1 2012-08-31 00:08:05 <denisx> DBordello: miners who do not support someting other than diff-1 got too much invalids
2 2012-08-31 00:09:13 <DBordello> denisx, ah,
3 2012-08-31 00:25:20 <amiller> midnightmagic, are your databases still corrupted
4 2012-08-31 00:25:42 <amiller> i wonder what would happen if we tried to diff your database against one that someone else prepares for you
5 2012-08-31 00:27:12 <amiller> i don't remember if your db ended up totally unreadable, maybe this isn't applicable
6 2012-08-31 00:27:40 <gmaxwell> amiller: when a contradiction with the chain in found bitcoin will log what the problem is.
7 2012-08-31 00:28:05 <gmaxwell> (e.g. when the chain spends a transaction which wasn't unspent bitcoin will tell you which txn it couldn't find in the database)
8 2012-08-31 00:28:45 <amiller> if your index lost a hundred random elements, for example, you would only notice the first txn that hits a utxo in that set
9 2012-08-31 00:29:08 <amiller> so you might not know about all the ones you've got wrong
10 2012-08-31 00:29:18 <gmaxwell> Thats true.
11 2012-08-31 01:28:50 <BlueMattBot> Project Bitcoin build #51: FAILURE in 4 hr 38 min: http://jenkins.bluematt.me/job/Bitcoin/51/
12 2012-08-31 01:29:00 <BlueMatt> awww shit
13 2012-08-31 01:29:24 <BlueMatt> dammit, guess the updated block checking thing is broken...
14 2012-08-31 01:29:35 <BlueMatt> meh, Ill fix it tomorrow - no builds for now
15 2012-08-31 01:32:33 <BlueMatt> dont think any pulls were effected
16 2012-08-31 01:32:47 <BlueMattBot> Project Bitcoin build #52: ABORTED in 3 min 39 sec: http://jenkins.bluematt.me/job/Bitcoin/52/
17 2012-08-31 01:35:50 <Luke-Jr> I need a testnet3 node :/
18 2012-08-31 01:40:38 <Luke-Jr> anyone awake? can't seem to get testnet3 connections..
19 2012-08-31 01:53:21 <Luke-Jr> ACTION really would like to test BIP 34 on testnet before mainnet???
20 2012-08-31 02:09:51 <jgarzik> Luke-Jr: us4.exmulti.net:18333
21 2012-08-31 02:10:28 <Luke-Jr> thanks
22 2012-08-31 02:18:56 <Luke-Jr> jgarzik: I found a bug in either BIP 34 or bitcoind's implemetnation thereof - got a minute to discuss?
23 2012-08-31 02:19:24 <jgarzik> Luke-Jr: I always have an ear open, sure
24 2012-08-31 02:20:02 <Luke-Jr> jgarzik: OK, so the problem amounts to bitcoind expecting an exact byte sequence at the start of scriptSig
25 2012-08-31 02:20:16 <Luke-Jr> whereas I am trying to just do \\x04 + LE 32-bit height
26 2012-08-31 02:20:37 <Luke-Jr> BIP 34 doesn't specify the shortest possible serialization is required
27 2012-08-31 02:22:42 <Luke-Jr> so, which is it? <.<
28 2012-08-31 02:25:14 <jgarzik> Luke-Jr: I'm surprised it does not already produce that. What does bitcoind produce in the bytestream, for "<< nHeight"?
29 2012-08-31 02:39:38 <Luke-Jr> jgarzik: bitcoind does, but nobody uses bitcoind to produce blocks..
