1 2013-03-26 00:00:27 <MC1984_> be nice if ms used it on xbox live, cos that still has a codec from when broadband was 3 times faster than dial up
  2 2013-03-26 00:02:18 <denisx> I use mumble with opus
  3 2013-03-26 00:02:21 <denisx> best groupchat ever
  4 2013-03-26 00:03:24 <pigeons> that must be where gmaxwell made his first set of admirers
  5 2013-03-26 00:04:10 <remotemass> sipa: would you consider doing a screecast tutorial of main.cpp ?
  6 2013-03-26 00:04:31 <remotemass> screencast
  7 2013-03-26 00:05:56 <sipa> remotemass: eh, i prefer irc over video to explain things
  8 2013-03-26 00:06:26 <sipa> it's also not really code to be proud of, in current form
  9 2013-03-26 00:06:38 <remotemass> not even for some money?
 10 2013-03-26 00:06:50 <gmaxwell> sipa: thats the idea, nothing should beat it at higher rates until you get to ~lossless either??? at least not on average???, but thats hard to measure.
 11 2013-03-26 00:07:36 <denisx> and best thing is it stops with this psychoacustic shit
 12 2013-03-26 00:08:14 <gmaxwell> Though that graph is a simplification, it's a composite of about 14 different tests, not all apples to applies, e.g. lower rate stuff is speech, higher rate stuff is music.
 13 2013-03-26 00:08:24 <MC1984_> whats wrong with psychoacoustics
 14 2013-03-26 00:08:47 <sipa> remotemass: i'd rather not
 15 2013-03-26 00:08:52 <etotheipi__> anyone know where to find the public key used for signing alerts?
 16 2013-03-26 00:09:16 <sipa> etotheipi__: use the source, luke!
 17 2013-03-26 00:09:25 <sipa> (alert.cpp, top)
 18 2013-03-26 00:09:30 <etotheipi__> oh duh, I guess it'd be hardcoded in the sourc
 19 2013-03-26 00:10:19 <etotheipi__> different keys for main and testnet?
 20 2013-03-26 00:11:32 <lianj> nope
 21 2013-03-26 00:11:59 <lianj> oh yes
 22 2013-03-26 00:12:02 <lianj> 
ACTION updates his code

 23 2013-03-26 00:12:36 <remotemass> sipa: for what money would you consider it, one youtube video tutorial about main.cpp?
 24 2013-03-26 00:13:00 <sipa> remotemass: i don't care about money :)
 25 2013-03-26 00:13:09 <denisx> that means something! ;)
 26 2013-03-26 00:13:20 <remotemass> alright! peace
 27 2013-03-26 00:13:22 <sipa> i guess i'm in the wrong business :p
 28 2013-03-26 00:13:23 <amiller> well... you must *care* about *money* in some sense
 29 2013-03-26 00:14:11 <amiller> surely main.cpp deserves some pride since it's held up well so far
 30 2013-03-26 00:14:24 <sipa> "held up" ?
 31 2013-03-26 00:14:41 <gmaxwell> Held up: has not caught fire and caused the end of the universe yet.
 32 2013-03-26 00:15:25 <amiller> like most of the files in openssl
 33 2013-03-26 00:16:18 <sipa> lol
 34 2013-03-26 00:16:41 <remotemass> Sipa: Who would you recommend for such, then?
 35 2013-03-26 00:17:27 <sipa> main.cpp is the remains of everything that was hacked into bitcoin since ages, and hasn't been able to be untangled in one way or another...
 36 2013-03-26 00:20:39 <phantomcircuit> sipa, perfect description :)
 37 2013-03-26 00:20:45 <remotemass> Would you say that if you completly understand main.cpp and main.h, you must understand at least 80% of it all?
 38 2013-03-26 00:21:02 <sipa> i guess so
 39 2013-03-26 00:21:11 <sipa> oh, no, not the wallet
 40 2013-03-26 00:21:19 <BlueMatt> remotemass: when I started with bitcoin, I spent half my time trying to understand main.cpp...dont, just decide on a project, and do it
 41 2013-03-26 00:21:21 <sipa> wallet stuff is pretty independent
 42 2013-03-26 00:21:45 <remotemass> nice :) I like the idea on focusing on main.cpp only
 43 2013-03-26 00:21:50 <remotemass> of
 44 2013-03-26 00:21:51 <sipa> you can't
 45 2013-03-26 00:22:05 <sipa> net and main interact in weird ways, for example
 46 2013-03-26 00:25:03 <remotemass> what about protocol.cpp .How does it play?
 47 2013-03-26 00:25:47 <remotemass> just trying to get as much info as possible because the code is so dense
 48 2013-03-26 00:27:55 <k9quaint> but CreateNewBlock seems so straightforward...
 49 2013-03-26 00:28:00 <BlueMatt> remotemass: seriously, pick a project and do it...
 50 2013-03-26 00:28:13 <BlueMatt> best way to learn is to dive in over your head and get through it
 51 2013-03-26 00:28:31 <k9quaint> implement bitcoin in lua
 52 2013-03-26 00:28:56 <remotemass> I dream of implementing it in LabVIEW
 53 2013-03-26 00:29:22 <BlueMatt> if you are doing your own implementation, read other clients too...most of them are easier
 54 2013-03-26 00:29:26 <phantomcircuit> s/dream/have nightmares/
 55 2013-03-26 00:29:28 <k9quaint> you guys should minify the CPP code
 56 2013-03-26 00:29:29 <remotemass> but first I need to understand it
 57 2013-03-26 00:29:42 <k9quaint> line feeds are over rated
 58 2013-03-26 00:30:37 <remotemass> I am confortable with Java but BitcoinJ, doesnt it to be what I am looking for to build my own client
 59 2013-03-26 00:31:11 <remotemass> seem
 60 2013-03-26 00:33:36 <remotemass> there's not much to choose, actually. Bitcoin-Qt is the real thing
 61 2013-03-26 00:34:32 <remotemass> the other projects are more of libraries that may be helpful, but are still far from the real thing
 62 2013-03-26 00:36:48 <remotemass> If only there were good books to help. Or more helpful comments along the code...
 63 2013-03-26 00:41:22 <remotemass> anyway, going to sleep. is getting late
 64 2013-03-26 00:45:08 <gfawkes> what is causing the rash of questions about lots of tiny inputs? are the coins really getting that small?
 65 2013-03-26 00:49:04 <user_corrupt> does anyone know how many bitcoins are bought and sold with various currencies by daily average?
 66 2013-03-26 00:49:23 <user_corrupt> or just like last 48 hours for example?
 67 2013-03-26 00:51:41 <Billdr> Surely someone does. Check in #mtgox
 68 2013-03-26 00:51:49 <K1773R> how about looking at the trading volumes @ exchanges? silly...
 69 2013-03-26 01:00:03 <user_corrupt> preface: im not really a developer
 70 2013-03-26 01:00:15 <user_corrupt> just a web design monkey
 71 2013-03-26 01:13:26 <helo> ProcessMessages(inv, 217 bytes) : CHECKSUM ERROR nChecksum=f97d39b7 hdr.nChecksum=769c4559
 72 2013-03-26 01:14:14 <helo> ProcessMessages(tx, 226 bytes) : CHECKSUM ERROR nChecksum=4ca614f1 hdr.nChecksum=54295560
 73 2013-03-26 01:14:22 <helo> PROCESSMESSAGE: INVALID MESSAGESTART
 74 2013-03-26 01:14:33 <helo> right after a misbehaving peer :/
 75 2013-03-26 01:17:59 <helo> sipa: this is on txoptim
 76 2013-03-26 01:26:26 <helo> the process aborted :/
 77 2013-03-26 01:31:23 <evan3> Hey, So, I'm using blockchain.info's payment button and it took over 5 minutes for the funds to make it through their system. Should I just use another service, or implement the funds receiving code myself?
