1 2013-07-18 02:39:38 <imton> guys gettxout
  2 2013-07-18 02:39:47 <imton> is only for unspent outputs?
  3 2013-07-18 02:44:21 <LLckfan> Hello
  4 2013-07-18 02:44:38 <imton> hi
  5 2013-07-18 05:12:19 <Luke-Jr> TheUni: I don't think the current stuff is holding up the merge
  6 2013-07-18 06:17:33 <warren> gmaxwell: hmm, have you looked into the type of deterministic keys is trezor?
  7 2013-07-18 06:21:33 <lianj> getrawtransaction also looks into the memorypool, right?
  8 2013-07-18 06:22:56 <Scrat> lianj: of course
  9 2013-07-18 06:24:15 <sipa> lianj: yes, but not in the wallet
 10 2013-07-18 06:24:32 <sipa> warren: i thoiught they'd use bip32?
 11 2013-07-18 06:24:37 <sipa> maybe not only
 12 2013-07-18 06:26:47 <warren> Is the concept of type 1 or type 2 obsolete now?
 13 2013-07-18 06:33:07 <lianj> thanks guys
 14 2013-07-18 06:33:27 <sipa> warren: no, bip32 does both
 15 2013-07-18 06:33:35 <warren> ah
 16 2013-07-18 06:43:45 <TD> sipa: question .... i thought i understood ecdsa but apparently not
 17 2013-07-18 06:43:59 <TD> sipa: yesterday i discovered that signatures produced by an uncompressed key don't seem to validate against a compressed key
 18 2013-07-18 06:44:07 <TD> is that expected or indicative of some other screwup in my code?
 19 2013-07-18 06:44:14 <TD> i thought they were equivalent
 20 2013-07-18 06:44:33 <TD> (though that code was kind of broken at the time so it's possible it was being misparsed, or whatever)
 21 2013-07-18 06:47:46 <warren> is that related to uncompressed address != compressed address?
 22 2013-07-18 06:47:51 <warren> same privkey
 23 2013-07-18 06:50:27 <_ingsoc> Is there any way I can ask if someone wants to get paid to work on Bitcoin code without getting laughed out of the room?
 24 2013-07-18 06:55:23 <TD> warren: no. addresses weren't involved here.
 25 2013-07-18 06:55:40 <TD> _ingsoc: why would people laugh at that? if you're recruiting the forums might work better though
 26 2013-07-18 06:56:08 <_ingsoc> I don't know. People get funny about these things.
 27 2013-07-18 06:56:53 <sipa> TD: the signatures should be interchangeable
 28 2013-07-18 06:57:00 <TD> yeah. that's what i thought.
 29 2013-07-18 06:57:06 <TD> there must have been some other issue
 30 2013-07-18 06:57:18 <sipa> TD: it's just the encoding of the public key that is different, the EC point is the same
 31 2013-07-18 06:57:23 <TD> yeah
 32 2013-07-18 06:57:29 <sipa> TD: or are you talking about message signatures?
 33 2013-07-18 06:57:34 <TD> no
 34 2013-07-18 06:57:36 <TD> tx sigs
 35 2013-07-18 06:57:40 <TD> must have been some other bug.
 36 2013-07-18 06:57:46 <sipa> in that case it shouldn't matter
 37 2013-07-18 07:44:04 <t7> is an elliptic curve over a binary field the same thing as over a prime field ?
 38 2013-07-18 07:48:53 <TD> same thing in what way? obviously they're not exactly the same thing
 39 2013-07-18 07:49:59 <t7> is it just an implementation thing?
 40 2013-07-18 07:50:06 <t7> do they use the same keys ?
 41 2013-07-18 07:55:00 <sipa> no
 42 2013-07-18 07:55:07 <sipa> keys are curve specific
 43 2013-07-18 07:55:30 <sipa> and if the field is different, the curve certainly is
 44 2013-07-18 07:55:52 <alexwaters> reading https://en.bitcoin.it/wiki/Wallet_import_format. is there a doc on how bitcoinQT generates the initial private key? i am going to just do a 2xsha256 of a string, but I'm wondering if there is something "better"
 45 2013-07-18 07:58:30 <sipa> it just generates a random 32-byte sequence
 46 2013-07-18 08:01:30 <jchp> alexwaters: you probably don't want to generate your keys using the same algorithm which has a global network trying to find collisions for (i believe they are called "bitcoin miners")
 47 2013-07-18 08:02:23 <alexwaters> jchp: how should I generate them if I want to use a deterministic seed?
