1 2012-04-13 00:44:18 <paulo_> hello
  2 2012-04-13 00:45:35 <paulo_> the merkle root are hashes of what, exactly?
  3 2012-04-13 00:47:52 <[Tycho]> Transactions.
  4 2012-04-13 00:48:18 <paulo_> "Because transactions aren't hashed directly, hashing a block with 1 transaction takes exactly the same amount of effort as hashing a block with 10,000 transactions." - https://en.bitcoin.it/wiki/Block_hashing_algorithm
  5 2012-04-13 00:48:33 <[Tycho]> Yes, exactly.
  6 2012-04-13 00:49:07 <midnightmagic> imagine a bunch of individual hashes. now every two hashes, put them together into one hash, connected to the two. now keep doing that, joining each level of hashes until you are left with one. it is a hash or a tree of a bunch of hashes, and by itself it is included in the hash calculation of the block so the entire transaction hash set doesn't have to be to do a speedy verification of the block chain..
  7 2012-04-13 00:49:10 <[Tycho]> Transactions are hashed into the merkle tree just once, and then this hash is hashed to find blocks.
  8 2012-04-13 00:49:46 <paulo_> another question,
  9 2012-04-13 00:49:49 <midnightmagic> that is the "root" of the tree of hashes.
 10 2012-04-13 00:49:59 <midnightmagic> the tree of hashes is called a merkle tree
 11 2012-04-13 00:50:21 <paulo_> transactions are made as a newly found block spreads across the network. how do you compensate for that?
 12 2012-04-13 00:50:59 <gmaxwell> Can you clarify what you're actually asking?
 13 2012-04-13 00:51:28 <paulo_> it's a second question, i'm researching what a merkle tree is.
 14 2012-04-13 00:52:02 <paulo_> i mean, new transactions are made, which aren't included in the newly found block.
 15 2012-04-13 00:52:23 <[Tycho]> paulo_: they may be included in the next one.
 16 2012-04-13 00:52:31 <luke-jr> paulo_: you might find Eloipool's source code useful
 17 2012-04-13 00:52:35 <gmaxwell> Right, they'll be included in a subsequent one.
 18 2012-04-13 00:52:43 <luke-jr> gitorious.org/bitcoin/eloipool
 19 2012-04-13 00:53:18 <midnightmagic> assuming the miners mining the new blocks "like" the new transaction..
 20 2012-04-13 00:54:06 <sethman895> Has Anybody here tried electrum?
 21 2012-04-13 00:54:16 <paulo_> but how will the client know that the new transactions are not supposed to be include in the new block? (since the block arrives late to the client)
 22 2012-04-13 00:55:33 <paulo_> by included i mean included in the hash
 23 2012-04-13 00:55:36 <gmaxwell> By the client you mean "some random mining node"?
 24 2012-04-13 00:55:58 <paulo_> no, the every bitcoin node.
 25 2012-04-13 00:56:43 <gmaxwell> I think you need to walk through what you're thinking more specifically and clearly. I'm guessing you're suffering a misunderstanding, but what you're saying is so vague I can't figure out what it could be.
 26 2012-04-13 00:57:05 <[Tycho]> paulo_: each block contains all those transactions, so you always see what it inside and what isn't
 27 2012-04-13 00:57:11 <[Tycho]> *is
 28 2012-04-13 00:57:12 <gmaxwell> Give me a step by step example.
 29 2012-04-13 00:58:48 <Diablo-D3> welp I guess that was fun
 30 2012-04-13 00:59:02 <Diablo-D3> I finished my stm impl, dunno if it works
 31 2012-04-13 00:59:06 <Diablo-D3> probably never will
 32 2012-04-13 01:00:23 <paulo_> gmaxwell: i'll try step by step.
 33 2012-04-13 01:00:49 <midnightmagic> paulo_: when the block arrives from other miners, the client will know whether or not any transaction it curently knows about has been included in that block, since right now, the block includes every transaction: so the client can recalculate the hash tree of the transactions, and verify for itself that the block is valid.
 34 2012-04-13 01:01:35 <Diablo-D3> wonder why I bother sometimes
 35 2012-04-13 01:01:53 <midnightmagic> if the transaction you're talking about hasn't been included in the new block, the clients knows this and it stays in limbo for the time being.
