1 2012-01-21 00:59:51 <gribble> New news from bitcoinrss: luke-jr opened pull request 774 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/774>
  2 2012-01-21 01:01:42 <luke-jr> I wish GitHub put the summary in those -.-
  3 2012-01-21 01:02:40 <TuxBlackEdo> luke-jr, i wrote something to the blockchain!
  4 2012-01-21 01:02:45 <TuxBlackEdo> but
  5 2012-01-21 01:02:53 <TuxBlackEdo> i didnt know it would turn the message backwards
  6 2012-01-21 01:02:54 <TuxBlackEdo> lol
  7 2012-01-21 01:02:56 <luke-jr> &
  8 2012-01-21 01:03:15 <luke-jr> that makes you a graffiti spammer
  9 2012-01-21 01:03:25 <TuxBlackEdo> http://blockchain.info/tx-index/14551557/f6707e09bcf0bd80884a7ffd40b92357144e50a7a426b6cade8fe329d71ee9f2
 10 2012-01-21 01:03:32 <TuxBlackEdo> dude you did it like a million times
 11 2012-01-21 01:03:36 <TuxBlackEdo> i wanted to do it just once
 12 2012-01-21 01:03:58 <TuxBlackEdo> i am waiting for it to confirm in the next block
 13 2012-01-21 01:04:13 <TuxBlackEdo> so i can see if it shows up using "strings -n 20 blk0001.dat"
 14 2012-01-21 01:06:43 <luke-jr> no, I didn't.
 15 2012-01-21 01:07:38 <TuxBlackEdo> yeah you did
 16 2012-01-21 01:11:17 <TuxBlackEdo> lol
 17 2012-01-21 01:11:20 <TuxBlackEdo> it showed up
 18 2012-01-21 01:11:24 <lianj> paying fees, pff
 19 2012-01-21 01:11:31 <TuxBlackEdo> !ti delipmoc yllaniF
 20 2012-01-21 01:11:32 <gribble> Error: "ti" is not a valid command.
 21 2012-01-21 01:12:16 <gribble> Error: "ti" is not a valid command.
 22 2012-01-21 01:12:16 <TuxBlackEdo> # tail -c 5000
 23 2012-01-21 01:12:17 <gribble> Error: "ti" is not a valid command.
 24 2012-01-21 01:12:34 <TuxBlackEdo> er
 25 2012-01-21 01:12:55 <TuxBlackEdo> luke-jr, tail -c 5000 ~/.bitcoin/blk0001.dat |strings -n 20
 26 2012-01-21 01:13:17 <luke-jr> TuxBlackEdo: mine aren't graffiti though
 27 2012-01-21 01:14:11 <TuxBlackEdo> yours isn't but mine is?
 28 2012-01-21 01:14:25 <luke-jr> correct
 29 2012-01-21 01:14:28 <luke-jr> yours is transaction spam
 30 2012-01-21 01:14:39 <luke-jr> mine are in the coinbase, which was designed for it
 31 2012-01-21 01:15:47 <TuxBlackEdo> you used up way more data in the blockchain then i have
 32 2012-01-21 01:16:52 <TuxBlackEdo> i think i should submit a patch for the windows makefile to make it compilable
 33 2012-01-21 01:19:01 <Diablo-D3> I thought up new idea
 34 2012-01-21 01:25:11 <CIA-76> bitcoin: Luke Dashjr watch * r3ade39605b43 supybot-bitcoin-marketmonitor-personal/plugin.py: hack to announce multi-output blocks in #eligius http://tinyurl.com/7vufxkd
 35 2012-01-21 03:01:27 <CIA-76> DiabloMiner: Patrick McFarland master * rea64dff / (2 files in 2 dirs): Improve kernel mangling parser, fix accidental disabling of BFI_INT - http://git.io/E_XO3Q https://github.com/Diablo-D3/DiabloMiner/commit/ea64dff10714794ffe805b8a4d53c3ed72c4c41b
 36 2012-01-21 04:00:39 <Joric> what's the story with the 1NS17iag9jJgTHD1VXjvLCEnZuQ3rJED9L ? who's address is that?
 37 2012-01-21 04:37:58 <gmaxwell> Joric: huh, what do you mean whats the story?
 38 2012-01-21 04:48:53 <etotheipi_> the Satoshi nodes require a fee for relay/mining for large size, btc*nconf is too low, or there's a dust output... what about a tx that spends outputs with zero-confirmations?
 39 2012-01-21 04:49:59 <gmaxwell> etotheipi_: it's just the same formula.
 40 2012-01-21 04:50:17 <gmaxwell> If the only input is zero confirm then the priority of that txn will be 0 and need a fee.
 41 2012-01-21 04:50:21 <etotheipi_> so as long as there's enough other outputs to exceed 1-BTC-at-144 confirmations, it's okay
 42 2012-01-21 04:50:33 <gmaxwell> if there are multiple inputs then the priority may be high enough.. right.
 43 2012-01-21 04:50:51 <etotheipi_> perfect... I had this all figured out before I integrated zero-conf transactions...
 44 2012-01-21 04:50:59 <gmaxwell> though it won't be forwarded or mined by any node which isn't accepting the first.
 45 2012-01-21 04:51:16 <etotheipi_> what does that mean?
 46 2012-01-21 04:51:20 <etotheipi_> the first what?
 47 2012-01-21 04:51:38 <gmaxwell> (er, if that wasn't clear: the txn which is not yet confirmed must be in the memory pool of any node that would forward or mine its child)
 48 2012-01-21 04:51:59 <etotheipi_> oh, right
 49 2012-01-21 04:52:20 <gmaxwell> so if the not yet confirmed transaction happens to be not confirmed because it's failing the fee rules and isn't getting forwarded, the second won't forward either if it meets the rules in isolation.
 50 2012-01-21 04:52:31 <gmaxwell> Of even if the composition of the two meets the rules, which is arguably a bug.
 51 2012-01-21 04:53:29 <etotheipi_> so then, it should be easy enough to request the first tx from the node, to determine if it has it in the memory pool
 52 2012-01-21 04:53:43 <etotheipi_> then send your tx if it does
 53 2012-01-21 04:56:22 <etotheipi_> and I'm sooo glad I lost my SelectCoins unit-test... ugh
 54 2012-01-21 04:56:38 <gmaxwell> cute idea. (the probe idea)
 55 2012-01-21 04:57:15 <etotheipi_> I know you don't like it, but for now I'm basically relying entirely on the localhost bitcoind to tell me that I did things right
 56 2012-01-21 04:57:40 <etotheipi_> so before I add a tx to my own memory pool, I'm probing bitcoind to see if the tx is in its memory pool
 57 2012-01-21 04:58:28 <etotheipi_> one day, I will cut the umbilical cord, and have independent networking, but right now it's pretty convenient to have bitcoin-qt/bitcoind between me and the network
 58 2012-01-21 04:58:53 <etotheipi_> prob better for the network too.. since everyone who uses Armory will have to maintain the full-validation node in the background
 59 2012-01-21 05:03:41 <etotheipi_> my understanding is, implementing full-validation is going to be a nightmare
 60 2012-01-21 05:05:12 <CIA-76> bitcoin: Luke Dashjr * r9a0bcc889084 cgminer/ (main.c miner.h): Bugfix: Iterate over devices rather than threads for watchdog (and ADL) http://tinyurl.com/74xd5h6
 61 2012-01-21 05:05:13 <CIA-76> bitcoin: Con Kolivas * r1b0b7e4a7ca6 cgminer/ (main.c miner.h): Merge pull request #78 from luke-jr/bugfix_adl http://tinyurl.com/77snr6z
 62 2012-01-21 05:22:21 <gmaxwell> etotheipi_: I don't dislike it, as much as it makes me sad because we really do need independant implementations of full validation.
 63 2012-01-21 05:23:49 <etotheipi_> gmaxwell, I agree with you in concept, but I feel I can add tremendously more value to Bitcoin (in general) by expanding functionality
 64 2012-01-21 05:24:58 <gmaxwell> We don't agree, but thats okay. I think that value is added by there being different perspectives on what adds value to bitcoin.
 65 2012-01-21 05:25:12 <etotheipi_> one day, I probably will do full-validation... but for now I believe that offline wallets, paper backups, and eventually multi-sig... all will help people use the network, which is what I think is needed
 66 2012-01-21 05:25:53 <gmaxwell> (mostly my non-aggreement is that it so much easier and fun to make features in the context of 'toy' software rather than the full production code, we'll probably see lots of features made in systems like yours but not become easily accessible to most users)
 67 2012-01-21 05:26:01 <etotheipi_> right now I see a big black hole with regards to security for companies that plan to handle LOTS of money... offline wallets solve that
 68 2012-01-21 05:27:11 <etotheipi_> Armory is in no way toy software... I recognize that a lot of projects out there are more like experiments
 69 2012-01-21 05:27:54 <etotheipi_> but it is production quality (or will be), and I have plans to bring system requirements down to "normal"... there's a reason I have separated "Standard" and "Advanced" usermodes (and "Developer"
 70 2012-01-21 05:27:58 <gmaxwell> The fact that you have to use it on top of a full node and that it has resources requirements _in excess_ of a full node on its own greatly reduces its usability. I think in some ways this degrades it to toy status.
