 08:22:24 phungus has joined
559 2013-02-23 08:23:35 jdnavarro has joined
560 2013-02-23 08:25:58 Guest32273 has quit (Read error: No route to host)
561 2013-02-23 08:27:42 slush1 has quit (Ping timeout: 245 seconds)
562 2013-02-23 08:28:17 mappum has quit (Ping timeout: 246 seconds)
563 2013-02-23 08:38:12 davout has joined
564 2013-02-23 08:38:12 davout has quit (Changing host)
565 2013-02-23 08:38:12 davout has joined
566 2013-02-23 08:42:12 slush has joined
567 2013-02-23 08:43:18 Guest32273 has joined
568 2013-02-23 08:52:35 grau has quit (Remote host closed the connection)
569 2013-02-23 08:53:07 grau has joined
570 2013-02-23 08:57:49 HM has quit (Read error: Connection reset by peer)
571 2013-02-23 09:00:13 davout has quit (Ping timeout: 245 seconds)
572 2013-02-23 09:03:03 toffoo has quit ()
573 2013-02-23 09:03:17 dust-otc has joined
574 2013-02-23 09:05:12 JDuke128 has joined
575 2013-02-23 09:06:08 sgornick has quit (Ping timeout: 256 seconds)
576 2013-02-23 09:07:16 davout has joined
577 2013-02-23 09:12:47 davout has quit (Remote host closed the connection)
578 2013-02-23 09:13:23 davout has joined
579 2013-02-23 09:13:23 davout has quit (Changing host)
580 2013-02-23 09:13:23 davout has joined
581 2013-02-23 09:14:32 <Luke-Jr> gmaxwell: what if we craft a rickroll txn such that it gets relayed, but not mined? :P
582 2013-02-23 09:14:42 <Luke-Jr> then getrawtransaction works without bloating
583 2013-02-23 09:17:39 davout has quit (Ping timeout: 248 seconds)
584 2013-02-23 09:18:55 abrkn has quit (Ping timeout: 264 seconds)
585 2013-02-23 09:22:56 ThomasV has joined
586 2013-02-23 09:25:40 <ThomasV> is there a bitcoind command that returns the public key for a wallet address?
587 2013-02-23 09:25:56 <Luke-Jr> not afai
588 2013-02-23 09:25:57 <Luke-Jr> k
589 2013-02-23 09:26:41 <ThomasV> that's a pity, because createmultisig needs pubkeys
590 2013-02-23 09:26:56 * Luke-Jr facepalms
591 2013-02-23 09:27:16 <Luke-Jr> maybe I'm wrong
592 2013-02-23 09:27:18 <ThomasV> Luke-Jr: huh, did I get something wrong?
593 2013-02-23 09:27:24 <Luke-Jr> that would be a stupid oversight
594 2013-02-23 09:28:00 <Luke-Jr> ThomasV: let's hope one of us is wrong ;)
595 2013-02-23 09:28:08 <sipa> validateaddress
596 2013-02-23 09:28:23 <ThomasV> sipa: thanks
597 2013-02-23 09:31:18 <ThomasV> sipa: since you're here, another question. why does signrawtransaction always need redeemScript, even if the tx is already partially signed?
598 2013-02-23 09:31:36 abrkn has joined
599 2013-02-23 09:31:38 wizkid057 has quit (Remote host closed the connection)
600 2013-02-23 09:32:05 <ThomasV> I'm cloning these commands into electrum, and trying to use as much as possible the same syntax
601 2013-02-23 09:32:28 sgornick has joined
602 2013-02-23 09:36:33 Jackneill has joined
603 2013-02-23 09:36:55 wizkid057 has joined
604 2013-02-23 09:39:49 nus has joined
605 2013-02-23 09:43:47 dust-otc has quit (Read error: Connection reset by peer)
606 2013-02-23 09:44:21 dust-otc has joined
607 2013-02-23 09:49:02 denisx has joined
608 2013-02-23 09:58:32 profmac has quit (K-Lined)
609 2013-02-23 10:02:28 tonikt has joined
610 2013-02-23 10:03:07 TD has joined
611 2013-02-23 10:05:54 B0g4r7 has quit (Ping timeout: 276 seconds)
612 2013-02-23 10:11:20 B0g4r7 has joined
613 2013-02-23 10:12:30 TD has quit (Read error: Connection reset by peer)
614 2013-02-23 10:12:42 TD has joined
615 2013-02-23 10:14:17 ovidiusoft has joined
616 2013-02-23 10:22:09 B0g4r7 has quit (Ping timeout: 276 seconds)
617 2013-02-23 10:23:11 [\\\] has quit (Remote host closed the connection)
618 2013-02-23 10:25:01 [\\\] has joined
619 2013-02-23 10:26:31 BurtyBB is now known as BurtyB
620 2013-02-23 10:29:14 RazielZ has joined
621 2013-02-23 10:29:44 Goonie has joined
622 2013-02-23 10:30:18 B0g4r7 has joined
623 2013-02-23 10:30:58 variousnefarious has quit (Ping timeout: 276 seconds)
624 2013-02-23 10:32:56 variousnefarious has joined
625 2013-02-23 10:50:21 saulimus has joined
626 2013-02-23 10:58:10 Jackneill has quit (Remote host closed the connection)
627 2013-02-23 11:02:09 one_zero has quit ()
628 2013-02-23 11:04:57 JDuke128 has quit (Quit: ["Textual IRC Client: www.textualapp.com"])
629 2013-02-23 11:20:24 Jackneill has joined
630 2013-02-23 11:23:45 pooler has joined
631 2013-02-23 11:23:54 BTCOxygen has quit (Ping timeout: 252 seconds)
632 2013-02-23 11:28:14 ralphtheninja has joined
633 2013-02-23 11:37:12 rbecker has quit (Quit: You care. You're there for me. You love me so much, and I never want to let it go. You are the one truly amazing person. MDR 3/6/11 <3)
634 2013-02-23 11:39:16 RBecker has joined
635 2013-02-23 11:39:16 RBecker has quit (Excess Flood)
636 2013-02-23 11:41:22 RBecker has joined
637 2013-02-23 11:41:26 <TD> sigh
638 2013-02-23 11:41:34 <TD> i hate the stupid block chain download protocol
639 2013-02-23 11:44:42 Hashdog has joined
640 2013-02-23 11:48:34 brwyatt is now known as brwyatt|Away
641 2013-02-23 11:51:54 profmac has joined
642 2013-02-23 11:56:45 BTCOxygen has joined
643 2013-02-23 11:56:45 slush1 has joined
644 2013-02-23 11:57:39 slush has quit (Ping timeout: 248 seconds)
645 2013-02-23 11:58:38 tonikt has quit (Read error: Connection reset by peer)
646 2013-02-23 12:05:45 kirill_42 has joined
647 2013-02-23 12:08:07 i2pRelay has quit (Ping timeout: 276 seconds)
648 2013-02-23 12:17:43 copumpkin has quit (Ping timeout: 264 seconds)
649 2013-02-23 12:18:18 copumpkin has joined
650 2013-02-23 12:18:50 Ken` has joined
651 2013-02-23 12:18:54 abrkn has quit (Ping timeout: 252 seconds)
652 2013-02-23 12:19:17 dvide has quit ()
653 2013-02-23 12:20:26 pecket has quit (Quit: I'm not stupid. I'm just unlucky when I think.)
654 2013-02-23 12:20:30 kirill42 has joined
655 2013-02-23 12:21:53 kirill_42 has quit (Ping timeout: 245 seconds)
656 2013-02-23 12:27:49 Jackneill has quit (Remote host closed the connection)
657 2013-02-23 12:28:39 bitafterbit has joined
658 2013-02-23 12:32:14 JDuke128 has joined
659 2013-02-23 12:38:26 J6So4W7Yb has joined
660 2013-02-23 12:44:15 pecket has joined
661 2013-02-23 12:47:15 TradeFortress_ has quit (Ping timeout: 248 seconds)
662 2013-02-23 12:52:43 copumpkin has quit (Ping timeout: 245 seconds)
663 2013-02-23 12:53:18 copumpkin has joined
664 2013-02-23 13:05:58 TD has quit (Ping timeout: 244 seconds)
665 2013-02-23 13:06:37 ralphtheninja has quit (Ping timeout: 276 seconds)
666 2013-02-23 13:07:54 ralphtheninja has joined
667 2013-02-23 13:11:15 denisx has quit (Quit: denisx)
668 2013-02-23 13:11:51 copumpkin has quit (Ping timeout: 252 seconds)
669 2013-02-23 13:12:25 copumpkin has joined
670 2013-02-23 13:17:31 daybyter has joined
671 2013-02-23 13:22:11 BTCOxygen has joined
672 2013-02-23 13:23:20 TradeFortress_ has joined
673 2013-02-23 13:24:10 BTCOxygen has quit (Ping timeout: 260 seconds)
674 2013-02-23 13:27:25 J6So4W7Yb has quit (Ping timeout: 276 seconds)
675 2013-02-23 13:28:18 J6So4W7Yb has joined
676 2013-02-23 13:44:04 asa1024 has quit (Quit: asa1024)
677 2013-02-23 13:57:30 copumpkin has quit (Ping timeout: 252 seconds)
678 2013-02-23 13:57:59 theorbtwo has quit (Read error: Connection reset by peer)
679 2013-02-23 13:58:04 copumpkin has joined
680 2013-02-23 13:59:41 theorbtwo has joined
681 2013-02-23 14:02:01 <treb> hi, i'm on version 0.7 and here's my ulimits: http://pastebin.com/dveuU8Ap
682 2013-02-23 14:02:23 <treb> my bitcoind did crash twice now, at least once with errors reported in memory allocation
683 2013-02-23 14:02:28 <treb> do you see anything i may need to adjust?