30 2012-08-31 02:40:12 <Luke-Jr> jgarzik: the problem is bitcoind is rejecting blocks with longer-than-shortest serialization
31 2012-08-31 02:41:18 <jgarzik> ACTION is still waiting for an answer to the open question
32 2012-08-31 02:41:43 <Luke-Jr> bitcoind produces \\x03, low 8 bits, next lowest 8 bits, then next lowest 8 bits
33 2012-08-31 02:42:15 <Luke-Jr> unless height requires a 4th octet, then \\x04 etc
34 2012-08-31 02:44:00 <Luke-Jr> so, if BIP34 merely neglects to mention that the shortest possible serialization is required, I will need to adapt Eloipool code (making it slower); if bitcoind is wrong, that should be fixed for rc2
35 2012-08-31 02:47:26 <jgarzik> Luke-Jr: you are misunderstanding how numbers are stored in scripts
36 2012-08-31 02:47:44 <Luke-Jr> oh
37 2012-08-31 02:47:46 <jgarzik> Luke-Jr: it is not the "shortest possible serialization", it is a wholly different serialization
38 2012-08-31 02:47:56 <jgarzik> Luke-Jr: numbers are bignums, serialized.
39 2012-08-31 02:48:07 <jgarzik> Luke-Jr: see script.h, push_[u]int64
40 2012-08-31 02:48:09 <Luke-Jr> \\x04\\x01\\x00\\x00\\x00 isn't valid bignum? :/
41 2012-08-31 02:49:38 <jgarzik> Luke-Jr: not according to script, no. This is the entire reason why CastToBigNum exists... to normalize this stuff
42 2012-08-31 02:50:24 <jgarzik> Luke-Jr: If you want to update bitcoind to pop a bignum off the stack and read it, rather than a byte comparison, I don't think that is too terrible though
43 2012-08-31 02:54:10 <jgarzik> script.GetOp(pc, opcode, vchPushValue)
44 2012-08-31 02:54:22 <jgarzik> int nHeight = CastToBigNum(vchPushValue).getint()
45 2012-08-31 02:54:28 <jgarzik> something like that
46 2012-08-31 02:54:50 <jgarzik> s/pop/inspect/
47 2012-08-31 02:55:22 <Luke-Jr> ok
48 2012-08-31 02:55:40 <midnightmagic> ACTION prods gmaxwell's eyeball.
49 2012-08-31 03:15:47 <Luke-Jr> hmm
50 2012-08-31 03:15:53 <Luke-Jr> I think testnet-in-a-box is broken :/
51 2012-08-31 03:17:30 <freewil> how so
52 2012-08-31 03:18:16 <Luke-Jr> getwork/getblocktemplate claim "Bitcoin is downloading blocks???" if there aren't any
53 2012-08-31 03:18:25 <jgarzik> Deepbit is smaller than slush now? btcguild is biggest known? wow, when did that happen.
54 2012-08-31 03:20:09 <Luke-Jr> when blockchain.info got crappy at identifying pools, probably :
55 2012-08-31 03:20:10 <Luke-Jr> :p
56 2012-08-31 03:20:39 <Luke-Jr> especially Deepbit ones; I think they banned bc.info's IPs
57 2012-08-31 03:23:05 <jgarzik> seems like an IP ban would result in zero results, not smaller hash counts
58 2012-08-31 03:23:37 <Luke-Jr> jgarzik: AFAIK they're using other algorithms to identify them now
59 2012-08-31 03:24:49 <freewil> Luke-Jr, you could try this, 130 block ready to go https://github.com/freewil/bitcoin-testnet-box
60 2012-08-31 03:27:27 <jgarzik> my my my
61 2012-08-31 03:27:30 <jgarzik> ERROR: ConnectInputs() : e6a136134e prev tx already used at (nFile=2, nBlockPos=685303088, nTxPos=685510477)
62 2012-08-31 03:27:39 <jgarzik> ...
63 2012-08-31 03:27:40 <jgarzik> ERROR: ConnectInputs() : 07c08d54c1 prev tx already used at (nFile=2, nBlockPos=699103095, nTxPos=699307082)
64 2012-08-31 03:28:21 <jgarzik> ERROR: ConnectInputs() : a94ca294bb prev tx already used at (nFile=2, nBlockPos=699103095, nTxPos=699317663)
65 2012-08-31 03:28:34 <jgarzik> double-spend attempts?