 71 2012-01-21 05:28:25 <etotheipi_> my goal was never to require it to suck lots of memory and require the full-node
 72 2012-01-21 05:28:31 <etotheipi_> this is transient
 73 2012-01-21 05:28:37 <gmaxwell> (the difficulty of running a full node is already a major usability problem in bitcoin)
 74 2012-01-21 05:28:45 <gmaxwell> etotheipi_: Okay, fine enough.
 75 2012-01-21 05:29:06 <gmaxwell> etotheipi_: I also still think your offline proposal has the wrong security properties.
 76 2012-01-21 05:29:09 <etotheipi_> I totally agree with you if I thought regular uses would use it, as is
 77 2012-01-21 05:29:47 <etotheipi_> but the fact is, I believe offline wallets are critical, and that people with shitloads of money won't care about buying a decent computer to do it... the fact is there is no other way to do it
 78 2012-01-21 05:29:59 <gmaxwell> If a comprimised interface node can fake out the value of the transaction in order to bankrupt the offline wallet, in an undetectable way, I think you might as well not have an offline wallet.
 79 2012-01-21 05:30:02 <etotheipi_> and eventually more people will have access when I switch
 80 2012-01-21 05:30:35 <etotheipi_> gmaxwell, you're focusing on the wrong part of it... this is not a problem so much as it is a precaution
 81 2012-01-21 05:30:48 <gmaxwell> etotheipi_: I say the same to you! :)
 82 2012-01-21 05:31:01 <gmaxwell> I've hacked your interface node. This is the risk you worried about which is why your private keys are offline.
 83 2012-01-21 05:31:02 <etotheipi_> it's like saying, don't use prime numbers, P whose P-1 has lots of small factors when picking RSA numbers
 84 2012-01-21 05:31:38 <gmaxwell> I wait for you to do a transaction from it and I swap out the inputs so that you instead give away all your coins as fees.
 85 2012-01-21 05:31:40 <etotheipi_> just because some prime numbers are insecure doesn't mean the whole system is broken, you just implement additional safeguards to make sure you don't "fall" for it
 86 2012-01-21 05:31:44 <gmaxwell> You process it never the wisers.
 87 2012-01-21 05:31:55 <gmaxwell> s/wisers/wiser/   until its too late.
 88 2012-01-21 05:32:34 <gmaxwell> etotheipi_: offline wallets that only provide highly speculative security improvements seem kind of worthless to me.
 89 2012-01-21 05:33:50 <gmaxwell> And by doing so would close real vulnerabilties.
 90 2012-01-21 05:33:54 <etotheipi_> you're talking about complete subsets of existing problems with online wallets
 91 2012-01-21 05:33:59 <gmaxwell> Not just speculative warm fuzzies.
 92 2012-01-21 05:34:22 <etotheipi_> there's one to two orders of magnitude less risk with offline wallets than online wallets:  sure there's still attack vectors, but there's only a fraction of them
 93 2012-01-21 05:35:11 <CIA-76> bitcoin: Luke Dashjr * rf634b68e7c93 cgminer/main.c: Bugfix: Only show each device once in exit summary http://tinyurl.com/84mu33e
 94 2012-01-21 05:35:14 <CIA-76> bitcoin: Con Kolivas * r9bf2028d9bfc cgminer/main.c: Merge pull request #79 from luke-jr/bugfix_summary http://tinyurl.com/7xwzwtu
 95 2012-01-21 05:35:14 <gmaxwell> No. I don't believe that you're correct. The interesting attack vector with online wallets which isn't closed by wallet encryption is that a hostile party gains control of the online wallet system. What other attacks are you concerned with?
 96 2012-01-21 05:35:45 <etotheipi_> any attack that compromises the offline wallet also compromises an online wallet
 97 2012-01-21 05:35:49 <_W_> etotheipi_, what is the average time a system stays infected undetected, once it has a virus or trojan?
 98 2012-01-21 05:36:49 <etotheipi_> this is one reason I do, eventually, want to make it even more secure, perhaps switch to purely-QR codes
 99 2012-01-21 05:37:04 <etotheipi_> but just like everything else in the world, you have to start somewhere and evolve...
100 2012-01-21 05:37:28 <gmaxwell> etotheipi_: Thats true, but misleading.  If every realistic attack that compromises the online(encrypted) wallet also compromises the offline one because the offline scheme is weak, then the offline system only added complexity. This complexity creates risks on its own, and as such your "strict subset of vulnerabilities" may actually be worse for users.
101 2012-01-21 05:38:08 <gmaxwell> etotheipi_: I don't understand why you don't just make it possible for the offline wallet to confirm the real values: Include the input transactions. It can hash them and confirm the values.
102 2012-01-21 05:38:17 <etotheipi_> gmaxwell, one thing we agree on for sure is that the lack of being able to verify tx-inputs in an offline transaction is bad
103 2012-01-21 05:38:41 <gmaxwell> The the offline system can simply say "I'm going to sign this thing which spends 5 btc to X and pays 0 in fees. Are you sure?"
104 2012-01-21 05:38:47 <etotheipi_> if you ahve a recommendation for a way to securely do that, I'd love to hear it, but I haven't thought of a way yet
105 2012-01-21 05:38:53 <gmaxwell> etotheipi_: !
106 2012-01-21 05:39:09 <gmaxwell> etotheipi_: sillyness. If the inputs aren't valid they just won't work. This isn't hard at all.
107 2012-01-21 05:39:37 <gmaxwell> etotheipi_: send a copy of the input transactions. Thats it.  A dishonest interface node could DOS, but they couldn't make you burn coin.
108 2012-01-21 05:40:00 <etotheipi_> I already include the tx-input values, but of course those could be manipulated, and the signed transaction will still be valid
109 2012-01-21 05:40:13 <k9quaint> I think at some fundamental level, offline confirmations are add odds with a web of trust model
110 2012-01-21 05:40:15 <etotheipi_> unfortunately, those values are not part of the signed message, which is what's killing me
111 2012-01-21 05:40:17 <gmaxwell> You need to include the whole input transactions so the offline node can compute the hashes themselves.
112 2012-01-21 05:40:39 <etotheipi_> gmaxwell, that's a fine idea
113 2012-01-21 05:40:50 <gmaxwell> The outputs are part of the txn id of the inputs! :) yea. it should JustWork"
114 2012-01-21 05:41:09 <etotheipi_> I was leaning towards a less-bloaty way to do it... but at least I have that which works
115 2012-01-21 05:41:48 <gmaxwell> Yea, I don't think there is a less bloaty way. Sadly. You could have it remember the inputs, so they might already be there, but you still have to get them there.
116 2012-01-21 05:42:11 <gmaxwell> but hey, a typical transaction is like.. what 250 bytes or so.. it's not that terrible.
117 2012-01-21 05:42:20 <etotheipi_> well if there's a dozen inputs...
118 2012-01-21 05:42:33 <etotheipi_> that won't be fitting in a QR code
119 2012-01-21 05:42:53 <gmaxwell> there are QR codes that can contain 200k IIRC
120 2012-01-21 05:43:01 <etotheipi_> and there you have it... after all this my time spent on offlien transactions may not have been a waste
121 2012-01-21 05:43:40 <etotheipi_> I just don't like the attitude that you'll just critisize my first shot at it as if it's all set in stone and I refuse to improve it...
122 2012-01-21 05:43:57 <gmaxwell> if QR falls down you could use sound.. some 1200 baud FSK modem could easily be made to work over open air.  :)
123 2012-01-21 05:44:16 <etotheipi_> IR tx/rx could work well
124 2012-01-21 05:44:26 <_W_> Data Matrix barcodes are infinitely expandable. The limit would be what the device supported
125 2012-01-21 05:44:30 <gmaxwell> or just RS232.
126 2012-01-21 05:44:53 <etotheipi_> in fact, a lot of laptops already have IR tx/rx hardware
127 2012-01-21 05:44:58 <k9quaint> why don't you just print circuits out of conductive ink, and hook your paper up to the internet
128 2012-01-21 05:45:04 <gmaxwell> (I mean, you can't really say RS232 is less secure than IR.. just because there is no 'wire' doesn't really matter)
129 2012-01-21 05:45:16 <gmaxwell> but sure, existing hardware is nice
130 2012-01-21 05:46:50 <gmaxwell> ideally you'd want to be able to have some harded secure hardware that looks like a tamagotchi. :) stores a determinstic wallet seed in encrypted form. you IRDA over the txn, it asks for a confirmation then asks you to punch in the decrpytion key, then it IRDAs back the signed transaction.
131 2012-01-21 05:47:18 <etotheipi_> well I'll keep in mind the idea of including full tx... I had glossed over it in my head before, but for some reason figured I'd need the full chain of tx, not just the most recent ones
132 2012-01-21 05:47:50 <etotheipi_> probably nothing as fancy as you suggest, though :)
133 2012-01-21 05:48:40 <etotheipi_> and for now, it's time for bed
134 2012-01-21 05:48:45 <etotheipi_> thanks gmaxwell
135 2012-01-21 05:48:46 <gmaxwell> goodnight!
136 2012-01-21 05:48:55 <gmaxwell> Thanks for listening to me rant. :)
137 2012-01-21 05:49:08 <Diablo-D3> gmaxwell: hate 2.6 so goddamned much
138 2012-01-21 05:51:29 <k9quaint> Diablo-D3: you should have a flag to print out the device list
139 2012-01-21 05:52:10 <Diablo-D3> k9quaint: it does it on startup anyhow
140 2012-01-21 05:53:00 <k9quaint> thats if you want to run everything at the same time
141 2012-01-21 05:53:20 <Diablo-D3> it lists it in order
142 2012-01-21 05:53:59 <k9quaint> if I just want to run a single device in isolation, I gotta run aticonfig and add 1 :P
143 2012-01-21 05:54:35 <Diablo-D3> no you dont
144 2012-01-21 05:54:38 <Diablo-D3> you look at the list and count.