684 2013-02-23 14:08:07 <treb> aha, my ulimit -v was 512MB, may be the issue, need to check out
685 2013-02-23 14:09:57 saulimus has quit (Quit: saulimus)
686 2013-02-23 14:10:19 BTCOxygen has quit (1!~BTCOxygen@unaffiliated/mroxy/bot/btcoxygen|Ping timeout: 276 seconds)
687 2013-02-23 14:30:25 valparaiso_ has joined
688 2013-02-23 14:30:49 valparaiso_ is now known as Guest20847
689 2013-02-23 14:30:49 valparaiso has quit (Ping timeout: 276 seconds)
690 2013-02-23 14:31:59 JDuke128 has quit (Quit: ["Textual IRC Client: www.textualapp.com"])
691 2013-02-23 14:32:10 Zarutian has joined
692 2013-02-23 14:35:57 Guest20847 has quit (Quit: Guest20847)
693 2013-02-23 14:36:17 valparaiso has joined
694 2013-02-23 14:41:54 ottulo has joined
695 2013-02-23 14:42:51 ThomasV has quit (Quit: Leaving)
696 2013-02-23 14:45:04 <ottulo> hello, I'm having a problem syncing the blockchain, anyone available?
697 2013-02-23 14:45:29 <SomeoneWeird> whats the problem?
698 2013-02-23 14:46:00 <ottulo> I have bitcoin-qt running on 2 separate computers, neither will sync past block 218275
699 2013-02-23 14:46:19 <ottulo> it's been stuck for a couple days now
700 2013-02-23 14:47:30 <MC1984> debug.log?
701 2013-02-23 14:49:03 <ottulo> seems to have plenty of stuff, nearly 200k lines... I'll upload it somewhere I suppose
702 2013-02-23 14:51:49 <ottulo> http://dl.dropbox.com/u/67651714/debug.log
703 2013-02-23 14:54:31 <MC1984> just look at the end
704 2013-02-23 14:55:28 <SomeoneWeird> ;;errno 10
705 2013-02-23 14:55:28 <SomeoneWeird> ;;errno 104
706 2013-02-23 14:55:28 <gribble> ECHILD (#10): No child processes
707 2013-02-23 14:55:29 <gribble> ECONNRESET (#104): Connection reset by peer
708 2013-02-23 14:55:40 <MC1984> oh fuck my browser just went down like the titanic
709 2013-02-23 14:55:58 <ottulo> aw, thanks to the log file?
710 2013-02-23 14:56:39 <MC1984> accepted connection 192.168.1.1:43027
711 2013-02-23 14:56:39 <MC1984> connected to self at 192.168.1.1:43027, disconnecting
712 2013-02-23 14:56:42 <MC1984> wut
713 2013-02-23 14:57:09 <MC1984> like its trying to connect out but the router is bouncing it back to itself or some shit?
714 2013-02-23 14:57:22 <ottulo> 2pcs in wlan, could they connect to each other?
715 2013-02-23 14:57:38 jdnavarro has quit (Remote host closed the connection)
716 2013-02-23 14:58:01 <MC1984> yes but not within the lan i dont think
717 2013-02-23 14:58:25 <MC1984> bitcoin doesnt have any sort of local peer resolution
718 2013-02-23 14:59:50 <MC1984> som routers bounce requests that go out the wan and straight back in around the lan by rewriting packet headers
719 2013-02-23 15:00:36 <MC1984> reset your router and see if it helps
720 2013-02-23 15:00:47 <ottulo> alright, I'll be back in a few
721 2013-02-23 15:02:24 CodeShark has quit (Remote host closed the connection)
722 2013-02-23 15:02:39 ottulo has quit (Quit: Leaving)
723 2013-02-23 15:02:46 ottulo has joined
724 2013-02-23 15:08:30 <ottulo> seems like it's running fine now
725 2013-02-23 15:08:38 <MC1984> both of them?
726 2013-02-23 15:08:48 daybyter has quit (Read error: Connection reset by peer)
727 2013-02-23 15:08:57 <ottulo> yep
728 2013-02-23 15:09:29 <MC1984> you can try going into your route and turning off "WAN request redirect" or words to that effect
729 2013-02-23 15:09:37 LainZ has quit (Remote host closed the connection)
730 2013-02-23 15:10:19 <MC1984> its a neat trick to save bandwidth but ive seen routers get confused with it before when the same service is running on two different hosts on the lan side
731 2013-02-23 15:10:42 <ottulo> ok, will see to that later, now I gotta go to sauna
732 2013-02-23 15:10:43 CodeShark has joined
733 2013-02-23 15:11:26 <ottulo> thanks for the help though, I'll know where to look first next time
734 2013-02-23 15:11:52 CodeShark has quit (Remote host closed the connection)
735 2013-02-23 15:14:11 SugarPink has joined
736 2013-02-23 15:14:13 SugarPink has left ()
737 2013-02-23 15:14:15 SugarPink has joined
738 2013-02-23 15:16:03 TradeFortress_ has quit (Ping timeout: 248 seconds)
739 2013-02-23 15:22:43 Luke-Jr has quit (Ping timeout: 245 seconds)
740 2013-02-23 15:23:29 jrmithdobbs has quit (Ping timeout: 276 seconds)
741 2013-02-23 15:27:49 ThomasV has joined
742 2013-02-23 15:27:57 LainZ has joined
743 2013-02-23 15:31:25 jrmithdobbs has joined
744 2013-02-23 15:33:06 [\\\] has quit (Remote host closed the connection)
745 2013-02-23 15:33:08 ralphtheninja has quit (Quit: leaving)
746 2013-02-23 15:33:20 Zarutian has quit (Quit: Zarutian)
747 2013-02-23 15:34:17 tonikt has joined
748 2013-02-23 15:34:54 [\\\] has joined
749 2013-02-23 15:40:42 MobiusL has quit (Remote host closed the connection)
750 2013-02-23 15:42:20 vigilyn has quit (Read error: Connection reset by peer)
751 2013-02-23 15:42:26 toffoo has joined
752 2013-02-23 15:42:36 vigilyn has joined
753 2013-02-23 15:42:36 vigilyn has quit (Changing host)
754 2013-02-23 15:42:36 vigilyn has joined
755 2013-02-23 15:46:14 MobiusL has joined
756 2013-02-23 15:56:29 <epscy> is there a way to have bitcoind notify when a transaction is received (unconfimed)
757 2013-02-23 15:59:55 egecko has quit (Quit: ~ Trillian Astra - www.trillian.im ~)
758 2013-02-23 16:00:59 SugarPink has quit (Read error: Connection reset by peer)
759 2013-02-23 16:09:07 Belkaar_ has joined
760 2013-02-23 16:09:12 HM has joined
761 2013-02-23 16:10:12 Belkaar has quit (Ping timeout: 255 seconds)
762 2013-02-23 16:10:12 Belkaar_ is now known as Belkaar
763 2013-02-23 16:10:41 <dario1> epscy: Will be in next release
764 2013-02-23 16:11:30 <Scrat> dario1: talking about walletnotify? if not paste commit #
765 2013-02-23 16:11:55 egecko has joined
766 2013-02-23 16:12:06 <dario1> Scrat: Yes, that commit has been recently merged
767 2013-02-23 16:16:01 <Scrat> since it triggers once on mempool and once on block inclusion it should trigger once while syncing? if the transaction took place while your bitcoind was offline
768 2013-02-23 16:17:04 andytoshi has quit (Ping timeout: 276 seconds)
769 2013-02-23 16:20:06 SmallTim has quit (Ping timeout: 260 seconds)
770 2013-02-23 16:20:16 Zarutian has joined
771 2013-02-23 16:29:40 ottulo has quit (Quit: Leaving)
772 2013-02-23 16:32:01 <epscy> dario1: thanks
773 2013-02-23 16:37:18 Zarutian has quit (Quit: Zarutian)
774 2013-02-23 16:39:25 davout has joined
775 2013-02-23 16:41:19 davout has quit (Remote host closed the connection)
776 2013-02-23 16:47:40 Ferroh_ is now known as Ferroh
777 2013-02-23 16:53:41 daybyter has joined
778 2013-02-23 17:05:20 owowo has joined
779 2013-02-23 17:05:46 fishfish has joined
780 2013-02-23 17:06:45 da2ce7 has joined
781 2013-02-23 17:08:49 da2ce7_d has quit (Ping timeout: 244 seconds)
782 2013-02-23 17:08:52 jdnavarro has joined
783 2013-02-23 17:10:05 mareo87 has joined
784 2013-02-23 17:14:53 andytoshi has joined
785 2013-02-23 17:19:22 andytoshi has quit (Client Quit)
786 2013-02-23 17:20:13 jdnavarro has quit (Remote host closed the connection)
787 2013-02-23 17:20:52 jdnavarro has joined
788 2013-02-23 17:20:52 jdnavarro has quit (Read error: error:1408F119:SSL routines:SSL3_GET_RECORD:decryption failed or bad record mac)
789 2013-02-23 17:22:09 <gavinandresen> ThomasV : signrawtransaction needs the redeemScript even for a partially-signed transaction because I never wrote any code to look in the scriptSig for the redeemScript. Should be pretty easy to add....