 48 2013-07-18 08:02:40 <jchp> use sha-512 or whatever it doesn't matter just don't use two rounds of sha-256
 49 2013-07-18 08:02:49 <alexwaters> ok cool
 50 2013-07-18 08:03:44 <alexwaters> sipa: thank you
 51 2013-07-18 08:03:55 <sipa> alexwaters: read BIP32
 52 2013-07-18 08:04:23 <alexwaters> oh snap :)
 53 2013-07-18 08:04:24 <TD> alexwaters: seeds are meant to be random data
 54 2013-07-18 08:05:28 <sipa> they're mostly meant to be secret
 55 2013-07-18 08:05:42 <sipa> being non-random makes that harder to guarantee
 56 2013-07-18 08:07:39 <matjeh> alexwaters: i've used sha256(dictionary_word) as a private key, sent some BTC to the corresponding address, to see how fast it would be stolen. it took 3 seconds.
 57 2013-07-18 08:08:03 <matjeh> so be careful
 58 2013-07-18 08:08:36 <coingenuity> matjeh: lol
 59 2013-07-18 08:11:17 <alexwaters> good to know. is a sha512(random_base64_21char_passphrase) as a private key decent?
 60 2013-07-18 08:11:55 <matjeh> i capitalised 50% of the letters and added a 3 digit number on the end, and did the same. and its still there after a month
 61 2013-07-18 08:12:03 <matjeh> so probably
 62 2013-07-18 08:12:07 <sipa> alexwaters: define 'random'
 63 2013-07-18 08:12:08 <matjeh> for a while anyway
 64 2013-07-18 08:12:12 <matjeh> ;)
 65 2013-07-18 08:12:49 <alexwaters> sipa: generated by keypass or 1password - so not perfectly random
 66 2013-07-18 08:12:52 <sipa> if you mean "uniformly random", sure, but just random_32_byte_sequence is even better
 67 2013-07-18 08:13:55 <sipa> or can you really remember a 21 character randomly generated passphrase?
 68 2013-07-18 08:14:02 <matjeh> just make sure its < 2^256 - 2^32 - 2^9 - 2^8 ... i forget the rest of the prime
 69 2013-07-18 08:14:45 <alexwaters> sipa: i definitely cannot. i am useless if they ever waterboard me for my passwords
 70 2013-07-18 08:14:57 <sipa> alexwaters: so you have to store it somewhere anyway
 71 2013-07-18 08:15:08 <sipa> alexwaters: so store just the private key itself
 72 2013-07-18 08:15:58 <alexwaters> it's more of a UX thing. I want to collect a username and a password from users and generate pub/priv for them. the password requirements are very strict though
 73 2013-07-18 08:16:07 <sipa> DO NOT
 74 2013-07-18 08:16:10 <sipa> PLEASE
 75 2013-07-18 08:16:16 <alexwaters> yikes
 76 2013-07-18 08:16:40 <alexwaters> due to security of the priv key storage? or lack of entropy?
 77 2013-07-18 08:16:58 <sipa> due to people being very bad at judging randomness
 78 2013-07-18 08:17:08 <sipa> and you can have some rules that prevent the worst offenders
 79 2013-07-18 08:17:13 <sipa> but people will find ways around it
 80 2013-07-18 08:17:38 <sipa> i don't consider myself capable of coming up with a passphrase that is strong enough to store money IN
 81 2013-07-18 08:18:02 <alexwaters> what if I make a random salt to keep them from being bad?
 82 2013-07-18 08:18:10 <sipa> then what's the point?
