1 2013-02-25 00:00:11 rbecker is now known as RBecker
   2 2013-02-25 00:02:45 Phoebus has quit (Ping timeout: 264 seconds)
   3 2013-02-25 00:04:00 X-Scale has left ()
   4 2013-02-25 00:08:32 b4tt3r135 has joined
   5 2013-02-25 00:12:51 Phoebus has joined
   6 2013-02-25 00:13:30 bitzi has joined
   7 2013-02-25 00:15:07 <HM> I don't know what I'd do without Boost
   8 2013-02-25 00:15:29 <bitzi> ii've been running bitcoind from git master and rebuilding once a month. after running today's build, i see the memory use has climbed from 250MB to 450MB. has something changed recently to cause this? i
   9 2013-02-25 00:15:42 <bitzi> i.e. is this the new expected size?
  10 2013-02-25 00:16:00 RBecker is now known as rbecker
  11 2013-02-25 00:18:21 Phoebus has quit (Ping timeout: 264 seconds)
  12 2013-02-25 00:18:37 bitzi has quit (Quit: Page closed)
  13 2013-02-25 00:18:59 <gmaxwell> doh
  14 2013-02-25 00:19:05 <gmaxwell> two minutes!
  15 2013-02-25 00:19:10 <gmaxwell> can people not wait two minutes?
  16 2013-02-25 00:19:25 <gmaxwell> In any case, if he comes back— ask if it could just be his connection count.
  17 2013-02-25 00:21:24 * JWU42 nods
  18 2013-02-25 00:22:37 emryss has quit (Ping timeout: 240 seconds)
  19 2013-02-25 00:25:28 frosks has joined
  20 2013-02-25 00:26:02 emryss has joined
  21 2013-02-25 00:26:39 rs0 has quit (Quit: leaving)
  22 2013-02-25 00:28:45 axhlf has quit (Ping timeout: 256 seconds)
  23 2013-02-25 00:29:08 rbecker is now known as RBecker
  24 2013-02-25 00:29:27 rabbit_ has quit (Ping timeout: 245 seconds)
  25 2013-02-25 00:31:07 mareo87 has quit (Ping timeout: 245 seconds)
  26 2013-02-25 00:34:35 Phoebus has joined
  27 2013-02-25 00:39:52 ProfMac has quit (Ping timeout: 245 seconds)
  28 2013-02-25 00:41:32 forloop has quit (Ping timeout: 245 seconds)
  29 2013-02-25 00:42:28 da2ce7_d has joined
  30 2013-02-25 00:42:39 ProfMac has joined
  31 2013-02-25 00:43:18 MC1984 has quit (Quit: Leaving)
  32 2013-02-25 00:43:35 monkeynipples has quit (Remote host closed the connection)
  33 2013-02-25 00:43:38 <Luke-Jr> how does a typo fix decrease test coverage? -.-
  34 2013-02-25 00:44:04 <BlueMatt> Luke-Jr: link?
  35 2013-02-25 00:44:16 da2ce7 has quit (Ping timeout: 260 seconds)
  36 2013-02-25 00:44:26 <Luke-Jr> BlueMatt: https://github.com/bitcoin/bitcoin/pull/2335
  37 2013-02-25 00:44:41 <BlueMatt> hmm...odd, not sure about that one
  38 2013-02-25 00:44:58 <BlueMatt> maybe i forgot to make it < instead of <=...Ill look into it later
  39 2013-02-25 00:45:49 FredEE has quit (Quit: FredEE)
  40 2013-02-25 00:49:59 blinkier has quit (Ping timeout: 246 seconds)
  41 2013-02-25 00:55:57 nouitfvf has quit (Read error: Connection timed out)
  42 2013-02-25 00:56:41 nouitfvf has joined
  43 2013-02-25 00:57:05 * HM mulls over making json output the default for his command line app or just sticking to something even easier to parse
  44 2013-02-25 00:57:19 * HM thinks he'll stick to lines
  45 2013-02-25 00:57:38 D34TH_ has joined
  46 2013-02-25 00:57:42 D34TH has quit (Read error: Connection reset by peer)
  47 2013-02-25 00:59:01 <CodeShark> for lists, lines are good enough :)
  48 2013-02-25 00:59:22 tg has quit (Read error: Connection reset by peer)
  49 2013-02-25 00:59:45 <CodeShark> or for unnested key:value pairs
  50 2013-02-25 01:00:11 <HM> well i can get away with deliimited lines
  51 2013-02-25 01:00:24 <HM> key->value pairs
  52 2013-02-25 01:00:33 <HM> detail detail for now
  53 2013-02-25 01:00:42 <CodeShark> JSON is more powerful once you start having nested structures
  54 2013-02-25 01:01:17 <HM> the problem is i'm much a believer in the unix philosophy of do one thing and do it well
  55 2013-02-25 01:01:26 <HM> adding json might be necessary but i'd rather wait and see
  56 2013-02-25 01:01:31 <HM> it's not usable for piping
  57 2013-02-25 01:01:47 <CodeShark> unless you use a json stream parser
  58 2013-02-25 01:03:04 <CodeShark> in any case, formatting JSON output is simple :)
  59 2013-02-25 01:03:26 tg has joined
  60 2013-02-25 01:03:41 <CodeShark> but parsing it is slightly trickier than just newlines
  61 2013-02-25 01:09:32 FredEE has joined
  62 2013-02-25 01:11:45 tg has quit (Read error: Connection reset by peer)
  63 2013-02-25 01:12:33 tg has joined
  64 2013-02-25 01:13:23 da2ce7_d has quit (Ping timeout: 252 seconds)
  65 2013-02-25 01:14:52 da2ce7 has joined
  66 2013-02-25 01:15:13 <CodeShark> doesn't sound like a fundamental design decision, though :)
  67 2013-02-25 01:17:26 <HM> how're you CodeShark?
  68 2013-02-25 01:17:41 <gmaxwell> Luke-Jr: might also be that some of the tests are not determinstic.
  69 2013-02-25 01:17:50 <CodeShark> HM: extraordinary
  70 2013-02-25 01:18:08 <CodeShark> (not really, but could be much worse)
  71 2013-02-25 01:19:35 tg has quit (Quit: Changing server)
  72 2013-02-25 01:20:21 tg has joined
  73 2013-02-25 01:25:12 da2ce7_d has joined
  74 2013-02-25 01:25:25 <HM> lol ok
  75 2013-02-25 01:25:44 copumpkin has quit (Ping timeout: 260 seconds)
  76 2013-02-25 01:25:57 da2ce7_d has quit (Read error: Connection reset by peer)
  77 2013-02-25 01:26:19 copumpkin has joined
  78 2013-02-25 01:26:50 da2ce7_d has joined
  79 2013-02-25 01:27:33 da2ce7 has quit (Ping timeout: 252 seconds)
  80 2013-02-25 01:29:04 <ainfo> what's the startup time for a "simplified verification" btc client?
  81 2013-02-25 01:29:09 dario1 is now known as rdponticelli1
  82 2013-02-25 01:29:28 da2ce7 has joined
  83 2013-02-25 01:30:06 ovidiusoft has quit (Quit: leaving)
  84 2013-02-25 01:30:33 <ainfo> i don't get how the transactions make bitcoin addresses yet; is it that the term address comes into existence through the transaction having a script that requires a specific public key to match, which is the address?
  85 2013-02-25 01:31:55 <CodeShark> no, an address is actually a hash of an ECDSA public key
  86 2013-02-25 01:32:20 da2ce7 has quit (Client Quit)
  87 2013-02-25 01:32:45 da2ce7_d has quit (Ping timeout: 264 seconds)
  88 2013-02-25 01:32:55 <ainfo> codeshark: right.  now, based on this and a copy of the blockchain, how do you determine the current balance for your bitcoin address, and what's the content of a transaction command as to transfer btc:s from your address to another one?
  89 2013-02-25 01:33:06 <CodeShark> addresses don'
  90 2013-02-25 01:33:11 <HM> you don't determine a balance
  91 2013-02-25 01:33:13 <CodeShark> addresses don't have balances
  92 2013-02-25 01:33:22 <CodeShark> each transaction has outputs
  93 2013-02-25 01:33:31 <CodeShark> by signing a transaction, you claim an output
  94 2013-02-25 01:33:37 <CodeShark> you claim it in its entirety
  95 2013-02-25 01:33:48 <jaakkos> ainfo: the balance is the sum of outputs that your private keys can potentially redeem
  96 2013-02-25 01:33:58 <CodeShark> if you don't want to spend it all in the same transaction, you need to send change in a new output
  97 2013-02-25 01:34:21 <CodeShark> each output can be claimed exactly once
  98 2013-02-25 01:34:59 <CodeShark> a "balance" is actually a sum over all the unspent outputs claimable by the holder of the private key corresponding to the public key that hashes to that address
  99 2013-02-25 01:35:10 <amiller> the "state" of bitcoin consists of a set of coins, each coin associates a value with a scriptpubkey
 100 2013-02-25 01:35:31 <amiller> a valid transaction removes some coins from this set, and adds some new ones, such that the sum of all the values of all the coins stays the same!
 101 2013-02-25 01:36:18 <ainfo> right. hmm.   claimability is defined by the unspent output transaction's script, no?
 102 2013-02-25 01:36:20 <amiller> the blockchain is like secure log of operations, but the UTXO set is like the state data that the operations modify
 103 2013-02-25 01:36:29 <CodeShark> yes, ainfo
 104 2013-02-25 01:37:12 <ainfo> hmhm right.
 105 2013-02-25 01:37:28 <HM> and all of it's glorious problems stem from this design :)
 106 2013-02-25 01:37:52 <ainfo> in "simplified payment verification mode", say that i want to import 1000 privkeys as to calculate their btc addresses' balances, and make transactions.
 107 2013-02-25 01:38:16 <ainfo> what's the mechanics/operatiosn/alike needed to make this happen? =)  (didn't get exactly how this simplified payment verif. mode steers throught his yet)
 108 2013-02-25 01:38:36 <CodeShark> you add up all unspent outputs claimable by these privkeys
 109 2013-02-25 01:38:47 <ainfo> (or if it can at all, but needs a non-broken wallet.dat kind of file to function w.o. needing to ruin in fully verif. mode9
 110 2013-02-25 01:39:04 <ainfo> codeshark: does that require a local copy of the blockchain`?
 111 2013-02-25 01:39:06 <jaakkos> CodeShark: but you don't know those in spv since you don't have the whole chain
 112 2013-02-25 01:39:17 <CodeShark> yes, ainfo
 113 2013-02-25 01:39:38 <CodeShark> correct, jaakkos - you would need the whole chain
 114 2013-02-25 01:39:45 <ainfo> aha. so "simplified payment verification mode" requires the user not to trash the wallet.dat kind of file. aha.
 115 2013-02-25 01:40:02 <jaakkos> you must never trash that...
 116 2013-02-25 01:40:09 <jaakkos> your private keys are gone if you do that
 117 2013-02-25 01:40:18 <ainfo> jaakkos: well it is regeneratable?
 118 2013-02-25 01:40:20 <jaakkos> no
 119 2013-02-25 01:40:22 <CodeShark> you can still reconstruct your transaction history from public keys
 120 2013-02-25 01:40:22 <jaakkos> it's not
 121 2013-02-25 01:40:26 <ainfo> well, i meant, if you have the privkey backedup elsewhere
 122 2013-02-25 01:40:27 <CodeShark> but you cannot reconstruct your accounts
 123 2013-02-25 01:40:52 <jaakkos> ainfo: backups of course always help
 124 2013-02-25 01:40:55 <CodeShark> so having the private keys is enough to reconstruct the wallet except for the accounts
 125 2013-02-25 01:40:58 <ainfo> waht is the unique information in a wallet.dat?
 126 2013-02-25 01:41:05 <jaakkos> the private keys
 127 2013-02-25 01:41:09 <ainfo> ah right, if in SPV and I have a one-week-old wallet.dat, what happens?
 128 2013-02-25 01:41:09 <CodeShark> private keys, transactions, and accounts
 129 2013-02-25 01:41:41 <ainfo> codeshark: the accounts are purely a bitcoind-internal abstraction isn't it, it has nothing to do with how the bitcoin protocol and network and system works between them right?
 130 2013-02-25 01:41:47 <CodeShark> correct, ainfo
 131 2013-02-25 01:42:16 FredEE has quit (Quit: FredEE)
 132 2013-02-25 01:42:18 <ainfo> the transactions in wallet.dat, is that *all* transactions made *ever* that go from and to all of the contained btc addresses?
 133 2013-02-25 01:42:25 <CodeShark> yes
 134 2013-02-25 01:42:43 <ainfo> so if the user made a Zillion transactions, then the wallet.dat will have a size corresponding to that. mhm.
 135 2013-02-25 01:42:48 <CodeShark> indeed
 136 2013-02-25 01:43:12 <CodeShark> the block chain storage in bitcoind is not indexed by address
 137 2013-02-25 01:43:21 DaQatz has joined
 138 2013-02-25 01:43:22 <CodeShark> so the wallet stores its own copy of transactions
 139 2013-02-25 01:43:40 <ainfo> technically what's the difference between a SPV clietn and a fully verifying client? i mean, do they do the same requests to peers or different? does the SPV filter input from the network another way?    how does the SPV tell the network what btc addresses it wants updates for?
 140 2013-02-25 01:43:56 <ainfo> codeshark: right
 141 2013-02-25 01:43:56 <CodeShark> SPV only checks proof of work
 142 2013-02-25 01:44:08 <CodeShark> fully verifying also checks transaction signatures
 143 2013-02-25 01:44:32 <jaakkos> ainfo: the SPV node tells a bloom filter to its neighbors so that the SPV node will only receive relevant traffic
 144 2013-02-25 01:44:58 <jaakkos> relevant as in, only new block headers and transactions that can affect the SPV node's accounts
 145 2013-02-25 01:45:17 <CodeShark> ainfo: to validate a block, you need to 1) check the header and make sure that it satisfies proof-of-work, 2) check the transactions to make sure the inputs connect
 146 2013-02-25 01:45:34 <CodeShark> SPV clients don't do (2) for all transactions
 147 2013-02-25 01:46:08 <ainfo> does a SPV client have *any* blockchain database copy (beyond exactly the data that's in the wallet dat)?
 148 2013-02-25 01:46:33 <CodeShark> it could - but it only really needs the block headers and a merkle tree for its own transactions
 149 2013-02-25 01:46:44 <ainfo> what kind of startup time is bad and good respectively for a SPV client to start being able to transact?
 150 2013-02-25 01:46:56 nanotube has quit (Remote host closed the connection)
 151 2013-02-25 01:46:56 gribble has quit (Read error: Connection reset by peer)
 152 2013-02-25 01:47:04 <ainfo> codeshark: aha, and all of that is in wallet dat right?
 153 2013-02-25 01:47:11 D34TH_ is now known as D34TH
 154 2013-02-25 01:47:19 D34TH has quit (Changing host)
 155 2013-02-25 01:47:19 D34TH has joined
 156 2013-02-25 01:48:03 rs0 has joined
 157 2013-02-25 01:48:28 <BlueMatt> gmaxwell: so, ignoring size and utxo_created, and assuming you include the utxo_spents in the sum, you need an f(a, b) to satisfy http://pastebin.com/1rEAbxSJ correct?
 158 2013-02-25 01:48:37 <BlueMatt> well, the last one is just to make me happy :)
 159 2013-02-25 01:49:02 <ainfo> jaakos: this instruction-information (i.e. asking for only new block h.s and trx.es that can affect the SPV node's accounts), what does it consists of?  ah, and noob q, what atre the bloom filters about? =)
 160 2013-02-25 01:49:07 <ainfo> this weekend has been btc weekend. :)
 161 2013-02-25 01:50:14 <jaakkos> ainfo: bloom filters are a sort of probabilistic set structures
 162 2013-02-25 01:50:14 nanotube has joined
 163 2013-02-25 01:50:38 <jaakkos> the filter itself is a bit string
 164 2013-02-25 01:51:14 <jaakkos> elements of some sort (data that can be hashed) can be tested against the filter, whether it seems to belong to the set or not
 165 2013-02-25 01:51:31 <jaakkos> and (infinite amount of!) elements can be added to the filter.
 166 2013-02-25 01:51:55 gribble has joined
 167 2013-02-25 01:51:59 <CodeShark> the purpose of the bloom filters is to allow clients to tell peers to only send them transactions relevant to them. but bloom filters also have false positives, so the client will get not just their own transactions - however, this also improves privacy
 168 2013-02-25 01:52:30 <CodeShark> so you get better privacy and the peer doesn't have to store the entire list of the client addresses
 169 2013-02-25 01:53:27 <ainfo> cool.
 170 2013-02-25 01:54:08 <ainfo> just for some simple but robust SPV implementation, which ought to be best and most used nowadays? BitcoinJ, BitcoinJS, BitcoinPY, some other?
 171 2013-02-25 01:54:16 <gmaxwell> BlueMatt: f() being the combining function?
 172 2013-02-25 01:54:57 <BlueMatt> gmaxwell: yes, a being value, b being utxo_size, sorry
 173 2013-02-25 01:56:05 moore_ has joined
 174 2013-02-25 01:56:13 <gmaxwell> BlueMatt: yes, the last is the criteria you want that I don't care about.  It was incompatible with my concept of line (2), but I see you've stated what I _really_ want there in a less constrained way.
 175 2013-02-25 01:56:35 <epscy> how do start a testnet node, -testnet?
 176 2013-02-25 01:57:57 B0g4r7 has quit (Ping timeout: 276 seconds)
 177 2013-02-25 01:59:10 <BlueMatt> gmaxwell: well, that requirement + the requirement that we divide by utxo_created or similar
 178 2013-02-25 01:59:31 <BlueMatt> sadly such a function is very non-obvious
 179 2013-02-25 02:01:18 <gmaxwell> BlueMatt: I'd be happy to just get that in the numerator.
 180 2013-02-25 02:01:21 <gmaxwell> epscy: yes.
 181 2013-02-25 02:02:08 <gmaxwell> BlueMatt: basically, I want 'creating more outputs never gives you more juice for future transactions'
 182 2013-02-25 02:03:25 b4tt3r135 has quit (Quit: Nettalk6 - www.ntalk.de)
 183 2013-02-25 02:04:19 da2ce7 has joined
 184 2013-02-25 02:04:44 <BlueMatt> gmaxwell: yea, plus the (obvious) spending utxo is "good" in the numerator
 185 2013-02-25 02:04:59 <BlueMatt> gmaxwell: also, the suggestion from earlier today is breaks #1
 186 2013-02-25 02:05:35 B0g4r7 has joined
 187 2013-02-25 02:06:27 <gmaxwell> How so? You get peanalized some multiple of your txouts you create via the denominator.
 188 2013-02-25 02:09:49 <ainfo> there's no uissue having one and the same wallet on two machines concurrently, is there? (except for some minior lag in updating of the respective BTC address' balance between the two machines)
 189 2013-02-25 02:11:00 <HM> just twice the risk of being hacked
 190 2013-02-25 02:11:04 <BlueMatt> gmaxwell: ex outputs 100,000 blocks deep, 10 bytes large: (10*100,000+10*100,000)/30 vs 10*100,000/20 (assuming even 10 bytes per txin/txout, very simplified but it shows the point
 191 2013-02-25 02:11:05 da2ce7_d has joined
 192 2013-02-25 02:11:39 <BlueMatt> gmaxwell: evenif you said 20 bytes of size per txin and 10 per txout, the same example holds
 193 2013-02-25 02:14:14 <BlueMatt> gmaxwell: admittedly it only breaks #1 for very old inputs, which I would say doesnt matter much (since you have high priority anyway) but....
