1 2012-02-06 00:00:21 <FROTUSCI> if youre a new node, you dont give up much by validating each tx in the background. the proof of work gives you enough assurance that the chain was difficult to forge, enough to start using it anyway under a yellow flag of caution
   2 2012-02-06 00:00:46 <tcatm> However, they wouldn't verify the transaction chains but instead ask other nodes to provide the merkle tree to verify that an output was indeed included in a block "long ago".
   3 2012-02-06 00:01:22 <gmaxwell> FROTUSCI: but if all nodes run your rules then, in fact it was trivial to forge.
   4 2012-02-06 00:01:47 <gmaxwell> I don't have any problems with SPV nodes, which are based on that "difficult to forge" assumption.
   5 2012-02-06 00:02:09 <gmaxwell> I have a problem with removing validation from full nodes which is the rational basis for the "difficult to forge" assumption.
   6 2012-02-06 00:02:13 <FROTUSCI> going forward you still do full validation, but you can "tentatively" accept past work and validate it in the background
   7 2012-02-06 00:02:35 <gmaxwell> FROTUSCI: it's much cheaper to validate the history once then to do it incrementally.
   8 2012-02-06 00:02:42 <gmaxwell> (and sure, if its in the background thats fine too)
   9 2012-02-06 00:02:53 Zarutian has quit (Read error: Operation timed out)
  10 2012-02-06 00:03:13 d4de has quit (Ping timeout: 260 seconds)
  11 2012-02-06 00:03:19 <FROTUSCI> what if a different data structure is used for the tx tree
  12 2012-02-06 00:03:47 <tcatm> Why? A merkle tree is a pretty good choice...
  13 2012-02-06 00:04:14 <gmaxwell> FROTUSCI: You mean like this? https://bitcointalk.org/index.php?topic=21995.0  (other people have since filled in some of the details in that concept-space)
  14 2012-02-06 00:04:33 d4de has joined
  15 2012-02-06 00:04:52 <gmaxwell> It still doesn't safe us from actually having lots and lots of people actually validate the full history in order to make bitcoin's properties hold true.
  16 2012-02-06 00:05:40 TD has quit (Quit: TD)
  17 2012-02-06 00:05:43 <gmaxwell> But thats okay, validating the full history is cheap and should always be fairly cheap relative to computing power (lets hope, it seems likely).  It's not great for good new user expirence, but thats why backgrounding it makes sense.
  18 2012-02-06 00:06:19 <gmaxwell> s/safe/save/
  19 2012-02-06 00:06:49 <FROTUSCI> yea a yellow flag saying "validating 27%" etc would be fine for the new user experience
  20 2012-02-06 00:07:17 <FROTUSCI> but dont rely on it for any multi-million btc transactions
  21 2012-02-06 00:07:24 <gmaxwell> Hell. I wouldn't bother with that much. I'd show all transactions as unconfirmed until the validation caught up and call it done.
  22 2012-02-06 00:07:42 <gmaxwell> The validation isn't really for the user themselves, it's for the faith of the network overall.
  23 2012-02-06 00:08:03 <FROTUSCI> if someone went to the trouble of creating a million-cpu-year chain with a bogus tx buried in it, to cheat you over 10 btc.. that is quite unlikely
  24 2012-02-06 00:08:11 <tcatm> There's nothing wrong with receiving a new transaction, waiting 6 blocks and then treat that transaction as 99.9994% verified.
  25 2012-02-06 00:08:31 <gmaxwell> tcatm: eh, if you also SPV validate the inputs.
  26 2012-02-06 00:08:39 <gmaxwell> oh I guess not.
  27 2012-02-06 00:08:42 <tcatm> Even if you don't.
  28 2012-02-06 00:08:48 <gmaxwell> Sure. Thats just SPV model trustl.
  29 2012-02-06 00:09:02 <tcatm> And that's fine for most users.
  30 2012-02-06 00:09:12 <gmaxwell> FROTUSCI: careful with your absolutes.
  31 2012-02-06 00:09:48 theorb has joined
  32 2012-02-06 00:10:39 theorbtwo has quit (Ping timeout: 240 seconds)
  33 2012-02-06 00:10:42 <FROTUSCI> how many addresses with >0 balance are ther?
  34 2012-02-06 00:10:47 theorb is now known as theorbtwo
  35 2012-02-06 00:11:58 <gmaxwell> tcatm: How about this: You put a node on a network I control. I will pay it 1000 BTC.  When that transaction 6 confirmations, you'll pay me 900 btc (from a seperate node, please). Easy 100 BTC profit for you? :)
  36 2012-02-06 00:12:16 <gmaxwell> On that subject. Miners using GPUMAX are now being paid 140-150% PPS for mining.
  37 2012-02-06 00:13:10 <gmaxwell> It sounds like it should soon be pretty easy to get the hashpower needed to make reasonably deep cuts for only a modest premium.
  38 2012-02-06 00:14:05 <gmaxwell> FROTUSCI: a half million or so IIRC.
  39 2012-02-06 00:14:19 <FROTUSCI> !bc,blocks
  40 2012-02-06 00:14:19 <gribble> 165539
  41 2012-02-06 00:14:25 <FROTUSCI> hmm
  42 2012-02-06 00:14:50 <tcatm> gmaxwell: In reality you wouldn't be able to control that network
  43 2012-02-06 00:14:52 Sedra- has quit (Remote host closed the connection)
  44 2012-02-06 00:15:32 <gmaxwell> tcatm: I control lots of networks, in reality.
  45 2012-02-06 00:16:08 h4ckm3 has quit (Ping timeout: 240 seconds)
  46 2012-02-06 00:16:44 <gmaxwell> If I were some criminal (and not just an engineer who wouldn't want to lose his job) I could perform isolation attacks against hundreds of thousands of people. There are lots of people who can.
  47 2012-02-06 00:16:51 iocor has quit (Quit: Computer has gone to sleep.)
  48 2012-02-06 00:17:00 <gmaxwell> In any case, I'm not dissing the SPV properties. It's good stuff. It has its limits though.
  49 2012-02-06 00:17:06 <tcatm> Sure, let's do that with one small modification: I'm allowed to forward a few coins from that 1000 BTC to a node on a network I control.
  50 2012-02-06 00:17:06 Sedra has joined
  51 2012-02-06 00:17:53 <gmaxwell> tcatm: can I mediate the forwarding?  (e.g. substitute the transaction with my own coins?) :)
  52 2012-02-06 00:18:19 BTC_Bear is now known as BTC_Bear|hbrntng
  53 2012-02-06 00:18:37 pingdrive has joined
  54 2012-02-06 00:18:45 <gmaxwell> actually fine you can do that.. I'd just make the block paying you impossible to validate. Then it doens't matter what you forward out, so long as I prevent you from seeing the longer chain.
  55 2012-02-06 00:18:47 JRWR has joined
  56 2012-02-06 00:19:36 <tcatm> You'd still have to fake blocks at a pretty high difficulty.
  57 2012-02-06 00:19:40 <FROTUSCI> TD=bitcoin master
  58 2012-02-06 00:20:55 <gmaxwell> tcatm: sure, but at that value I can just _buy_ the mining power and turn a profit.
  59 2012-02-06 00:21:53 <gmaxwell> It would cost me somewhere between 300-450 BTC (depending getting the blocks 'at cost' or having to use a competative market like gpumax) to mine the six bogus blocks.
  60 2012-02-06 00:22:38 <josephcp> does anyone have an example of a payout transaction from GPUMAX?
  61 2012-02-06 00:22:55 <josephcp> kind of curious if they're paying out directly from pools or not
  62 2012-02-06 00:23:31 twmz-otg has joined
  63 2012-02-06 00:23:56 <gmaxwell> FROTUSCI: oh, are you the same person as poiuh?  (people changing names always confuses me)
  64 2012-02-06 00:24:50 <gmaxwell> josephcp: good idea. I hadn't thought to look at those. I've got some people doing share logging against some of these >100% pps services now.
  65 2012-02-06 00:25:15 <gmaxwell> I've also got an idea that can be used to prevent them from being used for reorg attacks.. but I need to flesh it out some.
  66 2012-02-06 00:25:43 <josephcp> seems like you can statistically analyze it with a good degree of confidence, simply check whether it derives from an average of less than 4 transactions deep or not, no?
  67 2012-02-06 00:28:39 h4ckm3 has joined
  68 2012-02-06 00:29:32 <josephcp> protecting against block reorg attacks sounds difficult, but i suspect it's simple laundry
  69 2012-02-06 00:30:11 <gmaxwell> josephcp: actually its super simple.
  70 2012-02-06 00:31:21 <gmaxwell> josephcp: the rpc miners will remember the PREV/height in all the headers they see.. and then simply refuse to work on a new getwork that uses prev which is 'too far' back.
  71 2012-02-06 00:31:51 <josephcp> yeah that sounds plausible, it's "good enough" for nodes that run 24/7 at least (and i guess those are the ones that mater the most)
  72 2012-02-06 00:32:23 <gmaxwell> Yep. It's not perfect. for example.. once you have a one block fork going they'll be willing to work on that fork, sadly.
  73 2012-02-06 00:33:07 <gmaxwell> But yes, I _know_ these things are not currently being used for reorg attacks. The problem is that they create a ready market for them.
  74 2012-02-06 00:33:20 <josephcp> yeah, and the code has to explicitly know when to activate, but i think the problem may arise when you get a hard fork and there isn't a clear way for new clients to use a specific branch
  75 2012-02-06 00:33:45 <gmaxwell> with 150% PPS payments ... it seems pretty much inevitable that most of our hash power will be on these laundry markets within a few months.
  76 2012-02-06 00:35:17 <gmaxwell> I don't actually think there should normally ever be a case where a client will go back, at least wrt a single pool.
  77 2012-02-06 00:35:28 <FROTUSCI> the cost to do a double-spend against someone operating in spv mode vs. full mode is the same (except for a constant factor)
  78 2012-02-06 00:35:37 <gmaxwell> E.g. if the pool gives you work where PREV=A   then PREV=B .. you should _never_ see PREV=A again.
  79 2012-02-06 00:35:51 <FROTUSCI> therefore spv is secure
  80 2012-02-06 00:36:27 <gmaxwell> FROTUSCI: did you see my earlier question?
  81 2012-02-06 00:36:54 <FROTUSCI> about spv?
  82 2012-02-06 00:36:59 <FROTUSCI> i dont see it
  83 2012-02-06 00:37:16 <gmaxwell> 16:17 < gmaxwell> FROTUSCI: oh, are you the same person as poiuh?  (people changing names always confuses me)
  84 2012-02-06 00:37:32 <FROTUSCI> no im not
  85 2012-02-06 00:37:37 <gmaxwell> Ah okay.
  86 2012-02-06 00:39:11 <gmaxwell> FROTUSCI: nah, its not the same cost... For example, I could create a block paying you 100,000 BTC, and 5 extending it for a cost of 300-450 BTC.
  87 2012-02-06 00:39:21 <gmaxwell> A SPV node would believe it, a full node would not.
  88 2012-02-06 00:39:51 <gmaxwell> (er, to be clear, I don't have 100k btc)
  89 2012-02-06 00:39:58 <FROTUSCI> you would be generating 6 blocks at the current difficulty?
  90 2012-02-06 00:40:07 <gmaxwell> FROTUSCI: Yes. Sure.
  91 2012-02-06 00:40:27 <gmaxwell> The cost of that is 300 BTC, but I might have to pay a premium to get it done in acceptable time.
  92 2012-02-06 00:40:42 <FROTUSCI> which is just what you'd need to do to double-spend against full-validators (assuming you had the balance to begin with - on block 7 you overtake the chain and reverse)
  93 2012-02-06 00:40:48 twmz-otg has quit (Quit: twmz-otg)
  94 2012-02-06 00:41:40 <gmaxwell> Right, assuming. I don't have to have the balance to begin with for an SPV node which is actually a big improvement. And I also don't have the risk of the real network reorging onto my theft.
  95 2012-02-06 00:42:23 <gmaxwell> say I use my own 1000 btc to reverse/respend you. I'm successful. But now I have this 'tainted' 1000 btc that you can prove to anyone was involved with ripping you off.
  96 2012-02-06 00:42:30 h4ckm3 has quit (Quit: Leaving)
  97 2012-02-06 00:42:34 <gmaxwell> So then I have to try to somehow launder that btc.. which is costly.
  98 2012-02-06 00:42:37 <FROTUSCI> to success in both cases you need to outpace the network for 6 (or 7 blocks in the full case)
  99 2012-02-06 00:43:04 <gmaxwell> FROTUSCI: No, you don't— you can combine it with isolating a node without outpacing the network.
 100 2012-02-06 00:43:09 <FROTUSCI> yes theres more traceability in the full case, and risk your gamble wont pay off if you arent fast enough
 101 2012-02-06 00:43:21 <gmaxwell> right, I think we agree.
 102 2012-02-06 00:43:29 <FROTUSCI> isolate a node? in that case its easy. you just need to have the coins up front
 103 2012-02-06 00:43:41 <gmaxwell> The cost is the same.
 104 2012-02-06 00:44:00 <gmaxwell> (I mean it costs 300 btc in the non-outpacing case)
 105 2012-02-06 00:44:30 <gmaxwell> You can't currently buy outpacing on the open market, but if trends continue it seems likely that this will become possible.
 106 2012-02-06 00:44:35 <FROTUSCI> yeah so the computing power is the same for both attacks (but theres greater risk of being traced or losing the race against full validators)
 107 2012-02-06 00:45:41 <FROTUSCI> if you complained that someone isolated you and fed you false blocks, it probably wouldnt affect anything. afaik no one has been blacklisting addresses
 108 2012-02-06 00:46:28 <gmaxwell> FROTUSCI: Sure it would— all the exchanges take reports and turn over people exchanging stolen funds to law enforcement.
 109 2012-02-06 00:47:07 <FROTUSCI> and LE marks it as spam, going back to their cat videos
 110 2012-02-06 00:47:49 <gmaxwell> right. sorry. I forgot. *ploink*
 111 2012-02-06 00:47:56 <gmaxwell> Come back when you have patches.
 112 2012-02-06 00:48:17 <FROTUSCI> im sure the fbi has a taskforce just for investigating bitcoin crimes...
 113 2012-02-06 00:49:00 <josephcp> i'm sure the fbi cares about actual money laundering happening instead of theoretical bitcoins, i.e. visa gift cards
 114 2012-02-06 00:49:18 <FROTUSCI> afaik there has no been a single prosecution of any bitcoin user
 115 2012-02-06 00:51:00 ovidiusoft has quit (Ping timeout: 252 seconds)
 116 2012-02-06 00:51:02 <FROTUSCI> anyway its good to quantify the security tradeoff between spv and full validation. the attacker still needs the same compute power to pull it off. so you can be sure no one would burn a bunch of gpu cycles just to con you out of 10 btc
 117 2012-02-06 00:51:36 gjs278 has quit (Read error: Connection reset by peer)
 118 2012-02-06 00:51:42 gjs278 has joined
 119 2012-02-06 00:53:17 Cablesaurus has quit (Quit: Why is the alphabet in that order?   Is it because of that song?)
 120 2012-02-06 00:54:13 user_ has quit (Quit: Leaving)
 121 2012-02-06 00:54:55 <FROTUSCI> if the attacker can isolate you, then the attack is still the same.   the only difference is in the non-isolatng case where the double-spender has to risk his initial coins in the hope of reversing them later. which would only be a "risk" if they didn't have enouch cpu to outpace the network..
 122 2012-02-06 00:57:53 Cablesaurus has joined
 123 2012-02-06 00:57:54 Cablesaurus has quit (Changing host)
 124 2012-02-06 00:57:54 Cablesaurus has joined
 125 2012-02-06 00:59:45 roconnor has left ("Konversation terminated!")
 126 2012-02-06 01:00:40 graingert has quit (Remote host closed the connection)
 127 2012-02-06 01:13:11 osmosis has joined
 128 2012-02-06 01:19:51 <FROTUSCI> hmm
 129 2012-02-06 01:20:08 <FROTUSCI> how do you get the figure of 300btc to generate 6 confirmations?
 130 2012-02-06 01:20:16 <FROTUSCI> is there a calculator somehwere
 131 2012-02-06 01:20:28 Cablesaurus has quit (Quit: Make it idiot proof and someone will make a better idiot.)
 132 2012-02-06 01:21:55 eldentyrell has quit (Quit: eldentyrell)
 133 2012-02-06 01:22:30 occulta has quit (Quit: KVIrc 4.1.1 Equilibrium http://www.kvirc.net/)
 134 2012-02-06 01:24:02 copumpkin has joined
 135 2012-02-06 01:24:11 copumpkin has quit (Changing host)
 136 2012-02-06 01:24:11 copumpkin has joined
 137 2012-02-06 01:25:35 wood has quit (Ping timeout: 244 seconds)
 138 2012-02-06 01:30:12 pingdrive has quit (Quit: Leaving)
 139 2012-02-06 01:32:13 Turingi has quit (Read error: Connection reset by peer)
 140 2012-02-06 01:32:19 rdponticelli_ has joined
 141 2012-02-06 01:32:59 Cablesaurus has joined
 142 2012-02-06 01:32:59 Cablesaurus has quit (Changing host)
 143 2012-02-06 01:32:59 Cablesaurus has joined
 144 2012-02-06 01:32:59 rdponticelli has quit (Ping timeout: 240 seconds)
 145 2012-02-06 01:48:11 <Graet> 6 blocks @ 50btc ea
 146 2012-02-06 01:49:24 <FROTUSCI> right of course hehe.. i ws thinking $
 147 2012-02-06 01:49:52 user_ has joined
 148 2012-02-06 01:50:01 etotheipi_ has joined
 149 2012-02-06 01:50:13 <splatster> gtg
 150 2012-02-06 01:50:23 splatster is now known as splatster|away
 151 2012-02-06 01:50:38 <Graet> :)
 152 2012-02-06 01:50:40 <luke-jr> BlueMatt: might want to upgrade libpng in gitian desc, since 1.5.7 has a security vuln
 153 2012-02-06 01:51:59 <luke-jr> (shouldn't affect us since we just use bundled icons)
 154 2012-02-06 01:52:07 denisx has quit (Quit: denisx)
 155 2012-02-06 01:56:39 Nicksasa has quit (Ping timeout: 240 seconds)
 156 2012-02-06 01:59:52 h4ckm3 has joined
 157 2012-02-06 02:00:29 da2ce7 has joined
 158 2012-02-06 02:01:38 Nicksasa has joined
 159 2012-02-06 02:02:56 minimoose has left ()
 160 2012-02-06 02:03:17 da2ce7 has quit (Ping timeout: 276 seconds)
 161 2012-02-06 02:08:38 * luke-jr wonders why bitcoin's gitian installs wine and never uses it O.o
 162 2012-02-06 02:10:46 b4epoche_ has joined
 163 2012-02-06 02:11:27 b4epoche has quit (Ping timeout: 245 seconds)
 164 2012-02-06 02:11:27 b4epoche_ is now known as b4epoche
 165 2012-02-06 02:19:41 twmz-otg has joined
 166 2012-02-06 02:23:18 twmz-otg has quit (Client Quit)
 167 2012-02-06 02:23:42 twmz-otg has joined
 168 2012-02-06 02:27:32 p0s has quit (Remote host closed the connection)
 169 2012-02-06 02:32:57 <luke-jr> ;;bc,calc 1000 256
 170 2012-02-06 02:32:58 <gribble> Error: invalid syntax (<string>, line 1)
 171 2012-02-06 02:33:02 <luke-jr> ;;bc,calcd 1000 256
 172 2012-02-06 02:33:02 <gribble> The average time to generate a block at 1000 Khps, given the supplied difficulty of 256, is 1 week, 5 days, 17 hours, 25 minutes, and 11 seconds
 173 2012-02-06 02:36:50 hexTech has quit (Remote host closed the connection)
 174 2012-02-06 02:37:48 hexTech has joined
 175 2012-02-06 02:42:10 h4ckm3th32nd has joined
 176 2012-02-06 02:42:35 h4ckm3th32nd has quit (Client Quit)
 177 2012-02-06 02:43:07 user_ has quit (Quit: Leaving)
 178 2012-02-06 02:44:44 h4ckm3 has quit (Quit: Leaving)
 179 2012-02-06 02:45:59 hexTech has quit (Remote host closed the connection)
 180 2012-02-06 02:46:41 hexTech has joined
 181 2012-02-06 02:52:06 h4ckm3th32nd has joined
 182 2012-02-06 02:52:58 h4ckm3th32nd has quit (Read error: Connection reset by peer)
 183 2012-02-06 02:56:24 hexTech has quit (Remote host closed the connection)
 184 2012-02-06 02:57:11 hexTech has joined
 185 2012-02-06 02:59:10 bitlad has joined
 186 2012-02-06 03:01:21 BTC_Bear is now known as hbrntng!~BTC_Bear@unaffiliated/btc-bear/x-5233302|BTC_Bear
 187 2012-02-06 03:03:24 hexTech has quit (Remote host closed the connection)
 188 2012-02-06 03:04:12 hexTech has joined
 189 2012-02-06 03:04:23 bitlad has quit (Remote host closed the connection)
 190 2012-02-06 03:06:13 gfinn has quit (Remote host closed the connection)
 191 2012-02-06 03:13:24 hexTech has quit (Remote host closed the connection)
 192 2012-02-06 03:14:11 hexTech has joined
 193 2012-02-06 03:14:25 pingdrive has joined
 194 2012-02-06 03:15:03 BTC_Bear has quit (Remote host closed the connection)
 195 2012-02-06 03:16:54 m00p has quit (Ping timeout: 260 seconds)
 196 2012-02-06 03:20:55 gfinn has joined
 197 2012-02-06 03:21:12 twmz-otg has quit (Quit: twmz-otg)
 198 2012-02-06 03:22:24 TheSeven has quit (Disconnected by services)
 199 2012-02-06 03:22:36 [7] has joined
 200 2012-02-06 03:23:10 Xunie has quit (Quit: Can God microwave a taco so hot that not even *HE* can eat it without burns?)
 201 2012-02-06 03:23:56 hexTech has quit (Remote host closed the connection)
 202 2012-02-06 03:24:42 hexTech has joined
 203 2012-02-06 03:25:19 BTC_Bear has joined
 204 2012-02-06 03:25:59 splatster is now known as away!~splatster@unaffiliated/splatster|splatster
 205 2012-02-06 03:29:24 hexTech has quit (Remote host closed the connection)
 206 2012-02-06 03:30:18 hexTech has joined
 207 2012-02-06 03:33:39 bitlad has joined
 208 2012-02-06 03:33:50 m00p has joined
 209 2012-02-06 03:35:46 bitlad has quit (Remote host closed the connection)
 210 2012-02-06 03:37:25 JRWR has quit (Remote host closed the connection)
 211 2012-02-06 03:41:53 <BlueMatt> luke-jr: oh, wow didnt see that, thanks
 212 2012-02-06 03:42:30 <luke-jr> BlueMatt: does it actually need wine for something btw?
 213 2012-02-06 03:43:09 <luke-jr> (also, np)
 214 2012-02-06 03:43:18 <BlueMatt> luke-jr: the win32 build, yea it needs to run one minor thing iirc
 215 2012-02-06 03:43:27 <BlueMatt> like miniupnpc build utility or smth
 216 2012-02-06 03:43:51 <luke-jr> bleh
 217 2012-02-06 03:49:27 <gribble> New news from bitcoinrss: TheBlueMatt opened issue 802 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/802>
 218 2012-02-06 03:54:05 <jamescarr> when you issue a getbalance request to bitcoind with an account name, does it only look at accounts in the wallet?
 219 2012-02-06 03:56:17 <josephcp> yes but getbalance is mostly just cosmetic (i.e. when you move balances it doesn't actually show up on the blockchain), you're still spending from a single pool of keys
 220 2012-02-06 03:56:27 CornedBeefHash has joined
 221 2012-02-06 03:58:16 <jamescarr> I see
 222 2012-02-06 03:58:43 <jamescarr> yeah, if I request an address not on the server, it just gives me the server balance
 223 2012-02-06 03:59:36 <josephcp> you can't get balances of keys using stock bitcoind if you don't have the private key in your wallet
 224 2012-02-06 03:59:54 <josephcp> err balances of addresses
 225 2012-02-06 04:01:57 hexTech has quit (Remote host closed the connection)
 226 2012-02-06 04:01:57 <luke-jr> ugh
 227 2012-02-06 04:02:08 <luke-jr> too easy to click the little x on qemu by accident :/
 228 2012-02-06 04:02:15 <luke-jr> gotta do gitian all over again
 229 2012-02-06 04:02:50 Cablesaurus has quit (Quit: Pull the pin and count to what?)