 36 2012-04-13 01:02:05 <midnightmagic> Diablo-D3: stm?
 37 2012-04-13 01:02:12 <Diablo-D3> software transactional memory
 38 2012-04-13 01:06:15 <Diablo-D3> it was supposed to go in libmowgli2
 39 2012-04-13 01:06:40 <Diablo-D3> so I could get rid of the overhead of locks in a lot of things
 40 2012-04-13 01:06:47 <paulo_> midnightmagic: stays in limbo?
 41 2012-04-13 01:07:47 <[Tycho]> paulo_: memorypool
 42 2012-04-13 01:08:28 <[Tycho]> Still not sure what paulo_ is talking about.
 43 2012-04-13 01:08:34 <paulo_> midnightmagic: and how does the client "know" and not just discard the block as a fake block?
 44 2012-04-13 01:09:15 <gmaxwell> paulo_: why would it do that?
 45 2012-04-13 01:09:25 <gmaxwell> see this is why I wanted you to walk through it.
 46 2012-04-13 01:09:41 <paulo_> wait, i'll paste bin it.
 47 2012-04-13 01:09:44 <gmaxwell> Blocks are validated against the chain they extend and the rules of the system.
 48 2012-04-13 01:10:43 <gmaxwell> Transactions which aren't yet in blocks yet have no relevance rather if the blocks make those transactions invalid, they they are set aside. They don't influence the validation of new blocks.
 49 2012-04-13 01:12:30 <paulo_> I think I figured out where I misunderstood.
 50 2012-04-13 01:16:49 <paulo_> so you keep adding transactions to the tree until it matches the hash in the block?
 51 2012-04-13 01:17:36 <gmaxwell> No. Not at all.
 52 2012-04-13 01:17:50 <gmaxwell> The block comes with the set of transactions which are included in that block.
 53 2012-04-13 01:18:09 <gmaxwell> Given the set of transactions you can determinstically construct the tree and see that it matches.
 54 2012-04-13 01:18:41 <gmaxwell> And with the set of transactions you can also see that they themselves validate, and that they don't conflict with each other or any of the transactions in the prior blocks.
 55 2012-04-13 01:18:50 <paulo_> so when the blocks are spread across the network, the transactions themselves are included, and not just the hash?
 56 2012-04-13 01:19:06 <sethman895> This is the first time I've been on here and Bitcoin talk was going on.
 57 2012-04-13 01:20:50 <gmaxwell> sethman895: random sampling says this channel is mostly on-topic: http://pastebin.com/qQ9Bwc4R
 58 2012-04-13 01:21:16 <gmaxwell> paulo_: Yes.
 59 2012-04-13 01:21:38 <paulo_> I understand now. gtg & thanks.
 60 2012-04-13 02:06:54 <splatster> I'm still figuring stuff out as for my whole comp situation after what happened last night.
 61 2012-04-13 02:59:42 <paulo_> too bad I have an nvidia.
 62 2012-04-13 03:00:51 <gribble> New news from bitcoinrss: cardpuncher opened pull request 1088 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1088>
 63 2012-04-13 03:06:39 <midnightmagic> "Its sort of like BitCoin but with actual, government-backed value." ooookay
 64 2012-04-13 03:06:52 <Diablo-D3> except its not like bitcoin at all
 65 2012-04-13 03:07:00 <Diablo-D3> I mean, outside of that insanely stupid statement
 66 2012-04-13 03:08:00 <TuxBlackEdo> i think it is way more interesting for bitcoin to not have any backing at all and have value
 67 2012-04-13 03:08:21 <TuxBlackEdo> that makes it like 100x better then mintchip
 68 2012-04-13 03:08:34 <Diablo-D3> oh yeah?
 69 2012-04-13 03:08:41 <TuxBlackEdo> mintchip is just another fiat currency (even though its digital)
 70 2012-04-13 03:08:42 <Diablo-D3> its backed.
 71 2012-04-13 03:08:49 <Diablo-D3> bam.
 72 2012-04-13 03:09:28 <TuxBlackEdo> Diablo-D3, I guess...
 73 2012-04-13 03:09:44 <guruvan> the mintchip comes in different flavors that are in different current  fiat currencies (EUR, GBP, USD ,etc)
 74 2012-04-13 03:10:34 <Diablo-D3> does it come in strawberry?