 194 2013-02-25 02:14:15 <ainfo> to protect rom brokenr andom number generator or alike, is there any convenient way to check that your pubkey does not collide with any pubkey already used?
 195 2013-02-25 02:14:18 da2ce7 has quit (Ping timeout: 256 seconds)
 196 2013-02-25 02:14:50 <gmaxwell> BlueMatt: indeed, works with 20 / 10 in the denominator which is what I was thinking though it was wrong.
 197 2013-02-25 02:16:12 freewil has quit (Ping timeout: 245 seconds)
 198 2013-02-25 02:17:03 <BlueMatt> gmaxwell: frankly, I dont like having size in the priority algorithm at all
 199 2013-02-25 02:17:19 da2ce7 has joined
 200 2013-02-25 02:18:19 <gmaxwell> BlueMatt: fundimentally what the priority algorithim needs to do is balance finite resources the user is burning by transacting against the cost to the network to take their transaction.
 201 2013-02-25 02:18:36 <gmaxwell> And it should do so in a way that doesn't create perverse incentives.
 202 2013-02-25 02:18:57 <BlueMatt> gmaxwell: utxo size, ofc, but size doesnt matter all that much in the scheme of things, and priority is only used (well if we go with some of the suggestions in the fee gists stuff from like 9 months ago) to pick txn which are the "best" to overwrite txn which have high fee/size
 203 2013-02-25 02:19:45 <BlueMatt> gmaxwell: Id rather see a huge txn do that that spends a ton of utxo than otherwise
 204 2013-02-25 02:19:58 da2ce7_d has quit (Ping timeout: 256 seconds)
 205 2013-02-25 02:21:56 da2ce7_d has joined
 206 2013-02-25 02:22:33 <BlueMatt> gmaxwell: though realistically, if you divide by size, you get a neater algorithm to replace txn since you will then always get highest total priority replaced
 207 2013-02-25 02:22:45 FredEE has joined
 208 2013-02-25 02:22:48 da2ce7 has quit (Ping timeout: 256 seconds)
 209 2013-02-25 02:23:00 <BlueMatt> (as long as it makes sense to target highest total priority given the algorithm itself)
 210 2013-02-25 02:28:31 <HM> sha256 is slower than i thought on cpu
 211 2013-02-25 02:29:17 <BlueMatt> gmaxwell: in any case, I still kinda like the simplicity of the suggestion from earlier today, if I get around to it, Ill create some fictional transactions to see if I can trip it up
 212 2013-02-25 02:29:39 <ainfo> what's an expectable startup time for a SPV client that was offline for a week, before it can start making transactions?
 213 2013-02-25 02:31:17 <gmaxwell> ainfo: making transactions? as soon as you can get an rpc into it you can make a transaction
 214 2013-02-25 02:31:27 <gmaxwell> s/rpc/command/
 215 2013-02-25 02:31:37 <gmaxwell> You don't even need to be online to make a transaction with coins you already know about.
 216 2013-02-25 02:31:48 <ainfo> gmaxwell: what if the wallet dat is a bit out of sync with the network, so there were actually more transactions made for contained btc addresses?
 217 2013-02-25 02:32:11 <ainfo> gmaxwell: or, new btc:s arrived from other parties
 218 2013-02-25 02:32:20 <gmaxwell> ainfo: well, then you couldn't spend those until they arrived.
 219 2013-02-25 02:32:30 <gmaxwell> but you could spend what you already had.
 220 2013-02-25 02:32:43 <ainfo> gmaxwell: there's some kind of sycnrhonization driving mechanism right?
 221 2013-02-25 02:32:47 <ainfo> mm
 222 2013-02-25 02:35:42 <HM> heh
 223 2013-02-25 02:35:49 <HM> bitcoin atm story hit top of slashdot
 224 2013-02-25 02:37:08 <HM> i would say it has to be relatively easy to disassemble such an atm and steal the auth details needed to steal all the bitcoin as well as deposited notes.
 225 2013-02-25 02:37:37 <HM> hopefully each atm has its own wallet so that the value of breaking in remains constant
 226 2013-02-25 02:38:08 <HM> it looks like a tin can :|
 227 2013-02-25 02:38:46 tg has quit (Ping timeout: 264 seconds)
 228 2013-02-25 02:46:41 D34TH has quit (Read error: Connection reset by peer)
 229 2013-02-25 02:47:05 grau has joined
 230 2013-02-25 02:48:38 RazielZ has quit (Ping timeout: 246 seconds)
 231 2013-02-25 02:50:21 da2ce7_d is now known as da2ce7
 232 2013-02-25 02:54:26 HiroWhite has joined
 233 2013-02-25 02:59:17 <hasha> i love slashdot comments. never fail to entertain.
 234 2013-02-25 03:00:07 <hasha> HM, my immediate thoughts were "this thing looks too portable". thats prolly not a good thing.
 235 2013-02-25 03:00:26 RBecker is now known as rbecker
 236 2013-02-25 03:02:07 <BlueMatt> also, this isnt the first bitcoin ATM iirc
 237 2013-02-25 03:02:22 <BlueMatt> they're like a year and a half late...
 238 2013-02-25 03:04:01 <hasha> well, its the first one in ugly orange
 239 2013-02-25 03:04:23 <BlueMatt> lol, they didnt even google "bitcoin atm" theres one from September 2011 at like the 3rd result....
 240 2013-02-25 03:04:28 <BlueMatt> journalism really is dead
 241 2013-02-25 03:04:43 <hasha> sadly
 242 2013-02-25 03:04:47 <BlueMatt> oh, nope sorry, its the first fucking result
 243 2013-02-25 03:06:42 <BlueMatt> well, luckily it was just /.'s usual stupidity that added "world's first"
 244 2013-02-25 03:07:31 <hasha> i dont quite understand the value of a ATM.. is it for people who have trouble with mouseclicks?
 245 2013-02-25 03:07:54 <BlueMatt> well, you can use cash
 246 2013-02-25 03:08:21 <hasha> unlike everyone on localbitcoin
 247 2013-02-25 03:08:33 Elio19 has joined
 248 2013-02-25 03:08:34 <hasha> i guess that takes a little more thought and organization
 249 2013-02-25 03:08:41 <BlueMatt> more automated/convenient
 250 2013-02-25 03:08:42 <BlueMatt> yep
 251 2013-02-25 03:09:09 <Elio19> If i want to get into the nitty gritty of BTC protocol so i may understand it from the UI the packet level, what is a good way to do that? For example, a way discover possible hidden messages in addresses (or other places) of bitcoin transactions.
 252 2013-02-25 03:09:52 HiroWhite has quit (Ping timeout: 260 seconds)
 253 2013-02-25 03:09:52 <BlueMatt> if you want to find hidden messages... strings -n10 blk*.dat
 254 2013-02-25 03:10:30 <Elio19> Thanks BlueMatt, ill try that now
 255 2013-02-25 03:12:35 <Elio19> "The Times 03/Jan/2009 Chancellor on brink of second bailout for banks"
 256 2013-02-25 03:12:50 <BlueMatt> yes, that is in the genesis block
 257 2013-02-25 03:13:06 <BlueMatt> in fact, even hardcoded in the source
 258 2013-02-25 03:13:12 <hasha> nice one! i didnt know about that.
 259 2013-02-25 03:15:08 <HM> look for bernanke
 260 2013-02-25 03:15:08 <Elio19> Some ascii art in there
 261 2013-02-25 03:15:11 <HM> he's in there somewhere
 262 2013-02-25 03:15:30 <BlueMatt> ASCII bernake
 263 2013-02-25 03:16:06 <Elio19> Lots of religious stuff.
 264 2013-02-25 03:16:09 <Elio19> This is neat!
 265 2013-02-25 03:16:27 <Elio19> i did -n20 because 10 gave to much noise
 266 2013-02-25 03:16:29 <BlueMatt> the religious stuff is luke
 267 2013-02-25 03:16:35 Tycale is now known as Nietzschtml
 268 2013-02-25 03:17:22 <Elio19> luke who?
 269 2013-02-25 03:17:29 <BlueMatt> Luke-Jr
 270 2013-02-25 03:17:30 <hasha> oh, seriously?
 271 2013-02-25 03:17:45 <Luke-Jr> ⁇⁇⁇⁇
 272 2013-02-25 03:17:48 <BlueMatt> eliopool puts religious stuff in its coinbases, or used to
 273 2013-02-25 03:17:54 <BlueMatt> dont know
 274 2013-02-25 03:18:13 <Luke-Jr> Eloipool is software
 275 2013-02-25 03:18:18 <BlueMatt> ok, eligius
 276 2013-02-25 03:18:21 <Luke-Jr> Eligius puts various things suggested by miners
 277 2013-02-25 03:18:28 <Luke-Jr> or, these days, whatever the miners put in with GBT
 278 2013-02-25 03:18:36 <Luke-Jr> --coinbase-sig option in BFGMiner
 279 2013-02-25 03:18:45 HiroWhite has joined
 280 2013-02-25 03:18:58 <Elio19> Luke-Jr, apparently you wrote religious stuff in the block-chain. Is it true?
 281 2013-02-25 03:19:02 ProfMac has quit (Ping timeout: 245 seconds)
 282 2013-02-25 03:19:44 <Luke-Jr> Elio19: I cannot claim credit for being the author of it, but I chose to encode it
 283 2013-02-25 03:23:52 moore_ has quit (Ping timeout: 256 seconds)
 284 2013-02-25 03:24:04 <ainfo> if i have a SPV client and it crashes or alike between it has received or sent a transaction, and that it saved the wallet.dat file - on the next start, will the wallet.dat autorepair?
 285 2013-02-25 03:24:27 <Elio19> "like everything else that wiggles or jiggles"
 286 2013-02-25 03:25:22 <Luke-Jr> ainfo: in theory
 287 2013-02-25 03:25:41 <ainfo> luke-jr: what about in practice, for instance in bitcoinj?
 288 2013-02-25 03:25:54 <Luke-Jr> ainfo: I'm not aware of bitcoinj using wallet.dat
 289 2013-02-25 03:26:06 eennaam has joined
 290 2013-02-25 03:26:14 <ainfo> luke-jr: ah no, but something similar, a completely equivalent concept with the same data in it
 291 2013-02-25 03:27:08 <ainfo> luke-jr: what's needed for an autorepair to hapepn, does it happen spontaneously as new updates come in to the client and it realizes it's out of sync and issues read commands that way, or does the client actively need to ask the network to send a new copy of transactions it made..or sth.. or?
 292 2013-02-25 03:27:23 <Luke-Jr> I know nothing about bitcoinj
 293 2013-02-25 03:28:38 ProfMac has joined
 294 2013-02-25 03:35:36 <ainfo> k
 295 2013-02-25 03:35:53 <ainfo> luke-jr: in another bitcoin client like bitcoind, what would bne the mechansim of autorepair?
 296 2013-02-25 03:36:38 <Luke-Jr> …
 297 2013-02-25 03:36:54 <Luke-Jr> ainfo: have you written any software?
 298 2013-02-25 03:39:10 <ainfo> luke-jr: sure
 299 2013-02-25 03:39:23 <ainfo> not bitcoin software though. i'm very happy reading sourcecode.
 300 2013-02-25 03:39:35 <Luke-Jr> Bitcoin is nothing special with regard to data integrity.
 301 2013-02-25 03:39:46 Goonie has quit (Ping timeout: 252 seconds)
 302 2013-02-25 03:40:53 <ainfo> luke-jr: ah no, i was just thinkign like, what protocol or software mechanism is used to initiate and drive the autorepair
 303 2013-02-25 03:41:17 <Luke-Jr> bitcoind historically just uses Berkley DB
 304 2013-02-25 03:41:21 <ainfo> luke-jr: for instance, the other nodes do not have a list of the transactions a particular client did so it's not like they'll send over their query results for what data a particular client should have
 305 2013-02-25 03:41:31 <Luke-Jr> but because BDB sucks, 0.8 moves the main database to LevelDB
 306 2013-02-25 03:41:35 <ainfo> luke-jr: ah no no, by autorepair i meant, not fixing a broken file format, but an outdated walled.dat file
 307 2013-02-25 03:41:36 <Luke-Jr> 0.9 will likely replace BDB for the wallet
 308 2013-02-25 03:41:59 <Luke-Jr> ainfo: it just looks for known addresses in the blocks it receives, no different than realtime running
 309 2013-02-25 03:42:12 <Luke-Jr> ainfo: I think you need to read bitcoin.pdf
 310 2013-02-25 03:42:21 <ainfo> luke-jr: i did
 311 2013-02-25 03:42:30 <ainfo> ah right
 312 2013-02-25 03:42:34 <ainfo> wait
 313 2013-02-25 03:43:29 <ainfo> luke-jr: the general bitcoin client behavior is to ensure you got the longest blockchaincopied to you. but what about in SPV?
 314 2013-02-25 03:44:19 <ainfo> (getting the longest blockchain would bring with it any updates to your transactions that you might not have been aware of and thus have a fully syncing effect)
 315 2013-02-25 03:45:12 PhantomSpark has quit (3!~kvirc@pool-71-251-16-105.nycmny.fios.verizon.net|Ping timeout: 276 seconds)
 316 2013-02-25 03:45:15 nym has quit (Ping timeout: 245 seconds)
 317 2013-02-25 03:45:19 pierre` has quit (Read error: Connection reset by peer)
 318 2013-02-25 03:45:33 vigilyn2 has joined
 319 2013-02-25 03:45:47 vigilyn has quit (Read error: Connection reset by peer)
 320 2013-02-25 03:46:07 pierre` has joined
 321 2013-02-25 03:46:33 <gmaxwell> BlueMatt: hows that pull work? if I filter out all the txn I won't get the txn tree.
 322 2013-02-25 03:46:43 <gmaxwell> and if I don't have tree— shall I bruteforce the block? :P
 323 2013-02-25 03:47:11 <BlueMatt> gmaxwell: hmm??
 324 2013-02-25 03:47:33 <BlueMatt> gmaxwell: if you set a bloom filter to accept everything, it just skips sending you txn you've already seen
 325 2013-02-25 03:47:56 <BlueMatt> that pull changes nothing, it only short-circuits logic that would burn cpu for no reason
 326 2013-02-25 03:48:05 <gmaxwell> BlueMatt: oh I see. I thought you meant setting it to nothing.
 327 2013-02-25 03:51:15 fiesh has quit (Ping timeout: 255 seconds)
 328 2013-02-25 03:51:59 pierre` has quit (Ping timeout: 264 seconds)
 329 2013-02-25 03:52:14 pierre` has joined
 330 2013-02-25 03:55:38 fiesh has joined
 331 2013-02-25 04:05:17 b4tt3r135 has joined
 332 2013-02-25 04:13:22 okok_ has joined
 333 2013-02-25 04:15:49 okok has quit (Ping timeout: 244 seconds)
 334 2013-02-25 04:16:01 okok_ is now known as okok
 335 2013-02-25 04:17:07 moore_ has joined
 336 2013-02-25 04:21:58 ralphtheninja has quit (Ping timeout: 264 seconds)
 337 2013-02-25 04:28:33 saivann has quit (Quit: Quitte)
 338 2013-02-25 04:29:47 PRab has joined
 339 2013-02-25 04:29:48 <b4tt3r135> Hello PRab
 340 2013-02-25 04:29:50 saivann has joined
 341 2013-02-25 04:29:50 <b4tt3r135> Hello saivann
 342 2013-02-25 04:32:23 alexwaters has joined
 343 2013-02-25 04:32:23 <b4tt3r135> Hello alexwaters
 344 2013-02-25 04:33:09 <b4tt3r135> debugging this simple greet script. did you see that message upon join, alexwaters?
 345 2013-02-25 04:34:34 <alexwaters> yes
 346 2013-02-25 04:34:39 <b4tt3r135> ty
 347 2013-02-25 04:34:55 <BlueMatt> b4tt3r135: can you debug your greet script on a non-public channel?
 348 2013-02-25 04:35:03 <b4tt3r135> my bad but sure
 349 2013-02-25 04:35:17 <b4tt3r135> was looking for one that would have more traffic joining so i didnt have to keep using another clienbt
 350 2013-02-25 04:35:20 b4tt3r135 has left ()
 351 2013-02-25 04:37:10 eennaam has left ()
 352 2013-02-25 04:41:13 Cory has quit ()
 353 2013-02-25 04:41:39 robocoin has quit (Quit: http://25.media.tumblr.com/0b1c2187e1e6949f77eb4479cc0b78b0/tumblr_mi6ao9uIwX1qc0cxpo1_500.png)
 354 2013-02-25 04:58:05 TheSeven has quit (Disconnected by services)
 355 2013-02-25 04:58:14 [7] has joined
 356 2013-02-25 04:59:30 Cory has joined
 357 2013-02-25 05:02:08 owowo has quit (Quit: sayonara)
 358 2013-02-25 05:05:14 TradeFortress has joined
 359 2013-02-25 05:16:53 freewil has joined
 360 2013-02-25 05:17:24 tg has joined
 361 2013-02-25 05:19:44 okok_ has joined
 362 2013-02-25 05:22:59 okok has quit (Ping timeout: 244 seconds)
 363 2013-02-25 05:23:08 okok_ is now known as okok
 364 2013-02-25 05:26:31 cheebydi has quit (Ping timeout: 245 seconds)
 365 2013-02-25 05:31:51 cheebydi has joined
 366 2013-02-25 05:36:36 <gmaxwell> gavinandresen: Your bias accusation is impolite. And you have the chain of causality backwards, I want to see off chain transactions happen because I don't want bitcoin to become centeralized— I was advocating them long before I'd ever spoken to Peter Todd. https://bitcointalk.org/index.php?topic=28565.msg359948#msg359948
 367 2013-02-25 05:37:39 <Luke-Jr> gmaxwell: O.o?
 368 2013-02-25 05:37:56 <gmaxwell> gavinandresen: if you're going to adopt a rhetorical approch of ad hominem in the discussion I can match, but I'm doubting that will be productive.
 369 2013-02-25 05:38:13 <gmaxwell> Luke-Jr: https://bitcoinfoundation.org/forum/index.php?/topic/96-open-ended-discussion-of-various-facets-of-the-block-size-limit/#entry917
 370 2013-02-25 05:41:06 moore_ has quit (Ping timeout: 245 seconds)
 371 2013-02-25 05:41:24 ThomasV has joined
 372 2013-02-25 05:44:42 nym has joined
 373 2013-02-25 05:46:34 reeep has joined
 374 2013-02-25 05:47:04 CryptoCoin has joined
 375 2013-02-25 05:50:19 toffoo has quit ()
 376 2013-02-25 05:50:36 FredEE has quit (Quit: FredEE)
 377 2013-02-25 05:50:48 <doublec> Is that forum restricted to foundation members only?
 378 2013-02-25 05:56:53 <CodeShark> not accepting new members...oh well...
 379 2013-02-25 05:57:03 grau has quit (Remote host closed the connection)
 380 2013-02-25 05:57:37 <CodeShark> I'm also curious now :p
 381 2013-02-25 06:04:39 <CodeShark> alright, I sent my membership fee - now how do I log in?