66 2012-08-31 03:30:24 <Luke-Jr> O.o
67 2012-08-31 03:30:28 <Luke-Jr> jgarzik: https://github.com/bitcoin/bitcoin/pull/1764 look ok?
68 2012-08-31 03:30:39 <Luke-Jr> and did I explain the reasoning well enough?
69 2012-08-31 03:34:36 <keverw> Hey. I want to store TX and Address in to database. What sizes would I use for the varchar?
70 2012-08-31 03:35:57 <keverw> like are all addresses and TXIds the same length or ???
71 2012-08-31 03:41:25 <jgarzik> Luke-Jr: commented on pull request rather than IRC
72 2012-08-31 03:41:42 <keverw> Like this one TX I have is 64 characters.
73 2012-08-31 03:42:11 <jgarzik> keverw: txid is always a 256-bit hash value. addresses vary slightly in size.
74 2012-08-31 03:42:45 <keverw> So i can store tx id as 64 then. Address are what then? Like max? isn't it like 34. I forgot.
75 2012-08-31 03:44:02 <keverw> https://en.bitcoin.it/wiki/Address
76 2012-08-31 03:44:32 <keverw> "A Bitcoin address, or simply address, is an identifier of approximately 34 alphanumeric characters" is so varchar 34 would be safe then?
77 2012-08-31 03:45:00 <Luke-Jr> keverw: varchar is not neccesarily binary-safe, FYI
78 2012-08-31 03:45:14 <Luke-Jr> keverw: IIRC the theoretical max on addresses *today* is 35
79 2012-08-31 03:45:20 <keverw> oh??? what should I use then to store addresses?
80 2012-08-31 03:45:47 <Luke-Jr> probably best to store them as addresses
81 2012-08-31 03:46:06 <keverw> MySQL does not have that type...
82 2012-08-31 03:46:14 <Luke-Jr> I made the mistake of storing as the keyhash once, which breaks with script addrs
83 2012-08-31 03:46:23 <keverw> or I don't think I see it in the list.
84 2012-08-31 03:46:23 <Luke-Jr> keverw: yes, store addresses as varchar :p
85 2012-08-31 03:46:28 <keverw> ok
86 2012-08-31 03:46:53 <jgarzik> just make each field varchar(4096) and forget it. It is a VAR char, right? :)
87 2012-08-31 03:47:13 <Luke-Jr> :P
88 2012-08-31 06:55:41 <stikonas> I noticed that Lithuanian translation in 0.6.3 is just absolutely awful and terible. I have now updated it on Transifex. So I want to ask if it is not too late to be pulled in 0.7.0.
89 2012-08-31 07:01:57 <Luke-Jr> stikonas: it's not too late, probably ping Diapolo or wumpus
90 2012-08-31 07:18:40 <Diapolo> sipa: I'm overheating my brain, so how can I supply (via -proxy parameter) a normal proxy used for IPv4 peers and one that is used for IPv6 peers ... the code in init.cpp doesn't look like that is possible.
91 2012-08-31 07:20:09 <Diapolo> Any other dev can answer that one too :)?
92 2012-08-31 07:51:12 <MC-Eeepc> "I tried once to install the software. It took forever and didn't do anything," says Robert Hughes, a senior vice president with the consulting firm Speer & Associates who's been studying Bitcoin.
93 2012-08-31 07:51:15 <MC-Eeepc> burn
94 2012-08-31 07:54:24 <_dr> studying, nice
95 2012-08-31 08:03:12 <epscy> doesn't sound like he was desperate enough to purchase heroin on milk toad
96 2012-08-31 08:05:39 <jeremias> lol
97 2012-08-31 08:05:56 <jeremias> milk toad, great site
98 2012-08-31 10:49:31 <sipa> TD: I stripped your leveldb pullrequest to just the source tree import + makefile stuff
99 2012-08-31 10:50:16 <sipa> TD: i'm not sure whether leveldb will be pulled separately or together with ultraprune, so perhaps for a while we should try to keep them synchronized
100 2012-08-31 11:14:53 <egecko> anyone looking to employee a programmer for bitcoins?