 83 2013-07-18 08:18:23 <sipa> it will be unrememberable if it is secure
 84 2013-07-18 08:18:59 <sipa> (that's not entirely true, if you have 128 bits of true entropy it's fine)
 85 2013-07-18 08:19:12 <sipa> but that still means something inconveniently large
 86 2013-07-18 08:19:19 <coingenuity> alexwaters: why not just use the seed of username/password to generate enough entropy to create private keys from
 87 2013-07-18 08:19:46 <alexwaters> this is more of a proof of concept, which will be o/s and improved upon i'm sure. it's purely experimental - i'm not actually standing up any servers for a long while of peer review
 88 2013-07-18 08:19:46 <sipa> ...
 89 2013-07-18 08:20:33 <coingenuity> i.e. hash usernames against passwords via some one-way algo to give you enough randomness that the bitcoin privkey isnt easily obtainable via dictionary bruting
 90 2013-07-18 08:21:12 <sipa> i don't see how you're adding any entropy?
 91 2013-07-18 08:21:36 <alexwaters> concatinating the username makes is less rainbow-attackable
 92 2013-07-18 08:21:54 <alexwaters> but still a clever targeted attack has no added entropy from that
 93 2013-07-18 08:22:20 <sipa> alexwaters: ever saw https://bitcointalk.org/index.php?topic=102349.0 ?
 94 2013-07-18 08:22:56 <alexwaters> negative, it looks like I have a great deal of reading to do
 95 2013-07-18 08:23:16 <sipa> i don't think it's used anywhere, but i still think it's a pretty neat idea :)
 96 2013-07-18 08:23:20 <coingenuity> sipa: i mean, you hash "usernamepassword" via whirlpool and it yeilds d88b382a0af2ff2a8d1d419d4be2bc70b60e43661089bbeda382f3e6cfe5375b75a3065ad2d82c1c48869fb49838d5978b967d5d4636cb2ae8398d56bc60e426, you can grab some bytes of that to generate a privkey seed
 97 2013-07-18 08:23:33 <alexwaters> i think once this is put out there, it will make more sense - and i'm not actually jeopardizing any users. it's just a play thing for a while
 98 2013-07-18 08:23:45 <sipa> coingenuity: you understand the concept of "entropy" ?
 99 2013-07-18 08:24:14 <coingenuity> sipa: yeah, i'm saying that adding another
100 2013-07-18 08:24:23 <sipa> but you're not adding entropy at all
101 2013-07-18 08:24:25 <coingenuity> 'unknown' step to the mix increases randomness
102 2013-07-18 08:24:35 <sipa> how is the step unknown?
103 2013-07-18 08:24:40 <sipa> does it use a random source?
104 2013-07-18 08:24:48 <sipa> if it does, it's useless as a brainwallet
105 2013-07-18 08:24:55 <sipa> if it doesn't, it doesn't add entropy
106 2013-07-18 08:24:57 <coingenuity> because nobody on the network, aside from alexwaters, would know the steps that yeild the privkey seed
107 2013-07-18 08:25:10 <sipa> that's just security through obscurity
108 2013-07-18 08:25:27 <sipa> you can't assume people don't know your algorithm
109 2013-07-18 08:25:29 <coingenuity> i know, bad way to go about it, but it provides a more random seed from his desired starting inputs
110 2013-07-18 08:25:38 <sipa> no, it is not "more random"
111 2013-07-18 08:25:52 <sipa> it is exactly as random as the space of usernames and passwords
112 2013-07-18 08:26:00 <sipa> it may be harder to guess
113 2013-07-18 08:26:01 <alexwaters> sipa: alright to PM?
114 2013-07-18 08:26:03 <sipa> yes
115 2013-07-18 08:26:11 <sipa> but it's not more secure
116 2013-07-18 08:26:42 <coingenuity> sipa: yes, nothing is more secure than 100% random, but that's not quite his goal
117 2013-07-18 08:27:01 <SomeoneWeird> pfft 100% random
118 2013-07-18 08:27:06 <lianj> SomeoneWeird: haha yea
119 2013-07-18 08:27:08 <QuantumQrack> Anybody know which file to back up if you are using multibit?
