1 2013-04-07 00:01:34 <Luke-Jr> I hope i get a break from bugs soon so I can redo next-test
2 2013-04-07 01:28:13 <Kinnard> Is it usually quiet around this time?
3 2013-04-07 01:28:22 <lianj> yes
4 2013-04-07 01:29:22 <digitalmagus> *shhhh*
5 2013-04-07 01:29:46 <Phraust> they are feeding.
6 2013-04-07 01:30:45 <digitalmagus> ACTION Om nom,nom,nom
7 2013-04-07 01:31:23 <realazthat> rofl
8 2013-04-07 01:44:17 <azizLIGHTS> when you run bitcoin-qt and its uploading maximum speed, does anyone else get Error 15 (net::ERR_SOCKET_NOT_CONNECTED) errors on Google chrome, on https sites, even simple google searches fail with bitcoin-qt running. the problem stops after i quit bitcoin-qt
9 2013-04-07 01:46:49 <gmaxwell> sounds like a you have a broken router. Bummer.
10 2013-04-07 01:47:56 <azizLIGHTS> so far its only happening on https://google.com and mainly affecting my google searches
11 2013-04-07 01:59:43 <Diablo-D3> azizLIGHTS: your router is pretty shitty
12 2013-04-07 01:59:56 <Diablo-D3> does it happen when you torrent on a populated torrent too?
13 2013-04-07 02:00:02 <Diablo-D3> (like, connected to >100 peers)
14 2013-04-07 02:00:14 <azizLIGHTS> not sure... havent paid attention to that
15 2013-04-07 02:00:25 <Diablo-D3> find out
16 2013-04-07 02:00:30 <azizLIGHTS> what is it about my router
17 2013-04-07 02:00:37 <Diablo-D3> limited connection table size
18 2013-04-07 02:00:53 <Diablo-D3> the fix is to either quit using shitty routers or install openwrt/ddwrt on them
19 2013-04-07 02:01:18 <lianj> or decrease bitcoin-qts connections
20 2013-04-07 02:01:29 <Luke-Jr> ACTION peers. I say "I hate Google" in here and suddenly Google is trying to hire me again -.-
21 2013-04-07 02:01:46 <Diablo-D3> Luke-Jr: dude, they've repeatedly tried to hire nenolod
22 2013-04-07 02:01:48 <Diablo-D3> its hilarious
23 2013-04-07 02:02:00 <Luke-Jr> nenolod?
24 2013-04-07 02:02:24 <Diablo-D3> famous programmer guy
25 2013-04-07 02:02:32 <Diablo-D3> wrote the ircd you're connected to atm
26 2013-04-07 02:02:36 <Diablo-D3> wrote audacious
27 2013-04-07 02:02:41 <Diablo-D3> invented modern cloud computing
28 2013-04-07 02:02:50 <Diablo-D3> runs his own linux distro
29 2013-04-07 02:02:59 <Diablo-D3> was a multi-millionare at one point
30 2013-04-07 02:03:16 <lianj> no wonder he got no time for google
31 2013-04-07 02:03:22 <Diablo-D3> Luke-Jr: anyhow, the point is, he keeps turning them down
32 2013-04-07 02:03:33 <Diablo-D3> and they keep asking and asking
33 2013-04-07 02:03:38 <Diablo-D3> and its hilarious
34 2013-04-07 02:03:56 <Luke-Jr> I just thought it was a funny coincidence the timing with me ranting on them in here XD
35 2013-04-07 02:03:59 <Diablo-D3> he hates the corporate culture at google
36 2013-04-07 02:04:14 <tommygunner> evening
37 2013-04-07 02:04:17 <Luke-Jr> I hate the fact that I can't visit any Google website anymore, without loading up bloatware browsers
38 2013-04-07 02:04:20 <tommygunner> whats up with 0.8.1 qt
39 2013-04-07 02:04:28 <tommygunner> cant keep a blockchain to save its life
40 2013-04-07 02:04:34 <tommygunner> corrupt corrupt corrupt
41 2013-04-07 02:05:54 <tommygunner> http://pastebin.com/79A4zYXD
42 2013-04-07 02:07:23 <tommygunner> same thing happened maybe a week or two ago
43 2013-04-07 02:07:38 <tommygunner> with a grandfathered-in blockchain
44 2013-04-07 02:07:46 <tommygunner> re-did it then and now its happening again
45 2013-04-07 02:07:57 <tommygunner> (OS X 10.8.3)
46 2013-04-07 02:11:29 <realazthat> sipa: ping?
47 2013-04-07 02:11:38 <realazthat> sipa: you mentioned gettxout
48 2013-04-07 02:11:42 <realazthat> any docs on that?
49 2013-04-07 02:16:20 <Eneerge> anyone know of any cool open projects to get involved with?
50 2013-04-07 02:23:30 <Luke-Jr> Eneerge: what kind?
51 2013-04-07 02:24:04 <gmaxwell> realazthat: bitcoind help gettxout
52 2013-04-07 02:24:22 <realazthat> gmaxwell: ah ty
53 2013-04-07 03:31:20 <peawormsworth> someone has sent bitcoin to an ltc wallet. Would anyone like to help with this?
54 2013-04-07 03:35:12 <phantomcircuit> is that even possible
55 2013-04-07 03:35:14 <peawormsworth> "sohcpunk" said he moved 45 BTC to an LTC address. He is wondering if there is any way to get it back. He is in the #bitcoin-tech channel now.
56 2013-04-07 03:35:17 <peawormsworth> I dont know.
57 2013-04-07 03:35:36 <phantomcircuit> https://en.bitcoin.it/wiki/Litecoin#Different_Addresses
58 2013-04-07 03:36:15 <Diablo-D3> peawormsworth: wtf, thats impossible
59 2013-04-07 03:36:29 <Diablo-D3> ltc uses a different magic number
60 2013-04-07 03:36:43 <Diablo-D3> at least I thought it did
61 2013-04-07 03:36:45 <warren> Perhaps some alt client fails to reject the prefix?
62 2013-04-07 03:37:00 <Diablo-D3> well, you cant send it to another wallet, thats nonsensical
63 2013-04-07 03:37:10 <Diablo-D3> it sounds like he sent it to an address that is valid in bitcoin but no one has a private key for
64 2013-04-07 03:37:14 <Diablo-D3> ergo, they're lost forever
65 2013-04-07 03:37:57 <peawormsworth> Diablo-D3: would it be possible to extract the private key from an ltc wallet and import it into a btc client wallet ?
66 2013-04-07 03:38:56 <Diablo-D3> peawormsworth: sure, try it
67 2013-04-07 03:39:13 <Diablo-D3> do LTC addresses start with 1?
68 2013-04-07 03:39:19 <Diablo-D3> because if they do, LTC is broken
69 2013-04-07 03:40:30 <warren> Diablo-D3: L
70 2013-04-07 03:40:47 <warren> there are a few alts that begin with "1"
71 2013-04-07 03:40:57 <gmaxwell> peawormsworth: what software did this person use?
72 2013-04-07 03:40:58 <Diablo-D3> L would have never been valid as a bitcoin address
73 2013-04-07 03:41:04 <Diablo-D3> 1 would have been POSSIBLY
74 2013-04-07 03:41:34 <gmaxwell> peawormsworth: and does he know the transaction ID?
75 2013-04-07 03:41:47 <peawormsworth> he is talking on the #bitcoin-tech channel. He said it was sent thru btc-e
76 2013-04-07 03:42:12 <gmaxwell> peawormsworth: it's probably still in btc-e.
77 2013-04-07 03:42:24 <gmaxwell> "bugs in btce software? unheard of!"
78 2013-04-07 03:42:27 <Diablo-D3> yeah what gmaxwell said
79 2013-04-07 03:42:39 <peawormsworth> I encourage u to go to #bitcoin-tech if ur interested. I was just rying to get him some professional help.
80 2013-04-07 03:43:07 <peawormsworth> that sounds correct. btc-e probably has it and will return it.
81 2013-04-07 03:43:27 <ThomasV> or they sent to a black hole
82 2013-04-07 03:44:16 <phantomcircuit> peawormsworth, what's the address?