101 2012-08-31 11:20:11 <ThomasV> egecko: http://www.bitcoinjobs.com/
102 2012-08-31 11:23:59 <ErnestoJuarell> What kind of programming do u do?
103 2012-08-31 11:32:31 <TD> sipa: ok. so how much work was it to make the split db setup use leveldb? or did you not tackle that yet?
104 2012-08-31 11:33:04 <TD> bbiab
105 2012-08-31 11:39:49 <TD> back
106 2012-08-31 11:40:03 <sipa> TD: working on that now
107 2012-08-31 11:55:57 <firelegend> I have been looking for an aswer on the wiki, but so far I've found no solution. From a WIF private key, how do I obtain the Bitcoin address associated with it?
108 2012-08-31 11:57:30 <sipa> decode base58, do Elliptic Curve multiplication with base point, encode resulting point in SEC format, SHA256, RIPEMD160, base58 encode as bitcoin address
109 2012-08-31 11:59:53 <kjj_> there is a thread about generating keys in PHP. I think someone just posted code in there that covers it. and PHP is easy to read, no matter what language you intend to use
110 2012-08-31 12:02:49 <lianj> vErY_eAsY = "1"
111 2012-08-31 12:10:39 <firelegend> sipa:when you say base point, you are referring to?
112 2012-08-31 12:14:17 <TD> firelegend: G
113 2012-08-31 12:14:24 <TD> firelegend: or you can use some software to do it for you, of course
114 2012-08-31 12:14:33 <firelegend> TD:G?
115 2012-08-31 12:14:35 <TD> we are assuming you want to write software but that's not a safe assumption by us :-)
116 2012-08-31 12:14:47 <kjj_> haha. EC math by hand on 256 bit curves
117 2012-08-31 12:14:53 <TD> firelegend: G is the name of the generator point in the ECDSA specifications. it's a point on an elliptic curve, one of the chosen curve parameters
118 2012-08-31 12:15:01 <TD> i mean pre-existing software :-)
119 2012-08-31 12:15:14 <TD> if you have bitcoinj it's just a case of adding the private key to the wallet and then printing it out to the console
120 2012-08-31 12:15:26 <TD> what's WIF by the way?
121 2012-08-31 12:15:34 <firelegend> TD:wallet import format
122 2012-08-31 12:15:46 <TD> ok
123 2012-08-31 12:16:29 <firelegend> I am going to be writing it in java, and I do see base58 implementation in java for Bitcoin, which is helpful.
124 2012-08-31 12:17:07 <TD> take a look at bitcoinj.org
125 2012-08-31 12:17:09 <TD> the code is already written
126 2012-08-31 12:17:13 <gmaxwell> Luke-Jr: oh come on; we shouldn't have to optimize the software for your crappy programming languages. If you're worried about the packed heights, precompute the next hundred years worth at startup.
127 2012-08-31 12:17:34 <kjj_> and look into bouncy castle
128 2012-08-31 12:18:02 <Luke-Jr> gmaxwell: it's not an optimization, it's a bugfix.
129 2012-08-31 12:18:45 <Luke-Jr> gmaxwell: nowhere else in the protocol treats numbers this way.
130 2012-08-31 12:18:56 <gmaxwell> Luke-Jr: I'm referring to your isn't exactly cheap comment.
131 2012-08-31 12:19:21 <Luke-Jr> gmaxwell: that's just a reason why it's a practical problem
132 2012-08-31 12:20:37 <gmaxwell> Luke-Jr: canonical vs consistent is an okay thing to ponder; hypothetical performance optimizations in insanely slow programming languages is not.
133 2012-08-31 12:20:43 <Luke-Jr> gmaxwell: also note that 99% of miners use these "crappy programming languages", so being dismissing for something that requires their adoption is kinda silly
134 2012-08-31 12:21:08 <gmaxwell> Luke-Jr: irrelevant, as I pointed out if it actually was a performance problem you could just pre-compute it.