145 2012-01-21 05:54:44 <Diablo-D3> FOUND 5850
146 2012-01-21 05:54:45 <Diablo-D3> FOUND 5850
147 2012-01-21 05:54:49 <Diablo-D3> well, I want the third 5850
148 2012-01-21 05:54:52 <Diablo-D3> so -D 3
149 2012-01-21 05:55:32 <k9quaint> like I said, I have to run everything first
150 2012-01-21 05:55:33 <Mad7Scientist> bitcoin is locked up again on I/O to ~/.bitcoin/
151 2012-01-21 05:55:51 <Mad7Scientist> blank window. Miner disconnected.
152 2012-01-21 05:56:10 <gmaxwell> Mad7Scientist: Yes, mr. NFS man. Include backtraces with your complaints. :)
153 2012-01-21 05:56:12 <Mad7Scientist> the longer I leave bitcoin open the worse this gets
154 2012-01-21 05:57:11 <gmaxwell> (otherwise you're asking me to spent the time to setup a slow low memory NFS client on a slow network.. just to try to reproduce what you're already reproducing)
155 2012-01-21 05:57:35 <Mad7Scientist> You probably have enough disk cache to not see this problem as well as most bitcoin users
156 2012-01-21 05:57:37 <k9quaint> gmaxwell: well, typing on IRC is just going to slow you down, go reproduce it!
157 2012-01-21 05:57:41 <Mad7Scientist> I don't think it's related to NFS
158 2012-01-21 05:58:04 <Mad7Scientist> 12MB/sec is not slow
159 2012-01-21 05:58:12 <gmaxwell> Mad7Scientist: I believe your problem is real. I also believe it requires low cache _and_ some kind of slowish media.
160 2012-01-21 05:58:18 <gmaxwell> It's slow compared to anything I normally run.
161 2012-01-21 05:58:33 <gmaxwell> I'm not discounting your problem at all. I think it's very important.
162 2012-01-21 05:58:41 <k9quaint> I can write faster left-handed on 3M stick notes
163 2012-01-21 05:59:02 <gmaxwell> But I can't easily reproduce it.  Sooo.. it would be super useful if you could identify what code paths are causing it to get stuck most often.
164 2012-01-21 05:59:05 <Mad7Scientist> There is one thread that is "D" in top
165 2012-01-21 05:59:11 <Mad7Scientist> and it has been that way for 5 minutes
166 2012-01-21 05:59:16 <gmaxwell> Mad7Scientist: great, attack to it with gdb and get a backtrace.
167 2012-01-21 05:59:22 <gmaxwell> er attach.
168 2012-01-21 05:59:39 <Mad7Scientist> I might try later this week
169 2012-01-21 05:59:51 <Mad7Scientist> I hope the default binary has debugging enabled
170 2012-01-21 05:59:59 <Mad7Scientist> well it just unfroze
171 2012-01-21 06:00:07 <gmaxwell> OK, please do. With a backtrace the problem should be super obvious.
172 2012-01-21 06:00:33 <gmaxwell> Mad7Scientist: I believe it has symbols.
173 2012-01-21 06:00:50 <cjdelisle> love when it trashes the stack and it's like:  well... it was a memcpy() :)
174 2012-01-21 06:00:50 <gmaxwell> (er, well I think it has good odds of being obvious at least.)
175 2012-01-21 07:10:20 <CIA-76> bitcoin: Con Kolivas * r3cad815e0f58 cgminer/README: Fanspeed faq. http://tinyurl.com/7mkzv2g
176 2012-01-21 07:37:25 <helo> ssh bitcoin pubkey authentication?
177 2012-01-21 07:40:16 <cjdelisle> shhh
178 2012-01-21 07:41:53 <gmaxwell> helo: I don't think that makes much useful sense.
179 2012-01-21 07:42:18 <gmaxwell> Going to extract a bitcoin private key to feed to your ssh client? why bother. Just use regular ssh keys.
180 2012-01-21 07:43:05 <gmaxwell> If you have some inner hatred of RSA and want to use some other crypto, ... just because, then you want http://ed25519.cr.yp.to/
181 2012-01-21 07:54:12 <cjdelisle> <3 djb
182 2012-01-21 07:54:24 <cjdelisle> I saved one processor cycle today, #winning
183 2012-01-21 10:44:55 <sh69dow> how do I bitcoin mine to make monies?
184 2012-01-21 10:45:25 <nathan7> First you buy a ???1000 mining rig
185 2012-01-21 10:45:31 <nathan7> *or* find yourself a time machine
186 2012-01-21 10:45:34 <nathan7> and start mining
187 2012-01-21 10:56:13 <sh69dow> nathan7: I want to mine bro
188 2012-01-21 10:56:44 <nathan7> Then get a proper mining rig, will run you around ???600 to ???1000
189 2012-01-21 10:58:26 <nathan7> I wish the market would swing up and down so much
190 2012-01-21 11:00:41 <sh69dow> ok, I have a atom motherboard with 525 cpu
191 2012-01-21 11:00:47 <sh69dow> That good for mining?
192 2012-01-21 11:01:10 <nathan7> No.
193 2012-01-21 11:01:27 <nathan7> Well, you can give it a try and feed the bot the number of Mhash/s
194 2012-01-21 11:01:37 <nathan7> and it'll tell you how many years it'll take you to generate a block
195 2012-01-21 11:01:48 <nathan7> note that this is an average - if you're lucky, you could get a block right away
196 2012-01-21 11:01:54 <nathan7> or get a block after 100 years
197 2012-01-21 13:25:08 <CIA-76> bitcoin: Kano * r0eb5d5ecae36 cgminer/api-example.c: Fix send() for JSON strings http://tinyurl.com/6rp8pl4
198 2012-01-21 13:25:09 <CIA-76> bitcoin: Kano * rb2e15e493e92 cgminer/linux-usb-cgminer: Include HDD install details and related changes http://tinyurl.com/7omxkcv
199 2012-01-21 13:25:11 <CIA-76> bitcoin: Kano * r8e275d9f23e6 cgminer/api.c: Make all CPU code compile time optional - off by default http://tinyurl.com/772woa2
200 2012-01-21 13:25:12 <CIA-76> bitcoin: Con Kolivas * r6ab1f1c39546 cgminer/ (api-example.c api.c linux-usb-cgminer): Merge pull request #80 from kanoi/master http://tinyurl.com/7pocoj5
201 2012-01-21 13:30:11 <CIA-76> bitcoin: Con Kolivas * r0e5b6583f16b cgminer/main.c: qd_lock initialisation got dropped. Fix it. http://tinyurl.com/72cnlro
202 2012-01-21 13:30:13 <CIA-76> bitcoin: Con Kolivas * rdd4597081395 cgminer/ (api-example.c api.c linux-usb-cgminer): Merge branch 'master' of github.com:ckolivas/cgminer http://tinyurl.com/7bcqhdu
203 2012-01-21 13:50:11 <CIA-76> bitcoin: Con Kolivas * r2d5bbc7756d9 cgminer/miner.h: Provide nanosleep() equivalent for windows. http://tinyurl.com/6n633hq
204 2012-01-21 15:50:11 <CIA-76> bitcoin: p2k * rbde6c5196c56 ecoinpool/ (18 files in 3 dirs): Share Logging Framework; Extended Shares Logging http://tinyurl.com/7zcxexy
205 2012-01-21 15:50:12 <CIA-76> bitcoin: p2k * r246b3572ee94 ecoinpool/apps/ecoinpool/src/ (4 files): Share Logging Framework http://tinyurl.com/6rmcub8
206 2012-01-21 16:17:26 <kinlo> is there some kind of UML of the bitcoin source?
207 2012-01-21 16:25:59 <BlueMatt> heh yea right
208 2012-01-21 16:27:34 <sipa> there is doxygen documentation
209 2012-01-21 16:27:40 <kinlo> BlueMatt: it would help :)
210 2012-01-21 16:27:49 <kinlo> sipa: on a public url?
211 2012-01-21 16:28:08 <sipa> i believe so, wumpus maintains (maintained?) it
212 2012-01-21 16:28:26 <kinlo> lxr should be usefull too I guess
213 2012-01-21 16:28:55 <BlueMatt> having better code would also be useful
214 2012-01-21 16:29:35 <sipa> slowly getting there :)
215 2012-01-21 16:29:47 <kinlo> bitcoin is difficult to read :/
216 2012-01-21 16:30:43 <BlueMatt> oh, having comments would also be useful
217 2012-01-21 16:36:28 <diki> 40k blocks left
218 2012-01-21 16:37:15 <roconnor> what happens in 40k blocks?
219 2012-01-21 16:39:39 <sipa> ;;bc,blocks
220 2012-01-21 16:39:40 <gribble> 163241
221 2012-01-21 16:39:56 <sipa> no idea?
