1 2013-06-02 01:09:39 <Rebroad> Did anyone else see this? 2013-06-02 02:44:56 InvalidChainFound: invalid block=000000000000001dc876eb55c4ee75bbf2ae499519c3731c77d4294f4d6430e9  height=239189  log2_work=70.200444  date=2013-06-02 02:44:44
  2 2013-06-02 04:51:39 <warren> sipa: I have commits for secp256k1 gitian, want pulls?
  3 2013-06-02 04:52:07 <warren> sipa: I want to know how you handle src/secp256k1, submodule?  subtree like leveldb?
  4 2013-06-02 04:53:14 <warren> sipa: currently I have it in a separate tar file that extracts to src/secp256k1, but that's probably wrong.
  5 2013-06-02 05:12:43 <sipa> warren: for now, send them as pullreqs to sipa/bitcoin.git branch secp256k1
  6 2013-06-02 05:13:38 <gribble> The operation succeeded.
  7 2013-06-02 05:13:38 <sipa> ;;later tell oleganza CPrivKey is how private keys used to be serialized in the wallet, and it's retained for backward compatibility
  8 2013-06-02 05:16:38 <warren> sipa: OK.  Should I leave /src/secp256k1 out entirely?  no subtree or anything?
  9 2013-06-02 05:16:59 <warren> sipa: my gitian recipes won't work unless something copies the source in during gitian
 10 2013-06-02 05:23:17 <sipa> oh, i see
 11 2013-06-02 05:24:33 <warren> this I've been wondering how you want to handle src/secp256k1, like the new leveldb subtree?
 12 2013-06-02 05:24:39 <warren> thus*
 13 2013-06-02 05:25:26 <sipa> that seems like a good solution, but i'm not familiar with git-subtree (yet)
 14 2013-06-02 05:25:37 <sipa> i guess for now it
 15 2013-06-02 05:25:45 <sipa> i guess for now it's easy enough to just copy the source in
 16 2013-06-02 05:47:45 <oleganza> sipa: so if I write my own wallet from scratch, I can ignore CPrivKey?
 17 2013-06-02 06:03:55 <Aurigae> anybody here? Im suspicious on the bitcoin community and how it deals with scammers.
 18 2013-06-02 06:03:55 <Aurigae> Im mostly busy but it starts to provoke me, not bitclockers and its scammyness but the not caring by the bitcoin community, you hurt new users - who get dissapointed and hurt bitcoin, it contra productive
 19 2013-06-02 06:03:55 <Aurigae> I tried to rgister with the directory but still cant edit, cant figure an option todo so.
 20 2013-06-02 06:03:55 <Aurigae> SRSLY  speakin, there are plenty of options to pwn a site, from reverse SEO, community actions to preventing listings is popular directories.
 21 2013-06-02 06:03:55 <Aurigae> Why can bitclockers operate, why does nobody care? They have 1st pos in google for bitcoin mining pool! and still appear on https://en.bitcoin.it/wiki/Comparison_of_mining_pools .
 22 2013-06-02 06:04:39 <Aurigae> <--- just finished ranting
 23 2013-06-02 06:53:58 <gribble> The operation succeeded.
 24 2013-06-02 06:53:58 <sipa> ;;later tell oleganza yes, you don't need the CPrivKey's in that case
 25 2013-06-02 08:25:51 <sf92ug> hey guys
 26 2013-06-02 08:27:15 <sf92ug> just wanted to ask why 0.8.2 is not tagged yet and why it shows "beta" in the startup screen and not on the bitcoin.org website
 27 2013-06-02 08:27:57 <warren> sipa: copying the source in isn't how gitian wants it.  I guess I'll just put the secp256k1 git tree into a tarball in gitian/inputs/ until we have a better way.
 28 2013-06-02 08:28:26 <warren> sipa: the leveldb commit for 0.9 that uses subtree isn't the right solution here?
 29 2013-06-02 08:29:43 <BlueMatt> 
ACTION still votes for an option to compile secp256k1 as a simple openssl wrapper, merge it into bitcoin mainline and let releases use ssl-wrapper mode for a while

 30 2013-06-02 08:40:13 <warren> BlueMatt: I'd like that too.
 31 2013-06-02 10:01:53 <BlueMatt> warren: have you done a reimport with secp256k1 recently?
 32 2013-06-02 10:08:44 <BlueMatt> sipa: ^
 33 2013-06-02 12:32:58 <Vinnie_win> What does secp256k1 and openssl have to do with each other??
 34 2013-06-02 12:36:03 <Scrat> probably meant sipa's secp256k1 code
 35 2013-06-02 12:36:13 <Vinnie_win> Okay but what's that got to do with OpenSSL?
 36 2013-06-02 12:36:27 <Vinnie_win> OpenSSL is for transport layer security, while secp256k1 is for signature verification?
 37 2013-06-02 12:36:33 <Scrat> openssl is currently handling this part
 38 2013-06-02 12:36:57 <Vinnie_win> So an openssl wrapper lets calling code utilize the new routines transparently?
 39 2013-06-02 12:38:18 <Scrat> I guess so
 40 2013-06-02 14:20:10 <graingert> aw yeah
 41 2013-06-02 14:20:16 <graingert> I have myself an IPv6
 42 2013-06-02 14:20:20 <graingert> (some)
 43 2013-06-02 14:25:30 <graingert> oh my hidden host has hidden it
 44 2013-06-02 14:25:31 <graingert> oh well
 45 2013-06-02 15:12:54 <sipa> warren: i've used git-subtree to make my 'secp256k1' branch of bitcoin include the libsecp256k1 source tree
 46 2013-06-02 15:13:38 <Diablo-D3> woah wtf
 47 2013-06-02 15:13:44 <Diablo-D3> I just switched to 0.8.2
 48 2013-06-02 15:13:55 <Diablo-D3> I had bitcoin-qt off for a couple days
 49 2013-06-02 15:13:59 <Diablo-D3> why is it syncing so fast
 50 2013-06-02 15:16:01 <maaku> Diablo-D3: i wish my 0.8.2 was syncing faster
 51 2013-06-02 15:17:36 <sipa> maaku: what is the bottleneck?