83 2013-04-07 03:45:00 <doublec> he needs to find the owner of that ltc address. get them to export the private key from litecoin, import it into bitcoin and send the funds to him.
84 2013-04-07 03:45:07 <doublec> btce doesn't validate addresses
85 2013-04-07 03:45:18 <warren> btc-e should at least be able to supply a tx. If they can't, they didn't send it.
86 2013-04-07 03:45:24 <phantomcircuit> what version number is litecoin using?
87 2013-04-07 03:45:36 <doublec> this has sadly happened numerous times to my litecoin exchange
88 2013-04-07 03:45:58 <warren> doublec: huh? how do people send a "1" to "L" and vice versa?
89 2013-04-07 03:46:15 <doublec> warren: there's an equivalent bitcoin address it maps to
90 2013-04-07 03:46:22 <peawormsworth> LMk7wcGbX5MEgubyCA7LD8t62x7aVbgzci
91 2013-04-07 03:46:33 <warren> doublec: sending clients don't reject it?
92 2013-04-07 03:46:34 <doublec> old versions of namecoind did the same. You could happily send namecoins to bitcoin addresses because it didn't check
93 2013-04-07 03:46:40 <doublec> warren: apparently btce doesn't
94 2013-04-07 03:46:45 <warren> wow
95 2013-04-07 03:46:47 <peawormsworth> but once again... i am a middle man here. I invite any interested discussion to #bitcoin-tech... where he is talking about it.
96 2013-04-07 03:47:19 <phantomcircuit> PUBKEY_ADDRESS = 48, // Litecoin addresses start with L
97 2013-04-07 03:47:21 <peawormsworth> You guys are great. but i didnt want to pollute the dev channel with one issue.
98 2013-04-07 03:47:54 <phantomcircuit> so a normal litecoin address wouldn't be valid in bitcoinqt
99 2013-04-07 03:49:06 <phantomcircuit> doublec, except he said it was btc to an address ltc not the other way around
100 2013-04-07 03:49:21 <phantomcircuit> doublec, i could see sending ltc to a btc address, but the other way around should be impossible
101 2013-04-07 03:49:35 <Diablo-D3> phantomcircuit: well if they have 1 addresses
102 2013-04-07 03:49:38 <Diablo-D3> and it uses the same format
103 2013-04-07 03:49:39 <Diablo-D3> I can see it
104 2013-04-07 03:49:57 <Diablo-D3> this is why there are different magic numbers on addresses
105 2013-04-07 03:49:58 <phantomcircuit> they shouldn't since those aren't valid in litecoin
106 2013-04-07 03:49:59 <Diablo-D3> to prevent that
107 2013-04-07 03:50:19 <Diablo-D3> [01:39:55] <warren> there are a few alts that begin with "1"
108 2013-04-07 03:50:26 <warren> Diablo-D3: yes =(
109 2013-04-07 03:51:08 <phantomcircuit> https://github.com/litecoin-project/litecoin/blob/master/src/base58.h#L275
110 2013-04-07 03:51:13 <phantomcircuit> oh
111 2013-04-07 03:51:18 <phantomcircuit> well
112 2013-04-07 03:51:23 <phantomcircuit> litecoin isn't one of them though
113 2013-04-07 03:51:38 <phantomcircuit> except SCRIPT_ADDRESS is the same in litecoin and bitcoin
114 2013-04-07 03:51:40 <phantomcircuit> which is broken
115 2013-04-07 03:52:15 <phantomcircuit> either way it should be impossible to send bitcoins to a litecoin address
116 2013-04-07 03:52:58 <phantomcircuit> https://github.com/bitcoin/bitcoin/blob/master/src/base58.h#L270
117 2013-04-07 03:53:05 <Diablo-D3> phantomcircuit: I agree it should be impossible
118 2013-04-07 03:53:08 <phantomcircuit> see IsValid
119 2013-04-07 03:53:23 <phantomcircuit> so you'd have to send the bitcoins from a ridiculously broken client
120 2013-04-07 03:56:25 <peawormsworth> phantomcircuit: he was sending from btc-e to an address provided by btc-e (internally). I expect btc-e will probably have to sort this out... if the transfer was actually an internal book change.
121 2013-04-07 03:56:46 <Diablo-D3> yeah
122 2013-04-07 03:56:50 <Diablo-D3> if thats true
123 2013-04-07 03:56:50 <peawormsworth> I hope that makes sense... i dont know how btc-e operates at all.
124 2013-04-07 03:56:54 <Diablo-D3> its just something they fucked up
125 2013-04-07 03:57:04 <Diablo-D3> he should try sending ltc to an internal btc address
126 2013-04-07 03:57:07 <realazthat> gmaxwell: gettxout returns nothing for me though; do I need a setting turned on?
127 2013-04-07 03:57:08 <Diablo-D3> and then cash out ;)
128 2013-04-07 03:57:23 <peawormsworth> ha
129 2013-04-07 03:58:13 <gmaxwell> realazthat: bitcoind gettxout 2accb7d1d1bb44f4b4f44f23e823433503a414e4aab86a4e01aa04cb3109dce3 1
130 2013-04-07 03:58:43 <realazthat> ah got something
131 2013-04-07 03:58:51 <gmaxwell> realazthat: gettxout only returns on .. things in the unspent output set.. so they've got to be unspent.
132 2013-04-07 03:59:20 <realazthat> aha
133 2013-04-07 03:59:29 <realazthat> I tried a few, I figured one of them would be unspent
134 2013-04-07 04:00:58 <realazthat> gmaxwell: if two outputs have the same address though; this would only get one of them right?
135 2013-04-07 04:01:22 <realazthat> same address/challenge
136 2013-04-07 04:02:10 <gmaxwell> realazthat: huh? this returns a single output exactly, indexed by its txid and output index.
137 2013-04-07 04:02:21 <realazthat> yeah ok, I thought so
138 2013-04-07 04:02:22 <gmaxwell> it's not a lookup by address, we don't have one of those (yet)
139 2013-04-07 04:03:53 <realazthat> kk ty for your tmie
140 2013-04-07 04:03:58 <realazthat> time*
141 2013-04-07 04:11:04 <gmaxwell> sipa: any idea why valgrind says all the ECKEY objects aren't getting freeed after a stop? http://pastebin.com/DTHQ2FAk
142 2013-04-07 04:19:24 <realazthat> gmaxwell: do you prefer to additionally have the btc amounts aligned at the decimal?
143 2013-04-07 04:21:27 <gmaxwell> I think thats useful.
144 2013-04-07 04:41:41 <setkeh> does any one here know how to use curl to grab the json from the BTC-E API iknow https://btc-e.com/tapi/ i just dont knowhow toput the secret and the api onto it
145 2013-04-07 04:55:15 <rmwb> does somebody have to intervene to increase hashing difficulty or is it automatically determined by the software monitoring new block frequency?
146 2013-04-07 04:57:19 <realazthat> automatic
147 2013-04-07 04:57:32 <Perlboy> So before 0.8 bitcoind was berkeleydb
148 2013-04-07 04:57:53 <Perlboy> now since 0.8 it's leveldb?
149 2013-04-07 04:57:55 <Perlboy> is that right?
150 2013-04-07 04:59:18 <rmwb> cheers, realz
151 2013-04-07 04:59:27 <gmaxwell> rmwb: it's automatically set by the network, this is very fundimental to how bitcoin works. :)
152 2013-04-07 05:00:10 <gmaxwell> Perlboy: for the blockchain indexes, the wallet still uses BDB. 0.8 fundimentally rewrote how the databases are handled, it's not just replacing one database layer for another.
153 2013-04-07 05:00:23 <rmwb> I should read Satoshi's paper that is sitting on the desk beside me, shouldn't I...
154 2013-04-07 05:00:43 <Perlboy> gmaxwell, i guess my main question is, can I still use berkeleydb to setup 'watchers' on addresses?
155 2013-04-07 05:01:05 <gmaxwell> Perlboy: huh??!