 78 2013-03-26 01:42:15 <CodeShark> I wouldn't rely on someone else's service for anything other than casual personal stuff
 79 2013-03-26 01:42:47 <CodeShark> or testing
 80 2013-03-26 02:32:46 <MC1984_> full on holocaust denial going on in #bitcoin now
 81 2013-03-26 02:33:13 <MC1984_> i should probably be sleeping anyway
 82 2013-03-26 02:34:22 <Luke-Jr> MC1984_: nobody cares, and it's off-topic here
 83 2013-03-26 02:34:52 <MC1984_> not really gonna take an off topic telling from you luke
 84 2013-03-26 02:40:07 <helo> when a wallet is encrypted, are the unencrypted private keys encrypted, or just removed from the pool?
 85 2013-03-26 02:40:45 <MC1984_> vaguely remember that the keypool is regenerated?
 86 2013-03-26 02:42:23 <helo> i asked the wrong question... are the existing "visible" unencrypted private keys encrypted?
 87 2013-03-26 02:43:44 <MC1984_> isnt that what encrypting your wallet entails any way?
 88 2013-03-26 02:46:00 <helo> i thought the important part was that all new addresses would have encrypted keys
 89 2013-03-26 02:46:45 <helo> and it makes sense that the old unencrypted keys would become encrypted... i was just hoping they weren't for the sake of this guy who forgot his passphrase
 90 2013-03-26 02:51:45 <MC1984_> you know
 91 2013-03-26 02:51:59 <MC1984_> i wonder if wallet encryption if leading to moe lost coins, not less
 92 2013-03-26 02:52:20 <bitvoix> less stolen coins. more lost coins
 93 2013-03-26 02:52:32 <MC1984_> which is the lesser evil?
 94 2013-03-26 02:53:31 <necropumpkin> "if I don't get 'em, nobody should"
 95 2013-03-26 02:54:12 <SomeoneWeird> MC1984_, well people don't HAVE to use it
 96 2013-03-26 02:54:21 <SomeoneWeird> if you're not going to remember a password, you shouldn't
 97 2013-03-26 02:54:25 <osmosis> an option to encrypt the entire wallet would be nice, so that account balances are hidden as well.
 98 2013-03-26 02:54:45 <SomeoneWeird> what would that do?
 99 2013-03-26 02:54:54 <jgarzik> SomeoneWeird: privacy
100 2013-03-26 02:54:57 <SomeoneWeird> well
101 2013-03-26 02:55:02 <SomeoneWeird> yeah, I suppose
102 2013-03-26 02:55:25 <jgarzik> same effort as encrypting entire drive; adversaries would not know if you controlled keys or not
103 2013-03-26 02:55:37 <MC1984_> wallet size?
104 2013-03-26 02:55:53 <jgarzik> Movie Villain cannot put gun to your head, and say "I know you control key 1ABCD; spend it to me" :)
105 2013-03-26 02:56:31 <SomeoneWeird> lol
106 2013-03-26 02:56:32 <SomeoneWeird> ya
107 2013-03-26 02:58:02 <helo> there was some effort put into ensuring going from unencrypted wallet to encrypted wallet didn't leave unencrypted keys lying around on the drive, right?
108 2013-03-26 02:58:25 <MC1984_> lol if thats true this guy is fucked
109 2013-03-26 02:59:15 <helo> i think he's fucked... hopefully David__ who i guided through encrypting his wallet (which he was backing up on dropbox) doesn't suffer the same fate
110 2013-03-26 03:02:30 <MC1984_> bitcoin has so far to go before its atually usable
111 2013-03-26 03:02:40 <MC1984_> if it ever will be even
112 2013-03-26 03:03:11 <MC1984_> managing your wallet is too fucking scary
113 2013-03-26 03:03:46 <MC1984_> i dont even want to transact my coins for a very long time if possible
114 2013-03-26 03:04:16 <MC1984_> i thought i lost the passphrase once and my face felt like i was having a stroke
115 2013-03-26 03:04:24 <MC1984_> and i dont even have much
116 2013-03-26 03:05:08 <MC1984_> theres a finality with knowing that if you fuck up once theres not a power on earth that can help you
117 2013-03-26 03:05:29 <MC1984_> most peopledont even know that, they come on #bitcoin all like guys i lost my password halp
118 2013-03-26 03:05:34 <MC1984_> and someone has to break the bad news
119 2013-03-26 03:09:29 <MC1984_> people are used to forgetting and resetting thier facbook shit 3 times a month
120 2013-03-26 03:10:29 <MC1984_> when i worked network support for 500 ish users, and hour of my day every day was resetting user accounts for dumbasses
121 2013-03-26 03:19:58 <bitvoix> I encrypted my electrum wallet and wrote down the password.
122 2013-03-26 03:20:07 <bitvoix> but for some reason, I couldn't get it to work later
123 2013-03-26 03:20:13 <bitvoix> but luckily I had my seed
124 2013-03-26 03:20:19 <bitvoix> so I was able to simply recreate the wallet
125 2013-03-26 03:20:22 <bitvoix> phew
126 2013-03-26 03:20:50 <bitvoix> so that sold me on the whole brain wallet idea
127 2013-03-26 03:21:24 <bitvoix> however, on the other hand, there isn't much security in encrypting the wallet if the seed is stored in plain text :-)
128 2013-03-26 03:22:05 <dermoth> home|bitvoix, the point is that you normally print the seed and store it in a safe...
129 2013-03-26 03:22:16 <bitvoix> of course
130 2013-03-26 03:22:25 <HM> then take the combination to the safe and encrypt it
131 2013-03-26 03:22:29 <bitvoix> haha
132 2013-03-26 03:22:56 <dermoth> home|actually if you really have real value you may want to go with a bank safe
133 2013-03-26 03:23:57 <HM> just remember the damn seed
134 2013-03-26 03:24:15 <dermoth> home|or even better use the hedge fund's methos... encrypt the seed using keys, using secret sharing (ex need two out of three secrets to decrypt the seed), then store the encrypted seed along with one secret on a different bank
135 2013-03-26 03:25:19 <HM> interesting idea. it's only a matter of time before people start trading with these paper secrets
136 2013-03-26 03:25:51 <dermoth> home|you myght want to store them in diff. countryes of you fear armageddon-like scenarios, but at least if one bank gets hit by a meteor you still have the seed, and if someone manage to steel your save at one bank it won't be enough to get the coins
137 2013-03-26 03:26:39 <HM> they should put a gigantic TPM on the space station
138 2013-03-26 03:26:48 <HM> then we'll be safe from nuclear war
139 2013-03-26 03:26:56 <dermoth> home|no way
140 2013-03-26 03:27:11 <dermoth> home|way too much junk around it - it's actually pretty risky
141 2013-03-26 03:27:28 <dermoth> home|expecially if "they" nuke houston...
142 2013-03-26 03:27:39 <jgarzik> Or, no need for secret sharing.  Just have multiple keys.
143 2013-03-26 03:27:47 <dermoth> home|that's what I said
144 2013-03-26 03:27:47 <jgarzik> And multi-sig requires M of N to spend.