 230 2012-02-06 04:02:52 hexTech has joined
 231 2012-02-06 04:03:06 josephcp has quit (Read error: Connection reset by peer)
 232 2012-02-06 04:04:24 <luke-jr> BlueMatt: would it be difficult to split the other libraries into different gitian stages?
 233 2012-02-06 04:05:18 <BlueMatt> luke-jr: if you have the free-time, not really...
 234 2012-02-06 04:05:36 <BlueMatt> you just have to pretty carefully check for determinism which can take quite a while
 235 2012-02-06 04:06:22 gronager1 has joined
 236 2012-02-06 04:06:23 gronager has quit (Read error: Connection reset by peer)
 237 2012-02-06 04:06:25 <luke-jr> think it'd be better to do a deps-win32, or split each to its own?
 238 2012-02-06 04:07:12 <gmaxwell> so— Trivia time.
 239 2012-02-06 04:07:29 hexTech has quit (Remote host closed the connection)
 240 2012-02-06 04:07:32 <gmaxwell> Out of the last 192636 transactions, how many involve 1VayNert3x1KzbpzMGt2qdqrAThiRovi8?
 241 2012-02-06 04:07:56 <luke-jr> over 50%?
 242 2012-02-06 04:08:07 josephcp has joined
 243 2012-02-06 04:08:10 bitlad has joined
 244 2012-02-06 04:08:22 <gmaxwell> Not quite. 27%
 245 2012-02-06 04:08:44 hexTech has joined
 246 2012-02-06 04:09:41 gronager has joined
 247 2012-02-06 04:09:42 gronager1 has quit (Read error: Connection reset by peer)
 248 2012-02-06 04:09:53 bitlad has quit (Remote host closed the connection)
 249 2012-02-06 04:11:43 <gmaxwell> I'm dumping data for transaction graph analysis, and I for a bit I thought it was broken because it looked like the same data over and over again.
 250 2012-02-06 04:11:54 <gmaxwell> s/ I //
 251 2012-02-06 04:12:27 hexTech has quit (Remote host closed the connection)
 252 2012-02-06 04:12:28 <luke-jr> lol
 253 2012-02-06 04:13:17 hexTech has joined
 254 2012-02-06 04:13:25 <gmaxwell> I'm not sure what to do about it— we've asked him over and over again to use send many (and to consider paying out a bit less frequently)
 255 2012-02-06 04:13:39 <gmaxwell> If he were a smaller pool I'd circulate a patch to discourage blocks with these transactions.
 256 2012-02-06 04:14:30 bitlad has joined
 257 2012-02-06 04:14:31 <luke-jr> discouraging blocks can be dangerous
 258 2012-02-06 04:14:49 <luke-jr> Eligius already rejects them because they lack a reasonable fee, I think
 259 2012-02-06 04:15:20 <gmaxwell> Yes, they're zero fee. They started pissing me off because I saw times when they were the entire free area in my own blocks.
 260 2012-02-06 04:15:37 <gmaxwell> I don't relay them at all now.
 261 2012-02-06 04:15:58 <BlueMatt> easy to ban relay of the txes in the next default client, but it wont help
 262 2012-02-06 04:16:32 <gmaxwell> I don't understand why he's doing this.
 263 2012-02-06 04:16:51 <gmaxwell> He could easly cut the data volume to 1/4th by doing sendmany
 264 2012-02-06 04:17:04 bitlad_ has joined
 265 2012-02-06 04:17:24 <copumpkin> gmaxwell: who is it?
 266 2012-02-06 04:17:43 <gmaxwell> copumpkin: tycho / deepbit
 267 2012-02-06 04:18:13 CornedBeefHash has quit (Ping timeout: 255 seconds)
 268 2012-02-06 04:18:16 <copumpkin> oh
 269 2012-02-06 04:18:19 <gmaxwell> copumpkin: he does automatic payouts for his users with as little at 0.01 BTC and doesn't use sendmany.
 270 2012-02-06 04:18:25 <copumpkin> ugh
 271 2012-02-06 04:18:30 <copumpkin> how about https://blockchain.info/address/5ef8c8e6fe5364ba7e9d027119db36a483ce741a
 272 2012-02-06 04:18:40 bitlad has quit (Ping timeout: 240 seconds)
 273 2012-02-06 04:19:02 <gmaxwell> Okay thats stupid too. :)
 274 2012-02-06 04:19:10 <josephcp> i think that might be the faucet?
 275 2012-02-06 04:19:24 <copumpkin> josephcp: could be
 276 2012-02-06 04:19:32 <copumpkin> the addresses don't look random though
 277 2012-02-06 04:19:32 <gmaxwell> Is the faucet 0.001 now? Okay.
 278 2012-02-06 04:19:34 <josephcp> oh wait faucet pays 0.005
 279 2012-02-06 04:19:36 <copumpkin> I guess he could sort them
 280 2012-02-06 04:19:37 <josephcp> i thought it was 0.001
 281 2012-02-06 04:19:43 <gmaxwell> copumpkin: the faucet batches, and they're sorted.
 282 2012-02-06 04:19:53 <josephcp> i could've sworn it was 0.001
 283 2012-02-06 04:20:01 bitlad_ has quit (Remote host closed the connection)
 284 2012-02-06 04:20:10 <copumpkin> when I got my faucet back in june it was 0.001
 285 2012-02-06 04:20:16 <gmaxwell> It used to be 0.001 right.
 286 2012-02-06 04:20:25 <gmaxwell> (actually it _used_ to be 5 heh)
 287 2012-02-06 04:20:47 <BlueMatt> still, deepbit being 27% of the chain volume, that is just ridiculous
 288 2012-02-06 04:20:51 <cjd> heh don't store blocks with that in it
 289 2012-02-06 04:21:09 <cjd> store header, demand block from anyone who wants to spend out of it
 290 2012-02-06 04:21:18 <gmaxwell> BlueMatt: in his very slight defense, he did a bunch of work to do pay-to-pubkey there which somewhat reduced the data volume.
 291 2012-02-06 04:21:21 pirateat40 has joined
 292 2012-02-06 04:21:23 Cablesaurus has joined
 293 2012-02-06 04:21:23 Cablesaurus has quit (Changing host)
 294 2012-02-06 04:21:23 Cablesaurus has joined
 295 2012-02-06 04:23:14 <gmaxwell> if anyone wants to do any interesting transaction graph analysis, I'm currently dumping the complete history of bitcoin, with one line per transaction that looks like this:
 296 2012-02-06 04:23:18 <gmaxwell> TX: [[16367a8594108d4b9e381992621745fdc7c44ff2ed8318e37abaf93455d4051d:0/1VayNert3x1KzbpzMGt2qdqrAThiRovi8,2a8f
 297 2012-02-06 04:23:21 <gmaxwell> 81f4e34ad0dba9f2b46a13bc6f08f2647f0fcf939fee4ccecf939fdf6a1b:0/1VayNert3x1KzbpzMGt2qdqrAThiRovi8],[0.012210/1Va
 298 2012-02-06 04:23:24 <gmaxwell> yNert3x1KzbpzMGt2qdqrAThiRovi8,0.060000/1F7Lyeq8bp58iZUHmK6EeRsrQjiNEzr2xD]]
 299 2012-02-06 04:23:25 <BlueMatt> gmaxwell: let me solve two of his problems in one solution: a. his security is clearly broken: he has a live bitcoin node which holds large sums of bitcoin and is rpc-accessible to an internet-facing server doing automatic payouts, and 2. his sends are broken.  easy solution: sneakernet bulk payments every ~day or serial bulk tx sends instead of using rpc
 300 2012-02-06 04:24:19 <copumpkin> gmaxwell: apparently that thing isn't the faucet
 301 2012-02-06 04:24:26 <BlueMatt> (serial or any other overly restrictive method to pass data from web/miner frontend to high-bitcoin-holding backend node
 302 2012-02-06 04:24:36 <copumpkin> pirateat40 was asking about it in another channel, and apparently he's never used the faucet but just received two payments from that thing
 303 2012-02-06 04:24:42 <gmaxwell> (the format is TX: [[input/addr,input/addr,input/addr],[amountout/addr,amountout/addr...]])
 304 2012-02-06 04:24:43 <pirateat40> How would they have one of my addresses?
 305 2012-02-06 04:25:02 <copumpkin> pirateat40: any idea where you used that address?
 306 2012-02-06 04:25:03 <gmaxwell> pirateat40: pulling it from the blockchain.
 307 2012-02-06 04:25:13 <gmaxwell> or who knows what.
 308 2012-02-06 04:25:25 <pirateat40> I just created that address like 10min ago
 309 2012-02-06 04:25:31 <copumpkin> lol
 310 2012-02-06 04:25:32 <pirateat40> no transactions
 311 2012-02-06 04:25:34 <pirateat40> at all
 312 2012-02-06 04:25:37 <pirateat40> then .001
 313 2012-02-06 04:25:40 <gmaxwell> pirateat40: Law enforcement is onto you then buddy boy.
 314 2012-02-06 04:25:42 <BlueMatt> gmaxwell: also, post the fact that [Tycho] is eating ~27% of the chain space being added to the mailing list/forums
 315 2012-02-06 04:25:52 RobinPKR_ has joined
 316 2012-02-06 04:25:54 <copumpkin> gmaxwell: but seriously, that does seem odd :P
 317 2012-02-06 04:25:59 <BlueMatt> because that is very, very seriously poor
 318 2012-02-06 04:26:10 <josephcp> shut-down-everything.jpg
 319 2012-02-06 04:26:20 BlueMatt has quit (Quit: Ex-Chat)
 320 2012-02-06 04:26:21 <copumpkin> doesn't it?
 321 2012-02-06 04:26:26 <pirateat40> The first one was received 54min ago.
 322 2012-02-06 04:26:31 <pirateat40> not confirmed yet
 323 2012-02-06 04:26:34 <josephcp> sounds like a heavy dose of paranoia of moving everything into offline wallets for the time being
 324 2012-02-06 04:27:07 <josephcp> wait so you didn't use the address online at all ever?
 325 2012-02-06 04:27:15 <pirateat40> EVER
 326 2012-02-06 04:27:19 <luke-jr> pirateat40: btw, I heard some GPUMAX users were receiving confirmed laundry money? you know anything about that>
 327 2012-02-06 04:27:33 <pirateat40> lol
 328 2012-02-06 04:27:37 <gmaxwell> pirateat40: which address was yours? (I want to sanity check your never used claim)
 329 2012-02-06 04:27:42 <pirateat40> luke-jr, you would say something like that.
 330 2012-02-06 04:27:51 <copumpkin> I used some laundry money earlier today
 331 2012-02-06 04:27:54 <copumpkin> to do some laundry
 332 2012-02-06 04:27:59 <luke-jr> pirateat40: I would? only because I heard that for real
 333 2012-02-06 04:28:02 <pirateat40> 16qNabsf6yL41iuv5wVPzbXfp9xdAgVBBC
 334 2012-02-06 04:28:35 <pirateat40> Yea I'm laundering CosbyCoins.
 335 2012-02-06 04:28:42 <gmaxwell> pirateat40: welp, if you've really just generated that and never given it out— then it means your system security is compromised. Today is your lucky day, I guess.
 336 2012-02-06 04:28:43 RobinPKR has quit (Ping timeout: 272 seconds)
 337 2012-02-06 04:28:43 RobinPKR_ is now known as RobinPKR
 338 2012-02-06 04:29:09 <pirateat40> Good, I hope I get more!!!
 339 2012-02-06 04:29:20 <pirateat40> I could be rich in years.
 340 2012-02-06 04:29:48 <luke-jr> pirateat40: what confuses me is, I assumed you took payment from your customers, then paid the miners directly… so miners shouldn't be receiving money from your customers in any case, right?
 341 2012-02-06 04:30:20 <pirateat40> What are you talking about?
 342 2012-02-06 04:30:39 <pirateat40> People buy, the go into a wallet and then miners are paid from it.
 343 2012-02-06 04:30:49 <copumpkin> pirateat40: I'd be freaked out
 344 2012-02-06 04:30:49 <pirateat40> Why is that hard for you to understand?
 345 2012-02-06 04:31:14 <josephcp> yeah i'd be moving everything onto a bitcoind instance on a laptop instead of chatting in IRC
 346 2012-02-06 04:31:15 <pirateat40> I know its not a cash register with dollar bills but it works the same way.
 347 2012-02-06 04:31:26 Karmaon has quit (Quit: WeeChat 0.3.5)
 348 2012-02-06 04:31:46  has joined
 349 2012-02-06 04:31:53 <copumpkin> pirateat40: I'd move all your coins over to me (or to someone trusted :P)
 350 2012-02-06 04:32:15 <copumpkin> at least until you figure out what's going on
 351 2012-02-06 04:32:19 <gmaxwell> I suggest allinvain.
 352 2012-02-06 04:32:25 <pirateat40> lol
 353 2012-02-06 04:32:32 <copumpkin> pirateat40: but seriously
 354 2012-02-06 04:34:25 <gmaxwell> pirateat40: Is it true that some of your customers are paying 140-150% in bitcoins to purhase hashpower?
 355 2012-02-06 04:34:45 bitlad has joined
 356 2012-02-06 04:35:04 <pirateat40> Yes, to begin with... rates have some down a little but its still pretty high.
 357 2012-02-06 04:35:11 bitlad has quit (Remote host closed the connection)
 358 2012-02-06 04:35:26 <pirateat40> come*
 359 2012-02-06 04:36:01 <pirateat40> As of now http://screenshotuploader.com/s/01/Db0gJzc87
 360 2012-02-06 04:36:10 pingdrive has quit (Quit: Leaving)
 361 2012-02-06 04:37:37 <copumpkin> pirateat40: how are you generating the addresses? which client?
 362 2012-02-06 04:38:00 <pirateat40> That address was a correction address for my scripts.
 363 2012-02-06 04:38:04 <josephcp> check this address out, it was also part of the massive pending transaction to 16qNab... https://blockexplorer.com/address/19R4SNQAXRFP8Dax6Mo8CcwZn38uuJ1Ccz
 364 2012-02-06 04:38:08 <josephcp> looks like it's mining on eligius
 365 2012-02-06 04:38:09 <pirateat40> like we did before copumpkin
 366 2012-02-06 04:38:15 BlueMatt has joined
 367 2012-02-06 04:38:40 <gmaxwell> Correction address?
 368 2012-02-06 04:38:52 <copumpkin> pirateat40: hmm, well if it wasn't advertised and someone sent to it, either they were _very_ lucky, or someone can look at your addresses, which is disturbing. Especially if it happened twice
 369 2012-02-06 04:39:17 <pirateat40> gmaxwell, Yea just behind the scenes stuff for my system.
 370 2012-02-06 04:39:30 <gmaxwell> copumpkin: don't even bother saying very lucky. You'll just encourage people to think poorly about it.
 371 2012-02-06 04:39:41 <pirateat40> Yea ill monitor it, but just found it odd.
 372 2012-02-06 04:41:14 hexTech has quit (Ping timeout: 276 seconds)
 373 2012-02-06 04:42:02 pumpkin has joined
 374 2012-02-06 04:42:05 <pumpkin> pirateat40: is this the wallet with all the coins in it? :P
 375 2012-02-06 04:42:14 <pirateat40> lol
 376 2012-02-06 04:42:17 <pirateat40> no
 377 2012-02-06 04:42:23 hexTech has joined
 378 2012-02-06 04:42:29 <pumpkin> okay :)
 379 2012-02-06 04:43:04 spawn- has joined
 380 2012-02-06 04:43:24 <pumpkin> spawn-: nice host...
 381 2012-02-06 04:43:25 copumpkin has quit (Ping timeout: 255 seconds)
 382 2012-02-06 04:43:25 pumpkin is now known as copumpkin
 383 2012-02-06 04:43:52 <spawn-> thanks bro
 384 2012-02-06 04:43:56 <BlueMatt> ...
 385 2012-02-06 04:43:57 <spawn-> just a BNC :)
 386 2012-02-06 04:44:00 <spawn-> and im offering it :)
 387 2012-02-06 04:44:11 JRWR has joined
 388 2012-02-06 04:44:13 <BTC_Bear> luke-jr: That is a little unfair, saying pirateat40 has received laundered money. As MTGOX and TH have also received laundered money. So what is you implications there.
 389 2012-02-06 04:44:24 <BTC_Bear> are*
 390 2012-02-06 04:44:37 <luke-jr> BTC_Bear: I asked a question. Nothing more.
 391 2012-02-06 04:45:07 <pirateat40> Jumped to conclusions like always.... without knowing anything about my system.
 392 2012-02-06 04:45:19 <luke-jr> nonsense
 393 2012-02-06 04:45:32 <pirateat40> I know
 394 2012-02-06 04:46:14 <luke-jr> asking about your system I know nothing about is the furthest thing from giving any kind of conclusion
 395 2012-02-06 04:46:48 <copumpkin> pirateat40: have you stopped beating your wife?
 396 2012-02-06 04:47:24 <pirateat40> no, i just can't stop
 397 2012-02-06 04:47:34 <copumpkin> :)
 398 2012-02-06 04:49:52 <gmaxwell> BTC_Bear: wait. Are you really going to stand up in public and say that pirateat40 isn't running a laundering service?
 399 2012-02-06 04:50:23 <gmaxwell> BTC_Bear: because at 150% payout buying bitcoins for bitcoins, all anyone could reasonable conclude is that it's being used to wash coins or pay for attacks.
 400 2012-02-06 04:50:24 <pirateat40> gmaxwell, even if I was.  What is your point?
 401 2012-02-06 04:50:43 <gmaxwell> (And I think we're pretty sure it's not being used to pay for attacks)
 402 2012-02-06 04:51:02 <copumpkin> I don't like smelly coins
 403 2012-02-06 04:51:05 <copumpkin> I wash them twice a day
 404 2012-02-06 04:51:24 <pirateat40> I am still shocked that YOU don't understand what GPUMAX is for.
 405 2012-02-06 04:51:26 hexTech has quit (Remote host closed the connection)
 406 2012-02-06 04:51:33 <BTC_Bear> gmaxwell: I can see many scenarios in which his business model can pay out what it does. And given a little time, I think many others will also but letting the cat out of the bag, sort of speak, would hurt his bottom line currently.
 407 2012-02-06 04:51:38 <gmaxwell> pirateat40: What point should I have. People are mining for you and getting "dirty" coins, they should know about it. I assume with the kind of returns they get they _do_ know.
 408 2012-02-06 04:51:51 <pirateat40> LOL
 409 2012-02-06 04:51:56 <pirateat40> I AM NOT A POOL!
 410 2012-02-06 04:52:04 <pirateat40> I don't mine with it.
 411 2012-02-06 04:52:15 <gmaxwell> via you, come on. I know what you do.
 412 2012-02-06 04:52:27 hexTech has joined
 413 2012-02-06 04:52:51 <copumpkin> o.O
 414 2012-02-06 04:52:56 <cjd> 23:59 < pirateat40> I am still shocked that YOU don't understand what GPUMAX is for. <-- unloading the remaining MBC loot?
 415 2012-02-06 04:53:09 <copumpkin> wtf
 416 2012-02-06 04:53:12 <cjd> err was that solidcoin?
 417 2012-02-06 04:53:14 <luke-jr> gmaxwell: isn't it for enforcing BIP 17?
 418 2012-02-06 04:53:20 <gmaxwell> BTC_Bear: if there is an above board application, it's evading a fair number of smart people.
 419 2012-02-06 04:53:20 <BTC_Bear> lol
 420 2012-02-06 04:53:25 * cjd confuses easily
 421 2012-02-06 04:53:53 <luke-jr> at least, that's one example of legitimate use
 422 2012-02-06 04:54:13 <gmaxwell> pirateat40: people mine via your service. For every 1 BTC of value they mine in, you're paying out 1.40 BTC in very old coins.  ::shrugs::
 423 2012-02-06 04:54:21 <gmaxwell> I guess I'm just not rocket scientist enough to figure that out.
 424 2012-02-06 04:54:30 <pirateat40> No, i don't pay the people that pay me.
 425 2012-02-06 04:54:34 <pirateat40> I pay the miners.
 426 2012-02-06 04:54:35 <BTC_Bear> smart is a matter of perspective, look at the world of finance, there is one scenario there that would apply, especially given a little time to consolidate.
 427 2012-02-06 04:54:39 <gmaxwell> Maybe I'm supposted to believe that the idea is to fee-snipe the 0.02/block in transaction fees.
 428 2012-02-06 04:55:09 <gmaxwell> Or that it's to get coins on altchains, all of which add up to about .05 BTC/block worth of additional income.
 429 2012-02-06 04:55:38 <luke-jr> gmaxwell: rg used to pay above PPS as a form of gambling
 430 2012-02-06 04:55:52 <luke-jr> ie, he'd bet on getting lucky
 431 2012-02-06 04:56:04 <copumpkin> reeses has been gambling that way recently
 432 2012-02-06 04:56:16 <pirateat40> Its funny cause there are a lot of our miners that ask the same questions you do, then start purchases shares and cant STOP.  Its like an addiction.
 433 2012-02-06 04:56:23 <gmaxwell> luke-jr: doesn't work so well when the system knocks you onto standard pools.
 434 2012-02-06 04:56:34 <luke-jr> gmaxwell: considering Eligius's 1000 BTC buffers, such gambling could pay pretty well
 435 2012-02-06 04:56:50 bitlad has joined
 436 2012-02-06 04:57:31 <pirateat40> Plan and simple, GPUMAX is a gambling site.
 437 2012-02-06 04:58:18 <gmaxwell> pirateat40: er, I thought you confined the share purchasing to approved pools?
 438 2012-02-06 04:58:21 mizerydearia has quit (Ping timeout: 252 seconds)
 439 2012-02-06 04:58:33 <gmaxwell> meaning no high variance option.
 440 2012-02-06 04:59:12 <pirateat40> gmaxwell, no we have a system now in place that prevents users for using the purchased power on pools that either are not working on bitcoin or cant handle the speeds.
 441 2012-02-06 04:59:30 <pirateat40> They get one warning, then we suspend the account.
 442 2012-02-06 04:59:48 bitlad has quit (Remote host closed the connection)
 443 2012-02-06 04:59:57 <pirateat40> Regardless of them trying they can only use it for maybe 10 shares.
 444 2012-02-06 05:00:14 <gmaxwell> What do you mean by not working on bitcoin?
 445 2012-02-06 05:00:48 <luke-jr> pirateat40: btw, how's Eligius working out now that we've changed pool servers?
 446 2012-02-06 05:01:40 <pirateat40> If the pool or "what looks like a pool" sends work that does not match 95% of the rest of our network over a block change its automatically suspended.
 447 2012-02-06 05:01:50 bitlad has joined
 448 2012-02-06 05:02:17 <pirateat40> luke-jr, we have not tried but I can give it go if you want.
 449 2012-02-06 05:02:27 hexTech has quit (Remote host closed the connection)
 450 2012-02-06 05:02:37 <gmaxwell> pirateat40: cool.  I was planning on adding code to RPC miners to check that the PREV hash doesn't move backwards. This would make it much harder to use rpc miners to back cut a chain.
 451 2012-02-06 05:02:46 <gmaxwell> pirateat40: sounds like you're doing something kinda similar internally.
 452 2012-02-06 05:02:50 <luke-jr> pirateat40: would be interesting to see how it compares, if you don't mind…
 453 2012-02-06 05:03:00 bitlad has quit (Remote host closed the connection)
 454 2012-02-06 05:03:02 <pirateat40> The failover (private mining) was never the issues with eligius it was purchasing.  Which we don't get many for. :(
 455 2012-02-06 05:03:22 hexTech has joined
 456 2012-02-06 05:04:10 <pirateat40> gmaxwell, we are doing everything possible to make sure our system can not be used to harm bitcoin.  We have already caught some start up pools trying to game the system and have suspended their accounts.
 457 2012-02-06 05:04:40 <pirateat40> luke-jr, but im willing to stress test your system if you want.
 458 2012-02-06 05:04:49 <gmaxwell> I am _very_ interested in knowing what they were trying to do— since thats problem making that would be interesting to detect more generally.
 459 2012-02-06 05:05:12 <luke-jr> pirateat40: makes sense if you've mostly got gamblers - Eligius isn't really gamble-friendly
 460 2012-02-06 05:05:43 <pirateat40> Yea
 461 2012-02-06 05:06:03 <pirateat40> Without even trying we have become the police of shady pool ops. :)
 462 2012-02-06 05:06:17 <luke-jr> ?
 463 2012-02-06 05:06:25 <luke-jr> what do the shady pool ops do?
 464 2012-02-06 05:06:37 <pirateat40> Use there pool to attack other coins!