790 2013-02-23 17:22:12 robocoin has joined
791 2013-02-23 17:22:42 Hashdog has quit (Remote host closed the connection)
792 2013-02-23 17:22:44 <ThomasV> gavinandresen: thanks for the answer. I was wondering if there was a deeper reason
793 2013-02-23 17:23:03 <gavinandresen> nope, just the assumption that all parties will know the redeemscript a-priori.
794 2013-02-23 17:23:24 toffoo has quit ()
795 2013-02-23 17:24:01 gfinn has quit (Ping timeout: 276 seconds)
796 2013-02-23 17:24:03 <gavinandresen> ThomasV: can you file an issue so we don't forget? And⦠"patches welcome" ...
797 2013-02-23 17:24:20 <ThomasV> I will
798 2013-02-23 17:25:06 <ThomasV> gavinandresen: well, it's not really an issue for me. I was asking because I try to make electrum commands compatible
799 2013-02-23 17:25:44 <gavinandresen> Electrum has an RPC now? cool...
800 2013-02-23 17:26:02 <ThomasV> not really rpc for the moment, just cmd line
801 2013-02-23 17:52:10 snakie has joined
802 2013-02-23 17:55:53 <etotheipi_> sipa: ping
803 2013-02-23 17:57:08 <etotheipi_> sipa: I was thinking about the seed criteria, "N zero bits after 2^N iterations"
804 2013-02-23 17:58:10 <etotheipi_> sipa: if we assume that N is 16, then that's only 65536 iterations... we can easily keep a circular buffer of values and just keep hashing sequentially until we find N zero bits... then go to the beginning of the circular buffer and there's what's we were looking for
805 2013-02-23 17:59:24 <etotheipi_> we have to do at least 65536 iterations to fill the buffer, but after that we will find a sufficient seed within then next 65k iterations 50% of the time
806 2013-02-23 17:59:29 <etotheipi_> gmaxwell: ^^
807 2013-02-23 18:00:06 <etotheipi_> so searching is not actually N^4
808 2013-02-23 18:00:46 <etotheipi_> assuming the system has sufficient RAM to hold a circular buffer of hashes of size 2^N
809 2013-02-23 18:05:36 RBecker is now known as rbecker
810 2013-02-23 18:06:00 rbecker is now known as RBecker
811 2013-02-23 18:07:17 nus has quit (Ping timeout: 276 seconds)
812 2013-02-23 18:07:33 <HM> etotheipi_: what problem is that solving?
813 2013-02-23 18:10:23 RBecker is now known as rbecker
814 2013-02-23 18:11:08 nus has joined
815 2013-02-23 18:12:01 rbecker is now known as RBecker
816 2013-02-23 18:12:32 <etotheipi_> HM, we were talking about discouraging brainwallets (and encouraging good entropy), and adding stretching, in BIP 32 seeds
817 2013-02-23 18:12:57 <etotheipi_> by having apps require BIP32 seeds that have N zero bits after 2^N sequential hashes
818 2013-02-23 18:14:11 <HM> wait, how can you enforce N hashes
819 2013-02-23 18:14:22 <HM> 2^N rather
820 2013-02-23 18:14:53 <HM> N zero bits would be found in an average of 2^(N-1) hashes wouldn't it?
821 2013-02-23 18:15:18 <etotheipi_> Bitcoin application: "To recover your wallet, please enter the seed value:" <user enters "My name is bob">; Bitcoin Application <hashes it 2^N times sees it does not have zero bits, rejects it>
822 2013-02-23 18:16:24 <HM> where do you apply the salt to ensure that?
823 2013-02-23 18:16:27 <etotheipi_> HM: the concern was that the original *search* for such a seed to give to the user in the first place, might take a while
824 2013-02-23 18:16:28 <HM> the last hash?
825 2013-02-23 18:16:39 Hashdog has joined
826 2013-02-23 18:17:36 <etotheipi_> you can't have a salt... at least not an application-specific one
827 2013-02-23 18:17:49 <HM> so "my name is bob" is generated, not chosen by the user?
828 2013-02-23 18:17:56 <HM> it's a mnemonic
829 2013-02-23 18:18:06 <etotheipi_> the idea is that multiple programs all decide to agree on this criteria, and entering a valid seed in any of them will produce the same BIP 32 wallet
830 2013-02-23 18:18:22 <etotheipi_> and they all agree that only 1/2^N seeds will be valid
831 2013-02-23 18:19:03 <etotheipi_> HM, regular users who don't understand or care about brainwallets, will simply create a new wallet, and the application will search for a 128-bit seed that meets this criteria
832 2013-02-23 18:20:00 <etotheipi_> but if any seed is valid, then there's nothing stopping users from entering "Bob" as their seed in the wallet recovery, and creating a wallet that is bound to collide with another terrible user using "Bob' as thier seed
833 2013-02-23 18:20:23 <etotheipi_> and of course, attackers attacking all wallet apps at once by just searching through all seeds
834 2013-02-23 18:20:23 <HM> I'm not sure I see that
835 2013-02-23 18:20:55 <etotheipi_> *all seeds* means, a lot of very-low-entropy seed values that are likely to be produced by humans
836 2013-02-23 18:21:18 <HM> if you have a 1000 words in your dictionary and 5 words then you have ~50 bits of possible mnemonics
837 2013-02-23 18:21:22 <HM> your hashing requirement reduces that
838 2013-02-23 18:21:47 <HM> because only 1 in 2^something will produce the 0s you want
839 2013-02-23 18:23:17 <etotheipi_> HM, it's kind of like saying: we only accept seeds such that seed%65536==0.... it doesn't reduce entropy just becuase we've artificially reducded the seedspace by 65535/65536
840 2013-02-23 18:23:47 <etotheipi_> however, if you were using the resulting value is does have leading zeros (instead of the original seed, then yes, you've reduced entropy)
841 2013-02-23 18:24:06 <HM> oh i see
842 2013-02-23 18:24:06 <etotheipi_> err... parens are a little misplaced there
843 2013-02-23 18:24:16 <HM> errm
844 2013-02-23 18:24:20 <HM> no i don't
845 2013-02-23 18:24:43 <etotheipi_> damnit, I have to run ... sipa gmaxwell, <tag>
846 2013-02-23 18:24:48 <HM> :|
847 2013-02-23 18:25:19 <etotheipi_> HM, the point is that the original seed still has 128 bits of entropy... it's just an arbitrarily filtered one
848 2013-02-23 18:25:21 chmod755 has joined
849 2013-02-23 18:25:47 <etotheipi_> knowing that the 2^N'th hash of that value has N leading zeros doesn't help you narrow your search space
850 2013-02-23 18:26:28 <etotheipi_> in fact, it means you have to search 2^N times as many values (and do 2^N hashes for each) in order to search for that seed
851 2013-02-23 18:27:34 <HM> you have to put 2^N the hash operations to filter the keyspace
852 2013-02-23 18:27:48 <HM> but the keyspace is effectively reduced because only a subset of those keys will satisfy the requirement
853 2013-02-23 18:27:56 <HM> so i don't see what is really gained
854 2013-02-23 18:28:14 <etotheipi_> HM, yes, but you don't know, without doing 2^N hashes, which part of the seedspace each seed belongs to
855 2013-02-23 18:28:54 <HM> huh?