 75 2012-04-13 03:10:40 <TuxBlackEdo> but it's still fiat currency
 76 2012-04-13 03:10:48 <Diablo-D3> TuxBlackEdo: actually, its not even a currency
 77 2012-04-13 03:10:57 <Diablo-D3> its a digital certificate bearer bond... certificate.
 78 2012-04-13 03:11:11 <TuxBlackEdo> the reason i like bitcoin is because nobody can just make a million bitcoins in a day
 79 2012-04-13 03:11:35 <TuxBlackEdo> Diablo-D3, I get it
 80 2012-04-13 03:12:11 <TuxBlackEdo> Diablo-D3, here is a question to keep you busy: how much do 1 million bitcoins weigh?
 81 2012-04-13 03:12:41 <Diablo-D3> enough to crush a table, apparently.
 82 2012-04-13 03:12:49 <BlueMatt> a shitton, actually (when you calculate hard drive weight of people holding the chain...)
 83 2012-04-13 03:13:12 <TuxBlackEdo> BlueMatt, my laptop does get heavier the more i fill up my hard drive ;)
 84 2012-04-13 03:13:20 <BlueMatt> heh
 85 2012-04-13 03:13:36 <BlueMatt> it does if you have an ssd (a few electrons per 1-bit) ;)
 86 2012-04-13 03:14:16 <TuxBlackEdo> even mechanical hard drives apparently
 87 2012-04-13 03:14:22 <luke-jr> the problem is, to NORMAL PEOPLE, Mintchip *is* just like Bitcoin
 88 2012-04-13 03:14:38 <luke-jr> and they won't understand why it isn't.
 89 2012-04-13 03:14:44 <andytoshi> luke-jr: no, it has a completely different trust model
 90 2012-04-13 03:14:49 <andytoshi> that can be understood at a high level
 91 2012-04-13 03:14:50 <TuxBlackEdo> luke-jr, is it possible to use eligius to attack this alt chain? :)
 92 2012-04-13 03:14:56 <luke-jr> TuxBlackEdo: what chain?
 93 2012-04-13 03:15:01 <Diablo-D3> normal people are fucktards
 94 2012-04-13 03:15:02 <TuxBlackEdo> mintchip
 95 2012-04-13 03:15:06 <Diablo-D3> this is why I keep taking their money
 96 2012-04-13 03:15:08 <TuxBlackEdo> i wanted to make a joke luke-jr :(
 97 2012-04-13 03:17:40 <TuxBlackEdo> BlueMatt, turns out a laptop weighs more if it is fully charged vs. it's ssd being full
 98 2012-04-13 03:17:50 <BlueMatt> TuxBlackEdo: makes sense
 99 2012-04-13 03:19:27 <xenland> Apparently electricity carries weight these days...
100 2012-04-13 03:19:40 <Diablo-D3> hurray quantum chromodynamics.
101 2012-04-13 03:19:52 <TuxBlackEdo> xenland, electrons do have weight
102 2012-04-13 03:20:20 <xenland> Deffinatly not enough to be noticeable enough on a fully charged laptop
103 2012-04-13 03:20:28 <xenland> maybe a super battery.... i might believe you
104 2012-04-13 03:20:37 <Diablo-D3> something the size of the earth, really.
105 2012-04-13 03:20:59 <BlueMatt> xenland: no one actually said it was noticeable
106 2012-04-13 03:21:03 <TuxBlackEdo> even with a super battery the weight difference would be 1^-18 grams
107 2012-04-13 03:21:20 <Diablo-D3> TuxBlackEdo: or a street value of 2 million
108 2012-04-13 03:21:41 <xenland> Noticable/observable
109 2012-04-13 03:21:42 <BlueMatt> thats some expensive coke
110 2012-04-13 03:21:54 <BlueMatt> 2mill/1^-18 grams
111 2012-04-13 03:21:58 <Diablo-D3> and when they get that 1^-19 grams into the evidence locker, that 1^-20 grams will put you away for years
112 2012-04-13 03:22:08 <BlueMatt> heh
113 2012-04-13 03:22:47 <TuxBlackEdo> i meane 1 e-19
114 2012-04-13 03:22:53 <TuxBlackEdo> meant*
115 2012-04-13 03:23:07 <BlueMatt> so did the rest of us, I think
116 2012-04-13 03:23:34 <TuxBlackEdo> or 1x10^-19
117 2012-04-13 03:23:40 <Diablo-D3> I didnt, I thought that was the joke
118 2012-04-13 03:23:45 <BlueMatt> oh...