222 2012-01-21 16:40:21 <roconnor> oh maybe diki is downloading the blockchain
223 2012-01-21 17:03:59 <diki> no
224 2012-01-21 17:04:05 <diki> I am importing it into mysql
225 2012-01-21 17:04:14 <diki> its going to take a couple of days
226 2012-01-21 17:12:52 <diki> but in 40k from current block count, the block reward is halved
227 2012-01-21 17:17:14 <TuxBlackEdo> ;;bc,halfreward
228 2012-01-21 17:17:15 <gribble> Estimated time of bitcoin block reward halving: Tue Dec 11 02:57:00 2012 | Time remaining: 46 weeks, 2 days, 16 hours, 40 minutes, and 0 seconds
229 2012-01-21 17:17:26 <gribble> (bc,halfreward <an alias, 0 arguments>) -- Alias for "echo Estimated time of bitcoin block reward halving: [ctime [math calc [at [time]] + [bc,sectohalve]]] | Time remaining: [elapsed [bc,sectohalve]]".
230 2012-01-21 17:17:26 <TuxBlackEdo> ;;help bc,halfreward
231 2012-01-21 17:17:40 <gribble> (bc,sectohalve <an alias, 0 arguments>) -- Alias for "math calc (210000 - [bc,blocks])*60*10".
232 2012-01-21 17:17:40 <TuxBlackEdo> ;;help bc,sectohalve
233 2012-01-21 17:17:51 <gribble> 46756
234 2012-01-21 17:17:51 <TuxBlackEdo> ;;calc 210000 - [bc,blocks]
235 2012-01-21 17:18:06 <TuxBlackEdo> 46756 blocks left, diki, not 40k
236 2012-01-21 17:38:56 <diki> TuxBlackEdo:I was not specific. One thing people should know about me, if I write an exat number, no 41,65460856 then that means I always mean approximately
237 2012-01-21 17:39:06 <diki> *exact
238 2012-01-21 17:46:14 <luke-jr> 46756 isn't even approximately 40k
239 2012-01-21 17:47:00 <k9quaint> so we should also not pay attention when you type numbers (in addition to letters)
240 2012-01-21 17:47:12 <k9quaint> what about special and non-printable characters?
241 2012-01-21 18:25:34 <Diablo-D3> http://news.ycombinator.com/item?id=3494323
242 2012-01-21 18:25:40 <Diablo-D3> calling all HN users: time to mod that up
243 2012-01-21 18:26:59 <BlueMatt> what are you even referring to?
244 2012-01-21 18:30:05 <Diablo-D3> bluematt: hrm?
245 2012-01-21 18:30:36 <BlueMatt> I think everyone reading hn who doesnt live under a rock has heard of bitcoin, what is that gonna do?
246 2012-01-21 18:30:57 <Diablo-D3> HN has an ongoing war against bitcoin
247 2012-01-21 18:31:08 <Diablo-D3> but if pro bitcoin commnets get highly moderated, they cant deny it
248 2012-01-21 18:31:56 <k9quaint> Diablo-D3: butterfly labs, my prediction is if the niners win, it is not for real
249 2012-01-21 18:32:06 <k9quaint> if the giants win, those FPGAs are good as gold
250 2012-01-21 18:32:37 <Diablo-D3> who?
251 2012-01-21 18:32:40 <Diablo-D3> are those sports teams?
252 2012-01-21 18:32:41 <BlueMatt> heh, conditional predictions
253 2012-01-21 18:32:50 <BlueMatt> Diablo-D3: hes referring to the super bowl...
254 2012-01-21 18:32:52 <BlueMatt> (football)
255 2012-01-21 18:32:52 <Diablo-D3> k9quaint: I dont care if its real or not
256 2012-01-21 18:32:56 <BlueMatt> (american football)
257 2012-01-21 18:33:07 <Diablo-D3> I just need to kick people periodically to keep the flame wars down
258 2012-01-21 18:33:13 <Diablo-D3> as in, with my foot
259 2012-01-21 18:34:36 <k9quaint> Diablo-D3: I think goat likes you
260 2012-01-21 19:05:26 <hexx> Hi guys, does anyone have any advices for developing exchange? I plan to write one in .NET, I will make it open source.
261 2012-01-21 19:05:44 <BlueMatt> dont use floats
262 2012-01-21 19:06:39 <hexx> yea I know that
263 2012-01-21 19:06:48 <hexx> But I havent desided what to use
264 2012-01-21 19:06:54 <hexx> decimal or int/long ?
265 2012-01-21 19:07:06 <hexx> any pros/cons ?
266 2012-01-21 19:07:18 <BlueMatt> Id go int, but meh
267 2012-01-21 19:07:25 <BlueMatt> s/int/long/
268 2012-01-21 19:08:39 <hexx> is that how bitcoin internaly works ?
269 2012-01-21 19:10:58 <hexx> also does anyone have any security advices ?  I plan to have offline wallet. Also strict DB security. All interaction over stored procedures.
270 2012-01-21 19:15:15 <sipa> hexx: bitcoin internally uses 64-bit integer multiples of 0.00000001 BTC
271 2012-01-21 19:17:47 <Diablo-D3> k9quaint: meh, goats very lucky he isnt banned
272 2012-01-21 19:18:57 <hexx> then I will use that
273 2012-01-21 19:19:00 <sipa> hexx: note that nothing prevents your exchange from doing calculations and keeping balances at higher accuracy
274 2012-01-21 19:19:19 <Diablo-D3> SIPA lawlz
275 2012-01-21 19:21:10 <hexx> Why would I need that ?
276 2012-01-21 19:21:36 <hexx> that is allready much more then enough precision
277 2012-01-21 19:25:39 <sipa> rounding errors will be in the order of 0.00000001 BTC then, which is visible
278 2012-01-21 19:27:13 <TuxBlackEdo> ;;calc 1024/20
279 2012-01-21 19:27:14 <gribble> 51.2
280 2012-01-21 19:27:19 <gribble> 1.536
281 2012-01-21 19:27:19 <TuxBlackEdo> ;;calc 1024/20*0.03
282 2012-01-21 19:28:16 <gribble> 0.15872
283 2012-01-21 19:28:16 <TuxBlackEdo> ;;calc 1024/20*0.0031
284 2012-01-21 19:28:26 <gribble> 162.52928
285 2012-01-21 19:28:26 <TuxBlackEdo> ;;calc 0.15872*1024
286 2012-01-21 19:28:38 <TuxBlackEdo> 162.52928/0.03
287 2012-01-21 19:28:42 <gribble> 5417.64266667
288 2012-01-21 19:28:42 <TuxBlackEdo> ;;calc 162.52928/0.03
289 2012-01-21 19:29:12 <gribble> 0.00312771
290 2012-01-21 19:29:12 <TuxBlackEdo> ;;calc [ticker --last]*0.00050001
291 2012-01-21 19:29:35 <gribble> 0.0256
292 2012-01-21 19:29:35 <TuxBlackEdo> ;;calc (1024/20)*0.00050001
293 2012-01-21 19:29:46 <gribble> 26.2144
294 2012-01-21 19:29:46 <TuxBlackEdo> ;;calc 0.0256*1024
295 2012-01-21 19:30:10 <hexx> And does anybody know what could be possible attacking vectors beside SQL injection and XSS ?
296 2012-01-21 19:43:33 <BlueMatt> anyone want to guess as to the most common lock which blocks a bitcoin thread while downloading blocks?
297 2012-01-21 19:43:46 <BlueMatt> s/as to/what/
298 2012-01-21 19:43:57 <BlueMatt> s/?/is ?/
299 2012-01-21 19:45:31 <sipa> tell me
300 2012-01-21 19:46:00 <sipa> most common as in most often locking, or as in being most of the time locked?
301 2012-01-21 19:46:28 <BlueMatt> most common lock which has to be waited for
302 2012-01-21 19:46:54 <BlueMatt> cs_wallet if you have txes in the blocks or cs_vNodes
303 2012-01-21 19:48:39 <sipa> vNodes... that shouldn't happen
304 2012-01-21 19:48:49 <sipa> is this in master or in cblockstore?
305 2012-01-21 19:48:53 <BlueMatt> master
306 2012-01-21 19:49:05 <BlueMatt> not blocking block download, but its the only lock that really ever gets waited on
307 2012-01-21 19:49:36 <BlueMatt> its master + https://github.com/TheBlueMatt/bitcoin/commit/938ab33aaeaaac50c4d33e39842f4bef49540875
308 2012-01-21 20:10:19 <CIA-76> bitcoin: jedi95 * r8263f658c835 Phoenix-Miner/ (7 files in 5 dirs): Add X-Work-Identifier support to RPC for better compatibility with P2Pool. Tweaked kernel worksize validation. Bump version to 1.7.4 http://tinyurl.com/7yot6kt
309 2012-01-21 20:22:08 <gribble> New news from bitcoinrss: TheBlueMatt opened pull request 775 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/775>
310 2012-01-21 21:31:27 <diki> damn
311 2012-01-21 21:31:31 <diki> I need to rebuild the db now...
312 2012-01-21 21:31:33 <diki> cool....
313 2012-01-21 21:31:53 <diki> It took me over 2 days to get to blk 130k, now I have to do it all over again
314 2012-01-21 21:36:35 <Joric> i'm playing with multibit (multibit.org) how come the whole bitcoinj blockchain is 17 megs and downloading in 30 seconds?
315 2012-01-21 21:36:52 <sipa> because it is not the entire block chain
316 2012-01-21 21:36:55 <sipa> only headers
317 2012-01-21 21:37:17 <Joric> how do they send transactions then if they don't know inputs
318 2012-01-21 21:37:27 <sipa> you always know your own inputs
319 2012-01-21 21:37:42 <sipa> but you can't verify anything happening on the network
320 2012-01-21 21:38:00 <sipa> only trust that the chain with the highest work amount is legitimate
321 2012-01-21 21:39:04 <diki> which is rather unlikely to not be bitcoin so...