135 2012-08-31 12:21:09 <firelegend> kjj_:I am planning to write it for Android, and I think I saw some packages that already provide the tools I need.
136 2012-08-31 12:21:30 <sipa> firelegend: all code you need is probably already in bitcoinj, in that case
137 2012-08-31 12:21:58 <kjj_> firelegend: cool. I've been looking for a good app to abuse the unlimited cellular data on my tablet
138 2012-08-31 12:22:20 <Luke-Jr> gmaxwell: that may be a viable solution for the practical problem, but it would still be better to fix the bug IMO.
139 2012-08-31 12:22:36 <sipa> Luke-Jr: i don't see how a specification can be a bug?
140 2012-08-31 12:22:47 <Luke-Jr> sipa: the specification is fine
141 2012-08-31 12:22:56 <Luke-Jr> bitcoind is enforcing a rule not in it
142 2012-08-31 12:23:04 <sipa> sure, right now the implementation and specification are different
143 2012-08-31 12:23:09 <Luke-Jr> and which is inconsistent with the rest of the Bitcoin protocol
144 2012-08-31 12:23:40 <Luke-Jr> of course, these's already bigger inconsistencies in there, so meh
145 2012-08-31 12:23:44 <sipa> exactly
146 2012-08-31 12:24:02 <gmaxwell> Luke-Jr: right, and the redundant integer coding are themselves arguably a bug in the bitcoin design.
147 2012-08-31 12:24:20 <gmaxwell> s/coding/codings/
148 2012-08-31 12:24:24 <kjj_> I think Gavin has made it pretty clear that he wants to move everything to a canonical coding, as opportunities arise
149 2012-08-31 12:25:48 <Luke-Jr> kjj_: Gavin isn't supposed to be a dictator. (though he's probably justified on that particular goal)
150 2012-08-31 12:26:18 <kjj_> agreed, on both parts.
151 2012-08-31 12:26:29 <gavinandresen> I never said We Shalt Not, I said canonical encodings are less likely to open up security holes, so they should be the default.
152 2012-08-31 12:26:45 <gmaxwell> If people didn't agree with that goal, we all would have protested it before now. :)
153 2012-08-31 12:26:56 <Luke-Jr> gmaxwell: this is the first time it is affecting a protocol change ;)
154 2012-08-31 12:27:06 <Luke-Jr> up to now, it has just been IsStandard
155 2012-08-31 12:27:13 <TD> firelegend: there's already an android bitcoin wallet that talks p2p protocol. you probably want to start by looking at that
156 2012-08-31 12:27:17 <gavinandresen> No, the second. P2SH enforces a canonical encoding
157 2012-08-31 12:27:18 <gmaxwell> Luke-Jr: K. So? do you actually disagree with that goal? :)
158 2012-08-31 12:27:22 <kjj_> I'd say fix the specification
159 2012-08-31 12:27:27 <TD> firelegend: javadocs are here: http://plan99.net/~mike/bitcoinj/0.6/
160 2012-08-31 12:27:58 <Luke-Jr> gmaxwell: I don't think so, but I haven't really considered it much
161 2012-08-31 12:28:34 <Luke-Jr> anyhow, I came into this problem initially asking whether bitcoind or BIP34 had the bug; if consensus is that BIP34 needs to be fixed, that's tolerable
162 2012-08-31 12:28:59 <gavinandresen> BIP34 already says "(will be 0x03 on main net for the next 300 or so years)" but I agree we should be more explicit.
163 2012-08-31 12:29:08 <Luke-Jr> jgarzik did hold the opposite opinion though, so perhaps it should be discussed with him further first
164 2012-08-31 12:29:34 <gmaxwell> Luke-Jr: yea, I think the BIP should probably be fixed. I agree that encoding inconsistency is generally lame, but also think that non-canonical encoding is worse.
165 2012-08-31 12:29:37 <Luke-Jr> or maybe he meant to hold more of a neutral opinion *shrug*
166 2012-08-31 12:30:13 <gavinandresen> gmaxwell: what do you mean by encoding inconsistency? 0x03 <height> is consistent....