856 2013-02-23 18:29:08 <etotheipi_> HM, given the criteria I have given you, and you want to try to brute force a seed.... you are suggesting you only have to search 1/2^N as many keys
857 2013-02-23 18:29:29 <etotheipi_> what I'm saying is, you don't know how to search only the reduced key space
858 2013-02-23 18:29:32 <HM> if you had a 50 bit keyspace
859 2013-02-23 18:29:38 <HM> and 2 bytes had to be 0
860 2013-02-23 18:30:09 <HM> assuming a good hash function only 2^34 keys would prefix 0000
861 2013-02-23 18:30:18 <etotheipi_> HM, but the value that has only 34 non-zero bits is not the seed... it's just the proof that original seed chosen is 1/2^N
862 2013-02-23 18:30:34 <etotheipi_> *is part of that 1/2^N seedspace
863 2013-02-23 18:30:37 <HM> why does that improve security?
864 2013-02-23 18:31:07 <HM> sounds like a checksum that gives no additional strength
865 2013-02-23 18:31:08 <etotheipi_> for seed values that are not generated with proper entropy, it guarantees at least N bits of entropy be appended in order to be valid
866 2013-02-23 18:31:57 sgornick has quit (Ping timeout: 252 seconds)
867 2013-02-23 18:32:01 <etotheipi_> if I want to use "Bob" as my seed, well there's only a 1/2^N chance that the app will accept that... I'm going to have to search through seeds of value "Bob<N-bit random number>" in order to find one that the application will accept
868 2013-02-23 18:32:31 <HM> well you never mentioned using a nonce
869 2013-02-23 18:32:55 <etotheipi_> HM: the nonce is only an artifact of this scheme, for people that insist on using simple seeds
870 2013-02-23 18:33:11 <etotheipi_> in the grand scheme of things, for 98% of users, this won't be a problem
871 2013-02-23 18:33:23 <etotheipi_> the app generates 128 bits of entropy that meets this criteria
872 2013-02-23 18:33:33 <HM> i thought you were talking about a *fixed* iteration count of hashes
873 2013-02-23 18:33:35 <etotheipi_> and they only enter 128-bits of seed to recover their wallet
874 2013-02-23 18:34:33 <HM> so the user inputs "Bob" and the program must find the salt such that substr(h(salt+"Bob"), 0, 3) == 0x00 0x00 0x00
875 2013-02-23 18:34:39 <HM> is that the jist?
876 2013-02-23 18:34:55 <HM> then the full key is derived from salt+Bob
877 2013-02-23 18:34:57 <etotheipi_> HM: well there won't be a program that does it for them... they'll have to make it themselves
878 2013-02-23 18:35:21 <etotheipi_> my point was, it discourages them from using crappy entropy, and encourages them to just use the seed value generated by the app
879 2013-02-23 18:35:44 <HM> if the app generates this "seed" what is the point
880 2013-02-23 18:35:52 <HM> just randomly generate one to begin with, with good entropy
881 2013-02-23 18:36:12 <petertodd> the seed is shorter and easier to memorize
882 2013-02-23 18:36:39 <HM> right sure
883 2013-02-23 18:36:59 <petertodd> eaiser to handle too - less typing to put it into your off-line computer or whatever
884 2013-02-23 18:37:01 CodeShark has joined
885 2013-02-23 18:37:01 <HM> but if someone makes a crappy 1000 word mnemonic generator that outputs 5 words, e.g. easy to remember but only 50 bits
886 2013-02-23 18:37:04 <etotheipi_> HM: because without this check, the user can enter 2 bits of entropy into the "wallet recovery from seed value" window
887 2013-02-23 18:37:21 <HM> i can still precompute ~2^34 values
888 2013-02-23 18:37:28 <petertodd> like causacius's short private key format kinda
889 2013-02-23 18:37:31 <etotheipi_> gah, brb
890 2013-02-23 18:38:15 <HM> i understand how key stretching works btw
891 2013-02-23 18:38:31 <HM> i just don't see what the advantage is when the user isn't chosing their own password
892 2013-02-23 18:38:58 <HM> just use a bigger dictionary
893 2013-02-23 18:39:07 <HM> it's just as easy to remember
894 2013-02-23 18:39:13 <HM> 5 words is 5 words
895 2013-02-23 18:39:44 <petertodd> maybe you're right. we probably need a human interface designer...
896 2013-02-23 18:39:52 <petertodd> and some science tests
897 2013-02-23 18:40:22 <petertodd> more likely, human memorizable seeds are just bad, so don't support them, but people want to do stupid things
898 2013-02-23 18:40:23 <HM> i wouldn't like to arbitrarily make these choices for users, let them make their own mistakes
899 2013-02-23 18:40:33 <HM> if you're building a brainwallet in to the qt interface, then sure. go for it
900 2013-02-23 18:41:00 <HM> the thing about supposedly "memorable" phrases is they aren't
901 2013-02-23 18:41:04 <petertodd> meh, this argument will be around forever, to be honest I find it all kinda tiresome
902 2013-02-23 18:41:20 <petertodd> personally I have a few 128bit passwords memorized, but I'm a weirdo
903 2013-02-23 18:41:29 <HM> "correct horse battery staple" might seem memorable, but try remembering one for every site you use
904 2013-02-23 18:41:35 <HM> (or wallet in this case)
905 2013-02-23 18:41:41 <petertodd> (passwords were generated properly and randomly btw)
906 2013-02-23 18:41:56 J6So4W7Yb has quit (Ping timeout: 248 seconds)
907 2013-02-23 18:42:00 <petertodd> yeah, I use a few passwords for varying levels of security, and prefix them with the site name as a shitty salt
908 2013-02-23 18:42:13 <petertodd> mostly I use google auth whereever possible
909 2013-02-23 18:42:28 J6So4W7Yb has joined
910 2013-02-23 18:42:33 <HM> i use keepass but find it frustrating
911 2013-02-23 18:42:49 <HM> i'm tempted to move to chrome sync
912 2013-02-23 18:42:51 <petertodd> I don't like how keypass isn't git-rcs compat
913 2013-02-23 18:42:57 <HM> as horrific as that sounds :|
914 2013-02-23 18:43:31 <HM> i use ssh keys with github
915 2013-02-23 18:43:50 <HM> meh
916 2013-02-23 18:44:01 <petertodd> yeah, I got a hardware smartcard for gpg/ssh stuff
917 2013-02-23 18:44:13 <HM> i think brainwallets are more about not worrying about losing the keys than anything else
918 2013-02-23 18:44:17 <HM> data loss happens
919 2013-02-23 18:44:19 <petertodd> absolutely
920 2013-02-23 18:44:27 <petertodd> and also, ensuring that no-one can prove you have the keys too
921 2013-02-23 18:44:35 zxcanon has joined
922 2013-02-23 18:46:00 <HM> I don't see how keystretching helps for purely memorised passwords
923 2013-02-23 18:46:02 monkeynipples has joined
924 2013-02-23 18:46:22 <petertodd> why not? reduces the bits you need to memorize
925 2013-02-23 18:46:29 <petertodd> just not by as much as people would like
926 2013-02-23 18:47:01 JDuke128 has joined
927 2013-02-23 18:47:06 nouitfvf has quit (Quit: happy chatting)
928 2013-02-23 18:47:27 <HM> sorry i mean what eto was saying
929 2013-02-23 18:47:53 <HM> you can make your client stretch keys but i don't get the whole generate seeds that satisfy his prefix requirement thing
930 2013-02-23 18:48:10 <HM> it just sounds like an expensive checksum that wouldn't slow bruteforce
931 2013-02-23 18:48:22 <petertodd> gotta go, later
932 2013-02-23 18:48:27 <HM> later
933 2013-02-23 18:49:46 TD has joined
934 2013-02-23 18:51:24 TD is now known as TD[kiwi]
935 2013-02-23 18:51:53 RBecker is now known as rbecker
936 2013-02-23 18:54:38 chmod755 is now known as chmod-rwxr-xr-x
937 2013-02-23 18:55:06 TD has joined
938 2013-02-23 18:58:13 Maged has quit (Quit: ChatZilla 0.9.90 [Firefox 18.0.2/20130201065344])
939 2013-02-23 19:00:01 FredEE has joined
940 2013-02-23 19:01:17 Ramokk has quit (Read error: Connection reset by peer)
941 2013-02-23 19:03:30 Ramokk has joined
942 2013-02-23 19:05:34 TD[kiwi] has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client)
943 2013-02-23 19:11:02 <HM> Scrupt has been published as a draft by the IETF now, might be worth using to stretch keys over boring old PBKDF2
944 2013-02-23 19:12:51 <HM> Scrypt*
945 2013-02-23 19:13:02 chmod-rwxr-xr-x is now known as chmod755
946 2013-02-23 19:14:23 wizkid057 has quit (Quit: THE END IS NEAR)
947 2013-02-23 19:17:54 daybyter has quit (Quit: Konversation terminated!)