322 2012-01-21 21:40:11 <CIA-76> bitcoin: Con Kolivas * re15d57d729c8 cgminer/ (README main.c miner.h util.c): Implement socks4 proxy support. http://tinyurl.com/7rz8n7q
323 2012-01-21 21:40:44 <Joric> why they even need blocks then
324 2012-01-21 21:41:02 <sipa> because even bitcoinj downloads blocks
325 2012-01-21 21:41:08 <sipa> to know payments to yourself
326 2012-01-21 21:41:14 <sipa> it just doesn't store them
327 2012-01-21 21:42:27 <roconnor> where can I find amoury?
328 2012-01-21 21:42:32 <roconnor> etotheipi_: ^^
329 2012-01-21 21:42:42 <sipa> roconnor: see the forum
330 2012-01-21 21:42:57 <roconnor> the forum? >_>
331 2012-01-21 21:43:03 <sipa> bitcointalk.org
332 2012-01-21 21:43:15 <sipa> in the dev section, under alternative clients
333 2012-01-21 21:44:40 <diki> slush:I sent you a PM, but since the PM system is a bit wonky, I will ask you here as well. Can you upload a mysql dump of Abe, because I got the duplicate entry error, and the torrent has no seeds
334 2012-01-21 21:45:58 <slush> diki: torrent has the seed again, sorry, mistake on my side
335 2012-01-21 21:46:43 <diki> thanks :)
336 2012-01-21 21:47:15 <slush> hm, but there was one seed on my webserver running all the time
337 2012-01-21 21:47:36 <sipa> a seed "running"?
338 2012-01-21 21:48:49 <diki> just a quickie. When I import the dump, will abe be able to continue, or will it ask for a rebuild?
339 2012-01-21 21:49:34 <slush> diki: read the abe forum thread, there's some variable which need to be reset
340 2012-01-21 21:49:54 <slush> then you can start abe and it will continue without any hacking
341 2012-01-21 21:50:27 <slush> well, there will be one more seed in few minutes, downloading will be much faster then
342 2012-01-21 21:51:43 <slush> diki: how many seeds do you see?
343 2012-01-21 21:51:59 <etotheipi_> roconnor, http://bitcoinarmory.com/
344 2012-01-21 21:52:55 <diki> slush:1(1)
345 2012-01-21 21:53:08 <diki> ok speed jumped to 200 kilobytes/s
346 2012-01-21 21:54:49 <slush> diki: if you can, add manual seed from palatinus.cz:10001 . For some reason this client cannot connect to tracker.
347 2012-01-21 21:55:29 <BlueMatt> etotheipi_: does armory use satoshi client code or all from the ground up (or another client?)
348 2012-01-21 21:56:55 <diki> slush:added, however still no dice
349 2012-01-21 21:58:14 <etotheipi_> BlueMatt, it's all from the ground up
350 2012-01-21 21:58:23 <etotheipi_> it only uses the Satoshi client as an interface to the BTC network
351 2012-01-21 21:58:36 <sipa> why do you still need that?
352 2012-01-21 21:59:21 <gmaxwell> sipa: because it only implements a fraction of the validation logic.
353 2012-01-21 21:59:52 <sipa> well, as long as it only trusts the best chain, it's no less secure than a lightweight client
354 2012-01-21 22:00:33 <etotheipi_> though my top priority after Armory-alpha, is "normal" RAM req'ts and independent networking
355 2012-01-21 22:02:32 <Joric> does it use boost? )
356 2012-01-21 22:04:31 <BlueMatt> etotheipi_: the interface with the network is what, just the net code or the net code and block/tx verification or what?
357 2012-01-21 22:05:05 <etotheipi_> I use the Satoshi client to get zero-conf tx, and broadcast transactions
358 2012-01-21 22:05:35 <etotheipi_> Armory is just a very dumb node/peer on localhost with only those two functions... it dramatically simplifies the networking :)
359 2012-01-21 22:07:17 <BlueMatt> mmm
360 2012-01-21 22:10:14 <jjjrmy> "Runtime Error! Program C:Program Files (x86)Bitcoinitcoin-qt.exe This application has requested the Runtime to terminate it in a unusual way."
361 2012-01-21 22:10:16 <jjjrmy> please help
362 2012-01-21 22:10:19 <jjjrmy> "Runtime Error! Program C:Program Files (x86)Bitcoinitcoin-qt.exe This application has requested the Runtime to terminate it in a unusual way."
363 2012-01-21 22:16:51 <BlueMatt> last line in debug.log?
364 2012-01-21 22:18:29 <diki> slush:Don't know what you did, but the speed upped to 1.1mb/s
365 2012-01-21 22:18:42 <diki> and rising
366 2012-01-21 22:18:58 <slush> diki: you finally connected to my server :)
367 2012-01-21 22:24:13 <diki> ok got it
368 2012-01-21 22:24:28 <slush> can you seed it for some time, please?
369 2012-01-21 22:25:43 <diki> sure
370 2012-01-21 22:26:13 <diki> I've seeded torrents for less than 10 seconds after download time
371 2012-01-21 22:26:20 <diki> and i've seeded torrents for well over a year
372 2012-01-21 22:35:13 <roconnor> ah crap, the transaction that my rogue relayer relayed; turns out I mined it myself :(
373 2012-01-21 22:35:33 <BlueMatt> heh
374 2012-01-21 22:35:39 <roconnor> I should have noticed
375 2012-01-21 22:35:54 <BlueMatt> does your miner mine anything the regular client wouldnt have?
376 2012-01-21 22:36:05 <roconnor> yep
377 2012-01-21 22:36:10 <BlueMatt> how so?
378 2012-01-21 22:36:23 <roconnor> my rogue relayer is simply a modified standard client
379 2012-01-21 22:36:27 <BlueMatt> IsStandard() { return true; }
380 2012-01-21 22:36:36 <roconnor> on testnet, IsStandard is ignored
381 2012-01-21 22:36:52 <sipa> on testnet, IsStandard() always returned true already, no?
382 2012-01-21 22:36:58 <BlueMatt> I was asking what modifications you made?
383 2012-01-21 22:37:07 <BlueMatt> ie do you just ignore isstandard, or what?
384 2012-01-21 22:37:36 <BlueMatt> sipa: I know IsStandard isnt called on testnet, dont know what it returns though (havent looked at that code in a long time)
385 2012-01-21 22:37:43 <roconnor> I've modified CTransaction::AcceptToMemoryPool
386 2012-01-21 22:37:51 <BlueMatt> roconnor: with what?
387 2012-01-21 22:37:53 <roconnor> for (int i = 0; i < vin.size(); i++)
388 2012-01-21 22:37:55 <roconnor> {
389 2012-01-21 22:37:56 <roconnor> const char* latin = "Stack is innocent";
390 2012-01-21 22:37:58 <roconnor> vin[i].scriptSig = CScript(vector<unsigned char>((const unsigned char*)latin, (const unsigned char*)latin + strlen(latin))) + vin[i].scriptSig;
391 2012-01-21 22:37:59 <roconnor> }
392 2012-01-21 22:38:01 <roconnor> is added
393 2012-01-21 22:38:34 <Eliel> :D
394 2012-01-21 22:38:43 <BlueMatt> so in other words it doesnt mine anything that the regular client wouldnt
395 2012-01-21 22:38:50 <BlueMatt> (at least in terms of checks)
396 2012-01-21 22:39:08 <roconnor> BlueMatt: ya sorry I wasn't clear due to the negative of your question.
397 2012-01-21 22:39:09 <BlueMatt> is that addition before or after all the checks?
398 2012-01-21 22:39:18 <a_meteorite> That Bitcoin Armory client looks interesting
399 2012-01-21 22:39:22 <gmaxwell> well, it does his rogue relayer may not be successful without his own mining simply because it'll be late.
400 2012-01-21 22:39:32 <a_meteorite> etotheipi_: are you the developer on it?
401 2012-01-21 22:39:47 <BlueMatt> gmaxwell: unless you are ie the first one who gets it
402 2012-01-21 22:39:52 <roconnor> BlueMatt: it is before most checks, but after the non-excecuted IsStandard and after not sigop_counting
403 2012-01-21 22:40:04 <gmaxwell> BlueMatt: or something like that.
404 2012-01-21 22:40:08 <Eliel> a_meteorite: he is.
405 2012-01-21 22:40:19 <BlueMatt> wait, you did this on testnet, I thought it was on mainnet?
406 2012-01-21 22:40:25 <gmaxwell> (you don't have to be _first_ strictly)
407 2012-01-21 22:40:35 <BlueMatt> gmaxwell: well yea, just early
408 2012-01-21 22:40:38 <BlueMatt> (and better peered)
409 2012-01-21 22:40:52 <roconnor> BlueMatt: I'm told messing around on mainnet is frowned upon
410 2012-01-21 22:40:56 <a_meteorite> Do you know if anyone else works (or uses) on it? Using an alternative client does make me a bit concerned with security
411 2012-01-21 22:41:08 <roconnor> a_meteorite: it is in beta, or pre-beta
412 2012-01-21 22:41:09 <BlueMatt> roconnor: no, better that you did on testnet, I just thought it was main
413 2012-01-21 22:41:37 <BlueMatt> (the original satoshi client is also in beta for the record...)