167 2012-08-31 12:30:54 <gavinandresen> It's not a standard integer encoding, but it is consistent
168 2012-08-31 12:31:06 <gmaxwell> gavinandresen: I mean with scripts generally. An implementor may need to use two encoding functions, if their default one doesn't always produce the sammest form.
169 2012-08-31 12:31:51 <gavinandresen> their default one aught to produce the most compact form, if it doesn't I'd argue they have a bug.
170 2012-08-31 12:32:21 <Luke-Jr> anyhow, I haven't slept in over 24 hours, so I'll be back later and work on dealing with this encoding if I have to then???
171 2012-08-31 12:32:39 <gmaxwell> Luke-Jr: thanks for the good spotting.
172 2012-08-31 12:34:46 <Diapolo> sipa: Thanks for your clarification that we don't currently can specify a separate proxy for IPv6 peers ^^. Am I right, that it would suffice to place that via an extra address supplied and SetProxy(NET_IPV6, addrProxy, nSocksVersion); after all needed checks?
173 2012-08-31 12:35:10 <sipa> Diapolo: yes
174 2012-08-31 12:35:22 <sipa> Diapolo: the socks version needs to be 5 for IPv6 though
175 2012-08-31 12:35:32 <Diapolo> sipa: yes I know, I learned that from you ^^
176 2012-08-31 12:35:58 <Diapolo> sipa: are you fine with -proxy6 for such a proxy?
177 2012-08-31 12:36:02 <sipa> Diapolo: ok
178 2012-08-31 12:40:41 <firelegend> TD:Actually, my application's purposes are complete. I just want to add a small failsafe like if(bs58_address != getAddressFromPrivkey(wif_privkey)) (well, something like that anyway)
179 2012-08-31 12:40:44 <sipa> gmaxwell, gavinandresen, Luke-Jr: i think clarifying that the height is encoded as a script pushing an integer, but requiring minimum length encoding to make it canonical
180 2012-08-31 12:40:47 <TD> ok
181 2012-08-31 12:41:17 <sipa> right now the BIP does mention the 0x03 + 3 bytes thing, but doesn't really motivate
182 2012-08-31 12:42:41 <gavinandresen> sipa: agreed.
183 2012-08-31 12:48:42 <firelegend> TD:Despite the fact that I already have my android app working, I pretty much do not know anything about encryption/cryptography other than(as far as I read) that public/private keys are large integers that can be represented in many ways. So this means that while I would certainly take advantage of the pre-existing code, it does not mean I understand it.
184 2012-08-31 12:49:03 <TD> ok
185 2012-08-31 12:49:10 <TD> well, look at the bitcoinj ECKey class
186 2012-08-31 12:49:21 <TD> to see how to convert private to public and then look at the Address/Base58 classes to see how to encode
187 2012-08-31 12:50:27 <firelegend> TD:I will :)
188 2012-08-31 13:01:10 <firelegend> But I would sure like to understand more about the cryptography used in Bitcoin. Right now, using the existing code in my application would mean my app has no real value and telling me to write my own implementation of what I need would be like telling me to fix a car, without knowing how to fix it(nor understanding how it works). I've looked at Vanitygen and am astounded at how the person
189 2012-08-31 13:01:41 <firelegend> s/desired code/desired result
190 2012-08-31 13:02:27 <gmaxwell> If you're looking for a learning excercise asking us to spell out each step for you here won't help much either. IIRC it's possible to implement ECDSA using only information in Wikipedia, so you might want to try that.
