1 2012-04-04 00:02:06 Cablesaurus has quit (Quit: Download IceChat at www.icechat.net)
   2 2012-04-04 00:16:06 Cablesaurus has joined
   3 2012-04-04 00:16:06 Cablesaurus has quit (Changing host)
   4 2012-04-04 00:16:06 Cablesaurus has joined
   5 2012-04-04 00:23:26 barmstrong has quit (Remote host closed the connection)
   6 2012-04-04 00:24:06 barmstrong has joined
   7 2012-04-04 00:31:52 ThomasV has quit (Ping timeout: 244 seconds)
   8 2012-04-04 00:32:20 eoss has quit (Remote host closed the connection)
   9 2012-04-04 00:33:09 user_ has quit (Quit: Leaving)
  10 2012-04-04 00:39:43 brwyatt is now known as Away!~brwyatt@pool-71-252-154-11.dllstx.fios.verizon.net|brwyatt
  11 2012-04-04 00:40:36 <Jezzz> wooo
  12 2012-04-04 00:43:26 wood__ has joined
  13 2012-04-04 00:48:26 Clipse has quit (Ping timeout: 246 seconds)
  14 2012-04-04 00:49:11 da2ce7 has quit (Ping timeout: 260 seconds)
  15 2012-04-04 00:49:23 wood__ has left ()
  16 2012-04-04 00:54:25 mmoya has quit (Remote host closed the connection)
  17 2012-04-04 00:56:44 tracton has joined
  18 2012-04-04 01:08:27 kreal- has quit ()
  19 2012-04-04 01:09:49 barmstrong has quit (Remote host closed the connection)
  20 2012-04-04 01:10:10 <luke-jr> http://eligius.st/wiki/index.php/FAQ#Botnets.2FViruses <-- thoughts?
  21 2012-04-04 01:10:55 lh77 has quit ()
  22 2012-04-04 01:11:54 <Diablo-D3> that works
  23 2012-04-04 01:12:14 <brwyatt> So for those still caring about it, I have the official IP information for my network. 2012-03-06 18:50 GMT-5 my IP changed from 71.123.170.117 to 71.123.170.150. It changed again on 2012-04-01 14:43 to 71.252.154.11
  24 2012-04-04 01:13:14 <brwyatt> I don't know if that helps any of y'all looking at IP addresses and trying to track things down. But there it is.
  25 2012-04-04 01:18:45 <BlueMatt> jgarzik: sorry, ignore me
  26 2012-04-04 01:24:10 denisx has joined
  27 2012-04-04 01:24:10 denisx has quit (Client Quit)
  28 2012-04-04 01:30:19 barmstrong has joined
  29 2012-04-04 01:31:10 Hasbro has quit (Ping timeout: 240 seconds)
  30 2012-04-04 01:35:36 <riush> luke-jr: s/cooperate/cooperation/ .. also do you offer help for illegal botnets too? (doesn't specify ;)
  31 2012-04-04 01:37:18 <tracton> luke-jr: have you ever been contacted by LE about one of your users?
  32 2012-04-04 01:39:08 user_ has joined
  33 2012-04-04 01:39:17 <luke-jr> riush: I specify "no questions asked", so I don't know if it's legal or not :p
  34 2012-04-04 01:39:32 <luke-jr> riush: I'd rather help an illegal one, than get them breaking me
  35 2012-04-04 01:39:36 <luke-jr> tracton: no
  36 2012-04-04 01:40:29 cesio has joined
  37 2012-04-04 01:40:33 <riush> ah hehe
  38 2012-04-04 01:43:33 <luke-jr> tracton: but I *have* been getting a lot of "I'm infected" reports lately
  39 2012-04-04 01:44:30 <brwyatt> luke-jr: I'm guessing I'm not the only one getting emails, then?
  40 2012-04-04 01:44:31 <gmaxwell> Frankly, "no questions asked" sound like tacit endorcement. I'd omit that or find anyother way of saying that.
  41 2012-04-04 01:44:47 worlemp has joined
  42 2012-04-04 01:44:48 <luke-jr> brwyatt: ?
  43 2012-04-04 01:45:08 <luke-jr> gmaxwell: any suggestions? :/
  44 2012-04-04 01:45:43 <luke-jr> I suppose I could say "for your legal botnet" and let the illegal ones lie to me
  45 2012-04-04 01:45:45 <brwyatt> luke-jr: I got an email telling me I may be infected, and ended up here. Perhaps they got something similar and turned to you?
  46 2012-04-04 01:45:56 <luke-jr> brwyatt: maybe.
  47 2012-04-04 01:46:24 cesio has quit (Read error: Connection reset by peer)
  48 2012-04-04 01:46:31 <riush> you offer the service for legal botnets, but you can't be bothered to verify that claim
  49 2012-04-04 01:48:32 <gmaxwell> riush: it's the truth too. ::shrugs::
  50 2012-04-04 01:50:10 <Graet> luke-jr, i distinguish between "large cpu miners" =legit and "botnets" = illegal activity, always happy bto mtalk to large cpu miners that verify they arent botnets ;) but its an interesting/awkward subject ;
  51 2012-04-04 01:51:24 forsetofox_ has joined
  52 2012-04-04 01:51:54 <Graet> and in *some* ways its better they are "supporting" the network rather than attacking it :/
  53 2012-04-04 01:53:20 <luke-jr> riush: from my perspective, the legalities are not really my business - I am not an authority, nor does my authority expect me to investigate/report for them; but at the same time, I don't endorse criminal activity, so I will help LE lock them up
  54 2012-04-04 01:53:45 forsetofox_ has quit (Client Quit)
  55 2012-04-04 01:54:03 <luke-jr> if someone comes right out and tells me they're doing something criminal, then I'll probably pass that info along ASAP :P
  56 2012-04-04 01:54:19 <luke-jr> (someone actually did recently, but I couldn't find a contact for South Africa LE :/)
  57 2012-04-04 01:54:51 <luke-jr> (it was pretty ironic to have the victim in the IRC channel, and the criminal trying to downplay him)
  58 2012-04-04 01:56:42 <luke-jr> riush: probably more important though.. it's pretty likely that when/if I take action against an illegal botnet, they will try to DDoS me, which is very expensive
  59 2012-04-04 01:57:22 <gmaxwell> There are also all sorts of different potentially illegal botnets. E.g. someone who has lawful control of a large number of computers mines on them... but they aren't his computers. Is that an unlawful botnet?  Whatever it is, it's certantly different from someone who hacked a whole bunch of computers to mine on them.
  60 2012-04-04 01:57:30 <luke-jr> and/or they can just change bitcoin address and go from "known risk" to "unknown risk"
  61 2012-04-04 01:57:35 <BTC_Bear> There are a whole bunch of criminals in Syria right now, are you going to turn them in to LE?  Just saying this isn't a white and black issue.
  62 2012-04-04 01:58:17 <luke-jr> BTC_Bear: if someone is stupid enough to PM me and say "hey, I owned that guy, don't send him to LE!", I'm more than happy to report them
  63 2012-04-04 01:58:30 <gmaxwell> But in general, someone who is careful enough to put on the pretext of lawfulness is a lot less likely to be a troublemaker.
  64 2012-04-04 01:58:35 <BTC_Bear> I understand your side.
  65 2012-04-04 01:58:52 <BTC_Bear> Case by case tho... no generalization.
  66 2012-04-04 01:59:00 <luke-jr> I guess
  67 2012-04-04 01:59:10 <luke-jr> I'd prefer if LE made an arrest before I got DDoS'd
  68 2012-04-04 01:59:23 <luke-jr> that's basically what it comes down to in practical terms
  69 2012-04-04 01:59:28 <gmaxwell> The folks that are you-cant-catch-me-Im-the-gingerbread-man seem to be more likely to knock out the pool through carelessness and then DOS attack when they get blocked for DOSing the pool.
  70 2012-04-04 02:00:02 <luke-jr> XD
  71 2012-04-04 02:00:28 <luke-jr> when I see botnet-alike activity, my first reaction is usually to try to improve how I handle the higher load :P
  72 2012-04-04 02:00:52 <luke-jr> except in the case of a legit FPGA miner who opened a ton of LP connections -.-
  73 2012-04-04 02:01:05 <luke-jr> I blocked his LPs and sent him a PM. bad idea.
  74 2012-04-04 02:01:20 <luke-jr> I guess miner sw doesn't react well to forbidden LPs
  75 2012-04-04 02:02:13 sacarlson has quit (Read error: Connection reset by peer)
  76 2012-04-04 02:06:46 malaimo has quit (Ping timeout: 260 seconds)
  77 2012-04-04 02:08:20 <phantomcircuit> luke-jr, tarpit
  78 2012-04-04 02:08:24 <phantomcircuit> proceed to laugh
  79 2012-04-04 02:08:27 <luke-jr> …
  80 2012-04-04 02:08:35 m00p has quit (Read error: Operation timed out)
  81 2012-04-04 02:08:48 <XMPPwocky> phantomcircuit: better, just send invalid getworks
  82 2012-04-04 02:08:56 <XMPPwocky> he'll spend days debugging before he figures it out
  83 2012-04-04 02:09:14 <phantomcircuit> i had to tarpit someone who was polling intersango with 5 open connections every 10 ms
  84 2012-04-04 02:09:19 <luke-jr> >_<
  85 2012-04-04 02:09:32 <phantomcircuit> i dont even understand how you would think that's ok
  86 2012-04-04 02:19:02 sacarlson has joined
  87 2012-04-04 02:20:50 pavel__ has joined
  88 2012-04-04 02:26:13 m00p has joined
  89 2012-04-04 02:38:50 T_X_ has quit (Read error: Operation timed out)
  90 2012-04-04 02:43:21 [7] has quit (Disconnected by services)
  91 2012-04-04 02:43:28 TheSeven has joined
  92 2012-04-04 02:44:00 minimoose has left ()
  93 2012-04-04 02:44:29 minimoose has joined
  94 2012-04-04 02:45:24 T_X has joined
  95 2012-04-04 03:02:45 toffoo has joined
  96 2012-04-04 03:03:40 Karmaon has joined
  97 2012-04-04 03:06:07 <gmaxwell> sipa: what if any cases do you know still remain where a node can get stuck on a fork?
  98 2012-04-04 03:07:01 <gmaxwell> Azelphur: in #bitcoin has upgraded to 0.6 after finding himself stuck on a fork on old code.. but he's staying stuck. I had him addnode me, he fetches new blocks but I don't see him try to pull the connecting chain.
  99 2012-04-04 03:07:30 <gmaxwell> (he's stuck at 171819 )
 100 2012-04-04 03:08:11 <gmaxwell> This is his debug log from before he was connected to me: http://paste.ubuntu.com/914011/
 101 2012-04-04 03:09:51 * TuxBlackEdo blinks
 102 2012-04-04 03:11:52 forsetifox has quit (Quit: Page closed)
 103 2012-04-04 03:13:51 forsetifox has joined
 104 2012-04-04 03:16:29 <luke-jr> gmaxwell: did he -checkblocks?
 105 2012-04-04 03:19:09 paulo has quit (away!~foo@112.204.27.38|Ping timeout: 244 seconds)
 106 2012-04-04 03:19:14 <tracton> remember when i said rg and bitvps sucks, that the servers constantly hang, and that the service is horrible? dude i was kidding. it was all a lie, sorry. i never even used bitvps. tho i'm hearing from ppl they do have some awesome vps's that kick ass. rg did ban me for no reason tho. why did he do it?
 107 2012-04-04 03:19:15 <tracton> remember when i said rg and bitvps sucks, that the servers constantly hang, and that the service is horrible? dude i was kidding. it was all a lie, sorry. i never even used bitvps. tho i'm hearing from ppl they do have some awesome vps's that kick ass. rg did ban me for no reason tho. why did he do it?
 108 2012-04-04 03:19:15 <tracton> remember when i said rg and bitvps sucks, that the servers constantly hang, and that the service is horrible? dude i was kidding. it was all a lie, sorry. i never even used bitvps. tho i'm hearing from ppl they do have some awesome vps's that kick ass. rg did ban me for no reason tho. why did he do it?
 109 2012-04-04 03:19:15 tracton has left ()
 110 2012-04-04 03:19:37 <gmaxwell> Oh god, this idiot again?
 111 2012-04-04 03:19:50 <gmaxwell> luke-jr: nothing in his logs indicates block corruption at all.
 112 2012-04-04 03:20:19 <BlueMatt> yay static ips
 113 2012-04-04 03:20:49 <luke-jr> lol
 114 2012-04-04 03:22:26 da2ce7 has joined
 115 2012-04-04 03:26:17 PsiliPharm has joined
 116 2012-04-04 03:29:09 winterblack_ has quit (Read error: Connection reset by peer)
 117 2012-04-04 03:42:47 PsiliPharm has quit (Quit: Colloquy for iPad - http://colloquy.mobi)
 118 2012-04-04 03:42:53 user_ has quit (Quit: Leaving)
 119 2012-04-04 03:48:28 forsetifox has quit (Quit: Page closed)
 120 2012-04-04 03:51:52 forsetifox has joined
 121 2012-04-04 03:55:22 <nanotube> so when does the bitcoin-qt on windows critfix alert expire? it keeps sitting in my taskbar...
 122 2012-04-04 03:56:58 Slix` has quit (Remote host closed the connection)
 123 2012-04-04 03:58:30 <Cory> Update!
 124 2012-04-04 03:58:57 <BlueMatt> nanotube: when you upgrade
 125 2012-04-04 03:59:11 <nanotube> BlueMatt: ah... but i'm not even on windows! :)
 126 2012-04-04 03:59:21 <BlueMatt> upgrade anyway ;)
 127 2012-04-04 03:59:24 <nanotube> hehe yea
 128 2012-04-04 04:01:37 <luke-jr> nanotube: test 0.5.4 :D
 129 2012-04-04 04:03:04 <nanotube> heh
 130 2012-04-04 04:03:12 <nanotube> that's only bitcoind though innit?
 131 2012-04-04 04:04:56 <luke-jr> no
 132 2012-04-04 04:06:49 <nanotube> ah well... maybe. :)
 133 2012-04-04 04:11:21 bd___ is now known as bd_
 134 2012-04-04 04:28:44 molecular has quit (Quit: Leaving)
 135 2012-04-04 04:29:14 brwyatt is now known as brwyatt|Away
 136 2012-04-04 04:29:26 splatster has joined
 137 2012-04-04 04:31:53 jgarzik has quit (Read error: Operation timed out)
 138 2012-04-04 04:32:37 jgarzik has joined
 139 2012-04-04 04:33:02 jgarzik is now known as Guest41970
 140 2012-04-04 04:34:41 ThomasV has joined
 141 2012-04-04 04:35:51 osmosis has quit (Quit: Leaving)
 142 2012-04-04 04:36:59 copumpkin has quit (Remote host closed the connection)
 143 2012-04-04 04:37:14 copumpkin has joined
 144 2012-04-04 04:47:12 sacarlson has quit (Quit: Leaving.)
 145 2012-04-04 04:47:23 sacarlson has joined
 146 2012-04-04 04:47:38 ForceMajeure_ has joined
 147 2012-04-04 04:48:54 ForceMajeure has quit (Ping timeout: 246 seconds)
 148 2012-04-04 04:50:09 <Cory> Can alerts target specific OSs?
 149 2012-04-04 04:51:16 <phantomcircuit> no
 150 2012-04-04 04:54:32 <Cory> Has an invalid alert (not signed by the right pubkey) ever been broadcast?
 151 2012-04-04 04:54:44 <BlueMatt> it cant be
 152 2012-04-04 04:55:05 <BlueMatt> or...it wouldnt get past the first node
 153 2012-04-04 04:55:24 sacarlson has quit (Remote host closed the connection)
 154 2012-04-04 04:56:15 sacarlson has joined
 155 2012-04-04 04:56:33 <Cory> Ah, right. Makes sense.
 156 2012-04-04 04:57:41 winterblack has joined
 157 2012-04-04 05:01:05 Guest41970 has quit (Changing host)
 158 2012-04-04 05:01:05 Guest41970 has joined
 159 2012-04-04 05:01:12 Guest41970 is now known as jgarzik_
 160 2012-04-04 05:01:38 <BlueMatt> jgarzik_: sorry, /ignore me...
 161 2012-04-04 05:03:35 <nameless> !~root@mindjail.subluminal.net|BlueMatt: we do
 162 2012-04-04 05:03:52 <BlueMatt> good to hear someone has sense
 163 2012-04-04 05:04:39 ForceMajeure has joined
 164 2012-04-04 05:05:05 ForceMajeure is now known as Guest20001
 165 2012-04-04 05:06:46 ForceMajeure_ has quit (Ping timeout: 276 seconds)
 166 2012-04-04 05:13:03 gp5st has joined
 167 2012-04-04 05:32:11 dwon has quit (Quit: Leaving)
 168 2012-04-04 05:38:03 Blitzboom has quit (Remote host closed the connection)
 169 2012-04-04 05:40:40 fiddur has joined
 170 2012-04-04 05:41:24 <twobitcoins> If anyone is still looking for a copy of the bad P2SH transaction, I found it.
 171 2012-04-04 05:41:47 <twobitcoins> 01000000019dc23528f5a5f376da3f3f4efd45be8c5b551abdb8093940e0b313de459a
 172 2012-04-04 05:41:49 <twobitcoins> 53b00100000026255121029c7187ecea7f09146820075c3a8de5d33ffbc293b63228ea
 173 2012-04-04 05:41:54 <twobitcoins> 1667c8d3796aff3f51aeffffffff0130570500000000001976a9147288ca9e213c54cb
 174 2012-04-04 05:41:57 <twobitcoins> b2094f00bcf33bfbce691dbb88ac00000000
 175 2012-04-04 05:42:09 Zarutian has quit (Quit: Zarutian)
 176 2012-04-04 05:44:58 <dooglus> "// BIP16 didn't become active until Apr 1 2012 (Feb 15 on testnet)"
 177 2012-04-04 05:45:07 <dooglus> so it's only just started rejecting these blocks
 178 2012-04-04 05:46:21 <dooglus> won't this cause a fork between different client versions?
 179 2012-04-04 05:48:55 <twobitcoins> It has created a bunch of little forks, but they always resolve.  New nodes create blocks that old nodes consider valid and new nodes have the majority of hashing power, so eventually new nodes have the longest chain and old nodes switch to it.
 180 2012-04-04 05:49:46 Blitzboom has joined
 181 2012-04-04 05:49:46 Blitzboom has quit (Changing host)
 182 2012-04-04 05:49:46 Blitzboom has joined
 183 2012-04-04 05:50:10 <dooglus> http://blockchain.info/ tells me that deepbit is building on the invalid block - doesn't it?
 184 2012-04-04 05:53:57 <bitfoo> blockchain.info shows 2 blocks at height 174232
 185 2012-04-04 05:54:02 <bitfoo> and both relayed by deepbit :P
 186 2012-04-04 05:54:06 lh77 has joined
 187 2012-04-04 05:54:14 <twobitcoins> It claims to have heard about the last bad block from Deepbit.  Perhaps Deepbit runs an old node, but I doubt it mines using an old node.
 188 2012-04-04 05:55:44 <SomeoneWeird> ask tycho?
 189 2012-04-04 05:55:53 <BlueMatt> [Tycho]:
 190 2012-04-04 05:57:15 <gmaxwell> Stop.!
 191 2012-04-04 05:57:24 <gmaxwell> God is a @#$@## forest fire.
 192 2012-04-04 05:57:33 <gmaxwell> The reports of pools on blockchain.info are unreliable. In particular the relayed by field is unreliable as an indicator of block origin and there are a great many blocks reported as relayed by deepbit which are not from deepbit.
 193 2012-04-04 05:58:04 <gmaxwell> Many of the orphans produced by 1txn mining mytery have been relayed by deepbit.
 194 2012-04-04 05:58:15 <gmaxwell> Deepbits actual blocks are listed on their website.
 195 2012-04-04 05:58:41 sacarlson has quit (Ping timeout: 240 seconds)
 196 2012-04-04 06:00:38 <gmaxwell> dooglus: No, BIP16 doesn't cause a persistent fork. Old nodes accept the new txns happily.
 197 2012-04-04 06:01:01 da2ce7 has quit (Ping timeout: 252 seconds)
 198 2012-04-04 06:01:13 <dooglus> gmaxwell: so what causes this situation?
 199 2012-04-04 06:02:04 <gmaxwell> It's only a one way disagreement, and a significant supermajority of hashpower (>75%) enforces the P2SH rules— so that chain always overtakes and the old nodes happily reorganize onto it.
 200 2012-04-04 06:02:11 RazielZ has joined
 201 2012-04-04 06:02:12 Seventoes has joined
 202 2012-04-04 06:02:44 <gmaxwell> dooglus: BIP16 enabled nodes will reject some transactions which old nodes see as valid (invalid P2SH transactions). Someone produced one of these transactions.
 203 2012-04-04 06:02:59 <gmaxwell> (presumably intentionally— though it's constructed in a way which could have been an accident)
 204 2012-04-04 06:03:04 <dooglus> the 'bad P2SH tx' - was it created deliberately to mess things up?  or is it a valid transaction for old clients that arises 'naturally'?
 205 2012-04-04 06:03:10 <dooglus> ok
 206 2012-04-04 06:03:24 <gmaxwell> It's not a kind of transaction that anyone would ever have made before...
 207 2012-04-04 06:03:25 <gmaxwell> https://blockchain.info/tx-index/3618498/4005d6bea3a93fb72f006d23e2685b85069d270cb57d15f0c057ef2d5e3f78d2?show_adv=true
 208 2012-04-04 06:04:20 <gmaxwell> Though it could have been an honest attempt at a P2SH transaction that simply didn't account for the er 'surprising' behavior of CHECKMULTISIG.
 209 2012-04-04 06:04:34 <gmaxwell> But only custom code could produce it.
 210 2012-04-04 06:04:54 <gmaxwell> If someone wanted to make an intentionally bad transaction the could have used a much simpler script though.
 211 2012-04-04 06:08:57 splatster has quit (Quit: splatster)
 212 2012-04-04 06:09:44 <gmaxwell> twobitcoins: How'd you get a copy of it?
 213 2012-04-04 06:11:39 Seventoes has left ()
 214 2012-04-04 06:11:59 <twobitcoins> I modified ConnectInputs to dump transactions that fail VerifySignature to debug.log and waited for someone to generate another bad block and send it to me.
 215 2012-04-04 06:12:08 sacarlson has joined
 216 2012-04-04 06:13:18 ThomasV has quit (Ping timeout: 260 seconds)
 217 2012-04-04 06:18:16 chk has joined
 218 2012-04-04 06:22:35 <gmaxwell> Ah, good call to catch it in a bad block.