145 2013-03-26 03:27:58 <HM> multiple orbital keystores....hmmm...
146 2013-03-26 03:29:05 <dermoth> home|HM, orbital is a really bad idea - sooner or later you'll get hit by some junk flying at a few thousand miles/second...
147 2013-03-26 03:29:47 <dermoth> home|The ISS has to dodge junk pieces sometimes
148 2013-03-26 03:30:03 <HM> Yeah, best not keep anything valuable on anything orbiting at 10s of thousands of miles an hour
149 2013-03-26 03:30:23 <HM> especially if it's full of radioactive molten rock, surrounding a ball of burning hydrogen
150 2013-03-26 03:40:55 <gmaxwell> 21:20 < bitvoix> so that sold me on the whole brain wallet idea
151 2013-03-26 03:41:22 <gmaxwell> uh. that just refuted the brain wallet idea! the reason you couldn't decrypt it would be that the key wasn't what you thought it was!
152 2013-03-26 03:41:31 <gmaxwell> but you had the not-in-your-brain written down seed to save you!
153 2013-03-26 03:41:40 <bitvoix> haha
154 2013-03-26 03:41:44 <bitvoix> you're right.
155 2013-03-26 03:41:58 <bitvoix> I guess it sold me on the deterministic wallet
156 2013-03-26 03:42:46 <bitvoix> the problem wasn't with my memory though, I actually wrote down the encryption password. I must have just typed it in wrong during encryption
157 2013-03-26 03:42:48 <bitvoix> :-)
158 2013-03-26 03:42:54 <gmaxwell> (and normally thats what I recommend: offline hardcopy backups??? not depending on brain anything, because brains are less reliable than expected... and attackers don't find paper or usb keys in your house)
159 2013-03-26 03:43:03 <gmaxwell> Yea, that happens too.. persistant typo
160 2013-03-26 03:43:53 <MC1984_> subcutaneous key implant?
161 2013-03-26 03:44:03 <MC1984_> bit of metal with a passphrase on it?
162 2013-03-26 03:44:14 <MC1984_> bet you could get it done in a tatoo parlour
163 2013-03-26 03:46:30 <jgarzik> My dogs got a sub-dermal ID chip, when they were rescued as puppies
164 2013-03-26 03:46:38 <jgarzik> Has a bit of data on it
165 2013-03-26 03:47:19 <MC1984_> i dont want someone stealing my money just by pointing a parabola at me
166 2013-03-26 03:47:59 <jgarzik> there's always crystal-based keys, practically impossible to copy ;p
167 2013-03-26 03:48:13 <MC1984_> the idea is that to redeem you gonna have to cut it out of your flesh, and that is unlikely to be done by accident or without permission
168 2013-03-26 03:48:38 <MC1984_> crystal?
169 2013-03-26 03:49:26 <warren> MC1984_: won't that be destroyed (along with some of your human tissue) if you get a MRI?
170 2013-03-26 03:50:19 <MC-Droid> could be ceramic instead i suppose
171 2013-03-26 03:50:37 <MC-Droid> better if it is actually, if you want to keep it secret
172 2013-03-26 03:51:00 <MC-Droid> laser etched ceramic bead
173 2013-03-26 03:51:39 <MC-Droid> ame trust level as a casicus coin though
174 2013-03-26 03:53:23 <warren> MC-Droid: operate the laser yourself, somehow...
175 2013-03-26 04:48:12 <helo> bad node! not sure if anyone cares, but Misbehaving: 108.61.63.251:8333 (90 -> 100) DISCONNECTING
176 2013-03-26 04:49:58 <gmaxwell> helo: what was it doing?
177 2013-03-26 04:54:34 <jgarzik> thinking of P2SH
178 2013-03-26 04:54:45 <jgarzik> I wish there was a useful way to store the redeem script locally
179 2013-03-26 04:55:36 <gmaxwell> jgarzik: addmultisigaddress loads one...
180 2013-03-26 04:56:02 <gmaxwell> really there should be some kind of importp2sh that imports the redeemscript as well as additional metadata about how to satisify it.
181 2013-03-26 04:56:35 <gmaxwell> e.g. gives you urls and/or descriptions of all the involved keys in a multisig.
182 2013-03-26 04:57:21 <jgarzik> even if we wanted to use P2SH by default, for most spend transactions, it seems like -- from a user interface perspective -- you want to communicate the redeem script (via this wonderful payment protocol perhaps)
183 2013-03-26 04:57:48 <gmaxwell> jgarzik: huh? I don't think you do.
184 2013-03-26 04:58:08 <gmaxwell> If I'm asking you to pay me, it's none of your business how I manage my accounts.
185 2013-03-26 04:59:03 <gmaxwell> oh, perhaps I misunderstand what you mean by spend transactions.
186 2013-03-26 05:41:47 <TradeFortress> sorry for leaving earlier but bitcoind just keeps on exiting :P
187 2013-03-26 05:41:54 <TradeFortress> has being happening for multiple times
188 2013-03-26 05:43:11 <Luke-Jr> you use IRC with bitcoind? O.o
189 2013-03-26 05:43:44 <TradeFortress> nope, but had to go
190 2013-03-26 05:43:56 <TradeFortress> anyway, again, nothing suspicious in debug.log
191 2013-03-26 05:43:59 <TradeFortress> just accepting blocks
192 2013-03-26 05:56:27 <TradeFortress> hmm I guess I will have to get in the habit of restarting bitcoind every time it crashes
193 2013-03-26 05:56:56 <sivu> while true ; do bitcoind ; done
194 2013-03-26 05:57:36 <gmaxwell> TradeFortress: you're running out of virtual memory most likely. Stop listening to inbound connections and I bet it stays up.
195 2013-03-26 05:57:45 <Belxjander> Can anyone point me at a general overview document of the processes inside the "Wallet" and "miner" applications?
196 2013-03-26 05:58:04 <gmaxwell> I think linnode is running some weird kernel that reduces the VM space??? you're the second person who's shown up with a linnode node that was mysteriously crashing.
197 2013-03-26 05:59:13 <TradeFortress> gmaxwell, hmm, will more swap space help?
198 2013-03-26 06:00:35 <sivu> TradeFortress: it's not the memory thats the issue, it is the address space for the memory
199 2013-03-26 06:00:41 <sivu> so more swap will not help
200 2013-03-26 06:01:10 <TradeFortress> okay, so what can I do to make it stop crashing? reduce the number of connections?
201 2013-03-26 06:01:21 <gmaxwell> TradeFortress: ldpreloading tcmalloc will help, though not likely enough for you to have a full compliment of connections.
202 2013-03-26 06:01:40 <gmaxwell> TradeFortress: try setting listen=0  and see if it stays up.
203 2013-03-26 06:02:42 <TradeFortress> listen=0 in bitcoin.conf? Will that allow me to dl the blockchain?
204 2013-03-26 06:02:51 <TradeFortress> (and keep up with transactions, otherwise there's no point)
205 2013-03-26 06:02:59 <sivu> yes
206 2013-03-26 06:03:12 <sivu> it will make the connections itself, but does not accept incoming connections
207 2013-03-26 06:03:40 <TradeFortress> okay, thanks!