119 2012-04-13 03:23:51 <BlueMatt> well, that is also a valid joke...
120 2012-04-13 03:24:49 <TuxBlackEdo> does anyone here have bitcoin related dreams?
121 2012-04-13 03:24:59 <Diablo-D3> no
122 2012-04-13 03:25:07 <TuxBlackEdo> me neither...
123 2012-04-13 03:25:10 <Diablo-D3> but I once had a dream I was having sex with a hot alien woman.
124 2012-04-13 03:25:20 <TuxBlackEdo> Diablo-D3, pics or it didn't happen
125 2012-04-13 03:25:24 <Diablo-D3> >dream
126 2012-04-13 03:25:25 <Diablo-D3> >pics
127 2012-04-13 03:25:26 <luke-jr> &
128 2012-04-13 03:25:30 <xenland> hehe
129 2012-04-13 03:25:33 <BlueMatt> Diablo-D3: meh, I actually experienced that...
130 2012-04-13 03:25:38 <BlueMatt> (well, the drugs told me I did)
131 2012-04-13 03:25:48 <xenland> Lucid dreams are very entertaining
132 2012-04-13 03:26:20 <luke-jr> oh, I thought BlueMatt meant the sex with a hot alien woman.
133 2012-04-13 03:26:23 <Diablo-D3> what I dont get is, cat ears, right? how the fuck did the fleet admiral hat fit her head without ear holes
134 2012-04-13 03:26:41 <Diablo-D3> (needless to mention, that was the ONLY thing she was wearing)
135 2012-04-13 03:27:22 <TuxBlackEdo> i have nightmares about slender man :(
136 2012-04-13 03:27:47 <Diablo-D3> well dont worry TuxBlackEdo, Im not slender.
137 2012-04-13 03:28:20 <ferroh_> Any idea what might be causing the chain to DL so slow for this guy? http://www.reddit.com/r/Bitcoin/comments/s60sj/why_is_block_downloading_so_slow/
138 2012-04-13 03:28:35 <ferroh_> He is using a USB key to store the blockchain, I suggested that he was heavily IO bound then.
139 2012-04-13 03:28:47 <ferroh_> However, he reports downloading only 150 blocks in 3 hours.
140 2012-04-13 03:28:49 <Diablo-D3> yup thats it
141 2012-04-13 03:28:56 <Diablo-D3> thats the right answer
142 2012-04-13 03:29:00 <ferroh_> over 6 hours actually
143 2012-04-13 03:29:10 <ferroh_> 6 hours = 150 blocks?
144 2012-04-13 03:29:16 <Diablo-D3> on a slow enough IO device? yes
145 2012-04-13 03:29:34 <phantomcircuit> ferroh_, a usb flash drive is going to give you like
146 2012-04-13 03:29:39 <phantomcircuit> maybe 2 write ops/second
147 2012-04-13 03:29:51 <phantomcircuit> and once it's close to full you're talking seconds per write op
148 2012-04-13 03:30:55 <ferroh_> Is there something I can suggest for this guy, as a way to easily store the wallet.dat only on the flash drive?
149 2012-04-13 03:31:15 <ferroh_> I guess copy/paste it
150 2012-04-13 03:31:45 <Diablo-D3> just make sure the client is off when he does it
151 2012-04-13 03:32:38 <ferroh_> indeed, alrighty
152 2012-04-13 03:32:46 <ferroh_> How about using a USB HDD?
153 2012-04-13 03:32:53 <MC1984> wait can you really currupt a wallet by copying it
154 2012-04-13 03:32:58 <ferroh_> Is the bottleneck the USB interface, or the drive itself do you suppose?
155 2012-04-13 03:37:07 <MC1984> guys
156 2012-04-13 03:37:18 <MC1984> copying a live wallet corrupts it?