948 2013-02-23 19:18:56 ashams has joined
949 2013-02-23 19:18:56 ashams has quit (Changing host)
950 2013-02-23 19:18:56 ashams has joined
951 2013-02-23 19:19:44 Luke-Jr has joined
952 2013-02-23 19:21:12 <HM> another thing you could do is pilfer a chunk of the public key as a salt
953 2013-02-23 19:21:52 <HM> so generate hash(hash(password)+counter) until the the prefix of public key is what you need
954 2013-02-23 19:22:19 <HM> if a user loses their private key the public key (and therefore salt) should still be findable if the wallet has any value
955 2013-02-23 19:23:16 <HM> if counter isn't at least as many iterations as you want, just randomly select a new prefix the first time you generate the brainwallet
956 2013-02-23 19:24:32 <HM> actually that wouldn't slow down a brute force attack against a known wallet, it'd just reduce the chances of 2 people generating the same brainwallet
957 2013-02-23 19:31:06 <HM> that's basically vanitygen
958 2013-02-23 19:31:36 <HM> vanitygen could use used for brainwalleting if you seeded the PRNG from a password
959 2013-02-23 19:37:39 <grau> BlueMatt: What is the "Exploit the SigHash.SINGLE bug to avoid having to make more than one signature" I see in the blocktester ?
960 2013-02-23 19:41:54 ralphtheninja has joined
961 2013-02-23 19:43:44 <grau> BlueMatt: Never mind I got it
962 2013-02-23 19:45:02 ralphtheninja has left ()
963 2013-02-23 19:50:36 J6So4W7Yb has quit (Ping timeout: 252 seconds)
964 2013-02-23 19:50:56 J6So4W7Yb has joined
965 2013-02-23 19:53:06 freakazoid has joined
966 2013-02-23 19:55:42 J6So4W7Yb has quit (Ping timeout: 260 seconds)
967 2013-02-23 19:55:59 CodeShark has quit (Remote host closed the connection)
968 2013-02-23 19:59:51 etotheipi_ has quit (Quit: Ex-Chat)
969 2013-02-23 20:04:14 [\\\] has quit (Remote host closed the connection)
970 2013-02-23 20:05:57 [\\\] has joined
971 2013-02-23 20:05:58 Belkaar has quit (Ping timeout: 260 seconds)
972 2013-02-23 20:06:30 owowo has quit (Remote host closed the connection)
973 2013-02-23 20:06:46 Lone has joined
974 2013-02-23 20:06:48 tonikt has quit (Read error: Connection reset by peer)
975 2013-02-23 20:06:52 LainZ has quit (Read error: Connection reset by peer)
976 2013-02-23 20:07:40 Belkaar has joined
977 2013-02-23 20:16:10 <grau> BlueMatt: no I do not yet have it...
978 2013-02-23 20:20:10 <TD> sighash_single processing has a bug
979 2013-02-23 20:20:21 <TD> see bitcoinj, Transaction.hashTransactionForSignature
980 2013-02-23 20:22:05 <TD> in that case he appears to be using it as a shortcut
981 2013-02-23 20:24:20 Hashdog has quit (Ping timeout: 264 seconds)
982 2013-02-23 20:26:01 Belkaar has quit (Ping timeout: 276 seconds)
983 2013-02-23 20:26:11 Belkaar_ has joined
984 2013-02-23 20:26:11 Belkaar_ is now known as Belkaar
985 2013-02-23 20:26:16 wizkid057 has joined
986 2013-02-23 20:29:20 rbecker is now known as RBecker
987 2013-02-23 20:30:50 FredEE has quit (Quit: FredEE)
988 2013-02-23 20:39:49 FredEE has joined
989 2013-02-23 20:40:31 sgornick has joined
990 2013-02-23 20:46:46 JDuke128 has quit (Quit: ["Textual IRC Client: www.textualapp.com"])
991 2013-02-23 20:47:08 WolfAlex_ has joined
992 2013-02-23 20:47:33 FredEE has quit (Quit: FredEE)
993 2013-02-23 20:48:29 Ken` has quit (Quit: leaving)
994 2013-02-23 20:48:38 CodeShark has joined
995 2013-02-23 20:50:28 WolfAlex has quit (Ping timeout: 248 seconds)
996 2013-02-23 20:52:44 kirill42 has quit (Ping timeout: 245 seconds)
997 2013-02-23 20:57:39 mappum has joined
998 2013-02-23 21:01:52 Oswa has joined
999 2013-02-23 21:02:47 <grau> TD: Thanks that is it: "actually returns the constant "1" to indicate an error, which is never checked for. Oops."
1000 2013-02-23 21:03:34 Oswa has left ()
1001 2013-02-23 21:04:04 gruvfunk has quit (Quit: Leaving)
1002 2013-02-23 21:04:13 dazla has joined
1003 2013-02-23 21:04:33 starsoccer has joined
1004 2013-02-23 21:07:46 Oswa has joined
1005 2013-02-23 21:08:15 grau has quit (Remote host closed the connection)
1006 2013-02-23 21:08:30 <Oswa> does anybody know when the ubuntu ppa will be updated?
1007 2013-02-23 21:08:47 <dazla> anyone buying
1008 2013-02-23 21:12:17 TD has quit (Ping timeout: 260 seconds)
1009 2013-02-23 21:14:13 testnode9 has quit (Remote host closed the connection)
1010 2013-02-23 21:14:57 <BlueMatt> Oswa: today
1011 2013-02-23 21:15:40 <Oswa> bluematt: thanks
1012 2013-02-23 21:16:14 asuk is now known as asuk|afk
1013 2013-02-23 21:16:32 testnode9 has joined
1014 2013-02-23 21:16:58 EvanR3 is now known as EvanR
1015 2013-02-23 21:17:28 EvanR is now known as Guest87093
1016 2013-02-23 21:17:55 Guest87093 is now known as EvanR
1017 2013-02-23 21:17:59 EvanR has quit (Changing host)
1018 2013-02-23 21:18:00 EvanR has joined
1019 2013-02-23 21:20:11 coolsa_ has joined
1020 2013-02-23 21:20:44 grau has joined
1021 2013-02-23 21:22:19 coolsa has quit (Ping timeout: 252 seconds)
1022 2013-02-23 21:26:50 nus has quit (Ping timeout: 276 seconds)
1023 2013-02-23 21:27:06 grau has quit (Read error: Operation timed out)
1024 2013-02-23 21:27:54 nus has joined
1025 2013-02-23 21:37:52 fishfish has quit (Quit: Leaving...)
1026 2013-02-23 21:38:46 TD has joined
1027 2013-02-23 21:42:45 eoss has joined
1028 2013-02-23 21:43:04 ashams has quit (Ping timeout: 252 seconds)
1029 2013-02-23 21:46:04 sgornick has quit (Ping timeout: 245 seconds)
1030 2013-02-23 21:48:24 sgornick has joined
1031 2013-02-23 21:50:15 Oswa has quit (Ping timeout: 245 seconds)
1032 2013-02-23 21:52:12 TD has quit (Ping timeout: 255 seconds)
1033 2013-02-23 21:52:14 RBecker is now known as rbecker
1034 2013-02-23 21:53:12 fishfish has joined
1035 2013-02-23 22:02:57 J6So4W7Yb has joined
1036 2013-02-23 22:07:20 xorgate has quit (Read error: Connection reset by peer)
1037 2013-02-23 22:07:40 xorgate has joined
1038 2013-02-23 22:09:22 treb has quit (Ping timeout: 276 seconds)
1039 2013-02-23 22:09:39 treb has joined
1040 2013-02-23 22:10:01 random_cat has quit (Ping timeout: 276 seconds)
1041 2013-02-23 22:10:32 phungus has quit (Ping timeout: 264 seconds)
1042 2013-02-23 22:11:40 phungus has joined
1043 2013-02-23 22:15:38 brwyatt is now known as Away!~brwyatt@brwyatt.net|brwyatt
1044 2013-02-23 22:25:32 freakazoid has quit (Ping timeout: 244 seconds)
1045 2013-02-23 22:39:06 Diapolo has joined
1046 2013-02-23 22:41:38 <jaakkos> how about implementing a block/tx explorer in bitcoind/bitcoin-qt? the answer is probably no, but i think it's worth some discussion...
1047 2013-02-23 22:42:19 <jaakkos> maybe the reference client encourages users to trust random sites that may or may not provide correct information
1048 2013-02-23 22:42:45 <Scrat> not easy to store and process all those indices
1049 2013-02-23 22:43:01 <sipa> jaakkos: against: bitcoind should try to do to much, and just try to be the reference client; pro: people might otherwise use centralized solutions for that
1050 2013-02-23 22:43:01 <jaakkos> yeah, that is a point
1051 2013-02-23 22:43:10 <sipa> i've suggested it before, and it wouldn't be that much work actually
1052 2013-02-23 22:43:17 <sipa> *should not try
1053 2013-02-23 22:43:44 <Scrat> sipa: wouldn't be much work? hmm
1054 2013-02-23 22:43:47 <sipa> only an address->txpos index is missing
1055 2013-02-23 22:44:05 <sipa> (which i've already implemented, but not completely to my liking0
1056 2013-02-23 22:44:17 <jaakkos> (i am not yet familiar with bitcoind internal database structures)
1057 2013-02-23 22:44:56 <sipa> you could already write a limited one, which can't search by address or show the transactions affecting a given address
1058 2013-02-23 22:45:14 <sipa> but otherwise be able to search/show/browse through blocks and transactions
1059 2013-02-23 22:45:27 <jaakkos> perhaps this kind of thing should be in entirely different process to provide strong isolation to actual bitcoin operation
1060 2013-02-23 22:45:37 <jaakkos> perhaps another binary.