208 2013-03-26 06:03:43 <TradeFortress> I'll try that
209 2013-03-26 07:26:19 <grau> !ticker
210 2013-03-26 07:26:19 <gribble> BTCUSD ticker | Best bid: 76.20001, Best ask: 76.50000, Bid-ask spread: 0.29999, Last trade: 76.50000, 24 hour volume: 81153.30956201, 24 hour low: 71.52122, 24 hour high: 78.00000, 24 hour vwap: 74.80453
211 2013-03-26 07:36:52 <grau> !ticker
212 2013-03-26 07:36:57 <gribble> BTCUSD ticker | Best bid: 76.40000, Best ask: 76.50000, Bid-ask spread: 0.10000, Last trade: 76.50000, 24 hour volume: 81165.20482049, 24 hour low: 71.52122, 24 hour high: 78.00000, 24 hour vwap: 74.80844
213 2013-03-26 08:19:07 <qdii> hello
214 2013-03-26 08:20:45 <Billdr> oh, hi.
215 2013-03-26 08:22:40 <qdii> if I understand what happens during data mining well, a bunch of people are trying to find a block with a particular nonce so that the hash of the hash of the header verifies some weird condition
216 2013-03-26 08:23:16 <Billdr> well, data mining is a phrase that means something different than bitcoin mining.
217 2013-03-26 08:23:25 <qdii> *bitcoin mining
218 2013-03-26 08:23:27 <qdii> my bad :)
219 2013-03-26 08:23:32 <Billdr> no problem
220 2013-03-26 08:23:41 <qdii> does that mean that whenever a block is found, the whole miner fleet has to start over?
221 2013-03-26 08:23:56 <Belxjander> not that I can see
222 2013-03-26 08:23:56 <Billdr> yea
223 2013-03-26 08:24:10 <Belxjander> eh?
224 2013-03-26 08:24:11 <qdii> the reason is that the hash of the previous block is part of the header
225 2013-03-26 08:24:21 <Belxjander> so the entire block chain is processed every 10 minutes?
226 2013-03-26 08:24:32 <Billdr> The set of transactions in the current block have been commited to the block chain. We then make a new block with the next set of transactions
227 2013-03-26 08:25:37 <Billdr> of course, the other miners have to agree that the last commited block is valid before it's actually commited.
228 2013-03-26 08:26:01 <Belxjander> Billdr: does the block chain have to be obtained from the beginning?
229 2013-03-26 08:26:04 <n1c> (I could be way off here) but I think the idea of "start over" might be misleading, it's not necessarily a sequential process the miners are going through. They do however start with a new "top" block though. (afaik).
230 2013-03-26 08:26:23 <Billdr> ^
231 2013-03-26 08:26:59 <Belxjander> and you can link into the network from any point at all?
232 2013-03-26 08:27:00 <qdii> n1c: but it is somehow though, you start with a nonce, say 0, and then you increment it until the sha(sha(header)) is correct
233 2013-03-26 08:27:06 <Billdr> Belxjander, it used to. I'm not 100% sure on how it works since .8
234 2013-03-26 08:27:34 <n1c> What I'm saying qdii, I don't think you /have/ to go nonce++
235 2013-03-26 08:27:49 <n1c> You could generate random ones
236 2013-03-26 08:27:59 <Belxjander> Billdr: I think I should have started with coming here to ask more about the code aspects of bitcoin mining
237 2013-03-26 08:28:13 <n1c> Please keep in mind, I've neither read or written any bitcoin or miner code :)
238 2013-03-26 08:28:19 <Belxjander> 
ACTION has wasted a week trying to grok the docs and various sources

239 2013-03-26 08:28:27 <qdii> n1c: hm I think that would result in the same thing: say you start with 5, it doesn???t work, then everyone in the pool knows that 5 is not a correct nonce
240 2013-03-26 08:28:36 <qdii> so you have progressed
241 2013-03-26 08:28:45 <n1c> I don't think you communicate failures back to the network?
242 2013-03-26 08:28:49 <n1c> You all just go at it until a success.
243 2013-03-26 08:28:55 <n1c> Again, speculation!
244 2013-03-26 08:28:57 <n1c> :)
245 2013-03-26 08:28:58 <Billdr> I haven't done any work on the bitcoin spec itself. I'm just building things on top of it. My understanding isn't going to be a lot better than the average investor's :p
246 2013-03-26 08:29:01 <qdii> yea, here too
247 2013-03-26 08:29:02 <Belxjander> n1c: well I saw that the nonces are cycled so I was thinking of starting and then throwing blocks of nonces through an FPGA using that "mid-state" thing the docs talked about
248 2013-03-26 08:29:36 <n1c> Yeah I was thinking about similar stuff.
249 2013-03-26 08:29:53 <n1c> But lots of the crypto-stuff is over my head so I haven't done more digging.
250 2013-03-26 08:30:10 <Belxjander> hell... I even have an FPGA sitting here idle to play with built into my machine
251 2013-03-26 08:30:10 <qdii> I thought the purpose of pool was to distribute the nonces among the miners, so that once a nonce has been proven bad, no other miner computes it again
252 2013-03-26 08:30:48 <n1c> I was thinking the other day - if you got sneaky with the way you pick nonces, could you beat others to correct blocks? But surely there's a reason it's not that easy? Then I realised I know so little about stats I'm just guessing :)
253 2013-03-26 08:30:54 <Belxjander> n1c: I need a nice walkthrough like how blowfish and twofish are described for the steps in the algorithm
254 2013-03-26 08:31:15 <Belxjander> n1c: its brute-force or lucky chances
255 2013-03-26 08:31:19 <n1c> Belx if it's not on the wiki, it'll be in the white paper or on the internet somewhere I'm sure.
256 2013-03-26 08:31:23 <n1c> Or, just read the code?
257 2013-03-26 08:31:29 <n1c> I don't think it's /that/ much.
258 2013-03-26 08:31:40 <Belxjander> n1c: Ive been breaking my mind with the C++ and python clients
259 2013-03-26 08:31:47 <n1c> -_-
260 2013-03-26 08:32:06 <Belxjander> Ive never been very good at grokking other peoples code without some kind of interaction to work it out
261 2013-03-26 08:32:16 <n1c> It's a pity all the core people are in a way different timezone.
262 2013-03-26 08:33:13 <n1c> qdii - when you say pool you mean, pool mining like 50BTC, BTC Guild etc?
263 2013-03-26 08:33:22 <qdii> yes :)
264 2013-03-26 08:33:28 <n1c> Ah, I don't know how they work.
265 2013-03-26 08:33:37 <qdii> me neither, only speculating :)
266 2013-03-26 08:33:38 <n1c> It would be smart if they did do something like alert nodes of already tried/failed nonces.
267 2013-03-26 08:35:15 <n1c> I really wish we had proper answers to this dicussion
268 2013-03-26 08:35:17 <n1c> It keeps me up at night.
269 2013-03-26 08:35:53 <Belxjander> n1c: just reading some of the wiki about the blockchain...
270 2013-03-26 08:36:26 <qdii> n1c: I thought the whole purpose of poolmining was exactly that. A central authority says worker1 takes these nonces, worker2 takes those, ??? so that no 2 workers overlap nonces
271 2013-03-26 08:36:29 <Belxjander> would it still be valid to run an "AppEngine" instance as a "bitcoind" node and have the backend database off that handle the blockchain... but never run anything else on it?
272 2013-03-26 08:36:45 <sivu> n1c, timezones are good. that way the development will continue 24/7
273 2013-03-26 08:37:02 <n1c> :p
274 2013-03-26 08:37:16 <sivu> every 8 hours theres a switchover to next continent
275 2013-03-26 08:37:35 <n1c> qdii: I always just understood it as a way of getting smaller increments of the winnings. So it's less of a lottery and more spread out.