 382 2013-02-25 06:10:36 ThomasV has quit (Read error: Operation timed out)
 383 2013-02-25 06:11:06 ThomasV has joined
 384 2013-02-25 06:13:16 dvide has joined
 385 2013-02-25 06:14:34 swappermall_ has joined
 386 2013-02-25 06:14:47 swappermall_ is now known as Guest84621
 387 2013-02-25 06:15:58 Whitez- has joined
 388 2013-02-25 06:20:06 andytoshi has joined
 389 2013-02-25 06:35:56 meLon has quit (Ping timeout: 255 seconds)
 390 2013-02-25 06:38:43 Whitez- has quit ()
 391 2013-02-25 06:40:34 cheako has quit (Quit: Client exiting)
 392 2013-02-25 06:47:47 HM has quit (Ping timeout: 252 seconds)
 393 2013-02-25 06:48:16 QM has joined
 394 2013-02-25 06:48:18 HM has joined
 395 2013-02-25 06:49:44 <saivann> Ooooo, Bitcoin-qt 0.8.0 starts so fast now. Congrats to the development team ;)
 396 2013-02-25 06:52:54 <Diablo-D3> too bad bitcoin still doesnt use rate limiting
 397 2013-02-25 07:01:09 <gmaxwell> Luke-Jr: I responded... finally. https://bitcoinfoundation.org/forum/index.php?/topic/96-open-ended-discussion-of-various-facets-of-the-block-size-limit/#entry919
 398 2013-02-25 07:01:26 <gmaxwell> Diablo-D3: I'm looking and not seeing your pull request!
 399 2013-02-25 07:01:50 <gmaxwell> (if you'd like to work on one, I'd suggest you start with just a gui option to disable listening. That does 99.9% of the job on its own.
 400 2013-02-25 07:02:09 <Diablo-D3> gmaxwell: hurr. No, listening isnt the whole issue.
 401 2013-02-25 07:02:27 <Diablo-D3> gmaxwell: the network code is an absolute mess from what Ive seen, so Im probably the last person that should work on it
 402 2013-02-25 07:03:15 <gmaxwell> Diablo-D3: it's fairly straight forward— but listening indirectly addresses your concern.  The big bandwidth user is feeding blocks to newly bootstrapped peers. (esp now with people nuking their installs to install 0.8)
 403 2013-02-25 07:03:28 <gmaxwell> If you disable listening you'll almost never end up talking to one of those nodes.
 404 2013-02-25 07:03:48 <Diablo-D3> gmaxwell: yeah but I can just close the port on my router and get the same
 405 2013-02-25 07:04:08 ThomasV has quit (Ping timeout: 246 seconds)
 406 2013-02-25 07:04:26 <gmaxwell> If your problem was txn messages (pretty hard because of how efficient the INV systme is!) then the only way to improve that is to reduce your connections.. even if you rate limit out you'll get invs in in proportion to your links.
 407 2013-02-25 07:05:18 <Diablo-D3> gmaxwell: yeah, but that isnt really the correct response
 408 2013-02-25 07:05:33 <Diablo-D3> like, bittorrent does it this way, it pings remove clients, and then massively throttles if the latency is too high
 409 2013-02-25 07:06:12 reeep has quit ()
 410 2013-02-25 07:07:19 <gmaxwell> Diablo-D3: not sure that this helps much with the way we send traffic. Some expirementation would be useful.
 411 2013-02-25 07:07:40 <phantomcircuit> Diablo-D3, you're worried about bandwidth consumption?
 412 2013-02-25 07:07:44 <phantomcircuit> or something else
 413 2013-02-25 07:07:51 <Diablo-D3> phantomcircuit: yes
 414 2013-02-25 07:07:55 <Diablo-D3> its using too damned much
 415 2013-02-25 07:08:03 <Diablo-D3> even with wondershaper its really killing network performance
 416 2013-02-25 07:08:17 <phantomcircuit> if you're not accepting connections it uses almost no bandwidth once sync'd
 417 2013-02-25 07:08:38 <Diablo-D3> its not downstream bandwidth thats the problem, its upstream
 418 2013-02-25 07:08:45 <gmaxwell> shaping it like that will only make it worse.
 419 2013-02-25 07:08:49 <Diablo-D3> if I handicap the number of connections, then p2pool gets bent
 420 2013-02-25 07:09:03 <gmaxwell> well, it'll just abuse your peer, send the same amout of data, but do it over a longer period.
 421 2013-02-25 07:09:46 <Diablo-D3> hrm, p2pool feeds bitcoin stuff like txes right?
 422 2013-02-25 07:10:35 <Diablo-D3> Im wondering if I could just block bitcoin off completely and let p2pool feed it
 423 2013-02-25 07:11:09 <forrestv> Diablo-D3, it doesn't feed bitcoin blocks
 424 2013-02-25 07:11:29 <Diablo-D3> oh damn
 425 2013-02-25 07:11:43 <Diablo-D3> so maybe 1 connection? to a known central hub?
 426 2013-02-25 07:13:30 <phantomcircuit> Diablo-D3, there should be effectively zero upload
 427 2013-02-25 07:13:44 <phantomcircuit> unless you're sending transactions
 428 2013-02-25 07:14:10 <Diablo-D3> phantomcircuit: no, I always have at least a couple new clients leeching off of me to do entire chains
 429 2013-02-25 07:14:44 hekep has joined
 430 2013-02-25 07:15:13 coolfengyu has joined
 431 2013-02-25 07:15:44 Guest84621 has quit (Ping timeout: 256 seconds)
 432 2013-02-25 07:17:21 testnode9 has quit (Ping timeout: 245 seconds)
 433 2013-02-25 07:26:02 <Diablo-D3> okay so
 434 2013-02-25 07:26:10 <Diablo-D3> brand new clients shouldnt already know my IP, right?
 435 2013-02-25 07:27:10 <gmaxwell> Diablo-D3: if your IP has been public before then it will still circulate some.
 436 2013-02-25 07:28:27 <Diablo-D3> well, I dont want to turn off my port forwarding, but if thats the only way =/
 437 2013-02-25 07:30:30 <coolfengyu> v0.8 makes win7 freeze?
 438 2013-02-25 07:31:23 <gmaxwell> Diablo-D3: change your port number.
 439 2013-02-25 07:32:01 <Diablo-D3> gmaxwell: why? I can just stop forwarding the port
 440 2013-02-25 07:32:50 testnode9 has joined
 441 2013-02-25 07:34:33 HM has quit (Ping timeout: 252 seconds)
 442 2013-02-25 07:36:02 <hekep> my 0.8 & win7 on lenovo crashes after one hour of network syncing... too much processor usage ... I think lenovo is bad ...
 443 2013-02-25 07:36:09 Ramokk has quit ()
 444 2013-02-25 07:36:35 <coolfengyu> hekep: my asrock
 445 2013-02-25 07:38:20 HM has joined
 446 2013-02-25 07:38:40 <gmaxwell> Diablo-D3: you can, changing it might be simpler for you
 447 2013-02-25 07:38:53 <Diablo-D3> no, requires client restart for that
 448 2013-02-25 07:46:19 Nes-asleep is now known as Nesetalis
 449 2013-02-25 07:54:40 idstam has joined
 450 2013-02-25 07:58:07 veerboot has joined
 451 2013-02-25 07:59:00 alexwaters has quit (Quit: Leaving.)
 452 2013-02-25 08:06:00 ThomasV has joined
 453 2013-02-25 08:18:00 HM has quit (Ping timeout: 252 seconds)
 454 2013-02-25 08:18:21 Goonie has joined
 455 2013-02-25 08:18:29 HM has joined
 456 2013-02-25 08:18:50 techlife has quit (Read error: Connection reset by peer)
 457 2013-02-25 08:23:11 MobiusL has quit (Remote host closed the connection)
 458 2013-02-25 08:24:26 MobiusL has joined
 459 2013-02-25 08:25:53 grau has joined
 460 2013-02-25 08:28:08 ThomasV has quit (Ping timeout: 246 seconds)
 461 2013-02-25 08:28:39 techlife has joined
 462 2013-02-25 08:34:37 unknown45682 has quit ()
 463 2013-02-25 08:39:17 brwyatt is now known as brwyatt|Away
 464 2013-02-25 08:40:30 <sipa> right now, i'm actually against rate limiting, as it will decrease sync performamce for those fetching from you
 465 2013-02-25 08:41:04 <sipa> better have an option to disable serving altogether, and remove NODE_NETWORK as well
 466 2013-02-25 08:45:10 techlife has quit (Ping timeout: 256 seconds)
 467 2013-02-25 08:46:29 techlife has joined
 468 2013-02-25 08:49:21 unknown45682 has joined
 469 2013-02-25 08:54:31 ThomasV has joined
 470 2013-02-25 08:55:57 HM has quit (Ping timeout: 252 seconds)
 471 2013-02-25 08:56:15 reizuki__ has joined
 472 2013-02-25 08:56:15 reizuki__ has quit (Changing host)
 473 2013-02-25 08:56:15 reizuki__ has joined
 474 2013-02-25 08:58:19 HM has joined
 475 2013-02-25 09:00:49 grau has quit (Remote host closed the connection)
 476 2013-02-25 09:04:50 eoss has joined
 477 2013-02-25 09:04:52 DamascusVG has quit (Quit: I Quit - http://www.youtube.com/watch?v=9p97zsQ51Rw)
 478 2013-02-25 09:06:29 TradeFortress_ has joined
 479 2013-02-25 09:10:37 TradeFortress has quit (Ping timeout: 260 seconds)
 480 2013-02-25 09:11:20 JDuke128 has joined
 481 2013-02-25 09:19:45 mappum has quit (Ping timeout: 256 seconds)
 482 2013-02-25 09:21:58 TD_ has joined
 483 2013-02-25 09:24:31 TradeFortress has joined
 484 2013-02-25 09:25:21 m00p has joined
 485 2013-02-25 09:27:21 TradeFortress_ has quit (Ping timeout: 245 seconds)
 486 2013-02-25 09:34:19 <gmaxwell> ;;ticker
 487 2013-02-25 09:34:19 <gribble> BTCUSD ticker | Best bid: 30.20001, Best ask: 30.21688, Bid-ask spread: 0.01687, Last trade: 30.20000, 24 hour volume: 30174.00352563, 24 hour low: 29.20123, 24 hour high: 30.39877, 24 hour vwap: 29.86887
 488 2013-02-25 09:35:33 HM has quit (Ping timeout: 252 seconds)
 489 2013-02-25 09:38:18 HM has joined
 490 2013-02-25 09:44:48 TD_ has quit (Quit: TD_)
 491 2013-02-25 09:45:28 <gmaxwell> 28nm ? https://bitcointalk.org/index.php?topic=146371.0 crazy. the pricewar is going to be awesome.
 492 2013-02-25 09:51:15 Mad7Scientist has quit (Ping timeout: 255 seconds)
 493 2013-02-25 09:53:37 Mad7Scientist has joined
 494 2013-02-25 09:56:08 CaptainBlaze has joined
 495 2013-02-25 09:56:45 eoss has quit (Remote host closed the connection)
 496 2013-02-25 10:11:26 <topi`> are there any fixes for the vulnerability described here: http://bitslog.wordpress.com/2013/01/23/new-bitcoin-vulnerability-get-your-peer-public-addresses/
 497 2013-02-25 10:12:00 <topi`> it seems to me it's a pretty straightforward operation to find out the pubaddresses of a given peer
 498 2013-02-25 10:12:40 <sipa> fixed in 0.8, afaik
 499 2013-02-25 10:12:59 <topi`> so how does the penny flood prevention work in 0.8, then?
 500 2013-02-25 10:15:57 <topi`> well I guess my addresses are safe as long as I'm behind a nat...
 501 2013-02-25 10:17:21 <gmaxwell> being behind tor is better.
 502 2013-02-25 10:17:48 <gmaxwell> topi`: iirc the fix is to just not treat your own wallet addresses differently for that purpose.
 503 2013-02-25 10:21:15 LainZ has quit (Ping timeout: 260 seconds)
 504 2013-02-25 10:22:22 B0g4r7 has quit (Ping timeout: 276 seconds)
 505 2013-02-25 10:26:23 ThomasV has quit (Quit: Leaving)
 506 2013-02-25 10:27:30 B0g4r7 has joined
 507 2013-02-25 10:28:01 <sipa> topi`: nat is not a firewall :)
 508 2013-02-25 10:31:57 Mandrius has joined
 509 2013-02-25 10:34:20 <Mandrius> Good day everyone
 510 2013-02-25 10:35:42 <Mandrius> So i`m planing to lunch a online exchanger in my country. I`m planning to use Intersango framework
 511 2013-02-25 10:36:15 <Mandrius> Maybe someone already used it or could recommend something more?
 512 2013-02-25 10:39:49 <epscy> just as an aside, nodejs compiles from source on a raspberry pi, bitcoin 0.8 runs out of memory
 513 2013-02-25 10:40:50 <sipa> haha
 514 2013-02-25 10:42:23 <SomeoneWeird> lmao epscy
 515 2013-02-25 10:42:25 <SomeoneWeird> use bitcoinjs
 516 2013-02-25 10:42:32 <SomeoneWeird> seems fairly broken though
 517 2013-02-25 10:55:28 reizuki__ has quit (Quit: Konversation terminated!)
 518 2013-02-25 11:06:31 one_zero has quit ()
 519 2013-02-25 11:11:06 JDuke128 has quit (Quit: ["Textual IRC Client: www.textualapp.com"])
 520 2013-02-25 11:11:27 JDuke128 has joined
 521 2013-02-25 11:21:35 pooler has joined
 522 2013-02-25 11:21:35 pooler has quit (Changing host)
 523 2013-02-25 11:21:35 pooler has joined
 524 2013-02-25 11:41:36 monkeynipples has joined
 525 2013-02-25 11:42:25 rbecker is now known as RBecker
 526 2013-02-25 11:54:06 RBecker is now known as rbecker
 527 2013-02-25 12:00:11 ThomasV has joined
 528 2013-02-25 12:09:56 drizztbsd has joined
 529 2013-02-25 12:13:50 TradeFortress has quit (Ping timeout: 260 seconds)
 530 2013-02-25 12:14:44 ralphtheninja has joined
 531 2013-02-25 12:31:43 idstam has quit (Ping timeout: 276 seconds)
 532 2013-02-25 12:38:02 RazielZ has joined
 533 2013-02-25 12:42:07 JZavala has quit (Ping timeout: 276 seconds)
 534 2013-02-25 12:47:40 ralphtheninja has quit (Quit: Lost terminal)
 535 2013-02-25 12:48:35 asuk is now known as afk!~asuk@ec2-54-234-43-224.compute-1.amazonaws.com|asuk
 536 2013-02-25 13:02:14 agricocb has quit (Ping timeout: 252 seconds)
 537 2013-02-25 13:06:33 m00p has quit (Ping timeout: 255 seconds)
 538 2013-02-25 13:07:34 tg has quit (Ping timeout: 264 seconds)
 539 2013-02-25 13:10:56 grau has joined
 540 2013-02-25 13:11:13 JDuke128 has quit (Quit: ["Textual IRC Client: www.textualapp.com"])
 541 2013-02-25 13:12:28 nus-- has joined
 542 2013-02-25 13:12:32 daybyter has joined
 543 2013-02-25 13:16:56 nus- has quit (Ping timeout: 276 seconds)
 544 2013-02-25 13:17:18 tg has joined
 545 2013-02-25 13:19:27 m00p has joined
 546 2013-02-25 13:19:32 rdymac has joined
 547 2013-02-25 13:20:42 b4tt3r135 has joined
 548 2013-02-25 13:21:27 agricocb has joined
 549 2013-02-25 13:23:36 Zarutian has joined
 550 2013-02-25 13:24:20 rdymac has quit (Client Quit)
 551 2013-02-25 13:26:52 rdymac has joined
 552 2013-02-25 13:27:23 <HM> epscy: that'll be boosts fault
 553 2013-02-25 13:36:29 reizuki__ has joined
 554 2013-02-25 13:36:29 reizuki__ has quit (Changing host)
 555 2013-02-25 13:36:29 reizuki__ has joined
 556 2013-02-25 13:41:04 agricocb has quit (Quit: Leaving.)
 557 2013-02-25 13:42:39 m0mchil has quit (Ping timeout: 255 seconds)
 558 2013-02-25 13:42:49 datagutt has joined
 559 2013-02-25 13:46:17 monkeynipples has quit (Remote host closed the connection)
 560 2013-02-25 13:47:09 ralphtheninja has joined
 561 2013-02-25 13:48:06 PhantomSpark has joined
 562 2013-02-25 13:48:18 okok_ has joined
 563 2013-02-25 13:50:53 okok has quit (Ping timeout: 244 seconds)
 564 2013-02-25 13:50:57 okok_ is now known as okok
 565 2013-02-25 13:52:10 copumpkin has quit (Quit: Computer has gone to sleep.)
 566 2013-02-25 13:53:23 okok_ has joined
 567 2013-02-25 13:54:37 m0mchil has joined
 568 2013-02-25 13:54:37 m0mchil has quit (Changing host)
 569 2013-02-25 13:54:37 m0mchil has joined
 570 2013-02-25 13:56:03 okok has quit (Ping timeout: 244 seconds)
 571 2013-02-25 13:56:17 okok_ is now known as okok
 572 2013-02-25 14:10:47 agricocb has joined
 573 2013-02-25 14:10:57 nus-- is now known as nus
 574 2013-02-25 14:12:11 JDuke128 has joined
 575 2013-02-25 14:18:55 reizuki__ has quit (Quit: Konversation terminated!)
 576 2013-02-25 14:24:29 Cory has quit ()
 577 2013-02-25 14:25:07 idstam has joined
 578 2013-02-25 14:27:14 daybyter has quit (Quit: Konversation terminated!)
 579 2013-02-25 14:32:21 Hashdog has joined
 580 2013-02-25 14:34:47 idstam has quit ()
 581 2013-02-25 14:36:32 <jgarzik> <TD> It seems that there are some serious issues catching up with the chain that didn't used to exist. I was able to reproduce some problems where getheaders wasn't working well. I'll be working on an improvement to block chain handling this week so it's faster and uses less disk space.
 582 2013-02-25 14:36:41 <jgarzik> is TD talking about bitcoind or bitcoinj?
 583 2013-02-25 14:36:48 <jgarzik> https://bitcointalk.org/index.php?topic=145349.msg1558038#msg1558038
 584 2013-02-25 14:37:48 denisx has joined
 585 2013-02-25 14:39:16 drizztbsd has quit (Remote host closed the connection)
 586 2013-02-25 14:41:02 ralphtheninja has left ()
 587 2013-02-25 14:45:20 alexwaters has joined
 588 2013-02-25 14:51:57 jine has quit (Ping timeout: 245 seconds)
 589 2013-02-25 14:52:12 Zarutian has quit (Quit: Zarutian)
 590 2013-02-25 14:54:33 <gavinandresen> gmaxwell: sorry for being impolite, I'm just saying what I feel.  It seems like you and Peter are immediately against any idea that involves making MAX_BLOCKSIZE larger or eliminating it.
 591 2013-02-25 14:54:44 alexwaters2 has joined
 592 2013-02-25 14:55:00 alexwaters2 has left ()
 593 2013-02-25 14:55:55 <gavinandresen> gmaxwell: … and re: bias being an ad-hominen:  we are all biased.  I am biased.  I think it's silly to pretend otherwise.