 465 2012-02-06 05:06:40 bitlad has joined
 466 2012-02-06 05:06:41 <pirateat40> their*
 467 2012-02-06 05:07:00 <luke-jr> didn't realize there were any of those
 468 2012-02-06 05:07:15 <pirateat40> Nor did I ;)
 469 2012-02-06 05:07:27 <FROTUSCI> why not just run your own pool
 470 2012-02-06 05:07:58 <pirateat40> That's not our model and we are not here to step on pool ops toes.
 471 2012-02-06 05:08:22 <pirateat40> I said from the start that we would not be a pool only a service.
 472 2012-02-06 05:08:25 <luke-jr> pirateat40: in any case, when/if you run the test again, lemme know how it does
 473 2012-02-06 05:08:34 <pirateat40> Will do
 474 2012-02-06 05:11:50 <gmaxwell> pirateat40: If I run a "pool" that simply proxies other pools but withholds solutions to put them out of business, you're okay with that, right?
 475 2012-02-06 05:12:24 sytse has quit (Ping timeout: 252 seconds)
 476 2012-02-06 05:12:25 <pirateat40> You have to live with that, no me.
 477 2012-02-06 05:12:29 sytse has joined
 478 2012-02-06 05:12:38 <pirateat40> not*
 479 2012-02-06 05:13:14 <gmaxwell> Right. And if I happened to have 20,000 in stolen bitcoins that I used to buy shares on your service to mine on deebit to get clean coins. .. I'd have to live with that too? Right?
 480 2012-02-06 05:13:17 <FROTUSCI> pirateat40 is cool
 481 2012-02-06 05:13:48 <copumpkin> gmaxwell: what distinguishes a stolen bitcoin from a non-stolen one?
 482 2012-02-06 05:13:51 <FROTUSCI> i did some gpumax mining
 483 2012-02-06 05:14:01 <TuxBlackEdo> copumpkin, it's history?
 484 2012-02-06 05:14:11 <gmaxwell> pirateat40: You also won't mind me doing the analysis and publishing evidence of your customers doing that, right? Perhaps you'd like to help out? I only know a few addresses you've been paying to.
 485 2012-02-06 05:14:31 <pirateat40> Sure
 486 2012-02-06 05:14:51 <pirateat40> Stolen coins are still bitcoins.
 487 2012-02-06 05:14:56 <gmaxwell> Cool. I'm building transaction graphs right now. I can't believe no one has published tools for this already.
 488 2012-02-06 05:15:13 bitlad has quit (Remote host closed the connection)
 489 2012-02-06 05:15:40 ThomasV has joined
 490 2012-02-06 05:20:38 d4de has quit (Ping timeout: 245 seconds)
 491 2012-02-06 05:20:57 hexTech has quit (Remote host closed the connection)
 492 2012-02-06 05:22:17 <copumpkin> gmaxwell: them "money launderers" are getting frustrated now :P
 493 2012-02-06 05:23:15 <gmaxwell> I don't know much about gambling, are there many games where the house advantage is 1.4x?
 494 2012-02-06 05:23:45 <pirateat40> lol, you still don't get it?
 495 2012-02-06 05:23:50 hexTech has joined
 496 2012-02-06 05:24:43 <gmaxwell> No. I'm sorry. I'm not a very smart guy, and I often need stuff spelled out for me. I find the cloak and dagger stuff kinda confusing. :-/
 497 2012-02-06 05:25:06 <TuxBlackEdo> lol
 498 2012-02-06 05:25:24 <gmaxwell> Also, I'm pretty sure the butler did it.
 499 2012-02-06 05:25:45 <copumpkin> in the powder room, with the candlestick
 500 2012-02-06 05:26:03 <copumpkin> don't ask me how I know
 501 2012-02-06 05:26:18 <splatster> How do you know?
 502 2012-02-06 05:26:19 <pirateat40> gmaxwell, tell me what exactly you think I'm doing with GPUMAX.
 503 2012-02-06 05:26:19 <gmaxwell> copumpkin: graph analysis and zero knoweldge proofs.
 504 2012-02-06 05:27:17 d4de has joined
 505 2012-02-06 05:29:07 Staatsfeind has joined
 506 2012-02-06 05:29:15 <gmaxwell> pirateat40: Well, I know the facts: You're acting as a middleman where people are paid 1.4 for something with an expected value of 1. There is basically no time-arbitrage involved— they're paid without a long delay.
 507 2012-02-06 05:29:36 <pirateat40> ok
 508 2012-02-06 05:29:38 d4de has quit (Read error: Connection reset by peer)
 509 2012-02-06 05:30:06 d4de has joined
 510 2012-02-06 05:30:13 <pirateat40> So let make a bet.  If you can pick which pool will produce the next block ill pay you 10 BTC.
 511 2012-02-06 05:30:16 <copumpkin> gmaxwell: you're saying the miners are paid 40% above, up front?
 512 2012-02-06 05:30:25 <gmaxwell> copumpkin: in PPS, yes.
 513 2012-02-06 05:30:32 <BTC_Bear> SLUSH, deal
 514 2012-02-06 05:30:33 <copumpkin> gmaxwell: time value of money! :D
 515 2012-02-06 05:30:37 <pirateat40> lol
 516 2012-02-06 05:30:49 <gmaxwell> copumpkin: they're paid quickly, not e.g. weeks later.
 517 2012-02-06 05:30:55 <pirateat40> gmaxwell, did you know there are prop pools still?
 518 2012-02-06 05:31:02 <gmaxwell> pirateat40: what.. no way?
 519 2012-02-06 05:31:13 <pirateat40> yea.... the big ones
 520 2012-02-06 05:31:15 <gmaxwell> okay, so fine but that still makes no sense.
 521 2012-02-06 05:31:23 <pirateat40> ok, so like i said
 522 2012-02-06 05:31:28 <pirateat40> lets do the bet
 523 2012-02-06 05:31:48 <pirateat40> But instead I give you 200GH to throw at any pool you want.
 524 2012-02-06 05:32:04 <pirateat40> BOOM, slush hits 8 block in 30min
 525 2012-02-06 05:32:11 <pirateat40> Do you know how much BTC you get from that?
 526 2012-02-06 05:32:17 <pirateat40> Way more than you paid
 527 2012-02-06 05:32:21 <pirateat40> Im talking WAY more
 528 2012-02-06 05:33:23 <josephcp> isn't it a lot simpler just to offer hashing power to give directly to miners? i'm interested in mining in your pool if you can give good hopping results ^^
 529 2012-02-06 05:33:34 <midnightmagic> ...
 530 2012-02-06 05:33:44 <pirateat40> We are not a pool.
 531 2012-02-06 05:34:21 <FROTUSCI> so which pool is the luckiest
 532 2012-02-06 05:34:26 <pirateat40> lol
 533 2012-02-06 05:34:42 <FROTUSCI> always bet on blac
 534 2012-02-06 05:35:00 <pirateat40> we actually keep stats but don't release them. :(
 535 2012-02-06 05:36:08 <josephcp> yes but you're saying better results are from pool hopping, can't you just run the same service as a pool hopping proxy and offer the returns to miners (less a 20% commission of above-expected results)
 536 2012-02-06 05:36:44 <josephcp> i might be confused though i haven't looked into your service too closely :-P
 537 2012-02-06 05:36:47 <pirateat40> That's not our model.
 538 2012-02-06 05:36:52 <josephcp> oh ok never mind then
 539 2012-02-06 05:37:00 <pirateat40> People that purchase power are making the bets.
 540 2012-02-06 05:37:11 <pirateat40> The miners just reap the rewards.
 541 2012-02-06 05:38:04 onelineproof has quit (Quit: Leaving.)
 542 2012-02-06 05:38:15 <josephcp> just saying my way would be easier if you're just running a meta-pool trying to predict how to mine more coins
 543 2012-02-06 05:38:35 <josephcp> but again i could be misunderstanding haha
 544 2012-02-06 05:38:38 <pirateat40> We don't care, we make the same money regardless the way we are running now.
 545 2012-02-06 05:38:39 <Staatsfeind> Yeah, and running a restaurant where only the chef eats would be easier, too.
 546 2012-02-06 05:39:01 hexTech has quit (Remote host closed the connection)
 547 2012-02-06 05:39:41 hexTech has joined
 548 2012-02-06 05:39:42 <pirateat40> If we bet we risk losing.
 549 2012-02-06 05:40:06 <Staatsfeind> Information is not evenly distributed and if you thought it could be so easily centralized, you wouldn't be a capitalist.
 550 2012-02-06 05:40:54 <copumpkin> pirateat40: so what kind of shady stuff are you doing, anyway?! gmaxwell wants all the dirt
 551 2012-02-06 05:41:08 <copumpkin> gambling clearly isn't shady enough
 552 2012-02-06 05:41:11 <pirateat40> Oh just attacking bitcoin again.
 553 2012-02-06 05:41:19 <pirateat40> Got bills to pay.
 554 2012-02-06 05:41:31 <Staatsfeind> Right, stop that, please.
 555 2012-02-06 05:41:37 <FROTUSCI> droptop
 556 2012-02-06 05:41:38 * copumpkin sends his dirty laundry to pirateat40
 557 2012-02-06 05:41:38 <BTC_Bear> How much does a Gallon of gas cost?
 558 2012-02-06 05:41:45 <copumpkin> BTC_Bear: $9.50
 559 2012-02-06 05:41:46 <FROTUSCI> no pool has an advantage over another
 560 2012-02-06 05:41:50 <Staatsfeind> lol
 561 2012-02-06 05:41:59 <gmaxwell> copumpkin: well, to be clear— whats being described isn't just gambling. It's pool hopping and it rips off the non-hopping participants. My sympathy is pretty limited however.
 562 2012-02-06 05:42:01 <FROTUSCI> unless one pool is completely fucked up and not mining correct blocks
 563 2012-02-06 05:42:04 <pirateat40> lol
 564 2012-02-06 05:42:08 <Staatsfeind> BTC_Bear: Last I looked it was $3.6-something
 565 2012-02-06 05:42:20 <BTC_Bear> hmm.. wonder why?
 566 2012-02-06 05:42:26 <pirateat40> FROTUSCI, Ill bet deepbit does more blocks than eliguis any day.
 567 2012-02-06 05:42:28 <copumpkin> gmaxwell: yeah, I don't think anyone's ever denied the hopping allegations
 568 2012-02-06 05:42:28 <FROTUSCI> of course variance plays into it
 569 2012-02-06 05:42:48 <Staatsfeind> “rips off the non-hopping participants”  ← How the fuck does it do that?
 570 2012-02-06 05:42:49 <FROTUSCI> true pirateat40.. depends on your time horizon.. bet on the biggest pool
 571 2012-02-06 05:43:00 <FROTUSCI> but the biggest pool probably takes the most fees
 572 2012-02-06 05:43:11 <gmaxwell> copumpkin: I also can't throw stones there. I hopped the crap out of some pools, but I actually talked to the operators and got permission. :)
 573 2012-02-06 05:43:12 <pirateat40> That's why its fun.
 574 2012-02-06 05:43:31 <gmaxwell> Staatsfeind: Becuase the pool payouts are not memoryless.. even though blockfinding is.
 575 2012-02-06 05:43:31 <pirateat40> We had one guy bet on a smaller pool and hit 2 blocks with a net win of over 20 BTC
 576 2012-02-06 05:43:39 <pirateat40> Its a bet, plan and simple.
 577 2012-02-06 05:43:50 <Staatsfeind> gmaxwell: and …?
 578 2012-02-06 05:44:25 <gmaxwell> Staatsfeind: basically, on a straight proportional pool if you mine only in the .434 of the expected round duration you will increase your returns by 24%.
 579 2012-02-06 05:44:38 <copumpkin> :)
 580 2012-02-06 05:44:46 <Staatsfeind> OK, and …?
 581 2012-02-06 05:44:49 <gmaxwell> Staatsfeind: of course the rest of the pool users will have their returns lowered by that amount.
 582 2012-02-06 05:45:00 <copumpkin> gmaxwell: thus making the market more efficient!
 583 2012-02-06 05:45:02 <copumpkin> </perverse>
 584 2012-02-06 05:45:18 <Staatsfeind> Because information is not perfectly distributed, it has a value.
 585 2012-02-06 05:45:25 <gmaxwell> Staatsfeind: this is why sane pools switched to hop resistent payout schemes like PPLNS.
 586 2012-02-06 05:45:59 <gmaxwell> Staatsfeind: there is not information about future solutions, the process is straight random. It's just abusing the non-uniform marginal value of shares due to the memory added by stupid payout systems.
 587 2012-02-06 05:46:13 <copumpkin> damn right
 588 2012-02-06 05:46:45 <Staatsfeind> You can think of it then as an incentive for pool operators to adopt sane payout schemes.
 589 2012-02-06 05:47:04 <gmaxwell> Staatsfeind: indeed.
 590 2012-02-06 05:47:14 <FROTUSCI> there are still pool that are susceptible to pool hoppin? hmm
 591 2012-02-06 05:47:17 <gmaxwell> Though most pools are ghost ships today.
 592 2012-02-06 05:47:50 <gmaxwell> The competition in the pool market (and the lack of sensitivity of the remaining customers) doesn't make a lot of development attractive.
 593 2012-02-06 05:48:04 <gmaxwell> Also, the pool operator's fees come off the top— hopping doesn't hurt them.
 594 2012-02-06 05:48:18 <gmaxwell> (other than increasing their required system capacity due to burstyness)
 595 2012-02-06 05:48:19 <pirateat40> right
 596 2012-02-06 05:48:31 <copumpkin> people need to publicize the harm to other users of hopping, so other miners leave and discourage operators
 597 2012-02-06 05:48:39 <gmaxwell> copumpkin: It's been done.
 598 2012-02-06 05:48:42 <pirateat40> I get messages at least once a week from someone wanting to make a prop pool for gpumax users.
 599 2012-02-06 05:48:47 <copumpkin> obviously didn't work well enough :P
 600 2012-02-06 05:48:49 <Staatsfeind> Well, anything that reduces the volume of a pool hurts the operator.
 601 2012-02-06 05:49:02 <gmaxwell> copumpkin: Though it's _really really_ hard to convince people pool hopping is harmful to them.
 602 2012-02-06 05:49:25 da2ce7 has quit (2!~da2ce7@gateway/tor-sasl/da2ce7|Quit: KVIrc 4.1.3 Equilibrium http://www.kvirc.net/)
 603 2012-02-06 05:49:29 <gmaxwell> They go "but! memoryless!" (using whatever non-technical terms they care to use)
 604 2012-02-06 05:49:47 blishchrot has joined
 605 2012-02-06 05:50:08 <copumpkin> I used to think that too, until I realized that the payouts didn't necessarily work that way
 606 2012-02-06 05:50:25 <gmaxwell> I just shut up and wrote a simulation.
 607 2012-02-06 05:50:39 gjs278 has quit (Ping timeout: 244 seconds)
 608 2012-02-06 05:50:55 <gmaxwell> And then I tried to convince the pools I used to change. And they refused because they said hopping didn't work. And I then hopped the everlasting @#$#@ out of them.
 609 2012-02-06 05:51:07 <BlueMatt> are pools still not using payouts that discourage hopping?
 610 2012-02-06 05:51:27 hexTech has quit (Remote host closed the connection)
 611 2012-02-06 05:51:37 <pirateat40> Their will always be prop pools.
 612 2012-02-06 05:51:43 <pirateat40> There*
 613 2012-02-06 05:51:44 <BlueMatt> wow, thats bad
 614 2012-02-06 05:51:50 <Staatsfeind> pirateat40: there you go
 615 2012-02-06 05:51:53 <josephcp> isn't it still theoretically possible to hop deepbit if you try to detect block announce timings -- similar to what blockchain.info does?
 616 2012-02-06 05:51:58 <gmaxwell> BlueMatt: ahem. deepbit for example.
 617 2012-02-06 05:52:07 <BlueMatt> wait, you are kidding me?
 618 2012-02-06 05:52:11 <pirateat40> Pool ops want money, so if they all switch another one will pop up.
 619 2012-02-06 05:52:17 hexTech has joined
 620 2012-02-06 05:52:29 <BlueMatt> pirateat40: but poolops lose money if they run prop
 621 2012-02-06 05:52:45 <pirateat40> How?
 622 2012-02-06 05:52:48 <BlueMatt> s/lose/lose if they are also mining/
 623 2012-02-06 05:52:55 <pirateat40> its prop, they take a fee and its over.
 624 2012-02-06 05:53:14 <gmaxwell> BlueMatt: pool fees come off the top.
 625 2012-02-06 05:53:15 <pirateat40> Its the pools running like deepbit that are banking with PPS setups.
 626 2012-02-06 05:53:21 <BlueMatt> (a few smaller pools are used by the poolops to mine themselves, or atleast a few looked that way when I was mining)
 627 2012-02-06 05:53:34 <splatster> Wouldn't pool ops lose money if they mined in their own pool?
 628 2012-02-06 05:53:35 <gmaxwell> okay, in those cases they might lose out there.
 629 2012-02-06 05:53:40 <BlueMatt> dont think deepbit is, but...
 630 2012-02-06 05:54:23 <gmaxwell> As far as I know most poolops don't mine any any real scale.
 631 2012-02-06 05:54:34 <BlueMatt> well, ok maybe not
 632 2012-02-06 05:55:48 gjs278 has joined
 633 2012-02-06 05:56:31 da2ce7 has joined
 634 2012-02-06 06:01:57 hexTech has quit (Remote host closed the connection)
 635 2012-02-06 06:02:55 hexTech has joined
 636 2012-02-06 06:12:15 wizkid057_ has quit (Ping timeout: 276 seconds)
 637 2012-02-06 06:17:10 Cablesaurus has quit (Quit: Never put off till tomorrow, what you can do the day after tomorrow)
 638 2012-02-06 06:22:19 b4epoche has quit (Read error: Operation timed out)
 639 2012-02-06 06:22:58 hexTech has quit (Remote host closed the connection)
 640 2012-02-06 06:23:12 b4epoche has joined
 641 2012-02-06 06:24:13 hexTech has joined
 642 2012-02-06 06:27:09 jamescarr has quit (Ping timeout: 272 seconds)
 643 2012-02-06 06:28:35 wood has joined
 644 2012-02-06 06:28:46 Diablo-D3 has quit (Read error: Operation timed out)
 645 2012-02-06 06:29:58 hexTech has quit (Remote host closed the connection)
 646 2012-02-06 06:30:45 hexTech has joined
 647 2012-02-06 06:31:45 Diablo-D3 has joined
 648 2012-02-06 06:38:27 hexTech has quit (Remote host closed the connection)
 649 2012-02-06 06:39:27 hexTech has joined
 650 2012-02-06 06:44:17 Diablo-D3 has quit (Read error: Operation timed out)
 651 2012-02-06 06:45:57 Staatsfeind has left ()
 652 2012-02-06 06:46:28 hexTech has quit (Remote host closed the connection)
 653 2012-02-06 06:47:22 Diablo-D3 has joined
 654 2012-02-06 06:47:36 hexTech has joined
 655 2012-02-06 06:51:50 qw112 has joined
 656 2012-02-06 06:55:53 jamescarr has joined
 657 2012-02-06 06:58:36 gronager has quit (Quit: Leaving.)
 658 2012-02-06 07:03:25 jamescarr has quit (Ping timeout: 248 seconds)
 659 2012-02-06 07:05:15 BlueMatt has quit (Quit: Ex-Chat)
 660 2012-02-06 07:12:10 Diablo-D3 has quit (Ping timeout: 240 seconds)
 661 2012-02-06 07:24:37 Diablo-D3 has joined
 662 2012-02-06 07:28:36 ThomasV has quit (Ping timeout: 272 seconds)
 663 2012-02-06 07:29:05 underscor has quit (Quit: Leaving)
 664 2012-02-06 07:35:09 finway has joined
 665 2012-02-06 07:35:36 underscor has joined
 666 2012-02-06 07:36:25 <sipa> luke-jr: haven't touched addrman in a while
 667 2012-02-06 07:36:44 <finway> Once i open the client, all the blocks&transactions i missed was tagged the same "TIME"?
 668 2012-02-06 07:37:03 <finway> reciving time.
 669 2012-02-06 07:38:54 [Tycho] has joined
 670 2012-02-06 07:40:51 onelineproof has joined
 671 2012-02-06 07:43:13 finway has quit (Ping timeout: 245 seconds)
 672 2012-02-06 07:44:31 sgstair has quit (Read error: Connection reset by peer)
 673 2012-02-06 07:44:53 ThomasV has joined
 674 2012-02-06 07:45:01 sgstair has joined
 675 2012-02-06 07:46:03 BTC_Bear is now known as BTC_Bear|hbrntng
 676 2012-02-06 07:53:21 RazielZ has joined
 677 2012-02-06 07:54:35 sipa has quit (Quit: leaving)
 678 2012-02-06 07:57:58 npouillard has joined
 679 2012-02-06 08:08:21 finway has joined
 680 2012-02-06 08:08:50 <finway> So, roconnor found another bug?
 681 2012-02-06 08:08:58 <finway> double spending ?
 682 2012-02-06 08:09:54 <finway> Should we be careful of mining pools ?  Cause they can double_spending coinbase?
 683 2012-02-06 08:09:56 <Diablo-D3> finway: wtf is with you and your obsession with shit?
 684 2012-02-06 08:10:17 JRWR has quit (Read error: Connection reset by peer)
 685 2012-02-06 08:10:38 ThomasV has quit (Ping timeout: 248 seconds)
 686 2012-02-06 08:10:46 <finway> Diablo-D3: curiosity
 687 2012-02-06 08:11:21 <gmaxwell> finway: nothing that simple.
 688 2012-02-06 08:11:42 <gmaxwell> At the moment I don't feel adequately competent to try to walk through a comprehensible explination.
 689 2012-02-06 08:11:51 <Diablo-D3> finway: okay, hows this
 690 2012-02-06 08:11:58 <Diablo-D3> I can summarize the entirety of pretty much every attack
 691 2012-02-06 08:12:06 <Diablo-D3> you need 51% first, THEN the attack works
 692 2012-02-06 08:12:15 <Diablo-D3> but if you already have 51%, you already won
 693 2012-02-06 08:12:22 <Diablo-D3> theres no new game+ mode for bitcoin.
 694 2012-02-06 08:12:56 <finway> Diablo-D3: I don't think roconnor have 51% on the testnet ?
 695 2012-02-06 08:13:15 <gmaxwell> finway: what rconnor was doing on testnet wasn't an attack.
 696 2012-02-06 08:13:37 iocor has joined
 697 2012-02-06 08:13:53 <Diablo-D3> finway: dude, its testnet
 698 2012-02-06 08:13:57 <finway> gmaxwell: sure it's double spending?  and was adopted by BBE ?
 699 2012-02-06 08:14:02 <Diablo-D3> I could plug a 486 in and get 51%.
 700 2012-02-06 08:14:09 <gmaxwell> finway: It was just a demonstration that you could mineCoinbaseA sependA mineexactDupeofA spendA.  This isn't a double spend.
 701 2012-02-06 08:14:32 molecular has quit (Ping timeout: 252 seconds)
 702 2012-02-06 08:14:37 <gmaxwell> It's two seperate spends of two seperate coinbases that happen to share the same ID.
 703 2012-02-06 08:15:12 molecular has joined
 704 2012-02-06 08:15:28 <gmaxwell> However, it does appear that there is a bug here— but its related to very specific reorgs happening on top of that sequence. It's not actually been tested anywhere yet, however.
 705 2012-02-06 08:15:53 <gmaxwell> and BBE blows up when you sneeze on it slightly.
 706 2012-02-06 08:16:14 <gmaxwell> I blew up BBE last week doing nothing stranger than a simple reorg of testnet.
 707 2012-02-06 08:17:02 <[Tycho]> btw, blockchaininfo shows orphans incorrectly.
 708 2012-02-06 08:17:02 <finway> Diablo-D3: that's helpful.
 709 2012-02-06 08:17:24 <finway> gmaxwell：thanks for elaboration, always.
 710 2012-02-06 08:17:55 <gmaxwell> [Tycho]: it also shows payments from orphaned coinbases as real.
 711 2012-02-06 08:18:03 <gmaxwell> (e.g. still in the balances of the addresses you look at)
 712 2012-02-06 08:18:25 <finway>  mineCoinbaseA sependA mineexactDupeofA spendA
 713 2012-02-06 08:18:40 <finway> that means you should mine two blocks ? not just copy it?
 714 2012-02-06 08:18:55 <finway> mine twice ?
 715 2012-02-06 08:19:02 <gmaxwell> finway: yes, these would be completely seperate blocks, but the coinbase transactions would be identical.
 716 2012-02-06 08:19:15 <finway> Oh, that's silly.
 717 2012-02-06 08:19:24 <Diablo-D3> gmaxwell: okay so wait
 718 2012-02-06 08:19:32 <Diablo-D3> what is dupe of A?