 52 2013-06-02 15:18:13 <nsh> Diablo-D3, the NSA made some code improvments
 53 2013-06-02 15:18:20 <maaku> sipa: probably network, but honestly I haven't profiled
 54 2013-06-02 15:19:19 <maaku> I'm still on a ~500kbps DSL despite living in the middle of silicon valley
 55 2013-06-02 15:19:45 <Diablo-D3> wtf.
 56 2013-06-02 15:20:03 <Diablo-D3> if you're in SF, you can get 1gbps for like $25/mo or some shit
 57 2013-06-02 15:20:06 <gonffen> maaku: how much do you pay for that?
 58 2013-06-02 15:20:19 <maaku> $50/mo + taxes
 59 2013-06-02 15:20:23 <maaku> Diablo-D3: San Jose :(
 60 2013-06-02 15:20:30 <gonffen> ouch
 61 2013-06-02 15:21:24 <gonffen> I had dsl from verizon for a while. It was 1000/386, but I only paid ~$20/mo.
 62 2013-06-02 15:21:42 <Diablo-D3> I have 3/768 atm and I live in maine.
 63 2013-06-02 15:21:48 <Diablo-D3> and I pay $35/mo.
 64 2013-06-02 15:22:02 <maaku> well eventually Sonic will roll out gigabit for almost the same price: http://arstechnica.com/tech-policy/2012/02/gigabit-internet-for-80-the-unlikely-success-of-californias-sonicnet/
 65 2013-06-02 15:22:08 <maaku> they just haven't hit my neighborhood yet :(
 66 2013-06-02 15:22:59 <sipa> maaku: were you at the conference?
 67 2013-06-02 15:23:11 <maaku> yeah
 68 2013-06-02 15:23:19 <sipa> can't remember meeting you :)
 69 2013-06-02 15:23:35 <maaku> i'm not sure we ran into each other :(
 70 2013-06-02 15:23:53 <maaku> i was on the alt-chain panel, and attended most of the tech talks
 71 2013-06-02 15:23:53 <sipa> and if it's network, you mean bandwidth or bitcoin's peer selection mechanism?
 72 2013-06-02 15:24:16 <sipa> then we've likely been in the same room at least!
 73 2013-06-02 15:24:28 <nsh> maaku, are you freicoin?
 74 2013-06-02 15:24:28 <sipa> afk
 75 2013-06-02 15:24:37 <maaku> yes
 76 2013-06-02 15:24:40 <nsh> how is the new difficulty algo working out?
 77 2013-06-02 15:24:58 <maaku> sipa: just bandwidth - i max out when syncing
 78 2013-06-02 15:25:15 <nsh> (would be interested in some analysis once there's enough data)
 79 2013-06-02 15:25:56 <maaku> nsh: this is it's response so far: http://i.imgur.com/98Whj4r.png
 80 2013-06-02 15:26:14 <gonffen> maaku: $40/mo is right around the sweet spot for me. Do you know if it is 100/100?
 81 2013-06-02 15:26:34 <nsh> maaku, hmmm
 82 2013-06-02 15:27:02 <maaku> gonffen: it's about 100/40 (down/up) - the Sonic tech was cool and let me set my own ratio :)
 83 2013-06-02 15:27:23 <gonffen> heh that's awesome :)
 84 2013-06-02 15:28:10 <maaku> nsh: when the nethash randomly swings up or down 10x due to profitability, there's very little you can do :\\
 85 2013-06-02 15:28:32 <nsh> maaku, indeed
 86 2013-06-02 15:28:57 <maaku> the filter is tuned for steady-state, long term behavior, using bitcoin nethash data as the input
 87 2013-06-02 15:29:19 <maaku> but so far it's at least kept us from having to do any more drastic measures or hard forks, despite 10x nethash swings
 88 2013-06-02 15:29:27 <nsh> this is good
 89 2013-06-02 15:30:09 <nsh> i wonder if more work on merged mining tech and recruitment could help reduce altcoin nethash variances
 90 2013-06-02 15:31:48 <nsh> i'm not sure there's even consensus that this would be a Good Thing
 91 2013-06-02 15:33:36 <ProfMac> This miner and difficulty variance is happening with bytecoin now.  Blocks are as slow as 1 per day, and difficculty schedule to drop to 10% of it's current value.
 92 2013-06-02 15:34:18 <ProfMac> I wondered what the implication would be of making the difficulty be based on the past 1600 blocks, rather than evaluated after 1600 blocks.
 93 2013-06-02 15:43:07 <TheLordOfTime> anyone here want to scan their node to see if a specific transaction shows up in the blockchain?
 94 2013-06-02 15:44:15 <warren> sipa: git subtree work well for that purpose?
 95 2013-06-02 15:46:08 <TheLordOfTime> better question, is there a way to check the blockchain on my node to find a specific transaction?
 96 2013-06-02 15:46:17 <TheLordOfTime> using bitcoin-qt's RPC console in the debug window
 97 2013-06-02 15:47:58 <warren> sipa: hmm, I wonder if I can avoid git subtree or manual copying in the shell script with by adding another section to remotes:
 98 2013-06-02 15:50:00 <maaku> TheLordOfTime: getrawtransaction
 99 2013-06-02 15:53:33 <TheLordOfTime> maaku:  can i scan for transactions that aren't mine in that manner too?
100 2013-06-02 15:57:48 <BlueMatt> sipa: have you ever sync'd past 238797 on libsecp256k1?
101 2013-06-02 15:58:08 <BlueMatt> (last friday's block)
102 2013-06-02 15:58:20 <realazthat> TheLordOfTime: https://en.bitcoin.it/wiki/Original_Bitcoin_client/API_Calls_list
103 2013-06-02 15:58:29 <realazthat> I dunno about the debug window, never used it
104 2013-06-02 15:58:39 <realazthat> but you can run bitcoind on the commandline
105 2013-06-02 15:59:03 <realazthat> like, ./bitcoind gettransaction <txid>
106 2013-06-02 15:59:37 <TheLordOfTime> realazthat:  i thin kthey're the same, just that one's the RPC console for it
107 2013-06-02 15:59:42 <realazthat> TheLordOfTime: as for tx's that aren't yours, I am not sure if you can do that by default
108 2013-06-02 15:59:45 <realazthat> there is a txindex option
109 2013-06-02 15:59:53 <realazthat> in your conf file
110 2013-06-02 15:59:55 <realazthat> txindex=1
111 2013-06-02 15:59:59 <realazthat> or somesuch
112 2013-06-02 16:00:15 <realazthat> it requires you to run -reindex once
113 2013-06-02 16:00:23 <realazthat> I dunno if you can do that through the UI
114 2013-06-02 16:00:47 <realazthat> but once you reindex it should work in the debug window, if it is indeed the same
115 2013-06-02 16:05:01 <TheLordOfTime> realazthat:  if getrawtransaction returns nothing does that mean the transaction hasn't been picked up by any bitcoin nodes yet, or that it hasn't been broadcasted yet?