414 2012-01-21 22:41:52 <sipa> a_meteorite: armory depends on the satoshi client
415 2012-01-21 22:42:01 <sipa> for now
416 2012-01-21 22:42:19 <a_meteorite> sipa: so it's more of a frontend?
417 2012-01-21 22:42:23 <sipa> no
418 2012-01-21 22:42:35 <sipa> but it connects with the network through bitcoind
419 2012-01-21 22:42:40 <a_meteorite> ah
420 2012-01-21 22:42:48 <BlueMatt> its a full client, but it puts a satoshi client in front of itself
421 2012-01-21 22:42:58 <BlueMatt> (well full -networking)
422 2012-01-21 22:44:45 <a_meteorite> heh, full blockchain in RAM
423 2012-01-21 22:44:49 <Joric> it needs original client running side by side
424 2012-01-21 22:44:50 <a_meteorite> glad I got 8GB :)
425 2012-01-21 22:44:56 <Joric> just tried to launch it
426 2012-01-21 22:45:13 <BlueMatt> (and often multiple datadirs)
427 2012-01-21 22:47:19 <etotheipi_> a_meteorite, I'm the Armory developer
428 2012-01-21 22:48:08 <roconnor> does Armory really hold everything in memory?
429 2012-01-21 22:48:15 <etotheipi_> yeah, there's some caveats with armory, which I hope to remove when I go from alpha to beta... for now I decided that since the functionality works and is useful, I should make it avail to users that can run it
430 2012-01-21 22:48:17 <roconnor> I mean the whole blockchain
431 2012-01-21 22:48:26 <etotheipi_> roconnor, it does hold the whole blockchain in ram
432 2012-01-21 22:48:28 <roconnor> ie, should I turn on swap?
433 2012-01-21 22:48:47 <etotheipi_> which is why it's so damned fast:  I can scan a fresh set of addresses in 0.75s to get balances and unspent outputs
434 2012-01-21 22:48:48 <Joric> 'loading blockchain please wait'
435 2012-01-21 22:48:55 <Joric> i only have 2gb
436 2012-01-21 22:49:01 <a_meteorite> the sweeping and offline stuff looks killer
437 2012-01-21 22:49:10 <roconnor> oh wait, I just want to run it on testnet
438 2012-01-21 22:49:11 <etotheipi_> 2 GB is a stretch...
439 2012-01-21 22:49:14 <roconnor> so that is fine
440 2012-01-21 22:49:22 <etotheipi_> testnet can be run on any system
441 2012-01-21 22:49:26 <a_meteorite> I just get a slight antsy with my wallet in the hands of less-tested code
442 2012-01-21 22:49:47 <etotheipi_> a_meteorite, that's why it's not even alpha yet... though I've managed a lot of transactions so far and there's not a hint of a problem
443 2012-01-21 22:49:59 <a_meteorite> cool
444 2012-01-21 22:50:01 <etotheipi_> thoroughly tested myself, but I'm waiting for more users to test before I release alpha
445 2012-01-21 22:50:19 <Eliel> a_meteorite: you might want to use a new wallet and do remember to keep a backup if you put bitcoins in there :)
446 2012-01-21 22:50:28 <a_meteorite> I can test on OS X and Windows (don't use Linux much on desktops)
447 2012-01-21 22:50:41 <etotheipi_> however, if you use it on an offline computer, it doesn't load any blockchain so there's virtually no mem reqts
448 2012-01-21 22:50:42 <a_meteorite> though my wallet is on my Windows partition, would have to see about moving it over to OS X
449 2012-01-21 22:51:04 <etotheipi_> I actually just adopted an old system with 512 MB of RAM just for keeping an offline wallet
450 2012-01-21 22:51:33 <etotheipi_> a_meteorite, I currently don't know a thing about OSX or developing/compiling for it... I actually really need someone to help
451 2012-01-21 22:51:36 <a_meteorite> That offline bit is just awesome
452 2012-01-21 22:52:09 <etotheipi_> the offline wallets were my holy grail feature that inspired the whole project
453 2012-01-21 22:52:12 <a_meteorite> etotheipi_: I could help, but I'd need a little shove in the direction of moving my (pretty outdated) blockchain and wallet from Windows to OS X
454 2012-01-21 22:52:28 <a_meteorite> Though I'm set-up for compiling and can give it a go
455 2012-01-21 22:52:30 <etotheipi_> a_meteorite, I recommend not moving anything yet
456 2012-01-21 22:52:46 <etotheipi_> make a new wallet, print a paper backup, and transfer some coins in
457 2012-01-21 22:52:46 <sipa> etotheipi_: so, we could make you pretty pissed by adding that to the satoshi client? ;)
458 2012-01-21 22:52:50 <a_meteorite> k
459 2012-01-21 22:53:10 <roconnor> a_meteorite: print on acid-free thesis paper
460 2012-01-21 22:53:28 <roconnor> a_meteorite: and while you are at it, put it in your thesis, that way no one will ever find it.
461 2012-01-21 22:54:00 <a_meteorite> Just got some acid free heavy (not cardstock) paper with a laser printer
462 2012-01-21 22:54:00 <etotheipi_> sipa, I wouldn't be pissed... if anything I'd be glad to see the Satoshi client implement some of this stuff since it will maintain "monopoly" in the BTC world, I'd like to think I influenced the addition by adding competition :)
463 2012-01-21 22:54:28 <a_meteorite> I will be in a few years, but only Master's
464 2012-01-21 22:54:28 <sipa> roconnor: those that are in #haskell as well may be :)
465 2012-01-21 22:54:42 <etotheipi_> paper backups are also a holy-grail feature... I *hate* backing up to USB keys, because I never know if they're going to work when I plug them in 6 months later... so I have to keep multiple keys
466 2012-01-21 22:54:44 <a_meteorite> I really don't want a CS PhD :)
467 2012-01-21 22:55:00 <a_meteorite> etotheipi_: put in a bank vault, and bam!
468 2012-01-21 22:55:00 <Eliel> roconnor: yeah, bitcoin is very much not an academic only project :)
469 2012-01-21 22:55:01 <etotheipi_> at least you can visually verify the paper backup
470 2012-01-21 22:55:28 <etotheipi_> plus paper is cheaper than USB keys (though they're both pretty cheap :))
471 2012-01-21 22:55:35 <roconnor> what is sweeping?
472 2012-01-21 22:55:48 <sipa> roconnor: import + send to new address
473 2012-01-21 22:55:59 <roconnor> what is import?
474 2012-01-21 22:56:11 <sipa> add private key, and transactions spending to it
475 2012-01-21 22:56:15 <etotheipi_> sweeping scans the blockchain, finds all the unspent outputs, and then sends them to a new address
476 2012-01-21 22:56:36 <etotheipi_> importing actually adopts the private key to you wallet so you can continue to use it
477 2012-01-21 22:56:55 <etotheipi_> there was an interesting discussion about this on the forums, because I didn't realize at first the security risks involved with importing
478 2012-01-21 22:56:58 <roconnor> etotheipi_: how does address generation work? random, or the generated thing?
479 2012-01-21 22:57:04 <etotheipi_> deterministic
480 2012-01-21 22:57:08 <roconnor> :)
481 2012-01-21 22:57:15 <a_meteorite> so basically taking funds from another wallet? does the imported private key "control" those "old" bitcoins still?
482 2012-01-21 22:57:27 <sipa> a_meteorite: no
483 2012-01-21 22:57:38 <sipa> they've moved on
484 2012-01-21 22:57:40 <a_meteorite> Ah
485 2012-01-21 22:57:50 <etotheipi_> the security risk is this... someone gives you a private key with funds as a way of payment.  But you import not sweep
486 2012-01-21 22:58:09 <etotheipi_> then later, they say hey, I'll buy somethign from you for 100 BTC... they send the 100 BTC to that address which they actually have the private key for
487 2012-01-21 22:58:24 <etotheipi_> but it shows up in your wallet as your own
488 2012-01-21 22:58:51 <sipa> payment via private key exchange bypasses the trust issue that bitcoin tries to solve
489 2012-01-21 22:59:07 <etotheipi_> sipa, I don't approve of it, but there's lots of reasons
490 2012-01-21 22:59:12 <a_meteorite> Yeah I see it more of a way as migrating old wallets to one wallet
491 2012-01-21 22:59:13 <etotheipi_> how about casascius bitcoins?
492 2012-01-21 22:59:17 <sipa> sure it is useful
493 2012-01-21 22:59:23 <etotheipi_> someone else obviously has the private key
494 2012-01-21 22:59:51 <etotheipi_> I guess the point is... regardless of how you got it... if you aren't sure whether someone else has the private key, just sweep it
495 2012-01-21 23:00:03 <etotheipi_> untrusted keys:  sweep,  trusted keys: import
496 2012-01-21 23:00:29 <sipa> no
497 2012-01-21 23:00:42 <sipa> untrusted keys: don't use   trusted keys: sweep   own keys: import
498 2012-01-21 23:01:05 <etotheipi_> sipa, you're using a different definition for these things
499 2012-01-21 23:01:12 <Eliel> why no sweeping of untrusted keys?