 719 2012-02-06 08:19:35 <gmaxwell> There are some of them in the blockchain already due to software bugs in miners.
 720 2012-02-06 08:19:48 <finway> dupofA  coinbase tx
 721 2012-02-06 08:19:51 <finway> i guess
 722 2012-02-06 08:20:19 <gmaxwell> Diablo-D3: if you set the same extranonce and same payout address you can generate coinbases which have the same txn ID.
 723 2012-02-06 08:20:39 <Diablo-D3> but different nonce?
 724 2012-02-06 08:21:18 <gmaxwell> sure, different transactions in the block, different block nonce.. different prev different timestamps..
 725 2012-02-06 08:21:21 <gmaxwell> They'd be different blocks.
 726 2012-02-06 08:21:24 <finway> What's the odds of creating two blocks with same nonce?
 727 2012-02-06 08:21:29 <gmaxwell> But the transaction paying the miner is duplicated.
 728 2012-02-06 08:21:57 <Diablo-D3> gmaxwell: erm
 729 2012-02-06 08:21:59 <gmaxwell> This kind of duplication can't happen on its own, you need to have a software bug to cause it, pretty much.
 730 2012-02-06 08:22:00 <Diablo-D3> but...
 731 2012-02-06 08:22:18 <gmaxwell> It's a stupid thing to do because normally you can only spend one of them. So you just burned 50 BTC in the process.
 732 2012-02-06 08:22:21 <Diablo-D3> okay, tx id has ti be unique forever? because that sounds insane
 733 2012-02-06 08:22:39 <finway> It's a stupid thing to do because normally you can only spend one of them. So you just burned 50 BTC in the process.  -- That's good news
 734 2012-02-06 08:22:58 <gmaxwell> roconnor pointed out that if you do one, then spend it, then do another you can spend it too.
 735 2012-02-06 08:23:23 <gmaxwell> This is also good news, actually, because it means someone with a pruned node doesn't have to keep copies of coinbases forever to prevent dupe spending.
 736 2012-02-06 08:23:36 <Diablo-D3> gmaxwell: okay okay so wait
 737 2012-02-06 08:23:36 <Diablo-D3> because this sounds nuts
 738 2012-02-06 08:23:53 <Diablo-D3> I make a block, gen tx id is x... then I make another block and gen tx id is also x?
 739 2012-02-06 08:23:59 <midnightmagic> does the first spend *have* to happen prior to the second mine?
 740 2012-02-06 08:24:04 <gmaxwell> midnightmagic: Yes.
 741 2012-02-06 08:24:09 <midnightmagic> k
 742 2012-02-06 08:24:10 <gmaxwell> Diablo-D3: yes, you have to be stupid to make that happen.
 743 2012-02-06 08:24:25 <Diablo-D3> but that shouldnt matter
 744 2012-02-06 08:24:36 <gmaxwell> midnightmagic: if you do them A, DupA, SpendA, SpendDupA  the final txn fails (double spend).
 745 2012-02-06 08:24:40 <Diablo-D3> because the id of gen tx isnt x, its blockhash:x
 746 2012-02-06 08:25:40 <gmaxwell> The ID of the transaction is really just the transaction ID. Design oversight. The height or prev should have been added to the generation txn hash (as an input probably, the input is 32 bytes of zeros)
 747 2012-02-06 08:25:53 <gmaxwell> In any case all that stuff is harmless if a bit weird.
 748 2012-02-06 08:26:02 <[Tycho]> Did someone mined duplicate coinbases ?
 749 2012-02-06 08:26:08 dissipate has joined
 750 2012-02-06 08:26:08 dissipate has quit (Changing host)
 751 2012-02-06 08:26:08 dissipate has joined
 752 2012-02-06 08:26:09 <gmaxwell> [Tycho]: several times in the past, yes.
 753 2012-02-06 08:26:22 <gmaxwell> [Tycho]: I'm aware of at lest two pairs.
 754 2012-02-06 08:26:35 <gmaxwell> Custom miner code that was stupid.
 755 2012-02-06 08:26:41 <gmaxwell> (most likely)
 756 2012-02-06 08:26:48 <Diablo-D3> gmaxwell: well, its what I said then
 757 2012-02-06 08:26:57 <Diablo-D3> the actual name of the tx is blockid:x, not x.
 758 2012-02-06 08:27:14 <gmaxwell> No, it's really X.. it just happens to do the right thing.
 759 2012-02-06 08:27:24 <finway> [quote]two paires[/quote] so 100 bitcoins vanished?
 760 2012-02-06 08:27:26 <gmaxwell> Unfortunately, reorgs don't appear to do the right thing.
 761 2012-02-06 08:27:30 <gmaxwell> finway: yes.
 762 2012-02-06 08:27:48 qw112 has quit (Ping timeout: 245 seconds)
 763 2012-02-06 08:28:13 <Diablo-D3> gmaxwell: no, Im saying
 764 2012-02-06 08:28:16 <Diablo-D3> its blockid:x.
 765 2012-02-06 08:28:20 <Diablo-D3> as in, people have a last name.
 766 2012-02-06 08:28:44 <Diablo-D3> the code should be internally identifying tx in cross-block manners exactly that way
 767 2012-02-06 08:28:59 <gmaxwell> Diablo-D3: the input specification is TxnHash:output_index  the block isn't relevant.
 768 2012-02-06 08:29:10 doublec has quit (Changing host)
 769 2012-02-06 08:29:10 doublec has joined
 770 2012-02-06 08:29:17 <gmaxwell> You can make txn based on transactions that aren't in blocks yet (an important feature).
 771 2012-02-06 08:29:19 <finway> but i thought according to this it's(dupA) spendable ?  http://blockexplorer.com/testnet/block/0000000013aa9f67da178005f9ced61c7064dd6e8464b35f6a8ca8fabc1ca2cf
 772 2012-02-06 08:29:29 <Diablo-D3> gmaxwell:  the specification doesnt mean shit though
 773 2012-02-06 08:29:33 <Diablo-D3> gmaxwell: the code can do whatever it wants
 774 2012-02-06 08:29:37 <gmaxwell> finway: they're only spendable in that sequence.
 775 2012-02-06 08:29:52 <midnightmagic> does roconnor post details of the reorg problem somewhere?
 776 2012-02-06 08:29:57 <Diablo-D3> gmaxwell: its just an alternate internal representation
 777 2012-02-06 08:30:00 <gmaxwell> No. jesus. It's too early.
 778 2012-02-06 08:30:04 <gmaxwell> sorry. :)
 779 2012-02-06 08:30:16 <midnightmagic> :-(
 780 2012-02-06 08:30:31 <midnightmagic> lol well how do you know about it then?
 781 2012-02-06 08:31:23 <gmaxwell> In any case, it appears that if you have the sequence:  A ... [DupA SpendA] ... SpendA   < which is perfectly find and sane if a bit weird.
 782 2012-02-06 08:31:49 OneFixt has quit (Read error: Connection reset by peer)
 783 2012-02-06 08:31:50 <gmaxwell> and then reorg so that "[DupA SpendA] ... SpendA" didn't happen. Then nodes that saw the reorg don't correctly mark A as spendable again.
 784 2012-02-06 08:32:21 <gmaxwell> so nodes that saw the reorg will not allow a future SpendA but nodes that didn't see the reorg will (it hasn't been spent yet).
 785 2012-02-06 08:32:24 OneFixt has joined
 786 2012-02-06 08:32:31 <gmaxwell> But this hasn't actually been verified yet.
 787 2012-02-06 08:32:45 <gmaxwell> And the exact details required to make it happen may be different.
 788 2012-02-06 08:32:48 <Diablo-D3> gmaxwell: okay okay wait a second
 789 2012-02-06 08:32:48 <Diablo-D3> so
 790 2012-02-06 08:33:11 <Diablo-D3> you spend A, then dupe A then spend dupe A.... and its too stupid to realize a tx with that id was already spent?
 791 2012-02-06 08:33:12 <gmaxwell> midnightmagic: and I only know about it from talking in here, and reading the code.
 792 2012-02-06 08:33:41 <gmaxwell> Diablo-D3: Thats actually a good thing. Otherwise pruning wouldn't work as well.
 793 2012-02-06 08:34:04 <gmaxwell> Diablo-D3: and it's okay because DupeA makes it alright.. (prevents the inflation of the currency) e.g. you're not actually creating any more coin this way.
 794 2012-02-06 08:34:35 <Diablo-D3> gmaxwell: but... you are
 795 2012-02-06 08:34:38 <Diablo-D3> since you spent 50 twice
 796 2012-02-06 08:35:01 <gmaxwell> Diablo-D3: no, different 50.. one was created by block X one was created by later block >X.
 797 2012-02-06 08:35:10 <gmaxwell> It's two 50s .. which happen to share the same name.
 798 2012-02-06 08:35:24 <Diablo-D3> yes but like
 799 2012-02-06 08:35:35 <Diablo-D3> you're just saying what I said now
 800 2012-02-06 08:35:52 <gmaxwell> It's like if you have two kids- You could name them "Alice and Bob" or "John and John" You'd have two Johns in the second case, but no more children. :)
 801 2012-02-06 08:35:57 <Diablo-D3> tx id means shit since the actual tx is blockid:tx, the code just isnt efficient about it
 802 2012-02-06 08:36:10 <midnightmagic> lol, wait, we can do multisig right now if we mine our own blocks?
 803 2012-02-06 08:36:36 <gmaxwell> Diablo-D3: ah but it's important that it really isn't that way. because the mess made by reorgs wouldn't happen if there was some blockid in it.
 804 2012-02-06 08:37:03 <Diablo-D3> gmaxwell: then the reorg behavior needs to be improved
 805 2012-02-06 08:37:12 <Diablo-D3> I mean, how did this code get written this way to begin with?
 806 2012-02-06 08:37:24 <gmaxwell> Diablo-D3: also if you do  A then in the next block DupA ... you can only spend one of them.
 807 2012-02-06 08:37:59 <Diablo-D3> if txid can be legitimately duplicated, then tx cannot be refereed to txid alone, they need to be referred to by fqtxid
 808 2012-02-06 08:38:08 <gmaxwell> Diablo-D3: I don't think Satoshi realized that you could duplicate a coinbase, if he did he probably would have added the prev to the hash so you couldn't.
 809 2012-02-06 08:38:19 <Diablo-D3> gmaxwell: well
 810 2012-02-06 08:38:25 <Diablo-D3> blergh.
 811 2012-02-06 08:38:32 <gmaxwell> Yes, this could also blow up external systems that expect txid to be unique.
 812 2012-02-06 08:38:34 <Diablo-D3> just adding the rule requiring unique txid would probably fix it
 813 2012-02-06 08:38:43 <Diablo-D3> it doesnt break bitcoin
 814 2012-02-06 08:39:06 <gmaxwell> Also, roconner pointed out that because you can make duplicate coinbases you can probably make duplicate normal txn too. (based on duplicate coinbases)
 815 2012-02-06 08:39:24 <gmaxwell> And yes, this doesn't break bitcoin.
 816 2012-02-06 08:39:25 <Diablo-D3> oh probably
 817 2012-02-06 08:39:29 erle- has joined
 818 2012-02-06 08:39:41 <Diablo-D3> gmaxwell: I dont like this bug either way
 819 2012-02-06 08:39:52 <gmaxwell> It'll probably get fixed by either some kind of duplicate disallowance (ugh, means pruning can't completely prune coinbases) or by fixing the reorg logic to be smarter.
 820 2012-02-06 08:39:53 <Diablo-D3> because I would have automatically written code to do unique txen only
 821 2012-02-06 08:40:05 <gmaxwell> You and everyone else.
 822 2012-02-06 08:40:53 dissipate has quit (Quit: Leaving)
 823 2012-02-06 08:41:03 <gmaxwell> In any case, it sounds like someone with the hashpower to do a big reorg (I see how to do it with 100 block reorg, though perhaps there is a way with less) could potentially create a fork between nodes that were online and nodes that weren't online.
 824 2012-02-06 08:41:59 <gmaxwell> If the were-online nodes had more hashpower the weren't online nodes would merge back. This is the more likely case.
 825 2012-02-06 08:42:25 <gmaxwell> If it was the other way though— the weren't online nodes had more hash power.. then the fork would never resolve and the were-online nodes would be stuck.
 826 2012-02-06 08:42:41 <midnightmagic> i can't wait until the first big scary reorg happens. :)
 827 2012-02-06 08:42:45 <gmaxwell> But as I said: currently tested. And I might have the details wrong.
 828 2012-02-06 08:43:56 <gmaxwell> midnightmagic: hm? we've had a 53 block reorg on the mainnet once.
 829 2012-02-06 08:44:24 <midnightmagic> gmaxwell: That was a while ago, no? I mean "scary" :)
 830 2012-02-06 08:44:47 ThomasV has joined
 831 2012-02-06 08:44:53 <gmaxwell> Yea, that was the value overflow when some industrious hacker got his 184 billion bitcoin taken away.
 832 2012-02-06 08:45:04 <midnightmagic> =]
 833 2012-02-06 08:45:06 <finway> how to survive from "reorg" ?
 834 2012-02-06 08:45:33 <gmaxwell> finway: there are reorgs every day... though usually only 1 or 2 blocks deep.
 835 2012-02-06 08:46:12 <finway> So, 53 blocks_long reorg will make 6 confirmation transactions unsafe,right ?
 836 2012-02-06 08:46:17 devrandom has quit (Remote host closed the connection)
 837 2012-02-06 08:46:25 <gmaxwell> finway: hah yes.
 838 2012-02-06 08:46:30 <gmaxwell> But that was a special case.
 839 2012-02-06 08:46:48 <gmaxwell> A bug in bitcoin allowed someone to give themselves 184 billion bitcoin.
 840 2012-02-06 08:47:03 devrandom has joined
 841 2012-02-06 08:47:06 <gmaxwell> fixing the bug rewrote the chain after that point.
 842 2012-02-06 08:47:35 <gmaxwell> (fixed nodes did not believe the chain from the bad transaction on)
 843 2012-02-06 08:48:11 <gmaxwell> https://bitcointalk.org/index.php?topic=822.0
 844 2012-02-06 08:48:48 <finway> gmaxwell: that's huge bug, hah
 845 2012-02-06 08:49:26 <Diablo-D3> [03:40:40] <gmaxwell> Yea, that was the value overflow when some industrious hacker got his 184 billion bitcoin taken away.
 846 2012-02-06 08:49:27 <Diablo-D3> wat?
 847 2012-02-06 08:49:31 <finway> bug cracks confidence.
 848 2012-02-06 08:49:43 <Diablo-D3> dude
 849 2012-02-06 08:49:48 iocor has quit (Quit: Computer has gone to sleep.)
 850 2012-02-06 08:49:48 <Diablo-D3> if I had 184 billion bitcoins
 851 2012-02-06 08:49:49 FROTUSCI has quit (Ping timeout: 245 seconds)
 852 2012-02-06 08:49:51 <Diablo-D3> I could be happy forever
 853 2012-02-06 08:49:54 <gmaxwell> finway: Yes, but finding bugs also proves people are looking.
 854 2012-02-06 08:50:01 <Diablo-D3> seriously
 855 2012-02-06 08:50:05 <Diablo-D3> I could, like
 856 2012-02-06 08:50:10 <Diablo-D3> found a colony on mars
 857 2012-02-06 08:50:15 <Diablo-D3> with me and thousands of hot women
 858 2012-02-06 08:50:44 <Diablo-D3> and I can finally found my church of bacon.
 859 2012-02-06 08:50:47 <finway> Diablo-D3: you wish, go, get the bug out
 860 2012-02-06 08:50:59 <Diablo-D3> finway: fuck you, I cant hack on everything at once
 861 2012-02-06 08:51:05 <Diablo-D3> if I could, we'd have hard AI by now
 862 2012-02-06 08:54:23 MrTiggr has joined
 863 2012-02-06 08:56:18 <midnightmagic> Church Of Bacon
 864 2012-02-06 08:56:24 <Diablo-D3> THE CHURCH OF BACON!
 865 2012-02-06 08:56:35 <midnightmagic> I've learned how to smoke my own bacon. I should send you some. :)
 866 2012-02-06 08:56:55 <midnightmagic> (It's from half-wild special heritage stock pigs, they have razorbacks.)
 867 2012-02-06 08:57:53 <Diablo-D3> naw
 868 2012-02-06 08:57:55 <Diablo-D3> dude
 869 2012-02-06 08:57:58 <Diablo-D3> sell bacon for btc
 870 2012-02-06 08:58:08 <midnightmagic> But then it would be a job.
 871 2012-02-06 08:58:33 <gmaxwell> furry piggies.
 872 2012-02-06 09:00:12 finway has quit (Quit: Page closed)
 873 2012-02-06 09:00:13 <midnightmagic> gmaxwell: Dude, this pork tastes better than any pork I've ever eaten in my whole life. :)
 874 2012-02-06 09:00:32 <Diablo-D3> are you sure its pork?
 875 2012-02-06 09:00:36 <spawn-> i think thats a young pork :)
 876 2012-02-06 09:00:48 <midnightmagic> Yes, the farmer made me butcher it to get more..
 877 2012-02-06 09:05:11 <midnightmagic> 280lb sow isn't so young. :)
 878 2012-02-06 09:06:47 sipa1024 has joined
 879 2012-02-06 09:06:55 <Diablo-D3> going to bed
 880 2012-02-06 09:06:56 <Diablo-D3> night all
 881 2012-02-06 09:07:12 <Diablo-D3> maybe Ill finish the first stage of the new kernel tommorow
 882 2012-02-06 09:07:17 <Diablo-D3> then I can start in optimizing it
 883 2012-02-06 09:11:21 Clipse has quit (Quit: Clipse)
 884 2012-02-06 09:12:10 Diablo-D3 has quit (Ping timeout: 244 seconds)
 885 2012-02-06 09:22:06 sgstair has quit (Ping timeout: 252 seconds)
 886 2012-02-06 09:36:02 marf_away has joined
 887 2012-02-06 09:41:42 _Fireball has joined
 888 2012-02-06 09:41:43 Clipse has joined
 889 2012-02-06 09:42:27 mizerydearia has joined
 890 2012-02-06 09:42:43 cdecker has joined
 891 2012-02-06 09:43:43 wtfman[away] has quit (Read error: Operation timed out)
 892 2012-02-06 09:43:47 topi` has quit (Read error: Operation timed out)
 893 2012-02-06 09:43:50 ForceMajeure has quit (Read error: Connection reset by peer)
 894 2012-02-06 09:43:52 wtfman[away] has joined
 895 2012-02-06 09:43:53 topi` has joined
 896 2012-02-06 09:44:13 ovidiusoft has joined
 897 2012-02-06 09:44:23 aCZ-[e] has quit (Read error: Connection reset by peer)
 898 2012-02-06 09:44:59 paul0 has joined
 899 2012-02-06 09:45:02 ForceMajeure has joined
 900 2012-02-06 09:45:28 ForceMajeure is now known as Guest20048
 901 2012-02-06 09:50:16 dr_win has quit (Read error: Connection reset by peer)
 902 2012-02-06 09:50:25 dr_win has joined
 903 2012-02-06 09:57:54 splatster has quit (Quit: Up up and away.)
 904 2012-02-06 10:03:04 ahihi2 has quit (Ping timeout: 252 seconds)
 905 2012-02-06 10:04:46 booo has joined
 906 2012-02-06 10:08:56 BGL has joined
 907 2012-02-06 10:10:22 ahihi2 has joined
 908 2012-02-06 10:14:32 ThomasV has quit (Ping timeout: 272 seconds)
 909 2012-02-06 10:30:17 wizkid057 has joined
 910 2012-02-06 10:30:44 wizkid057 is now known as Guest84384
 911 2012-02-06 10:31:14 zeiris has quit (Ping timeout: 244 seconds)
 912 2012-02-06 10:34:59 b4epoche has quit (Read error: Operation timed out)
 913 2012-02-06 10:36:12 b4epoche has joined
 914 2012-02-06 10:37:57 marf_away has quit (Ping timeout: 245 seconds)
 915 2012-02-06 10:42:20 erle- has quit (Quit: erle-)
 916 2012-02-06 10:44:33 marf_away has joined
 917 2012-02-06 10:47:19 da2ce7 has joined
 918 2012-02-06 11:00:54 booo has quit (Ping timeout: 256 seconds)
 919 2012-02-06 11:21:32 ovidiusoft has quit (Ping timeout: 265 seconds)
 920 2012-02-06 11:25:26 dikidera has joined
 921 2012-02-06 11:26:31 diki has quit (Ping timeout: 244 seconds)
 922 2012-02-06 11:30:10 dwon has quit (Read error: Connection reset by peer)
 923 2012-02-06 11:33:34 dwon has joined
 924 2012-02-06 11:40:52 nameless has joined
 925 2012-02-06 11:40:52 jgarzik has joined
 926 2012-02-06 11:40:52 bitfoo has joined
 927 2012-02-06 11:40:52 d4de has joined
 928 2012-02-06 11:42:24 MrTiggr has joined
 929 2012-02-06 12:01:16 barmstrong has joined
 930 2012-02-06 12:02:08 wood has joined
 931 2012-02-06 12:07:00 sje has joined
 932 2012-02-06 12:07:00 sje has quit (Changing host)
 933 2012-02-06 12:07:00 sje has joined
 934 2012-02-06 12:12:59 marf_away has quit (Ping timeout: 272 seconds)
 935 2012-02-06 12:17:36 iocor has joined
 936 2012-02-06 12:18:58 marf_away has joined
 937 2012-02-06 12:20:44 booo has joined
 938 2012-02-06 12:22:00 dikidera has quit (Ping timeout: 255 seconds)
 939 2012-02-06 12:22:13 diki has joined
 940 2012-02-06 12:24:45 dikidera has joined
 941 2012-02-06 12:26:41 diki has quit (Ping timeout: 240 seconds)
 942 2012-02-06 12:32:32 sshirokov has quit (Ping timeout: 252 seconds)
 943 2012-02-06 12:34:09 sshirokov has joined
 944 2012-02-06 12:43:17 hexTech has quit (Quit: -)
 945 2012-02-06 12:44:16 midnightmagic has quit (Ping timeout: 252 seconds)
 946 2012-02-06 12:44:45 midnightmagic has joined
 947 2012-02-06 12:44:46 midnightmagic has quit (Excess Flood)
 948 2012-02-06 12:45:14 midnightmagic_ has joined
 949 2012-02-06 12:45:15 midnightmagic_ has quit (Excess Flood)
 950 2012-02-06 12:45:20 hexTech has joined
 951 2012-02-06 12:45:44 midnightmagic has joined
 952 2012-02-06 12:45:51 josephcp has quit (Ping timeout: 248 seconds)
 953 2012-02-06 12:46:22 ThomasV has joined
 954 2012-02-06 12:48:11 bitlad has joined
 955 2012-02-06 12:55:55 gfinn has quit (Ping timeout: 276 seconds)
 956 2012-02-06 12:56:09 iocor has quit (Quit: Computer has gone to sleep.)
 957 2012-02-06 12:56:34 da2ce7 has quit (2!~da2ce7@gateway/tor-sasl/da2ce7|Ping timeout: 276 seconds)
 958 2012-02-06 13:04:39 da2ce7 has joined
 959 2012-02-06 13:05:45 erle- has joined
 960 2012-02-06 13:06:29 cdecker has quit (Remote host closed the connection)
 961 2012-02-06 13:10:29 josephcp has joined
 962 2012-02-06 13:12:02 Joric has joined
 963 2012-02-06 13:12:02 Joric has quit (Changing host)
 964 2012-02-06 13:12:02 Joric has joined
 965 2012-02-06 13:13:19 slush has joined
 966 2012-02-06 13:13:48 gfinn has joined
 967 2012-02-06 13:17:50 ovidiusoft has joined
 968 2012-02-06 13:19:23 userhhj has joined
 969 2012-02-06 13:19:40 Moron__ has joined
 970 2012-02-06 13:24:56 minimoose has joined
 971 2012-02-06 13:25:52 gjs278 has quit (Ping timeout: 244 seconds)
 972 2012-02-06 13:26:44 gjs278 has joined
 973 2012-02-06 13:28:09 Prattler has joined
 974 2012-02-06 13:29:35 datagutt has joined
 975 2012-02-06 13:29:38 datagutt has quit (Changing host)
 976 2012-02-06 13:29:38 datagutt has joined
 977 2012-02-06 13:29:44 Zarutian has joined
 978 2012-02-06 13:30:49 danbri has joined
 979 2012-02-06 13:33:32 <gmaxwell> [Tycho]: Ah, you are on now. I was running some transaction analysis last night and at first I thought the output was broken because I kept seeing some address over and over again.