116 2013-06-02 16:06:28 <gonffen> it doesn't necessarily mean either of those things
117 2013-06-02 16:06:57 <gonffen> just because some nodes know about it doesn't mean your node does
118 2013-06-02 16:07:35 <realazthat> TheLordOfTime: like I said, I am not sure if you have to have txindex on, but also, it might not have reached you yet
119 2013-06-02 16:07:45 <realazthat> but it is supposed to spread really quick though
120 2013-06-02 16:08:49 <realazthat> This release no longer maintains a full index of historical transaction ids
121 2013-06-02 16:08:50 <realazthat> by default, so looking up an arbitrary transaction using the getrawtransaction
122 2013-06-02 16:08:50 <realazthat> details).
123 2013-06-02 16:08:50 <realazthat> RPC call will not work. If you need that functionality, you must run once
124 2013-06-02 16:08:50 <realazthat> with -txindex=1 -reindex=1 to rebuild block-chain indices (see below for more
125 2013-06-02 16:09:07 <realazthat> from bitcoin-0.8.0 release notes
126 2013-06-02 16:09:36 <realazthat> TheLordOfTime: ^^
127 2013-06-02 16:11:06 <TheLordOfTime> bleh
128 2013-06-02 16:11:41 <TheLordOfTime> realazthat:  i might do that later, main issue is that the transaction hasn't shown up in blockchain.info or blockexplorer, and  is sitting "unconfirmed" in the wallet
129 2013-06-02 16:11:55 <TheLordOfTime> so i'm trying to figure out if it's just not picked up in the blockchain yet, or if something else is going on
130 2013-06-02 16:12:03 <BlueMatt> dust output?
131 2013-06-02 16:12:21 <realazthat> mmm dunno :/
132 2013-06-02 16:12:24 <TheLordOfTime> no, ~1BTC.
133 2013-06-02 16:12:45 <BlueMatt> what about change value?
134 2013-06-02 16:13:15 <TheLordOfTime> BlueMatt:  its possible to see that on a receive transaction?
135 2013-06-02 16:13:30 <TheLordOfTime> (it was received and has been sitting as unconfirmed, i didn't send the btc)
136 2013-06-02 16:13:41 <BlueMatt> are you on 0.8.2?
137 2013-06-02 16:19:19 <Julius129> how secure is "verifymessage" should it be considered for a login or authentication scheme?
138 2013-06-02 16:20:30 <TheLordOfTime> BlueMatt:  mhm
139 2013-06-02 16:20:35 <TheLordOfTime> BlueMatt:  doesn't matter, it's in the chain now
140 2013-06-02 16:33:12 <Vinnie_win> So it looks like the market absorbed that big sale and all the hubbaballoo is over
141 2013-06-02 16:54:40 <BlueMatt> petertod1: nope, good now, thanks
142 2013-06-02 16:55:35 <petertod1> BlueMatt: np
143 2013-06-02 16:56:10 <BlueMatt> found another bug in bitcoinj script verification :(
144 2013-06-02 16:56:30 <BlueMatt> this time because values were being serialized as unsigned integers, when they're really signed
145 2013-06-02 16:56:39 <petertodd> BlueMatt: ha lovely
146 2013-06-02 16:56:48 <petertodd> BlueMatt: did sipa mention the possible one I found to you?
147 2013-06-02 16:56:55 <BlueMatt> in bitcoinj?
148 2013-06-02 16:56:55 <BlueMatt> no
149 2013-06-02 16:57:49 <petertodd> good, he probably found a reason it didn't matter
150 2013-06-02 16:58:06 <ecoloco> Why is the course so unstable right now? It has fallen from 135$-140$, to 120$ now in just a few days?
151 2013-06-02 16:58:18 <BlueMatt> ecoloco: wrong channel
152 2013-06-02 16:58:40 <ecoloco> BlueMatt: what channel should I ask in?
153 2013-06-02 16:58:51 <BlueMatt> probably #bitcoin-otc
154 2013-06-02 16:58:58 <BlueMatt> or really #bitcoin-pricetalk
155 2013-06-02 16:59:15 <chmod755> ^^^
156 2013-06-02 17:02:59 <nsh> BlueMatt, repercussions?
157 2013-06-02 17:03:13 <nsh> (of script validation bug)
158 2013-06-02 17:06:00 <BlueMatt> nsh: anyone who is using bitcoinj full validation would have failed to validate the chain a few days ago
159 2013-06-02 17:06:22 <BlueMatt> nsh: but since bitcoinj full verification is still very much beta, its discouraged and I dunno of too many people using it
160 2013-06-02 17:06:22 <nsh> oh, right, that the block you mentioned earlier
161 2013-06-02 17:06:26 <nsh> 
ACTION nods

162 2013-06-02 17:06:30 <nsh> *at
163 2013-06-02 17:06:51 <BlueMatt> (the only time negative values are serialized as such is in SIGHASH_SINGLE, so any non-verification tasks should be fine)
164 2013-06-02 17:06:57 <petertodd> for most more mundane bitcoinj uses, merchants and stuff, I assume that probably wouldn't have affected anything right?
165 2013-06-02 17:08:19 <runeks> WOO! I *finally* figured out how to use tc and iptables to limit traffic to and from port 8333.
166 2013-06-02 17:08:28 <runeks> Let me know if you are interested in the script.
167 2013-06-02 17:08:43 <runeks> Means that I can have bitcoind running at all times, without worrying about it consuming all my outbound bandwidth.