 219 2012-04-04 06:25:48 Guest20001 is now known as ForceMajeure
 220 2012-04-04 06:37:41 worlemp has quit (Remote host closed the connection)
 221 2012-04-04 06:45:13 mmoya has joined
 222 2012-04-04 07:05:22 Diablo-D3 has quit (Remote host closed the connection)
 223 2012-04-04 07:05:32 ThomasV has joined
 224 2012-04-04 07:09:25 <[Tycho]> What ?
 225 2012-04-04 07:10:07 darkskiez has quit (Ping timeout: 260 seconds)
 226 2012-04-04 07:13:22 <chk> can anyone here assist me with using gmp-proxy?
 227 2012-04-04 07:15:59 <gribble> New news from bitcoinrss: Diapolo opened pull request 1032 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1032>
 228 2012-04-04 07:25:55 ThomasV has quit (Read error: Operation timed out)
 229 2012-04-04 07:29:42 ovidiusoft has joined
 230 2012-04-04 07:30:49 chk has quit (Excess Flood)
 231 2012-04-04 07:31:07 chk has joined
 232 2012-04-04 07:37:06 datagutt has joined
 233 2012-04-04 07:40:43 mmoya has quit (Quit: Saliendo)
 234 2012-04-04 07:49:58 darkskiez has joined
 235 2012-04-04 07:51:01 forsetifox has quit (Quit: Page closed)
 236 2012-04-04 07:57:56 copumpkin has quit (Ping timeout: 260 seconds)
 237 2012-04-04 07:58:29 copumpkin has joined
 238 2012-04-04 08:00:26 t7 has joined
 239 2012-04-04 08:00:33 <BTC_Bear> Don't know if anyone brought this up. I don't have windows, can you 'Clear' the urgent update message.
 240 2012-04-04 08:02:32 t7_ has joined
 241 2012-04-04 08:04:08 Clipse has joined
 242 2012-04-04 08:05:19 t7 has quit (Ping timeout: 252 seconds)
 243 2012-04-04 08:05:23 t7_ is now known as t7
 244 2012-04-04 08:10:07 Snapman is now known as Snapman[afkers]
 245 2012-04-04 08:13:49 da2ce7 has joined
 246 2012-04-04 08:18:06 ovidiusoft2 has joined
 247 2012-04-04 08:19:37 ovidiusoft has quit (Ping timeout: 252 seconds)
 248 2012-04-04 08:21:07 copumpkin has quit (Remote host closed the connection)
 249 2012-04-04 08:21:49 copumpkin has joined
 250 2012-04-04 08:31:47 <t7> have you guys heard of mental poker?
 251 2012-04-04 08:32:39 <t7> i need a communative encryption scheme
 252 2012-04-04 08:35:59 cdecker has joined
 253 2012-04-04 08:37:13 <wumpus> BTC_Bear: you'll have to put up with it until a new release for linux, sorry :)
 254 2012-04-04 08:42:59 ThomasV has joined
 255 2012-04-04 08:50:16 toxicFork has joined
 256 2012-04-04 08:50:25 <toxicFork> has anyone made a desktop wallet app which'd store only the wallet.dat and fetch "wallet status" etc from internet ( e.g. blockexplorer ) ?
 257 2012-04-04 08:50:39 <toxicFork> instead of storing blockchain
 258 2012-04-04 08:52:34 <sturles> Yes.
 259 2012-04-04 08:52:42 <sturles> Electrum.
 260 2012-04-04 08:53:15 <toxicFork> thanks
 261 2012-04-04 08:54:13 <toxicFork> looks great!
 262 2012-04-04 08:55:57 <wumpus> we really need an overview of clients on bitcoin.org
 263 2012-04-04 08:57:27 copumpkin has quit (Ping timeout: 245 seconds)
 264 2012-04-04 08:58:07 copumpkin has joined
 265 2012-04-04 09:09:53 BTC_Bear is now known as BTC_Bear|hbrntng
 266 2012-04-04 09:17:14 toxicFork has left ()
 267 2012-04-04 09:20:16 Fnar has quit (Ping timeout: 245 seconds)
 268 2012-04-04 09:20:56 Fnar has joined
 269 2012-04-04 09:21:58 FellowTraveler has joined
 270 2012-04-04 09:22:23 FellowTraveler has left ()
 271 2012-04-04 09:24:38 Clipse has quit (Ping timeout: 260 seconds)
 272 2012-04-04 09:28:06 Clipse has joined
 273 2012-04-04 09:29:01 [Tycho] has quit (Ping timeout: 245 seconds)
 274 2012-04-04 09:32:50 Turingi has joined
 275 2012-04-04 09:37:36 dvide has joined
 276 2012-04-04 09:41:33 sacarlson has quit (Ping timeout: 260 seconds)
 277 2012-04-04 09:51:48 Diablo-D3 has joined
 278 2012-04-04 09:56:41 _Fireball has quit (Read error: Connection reset by peer)
 279 2012-04-04 09:57:21 _Fireball has joined
 280 2012-04-04 10:00:02 JFK911_ has joined
 281 2012-04-04 10:00:38 _Fireball has quit (Read error: Connection reset by peer)
 282 2012-04-04 10:01:20 _Fireball has joined
 283 2012-04-04 10:01:30 JFK911 has quit (Read error: Connection reset by peer)
 284 2012-04-04 10:07:20 BTC_Bear is now known as hbrntng!~BTC_Bear@unaffiliated/btc-bear/x-5233302|BTC_Bear
 285 2012-04-04 10:18:05 Archer has joined
 286 2012-04-04 10:18:31 Archer is now known as Guest43823
 287 2012-04-04 10:18:56 vextr_ has joined
 288 2012-04-04 10:19:25 barmstro_ has joined
 289 2012-04-04 10:19:54 <sipa> gmaxwell: so you see a getblocks, you reply with invs, but no getdata follows (RE: Azelphur)?
 290 2012-04-04 10:21:00 Turingi has quit (Quit: Leaving)
 291 2012-04-04 10:21:56 midnightmagic_ has joined
 292 2012-04-04 10:21:56 midnightmagic_ has quit (Changing host)
 293 2012-04-04 10:21:56 midnightmagic_ has joined
 294 2012-04-04 10:22:39 _wizkid057 has joined
 295 2012-04-04 10:22:51 Ken`_ has joined
 296 2012-04-04 10:22:52 Staatsfeind has joined
 297 2012-04-04 10:23:19 riush_ has joined
 298 2012-04-04 10:28:11 gp5st1 has joined
 299 2012-04-04 10:29:14 MBS has joined
 300 2012-04-04 10:31:35 nameless has joined
 301 2012-04-04 10:32:25 ski_ has joined
 302 2012-04-04 10:34:03 wood has joined
 303 2012-04-04 10:35:19 Staatsfeind is now known as vragnaroda
 304 2012-04-04 10:37:20 BTC_Bear is now known as BTC_Bear|hbrntng
 305 2012-04-04 10:43:15 iHerb has joined
 306 2012-04-04 10:43:16 <iHerb> Hello, and please sorry for that =) Have a nice day!
 307 2012-04-04 10:43:17 <iHerb> iHerb.com - Vitamins, Supplements & Natural Health Products & etc!
 308 2012-04-04 10:43:17 <iHerb> please used and saved $10 with the iHerb coupon code VAD515 or please visit my link http://iherb.com?rcode=vad515, and free Shipping on orders over $20. And FREE SAMPLE for every purchase!
 309 2012-04-04 10:45:31 <Graet> sorry for spamming offtop[ic advertising, why not just dont do it?
 310 2012-04-04 10:45:47 iHerb has left ()
 311 2012-04-04 10:45:50 <upb> heh
 312 2012-04-04 10:58:45 sje has joined
 313 2012-04-04 10:58:46 sje has quit (Changing host)
 314 2012-04-04 10:58:46 sje has joined
 315 2012-04-04 11:02:48 stranger_ has joined
 316 2012-04-04 11:11:29 theorb has joined
 317 2012-04-04 11:16:10 theorbtwo has quit (Ping timeout: 264 seconds)
 318 2012-04-04 11:16:17 theorb is now known as theorbtwo
 319 2012-04-04 11:17:00 Nicksasa has joined
 320 2012-04-04 11:23:18 stranger_ has quit (Quit: Page closed)
 321 2012-04-04 11:29:32 TD has joined
 322 2012-04-04 11:36:01 toffoo has quit ()
 323 2012-04-04 11:42:29 davout has joined
 324 2012-04-04 11:42:30 davout has quit (Read error: Connection reset by peer)
 325 2012-04-04 11:43:35 <wumpus> vitamin spam on freenode, we're becoming too mainstream :/
 326 2012-04-04 11:43:45 Joric has joined
 327 2012-04-04 11:43:45 Joric has quit (Changing host)
 328 2012-04-04 11:43:45 Joric has joined
 329 2012-04-04 11:44:07 <Graet> at least he could have given prices in BTC :P
 330 2012-04-04 11:45:14 <lh77> :P
 331 2012-04-04 11:45:20 <wumpus> hehe
 332 2012-04-04 11:59:18 ThomasV has quit (Quit: Leaving)
 333 2012-04-04 12:07:42 denisx has joined
 334 2012-04-04 12:07:56 <denisx> what happened to blockchain?
 335 2012-04-04 12:08:04 <lh77> maintanence
 336 2012-04-04 12:08:08 <lh77> from 12am to 2pm
 337 2012-04-04 12:08:10 <lh77> gmt
 338 2012-04-04 12:09:39 <denisx> is there a city which als always GMT? ;)
 339 2012-04-04 12:09:42 <denisx> is
 340 2012-04-04 12:10:06 <egecko> london.
 341 2012-04-04 12:10:23 <denisx> but they have summertime
 342 2012-04-04 12:10:46 <sipa> afaik no place on earth ha GMT all year long
 343 2012-04-04 12:10:49 <sipa> *has
 344 2012-04-04 12:14:33 agricocb has quit (Quit: Leaving.)
 345 2012-04-04 12:26:47 slush has joined
 346 2012-04-04 12:33:10 m00p has quit (Quit: Leaving)
 347 2012-04-04 12:34:49 paulo has joined
 348 2012-04-04 12:35:54 agricocb has joined
 349 2012-04-04 12:38:11 ThomasV has joined
 350 2012-04-04 12:39:39 Zarutian has joined
 351 2012-04-04 12:40:12 phma has quit (Remote host closed the connection)
 352 2012-04-04 12:40:51 Nicksasa has quit (Ping timeout: 265 seconds)
 353 2012-04-04 12:47:22 paulo is now known as away!~foo@112.204.27.38|paulo_
 354 2012-04-04 12:47:54 gp5st1 has left ()
 355 2012-04-04 12:48:23 erle- has joined
 356 2012-04-04 12:48:47 rebroad has joined
 357 2012-04-04 12:50:39 <rebroad> hi..... I'm looking at bitcoin-qt, and noticing that DBFlush(false) is taking over 4 minutes to run, even if the client has been running only a few seconds... does anyone have any idea what it's doing that takes so long please?
 358 2012-04-04 12:51:32 <rebroad> and given that so many people are apparently having it kill -9ed, then does this mean the DBFlush could be skipped altogether?
 359 2012-04-04 12:52:25 <sipa> it's cleaning the log files
 360 2012-04-04 12:52:38 <sipa> not sure how much can be done about that, but 4 minutes sounds incredibly long
 361 2012-04-04 12:52:47 <sipa> slow disk, and right when you were downloading blocks?
 362 2012-04-04 12:53:06 <rebroad> sipa, 268451ms to be precise...
 363 2012-04-04 12:53:32 <rebroad> it's when I quit (when it runs DBFlush)
 364 2012-04-04 12:53:38 <sipa> yes
 365 2012-04-04 12:53:58 <sipa> imho, a "Shutting dow (please don't turn off your computer)" should be shown during shutdown
 366 2012-04-04 12:54:02 <sipa> in the gui
 367 2012-04-04 12:54:54 <rebroad> sipa.... but if someone chooses to shutdown their OS with bitcoin running, then generally it ends up being kill -9ed I think, because it takes too long by most application standards
 368 2012-04-04 12:55:26 <rebroad> (sorry, no idea why I put the four dots after your name there)
 369 2012-04-04 12:55:28 malaimo has joined
 370 2012-04-04 12:56:16 <rebroad> if I edited DBFlush to simply return before the loop... I wonder, what would be the harm...?
 371 2012-04-04 12:56:21 <sipa> no
 372 2012-04-04 12:56:32 <sipa> but the database log files wouldn't be flushed
 373 2012-04-04 12:56:44 <sipa> so you could not copy a wallet.dat file in or out, for example
 374 2012-04-04 12:57:58 <rebroad> Shall I raise an issue for this? I think >4 minutes when told to shutdown is way too long... and it seems to be getting longer... whatever it's doing should be done during running, not during shutdown, IMHO
 375 2012-04-04 12:58:31 <sipa> maybe the solution is only flushing the lsn's for the wallet file, and not for the transaction database
 376 2012-04-04 12:58:52 <rebroad> it is the blkindex.dat flish...
 377 2012-04-04 12:58:59 <sipa> yes, i know
 378 2012-04-04 12:59:06 <rebroad> (or so it says)
 379 2012-04-04 12:59:27 <jgarzik_> IMO a nice increment step is simply having separate db environments for wallet vs. "other stuff"
 380 2012-04-04 12:59:28 <rebroad> I wasn't aware DBFlush was used for wallet.dat, debug.log doesn't seem to suggest it's used for the wallet.dat
 381 2012-04-04 12:59:33 <jgarzik_> *incremental
 382 2012-04-04 12:59:42 <jgarzik_> solves _many_ problems
 383 2012-04-04 12:59:45 jgarzik_ is now known as jgarzik
 384 2012-04-04 12:59:50 <sipa> jgarzik_: IMO the solution is not using bdb for wallets :)
 385 2012-04-04 13:00:21 <jgarzik> sipa: sure, but note word "incremental"  all these flush and log problems simply go away
 386 2012-04-04 13:00:53 <jgarzik> sipa: you would have the same problem with sqlite or another db system, because we're simply trying to force one ACID environment for all our data
 387 2012-04-04 13:01:25 <jgarzik> db usage mismatch with data
 388 2012-04-04 13:01:25 <sipa> since addrman, addr.dat is always written in its entirety now; they're still using bdb, but there are easier ways if you keep everything in memory
 389 2012-04-04 13:01:50 <sipa> wallets are loaded entirely into memory, and are actually only read at startup
 390 2012-04-04 13:02:00 <rebroad> so, if I get DBFlush to return when it notices it's blkindex.dat, and only blkindex.dat, would that be safe to do?
 391 2012-04-04 13:02:00 <sipa> the only thing that really needs a database, is the blockchain info
 392 2012-04-04 13:02:02 <jgarzik> yep
 393 2012-04-04 13:02:03 <sipa> bdb is fine for that
 394 2012-04-04 13:02:33 <sipa> rebroad: it would be safe, yes, but you would be left with log files that are part of the database
 395 2012-04-04 13:02:39 <sipa> and cannot be removed
 396 2012-04-04 13:03:03 <sipa> + start-up would take longer the next time
 397 2012-04-04 13:03:08 <rebroad> sipa, would more disk space be used up as a result?
 398 2012-04-04 13:03:32 <sipa> depends on what you mean by that
 399 2012-04-04 13:03:42 <sipa> more disk space when? on average or maximally?
 400 2012-04-04 13:03:45 <rebroad> but how come the flush takes longer than the amount of time the client has been running? Sometimes I exit after running only for 30 seconds and it takes 4 minutes...
 401 2012-04-04 13:04:13 <sipa> writing log files = sequential writes; flushing log files = random writes
 402 2012-04-04 13:04:22 <rebroad> I would have thought the flush would take longer the more stuff that had happened..
 403 2012-04-04 13:04:52 <rebroad> and sometimes the whole OS seems to hang during a flush too...
 404 2012-04-04 13:04:58 <sipa> what OS?
 405 2012-04-04 13:05:02 <rebroad> ubuntu
 406 2012-04-04 13:05:10 <sipa> strange, no problems here
 407 2012-04-04 13:05:34 <rebroad> well, i mean... disk I/O related things seem to hang/go_slow
 408 2012-04-04 13:06:09 Turingi has joined
 409 2012-04-04 13:06:09 Turingi has quit (Changing host)
 410 2012-04-04 13:06:09 Turingi has joined
 411 2012-04-04 13:06:32 <sipa> jgarzik: what i mean is: if the only thing that really needs a db environment is the block index, there's no point in adding the extra complexity for maintaining multiple db env, if the next step is going back to only a db env for the block chain anyway
 412 2012-04-04 13:06:46 <sipa> *single
 413 2012-04-04 13:06:51 <rebroad> what is the advantage in having the data in blkindex.dat instead of in the database files? and why would it take longer to start up if it's not flushed at shutdown..?
 414 2012-04-04 13:07:10 <sipa> i believe it will flush them on startup anyway, otherwise
 415 2012-04-04 13:07:22 <sipa> log files are slow to access, but fast to write
 416 2012-04-04 13:07:34 <sipa> database files are faster to access, but slower to write
 417 2012-04-04 13:08:00 <sipa> so the log functions as a temporary cache for queued changes that must be written to the database still
 418 2012-04-04 13:08:20 <rebroad> sipa, can it write them to the database during runtime rather than shutdown?
 419 2012-04-04 13:08:32 <sipa> of course, it does so continuously
 420 2012-04-04 13:08:35 <rebroad> or do something to reduce the load at shutdown?
 421 2012-04-04 13:08:47 <rebroad> it'd be better at startup than shutdown, I'd have thought
 422 2012-04-04 13:08:51 <sipa> it's strange that it is so *so* slow for you
 423 2012-04-04 13:09:02 <sipa> which version are you running, by the way?
 424 2012-04-04 13:09:18 <sipa> 0.6.0rc5 had a problem that it left way too large log files
 425 2012-04-04 13:09:23 <rebroad> 0.6.0.4-beta
 426 2012-04-04 13:09:27 <sipa> right
 427 2012-04-04 13:09:34 <sipa> just upgrade to the latest version then, please
 428 2012-04-04 13:09:43 <rebroad> ah, ok..
 429 2012-04-04 13:09:46 <rebroad> thanks
 430 2012-04-04 13:11:29 mmoya has joined
 431 2012-04-04 13:19:01 rebroad has quit (Ping timeout: 246 seconds)
 432 2012-04-04 13:22:05 gavinandresen has joined
 433 2012-04-04 13:25:23 word_ has joined
 434 2012-04-04 13:25:33 Tykling has quit (Read error: Connection reset by peer)
 435 2012-04-04 13:28:06 word has quit (Ping timeout: 260 seconds)
 436 2012-04-04 13:30:27 Tykling has joined
 437 2012-04-04 13:38:43 rebroad has joined
 438 2012-04-04 13:40:18 Ukyo has joined
 439 2012-04-04 13:40:27 Ukyo is now known as Ukto
 440 2012-04-04 13:44:14 <Perlboy> So, computationally, how hard is it to brute force a wallet key?
 441 2012-04-04 13:44:43 <rebroad> Perlboy, probably a question for #bitcoin I suspect
 442 2012-04-04 13:45:02 <Perlboy> ok
 443 2012-04-04 13:45:16 pavel__ has quit (Ping timeout: 245 seconds)
 444 2012-04-04 13:45:45 <MagicalTux> Perlboy: quite
 445 2012-04-04 13:45:56 <Joric> 2^256
 446 2012-04-04 13:46:32 <Joric> except a few special points, it was already discussed here
 447 2012-04-04 13:47:01 <Perlboy> got logs?
 448 2012-04-04 13:47:23 <SomeoneWeird> Joric; the hex conversion thing?
 449 2012-04-04 13:47:34 <MagicalTux> 0~FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFEBAAEDCE6AF48A03BBFD25E8CD0364141
 450 2012-04-04 13:49:16 <MagicalTux> need some extra ecdsa, sha256 and ripemd160 to get the actual bitcoin address from a private addr to know if you found the right one
 451 2012-04-04 13:49:53 * luke-jr facepalms.
 452 2012-04-04 13:50:03 <SomeoneWeird> luke-jr, ?
 453 2012-04-04 13:50:20 <luke-jr> found out why my node reacted differently to the viral P2SH redemption
 454 2012-04-04 13:50:26 <Joric> <sipa> ok, so bitcoin secret keys have 255.999999999999999999999999999999999999995 bits of entropy
 455 2012-04-04 13:50:27 copumpkin has quit (Quit: Computer has gone to sleep.)
 456 2012-04-04 13:51:00 <Joric> ;;calc log(0xFFFFFFFF00000000FFFFFFFFFFFFFFFFBCE6FAADA7179E84F3B9CAC2FC632551)/log(2)
 457 2012-04-04 13:51:00 <gribble> 255.999998888
 458 2012-04-04 13:52:12 <SomeoneWeird> there ya go
 459 2012-04-04 13:52:12 <SomeoneWeird> lol
 460 2012-04-04 13:52:34 <luke-jr> it was because I was testing with 0.4.5, and left out [db9f2e01 Do not invoke anti-DoS system for invalid BIP16 transactions] for 0.4.x since it has no anti-DoS :/
 461 2012-04-04 13:53:38 <SomeoneWeird> heh
 462 2012-04-04 13:56:58 <sipa> Perlboy: brute-forcing a public key is 2^256; brute-forcing an address is only 2^160
 463 2012-04-04 13:57:13 <sipa> Perlboy: if all you want is being able to spend funds sent to a particular address, the latter suffices
 464 2012-04-04 13:57:40 <Perlboy> 2^160 * hex?
 465 2012-04-04 13:59:27 <Perlboy> 1 cpu with 1 hz on 2^160 is still 4.8 with 42 zeros.
 466 2012-04-04 13:59:53 <Perlboy> divide it by 12 cores (pure cpu brute)
 467 2012-04-04 14:01:17 <Perlboy> lowest i could get to was 16 zeros
 468 2012-04-04 14:01:33 <Perlboy> and that was assuming like 100 machines
 469 2012-04-04 14:01:35 <Perlboy> ie. hard. :)
 470 2012-04-04 14:01:39 <Perlboy> goodo
 471 2012-04-04 14:01:49 <Perlboy> thanks </end spam>
 472 2012-04-04 14:01:58 <SomeoneWeird> lol
 473 2012-04-04 14:03:48 <Joric> 2^160 * hex? wtf is hex? :D
 474 2012-04-04 14:04:28 <Joric> 2d^160d
 475 2012-04-04 14:05:25 <SomeoneWeird> lol
 476 2012-04-04 14:06:29 <forrestv> it would cost about 10^34 bitcoins to bruteforce a single address (assuming ripemd160 is about as hard as sha256)
 477 2012-04-04 14:06:37 <Joric> https://bitcointools.appspot.com/?t=Sddm
 478 2012-04-04 14:07:14 <helo> sipa: don't you have to have the private key to spend coin at an address?