 594 2013-02-25 14:56:09 <gavinandresen> (but maybe I've been reading the Overcoming Bias blog too long)
 595 2013-02-25 14:57:17 idstam has joined
 596 2013-02-25 14:57:44 alexwaters has quit (Ping timeout: 252 seconds)
 597 2013-02-25 14:58:45 jine has joined
 598 2013-02-25 14:58:47 <sipa> gavinandresen: afaik peter is not against increasing the blocksize, only against making miners decide the size
 599 2013-02-25 14:59:03 zooko has quit (Remote host closed the connection)
 600 2013-02-25 14:59:39 <sipa> (but that's perhaps my own bias)
 601 2013-02-25 15:00:08 <gavinandresen> sipa: great.  I've only been half-paying attention, is there a proposal that increases the blocksize that doesn't involve miners deciding?
 602 2013-02-25 15:02:11 <sipa> gavinandresen: in some forum thread i suggested increasing the limit once + allow it to grow at most slowly exponentially after that
 603 2013-02-25 15:02:54 <sipa> then again, i don't feel like joining all discussion heating now
 604 2013-02-25 15:03:06 <gavinandresen> sipa: yes, too much heat, not enough light.
 605 2013-02-25 15:03:10 <Luke-Jr> gavinandresen: jgarzik suggested having it increase as the subsidy drops, or something like that
 606 2013-02-25 15:03:42 <gavinandresen> it would be nice if we could start from development team consensus:  "Yes, the blocksize will, at some point, be increased."
 607 2013-02-25 15:03:53 <gavinandresen> Then we can argue over how/when....
 608 2013-02-25 15:04:45 <sipa> Luke-Jr: i actually like that, because of its simplicity- but perhaps it's too simple
 609 2013-02-25 15:04:47 bakh has joined
 610 2013-02-25 15:05:07 <Luke-Jr> sipa: well, no matter what scenario happens, miners can always cap it lower
 611 2013-02-25 15:05:28 <sipa> of course, and reducing the limit later is only a softfork
 612 2013-02-25 15:05:45 Guest32273 is now known as topace
 613 2013-02-25 15:05:50 topace has quit (Changing host)
 614 2013-02-25 15:05:50 topace has joined
 615 2013-02-25 15:06:09 <sipa> _however_ the argument (whether or not it is justified), is that an unlimited block size gives (the largest) miners an incentive for increasing the size
 616 2013-02-25 15:06:33 <sipa> which a soft fork will not help against, as that would require consensus from the same miners
 617 2013-02-25 15:06:38 <Luke-Jr> gavinandresen: FWIW, I think at some point it's reasonable to increase it, but I've pretty much ignored the discussion since it seems to me the time that needs to happen is a number of years off still
 618 2013-02-25 15:07:57 <gavinandresen> I'm posting in the Foundation thread about my conversation with my economist friend.  I think the first thing we need to figure out is the marginal cost of a transaction to a typical miner (I've wanted to know that for a while, anyway).
 619 2013-02-25 15:08:19 <gavinandresen> … because people keep on saying that "adding another transaction to a block is free", and that is simply not true.
 620 2013-02-25 15:08:25 okok_ has joined
 621 2013-02-25 15:08:28 <Luke-Jr> gavinandresen: that's a very complicated topic to study, IMO
 622 2013-02-25 15:08:50 <gavinandresen> mmm.  so is cryptography....
 623 2013-02-25 15:09:10 <Luke-Jr> it should probably include indirect costs/benefits like the affect on adoption and from there value of each Bitcoin mined
 624 2013-02-25 15:09:32 <Luke-Jr> effect*?
 625 2013-02-25 15:09:55 <Luke-Jr> (otherwise I'm certain the conclusion would be to mine empty blocks only!)
 626 2013-02-25 15:10:06 <gavinandresen> nah, I want a lower bound, I don't want to spend time thinking about indirect costs/benefits.
 627 2013-02-25 15:10:42 okok has quit (Ping timeout: 245 seconds)
 628 2013-02-25 15:10:50 okok_ is now known as okok
 629 2013-02-25 15:11:26 <gavinandresen> ???  if the benefit in fees is greater than the cost to the miner then OF COURSE they'll include a transaction...
 630 2013-02-25 15:11:33 MrMeowork has joined
 631 2013-02-25 15:11:34 <gavinandresen> I agree there's no incentive to include free transactions.
 632 2013-02-25 15:11:35 <Luke-Jr> gavinandresen: well, on that topic also.. Eligius is now submitblock-ing to 3 diversely spread 0.8 nodes; the other day, we observed that 1 of the remote nodes received the block over p2p before the RPC - yet at the same time my own wallet took 8 minutes(!) to get it
 633 2013-02-25 15:11:53 CodeShark has quit (Remote host closed the connection)
 634 2013-02-25 15:12:18 <Luke-Jr> gavinandresen: I somewhat doubt any of the current fees outweigh the practical costs, just from my observations
 635 2013-02-25 15:12:38 <gavinandresen> Has anybody figured out the average in-reality, cross-pool orphan rate?  And the theoretical minimum orphan rate?
 636 2013-02-25 15:13:15 <gavinandresen> if the fees outweigh the practical MARGINAL costs (I'm not talking about fixed hardware costs), then they're too low.
 637 2013-02-25 15:14:26 rdymac has quit (Quit: This computer has gone to sleep)
 638 2013-02-25 15:18:07 PhantomSpark has quit (3!~kvirc@pool-71-251-16-105.nycmny.fios.verizon.net|Ping timeout: 276 seconds)
 639 2013-02-25 15:18:20 denisx_ has joined
 640 2013-02-25 15:18:36 robocoin has joined
 641 2013-02-25 15:19:07 <gavinandresen> amended statement:  I agree there is no DIRECT incentive to include free transactions....
 642 2013-02-25 15:20:04 denisx has quit (Ping timeout: 276 seconds)
 643 2013-02-25 15:20:05 denisx_ is now known as denisx
 644 2013-02-25 15:23:24 jackyone has joined
 645 2013-02-25 15:24:38 <jackyone> hello all, i need source code for my web for read my wallet to see payments and send payments, where is source code?
 646 2013-02-25 15:25:50 <b4tt3r135> hm
 647 2013-02-25 15:26:44 rdymac has joined
 648 2013-02-25 15:29:42 jaakkos has quit (Ping timeout: 252 seconds)
 649 2013-02-25 15:30:43 ProfMac has quit (Quit: Page closed)
 650 2013-02-25 15:31:07 <helo> if the starting point is "not if, how", then hopefully it won't ever be true that success is proportional to decentralization
 651 2013-02-25 15:31:38 rdymac has quit (Ping timeout: 244 seconds)
 652 2013-02-25 15:32:32 jaakkos has joined
 653 2013-02-25 15:33:13 d4de_ has joined
 654 2013-02-25 15:34:28 d4de has quit (Ping timeout: 264 seconds)
 655 2013-02-25 15:34:40 Mrcheesenips has quit (Read error: Connection reset by peer)
 656 2013-02-25 15:35:06 Mrcheesenips has joined
 657 2013-02-25 15:35:06 Mrcheesenips has quit (Changing host)
 658 2013-02-25 15:35:06 Mrcheesenips has joined
 659 2013-02-25 15:39:16 rdymac has joined
 660 2013-02-25 15:39:18 <jackyone> hello all, i need source code for my web for read my wallet to see payments and send payments, where is source code?
 661 2013-02-25 15:39:30 <Happzz> google?
 662 2013-02-25 15:39:56 <SomeoneWeird> you can't just ask and expect it to be given to you
 663 2013-02-25 15:40:02 <jackyone> i would like 'a professional code', not Google....
 664 2013-02-25 15:40:06 alexwaters3 has joined
 665 2013-02-25 15:40:31 andytoshi has quit (Quit: Page closed)
 666 2013-02-25 15:41:49 <gavinandresen> jackyone: would you walk into a car repair shop and ask "hello all, I need new transmission for my car to drive, where is transmission?"
 667 2013-02-25 15:42:30 <gavinandresen> jackyone: if you yourself are a programmer, start here:  https://en.bitcoin.it/wiki/API_reference_(JSON-RPC)
 668 2013-02-25 15:42:47 <gavinandresen> jackyone: if you are not yourself a programmer, then you need to find and pay one….
 669 2013-02-25 15:43:05 <helo> or kidnap
 670 2013-02-25 15:43:15 <gavinandresen> no, don't kidnap.  That would be bad.
 671 2013-02-25 15:44:52 * SomeoneWeird glares at helo
 672 2013-02-25 15:44:56 <jackyone> thanks gavinandresen....
 673 2013-02-25 15:45:14 <helo> well it could conceivably get the job done, is all i'm saying :P
 674 2013-02-25 15:46:24 <alexwaters3> if I wanted to know what the top recent contributors to the satoshi client and bitcoin protocol were working on; what would be the most efficient way to do that?
 675 2013-02-25 15:46:34 vigilyn2 has quit (Read error: Connection reset by peer)
 676 2013-02-25 15:46:44 <alexwaters3> sorry my grammar is horrible
 677 2013-02-25 15:46:47 vigilyn2 has joined
 678 2013-02-25 15:47:31 <helo> look at their branches on github
 679 2013-02-25 15:48:06 b4tt3r135 has quit (Ping timeout: 255 seconds)
 680 2013-02-25 15:48:16 <QM> Regarding theoretical minimum orphan rate: can we assume miners know each others' hash rates and download capacity, and can connect to each other at will?
 681 2013-02-25 15:49:22 <gavinandresen> QM: I'd say no, but I'd be curious what other people say.
 682 2013-02-25 15:49:22 daybyter has joined
 683 2013-02-25 15:49:51 <gavinandresen> Example: I tried to get the big mining pools / solo miners to collaborate on a mailing list, so when there were exploits I could email them all at once easily.
 684 2013-02-25 15:50:09 <gavinandresen> … never happened, there seemed to be no desire from them to collaborate at all.
 685 2013-02-25 15:50:58 <gavinandresen> I think since mining is a zero-sum game they consider each other competitors.  So I'm always skeptical that a huge Evil Mining Consortium will spontaneously arise….  but I am probably biased.
 686 2013-02-25 15:51:03 <jouke> Talking about that, how to get on the list of people that get notified on security issues?
 687 2013-02-25 15:51:28 <gavinandresen> jouke: send me an email, convincing me that you can be trusted with early disclosure
 688 2013-02-25 15:51:40 <kinlo> gavinandresen: I don't think that's the case
 689 2013-02-25 15:52:15 <kinlo> gavinandresen: I have no problems with several mining pool operators actually, and we do have an irc channel, with several poolowners on the same channel
 690 2013-02-25 15:52:20 <gavinandresen> jouke: if you're not asking for early disclosure, then just watch the Important Announcement section of the bitcointalk forums or subscribe to my twitter feed or… umm....
 691 2013-02-25 15:52:33 <gavinandresen> jouke: if you're not asking for early disclosure, then just watch the Important Announcement section of the bitcointalk forums or subscribe to my twitter feed or… umm….
 692 2013-02-25 15:52:45 <gavinandresen> kinlo:  cool.  Y'all should create a mailing list so I can talk to you easier.
 693 2013-02-25 15:53:54 <jouke> I was asking about early disclosure.
 694 2013-02-25 15:54:23 <kinlo> gavinandresen: I don't mind the current setup tough, the disclosures being sent to us directly....  I get them, the other large pools have them...
 695 2013-02-25 15:54:36 <kinlo> and we communicate the way we want
 696 2013-02-25 15:55:13 <kinlo> also note that there do are pools that are impossible to communicate with, I've tried to contact some pools for my blockorigin project with no response at all
 697 2013-02-25 15:55:20 <gavinandresen> kinlo: mmm.  it's painful for me-- I have no idea if my list is up to date, and have little desire to keep track of who the big pool are to keep it up to date
 698 2013-02-25 15:56:21 <gavinandresen> so kinlo:  for the pool operators you know, do you know each others' hash rates and download capacity, and can connect to each other at will?
 699 2013-02-25 15:57:59 <kinlo> the hashrates are pretty straight forward, every pool publishes them, and my blockorigin project shows a top 20, but what do you mean by download capacity
 700 2013-02-25 15:59:19 <QM> kinlo: I meant bandwidth available for downloading blocks
 701 2013-02-25 16:00:05 <QM> So a miner can estimate time of arrival.
 702 2013-02-25 16:00:12 <kinlo> QM: bandwith for downloading blocks?  Pools are and should be uptodate with their blocks, so that's usually not an issue
 703 2013-02-25 16:00:48 freewil has quit (Ping timeout: 245 seconds)
 704 2013-02-25 16:00:52 <kinlo> being well connected to receive the blocks first is much more important then having a big pipe I think
 705 2013-02-25 16:01:15 <QM> kinlo: Say I find a block, then I want to know how long it'll take for you to receive it and start working on top of it.
 706 2013-02-25 16:01:32 <kinlo> QM: exactly, so you want to be well connected
 707 2013-02-25 16:01:55 freewil has joined
 708 2013-02-25 16:02:18 <QM> Yeah, but I want to quantify this to calculate a theoretical minimum orphan rate.
 709 2013-02-25 16:02:45 <kinlo> It's difficult for me, I've started to use multiple bitcoind's
 710 2013-02-25 16:05:28 jackyone has quit (Quit: Page closed)
 711 2013-02-25 16:08:26 <QM> kinlo: do the miners all tend to connect directly with eachother to reduce their orphan rates?
 712 2013-02-25 16:09:34 nouitfvf_ has joined
 713 2013-02-25 16:09:41 <kinlo> QM: not that I know of
 714 2013-02-25 16:11:03 Zarutian has joined
 715 2013-02-25 16:11:56 JDuke128 has quit (Quit: ["Textual IRC Client: www.textualapp.com"])
 716 2013-02-25 16:12:06 nouitfvf has quit (Ping timeout: 252 seconds)
 717 2013-02-25 16:12:14 <gavinandresen> Benefit of telling all your competitors how to directly contact you:  perhaps slightly lower orphan rates.  Potential cost:  they turn out to be evil and DDoS you.
 718 2013-02-25 16:12:41 <QM> Oh right...
 719 2013-02-25 16:12:42 <gavinandresen> ^contact^connect to
 720 2013-02-25 16:13:33 <kinlo> well, there are a few pools that actually make any money, most pools are hardly running break even so it's mostly people who are in the same shit :)
 721 2013-02-25 16:13:57 b4tt3r135 has joined
 722 2013-02-25 16:14:12 <kinlo> at least I have to spend more then I ever get in return
 723 2013-02-25 16:14:21 MrMeowork has quit (Quit: MrMeowork)
 724 2013-02-25 16:14:59 <gavinandresen> what are the incentives for running a pool?  Seems like a lot of work and abuse for not a lot of gain-- is it the "I'll make lots of money one of these days…." wishful thinking, or "well, I've already gone THIS far, might as well go a little farther...."
 725 2013-02-25 16:15:12 nouitfvf_ has quit (Quit: happy chatting)
 726 2013-02-25 16:15:52 <kinlo> gavinandresen: call it a hobby :)
 727 2013-02-25 16:16:08 <gavinandresen> cool
 728 2013-02-25 16:16:08 <Graet> some of us actually think bitcoin is  a good thing and want to make it available to everyone, not just high hashrate solo miners :)
 729 2013-02-25 16:16:19 <kinlo> indeed
 730 2013-02-25 16:16:27 <Graet> so pools are logical way to "co-operatively" mine
 731 2013-02-25 16:16:41 <Graet> obviously the market is changing
 732 2013-02-25 16:16:48 <kinlo> so while on paper Graet should be mortal enemies as we are competition, we are not :)
 733 2013-02-25 16:16:48 <Graet> less cpu/gpu/hobby miners
 734 2013-02-25 16:16:52 <gavinandresen> good for you-- I admire your persistence in the face of… well all the crap that seems to get thrown at pool operators.
 735 2013-02-25 16:17:10 <kinlo> gavinandresen: you have no idea :)
 736 2013-02-25 16:17:13 <Graet> ahh well, in my case i'm off work, so its better than playing games all day
 737 2013-02-25 16:17:37 <Graet> plus i have lernt a loit , and met some awesome people, apart from mined and enjoyed bitcoins:)
 738 2013-02-25 16:17:40 <kinlo> besides, it is actually a good test to see if you can run something this big successfully
 739 2013-02-25 16:17:52 <kinlo> it is not easy to run a pool, do not underestimate it
 740 2013-02-25 16:18:24 <gavinandresen> I know it's not easy to run a pool, which is why I'm surprised somebody is willing to do it as a hobby
 741 2013-02-25 16:18:52 <gavinandresen> I suppose there are pleny of hobbies that aren't easy, though.
 742 2013-02-25 16:19:15 <gavinandresen> … just golf is a different kind of "not easy"
 743 2013-02-25 16:19:29 <kinlo> :)
 744 2013-02-25 16:19:53 <Graet> yeah, and i suppose, my goal for running the  pool was to offer low cost mining to miners and cover costs, my interest in making btc are outside the pool :) other ventures
 745 2013-02-25 16:20:14 <Graet> but i have got some good contacts from running the pool :)
 746 2013-02-25 16:20:18 <kinlo> I must say that running blockorigin is a lot more rewarding, people apreciate it and I don't get much thrown at me for doing it
 747 2013-02-25 16:20:25 WolfAlex has quit (Read error: Connection reset by peer)
 748 2013-02-25 16:20:41 WolfAlex has joined
 749 2013-02-25 16:20:49 <Graet> it is a good resource kinlo , i'm keenly watching atm :P
 750 2013-02-25 16:22:02 <denisx> ui, my pool is in the top20 ;)
 751 2013-02-25 16:23:10 rdymac has quit (Ping timeout: 264 seconds)
 752 2013-02-25 16:23:21 <Graet> nice denisx :) mine is about to overtake deepbit lol
 753 2013-02-25 16:23:39 rdymac has joined
 754 2013-02-25 16:23:48 <denisx> Graet: time it is! ;)
 755 2013-02-25 16:23:53 <Graet> :D
 756 2013-02-25 16:25:26 da2ce7_d has joined
 757 2013-02-25 16:25:43 <denisx> the top5 pools still generating version blocks?
 758 2013-02-25 16:25:44 rdymac has quit (Client Quit)
 759 2013-02-25 16:25:45 <denisx> lame
 760 2013-02-25 16:27:09 da2ce7 has quit (Ping timeout: 256 seconds)
 761 2013-02-25 16:27:40 <Graet> gavinandresen, thanks for asking though, not many people do, assuming is easy :D
 762 2013-02-25 16:28:06 rdymac has joined
 763 2013-02-25 16:29:03 bakh has quit (Ping timeout: 252 seconds)
 764 2013-02-25 16:31:29 <denisx> version 1 blocks I meant
 765 2013-02-25 16:31:47 b4tt3r135 has left ()
 766 2013-02-25 16:35:11 dhill has joined
 767 2013-02-25 16:35:37 <dhill> so how does bitcoind get a list of initial peers to connect to if I don't use irc?
 768 2013-02-25 16:35:55 <sipa> dns seeds
 769 2013-02-25 16:36:05 <dhill> ahh
 770 2013-02-25 16:36:51 <dhill> oh, just needed to scroll down more on the wiki
 771 2013-02-25 16:38:00 meLon has joined
 772 2013-02-25 16:39:19 <kinlo> gavinandresen: I've been discussing with Graet, we'll work on a mailinglist
 773 2013-02-25 16:39:26 <QM> gavinandresen: Is it too simplistic a model to just assume miners connect to peers at random?  Would this "theoretical mimum orphan rate" even be useful (especially with large blocks where connectivity really matters)?