191 2012-08-31 13:08:44 <gavinandresen> "Require canonical, minimum-length encoding of height" -- https://github.com/genjix/bips/pull/12
192 2012-08-31 13:13:37 <gavinandresen> firelegend: http://www.certicom.com/index.php/ecc-tutorial might be a good place to start
193 2012-08-31 13:15:58 <kjj_> I always find those tutorials amusing. I've read code for doing EC math, and it sure as hell is not drawing lines on a graph
194 2012-08-31 13:17:01 <firelegend> gmaxwell:That would've made life much easier :). I've actually read the Wiki articles. I started with what connected Bitcoin to cryptography, the article on ECDSA, but it contained links to ECC which further contains links to public-key cryptography, and that article further contained terms which I did not know like discrete logarithm, prime factorization. It's like an endless circle.
195 2012-08-31 13:17:20 <gmaxwell> kjj_: different abstraction layers. If you're an x86 asm programmer and I show you some lisp code, you might say your code 'sure as hell is not running car and cdr on registers' :)
196 2012-08-31 13:18:02 <gavinandresen> yeah, highly optimizes crypto code is unlikely to bear any resemblance to the original problem.
197 2012-08-31 13:19:03 <firelegend> gavinandresen:thank you for the link, for some reason I had missed it.
198 2012-08-31 13:20:15 <kjj_> gmaxwell: heh. lots of technical subjects have that problem. digital logic tutorials always handwave over the transition from "this is how gates work" to "and this is a CPU"
199 2012-08-31 13:20:47 <kjj_> which sucks if you want to know how to make a CPU, or in this case, if you want to actually do some EC math on your own
200 2012-08-31 13:21:26 <kjj_> but again, I'll recommend the thread on PHP. there is a PHP EC math library that is pretty easy to read and follow, even for Java guys
201 2012-08-31 13:22:29 <firelegend> kjj_:I hope I can become a Java guy, my android app is indeed in java, but that is because Google have really made it easy to use their APIs.
202 2012-08-31 13:23:00 <firelegend> In truth, I wrote it without any previous knowledge of java, simply because android does everything for you.
203 2012-08-31 13:26:51 <kjj_> http://matejdanter.com/2010/12/elliptic-curve-php-oop-dsa-and-diffie-hellman/
204 2012-08-31 13:27:56 <firelegend> kjj_:thank you.
205 2012-08-31 13:28:20 <kjj_> his paper is light on how things work, but the library is linked at the bottom
206 2012-08-31 13:30:45 <kjj_> I take that back, scroll about halfway down that page to see how the code works
207 2012-08-31 13:46:56 <MC-Eeepc> ultraprune still chuggin
208 2012-08-31 13:47:05 <MC-Eeepc> cpu usage almost zero
209 2012-08-31 13:47:28 <MC-Eeepc> 175k blox
210 2012-08-31 13:49:06 <gmaxwell> MC-Eeepc: what is a blox?
211 2012-08-31 13:49:09 <gmaxwell> :P
212 2012-08-31 13:50:04 <MC-Eeepc> its my word deal with it
213 2012-08-31 13:53:40 <MC-Eeepc> its clear this is significantly faster
214 2012-08-31 13:54:03 <MC-Eeepc> took about 5 days to get this far last time
215 2012-08-31 13:54:37 <firelegend> MC-Eeepc:ABE?
216 2012-08-31 13:54:40 <maaku> my bitcoind is crashing on mac os x with just ???terminate called throwing an exception??? in the debug.log; any ideas on how I can get more info about the crashes?
217 2012-08-31 14:01:06 <MC-Eeepc> firelegend u wot m8
218 2012-08-31 14:06:53 <shamoon> are there any open source libraries that will let me read data from the blockchain? i want to see transactions going in and out of arbitrary addresses
219 2012-08-31 14:07:09 <ThomasV_> libbitcoin
220 2012-08-31 14:07:57 <shamoon> libbitcoin will let me read the blockchain?
221 2012-08-31 14:09:00 <ThomasV_> yes
222 2012-08-31 14:09:07 <shamoon> awesome
223 2012-08-31 14:11:26 <gmaxwell> The bitcoind rpc will happily let you see transactions for addresses, you just can look up by address.
224 2012-08-31 14:11:38 <shamoon> oh really?
225 2012-08-31 14:11:46 <shamoon> i can look up by any arbitrary address?