156 2013-04-07 05:01:56 <Perlboy> gmaxwell, if I want to setup offline wallets but maintain their transaction history using a hot wallet/live bitcoind
157 2013-04-07 05:02:13 <Perlboy> (which bitcoin armory has implemented but i'm trying to avoid that as a dependency)
158 2013-04-07 05:02:21 <Perlboy> what's the easiest way to do it?
159 2013-04-07 05:02:25 <Perlboy> well actually
160 2013-04-07 05:02:37 <Perlboy> what's 'the way' to do it using only what bitcoind gives me
161 2013-04-07 05:04:14 <gmaxwell> Perlboy: easiest way to do that is to encrypt the wallet on the offline machine, then take the encrypted wallet to the online machine and _never decrypt it online_
162 2013-04-07 05:04:27 <Perlboy> gmaxwell, i'm not talking about spending.
163 2013-04-07 05:04:34 <Perlboy> i'm talking about monitoring receives to it.
164 2013-04-07 05:05:06 <gmaxwell> sure.
165 2013-04-07 05:05:26 <Perlboy> pre 0.8 i could have just used the various scripts on github that allowed for exporting the transaction db and monitoring that.
166 2013-04-07 05:05:33 <gmaxwell> and as I said??? put a copy of it online but encrypted, then you can montor recieves to it.
167 2013-04-07 05:05:36 <Perlboy> but now.. not sure about their compatibility.
168 2013-04-07 05:05:51 <Perlboy> gmaxwell, that doesn't help me if I have 50 wallets to monitor :)
169 2013-04-07 05:05:58 <Perlboy> and yes, 50 unique keys.
170 2013-04-07 05:06:04 <gmaxwell> Perlboy: "exporting the transaction db" huh? I don't know what you're talking about there.
171 2013-04-07 05:06:04 <Perlboy> not 50 addresses 1 wallet.
172 2013-04-07 05:06:32 <Perlboy> https://github.com/gavinandresen/bitcointools#readme
173 2013-04-07 05:06:35 <gmaxwell> if you want you can just copy over the blockchain. thats not any different, but ... perhaps it would help if you link to the tools you're already using.
174 2013-04-07 05:06:44 <Perlboy> essentially that. without the obsolesence :)
175 2013-04-07 05:06:50 <gmaxwell> and what were you doing with that?
176 2013-04-07 05:07:10 <Perlboy> dbdump.py --search-blocks=15VjRaDX9zpbA8LVnbrCAFzrVzN7ixHNsC
177 2013-04-07 05:07:10 <Perlboy> Print out all blocks involving transactions to the Bitcoin Faucet:
178 2013-04-07 05:07:14 <Perlboy> essentially. that.
179 2013-04-07 05:07:21 <Perlboy> listing all transactions for a specific address.
180 2013-04-07 05:07:47 <gmaxwell> ah. hm. since that should just read the block files it may work even on the new stuff the block files are not in a database. I haven't tried.
181 2013-04-07 05:08:04 <gmaxwell> Though thats not really a secure way to see activity, since IIRC it will tell you about transactions which are in dead forks.
182 2013-04-07 05:12:38 <Perlboy> gmaxwell, so what would be the 'secure way' ?
183 2013-04-07 05:16:08 <Insu> hmm trc gave me a idea
184 2013-04-07 05:16:15 <Insu> ACTION runs off with some gassoil in search of asicminer's farm
185 2013-04-07 05:17:56 <gmaxwell> sipa: ::sigh:: 2479 I'm super glad someone is actually writing bandwidth management code instead of just complaining. Super not glad that it appears to work by dos attacking the network (it looks like it just stops relaying you blocks when it reaches its limit) ... and sad that the "ibd logic must be fixed first" comments were ignored.
186 2013-04-07 05:23:50 <K1773R> gmaxwell: i hope this will be implemented
187 2013-04-07 05:24:32 <gmaxwell> K1773R: because you want bitcoin to fall over and stop working for people?
188 2013-04-07 05:25:12 <Perlboy> gmaxwell, so, reading block chain as berkeley db should be possible
189 2013-04-07 05:25:14 <K1773R> no, should be tested first of course. its a long request feature for a long active issue
190 2013-04-07 05:25:18 <gmaxwell> If you want to reduce your bandwidth usage, just set listen=0 and this will remove most of it without presenting any hazard to the network.
191 2013-04-07 05:25:21 <Perlboy> and there isn't a lot of alternatives?
192 2013-04-07 05:25:44 <gmaxwell> Perlboy: the blockchain itself isn't a BDB and has never been one.
193 2013-04-07 05:26:33 <Perlboy> i was referencing this: https://github.com/gavinandresen/bitcointools#readme
194 2013-04-07 05:26:47 <Perlboy> which while i haven't read his code appears to read bdb's?
195 2013-04-07 05:35:36 <lobhater> hello all
196 2013-04-07 07:46:50 <realazthat> Perlboy: you can iterate the blockchain using bitcoind
197 2013-04-07 07:46:55 <realazthat> the RPC api
198 2013-04-07 07:47:01 <realazthat> and do w/e you want with the results
199 2013-04-07 07:47:28 <realazthat> however, I am not sure how you'd handle it if bitcoin decided to switch at a fork
200 2013-04-07 07:48:00 <realazthat> if you just want to do this once,
201 2013-04-07 07:48:08 <realazthat> you can get your results, though it would take some time
202 2013-04-07 07:51:28 <jrmithdobbs> realazthat: load the chain on a disconnected node and then query it
203 2013-04-07 07:51:30 <jrmithdobbs> basically
204 2013-04-07 07:51:36 <Belxjander> Maybe a dumb question.... would it be valid to run an exchange using a localized blockchain?
205 2013-04-07 07:52:04 <Belxjander> Maybe a dumb question.... would it be valid to run an exchange using a localized blockchain?
206 2013-04-07 07:52:54 <jrmithdobbs> if you mean a merge mined chain to publicly log and verify the exchanges' action in some way
207 2013-04-07 07:53:04 <jrmithdobbs> no it's not stupid, but it's not overly simple either
208 2013-04-07 07:55:10 <Belxjander> I was more thinking they exchange itself running an internal blockchain with minor modifications and only pushing bulk purchase/sale orders to the master blockchain of bitcoin...
209 2013-04-07 07:56:48 <Belxjander> That would allow the minor? Changes needed for account details to run always to a fixed decimal place with any overspill reverting to a slushing fund as exchange fee
210 2013-04-07 07:57:02 <tcatm> What's the point of an internal blockchain?
211 2013-04-07 07:59:05 <Belxjander> To handle internal user acct transactions and allow for relocation of overspill after 8 decimal places to the slush acct
212 2013-04-07 07:59:17 <gmaxwell> Belxjander: you just use a database for that. :P
213 2013-04-07 07:59:35 <gmaxwell> tcatm: though there are interesting things you can do in this space, e.g. you could prove an exchange wasn't fractional reserve.
214 2013-04-07 07:59:35 <jrmithdobbs> ya you don't gain much
215 2013-04-07 07:59:48 <jrmithdobbs> ya that's what I thought he was going for
216 2013-04-07 08:00:12 <Belxjander> Gmaxwell: more wanting the anti-hacking properties of the blockchain
217 2013-04-07 08:00:27 <jrmithdobbs> then you don't understand how/why it works
218 2013-04-07 08:00:30 <gmaxwell> Belxjander: blockchains are not magic. :P
219 2013-04-07 08:00:38 <tcatm> gmaxwell: Wouldn't signatures suffice for that (with some external signed time source or something)?
220 2013-04-07 08:01:18 <gmaxwell> tcatm: no, you can just lie lie lie. "I have 100 in balances, and here is proof that I hold 100 in bitcoins" (but really you have 10000 in balances)
221 2013-04-07 08:01:22 <Belxjander> Jrmithdobbs definitely don't understand it beyond an emotional level
222 2013-04-07 08:01:26 <gmaxwell> tcatm: What you do is arrange all customer's balances in a merkle tree where every node also as the sum of its child balances. Then you publish the hash root with the sum very publically.