168 2013-06-02 17:09:07 <nsh> runeks, worth posting imho
169 2013-06-02 17:09:12 <BlueMatt> yes, unless you call script verification on a SIGHASH_SINGLE script, nothing is effected
170 2013-06-02 17:09:29 <BlueMatt> (and even then some _SINGLE signatures arent effected)
171 2013-06-02 17:10:39 <petertodd> runeks: consider writing it up as a pull-req to go into the contrib/ directory in bitcoin-qt
172 2013-06-02 17:13:42 <runeks> nsh: Here you go: http://pastebin.com/BqvMnqAE
173 2013-06-02 17:13:55 <nsh> ty
174 2013-06-02 17:14:08 <nsh> dem incantations!
175 2013-06-02 17:14:30 <runeks> It also makes sure not to limit LAN traffic. So if you have an always-online bitcoind instance running, and you connect your local bitcoind to that, traffic won't be limited to you.
176 2013-06-02 17:15:16 <runeks> BEWARE
177 2013-06-02 17:15:21 <runeks> I forgot to say this
178 2013-06-02 17:15:32 <runeks> It deleted ALL existing rules in the iptables OUTPUT chain
179 2013-06-02 17:15:59 <runeks> Was mostly used for testing though. Perhaps I should remove that.
180 2013-06-02 17:16:20 <runeks> nsh: What do you mean by "incantations?"
181 2013-06-02 17:17:00 <runeks> petertodd: I think I will do that.
182 2013-06-02 17:17:22 <runeks> Took me a while to figure out
183 2013-06-02 17:18:28 <petertodd> runeks: Thanks. Don't be surprised when it gets a bunch of criticism though. :) Take the opportunity to get all the bugs out of the script.
184 2013-06-02 17:19:36 <nsh> runeks, linux commands of a mystical and abstruse nature
185 2013-06-02 17:19:37 <runeks> petertodd: We'll see. I can't see any bugs for now. Mostly because it's really simple. Two classes. One unlimited (1gbit), one limited. Traffic to/from port 8333, but not to a LAN address, is limited. End of story.
186 2013-06-02 17:19:40 <nsh> (iptables)
187 2013-06-02 17:19:46 <runeks> nsh: Ah :)
188 2013-06-02 17:19:49 <petertodd> runeks: Sounds good. Write it up!
189 2013-06-02 17:20:18 <petertodd> runeks: Include a decent README or similar to document how it works too - with firewalls people are probably not going to run your exact script.
190 2013-06-02 17:20:20 <runeks> nsh: It makes a lot more sense when you read through the guides. Took me a while though. Precisely because it looks so cryptic on first inspection.
191 2013-06-02 17:20:51 <runeks> petertodd: I'll do that. But the script itself contains a lot of comments too,
192 2013-06-02 17:21:15 <petertodd> runeks: Oh, yeah comments in the script are the perfect way to do it.
193 2013-06-02 17:22:15 <nsh> runeks, indeed. i just have trouble remembering any of the things i learnt from one instance of reading the manpage to another
194 2013-06-02 17:22:58 <runeks> nsh: I never learn by manpages. Or at least only to refine what I've learned through simpler guides. I think the tc/iptables manpages would take a week to read (and understand).
195 2013-06-02 17:38:41 <nsh> runeks, indeed
196 2013-06-02 17:43:11 <nsh> has there been much research into address cache pollution attacks?
197 2013-06-02 17:46:06 <runeks> Done: https://github.com/bitcoin/bitcoin/pull/2728
198 2013-06-02 17:52:42 <nsh> Luke-Jr, how are tonal numbers going to save the world and give me eternal caviar and felatio?
199 2013-06-02 17:52:51 <nsh> 
ACTION is somewhat confused

200 2013-06-02 17:53:17 <Luke-Jr> nsh: trolling is off-topic here
201 2013-06-02 17:53:50 <nsh> sorry, i was being somewhat serious, but you're right it's not ontopic here
202 2013-06-02 17:54:08 <nsh> 
ACTION finishes reading posts/wiki

203 2013-06-02 17:54:59 <gwillen> nsh: you also misspelled fellatio
204 2013-06-02 17:55:27 <nsh> 
ACTION performs penance

205 2013-06-02 18:35:51 <BlueMatt> ;;bc,blocks
206 2013-06-02 18:35:51 <gribble> 239338
207 2013-06-02 18:38:52 <dansmith_btc> Hi, I followed the instructions ( https://gist.github.com/gavinandresen/3966071 )  of creating a 2-of-3 address, then first sending BTC to it and then spending them.  However the spend-from-multisig tx is not shown on blockchain.info. What happened to it?
208 2013-06-02 18:38:57 <dansmith_btc> txid 9079a33a220afd75cfaf4d155d1a14f655aab4e781f99a3c405921389c7e9864
209 2013-06-02 18:42:00 <petertodd> dansmith_btc: blockchain.info doesn't show multisig until they are confirmed
210 2013-06-02 18:42:14 <dansmith_btc> petertodd, thanks
211 2013-06-02 18:50:35 <warren> Hmm, the 0.7 backport with the May 15th hardfork isn't listed on bitcoin.org
212 2013-06-02 18:50:37 <warren> was it unofficial?
213 2013-06-02 18:53:27 <Luke-Jr> warren: ?
214 2013-06-02 18:54:03 <Luke-Jr> backports are for people already using the old versions who can't upgrade; we don't want new users to use them
215 2013-06-02 18:54:55 <warren> Luke-Jr: I know.  I'm just looking for source for backports for the silly alt coin.
216 2013-06-02 19:03:14 <ecoloco> The course is like a roller coaster now
217 2013-06-02 19:06:02 <MC1984_> who on earth could be locked into an old beta version already?
218 2013-06-02 19:08:10 <Luke-Jr> MC1984_: anyone who uses patches to get specific functionality not mainlined
219 2013-06-02 19:21:01 <BlueMatt> ;;bc,blocks
220 2013-06-02 19:21:02 <gribble> 239346
221 2013-06-02 19:21:31 <owowo> hm, I noticed something odd, I just snyced my wallet and now my spend transactions are not shown in the list anymore, but they were showing when I started it
222 2013-06-02 19:21:58 <owowo> Is it meant to be this way?
223 2013-06-02 19:43:24 <passepartout> Hi everyone - I have a question: how might I recover BitCoins from a Change address that is in a corrupted wallet?
224 2013-06-02 19:49:27 <midnightmagic> passepartout: salvagewallet maybe.
225 2013-06-02 19:50:35 <passepartout> OK thanks midnightmagic will try that.