 980 2012-02-06 13:35:24 <gmaxwell> [Tycho]: Out of the last 192636 transactions, 52265 —or 27%— involve 1VayNert3x1KzbpzMGt2qdqrAThiRovi8. I think thats more than a little extreme.
 981 2012-02-06 13:35:57 bitlad has quit (Remote host closed the connection)
 982 2012-02-06 13:35:59 Joric has quit (Ping timeout: 248 seconds)
 983 2012-02-06 13:36:02 <lianj> ^^
 984 2012-02-06 13:39:40 <gmaxwell> [Tycho]: Adopting sendmany would dramatically reduce your volume, so would being a little more agressive in delaying and batching small payouts: many of these payouts are just bitcents.
 985 2012-02-06 13:39:41 marf_away has quit (Quit: Nettalk6 - www.ntalk.de)
 986 2012-02-06 13:40:25 <gmaxwell> Excessively small payouts don't benefit your customers particularly either— as they end up with bloated wallets that suffer higher transaction fees when they spend from them.
 987 2012-02-06 13:41:08 <gmaxwell> And, apparently, reducing the volume of your transactions would materially reduce the size of the bitcoin blockchain growth— something beneficial to all bitcoin users.
 988 2012-02-06 13:42:11 <lianj> does he include these small payouts in his mined blocks only, and thus paying no fees on them?
 989 2012-02-06 13:42:20 <slush> lianj: no
 990 2012-02-06 13:42:54 <gmaxwell> lianj: he announces transactions, sometimes the whole default free area is used up by his transactions.
 991 2012-02-06 13:43:18 <slush> gmaxwell: the way how to force bitcoin users to pay fees? ;-)
 992 2012-02-06 13:43:31 Dyaheon has joined
 993 2012-02-06 13:43:54 <gmaxwell> I haven't figured out if he's slowing down free transaction processing in any meaningful way— but it wouldn't surprise me if he was.
 994 2012-02-06 13:44:01 <Moron__> tsssk
 995 2012-02-06 13:44:06 <Moron__> bad [Tycho]
 996 2012-02-06 13:45:46 <gmaxwell> I could start posting patches to suppress relaying/mining these transactions unless they sendmany or include a fee since he's made them so reliably identifyable.. but that seems a little petty.
 997 2012-02-06 13:46:02 <gmaxwell> Then again, if its slowing down transactions, perhaps not.
 998 2012-02-06 13:46:34 bitlad has joined
 999 2012-02-06 13:51:36 <Moron__> do it, he needs to be taught a lesson in respect for the almighty blockchain!
1000 2012-02-06 13:51:37 <Moron__> :P
1001 2012-02-06 13:52:44 <lianj> then he just changes the fingerprint
1002 2012-02-06 13:53:00 <gmaxwell> Moron__: this would not likely have that effect, but perhaps it would reduce the slowdown harm.
1003 2012-02-06 13:53:12 paul0 has quit (Ping timeout: 252 seconds)
1004 2012-02-06 13:53:38 <[Tycho]> Hello, gmaxwell.
1005 2012-02-06 13:53:42 <nanotube> or perhaps... just talk to him and get him to batch/sendmany. :)
1006 2012-02-06 13:53:49 <gmaxwell> nanotube: yup.
1007 2012-02-06 13:53:57 <gmaxwell> [Tycho]: You evil spammer you. :)
1008 2012-02-06 13:54:14 <[Tycho]> No, all my TXes are meaningful.
1009 2012-02-06 13:54:18 paul0 has joined
1010 2012-02-06 13:54:48 <[Tycho]> They are keeping network healthy and forcing adoption of light clients and stuff.
1011 2012-02-06 13:54:59 <Moron__> lol
1012 2012-02-06 13:55:20 <sipa1024> [Tycho]: is that your excuse?
1013 2012-02-06 13:55:23 <lianj> haha, thats a viewpoint
1014 2012-02-06 13:55:23 sipa1024 is now known as sipa
1015 2012-02-06 13:56:13 bitlad has quit (Ping timeout: 272 seconds)
1016 2012-02-06 13:56:16 <[Tycho]> As I remember, I had some very important reason not to use sendmany. But I can't remember what is was exactly.
1017 2012-02-06 13:56:51 <lianj> handful of miningpool owners, handful of full clients - bitcoin is screwed
1018 2012-02-06 13:56:55 <gmaxwell> I think around here we've been continually expecting you to switch to sendmany. I don't believe you've disclosed any such reason before.
1019 2012-02-06 13:57:36 <[Tycho]> But those TXes are bothering you, I can make them less noticeable.
1020 2012-02-06 13:58:10 <gmaxwell> [Tycho]: I'd like you to make them less numerous, not less noticable.
1021 2012-02-06 13:58:18 <riush> but what would be the disadvantage in avoiding them?
1022 2012-02-06 13:58:40 <[Tycho]> I'm including most of those TXes in my own blocks anyway.
1023 2012-02-06 13:59:08 <[Tycho]> Also, you are counting percentage of quantity, not size, that's a bit wrong.
1024 2012-02-06 13:59:21 iocor has joined
1025 2012-02-06 14:00:23 <[Tycho]> Also, I'm including most free TXes in my blocks (my free zone is bigger).
1026 2012-02-06 14:00:36 <sipa> That doesn't mean the rest doesn't need to store them.
1027 2012-02-06 14:00:38 <sipa> Would you please at least consider switching to sendmany again?
1028 2012-02-06 14:00:57 Phoebus has joined
1029 2012-02-06 14:01:16 <gmaxwell> Yes, you're using pay to pubkey, so they're a little smaller than they would be otherwise— but do you think it would really change the outcome?  I guess by sizes its 20% instead of 27%.
1030 2012-02-06 14:01:48 <[Tycho]> Sending change to pubkey makes the round trip 9% smaller.
1031 2012-02-06 14:02:31 <gmaxwell> [Tycho]: but you have a dozen times more uses of your own address than you would if you were using sendmany. 9% is small by comparison.
1032 2012-02-06 14:02:41 iocor has quit (Read error: Connection reset by peer)
1033 2012-02-06 14:03:37 gavinandresen has joined
1034 2012-02-06 14:03:48 <gmaxwell> With sendmany you would have 1-2 of your own inputs, N customers plus your own output. Instead of one input, one customer, and one back to yourself.
1035 2012-02-06 14:03:50 <sipa> an extra txout is 34 bytes
1036 2012-02-06 14:04:04 <sipa> an extra transaction is is 250 bytes
1037 2012-02-06 14:04:09 <Eliel> gmaxwell: then again, if the VayNert address is present in every one of those transactions, they'll compress pretty well once we include blockchain compression in the client
1038 2012-02-06 14:05:02 <sipa> Eliel: not sure how that matters
1039 2012-02-06 14:05:47 <gmaxwell> Eliel: That would be just as true if he had ten times fewer due to sendmany, and half the number of outputs due to better batching.
1040 2012-02-06 14:06:14 copumpkin has quit (Quit: Computer has gone to sleep.)
1041 2012-02-06 14:06:31 <gmaxwell> I'm guessing that pubkey isn't a compressed pubkey either, but I didn't bother looking.
1042 2012-02-06 14:07:24 BurtyB has quit (Read error: Connection reset by peer)
1043 2012-02-06 14:07:24 BaltarNZ has quit (Read error: Operation timed out)
1044 2012-02-06 14:07:41 BurtyB has joined
1045 2012-02-06 14:08:35 Joric has joined
1046 2012-02-06 14:10:13 <gmaxwell> https://people.xiph.org/~greg/last.192636.txt.gz  < if anyone would like to do some analysis on how much his outputs could be reduced by better batching or the like...
1047 2012-02-06 14:11:12 <gmaxwell> The format is TX: [[inputtxn/addr,...],[outvalue/addr,...]] with one line per transaction.
1048 2012-02-06 14:11:29 <gmaxwell> (35MiB file warning :) )
1049 2012-02-06 14:12:01 <Eliel> what time period does that file contain?
1050 2012-02-06 14:12:42 <gmaxwell> It's the last 192636 transactions.. whatever that was.. let me check.
1051 2012-02-06 14:13:06 <gmaxwell> (well last from block 165564, when I started this dump, down)
1052 2012-02-06 14:13:07 * Eliel wonders about the strange number.
1053 2012-02-06 14:13:55 da2ce7 has quit (2!~da2ce7@gateway/tor-sasl/da2ce7|Ping timeout: 276 seconds)
1054 2012-02-06 14:13:57 <gmaxwell> BLOCK_161186
1055 2012-02-06 14:14:59 <gmaxwell> Eliel: I'm dumping all txn in this form right now. ... I wanted to do some analysis on the data I had so far after it had run for a few minutes, so 192636 was simply where it was at the time.
1056 2012-02-06 14:15:00 <[Tycho]> better batching == sendmany ?
1057 2012-02-06 14:15:57 <Eliel> oh ok
1058 2012-02-06 14:16:13 <gmaxwell> [Tycho]: By better batching I meant further delaying payments to very small miners, e.g. not paying them every day.  Though using that data, someone could also figure out the send many data size savings.
1059 2012-02-06 14:16:53 <gmaxwell> Eliel: they're ordered a block at a time. You can implicitly see the block boundaries due to the coinbase txn which have the BLOCK_nnn as the input.
1060 2012-02-06 14:19:54 userhhj has quit (Ping timeout: 255 seconds)
1061 2012-02-06 14:25:07 <gmaxwell> slush: any idea how many payout transactions you're doing per week, how many total outputs you're doing per week?
1062 2012-02-06 14:26:04 <slush> gmaxwell: I'm using sendmany with up to 150 outputs per tx
1063 2012-02-06 14:26:18 <gmaxwell> So obviously not that many transactions! :)
1064 2012-02-06 14:26:21 iocor has joined
1065 2012-02-06 14:27:20 da2ce7 has joined
1066 2012-02-06 14:27:36 <slush> gmaxwell: and average output is for 1.6 BTC
1067 2012-02-06 14:29:22 <gmaxwell> ‽
1068 2012-02-06 14:29:23 <gmaxwell> wow.
1069 2012-02-06 14:30:36 <slush> gmaxwell: oh, sorry, 0.95BTC
1070 2012-02-06 14:30:42 <slush> gmaxwell: I forgot to filter out nmc payouts
1071 2012-02-06 14:32:47 booo has quit (Ping timeout: 245 seconds)
1072 2012-02-06 14:34:12 <gmaxwell> slush: what percantage of the outputs are less than .1?
1073 2012-02-06 14:34:42 userhhj has joined
1074 2012-02-06 14:34:50 Clipse has quit (Quit: Clipse)
1075 2012-02-06 14:35:40 <gmaxwell> Looks like the average output from [Tycho] is 1.61704 but thats pumped up by a few very large ones.
1076 2012-02-06 14:35:51 <slush> gmaxwell: 33%
1077 2012-02-06 14:37:45 <gmaxwell> Ah. More like 20% in his case, in fact. — unless my very quick ghetto awk parsing that output is incorrect (a distinct possibility :) )
1078 2012-02-06 14:38:39 danbri has quit (Remote host closed the connection)
1079 2012-02-06 14:47:45 Clipse has joined
1080 2012-02-06 14:47:51 b4epoche_ has joined
1081 2012-02-06 14:48:42 b4epoche has quit (Ping timeout: 240 seconds)
1082 2012-02-06 14:48:42 b4epoche_ is now known as b4epoche
1083 2012-02-06 14:50:56 baz has joined
1084 2012-02-06 14:51:36 att has joined
1085 2012-02-06 14:55:42 iocor has quit (Ping timeout: 240 seconds)
1086 2012-02-06 14:56:03 iocor_ has joined
1087 2012-02-06 14:57:30 gp5st has joined
1088 2012-02-06 14:59:13 vigilyn has quit (Quit: Leaving)
1089 2012-02-06 15:05:35 iocor_ has quit (Ping timeout: 248 seconds)
1090 2012-02-06 15:09:36 iocor has joined
1091 2012-02-06 15:09:36 iocor has quit (Changing host)
1092 2012-02-06 15:09:36 iocor has joined
1093 2012-02-06 15:10:40 copumpkin has joined
1094 2012-02-06 15:11:13 userhhj has quit (Read error: Connection reset by peer)
1095 2012-02-06 15:11:19 tower has quit (Quit: | ReactOS - The FOSS alternative to MS Windows! | http://www.reactos.org/ | join #ReactOS |)
1096 2012-02-06 15:12:53 userhf has joined
1097 2012-02-06 15:14:57 devrandom has quit (Remote host closed the connection)
1098 2012-02-06 15:15:51 <sipa> Joric: i see on stackexchange someone mentioning that after importing a key using pywallet, they should -rescan
1099 2012-02-06 15:17:12 dr_win has quit (Remote host closed the connection)
1100 2012-02-06 15:17:25 iocor has quit (Ping timeout: 248 seconds)
1101 2012-02-06 15:17:54 <sipa> is this true? if so, you can force bitcoin to do a rescan by itself, by touching the "bestblock" record
1102 2012-02-06 15:18:49 iocor has joined
1103 2012-02-06 15:21:11 devrandom has joined
1104 2012-02-06 15:21:59 booo has joined
1105 2012-02-06 15:25:06 wasabi has joined
1106 2012-02-06 15:25:20 denisx has joined
1107 2012-02-06 15:31:22 BTC_Bear is now known as hbrntng!~BTC_Bear@unaffiliated/btc-bear/x-5233302|BTC_Bear
1108 2012-02-06 15:33:12 iocor has quit (Ping timeout: 240 seconds)
1109 2012-02-06 15:44:40 erle- has quit (Quit: erle-)
1110 2012-02-06 15:45:34 tower has joined
1111 2012-02-06 15:52:58 pusle has joined
1112 2012-02-06 15:53:37 <Joric> sipa, frankly i have no idea ) i removed it from the readme
1113 2012-02-06 15:54:02 <Joric> i don't touch the bestblock record maybe jackjack does
1114 2012-02-06 15:55:36 <Joric> what kind of touching is needed by the way? )
1115 2012-02-06 15:55:59 <sipa> reset it to the genesis block, for example
1116 2012-02-06 15:56:07 <sipa> that will make it rescan from the genesis block :)
1117 2012-02-06 15:56:17 erle- has joined
1118 2012-02-06 15:57:06 <Moron__>   :/
1119 2012-02-06 16:01:20 BTC_Bear is now known as BTC_Bear|hbrntng
1120 2012-02-06 16:01:23 <gmaxwell> gavinandresen: Think CAddrMan is hopeless out of the window for .6 now? I'd also like to see it get in sooner rather than later.
1121 2012-02-06 16:01:43 <sipa> i haven't found the problem with it yet
1122 2012-02-06 16:01:49 JRWR has joined
1123 2012-02-06 16:02:03 <sipa> valgrind shows me an error comparable to the stacktrace gavin gave
1124 2012-02-06 16:02:05 <luke-jr> gmaxwell: it has some kind of corruption bug right now, sipa is looking into ^
1125 2012-02-06 16:02:06 <gavinandresen> gmaxwell: it's in the "kinda big-scary change to get in at the last second" category....
1126 2012-02-06 16:02:24 <sipa> but valgrind shows the same problem with main
1127 2012-02-06 16:02:28 <gavinandresen> and the "it ain't completely broke right now" category
1128 2012-02-06 16:02:31 <gmaxwell> sipa: well thats great news then. oh but thats not great news.
1129 2012-02-06 16:02:45 <sipa> so i have no idea where to look
1130 2012-02-06 16:03:02 <gavinandresen> sipa: you're getting a valgrind warning with git HEAD?
1131 2012-02-06 16:03:39 <sipa> yes
1132 2012-02-06 16:03:39 <gmaxwell> If you don't modify your openssl bitcoin will be pure valgrind errors, though I assume you know that.
1133 2012-02-06 16:03:43 <sipa> wel, main-before-addrman
1134 2012-02-06 16:03:50 <sipa> but it's inside bdb
1135 2012-02-06 16:03:57 <sipa> so it may well be a spurious warning
1136 2012-02-06 16:04:09 <gavinandresen> you have to config bdb with --enable-umrw
1137 2012-02-06 16:04:21 <gavinandresen> (it is spurious)
1138 2012-02-06 16:04:31 <gmaxwell> That bdb one is old.
1139 2012-02-06 16:04:51 <gavinandresen> bdb and openssl both have to be compiled to be valgrind-clean
1140 2012-02-06 16:05:16 <gmaxwell> (it seemed harmless when I looked at it— at least before it only ever fired once per execution so it was easy to ignore)
1141 2012-02-06 16:05:55 <gavinandresen> (I left three bitcoind's running under valgrind last Thurdsday night with the BIP 16 transaction fuzzer doing it's thing, all was clean)
1142 2012-02-06 16:05:59 <sipa> strange, i've ran bitcoin in valgrind before, didn't see it back then
1143 2012-02-06 16:06:45 baz has quit (Quit: Leaving)
1144 2012-02-06 16:09:15 <gmaxwell> gavinandresen: Our custom allocator probably greatly reduces valgrind's sensitivity. It really needs a VALGRIND_MAKE_MEM_UNDEFINED(ptr,size) e.g. in secure allocator after the memset.
1145 2012-02-06 16:09:46 <gavinandresen> Inside an #ifdef VALGRIND ?
1146 2012-02-06 16:10:06 <gmaxwell> Yes.
1147 2012-02-06 16:10:08 <gmaxwell> also
1148 2012-02-06 16:10:08 <gmaxwell> #include <valgrind/memcheck.h>
1149 2012-02-06 16:10:17 <gmaxwell> (which defines those headers)
1150 2012-02-06 16:10:20 <gavinandresen> Sounds like a very good idea to me.
1151 2012-02-06 16:11:44 FROTUSCI has joined
1152 2012-02-06 16:13:22 Guest20048 is now known as ForceMajeure
1153 2012-02-06 16:16:28 <gmaxwell> (there is a way to just tell valgrind about your custom allocator on the commandline— but I never could figure out how to get that to work with C++ mangling)
1154 2012-02-06 16:16:36 bitcoinbulletin has quit (Remote host closed the connection)
1155 2012-02-06 16:18:39 <gavinandresen> Do the "-nofoo" options bother anybody else?  I hate double-negatives; I'm thinking of modifying GetBoolArg() so it looks for the positive version (e.g. -irc=0/1) but if not set automatically also looks for -no (e.g. -noirc means -irc=0 if irc isn't set).  And changing all of the "-nofoo" to GetBoolArg("-foo", true)
1156 2012-02-06 16:18:44 erle- has quit (Ping timeout: 272 seconds)
1157 2012-02-06 16:19:32 <JFK911> it should be -with-noirc
1158 2012-02-06 16:20:00 OneFixt has quit (Ping timeout: 272 seconds)
1159 2012-02-06 16:20:01 <gmaxwell> gavinandresen: thats what my makeIRC suck less patch does.
1160 2012-02-06 16:20:57 <gavinandresen> gmaxwell: what's the state of that patch?
1161 2012-02-06 16:21:38 <Joric> what it would support? anything from -irc -noirc -irc=0 -irc=1 ?
1162 2012-02-06 16:22:01 <gavinandresen> -irc -noirc -irc=0 -irc=1 -noirc=0 -noirc=1
1163 2012-02-06 16:22:03 <Joric> -irc=2 :)
1164 2012-02-06 16:22:19 <gavinandresen> yes, 2/3/4/foo/ircsucks/....
1165 2012-02-06 16:22:38 da2ce7 has quit (2!~da2ce7@gateway/tor-sasl/da2ce7|Ping timeout: 276 seconds)
1166 2012-02-06 16:24:03 <nanotube> it should be --, dammit!
1167 2012-02-06 16:24:11 <nanotube> :)
1168 2012-02-06 16:24:22 <gavinandresen> Right, I was thinking of changing that, too...
1169 2012-02-06 16:24:27 <Joric> yeah how about double dashes? single dash is pretty unusial for long keys
1170 2012-02-06 16:24:45 <gavinandresen> ^changing^supporting
1171 2012-02-06 16:24:46 <nanotube> well, probably to start with, just get - and -- to work, for backwards compat.
1172 2012-02-06 16:24:50 <nanotube> ya
1173 2012-02-06 16:24:57 <gavinandresen> ya
1174 2012-02-06 16:25:17 <gmaxwell> gavinandresen: I've been testing the current version for a while... I think it's fine. I haven't submitted it because I wanted to sit down and review it with a clean mind.
1175 2012-02-06 16:25:25 <gmaxwell> gavinandresen: https://people.xiph.org/~greg/irc-suckless.patch
1176 2012-02-06 16:25:35 att has quit (Ping timeout: 248 seconds)
1177 2012-02-06 16:26:19 <gmaxwell> Basically it turns off irc, and if IRC is on it improves the behavior to make IRC less partitioning by, if the node has incomining connections, joining 4 channels and jumping randomly between them every once in a while.
1178 2012-02-06 16:26:56 bitcoinbulletin has joined
1179 2012-02-06 16:27:32 cryptoxchange has quit (Read error: Connection reset by peer)
1180 2012-02-06 16:27:39 <gmaxwell> (sorry I'm slow with actual code— I'm a competent C coder, in C++ I feel like I'm a hazard)
1181 2012-02-06 16:29:04 cryptoxchange has joined
1182 2012-02-06 16:29:19 <Moron__> gmaxwell: what irc channel is used to find peers by the bitcoin client?
1183 2012-02-06 16:30:59 <gmaxwell> #bitcoin00 - #bitcoin99 on lfnet.
1184 2012-02-06 16:31:24 Joric has quit (Ping timeout: 272 seconds)
1185 2012-02-06 16:31:27 m00p has quit (Ping timeout: 248 seconds)
1186 2012-02-06 16:33:00 <Moron__> does the client use a strange wierd protocol or can I go in the channel and understand what theyre saying?
1187 2012-02-06 16:33:09 <gmaxwell> They don't say anything at all.
1188 2012-02-06 16:33:31 <Moron__> oh
1189 2012-02-06 16:33:38 <Moron__> how do the clients communicate?
1190 2012-02-06 16:33:41 topace_ has quit (Read error: Connection reset by peer)
1191 2012-02-06 16:33:46 <nanotube> gmaxwell: implement a c client! :)
1192 2012-02-06 16:33:46 <nanotube> Moron__: node ips are encoded in nicks
1193 2012-02-06 16:33:46 <nanotube> you do a /who and see who's on
1194 2012-02-06 16:33:52 <nanotube> then pick some nodes to try to connect to
1195 2012-02-06 16:34:03 <Moron__> oh clever
1196 2012-02-06 16:34:23 <gmaxwell> nanotube: I don't disagree with the choice of C++ in bitcoin. I'm just not very C++ clueful— and probably never will be. :)
1197 2012-02-06 16:34:57 Joric has joined
1198 2012-02-06 16:34:58 Joric has quit (Changing host)
1199 2012-02-06 16:34:58 Joric has joined
1200 2012-02-06 16:35:14 <denisx> a little more freebsd love would be nice
1201 2012-02-06 16:36:01 <Moron__> C++ is just basically C with a little object orientation through classes gmaxwell
1202 2012-02-06 16:36:06 <denisx> like adding netinet/in.h and sys/socket.h to protocol.cpp
1203 2012-02-06 16:36:06 <gmaxwell> hahahaha
1204 2012-02-06 16:36:23 SomeoneWeird is now known as SomeoneWeirdzzzz
1205 2012-02-06 16:36:39 Diablo-D3 has joined
1206 2012-02-06 16:36:40 OneFixt has joined
1207 2012-02-06 16:37:07 <gmaxwell> Moron__: thats like saying that C is some syntatic sugar on an abstract Turing machine.
1208 2012-02-06 16:37:22 beppu has quit (Read error: Operation timed out)
1209 2012-02-06 16:37:24 niggaholdthat has quit (Read error: Operation timed out)
1210 2012-02-06 16:37:28 beppu has joined
1211 2012-02-06 16:37:30 niggaholdthat has joined
1212 2012-02-06 16:37:46 <nanotube> gmaxwell: not saying cpp is bad, just saying that it would be nice to have more alt clients from people who know what's up, as a general point of desirability.
1213 2012-02-06 16:40:39 <lianj> well, parsing and signature verification is easy. i find it hard to implement the blockhain rules and script runner. or lets say, to keep up with it
1214 2012-02-06 16:40:53 Nicksasa has quit (Read error: Connection reset by peer)
1215 2012-02-06 16:41:03 MrTiggr has quit (Ping timeout: 248 seconds)
1216 2012-02-06 16:41:33 Nicksasa has joined
1217 2012-02-06 16:41:43 <Moron__> i might try to implement a basic bitcoin client using brainfuck
1218 2012-02-06 16:41:49 * luke-jr kicks self
1219 2012-02-06 16:42:17 <luke-jr> sipa: I'm sorry, I stupidly just deleted the corrupt addr.dat again x.x
1220 2012-02-06 16:42:32 <Joric> i'd rather do a basic brainfuck using bitcoin client
1221 2012-02-06 16:43:42 <Moron__> lol now iz confused
1222 2012-02-06 16:44:43 osmosis has quit (Quit: Leaving)
1223 2012-02-06 16:45:24 Prattler has quit (Remote host closed the connection)
1224 2012-02-06 16:49:09 <luke-jr> gavinandresen: btw, I am working on a minor tweak to the gitian descriptors to make it significantly faster to build new win32 binaries (by not rebuilding the deps every time). just thought I'd let you know in case you want to wait for that.