 479 2012-04-04 14:07:37 <helo> brute forcing the ripe would only give the public key wouldn't it?
 480 2012-04-04 14:08:12 <forrestv> no, you'd try to find a private key whose pubkey matches the hash
 481 2012-04-04 14:08:21 <forrestv> so it would give you a private key, not necessarily the original one
 482 2012-04-04 14:08:35 <sipa> exactly
 483 2012-04-04 14:09:16 <helo> so by "brute force an address" you mean "generate private key -> derive public key -> ripe160 to address"
 484 2012-04-04 14:09:46 <SomeoneWeird> yes
 485 2012-04-04 14:10:51 <luke-jr> don't forget the sha256 in there :P
 486 2012-04-04 14:13:02 copumpkin has joined
 487 2012-04-04 14:16:31 denisx_ has joined
 488 2012-04-04 14:16:35 <gavinandresen> Joric: neat!  Are you doing flood control to prevent some jerk from costing you a bunch of your appengine budget?
 489 2012-04-04 14:18:09 denisx has quit (Ping timeout: 272 seconds)
 490 2012-04-04 14:18:09 denisx_ is now known as denisx
 491 2012-04-04 14:22:00 <brokenwallet> oh shit!
 492 2012-04-04 14:22:16 <brokenwallet> sipa I found the correct pass, and it wasnt the same pass that was crashing my client
 493 2012-04-04 14:22:55 <brokenwallet> gmaxwell was right i think
 494 2012-04-04 14:23:25 <brokenwallet> do you want my wallet to analyze why it was causing a crash on the incorrect pass?
 495 2012-04-04 14:24:48 <brokenwallet> and PM me an address sipa to send you some coins for the help (even though it was my stupidness that caused the problem)
 496 2012-04-04 14:25:47 <sipa> brokenwallet: whatever happens, a bad passphrase should not cause your client to crash
 497 2012-04-04 14:26:23 <sipa> if you use the right passphrase, does the client just work (and able to send coins), or are you still using my patched version?
 498 2012-04-04 14:26:24 topace has quit (Read error: Connection reset by peer)
 499 2012-04-04 14:27:26 <brokenwallet> i'm using the latest and i got it to work, basically i typed out every permutation that could be the password that i was using and ran through them
 500 2012-04-04 14:28:04 <SomeoneWeird> lol
 501 2012-04-04 14:28:24 <brokenwallet> first i ran through them with your modified client and nothing worked. Could this be due to it stripping the incorrectly decrypted keys? Meaning if i try a wrong pass then the right pass with your client would it attempt to re-read the wallet?
 502 2012-04-04 14:28:25 <SomeoneWeird> ... thats one way
 503 2012-04-04 14:29:01 <brokenwallet> or use the one in memory that at that point had the keys removed (at that point)
 504 2012-04-04 14:29:21 <brokenwallet> let me try again to verify
 505 2012-04-04 14:29:41 Zarutian has quit (Ping timeout: 240 seconds)
 506 2012-04-04 14:32:10 z310 has quit (Ping timeout: 252 seconds)
 507 2012-04-04 14:33:49 <brokenwallet> off topic, why does the "pay to" field allow 35 chars when it looks like bitcoin addresses are always 34?
 508 2012-04-04 14:34:16 <gmaxwell> brokenwallet: congrats!
 509 2012-04-04 14:35:21 Zarutian has joined
 510 2012-04-04 14:36:13 <luke-jr> wumpus: ping
 511 2012-04-04 14:36:15 fiddur has quit (Quit: Leaving.)
 512 2012-04-04 14:36:56 <luke-jr> wumpus: 5a60b66 Use a messagebox to display the error when -server is provided without providing a rpc password <-- does this INTENTIONALLY change behaviour so that no-rpcpassword is fatal for Bitcoin-Qt? O.o
 513 2012-04-04 14:37:18 <helo> brokenwallet: good question... as far as i know they can never be 35
 514 2012-04-04 14:37:31 <gmaxwell> brokenwallet: Rather than your coins... perhaps we could have your broken backup? e.g. transfer all your coins to a new wallet, and gis us your old backup so we can reproduce?  (This won't be viable if you have addresses published anywhere)
 515 2012-04-04 14:37:34 <gavinandresen> multisig (BIP16) addresses on testnet can be 35.
 516 2012-04-04 14:37:54 <gavinandresen> ( I don't remember if mainnet BIP16 addresses can be 35 or not)
 517 2012-04-04 14:37:58 <brokenwallet> gmaxwell that is fine with me, anything to support the dev work
 518 2012-04-04 14:38:24 <brokenwallet> and i use unique passes in most places including my wallet so i dont care if you know them
 519 2012-04-04 14:38:47 <brokenwallet> i'm just going to re-verify everything and document it for #1024
 520 2012-04-04 14:39:30 z310 has joined
 521 2012-04-04 14:40:17 <gmaxwell> Great.
 522 2012-04-04 14:40:21 <brokenwallet> gmaxwell what would be the best way to get you the wallet? email?
 523 2012-04-04 14:41:38 <gmaxwell> Email would be fine. Make sure to send your funds to a totally new wallet before emailing me— (Not that I'm going to steal your coins, but if email goblins get them I don't want to deal with doubts!)
 524 2012-04-04 14:42:07 <Joric> gavinandresen, not really, i have no flood control atm, and i got a feeling i have to rewrite all that to js after being featured in Forbes Magazine http://www.forbes.com/sites/jonmatonis/2012/03/12/brainwallet-the-ultimate-in-mobile-money
 525 2012-04-04 14:42:21 topace has joined
 526 2012-04-04 14:42:36 <brokenwallet> yep i have a very elementary understanding of bitcoin and how there are 100 pregenerated keys per wallet
 527 2012-04-04 14:42:48 <gmaxwell> Great.
 528 2012-04-04 14:43:23 TheSeven is now known as [7]
 529 2012-04-04 14:45:12 <gavinandresen> Joric: if you're using django, something like this:  http://django-ratelimit-backend.readthedocs.org/en/latest/    can be easy to install and will keep the jerks at bay
 530 2012-04-04 14:46:06 <Joric> yeah it's all django/python
 531 2012-04-04 14:47:54 <gavinandresen> Joric: code that the Faucet uses to ratelimit requests:  https://gist.github.com/63bae3ff3667880719c1
 532 2012-04-04 14:49:00 <Joric> gavinandresen, did you think about storing the blockchain in the big table?
 533 2012-04-04 14:49:31 <Joric> it allows up to 5 gb per app atm
 534 2012-04-04 14:49:44 <Joric> no sockets though :(
 535 2012-04-04 14:49:52 <gavinandresen> Joric: sure, you could. But you still need a persistent bitcoind because of the no sockets....
 536 2012-04-04 14:49:53 <Joric> if only there were http nodes
 537 2012-04-04 14:50:45 <gavinandresen> Wouldn't be too hard to create a HTTP node that just send new blocks or transactions to registered HTTP addresses.
 538 2012-04-04 14:51:04 <gavinandresen> Might even be a good business if there are people who don't want to run bitcoind's for some reason.
 539 2012-04-04 14:51:44 <gavinandresen> (I doubt it'd be a good business yet, I doubt there are many people who'd be interested and willing to pay)
 540 2012-04-04 14:56:11 ThomasV has quit (Ping timeout: 265 seconds)
 541 2012-04-04 14:58:01 nonick has quit (Remote host closed the connection)
 542 2012-04-04 14:58:37 nonick has joined
 543 2012-04-04 15:09:04 ThomasV has joined
 544 2012-04-04 15:10:14 Cablesaurus has quit (Quit: Now if you will excuse me, I have a giant ball of oil to throw out my window)
 545 2012-04-04 15:14:22 MBS has joined
 546 2012-04-04 15:16:26 <luke-jr> any last-minute bugfixes for stable? :P
 547 2012-04-04 15:18:07 <gavinandresen> 0.6 is the stable release right now.
 548 2012-04-04 15:19:50 da2ce7 has quit (Ping timeout: 252 seconds)
 549 2012-04-04 15:20:22 paraipan has quit (Ping timeout: 276 seconds)
 550 2012-04-04 15:23:04 paraipan has joined
 551 2012-04-04 15:27:59 ovidiusoft2 has quit (Quit: Ex-Chat)
 552 2012-04-04 15:41:56 vigilyn has quit (Ping timeout: 245 seconds)
 553 2012-04-04 15:42:43 <gribble> New news from bitcoinrss: sipa opened pull request 1033 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1033>
 554 2012-04-04 15:45:12 slush has quit (Ping timeout: 260 seconds)
 555 2012-04-04 15:50:07 <gavinandresen> sipa: did you consider getting rid of the CRITICAL_BLOCK macro entirely ?
 556 2012-04-04 15:50:18 <gavinandresen> (and just using objects scoped in blocks)
 557 2012-04-04 15:50:22 Nicksasa has joined
 558 2012-04-04 15:50:22 Nicksasa has quit (Changing host)
 559 2012-04-04 15:50:22 Nicksasa has joined
 560 2012-04-04 15:50:44 <sipa> gavinandresen: sure, but that's a larger change
 561 2012-04-04 15:51:23 <gavinandresen> Larger in terms of lines of code changed... but cleaner
 562 2012-04-04 15:51:31 <gavinandresen> and more in the spirit of C++
 563 2012-04-04 15:51:39 <gavinandresen> and easier to debug
 564 2012-04-04 15:51:39 <sipa> true
 565 2012-04-04 15:52:47 <sipa> i'm kinda used to CRITICAL_BLOCK nowadays, identifying places where locks were taken was a lot harder in the libcoin source :)
 566 2012-04-04 15:52:58 <sipa> no, i agree, it would be cleaner to get rid of them
 567 2012-04-04 15:53:26 <gavinandresen> Not a high priority if it is a ton of work.
 568 2012-04-04 15:53:43 <gavinandresen> By the way, I did some preliminary looking at using boost::thread instead of CREATE_THREAD
 569 2012-04-04 15:54:02 <gavinandresen> ... and the ugly nasty vnThreadsRunning[] array
 570 2012-04-04 15:55:19 <gavinandresen> And I don't think it would be terribly hard to do it The Right Way -- using boost::thread everywhere, using thread.join at shutdown to wait for threads to exit after using the boost::thread interruption mechanism, etc....
 571 2012-04-04 15:56:27 <gavinandresen> I believe the only reason Satoshi didn't use boost::thread is because he couldn't figure out how to set the thread priority lower for mining threads.  And that's not an issue any more
 572 2012-04-04 15:58:04 t7 has quit (Quit: ChatZilla 0.9.88.1 [Firefox 12.0/20120328051619])
 573 2012-04-04 15:59:33 xenland has joined
 574 2012-04-04 15:59:41 <sipa> gavinandresen: nice, so all that would be required is making sure a thread deals with exceptions thrown from sleep or wait
 575 2012-04-04 15:59:49 agricocb has quit (Quit: Leaving.)
 576 2012-04-04 15:59:59 <sipa> and i believe most are in a clean state before sleeping anyway
 577 2012-04-04 16:01:01 <gavinandresen> sipa: yes, fShutdown ugliness mostly goes away, too
 578 2012-04-04 16:02:10 <gavinandresen> sipa: I've been busy doing taxes and other non-coding stuff, but plan on tackling it soon.  But feel free to do it yourself if you feel inspired....
 579 2012-04-04 16:03:35 <gavinandresen> (I sidetracked myself experimenting with a template/boost::functional replacement to IMPLEMENT_RANDOMIZE_STACK... I really don't like macros....)
 580 2012-04-04 16:03:39 sje has quit (Read error: Connection reset by peer)
 581 2012-04-04 16:06:17 TD has quit (Quit: TD)
 582 2012-04-04 16:08:12 <rebroad> sorry for the git-noob question here.. but is there a way I can use git to "import" the latest patches, but without it removing the changes I've also made, please?
 583 2012-04-04 16:08:40 <sipa> rebroad: merge or rebase
 584 2012-04-04 16:08:44 <sipa> depending on what you want
 585 2012-04-04 16:09:05 TD has joined
 586 2012-04-04 16:09:15 <rebroad> sipa, well, I'd like to bring in the changes since rc4.. but there are some minor changes I've made (e.g. commenting out from of the debug printfs) that I'd like to retain if poss
 587 2012-04-04 16:09:45 <sipa> rebasing allows you to modify your own commits so that they look like they were made to another base
 588 2012-04-04 16:09:57 <sipa> merging just combines your commits with other commits
 589 2012-04-04 16:10:01 <luke-jr> …
 590 2012-04-04 16:10:07 <luke-jr> C++ templates are just super macros
 591 2012-04-04 16:10:32 <rebroad> also, I'm making various changes for different reasons, and I could do with some way of grouping those changes together, so I can keep them, but they'll be stored as separate patches/subprojects in github, so I can perhaps do pulls(?) on them separately at some future stage
 592 2012-04-04 16:11:07 <rebroad> sipa... I think rebasing is probably what I need then. (although I'm not entirely sure what a "base" is)
 593 2012-04-04 16:11:19 <sipa> rebroad: you made a patch against rc4
 594 2012-04-04 16:11:26 <rebroad> sipa, yes, well, several
 595 2012-04-04 16:11:38 <sipa> you want a patch that does functionally the same thing, but "applies to" the current head
 596 2012-04-04 16:11:40 <luke-jr> [12:05:07] <sunbird> anyonne had trouble senidng btc using latest bitcoin-qt on ubuntu lts? when i send to a known-good address, the client hangs.
 597 2012-04-04 16:12:07 <rebroad> I'd like to do various patches, which I'd like to be able to submit for others to use, but each time the main branch moves along, I'd like to keep up, but without undoing the patches I've done
 598 2012-04-04 16:12:38 <rebroad> sipa, yes
 599 2012-04-04 16:12:42 <luke-jr> rebroad: like this? https://github.com/luke-jr/bitcoin/branches
 600 2012-04-04 16:12:55 sunbird has joined
 601 2012-04-04 16:13:15 <sipa> rebroad: ok, check out the branch you want to rebase
 602 2012-04-04 16:13:35 <rebroad> luke-jr, yes, I think so.. well, also like sipa's branches too... but, will I have to keep rebasing them all so that they continue to be useable by people using the latest head(?)?
 603 2012-04-04 16:13:37 <sipa> and do a git fetch upstream, so you have the latest upstream master
 604 2012-04-04 16:13:57 <luke-jr> rebroad: if you want to rebase every commit. usually it's best to just rebase when something conflicts
 605 2012-04-04 16:14:04 <rebroad> sipa, do I need to do the checkout into a different directory to the directory I'm storing my patches in?
 606 2012-04-04 16:14:06 <luke-jr> rebroad: then people using it can just merge
 607 2012-04-04 16:14:10 <sipa> rebroad: no
 608 2012-04-04 16:14:36 <sipa> i suspect every patch you want to make is in a separate branch?
 609 2012-04-04 16:14:43 * luke-jr stays out of sipa's way, since he is sure to do rebasing differently.
 610 2012-04-04 16:14:43 <sunbird> hi all. i'm having difficulty sending btc on latest stable bitcoin-qt for ununtu lts. client hangs on sending to a known-good address. thoughts on troubleshooting? i will try latest rc2 as first stepp.
 611 2012-04-04 16:14:45 <sipa> or you only have one, and it's in master?
 612 2012-04-04 16:14:56 <rebroad> sipa, erm... not so far, but I plan to rejig them so that they are....
 613 2012-04-04 16:15:11 <sipa> rebroad: ok, so how do you have them now?
 614 2012-04-04 16:15:29 <rebroad> sipa, well, i did create a branch for the patch I uploaded to github... brb
 615 2012-04-04 16:15:31 <luke-jr> sunbird: fwiw, 0.5.4rc2 is a downgrade from 0.6.0; but it helps determine if the bug is new, or old
 616 2012-04-04 16:16:06 <sipa> sunbird: anything that appears in debug.log when sending?
 617 2012-04-04 16:20:45 t7 has joined
 618 2012-04-04 16:21:49 xenland has quit (Quit: Leaving)
 619 2012-04-04 16:28:46 <Blitzboom> have you guys seen this? http://mintchipchallenge.com/
 620 2012-04-04 16:29:00 <Diablo-D3> yes.
 621 2012-04-04 16:29:08 <Blitzboom> is it a threat?
 622 2012-04-04 16:29:23 <Blitzboom> i mean, will people actually favor this?
 623 2012-04-04 16:30:39 agricocb has joined
 624 2012-04-04 16:30:41 <wumpus> luke-jr: yes, providing no rpcpasswd but specifying -server is fatal (it used to be fatal, but this was commented out conditionally because it resulted in a segfault)
 625 2012-04-04 16:31:38 <luke-jr> used to be fatal? when? O.o
 626 2012-04-04 16:32:00 <luke-jr> v0.5.0 has it non-fatal for Bitcoin-Qt at least, glancing at the code
 627 2012-04-04 16:32:34 <Blitzboom> >the supply of coin is not fixed
 628 2012-04-04 16:32:38 <Blitzboom> ok, forget that crap then
 629 2012-04-04 16:32:58 <sipa> Blitzboom: is it a separate currency, or tied to another?
 630 2012-04-04 16:33:41 <Blitzboom> i dunno http://developer.mintchipchallenge.com
 631 2012-04-04 16:33:48 <Blitzboom> you devs make sense of it
 632 2012-04-04 16:33:53 <Diablo-D3> Blitzboom: no
 633 2012-04-04 16:33:56 <Diablo-D3> it doesnt do anything
 634 2012-04-04 16:34:45 * Graet submits the bitcoin whitepaper
 635 2012-04-04 16:34:46 <Graet> :P
 636 2012-04-04 16:34:53 <Diablo-D3> they're not even coins
 637 2012-04-04 16:34:59 <Blitzboom> what are they
 638 2012-04-04 16:35:03 <Diablo-D3> they're just cryptograhic transactions in CAD
 639 2012-04-04 16:35:07 <Blitzboom> lol
 640 2012-04-04 16:35:10 <Diablo-D3> completely offline
 641 2012-04-04 16:35:16 <Diablo-D3> they're basically trying to make secure credit cards
 642 2012-04-04 16:35:24 <Blitzboom> ok
 643 2012-04-04 16:35:30 <Diablo-D3> it doesnt work because the keys are created by a central authority
 644 2012-04-04 16:35:37 <Diablo-D3> its broken like every other DRM system
 645 2012-04-04 16:35:52 <Blitzboom> The Royal Canadian Mint, for example, is exploring how to issue digital currency in the future. Its chief financial officer Marc Brule said Bitcoin's biggest problem was that it is not backed by anything.
 646 2012-04-04 16:35:52 <Blitzboom> "The system we would bring in would be backed by a fund," he told Reuters. "Bitcoin may work for the small group of people that believe in its value, but that could change very suddenly."
 647 2012-04-04 16:36:06 <Blitzboom> now that part from the reuters article makes sense
 648 2012-04-04 16:36:14 <Blitzboom> yeah, it’s not backed by anything
 649 2012-04-04 16:36:18 <Blitzboom> wait, let’s back it with FIAT MONEY
 650 2012-04-04 16:36:19 <Diablo-D3> yeah, they're trying to cash in on the bitcoin craze
 651 2012-04-04 16:36:20 rebroad has quit (Read error: Connection reset by peer)
 652 2012-04-04 16:36:21 <Diablo-D3> morons
 653 2012-04-04 16:36:28 <Diablo-D3> they admit it up front
 654 2012-04-04 16:36:41 <Graet> omg gold may work for the small group of people that believe in its value, but that could change very suddenly.
 655 2012-04-04 16:36:42 <Graet> ....
 656 2012-04-04 16:36:49 <Graet> :P
 657 2012-04-04 16:37:01 <Blitzboom> gold doesn’t even work
 658 2012-04-04 16:37:05 <Blitzboom> noone uses that shit
 659 2012-04-04 16:37:06 <Graet> :)
 660 2012-04-04 16:37:14 <Blitzboom> no seriously, it’s not a currency
 661 2012-04-04 16:37:15 <Diablo-D3> https://news.ycombinator.com/item?id=3797977
 662 2012-04-04 16:37:19 <Graet> i know
 663 2012-04-04 16:37:43 <luke-jr> the only reason gold can't function as a currency is that its value is too high
 664 2012-04-04 16:37:56 <Blitzboom> are you stupid
 665 2012-04-04 16:38:03 <luke-jr> I mean, who wants to be trading coins worth $1500 ea?
 666 2012-04-04 16:38:07 <sunbird> luke-jr: ah, thanks for clarifying.
 667 2012-04-04 16:38:09 <luke-jr> Blitzboom: no u
 668 2012-04-04 16:38:11 <Blitzboom> what’s the difference?
 669 2012-04-04 16:38:15 <Blitzboom> make smaller denominations
 670 2012-04-04 16:38:19 <sipa> In Belgium we have something called Proton: it's cash on a card, that does not require online transactions to pay with
 671 2012-04-04 16:38:25 <sipa> Is it just that?
 672 2012-04-04 16:38:28 <luke-jr> Blitzboom: then it gets too small to handle sanely
 673 2012-04-04 16:38:44 <Blitzboom> luke-jr: how about an alloy
 674 2012-04-04 16:38:57 <sunbird> sipa: no, nothing! it's weird. when i get home from work, i'll try downgrading to 0.5.4rc2 and also sending through command line.
 675 2012-04-04 16:39:00 <luke-jr> Blitzboom: then you lose gold's anti-counterfieting function
 676 2012-04-04 16:39:13 <Blitzboom> how about silver
 677 2012-04-04 16:39:15 <sipa> sunbird: there must be something in debug.log
 678 2012-04-04 16:39:20 <sipa> gtg now
 679 2012-04-04 16:39:23 <luke-jr> Blitzboom: I use silver as currency.
 680 2012-04-04 16:39:35 <Blitzboom> can you buy your groceries with silver?
 681 2012-04-04 16:39:45 <luke-jr> Blitzboom: can you buy your groceries with Bitcoin?
 682 2012-04-04 16:39:48 <Blitzboom> can you pay your rent with it?
 683 2012-04-04 16:39:53 <Blitzboom> i never claimed bitcoin was a currency
 684 2012-04-04 16:39:57 <luke-jr> until this week, I could pay rent with it.
 685 2012-04-04 16:40:02 <luke-jr> (silver)
 686 2012-04-04 16:40:07 <Blitzboom> ok
 687 2012-04-04 16:40:21 <luke-jr> I haven't asked new landlord tho
 688 2012-04-04 16:40:23 <sunbird> sipa: i'll check later. not near that machine now.