226 2013-06-02 19:53:11 <warren> bitcointalk is down.
227 2013-06-02 19:57:30 <phantomcircuit> passepartout, make sure you keep a backup of course
228 2013-06-02 20:06:23 <nsh> warren, up for me
229 2013-06-02 20:11:14 <warren> ah, back
230 2013-06-02 20:27:32 <dansmith_btc> Hi, is it possible to sign a transaction on an unsynced bitcoind (having the correct priv key, of course)?
231 2013-06-02 20:28:11 <gwillen> dansmith_btc: you need to sync to the point where the coins were received
232 2013-06-02 20:28:22 <gwillen> but note that if some of the coins were later spent
233 2013-06-02 20:28:33 <nsh> you're be shot.
234 2013-06-02 20:28:36 <gwillen> you might get in trouble by signing a transaction that spends coins it doesn't know are already spent.
235 2013-06-02 20:28:36 <nsh> *you'll
236 2013-06-02 20:28:40 <gwillen> right.
237 2013-06-02 20:29:03 <dansmith_btc> wow, I thought signrawtransaction can be used in an offline mode
238 2013-06-02 20:30:46 <dansmith_btc> What is the rationale for not letting just sign a transaction and sending it? If I'm not the owner of the output I signed, the tx will be rejected by the network.
239 2013-06-02 20:30:59 <nsh> it can, but in terms of "doin' it right" for the network, it makes sense to sign transactions only when you have an up-to-date blockchain
240 2013-06-02 20:32:18 <dansmith_btc> nsh, so bitcoind refuses to sign outputs that are not in the wallet then?
241 2013-06-02 20:33:35 <nsh> 
ACTION is not certain, defers to higher authority

242 2013-06-02 20:35:07 <dansmith_btc> OK, I'll launch bitcoind with a clean datadir and check myself
243 2013-06-02 20:36:56 <gmaxwell> dansmith_btc: you have to know the pubkey for a transaction in order to identify the right keys to use and the right script to write
244 2013-06-02 20:37:16 <gmaxwell> dansmith_btc: the sign command allows you to specify the scriptpubkey for the inputs however.
245 2013-06-02 20:37:56 <dansmith_btc> gmaxwell, OK, so you're saying it can be done in an unsynced mode?
246 2013-06-02 20:38:07 <sipa> BlueMatt: hmm, maybe, why?
247 2013-06-02 20:38:14 <gmaxwell> Yes.
248 2013-06-02 20:38:27 <gmaxwell> dansmith_btc: http://people.xiph.org/~greg/signdemo.txt
249 2013-06-02 20:39:53 <nsh> gmaxwell, you should convince/bribe someone to turn that textfile into a pretty-thing
250 2013-06-02 20:40:39 <gmaxwell> I'm perfectly capable of formating it myself, but I doubt anyone who's turned off by the text will actually complete the procedure.
251 2013-06-02 20:41:25 <gmaxwell> s/doubt/doubtful
252 2013-06-02 20:41:30 <nsh> well, it's another matter if you're actively trying to solve a problem. but i've had the page open before and just gone glossybrained from hexstring:english ratio
253 2013-06-02 20:42:10 <nsh> it's comprehensible second-time around with additional motivation
254 2013-06-02 20:42:11 <nsh> 
ACTION shrugs

255 2013-06-02 20:42:55 <gmaxwell> nsh: no matter how many pretty monospaced divs you draw around it??? it'll still be "oh my god, it's full of hex"
256 2013-06-02 20:42:56 <dansmith_btc> gmaxwell, it says there " We need to provide the scriptpubkey of the inputs we're signing so our offline wallet knows which of its keys to use." What if I know already which priv key to use, can I forgo scriptPubKey?
257 2013-06-02 20:42:59 <gmaxwell> :P
258 2013-06-02 20:43:11 <nsh> 
ACTION smiles

259 2013-06-02 20:43:34 <gmaxwell> dansmith_btc: No. It still needs it in order to figure out how to write the script that satisfies it.
260 2013-06-02 20:44:03 <gmaxwell> dansmith_btc: you might be able to get away with copying one from another transaction that pays that address though.
261 2013-06-02 20:44:18 <gmaxwell> (including one you just createrawtransactioned for that express purpose
262 2013-06-02 20:44:19 <gmaxwell> )
263 2013-06-02 20:44:23 <nsh> with a 10% probability of immanentizing the eschaton
264 2013-06-02 20:45:23 <gmaxwell> gp
265 2013-06-02 20:45:25 <gmaxwell> oops
266 2013-06-02 20:45:51 <phantomcircuit> dansmith_btc, what's your use case
267 2013-06-02 20:46:28 <dansmith_btc> trying to create an escrow for users of blockchain.info
268 2013-06-02 20:47:03 <dansmith_btc> the users will have to fall back on bitcoind's RPC in n unsynced mode
269 2013-06-02 20:48:22 <gmaxwell> dansmith_btc: someone should _never_ spend an transaction paying to them without a copy of the transaction. If they're willing to do that you can trick them into diverting most of the funds to fees.
270 2013-06-02 20:49:12 <phantomcircuit> every year charles schwab sends me their annual report
271 2013-06-02 20:49:15 <phantomcircuit> i have 2 shares
272 2013-06-02 20:49:22 <phantomcircuit> collectively worth like
273 2013-06-02 20:49:23 <phantomcircuit> $40
274 2013-06-02 20:49:25 <phantomcircuit> WHYY
275 2013-06-02 20:50:26 <petertodd> BlueMatt: What's the state of re-organization tests right now?
276 2013-06-02 20:51:10 <petertodd> BlueMatt: Looking at my child-pays-for-parent mempool stuff again, and reorgs will be one of the more complex cases.
277 2013-06-02 20:51:44 <dansmith_btc> gmaxwell, ic, but in my case they will be spending their own transaction which is sitting in their blockchain.info wallet (blockchain.info doesn't allow sending BTC into multi-sig addresses, hence this workaround)
278 2013-06-02 20:54:14 <nsh> phantomcircuit, to justify you dressing and acting like a victorian steel magnate, obviously
279 2013-06-02 20:54:15 <gmaxwell> dansmith_btc: "blockchain.info doesn't allow sending BTC into multi-sig addresses" < hm? what does that mean?