276 2013-03-26 08:37:37 <sivu> its like working in three shifts
277 2013-03-26 08:38:52 <Scrat> Belxjander: can't decipher what you said. appengine can only run specific databases
278 2013-03-26 08:39:01 <hardsoft> Habla espa?ol alguien?
279 2013-03-26 08:39:39 <n1c> :x
280 2013-03-26 08:40:50 <qdii> yo si
281 2013-03-26 08:41:12 <sivu> uno cerveza por favor
282 2013-03-26 08:46:03 <da2ce7> G'day all
283 2013-03-26 08:47:31 <qdii> good day
284 2013-03-26 08:53:44 <sipa> helo: thx, i know there is a bug in txoptim; the last debug.log lines may be useful
285 2013-03-26 09:22:03 <n1c> Any of you guys know much about wallet "cold storage"
286 2013-03-26 09:22:04 <n1c> ?
287 2013-03-26 09:22:18 <Billdr> I use one, I think I understand it.
288 2013-03-26 09:22:24 <Billdr> I could be completely wrong though.
289 2013-03-26 09:23:08 <n1c> Haha
290 2013-03-26 09:23:19 <n1c> So it basically means - taking your wallet off the internet right?
291 2013-03-26 09:24:31 <Billdr> yea
292 2013-03-26 09:24:55 <Billdr> the wallet I'm using as my cold storage has never been imported to a pc that is online.
293 2013-03-26 09:25:31 <Guest70513> dude - read here - https://bitcoinarmory.com/ and then here - https://bitcoinarmory.com/using-offline-wallets-in-armory/
294 2013-03-26 09:25:40 <n1c> And, a wallet is basically just a collection of private keys right?
295 2013-03-26 09:25:41 <Guest70513> tutorial / presentation - everything....
296 2013-03-26 09:26:10 <n1c> Thanks, I'm more asking for the point of discussion.
297 2013-03-26 09:26:18 <Billdr> as I understand it, it's one private key and a bunch of associated public keys
298 2013-03-26 09:26:20 <qdii> does angela merkel like trees or something?
299 2013-03-26 09:26:34 <Billdr> Once we start talking cryptography I get really vague real fast.
300 2013-03-26 09:26:45 <n1c> Yeah me too -_-
301 2013-03-26 09:26:49 <n1c> Hence the vague questions
302 2013-03-26 09:27:12 <n1c> Anyway, where I was going with this...
303 2013-03-26 09:27:36 <n1c> It just seems like a cool idea that you would be able to 3d-print something, that contains your private key for offline storage.
304 2013-03-26 09:27:39 <Guest70513> sent u a pm with tutorial
305 2013-03-26 09:27:54 <jouke> well, actaully, my definition of a wallet is that it contains the private keys, but also information about transactions regarding those keys.
306 2013-03-26 09:28:11 <Billdr> That makes more sense
307 2013-03-26 09:28:33 <Billdr> but, is that plural? Private key*S*?
308 2013-03-26 09:28:40 <sipa> n1c: what do you need?
309 2013-03-26 09:28:43 <n1c> But there's no value in keeping the transaction history "offline"
310 2013-03-26 09:29:14 <n1c> sipa - just curious about if there would ever be a need to keep your wallet offline in some physical form.
311 2013-03-26 09:29:17 <n1c> that could be 3d printed.
312 2013-03-26 09:29:23 <Billdr> The transaction history is on the chain. You really just need the private key to find the relevant transactions.
313 2013-03-26 09:29:28 <jouke> depends. If you want to send those coins you received, you need to have that transaction history.
314 2013-03-26 09:29:35 <qdii> from bitcoin-qt's gettransaction command, I??cannot see the address from which the transaction originated
315 2013-03-26 09:29:50 <qdii> why is that?
316 2013-03-26 09:30:10 <sipa> qdii: because bitcoin transactions do not have a 'from' address
317 2013-03-26 09:30:36 <Billdr> n1c, it'd have to be cooler than this: https://www.casascius.com/
318 2013-03-26 09:31:00 <n1c> It's just been rolling around in my head for a few days.
319 2013-03-26 09:31:00 <sipa> they have inputs, and each input may have an identifiable previous address that coins were previously assigned to
320 2013-03-26 09:31:14 <n1c> If there was a /somewhat/ standard way of printing them out, it could be pretty awesome.
321 2013-03-26 09:31:16 <sipa> but even then, it probablyndoesn't do what you want
322 2013-03-26 09:31:35 <jouke> n1c: I use a pen.
323 2013-03-26 09:31:39 <n1c> haha yeah Billdr
324 2013-03-26 09:31:40 <Scrat> n1c: a QR encoded privkey would do the trick
325 2013-03-26 09:31:52 <n1c> Yeah, I figured as much.
326 2013-03-26 09:32:00 <qdii> sipa: from gettransaction, I??can???t see the input either
327 2013-03-26 09:32:08 <n1c> not very easy to read that, but printing would be easy
328 2013-03-26 09:32:23 <sipa> (in particular, sending transactions there will not guarantee it arrives at the sender, in addition to being blatant address rwuse)
329 2013-03-26 09:32:26 <Scrat> n1c: it is easy if you also print the privkey in base58 below it :p
330 2013-03-26 09:32:37 <sipa> qdii: you may need getrawtransaction for that
331 2013-03-26 09:32:42 <n1c> hehehe
332 2013-03-26 09:32:49 <sipa> qdii: which need -txindex
333 2013-03-26 09:34:12 <qdii> sipa: sorry what is txindex?
334 2013-03-26 09:34:41 <sipa> qdii: a transaction index, to be able to look up transactions in the block chain by txid
335 2013-03-26 09:34:56 <sipa> qdii: since 0.8, that's off by default
336 2013-03-26 09:35:52 <qdii> hm, the help page says getrawtransaction <txid>, are you sure I need the index?
337 2013-03-26 09:36:03 <sipa> yes
338 2013-03-26 09:36:28 <sipa> it will work without, for not-entirely-spent transactions, but more slowly
339 2013-03-26 09:41:01 <qdii> so basically, for a regular, address-to-address, transaction, it has one input. When the transaction generates bitcoins (like, mining), it has no input and one output, right?
340 2013-03-26 09:41:55 <sipa> there is no such thing as an address to address transaction
341 2013-03-26 09:42:13 <sipa> bitcoin does not function in terms of addresses but in terms of coins
342 2013-03-26 09:42:29 <sipa> there are no address balances or anything at the protocol level
343 2013-03-26 09:42:42 <sipa> you spend a previous transaction output
344 2013-03-26 09:42:49 <sipa> andntypically more than one
345 2013-03-26 09:44:54 <qdii> hm, so is my "balance" is the sum of all the transaction I have received and not yet spent?
346 2013-03-26 09:45:25 <Eliel> qdii: yes
347 2013-03-26 09:47:58 <jaakkos> qdii: to be exact, it's the sum of outputs you've not yet spent - the transactions may have multiple outputs, to many persons for instance
348 2013-03-26 09:48:44 <sipa> "coin" is often used as a synonym for "unspent transaction output"
349 2013-03-26 09:48:48 <qdii> okay. so if I??wanted to write sth to display my balance, I would crawl all the transactions one by one and filter those which have an output to my address
350 2013-03-26 09:49:21 <sipa> that would give you the "balance of an address", which is mostly useless
351 2013-03-26 09:49:38 <sipa> as you typically have a combination of addresses that constitute a wallet
352 2013-03-26 09:50:30 <qdii> okay, but there is no way to know which addresses are part of a wallet? that's the beauty of it right
353 2013-03-26 09:51:19 <jaakkos> others than yourself might not know, but looking at the transaction history they might be able to deduce a thing or two.