 774 2013-02-25 16:40:32 <QM> peers = not necessarily other miners
 775 2013-02-25 16:41:00 <gavinandresen> QM: the argument is that a 51%, well-connected cabal of miners might conspire to try to produce huge blocks so the other 49% suffer from higher orphan rates.
 776 2013-02-25 16:41:18 <gavinandresen> … driving them out, and getting a monopoly on mining.  HAHA!
 777 2013-02-25 16:41:55 <gavinandresen> … then half of them turn on the other half, and drive THEM out….
 778 2013-02-25 16:42:01 <gavinandresen> … and eventually we get one miner.
 779 2013-02-25 16:42:05 <QM> and ruining the value of their future rewards by ruining confidence?
 780 2013-02-25 16:42:14 <gavinandresen> yes.  I guess.
 781 2013-02-25 16:42:24 monkeynipples has joined
 782 2013-02-25 16:42:43 <QM> taking game theory a bit too seriously here I think...
 783 2013-02-25 16:43:16 <QM> or simplifying the game too much, at least
 784 2013-02-25 16:43:20 frosks has quit (Remote host closed the connection)
 785 2013-02-25 16:43:23 <gavinandresen> I tend to agree, but I'd like to think about worst-possible-case scenarios
 786 2013-02-25 16:43:43 axhlf has joined
 787 2013-02-25 16:43:46 <gavinandresen> or maybe: worst-likely-case scenarios
 788 2013-02-25 16:44:34 <gavinandresen> I'm not going to worry about never-going-to-happen things like "what if I get REALLY unlucky and my 160-bit address collides with somebody else's..."
 789 2013-02-25 16:45:59 <jgarzik> gavinandresen, Luke-Jr, gmaxwell: thoughts at the moment:  (a) I once posted a patch to change max block size, so I thought about this long before forum trolls ever woke up, (b) I have since backed down from that radical position, (c) it seems likely that max block size will change sometime in bitcoin's future, (d) block size is VERY MUCH like bitcoin's 21M limit, so a lot of care must be taken when changing MAX_BLOCK_SIZE logic
 790 2013-02-25 16:46:00 <jgarzik> .  Block size is an economically limited resource whose production is tightly defined and controlled by algorithm, with an intentionally steady production rate. (e) I lean against solutions that are feedback-based (average of last 1000 block sizes, etc.), as they can be gamed too easily
 791 2013-02-25 16:46:50 <jgarzik> (f) implementation will likely be:  if (now > future date) { do it }
 792 2013-02-25 16:47:23 <jgarzik> (g) my default rhetorical position will be to push back against changes, for now, since there is no demonstrated need for hard fork.
 793 2013-02-25 16:47:58 <jgarzik> (h) when I ran asic + p2pool, I set max block size at 900k and free tx at 300k
 794 2013-02-25 16:48:06 <jgarzik> that reflects my personal preferences
 795 2013-02-25 16:48:31 <jgarzik> I filtered out 1e8 outputs
 796 2013-02-25 16:48:36 <jgarzik> from mempool
 797 2013-02-25 16:48:37 dhill has quit (Quit: leaving)
 798 2013-02-25 16:48:50 <gavinandresen> how big were your blocks?
 799 2013-02-25 16:49:01 <jgarzik> pretty big
 800 2013-02-25 16:49:10 <jouke> Can someone explain to me what is wrong to have nodes delay large blocks relay? I have the understanding that that is the solution to the blocksize problem.
 801 2013-02-25 16:49:40 <jgarzik> gavinandresen: 413k here: http://blockexplorer.com/block/00000000000001528a3fa72b86032459e1fb6ab38720e19a26e3a1f4a64e461a
 802 2013-02-25 16:49:42 <gavinandresen> pretty big… like 900k big?  or pretty big like 500k big ?  Were they full?
 803 2013-02-25 16:49:51 <jgarzik> gavinandresen: nowhere near full
 804 2013-02-25 16:49:56 <gavinandresen> ok, cool.
 805 2013-02-25 16:50:00 <jgarzik> gavinandresen: but I made sure that it was easy to go up to 900k
 806 2013-02-25 16:50:08 <jgarzik> as a test
 807 2013-02-25 16:51:09 <jgarzik> oh and
 808 2013-02-25 16:51:12 <gavinandresen> I agree with no change to the limit, it isn't needed right now.  But I do think we need to think about what data we need to gather.  And we definitely need to make clients smarter about fees before we hit the limit.
 809 2013-02-25 16:51:20 <jgarzik> very importantly,
 810 2013-02-25 16:51:59 freakazoid has joined
 811 2013-02-25 16:52:24 <jgarzik> (i) it is a mistake to increase block size simply because people are too lazy to implement layers on top of bitcoin.  Bitcoin will forever be a zen BALANCE of applications and layers that sit on top of the blockchain, and those that directly use the blockchain itself as their comm/functional layer (c.f. SatoshiDICE)
 812 2013-02-25 16:52:38 <jgarzik> so I generally agree with a lot of gmaxwell's points
 813 2013-02-25 16:53:06 <jgarzik> and it is important therefore to _not_ increase blocksize, simply because that makes some apps easier because they free-ride on the blockchain itself
 814 2013-02-25 16:53:31 <jgarzik> maybe one future state of The Mainnet Blockchain is simply a high security merged mining root for several important chains
 815 2013-02-25 16:53:45 <sipa> right, temporarily keeping the block size low will encourage development of solutions which are more blockchain-efficient
 816 2013-02-25 16:53:46 <gavinandresen> sure… I'm working hard on the payment protocol right now partly because it will give SatoshiDice a better way of doing what they want to do.
 817 2013-02-25 16:53:55 <jgarzik> therefore
 818 2013-02-25 16:54:12 <jgarzik> my GUESS (may be wrong) for a solution was:  MAX_BLOCK_SIZE *= 2, every 1 years
 819 2013-02-25 16:54:16 <jgarzik> or something like that
 820 2013-02-25 16:54:23 <sipa> when that has happened, and there is enough economic groth to warrant a block size agree, go for it, but make it not fully controllable by miners
 821 2013-02-25 16:54:28 <jgarzik> thus, not feedback based
 822 2013-02-25 16:54:40 <jgarzik> and not controlled by miners at all
 823 2013-02-25 16:54:42 <gavinandresen> jgarzik: shhh, you just said your rhetorical position was no changes
 824 2013-02-25 16:54:42 <jgarzik> 100% algorithm
 825 2013-02-25 16:55:27 <sipa> how long is the minimal timeframe you consider for a hard fork?
 826 2013-02-25 16:55:30 <jgarzik> gavinandresen: current position == ACK or NAK in pull requests.  I would NAK a block size change pull req today :)
 827 2013-02-25 16:56:08 <jgarzik> (revised) my GUESS (may be wrong) for a solution was:  if (todays_date > SOME_FUTURE_DATE) { MAX_BLOCK_SIZE *= 2, every 1 years }
 828 2013-02-25 16:56:36 <sipa> jgarzik: wow, *2 every year? that's ridiculously fast
 829 2013-02-25 16:56:57 <denisx> faster than moore law
 830 2013-02-25 16:57:03 <sipa> yes, that
 831 2013-02-25 16:57:11 <jgarzik> sipa: <shrug> it was just an example metric.  pick a better one.
 832 2013-02-25 16:57:19 <jgarzik> sipa: the point is, not feedback based nor miner controlled
 833 2013-02-25 16:57:20 <gavinandresen> I'd do min(bandwidth,cpu,disk) growth per year maybe.
 834 2013-02-25 16:57:34 <sipa> jgarzik: right, sure, agree about that
 835 2013-02-25 16:57:41 <jgarzik> and 100% internally algorithmic, like the subsidy
 836 2013-02-25 16:57:48 <jgarzik> depends on no external factors
 837 2013-02-25 16:57:54 <sipa> jgarzik: i don't mind miner feedback, by the way, as long as it's not unlimited apart from that
 838 2013-02-25 16:58:08 * jgarzik is skeptical but willing to be convinced :)
 839 2013-02-25 16:58:27 <jgarzik> IMNSHO block size is a core economic resource, like the number of bitcoins itself
 840 2013-02-25 16:58:45 <jgarzik> not merely the number of transactions we can support... it influences fees and many other factors
 841 2013-02-25 16:58:58 * jgarzik emphasizes the importance
 842 2013-02-25 16:59:16 <sipa> jgarzik: well, there are several aspects to consider
 843 2013-02-25 16:59:31 freakazoid has quit (Ping timeout: 276 seconds)
 844 2013-02-25 17:01:00 <QM> How about with miner feedback, but with the max growth they can achieve limiting out to jgarzik's formula (or whatever's better).  That way it doesn't have to rise unnecessarily.
 845 2013-02-25 17:01:04 <sipa> jgarzik: anyway, i haven't made up my mind about miner feedback or not - it's just something i consider because many seem to favor it
 846 2013-02-25 17:01:30 <jgarzik> default: against :)
 847 2013-02-25 17:01:34 <sipa> i like the simplicity of a deterministic rule, but that almost inevitably means the rule will have to be adjusted again later
 848 2013-02-25 17:01:51 * jgarzik is an engineer, and always qualifies his answers :)
 849 2013-02-25 17:02:09 <QM> :)
 850 2013-02-25 17:02:52 <jgarzik> sipa: simple rule + let the free market figure out the rest (which might include a more complex rule years later, when the picture is more clear)
 851 2013-02-25 17:02:57 <gavinandresen> i like to eliminate the as-simple-as-possible answer first before considering something more complex.
 852 2013-02-25 17:03:16 rdymac has quit (Quit: This computer has gone to sleep)
 853 2013-02-25 17:03:32 <jgarzik> IOW, iff we need a rule, simple and not feedback based.  by the time a rule change is needed, the picture will be clear about how to next change the rule.
 854 2013-02-25 17:03:45 <jgarzik> and we don't need a change now.
 855 2013-02-25 17:03:46 <jgarzik> :)
 856 2013-02-25 17:07:57 <jgarzik> That was more than I intended to type, about block size.  It seems more like The Question Of The Moment on the web, than a real engineering need.
 857 2013-02-25 17:08:11 <jgarzik> just The Thing people are talking about right now
 858 2013-02-25 17:08:29 <gmaxwell> _everywhere_ ugh. Make it go away.
 859 2013-02-25 17:08:46 <gmaxwell> just reading all of it is a full time job right now.
 860 2013-02-25 17:10:29 <jgarzik> making block size unlimited is a hugely bad idea
 861 2013-02-25 17:12:32 freakazoid has joined
 862 2013-02-25 17:12:53 * gmaxwell reads scrollback
 863 2013-02-25 17:13:20 <helo> is happy with the way this discussion is going
 864 2013-02-25 17:15:16 <gmaxwell> jgarzik: yes, I agree with you completely. On (c)  perhaps multiple times. Computers get faster, transaction loads change. The only available argument that involves under-no-condition-should-it-ever-change is one that— thank god— no one is making.  I've been holding back a lot in these discussions because I am seriously afraid of inspiring people to make the bitcoin-is-a-suicide-pact argument.
 865 2013-02-25 17:16:15 freakazoid_ has joined
 866 2013-02-25 17:17:08 toffoo has joined
 867 2013-02-25 17:17:43 zooko has joined
 868 2013-02-25 17:18:30 freakazoid has quit (Quit: Leaving)
 869 2013-02-25 17:18:49 freakazoid_ is now known as freakazoid
 870 2013-02-25 17:20:48 bitnumus has left ("Leaving")
 871 2013-02-25 17:21:56 bitnumus has joined
 872 2013-02-25 17:21:58 <bitnumus> bitnumus> Tiny bug on 0.8.0, when launching the message 'Verifying database block integrity' is cut off in the splashscreen, i am only guessing thats what its saying :)
 873 2013-02-25 17:23:12 <helo> the block size debate will rage on until people become aware of dev consensus
 874 2013-02-25 17:31:01 OneFixt has quit (Read error: Connection reset by peer)
 875 2013-02-25 17:31:16 OneFixt has joined
 876 2013-02-25 17:31:49 <gavinandresen> gmaxwell: have you calculated the marginal cost of a transaction?
 877 2013-02-25 17:32:04 <gavinandresen> gmaxwell: … or, better, marginal cost to a miner of making their block 1K bigger ?
 878 2013-02-25 17:32:53 superman2016 has quit (Ping timeout: 245 seconds)
 879 2013-02-25 17:35:48 makomk has quit (Ping timeout: 245 seconds)
 880 2013-02-25 17:36:45 makomk has joined
 881 2013-02-25 17:36:48 rdymac has joined
 882 2013-02-25 17:39:58 m00p has quit (Ping timeout: 252 seconds)
 883 2013-02-25 17:41:07 <jgarzik> the initial block reward will cloud true transaction costs for months, perhaps years to come
 884 2013-02-25 17:45:22 denisx has quit (Quit: denisx)
 885 2013-02-25 17:50:57 alexwaters1 has quit (Ping timeout: 260 seconds)
 886 2013-02-25 17:51:18 PiZZaMaN2K has joined
 887 2013-02-25 17:56:31 m00p has joined
 888 2013-02-25 17:58:38 brwyatt is now known as Away!~brwyatt@brwyatt.net|brwyatt
 889 2013-02-25 18:00:06 Hashdog has quit (Remote host closed the connection)
 890 2013-02-25 18:00:34 Hashdog has joined
 891 2013-02-25 18:00:58 Hashdog has quit (Read error: Connection reset by peer)
 892 2013-02-25 18:01:06 Hashdog has joined
 893 2013-02-25 18:01:25 m00p has quit (Ping timeout: 252 seconds)
 894 2013-02-25 18:03:14 FredEE has joined
 895 2013-02-25 18:04:13 freakazoid has quit (Remote host closed the connection)
 896 2013-02-25 18:08:46 freakazoid has joined
 897 2013-02-25 18:10:13 <gmaxwell> gavinandresen: sure.
 898 2013-02-25 18:10:14 <gmaxwell> Yes, adding a byte to a block costs roughly (236.51ms/250341txn/1000.)*35watt/(60*60second)/1000*$0.30($/kwh) = 2.7555e-12 in CPU
 899 2013-02-25 18:10:17 <gmaxwell> using conservative numbers on my laptop. Plus (1*(1/10)*8) = 0.8bps (to transmit the 1 byte in <= 10 seconds to 8 peers), or
 900 2013-02-25 18:10:21 <gmaxwell> $0.0000016000 at $2/mbit/sec/month assuming you pay full usage even when you don't solve a block (well, you'll pay it for recieving
 901 2013-02-25 18:10:24 <gmaxwell> other people's blocks— your inclusion isn't what costs you, its what other people include that costs you)
 902 2013-02-25 18:10:34 Belkaar has quit (Ping timeout: 264 seconds)
 903 2013-02-25 18:10:55 <gmaxwell> I'm assuming just 10 second times then because that makes orphaning very small, actually figuring out the cost requiries knowing what is at stake.
 904 2013-02-25 18:11:26 Belkaar has joined
 905 2013-02-25 18:12:02 PhantomSpark has joined
 906 2013-02-25 18:12:38 <gavinandresen> gmaxwell: I'm interested in the extra-chance-my-block-will-be-orphaned cost
 907 2013-02-25 18:12:43 <gmaxwell> In any case— I'm not sure that this actually matters. The point that I've been making is that if the marginal cost of processing a transaction is what determines the required fee then bitcoin is insecure— because an attacker who only processes a few transactions or validates nothing can just replace a chain built by miners competing for marginal costs.
 908 2013-02-25 18:13:11 <gavinandresen> … and mike and I keep saying that security may be paid for indirectly…   but lets put that aside.
 909 2013-02-25 18:13:32 <gmaxwell> You can keep saying that, but if thats all you say no one will hear you. Indirectly how?
 910 2013-02-25 18:13:34 <gavinandresen> I'm actually interested in the orphaning cost aside from the blocksize debate
 911 2013-02-25 18:14:00 <gmaxwell> The simplistic orphaning cost is easy but maybe not useful.
 912 2013-02-25 18:14:01 <gavinandresen> indirectly by big merchants running or subsidizing big mining farms
 913 2013-02-25 18:14:33 <gavinandresen> not useful for what?  I'd like to tell miners NOW:  if you make 900K blocks, this is what it will cost you versus making 200K blocks.
 914 2013-02-25 18:14:47 m00p has joined
 915 2013-02-25 18:14:51 <gavinandresen> (cost you in terms of theoretically increased orphan rates)
 916 2013-02-25 18:15:06 <jgarzik> I think each miner will figure that out on their own
 917 2013-02-25 18:15:12 <gavinandresen> I'll do the calculations myself, I'm just trying to be lazy if somebody else has already done the work.
 918 2013-02-25 18:15:52 <gmaxwell> It's just "how likely is it that another block is found before mine reaches 50% of the hashpower",  CDF of the exponential distribution tells you how likely it is for a block to be found in the next X seconds = 1-e^(-(1/600)*X)
 919 2013-02-25 18:17:17 <gmaxwell> So you then just need a model for how bytes turn into X. The relaying is exponential, so you can likely approximate that to "how long to send and validate one copy" times some small number (like 3).
 920 2013-02-25 18:17:39 <gavinandresen> right.  which is why I was asking about theoretical and actual orphan rates.  I suppose for theoretical X is one-quarter the earth's circumference times speed-of-light-in-fiber, assuming hashing power is reasonably geographically distributed
 921 2013-02-25 18:17:57 <gmaxwell> Validation time on my laptop on current blocks is 9.4475e-04 ms per byte
 922 2013-02-25 18:18:02 <gavinandresen> … although we shouldn't assume infinite bandwidth
 923 2013-02-25 18:18:53 <gavinandresen> … and we SHOULD probably assume pre-validation has been done for the vast majority of transactions in blocks.  Although that's another thing I'd like to know if I'm a miner:  how costly is it to put never-seen-on-the-network transactions into my blocks?
 924 2013-02-25 18:19:55 <gmaxwell> gavinandresen: the only increase is in validation time right now— and the figure I gave of 9.4475e-04 ms/byte is for never seen. Though this is with 0.8, it used to be slowerl
 925 2013-02-25 18:20:37 <gmaxwell> with 0.8 it's pretty likely that the coins database is in ram on any node which hasn't been under memory pressure: it's under 200mb.
 926 2013-02-25 18:22:11 <gmaxwell> gavinandresen: I would note that we don't have market evidence that miners care: they could greatly reduce these issues by directly peering (but they won't because they believe they'll dos attack each other or don't care) or by having neutral meeting points (solves the DOS problem) which they don't bother with.
 927 2013-02-25 18:22:28 <gavinandresen> gmaxwell: 9.4e-04 ms/byte:  so for a 250byte transaction you're validating in 0.2 ms  ?  That's a lot faster than when we last estimated… (6 ms per transaction)
 928 2013-02-25 18:24:45 <gmaxwell> block 223085 took my laptop after a cold start of bitcoin 236.72ms to validate, it was 250341 bytes. I could get figures from more blocks if you like.
 929 2013-02-25 18:25:33 <sipa> 0.2ms per txin is reasonable on decent hardware with 4 cores
 930 2013-02-25 18:25:46 <sipa> per tx... sounds low
 931 2013-02-25 18:26:59 <gmaxwell> (I only grabbed that for my power estimates before, which I was then going to multiply by the crazy 35 watt TDP number, so I didn't really care if it was _that_ accurate).