280 2013-06-02 20:55:03 <nsh> (unrelated charles schwab, but nevertheless...)
281 2013-06-02 20:55:17 <dansmith_btc> gmaxwell, try sending BTC into addresses like 3JNaZ9A1yMpRTGyoG9Vq9MvvjKcECYbSoo , they won't allow
282 2013-06-02 20:55:24 <phantomcircuit> nsh, lol
283 2013-06-02 20:55:39 <ielo> phantomcircuit, hey
284 2013-06-02 20:56:00 <gmaxwell> dansmith_btc: what?!
285 2013-06-02 20:56:13 <gmaxwell> thats insane!
286 2013-06-02 20:56:28 <phantomcircuit> ielo, hello
287 2013-06-02 20:56:36 <gmaxwell> dansmith_btc: has anyone asked them to fix that?
288 2013-06-02 20:56:58 <petertodd> gmaxwell: I did almost a year ago
289 2013-06-02 20:57:07 <sipa> wait, they don't allow P2SH addresses?
290 2013-06-02 20:57:14 <petertodd> sipa: yup
291 2013-06-02 20:57:19 <sipa> perhaps not remarkable given that they're hardly used...
292 2013-06-02 20:57:25 <petertodd> sipa: most services don't
293 2013-06-02 20:57:36 <phantomcircuit> lots of stuff doesn't...
294 2013-06-02 20:57:39 <gmaxwell> or conversely, not remarkable that they're hardly used.
295 2013-06-02 20:57:52 <petertodd> easywallet is another example that doesn't support them
296 2013-06-02 20:57:58 <nsh> there's this: https://blockchain.info/wallet/escrow
297 2013-06-02 20:58:08 <gmaxwell> dansmith_btc: do you really want to do business with someone who can't be bothered to make a ~two line change to their software in two years?
298 2013-06-02 20:58:10 <petertodd> yup, and why would you use them if they aren't supported...
299 2013-06-02 20:58:48 <nsh> seems to indicate support for multisig transactions, but what do i know
300 2013-06-02 20:59:42 <dansmith_btc> gmaxwell, well it's not me doing business, I'm simply creating a tool which will enable blockchain.info users. But I see your point.
301 2013-06-02 21:00:05 <sipa> dansmith_btc: how will it work?
302 2013-06-02 21:01:08 <dansmith_btc> sipa, a user with a blockchain account and an unsynced bitcoind will create multisig, fund it and redeem it.
303 2013-06-02 21:01:44 <sipa> redeem it how?
304 2013-06-02 21:02:32 <dansmith_btc> sipa, I mean create a multisig address. Redeem when transaction goes through OK, both buyer and seller signrawtransaction  spend-from-multisig
305 2013-06-02 21:03:09 <sipa> how does the b.i user do a signrawtransaction?
306 2013-06-02 21:03:16 <sipa> do they have an API for that?
307 2013-06-02 21:03:21 <dansmith_btc> His bitcoind will do
308 2013-06-02 21:03:36 <dansmith_btc> No, they dont have API for that
309 2013-06-02 21:03:43 <sipa> i'm not following
310 2013-06-02 21:03:51 <sipa> where are the private keys?
311 2013-06-02 21:03:58 <dansmith_btc> it's a complicated scenario, I know
312 2013-06-02 21:04:13 <dansmith_btc> THe private key is both in bitcoind and on b.i
313 2013-06-02 21:04:22 <dansmith_btc> The user importprivkey to b.i
314 2013-06-02 21:05:46 <gribble> The operation succeeded.
315 2013-06-02 21:05:46 <sipa> ;;later tell BlueMatt yes, synced past 238797 just fine on libsecp256k1
316 2013-06-02 21:05:51 <dansmith_btc> I'll write up a bitcointalk post for peer review when I'm finished figuring out the details
317 2013-06-02 21:06:03 <sipa> dansmith_btc: why is b.i involved at all?
318 2013-06-02 21:06:40 <dansmith_btc> sipa, it will eventually cover other lightweight clients. It's just that b.i is kinda most popular
319 2013-06-02 21:06:57 <sipa> no, what's the point of having b.i in the loop if the user needs bitcoind too?
320 2013-06-02 21:07:11 <dansmith_btc> Because bitcoind is UNsynced
321 2013-06-02 21:07:20 <dansmith_btc> it's just a 10MB download
322 2013-06-02 21:07:32 <sipa> ah
323 2013-06-02 21:07:37 <dansmith_btc> So, I'm offering a lightweight way of doing escrow
324 2013-06-02 21:08:40 <sipa> k
325 2013-06-02 21:09:15 <sipa> warren: you don't need to do anything anymore, secp256k1 is in the bitcoin source now
326 2013-06-02 21:09:23 <sipa> warren: in my branch at least
327 2013-06-02 21:09:45 <warren> sipa: ok, should I also modify the remote address to point at your repo temporarily?
328 2013-06-02 21:09:48 <BlueMatt> sipa: yea, sorry
329 2013-06-02 21:09:58 <BlueMatt> I wanted to blame secp256k1 for a bug of mine
330 2013-06-02 21:10:03 <warren> sipa: for the gitian .yml file
331 2013-06-02 21:10:13 <sipa> warren: i always just build from local sources
332 2013-06-02 21:10:21 <sipa> don't use the uri in the repo
333 2013-06-02 21:10:22 <BlueMatt> petertodd: afaik there are no tests for mempool
334 2013-06-02 21:10:27 <BlueMatt> petertodd: PLEASE write one
335 2013-06-02 21:11:07 <warren> sipa: trouble is, gitian uses whatever URL is in the .yml
336 2013-06-02 21:11:21 <sipa> warren: unless you manually put the source tree in inputs/bitcoin
337 2013-06-02 21:12:04 <warren> oh, it bypasses the URL if you do?
338 2013-06-02 21:12:06 <sipa> yes
339 2013-06-02 21:12:11 <warren> and it clones directly from that?
340 2013-06-02 21:12:28 <warren> ok, didn't realize this.