 689 2012-04-04 16:40:29 <luke-jr> and I'd prefer to get rid of my USD first
 690 2012-04-04 16:42:08 * luke-jr is beginning to think inflationary currency has an inherent advantage over deflationary.
 691 2012-04-04 16:43:01 <BlueMatt> you're just now beginning to think that?
 692 2012-04-04 16:43:30 <luke-jr> BlueMatt: well, the subjective value of the advantage, I mean
 693 2012-04-04 16:43:41 Sedra has joined
 694 2012-04-04 16:43:58 <luke-jr> pretty much, so long as Joe has USD, he has no reason to spend Bitcoins
 695 2012-04-04 16:44:18 <BlueMatt> yea, ofc
 696 2012-04-04 16:44:32 <luke-jr> …
 697 2012-04-04 16:44:34 * BlueMatt has always maintained bitcoin would never function as a proper currency
 698 2012-04-04 16:44:49 <BlueMatt> its great for making payments, not as a holding currency
 699 2012-04-04 16:45:04 <luke-jr> the exact opposite.
 700 2012-04-04 16:45:15 <BlueMatt> until real deflation kicks in (which requires such a large backing...)
 701 2012-04-04 16:45:24 <luke-jr> making payments = proper currency
 702 2012-04-04 16:45:37 <BlueMatt> proper currency has a lot of aspects to it
 703 2012-04-04 16:46:51 <BlueMatt> bitcoin is great for eg buying/selling digital goods
 704 2012-04-04 16:46:53 <luke-jr> my point is that in general, there is no reason to use Bitcoin to make payments if you can pay with USD instead
 705 2012-04-04 16:46:56 Sedra- has quit (Ping timeout: 245 seconds)
 706 2012-04-04 16:47:03 <BlueMatt> yea
 707 2012-04-04 16:47:22 <Blitzboom> it’s called gresham’s law
 708 2012-04-04 16:47:23 <BlueMatt> but if Im a merchant selling digital goods (and I, sanely, want to accept any customer anywhere in the world)
 709 2012-04-04 16:47:33 <BlueMatt> its 100x easier to accept bitcoin and move on
 710 2012-04-04 16:47:36 <Blitzboom> the shit money drives out the good one
 711 2012-04-04 16:47:49 vextr has joined
 712 2012-04-04 16:48:04 <luke-jr> Blitzboom: no, it's different..
 713 2012-04-04 16:48:15 <Graet> i cant pay with usd...
 714 2012-04-04 16:48:29 cdecker has quit (Quit: Leaving.)
 715 2012-04-04 16:48:35 vextr_ has quit (Ping timeout: 264 seconds)
 716 2012-04-04 16:48:43 <Blitzboom> bitcoin will just evolve to be gold 2.0
 717 2012-04-04 16:48:55 <BlueMatt> heh
 718 2012-04-04 16:48:58 <Blitzboom> as long as the legal and technical matters stay ok
 719 2012-04-04 16:49:32 <BlueMatt> bitcoin will never be as much of a safety net in anyone here's lifetime
 720 2012-04-04 16:49:38 <BlueMatt> (as gold)
 721 2012-04-04 16:49:38 TD has quit (Quit: TD)
 722 2012-04-04 16:49:51 <Blitzboom> why not
 723 2012-04-04 16:50:17 <BlueMatt> its too ingrained in people, plus consider how much the average person trusts technology in general
 724 2012-04-04 16:50:40 <Blitzboom> i claim tech will be more ingrained than the belief in gold soon
 725 2012-04-04 16:50:48 <BlueMatt> even if it has the backing, people's lack of trust in all things technical is...hard to overcome
 726 2012-04-04 16:50:57 <Blitzboom> also you probably underestimate how long we will probably live :P
 727 2012-04-04 16:51:14 <BlueMatt> ok, well for sane human lifespans
 728 2012-04-04 16:51:14 <Blitzboom> they don’t need trust
 729 2012-04-04 16:51:19 <Blitzboom> all they need is brainwashing over TV
 730 2012-04-04 16:51:20 <BlueMatt> what?
 731 2012-04-04 16:51:24 <BlueMatt> heh
 732 2012-04-04 16:51:29 <Blitzboom> some ads
 733 2012-04-04 16:51:34 <Blitzboom> like it is done for gold
 734 2012-04-04 16:51:36 <BlueMatt> you need way more than that
 735 2012-04-04 16:51:47 <Blitzboom> you actually don’t
 736 2012-04-04 16:51:55 <BlueMatt> gold has been ingrained in people as having monetary value for thousands of years
 737 2012-04-04 16:52:01 <BlueMatt> its rare, but also shiny
 738 2012-04-04 16:52:04 <BlueMatt> people like shiny
 739 2012-04-04 16:52:08 <BlueMatt> bitcoin...not so much
 740 2012-04-04 16:52:09 <Blitzboom> so? gold has problems
 741 2012-04-04 16:52:32 <Blitzboom> did you read about the tungsten gold bars?
 742 2012-04-04 16:52:36 <BlueMatt> even if you ignore people's lack of technical trust, the lack of physical object here would play a role
 743 2012-04-04 16:52:47 <BlueMatt> heh, someone making fake gold bars?
 744 2012-04-04 16:53:06 <Blitzboom> also, how much gold do common people have?
 745 2012-04-04 16:53:12 <Blitzboom> i’ve read that maybe 10% own gold at all
 746 2012-04-04 16:53:30 <BlueMatt> since when are we talking about common people?
 747 2012-04-04 16:53:30 <Blitzboom> it’ll be rich people who want BTC for its properties
 748 2012-04-04 16:53:53 rebroad has joined
 749 2012-04-04 16:53:57 <BlueMatt> rich people have very similar technological ideas as common people afaik
 750 2012-04-04 16:54:01 <Blitzboom> you did? the distrust of tech etc.
 751 2012-04-04 16:54:15 <Blitzboom> that’s changing since a few decades
 752 2012-04-04 16:54:28 <BlueMatt> rich people have the same distrust
 753 2012-04-04 16:54:41 <Blitzboom> we’ll see who is right
 754 2012-04-04 16:54:43 <BlueMatt> just because they have money doesnt mean they spend more time with tech and trust it more
 755 2012-04-04 16:54:50 <BlueMatt> yea, no way to know for years to come
 756 2012-04-04 16:54:59 <BlueMatt> anyway, getting the backing is hard enough
 757 2012-04-04 16:55:20 <BlueMatt> and is a prereq for any kind of btc becoming a safety "currency"
 758 2012-04-04 16:55:33 <Blitzboom> kim dotcom could have made good use of bitcoin instead of having his assets seized
 759 2012-04-04 16:55:40 <BlueMatt> hah
 760 2012-04-04 16:55:47 <BlueMatt> oh kim dotcom...
 761 2012-04-04 16:56:07 <gavinandresen> mmm.... mint....
 762 2012-04-04 16:56:08 Diapolo has joined
 763 2012-04-04 16:56:15 <Diapolo> good evening all
 764 2012-04-04 16:56:17 <Blitzboom> backing? gold doesn’t have backing
 765 2012-04-04 16:56:43 <BlueMatt> sorry, not backing
 766 2012-04-04 16:56:56 <BlueMatt> but large use
 767 2012-04-04 16:57:02 <BlueMatt> or whatever the word is im looking for
 768 2012-04-04 16:57:03 <Blitzboom> gold doesn’t have large use either
 769 2012-04-04 16:57:11 <BlueMatt> yes it does
 770 2012-04-04 16:57:15 <BlueMatt> the market for gold is huge
 771 2012-04-04 16:57:17 <Blitzboom> no. it sits there in vaults
 772 2012-04-04 16:57:38 <Blitzboom> the bitcoin market is also huge relative to its userbase
 773 2012-04-04 16:58:32 <BlueMatt> yea, but still doesnt have volume even remotely close to any "real" market
 774 2012-04-04 16:58:39 <BlueMatt> esp the gold market
 775 2012-04-04 16:58:45 <Blitzboom> what a surprise
 776 2012-04-04 16:58:55 <Blitzboom> i claim that this is about to change
 777 2012-04-04 16:59:11 vextr_ has joined
 778 2012-04-04 16:59:25 <BlueMatt> we have to get uptake in everyday use before we get a large market like that...
 779 2012-04-04 16:59:32 <Blitzboom> as long as BTC stays functional/legal, it’s pretty much inevitable
 780 2012-04-04 16:59:32 <BlueMatt> at least, that is the most direct way I see
 781 2012-04-04 16:59:56 <Blitzboom> also, regarding the physical
 782 2012-04-04 17:00:03 <Blitzboom> i got some casascius coins today
 783 2012-04-04 17:00:04 <BlueMatt> its not currently (really considered) legal
 784 2012-04-04 17:00:12 <Blitzboom> not considered illegal either
 785 2012-04-04 17:00:18 <Blitzboom> so legal :P
 786 2012-04-04 17:00:24 vextr has quit (Ping timeout: 246 seconds)
 787 2012-04-04 17:00:28 <BlueMatt> yea, but no company is gonna back something that isnt legal afaik
 788 2012-04-04 17:00:35 <BlueMatt> s/afaik/afa(it)k
 789 2012-04-04 17:00:37 vextr_ is now known as vextr
 790 2012-04-04 17:00:40 <BlueMatt> /
 791 2012-04-04 17:00:41 <Blitzboom> who cares about companies
 792 2012-04-04 17:00:44 <Blitzboom> we don’t need them
 793 2012-04-04 17:00:53 <BlueMatt> uh...they have the backing we need to get moving quickly
 794 2012-04-04 17:01:00 <BlueMatt> just p2p txes really doesnt generate much
 795 2012-04-04 17:01:12 <Blitzboom> it’s not about transactions but store of vale
 796 2012-04-04 17:01:15 <Blitzboom> value*
 797 2012-04-04 17:01:16 <gavinandresen> mintchip looks really interesting. Not sure relying on trusted, cannot-be-backed-up hardware will work, though.
 798 2012-04-04 17:01:50 <Blitzboom> sure, it would help if amazon accepted BTC
 799 2012-04-04 17:01:53 <Blitzboom> or steam
 800 2012-04-04 17:02:07 <freewil> how does this mintchip work - does it go through a central server
 801 2012-04-04 17:02:31 <gavinandresen> no, looks like value is stored in the secure hardware.
 802 2012-04-04 17:02:59 <gavinandresen> http://developer.mintchipchallenge.com/devguide/ecosystem.html
 803 2012-04-04 17:03:18 <freewil> so the canadian gov is going to issue as many sdcards as they want?
 804 2012-04-04 17:03:40 <freewil> how does one become a mintchip minter i wonder
 805 2012-04-04 17:03:43 <gavinandresen> I'm sure they'll just certify sd cards as "trusted and secure"
 806 2012-04-04 17:03:58 <gavinandresen> No miners, the canadian mint decides how much value there is
 807 2012-04-04 17:04:16 <Blitzboom> is it anon?
 808 2012-04-04 17:04:26 <BlueMatt> yuck, tamper resistant hardware to hold $$$, thats not gonna work
 809 2012-04-04 17:04:31 <Blitzboom> can silkroad use it?
 810 2012-04-04 17:04:52 <gavinandresen> BlueMatt: why not?  If it's holding less than $100 then losing it isn't a HUGE deal
 811 2012-04-04 17:05:06 <BlueMatt> well, ok if it only holds semi-small amounts
 812 2012-04-04 17:05:17 <luke-jr> gavinandresen: bypass the tamper-resistance crap, and add a few zeros on the end…
 813 2012-04-04 17:05:18 <BlueMatt> if its anything big though, you know someone will crack it no problem
 814 2012-04-04 17:05:20 <freewil> i said "minter" - with a T - The MintChip Minter is the trusted entity that creates the MintChip value and which it puts into circulation by selling value to the Broker.
 815 2012-04-04 17:05:30 <freewil> so how gets to be a "minTer"
 816 2012-04-04 17:05:32 <freewil> who*
 817 2012-04-04 17:05:42 <BlueMatt> freewil: probably the canadian mint, Id assume
 818 2012-04-04 17:05:55 <gavinandresen> luke-jr: their security model is "you cannot bypass the tamper-resistance crap."  Might work....
 819 2012-04-04 17:06:07 <luke-jr> gavinandresen: nothing is perfect.
 820 2012-04-04 17:06:14 <gavinandresen> ... especially if transactions are limited to amounts less than X-hundred dollars
 821 2012-04-04 17:06:39 <BlueMatt> gavinandresen: the best tamper resistant chips have been broken before, but yea, if its limited its not worth it
 822 2012-04-04 17:06:42 <gavinandresen> Yes, the canadian mint is the minter.   It looks at first glance a little like M-Pesa
 823 2012-04-04 17:06:45 <freewil> as long as it's hey, "we're going to centrally plan and arbitrarily decide how many to issue" - im not interested
 824 2012-04-04 17:06:46 Joric_ has joined
 825 2012-04-04 17:07:14 <Blitzboom> yes, 21 million is the perfect number!
 826 2012-04-04 17:07:20 <Blitzboom> hehe
 827 2012-04-04 17:07:21 <BlueMatt> I really dont see how this works on and offline, how do you prevent double-spends?
 828 2012-04-04 17:07:29 <gavinandresen> freewil: if it gives people an instant, non-refundable way of transferring money then it is very interesting
 829 2012-04-04 17:07:44 <gavinandresen> ... at the very least as a good way of buying bitcoins
 830 2012-04-04 17:08:18 <Diablo-D3> 4333aaaaaZZZZZ`````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````````` zcvbbbbvV CCCCCCCCCCCCCCX99999
 831 2012-04-04 17:08:18 <Diablo-D3> 999          [[[[[[[[[[ZX7?????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????
 832 2012-04-04 17:08:19 <Diablo-D3> ???????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????
 833 2012-04-04 17:08:23 Joric has quit (Ping timeout: 264 seconds)
 834 2012-04-04 17:08:24 <Diablo-D3> ???????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????
 835 2012-04-04 17:08:25 <BlueMatt> uhhhh...?
 836 2012-04-04 17:08:28 <gavinandresen> BlueMatt: I'm sure the online merchant will have to plug some hardware into their server.  And the customer will have to plug hardware into their computer/phone/whatever
 837 2012-04-04 17:08:29 <Diablo-D3> ????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????
 838 2012-04-04 17:08:34 <wumpus> cat?
 839 2012-04-04 17:08:35 <Blitzboom> ban this retard
 840 2012-04-04 17:08:38 Diablo-D3 has joined
 841 2012-04-04 17:09:03 <BlueMatt> gavinandresen: thus, I dont see how this is useful offline
 842 2012-04-04 17:09:14 <BlueMatt> gavinandresen: you might as well use ious at that point
 843 2012-04-04 17:09:40 <gavinandresen> BlueMatt: offline?  card plugged into your cellphone talks to the card at the unconnected merchant in the Yukon
 844 2012-04-04 17:09:47 <Diablo-D3> goddamnit
 845 2012-04-04 17:09:52 <Diablo-D3> cat was sitting on keyboard
 846 2012-04-04 17:09:55 <gavinandresen> heheh
 847 2012-04-04 17:09:56 <Graet> LOL
 848 2012-04-04 17:09:57 <Diapolo> LOL
 849 2012-04-04 17:10:01 <BlueMatt> well dont let the cat sit on the keyboard
 850 2012-04-04 17:10:05 <freewil> Blitzboom, at least with 21 million you know the rules - here it's like something is being issued with no limit on it, backed by an already fiat currency
 851 2012-04-04 17:10:05 <luke-jr> shoot the cat.
 852 2012-04-04 17:10:11 <Diapolo> eat it
 853 2012-04-04 17:10:15 <Diablo-D3> how about you go fuck yourself luke
 854 2012-04-04 17:10:22 <gavinandresen> i like cats
 855 2012-04-04 17:10:32 <BlueMatt> gavinandresen: ok, so in other words, it doesnt work offline
 856 2012-04-04 17:10:34 <gavinandresen> they're tasty (KIDDING!)
 857 2012-04-04 17:10:42 <Diapolo> ^^
 858 2012-04-04 17:10:47 <Blitzboom> freewil: yeah, i was just kidding how there are arbitrary values in bitcoins which weren’t explained
 859 2012-04-04 17:10:50 <BlueMatt> gavinandresen: defeats the purpose if you put kidding in the same line
 860 2012-04-04 17:10:52 <gavinandresen> BlueMatt: hmm?  it works offline as well as cash....
 861 2012-04-04 17:11:12 <freewil> this reminds me of IMF SDRs - basket of fiat currencies
 862 2012-04-04 17:11:25 * luke-jr wonders how cats taste.
 863 2012-04-04 17:11:33 <BlueMatt> gavinandresen: how do they prevent double spends as long as you are offline?
 864 2012-04-04 17:11:37 <gavinandresen> If I have some cash in my pocket, I can spend it.  If I have an SD card with mintychiploonies on it then I can spend them
 865 2012-04-04 17:11:53 <luke-jr> BlueMatt: the SD card has a trusted chip in it that controls the integer value
 866 2012-04-04 17:11:55 TiggrBot has quit (Ping timeout: 260 seconds)
 867 2012-04-04 17:12:00 <gavinandresen> BlueMatt: value is stored securely on the card.  They assume it is tamper-proof....
 868 2012-04-04 17:12:01 <BlueMatt> wow thats worthless
 869 2012-04-04 17:12:09 <luke-jr> lol
 870 2012-04-04 17:12:17 <BlueMatt> then people are gonna be double-spending like fucking crazy
 871 2012-04-04 17:12:29 <luke-jr> yeah, it's so dumb that you'd never expect it!
 872 2012-04-04 17:12:44 <gavinandresen> BlueMatt: you're assuming the hardware will get hacked?
 873 2012-04-04 17:12:49 <BlueMatt> gavinandresen: absolutely
 874 2012-04-04 17:13:05 <BlueMatt> I dont care how "unhackable" your hardware is, it isnt
 875 2012-04-04 17:13:06 <luke-jr> BlueMatt: double-spending assumes there's unique tokens.
 876 2012-04-04 17:13:06 <freewil> couldnt the same thing eventually happen with bitcoin though
 877 2012-04-04 17:13:19 <luke-jr> BlueMatt: this sounds like a hacker can just change the value to a higher number
 878 2012-04-04 17:13:21 <BlueMatt> freewil: if you are offline, otherwise we have a blockchain for a reason
 879 2012-04-04 17:13:27 <freewil> ...quantum computing or something i mean
 880 2012-04-04 17:13:42 <luke-jr> freewil: quantum computing can't double-spend, just steal
 881 2012-04-04 17:14:02 <freewil> well isnt that even worse
 882 2012-04-04 17:14:04 <BlueMatt> luke-jr: maybe, maybe not, even if they cant change the value to something higher than the max (because they properly signed or whatever), you can still double spend like crazy
 883 2012-04-04 17:14:05 <gavinandresen> I don't know nuthin about secure hardware.  But I suppose I have more faith that it IS possible to create very-expensive-to-hack devices.  LIke IronKey
 884 2012-04-04 17:14:28 <luke-jr> freewil: yes :P
 885 2012-04-04 17:14:28 <BlueMatt> gavinandresen: even if it is very expensive to hack, its still worth it
 886 2012-04-04 17:14:36 <BlueMatt> it gives you unlimited money if you succeed
 887 2012-04-04 17:14:58 <gavinandresen> If they do the security right a succesful hack gives you exactly one SD-card's-value-worth
 888 2012-04-04 17:15:14 <luke-jr> gavinandresen: I see no evidence the SD card is the value itself.
 889 2012-04-04 17:15:27 <BlueMatt> gavinandresen: except that you could keep resetting the sdcard to full and double spend
 890 2012-04-04 17:15:31 <freewil> i dont think these sdcards are meant to be for long-term storage though - just quick buy and spend - the security of them can be upgraded over time
 891 2012-04-04 17:15:41 <Diapolo> Can someone tell me, how "fileout << block" handles the position indicator in a file? Can I position it somehow? Talking about CAutoFile.
 892 2012-04-04 17:15:41 <luke-jr> gavinandresen: I'm reading this as "the SD card's chip protects a value number, and only signs a transaction when it's decremented the local value"
 893 2012-04-04 17:15:49 <BlueMatt> maybe they expire, but you can still double spend like crazy
 894 2012-04-04 17:16:19 <gavinandresen> Y'all are assuming that they're idiots who haven't thought of all this....
 895 2012-04-04 17:16:23 * luke-jr imagines a hacked SD card that just signs any amount without limit
 896 2012-04-04 17:16:38 <luke-jr> gavinandresen: no, I'm assuming they're trusting their chip to be unhackable. which is impossible.
 897 2012-04-04 17:16:39 <wumpus> luke-jr: I agree though that it doesn't have to be fatal though, could just give the error message and shut down the RPC thread
 898 2012-04-04 17:16:49 <Blitzboom> the government has thought of everything
 899 2012-04-04 17:16:52 <Blitzboom> ^_^
 900 2012-04-04 17:16:54 <BlueMatt> gavinandresen: oh, Im sure they have, but that doesnt mean their solution works
 901 2012-04-04 17:16:55 <luke-jr> wumpus: 0.5.0 through 0.6.0 all do just that - not fatal
 902 2012-04-04 17:17:23 <gavinandresen> I WOULD like to know what types of policies they'll have for maximum value allowed to be stored on a chip, max transaction size, etc etc
 903 2012-04-04 17:17:40 <gavinandresen> And if they have any solution for backup in case of hardware failure
 904 2012-04-04 17:17:42 <wumpus> luke-jr: yes, as I said, the shutdown was commented out because it crashed out with a segfault
 905 2012-04-04 17:17:49 <BlueMatt> they can limit the vulnerabilities, sure, but making offline digital cash isnt as easy as they make it out to be
 906 2012-04-04 17:18:09 <BlueMatt> I know Im a dumbass, but I cant think of a way to defeat double spending offline
 907 2012-04-04 17:18:18 <luke-jr> wumpus: hmm, so fatal was the *desired* behaviour, just not practical?
 908 2012-04-04 17:18:37 <gavinandresen> If you can't copy the state of the SD card then you can't double-spend.  That's the key to their security model.
 909 2012-04-04 17:18:37 <wumpus> luke-jr: yes...
 910 2012-04-04 17:18:37 <luke-jr> wumpus: btw, do you know what actually fixed the crash on the messagebox?
 911 2012-04-04 17:18:46 imsaguy2 has quit (Ping timeout: 265 seconds)
 912 2012-04-04 17:18:51 <wumpus> luke-jr: QueueShutdown()
 913 2012-04-04 17:18:51 <luke-jr> I tried bisecting it, and didn't get anywhere :/
 914 2012-04-04 17:19:12 <BlueMatt> gavinandresen: thats a very, very broken security model....