120 2013-07-18 08:27:20 <coingenuity> it's just degrees of safeness that we're really talking about, in terms of generating privkey seeds from 'known' data
121 2013-07-18 08:27:40 <sipa> coingenuity: there are much better ways to do that than using a secret algorithm
122 2013-07-18 08:29:28 <Scrat> I use Pi digits for my privkeys
123 2013-07-18 08:40:10 <alexwaters> Scrat: I want to use my cat's brainwave patterns
124 2013-07-18 08:55:10 <lianj> is sendrawtransaction returning -22 if its already in memorypool fixed in .3 now?
125 2013-07-18 09:14:11 <jouke> In my debug log I see: "Misbehaving 1 address from <ip> (0->0)" Does anyone know what that means?
126 2013-07-18 09:15:02 <sipa> just something invalid was detected, but which doesn't trigger the DoS score to go up
127 2013-07-18 09:15:13 <sipa> the log message could be clearer
128 2013-07-18 09:17:29 <jouke> What kind of behaviour could trigger such a thing?
129 2013-07-18 09:17:45 <sipa> look at the line before :)
130 2013-07-18 09:23:17 <Diablo-D3> gmaxwell: >Elric brothers
131 2013-07-18 09:23:18 <Diablo-D3> gmaxwell: :3
132 2013-07-18 09:24:59 <t7> i roll my dice 3 times to get 3 bytes of randomness
133 2013-07-18 09:25:04 <t7> 4 times*
134 2013-07-18 09:26:03 <t7> hang on the sum of 2 dice is more likely to be 6/7 than other values ...
135 2013-07-18 09:26:08 <t7> right?
136 2013-07-18 09:26:21 <matjeh> they'd have to be 16-sided to fill the entropy needed, then
137 2013-07-18 09:26:49 <t7> i can flip a coin for each bit
138 2013-07-18 09:27:04 <matjeh> wtf am i talking about, 256-sided
139 2013-07-18 09:27:24 <t7> i think 16 is ok, because you can append the nibbles
140 2013-07-18 09:27:32 <t7> rather than addition and changing base
141 2013-07-18 09:27:47 <matjeh> you'd have to roll a 6-sided die 10 times
142 2013-07-18 09:28:13 <matjeh> for 3 bytes
143 2013-07-18 09:28:39 <matjeh> or 6 dices once. parallel computation.
144 2013-07-18 09:28:58 <sipa> t7: roll two dice, compute dice_1+dice_2*6-7
145 2013-07-18 09:29:05 <sipa> t7: if result is >= 32, repeat
146 2013-07-18 09:29:10 <sipa> if not, you have 5 bits of randomness
147 2013-07-18 10:12:26 <t7> are you sure...
148 2013-07-18 11:17:42 <sipa> t7: yes
149 2013-07-18 11:18:06 <t7> oh i had order of operations wrong
150 2013-07-18 11:18:12 <t7> thats makes sense
151 2013-07-18 11:21:15 <t7> sipa is a maths wizard
152 2013-07-18 11:43:08 <helo> i thought he's the kind of wizard that morphs into turtles, rabbits, caterpillars, walruses, voles, pox, etc
153 2013-07-18 11:52:38 <t7> he maths so much math he should be called matth-ew
154 2013-07-18 11:53:08 <sipa> you must be talking about BlueMatt
155 2013-07-18 11:53:27 <BlueMatt> heh, sadly I dont maths that much
156 2013-07-18 11:55:17 <bmcgee> hey guys i'm having difficulties understanding the hex representations for target. Admittedly it seems my hex knowledge has run out. In these 2 examples 0x1b0404cb, 0x1d00ffff could someone explain the significance of the 0x1b and 0x1d portions
157 2013-07-18 11:56:37 <t7> i think its a bespoke encoding for an into
158 2013-07-18 11:56:38 <t7> int*
159 2013-07-18 11:56:45 <t7> its on the wiki i think
160 2013-07-18 11:56:53 <bmcgee> yeah im looking through the wiki
161 2013-07-18 11:57:08 <bmcgee> when unpacking the target the 1d and 1b portions are significant in the multiplier
162 2013-07-18 11:59:06 <t7> last_three_bytes * 2**(8*(firstbyte - 3))
163 2013-07-18 11:59:11 <t7> it looks like
164 2013-07-18 12:00:00 <bmcgee> yup just want to be sure, in case i'm doing something automagical
165 2013-07-18 13:08:29 <xavier23> Anybody know how big (in bytes) an individual transaction is on average? And also, is there a limit on the size?