 932 2013-02-25 18:27:47 * gmaxwell benchmarks
 933 2013-02-25 18:27:49 <sipa> my numbers are for IBD only by the way
 934 2013-02-25 18:28:05 <sipa> cache usage is a lot worse afterwards (i plan to work on that)
 935 2013-02-25 18:28:20 PhantomSpark has joined
 936 2013-02-25 18:28:30 PhantomSpark has quit (3!~kvirc@pool-71-251-16-105.nycmny.fios.verizon.net|Read error: Connection reset by peer)
 937 2013-02-25 18:31:47 <helo> is the long term impact of block size on the number of full nodes relevant?
 938 2013-02-25 18:32:36 ThomasV has quit (Quit: Leaving)
 939 2013-02-25 18:34:10 <sipa> very relevant, i think :)
 940 2013-02-25 18:35:08 <helo> so messy
 941 2013-02-25 18:35:30 <helo> how would we even know how many full nodes there should be?
 942 2013-02-25 18:37:47 <gmaxwell> helo: I don't think thats a relevant question in any case, because you can't take N nodes and get some parameter out of it in any obvious way.
 943 2013-02-25 18:39:35 <gmaxwell> I think the easier question is what kind of convenience / security tradeoff each user should have to make. People talk about having freedom, but freedom always has a price. Being free isn't— in practice— isn't about an absolute "freedom" it's about having freedom be low cost.  Likewise, bitcoin should be trustless. But absolutely trustless is abstract and useless— instead? what is the cost?
 944 2013-02-25 18:41:05 rdymac has quit (Quit: This computer has gone to sleep)
 945 2013-02-25 18:41:17 theymos has joined
 946 2013-02-25 18:50:51 <helo> either the user pays for zero trust with (likely prohibitively high) transaction fees, or with (likely prohibitively high) hardware demands.
 947 2013-02-25 18:51:13 Cylta has joined
 948 2013-02-25 18:51:17 <Cylta> what is the gpg keys folder on linux?
 949 2013-02-25 18:51:27 <Cylta> something like ~/.gpg but not it.
 950 2013-02-25 18:51:35 <helo> ~/.gnupg
 951 2013-02-25 18:52:19 <helo> if hardware demands are too high, they won't be able to use zero trust at all.
 952 2013-02-25 18:52:47 <gmaxwell> Yea, you also have to think about cost exteranlization and freeloading.
 953 2013-02-25 18:53:08 <theymos> Regarding the max block size, what do you guys think about setting some regular schedule for changing the max block size to something reasonable and making other hardfork changes? Do it every 5 years on a particular day, for example. Then the max block size will be set properly, it will be done in a decentralized way (since people will have to voluntarily choose their rules), and there won't be so much argument each time since the change is expe
 954 2013-02-25 18:53:08 <theymos> cted.
 955 2013-02-25 18:53:16 <Cylta> helo: that's it! thank you
 956 2013-02-25 18:53:43 bitafterbit has joined
 957 2013-02-25 18:53:53 <gmaxwell> Zero trust is mostly useful because of the kind of security it gives, but if security is the objective.. you can be totally secure, but do no validation— so long as everyone else does... and then you have fun result when you iterate that rational thinking.
 958 2013-02-25 18:54:02 <gmaxwell> theymos: see jeff's comments in the log from a few hours ago.
 959 2013-02-25 18:54:41 <gavinandresen> gmaxwell: was it you who suggested quite a while ago that the client have a built-in expiration date?
 960 2013-02-25 18:55:21 Hasimir has quit (Ping timeout: 260 seconds)
 961 2013-02-25 18:55:22 <gmaxwell> Maybe? I might have.  Actually, I'm pretty sure I suggested it for RCs and development builds.
 962 2013-02-25 18:55:43 <gmaxwell> (Since thats something prior-employer did for special builds and I thought it was interesting)
 963 2013-02-25 18:56:13 <gmaxwell> On production versions, I'm less sure— but we already have a mechenism like that: the client whines when it doesn't understand the block version.
 964 2013-02-25 18:56:17 <gavinandresen> I like "this software expires in 5 years" better than "We Will Have A Master Hard Fork Date Every N Years"
 965 2013-02-25 18:56:44 Cylta has left ()
 966 2013-02-25 18:56:59 <gavinandresen> and yes, not so necessary now that it whines if the chain block.versions change
 967 2013-02-25 18:57:18 bitafterbit has quit (Remote host closed the connection)
 968 2013-02-25 18:58:38 bitafterbit has joined
 969 2013-02-25 18:58:48 <gavinandresen> does anybody have the URL to the "how people screw up ssl certificate validation" paper that came out a few months ago?
 970 2013-02-25 18:59:17 <gavinandresen> I'm trying to figure out if I missed any ways writing unit tests :  https://github.com/gavinandresen/bitcoin-git/blob/paymentrequest/src/qt/test/paymentservertests.cpp
 971 2013-02-25 18:59:30 <gmaxwell> https://crypto.stanford.edu/~dabo/pubs/abstracts/ssl-client-bugs.html
 972 2013-02-25 18:59:38 <gavinandresen> thanks
 973 2013-02-25 18:59:57 <gmaxwell> keep in mind that openssl just doesn't implement hostname validation, so you're on your own.
 974 2013-02-25 19:00:08 <jaakkos> does anyone have a fix for the LLVM problems with 'oclvanitygen' + AMD Catalyst?
 975 2013-02-25 19:00:34 <gavinandresen> gmaxwell: happily payment requests will let the user decide whether they want to accept a *.foo.com certificate
 976 2013-02-25 19:01:50 <gmaxwell> gavinandresen: ah, so there is no closed loop inspection of the apparent url.  A little ugly, but perhaps unsolvable. The issue is I MITM FOO.com and then server you a FO0.com cert or the like.
 977 2013-02-25 19:02:22 PhantomSpark has quit (Ping timeout: 276 seconds)
 978 2013-02-25 19:02:47 Hasimir has joined
 979 2013-02-25 19:02:56 <gavinandresen> gmaxwell: suggestions for how to solve that welcome, but aside from telling certificate authorites not to sign lookalike domains I don't see a solution
 980 2013-02-25 19:03:01 Hasimir is now known as Guest5817
 981 2013-02-25 19:04:15 coolsa_ has quit (Ping timeout: 252 seconds)
 982 2013-02-25 19:04:16 <gmaxwell> I don't know what information we get from the browser on those links— can we discover the URL of the page that it was on? If so we could refuse a browser sourced payment request where the domains don't agree.
 983 2013-02-25 19:04:41 <gavinandresen> gmaxwell: nope, we don't get any information about where the request came from.  Just the request itself.
 984 2013-02-25 19:05:21 <gavinandresen> gmaxwell: might not be a browser, anyway-- might have come in via email attachment.
 985 2013-02-25 19:05:36 <gmaxwell> Indeed, in email you're hosed.
 986 2013-02-25 19:06:06 <gavinandresen> in any case, the same mechanism is used if you double-click on a .bitcoinpaymentrequest file, click on a link, or open an attachement.
 987 2013-02-25 19:06:22 <gmaxwell> but a situation where there is already a clear domain 'context' is not the same as one where there isn't. But it sounds like we can't distinguish them, alas.
 988 2013-02-25 19:07:10 <gavinandresen> nope. Not sure how that helps, anyway:  user could be on the fake FO0 domain to start
 989 2013-02-25 19:07:26 dust-otc has quit (Remote host closed the connection)
 990 2013-02-25 19:08:16 <gmaxwell> They could be, of course. But they wouldn't have found that URL from e.g. inside mtgox.
 991 2013-02-25 19:09:10 toffoo has quit ()
 992 2013-02-25 19:10:38 nus has quit (Read error: Connection reset by peer)
 993 2013-02-25 19:11:09 nus has joined
 994 2013-02-25 19:11:17 meLon has left (""PART BARK BARK BARK"")
 995 2013-02-25 19:15:02 rdymac has joined
 996 2013-02-25 19:16:36 theymos has quit (Remote host closed the connection)
 997 2013-02-25 19:18:58 daybyter has quit (Read error: Connection reset by peer)
 998 2013-02-25 19:21:15 zooko has quit (Ping timeout: 246 seconds)
 999 2013-02-25 19:28:00 coolsa has joined
1000 2013-02-25 19:34:25 ThomasV has joined
1001 2013-02-25 19:47:58 zooko has joined
1002 2013-02-25 19:52:41 none has joined
1003 2013-02-25 19:53:01 none is now known as BTCTrader2
1004 2013-02-25 19:53:18 BTCTrader2 has quit (Client Quit)
1005 2013-02-25 19:53:26 BTCTrader2 has joined
1006 2013-02-25 19:54:10 [\\\] has quit (Read error: error:1408F10B:SSL routines:SSL3_GET_RECORD:wrong version number)
1007 2013-02-25 19:54:15 Nachtwind has joined
1008 2013-02-25 19:54:33 <gmaxwell> jgarzik: I suspect the people who are intentionally trolling for god-knows-what indecipherable purpose have picked up the block size thing as something that is fun to play with. ... in any case, after you mentioned it here they found your old block size thread and are now having fun with it.
1009 2013-02-25 19:55:36 <Nachtwind> hi.. i have a problem with my bitcoind.. all of a sudden it only responds to rpc calls with request id '' instead of the number wanted.. did something change to bitcoind latey in the rpc code or is 0.8 in any other way known to have problems with thatT?
1010 2013-02-25 19:55:55 [\\\] has joined
1011 2013-02-25 19:56:31 rdymac has quit (Remote host closed the connection)
1012 2013-02-25 19:57:10 rdymac has joined
1013 2013-02-25 19:57:23 discrete has quit ()
1014 2013-02-25 19:58:40 <gmaxwell> since when do RPCs return a request id?
1015 2013-02-25 19:59:23 vigilyn2 is now known as vigilyn
1016 2013-02-25 19:59:26 <gavinandresen> responses should return the same request id in the JSON-RPC call that asked for them
1017 2013-02-25 19:59:31 vigilyn has quit (Changing host)
1018 2013-02-25 19:59:31 vigilyn has joined
1019 2013-02-25 20:03:03 <gavinandresen> Nachtwind: working for me:  http://pastebin.com/uSUBQ333
1020 2013-02-25 20:05:52 <Nachtwind> sorry meant json-ropc
1021 2013-02-25 20:06:12 PiZZaMaN2K is now known as PiZZaMaN2K|away
1022 2013-02-25 20:06:19 <Nachtwind> gavin: i will try this
1023 2013-02-25 20:07:42 <Nachtwind> that way it works..
1024 2013-02-25 20:07:45 <Nachtwind> hmm
1025 2013-02-25 20:07:54 <Nachtwind> interestering
1026 2013-02-25 20:08:06 <Nachtwind> didnt change a thing on the php side of the project
1027 2013-02-25 20:08:09 rbecker is now known as RBecker
1028 2013-02-25 20:08:26 TD has joined
1029 2013-02-25 20:11:16 Belkaar has quit (Ping timeout: 276 seconds)
1030 2013-02-25 20:12:30 Belkaar has joined
1031 2013-02-25 20:19:21 rdymac has quit (Ping timeout: 246 seconds)
1032 2013-02-25 20:22:35 makomk has quit (Ping timeout: 264 seconds)
1033 2013-02-25 20:23:09 makomk has joined
1034 2013-02-25 20:23:46 BurtyBB has joined
1035 2013-02-25 20:24:10 rdymac has joined
1036 2013-02-25 20:24:50 <QM> Is there a writeup explaining how PoM works somewhere?
1037 2013-02-25 20:26:48 <QM> (proof-of-memory)
1038 2013-02-25 20:27:26 datagutt has quit (Quit: kthxbai)
1039 2013-02-25 20:28:10 BurtyB has quit (Ping timeout: 276 seconds)
1040 2013-02-25 20:30:35 <QM> ah nm, think I found it
1041 2013-02-25 20:33:23 Nachtwind has quit (Quit: ZNC - http://znc.in)
1042 2013-02-25 20:34:00 PhantomSpark has joined
1043 2013-02-25 20:35:07 <jgarzik> gmaxwell: indeed :)  such is life, when one publicly changes a stated position
1044 2013-02-25 20:35:47 <jgarzik> gmaxwell: Ten years ago, trolls were suggesting that I would destroy Linux
1045 2013-02-25 20:35:48 <gmaxwell> Does seem kinda odd that you mentioned changing your position here and then .. boom necropost.
1046 2013-02-25 20:37:01 <jgarzik> it had been alluded to a little while before the necropost.  necropost was merely executing at the level of a script kiddie at that point
1047 2013-02-25 20:37:51 <jgarzik> the ball had been teed up, as it were
1048 2013-02-25 20:40:24 dvide has quit ()
1049 2013-02-25 20:40:37 whizter has joined
1050 2013-02-25 20:46:14 brwyatt is now known as brwyatt|Away
1051 2013-02-25 20:46:29 mappum has joined
1052 2013-02-25 20:46:55 <TD> jgarzik: well, you have to admit, you got involved with linux and then …. the year of linux on the desktop never came. 2+2 if you ask me ;)
1053 2013-02-25 20:47:09 <jgarzik> hehe
1054 2013-02-25 20:47:21 WolfAlex_ has joined
1055 2013-02-25 20:47:24 rdymac has quit (Remote host closed the connection)
1056 2013-02-25 20:47:34 WolfAlex has quit (Read error: Operation timed out)
1057 2013-02-25 20:48:50 <Luke-Jr> oh, now I know who to blame.. ;)
1058 2013-02-25 20:53:20 grau has quit (Remote host closed the connection)
1059 2013-02-25 20:55:01 * sipa has Linux on his desktop
1060 2013-02-25 20:55:24 * Luke-Jr as well
1061 2013-02-25 20:55:33 <gmaxwell> Me too.
1062 2013-02-25 20:55:42 <sipa> i don't really care whether people think about "the year of" - it works for me and for many others
1063 2013-02-25 20:56:26 alexwaters has joined
1064 2013-02-25 20:56:33 alexwaters has left ()
1065 2013-02-25 20:57:26 <RBecker> so why is it that the 0.8 client pegs my cpu at 100% as it downloads new blocks
1066 2013-02-25 20:57:39 <TD> rbecker: that's a question more suited to #bitcoin
1067 2013-02-25 20:57:46 <RBecker> this is the development channel
1068 2013-02-25 20:57:46 <TD> rbecker: but briefly, because it's checking signatures
1069 2013-02-25 20:57:51 <gmaxwell> (and I can see two other people's linux deskops from where I'm sitting, and besides if you care about "linux" instead of GNU, android some crazy 75% market share now or something in mobile— in some ways it's like alchemy, we got the ability to turn lead into gold a long time back— though cost prohibitively— and no one even bothered with a headline "scientists turn lead in to gold!", subtle changes are more powerful)
1070 2013-02-25 20:57:55 <TD> yes. and your question is not a development question
1071 2013-02-25 20:58:43 <Luke-Jr> gmaxwell: meh, Android isn't Linux
1072 2013-02-25 20:59:22 alexwaters3 has quit (Ping timeout: 256 seconds)
1073 2013-02-25 20:59:22 <gmaxwell> Luke-Jr: I agree, other people don't. When they say Linux they mean GNU which is what rms has been whining about for all these years. :P  but I'm being naughty with the OT banter.
1074 2013-02-25 20:59:52 freakazoid has quit (Ping timeout: 245 seconds)
1075 2013-02-25 20:59:58 makomk has quit (Ping timeout: 255 seconds)
1076 2013-02-25 21:00:25 i2pRelay has quit (Remote host closed the connection)
1077 2013-02-25 21:01:31 <sipa> RBecker: the best answer is that it's because it's faster
1078 2013-02-25 21:01:53 <RBecker> it made my computer completely unusable when it was doing conversion from the old format
1079 2013-02-25 21:02:12 <sipa> RBecker: there is some unavoidable CPU time for verification of the chain, so the faster the client is, the higher the CPU usage will be
1080 2013-02-25 21:02:20 <RBecker> right
1081 2013-02-25 21:02:25 <sipa> RBecker: really, what OS/specs?
1082 2013-02-25 21:02:34 <RBecker> My mouse cursor lagged accross the screen
1083 2013-02-25 21:02:44 <RBecker> Windows 7 x64, quad-core AMD 3.2GHz, 16GB RAM
1084 2013-02-25 21:03:08 <sipa> gavinandresen: ^ i consider that an argument in favor of thread priorities on windows
1085 2013-02-25 21:03:32 <TD> rbecker: what storage type?
1086 2013-02-25 21:03:33 <sipa> RBecker: you can limit the number of threads used for verification (put par=1 in the config file, or use -par=1 on the command line)
1087 2013-02-25 21:03:36 <TD> rbecker: hdd?
1088 2013-02-25 21:03:38 <RBecker> TD: SSD
1089 2013-02-25 21:04:03 <gavinandresen> sipa: ok, as long a somebody actually tests it to make sure setting the thread priority fixes the problem.
1090 2013-02-25 21:04:07 <TD> lagging mouse is normally caused by issues with storage or bad drivers. high userland cpu usage shouldn't be causing that regardless of the thread priority
1091 2013-02-25 21:04:10 <sipa> gavinandresen: ACK
1092 2013-02-25 21:04:21 <RBecker> TD: if my CPU is pegged with an intensive task it slows
1093 2013-02-25 21:05:22 * gavinandresen wonders why Windows hasn't caught up to SGI graphics in 1989, which had hardware/kernel support for the mouse cursor….
1094 2013-02-25 21:05:27 <TD> if a regular program is able to make the OS UI unusable there's a serious problem with the OS, regardless of what the app i sdoing
1095 2013-02-25 21:05:30 <sipa> TD: i made the same assumption before when someone on the forum made a similar complaint, but afaik that person actually verified that changing the thread (perhaps process) CPU priority improved the situation
1096 2013-02-25 21:05:32 <Scrat> RBecker: hasn't happened to me in the last 6 years
1097 2013-02-25 21:05:58 <TD> gavinandresen: it has. that's why i'm questioning whether it's really CPU
1098 2013-02-25 21:06:19 <TD> huh
1099 2013-02-25 21:06:21 <TD> ok
1100 2013-02-25 21:06:52 <TD> well, if they did that and it fixed it, i guess that's unarguable. windows never _used_ to work like that. i suspect it's actually something to do with IO scheduling on particular devices and IOPs inherit the priority of their invoking thread
1101 2013-02-25 21:07:53 <RBecker> BOINC runs my CPU at 100% all the time but I guess it's not as intensive as a task, it doens't cause any slowdown at all
1102 2013-02-25 21:08:02 <RBecker> and yes, it was suspended while bitcoin-qt was doing its thing
1103 2013-02-25 21:08:05 sgornick has joined
1104 2013-02-25 21:08:19 <TD> yeah. i strongly suspect the drivers for your SSD device are bogus
1105 2013-02-25 21:08:23 <TD> but i can't prove it
1106 2013-02-25 21:08:45 <RBecker> i'm pretty sure they came from the manufacturer
1107 2013-02-25 21:09:04 <RBecker> i could check device manager to see if they're generic MS
1108 2013-02-25 21:09:38 <TD> heh
1109 2013-02-25 21:09:47 <TD> i'd trust microsoft to write drivers a lot more than some random OE
1110 2013-02-25 21:09:48 <TD> OEM
1111 2013-02-25 21:09:52 <RBecker> it is a MS driver
1112 2013-02-25 21:10:09 <RBecker> write cache is on
1113 2013-02-25 21:10:26 okok has quit (Quit: ChatZilla 0.9.90 [Firefox 19.0/20130215130331])
1114 2013-02-25 21:10:33 <TD> ok. if you want to investigate you could try different drivers, see if it makes any difference. it could be a wild goose chase though
1115 2013-02-25 21:10:34 <RBecker> it's a sata iii drive connected to a iii port and using a iii cable
1116 2013-02-25 21:10:50 <RBecker> it's the higest scoring device in the windows performance index
1117 2013-02-25 21:11:13 <TD> rbecker: http://www.sevenforums.com/hardware-devices/199672-intermittent-but-extreme-ssd-lag.html
1118 2013-02-25 21:11:24 <TD> try checking the SSD access time
1119 2013-02-25 21:11:26 <TD> http://www.alex-is.de/PHP/fusion/downloads.php?download_id=9
1120 2013-02-25 21:11:41 <gmaxwell> RBecker: what height are you at in the sync?