223 2013-04-07 08:01:47 <jrmithdobbs> Belxjander: what does that even mean
224 2013-04-07 08:01:49 <gmaxwell> tcatm: then when someone logs in you give them the fragment that allows them to prove their balance is in the public hash root.
225 2013-04-07 08:01:55 <jrmithdobbs> how do you understand math on an emotional level
226 2013-04-07 08:02:25 <gmaxwell> tcatm: and then you do signmessages on the bitcoins held to show they match your balances. tada.
227 2013-04-07 08:02:32 <tcatm> gmaxwell: Ah that would work, but it's still not a block chain.
228 2013-04-07 08:03:08 <Belxjander> Jrmithdobbs I only have an approximate feeling of how thish works despite not grokking the concept with inyellectual understanding
229 2013-04-07 08:03:10 <gmaxwell> tcatm: you can do this with a blockchain too, at least one that hash the utxo set in a commited sum-merkle tree.. but .. why, the rest would be pointless. :)
230 2013-04-07 08:03:29 <jrmithdobbs> tcatm: you store the progression in a chain that's merge mined is how a blockchain could be utilized
231 2013-04-07 08:03:46 <gmaxwell> tcatm: I was responding to the thing I thought Belxjander wanted to accomplish, not the mechenism which seemed obviously wrongheaded to me. :P
232 2013-04-07 08:03:52 <jrmithdobbs> gmaxwell: i don't know that it'd be entirely useless
233 2013-04-07 08:04:12 <jrmithdobbs> gmaxwell: variations could be used for a p2market, for instance
234 2013-04-07 08:04:30 <jrmithdobbs> (you need a bit more, obviously)
235 2013-04-07 08:05:03 <Belxjander> Gmaxwell a second blockchain for a distributed "exchange" service buffering the master blockchain from day-to-day transactions
236 2013-04-07 08:05:33 <gmaxwell> Belxjander: blockchains are a pretty inefficient way to track transactions... I'm not sure why'd you'd invoke one where it isn't really needed.
237 2013-04-07 08:05:59 <gmaxwell> Belxjander: if you have to trust the exchange service to hold the non-digital assets and remove counterparty risk, then why not also trust them to do the other recordkeeping required?
238 2013-04-07 08:06:12 <Belxjander> Gmaxwell sometimes the efficient way is not optimal
239 2013-04-07 08:07:33 <Belxjander> Gmaxwell working on the assumption standard security is breached
240 2013-04-07 08:07:35 <gmaxwell> Belxjander: I did say "why not" for a reason there. ... like ... if you're going to suggest someone do something inefficient there ought to be a good reason. :P
241 2013-04-07 08:07:50 <gmaxwell> Belxjander: blockchains don't create security themselves.
242 2013-04-07 08:08:45 <jrmithdobbs> Belxjander: i think what you're misunderstanding why the bitcoin blockchain works for bitcoin
243 2013-04-07 08:09:13 <jrmithdobbs> Belxjander: you just want an internal ledger with authenticated communications, you gain nothing using a blockchain for that scenario
244 2013-04-07 08:09:13 <realazthat> what you really need is 30 oompa loompas ferrying bitcoins to and from the blockchain :P
245 2013-04-07 08:09:17 <Belxjander> Gmaxwell I know I'm a blockhead about this which is why I asked here...
246 2013-04-07 08:09:28 <jrmithdobbs> in fact, you probably end up with something less secure than doing it any other way
247 2013-04-07 08:09:44 <jrmithdobbs> (in most cases)
248 2013-04-07 08:14:12 <Belxjander> Sorry about bouncing...
249 2013-04-07 08:17:46 <Belxjander> I was thinking of the master international blockchain and having a local blockchain system with seperation of purpose and only seperating the services using a volunteer member directory for the localized service
250 2013-04-07 08:19:17 <Belxjander> Distribute out the local service to participants as well as internal network
251 2013-04-07 08:42:27 <andkore> hmm, is there no channel for dev involving bitcoin, but not bitcoin development itself?
252 2013-04-07 08:43:05 <andkore> the level of discourse in #bitcoin is... not high "<pZombie> throughout history those behind the global banking mafia have used many as tools. Their strategy is to make certain groups think they are special and separate them from others by subjecting groups to isms which are contradicting and conflicting. Divide and conquer"
253 2013-04-07 08:44:35 <muhoo> andkore: generally, discussions involving development of bitcoin seem to happen here
254 2013-04-07 08:45:38 <muhoo> andkore: and there's another channel somewhere that the blue-sky, more far-reaching, less mundane, more speculative dev discussions happen on
255 2013-04-07 08:46:22 <andkore> muhoo: lol. confusing, but ok
256 2013-04-07 08:47:01 <muhoo> andkore: don't worry. just ask what you have to ask.
257 2013-04-07 08:47:19 <muhoo> if you get an answer, or discussion, fine. if it's OT, someone will ding you and tell you where to take the discussion.
258 2013-04-07 08:48:01 <KOLANICH> hi all
259 2013-04-07 08:48:20 <andkore> ok. are there currently any payment services set up that people can use to buy things without manually making a transfer from their wallet?
260 2013-04-07 08:48:42 <andkore> like they put money there and then merchants using the service can receive money
261 2013-04-07 08:48:47 <KOLANICH> i think it is secure to store not all transaction history
262 2013-04-07 08:48:49 <andkore> I'm kind of out of the loop
263 2013-04-07 08:49:22 <KOLANICH> but only the most recent part of it
264 2013-04-07 08:50:33 <KOLANICH> a sends money to b
265 2013-04-07 08:50:34 <KOLANICH> b - to c
266 2013-04-07 08:50:36 <KOLANICH> so we dont need to store a to b transaction
267 2013-04-07 08:55:36 <sipa> gmaxwell: strange, i wonder why that didn't happen for me; afaik valgribd reports 0 bytes lost for me
268 2013-04-07 08:59:34 <wumpus> <andkore> hmm, is there no channel for dev involving bitcoin, but not bitcoin development itself? <- not that I know of, may be a good idea
269 2013-04-07 09:00:14 <wumpus> this channel is focused mainly on bitcoin *client* dev, not so much the ecosystem around it
270 2013-04-07 09:00:35 <andkore> yeah that's what I figured
271 2013-04-07 09:00:44 <andkore> bitcoin is just a shithole
272 2013-04-07 09:00:48 <andkore> and a noisy one
273 2013-04-07 09:00:52 <andkore> #bitcoin*
274 2013-04-07 09:01:02 <wumpus> yes, it is, but it keeps the trolls out of here I think that's the point :)
275 2013-04-07 09:01:20 <andkore> lol
276 2013-04-07 09:01:43 <sipa> apart from some occasional spam, we've been quite lucky regarding trolling here :)
277 2013-04-07 09:02:07 <wumpus> indeed
278 2013-04-07 09:04:30 <iwilcox> andkore: #bitcoin-tech
279 2013-04-07 09:05:01 <andkore> iwilcox: thanks
280 2013-04-07 09:05:12 <andkore> 'friendlier than #bitcoin' is an encouraging promise
281 2013-04-07 09:05:13 <andkore> lol
282 2013-04-07 09:05:19 <iwilcox> At least, that's the idea there. Less spam for here, less noise than #bitcoin.
283 2013-04-07 09:06:45 <iwilcox> I'd like it to have the blessing of #bitcoin-dev folk though; midnightmagic didn't seem to keen on the idea of another channel
284 2013-04-07 09:07:23 <Insu> 3 programmers in a row that declined, no one wants to work with this code base :(
285 2013-04-07 09:07:38 <iwilcox> I just feel many technical-but-not-client-dev things would get in the way here. This channel is pretty low-level and businesslike.
286 2013-04-07 09:09:22 <sipa> i don't mind having the "how do i integrate bitcoin in my l33t php site??!?" questions elsewhere, if that is what you mean :)
287 2013-04-07 09:10:24 <wumpus> yes, if this channel extended to that scope it's going to be much too busy
288 2013-04-07 09:10:40 <iwilcox> Well, yeah, API stuff like that. As MM said, #bitcoin can often be a "cesspool" of religious/political debates. 1300 nicks.