354 2013-03-26 09:54:44 <jaakkos> qdii: for instance, if you make a transaction that spends two inputs, then you must have known the private key to both, so they may both have been your addresses
355 2013-03-26 09:55:19 <Scrat> at which point does getbalance * get slow? ie. > 100ms. I see that CWalletTX is using a cache so subsequent calls to getbalance should be pretty fast?
356 2013-03-26 09:56:03 <jaakkos> qdii: further, since the outputs are always spent 100%, your change is sent back to you, to a new address - one might be able to deduce which output is your change, and associate the change address to you before you spend it
357 2013-03-26 09:57:17 <qdii> jaakkos: I create a transaction to spend two inputs ??? why do I need the private key to both ?
358 2013-03-26 09:57:37 <jaakkos> my bad - s/inputs/outputs/
359 2013-03-26 09:57:49 <Eliel> qdii: you need to prove you're authorized to spend both.
360 2013-03-26 09:58:07 <Eliel> not enough if you're just authorized to spend one of them
361 2013-03-26 10:00:01 <qdii> ok jaakkos
362 2013-03-26 10:00:19 <qdii> Eliel: yea, but that would be the same one private key of the address
363 2013-03-26 10:00:41 <sipa> qdii: why do you assume both inputs spend something assigned to the same address?
364 2013-03-26 10:00:53 <Eliel> of course, if a wallet becomes common that allows a secure/low priority transaction and when those are made, it tries to find other nodes making similar transactions and negotiates with them to combine the transactions, then it becomes more difficult to interpret the blockchain data this way.
365 2013-03-26 10:01:06 <sipa> that's possible, but not really common operation, as you're supposed to use a new key per transaction anywa
366 2013-03-26 10:01:46 <sipa> (though current infrastructure makes that somewhat hard, so we do see some address reuse today)
367 2013-03-26 10:01:49 <jaakkos> qdii: notice that output/input is interchangeable here - it's a previous tx's output, but now used as an input in the hypotethical transaction we're discussing
368 2013-03-26 10:10:49 <qdii> now I am a bit confused, sorry. Am I right to think that an input is a transaction itself?
369 2013-03-26 10:11:03 <_dr> sorry if it's OT, but regarding avalon's 3rd batch.. why can i still order them? i thought they sold out like 10 minutes after they announced them
370 2013-03-26 10:11:44 <Graet> ppl put orders in that didnt get paid for
371 2013-03-26 10:11:53 <Graet> so the units go bak ibto "stock"
372 2013-03-26 10:12:17 <jaakkos> qdii: no. a transaction has 2 lists: inputs and outputs.
373 2013-03-26 10:12:29 <_dr> ah, seems like the initial 'people will buy them at any price' assumption was wrong :)
374 2013-03-26 10:12:52 <Graet> well ppl are still buying :P
375 2013-03-26 10:12:58 <qdii> jaakkos: yea, I never said otherwise :)
376 2013-03-26 10:13:13 <Eliel> _dr: more likely they had some problems paying for them.
377 2013-03-26 10:13:24 <qdii> but is the input of a transaction a transaction itself?
378 2013-03-26 10:13:26 <Eliel> _dr: like forgetting passphrases, paying the wrong amount... that kind of stuff.
379 2013-03-26 10:13:44 <Eliel> or even paying the wrong address
380 2013-03-26 10:13:44 <qdii> or rather, are the inputs of a transaction transactions themselves
381 2013-03-26 10:13:55 <jaakkos> qdii: no, the input is a reference to an output of a previous transaction
382 2013-03-26 10:14:30 <_dr> paying 75btc to the wrong address, heh. people handling such amounts should know better :)
383 2013-03-26 10:14:58 <jaakkos> qdii: the input refers a previous output, ie. it refers previous tx plus an index to the previous tx's output list, and gives the signature to redeem that output
384 2013-03-26 10:15:03 <qdii> okay, so a transaction could really have two inputs, which are references to 2 outputs to different addresses
385 2013-03-26 10:15:25 <jaakkos> yes that is common.
386 2013-03-26 10:15:27 <qdii> 
ACTION trying to make sense of all what have been said

387 2013-03-26 10:16:10 <qdii> ahhh
388 2013-03-26 10:16:49 <qdii> I didn???t see that bitcoin-qt didn???t let me choose which address to send the money FROM
389 2013-03-26 10:17:00 <qdii> it acts as a real wallet
390 2013-03-26 10:17:50 <jaakkos> that could be interesting functionality but i think most people wouldn't need it
391 2013-03-26 10:20:51 <qdii> well, I thought it was important: if a company wants to do some serious business and have only ONE address to show to its customers
392 2013-03-26 10:20:56 <jaakkos> if the recipient knows one of your prior addresses, you might use such function to send them coins and they would know where they came from
393 2013-03-26 10:21:22 <qdii> (one address, to make it less error-prone)
394 2013-03-26 10:23:14 <jaakkos> well, you could do that if you really want
395 2013-03-26 10:23:19 <jaakkos> but not with the reference wallet
396 2013-03-26 10:24:06 <jaakkos> these days the standard is that if you want to know who is sending you money, the recipient gives a unique address only to the sender, and that way they don't need to care about the 'from' addresses.
397 2013-03-26 10:31:42 <qdii> which makes it hard to do accounting based on the transactions only
398 2013-03-26 10:35:24 <qdii> is that butterflylab a scam? are there people who have received boxes from batch 1 and 2???
399 2013-03-26 10:36:43 <jaakkos> this is OT on -dev, but perhaps you're referring to Avalon, who have certainly delivered. butterflylabs has not shipped anything yet, nor have they proved they have working units, AFAIK.
400 2013-03-26 10:38:23 <SomeoneWeird> correct
401 2013-03-26 10:38:41 <Billdr> I think Luke has seen working units.
402 2013-03-26 10:39:24 <Billdr> in their production run test rig
403 2013-03-26 10:39:40 <sipa> qdii: you really shouldn't want to have only one address, the privacy implications are terrible
404 2013-03-26 10:40:12 <qdii> I personally??don???t, but a company might want to display clear accounting
405 2013-03-26 10:40:33 <sipa> i don't mean their privacy, i mean the privacy of everyone using the system
406 2013-03-26 10:40:53 <sipa> clear accounting is certainly beneficial, but it doesn't mean revealing everything to the whole world
407 2013-03-26 10:44:42 <sipa> qdii: the payment protocol will hopefully make it soon a lot easier to have separate keys per transaction
408 2013-03-26 10:50:41 <cads> hey sipa, is there a #bitcoin-blah for laid back and intellectual discussions free of trolling and disrespect?
409 2013-03-26 10:57:39 <SomeoneWeird> cads, #bitcoin-offtopic?
410 2013-03-26 11:23:05 <Diablo-D3> https://bitmessage.org/wiki/Main_Page
411 2013-03-26 11:23:08 <Diablo-D3> that looks interesting
412 2013-03-26 11:23:14 <SomeoneWeird> yeah, I saw that posted on hn
413 2013-03-26 11:23:17 <SomeoneWeird> i looks over it a bit
414 2013-03-26 11:23:25 <SomeoneWeird> seems like it could work
415 2013-03-26 11:23:30 <SomeoneWeird> s/looks/looked/
416 2013-03-26 11:24:18 <Diablo-D3> I dunno, maybe I really should just make a framework that does bitcoin-type POW chains
417 2013-03-26 11:24:44 <Diablo-D3> does it correctly, securely, and easily
418 2013-03-26 11:25:18 <Diablo-D3> and have source and destination, not just destination
419 2013-03-26 11:25:49 <SomeoneWeird> well, wouldn't you spam the chain
420 2013-03-26 11:27:16 <Diablo-D3> SomeoneWeird: kind of.