 915 2012-04-04 17:19:12 <luke-jr> wumpus: yeah, I'm talking about the fact that wxMessageBox crashed before..
 916 2012-04-04 17:19:33 <wumpus> wxmessagebox crashed? no, not that I know of
 917 2012-04-04 17:19:41 <gavinandresen> BlueMatt: it is exactly the security model of cash-- if you can't counterfeit....
 918 2012-04-04 17:19:46 <luke-jr> tons of QPixmap: It is not safe to use pixmaps outside the GUI thread
 919 2012-04-04 17:20:05 <wumpus> no pixmaps are being used outside the GUI thread
 920 2012-04-04 17:20:10 <luke-jr> gavinandresen: except cash anti-counterfiet checks are user-checkable
 921 2012-04-04 17:20:19 <BlueMatt> what luke-jr said
 922 2012-04-04 17:20:31 <wumpus> all the gui stuff is limited to one thread
 923 2012-04-04 17:20:49 <wumpus> the other threads can only queue signals
 924 2012-04-04 17:21:15 imsaguy2 has joined
 925 2012-04-04 17:21:16 <gavinandresen> okey doke.  Y'all don't like mintchip; I think it has a pretty good chance of being really important.
 926 2012-04-04 17:21:18 <luke-jr> wumpus: up until 7cfbe1fee, calling wxMessageBox from the -rpcpassword error will give those errors and crash
 927 2012-04-04 17:21:28 Joric_ is now known as Joric
 928 2012-04-04 17:21:43 <luke-jr> gavinandresen: more than don't like it, I think it will fail miserably within a few years
 929 2012-04-04 17:21:46 <BlueMatt> gavinandresen: its a cool idea, but I prefer a system that is more secure than that
 930 2012-04-04 17:21:53 <BlueMatt> (which isnt hard to do, given being on-line)
 931 2012-04-04 17:22:01 <BlueMatt> and hey, maybe their system will be really well done if its online
 932 2012-04-04 17:22:22 <wumpus> luke-jr:  wxMessageBox?! it's from a thread, you should use ThreadSafeMessageBox
 933 2012-04-04 17:22:44 <luke-jr> wumpus: wxMessageBox was renamed to ThreadSafeMessageBox in 7cfbe1fee, no?
 934 2012-04-04 17:22:50 <wumpus> luke-jr: you have no other options anymore after the last commits though, as there is only one message box function left
 935 2012-04-04 17:23:07 <wumpus> luke-jr: no, ThreadSafeMessageBox always existed
 936 2012-04-04 17:23:13 <luke-jr> O.o
 937 2012-04-04 17:23:16 <wumpus> there were two message box functions
 938 2012-04-04 17:23:21 <wumpus> the unsafe one is gone now
 939 2012-04-04 17:23:43 <luke-jr> hmm
 940 2012-04-04 17:24:03 <wumpus> wxMessageBox was only used in AppInit2 though, so it was never really unsafe
 941 2012-04-04 17:24:23 <wumpus> (as AppInit2 is currently called from the GUI thread)
 942 2012-04-04 17:24:37 <gribble> New news from bitcoinrss: jevonx opened issue 1034 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1034>
 943 2012-04-04 17:25:04 <luke-jr> wumpus: ThreadSafeMessageBox didn't actually do a Qt messagebox before, though ;p
 944 2012-04-04 17:25:21 <wumpus> I know, that's why I added wxMODAL
 945 2012-04-04 17:25:45 <luke-jr> >_<
 946 2012-04-04 17:25:56 <wumpus> for errors that are fatal or really require the users long term attention
 947 2012-04-04 17:26:04 <BlueMatt> heh, yet another request for sending double spend notifications...
 948 2012-04-04 17:26:06 <BlueMatt> wow...
 949 2012-04-04 17:26:45 <gavinandresen> while we're cleaning up code... wxAnything aught to be renamed
 950 2012-04-04 17:26:53 <jgarzik> heh
 951 2012-04-04 17:27:21 <wumpus> gavinandresen: yes
 952 2012-04-04 17:29:14 <lh77> What is mintchip?
 953 2012-04-04 17:29:22 <dvide> the best ice cream flavour
 954 2012-04-04 17:29:25 erle- has quit (Quit: erle-)
 955 2012-04-04 17:29:41 <Graet> its the evolution of currency.
 956 2012-04-04 17:29:48 <BlueMatt> heh
 957 2012-04-04 17:29:50 <Graet> http://mintchipchallenge.com/
 958 2012-04-04 17:29:52 <Graet> :P
 959 2012-04-04 17:29:58 <lh77> is it like bitcoin?
 960 2012-04-04 17:30:00 <lh77> peer to peer?
 961 2012-04-04 17:30:04 <Graet> not at all
 962 2012-04-04 17:30:12 <lh77> what does it do exactly?
 963 2012-04-04 17:30:15 <Graet> mint = canadian mint
 964 2012-04-04 17:30:20 <Graet> read link?
 965 2012-04-04 17:30:26 <Graet> i'm going bed :)
 966 2012-04-04 17:30:38 Joric_ has joined
 967 2012-04-04 17:30:38 Joric_ has quit (Changing host)
 968 2012-04-04 17:30:38 Joric_ has joined
 969 2012-04-04 17:31:42 coingenuity has quit (Read error: Connection reset by peer)
 970 2012-04-04 17:31:50 <freewil> well one thing mintchip has that bitcoin doesnt is a large organization behind it
 971 2012-04-04 17:32:04 Joric has quit (Ping timeout: 246 seconds)
 972 2012-04-04 17:32:20 <Diablo-D3> no one has even proven its from the canadian mint
 973 2012-04-04 17:32:21 <Diablo-D3> also
 974 2012-04-04 17:32:25 <Diablo-D3> its not distributed nor secure
 975 2012-04-04 17:32:40 <Diablo-D3> its DRMed virtual money using a central source to issue keys
 976 2012-04-04 17:32:42 <BlueMatt> wait, do I agree with Diablo-D3 for once?
 977 2012-04-04 17:32:56 <wumpus> that bitcoin has no large organization behind it is a good thing imo
 978 2012-04-04 17:32:56 <BlueMatt> damn im getting cynical
 979 2012-04-04 17:33:03 barmstro_ has quit (Remote host closed the connection)
 980 2012-04-04 17:33:33 <Blitzboom> no it’s not
 981 2012-04-04 17:33:39 <gavinandresen> awww.... we don't get no respect, BTC isn't one of the currency codes supported by MintChip
 982 2012-04-04 17:33:40 Joric_ is now known as Joric
 983 2012-04-04 17:33:53 <Blitzboom> it would be good to have funding for developmenta, PR and legal matters
 984 2012-04-04 17:34:04 <wumpus> bitcoin is neutral, which is a good thing, and the only way it can work
 985 2012-04-04 17:34:23 <gavinandresen> PR is lovebitcoins.org
 986 2012-04-04 17:34:26 <wumpus> if it was controlled by an organisation it'd just be another virtual object type
 987 2012-04-04 17:34:34 <wumpus> and I would have zero interest in it
 988 2012-04-04 17:34:53 <freewil> im just talking about a non-profit to fund development
 989 2012-04-04 17:35:01 <gavinandresen> I'm talking with somebody tomorrow about a non-profit for legal stuff about bitcoin
 990 2012-04-04 17:35:25 <freewil> gavinandresen, what about paying you to develop full time :)
 991 2012-04-04 17:35:42 <wumpus> that's good, as long as it doesn't become a central point of failure
 992 2012-04-04 17:35:48 <BlueMatt> gavinandresen pretty much does develop full-time
 993 2012-04-04 17:35:54 <gavinandresen> freewil: feel free.  TruCoin was paying me until their funding dried up...
 994 2012-04-04 17:35:55 <BlueMatt> (and we should pay him)
 995 2012-04-04 17:36:43 <gavinandresen> There are several workable models for funding open source projects, if I was more of a business person I'd spend more time thinking about which one makes sense for bitcoin
 996 2012-04-04 17:37:20 <freewil> the only ones i really know are donations and support-based revenue
 997 2012-04-04 17:37:31 <Blitzboom> satoshi should simply throw in a few hundred k BTC
 998 2012-04-04 17:37:49 <gavinandresen> throw in... where ?
 999 2012-04-04 17:38:00 <Graet> the "kitty"
1000 2012-04-04 17:38:01 <wumpus> in gavinandresens account
1001 2012-04-04 17:38:14 <Blitzboom> that non-profit’s bank
1002 2012-04-04 17:39:00 Diapolo has quit (Quit: Page closed)
1003 2012-04-04 17:39:03 <Blitzboom> yeah, donations should work. there are plenty of large holders who have a healthy interest in it
1004 2012-04-04 17:39:43 <BlueMatt> luckily we are a "currency"
1005 2012-04-04 17:39:52 <BlueMatt> getting eg mtgox to donate a bit wouldnt be hard
1006 2012-04-04 17:40:43 <denisx> isnt deepbit the biggest profitmaker?
1007 2012-04-04 17:40:48 <wumpus> support-based sounds interesting, maybe we should add a payment to open issues :-)
1008 2012-04-04 17:40:58 mmoya has quit (Ping timeout: 244 seconds)
1009 2012-04-04 17:41:21 <Joric> what do you think about blocking mtgox accounts due to 'tainted' coins? that bitcoin AIDS is making me nervous
1010 2012-04-04 17:41:40 <BlueMatt> Joric: welcome to the world of finance regulation
1011 2012-04-04 17:41:46 [Tycho] has joined
1012 2012-04-04 17:42:01 <wumpus> well mtgox can decide what they do for themselves right?
1013 2012-04-04 17:42:20 <Joric> it's the biggest exchange atm
1014 2012-04-04 17:42:36 <Joric> most coins end there
1015 2012-04-04 17:43:09 <BlueMatt> wumpus: sort of, though bitcoin isnt technically regulated as a currency, they still have to try to regulate it as if it were to avoid aml issues
1016 2012-04-04 17:43:41 <wumpus> BlueMatt: yes, that's the problem with it being so easy to trace
1017 2012-04-04 17:44:15 <jgarzik> bitcoin AIDS?
1018 2012-04-04 17:44:24 <wumpus> though a global ledger is probably the only way a distributed currency can work
1019 2012-04-04 17:44:37 <jgarzik> does mtgox have a list of tainted coins?
1020 2012-04-04 17:44:42 <wumpus> yes
1021 2012-04-04 17:44:51 <wumpus> they have a naughty list
1022 2012-04-04 17:45:02 barmstrong has joined
1023 2012-04-04 17:46:28 <wumpus> I think the bigger problem is not mtgox specifically but that bitcoin is so dependent on a few exchanges, they are centralized bottlenecks
1024 2012-04-04 17:46:30 <Joric> how laundered they should be to stop being 'tainted'?
1025 2012-04-04 17:47:00 <BlueMatt> Joric: thats the point of laundering
1026 2012-04-04 17:47:07 <Joric> they still will be traceable back to the source at some level
1027 2012-04-04 17:47:12 <wumpus> yes, but how much should they be diluted?
1028 2012-04-04 17:47:20 <BlueMatt> not if you put them in an mtgox account, and pull them out
1029 2012-04-04 17:47:21 <wumpus> I don't suppose they give that information :-)
1030 2012-04-04 17:47:24 <BlueMatt> they will then be different coins
1031 2012-04-04 17:48:06 <wumpus> what if someone sends you 1 tainted satoshi.. are all the coins on that address tainted for ever? 
1032 2012-04-04 17:48:23 <wumpus> it's a witchhunt
1033 2012-04-04 17:48:32 <BlueMatt> its a complicated issue
1034 2012-04-04 17:48:41 <BlueMatt> but that is the world of alm
1035 2012-04-04 17:48:46 <BlueMatt> s/alm/sml/
1036 2012-04-04 17:48:49 <BlueMatt> s/sml/aml/
1037 2012-04-04 17:50:23 <[Tycho]> All coins should be equal.
1038 2012-04-04 17:50:52 <wumpus> we need more, smaller, local exchanges
1039 2012-04-04 17:50:54 <Joric> if only MagicalTux agreed to that
1040 2012-04-04 17:51:01 TD has joined
1041 2012-04-04 17:51:20 <[Tycho]> Well, we can just start more exchanges.
1042 2012-04-04 17:51:35 <[Tycho]> With better service and support.
1043 2012-04-04 17:51:40 <wumpus> I think it's what will naturally happen though
1044 2012-04-04 17:51:53 <[Tycho]> We are already doing this.
1045 2012-04-04 17:51:58 <phantomcircuit> [Tycho], and magical pixie dust
1046 2012-04-04 17:52:12 <wumpus> mtgox can't be king for ever, otherwise it would be huge market failure
1047 2012-04-04 17:52:25 <Eliel> ideally, there would be a decentralized exchange network :)
1048 2012-04-04 17:52:31 <wumpus> yes Eliel
1049 2012-04-04 17:53:44 <BlueMatt> actually, I like mtgox's approach to aml
1050 2012-04-04 17:54:04 <BlueMatt> to get off the ground in an uncertain legal environment, bitcoin companies need to go a bit overboard on the regulation
1051 2012-04-04 17:54:19 <wumpus> yes, I can certainly understand *why* they do it
1052 2012-04-04 17:54:24 <BlueMatt> or we actually become a ml target and then we get screwed
1053 2012-04-04 17:54:54 graingert has joined
1054 2012-04-04 17:55:27 forsetifox has joined
1055 2012-04-04 17:55:33 <wumpus> regulating it to death will screw it just as much though by creating a lot of uncertainly (can I accept these coins? won't they taint myaccount? etc)
1056 2012-04-04 17:55:44 <dvide> but if bitcoin becomes non-fungible we're also screwed
1057 2012-04-04 17:55:48 <wumpus> right
1058 2012-04-04 17:56:06 <BlueMatt> wumpus: absolutely, its a tight rope to waslk
1059 2012-04-04 17:56:08 <dvide> nobody will accept bitcoin if they have no way to tell if it has any exchange value
1060 2012-04-04 17:56:39 <wumpus> well it will always have exchange value, but some coins will have more value as others as they can be more easily sold
1061 2012-04-04 17:56:45 <Joric> if only i had some of those i'd send them to every forum signature. taint the planet!
1062 2012-04-04 17:56:50 <dvide> yeah
1063 2012-04-04 17:57:07 JFK911_ is now known as JFK911
1064 2012-04-04 17:57:07 <wumpus> which is... strange, like you have copper and gold coins mixed but you cannot distinguish them
1065 2012-04-04 17:57:33 <[Tycho]> Hmm, may be tainted coins can be more valuable for this reason :)
1066 2012-04-04 17:57:46 <Eliel> wumpus: I thought mtgox is just insisting on knowing who you are when you've sent them some of the marked coins.
1067 2012-04-04 17:57:46 vigilyn has joined
1068 2012-04-04 17:57:50 <wumpus> some people will want to collect them :-)
1069 2012-04-04 17:58:11 <[Tycho]> wumpus: for donations.
1070 2012-04-04 17:58:19 Ken`_ has quit (Ping timeout: 276 seconds)
1071 2012-04-04 17:59:44 <dvide> also it's vulnerable to liars as I can just point at any transaction and claim it was stolen from me -- will it be added to the tainted list?
1072 2012-04-04 18:00:55 <BlueMatt> does mtgox not allow disputes on such frozen accounts?
1073 2012-04-04 18:01:04 <luke-jr> dvide: MtGox requires a police report to add to the list
1074 2012-04-04 18:01:19 <lh77> oh great
1075 2012-04-04 18:01:24 <lh77> so if i get a virus and my bitcoins are hacked
1076 2012-04-04 18:01:27 <lh77> i need to report it to the police?
1077 2012-04-04 18:01:28 <lh77> lol
1078 2012-04-04 18:01:32 <luke-jr> lh77: yes, please do.
1079 2012-04-04 18:01:40 <JFK911> can you imagine trying to explain to a cop what a bitcoin is and how to write this report
1080 2012-04-04 18:01:44 <lh77> and what, they come over sniffing around my pc to check ?
1081 2012-04-04 18:01:59 <JFK911> you'd be called crazy and possibly get in trouble for wasting the cops time
1082 2012-04-04 18:02:15 <dvide> how do the cops verify it, they won't know what bitcoin is
1083 2012-04-04 18:02:23 paraipan_ has joined
1084 2012-04-04 18:02:28 <luke-jr> JFK911: unlikely.
1085 2012-04-04 18:02:30 <BlueMatt> you got hacked, shit got stolen, how are you wasting the cops time?
1086 2012-04-04 18:02:31 <lh77> the average cop just doesnt have that level of technical knowledge
1087 2012-04-04 18:02:39 <luke-jr> dvide: they don't need to.
1088 2012-04-04 18:03:01 <luke-jr> dvide: someone hacked into your computer and accessed your financial accounts.
1089 2012-04-04 18:03:24 <dvide> luke-jr: you say MtGox require a police report to stop me from pointing at a random transaction and claiming theft to add it to the taint list
1090 2012-04-04 18:03:26 <luke-jr> why would they care that there's no bank?
1091 2012-04-04 18:03:38 <JFK911> i think you have a very idealistic, wishful thinking, about as reconciled with reality as your view of the catholic church, view of law enforcement
1092 2012-04-04 18:03:42 <dvide> but how does a police report help if they can't verify my theft?
1093 2012-04-04 18:03:53 <luke-jr> JFK911: so pretty accurate, eh?
1094 2012-04-04 18:04:05 <Optimo> libbitcoin yay
1095 2012-04-04 18:04:17 <luke-jr> dvide: it helps because you can provide the report to MtGox
1096 2012-04-04 18:04:18 <Optimo> does this genjix person chat here?
1097 2012-04-04 18:04:22 <luke-jr> Optimo: not often
1098 2012-04-04 18:04:27 <wumpus> I wonder if a police report from any country is good
1099 2012-04-04 18:04:34 <lh77> thats funny
1100 2012-04-04 18:04:50 <wumpus> or does it have to be filed in japan? :P
1101 2012-04-04 18:04:51 Ken` has joined
1102 2012-04-04 18:05:11 <BlueMatt> Optimo: try #bitcoinconsultancy hes there more often
1103 2012-04-04 18:05:28 paraipan has quit (Ping timeout: 276 seconds)
1104 2012-04-04 18:05:43 <luke-jr> wumpus: for Eligius, I accept police reports from any jurisdiction
1105 2012-04-04 18:05:54 gavinandresen has quit (Quit: gavinandresen)
1106 2012-04-04 18:06:03 <wumpus> how do you check if they're real?
1107 2012-04-04 18:06:40 <freewil> what is to report for Eligius?
1108 2012-04-04 18:06:46 <freewil> stolen hashing power?
1109 2012-04-04 18:07:13 <Graet> thats exactly what a botnet is....
1110 2012-04-04 18:07:16 <Joric> electricity bills
1111 2012-04-04 18:08:34 <luke-jr> wumpus: who cares
1112 2012-04-04 18:08:37 <luke-jr> freewil: yeah, botnets and such
1113 2012-04-04 18:08:55 <freewil> so what is a police report show? you're assuming guilty until proven innocent?
1114 2012-04-04 18:09:08 <luke-jr> freewil: I'm assuming reasonable evidence of criminal involvement.
1115 2012-04-04 18:09:26 <luke-jr> I'm also assuming privacy is a convenience/privilege, not an inalienable right
1116 2012-04-04 18:09:34 <upb> lol
1117 2012-04-04 18:09:46 <upb> yet you hide behind a nickname
1118 2012-04-04 18:09:53 <luke-jr> upb: what nickname?
1119 2012-04-04 18:10:01 <upb> 'luke-jr'
1120 2012-04-04 18:10:15 <luke-jr> upb: Luke Dashjr is my full name
1121 2012-04-04 18:10:22 <upb> ofcourse
1122 2012-04-04 18:10:40 <luke-jr> besides, even if I went by joebob
1123 2012-04-04 18:10:53 <luke-jr> just because privacy is a convenience doesn't mean you don't get it normally
1124 2012-04-04 18:10:54 <luke-jr> :p
1125 2012-04-04 18:11:06 <luke-jr> just that it can be dismissed if there's reason to believe a crime is being committed
1126 2012-04-04 18:11:12 <BlueMatt> the eu disagrees with you
1127 2012-04-04 18:11:17 <wumpus> having oxygen to breathe is also a convenience
1128 2012-04-04 18:11:35 <luke-jr> BlueMatt: the EU is screwed up in many ways
1129 2012-04-04 18:12:09 <luke-jr> BlueMatt: if they want to get a warrant, they can feel free to do so :P
1130 2012-04-04 18:13:55 gfinn has quit (Ping timeout: 276 seconds)
1131 2012-04-04 18:14:50 paulo_ has quit (Remote host closed the connection)
1132 2012-04-04 18:15:17 paulo_ has joined
1133 2012-04-04 18:17:41 chk has quit (Excess Flood)
1134 2012-04-04 18:17:41 Joric_ has joined
1135 2012-04-04 18:17:41 Joric_ has quit (Changing host)
1136 2012-04-04 18:17:41 Joric_ has joined
1137 2012-04-04 18:17:56 chk has joined
1138 2012-04-04 18:19:11 Joric has quit (Ping timeout: 260 seconds)
1139 2012-04-04 18:27:27 h4ckm3 has quit (Quit: changing servers)
1140 2012-04-04 18:30:28 smoothie has quit (Quit: changing servers)
1141 2012-04-04 18:31:19 smoothie has joined
1142 2012-04-04 18:31:35 smoothie has quit (Max SendQ exceeded)
1143 2012-04-04 18:33:42 Diablo-D3 has quit (Ping timeout: 240 seconds)
1144 2012-04-04 18:34:37 h4ckm3 has joined
1145 2012-04-04 18:36:56 Joric has joined
1146 2012-04-04 18:36:56 Joric has quit (Changing host)
1147 2012-04-04 18:36:56 Joric has joined
1148 2012-04-04 18:37:09 Joric_ has quit (Ping timeout: 248 seconds)
1149 2012-04-04 18:38:19 smoothie has joined
1150 2012-04-04 18:40:56 RazielZ has quit (Ping timeout: 265 seconds)
1151 2012-04-04 18:42:16 BTC_Bear is now known as hbrntng!~BTC_Bear@unaffiliated/btc-bear/x-5233302|BTC_Bear
1152 2012-04-04 18:42:20 rebroad has quit (Read error: Connection reset by peer)
1153 2012-04-04 18:45:18 gfinn has joined
1154 2012-04-04 18:49:56 user__ has joined
1155 2012-04-04 18:55:03 <gmaxwell> Is someone here a windows filesystem guru, and can you go calm down the people who are spazzing out on issu #776
1156 2012-04-04 18:55:32 <gmaxwell> Some people have noticed that their defragmentation tool shows fragmentation in block001.dat and they're very concerned about it.