1225 2012-02-06 16:50:14 <gavinandresen> luke-jr: cool. Not in the "critical enough to wait for" category
1226 2012-02-06 16:50:37 <luke-jr> yeah, just thought you might want it to make the job easier
1227 2012-02-06 16:55:12 echelon has joined
1228 2012-02-06 17:04:34 BlueMatt has joined
1229 2012-02-06 17:09:40 Joric has quit ()
1230 2012-02-06 17:11:25 btc_novice has joined
1231 2012-02-06 17:24:03 booo has quit (Ping timeout: 245 seconds)
1232 2012-02-06 17:25:38 <userhf> hi, anyone know how many address with non zero btc exist?
1233 2012-02-06 17:25:50 <gmaxwell> Roughtly a half million.
1234 2012-02-06 17:26:30 <userhf> ok, thanks
1235 2012-02-06 17:27:45 BTC_Bear is now known as hbrntng!~BTC_Bear@unaffiliated/btc-bear/x-5233302|BTC_Bear
1236 2012-02-06 17:28:18 <BlueMatt> since when can you spend a dup coinbase twice? I could have sworn that was not allowed...
1237 2012-02-06 17:29:21 <BlueMatt> ;;seen roconnor
1238 2012-02-06 17:29:21 <gribble> roconnor was last seen in #bitcoin-dev 1 day, 2 hours, 37 minutes, and 12 seconds ago: <roconnor> sipa, gmaxwell: I think I know how to turn 1000 confirmation coins back into 1 confirmation coins using duplicate transactions.
1239 2012-02-06 17:31:21 <gmaxwell> BlueMatt: MineA SpendA MineA SpendA works, demonstrated on testnet. (Also not surprising)
1240 2012-02-06 17:31:40 <gmaxwell> BlueMatt: MineA MineA SpendA SpendA would not work (latter spenda would fail)
1241 2012-02-06 17:31:46 <BlueMatt> ok, yea
1242 2012-02-06 17:32:24 <kinlo> that's kinda dangerous...
1243 2012-02-06 17:32:35 <BlueMatt> not really
1244 2012-02-06 17:32:38 <kinlo> that makes things very implementation specific
1245 2012-02-06 17:32:39 <luke-jr> and SpendA could have multiple outputs ;)
1246 2012-02-06 17:32:51 <gmaxwell> kinlo: Thats the _least_ of our implementation specific behavior.
1247 2012-02-06 17:32:54 <luke-jr> BlueMatt: yes, you can use this to cheaply fork the Satoshi clients
1248 2012-02-06 17:33:11 <gmaxwell> Denying the second would slightly defeat pruning.
1249 2012-02-06 17:33:12 <BlueMatt> luke-jr: how?
1250 2012-02-06 17:33:25 <gmaxwell> BlueMatt: you construct a reorg that switches between the two cases basically.
1251 2012-02-06 17:33:31 <kinlo> it will require everyone willing to use the blockchain directly to also include that behavious
1252 2012-02-06 17:33:34 <kinlo> r
1253 2012-02-06 17:33:41 <luke-jr> BlueMatt: because reorgs don't handle it sanely
1254 2012-02-06 17:33:44 <gmaxwell> BlueMatt: it _appears_ that the reorg code will not restore the spendability of A when it needs to.
1255 2012-02-06 17:33:51 <BlueMatt> well yea, but then you have to mine your ass off
1256 2012-02-06 17:34:07 <luke-jr> BlueMatt: roconnor had a way to do it by mining 2 blocks over an unlimited time period
1257 2012-02-06 17:34:10 <gmaxwell> So, if you do this using coinbases it requires 100 block rorgs, yes.
1258 2012-02-06 17:34:30 <gmaxwell> But if you make duplicate _transactions_ using the duplicate coinbases, I think you can do it with a two block reorg. :(
1259 2012-02-06 17:34:31 <luke-jr> using transactions
1260 2012-02-06 17:34:45 <luke-jr> gmaxwell: only 1, I thought
1261 2012-02-06 17:34:52 <BlueMatt> oh, yea...
1262 2012-02-06 17:34:54 <gmaxwell> or one, I actually haven't walked through it carefully.
1263 2012-02-06 17:38:12 luke-jr has quit (Excess Flood)
1264 2012-02-06 17:38:18 <gmaxwell> Totally OT— it seems someone actually performed a network time attack against that totally crazy LQC coin and caused users a lot of trouble.
1265 2012-02-06 17:38:30 luke-jr has joined
1266 2012-02-06 17:38:36 <gmaxwell> Of course, since everyone involved is stupid they didn't do so much as gather useful data about the attack.
1267 2012-02-06 17:38:44 <luke-jr> [12:29:07] <luke-jr> I think if you split the outputs up, you can even make both branches of the fork invalid
1268 2012-02-06 17:38:46 <luke-jr> [12:29:12] <luke-jr> so new clients create a 3rd fork
1269 2012-02-06 17:39:09 <gmaxwell> I sounds like someone spun up a ton of nodes giving crazy timestamps in a sybil attack to cause nodes to misbehave.
1270 2012-02-06 17:39:17 <BlueMatt> so in other words, someone needs to spend some time with reorg code...
1271 2012-02-06 17:39:44 <BlueMatt> so in other words, someone needs to spend some time with the network time code
1272 2012-02-06 17:40:04 <BlueMatt> ;)
1273 2012-02-06 17:40:43 <gmaxwell> For network time code— I'm contemplating submiting a pull request for code I've been running awhile which basically amounts to -ignorenetworktime or -fuckyoumyclockisrightdamnit
1274 2012-02-06 17:40:55 * luke-jr is scared of touching reorg code
1275 2012-02-06 17:41:05 <luke-jr> gmaxwell: lol
1276 2012-02-06 17:41:09 <gmaxwell> BlueMatt: wrt reorg, I think we should discourage blocks with duplicate coinbases in the short term.
1277 2012-02-06 17:41:15 <gavinandresen> gmaxwell: ++
1278 2012-02-06 17:41:39 <gmaxwell> gavinandresen: was that regarding time or reorg?
1279 2012-02-06 17:41:44 <gavinandresen> both
1280 2012-02-06 17:42:56 <gmaxwell> BlueMatt: long term fixes are: prevent dupes entirely which paritally breaks pruning (can't prune them completely), or fix reorg.
1281 2012-02-06 17:43:32 <gmaxwell> The latter doesn't break pruning, but it may expose external software that assumes txn ids are unique to exploitation.
1282 2012-02-06 17:43:55 sje has quit (Remote host closed the connection)
1283 2012-02-06 17:43:58 <BlueMatt> would be nice for critical nodes ie pools/merchants/etc, but for regular users you still need something smarter
1284 2012-02-06 17:44:01 <BlueMatt> and make it default in ubuntu packages since they all ntp when connecting to a network by default ;)
1285 2012-02-06 17:44:34 <BlueMatt> I dont think its worth a hard fork just to makeit easier for alt implementations to work right
1286 2012-02-06 17:44:45 <gmaxwell> BlueMatt: At least if someone starts actively attacking we can tell people to set -ignorenetworktime and the code for it is already deployed.
1287 2012-02-06 17:44:58 <BlueMatt> very much so
1288 2012-02-06 17:45:40 <gmaxwell> We _could_ default to -ignorenetworktime iff we query local NTP and it claims good health, but I feel a bit uneasy about that... even if thats where we go longterm I'd prefer a phased approach.
1289 2012-02-06 17:45:54 <BlueMatt> agreed
1290 2012-02-06 17:46:00 <BlueMatt> dont think it should ever be default
1291 2012-02-06 17:46:07 <BlueMatt> (I was joking about the ubuntu thing)
1292 2012-02-06 17:46:19 <gmaxwell> (Uneasy because it means "exploit the soft and chewy ntp infrastructure to exploit bitcoin", though timing attacks are mostly pretty narrow)
1293 2012-02-06 17:46:39 <luke-jr> gmaxwell: someone suggested define duplicate txnids to be the same txn
1294 2012-02-06 17:46:45 <luke-jr> gmaxwell: any reason that wouldn't work?
1295 2012-02-06 17:46:57 <BlueMatt> it would mean forking old clients
1296 2012-02-06 17:47:12 <luke-jr> well, could use a starting block number
1297 2012-02-06 17:47:18 <BlueMatt> yea, ofc
1298 2012-02-06 17:48:03 <luke-jr> IMO, identical-but-distinct transactions are asking for bugs
1299 2012-02-06 17:48:15 phantomfake has quit (Read error: Connection reset by peer)
1300 2012-02-06 17:48:29 <gmaxwell> luke-jr: breaks pruning.
1301 2012-02-06 17:48:30 <BlueMatt> i see it as easier to just fix the reorg code
1302 2012-02-06 17:49:04 <gmaxwell> because you'd have to remember a TXN ID forever to know it was a duplicate. We could only do that for coinbases and then it becomes nearly impossible to make duplicates.
1303 2012-02-06 17:49:15 <luke-jr> gmaxwell: why do you care if it's a duplicate?
1304 2012-02-06 17:49:33 <luke-jr> if they're the same txn
1305 2012-02-06 17:49:34 <gmaxwell> luke-jr: if it's defined to be the same transaction, and that transaction is already spent then you can't mine it.
1306 2012-02-06 17:49:49 <gmaxwell> so you need to remember that that transaction existed and was already spent.
1307 2012-02-06 17:50:07 <luke-jr> hmm
1308 2012-02-06 17:50:19 phantomfake has joined
1309 2012-02-06 17:51:00 <luke-jr> could include block hash in the txnid for coinbases iff we force all miners to update
1310 2012-02-06 17:51:28 <gmaxwell> luke-jr: yes, prev-hash rather.
1311 2012-02-06 17:51:33 <luke-jr> sure
1312 2012-02-06 17:51:41 <gmaxwell> Or just the height really.
1313 2012-02-06 17:51:43 <sipa> luke-jr: your corrupted addr.dat was caused by a crashed client, right?
1314 2012-02-06 17:51:54 <luke-jr> sipa: at least they occurred at the same time
1315 2012-02-06 17:52:12 <sipa> a stacktrace would be very useful
1316 2012-02-06 17:52:16 <luke-jr> sorry :<
1317 2012-02-06 17:52:34 <sipa> i ran an addrman enabled client here all night, every 5 minutes doing a ./bitcoind stop and restart
1318 2012-02-06 17:52:41 <luke-jr> actually, it was in gdb!
1319 2012-02-06 17:52:48 <sipa> (without problems)
1320 2012-02-06 17:53:01 <luke-jr> sipa: I think both times, it had been running well over 24 hours
1321 2012-02-06 17:53:01 HostFat has joined
1322 2012-02-06 17:53:09 <sipa> ah
1323 2012-02-06 17:53:25 <luke-jr> looks like it was killed by a SIGABRT
1324 2012-02-06 17:53:35 <luke-jr> unfortuanteyl, I forgot I had gdb on it and kill -9'd it
1325 2012-02-06 17:53:36 <BlueMatt> forcing coinbase seems wrong...
1326 2012-02-06 17:54:21 <luke-jr> #0  0x00007ffff5a71855 in raise () from /lib64/libc.so.6
1327 2012-02-06 17:54:30 <luke-jr> ^ suggests C++ exception raised
1328 2012-02-06 17:54:44 <echelon> why isn't my bitcoin client listening on port 8333?
1329 2012-02-06 17:54:53 <echelon> instead it's listening on 8332
1330 2012-02-06 17:54:54 <luke-jr> otoh, ⁂ glibc detected ⁂ /home/bitcoinpool/src/bitcoind-pool/src/bitcoind: free(): invalid next size (fast): 0x0000000003b96d20 ⁂
1331 2012-02-06 17:55:00 <BlueMatt> echelon: 8332 == rpc
1332 2012-02-06 17:55:02 smickles has joined
1333 2012-02-06 17:55:04 <luke-jr> echelon: bitcoin.conf or CLI options?
1334 2012-02-06 17:55:16 <echelon> bitcoin.conf
1335 2012-02-06 17:55:36 <echelon> i just set the rcpuser/pass server=1, added some fallback nodes
1336 2012-02-06 17:55:49 <gmaxwell> gavinandresen: hmph. We should avoid learning or relaying addresses which are currently setBanned.
1337 2012-02-06 17:55:55 <echelon> also, noirc=1
1338 2012-02-06 17:55:57 <gmaxwell> (and probably totally evict them in addrman)
1339 2012-02-06 17:56:14 <gavinandresen> gmaxwell: good idea
1340 2012-02-06 17:56:16 ovidiusoft has quit (Ping timeout: 272 seconds)
1341 2012-02-06 17:56:23 <BlueMatt> echelon: added fallback nodes as addnode?
1342 2012-02-06 17:56:34 <echelon> no, using connect
1343 2012-02-06 17:56:35 <smickles> would i be able to get a touch of help with a pushpool issue from here or directions to where a good place to ask is?
1344 2012-02-06 17:56:39 barmstrong has quit (Remote host closed the connection)
1345 2012-02-06 17:56:47 <luke-jr> sipa: http://pastebin.com/wx7yFTxF
1346 2012-02-06 17:56:49 <BlueMatt> echelon: yuck, dont do that
1347 2012-02-06 17:56:55 <echelon> why
1348 2012-02-06 17:56:59 <luke-jr> echelon: connect disables any other connection
1349 2012-02-06 17:57:02 <luke-jr> ie, listening
1350 2012-02-06 17:57:11 <BlueMatt> it doesnt disable listening
1351 2012-02-06 17:57:14 <echelon> luke-jr: that's what i want
1352 2012-02-06 17:57:17 <BlueMatt> but it disables other outgoing connections
1353 2012-02-06 17:57:21 <luke-jr> smickles: you can ask, but I don't think many people use pushpool by now
1354 2012-02-06 17:57:23 <BlueMatt> and thats not what you want
1355 2012-02-06 17:57:32 <echelon> i'm running a hidden service seed node
1356 2012-02-06 17:57:47 <BlueMatt> make normal outgoing connections
1357 2012-02-06 17:57:59 <echelon> so connect disables listening?
1358 2012-02-06 17:58:03 <BlueMatt> no
1359 2012-02-06 17:58:13 <BlueMatt> or, it didnt last time I checked
1360 2012-02-06 17:58:22 <smickles> luke-jr: ok, you seem to know what you are doing sometimes; whats a good alternative to pushpool?
1361 2012-02-06 17:58:31 <gmaxwell> fUseProxy probably does though.
1362 2012-02-06 17:58:37 <gmaxwell> (Because we're stupid)
1363 2012-02-06 17:59:38 <BlueMatt> we need a -backupnode option
1364 2012-02-06 17:59:50 <gmaxwell> Seems not.
1365 2012-02-06 18:00:06 <luke-jr> smickles: I wrote Eloipool specifically for that purpose
1366 2012-02-06 18:00:07 <BlueMatt> (because I have a feeling addnode=every backup node on the wiki is all too common
1367 2012-02-06 18:00:08 <BlueMatt> )
1368 2012-02-06 18:00:10 <echelon> gmaxwell: does witcoin still have our balance?
1369 2012-02-06 18:00:45 <gmaxwell> echelon: got me.
1370 2012-02-06 18:01:49 <smickles> luke-jr: ooh, agpl, i like that :)
1371 2012-02-06 18:02:44 <gmaxwell> echelon: if you've set things via the gui then it will also remember those setting and they are default.
1372 2012-02-06 18:02:52 <Diablo-D3> okay so
1373 2012-02-06 18:03:00 <gmaxwell> echelon: so, e.g. setting nolisten in the gui (I presume there is a setting for that?)
1374 2012-02-06 18:03:03 <Diablo-D3> can anyone actually tell me who came in second in arizona?
1375 2012-02-06 18:03:25 <Diablo-D3> oh wait
1376 2012-02-06 18:03:26 <Diablo-D3> hahah
1377 2012-02-06 18:03:27 <Diablo-D3> I forgot
1378 2012-02-06 18:03:30 <Diablo-D3> I can just ask google now
1379 2012-02-06 18:03:57 <Diablo-D3> gingrich and ron paul basically tied
1380 2012-02-06 18:04:01 <Diablo-D3> for 2nd
1381 2012-02-06 18:04:10 <BlueMatt> gmaxwell: nolisten is not in the gui, only proxy is
1382 2012-02-06 18:04:11 <echelon> gmaxwell: nah, i don't recall any such thing
1383 2012-02-06 18:04:19 <BlueMatt> but proxy will set nolisten
1384 2012-02-06 18:04:25 <BlueMatt> (ie tor proxy will disable listen)
1385 2012-02-06 18:04:28 <echelon> yeah, i'm using a proxy
1386 2012-02-06 18:04:32 <echelon> what, why
1387 2012-02-06 18:04:42 <BlueMatt> potential identity leak otherwise
1388 2012-02-06 18:04:43 <Diablo-D3> hrm
1389 2012-02-06 18:04:49 <Diablo-D3> bitcoin should support listen over tor
1390 2012-02-06 18:04:58 <echelon> but i'm trying to run a hidden service
1391 2012-02-06 18:05:03 <Diablo-D3> yeah what echelon said
1392 2012-02-06 18:05:07 <Diablo-D3> theres only a small problem
1393 2012-02-06 18:05:11 <BlueMatt> yes, it should support that
1394 2012-02-06 18:05:15 <Diablo-D3> bitcoin stores IPs
1395 2012-02-06 18:05:17 <Diablo-D3> not names
1396 2012-02-06 18:05:43 <gmaxwell> echelon: you really need phantomcircuit's tor stuff to properly work as a hidden service.
1397 2012-02-06 18:05:53 <echelon> hmm
1398 2012-02-06 18:06:04 <BlueMatt> onioncat needs to be merged upstream into tor so we can just rely on that and not have to worry about anything else
1399 2012-02-06 18:06:09 <echelon> can't i just do.. nolisten=0 or something?
1400 2012-02-06 18:06:12 <gmaxwell> Diablo-D3: we store IPv6 IPs in fact, so problem solved. :) see phantomcircuit's code
1401 2012-02-06 18:06:16 <phantomcircuit> still waiting on sipas address manager to be integrated
1402 2012-02-06 18:06:21 MobiusL has quit (Remote host closed the connection)
1403 2012-02-06 18:06:26 <phantomcircuit> as soon as that's done i'll be able to finish up
1404 2012-02-06 18:06:30 <BlueMatt> phantomcircuit: what stuff did you add?
1405 2012-02-06 18:06:44 <Diablo-D3> gmaxwell: no
1406 2012-02-06 18:06:50 <phantomcircuit> BlueMatt, .onion <-> ipv6 mapper
1407 2012-02-06 18:06:51 <gmaxwell> Diablo-D3: Yup.
1408 2012-02-06 18:06:51 <Diablo-D3> gmaxwell: because tor doesnt use IPs
1409 2012-02-06 18:06:58 <BlueMatt> phantomcircuit: ie onioncat?
1410 2012-02-06 18:06:58 <phantomcircuit> more or less
1411 2012-02-06 18:07:02 <gmaxwell> Diablo-D3: I know. MAGIC.
1412 2012-02-06 18:07:10 <Diablo-D3> well what Im saying is
1413 2012-02-06 18:07:12 <echelon> bitcoin has ipv6 support?
1414 2012-02-06 18:07:16 <Diablo-D3> if it thinks all nodes are 127.0.0.1 due to tor
1415 2012-02-06 18:07:18 <gmaxwell> Diablo-D3: I know. And you are wrong. Because MAGIC.
1416 2012-02-06 18:07:18 <BlueMatt> echelon: no
1417 2012-02-06 18:07:19 <Diablo-D3> we're kind of fucked
1418 2012-02-06 18:07:27 <echelon> hrm
1419 2012-02-06 18:07:29 <Diablo-D3> gmaxwell: magic my overgrown dick.
1420 2012-02-06 18:07:50 <BlueMatt> phantomcircuit: more or less, so what, our own onioncat-like address creation?
1421 2012-02-06 18:07:50 MobiusL has joined
1422 2012-02-06 18:07:56 <BlueMatt> that just seems wrong
1423 2012-02-06 18:08:05 <gmaxwell> Diablo-D3: tor hidden service identifiers are only 80 bits long, they can be packed into IPv6 addresses.. thats what we're doing.
1424 2012-02-06 18:08:15 <Diablo-D3> gmaxwell: _lawlz_
1425 2012-02-06 18:08:18 <Diablo-D3> thats clever
1426 2012-02-06 18:08:23 <phantomcircuit> BlueMatt, onioncat actually has broken base32 decoding, which is pretty hilarious
1427 2012-02-06 18:08:27 <BlueMatt> what prefix do we use?
1428 2012-02-06 18:08:37 <phantomcircuit> BlueMatt, so the range we map to is off by 1 bit from the range they map to
1429 2012-02-06 18:08:52 <BlueMatt> are you kidding me?
1430 2012-02-06 18:08:55 <Diablo-D3> so they start at 1/
1431 2012-02-06 18:08:57 <phantomcircuit> not sure if we should break the encode/decode in the same way or not
1432 2012-02-06 18:09:03 <phantomcircuit> BlueMatt, im not
1433 2012-02-06 18:09:06 <BlueMatt> yes, we should
1434 2012-02-06 18:09:11 <Diablo-D3> yes you should
1435 2012-02-06 18:09:12 <Diablo-D3> ooh ooh
1436 2012-02-06 18:09:17 <Diablo-D3> how do I set this up?
1437 2012-02-06 18:09:19 <sipa> phantomcircuit: i'd try to be compatible with onioncat
1438 2012-02-06 18:09:24 <gmaxwell> I don't know that being compatible with onioncat matters, but if we're not we should move into a seperate linklocal (or whatever unroutable) block.
1439 2012-02-06 18:09:34 <gmaxwell> if we're in the same block we should be compatible for sure.
1440 2012-02-06 18:09:40 <phantomcircuit> gmaxwell, i did that actually
1441 2012-02-06 18:09:48 <phantomcircuit> incremented the prefix by 1
1442 2012-02-06 18:09:50 <BlueMatt> if I have an ipv6 node and onioncat enabled on my system, will an ipv6 bitcoin be able to just connect to onioncat w/o any tor stuff in bitcoin?
1443 2012-02-06 18:09:59 <Diablo-D3> hey guys, how do I turn bitcoin's tor mode on?
1444 2012-02-06 18:10:00 <gmaxwell> sipa: some crazy person might want to combine onioncat and bitcoin-hs-mode.. ::shrugs::
1445 2012-02-06 18:10:16 <BlueMatt> phantomcircuit: so you used a prefix that isnt ours?
1446 2012-02-06 18:10:16 james has joined
1447 2012-02-06 18:10:19 <sipa> Diablo-D3: tor mode, or onioncat?
1448 2012-02-06 18:10:21 <Diablo-D3> because I'd like to support incoming connections over tor
1449 2012-02-06 18:10:24 <echelon> why not just bind to localhost when a proxy is set instead of just disabling listen entirely
1450 2012-02-06 18:10:33 <sipa> Diablo-D3: you need phantomcircuit's stuff for that
1451 2012-02-06 18:10:37 <phantomcircuit> BlueMatt, it's linklocal stuff
1452 2012-02-06 18:10:38 <sipa> Diablo-D3: but nobody will connect to you
1453 2012-02-06 18:10:42 james is now known as Guest599
1454 2012-02-06 18:10:43 <Diablo-D3> oh =/
1455 2012-02-06 18:10:46 <Diablo-D3> dont want then
1456 2012-02-06 18:10:47 <sipa> for now at least
1457 2012-02-06 18:10:48 <gmaxwell> Onioncat itself is pretty awful and scary, and I'm disinclined to support it.
1458 2012-02-06 18:10:57 <BlueMatt> gmaxwell: how is it done?
1459 2012-02-06 18:11:07 <gmaxwell> BlueMatt: ip in tcp.
1460 2012-02-06 18:11:09 <Diablo-D3> lol
1461 2012-02-06 18:11:10 <phantomcircuit> the important part is that addresses are propagated throughout the network
1462 2012-02-06 18:11:11 <BlueMatt> I was under the impression they got their own prefix
1463 2012-02-06 18:11:11 <Diablo-D3> theres a garlicat too
1464 2012-02-06 18:11:12 <sipa> right, we're just reusing their address mapping, not actually using onioncat
1465 2012-02-06 18:11:16 <BlueMatt> gmaxwell: oh god...