341 2013-06-02 21:12:30 <sipa> i suppose it goes a git checkout <commit> in that dir, and then copies the data over
342 2013-06-02 21:12:37 <sipa> i use a wrapper script
343 2013-06-02 21:12:50 <warren> sorry, dealing with three problems at once
344 2013-06-02 21:13:03 <sipa> http://bitcoin.sipa.be/builds/bitcoin-build.sh.txt
345 2013-06-02 21:13:04 <sipa> :)
346 2013-06-02 21:13:27 <BlueMatt> 
ACTION wonders why the bitcoin 2013 playlist has one that is just "[Private video]"

347 2013-06-02 21:13:51 <sipa> ?
348 2013-06-02 21:13:57 <BlueMatt> what are the devs hiding? hmmm...
349 2013-06-02 21:14:06 <sipa> what? where? link?
350 2013-06-02 21:14:10 <BlueMatt> http://www.youtube.com/playlist?list=PLUOP0P68GJ3BGjfqoLLnzAefk3ZzXQtJ7
351 2013-06-02 21:14:27 <owowo> bitcoinpr0n :P
352 2013-06-02 21:14:33 <sipa> heh
353 2013-06-02 21:14:43 <nsh> did you miss the NSA panel?
354 2013-06-02 21:14:46 <nsh> :)
355 2013-06-02 21:14:48 <BlueMatt> "NSA buys bitcoin"
356 2013-06-02 21:15:02 <BlueMatt> ohhhhh, thats how you pronounce Wuille
357 2013-06-02 21:15:17 <sipa> haha
358 2013-06-02 21:15:37 <BlueMatt> 
ACTION has to admit he wasnt sure how to pronounce Pieter either until mike said it

359 2013-06-02 21:15:52 <BlueMatt> (mostly do you pronounce the i)
360 2013-06-02 21:16:13 <sipa> 'ie' in dutch is a long "ee" sound as in "beep"
361 2013-06-02 21:16:23 <sipa> it's not a diphtongue
362 2013-06-02 21:16:50 <sipa> diphthong, whatever
363 2013-06-02 21:17:13 <BlueMatt> that just went waaay over my head
364 2013-06-02 21:17:22 <BlueMatt> my c is better than my english...and my c aint that great
365 2013-06-02 21:17:58 <sipa> diphtong = pronounciation which gradually goes from one sound to another
366 2013-06-02 21:18:09 <sipa> like 'ay' in 'way'
367 2013-06-02 21:18:14 <BlueMatt> ok...still no idea
368 2013-06-02 21:18:24 <sipa> anyway, doesn't matter
369 2013-06-02 21:18:38 <sipa> just say as it was 'Peter' in English and you're close enough
370 2013-06-02 21:18:49 <BlueMatt> ok, thats what I had been doing...sorry
371 2013-06-02 21:19:02 <sipa> sorry for what?
372 2013-06-02 21:19:09 <BlueMatt> for not getting your name?
373 2013-06-02 21:19:10 <BlueMatt> I dunno
374 2013-06-02 21:19:22 <sipa> if you say 'peter', you're very close :D
375 2013-06-02 21:19:30 <BlueMatt> well, ok :)
376 2013-06-02 21:19:48 <sipa> the t is somewhat harder than in english, and the r is different
377 2013-06-02 21:19:50 <sipa> but that's about it
378 2013-06-02 21:20:15 <BlueMatt> 
ACTION pronounces it with a german accent and probably gets it closer

379 2013-06-02 21:23:07 <nsh> just refer to everyone by individual vulgar expletives
380 2013-06-02 21:23:15 <nsh> it's much better for community relations
381 2013-06-02 21:23:26 <BlueMatt> yes, thats a good idea fucker
382 2013-06-02 21:23:57 <nsh> cheers, cuntcheese :)
383 2013-06-02 21:29:19 <petertodd> petertodd: Yeah, I'm writing a basic one as I go along.
384 2013-06-02 21:30:01 <petertodd> BlueMatt: Yeah, I'm writing a basic one as I go along.
385 2013-06-02 21:31:21 <petertodd> BlueMatt: So do we even have re-org testcases available?
386 2013-06-02 22:17:11 <berndj> has anyone looked at whether the feedback mechanism that holds block generation time close to 10 minutes is stable? (control theory wannabe nerd here)
387 2013-06-02 22:21:23 <gmaxwell> berndj: if you exclude human psychology from the process it's trivially and obviously so, if you don't, well good luck finding the eigenvectors of the minds of the populous.
388 2013-06-02 22:22:14 <berndj> i would model human psychology as noise. not sure if that'd be pink, white, or brown noise
389 2013-06-02 22:22:30 <gmaxwell> that wouldn't be a good model at all.
390 2013-06-02 22:22:42 <berndj> trivially and obviously though? got a nyquist diagram somewhere?
391 2013-06-02 22:23:14 <sipa> we've at least seen a ~3 week-delayed feedback from exchange increases to hash power increases
392 2013-06-02 22:23:20 <sipa> we've seen pool hoppinh
393 2013-06-02 22:23:30 <berndj> sipa, that's exactly the thing i'm wondering about
394 2013-06-02 22:23:44 <sipa> no way you can model human behaviour as uncorrelated
395 2013-06-02 22:23:47 <berndj> 3 weeks is like 270 degrees phase shift
396 2013-06-02 22:24:03 <sipa> it's certainly not the only effect
397 2013-06-02 22:24:04 <gmaxwell> seriously it's a box filter, if you don't count the superduper highly non-linear humans turning on and off their miners, it's just a box filter with a gain of 1 exactly.
398 2013-06-02 22:24:23 <berndj> wait no, it's more: it's 540 degrees - effectively 180
399 2013-06-02 22:24:53 <berndj> gmaxwell, i see what you're getting at at noise being a bad model for humans
400 2013-06-02 22:25:10 <berndj> to clarify: i'd model human psychology as a gain>1 amplifier with some noise added
401 2013-06-02 22:25:24 <maaku> berndj: that'd be a terrible model
402 2013-06-02 22:25:37 <gmaxwell> But that doesn't make any sense, humans are just nonlinear things.
403 2013-06-02 22:26:09 <gmaxwell> and without the humans the system is a simple linear function that instantly depends on the difficulty parameter.