166 2013-07-18 13:30:47 <xavier23> If anybody's interested.. the following might be a good approximation: txin,txout: 60 bytes + 139 bytes (scriptSig) + 25 (pk_script)
167 2013-07-18 13:33:33 <xavier23> Oh wow. Actually it seems a little more complicated than that
168 2013-07-18 13:33:52 <xavier23> Just talking to myself here??? lol
169 2013-07-18 13:38:01 <sipa> xavier23: look at a random block?
170 2013-07-18 13:38:03 <sipa> http://blockchain.info/block-index/399685/000000000000009343cac50275e4d75c390f2969cf5c3185974265d665e03a47
171 2013-07-18 13:38:35 <sipa> 161, 258, 258, 376, 257, 818, 259, 225, 966, 226, 227, 257, 293, ...
172 2013-07-18 13:40:25 <sipa> xavier23: having just a single txin and a single txout is very rare
173 2013-07-18 13:44:45 <xavier23> sipa: Ah thanks. I didn't realize they have `size` on blockchain.info.
174 2013-07-18 14:09:14 <xavier23> After a series of calculations, I estimate avg tx size: 0.3768kb. PM me if you 'd like an explanation... lol
175 2013-07-18 14:11:28 <sipa> xavier23: more common, that value would be expressed as '376.8 bytes'
176 2013-07-18 14:13:05 <gmaxwell> calculations? if you're talking about the average presumably you just went and measured the recent blocks?
177 2013-07-18 14:13:08 <Luke-Jr> sipa: I get 48.2384 bytes :p
178 2013-07-18 14:14:23 <xavier23> gmaxwell: Yes, I measured the one given by sipa
179 2013-07-18 14:15:14 <xavier23> Block Size / N Tx
180 2013-07-18 14:16:08 <xavier23> Then I made some adjustments
181 2013-07-18 14:18:51 <sipa> "add 3, subtract cosine of the phase of the moon, multiply by zero, add the square root of the block size"
182 2013-07-18 14:29:53 <xavier23> sipa: Lol
183 2013-07-18 14:30:03 <xavier23> Something like that??? ;)
184 2013-07-18 14:30:52 <xavier23> A believe the phrase is: there's lies, damn lies and then there's statistics
185 2013-07-18 15:44:40 <jedunnigan> ~ how large is the testnet blockchain atm?
186 2013-07-18 17:26:50 <BlueMatt> gavinandresen: any idea why the jenkins server got reset?
187 2013-07-18 17:28:35 <gmaxwell> BlueMatt: needed to reboot for the kernel rootkit to take effect.
188 2013-07-18 17:28:41 <BlueMatt> ahhhhhh
189 2013-07-18 17:31:05 <BlueMatt> (if anyone was wondering why pull-tester died 7 days ago... 21:26:22 up 7 days, 19:29,  1 user,  load average: 0.55, 0.16, 0.06                                           )
190 2013-07-18 20:12:09 <sipa> ;;nethash
191 2013-07-18 20:12:10 <gribble> 223553.981326
192 2013-07-18 20:12:28 <gribble> 139.721238329
193 2013-07-18 20:12:28 <sipa> ;;calc [nethash]*0.015/24
194 2013-07-18 20:12:37 <sipa> crap
195 2013-07-18 20:13:08 <sipa> the network hashing rate now grows more _per hour_ then it was in total when i mined my first block...
196 2013-07-18 20:13:27 <sipa> *than
197 2013-07-18 20:15:07 <nsh> them cycles...
198 2013-07-18 20:40:50 <gavinandresen> BlueMatt: jenkins server reset a few days ago?  Scheduled hardware maintenance, I think