289 2013-04-07 09:11:36 <andkore> sipa: too close to home...
290 2013-04-07 09:11:43 <wumpus> on the other hand API integration is very important, so it has to have a place
291 2013-04-07 09:12:21 <wumpus> and #bitcoin is not that place, it's the go-to place for hypers and bashers, not for people really trying to make something
292 2013-04-07 09:12:40 <andkore> yeah, exactly. I'll check out bitcoin-tech
293 2013-04-07 09:12:43 <wumpus> so agree with the #bitcoin-tech idea
294 2013-04-07 09:16:40 <iwilcox> Straying off-topic now but I'd really prefer a #bitcoin-newbies or similar to be people's first port of call. #bitcoin is embarrassing for the community at times.
295 2013-04-07 09:20:26 <wumpus> good luck with that, but people are really going to join #bitcoin first as it's the most straightforward, creating a channel is more of a psychological and marketing thing than technical, and I think every newbie channel will change into #bitcoin, unless you find some ops that are seriously into providing organized tech support
296 2013-04-07 09:21:35 <sipa> and withstand the shitstorm rhat breaks lose if they decide to +m the channel
297 2013-04-07 09:21:41 <wumpus> (and very strict in kicking out people goofing off into arguments)
298 2013-04-07 09:22:21 <wumpus> heh
299 2013-04-07 09:37:16 <Scrat> +1 #bitcoin-tech, a great deal of questions here are integration related
300 2013-04-07 09:37:21 <Scrat> which is kinda offtopic
301 2013-04-07 09:38:51 <wumpus> yep, the challenge will be getting people there, it's currently kind of small, so let's start refering people with integration questions there
302 2013-04-07 09:45:07 <gmaxwell> I do worry that bitcoin will become even more of a cesspool if more gets moved out of it.
303 2013-04-07 09:46:35 <wumpus> that's true, but I don't think there is solution for that, there is demand for a cesspool so it is always moving somewhre
304 2013-04-07 09:49:40 <diki> Are compressed public keys preferred in order to reduce blockchain size?
305 2013-04-07 09:52:54 <sipa> diki: indeed
306 2013-04-07 09:53:12 <diki> does it require more computational power to compute them?
307 2013-04-07 09:53:37 <wumpus> only to uncompress them afaik
308 2013-04-07 09:53:47 <wumpus> you need to derive the y coordinate again
309 2013-04-07 09:53:53 <Scrat> arent all transactions currently on the network besides SD using compressed keys?
310 2013-04-07 09:54:23 <sipa> compressing is trivial
311 2013-04-07 09:54:44 <sipa> you need to decompress before signature validation, though
312 2013-04-07 09:55:27 <sipa> wumpus: plan is for a 0.8.2 based on current head, before we do the larger wallet changes in 0.9
313 2013-04-07 09:55:43 <sipa> well, that's my understanding at least
314 2013-04-07 09:57:06 <wumpus> huh?!
315 2013-04-07 09:57:28 <wumpus> why call it 0.8.2 then?
316 2013-04-07 09:57:51 <sipa> why not?
317 2013-04-07 09:58:09 <wumpus> right, ok...
318 2013-04-07 09:58:44 <sipa> gmaxwell: i don't understand how it's possible that EC_KEY_free is not called...
319 2013-04-07 10:00:11 <sipa> wumpus: i think the wallet improvements (multiwallet, watch-only, coin control, maybe deterministic, ...) are a nice target for 0.9, but current head has improvements enough for an own release despite that
320 2013-04-07 10:00:58 <wumpus> yes I suppose that'd be possible, but then I'll delay the qt5 pull until then
321 2013-04-07 10:01:01 <Scrat> sipa: care to list current head improvements?
322 2013-04-07 10:01:10 <wumpus> I really think we should discuss release policies etc on the mailing list, so everyone is up to date
323 2013-04-07 10:01:21 <sipa> yeah, agree
324 2013-04-07 10:01:38 <sipa> i'll send a mail
325 2013-04-07 10:01:54 <wumpus> for example, I pulled CodeShark's gui refactor for multiple wallets, which I wouldn't have done if I knew another 0.8.x was about to be spun off
326 2013-04-07 10:02:48 <CodeShark> oh, speaking of which, were you able to solve those issues, wumpus? I've been so incredibly busy this week - I'm sorry I didn't get a chance to look
327 2013-04-07 10:02:51 <wumpus> most kinks in it have been worked out now, though, so it's not a big issue
328 2013-04-07 10:03:24 <wumpus> yes we were able to resolve them, but I'd appreciate it if you could stand by yourself next time we merge a big thing by you :)
329 2013-04-07 10:03:52 <CodeShark> I would have liked to do a little more testing prior to merging - but last week I had so much stuff going on
330 2013-04-07 10:04:22 <CodeShark> you asked me to rebase it, so I did - that didn't mean I had actually done thorough testing :p
331 2013-04-07 10:04:28 <wumpus> I did some testing, but I don't usually test things such as export, some things are always bound to be found only after merging :p
332 2013-04-07 10:04:40 <wumpus> that's true, I'm not blaming anything on you
333 2013-04-07 10:06:48 <mughat_2> how to join the mailing lists
334 2013-04-07 10:06:50 <mughat_2> ?
335 2013-04-07 10:10:33 <wumpus> http://sourceforge.net/p/bitcoin/mailman/?source=navbar
336 2013-04-07 10:15:34 <joeykrim> thx for the link wumpus
337 2013-04-07 10:17:09 <sipa> Scrat: -walletnotify, -alertnotify, network/memory usage improvements, testnet seeds, a few small block validation improvements, may 15 changes
338 2013-04-07 10:17:55 <sipa> (not including any GUI changes, i'm less up-to-date with those)
339 2013-04-07 10:18:01 <HM> "overloaded function with no contextual type information"
340 2013-04-07 10:18:04 <HM> got to love GCC
341 2013-04-07 10:18:17 <HM> the function isn't overload damnit, it's a non-member function with no template params ;-;
342 2013-04-07 10:18:21 <HM> *overloaded
343 2013-04-07 10:18:33 <sipa> Scrat: oh, and thread management cleanup, and a better progress bar, and new icons!
344 2013-04-07 10:18:38 <wumpus> we should at least merge the new icon https://github.com/bitcoin/bitcoin/pull/2477
345 2013-04-07 10:18:40 <wumpus> yep
346 2013-04-07 10:36:15 <jn> hey, i'm working on a visualization using blockchain.info's websocket api. and i want to count the total amount of bitcoins transfered
347 2013-04-07 10:36:43 <jn> but as i understand it most transaction outputs are "change" going back to the sender
348 2013-04-07 10:36:48 <jn> can i filter this out somehow?
349 2013-04-07 10:37:08 <sipa> jn: you can probably guess in many cases, but in general it's designed not to be (trivially) possible
350 2013-04-07 10:37:21 <sipa> the privacy of the system sort-of depends on not being able to find that out
351 2013-04-07 10:37:50 <jn> okay.. how would i go about guessing?
352 2013-04-07 10:38:06 <sipa> if one output is 0.1 BTC and the other is 0.03818466
353 2013-04-07 10:38:14 <sipa> what do you think the change is likely to be?
354 2013-04-07 10:38:35 <sipa> or if one output goes back to one of the prevout addresses
355 2013-04-07 10:38:46 <sipa> (a bad practice in the first place, precisely for that reason)
356 2013-04-07 10:39:19 <jn> hmm, yeah
357 2013-04-07 10:39:21 <TD> jn: yeah you can't really know that
358 2013-04-07 10:39:32 <TD> jn: besides, blockchain already offers a graph that tries to calculate it
359 2013-04-07 10:40:11 <sipa> jn: then again, it's perfectly possible to do 'mix' transactions that just mix coins from different users together, and split them out again to new address of those same users
360 2013-04-07 10:40:19 <sipa> jn: and you'll likely make a very incorrect guess for those
361 2013-04-07 10:40:30 <sipa> as technically 0 is transferred
362 2013-04-07 10:41:25 <jn> TD: http://blockchain.info/charts/estimated-transaction-volume?daysAverageString=7 that one?