1061 2013-02-23 22:45:51 <sipa> oh, it could be implemented as a simple python script using the RPC interface
1062 2013-02-23 22:45:54 <sipa> right now, even
1063 2013-02-23 22:46:05 <jaakkos> just query stuff from bitcoind? well, ok
1064 2013-02-23 22:46:20 <jaakkos> i was thinking something like a small web server
1065 2013-02-23 22:46:25 <sipa> yes, sure
1066 2013-02-23 22:47:20 freakazoid has joined
1067 2013-02-23 22:48:56 <jaakkos> so atm bitcoind offers everything necessary but some lookups are expensive atm?
1068 2013-02-23 22:49:25 <sipa> it just can't search by address
1069 2013-02-23 22:49:32 <sipa> it can search by txid if you have txindex enabled
1070 2013-02-23 22:49:48 <sipa> (which is off by default in 0.8)
1071 2013-02-23 22:50:02 <jaakkos> ok
1072 2013-02-23 22:51:24 <jaakkos> i came across https://github.com/jtobey/bitcoin-abe
1073 2013-02-23 22:51:29 eoss has quit (Read error: Connection reset by peer)
1074 2013-02-23 22:52:15 <Scrat> Abe parses the blockchain and populates an SQL db
1075 2013-02-23 22:52:44 <jaakkos> also https://github.com/bitcoinjs/node-bitcoin-explorer
1076 2013-02-23 22:52:56 <jaakkos> yeah, not too nice
1077 2013-02-23 22:54:00 <Scrat> it aint bad, SQL can give you the ability to make complex queries which would be bothersome otherwise
1078 2013-02-23 22:54:03 bitafterbit has quit (Read error: Connection reset by peer)
1079 2013-02-23 22:54:26 <sipa> yeah, it's very flexible for queries
1080 2013-02-23 22:55:02 <sipa> but the data isn't such a good fit imho, it's more structured (txouts/txn/blocks) than relational
1081 2013-02-23 22:55:29 <sipa> so you need to break it up to some degree to make it accessible through sql
1082 2013-02-23 22:56:56 <jaakkos> the idea of duplicating databases doesn't sound too good to me
1083 2013-02-23 22:56:59 Ramokk has quit ()
1084 2013-02-23 22:57:33 <sipa> well, once pruning would be implemented in the reference client, it makes sense to move rich indexed blockchain data to another process entirely
1085 2013-02-23 22:57:49 <sipa> as bitcoind won't be storing the blockchain anymore
1086 2013-02-23 22:58:10 <jaakkos> reference client will do SPF by default in the future?
1087 2013-02-23 22:58:29 <sipa> SPV you mean, and no
1088 2013-02-23 22:58:37 <sipa> well maybe, but that's not related to pruning
1089 2013-02-23 22:58:47 <sipa> SPV means you just don't validate the blockchain
1090 2013-02-23 22:58:58 <sipa> you can validate it, but not store it
1091 2013-02-23 22:59:07 ThomasV has quit (Ping timeout: 252 seconds)
1092 2013-02-23 22:59:38 <jaakkos> (yes, SPV - i've spent too much time with mail servers lately...)
1093 2013-02-23 23:01:05 <sipa> it's sort of: validate transactions? (yes: store transactions? (yes: full node, no: pruning node)) (no: SPV node)
1094 2013-02-23 23:03:12 <jaakkos> so pruning nodes receive the transactions and perhaps broadcast them further?
1095 2013-02-23 23:03:18 dazla has quit (Quit: Page closed)
1096 2013-02-23 23:03:20 <jaakkos> but don't store them in the long run?
1097 2013-02-23 23:03:24 <sipa> indeed
1098 2013-02-23 23:03:50 kirill42 has joined
1099 2013-02-23 23:04:56 <HM> eto's posts on bitcointalk are epic
1100 2013-02-23 23:06:09 <jaakkos> then, how does a node verify that they really received coins if they can't look up the input tx and even their neighbors can't provide the merkle branch?
1101 2013-02-23 23:06:35 <jaakkos> i think some of my questions might be answered in the wiki...
1102 2013-02-23 23:06:40 <sipa> a pruning (but verifying) node does maintain a full UTXO database
1103 2013-02-23 23:06:49 freakazoid has quit (Ping timeout: 255 seconds)
1104 2013-02-23 23:07:04 <sipa> so it knows which outputs are unspent
1105 2013-02-23 23:07:09 <sipa> and it can look them up, locally
1106 2013-02-23 23:07:19 <sipa> it just doesn't remember the full transactions those outputs came from
1107 2013-02-23 23:07:38 <jaakkos> ah okay unspent db.
1108 2013-02-23 23:08:28 <jaakkos> btw, does SPV rely on neighbors in that they would tell the SPV node if a received transaction was double spent?
1109 2013-02-23 23:08:37 <jaakkos> i should read https://en.bitcoin.it/wiki/Thin_Client_Security
1110 2013-02-23 23:08:53 <sipa> no, it relies on the best chain not containing any double spends
1111 2013-02-23 23:08:59 grau has joined
1112 2013-02-23 23:10:09 <jaakkos> hmm. what if an attacker sends coins from tx1->tx2 from&to his own address
1113 2013-02-23 23:10:45 <jaakkos> then the attacker sends tx1->tx3 to alice (tx3 is reclaimable by alice)
1114 2013-02-23 23:11:10 <jaakkos> then why can't a neighbor choose not to tell about tx2? alice would only ask the merkle branch of tx1, right?
1115 2013-02-23 23:11:25 <sipa> you never 'ask' for a merkle branch
1116 2013-02-23 23:11:27 <jaakkos> i think it's because alice doesn't accept tx3 unconfirmed
1117 2013-02-23 23:11:30 <sipa> you here about it
1118 2013-02-23 23:11:50 <sipa> and with SPV you shouldn't ever ever ever think about accepting unconfirmed transactions (unless you trust the sender)
1119 2013-02-23 23:11:54 Diapolo has left ()
1120 2013-02-23 23:12:23 <sipa> SPV relies on transactions be in the best chain, and validating this chain
1121 2013-02-23 23:12:38 <sipa> without chain, there is no security guarantee at all
1122 2013-02-23 23:13:16 grau has quit (Ping timeout: 252 seconds)
1123 2013-02-23 23:14:09 <HM> jaakkos: a SPV node basically checks the apparently authenticity of a tx (like looking at the hologram or watermark on a banknote), and then the age of a transaction, and if it's old enough trusts that it must be correct because the network would have chucked it away long ago if it wasn't
1124 2013-02-23 23:14:24 <HM> *apparent
1125 2013-02-23 23:15:21 <sipa> and accepting an unconfirmed transaction in SPV would be trusting a client who tells you "yeah, you don't know this sort of watermark yet, because it's brand new, but it's very trustworthy!"
1126 2013-02-23 23:15:46 <HM> shiny = trustworthy :)
1127 2013-02-23 23:16:05 <jaakkos> do the SPV neighbors send the merkle branches for txes who produce a hit in the bloom filter?
1128 2013-02-23 23:16:12 <sipa> yes
1129 2013-02-23 23:16:58 <HM> sipa: i guess the more connections an SPV node has the better right?
1130 2013-02-23 23:17:04 <jaakkos> and the bloom filter is of all addresses (pubkeyhashes) owned by the SPV?
1131 2013-02-23 23:17:10 <jaakkos> ok, less qeustions, more wiki...
1132 2013-02-23 23:17:19 <sipa> jaakkos: read BIP 37 specifically
1133 2013-02-23 23:17:29 Ramokk has joined
1134 2013-02-23 23:19:47 valparaiso is now known as valparaiso_afk
1135 2013-02-23 23:20:26 <jaakkos> about the pruning nodes...
1136 2013-02-23 23:20:41 <jaakkos> what about bootstrap? all data is required at that time?
1137 2013-02-23 23:21:11 <sipa> a pruning full node still needs to validate all history
1138 2013-02-23 23:21:19 <sipa> in theory
1139 2013-02-23 23:22:27 <jaakkos> i thought 'pruning' and 'full' were complementary :)
1140 2013-02-23 23:22:55 <HM> sipa: checking block frequency seems advisable to an SPV node
1141 2013-02-23 23:22:58 <sipa> s/full/fully verifying/
1142 2013-02-23 23:23:27 <jaakkos> okay. and what about helping neighbors to bootstrap...
1143 2013-02-23 23:23:50 <jaakkos> a new node needs to find (rare) full nodes?