1466 2012-02-06 18:11:31 <phantomcircuit> currently you can connect and receive connections but won't be able to send through the network
1467 2012-02-06 18:11:34 <Ukyo> there is garlic ?
1468 2012-02-06 18:11:37 <Ukyo> sounds yummy :)
1469 2012-02-06 18:11:40 phantomfake has quit (Quit: KVIrc 4.0.4 Insomnia http://www.kvirc.net/)
1470 2012-02-06 18:11:59 <BlueMatt> phantomcircuit: well in any case, you shouldnt use a link-local prefix in case a non-tor bitcoin client shows up and starts trying to connect to local addresses
1471 2012-02-06 18:11:59 <Diablo-D3> well, bitcoin needs to learn how to connect to both tor and non-tor nodes
1472 2012-02-06 18:12:04 <gmaxwell> BlueMatt: they're reusing one of many link-local blocks that should be minimally harmful, I don't think anything was formally advised.
1473 2012-02-06 18:12:25 <Diablo-D3> so I can like provide a .onion address, and let people connect to me through tor
1474 2012-02-06 18:12:30 <gmaxwell> BlueMatt: we're basically special casing these IPv6 addresses to mean something particular to bitcoin "use tor".
1475 2012-02-06 18:12:31 <Diablo-D3> but I can also connect to both tor and non-tor nodes
1476 2012-02-06 18:12:40 <gmaxwell> Diablo-D3: yes, we'll support that eventually.
1477 2012-02-06 18:12:42 <phantomcircuit> BlueMatt, it's the same strategy that onioncat is using
1478 2012-02-06 18:13:05 <Ukyo> hmm, if you go out of the way to ensure bitcoin client supports tor specially, or at least make a big deal that it 'specially' supports tor, it could hurt butcoins images in the eyes of the masses :)
1479 2012-02-06 18:13:11 <echelon> or "run in tor hidden service mode"
1480 2012-02-06 18:13:27 <echelon> which will bind the listen to localhost
1481 2012-02-06 18:13:29 <Ukyo> bitcoin already has a bad enough rep with SL
1482 2012-02-06 18:13:31 <gmaxwell> Ukyo: we support people in iran who want to run bitcoin.
1483 2012-02-06 18:13:34 <gmaxwell> :)
1484 2012-02-06 18:13:38 <gmaxwell> s/we/I/
1485 2012-02-06 18:13:45 <Ukyo> lol
1486 2012-02-06 18:13:48 <sipa> phantomcircuit: did you change IsRoutable() to include the onioncat range?
1487 2012-02-06 18:13:51 <Ukyo> I support anyone who wants to run bitcoin
1488 2012-02-06 18:13:52 <BlueMatt> gmaxwell: and my point is we shouldnt special case address prefixes unless they are reserved for such uses
1489 2012-02-06 18:13:56 <BlueMatt> phantomcircuit: so?
1490 2012-02-06 18:14:01 <BlueMatt> they are doing it wrong, lets do it wrong too
1491 2012-02-06 18:14:04 <Ukyo> I am just stating public appearnce:)
1492 2012-02-06 18:14:17 <phantomcircuit> BlueMatt, there isn't a right way afaict
1493 2012-02-06 18:14:32 <BlueMatt> phantomcircuit: there arent reserved prefixes in ipv6?
1494 2012-02-06 18:14:33 <gmaxwell> BlueMatt: We're not going to manage to get a block of space reserved for that, using a chunk of link local is minimally harmful.. and these addresses won't be seen outside of the bitcoin system.
1495 2012-02-06 18:15:04 <gmaxwell> BlueMatt: We're using the part which is reserved for locally specific use. This is way better than using some reserved for future use segment.
1496 2012-02-06 18:15:19 <echelon> but regardless, instead of setting nolisten upon enabling proxy, it should've just set noirc
1497 2012-02-06 18:15:20 <BlueMatt> yea, but are there not more reserved blocks?
1498 2012-02-06 18:15:36 <gmaxwell> BlueMatt: there are damn near infinite reserved. But reserved for future use.
1499 2012-02-06 18:15:41 <sipa> there are tons or reserved ranges for various things
1500 2012-02-06 18:15:55 <sipa> but why use something else if someone already made a reasonable choice?
1501 2012-02-06 18:16:26 <gmaxwell> BlueMatt: of the existing reserved ranges, I believe the link local stuff is the most reasonable.. It's for addresses which are basically specific to a single wire. So it's not like you'll conflict with someone's campus network.
1502 2012-02-06 18:17:15 <BlueMatt> mmm, I was hoping there was some reserved crap for situations like this, I mean its not like we dont have enough ipv6 addresses for them to reserve for all kinds of crazy usage
1503 2012-02-06 18:17:32 <gmaxwell> and by using the same one as onioncat we don't make any baddness from their decision any worse.
1504 2012-02-06 18:17:47 <sipa> the range comes from this RFC: http://www.rfc-editor.org/rfc/rfc4193.txt
1505 2012-02-06 18:18:15 danbri has joined
1506 2012-02-06 18:18:18 <sipa> plus, link-local addresses are useless on the bitcoin network anyway
1507 2012-02-06 18:18:38 sgstair has joined
1508 2012-02-06 18:18:50 <gmaxwell> This is also internal to bitcoin only. When talking to the outside world bitcoin should always convert those addresses back to onion addresses.
1509 2012-02-06 18:19:09 <gmaxwell> (well internal to bitcoin the client and bitcoin the network)
1510 2012-02-06 18:19:10 <BlueMatt> yep, and other implementations now have to check each address they get...
1511 2012-02-06 18:19:16 <BlueMatt> though I suppose conencting to them will fail anyway
1512 2012-02-06 18:19:44 <gmaxwell> Right, it's harmless even if you're stupid.
1513 2012-02-06 18:19:47 <phantomcircuit> BlueMatt, they dont have to check the address will just fail
1514 2012-02-06 18:20:21 <BlueMatt> yea...
1515 2012-02-06 18:20:42 <phantomcircuit> not to mention a basic check for routable addresses will cause them to be excluded
1516 2012-02-06 18:20:51 <BlueMatt> yep
1517 2012-02-06 18:21:30 <sipa> by the way, we are already special-casing a certain range
1518 2012-02-06 18:22:00 <sipa> namely ipv4
1519 2012-02-06 18:22:13 <gmaxwell> Yes, but thats not our special casing. (unless we're doing it non-standardly)
1520 2012-02-06 18:22:21 <gmaxwell> All zeros then v4 right?
1521 2012-02-06 18:22:30 <sipa> no
1522 2012-02-06 18:22:34 <luke-jr> sipa: ::ffff:<IP4> is standard
1523 2012-02-06 18:22:57 <sipa> i didn't say it was non-standard
1524 2012-02-06 18:23:03 <sipa> only that it was a special-casing
1525 2012-02-06 18:23:10 MobiusL has quit (Remote host closed the connection)
1526 2012-02-06 18:23:40 <phantomcircuit> sipa, it's not special case at all
1527 2012-02-06 18:24:00 <phantomcircuit> it works by default on most linux systems if you open the socket as inet6
1528 2012-02-06 18:24:13 <gmaxwell> (you can, at least in theory, pass those directly to the local network stack and the right thing will happen)
1529 2012-02-06 18:24:16 <phantomcircuit> my python client actually uses inet6 sockets exclusively
1530 2012-02-06 18:24:17 <gmaxwell> right.
1531 2012-02-06 18:24:21 MobiusL has joined
1532 2012-02-06 18:24:23 <sipa> true, indeed
1533 2012-02-06 18:24:33 phantomfake has joined
1534 2012-02-06 18:24:58 <gmaxwell> sipa: does addrman have any efficient way to go evict all instances of an address?
1535 2012-02-06 18:25:08 <gmaxwell> (I think the blacklisting of an address should trigger that)
1536 2012-02-06 18:25:12 <sipa> gmaxwell: no
1537 2012-02-06 18:25:20 <gmaxwell> alas.
1538 2012-02-06 18:25:45 <gmaxwell> I guess skipping it is almost as good.
1539 2012-02-06 18:26:02 <sipa> O(k*log(n)), with k the number of unknown buckets, and n their log-average fill
1540 2012-02-06 18:26:27 <sipa> which isn't too bad
1541 2012-02-06 18:26:29 <phantomcircuit> sipa, btw i've been running your addrman branch and restarting bitcoind in a loop
1542 2012-02-06 18:26:38 <phantomcircuit> it doesn't seem to be having trouble
1543 2012-02-06 18:27:50 <phantomcircuit> potentially because it's running on a tmpfs
1544 2012-02-06 18:28:05 <gmaxwell> The node it was crashing on for luke I believe is running some patches that make the IRC behavior more agressive.
1545 2012-02-06 18:28:55 <luke-jr> gmaxwell: yes, at startup
1546 2012-02-06 18:29:04 <luke-jr> IIRC Gavin also reproduced, no?
1547 2012-02-06 18:29:34 <phantomcircuit> seems stable for me
1548 2012-02-06 18:29:48 <phantomcircuit> luke-jr, do you have a trace?
1549 2012-02-06 18:30:02 danbri has quit (Remote host closed the connection)
1550 2012-02-06 18:30:10 <luke-jr> phantomcircuit: no, just http://pastebin.com/wx7yFTxF
1551 2012-02-06 18:30:29 <luke-jr> also, both times was long after 24 hours uptime
1552 2012-02-06 18:30:42 <luke-jr> so it's no surprise to me that restarting a bunch won't reproduce
1553 2012-02-06 18:30:51 <gmaxwell> malloc pool corruption.
1554 2012-02-06 18:31:03 <gmaxwell> Probably a use after free.
1555 2012-02-06 18:31:42 <sipa> addrman doesn't do any manual memory-management
1556 2012-02-06 18:31:58 <gmaxwell> race condition on a non-thread-safe object?
1557 2012-02-06 18:32:31 FROTUSCI has quit (Read error: Connection reset by peer)
1558 2012-02-06 18:32:44 FROTUSCI has joined
1559 2012-02-06 18:33:17 Nick_ has joined
1560 2012-02-06 18:33:19 Nick_ has quit (Client Quit)
1561 2012-02-06 18:33:20 <gmaxwell> (and this is where my C++ incompetence comes into play, I have absolutely no clue what the normal thread safty semantics are of automatically memorymanaged objects are)
1562 2012-02-06 18:33:21 <sipa> all public methods grab a CS
1563 2012-02-06 18:33:24 <luke-jr> gmaxwell: the crash occurred within raise() apparently
1564 2012-02-06 18:33:24 <phantomcircuit> yeah that appears to not be related to addrman actually
1565 2012-02-06 18:33:36 <gmaxwell> sipa: ah...
1566 2012-02-06 18:33:59 <sipa> my guess is it exposed a previously-existing bug
1567 2012-02-06 18:34:01 <gmaxwell> phantomcircuit: well, if something is corrupting malloc's data structure you usually see the crash _someplace else_
1568 2012-02-06 18:34:01 Nicksasa has quit (Read error: No route to host)
1569 2012-02-06 18:34:29 <phantomcircuit> gmaxwell, addrman doesn't use malloc/free
1570 2012-02-06 18:34:35 <phantomcircuit> but there could be a bug somewhere else
1571 2012-02-06 18:34:54 <sipa> i'm currently unable to run a bitcoind for testing for a long time
1572 2012-02-06 18:35:09 <sipa> anyone can try to reproduce with core dumps turned on or so?
1573 2012-02-06 18:35:49 <gmaxwell> Reproducing inside valgrind may be more informative.
1574 2012-02-06 18:36:26 <sipa> even better
1575 2012-02-06 18:37:27 Nicksasa has joined
1576 2012-02-06 18:37:32 dfgh566 has joined
1577 2012-02-06 18:43:49 p0s has joined
1578 2012-02-06 18:51:14 Dyaheon has quit ()
1579 2012-02-06 18:53:03 phantomfake has quit (Read error: Connection reset by peer)
1580 2012-02-06 18:53:21 molecular has quit (Remote host closed the connection)
1581 2012-02-06 18:53:32 molecular has joined
1582 2012-02-06 18:54:34 barmstrong has joined
1583 2012-02-06 18:54:50 phantomfake has joined
1584 2012-02-06 18:59:08 b4epoche has quit (Read error: Operation timed out)
1585 2012-02-06 18:59:42 <p0s> any chance we see a wallet export/import function soon so once can merge wallets?
1586 2012-02-06 19:00:34 <helo> pywallet can be used to merge wallets afaik
1587 2012-02-06 19:00:34 b4epoche has joined
1588 2012-02-06 19:00:51 <sipa> i should get back to finishing that
1589 2012-02-06 19:00:57 Guest599 is now known as topace_
1590 2012-02-06 19:01:46 <p0s> helo: pywallet isn't part of the official bitcoin tar.gz yet, is it? i don't want to use third party tools on my wallet.
1591 2012-02-06 19:08:44 T_X has quit (Changing host)
1592 2012-02-06 19:08:44 T_X has joined
1593 2012-02-06 19:10:37 ovidiusoft has joined
1594 2012-02-06 19:10:52 MC1984 has quit (Read error: Connection reset by peer)
1595 2012-02-06 19:10:59 gwillen has joined
1596 2012-02-06 19:11:13 MC1984 has joined
1597 2012-02-06 19:11:51 TD[gone] has joined
1598 2012-02-06 19:12:18 mortikia has joined
1599 2012-02-06 19:12:51 dfgh566 has quit (Ping timeout: 245 seconds)
1600 2012-02-06 19:14:42 fahadsadah has joined
1601 2012-02-06 19:16:52 bitcoinbulletin has joined
1602 2012-02-06 19:20:45 pirateat40 has quit (Ping timeout: 276 seconds)
1603 2012-02-06 19:21:44 pingdrive has joined
1604 2012-02-06 19:23:41 cyberdo has joined
1605 2012-02-06 19:24:13 <helo> is there a good python tool to dig through a local dump of the blockchain? i.e. iterate over all blocks
1606 2012-02-06 19:26:24 phantomfake has quit (Quit: KVIrc 4.0.4 Insomnia http://www.kvirc.net/)
1607 2012-02-06 19:27:02 JimRogers has joined
1608 2012-02-06 19:31:42 <helo> bitcointools ftw
1609 2012-02-06 19:33:05 phantomfake has joined
1610 2012-02-06 19:42:06 freewil has quit (Read error: Connection reset by peer)
1611 2012-02-06 19:43:01 h4ckm3th32nd has joined
1612 2012-02-06 19:44:03 h4ckm3th32nd has quit (Client Quit)
1613 2012-02-06 19:46:16 iocor has joined
1614 2012-02-06 19:47:52 danbri has joined
1615 2012-02-06 19:48:49 hexTech has quit (Quit: -)
1616 2012-02-06 19:50:01 PK_ has joined
1617 2012-02-06 19:50:46 PK_ is now known as PK
1618 2012-02-06 19:50:59 erle- has joined
1619 2012-02-06 19:50:59 genjix has joined
1620 2012-02-06 19:51:32 abbe has quit (Quit: Aah IRC, where men are men, women are men)
1621 2012-02-06 19:52:17 `2Fast2BCn has joined
1622 2012-02-06 19:54:57 <gribble> New news from bitcoinrss: gavinandresen opened pull request 803 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/803>
1623 2012-02-06 19:55:54 <sipa> gmaxwell: i remember you used a node at luke-jr's to investigate some things, was it related to knowing whether default noirc would be fine?
1624 2012-02-06 19:57:59 blomqvist has joined
1625 2012-02-06 20:00:38 dikidera has quit (Ping timeout: 244 seconds)
1626 2012-02-06 20:04:47 <gmaxwell> sipa: Yes. making sure a totally unseen node would get incoming connections without any irc. It did.
1627 2012-02-06 20:05:38 <sipa> good, considering gavin's last pull
1628 2012-02-06 20:06:01 <gavinandresen> pull request... ACK's welcome
1629 2012-02-06 20:06:32 <luke-jr> gavinandresen: it looks good to me, but I didn't test it yet
1630 2012-02-06 20:06:34 <sipa> gavinandresen: i like the [no] prefixes
1631 2012-02-06 20:06:54 <luke-jr> gavinandresen: only thing I was uncertain of, is if -noupnp will be clear to people from the help
1632 2012-02-06 20:07:17 <gavinandresen> luke-jr: recommended form will be  -upnp=0
1633 2012-02-06 20:07:18 <sipa> will -irc=no work?
1634 2012-02-06 20:07:30 <luke-jr> hmm, will -noupnp=1 still work?
1635 2012-02-06 20:07:34 <luke-jr> ie, in config files
1636 2012-02-06 20:07:54 <gavinandresen> -irc=no  will atoi("no"),  and be same as -irc=0
1637 2012-02-06 20:08:09 <gavinandresen> -nounpn=1 will turn into -upnp=0
1638 2012-02-06 20:08:18 pingdrive has quit (Quit: Leaving)
1639 2012-02-06 20:08:19 <luke-jr> cool
1640 2012-02-06 20:08:44 [1]wirehead has joined
1641 2012-02-06 20:09:05 <genjix> luke-jr, gavinandresen, sipa: http://bitcoin.org/ <- how do you like the bitcoin media links i put there
1642 2012-02-06 20:09:09 wirehead has quit (Ping timeout: 248 seconds)
1643 2012-02-06 20:09:19 <gavinandresen> sipa:  only weird non-intuitive boolean case remains  -irc=true (atoi("true") becomes zero, so you get the opposite of what you want, same as before)
1644 2012-02-06 20:09:21 <genjix> i think they're informative and useful/non-spammy
1645 2012-02-06 20:09:39 <Moron__> oh my god, someones defaced the site
1646 2012-02-06 20:09:41 <Moron__> j/k
1647 2012-02-06 20:10:06 <sipa> genjix: i don't see anything?
1648 2012-02-06 20:10:18 <luke-jr> sipa: bottom of Resources
1649 2012-02-06 20:10:53 <sipa> oh, right
1650 2012-02-06 20:11:20 Turingi has joined
1651 2012-02-06 20:11:23 zgr has joined
1652 2012-02-06 20:11:30 <genjix> history, bullet list of advantages, and a merchant developer tutorial
1653 2012-02-06 20:11:32 <gavinandresen> genjix: I expect there will be an uproar from ... somebody ... about that
1654 2012-02-06 20:11:33 <luke-jr> genjix: I don't think it's good to link political stuff as "Bitcoin pre-history"; not everyone using (or considering using) bitcoin will agree with your politics
1655 2012-02-06 20:12:00 <gavinandresen> genjix: maybe Matthew Wright-- "why not link to the bitcoin magazine site" etc
1656 2012-02-06 20:12:01 <luke-jr> and linking it as such might turn people away
1657 2012-02-06 20:12:21 zgr has left ()
1658 2012-02-06 20:12:59 <XMPPwocky> woohoo
1659 2012-02-06 20:13:20 <XMPPwocky> I think I've got a complete protocol parser (except for the IP transaction stuff)
1660 2012-02-06 20:13:37 <sipa> XMPPwocky: ignore ip transactions
1661 2012-02-06 20:13:44 <gavinandresen> XMPPwocky: nice job.  It'll probably break on the 20'th....
1662 2012-02-06 20:14:31 <luke-jr> "Advantages" also has some issues; Bitcoin shouldn't be falsely advertised as deregulated or "free from control"; Bitcoin is not anonymous; for "Fast", I'd note that the 6 confirms is equivalent to the 6 months on a CC - otherwise, an hour isn't very fast
1663 2012-02-06 20:14:34 diki has joined
1664 2012-02-06 20:14:44 <luke-jr> genjix: also, using cusswords isn't professional
1665 2012-02-06 20:14:56 <XMPPwocky> sipa: yeah
1666 2012-02-06 20:15:04 <XMPPwocky> gavinandresen: I coded that in
1667 2012-02-06 20:15:09 <Moron__> i think more cusswords should be put in
1668 2012-02-06 20:15:25 <gavinandresen> XMPPwocky: Then just: nice!
1669 2012-02-06 20:15:25 booo has joined
1670 2012-02-06 20:15:27 <Moron__> otherwise its a reasonable job well done
1671 2012-02-06 20:15:37 <genjix> XMPPwocky: what language?
1672 2012-02-06 20:15:48 <XMPPwocky> python with Construct
1673 2012-02-06 20:15:53 <genjix> :o
1674 2012-02-06 20:16:00 <genjix> XMPPwocky: maybe try http://bitcoinmedia.com/libbitcoin-first-steps/
1675 2012-02-06 20:16:10 <XMPPwocky> http://xmppwocky.net/bitkit/.git
1676 2012-02-06 20:16:20 RazielZ has quit (Ping timeout: 255 seconds)
1677 2012-02-06 20:16:26 <sipa> luke-jr: i have no problem with a claim that bitcoin is free from control over its monetary policy
1678 2012-02-06 20:16:31 <sipa> free from anycontrol is another issue
1679 2012-02-06 20:16:32 <Moron__> genjix: is bitcoinmedia your site?
1680 2012-02-06 20:16:59 <XMPPwocky> genjix: this is mostly for a "directed fuzzer"
1681 2012-02-06 20:17:25 <genjix> Moron__: yep
1682 2012-02-06 20:17:30 <genjix> XMPPwocky: url 404s
1683 2012-02-06 20:17:33 <Moron__> oh nice design
1684 2012-02-06 20:17:36 <genjix> but this is funny: http://xmppwocky.net/libertarianslogans.html
1685 2012-02-06 20:17:56 <genjix> bitcoin libertarians can act a bit nutty like that
1686 2012-02-06 20:17:59 <XMPPwocky> genjix: git clone http://bitcoinmedia.com/libbitcoin-first-steps/
1687 2012-02-06 20:18:00 <XMPPwocky> er
1688 2012-02-06 20:18:11 <XMPPwocky> git clone  http://xmppwocky.net/bitkit/.git
1689 2012-02-06 20:18:22 iocor has quit (Quit: Computer has gone to sleep.)
1690 2012-02-06 20:18:29 <genjix> there's a big forum argument slandering catherine flick as a 'statist' https://bitcointalk.org/index.php?topic=62836.0
1691 2012-02-06 20:18:46 <genjix> cmon it's just some chick studying ethics that likes bitcoin and did a study for fun :)
1692 2012-02-06 20:19:02 BlueMatt has quit (Ping timeout: 255 seconds)
1693 2012-02-06 20:19:14 <gavinandresen> "chick" ?
1694 2012-02-06 20:19:37 <gavinandresen> you're gonna get yourself in trouble....
1695 2012-02-06 20:20:39 <XMPPwocky> genjix: nice thing about construct
1696 2012-02-06 20:20:48 <XMPPwocky> I'm going to write a whole set of adapters for it
1697 2012-02-06 20:22:14 <genjix> XMPPwocky: it's a fuzzer right?
1698 2012-02-06 20:22:20 <XMPPwocky> genjix: it's a protocol implementatiobn
1699 2012-02-06 20:22:24 <XMPPwocky> *implementation
1700 2012-02-06 20:22:35 <genjix> aha ok
1701 2012-02-06 20:23:10 <genjix> XMPPwocky: ohh a parser, right?
1702 2012-02-06 20:23:15 <XMPPwocky> parser and builder
1703 2012-02-06 20:23:19 <XMPPwocky> construct goes both ways
1704 2012-02-06 20:24:04 <genjix> if you need a reference then use this:
1705 2012-02-06 20:24:05 <genjix> https://gitorious.org/libbitcoin/libbitcoin/blobs/master/src/exporter.cpp
1706 2012-02-06 20:24:15 <genjix> https://gitorious.org/libbitcoin/libbitcoin/blobs/master/src/utility/serializer.cpp
1707 2012-02-06 20:24:50 <XMPPwocky> genjix: eeew. :P
1708 2012-02-06 20:25:19 <genjix> gavinandresen: btw i'm generating a few fake chains. do you want me to send you them after?
1709 2012-02-06 20:26:02 <XMPPwocky> so if a client ignores checkpoints, could someone do a really simple DoS by sending thousands of blocks at height 2 (since they're pretty trivial to generate)?
1710 2012-02-06 20:26:14 <genjix> i'm going to put them somewhere, but maybe you wanna put them in the fuzzer
1711 2012-02-06 20:26:22 <gavinandresen> genjix: fake valid or fake invalid?
1712 2012-02-06 20:26:28 <genjix> fake valid
1713 2012-02-06 20:26:53 <genjix> the whole idea is your implementation takes them and inserts them in funny ways to generate re-orgs
1714 2012-02-06 20:27:26 <genjix> XMPPwocky: no you cannot.