1157 2012-04-04 18:56:00 <gmaxwell> I think their concern is misplaced— but I'm no windows expert, and perhaps NTFS is so broken that it's actually a problem, though I doubt that.
1158 2012-04-04 18:57:05 <luke-jr> lol
1159 2012-04-04 18:57:52 <phantomcircuit> gmaxwell, they're dumb
1160 2012-04-04 18:57:54 <phantomcircuit> next question
1161 2012-04-04 18:59:00 <gmaxwell> Be nice— they're trying, and reporting issues.. Which is better than 99.999% of the users. :)
1162 2012-04-04 18:59:08 <phantomcircuit> true
1163 2012-04-04 18:59:20 rebroad has joined
1164 2012-04-04 18:59:28 <phantomcircuit> s/dumb/slightly above average/
1165 2012-04-04 19:00:10 <gmaxwell> I suspect that normal and expected high disk load of synchronization is upsetting them.  If we were really smart we'd have made the libdb settings commit say "fix fragementation issues" :)
1166 2012-04-04 19:00:21 <phantomcircuit> lol
1167 2012-04-04 19:00:30 <wumpus> maybe their disk is almost full?
1168 2012-04-04 19:00:59 <phantomcircuit> gmaxwell, really everything but the wallet should be considered volatile
1169 2012-04-04 19:01:30 <phantomcircuit> but when you use a WAL the disk sync operations become much less costly so it probably doesn't matter
1170 2012-04-04 19:07:44 user__ has quit (Quit: Leaving)
1171 2012-04-04 19:10:03 <Joric> looks like fragmentation may be disabled in DB_CONFIG - eg. 'set_cachesize 0 5242880 1  # sets a database cache of 5M and allows fragmentation'
1172 2012-04-04 19:10:46 <gmaxwell> Joric: that doesn't make sense.
1173 2012-04-04 19:11:03 <gmaxwell> I assume whatever you're reading is talking about memory allocation.
1174 2012-04-04 19:11:15 <gmaxwell> FS layout is not visible to applications.
1175 2012-04-04 19:11:35 <Graet> hmm on win xp, closed wallet, ran defrag analisis. says 13,768 fragments for 1.08Gb blk0001.dat. and thats 10x more fragments than next listing
1176 2012-04-04 19:12:11 <Graet> and thats my netbook thats pretty much a wallet and not much else
1177 2012-04-04 19:12:16 <gmaxwell> Graet: sure. But that isn't problematic by itself— sounds like the blockchain is probably ending up interleaved on disk with the index file.
1178 2012-04-04 19:12:34 <Graet> yes, propably lack of disk space
1179 2012-04-04 19:12:41 <gmaxwell> None of this is accessed sequentially. So unless there is something really weird about NTFS, it shouldn't matter.
1180 2012-04-04 19:12:49 <Graet> cool
1181 2012-04-04 19:12:54 * Graet goes to bed then :)
1182 2012-04-04 19:13:04 <wumpus> unless you make and delete a lot of small files, until the disk is almost full, application behaviour really shouldn't affect disk fragmentation much
1183 2012-04-04 19:13:07 <luke-jr> #winapi are telling me to make the program write 0s for the reserved space in advance, and seek-overwrite
1184 2012-04-04 19:13:13 <luke-jr> IMO, this is braindead
1185 2012-04-04 19:13:55 <gmaxwell> I gues it doesn't do block-group allocation like unix filesystems do.
1186 2012-04-04 19:13:56 <jgarzik> Most filesystem suck at slowly-growing files
1187 2012-04-04 19:14:00 <jgarzik> on unix, it's mbox
1188 2012-04-04 19:14:09 <gmaxwell> s/gues/guess/
1189 2012-04-04 19:14:22 <jgarzik> NTFS uses extents, just like ext4 and btrfs
1190 2012-04-04 19:14:33 <jgarzik> and allocation zones, to try and keep files together
1191 2012-04-04 19:14:55 <gmaxwell> In any case, I still don't see the file. The blockchain _is not read sequentially_ — the performance accessing it sucks, yes, but it sucks because the access is random. :)
1192 2012-04-04 19:15:00 <luke-jr> jgarzik: you're saying Windows users *shouldn't* be seeing 5000 fragments?
1193 2012-04-04 19:15:18 <jgarzik> slow-growing files, where requests for new blocks are interspersed within the request for new blk0001.dat space, are the worst for a filesystem to deal with
1194 2012-04-04 19:15:57 <jgarzik> a fragment for every new block or 3 would not be surprising at all
1195 2012-04-04 19:16:00 Cablesaurus has joined
1196 2012-04-04 19:16:00 Cablesaurus has quit (Changing host)
1197 2012-04-04 19:16:00 Cablesaurus has joined
1198 2012-04-04 19:16:06 <jgarzik> even in Linux, honestly
1199 2012-04-04 19:16:18 <luke-jr> ._.
1200 2012-04-04 19:16:25 <wumpus> but is it a problem?
1201 2012-04-04 19:16:45 <jgarzik> highly fragmented files take additional CPU and disk seeks to get data
1202 2012-04-04 19:16:55 <luke-jr> I have my doubts. At least one of the guys is ranting about how it's destroying his hardware, which is obviously bs.
1203 2012-04-04 19:17:12 <wumpus> not only if the file is accessed sequentially?
1204 2012-04-04 19:17:16 <luke-jr> jgarzik: even for highly fragmented reads?
1205 2012-04-04 19:17:28 <jgarzik> if you have low RAM, so that fs metadata is poorly cached, your disk takes a real pounding
1206 2012-04-04 19:18:00 <wumpus> database files are always slow growing files, it'd be strange if it was a big issue
1207 2012-04-04 19:18:19 <luke-jr> and logs too
1208 2012-04-04 19:18:31 <jgarzik> wumpus: well our database files are not accessed sequentially.  the tail is accessed randomly, with older blocks touched rarely
1209 2012-04-04 19:18:47 <jgarzik> index access is all over the place, and our indices are huge
1210 2012-04-04 19:19:09 <jgarzik> logs are not really a problem.  bdb preallocates logs and does other tricks
1211 2012-04-04 19:19:22 <jgarzik> one solution is fallocate()
1212 2012-04-04 19:19:37 <jgarzik> or the dumb equivalent thereof (write several MB of zeroes, then seek back)
1213 2012-04-04 19:19:59 <jgarzik> prezeroing large zones forces sequential block allocation
1214 2012-04-04 19:20:06 <jgarzik> avoiding the slow-growing-file problem
1215 2012-04-04 19:20:28 <jgarzik> anyway, time for Mexican food
1216 2012-04-04 19:20:32 <luke-jr> (I meant normal syslog logs :p)
1217 2012-04-04 19:21:34 <wumpus> normal syslog logs are not optimized for performance
1218 2012-04-04 19:22:24 <wumpus> and log rotation with compression probably makes it somewhat better, as the gz is written in one go
1219 2012-04-04 19:22:44 <jgarzik> yep
1220 2012-04-04 19:23:22 <jgarzik> makes it a _lot_ better, when you rewrite all in one go.  1000x better or more, depending on log file size :)
1221 2012-04-04 19:23:24 <luke-jr> any idea if fallocate's effect goes away when the fd closes?
1222 2012-04-04 19:23:43 <jgarzik> luke-jr: it does not go away when you close the file.
1223 2012-04-04 19:24:17 <luke-jr> jgarzik: what if the disk is otherwise full? :|
1224 2012-04-04 19:25:03 <midnightmagic_> preallocation is an *enormous* win on Windows, btw.
1225 2012-04-04 19:25:11 midnightmagic_ is now known as midnightmagic
1226 2012-04-04 19:25:36 <midnightmagic> less so on Linux.
1227 2012-04-04 19:26:17 p0s has joined
1228 2012-04-04 19:27:04 amiller is now known as amiller`1240
1229 2012-04-04 19:28:07 <wumpus> the question that remains is, how to get bdb to do that?
1230 2012-04-04 19:28:14 bitvampire has joined
1231 2012-04-04 19:28:32 <luke-jr> well, fallocate doesn't exist on Windows afaict, and blk0001 isn't bdb
1232 2012-04-04 19:29:24 Joric_ has joined
1233 2012-04-04 19:29:37 <wumpus> also not a similar call?
1234 2012-04-04 19:29:51 <luke-jr> not that I could find
1235 2012-04-04 19:30:38 Joric has quit (Ping timeout: 276 seconds)
1236 2012-04-04 19:30:41 <wumpus> of course, we could simply pad it manually with zeros up until the next N MB every time...
1237 2012-04-04 19:31:42 <wumpus> http://stackoverflow.com/questions/7970333/how-do-you-pre-allocate-space-for-a-file-in-c-c-on-windows
1238 2012-04-04 19:32:15 <wumpus> set the file pointer then set end of file, makes sense
1239 2012-04-04 19:33:08 <luke-jr> won't that make a sparse file? O.o
1240 2012-04-04 19:34:05 <wumpus> and there's also _chsize, it seems
1241 2012-04-04 19:34:14 <gmaxwell> Yea, I was about to link to: http://msdn.microsoft.com/en-us/library/dk925tyb%28v=vs.80%29.aspx
1242 2012-04-04 19:34:17 <wumpus> does windows even have sparse files?
1243 2012-04-04 19:35:09 <wumpus> only thing is that it affects the file size and makes actual zeros, so naive appending will no longer work (no idea if we use that)
1244 2012-04-04 19:35:16 gavinandresen has joined
1245 2012-04-04 19:35:37 <phantomcircuit> wumpus, that wont preallocate anything
1246 2012-04-04 19:35:48 <luke-jr> I suppose blkindex could always track "next block"…
1247 2012-04-04 19:35:49 <phantomcircuit> the only way to preallocate space is to write to it
1248 2012-04-04 19:36:19 datagutt is now known as datagutt|affk
1249 2012-04-04 19:36:20 <wumpus> how are you so sure phantomcircuit?
1250 2012-04-04 19:36:21 datagutt is now known as affk!~datagutt@unaffiliated/datagutt|datagutt|afk
1251 2012-04-04 19:36:25 datagutt is now known as afk!~datagutt@unaffiliated/datagutt|datagutt
1252 2012-04-04 19:37:52 <wumpus> "Most applications are not aware of sparse files and will not create sparse files. "
1253 2012-04-04 19:38:11 <wumpus> seems that in windows, you have to specifically request it http://msdn.microsoft.com/en-us/library/windows/desktop/aa365566(v=vs.85).aspx
1254 2012-04-04 19:38:37 <wumpus> so as I understand the _chsize would work on windows, fallocate on linu
1255 2012-04-04 19:39:15 bitvampire has quit (Remote host closed the connection)
1256 2012-04-04 19:40:18 <phantomcircuit> wumpus, i guess it doesn't
1257 2012-04-04 19:40:22 <phantomcircuit> color me wrong
1258 2012-04-04 19:41:11 <luke-jr> wumpus: probably want something async
1259 2012-04-04 19:43:30 <wumpus> yes it could be a background task I guess
1260 2012-04-04 19:44:24 <wumpus> depends on how fast the call is
1261 2012-04-04 19:44:38 Sedra- has joined
1262 2012-04-04 19:47:42 Sedra has quit (Ping timeout: 240 seconds)
1263 2012-04-04 19:52:32 <gribble> New news from bitcoinrss: gavinandresen opened pull request 1035 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1035>
1264 2012-04-04 19:54:28 <luke-jr> wumpus: src/qt/bitcoingui.cpp has a typo 20011 in the header fwiw
1265 2012-04-04 19:54:41 <wumpus> even twice I think
1266 2012-04-04 19:57:51 <luke-jr> phew, just skimmed all 88k lines of diff between 0.5.x and 0.6.0
1267 2012-04-04 19:58:02 <luke-jr> only thing obvious missing from backports was wumpus's PGP key :P
1268 2012-04-04 20:00:22 <luke-jr> Any opinions on whether this is a bugfix for 0.5.x, or not? 2675fe6 Increase time ago of last block for "up to date" status from 30 to 90 minutes
1269 2012-04-04 20:00:50 <wumpus> it's a bugfix, 30 minutes causes too many false positives
1270 2012-04-04 20:01:10 barmstrong has quit (Remote host closed the connection)
1271 2012-04-04 20:01:18 <luke-jr> k ty
1272 2012-04-04 20:01:48 barmstrong has joined
1273 2012-04-04 20:01:59 <jgarzik> back
1274 2012-04-04 20:02:08 <luke-jr> wb
1275 2012-04-04 20:02:25 <jgarzik> wumpus: the most portable and most effective and most reliable method is writing data (zeroes or whatever) to a file
1276 2012-04-04 20:02:40 <gribble> New news from bitcoinrss: gavinandresen opened pull request 1036 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1036>
1277 2012-04-04 20:02:47 <jgarzik> wumpus: system calls _do_ exist, but you want to avoid sparse files.  we want to actually allocate the data on disk.
1278 2012-04-04 20:02:57 <jgarzik> fallocate() will do that.... if there is FS support
1279 2012-04-04 20:03:07 <jgarzik> otherwise, it either fails or writes zeroes, depending on your libc
1280 2012-04-04 20:03:32 <jgarzik> Windows has a behavior like fallocate(), but there too are caveats
1281 2012-04-04 20:03:38 <wumpus> the only drawback of writing zeroes is that it is the slowest option
1282 2012-04-04 20:03:41 <jgarzik> easier to simply write 4MB at a time
1283 2012-04-04 20:03:48 <jgarzik> wumpus: not if you do it well in advance
1284 2012-04-04 20:03:59 <luke-jr> gavinandresen: aww, we don't want a handy multicast system in Bitcoin? :P
1285 2012-04-04 20:04:13 <wumpus> luke-jr: it can always be brought back, that's what git is for
1286 2012-04-04 20:04:31 <luke-jr> wumpus: heh, I was joking
1287 2012-04-04 20:04:34 <wumpus> hehe oke
1288 2012-04-04 20:04:39 <wumpus> you never know with you :)
1289 2012-04-04 20:04:40 <luke-jr> inv is basically a more efficient multicast anyway
1290 2012-04-04 20:05:38 <jgarzik> I wouldn't necessarily write zeroes, either.  Whatever is guaranteed to fail during de-serialize, such as 0xff's
1291 2012-04-04 20:06:13 <jgarzik> because with manual prealloc, blk0001.dat will be 2000 MB, but only contain 1999 MB of data.
1292 2012-04-04 20:06:25 <jgarzik> must be able to notice where the real data stops, upon program init
1293 2012-04-04 20:08:28 <wumpus> yes, that's what I meant with naive append no longer working
1294 2012-04-04 20:09:16 <gmaxwell> This grows uglier with every message.
1295 2012-04-04 20:09:23 * jgarzik already has a process on his bitcoind servers, which (a) stop bitcoind, (b) "cp -a data data.defrag ; rm -rf data ; mv data.defrag data" and (c) restart bitcoind
1296 2012-04-04 20:09:40 <jgarzik> solves defrag issues, which _are_ noticeable even on Linux
1297 2012-04-04 20:09:53 barmstrong has quit (Remote host closed the connection)
1298 2012-04-04 20:09:54 <jgarzik> I run that bitcoind defrag once per month
1299 2012-04-04 20:10:25 <gmaxwell> How is it noticable?
1300 2012-04-04 20:10:35 <luke-jr> I thought Linux automatically defragmentted in the background?
1301 2012-04-04 20:10:38 <gmaxwell> No.
1302 2012-04-04 20:10:39 <gmaxwell> ha
1303 2012-04-04 20:10:44 barmstrong has joined
1304 2012-04-04 20:10:44 <jgarzik> luke-jr: nope
1305 2012-04-04 20:10:57 <jgarzik> gmaxwell: startup time reduced by 5-6 seconds
1306 2012-04-04 20:11:00 <denisx> OSX does that, but only files up to 20MB
1307 2012-04-04 20:11:02 <jgarzik> last I measured
1308 2012-04-04 20:11:10 <forsetifox> Hm.. I thought linux just "knew" the best way to put files so they weren't fragmented.
1309 2012-04-04 20:11:23 <gmaxwell> forsetifox: the filesystems avoid it, but it's not magic.
1310 2012-04-04 20:11:38 <jgarzik> forsetifox: slow growing files are the worst, for any filesystem, Linux or Windows
1311 2012-04-04 20:11:43 <gavinandresen> I bet putting the database logs, database .dat files, debug.log, and blk0000*.dat on separate hard drives would be nice and fast and un-fragmented....
1312 2012-04-04 20:11:56 <jgarzik> gavinandresen: nope
1313 2012-04-04 20:12:00 <gavinandresen> nope?
1314 2012-04-04 20:12:21 <jgarzik> gavinandresen: well, if you mean blk00001.dat and nothing else on the hard drive at all... yes that would solve it :)
1315 2012-04-04 20:12:22 <gmaxwell> jgarzik: the only reason startup does anything with blk0000*.dat is just the checkblocks, which is currently nearly pointless.
1316 2012-04-04 20:12:29 <gavinandresen> jgarzik: yes, that's what I meant
1317 2012-04-04 20:12:40 <jgarzik> :)
1318 2012-04-04 20:12:49 <gribble> New news from bitcoinrss: gavinandresen opened pull request 1037 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1037>
1319 2012-04-04 20:12:54 <forsetifox> Tell those speed freaks to un-install their defragmenting programs so they don't see it or get SSD's. =P
1320 2012-04-04 20:12:58 <gavinandresen> I think it is always worth considering the "throw lots of hardware at the problem" solution
1321 2012-04-04 20:12:58 <gmaxwell> If you want to save time starting ... forget defragmenting the file, just make the statup check check the last dozen blocks instead of a few thousand.
1322 2012-04-04 20:13:56 <gmaxwell> (really— that check should be backgrounded, should check the whole chain, and do enough checking (hashes) to detect corruption)
1323 2012-04-04 20:14:03 <jgarzik> gavinandresen: hard drive per file is a lot of hardware, indeed ;-)
1324 2012-04-04 20:14:25 <jgarzik> gmaxwell: really a lot of startup could be improved
1325 2012-04-04 20:14:26 <gavinandresen> I was just suggesting 4 hard drives
1326 2012-04-04 20:14:27 <luke-jr> gmaxwell++
1327 2012-04-04 20:15:00 <gmaxwell> ln -s /dev/sda3 block00001.dat
1328 2012-04-04 20:15:12 <gavinandresen> (really, database .dat and log files aught to be on separate spindles for performance, and if you're going to do 2 you might as well do 4....)
1329 2012-04-04 20:15:28 <luke-jr> gmaxwell: not sure that works
1330 2012-04-04 20:15:28 <jgarzik> gavinandresen: we don't hit the database that hard
1331 2012-04-04 20:15:34 <gmaxwell> luke-jr: it wouldn't I was being silly.
1332 2012-04-04 20:15:35 <gavinandresen> not yet
1333 2012-04-04 20:16:01 <gavinandresen> although by the time we do hit the database that hard everybody might be running with a couple of terabytes of memory
1334 2012-04-04 20:16:45 <gmaxwell> It would be fun to try with an isolated testnet node to see how far it scales— unfortunately stupid bugs break things first. e.g. coin selection gets really slow really fast if you ./bitcoind sendtoaddress `./bitcoind getnewaddress` 100.
1335 2012-04-04 20:17:03 Joric has joined
1336 2012-04-04 20:17:03 Joric has quit (Changing host)
1337 2012-04-04 20:17:03 Joric has joined
1338 2012-04-04 20:17:05 <jgarzik> honestly our db needs are quite modest... a few large indices
1339 2012-04-04 20:17:45 <gmaxwell> Indeed. Too bad its already a performance problem in spite of how modest it is...
1340 2012-04-04 20:17:52 <gmaxwell> though much less of one now.
1341 2012-04-04 20:17:54 <jgarzik> and some stuff and really doesn't mesh well with (a) key/value, like wallet data or (b) being in the same bdb environment
1342 2012-04-04 20:18:56 <jgarzik> wallet data is better stored in some structured format than forced into a flat hierarchy by k/v
1343 2012-04-04 20:19:06 Joric_ has quit (Ping timeout: 250 seconds)
1344 2012-04-04 20:19:28 <jgarzik> address data could be a flat file, periodically dumped to disk, since we only read it at startup
1345 2012-04-04 20:19:37 <jgarzik> of course, need to start pruning address data intelligently
1346 2012-04-04 20:19:45 <gmaxwell> ...
1347 2012-04-04 20:19:56 <wumpus> key data would be better stored as ascii append only files 
1348 2012-04-04 20:20:02 * gmaxwell invites jgarzik to join the present. :)
1349 2012-04-04 20:20:10 <jgarzik> gmaxwell: we are pruning now?
1350 2012-04-04 20:20:12 <gmaxwell> Yep.
1351 2012-04-04 20:20:15 <jgarzik> woo!!
1352 2012-04-04 20:21:18 <gmaxwell> (Litecoin was attacked by some fruitcakes running them out of disk by flooding addresses, turning them into guinea pigs^w^wtest subjects for sipa's addrman code)
1353 2012-04-04 20:21:18 <jgarzik> gmaxwell: got a code pointer or grep string?
1354 2012-04-04 20:21:45 <midnightmagic> ;;seen lfm
1355 2012-04-04 20:21:45 <gribble> lfm was last seen in #bitcoin-dev 9 weeks, 4 days, 0 hours, 2 minutes, and 9 seconds ago: <lfm> adictive ms lock in
1356 2012-04-04 20:21:46 <gmaxwell> jgarzik: commit a6b4a11385bf44e695c3e47cbd0de6e40eea0b23 merged it
1357 2012-04-04 20:21:55 <midnightmagic> 9 weeks. shit.
1358 2012-04-04 20:22:26 <gmaxwell> jgarzik: in particular, read the comments in addrman.h
1359 2012-04-04 20:23:38 <jgarzik> well shit, let's drop addr.dat then :)
1360 2012-04-04 20:24:10 <gmaxwell> right now addr.dat is getting the addrman data as a blob basically.
1361 2012-04-04 20:27:11 Snapman[afkers] is now known as Snapman
1362 2012-04-04 20:27:46 barmstrong has quit (Remote host closed the connection)
1363 2012-04-04 20:28:11 barmstrong has joined
1364 2012-04-04 20:29:09 <jgarzik> yep
1365 2012-04-04 20:31:20 <rebroad> I have a question... at 19:54 UTC my debug.log notices an invalid block at height 174313, so saying best = 173412... between then and 20:27UTC it connects to 8 nodes, all of which report that 174313 is the highest block so far, and at 20:27 recognises a new 174313 block which is valid.... is this correct behaviour?