421 2013-03-26 11:31:11 <Diablo-D3> SomeoneWeird: in a public chain, you'd probably have a basic currency
422 2013-03-26 11:31:18 <Diablo-D3> not all chains would have to be public
423 2013-03-26 11:31:25 <SomeoneWeird> mm
424 2013-03-26 11:32:08 <Diablo-D3> that, and you could have a subscriber type model
425 2013-03-26 11:32:19 <Diablo-D3> and auto-forking chains
426 2013-03-26 11:32:37 <abadr> How can I tell whether bitcoind is done syncing?
427 2013-03-26 11:32:51 <Diablo-D3> abadr: bitcoin getinfo
428 2013-03-26 11:32:52 <Diablo-D3> er
429 2013-03-26 11:32:54 <Diablo-D3> bitcoind getinfo
430 2013-03-26 11:32:58 <Diablo-D3> the blocks match the number of actual blocks
431 2013-03-26 11:33:24 <abadr> you're saying to get the number of actual blocks from somewhere external?
432 2013-03-26 11:33:33 <Diablo-D3> abadr: yeah
433 2013-03-26 11:33:34 <Diablo-D3> ;;blocks
434 2013-03-26 11:33:34 <gribble> 228106
435 2013-03-26 11:33:56 <Diablo-D3> I dont think theres an rpc command to tell if its done yet
436 2013-03-26 11:33:58 <Diablo-D3> the GUI does it, though
437 2013-03-26 11:34:06 <sipa> it is never done
438 2013-03-26 11:34:11 <sipa> there will always be new blocks
439 2013-03-26 11:34:15 <Diablo-D3> sipa: hurrr =P
440 2013-03-26 11:34:40 <sipa> there are some heuristics for guessing whether you're certainly not done yet
441 2013-03-26 11:34:56 <sipa> like looking at the age of the last block, or comparing with your peer's heights
442 2013-03-26 11:35:02 <Diablo-D3> sipa: a new client asks for peer heights
443 2013-03-26 11:35:03 <sipa> but there is no way to know you are done
444 2013-03-26 11:35:09 <Diablo-D3> so it knows when its caught up
445 2013-03-26 11:35:12 <abadr> I see
446 2013-03-26 11:35:16 <sipa> peers can lie
447 2013-03-26 11:35:32 <sipa> or can be not-yet-synced themself
448 2013-03-26 11:35:45 <Diablo-D3> sipa: true, but if ALL your peers are lying, just give up now
449 2013-03-26 11:36:19 <sipa> i'm saying that any check to see whether you're up-to-date is always a guess
450 2013-03-26 11:36:53 <sipa> because if there was a way of knowing that, we wouldn't need the blockchain at all, and we could use that mechanism to order transactions too
451 2013-03-26 11:37:43 <Diablo-D3> sipa: yeah, but you're just describing the issue with asynchronous messaging
452 2013-03-26 11:37:50 <sipa> indeed
453 2013-03-26 11:37:51 <Diablo-D3> which was already solved in the 60s by IBM
454 2013-03-26 11:38:18 <sipa> by? central clock? quorum?
455 2013-03-26 11:38:20 <Diablo-D3> Im actually wondering why it took so long for someone to invent Bitcoin
456 2013-03-26 11:38:25 <sipa> neither of these apply to us
457 2013-03-26 11:38:34 <Diablo-D3> sipa: all three
458 2013-03-26 11:38:42 <abadr> heh
459 2013-03-26 11:38:56 <Diablo-D3> central clock, quorum, and racing chain lengths
460 2013-03-26 11:39:18 <sipa> abadr: to answer your question in practice: look at the age of the last block
461 2013-03-26 11:39:27 <abadr> got it. thanks sipa.
462 2013-03-26 11:41:38 <abadr> what's the value of "nextblockhash" from a "getblock" command if there are multiple next-blocks? (and where can i look to learn this kind of thing on my own?)
463 2013-03-26 11:42:47 <Diablo-D3> abadr: there are never multiple next blocks if you ignore chain forks that are not the selected one
464 2013-03-26 11:43:47 <abadr> Diablo-D3: how would you recommend detecting chain forks?
465 2013-03-26 11:43:56 <Diablo-D3> abadr: I wouldn't.
466 2013-03-26 11:43:58 <Diablo-D3> I
467 2013-03-26 11:44:02 <Diablo-D3> I'd leave the client do it's job
468 2013-03-26 11:44:37 <abadr> Diablo-D3: what bitcoin-related software have you developed?
469 2013-03-26 11:44:51 <Diablo-D3> abadr: wrote DiabloMiner, the worlds most popular GPU miner
470 2013-03-26 11:46:06 <sipa> abadr: nextblockhash refers to the next block in the currently active chain
471 2013-03-26 11:46:41 <Diablo-D3> sipa: yeah but you cant fork surf with it
472 2013-03-26 11:48:34 <sipa> indeed
473 2013-03-26 11:48:59 <Diablo-D3> I think he wants to write a blockchain.info clone
474 2013-03-26 11:49:01 <SomeoneWeird> <Diablo-D3> abadr: wrote DiabloMiner, the worlds most popular GPU miner < that might be a slight stretch
475 2013-03-26 11:49:16 <Diablo-D3> SomeoneWeird: remember, cgminer still uses my kernel =P
476 2013-03-26 11:49:31 <sipa> SomeoneWeird: everything that Diablo-D3 says is a slight stretch, don't worry about it
477 2013-03-26 11:49:46 <SomeoneWeird> Diablo-D3, oh well then
478 2013-03-26 11:49:46 <SomeoneWeird> :P
479 2013-03-26 11:49:51 <SomeoneWeird> oh, well then*
480 2013-03-26 12:00:57 <lodse> what happens during the "confirmation" of bitcoins to the wallet?
481 2013-03-26 12:01:09 <n1c> Could you be more specific?
482 2013-03-26 12:01:42 <lodse> when someone wires bitcoins to me it says they are not yet confirmed
483 2013-03-26 12:01:51 <lodse> so there has to be some kind of confirmation
484 2013-03-26 12:02:19 <n1c> Yes.
485 2013-03-26 12:02:36 <lodse> what exactly happens there
486 2013-03-26 12:02:38 <n1c> https://en.bitcoin.it/wiki/Confirmation
487 2013-03-26 12:02:42 <lodse> thank you
488 2013-03-26 12:02:53 <n1c> np
489 2013-03-26 12:11:54 <moarrr> 2zo4rgRtM7ZjiRYUqHyg3mRLJiK
490 2013-03-26 12:34:13 <rdponticelli> hardsoft: En #bitcoin-ar hablamos en espa??ol :)
491 2013-03-26 12:59:36 <sneak> zooko: hi
492 2013-03-26 13:20:59 <E1ven> I'm trying to understand the proof-of-work algorithm, as described at https://en.bitcoin.it/wiki/Proof_of_work, but I'm getting a different number of 0s required than the website indicates. http://pastebin.ca/raw/2342676  I'm sure I'm doing something stupid, but I was hoping someone might advise to what it is. Perhaps the encoding?
493 2013-03-26 13:22:53 <Scrat> E1ven: shouldn't you reset the hash inside the loop?