157 2012-04-13 03:37:28 <BlueMatt> ofc
158 2012-04-13 03:37:33 <Diablo-D3> yes
159 2012-04-13 03:37:33 <MC1984> how, since copying is a read only
160 2012-04-13 03:37:46 <Diablo-D3> MC1984: because you're looking at it non-transactionally
161 2012-04-13 03:37:53 <BlueMatt> its a database, you dont copy a database while its open
162 2012-04-13 03:38:02 <Diablo-D3> MC1984: writes to disk are not atomic
163 2012-04-13 03:38:25 <splatster> Yes, quit your client before trying to backup/replace/copy/whatever your wallet.
164 2012-04-13 03:38:39 <MC1984> so you can end up with a corrupted copy, but the original is fine?
165 2012-04-13 03:38:44 <Diablo-D3> yes
166 2012-04-13 03:38:47 <Diablo-D3> the original is fine
167 2012-04-13 03:38:47 <splatster> yup
168 2012-04-13 03:38:53 <MC1984> thats ok then
169 2012-04-13 03:39:01 <Diablo-D3> thats not okay either
170 2012-04-13 03:39:04 <Diablo-D3> you depend on the copy
171 2012-04-13 03:39:13 <MC1984> well no its not ok
172 2012-04-13 03:39:28 <MC1984> but its more ok than breaking your original too
173 2012-04-13 03:39:49 <MC1984> so the wallet export thing will let you do it without corruption
174 2012-04-13 03:40:24 <Diablo-D3> yeah
175 2012-04-13 03:40:30 <Diablo-D3> it produces an atomic copy
176 2012-04-13 03:41:23 <MC1984> duly noted
177 2012-04-13 03:42:56 <MC1984> i was thinking given what happened to splatster
178 2012-04-13 03:43:03 <MC1984> how about some anti keylogging stuff
179 2012-04-13 03:43:40 <Diablo-D3> he was the one with the missing 100btc right?
180 2012-04-13 03:43:43 <MC1984> like password input via on screen keyboard, which moves to a new screen X, Y position after every character
181 2012-04-13 03:43:58 <MC1984> yes
182 2012-04-13 03:44:03 <Diablo-D3> did anyone find out what happened?
183 2012-04-13 03:44:15 <splatster> Diablo-D3: I'm working on it.
184 2012-04-13 03:44:17 <MC1984> probably got a keylogger on his mac, no other way
185 2012-04-13 03:44:30 <Diablo-D3> yeah, but what would a keylogger do here
186 2012-04-13 03:44:43 <Diablo-D3> it'd still have to grab his wallet.dat
187 2012-04-13 03:44:49 <MC1984> log his wallet password
188 2012-04-13 03:44:55 <MC1984> well thats not hard
189 2012-04-13 03:44:55 <splatster> Again, my main focus is to get my operations running again on this fresh comp and then when I have free time I will dissect my clone.
190 2012-04-13 03:44:58 <Diablo-D3> but what use is a wallet password without a wallet
191 2012-04-13 03:45:24 <MC1984> Diablo-D3 if it can keylog you and send back to base, it can also grad a file and send it
192 2012-04-13 03:45:34 <MC1984> one piece of malware can do both bro...
193 2012-04-13 03:45:47 <Diablo-D3> this is why I use linux
194 2012-04-13 03:46:13 <MC1984> ha, thats what mac users were saying up until last week you smug prick :P
195 2012-04-13 03:46:37 <Diablo-D3> last week? why last week? osx has been known to be not secure for about 5 years
196 2012-04-13 03:47:02 <MC1984> yeah but it didnt matter until the 600k botnet story
197 2012-04-13 03:47:13 <Diablo-D3> and thats why osx cant be trusted
198 2012-04-13 03:47:15 <Diablo-D3> "it didnt matter"
199 2012-04-13 03:47:23 <MC1984> quite
200 2012-04-13 03:47:25 <Diablo-D3> you fix security holes BEFORE they happen, not after
201 2012-04-13 03:48:01 <splatster> Diablo-D3: It's my fault.  I kept funds on a hot wallet.  It was my mistake, so I took the hit.
202 2012-04-13 03:48:30 <Diablo-D3> splatster: that shouldnt actually matter, really
203 2012-04-13 03:48:52 <splatster> Diablo-D3: Well, it did.