1366 2012-04-04 20:31:56 Joric_ has joined
1367 2012-04-04 20:31:56 Joric_ has quit (Changing host)
1368 2012-04-04 20:31:56 Joric_ has joined
1369 2012-04-04 20:32:23 tower has quit (Ping timeout: 264 seconds)
1370 2012-04-04 20:32:33 localhost has quit (Remote host closed the connection)
1371 2012-04-04 20:32:46 <luke-jr> don't suppose anyone knows the nature of 3f1bb1a Proper support for Growl 1.3 notifications ? ie, does the old code not work with both?
1372 2012-04-04 20:33:24 <wumpus> I have no idea, the mac stuff scares me shitless :-)
1373 2012-04-04 20:33:43 Joric has quit (Ping timeout: 245 seconds)
1374 2012-04-04 20:35:20 denisx_ has joined
1375 2012-04-04 20:35:22 <luke-jr> XD
1376 2012-04-04 20:36:13 localhost has joined
1377 2012-04-04 20:39:11 denisx has quit (Ping timeout: 260 seconds)
1378 2012-04-04 20:39:11 denisx_ is now known as denisx
1379 2012-04-04 20:45:50 <gavinandresen> luke-jr: the old code didn't work with the new Growl, if I remember correctly.  I don't run Growl, though.  p2k is the person to ask.
1380 2012-04-04 20:46:46 <BlueMatt> there was an issue with specifics iirc
1381 2012-04-04 20:52:18 osmosis has joined
1382 2012-04-04 20:53:15 osmosis has left ()
1383 2012-04-04 20:53:16 _wizkid057 is now known as wizkid057
1384 2012-04-04 20:53:16 wizkid057 has quit (Changing host)
1385 2012-04-04 20:53:16 wizkid057 has joined
1386 2012-04-04 20:55:39 <jgarzik> nice
1387 2012-04-04 20:55:47 <jgarzik> with the new addrman stuff, using a flat file should be trivial
1388 2012-04-04 20:57:31 <sipa> you need a way for preventing failed partial writes
1389 2012-04-04 20:57:41 <sipa> that's not hard, and far easier than using bdb
1390 2012-04-04 20:57:48 <sipa> but it's still needed
1391 2012-04-04 21:00:40 denisx has quit (Quit: denisx)
1392 2012-04-04 21:06:17 datagutt has quit (Quit: Computer has gone to sleep.)
1393 2012-04-04 21:13:42 Slix` has joined
1394 2012-04-04 21:13:51 Turingi has quit (Read error: Connection reset by peer)
1395 2012-04-04 21:14:07 Slix`_ has joined
1396 2012-04-04 21:16:25 Nicksasa has quit (Read error: No route to host)
1397 2012-04-04 21:18:46 Slix`_ has quit (Read error: Connection reset by peer)
1398 2012-04-04 21:18:47 <Cory> Why can't I control v an address into the "
1399 2012-04-04 21:18:52 <Cory> pat to" field? :(
1400 2012-04-04 21:19:59 <sipa> what is a "pat to" field?
1401 2012-04-04 21:20:11 <sipa> ah, pay to
1402 2012-04-04 21:21:34 <Cory> Yeah. I don't know what's up with my typing on this keyboard. >:[
1403 2012-04-04 21:22:23 <Cory> 0.6 Windows 7
1404 2012-04-04 21:22:40 <sipa> hmm, pasting works perfectly here (ubuntu)
1405 2012-04-04 21:23:09 <Cory> I can paste into the label and amount field.
1406 2012-04-04 21:23:14 topace has quit (Changing host)
1407 2012-04-04 21:23:14 topace has joined
1408 2012-04-04 21:23:26 <sipa> wumpus: ^
1409 2012-04-04 21:24:11 Clipse has quit (Quit: Clipse)
1410 2012-04-04 21:25:11 <Cory> A restart of the client fixed it. :P
1411 2012-04-04 21:25:29 * sipa blames windows
1412 2012-04-04 21:25:29 Joric_ has quit (Read error: Connection reset by peer)
1413 2012-04-04 21:26:30 Clipse has joined
1414 2012-04-04 21:40:08 ThomasV has quit (Quit: Quitte)
1415 2012-04-04 21:41:02 erle- has joined
1416 2012-04-04 21:42:16 tower has joined
1417 2012-04-04 21:48:45 <jgarzik> sipa: sure, just write and flush a temp file
1418 2012-04-04 21:49:09 agricocb has quit (Remote host closed the connection)
1419 2012-04-04 21:49:46 agricocb has joined
1420 2012-04-04 21:53:04 Cablesaurus has quit (Quit: Hard work pays off in the future, laziness pays off now)
1421 2012-04-04 21:56:02 ashish has quit (Quit: Aah IRC, where men are men, women are men, and 14 year old girls are FBI agents.)
1422 2012-04-04 21:56:05 phma has joined
1423 2012-04-04 22:00:37 barmstrong has quit (Remote host closed the connection)
1424 2012-04-04 22:02:07 agricocb has quit (Quit: Leaving.)
1425 2012-04-04 22:02:20 <etotheipi_> gavinandresen, I finally got around to organizing my escrow thoughts into a gist ... there's a lot of different ways to do this, but I think this works ... I'm sure we can figure out a way to do it with less steps, too
1426 2012-04-04 22:02:21 <etotheipi_> https://gist.github.com/2305966
1427 2012-04-04 22:03:19 <etotheipi_> gavinandresen, although I think we need a wider discussion about the "risk deposit" idea, up front... you're right it's weird and different -- I wonder if there's a different paradigm that doesn't induce too much risk or complexity for either party
1428 2012-04-04 22:03:25 abbe has joined
1429 2012-04-04 22:03:46 <sipa> jgarzik: write to addr.dat.new, delete addr.dat, rename?
1430 2012-04-04 22:04:47 <sipa> or write to addr.dat.bak, overwrite addr.dat (with a marker byte saying "incomplete"), mark addr.dat complete, delete addr.dat.bak
1431 2012-04-04 22:05:04 <sipa> and when loading, if addr.dat doesn't exist or is incomplete, try addr.dat.bak instead
1432 2012-04-04 22:06:02 <sipa> gmaxwell: i got brokenwallet's wallet
1433 2012-04-04 22:06:07 <sipa> seems your theory could be correct
1434 2012-04-04 22:06:40 <sipa> for a small percentage of incorrect passphrases, the decrypted secret seems to be not 32 bytes
1435 2012-04-04 22:07:02 <gmaxwell> It was all I could come up with that satisfied all of the facts and required the least in addition.
1436 2012-04-04 22:07:08 <gmaxwell> Ah. tada!
1437 2012-04-04 22:07:31 <sipa> which makes sense, because there is (unintentionally) a padding system in effect
1438 2012-04-04 22:07:45 barmstrong has joined
1439 2012-04-04 22:08:05 <sipa> though i would expect it to be not 32 bytes in the majority of cases, instead of rarely
1440 2012-04-04 22:10:05 <sipa> i mean, the encrypted secret is 48 bytes
1441 2012-04-04 22:10:22 <etotheipi_> ahh yes... htat stupid 48-byte secret...
1442 2012-04-04 22:10:29 <sipa> and the padding scheme is supposed to be N types byte N at the end
1443 2012-04-04 22:10:39 <sipa> in the majoriy of cases, this padding scheme would fail
1444 2012-04-04 22:10:40 barmstrong has quit (Remote host closed the connection)
1445 2012-04-04 22:10:54 minimoose has quit (Quit: minimoose)
1446 2012-04-04 22:10:57 <sipa> a decent crypto implementation would complain loudly about this
1447 2012-04-04 22:11:41 <sipa> so apparently its default case is assuming 16 bytes padding (the maximum) if what it sees makes no sense?
1448 2012-04-04 22:11:48 barmstrong has joined
1449 2012-04-04 22:14:00 <etotheipi_> sipa, I never understood the justification to begin with:  why does a 32-byte secret need an extra 16 bytes of padding?  to me it just looks like an erroneous equation
1450 2012-04-04 22:14:33 <sipa> etotheipi_: it's part of the CBC specification
1451 2012-04-04 22:14:36 <etotheipi_> i.e. the equation is incorrect for data sizes that are exact multiples of the blocksize
1452 2012-04-04 22:14:57 <sipa> we should have used unpadded encryption because we knew the data would always be a multiple of the blocksize
1453 2012-04-04 22:15:24 <sipa> but because it's apparently implemented poorly in EVP, we never noticed this
1454 2012-04-04 22:15:51 <gribble> New news from bitcoinrss: Diapolo opened pull request 1038 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1038>
1455 2012-04-04 22:16:20 <sipa> you need to pad that is accidentally a multiple of the blocksize, because there is no place to store the length information otherwise
1456 2012-04-04 22:16:30 <sipa> s/that/data/
1457 2012-04-04 22:16:36 barmstrong has quit (Ping timeout: 260 seconds)
1458 2012-04-04 22:16:52 <sipa> hmm, trying 2^16 walletpassphrase rpc calls takes long
1459 2012-04-04 22:17:46 Joric has joined
1460 2012-04-04 22:19:00 <etotheipi_> sipa, gmaxwell:  can you look at the escrow proposal (https://gist.github.com/2305966) and tear it apart for me?  :)
1461 2012-04-04 22:19:34 <etotheipi_> I wanted to get the ball rolling on, at least, the paradigm
1462 2012-04-04 22:20:39 paulo_ is now known as paulo_|away
1463 2012-04-04 22:21:19 <etotheipi_> gavin didn't like the "risk deposit" concept, but I don't see a way to do it without:  I believe both parties need BTC held up in the tx that they get back when it's done, so both have an incentive to complete it
1464 2012-04-04 22:21:53 barmstrong has joined
1465 2012-04-04 22:21:56 barmstrong has quit (Remote host closed the connection)
1466 2012-04-04 22:23:24 agricocb has joined
1467 2012-04-04 22:23:56 <sipa> it could always be negotiated to be zero
1468 2012-04-04 22:24:39 <sipa> i like the general idea, but i've little opinion about the economic aspects
1469 2012-04-04 22:25:06 <etotheipi_> Malicious seller could offer to sell someone something for 100 BTC -- wait for the 2-of-2 or 2-of-3 tx to go through, then disappear, screwing over the buyer
1470 2012-04-04 22:25:57 <etotheipi_> sipa, I have a theory about how this could work, economically (in a win-win way)
1471 2012-04-04 22:26:16 <etotheipi_> as long as buyer and seller are registered with the third party
1472 2012-04-04 22:26:25 <etotheipi_> (it would be tough without registration)
1473 2012-04-04 22:26:41 <etotheipi_> both parties can retrieve and verify keys from the third-party, as many as they want for free
1474 2012-04-04 22:26:53 <etotheipi_> the third-party doesn't even have to be notified that they are part of the tx
1475 2012-04-04 22:26:53 erle- has quit (Quit: erle-)
1476 2012-04-04 22:27:22 <etotheipi_> it's only when a dispute happens, that one party contacts them with the P2SH script, and the third-party helps as long as there is 15%+ risk deposit for them to take as fee
1477 2012-04-04 22:27:42 <etotheipi_> (or whatever guidelines are set out from the start)
1478 2012-04-04 22:27:52 <etotheipi_> that way using the third-party is absolutely "free" if there is no dispute
1479 2012-04-04 22:28:27 <etotheipi_> and third-party only deals with tx already in dispute: they don't need to deal with registration of each transaction
1480 2012-04-04 22:29:40 <etotheipi_> ....maybe I should put that explanation in the gist
1481 2012-04-04 22:34:16 Z0rZ0rZ0r1 has joined
1482 2012-04-04 22:34:16 Z0rZ0rZ0r has quit (Disconnected by services)
1483 2012-04-04 22:34:43 eryngi has joined
1484 2012-04-04 22:34:49 <eryngi> hello
1485 2012-04-04 22:34:52 <eryngi> may I speak here?
1486 2012-04-04 22:35:02 <sipa> yes
1487 2012-04-04 22:35:04 <etotheipi_> eryngi, you are allowed 150 characters
1488 2012-04-04 22:35:19 <etotheipi_> eryngi, you will be given permission at that time for more, if you are interesting enough
1489 2012-04-04 22:35:26 <eryngi> I have an idea how to make 51% attack impossible
1490 2012-04-04 22:35:30 <eryngi> (I think)
1491 2012-04-04 22:35:41 t7 has quit (Remote host closed the connection)
1492 2012-04-04 22:35:53 <sipa> always fun to hear people come up with a solution to a theoretically impossible problem
1493 2012-04-04 22:35:57 <sipa> but let's hear it!
1494 2012-04-04 22:35:57 <gmaxwell> No worries, I'm standing buy to crush your dreams.
1495 2012-04-04 22:36:00 <eryngi> I don't know if 150 chars is enough to explain it :D
1496 2012-04-04 22:36:02 <gmaxwell> s/buy/by/
1497 2012-04-04 22:36:14 <etotheipi_> eryngi, j/k; please proceed
1498 2012-04-04 22:36:19 <gmaxwell> eryngi: etotheipi_ was kidding—
1499 2012-04-04 22:36:25 <etotheipi_> (and let gmaxwell crush your dreams, he always crushes mine)
1500 2012-04-04 22:37:29 <eryngi> as I understand it, the 51% attack involves a hostile party, using tremendous power to mine separate from the rest of the blockchain.
1501 2012-04-04 22:37:44 <eryngi> am I correct?
1502 2012-04-04 22:37:52 BTC_Bear has quit (Ping timeout: 272 seconds)
1503 2012-04-04 22:38:09 p0s has quit (Remote host closed the connection)
1504 2012-04-04 22:39:24 <etotheipi_> eryngi, essentially -- they have to be able to mine blocks faster than the rest of the network combined
1505 2012-04-04 22:39:49 <eryngi> exactly
1506 2012-04-04 22:40:02 <eryngi> and driving their own dif up in the process
1507 2012-04-04 22:40:12 <sipa> not necessarily
1508 2012-04-04 22:40:29 <eryngi> but that wouldn't matter given the rich and able pigs with ASIC is what they are
1509 2012-04-04 22:40:44 <eryngi> so..
1510 2012-04-04 22:40:51 <eryngi> what if
1511 2012-04-04 22:40:58 <sipa> if they started off a fork in the past, they could regenerate a mirror chain from that point on, with the exact same timestamps as the real network
1512 2012-04-04 22:41:11 <sipa> and exact same difficulty
1513 2012-04-04 22:42:08 <eryngi> mut what if they would need to broadcast the block they're working on every time they receive work?
1514 2012-04-04 22:42:19 <eryngi> work = transaction
1515 2012-04-04 22:42:32 <eryngi> and what if the block with most transactions wins
1516 2012-04-04 22:42:47 <sipa> then miners would just create dummy transactions
1517 2012-04-04 22:42:53 <sipa> as that is easier than mining itself
1518 2012-04-04 22:43:22 <gmaxwell> Thats like picking the person who sets the most fires to run the fire department. :)
1519 2012-04-04 22:43:30 <eryngi> but it costs?
1520 2012-04-04 22:43:32 <gmaxwell> (obviously he's an expert at fires!)
1521 2012-04-04 22:43:47 <gmaxwell> Processing transactions? Pratically nothing.
1522 2012-04-04 22:44:30 <eryngi> hmm..
1523 2012-04-04 22:44:34 <eryngi> thing is
1524 2012-04-04 22:45:14 <eryngi> when a transaction gets broadcasted, every miner gets it relatively quick.
1525 2012-04-04 22:45:23 <sipa> not necessarily
1526 2012-04-04 22:45:43 <eryngi> as they get it, they broadcast the block they're working on
1527 2012-04-04 22:46:08 <eryngi> signed by the last transaction received
1528 2012-04-04 22:46:21 <gmaxwell> 0_o
1529 2012-04-04 22:46:21 <sipa> gmaxwell: the return value of EVP_Decrypt is never checked :S
1530 2012-04-04 22:46:50 <gmaxwell> sipa: to be fair.. that isn't a function I would naturally expect to able to fail!
1531 2012-04-04 22:46:59 <sipa> gmaxwell: agree
1532 2012-04-04 22:47:01 <eryngi> is that a good 0_o or a bad =_o
1533 2012-04-04 22:47:03 Hasbro has joined
1534 2012-04-04 22:47:12 <gmaxwell> Usually people assume the cipher to be a pseudorandom permutation....
1535 2012-04-04 22:47:17 <sipa> gmaxwell: assuming it was unpadded encryption, it cannot fail
1536 2012-04-04 22:47:28 <gmaxwell> eryngi: "signed by the last transaction received" doesn't make sense, and I can't guess what you mean.
1537 2012-04-04 22:47:42 <gmaxwell> sipa: oh! it checks the padding!
1538 2012-04-04 22:47:52 <sipa> exactly
1539 2012-04-04 22:47:57 <sipa> the last block always fails
1540 2012-04-04 22:48:11 <sipa> and so we end up with some 32 bytes remaining
1541 2012-04-04 22:48:17 <gmaxwell> I feel so much better that we're not quietly corrupting wallets.
1542 2012-04-04 22:48:26 <sipa> and CSecret::SetKey accepts it
1543 2012-04-04 22:48:35 <sipa> but *sometimes* the padding is valid
1544 2012-04-04 22:48:37 <brokenwallet> gmaxwell you'll still have a few people like me claiming that you are if that makes you feel better
1545 2012-04-04 22:49:01 <sipa> and we get a decrypted key of (most commonly) 47 bytes
1546 2012-04-04 22:49:06 <sipa> with a chance of 1/256
1547 2012-04-04 22:49:20 <sipa> and 46 bytes with a chance of 1/65536
1548 2012-04-04 22:49:23 <sipa> and so on
1549 2012-04-04 22:49:58 <gmaxwell> brokenwallet: I'd tried to convince the second person that was reporting this that he has a bad passphrase unsuccessfully.
1550 2012-04-04 22:50:11 <gmaxwell> but thats okay, we can make the software return the invalid passphrase notice in these cases.
1551 2012-04-04 22:50:21 <gmaxwell> And then people will stop convincing themselves. :)
1552 2012-04-04 22:50:24 <brokenwallet> yeah that would go a long way
1553 2012-04-04 22:51:06 <gmaxwell> It'll probably never convince the people who've already hit it— except you! :(
1554 2012-04-04 22:51:35 <brokenwallet> well at least i help cut off the flow of complaints (right after someone does all the actual work and fixes it)
1555 2012-04-04 22:51:47 <brokenwallet> someone = not me
1556 2012-04-04 22:52:20 <eryngi> gmaxwell, I have to contemplate more on my idea #1
1557 2012-04-04 22:52:26 <eryngi> here's idea #2
1558 2012-04-04 22:52:54 barmstrong has joined
1559 2012-04-04 22:53:36 <eryngi> encourage all miners to move to FPGA and create algo that changes from time to time thus incapacitating any hostile large scale ASIC operation
1560 2012-04-04 22:55:26 <gmaxwell> eryngi: FPGA's (and GPUs) are ASICs too you know. :)  If you have bounded variability then an asic can be made to encompass that. If the variability is not bounded then it's hard to know its secure against cryptographic attacks— and some versions may not fit (well) on some fpgas.
1561 2012-04-04 22:55:56 <sipa> also, why is a large scale ASIC operation a problem?
1562 2012-04-04 22:55:59 <gmaxwell> eryngi: why do you assume only bad users will use asics?
1563 2012-04-04 22:56:28 <eryngi> gmaxwell no, but that would be the most feaible attack
1564 2012-04-04 22:56:28 <gmaxwell> I'd happily buy a largecoin unicorn miner— if I didn't think it was probably a scam. :)
1565 2012-04-04 22:56:46 <eryngi> I have one ordered :)
1566 2012-04-04 22:56:51 <eryngi> 3 actually
1567 2012-04-04 22:57:02 <eryngi> 1 for me and 2 for friends
1568 2012-04-04 22:57:19 <forsetifox> Haven't seen any ASIC's for sale on the boards. Where?
1569 2012-04-04 22:57:37 <eryngi> but thinking about my investment made me think about 51% attack
1570 2012-04-04 22:57:39 <gmaxwell> eryngi: at the moment, because most of the users haven't made the transition— but once asic miners are common then it won't be a useful attack.. then you have to worry about people with atomic power plants or large dams. :)
1571 2012-04-04 22:58:10 <gmaxwell> eryngi: what did you actually buy?
1572 2012-04-04 22:58:15 <eryngi> forsetifox https://bitcointalk.org/index.php?topic=67505.0;all
1573 2012-04-04 22:58:26 <gmaxwell> oh indeed.
1574 2012-04-04 22:58:29 <eryngi> just signed up
1575 2012-04-04 22:58:33 <forsetifox> I'm I correct to assume that the only people that want to kill bitcoin would be governemnts and they could already do it?
1576 2012-04-04 22:58:53 <eryngi> I'd say yes
1577 2012-04-04 22:58:56 <forsetifox> Thanks for linky. =)
1578 2012-04-04 22:59:06 <eryngi> but not governments only
1579 2012-04-04 22:59:12 <gmaxwell> forsetifox: at best.. though I don't agree that governments would want to do that. (and if they did, technical attacks are not the best choices)
1580 2012-04-04 22:59:31 <eryngi> anyone who sees bitcoins as a threat and has the meanings to attack it
1581 2012-04-04 22:59:42 <forsetifox> Not technical attacks. Super computers that they could just borrow?
1582 2012-04-04 23:00:08 <eryngi> forsetifox, too inefficient
1583 2012-04-04 23:00:18 <brokenwallet> anyone have a estimate of GFlops required for 51% and compare that to top500.org?
1584 2012-04-04 23:00:22 <forsetifox> Um.. it wouldn't take too much time I don't think.
1585 2012-04-04 23:00:24 <gmaxwell> forsetifox: most supercomputers aren't so useful.. e.g. I did numbers for jaguar supercomputer, and concluded it would do only 1TH/s or something like that.
1586 2012-04-04 23:00:34 <eryngi> you'd need a large-scale 28nm ASIC production
1587 2012-04-04 23:00:51 <forsetifox> nod gmaxwell
1588 2012-04-04 23:01:04 <midnightmagic> lol why buy LargeCoin appliance? you're paying for their NRE.
1589 2012-04-04 23:01:18 <sipa> NRE?
1590 2012-04-04 23:01:28 <midnightmagic> the cost to develop an asic.