404 2013-06-02 22:26:36 <berndj> continuity is perhaps more interesting than linearity - you can zoom in far enough until you have an approximately linear segment
405 2013-06-02 22:27:03 <maaku> berndj: as soon as you come up with a model of human behavior, assume that another human has done the same thing. and then figured out how to game that system
406 2013-06-02 22:27:10 <maaku> welcome to the social sciences
407 2013-06-02 22:27:44 <gmaxwell> berndj: uh, humans are not continuously differentiable, probably not even differentiable anywhere. :P
408 2013-06-02 22:27:46 <berndj> maaku: that description sounds like a noisy opamp
409 2013-06-02 22:27:58 <gmaxwell> ..
410 2013-06-02 22:28:07 <berndj> gmaxwell: there's an old joke about curly pi the smooth operator...
411 2013-06-02 22:28:09 <maaku> berndj: do opamps have free will?
412 2013-06-02 22:28:47 <gmaxwell> Your toolbox needs some more tools if you think you can just model people as linear devices!  like .. uh ... some non-linear tools.
413 2013-06-02 22:28:48 <berndj> no, but neither do humans, as a collective entity ;)
414 2013-06-02 22:30:43 <berndj> gmaxwell, in what way, specifically and in this context, are humans (collectively) "nonlinear"?
415 2013-06-02 22:31:44 <berndj> if we separate out the whims of fashion as "noise", what we're left with is an economic decision (i think): higher mining profitability -> higher investment in mining equipment
416 2013-06-02 22:32:00 <gmaxwell> ...
417 2013-06-02 22:32:27 <maaku> berndj: if only it was that easy
418 2013-06-02 22:32:40 <berndj> i'm sure it is a nonlinear relationship, but i'm not convinced that it's nonlinear in a way that matters
419 2013-06-02 22:32:50 <gmaxwell> Well good luck to you then.
420 2013-06-02 22:32:55 <SteveDekorte> berndj: you don't think fashion (and all other decisions) are ultimately economic (ie utility towards survival/replication) decisions?
421 2013-06-02 22:33:46 <berndj> SteveDekorte, yes, and granted, i'm using "economic" in a more limited sense here that counts only $$$ (of whatever denomination you like)
422 2013-06-02 22:34:28 <SteveDekorte> berndj: isn't money just a more liquid asset than clothing, art, etc?
423 2013-06-02 22:35:02 <berndj> i hear you objections that humans are more complicated than just gain + noise, but i'm wondering about what characteristics the feedback network has
424 2013-06-02 22:35:23 <berndj> if it would oscillate even if humans were some ideal linear (but noisy) amplifier, i'd say that's a problem
425 2013-06-02 22:35:38 <berndj> SteveDekorte, yes, the rhetorical victory goes to you
426 2013-06-02 22:35:50 <berndj> sadly, it doesn't address what i'm interested in
427 2013-06-02 22:35:59 <nsh> gmaxwell++ (re: good luck finding the eigenvectors of the minds of the populus)
428 2013-06-02 22:36:03 <gmaxwell> What you're saying just makes no sense, and suggests that you haven't even spent more than a moment thinking about it.
429 2013-06-02 22:36:20 <berndj> ok then i'll just stfu now
430 2013-06-02 22:36:25 <berndj> thanks for listening
431 2013-06-02 22:37:00 <SteveDekorte> berndj: I think it's relevant that there are no non-economic decisions wrt human behavior
432 2013-06-02 22:40:32 <warren> sigh.  pools shouldn't copy random code from github, especially if they are PPS...
433 2013-06-02 22:40:44 <gmaxwell> more halarity?
434 2013-06-02 22:40:53 <warren> gmaxwell: yeah.  yet another exploit going on.
435 2013-06-02 22:41:00 <gmaxwell> amuse me
436 2013-06-02 22:41:10 <warren> haven't figured it out yet
437 2013-06-02 22:41:28 <sipa> in the beginning, people thought bitcoin was free money
438 2013-06-02 22:41:34 <sipa> i think we're past that
439 2013-06-02 22:41:43 <sipa> but they still think that dealing with it is free...
440 2013-06-02 22:43:10 <SteveDekorte> berndj: FWIW I'm interested in your line of thought but I guess this isn't the best place to discuss it
441 2013-06-02 22:49:36 <nsh> sipa, free is not necessarily equal to zero cost, in practice. it's can be effectively below-some-threshold-worth-caring-about
442 2013-06-02 22:49:42 <nsh> *it
443 2013-06-02 22:50:09 <sipa> sure, and hopefully that's what bitcoin perhaps actually can do: make handling money cheaper
444 2013-06-02 22:50:30 <nsh> and more flexible, which is where i believe the greater promise lies
445 2013-06-02 22:50:36 <sipa> agree
446 2013-06-02 22:52:55 <ne0futur> gmaxwell: we re talking of ideas to make a tipbot on -otc, your opinion on how to make it well would be welcome
447 2013-06-02 22:53:19 <ne0futur> ;;tip gmaxwell thanks for your ideas on dust management for tips and small donations
448 2013-06-02 22:53:20 <gribble> Error: "tip" is not a valid command.
449 2013-06-02 22:54:18 <nsh> ne0futur, reddit's btc-tipbot might be a good model
450 2013-06-02 22:55:14 <nsh> main modification would be to use services idents rather than reddit accounts, obviously, then if it can tie into gribble's web-of-trust that might be a bonus (?)
451 2013-06-02 22:56:02 <nsh> also if the bitcoins are delivered by carrier pigeon in microfiche format
452 2013-06-02 22:56:10 <nsh> yup, i'm out of sensible input
453 2013-06-02 22:56:16 <ne0futur> yes, or even better in atheme-services ( nickserv, chanser . . . .tipserv ) so any irc network can set it up easily ;)
454 2013-06-02 22:56:37 <nsh> you'd maybe have a job convincing freenode to run another service
455 2013-06-02 22:57:03 <nsh> but optional network level integration could provide some benefits
456 2013-06-02 22:57:23 <nsh> someone who codes for atheme chats here (is it you?)
457 2013-06-02 22:57:41 <ne0futur> you d have to convince atheme , not freenode ;)
458 2013-06-02 22:57:59 <ne0futur> but both are pretty much supporters of bitcoin afaik
459 2013-06-02 22:58:32 <nsh> yes, but allowing unbattle-hardened code to access the network at that level would require a lot of faith
460 2013-06-02 23:00:26 <ne0futur> yeah its always risky to maintain a bitcoind
461 2013-06-02 23:00:51 <ne0futur> any ways to make it use a partner providing the bitcoin tipping account could be useful