1715 2012-02-06 20:27:39 <XMPPwocky> genjix: why not?
1716 2012-02-06 20:27:51 <gavinandresen> genjix: keep them for now, I'm still thinking about getting more organized RE: cross-implementation testing
1717 2012-02-06 20:28:05 <genjix> ok
1718 2012-02-06 20:28:26 datagutt has quit (Quit: kthxbai)
1719 2012-02-06 20:28:44 <XMPPwocky> either the client throws away blocks that aren't longer than the current chain
1720 2012-02-06 20:28:54 <XMPPwocky> which would prevent any sort of re-org
1721 2012-02-06 20:29:18 <Moron__> whats a cerealiser?
1722 2012-02-06 20:29:31 <genjix> k afk now
1723 2012-02-06 20:29:35 <genjix> nice chatting yall
1724 2012-02-06 20:29:35 barmstrong has quit (Read error: Connection reset by peer)
1725 2012-02-06 20:29:36 <genjix> cya
1726 2012-02-06 20:29:40 genjix has quit (Quit: leaving)
1727 2012-02-06 20:29:46 barmstrong has joined
1728 2012-02-06 20:30:24 barmstro_ has joined
1729 2012-02-06 20:30:59 marf_away has joined
1730 2012-02-06 20:33:08 pickett has quit (Remote host closed the connection)
1731 2012-02-06 20:34:04 [1]wirehead has quit (Ping timeout: 245 seconds)
1732 2012-02-06 20:34:14 barmstrong has quit (Ping timeout: 245 seconds)
1733 2012-02-06 20:35:58 smickles is now known as smickles|idle
1734 2012-02-06 20:36:08 smickles is now known as idle!~michael@cpe-071-070-169-083.nc.res.rr.com|smickles
1735 2012-02-06 20:39:58 pickett has joined
1736 2012-02-06 20:40:27 barmstrong has joined
1737 2012-02-06 20:40:36 att has joined
1738 2012-02-06 20:41:42 Wack0 has joined
1739 2012-02-06 20:41:42 wirehead has joined
1740 2012-02-06 20:41:53 Crshman has joined
1741 2012-02-06 20:42:42 nexes has joined
1742 2012-02-06 20:43:44 barmstro_ has quit (Ping timeout: 248 seconds)
1743 2012-02-06 20:46:17 smickles is now known as smickles|idle
1744 2012-02-06 20:49:29 rdponticelli_ has quit (Ping timeout: 245 seconds)
1745 2012-02-06 20:49:44 rdponticelli has joined
1746 2012-02-06 20:49:54 smickles is now known as idle!~michael@cpe-071-070-169-083.nc.res.rr.com|smickles
1747 2012-02-06 20:51:59 slush has quit (Ping timeout: 245 seconds)
1748 2012-02-06 20:54:24 rdponticelli has quit (Ping timeout: 240 seconds)
1749 2012-02-06 21:06:08 BlueMatt has joined
1750 2012-02-06 21:11:29 barmstrong has quit (Remote host closed the connection)
1751 2012-02-06 21:13:41 barmstrong has joined
1752 2012-02-06 21:13:50 zeiris has joined
1753 2012-02-06 21:15:00 barmstro_ has joined
1754 2012-02-06 21:15:00 barmstrong has quit (Read error: Connection reset by peer)
1755 2012-02-06 21:16:07 smickles is now known as smickles|idle
1756 2012-02-06 21:16:15 smickles is now known as idle!~michael@cpe-071-070-169-083.nc.res.rr.com|smickles
1757 2012-02-06 21:16:55 m00p has joined
1758 2012-02-06 21:18:01 <midnightmagic> sipa: Does the wallet import/merge work yet with addresses with coins in them? (isCompress etc)? I'd love to know when.. should I just watch that pull req?
1759 2012-02-06 21:19:03 <sipa> midnightmagic: untested, but it should
1760 2012-02-06 21:19:29 <sipa> i recently rebased it after compressedpubkeys
1761 2012-02-06 21:19:38 <midnightmagic> ah, wonderful!
1762 2012-02-06 21:19:46 <midnightmagic> i'm so happy to hear that.
1763 2012-02-06 21:20:01 <sipa> private key import/export should be in 0.6
1764 2012-02-06 21:20:41 <midnightmagic> are they keeping the signmessage function in it too?
1765 2012-02-06 21:21:19 <sipa> yes
1766 2012-02-06 21:21:22 RazielZ has joined
1767 2012-02-06 21:21:22 <midnightmagic> yay!!!
1768 2012-02-06 21:21:33 <midnightmagic> I would offer you coffee right now if I could.
1769 2012-02-06 21:21:35 <sipa> or wait, what do you mean?
1770 2012-02-06 21:21:49 <midnightmagic> signmessage <bitcoinaddress> <message> & verifymessage <bitcoinaddress> <signature> <message>
1771 2012-02-06 21:22:09 <sipa> 0.6 will have signmessage, privkey import/export, and compressed pubkeys, and all should be compatible with eachother
1772 2012-02-06 21:22:21 <midnightmagic> then still yay!
1773 2012-02-06 21:22:35 <Moron__> hey there midnightmagic
1774 2012-02-06 21:22:36 <sipa> and signmessage via gui
1775 2012-02-06 21:22:50 <midnightmagic> hey Moron__, still haven't picked a better nick yet, huh
1776 2012-02-06 21:23:01 <Moron__> well i was gonna go with Retard__
1777 2012-02-06 21:23:12 <Moron__> since so many people have blocked me in the main channel :P
1778 2012-02-06 21:23:27 <Moron__> they all haterz tho so its no problem
1779 2012-02-06 21:23:34 <midnightmagic> not a great improvement. I almost think you revel in it.
1780 2012-02-06 21:24:08 <Moron__> nah they just jelous me because im better than them in the thinking department :P
1781 2012-02-06 21:24:26 <midnightmagic> Are you just cocktopus in disguise?
1782 2012-02-06 21:25:05 <Moron__> cocktopus?
1783 2012-02-06 21:25:18 <Moron__> whos that?
1784 2012-02-06 21:27:13 smickles is now known as smickles|idle
1785 2012-02-06 21:27:19 smickles is now known as idle!~michael@cpe-071-070-169-083.nc.res.rr.com|smickles
1786 2012-02-06 21:32:23 iocor has joined
1787 2012-02-06 21:32:29 iocor has quit (Client Quit)
1788 2012-02-06 21:32:46 iocor has joined
1789 2012-02-06 21:34:05 smickles has quit (Remote host closed the connection)
1790 2012-02-06 21:36:06 dr_win has joined
1791 2012-02-06 21:36:31 ThomasV_ has joined
1792 2012-02-06 21:42:23 splatster has joined
1793 2012-02-06 21:42:58 erle- has quit (Quit: erle-)
1794 2012-02-06 21:43:28 erle- has joined
1795 2012-02-06 21:44:52 Clipse has quit (Quit: Clipse)
1796 2012-02-06 21:45:02 Clipse has joined
1797 2012-02-06 21:45:52 smtmnyz_ has joined
1798 2012-02-06 21:46:34 smtmnyz has quit (Ping timeout: 245 seconds)
1799 2012-02-06 21:49:57 <XMPPwocky> http://pastebin.com/JxJ7GcQG
1800 2012-02-06 21:50:04 <XMPPwocky> it works! (at least for version)
1801 2012-02-06 21:50:24 nx5 has joined
1802 2012-02-06 21:50:49 nx5 is now known as Guest50325
1803 2012-02-06 21:52:50 smtmnyz_ has quit ()
1804 2012-02-06 21:53:01 smtmnyz has joined
1805 2012-02-06 21:53:47 gjs278 has quit (Ping timeout: 272 seconds)
1806 2012-02-06 21:53:56 Guest14267 has joined
1807 2012-02-06 21:53:58 Guest14267 is now known as jgarzik
1808 2012-02-06 21:54:34 minimoose has quit (Quit: minimoose)
1809 2012-02-06 21:55:15 smickles has joined
1810 2012-02-06 21:55:15 gjs278 has joined
1811 2012-02-06 21:57:10 gavinandresen has quit (Quit: gavinandresen)
1812 2012-02-06 21:57:30 <gribble> New news from bitcoinrss: gavinandresen opened pull request 804 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/804>
1813 2012-02-06 21:58:06 PK has quit ()
1814 2012-02-06 21:59:01 paraipan has quit (Remote host closed the connection)
1815 2012-02-06 22:02:19 erle- has quit (Quit: erle-)
1816 2012-02-06 22:02:39 JRWR has quit (Ping timeout: 272 seconds)
1817 2012-02-06 22:05:19 baz has joined
1818 2012-02-06 22:07:13 smtmnyz_ has joined
1819 2012-02-06 22:07:54 smtmnyz has quit (Ping timeout: 256 seconds)
1820 2012-02-06 22:08:10 <splatster> an #bitcoin-auction OPs want to help me get an auction going?
1821 2012-02-06 22:08:14 <splatster> any*
1822 2012-02-06 22:13:27 smickles has quit (Remote host closed the connection)
1823 2012-02-06 22:18:54 gp5st has quit (Quit: Leaving.)
1824 2012-02-06 22:22:15 pusle has quit ()
1825 2012-02-06 22:22:24 <XMPPwocky> http://pastebin.com/3EJ6btJW
1826 2012-02-06 22:22:30 <XMPPwocky> now with pretty IPs!
1827 2012-02-06 22:23:18 <sipa> nice
1828 2012-02-06 22:24:01 * XMPPwocky wonders why bitcoin puts user_agent in slashes
1829 2012-02-06 22:24:17 <luke-jr> XMPPwocky: it's a path
1830 2012-02-06 22:24:22 <XMPPwocky> luke-jr: ... to what?
1831 2012-02-06 22:24:24 <luke-jr> BIP 14
1832 2012-02-06 22:24:42 <luke-jr> XMPPwocky: bitcoind is non-compliant, but it's multiple levels of the client
1833 2012-02-06 22:24:53 <luke-jr> eg, /Satoshi:0.6.0/bitcoind:0.6.0/
1834 2012-02-06 22:25:04 <luke-jr> eg, /Satoshi:0.6.0/Bitcoin-Qt:0.6.0/
1835 2012-02-06 22:25:29 <XMPPwocky> ah
1836 2012-02-06 22:25:32 <luke-jr> eventually, mix-and-match should be possible
1837 2012-02-06 22:26:46 <sipa> XMPPwocky: any messages beyond version implemented yet?
1838 2012-02-06 22:26:51 <XMPPwocky> sipa: all of them
1839 2012-02-06 22:26:59 _Fireball has quit (Quit:  Try HydraIRC -> http://www.hydrairc.com <-)
1840 2012-02-06 22:27:01 <XMPPwocky> but version's easiest to test
1841 2012-02-06 22:27:27 <splatster> Actioning off my White iPhone 3gs 16 GB in #bitcoin-auction!
1842 2012-02-06 22:27:30 <splatster> oops
1843 2012-02-06 22:27:37 <XMPPwocky> if anyone happens to have a bunch of packets
1844 2012-02-06 22:28:07 <XMPPwocky> a binary dump of /one side/ of a connection would be great
1845 2012-02-06 22:28:21 <sipa> XMPPwocky: that's not hard to find
1846 2012-02-06 22:29:24 <XMPPwocky> sipa: yeah, I'll probably hack something together with nc
1847 2012-02-06 22:29:41 <luke-jr> XMPPwocky: full python node?
1848 2012-02-06 22:29:49 <XMPPwocky> luke-jr: maybe eventually
1849 2012-02-06 22:30:26 <forrestv> XMPPwocky, i've already implemented the entire bitcoin protocol in python in p2pool.. btw
1850 2012-02-06 22:31:31 <XMPPwocky> forrestv: this is meant to be a set of tools for bitcoin
1851 2012-02-06 22:31:42 <luke-jr> forrestv: entire?
1852 2012-02-06 22:33:41 <lianj> i guess he means parsing
1853 2012-02-06 22:34:38 <sipa> XMPPwocky: i have 4-line patch to bitcoind to make it dump all its created packets in hex to debug log
1854 2012-02-06 22:38:01 <XMPPwocky> sipa: I'll just write the networking code
1855 2012-02-06 22:38:11 <XMPPwocky> at least for responding to version with verack
1856 2012-02-06 22:39:29 att has quit (Quit: Leaving)
1857 2012-02-06 22:42:00 <XMPPwocky> hey
1858 2012-02-06 22:42:13 <XMPPwocky> how the heck do you know, in a version packet, which services the remote host supports?
1859 2012-02-06 22:42:49 <sipa> look at its addr_from ?
1860 2012-02-06 22:43:07 <XMPPwocky> sipa: if you haven't gotten their version
1861 2012-02-06 22:43:31 <sipa> eh?
1862 2012-02-06 22:43:59 <sipa> it's in addr messages as well
1863 2012-02-06 22:44:06 <sipa> so you'd know before trying to connect
1864 2012-02-06 22:44:15 <XMPPwocky> okay, but if you're connecting "naively" (via -addnode, perhaps)
1865 2012-02-06 22:44:50 <sipa> don't -addnode an SPV node ;)
1866 2012-02-06 22:45:12 <XMPPwocky> SPV?
1867 2012-02-06 22:45:20 barmstro_ has quit (Remote host closed the connection)
1868 2012-02-06 22:45:26 <sipa> simplified payment verification, like bitcoinj
1869 2012-02-06 22:45:56 <XMPPwocky> oh
1870 2012-02-06 22:46:00 <XMPPwocky> that-- kinda sucks
1871 2012-02-06 22:46:17 <sipa> why?
1872 2012-02-06 22:46:27 <XMPPwocky> hm, does a node give a crap if you give it the wrong services in addr_recv?
1873 2012-02-06 22:46:30 <sipa> they don't listen anyway
1874 2012-02-06 22:47:03 <XMPPwocky> they ignore addr_recv?
1875 2012-02-06 22:47:16 <sipa> who?
1876 2012-02-06 22:47:21 <XMPPwocky> nodes
1877 2012-02-06 22:47:32 <XMPPwocky> okay, here's the question
1878 2012-02-06 22:47:52 <XMPPwocky> is it safe to set services to NODE_NETWORK in the addr_recv part of a version message to a node you're connecting to
1879 2012-02-06 22:47:57 <XMPPwocky> bitcoind seems to
1880 2012-02-06 22:48:33 <sipa> indeed
1881 2012-02-06 22:48:43 <sipa> but why would you?
1882 2012-02-06 22:48:52 [Tycho] has quit (Remote host closed the connection)
1883 2012-02-06 22:49:28 <XMPPwocky> for testing
1884 2012-02-06 22:49:37 RazielZ has quit (Quit: Leaving)
1885 2012-02-06 22:50:04 barmstrong has joined
1886 2012-02-06 22:58:10 copumpkin has quit (Quit: Computer has gone to sleep.)
1887 2012-02-06 22:59:41 devrandom has quit (Remote host closed the connection)
1888 2012-02-06 23:06:06 devrandom has joined
1889 2012-02-06 23:09:39 copumpkin has joined
1890 2012-02-06 23:10:13 baz has quit (Ping timeout: 252 seconds)
1891 2012-02-06 23:10:25 sneak has quit (Ping timeout: 272 seconds)
1892 2012-02-06 23:11:44 marf_away has quit (Ping timeout: 245 seconds)
1893 2012-02-06 23:13:04 booo has quit (Ping timeout: 240 seconds)
1894 2012-02-06 23:13:48 b4epoche_ has joined
1895 2012-02-06 23:14:15 b4epoche has quit (Ping timeout: 245 seconds)
1896 2012-02-06 23:14:15 b4epoche_ is now known as b4epoche
1897 2012-02-06 23:14:23 ThomasV_ has quit (Ping timeout: 244 seconds)
1898 2012-02-06 23:15:01 diki has quit (Remote host closed the connection)
1899 2012-02-06 23:15:15 diki has joined
1900 2012-02-06 23:15:25 tower has quit (Read error: Connection reset by peer)
1901 2012-02-06 23:15:37 sneak has joined
1902 2012-02-06 23:15:37 sneak has quit (Changing host)
1903 2012-02-06 23:15:37 sneak has joined
1904 2012-02-06 23:17:16 tower has joined
1905 2012-02-06 23:22:35 dr_win has quit (Remote host closed the connection)
1906 2012-02-06 23:22:58 paraipan has joined
1907 2012-02-06 23:24:44 Zarutian has quit (Ping timeout: 240 seconds)
1908 2012-02-06 23:25:42 <denisx> I thought the new versions of bitcoind are faster at block loading
1909 2012-02-06 23:25:57 <denisx> my client loads maybe one block per second
1910 2012-02-06 23:26:05 <denisx> 0.5.2
1911 2012-02-06 23:26:14 <BlueMatt> on linux, yea
1912 2012-02-06 23:26:22 <BlueMatt> on windows the difference is minor
1913 2012-02-06 23:26:25 <denisx> no, on freebsd
1914 2012-02-06 23:26:28 <BlueMatt> nfc
1915 2012-02-06 23:26:58 <BlueMatt> I have no idea how quick freebsd's mlock call is
1916 2012-02-06 23:28:16 <sipa> put the block chain on a tmpfs... and watch it fly
1917 2012-02-06 23:31:09 <denisx> ah, I see, it does heavy writing
1918 2012-02-06 23:33:32 AAA_awright has quit (Read error: Connection reset by peer)
1919 2012-02-06 23:33:48 AAA_awright has joined
1920 2012-02-06 23:39:44 Diablo-D3 has quit (Read error: Connection reset by peer)
1921 2012-02-06 23:41:08 Diablo-D3 has joined
1922 2012-02-06 23:42:51 forrestv has quit (Quit: Coyote finally caught me)
1923 2012-02-06 23:43:20 AAA_awright_ has joined
1924 2012-02-06 23:44:23 <denisx> hmm, setting my ZFS to async helped a little
1925 2012-02-06 23:44:28 <luke-jr> anyone know why bitcoin.cz is flooding Eligius? -.-
1926 2012-02-06 23:44:39 <BlueMatt> ???
1927 2012-02-06 23:44:42 ultra_ has joined
1928 2012-02-06 23:44:55 <gmaxwell> luke-jr: whats the traffic look like?
1929 2012-02-06 23:45:02 <luke-jr> BlueMatt: DDoS attack, except from one IP. someone noticed it's the same IP that bitcoin.cz resolves to
1930 2012-02-06 23:45:06 <luke-jr> gmaxwell: UDP to port 80
1931 2012-02-06 23:45:13 <luke-jr> 23:05:35.046937 IP 195.140.255.78.42586 > 78.47.187.254.80: UDP, length 150
1932 2012-02-06 23:45:14 <luke-jr> 23:05:35.046938 IP 195.140.255.78.42590 > 78.47.187.254.80: UDP, length 150
1933 2012-02-06 23:45:15 <Diablo-D3> "DDoS, except from one IP" = DoS
1934 2012-02-06 23:45:16 <luke-jr> 23:05:35.047002 IP 195.140.255.78.47471 > 78.47.187.254.80: UDP, length 150
1935 2012-02-06 23:45:17 <luke-jr> 23:05:35.047005 IP 195.140.255.78.47473 > 78.47.187.254.80: UDP, length 150
1936 2012-02-06 23:45:22 <Diablo-D3> luke-jr: udpdump?
1937 2012-02-06 23:45:24 <gmaxwell> whats the actual payload?
1938 2012-02-06 23:45:34 <luke-jr> gmaxwell: I can't tell. packet loss too high to do anything
1939 2012-02-06 23:45:51 <josephcp> UDP can spoof their source IP
1940 2012-02-06 23:45:58 <gmaxwell> yep.
1941 2012-02-06 23:45:59 <BlueMatt> probably not bitcoin.cz
1942 2012-02-06 23:46:11 <gmaxwell> josephcp: though most edge networks doe URPF... I guess not enough, enh?
1943 2012-02-06 23:46:28 <luke-jr> [15:45:24] <-- slush has left this server (Ping timeout: 245 seconds).
1944 2012-02-06 23:46:30 <josephcp> not if they're behind similar core routers as bitcoin.cz
1945 2012-02-06 23:46:38 <denisx> luke-jr: are you sending replys to bitcoin.cz? ;)
1946 2012-02-06 23:46:42 <ultra_> xD
1947 2012-02-06 23:46:42 <gmaxwell> luke-jr: the attacker might want backscatter...
1948 2012-02-06 23:46:43 <josephcp> i assume if the owner has a botnet they can find a couple IPs that might work
1949 2012-02-06 23:46:53 <luke-jr> denisx: didn't look like it
1950 2012-02-06 23:47:08 danbri has quit (Remote host closed the connection)
1951 2012-02-06 23:47:09 <denisx> that would be called a smurf attack
1952 2012-02-06 23:47:15 AAA_awright has quit (Ping timeout: 252 seconds)
1953 2012-02-06 23:47:28 <BlueMatt> heh get luke to dos slush
1954 2012-02-06 23:47:29 <denisx> had that in the late nineties
1955 2012-02-06 23:47:32 <BlueMatt> by dosing luke
1956 2012-02-06 23:48:10 <gmaxwell> In any case. iptables -I input -P UDP -dport 80 -j DROP   or something like that.. just in case.
1957 2012-02-06 23:48:11 <josephcp> yeah you can also sometimes use UDP to bounce attacks, odesn't have to be a smurf necessarily
1958 2012-02-06 23:49:05 <luke-jr> gmaxwell: already blocking that IP in iptables now
1959 2012-02-06 23:49:14 <BlueMatt> luke-jr: your pool isnt behind dos protection services?
1960 2012-02-06 23:49:53 <denisx> who can pay for that?
1961 2012-02-06 23:50:01 <BlueMatt> someone who runs a pool
1962 2012-02-06 23:50:12 <denisx> someone with a very big pool
1963 2012-02-06 23:50:14 * luke-jr wonders where BlueMatt gets the idea that pools have income
1964 2012-02-06 23:50:25 p0s has quit (Remote host closed the connection)
1965 2012-02-06 23:50:40 <NxTitle> ^ deepbit
1966 2012-02-06 23:50:42 <NxTitle> nuff said
1967 2012-02-06 23:50:49 <gmaxwell> Besides, why bother spending money on (very expensive!) anti-DOS services when you could just spend it to DOS the compeition equally? :)
1968 2012-02-06 23:50:50 <BlueMatt> I was under the impression luke-jr kept fees? or is that just deepbit
1969 2012-02-06 23:51:00 <BlueMatt> gmaxwell: ok, good point there
1970 2012-02-06 23:51:06 <gmaxwell> BlueMatt: fees amount to 0.02 BTC/block on average in any case.)
1971 2012-02-06 23:51:14 <BlueMatt> thats true
1972 2012-02-06 23:51:18 <denisx> BlueMatt: sure, but even small pools need to pay the ISP
1973 2012-02-06 23:51:37 <cjd> UDP w/ source spoofing, we can be thankful they don't know what they're doing.
1974 2012-02-06 23:51:42 <BlueMatt> yea, I just thought ddos on pools was more common
1975 2012-02-06 23:51:47 <BlueMatt> seems like it would be an ideal target
1976 2012-02-06 23:52:06 torsthaldo has joined
1977 2012-02-06 23:52:53 <gmaxwell> BlueMatt: how goes the UPNP fixes.
1978 2012-02-06 23:53:04 <BlueMatt> I thought they were merged?
1979 2012-02-06 23:53:15 <BlueMatt> or were there more I cant remember
1980 2012-02-06 23:53:16 <BlueMatt> ?
1981 2012-02-06 23:53:18 <luke-jr> BlueMatt: txn fees are nothing
1982 2012-02-06 23:53:18 <denisx> I know a guy who offers ddos protector for sport bet sites. he charges several thousand euros per month...
1983 2012-02-06 23:53:27 <denisx> protection
1984 2012-02-06 23:54:13 <ultra_> 'protection', great... gangsters...
1985 2012-02-06 23:54:32 p0s has joined
1986 2012-02-06 23:54:40 <ultra_> I prefer a rubber
1987 2012-02-06 23:54:51 <ultra_> lots cheaper
1988 2012-02-06 23:55:01 <BlueMatt> have fun plugging ethernet in through one (believe me, its not easy)
1989 2012-02-06 23:55:23 <NxTitle> ^ plugs ethernet through rubber
1990 2012-02-06 23:55:44 <ultra_> BlueMatt: works fine, just you don't get a connection. best DDoS protection around.
1991 2012-02-06 23:56:06 occulta has joined
1992 2012-02-06 23:56:37 <NxTitle> ultra_: I'm still fairly certain that's a *denial* of service
1993 2012-02-06 23:56:39 <NxTitle> bu yeah, sure
1994 2012-02-06 23:56:42 <BTC_Bear> lol
1995 2012-02-06 23:58:16 pingdrive has joined