1591 2012-04-04 23:01:32 <gmaxwell> Non-reoccuring/recoverable engineering (Costs)
1592 2012-04-04 23:01:33 <brokenwallet> it is in escrow
1593 2012-04-04 23:01:50 <midnightmagic> after that's paid-off, they have $2/chip and you've just handed them a monopoly
1594 2012-04-04 23:02:03 <eryngi> midnightmagic, is nothing for anyone who has the means
1595 2012-04-04 23:02:04 <sipa> ic
1596 2012-04-04 23:02:21 <sipa> gmaxwell: "only" 1TH/s
1597 2012-04-04 23:02:21 <sipa> ...
1598 2012-04-04 23:02:33 <sipa> (ok, i have no idea about the price of that beast)
1599 2012-04-04 23:02:36 <gmaxwell> midnightmagic: because even so— at those prices the $/MH MH/j sound attractive vs gpus/fpgas... and you believe that other people will give them enough money anyways.
1600 2012-04-04 23:02:59 <gmaxwell> sipa: over a hundred million dollars.
1601 2012-04-04 23:03:28 <gmaxwell> (and thats not considering what it costs to cool and power.. a quarter million amd64 cores)
1602 2012-04-04 23:03:32 <eryngi> if someone wants to destroy  bitcoin, they'd need like half the current market cap
1603 2012-04-04 23:03:48 <Ahimoth> would be cheaper to buy 25K gpus
1604 2012-04-04 23:03:56 <midnightmagic> gmaxwell: they might as well just hand deepbit >51%, and pretend the next-gen fpga aren't on the horizon to make current-gen fpga cheap enough.
1605 2012-04-04 23:04:35 <eryngi> 4MUSD would buy you enough ASIC to overpower the network for months ( before dif adj
1606 2012-04-04 23:04:57 <brokenwallet> uhh... "When you receive the unit and connect it to the network, you will be granted a temporary mining license enabling the system to mine for a period of 30 days. When we receive the balance of your payment, a permanent mining license will be issued to you."
1607 2012-04-04 23:05:09 <gmaxwell> brokenwallet: yea super duper scarry.
1608 2012-04-04 23:05:55 <gmaxwell> I emailed them and asked if they were really planning on centerally controlling these devices they 'sell'. They didn't respond.
1609 2012-04-04 23:06:15 <forsetifox> Wow. Didn't know that worked like that. Heh.
1610 2012-04-04 23:06:15 <midnightmagic> gmaxwell: for similar price, you get the same hashrate in fpga..  and you actually get to control your own units (you don't with the LC unit)
1611 2012-04-04 23:06:49 <gmaxwell> midnightmagic: hm? they halved their price.
1612 2012-04-04 23:06:53 mmoya has joined
1613 2012-04-04 23:06:55 <brokenwallet> they should just sell hosted mining at that point. who cares if i'm holding the box if they fully control it anyway
1614 2012-04-04 23:07:01 barmstrong has quit (Remote host closed the connection)
1615 2012-04-04 23:07:03 <brokenwallet> they should just send me a power bill each month and keep the thing
1616 2012-04-04 23:07:05 <gmaxwell> you can get 20GH/s for $15k in FPGA?
1617 2012-04-04 23:07:12 <forsetifox> You're paying for power.
1618 2012-04-04 23:07:16 <jgarzik> sipa: correct
1619 2012-04-04 23:07:19 paulo_ is now known as away!~foo@112.204.27.38|paulo_
1620 2012-04-04 23:07:20 <eryngi> actually they clarified that issue
1621 2012-04-04 23:07:28 <midnightmagic> $15K?  last I checked it was $30. where's it say $15k?
1622 2012-04-04 23:07:32 <[Tycho]> etotheipi_: what's wrong with 2-of-3 multisigs ?
1623 2012-04-04 23:07:34 <eryngi> but still, I agree it's scary
1624 2012-04-04 23:07:40 <jgarzik> sipa: slightly more complex than that, but yes, that is basically it.  write, flush/sync/etc. to a temp file, then atomically rename
1625 2012-04-04 23:07:41 <brokenwallet> gmaxwell does this count http://www.butterflylabs.com/products/
1626 2012-04-04 23:07:59 TD has quit (Quit: TD)
1627 2012-04-04 23:08:22 <midnightmagic> gmaxwell: I see this: http://www.reddit.com/r/Bitcoin/comments/qj9ov/somebody_please_perform_an_roi_on_the_30k/
1628 2012-04-04 23:08:23 <etotheipi_> [Tycho], I don't understand your question -- *actually* executing multi-sig tx will be complicated
1629 2012-04-04 23:08:29 <gmaxwell> brokenwallet: BFL's history is a little bit of over promise under deliver.
1630 2012-04-04 23:08:30 <brokenwallet> 50GH/s for 30k and they have shipped products
1631 2012-04-04 23:08:45 <brokenwallet> true but i thought their current products at least work
1632 2012-04-04 23:08:50 <brokenwallet> if you can find them
1633 2012-04-04 23:09:12 <[Tycho]> I was talking about "etotheipi_: Malicious seller could offer to sell someone something for 100 BTC -- wait for the 2-of-2 or 2-of-3 tx to go through, then disappear, screwing over the buyer"etotheipi_: Malicious seller could offer to sell someone something for 100 BTC -- wait for the 2-of-2 or 2-of-3 tx to go through, then disappear, screwing over the buyer"
1634 2012-04-04 23:09:25 <gmaxwell> brokenwallet: six months after their initial announcement, in tiny quantities, with higher prices and lower performance.. yea... I mean sure, but that 50GH/$30k number should be taken with lots of salt.
1635 2012-04-04 23:09:50 <eryngi> gmaxwell, is it a totally idiotic assumption that we could change the algo so ASIC could not rule us?
1636 2012-04-04 23:09:58 <brokenwallet> i wonder if that number went down in coorelation to the shipped product's numbers
1637 2012-04-04 23:10:00 barmstrong has joined
1638 2012-04-04 23:10:07 <etotheipi_> [Tycho], the third party has no obligation to help you collect that money without charging the fee
1639 2012-04-04 23:10:21 <jgarzik> sipa: rename() on *nix or ReplaceFile() on Windows
1640 2012-04-04 23:10:22 <[Tycho]> It's called "trusted" for a reason :)
1641 2012-04-04 23:10:39 <gmaxwell> eryngi: it could be, but it would be a difficult change.
1642 2012-04-04 23:10:50 <gmaxwell> It would have to have a good motivation behind it.
1643 2012-04-04 23:11:09 <etotheipi_> [Tycho], how do they know there wasn't a legitimate dispute?  they have to investigate the other party, first
1644 2012-04-04 23:11:19 <eryngi> gmaxwell like if we had a apparent 51% attack
1645 2012-04-04 23:11:45 <gmaxwell> eryngi: It would have to be sustained and very problematic for a while, I expect.
1646 2012-04-04 23:11:51 <[Tycho]> etotheipi_: who are "they" ?
1647 2012-04-04 23:12:11 <etotheipi_> [Tycho], the third party
1648 2012-04-04 23:12:34 <eryngi> gmazwell, but what it was built into the system
1649 2012-04-04 23:12:42 <eryngi> that the algo changes
1650 2012-04-04 23:12:45 <eryngi> just like diff
1651 2012-04-04 23:12:57 <eryngi> ASIC couldn't cope
1652 2012-04-04 23:13:15 Hasbro has quit (Changing host)
1653 2012-04-04 23:13:15 Hasbro has joined
1654 2012-04-04 23:13:20 <etotheipi_> [Tycho], the third party can't just unlock the money for one party:  once it's in the 2-of-3 tx, they have an obligation to "arbitrate" which involves contacting the other party before unlocking the money -- they have to do their job and take the fee regardless of what either party says
1655 2012-04-04 23:13:23 <midnightmagic> but..  why punish an asic?
1656 2012-04-04 23:13:52 <eryngi> midnightmagic, latest tech can be bought with money
1657 2012-04-04 23:14:11 <gmaxwell> eryngi: sure it could. CPUs, GPUs, FPGAs, are all ASICs themselves. You could potentially level the playing field some. But as midnightmagic points out, thats pointless except if something odd happens in the temporary transition period.
1658 2012-04-04 23:14:16 <eryngi> if someone in power had a problem with bitcoin they could have that
1659 2012-04-04 23:14:20 barmstro_ has joined
1660 2012-04-04 23:14:34 barmstrong has quit (Read error: Connection reset by peer)
1661 2012-04-04 23:14:47 <gmaxwell> eryngi: once you're on a direct modern asic process the futher improvements come fairly slowly.
1662 2012-04-04 23:15:14 <eryngi> gmaxwell, I disagree on your conclusion that CPU, GPU, FPGA are all alike ASIC
1663 2012-04-04 23:15:31 <eryngi> CPU, GPU, custom ASIC are ASIC
1664 2012-04-04 23:15:34 <eryngi> FPGA is not
1665 2012-04-04 23:15:52 <gmaxwell> Sure they are.
1666 2012-04-04 23:16:08 <graingert> you could make an ASIC that was an fpga
1667 2012-04-04 23:16:19 <eryngi> I know that
1668 2012-04-04 23:16:23 <gmaxwell> Do you think they're made from magical pixie dust? :)  It's just an ASIC with a particular (wasteful) design.
1669 2012-04-04 23:16:33 <gmaxwell> But one which allows flexibility.
1670 2012-04-04 23:16:50 <eryngi> err
1671 2012-04-04 23:17:15 <eryngi> that kinda conflicts with the definition
1672 2012-04-04 23:17:30 <sipa> ASIC is an application-specific integrated circuit
1673 2012-04-04 23:17:36 <eryngi> I know that :)
1674 2012-04-04 23:17:42 <sipa> so yes, if it's not application-specific, it's not an ASIC
1675 2012-04-04 23:17:47 <sipa> but that doesn't matter
1676 2012-04-04 23:17:47 <midnightmagic> can you do all those layers of interconnects in a normal asic design?
1677 2012-04-04 23:17:51 <sipa> they're all IC's
1678 2012-04-04 23:18:05 <eryngi> sipa, yes, correct :)
1679 2012-04-04 23:19:02 <eryngi> the thing is.. if we had a way to modify the algo so that miners that are not ASIC would be able to adjust, it would make it a lot harder for anyone malicious to attack the network
1680 2012-04-04 23:19:17 <eryngi> right now all I need is about 4MUSD and I have it
1681 2012-04-04 23:19:35 <eryngi> and that's just half of the market cap
1682 2012-04-04 23:19:47 <sipa> but they would know which set of algorithms the next one is from, no?
1683 2012-04-04 23:19:47 <eryngi> err
1684 2012-04-04 23:19:50 <midnightmagic> the return is a disincentive
1685 2012-04-04 23:19:55 <eryngi> make it 1/10 of market cap
1686 2012-04-04 23:20:06 <gavinandresen> If you are willing to spend more money than the rest of the miners combined then you win, no matter what we do.
1687 2012-04-04 23:20:07 <sipa> market cap is meaningless for bitcoin
1688 2012-04-04 23:20:20 <eryngi> midnight, making new ASIC is not trivial
1689 2012-04-04 23:20:35 <eryngi> there's a delay due to design and production
1690 2012-04-04 23:20:46 <eryngi> flashing FPGA takes 3000ms
1691 2012-04-04 23:21:01 <brokenwallet> wouldnt they just infiltrate the committee that is deciding the next algorithm?
1692 2012-04-04 23:21:15 <brokenwallet> or pay them off
1693 2012-04-04 23:21:19 <gmaxwell> eryngi: yes, and draws 10x the power and has 1/10th the density of a dedicated design.. ::shrugs::
1694 2012-04-04 23:21:21 <eryngi> brokenwallet, that doesn't matter
1695 2012-04-04 23:21:29 <sipa> ok, i thought i understood the bug, did a patch, realized that this patch cannot have fixed the problem, but i am no longer able to reproduce it :(
1696 2012-04-04 23:21:55 <brokenwallet> so you accidently fixed it?
1697 2012-04-04 23:21:57 <gmaxwell> sipa: you're starting to sound like our users.
1698 2012-04-04 23:22:13 <eryngi> brokenwallet, if the algo is changing, they'd need a new ASIC for every new one
1699 2012-04-04 23:22:29 <midnightmagic> eryngi: that's not an issue.. if someone were to do that, what do you think the current miners' response would be? :-)
1700 2012-04-04 23:22:32 <sipa> eryngi: who will decide the algorithm?
1701 2012-04-04 23:22:38 <sipa> a committee? that is very non-bitcoin
1702 2012-04-04 23:22:51 <eryngi> sipa, it could bealgorythm-based
1703 2012-04-04 23:23:12 <gmaxwell> then you just make the asic able to cover the whole design space.
1704 2012-04-04 23:23:22 <sipa> eryngi: if it's algorithm-based, people will know which set of hashing algorithm will be from
1705 2012-04-04 23:23:32 <gmaxwell> eryngi: why did you assume FPGA here and not just say CPU, btw?
1706 2012-04-04 23:23:35 <sipa> and create a slightly more expensive asic that can do them all
1707 2012-04-04 23:23:45 * gmaxwell tests for understanding the universality of computation
1708 2012-04-04 23:25:09 <eryngi> gmaxwell, CPU can not beat FPGA in efficiency
1709 2012-04-04 23:25:23 <eryngi> ASIC can not beat FPGA in adaptability
1710 2012-04-04 23:26:04 <eryngi> hence, constatly changing algo, would protect from ASIC attack, but not ofcouse from FPGA attack
1711 2012-04-04 23:26:13 <gmaxwell> eryngi: sure, it can. Depends on what you're doing.
1712 2012-04-04 23:26:31 <eryngi> CPU vs FPGA?
1713 2012-04-04 23:26:33 <eryngi> ofc
1714 2012-04-04 23:26:42 <eryngi> but not if designed properly
1715 2012-04-04 23:26:59 <eryngi> not with the CPUs we have now for sure
1716 2012-04-04 23:27:08 <gavinandresen> I don't understand.  Why wouldn't The Bad Guys just spend millions of dollars on FPGAs to get 50+% of hashing power?
1717 2012-04-04 23:27:26 <eryngi> gavinandersen, to destroy bitcoin?
1718 2012-04-04 23:27:35 <gavinandresen> yes, that's the attack you're talking about right?
1719 2012-04-04 23:27:40 <midnightmagic> eryngi: the defense is pointless.. destroying bitcoin merely requires a couple acts of legislation.
1720 2012-04-04 23:27:55 <eryngi> midnightmagic, I don't see it that way
1721 2012-04-04 23:28:13 <midnightmagic> oh no? you've never sen a government use the taxes of its people against them? :)
1722 2012-04-04 23:28:15 <eryngi> we haven't destroyed anything that way
1723 2012-04-04 23:28:19 <gmaxwell> gavinandresen: He's mostly concerned that there is a gap here because the GoodGuys are not yet using the most efficient means possible that the BadGuys can do their attack at a minor discount.
1724 2012-04-04 23:28:34 <eryngi> gmaxwell +1
1725 2012-04-04 23:28:43 <forsetifox> Heh.
1726 2012-04-04 23:28:51 <gavinandresen> gmaxwell: okey doke.  In my experience, The Good Guys are moving much faster these days.
1727 2012-04-04 23:28:57 <[Tycho]> eryngi: you are wrong
1728 2012-04-04 23:29:03 brwyatt is now known as Away!~brwyatt@pool-71-252-154-11.dllstx.fios.verizon.net|brwyatt
1729 2012-04-04 23:29:10 <eryngi> [Tycho] please elaborate
1730 2012-04-04 23:29:23 <[Tycho]> Why ?
1731 2012-04-04 23:29:29 <copumpkin> o.O
1732 2012-04-04 23:30:02 <eryngi> you just proclaimed me wrong. please elaborate
1733 2012-04-04 23:30:02 <gmaxwell> I agree. ... though some things are a bit concerning.. E.g. 'Largecoin' apparently raising NRE for an ASIC which you'll host but they'll centrally control?
1734 2012-04-04 23:30:07 <eryngi> I've spoken a lot here
1735 2012-04-04 23:30:22 <gmaxwell> eryngi: [Tycho] spent a lot of time researching making bitcoin chips.
1736 2012-04-04 23:30:35 <eryngi> gmaxwell, you should read the whole thread
1737 2012-04-04 23:30:47 <copumpkin> gmaxwell: it still seems kind of weird to just say "you're wrong" without bothering to elaborate
1738 2012-04-04 23:30:54 <[Tycho]> eryngi: may be we should punish GPU users first ? Bad Guys With Money can buy a lot of GPUs for 51% attack.
1739 2012-04-04 23:30:59 <midnightmagic> eryngi: what's the incentive to destroy bitcoin?  there is none.
1740 2012-04-04 23:31:24 splatster has joined
1741 2012-04-04 23:31:33 <eryngi> midnightmagic, f ex. no being very happy about people gaining power over currency?
1742 2012-04-04 23:31:47 <midnightmagic> eryngi: Second Life has a bigger economy.
1743 2012-04-04 23:31:52 <midnightmagic> eryngi: Bitcoin is small.
1744 2012-04-04 23:32:19 <gmaxwell> eryngi: I don't see anything in that thread reversing on the central control thing.
1745 2012-04-04 23:32:28 <eryngi> [Tycho] if it's designed correctly there's no punishment to anyone who is using unoptimized miners
1746 2012-04-04 23:32:50 <eryngi> gmaxwell, I'll look it up for you
1747 2012-04-04 23:32:56 <gmaxwell> eryngi: thanks!
1748 2012-04-04 23:32:59 <[Tycho]> eryngi: so legitimate users SHOULD use only unoptimized miners ?
1749 2012-04-04 23:33:46 <eryngi> https://bitcointalk.org/index.php?topic=67505.msg785352#msg785352
1750 2012-04-04 23:34:07 <eryngi> [Tycho] I would say that would be the way
1751 2012-04-04 23:34:23 RainbowDashh has joined
1752 2012-04-04 23:34:45 <eryngi> anyone close to the powers in control has greater chance of having access to tech that is still unavailable to the public
1753 2012-04-04 23:35:11 <eryngi> like quantum computing or new process
1754 2012-04-04 23:35:16 <sipa> of course, the NSA has a quantum computer that cracks SHA-256 in a second already
1755 2012-04-04 23:35:47 <eryngi> sipa, from what I've read, it actually only gets 200% benefit in BTC
1756 2012-04-04 23:36:17 <gmaxwell> er. sqrt() != 0.5
1757 2012-04-04 23:36:25 <sipa> if a general purpose quantum computer exists of sufficient size
1758 2012-04-04 23:36:29 <sipa> bitcoin has a problem
1759 2012-04-04 23:36:42 <lianj> http://i0.kym-cdn.com/photos/images/newsfeed/000/203/685/conspiracy-keanu.jpg
1760 2012-04-04 23:37:02 <sipa> the signing algorithm would be trivial to crack, and the sha256 hash function would only retain 128 bits of security
1761 2012-04-04 23:37:09 <sipa> the latter is not a problem, but the former is
1762 2012-04-04 23:37:12 <Eliel> hmm? I was under the impression that quantum computers aren't really helpful against cryptographic hash functions.
1763 2012-04-04 23:37:20 <sipa> Eliel: they aren'
1764 2012-04-04 23:37:21 <sipa> t
1765 2012-04-04 23:37:23 <eryngi> eliel +1
1766 2012-04-04 23:37:38 RainbowDashh has quit (Read error: Connection reset by peer)
1767 2012-04-04 23:37:45 <gmaxwell> sipa: well, 'sufficiently large' is quite big for the ecdsa, even ignoring error correction. But ::nods::
1768 2012-04-04 23:37:53 <gmaxwell> There is signing that is QC hard, ECDSA is not it.
1769 2012-04-04 23:37:54 RainbowDashh has joined
1770 2012-04-04 23:38:10 <gmaxwell> Unfortunately the QC hard signing options are bloaty.
1771 2012-04-04 23:39:12 <etotheipi_> QC's cut down any guess-and-check problem from:  gotta make N guesses, to:  gotta make approx sqrt(N) guesses
1772 2012-04-04 23:39:39 <etotheipi_> sha256 gives you 256 bits of security... but it's only 128-bits of security to a QC
1773 2012-04-04 23:39:50 <etotheipi_> which is still completely infeasible
1774 2012-04-04 23:39:54 minimoose has joined
1775 2012-04-04 23:40:04 <gmaxwell> etotheipi_: we said all this with different words. :)
1776 2012-04-04 23:41:10 <etotheipi_> gmaxwell, sometimes that's what people need to understand :)
1777 2012-04-04 23:41:20 <eryngi> gmaxwell, returning to model #1
1778 2012-04-04 23:41:25 <sipa> i really don't understand why my patch seems to have fixed the problem
1779 2012-04-04 23:42:38 <eryngi> ah..
1780 2012-04-04 23:42:54 <eryngi> it's too fucking complicated to write in irc :D
1781 2012-04-04 23:43:37 <eryngi> but the idea is to have a transaction tell the prev block it has. wouldn't that protect?
1782 2012-04-04 23:43:57 <brokenwallet> sipa was the wallet modified?
1783 2012-04-04 23:44:03 <brokenwallet> did you try the one from my email again
1784 2012-04-04 23:44:10 <eryngi> or client issuing the transaction manifest the prev block
1785 2012-04-04 23:44:11 <sipa> brokenwallet: of course
1786 2012-04-04 23:44:35 <sipa> eryngi: it's essential that transactions are not tied to a particular block
1787 2012-04-04 23:44:46 <sipa> during a reorgnisation they need to be able to move the a different chain entirely
1788 2012-04-04 23:44:58 <eryngi> I see
1789 2012-04-04 23:45:40 <eryngi> I'm just thinking of ways to monitor forking
1790 2012-04-04 23:45:56 <sipa> we already do
1791 2012-04-04 23:46:11 <eryngi> ok. Maybe I should just leave it to you then :)
1792 2012-04-04 23:49:30 <eryngi> I'm a designer, not a programmer
1793 2012-04-04 23:50:18 copumpkin has quit (Quit: Computer has gone to sleep.)
1794 2012-04-04 23:50:25 RedEmerald has joined
1795 2012-04-04 23:51:21 Rabbit67890 has joined
1796 2012-04-04 23:51:31 chk2 has joined
1797 2012-04-04 23:51:45 RainbowDashh has quit (Disconnected by services)
1798 2012-04-04 23:51:53 Rabbit67890 is now known as RainbowDashh
1799 2012-04-04 23:52:23 Snapman is now known as Snapman[afkers]
1800 2012-04-04 23:53:18 chk has quit (Ping timeout: 245 seconds)
1801 2012-04-04 23:53:23 <OneFixt> gmaxwell: https://bitcointalk.org/index.php?topic=67505.msg785449#msg785449
1802 2012-04-04 23:58:34 Snapman[afkers] is now known as Snapman