204 2012-04-13 03:49:12 <Diablo-D3> yes, if your machine isnt secure, it doesnt if the wallet is hot or luke warm or whatever
205 2012-04-13 03:50:14 <MC1984> ok so anti keylogging ideas
206 2012-04-13 03:50:24 <MC1984> the randomised on screen keyboard?
207 2012-04-13 03:50:30 <Diablo-D3> no
208 2012-04-13 03:50:36 <MC1984> why
209 2012-04-13 03:50:42 <Diablo-D3> thats extremely hard to implement anyhow
210 2012-04-13 03:51:09 <MC1984> i dont think so
211 2012-04-13 03:51:11 <Diablo-D3> use overlaid graphics and require the user to drag them around to produce a certain color by filtering.
212 2012-04-13 03:51:49 <MC1984> whut
213 2012-04-13 03:51:58 <Diablo-D3> hard mode: you do the dragging on an iphone or android device using touch.js in a browser.
214 2012-04-13 03:54:36 <splatster> luke-jr: What should I do about replacing my old GPG key with my new one?
215 2012-04-13 03:55:02 <Diablo-D3> splatster: produce a shitlisting certificate for it
216 2012-04-13 03:55:04 <splatster> (I have both, but I am unsure as to whether the old one was compromised as well)
217 2012-04-13 03:55:12 <Diablo-D3> do not sign your new with your old
218 2012-04-13 03:59:50 <mod6> Thats funny, he got exactly the opposite advice eariler.
219 2012-04-13 04:00:00 <mod6> *earlier
220 2012-04-13 04:00:14 <Diablo-D3> this is a faq in the gpg manual iirc
221 2012-04-13 04:00:58 <mod6> oh, maybe im remembering wrong, I thought it was, sign new with old to verify identity, then revoke old key.
222 2012-04-13 04:01:07 <Diablo-D3> no.
223 2012-04-13 04:01:12 <Diablo-D3> revoke, do NOT sign
224 2012-04-13 04:01:20 <Diablo-D3> and then have everyone that knows you sign your new one
225 2012-04-13 04:08:09 <mod6> ok, srry. looked but couldn't find, weak.
226 2012-04-13 04:08:29 <mod6> scrollback only goes so far :/
227 2012-04-13 04:08:48 <MC1984> hmm
228 2012-04-13 04:09:19 <MC1984> https://github.com/coblee/litecoin/wiki/Mining-hardware-comparison according to this, ati gpu stil far outstrips a gpy for litecoin mining
229 2012-04-13 04:09:32 <MC1984> i thought litecoin was supposed to be gpu hostile
230 2012-04-13 04:09:44 <Diablo-D3> you mean cpu vs gpu?
231 2012-04-13 04:09:48 <Diablo-D3> litecoin was designed retardedly.
232 2012-04-13 04:09:50 <MC1984> yes
233 2012-04-13 04:09:59 <MC1984> howso
234 2012-04-13 04:10:04 <Diablo-D3> it isnt gpu hostile.
235 2012-04-13 04:10:23 <MC1984> but its supposed to be?
236 2012-04-13 04:10:29 <Diablo-D3> yes
237 2012-04-13 04:10:38 <MC1984> how did they fuck up then
238 2012-04-13 04:10:52 <Diablo-D3> their algo uses memory, it doesnt use ENOUGH memory
239 2012-04-13 04:11:32 <MC1984> its just the memory requirement that was meant to keep in on cpu only?
240 2012-04-13 04:11:51 <Diablo-D3> basically
241 2012-04-13 04:11:56 <Diablo-D3> theres nothing suitably tricky going on
242 2012-04-13 04:12:39 <MC1984> hmm thats not so good
243 2012-04-13 04:13:30 <MC1984> is it even possible to make an algo that will only be good to mine on say, x86 cpus and nothing else for the foreseeable future
244 2012-04-13 04:13:53 <Diablo-D3> no
245 2012-04-13 04:14:16 <Diablo-D3> unless it uses an absolute fuckton of memory (which makes it unusable on smaller computers too), it can be gpu and fpga mined
246 2012-04-13 04:14:23 <MC1984> goddamit thats a big problem
247 2012-04-13 04:14:31 <Diablo-D3> well the stupidity is
248 2012-04-13 04:14:36 <Diablo-D3> bitcoin ALREADY solved the problem
249 2012-04-13 04:14:42 <Diablo-D3> just keep cranking difficulty up