494 2013-03-26 13:23:06 <Scrat> 
ACTION doesn't know python

495 2013-03-26 13:23:14 <E1ven> Ah, damnit. I found it. I misread as appending 0s to the end, with an increasing number of them. Instead, I should be appending 1, then 2, then 3, etc.
496 2013-03-26 13:23:56 <gavinandresen> E1ven: https://github.com/bitcoin/bitcoin/blob/master/contrib/pyminer/pyminer.py#L111   might give insight
497 2013-03-26 13:24:59 <E1ven> That's much better written than my test, and helps explain it. Thanks!
498 2013-03-26 13:42:03 <E1ven> Thanks again!
499 2013-03-26 14:09:31 <abadr> is there a way to get the mac client to act like bitcoind (i.e. so I can send it commands) or do I have to compile my own?
500 2013-03-26 14:09:42 <sipa> run with -server
501 2013-03-26 14:11:15 <abadr> sipa: thanks again
502 2013-03-26 14:18:35 <gmaxwell> https://bitcointalk.org/index.php?topic=155476.msg1674347#msg1674347
503 2013-03-26 14:20:09 <SomeoneWeird> bad nodes!
504 2013-03-26 14:22:44 <Diablo-D3> bad nodes bad nodes, whatcha gonna do when they come for you
505 2013-03-26 14:23:27 <Diablo-D3> wait hold up
506 2013-03-26 14:23:34 <Diablo-D3> someone ported bitcoin to erlang?
507 2013-03-26 14:25:29 <SomeoneWeird> that they did
508 2013-03-26 14:30:27 <abadr> that's awesome. i'ma throw some btc at them.
509 2013-03-26 15:34:31 <qdii> I wonder: would it be possible to switch from SHA256 to SHA512
510 2013-03-26 15:34:34 <qdii> or sth similar
511 2013-03-26 15:34:56 <qdii> as the longest chain of block is the valid one, it would require that a majority of clients update
512 2013-03-26 15:35:11 <sipa> qdii: no, it would require every client to upgrade
513 2013-03-26 15:35:18 <qdii> and that cannot be done simultaneously
514 2013-03-26 15:35:33 <qdii> sipa, actually, only 50% of the mining float, isn???t it?
515 2013-03-26 15:35:37 <qdii> *fleet
516 2013-03-26 15:35:38 <sipa> no, 100%
517 2013-03-26 15:35:46 <sipa> of every single client
518 2013-03-26 15:35:52 <qdii> I don???t see why
519 2013-03-26 15:35:56 <sipa> bitcoin is not a democracy
520 2013-03-26 15:36:20 <sipa> it's a consensus system where every node verifies every single rule individually, and each must reach the exact same conclusion
521 2013-03-26 15:36:28 <qdii> I know, but if more than 50% of the clients and mining chain switch to the new version, then the longest chain of blocks is preserved
522 2013-03-26 15:36:44 <qdii> so eventually, any client who upgrades will get it
523 2013-03-26 15:36:44 <sipa> but the new chain won't be valid to any old client
524 2013-03-26 15:36:48 <sipa> they'll just ignore it
525 2013-03-26 15:36:52 <qdii> yea, until they upgrade
526 2013-03-26 15:36:53 <sipa> and you have a fork
527 2013-03-26 15:37:01 <grapevine> forks are bad mkay
528 2013-03-26 15:37:10 <abadr> sipa: isn't that what's happening with the block size increase?
529 2013-03-26 15:37:13 <sipa> qdii: from their point of view, the others are simpy buggy
530 2013-03-26 15:37:33 <sipa> abadr: it's similar, but worse
531 2013-03-26 15:38:25 <sipa> abadr: in that switching hashing function is incompatible in both ways, immediately (old don't accept new, new don't accept old)
532 2013-03-26 15:38:47 <sipa> abadr: increasing the block limit would make blocks mined by old clients still accepted by new ones
533 2013-03-26 15:39:03 <abadr> got it
534 2013-03-26 15:39:24 <sipa> qdii: but 'majority' doesn't matter here: as soon as some nodes switch, they'll start a chain that is ignored by the others
535 2013-03-26 15:39:38 <sipa> independent of how much hash power there is on either side, there will be two sides
536 2013-03-26 15:40:14 <sipa> and if the minority (whatever side that is) chooses not to switch to the rules of the majority, there will be two chains
537 2013-03-26 15:40:28 <sipa> so no, the only way to change something like that is when everyone upgrades
538 2013-03-26 15:40:29 <qdii> sipa: okay, so say there is an upgrade that changes SHA256 to SHA512. The new clients start a chain that is ignored by the rest. Is that chain discarded?
539 2013-03-26 15:40:43 <sipa> define 'discarded' ?
540 2013-03-26 15:41:16 <qdii> the new clients will keep it, the old clients will discard it
541 2013-03-26 15:41:24 <qdii> consider it false
542 2013-03-26 15:41:32 <sipa> yes
543 2013-03-26 15:41:50 <sipa> old clients will simply consider those blocks are weird pieces of data they see on the network
544 2013-03-26 15:41:57 <sipa> and the other way around
545 2013-03-26 15:42:06 <Diablo-D3> sha512? no
546 2013-03-26 15:42:09 <Diablo-D3> if we go that route
547 2013-03-26 15:42:21 <sipa> also, there's absolutely no reason :)
548 2013-03-26 15:42:23 <Diablo-D3> I make my own hash that does 1024
549 2013-03-26 15:42:29 <sipa> moar bits!!!
550 2013-03-26 15:42:42 <Diablo-D3> sipa: anti-qc :<
551 2013-03-26 15:43:21 <sipa> we have no idea what size of state quantum computers will ever be able to manage
552 2013-03-26 15:43:22 <qdii> okay let's go one step deeper:  if there ever WAS to be an update, and we found ourselves with two coexisting versions of bitcoin. What happen to the transactions
553 2013-03-26 15:43:47 <qdii> theoretically. kick me off the chan if that's too boring :p
554 2013-03-26 15:44:08 <sipa> qdii: if there are two chains due to a disagreement between two sides of the network of nontrivial size, imho, bitcoin is dead
555 2013-03-26 15:44:15 <sipa> as it means the consensus fails
556 2013-03-26 15:44:33 <qdii> sipa: it only fails until all the clients upgrade
557 2013-03-26 15:44:40 <qdii> this is a temporary situation
558 2013-03-26 15:44:53 <sipa> of course, so people have a very strong incentive to agree
559 2013-03-26 15:44:58 <sipa> and compromise
560 2013-03-26 15:45:10 <sipa> but if a disagreement persists, we're done
561 2013-03-26 15:45:18 <gavinandresen> When we have to hard-fork the network, we'll make sure there is consensus to upgrade and won't switch until there is consensus.
562 2013-03-26 15:45:43 <qdii> sure, I am talking about this case??:)
563 2013-03-26 15:45:57 <qdii> what happen to the transaction occurring during the upgrade time
564 2013-03-26 15:46:04 <sipa> qdii: anyway, what would happen technically, is that you literally get two chains
565 2013-03-26 15:46:13 <sipa> transactions can exist on either or both sides independently
566 2013-03-26 15:46:28 <qdii> okay, so that makes the upgrade possible
567 2013-03-26 15:46:44 <sipa> in case there are two conflicting transactions (double spends) one version can be confirmed on one side, and the other on another side
568 2013-03-26 15:46:59 <sipa> every coin that existed before the split can be spent independently on both sides