1144 2013-02-23 23:23:59 <sipa> you need an archive of history
1145 2013-02-23 23:24:11 <petertodd> blockchainbymail.com
1146 2013-02-23 23:24:16 <sipa> nice thing: the archive doesn't really need to be a bitcoin node or anything
1147 2013-02-23 23:24:22 <sipa> it's just raw data
1148 2013-02-23 23:24:33 <sipa> and the internet these days is pretty good at storing and transferring bulk raw data
1149 2013-02-23 23:25:02 <sipa> of course, you can copy the UTXO database of anything node, if you *really* trust it
1150 2013-02-23 23:25:12 <jaakkos> yes. but likely that data won't be 100% up-to-date, so there is some time window that needs to be fixed somehow?
1151 2013-02-23 23:25:29 <HM> If an SPV node waits for 10 blocks, but doesn't check the time, then surely a SPV with a hijacked connection could be fed blocks produced by a malicious network with say 10% of the hash power and would still accept the tx after 17 hours.
1152 2013-02-23 23:25:30 <petertodd> nodes will have to maintain at least some archival history to handle re-orgs
1153 2013-02-23 23:25:38 <sipa> i expect pruning nodes to keep some minimal amount of recent history around
1154 2013-02-23 23:25:44 <petertodd> so getting history to the last checkpoint or something shouldn't be too hard
1155 2013-02-23 23:25:54 <sipa> (personally: "really trust" for me here would mean 'i operate it myself')
1156 2013-02-23 23:26:11 <jaakkos> damn - i though UTXO was for "micro transactions" or something ;(
1157 2013-02-23 23:26:25 <sipa> utxo = unspent transaction output database
1158 2013-02-23 23:26:28 <petertodd> UTXO *proofs* help scalability, but that's another thing
1159 2013-02-23 23:26:45 <sipa> petertodd: UTXO proofs also make miners less scalable
1160 2013-02-23 23:26:51 <petertodd> basially it's a hard-forking change where nodes must include a merkle tree hash of the UTXO set
1161 2013-02-23 23:26:58 <sipa> or the requirement to build them
1162 2013-02-23 23:27:11 <petertodd> sipa: yes, another reason for small blocks... At least it's an O(1) thing with radix trees.
1163 2013-02-23 23:27:24 unknown45682 has joined
1164 2013-02-23 23:27:27 <petertodd> sipa: O(1) per tx in a block that is
1165 2013-02-23 23:27:57 <sipa> petertodd: it's not, technically
1166 2013-02-23 23:28:07 <sipa> as the radix tree will overflow one day
1167 2013-02-23 23:28:22 <petertodd> sipa: yeah, yeah O(log2(N)) but the definition of Bitcoin is for SHA256, so...
1168 2013-02-23 23:28:22 <sipa> (at which time also tx lookups and PoW break down...)
1169 2013-02-23 23:28:26 <sipa> sure
1170 2013-02-23 23:28:53 <sipa> just pointing out that describing asymptotic behaviour in the presence of hard limits is actually meaningless :)
1171 2013-02-23 23:29:10 <petertodd> sipa: I'm more worried about photon decay
1172 2013-02-23 23:29:58 <sipa> ha
1173 2013-02-23 23:30:29 CodeShark has quit (Remote host closed the connection)
1174 2013-02-23 23:33:26 <petertodd> (I wonder what's the poincare recurrence time for never gonna give you up to reappear in the blockchain due to random statistical fluctuations?)
1175 2013-02-23 23:33:51 <amiller> what do you mean radix tree
1176 2013-02-23 23:34:03 <petertodd> http://en.wikipedia.org/wiki/Radix_tree
1177 2013-02-23 23:35:13 <HM> patricia tree is a nicer name
1178 2013-02-23 23:35:35 <petertodd> radix didn't dump me in highschool
1179 2013-02-23 23:35:40 <HM> all data structures should be named after significant others
1180 2013-02-23 23:35:46 <amiller> those are poorer choices for committing to a UTXO tree than a balanced tree
1181 2013-02-23 23:35:48 ovidiusoft has quit (Ping timeout: 244 seconds)
1182 2013-02-23 23:35:49 <HM> like judy arrays, that guy knows how to name a datastructure
1183 2013-02-23 23:36:03 <amiller> and the argument that one is O(1) while the other is O(log N) is entirely incoherent
1184 2013-02-23 23:36:15 <petertodd> amiller: Why? Worst case in radix is nicely bounded, and it'd dead simple to implement.
1185 2013-02-23 23:36:38 <petertodd> amiller: Balanced trees are effectively O(1) too with the same argument, but the implementation is more complex.
1186 2013-02-23 23:36:54 <amiller> the implementation is not more complex and in either case this is about a committed protocol rather than an implementation
1187 2013-02-23 23:37:05 <petertodd> amiller: You also want the tree to be deterministic from the set of UTXO's, rather than the order in whcih they were created.
1188 2013-02-23 23:37:11 <amiller> the worst case in radix is a constant factor worse than the worst case with a balance dtree
1189 2013-02-23 23:37:37 <amiller> petertodd, there's no reason to want that if you're committing to the whole tree in each block, which is the point of UTXO merkle tree commitment in a block anyway
1190 2013-02-23 23:37:41 <petertodd> amiller: Protocols have to be implemented, and because it's a consensus issue it must be done correctly.
1191 2013-02-23 23:38:17 Hashdog has joined
1192 2013-02-23 23:38:18 <amiller> a balanced binary tree is not "more difficult to implement" than a radix tree and that's a silly reason to argue in favor of a worse binding protocol!
1193 2013-02-23 23:38:20 <petertodd> amiller: Why require order info if you don't have to? Makes bootstrapping from existing UTXO info easier.
1194 2013-02-23 23:38:47 <amiller> bootstrapping from existing UTXO info is also a poor reason to implement a worse protocol for something we're stuck with afterwards
1195 2013-02-23 23:39:00 ovidiusoft has joined
1196 2013-02-23 23:39:05 <amiller> it's not much harder than validating the chain from scratch which you have to do anyway
1197 2013-02-23 23:39:10 <amiller> unless you want to download someone's index
1198 2013-02-23 23:39:13 <HM> this is the address->unspent tx hash database right?
1199 2013-02-23 23:39:20 <amiller> HM, yes
1200 2013-02-23 23:39:31 <petertodd> amiller: Ok, so name a balanced tree algorithm?
1201 2013-02-23 23:39:37 <amiller> petertodd, 2-3 tree
1202 2013-02-23 23:39:47 <HM> is this about an in memory datastructure?
1203 2013-02-23 23:39:55 <amiller> it's not about an in memory data structure
1204 2013-02-23 23:40:02 <petertodd> First of all, 2-3 trees lead to larger UTXO proofs, the real reaosn we want UTXO in the first place.
1205 2013-02-23 23:40:10 <amiller> because the point of this is that you can be a full-validator without storing even the entire UTXO in memory
1206 2013-02-23 23:40:11 <petertodd> Binary trees are the best here.
1207 2013-02-23 23:40:32 <amiller> petertodd, 2-3 trees are isomorphic to red black tree it wouldn't make any difference
1208 2013-02-23 23:40:35 <petertodd> No, that's just one reason, being able to prove the state of the UTXO set to other nodes is extremely valuable for a whole lot of reasons.
1209 2013-02-23 23:40:55 <petertodd> Ah, isomorphic, so where on wiki is that explained?
1210 2013-02-23 23:41:11 <amiller> http://stackoverflow.com/questions/8765689/how-are-red-black-trees-isomorphic-to-2-3-4-trees
1211 2013-02-23 23:41:12 <petertodd> (look, I'm being delibrately stupid here, the solution should be implementable by people who aren't that smart)
1212 2013-02-23 23:41:32 <amiller> http://paloma.eng.tau.ac.il/~yash/infosec-seminar/2005/naor-nissim-1998.pdf also if you're interested, this is the paper that basically describes this technique
1213 2013-02-23 23:41:46 <petertodd> And now you've linked me to an article, which goes to a paper which... You see what I mean? It's not a weekend thing.
1214 2013-02-23 23:41:50 <amiller> it's from 1998 and it was intended as a way of maintaining an outsourced set of PKI certificates
1215 2013-02-23 23:42:04 <amiller> petertodd, this is about a protocol change, not some particular implementation
1216 2013-02-23 23:42:23 <HM> I've been reading etotheipi's ultimate blockchain compression post
1217 2013-02-23 23:42:28 <HM> I'm not really following the motivation
1218 2013-02-23 23:42:29 <amiller> petertodd, i made a prototype in python months ago and now i'm prototyping a much better reference implementation in haskel
1219 2013-02-23 23:42:52 valparaiso_afk is now known as valparaiso
1220 2013-02-23 23:43:00 <petertodd> Yes I know, I'm impressed, yadda yadda. The point is you've created something brillent... and will get fucked up.