500 2012-01-21 23:01:42 <sipa> Eliel: i rather mean, you shouldn't trust a private key as a means of payment
501 2012-01-21 23:01:43 <etotheipi_> whatever definition you use.... a key that you know for sure no one else has:  importing is safe
502 2012-01-21 23:01:51 <sipa> if the sweep works, you're fine
503 2012-01-21 23:02:05 <sipa> but you have no way of knowing in advance that the funds are there, or the key is valid
504 2012-01-21 23:02:11 <etotheipi_> if there's ANY chance that someone else saw the private key, just sweep it (which throws it away)
505 2012-01-21 23:02:40 <Eliel> sipa: of course, private keys aren't to be trusted until the sweep is complete.
506 2012-01-21 23:03:31 <sipa> case in point: i may trust a payment via casascius coin because i trust casascius (and i can check for physical integrity)
507 2012-01-21 23:03:52 <sipa> i will not trust a payment via a paper with a private key on it, unless it is from a friend
508 2012-01-21 23:04:09 <etotheipi_> sipa... the sweep happens immediately... Armory scans the blockchain for unspent outputs, jams them all into a tx to yourself, broadcasts, then throws it away
509 2012-01-21 23:04:19 <diki> what is the max size in bytes of an address?
510 2012-01-21 23:04:30 <diki> bitcoin address
511 2012-01-21 23:04:33 <etotheipi_> I don't recommend anyone accept private keys if they can't check the blockchain and sweep right away
512 2012-01-21 23:04:39 <sipa> etotheipi_: "immediately" is only as soon as it reaches the digital world (your program)
513 2012-01-21 23:04:50 <sipa> we're saying the same thing :)
514 2012-01-21 23:04:53 <etotheipi_> sipa, we're in agreement
515 2012-01-21 23:04:55 <etotheipi_> :)
516 2012-01-21 23:06:07 <etotheipi_> I'm very close to my zero-confirmation-tx integration.... so I might get alpha out soon!
517 2012-01-21 23:06:28 <etotheipi_> I stupidly forgot to consider zero-conf txs when I built the original blockdata-manager
518 2012-01-21 23:06:45 <etotheipi_> I guess my goal at the time was more for doing statistical tests on the blockchain... not building a client
519 2012-01-21 23:07:33 <etotheipi_> and swift-geek helped me setup a cmake build system... so it might become a lot easier to support all different platforms
520 2012-01-21 23:07:36 <Joric> i checked the blockchain against 1.7m of most used passwords - the only private key i found is sha256('fuckyou') with 0.25 btc on it )
521 2012-01-21 23:08:07 <a_meteorite> Hah
522 2012-01-21 23:08:21 <a_meteorite> I didn't know you could make "designer" private keys or such
523 2012-01-21 23:09:18 <etotheipi_> btw, for most people here, if you fire up Armory, I recommend switchign to "Developer" mode
524 2012-01-21 23:09:43 <etotheipi_> I'll be adding some extra tools to that mode, soon, but for now it gives you some extra stuff
525 2012-01-21 23:09:55 <diki> Joric:you did what??
526 2012-01-21 23:10:08 <sipa> nanotube: i modified my bitcoin-seeder program to spit out a list of IPs with version number and uptime stats
527 2012-01-21 23:10:20 <sipa> nanotube: want to compare with what you're using for the seednode list?
528 2012-01-21 23:10:26 <a_meteorite> Bwahahaha. I'm using an old wallet with 0.05 BTC (from that one site that was giving them away) and the blockchain is 408 days old.
529 2012-01-21 23:10:51 <sipa> a_meteorite: ... that means...
530 2012-01-21 23:10:53 <diki> Joric:so you tried to what...find a collision or something?
531 2012-01-21 23:10:59 <sipa> you started using bitcoin before me :(
532 2012-01-21 23:11:02 <a_meteorite> etotheipi_: Guess you'll get to see how well Armory does with an old wallet once I eventually catch up
533 2012-01-21 23:11:09 <user__> If armory go to stable and be better than satoshi client. is it possible be the 'official' client?
534 2012-01-21 23:11:11 <a_meteorite> sipa: Oh, I'm before that even
535 2012-01-21 23:11:12 <Joric> diki, it's called deterministic keys
536 2012-01-21 23:11:28 <sipa> user__: there is no "official" client really
537 2012-01-21 23:11:36 <etotheipi_> a_meteorite, there is no wallet import for now
538 2012-01-21 23:11:40 <Eliel> user__: the whole point of bitcoin is that there is no "official" client.
539 2012-01-21 23:11:48 <a_meteorite> etotheipi_: Oh...
540 2012-01-21 23:11:55 <diki> Joric:you compared it to 1.7 what exactly?
541 2012-01-21 23:12:04 <user__> ok
542 2012-01-21 23:12:08 <diki> addresses in their final form?
543 2012-01-21 23:12:10 <a_meteorite> So I really need a fresh Bitcoin directory then?
544 2012-01-21 23:12:10 <Joric> the whole point of bitcoin is that there is "no" official client
545 2012-01-21 23:12:17 <gmaxwell> Eliel: I prefer to use the phrase "reference client", which I think is accurate.
546 2012-01-21 23:12:22 <a_meteorite> Or just no wallet.dat?
547 2012-01-21 23:12:22 <etotheipi_> a_meteorite, the issue is that it's stupid-hard to get private keys out of Satoshi wallet
548 2012-01-21 23:12:23 <Eliel> user__: there can be a de facto client, though.
549 2012-01-21 23:12:34 <etotheipi_> a_meteorite, each wallet is stored in a different file in your .armory directory
550 2012-01-21 23:12:41 <sipa> etotheipi_: 0.6 will have key export
551 2012-01-21 23:12:49 <Eliel> gmaxwell: yes, reference client is a good way to put it
552 2012-01-21 23:12:59 <etotheipi_> Armory uses the Satoshi blockchain (blk0001.dat), but nothing else from it
553 2012-01-21 23:13:10 <nanotube> sipa: sure, we could check to see if my seednode list has an intersection with your list?
554 2012-01-21 23:13:14 <etotheipi_> sipa, I guess that means I should support bulk-key import, soon
555 2012-01-21 23:13:21 <diki> etotheipi_:it does support...mining, right?
556 2012-01-21 23:13:24 <a_meteorite> etotheipi_: importing wallets sure would be nice...
557 2012-01-21 23:13:28 <diki> as in it can go into server mode
558 2012-01-21 23:13:31 <etotheipi_> but I won't be doing it right away...
559 2012-01-21 23:14:04 <etotheipi_> diki, I don't understand
560 2012-01-21 23:14:19 <etotheipi_> I have no intention to do anything mining related with Armory
561 2012-01-21 23:14:30 <diki> it doesnt support rpc commands?
562 2012-01-21 23:14:33 <diki> such as getwork?
563 2012-01-21 23:14:33 <Joric> diki, also there was a key made from "Satoshi Nakamoto" with 0.00000001 on it but someone wiped it
564 2012-01-21 23:15:29 <etotheipi_> diki, Armory would make a *damned good* server... given that it can do ridiculously fast blockchain scans
565 2012-01-21 23:15:36 <etotheipi_> but it's not implemented in anyway
566 2012-01-21 23:16:21 <etotheipi_> one day I want to allow some kind of socket-interface to Armory... but for now I have other priorities
567 2012-01-21 23:16:24 <gmaxwell> etotheipi_: Don't brag about how fast you are when you don't actually implement validation. :)
568 2012-01-21 23:16:38 <a_meteorite> etotheipi_: Can I catch up my blockchain then and just nuke everything but blk0001.dat?
569 2012-01-21 23:16:48 <a_meteorite> Or do I have to get a blockchain with no wallet or anything?
570 2012-01-21 23:16:49 <etotheipi_> gmaxwell, that's mostly fair...
571 2012-01-21 23:18:24 <etotheipi_> a_meteorite, the only thing Armory needs is the blk0001.dat file to be in the default bitcoin directory
572 2012-01-21 23:18:31 <a_meteorite> k
573 2012-01-21 23:18:50 <etotheipi_> but Armory can't run (yet) without the satoshi client also running
574 2012-01-21 23:19:09 <etotheipi_> it won't even pick up new blockdata without it out (it watches the blk0001.dat for updates)
575 2012-01-21 23:19:53 <a_meteorite> etotheipi_: thought it depended on it?
576 2012-01-21 23:20:11 <a_meteorite> Oh
577 2012-01-21 23:20:14 <a_meteorite> I misread you
578 2012-01-21 23:23:00 <etotheipi_> okay, I'm going undercover for a while... I gotta continue testing the new zero-conf code
579 2012-01-21 23:23:42 <etotheipi_> seco, thanks :)
580 2012-01-21 23:23:56 <seco> good luck, you can be shure i will have a criticism view on it :) - even if it does not mean anything hehe
581 2012-01-21 23:24:18 <etotheipi_> all comments/suggestions/complaints/donations are welcome!  :)
582 2012-01-21 23:24:30 <Joric> etotheipi_, how do you parse bdb? i assume you're loading the blockchain as is?
583 2012-01-21 23:24:39 <etotheipi_> the blockchain is not bdb
584 2012-01-21 23:24:51 <Joric> oh
585 2012-01-21 23:24:54 <sipa> wallet.dat, addr.dat and blkindex.dat are bdb
586 2012-01-21 23:25:05 <sipa> blk0001.dat is just a binary concatenation of blocks
587 2012-01-21 23:25:07 <etotheipi_> it's a flat binary file with just [magic | numBlkBytes | header | numTx | Tx1 | Tx2 |... ]
588 2012-01-21 23:25:39 <etotheipi_> yeah, I haven't touched bsddb, which is why I can't do any kind of wallet import
589 2012-01-21 23:25:52 <etotheipi_> but I didn't want to support that yet, anyway... at least not until the client gets some more testing
590 2012-01-21 23:27:49 <seco> at least the last suggestion is done right now :)
591 2012-01-21 23:28:16 <etotheipi_> oh wow, thanks seco!