363 2013-04-07 10:41:46 <jn> i think i'd better rethink my approach to this
364 2013-04-07 10:42:06 <TD> something like that
365 2013-04-07 10:42:14 <TD> what is your actual goal??
366 2013-04-07 10:42:28 <TD> this is a decent measurement of "is bitcoin being used more": http://blockchain.info/charts/n-transactions-excluding-popular
367 2013-04-07 10:42:36 <TD> http://blockchain.info/charts/n-unique-addresses
368 2013-04-07 10:42:39 <TD> that one is also not bad
369 2013-04-07 10:43:12 <jn> i want a ticker, saying X btc has changed hands since you started viewing this page
370 2013-04-07 10:44:09 <sipa> interesting from a publicity standpoint, but i guess you need a disclaimer that your numbers are based on potentially incorrect assumptions, as this data is not intended to be public
371 2013-04-07 10:46:24 <sipa> anyone an idea how many signature checks the block chain now contains?
372 2013-04-07 10:47:26 <jn> i'll whip up some code to try to estimate the actual amount transfered
373 2013-04-07 10:48:14 <lianj> sipa: estimate? 28M
374 2013-04-07 10:48:31 <sipa> lianj: seems reasonable, yes
375 2013-04-07 10:49:02 <lianj> 30m inputs, some no sig scripts, some multisig
376 2013-04-07 10:49:49 <sipa> i'll add a nSigChecks to script.cpp and do a reindex :)
377 2013-04-07 10:51:00 <lianj> :)
378 2013-04-07 10:59:31 <diki> Quickie: If I wanted to write my own program to import private keys in a wallet, by directly editing the wallet.dat file, where do I begin?
379 2013-04-07 11:00:17 <sipa> do db4.8_dump to convert the key/value pairs in the file to a usable form, edit those, and db4.8_load it into a new wallet.dat file
380 2013-04-07 11:00:39 <sipa> unless you have bindings for BDB in your language
381 2013-04-07 11:01:15 <diki> my language of choice is C, and I plan to insert, rather than edit.
382 2013-04-07 11:01:24 <sipa> still :)
383 2013-04-07 11:01:38 <sipa> you can use the BDB library directly in that case
384 2013-04-07 11:01:49 <sipa> just add the necessary key entries
385 2013-04-07 11:02:01 <diki> Is it not possible without the bdb library?
386 2013-04-07 11:02:08 <sipa> no
387 2013-04-07 11:02:27 <sipa> as even bdb changes its file format between releases
388 2013-04-07 11:02:42 <diki> that is plain stupid by bdb
389 2013-04-07 11:02:53 <sipa> the file format is really just "dump memory to a file"
390 2013-04-07 11:03:01 <sipa> well, it's ancient technology
391 2013-04-07 11:03:08 <sipa> it worked well for certain purposes in the past
392 2013-04-07 11:03:15 <sipa> but we have much more manageable solutions now
393 2013-04-07 11:04:00 <upb> like mongodb
394 2013-04-07 11:06:16 <melvster> so it's possible to have an output with a script that requires NO key to spend it, am i right to say that?
395 2013-04-07 11:06:24 <sipa> indeed
396 2013-04-07 11:06:45 <melvster> this is interesting, so i could have no key, and add a PUZZLE?
397 2013-04-07 11:06:50 <sipa> yes
398 2013-04-07 11:06:55 <melvster> wow
399 2013-04-07 11:06:56 <sipa> but any miner could steal it
400 2013-04-07 11:07:08 <lianj> "OP_TRUE" as output script is most secure
401 2013-04-07 11:07:21 <sipa> as you provide the solution to spend it, so they can create a new transaction that reuses the same solution
402 2013-04-07 11:07:26 <melvster> what else could i add other than a puzzle? additional keys?
403 2013-04-07 11:07:31 <sipa> and drop the origin spending transaction
404 2013-04-07 11:07:35 <sipa> melvster: that's called multisig
405 2013-04-07 11:08:30 <melvster> sipa: are there any good pointers to this other than the wiki, it's quite hard to get your head round ...
406 2013-04-07 11:09:06 <sipa> melvster: there is a standard OP_CHECKMULTISIG ...
407 2013-04-07 11:09:44 <sipa> which requires N successes out of a number of signatures given against M public keys given
408 2013-04-07 11:10:08 <melvster> fascinating
409 2013-04-07 11:10:14 <melvster> satoshi invented all of this?
410 2013-04-07 11:10:17 <sipa> typically, you'll use P2SH though; read BIP16 and BIP13
411 2013-04-07 11:10:30 <sipa> yes, OP_CHECKMULTISIG was present in the very first release
412 2013-04-07 11:10:41 <sipa> (though with a bug, we still have to workaround today)
413 2013-04-07 11:10:49 <sipa> BIP13 and BIP16 obviously came later
414 2013-04-07 11:11:09 <melvster> wow
415 2013-04-07 11:11:22 <melvster> impressive
416 2013-04-07 11:11:56 <TD> melvster: you could have a puzzle if it was combined with a signature, too
417 2013-04-07 11:12:00 <TD> though at that point it's unclear why you would
418 2013-04-07 11:12:22 <TD> in practice, a puzzle is probably OK unless it becomes a common thing. then indeed you might get dishonest miners who are stealing the coins
419 2013-04-07 11:12:29 <lianj> or a silly password, output = "OP_SHA256 2c26b46b68ffc68ff99b453c1d30413413422d706483bfa0f98a5e886266e7ae OP_EQUAL" input = "foo"
420 2013-04-07 11:12:39 <sipa> lianj: that's what i understand by 'puzzle'
421 2013-04-07 11:12:44 <TD> yes
422 2013-04-07 11:13:14 <TD> the issue is that the solution to the puzzle doesn't mix in the transaction hash. so the solution is valid for any transaction. the signature, on the other hand, is only valid for that transaction
423 2013-04-07 11:13:29 <lianj> true
424 2013-04-07 11:13:54 <melvster> thank you will digest
425 2013-04-07 11:14:04 <gribble> 230124
426 2013-04-07 11:14:04 <melvster> (no puns intended!)
427 2013-04-07 11:14:04 <sipa> ;;bc,blocks
428 2013-04-07 11:14:14 <sipa> melvster: hash it up first
429 2013-04-07 11:14:24 <melvster> :)
430 2013-04-07 11:15:47 <lianj> sipa: oh the wiki even mentions it as 'puzzle tx', didn't read that page in a while
431 2013-04-07 11:15:49 <wumpus> lianj: that's like a simpler take on brainwallets :P
432 2013-04-07 11:16:04 <sipa> ha
433 2013-04-07 11:16:28 <lianj> wumpus: but like TD said, only works once. as in after that ever one knows the password
434 2013-04-07 11:16:38 <melvster> can all the clients work with the whole scripting suite ... have they been unit tested? would be a shame to lose btc on a failed experiment
435 2013-04-07 11:16:40 <wumpus> oh, indeed
436 2013-04-07 11:16:57 <lianj> melvster: use testnet
437 2013-04-07 11:17:03 <melvster> ah ha
438 2013-04-07 11:17:04 <melvster> thx
439 2013-04-07 11:17:06 <wumpus> that's very good to remember
440 2013-04-07 11:17:35 <wumpus> once one person has found the pirate's stash, everyone will go looking for the rest :p
441 2013-04-07 11:17:40 <sipa> melvster: they don't have to
442 2013-04-07 11:17:46 <lianj> ACTION starts to watch for testnet puzzle txs that are later reused on mainnet
443 2013-04-07 11:18:01 <melvster> lol
444 2013-04-07 11:18:06 <sipa> melvster: as clients create their own receive addresses, they only need to accept transactions coming in to a script they support themself
445 2013-04-07 11:18:12 <jn> http://hastebin.com/hecokomija.coffee anoyne have some ideas to improve this?
446 2013-04-07 11:18:56 <melvster> sipa: example could I hash a password and then use that to transfer bitcoins to someone ... they would only need to know the password and tx number to gain the coins?