1121 2013-02-25 21:11:47 <RBecker> it's finished now
1122 2013-02-25 21:11:58 <gmaxwell> RBecker: was slow in the very beginning?
1123 2013-02-25 21:12:11 <RBecker> my computer was slow the whole time it was processing
1124 2013-02-25 21:12:33 <TD> it sounds like other people have seen this issue as well. if the SSD benchmarking tool causes the same issue, i guess that's the problem
1125 2013-02-25 21:12:39 <TD> in which case the SSD might need to be repaired/replaced
1126 2013-02-25 21:12:39 <gmaxwell> If true— thats an argument for an IO problem, not something thread priorities would likely fix.
1127 2013-02-25 21:12:47 <RBecker> TD: it's less than a year old
1128 2013-02-25 21:12:50 <sipa> RBecker: i wonder if you'd be willing to retry the reindexing, but with lower number of sigthreads
1129 2013-02-25 21:12:52 <TD> :(
1130 2013-02-25 21:12:52 <RBecker> but i'll check alighment and write times
1131 2013-02-25 21:12:58 <RBecker> alignment*
1132 2013-02-25 21:12:58 <TD> ok
1133 2013-02-25 21:12:59 <TD> cool
1134 2013-02-25 21:13:20 <RBecker> sipa: how would I? it already converted the entire chain
1135 2013-02-25 21:13:38 <sipa> RBecker: start with -reindex -par=1
1136 2013-02-25 21:13:47 <RBecker> ok
1137 2013-02-25 21:13:54 <gmaxwell> ^ when you do that, it'll take a little while, like the import again.
1138 2013-02-25 21:13:59 <sipa> RBecker: that will wipe your databases, so you may want to make a backup (chainstate and blocks/index directories)
1139 2013-02-25 21:14:05 <RBecker> it's backed up
1140 2013-02-25 21:14:07 <sipa> RBecker: it won't wipe your block data though
1141 2013-02-25 21:14:49 <RBecker> hm
1142 2013-02-25 21:14:58 <RBecker> the 250MB system reserved partition has a 1024KB offset
1143 2013-02-25 21:15:08 <RBecker> but the actual in-use partition is 252MB
1144 2013-02-25 21:15:08 <sipa> that won't hurt
1145 2013-02-25 21:15:15 <TD> it's really the access times that matter
1146 2013-02-25 21:15:28 <RBecker> yeah, i'll check that next
1147 2013-02-25 21:17:53 <RBecker> it's testing
1148 2013-02-25 21:18:26 zooko has left ("#tahoe-lafs the secure, decentralized storage system")
1149 2013-02-25 21:19:36 <RBecker> it's definitely not access times: http://grabs.thetechgeek.org/desktop1162.png
1150 2013-02-25 21:20:25 <sipa> perhaps it's write delays
1151 2013-02-25 21:20:55 <TD> did you see the same lag?
1152 2013-02-25 21:21:05 <RBecker> nope
1153 2013-02-25 21:21:18 <RBecker> i wasn't really moving my mouse around, but i didn't notice anything
1154 2013-02-25 21:22:00 makomk has joined
1155 2013-02-25 21:22:13 <RBecker> i can run it again with constant mouse movement to see if i notice it if you want
1156 2013-02-25 21:22:31 <sipa> i doubt you will
1157 2013-02-25 21:22:34 <RBecker> as do i
1158 2013-02-25 21:23:03 <sipa> if it is related to your SSD, i expect it to be delays because of wiping blocks
1159 2013-02-25 21:23:15 <RBecker> well i'll try reindexing with the lower threads
1160 2013-02-25 21:24:29 <gmaxwell> good. this is a useful test. If thats less obstructive then setting thread priorities may be helpful...
1161 2013-02-25 21:24:54 <RBecker> just making a backup right now
1162 2013-02-25 21:27:05 <sipa> jgarzik, gmaxwell: my turbo branch currently has 0.8.0 + #2016,#2017,#2061,#2199,#2272,#2275,#2282,#2333, and i'm testing that on my vps and my mining node; so far no problems
1163 2013-02-25 21:27:15 <sipa> well, git head + those
1164 2013-02-25 21:27:39 <RBecker> here goes
1165 2013-02-25 21:27:58 <RBecker> CPU is staying low, no system lag
1166 2013-02-25 21:27:58 <sipa> RBecker: by the way, that lag you noticed, was that before or after the last checkpoint?
1167 2013-02-25 21:28:01 <sipa> ;;bc,blocks
1168 2013-02-25 21:28:01 <gribble> 222167
1169 2013-02-25 21:28:09 <RBecker> what checkpoint?
1170 2013-02-25 21:28:30 <sipa> the last checkpoint happens when you're at about "6000 blocks left"
1171 2013-02-25 21:28:36 <RBecker> before
1172 2013-02-25 21:28:41 <RBecker> and it kept going after that
1173 2013-02-25 21:28:45 <jgarzik> sipa: including a valgrind run?
1174 2013-02-25 21:28:49 <sipa> ok, then the -par thing won't make a difference
1175 2013-02-25 21:28:51 <sipa> jgarzik: not yet
1176 2013-02-25 21:28:56 <RBecker> it was pretty much the entire reindex process
1177 2013-02-25 21:29:09 <RBecker> -par is making a difference, CPU is only at 50 something percent and there's no lag
1178 2013-02-25 21:29:19 <sipa> it can't make a difference yet
1179 2013-02-25 21:29:23 <gmaxwell> sipa: Thats why I asked if the slowness was right from the start.
1180 2013-02-25 21:29:24 <RBecker> mk
1181 2013-02-25 21:29:44 <sipa> RBecker: but the usage pattern throughout the reindexing changes significantly
1182 2013-02-25 21:29:49 <gmaxwell> RBecker: unless there is some very mysterious and unlikely bug, par does nothing at all until 6000 or so blocks before the end.
1183 2013-02-25 21:29:54 <RBecker> we'll see what happens at the checkpoint
1184 2013-02-25 21:30:02 <RBecker> unless i'm just remembering wrong
1185 2013-02-25 21:30:15 <sipa> observational bias :)
1186 2013-02-25 21:30:22 clr_ has joined
1187 2013-02-25 21:30:25 <RBecker> wouldn't be surprised
1188 2013-02-25 21:31:09 <RBecker> 11k left
1189 2013-02-25 21:32:31 <RBecker> just kidding, i read that wrong, heh
1190 2013-02-25 21:32:36 <RBecker> 94k left
1191 2013-02-25 21:33:02 grau has joined
1192 2013-02-25 21:39:17 grau has quit (Read error: Operation timed out)
1193 2013-02-25 21:41:01 makomk has quit (Ping timeout: 252 seconds)
1194 2013-02-25 21:43:28 <sipa> jgarzik: heh, only took a few minutes to crash when running under valgrind :D
1195 2013-02-25 21:43:38 QM has quit (Ping timeout: 245 seconds)
1196 2013-02-25 21:44:40 fishfish has joined
1197 2013-02-25 21:45:11 clr_ is now known as c00w
1198 2013-02-25 21:46:40 fishfish has quit (Client Quit)
1199 2013-02-25 21:50:06 andytoshi has joined
1200 2013-02-25 21:50:09 brwyatt is now known as Away!~brwyatt@brwyatt.net|brwyatt
1201 2013-02-25 21:52:38 makomk has joined
1202 2013-02-25 21:56:25 <gjs278> RBecker you are in IDE mode
1203 2013-02-25 21:56:38 <gjs278> your drivers are not bogus, but you really should switchover to AHCI one day
1204 2013-02-25 21:56:51 <RBecker> how'd you come to that conclusion
1205 2013-02-25 21:56:56 <gjs278> your driver and your 4k 64
1206 2013-02-25 21:57:00 <gjs278> pciide
1207 2013-02-25 21:57:01 <gjs278> and
1208 2013-02-25 21:57:04 <RBecker> ah
1209 2013-02-25 21:57:05 <gjs278> your 4k 64 is too low
1210 2013-02-25 21:57:11 <RBecker> should be a bios option, right?
1211 2013-02-25 21:57:15 <gjs278> yes but here is the thing
1212 2013-02-25 21:57:22 <gjs278> when you install under IDE, windows expects IDE for the next boot
1213 2013-02-25 21:57:36 <gjs278> if you switch to AHCI in the bios, you may bluescreen on the next boot
1214 2013-02-25 21:57:46 <sipa> fun fun fun
1215 2013-02-25 21:58:01 <gjs278> you can switch back and be 100% fine, but if you blue screen, you have to do some sort of registry hack in ide mode, then reboot, then flip to ahci
1216 2013-02-25 21:58:13 <gmaxwell> If only someone here knew something about SATA drivers.
1217 2013-02-25 21:58:18 <RBecker> H
1218 2013-02-25 21:58:20 <RBecker> ah&
1219 2013-02-25 21:58:21 <RBecker> **
1220 2013-02-25 21:58:46 <RBecker> i'll try it after reindex
1221 2013-02-25 21:59:00 <gjs278> as for your stuttering, I was able to replicate the stuttering on c300 drives (the generation before m4 drives) by maxing out the write channel during benchmarks
1222 2013-02-25 21:59:13 <RBecker> c300 sounds familiar
1223 2013-02-25 21:59:16 <RBecker> i used to have one of those i think
1224 2013-02-25 21:59:28 <gjs278> I didn't have the opportunity to do the blockchain on them, but if it ever tried writing 75mb/s at one time, it would indeed cause mouse lag
1225 2013-02-25 21:59:50 <RBecker> confirmed, it was a c300
1226 2013-02-25 22:00:11 <RBecker> it's in my laptop now
1227 2013-02-25 22:01:18 <gjs278> the other thing is that when you are in ide mode, you don't have trim
1228 2013-02-25 22:01:33 <gjs278> without trim, m4s degrade pretty quickly under consistent writes if you have little free space
1229 2013-02-25 22:02:05 <gjs278> http://img832.imageshack.us/img832/9485/m4vsforcegt.png is a personal test done on an m4 drive that was nearly full without trim, left side is relevant here
1230 2013-02-25 22:02:25 Gladamas has quit (Read error: Connection reset by peer)
1231 2013-02-25 22:02:56 <gjs278> when you get your drive into ahci mode, your 4k 64 read will match mine, and your write should exceed mine if you enable trim
1232 2013-02-25 22:03:02 <RBecker> alright
1233 2013-02-25 22:03:36 <RBecker> Windows reports that TRIM is enabled
1234 2013-02-25 22:04:20 <gjs278> yeah, it's enabled, but it has no target to trim yet
1235 2013-02-25 22:04:30 <RBecker> ah
1236 2013-02-25 22:04:56 <gjs278> the pciide driver thing will change to something like amd ahci or msachi - GOOD when it's flipped
1237 2013-02-25 22:05:17 <gjs278> er - OK
1238 2013-02-25 22:06:08 <RBecker> gonna hit the checkpoint soon
1239 2013-02-25 22:06:18 <RBecker> cpu use has gone up
1240 2013-02-25 22:06:23 <sipa> RBecker: what block are you at?
1241 2013-02-25 22:06:28 <RBecker> 7620
1242 2013-02-25 22:06:33 <RBecker> remaining
1243 2013-02-25 22:06:40 <sipa> no, what block are you at
1244 2013-02-25 22:06:46 <sipa> (hover over the sync icon)
1245 2013-02-25 22:07:03 <RBecker> 216051
1246 2013-02-25 22:07:09 juchmis has quit (Ping timeout: 246 seconds)
1247 2013-02-25 22:07:13 <sipa> not yet at the checkpoint then, but very close
1248 2013-02-25 22:07:17 <sipa> it's at 216116
1249 2013-02-25 22:07:19 <RBecker> i said about to be
1250 2013-02-25 22:07:24 <sipa> right :)
1251 2013-02-25 22:07:29 <RBecker> it hit it
1252 2013-02-25 22:07:47 <RBecker> block 216147
1253 2013-02-25 22:07:52 <RBecker> no performance drop, no lag
1254 2013-02-25 22:08:20 <sipa> no performance drop??
1255 2013-02-25 22:08:28 CaptainBlaze has quit (Quit: CaptainBlaze)
1256 2013-02-25 22:08:33 <RBecker> nope
1257 2013-02-25 22:08:33 <sipa> with -par=1 it should slow down a factor 5-10 or so
1258 2013-02-25 22:08:38 <RBecker> cpu isn't pegged
1259 2013-02-25 22:08:47 <sipa> heh
1260 2013-02-25 22:08:58 <RBecker> it's reindexing these blocks a lot slower tho
1261 2013-02-25 22:09:12 <Luke-Jr> …
1262 2013-02-25 22:09:13 <sipa> oh
1263 2013-02-25 22:09:19 <sipa> that's what i meant with performance :)
1264 2013-02-25 22:09:29 <RBecker> i was talking about system performance
1265 2013-02-25 22:09:38 <sipa> ok
1266 2013-02-25 22:17:20 toffoo has joined
1267 2013-02-25 22:18:16 ThomasV has quit (Ping timeout: 252 seconds)
1268 2013-02-25 22:19:32 bitafterbit has quit (Remote host closed the connection)
1269 2013-02-25 22:31:58 <muhoo> TD: what refactoring of Wallet would you be imterested in merging?
1270 2013-02-25 22:32:12 <TD> good question
1271 2013-02-25 22:32:20 <TD> what was your goal again? sorry, i know you told me
1272 2013-02-25 22:32:33 <muhoo> storefront
1273 2013-02-25 22:32:36 <TD> right now i seem to have people saying they want to refactor things almost every day and it's hard to keep up the mapping between email names and irc nicks
1274 2013-02-25 22:32:39 <TD> right
1275 2013-02-25 22:32:48 <TD> so right now the Wallet is a huge class.
1276 2013-02-25 22:33:06 <TD> a good place to start is just breaking it out into a separate package and making its subclasses real top level classes
1277 2013-02-25 22:33:10 <TD> just to make the file smaller
1278 2013-02-25 22:33:25 <muhoo> ok
1279 2013-02-25 22:33:48 <TD> after that, you'd need to introduce a new abstraction for holding the transactions and keeping them in the right pools
1280 2013-02-25 22:34:21 <HM> http://rsa.edgeboss.net/download/rsa/rsaconference/2013/us/podcasts/rsac_02-12-13-exp-t17.mp3 <-- Schneier interview with RSA talking about  "feudal" security, and how technology is moving toward centralised trust. He's talking tomorrow at RSA2013 about the same thing
1281 2013-02-25 22:34:22 <TD> as a first step you'd just keep the in-memory based code that can save to a file.
1282 2013-02-25 22:34:42 <muhoo> as oposed to WalletTransaction?
1283 2013-02-25 22:34:45 root2 has joined
1284 2013-02-25 22:35:03 <TD> once you've got a clean split of "code that stores stuff" from "code that manages stuff" (the old class), you can start to rewrite the code to be transactional in nature. right now the code just assumes it can move transactions around by adding and removing from hashmaps
1285 2013-02-25 22:35:20 <TD> so you'd need to introduce a concept of a wallet store transaction which can be mapped through to a database transaction
1286 2013-02-25 22:35:44 <TD> and then make all the code in, eg, receive() or reorganize() operate in terms of batches of transactions.
1287 2013-02-25 22:36:14 <TD> it's a bit confusing because of course the code already uses the word transaction to mean a change to the global Bitcoin database. satoshis code was confusing to read for the same reason. so i'd suggest trying to find an alternative name, even if it's a bit unusual
1288 2013-02-25 22:36:28 <muhoo> i see why this hasn't ben done yet :-)
1289 2013-02-25 22:36:37 <TD> muhoo: well a WalletTransaction is just a Transaction + what pool it's in. what you need is an object that holds groups of transactions
1290 2013-02-25 22:36:40 <TD> like the wallet does today
1291 2013-02-25 22:36:43 <TD> yes
1292 2013-02-25 22:36:49 <TD> it's not going to be easy.
1293 2013-02-25 22:37:02 <TD> the current code gains a lot of simplicity from assuming everything is in memory all the time
1294 2013-02-25 22:37:49 <TD> in theory you could just rewrite the Wallet code to use SQL directly for everything, but then every app would need to use  a SQL database for its wallet
1295 2013-02-25 22:37:55 ThomasV has joined
1296 2013-02-25 22:38:09 <HM> that's not so burdensome, you have light options like SQLlite
1297 2013-02-25 22:38:23 <muhoo> or, an interface, withna memory backend
1298 2013-02-25 22:38:43 <TD> muhoo: yes that's what i'm suggesting. first split the current code that just uses in memory hashmaps and saves them to a file in one go
1299 2013-02-25 22:38:55 <muhoo> in-memory "db"
1300 2013-02-25 22:38:57 <TD> muhoo: and then later start moving it in the direction of a real database
1301 2013-02-25 22:39:42 <muhoo> seems reasonable.
1302 2013-02-25 22:39:58 <TD> HM: yeah, it's possible, but it'd mean having to convert everyones wallets from the old format and then of course, some people would say "oh i want to use a non-sql database", etc.
1303 2013-02-25 22:40:12 <HM> different users have different query requirements
1304 2013-02-25 22:40:21 <TD> HM: so i'd rather keep the core wallet code independent of how exactly it's stored, if possible.
1305 2013-02-25 22:40:27 <HM> you see this already with txindex=0
1306 2013-02-25 22:41:06 <TD> if that's not going to work out well then we can consider making SQLite the canonical db file format and say anyone who wants more needs something that supports some subset of SQL
1307 2013-02-25 22:41:12 <HM> you could argue the fact that you even have txindex is an argument for a proper full DBMS, because then you can add or remove indexes at will
1308 2013-02-25 22:41:33 <TD> HM: this is specifically for the wallet, we're talking about. so you don't really need a tx index there.
1309 2013-02-25 22:41:33 <HM> like postgres has "partial indexes" so i can index all tx's that start with 67e8 or whatever i fancy
1310 2013-02-25 22:41:58 <muhoo> i won't be usingnsqlite or even sql possibly
1311 2013-02-25 22:42:20 <muhoo> so i'd be disinclined to tie it to that
1312 2013-02-25 22:42:24 <TD> however, you will need a storage system that can modify state atomically
1313 2013-02-25 22:42:35 <TD> that's probably your first job, really
1314 2013-02-25 22:42:41 <TD> figure out what constraints will be imposed on the storage layer
1315 2013-02-25 22:43:01 <HM> i guess
1316 2013-02-25 22:43:06 <TD> like, can you implement a wallet purely on top of a BigTable style storage model? what about LevelDB which provides cross-row atomicity?
1317 2013-02-25 22:43:21 <TD> or do you need more advanced indexing and querying than sstables can provide?
1318 2013-02-25 22:43:24 <TD> etc
1319 2013-02-25 22:43:29 <muhoo> will i need transactions?
1320 2013-02-25 22:43:46 <TD> that depends a lot on how you represent the data in your storage system doesn't it.