592 2012-01-21 23:28:21 <seco> the first 2 i will come in the next time to xD
593 2012-01-21 23:28:40 <seco> another reas why i love bitcoin: you can express your likes very fast :)
594 2012-01-21 23:28:43 <seco> reason*
595 2012-01-21 23:28:49 <etotheipi_> please do... I'm very open to recommendations
596 2012-01-21 23:29:11 <seco> :)
597 2012-01-21 23:29:13 <etotheipi_> though my feature list is already based on months of browsing the forums...
598 2012-01-21 23:29:29 <etotheipi_> so I expect mostly bug reports right now :)
599 2012-01-21 23:31:14 <etotheipi_> what does "see all addresses, including change" mean?
600 2012-01-21 23:31:51 <etotheipi_> I do have the full input/output list in the transaction view if you are in Advanced or Developer mode
601 2012-01-21 23:33:14 <seco> just have a look on his video linked; what i miss from the 1st day at bitcoin is the ability to choose which addresses to use from the wallets current coinbase, when you send some out to a new address
602 2012-01-21 23:35:31 <etotheipi_> btw, Armory codebase does have a SelectCoins algorithm that choose optimization based on minimal-tx-fee or anonymity
603 2012-01-21 23:36:03 <seco> better flowcontrol... but in case you dont have it in your featureplans, just add it somewhere; every man knows only patience leads to success :-)
604 2012-01-21 23:36:41 <etotheipi_> so if you don't mind paying transaction fees, it will find SelectCoins solutions that (1) minimize number of input addrs (2) maximize uniformity of outputs and (3) looks at the precision of the change and recip outputs
605 2012-01-21 23:37:20 <etotheipi_> for instance... if your outputs looks like (1.0 and 0.183922), it's pretty obvious which is change and which is recip)
606 2012-01-21 23:38:33 <roconnor> etotheipi_: which one is which?
607 2012-01-21 23:38:33 <seco> and somewhere in future, someone enales it to let the user decide which input addr to use manually for the outgoing addr *g*
608 2012-01-21 23:38:40 <seco> enables*
609 2012-01-21 23:38:41 <etotheipi_> however, if the SelectCoins algorithm finds a solution that is Chg: 1.0 and recip:0.183922... major bonus extra points for anonymity
610 2012-01-21 23:39:00 <seco> :)
611 2012-01-21 23:39:08 <etotheipi_> roconnor, usually you send someone 1.0 BTC, and then the changei s whatever is leftover, which may be something stupid like 0.183722
612 2012-01-21 23:39:13 <roconnor> okay
613 2012-01-21 23:39:31 <TuxBlackEdo> how does the bitcoin scripts work?
614 2012-01-21 23:39:32 <etotheipi_> so part of the optimization takes into account the number of trailing zeros in the chg and recip parts
615 2012-01-21 23:39:47 <TuxBlackEdo> is there a place on the wiki that says what all the OP_ codes do?
616 2012-01-21 23:40:03 <sipa> TuxBlackEdo: yes, the Script page :)
617 2012-01-21 23:40:29 <etotheipi_> if the change output actually has less trailing zeros than the recip output, then it gets "extra points" in the optimization for being deceptive
618 2012-01-21 23:40:44 <sipa> haha, nice one
619 2012-01-21 23:41:09 <roconnor> etotheipi_: how did you build the amoury website?
620 2012-01-21 23:41:14 <roconnor> armory
621 2012-01-21 23:41:25 <etotheipi_> roconnor, a friend helped me... set it up with WordPress
622 2012-01-21 23:41:39 <seco> roconnor,  <meta name="generator" content="WordPress 3.3.1" />
623 2012-01-21 23:41:47 <etotheipi_> I'm really impressed with WordPress
624 2012-01-21 23:42:45 <etotheipi_> I was dreading doing the website, but it's actually quite pleasant
625 2012-01-21 23:43:07 <gmaxwell> sipa: I had a patch for bitcoin that ran post selection and added inputs if doing to reduced the priority but kept it over the threshold and made the change more round.
626 2012-01-21 23:43:29 <gmaxwell> sipa: but I couldn't figure out how to keep it from using up too many of your inputs. e.g. making it less likely that you'll get exact matches.
627 2012-01-21 23:43:45 <roconnor> etotheipi_: no openssl dependency?
628 2012-01-21 23:43:54 <etotheipi_> roconnor, cryptopp
629 2012-01-21 23:44:32 <etotheipi_> gmaxwell, that is almost exactly what I did (if I understand what you said). ... after the SelectCoins is done, I look to see if there's any low-priority outputs from addresses already being used and tack them on
630 2012-01-21 23:44:49 <etotheipi_> as long as it doesn't break allowFree (if allowFree was satisfied to begin with)
631 2012-01-21 23:45:11 <roconnor> etotheipi_: cryptopp supports ecdsa, or you built it out of their primitives?
632 2012-01-21 23:45:31 <etotheipi_> roconnor, it supports ECDSA (though it's not nearly as fast as openssl)
633 2012-01-21 23:45:45 <etotheipi_> though I had to use lower-level integer math for the determinstic wallet code
634 2012-01-21 23:45:47 <gmaxwell> etotheipi_: yea, there you go. GMTA. But yea, what I was worried about is that it would gobble up all the inputs you have in varrious amount ranges, causing later transactions to produce change more often.
635 2012-01-21 23:46:30 <gmaxwell> etotheipi_: did you ever create simulation code to figure out how often it mattered?
636 2012-01-21 23:46:50 <sipa> etotheipi_: i believe i already asked you, but how fast is a signature verification in cryptopp's ECDSA?
637 2012-01-21 23:46:55 <etotheipi_> gmaxwell, I have to double-check, but it's pretty strict criteria for when it will tack them on... only if it's low priority and tiny
638 2012-01-21 23:47:56 <etotheipi_> gmaxwell, unfortunately I somehow lost my selectcoins unit test... but I did test a lot of different scenarios and was surprised how often it grabbed one or two extra inputs in that code
639 2012-01-21 23:48:04 <etotheipi_> sipa, I think it's like 100-150 ECDSA verifies/s
640 2012-01-21 23:48:13 <edcba> i doubt edcsa is bottleneck anywhere in bitcoin
641 2012-01-21 23:48:28 <etotheipi_> edcba, it would be a bottleneck if I was doing full verification
642 2012-01-21 23:48:55 <TuxBlackEdo> cant someone figure this out by putting in some nanosecond timers between routines?
643 2012-01-21 23:49:03 <edcba> indeed
644 2012-01-21 23:49:12 <gmaxwell> etotheipi_: my motivation was limited because I didn't think it would matter that often... and didn't care enough to write a simulation to find out.
645 2012-01-21 23:49:14 <sipa> etotheipi_: what kind of CPU?
646 2012-01-21 23:49:16 <edcba> or having some profiling tool running
647 2012-01-21 23:49:25 <etotheipi_> i5-2500k
648 2012-01-21 23:49:51 <etotheipi_> cryptopp is actually a bit faster, but I have some overhead to the function call for converting data types
649 2012-01-21 23:50:03 <gmaxwell> etotheipi_: Its interesting to me that you say it's run often. Perhaps I'll do it again then for the refernce client. Still dunno how to avoid the issue with making future txn more change heavy.
650 2012-01-21 23:50:07 <sipa> shouldn't be much difference
651 2012-01-21 23:50:09 <CIA-76> bitcoin: Con Kolivas * r10378f878a43 cgminer/bitforce.c: Fix windows build of bitforce blindly. Hopefully it works. http://tinyurl.com/6rvxwum
652 2012-01-21 23:50:41 <etotheipi_> gmaxwell, I had a lot of fun figuring out how to optimize "future unspent-output fields" or whatever you want to call it
653 2012-01-21 23:50:59 <etotheipi_> I mean, I didn't get very far beyond what I just described with tacking on extras
654 2012-01-21 23:51:41 <etotheipi_> however, if the tx is for X BTC, I try to make the output 2X so that I get back an almost-identical change output
655 2012-01-21 23:52:20 <roconnor> etotheipi_: 01de23d3997571c5d8f34955e45ecc56500097e8 is the latest revision?
656 2012-01-21 23:52:48 <etotheipi_> roconnor, whatever is on the qtdev branch
657 2012-01-21 23:52:56 <roconnor> oh
658 2012-01-21 23:53:32 <etotheipi_> my first release will be marked by merging into master
659 2012-01-21 23:53:41 <etotheipi_> but I haven't actually commited anythign to master for 1.5 montsh
660 2012-01-21 23:53:42 <edcba> ok how much transactions per second handles bitcoin network ?
661 2012-01-21 23:53:54 <etotheipi_> edcba, less than 1 tx/s
662 2012-01-21 23:53:58 <etotheipi_> (right now)
663 2012-01-21 23:55:06 <roconnor> does step 8 "python ArmoryQt.py" compile or run Amory?
664 2012-01-21 23:55:13 <etotheipi_> that runs it
665 2012-01-21 23:55:19 <edcba> indeed more like 1 tx/min
666 2012-01-21 23:55:19 <etotheipi_> and use "--testnet"