447 2013-04-07 11:19:15 <sipa> melvster: there is no client that will accept such a transaction
448 2013-04-07 11:19:24 <sipa> you'll need to write your own code to spend it
449 2013-04-07 11:19:27 <sipa> but yes, it would work
450 2013-04-07 11:19:31 <melvster> wow
451 2013-04-07 11:20:01 <sipa> there's no point in doing so
452 2013-04-07 11:20:26 <sipa> the public-key crypto we have is much better at controlling who can spend something
453 2013-04-07 11:20:52 <melvster> sure i get that, tho a password is easier to remember
454 2013-04-07 11:21:06 <sipa> then have a private key based on that password?
455 2013-04-07 11:21:09 <sipa> and send to its address
456 2013-04-07 11:21:14 <lianj> and weaker, but its a fun exercise
457 2013-04-07 11:21:20 <sipa> (btw: terrible idea)
458 2013-04-07 11:21:41 <sipa> unless you generate the password really randomly (i.e., not by a human)
459 2013-04-07 11:21:55 <sipa> at which point it is no easier to remember than any private key
460 2013-04-07 11:22:00 <melvster> true
461 2013-04-07 11:22:11 <maikie> Hi
462 2013-04-07 11:22:20 <melvster> im thinking about the possibility of an easy to remember brain wallet
463 2013-04-07 11:22:26 <sipa> melvster: don't
464 2013-04-07 11:22:35 <maikie> Habla espa?ol alguno?
465 2013-04-07 11:22:39 <lianj> Address.silly_find_key('42') => [:found, "14ZtDMZF1k2tFfVE1gzmTHMxyq8Maa6Hj9", "73475cb40a568e8da8a045ced110137e159f890ac4da883b6b17dc651b3a8049", "42"]
466 2013-04-07 11:22:50 <sipa> maikie: #bitcoin-es ?
467 2013-04-07 11:23:13 <melvster> ahh
468 2013-04-07 11:27:00 <melvster> this is the problem i really want to solve ... Alice has an RSA key pair n/m, but no bitcoin client or address ... Bob wants to transfer 1 BTC to alice, and communicate that to her so that she is the only one able to 'unlock' the sent coin
469 2013-04-07 11:27:34 <melvster> Alice publishes the public key n
470 2013-04-07 11:27:44 <melvster> but keeps the private key m secret
471 2013-04-07 11:27:47 <sipa> bitcoin scripts can't do RSA
472 2013-04-07 11:28:14 <melvster> yes i see that
473 2013-04-07 11:28:21 <melvster> but maybe some clever crypto work around
474 2013-04-07 11:28:25 <sipa> no
475 2013-04-07 11:28:56 <sipa> you can't do RSA without support for modular exponentiation of large number
476 2013-04-07 11:31:08 <melvster> what about a 256 bit rsa subkey
477 2013-04-07 11:31:13 <sipa> even that
478 2013-04-07 11:31:24 <melvster> ok thanks ... will think it over a bit more
479 2013-04-07 11:31:31 <sipa> the only crypto primitive in bitcoin is ECDSA on the secp256k1 curve
480 2013-04-07 11:31:42 <sipa> and SHA256 and RIPEMD160
481 2013-04-07 11:32:05 <sipa> lianj: 32.5M
482 2013-04-07 11:32:31 <lianj> melvster: best you can do is generate a bitcoin address out of the hashed rsa key as ecdsa privkey, so they then have a bitcoin address
483 2013-04-07 11:32:39 <lianj> sipa: ha
484 2013-04-07 11:33:09 <melvster> thanks
485 2013-04-07 11:33:16 <lianj> melvster: but dont do it ;)
486 2013-04-07 11:33:25 <melvster> ok only on test net :)
487 2013-04-07 11:34:13 <sipa> TD: so in practice i get my bitcoind (with libsecp256k1...) to do 13400 sigchecks/s - in theory this CPU should be able to do 3x-4x as much, but the synchronization overhead per-block slows things down
488 2013-04-07 11:36:04 <sipa> just writing the block index entry for each block already becomes significant compared to signature checking, it seems
489 2013-04-07 11:40:12 <maikie> Some bitcoin developer Here?
490 2013-04-07 11:40:27 <phantomcircuit> ask your question
491 2013-04-07 12:01:29 <jaakkos> what might be an easy way to dump the utxo set? (i need the output amount and block timestamp of the block including that tx)
492 2013-04-07 12:01:36 <cyberdo> I would like some testcoins if anyone has any to spare...
493 2013-04-07 12:01:42 <cyberdo> moMumSh9NzNpiLerPtRjD7XES99kzbxbtx
494 2013-04-07 12:02:17 <cyberdo> (1 or so should suffice.... I can experiment with microcoins to begin with)
495 2013-04-07 12:08:41 <TD> sipa: ok, good, i was hoping we'd get to the point where "other" dominated the profiles :)
496 2013-04-07 12:09:07 <TD> sipa: probably that's a sign it's time to stop on optimising ecdsa and that the biggest bang for buck is now elsewhere
497 2013-04-07 12:09:15 <TD> (well, assuming you get the code integrated??? would be a shame to lose it over time)
498 2013-04-07 12:13:31 <diki> Anyone managed to get oclvanitygen to work with catalyst 13?
499 2013-04-07 12:13:53 <diki> For some reason I(and many others) keep getting "Compiling kernel, can take minutes...LLVM ERROR: Cannot" with more errors when running it
500 2013-04-07 12:23:12 <Happzz> can someone send me testnet coins? ms99dz2NHfcTBaMquhsPGGXRwZd8jTBHiA
501 2013-04-07 12:30:52 <TD> sipa: when will you finalise the deterministic wallets spec?
502 2013-04-07 12:31:01 <TD> sipa: it's probably one of the next most important things to do in bitcoinj ...
503 2013-04-07 12:31:07 <TD> Happzz: check tpfaucet.appspot.com
504 2013-04-07 12:32:41 <Happzz> TD looks broken
505 2013-04-07 12:33:03 <TD> how so? it loads for me
506 2013-04-07 12:33:18 <Happzz> it doesn't send
507 2013-04-07 12:33:32 <Happzz> aw here
508 2013-04-07 12:33:38 <kermit_> 155, not bad
509 2013-04-07 12:33:58 <kermit_> 155, holy crap
510 2013-04-07 12:34:05 <TD> Happzz: you got them? i can send you some if you want
511 2013-04-07 12:34:11 <TD> but if the faucet works, there's no need
512 2013-04-07 12:34:21 <Happzz> meh i emptied it
513 2013-04-07 12:34:51 <Happzz> i guess it's so kind i gotta fill it up with some 300
514 2013-04-07 12:35:54 <TD> oh, don't empty the faucet! you don't need so many coins, really
515 2013-04-07 12:35:59 <TD> remember they are subdivisible to small amounts
516 2013-04-07 12:36:20 <Happzz> i know i know
517 2013-04-07 12:36:31 <Happzz> i took like 20 and sent in like 300
518 2013-04-07 12:36:37 <Happzz> that should hold for a while
519 2013-04-07 12:36:39 <TD> ok
520 2013-04-07 12:36:40 <TD> cool
521 2013-04-07 13:36:04 <sipa> Hello everyone,
522 2013-04-07 13:36:06 <sipa> eh
523 2013-04-07 13:36:12 <sipa> how did that get pasted here
524 2013-04-07 13:36:35 <olyd88> hello sipa
525 2013-04-07 14:04:17 <graingert> what happens if someone manages to build a block chain that forks below a checkpoint and reaches a blockheight above the current value
526 2013-04-07 14:05:14 <graingert> is the checkpoint ignored for the higher hash power block that was mined as an attack?
527 2013-04-07 14:09:35 <graingert> also who here has the highest 1/(t(d^2)) where d is distance from Winchester UK and t is number of PGP web of trust hops from the bitcoin developers
528 2013-04-07 14:10:34 <denisx> graingert: I would assume everything below the checkpoint is fixed
529 2013-04-07 14:11:37 <graingert> gmaxwell: ^
530 2013-04-07 14:12:23 <TD> the purpose of a checkpoint is to prevent re-orgs below it
531 2013-04-07 14:21:17 <gesell> is there a wiki concerning the directory and file structure for bitcoin-qt client? couldnt find one
532 2013-04-07 14:21:52 <TD> there isn't
533 2013-04-07 14:22:07 <gesell> something on bitcointalk maybe?