1221 2013-02-23 23:43:24 <amiller> i'm not telling you it's brilliant, i'm trying to tell you how it can help
1222 2013-02-23 23:43:27 <petertodd> UTXO *proofs* require code to be written over and over again for all sorts of implementations, not just in the Bitcoin codebase, so making writing those implementations as easy as possible, with really easy protocols.
1223 2013-02-23 23:43:38 <amiller> petertodd, i'm making a library that lets you write an algorihtm once
1224 2013-02-23 23:43:44 <petertodd> And I'm telling you how radix, or hell, pure prefix trees are *good* enough.
1225 2013-02-23 23:43:44 <amiller> and it automatically generates the different parts of the protocol
1226 2013-02-23 23:44:00 <petertodd> Yes, in fucking Haskell. Hey, it's a lovely idea, but I don't see it scaling.
1227 2013-02-23 23:44:07 <petertodd> (scaling as in programmer time)
1228 2013-02-23 23:44:15 phma has joined
1229 2013-02-23 23:44:17 <amiller> whether it's in haskell or not is irrelevant for deciding on a protocol....
1230 2013-02-23 23:44:28 <HM> petertodd is right, you should write it in Lisp so it's accessible
1231 2013-02-23 23:44:54 <petertodd> Yes it is. The protocol needs to be implemented multiple times over, so if you can write it easilly in Haskel, because Haskell is an incredibly pwoerful language, you're writing something too complex.
1232 2013-02-23 23:45:01 <amiller> it's a lot easier to right any implementation of a protocol when you have a correct reference implementation
1233 2013-02-23 23:45:20 <petertodd> If you understand that reference implementation. 99% of programmers don't get functional programming.
1234 2013-02-23 23:45:21 <amiller> petertodd, getting this stuff right is already complex and no one has made even a toy version of a Merkleized radix tree
1235 2013-02-23 23:45:47 <petertodd> Yes, which makes me think a merkleized radix tree is the most complex we can hope for.
1236 2013-02-23 23:45:57 <amiller> it's the merkleized parrt that's hard
1237 2013-02-23 23:45:59 <amiller> not the radix tree
1238 2013-02-23 23:45:59 <petertodd> Incidentally, I've got a toy version of a prefix tree I'm nearly done with so.
1239 2013-02-23 23:46:01 <jaakkos> do some of you guys develop bitcoin full-time?
1240 2013-02-23 23:46:04 <amiller> petertodd, is it merkleize
1241 2013-02-23 23:46:08 <amiller> is it a radix hash tree
1242 2013-02-23 23:46:25 <petertodd> Yes, the prefix tree has a fixed order, which makes merkleizing it trivial.
1243 2013-02-23 23:46:29 toffoo has joined
1244 2013-02-23 23:46:30 <petertodd> Same for the radix tree version.
1245 2013-02-23 23:46:59 <petertodd> The prefix tree winds up with more boiler plate, because you then need per-txout (rather than per-tx+vout#) information.
1246 2013-02-23 23:47:47 <petertodd> IE in the latter, your key is just txhash and a separate per-tx vout thing, while the latter can handle txout|vout fine and with one tree modification per tx changed.
1247 2013-02-23 23:48:02 <amiller> do you think it would be easy to modify it so that you support address->utxo mappings
1248 2013-02-23 23:48:27 <amiller> in addition to just tx:idx->txout
1249 2013-02-23 23:48:30 <petertodd> Very easy. address->utxo is just a separate thing, using the same prefix|radix tree class.
1250 2013-02-23 23:48:45 <petertodd> tx:idx->txout isn't a UTXO thing
1251 2013-02-23 23:48:59 kirill42 has quit (Ping timeout: 245 seconds)
1252 2013-02-23 23:48:59 <amiller> ok well cool :)
1253 2013-02-23 23:49:04 <petertodd> Oh, wait, you mean tx:idx -> scriptPubKey?
1254 2013-02-23 23:49:08 <amiller> if you make one, i'll make a compatible version
1255 2013-02-23 23:49:12 <amiller> yeah tx:idx->scriptpubkey
1256 2013-02-23 23:49:21 <amiller> really you want scriptpubkey+length+something else
1257 2013-02-23 23:49:26 <petertodd> I also want to implement merkle sums and stuff too, which is really important for a lot of things. Makes large-blocks possible too.
1258 2013-02-23 23:49:27 <amiller> whatever you need to process the script
1259 2013-02-23 23:49:38 <amiller> petertodd, okay well cool.
1260 2013-02-23 23:49:39 ovidiusoft has quit (Ping timeout: 255 seconds)
1261 2013-02-23 23:49:40 <petertodd> Yeah, same info returned by gettxout basically.
1262 2013-02-23 23:50:04 <amiller> if it's easier to do it with a radix tree then that's not too big a problem i agree
1263 2013-02-23 23:50:17 <petertodd> As I say, don't get me wrong, your work is brilliant, but I'm the guy without the comp-sci degree, and if even I can understand it, I think it's a better solution *if* it works.
1264 2013-02-23 23:51:02 * HM wonders if the wiki has a glossary
1265 2013-02-23 23:51:07 <amiller> i don't expect anyone to use my library i just plan to use it to make simpler arguments about which protocols are safe etc.
1266 2013-02-23 23:51:23 <amiller> so if you prefer to implement something in C++, hopefully i can model it my way and get the same root hashes!
1267 2013-02-23 23:51:52 <petertodd> Oh I'm sure you will, probably in like 5 lines of Haskell...
1268 2013-02-23 23:51:54 <amiller> also keep in mind that once you pick the protocol you can make more efficient implementations later
1269 2013-02-23 23:52:02 <amiller> suppose this goes well
1270 2013-02-23 23:52:03 <petertodd> Exactly
1271 2013-02-23 23:52:09 <amiller> and a year from now everyone is using merkle radix trees
1272 2013-02-23 23:52:41 <amiller> and now the clients say hey this would be a bit faster if ...... nvm that wouldn't apply anyway it's really only a somewhat unlikely worst case
1273 2013-02-23 23:52:46 <petertodd> One thing in particular that worries me, is I know for fidelity bonded banks' TPM hardware, you'll have all this semi-consensus code that absolutely must work...
1274 2013-02-23 23:53:43 <petertodd> Heck, as it is, I thought radix trees would be a weekend job, but I found myself running into bugs with the jsonrpc library even. (isn't closing sockets it seems)
1275 2013-02-23 23:54:29 <amiller> wait i put it back in
1276 2013-02-23 23:54:38 <amiller> radix trees are usually going to be a constant factor slower
1277 2013-02-23 23:54:54 <HM> hmm
1278 2013-02-23 23:54:55 <amiller> especially if the addresses are long enough
1279 2013-02-23 23:55:01 <amiller> so a year frm now if everyone is using this
1280 2013-02-23 23:55:07 <petertodd> Why exactly?
1281 2013-02-23 23:55:09 <amiller> they're going to say man we could have 5x faster clients if we just used binary trees
1282 2013-02-23 23:55:28 <petertodd> Radix can stop when the prefix is unique.
1283 2013-02-23 23:55:29 <amiller> because it's O(log 2^k) to insert in a radix tree
1284 2013-02-23 23:55:35 <amiller> where k is your security factor
1285 2013-02-23 23:56:06 <amiller> whereas the balanced tree is only O(log M)
1286 2013-02-23 23:56:10 <amiller> where M is the number of elements in the tree
1287 2013-02-23 23:56:14 <amiller> notice that M is substantially smaller than 2^k
1288 2013-02-23 23:56:23 <amiller> otherwise you would have an intolerable risk of collisions
1289 2013-02-23 23:57:00 <petertodd> But as I say, addresses are essentially randomly distributed, with huge cost to attackers to make them non-randomly distributed, so it's very close to as good.
1290 2013-02-23 23:57:38 <amiller> uh hm.
1291 2013-02-23 23:58:10 <HM> hmm
1292 2013-02-23 23:58:14 <petertodd> Like, say there are n honest transactions in the UTXO set, for an attacker to extend the longest unique prefix tree, they need to create n+log2(m) transactions.
1293 2013-02-23 23:58:23 <amiller> hm, i suppose that's true
1294 2013-02-23 23:58:25 <petertodd> (er, extend tree by m)
1295 2013-02-23 23:58:38 rbecker is now known as RBecker
1296 2013-02-23 23:58:39 <amiller> i guess for the prefix or patricia tree then that would be just as good
1297 2013-02-23 23:58:39 <petertodd> And for what, to make it slightly, slightly slower?
1298 2013-02-23 23:58:42 <petertodd> Not a big deal.
1299 2013-02-23 23:59:06 <petertodd> Yeah, radix is nicer, given you can just use txhash:vout as a key, but prefix just means more boilerplate.
1300 2013-02-23 23:59:34 <petertodd> (hell, hash(txhash:vout) works in the prefix case... but then the general usage of a new tx with a lot of txouts is silly, may not matter though)