1321 2013-02-25 22:43:52 <muhoo> i.e. will wallet be making several calls which need to all be treated as one atomic operation?
1322 2013-02-25 22:43:56 <TD> all these things have to be considered
1323 2013-02-25 22:44:10 <TD> well ok then, your actual first job is to read and understand the wallet code so you can tell me ;)
1324 2013-02-25 22:44:31 <muhoo> yeah, i've done the first part of that but not the second.
1325 2013-02-25 22:44:34 <TD> but i think the answer is yes, because during a re-organize the state of many transactions needs to change simultaneously
1326 2013-02-25 22:44:41 <HM> it was mentioned yesterday that importing a private key is expensive
1327 2013-02-25 22:45:03 COGSMITH has joined
1328 2013-02-25 22:45:32 <TD> HM: if you don't have a pre-calculated index of scripts to transactions then yes, you have to scan the chain since the keys creation time
1329 2013-02-25 22:45:58 <TD> muhoo: do you need the ability for multiple servers to read/write the same wallet in parallel?
1330 2013-02-25 22:46:00 <HM> yes, the whole chain
1331 2013-02-25 22:46:07 <muhoo> but there's no way to replay the chain in bitcoinj right?
1332 2013-02-25 22:46:22 <TD> muhoo: you can use wallet.clearTransactions() and then just delete the chain file and resync
1333 2013-02-25 22:46:30 <TD> that's how multibit/bitcoin wallet do it
1334 2013-02-25 22:46:38 <muhoo> ah yes, i think i've seen that in there.
1335 2013-02-25 22:46:44 <HM> https://www.sqlite.org/famous.html <-- lol @ "Flame"
1336 2013-02-25 22:46:45 <TD> muhoo: what you can't do is replay just part of the chain selectively.
1337 2013-02-25 22:46:56 <TD> HM: haha
1338 2013-02-25 22:46:59 <TD> classy
1339 2013-02-25 22:47:12 <TD> muhoo: though that's not very hard to fix.
1340 2013-02-25 22:47:20 <muhoo> TD: i was not planning to have multiple servers hitting one wallet
1341 2013-02-25 22:47:35 <muhoo> but there may be multiple threads oing that
1342 2013-02-25 22:47:40 <sipa> HM: no such thing as bad publicity :D
1343 2013-02-25 22:47:57 <muhoo> in fact, it's 100% certain i will have multiple threads doing that.
1344 2013-02-25 22:48:00 <HM> Nigerian scammers: trusting Bitcoin since 2009
1345 2013-02-25 22:48:01 whizter has quit ()
1346 2013-02-25 22:48:28 <TD> muhoo: you'd need to serialize some operations. like, you can't have two threads simultaneously trying to create spends.
1347 2013-02-25 22:48:43 <TD> muhoo: of course the wallet already serializes everything already. just be aware in case you decide to relax that.
1348 2013-02-25 22:49:18 <muhoo> ok, well this is in my path, so i suppose there is no way to avoid tha pain
1349 2013-02-25 22:49:18 <TD> muhoo: one question in my mind is whether you can actually scale up a large storefront without using a real db for the wallet at all. right now the biggest problem with loading the wallet into memory is that we store all spent transactions in it
1350 2013-02-25 22:49:29 <TD> muhoo: which is kind of pointless if you aren't building a user-facing wallet app
1351 2013-02-25 22:49:43 <muhoo> right, there is stuff i'll need to no-op
1352 2013-02-25 22:49:58 <TD> muhoo: if you remove spent transactions, then the wallet only has to be large enough to store transactions with unspent outputs.
1353 2013-02-25 22:50:29 <HM> unspent indexes presumably become more complex when non-standard scripts are involved as well
1354 2013-02-25 22:50:35 <TD> muhoo: assuming you actually spend the money you earn, it's possible that your wallets (if they didn't store spend txns) would never get larger than can fit in memory. in which case your whole mission becomes radically simplified because then you can just use the existing code that saves wallets to a file
1355 2013-02-25 22:51:02 <muhoo> would you be open for a more brute force approach such as, creating a wallet interface, having the existing wallet class implment it, and letting me create a totally separate FooWallet that plugs in and replaces it for my app?
1356 2013-02-25 22:51:04 <TD> muhoo: so you could just write a patch that moves spent transactions out into some other kind of stable storage, like a database or just a log file.
1357 2013-02-25 22:51:33 <TD> muhoo: you can already do that - Wallet implements BlockChainListener. You are free to create your own BlockChainListener. if you think it'd help we could make an interface for parts of the wallets public API sure.
1358 2013-02-25 22:51:48 <TD> muhoo: though if you look at the wallet, the code is fairly complicated. it'd suck to end up with a duplicate.
1359 2013-02-25 22:52:05 <muhoo> transactionlistener is another one i need. and there'd be a lot of cut and paste
1360 2013-02-25 22:52:21 <TD> :(
1361 2013-02-25 22:52:24 <muhoo> not transactionlistener, the peereventlistener, is what i'm using.
1362 2013-02-25 22:52:34 <TD> why don't you start by fixing the existing wallet implementation so saving of spent transactions is optional
1363 2013-02-25 22:52:48 <TD> and there's some way to plug in spent transaction evictors …. so you can keep the in-memory wallet small
1364 2013-02-25 22:53:12 <TD> if you run into scalability problems after that and can't optimize them away, then keeping the wallets working set in a database may well make sense
1365 2013-02-25 22:53:55 <muhoo> it's a balance, really. i could easilly see getting very distracted by major work on Wallet and not getting my app done, too.
1366 2013-02-25 22:54:13 <TD> sure. adding a spent tx evictor feature would be quite easy.
1367 2013-02-25 22:54:31 <TD> as all you're really doing is removing Transaction objects from the spent hashmap and stuffing them into some kind of record keeping system from time to time
1368 2013-02-25 22:54:49 <TD> heck, maybe you don't even care about recording spent transactions.
1369 2013-02-25 22:54:52 <muhoo> oh.
1370 2013-02-25 22:55:03 <TD> in which case your evictor could just be a no-op
1371 2013-02-25 22:55:24 <muhoo> i kind of like that. and do all my store accounting outside of wallet completely?
1372 2013-02-25 22:55:52 <muhoo> that seems much better.
1373 2013-02-25 22:56:07 <TD> i guess. i mean the purpose of the wallet is to keep transactions, let you know when they change, update their confidence data, and let you create spends.
1374 2013-02-25 22:56:49 <TD> when somebody buys from you a new transaction enters the wallet and is available for spending. once you move the coins (like to an exchange) the transaction becomes basically useless. after it's buried deep enough you can just throw it away, unless you have a burning desire to keep it around for record keeping purposes
1375 2013-02-25 22:57:28 <TD> same thing for keys. the current wallet won't actually let you delete keys, because that's rather dangerous, but for a storefront you could assume that after a while keys won't be re-used anymore and those too can be evicted from the wallet and written to a log file
1376 2013-02-25 22:57:54 <TD> if somebody does decide it's a good idea to send money to an inactive address, you can always go into the log and get the key back, but it wouldn't have to be sitting in RAM on your server all the time.
1377 2013-02-25 22:57:54 <muhoo> the only think i really need wallet for is creating spends, now that i think of it
1378 2013-02-25 22:58:02 <TD> well, that's not quite true
1379 2013-02-25 22:58:06 <TD> remember double spending attacks
1380 2013-02-25 22:58:17 <TD> if you think you received money and then it gets double spent away, you need to know about that.
1381 2013-02-25 22:58:28 <TD> that's what the TransactionConfidence data is for
1382 2013-02-25 22:58:32 <TD> (amongst other things)
1383 2013-02-25 22:58:36 <muhoo> in this case, i'm watching on a peereventlistener, just to see when the tx is broadcast, i'm not waiting for the chain
1384 2013-02-25 22:58:58 <muhoo> yes, transactionconfidence is very useful.
1385 2013-02-25 22:59:16 <TD> peereventlistener isn't the right place for that, imho. you can just register a WalletEventListener and it'll fire when you receive a pending transaction that's broadcast
1386 2013-02-25 22:59:38 <TD> you can then add a listener to its confidence object to find out when you are happy that the money is really yours, so you can …. i dunno …. ship your widgets
1387 2013-02-25 23:00:11 <TD> for a storefront the most secure setup is to not even have your web server able to create spends at all
1388 2013-02-25 23:00:16 <muhoo> yep, it's been a few weeks since i really worked seriously on this, so my memory is fuzzy, but that is exactly what i was planning to do.
1389 2013-02-25 23:00:19 <TD> all the private keys are held somewhere else.
1390 2013-02-25 23:00:34 <TD> and then all the wallet does, really, is keep track of the state of transactions
1391 2013-02-25 23:01:18 <muhoo> good point. since all tx'es are public, it's easy to decouple the two. nice
1392 2013-02-25 23:01:34 stalled has quit (Ping timeout: 255 seconds)
1393 2013-02-25 23:02:19 <muhoo> well, i think i'm somewhat less intimidated by this now than i have been.
1394 2013-02-25 23:02:24 [\\\] has quit (Remote host closed the connection)
1395 2013-02-25 23:02:39 <TD> yeah. i think it's not so hard to scale a bitcoinj based store front, now we talk about it.
1396 2013-02-25 23:02:51 <TD> even if you need 10 frontend servers, nothing stops you having ten in memory wallets.
1397 2013-02-25 23:03:04 <TD> there's nothing that says one website == one wallet
1398 2013-02-25 23:03:36 <muhoo> kindof like the old days when people scaled mysql by having read-only databases syncing to on write-only database
1399 2013-02-25 23:03:59 <HM> it's either a replication or a routing problem
1400 2013-02-25 23:04:49 <muhoo> so i do not need to replace the wallet class, just do that small mod to expire out transactions?
1401 2013-02-25 23:05:13 <muhoo> so the thing doesn't become a memory leak
1402 2013-02-25 23:05:30 <TD> yeah, when the current code inserts stuff into the sent map today, replace that with something like if (!spentTxEvictor.take(tx)) spent.put(...);
1403 2013-02-25 23:05:50 <TD> ie, keep the current behaviour unless someone configured a separate object that takes responsibility for doing something with spent transactions
1404 2013-02-25 23:05:52 [\\\] has joined
1405 2013-02-25 23:06:02 <TD> your code could just delete them, but i suppose having a convenient audit trail is worth a bit of extra coding
1406 2013-02-25 23:06:12 <muhoo> i'll try that. and yes i do want to save the history
1407 2013-02-25 23:06:20 <TD> you might also want to consider doing the same thing for keys, but that's a bit trickier
1408 2013-02-25 23:06:29 <TD> as there is no direct equivalent of being "used up" there.
1409 2013-02-25 23:06:45 <muhoo> that is a problem. i will be generating keys like there's no tomorrow
1410 2013-02-25 23:06:50 <muhoo> each sale is a eckey
1411 2013-02-25 23:07:08 <muhoo> each POTENTIAL sale is an eckey. even if the user doesn't complete it
1412 2013-02-25 23:07:30 <TD> indeed
1413 2013-02-25 23:07:51 <TD> the notion of "completed" is rather app-specific, isn't it.
1414 2013-02-25 23:08:10 <muhoo> and i am tracking them so i know what they bought
1415 2013-02-25 23:08:23 <TD> this is something we'll have to think about when doing deterministic wallets -- how to throw away old keys in a safe way.
1416 2013-02-25 23:09:12 <TD> muhoo: yeah the risk is that someone will try and send coins to an old address for some random reason. i suppose if your web server only has public keys it's not a big deal, whatever computer you have that's collecting your received funds and moving them just needs to be able to hold all keys in rAM
1417 2013-02-25 23:09:26 <TD> hmmm
1418 2013-02-25 23:09:27 c00w has quit (Ping timeout: 245 seconds)
1419 2013-02-25 23:09:32 <TD> i can think of a few ways to scale that
1420 2013-02-25 23:09:49 <muhoo> yep that could get sticky
1421 2013-02-25 23:10:00 <TD> still. for now just evicting pubkeys that were "used" is simple enough
1422 2013-02-25 23:10:15 <TD> worst case, you have to go into the log and fetch an old key if somebody sends money to it and then complains nothing happened
1423 2013-02-25 23:10:15 <muhoo> they are indeed one-time use
1424 2013-02-25 23:10:36 <muhoo> easier to evict the used ones than to try to guess which ones are old enough to remove
1425 2013-02-25 23:10:37 <TD> muhoo: well, "use" in this case means "achieved a balance equal to the invoiced amount"
1426 2013-02-25 23:11:11 ThomasV has quit (Ping timeout: 264 seconds)
1427 2013-02-25 23:11:21 <muhoo> i'm not going to make it easy for someone to buy something with multiple tx'es
1428 2013-02-25 23:11:26 <muhoo> wayy to complex
1429 2013-02-25 23:11:41 <HM> why? do you care who paid as long as it's paid?
1430 2013-02-25 23:11:47 <muhoo> UI
1431 2013-02-25 23:12:05 <muhoo> i'm trying to make this dog-simple..
1432 2013-02-25 23:12:09 JDuke128 has joined
1433 2013-02-25 23:12:35 <sipa> TD: with a payment protocol, you could much more easily decide to throw away old keys (or at least move them to less-readily accessible storage), as you can give each key a TTL, and if a transaction for an expired key arrives, simply refuse it
1434 2013-02-25 23:13:13 <muhoo> HM: so i just show them an address, with a QR and a bitcoin: url, and say "pay here'.
1435 2013-02-25 23:13:27 [\\\] has quit (Remote host closed the connection)
1436 2013-02-25 23:14:12 <sipa> TD: or, less dramatically, decide to reuse keys that didn't get a succesful transaction within a certain timeframe
1437 2013-02-25 23:14:48 D34TH has joined
1438 2013-02-25 23:14:49 D34TH has quit (Changing host)
1439 2013-02-25 23:14:49 D34TH has joined
1440 2013-02-25 23:14:57 <TD> muhoo: it's pretty easy to keep track of how much money an address has received.
1441 2013-02-25 23:15:14 <muhoo> true. and i may just keep a timestamp outside of bitcoinj
1442 2013-02-25 23:15:15 <TD> sipa: yeah. you probably don't want to reuse unused keys to avoid privacy leaks
1443 2013-02-25 23:15:16 [\\\] has joined
1444 2013-02-25 23:15:34 <muhoo> and i can hopefully find a way to wipe old keys that way
1445 2013-02-25 23:15:47 <TD> muhoo: i think paying with multiple txns will become quite common in future, btw, so it's worth handling properly.
1446 2013-02-25 23:16:19 <TD> muhoo: whenever you receive a transaction you can either (1) keep some external count of how much money is left on that particular sale, or (2) use getBalace() with a coin selector that only selects coins sent to that address
1447 2013-02-25 23:16:23 <TD> muhoo: to find out the balance of that address
1448 2013-02-25 23:16:35 <TD> the latter is a bit slower as it has to iterate over all the unspent outputs
1449 2013-02-25 23:16:44 <muhoo> TD: fair enough. i'd do that in the transationconfidencelistener
1450 2013-02-25 23:17:00 <TD> yeah, that's fine. are you going to wait for blocks or just accept pending txns?
1451 2013-02-25 23:17:17 <muhoo> pending with > 4 or so peers
1452 2013-02-25 23:17:25 <TD> what are you selling?
1453 2013-02-25 23:17:31 <muhoo> music
1454 2013-02-25 23:17:35 <TD> this is a server that can't handle full mode, right?
1455 2013-02-25 23:17:39 <TD> yeah, ok
1456 2013-02-25 23:17:46 <HM> merchants should probably afford to outsource some of this stuff provided they understand and limit their exposure
1457 2013-02-25 23:17:56 <TD> that's probably good enough, especially if you pick the nodes you use by hand (not using discovery)
1458 2013-02-25 23:18:04 <TD> if we had double-spend alerts it'd be better
1459 2013-02-25 23:18:10 <TD> (btw if you pick by hand don't tell anyone which nodes you picked)
1460 2013-02-25 23:18:17 <TD> (it makes it easier to do timing attacks on you)
1461 2013-02-25 23:18:27 <TD> HM: well that's what bitpay is for
1462 2013-02-25 23:18:28 <muhoo> HM: what i'm creating, if i do it right, will be something merchants can outsource to :-)
1463 2013-02-25 23:19:35 <HM> TD, exactly.
1464 2013-02-25 23:19:38 <TD> for a server, you could probably use a lot more than 4 nodes
1465 2013-02-25 23:19:48 <muhoo> TD: i think i have 8 currently
1466 2013-02-25 23:19:56 <muhoo> i could wait for all 8.
1467 2013-02-25 23:20:51 <TD> i'd be tempted to use more like 20 or 30 and the wait for 80%+ to acknowledge
1468 2013-02-25 23:20:56 <TD> but you can experiment with what works
1469 2013-02-25 23:21:07 slush1 has quit (Ping timeout: 245 seconds)
1470 2013-02-25 23:21:18 grau has joined
1471 2013-02-25 23:21:26 <TD> the trick is that if you don't use very many, and somebody finds out which nodes you use, they could try double-spending against you by sending two transactions out simultaneously
1472 2013-02-25 23:21:59 <muhoo> sure, and it wouldn't be hard to figure out my ip
1473 2013-02-25 23:22:25 <muhoo> then again, maybe not
1474 2013-02-25 23:22:48 DamascusVG has joined
1475 2013-02-25 23:22:48 DamascusVG has quit (Changing host)
1476 2013-02-25 23:22:48 DamascusVG has joined
1477 2013-02-25 23:22:49 <muhoo> maybe the machine that's sending tx'es is not the same one listening to determine if the sale is good
1478 2013-02-25 23:23:30 Internet13 has quit (Quit: Leaving)
1479 2013-02-25 23:23:33 <muhoo> like you were saying, 10 servers each with their own wallets. i can split up these tasks, use some RPC, etc
1480 2013-02-25 23:24:51 <TD> yeah
1481 2013-02-25 23:25:33 <muhoo> TD: thank you very much for your help. i think i have enough to start with, which i will do.
1482 2013-02-25 23:25:43 grau has quit (Ping timeout: 248 seconds)
1483 2013-02-25 23:25:47 <muhoo> very relieved at not having to do major surgery on Wallet too
1484 2013-02-25 23:26:11 <TD> ok
1485 2013-02-25 23:26:40 <muhoo> have to take off for now. again, thanks very much.
1486 2013-02-25 23:29:37 agricocb has quit (Ping timeout: 252 seconds)
1487 2013-02-25 23:29:47 Toresh has quit (Read error: Connection reset by peer)
1488 2013-02-25 23:30:16 Toresh has joined
1489 2013-02-25 23:32:52 [\\\] has quit (Remote host closed the connection)
1490 2013-02-25 23:33:33 BurtyBB is now known as BurtyB
1491 2013-02-25 23:33:41 Internet13 has joined
1492 2013-02-25 23:34:40 [\\\] has joined
1493 2013-02-25 23:42:03 Internet13 has quit (Quit: Leaving)
1494 2013-02-25 23:47:38 [\\\] has quit (Remote host closed the connection)
1495 2013-02-25 23:48:09 Internet13 has joined
1496 2013-02-25 23:49:07 ielo has joined
1497 2013-02-25 23:53:44 [\\\] has joined
1498 2013-02-25 23:56:19 robocoin has quit (Quit: nope)
1499 2013-02-25 23:58:42 HM2 has joined
1500 2013-02-25 23:58:42 HM has quit (Read error: Connection reset by peer)