534 2013-04-07 14:22:19 <gesell> remember there being a discussion when 0.8 was released but cant find it now
535 2013-04-07 14:22:31 <sipa> gesell: read doc/files.txt
536 2013-04-07 14:22:37 <gesell> aight
537 2013-04-07 14:22:43 <gesell> thanks
538 2013-04-07 14:31:46 <wallet42> greetings, can i acces the blockchain data immediatly (leveldb api) ?
539 2013-04-07 14:32:02 <graingert> wallet42: not immediatly
540 2013-04-07 14:32:07 <graingert> wallet42: it will take some time
541 2013-04-07 14:32:38 <wallet42> i dont want to go trough jsonrpc->bitcoind->blockchain.dat
542 2013-04-07 14:33:39 <fishfish> hello everyone :) quick q - if the bitcoind server has been down for a while, it obviously restarts by catching up on the blockchain. Is it however immediately aware of the current highest block #? (i'm trying to programatically define if the btcd server has been downed and for how long)
543 2013-04-07 14:34:22 <sipa> wallet42: the blockchain isn't stored in leveldb
544 2013-04-07 14:34:28 <sipa> it's just a chain of blocks :)
545 2013-04-07 14:34:47 <wallet42> but there is an index right?
546 2013-04-07 14:35:00 <sipa> there is an index (i.e. which block on which disk position), and a database with unspent transaction outputs in leveldb
547 2013-04-07 14:35:06 <wallet42> or how would the client find fast the output of tx 234121234123412341234
548 2013-04-07 14:35:13 <sipa> it can't
549 2013-04-07 14:35:24 <sipa> unless you enable the optional transaction index
550 2013-04-07 14:35:25 <wallet42> hm...
551 2013-04-07 14:35:36 <sipa> (which remembers which transaction is at which disk position)
552 2013-04-07 14:35:43 <sipa> but by default, it doesn't need that
553 2013-04-07 14:35:58 <wallet42> oke i need that :)
554 2013-04-07 14:36:04 <sipa> what for?
555 2013-04-07 14:36:06 <Belxjander> wallet42: t4ls0 x calculations from the local copy of the blockchain which it has to update from the live network and re-index?
556 2013-04-07 14:36:36 <wallet42> i want to create a webservice with statistical information about the chain
557 2013-04-07 14:36:37 <sipa> fishfish: there is no 'knowing' until it has actually downloaded and verified all blocks it didn't know about
558 2013-04-07 14:36:49 <sipa> fishfish: but it guesses based on what other peers report
559 2013-04-07 14:37:34 <sipa> wallet42: leveldb is single-process however, so you may be able to write something that can fetch information from it, but only while bitcoind itself isn't running
560 2013-04-07 14:37:51 <wallet42> but since i have to walk the while chain myself
561 2013-04-07 14:37:52 <fishfish> Thanks Sipa. Do you know if it emits a flag that state that it has now successfully synced?
562 2013-04-07 14:38:02 <sipa> fishfish: there is no 'synced'
563 2013-04-07 14:38:09 <wallet42> i can do it probably faster if i build my own index
564 2013-04-07 14:38:14 <sipa> fishfish: it just fetches blocks it didn't know about yet, continuously
565 2013-04-07 14:38:16 <wallet42> thx anyway :)
566 2013-04-07 14:38:49 <sipa> wallet42: that means you risk parsing side chains as well
567 2013-04-07 14:38:58 <sipa> or invalid ones
568 2013-04-07 14:39:01 <graingert> fishfish: you can work that out easy
569 2013-04-07 14:39:03 <wallet42> i'm aware
570 2013-04-07 14:39:10 <fishfish> graingert: how? :)
571 2013-04-07 14:39:16 <sipa> fishfish: the best guess you can make is look at the timestamp of the last block
572 2013-04-07 14:39:17 <graingert> fishfish: simply check the current block height/time
573 2013-04-07 14:39:23 <graingert> fishfish: look at the current time
574 2013-04-07 14:39:27 <sipa> if it's more than a few hours ago, you're very likely behind
575 2013-04-07 14:39:36 <graingert> fishfish: subtract and divide by 10
576 2013-04-07 14:39:38 <graingert> minutes
577 2013-04-07 14:39:40 <fishfish> gotcha
578 2013-04-07 14:39:40 <wallet42> but if i take a valid block hash and walk the way back it should be fine right?
579 2013-04-07 14:39:48 <graingert> that's how many blocks you're behind
580 2013-04-07 14:39:51 <sipa> wallet42: indeed
581 2013-04-07 14:39:57 <fishfish> it's not super clean but it will work! thank you graingert, sipa
582 2013-04-07 14:40:02 <sipa> graingert: meh, too many statistical variation for that
583 2013-04-07 14:40:14 <graingert> fishfish: it is possible to ask for the block height from your peers
584 2013-04-07 14:40:17 <graingert> but they might lie
585 2013-04-07 14:40:42 <graingert> fishfish: it's you get a very raw estimate unless you are days/weeks behind
586 2013-04-07 14:40:58 <wallet42> fishfish: im not sure but i think i saw a LOC the other day that simply takes the median of all connected peers to find a value for blockcount
587 2013-04-07 14:41:13 <sipa> wallet42: yup, that's one of the heuristics
588 2013-04-07 14:41:26 <wallet42> main.cpp:51
589 2013-04-07 14:41:30 <wallet42> CMedianFilter<int> cPeerBlockCounts(8, 0); // Amount of blocks that other nodes claim to have
590 2013-04-07 14:41:37 <sipa> unless there's a fenomenal drop in hashrate, the chance of the last block being 3-4 hours old is ridiculously small
591 2013-04-07 14:41:59 <sipa> i believe that's a better heuristic that asking your peers (who may not be entirely up-to-date themself)
592 2013-04-07 14:43:34 <graingert> fishfish: basically ask everyone that claims to know, then work out a rough estimate yourself
593 2013-04-07 14:43:44 <fishfish> right, that makes sense
594 2013-04-07 14:44:23 <graingert> fishfish: also if you just want to know down-time you could ask your OS
595 2013-04-07 14:44:49 <jackass_> lol
596 2013-04-07 14:45:57 <sipa> technically, the downtime at any time you're able to ask will be 0 :p
597 2013-04-07 14:51:06 <Happzz> i need some papers on bitcoin addresses and such. i wanna make a PHP class for it.
598 2013-04-07 14:52:01 <iwilcox> Happzz: https://en.bitcoin.it/wiki/Technical_background_of_Bitcoin_addresses
599 2013-04-07 14:53:01 <HM> ACTION thinks it's unfortunate that "std::string s; s = bool(false);" is valid C code
600 2013-04-07 14:53:20 <HM> *C++ code
601 2013-04-07 14:57:29 <franl> HM, is the bool() needed? Isn't the type of false already bool?
602 2013-04-07 15:00:18 <HM> franl, it's not. it's just for emphasis on the madness
603 2013-04-07 15:00:35 <franl> Heh. ok.
604 2013-04-07 15:01:02 <HM> string has an operator=(char), but no constructor taking a single char.
605 2013-04-07 15:01:31 <HM> so they thought it was a good idea to potentially allow any integer to be assigned to a string, as a raw character
606 2013-04-07 15:01:44 <franl> G++ 4.5.3 won't compile that: error: ambiguous overload for 'operator=' in 's = false'
607 2013-04-07 15:01:58 <HM> hmm
608 2013-04-07 15:02:49 <HM> that's very strange
609 2013-04-07 15:02:55 <Happzz> is the ECDSA a randomly generated number, or does it have make some sense?
610 2013-04-07 15:02:56 <franl> But it takes this: std::string s; s = 95;
611 2013-04-07 15:03:11 <HM> false shouldn't be ambiguous
612 2013-04-07 15:03:28 <HM> oh