226 2012-08-31 14:13:34 <gmaxwell> GAH.
227 2012-08-31 14:13:44 <gmaxwell> I meant to say _can't_.
228 2012-08-31 14:14:08 <shamoon> gotcha
229 2012-08-31 14:17:49 <gmaxwell> shamoon: e.g. you can look at all the transactions in all the blocks.
230 2012-08-31 14:18:01 <gmaxwell> You just can't currently ask it, give me all transactions paying to address X.
231 2012-08-31 14:18:01 <shamoon> ahhh
232 2012-08-31 14:18:09 <shamoon> so i need to maintain that DB separately
233 2012-08-31 14:18:16 <gmaxwell> Right, which is what blockexplorer does.
234 2012-08-31 14:18:16 <shamoon> i can go block by block?
235 2012-08-31 14:18:21 <shamoon> gotcha
236 2012-08-31 14:19:24 <gmaxwell> Yes, getblockhash / getblock / decoderawtransaction.
237 2012-08-31 14:19:29 <shamoon> thanks
238 2012-08-31 14:19:31 <shamoon> you da man gmaxwell
239 2012-08-31 14:19:42 <shamoon> count me out for lunch
240 2012-08-31 14:23:32 <maaku> ok, it seems my bitcoind is terminating after any successfully authenticated RPC call
241 2012-08-31 14:23:41 <maaku> any ideas why this might happen?
242 2012-08-31 14:24:17 <gmaxwell> maaku: what does the log say?
243 2012-08-31 14:24:36 <maaku> terminate called throwing an exception
244 2012-08-31 14:24:42 <gmaxwell> maaku: also try running bitcoind in the foreground and see if you get any console output.
245 2012-08-31 14:24:46 <gmaxwell> pastebin the exception.
246 2012-08-31 14:24:50 <maaku> k
247 2012-08-31 14:25:21 <maaku> terminate called throwing an exceptionAbort trap: 6
248 2012-08-31 14:25:31 <maaku> still trying to figure out how to get the exception stack trace
249 2012-08-31 14:25:42 <maaku> this is on an osx server
250 2012-08-31 14:25:53 <jrmithdobbs> maaku: if you can reliably trigger it just start it up in gdb
251 2012-08-31 14:26:00 <jrmithdobbs> then you'll get a stack trace
252 2012-08-31 14:30:43 <maaku> http://pastebin.com/uTHsPTxP
253 2012-08-31 14:35:34 <gmaxwell> maaku: what are you using to call the rpc?
254 2012-08-31 14:37:01 <maaku> In this case nothing--curl with basic HTTP auth (so it should return an error). I originally encountered the problem with a cgminer calling getwork. Let me see if that gives a different backtrace...
255 2012-08-31 14:37:27 <gmaxwell> maaku: can you give me an example commandline? (obviously leave out your key)
256 2012-08-31 14:37:47 <sipa> maaku: bitcoind or freicoind?
257 2012-08-31 14:37:52 <maaku> of cgminer?
258 2012-08-31 14:38:16 <maaku> freicoind, but bitcoinrpc.cpp is unchanged
259 2012-08-31 14:39:50 <sipa> it crashes in a string constructor?
260 2012-08-31 14:39:57 <maaku> gmaxwell: command line of cgminer?
261 2012-08-31 14:40:12 <gmaxwell> maaku: of curl.
262 2012-08-31 14:40:51 <maaku> curl --basic http://freicoinrpc:<password>@127.0.0.1:6000/
263 2012-08-31 14:42:34 <gmaxwell> Works for me.
264 2012-08-31 14:43:07 <gmaxwell> Sounds like freicoin is broken, whatever it is.
265 2012-08-31 14:45:05 <maaku> that's probably the case. i'll see if i can extract a more informative stack trace. thanks for looking into it
266 2012-08-31 15:47:46 <maaku> gmaxwell: FWIW, the problem was isolated to the gitian build. build natively on mac os x and the issue went away. not sure if this affects Luke's gitian scripts for Mac OS X bitcoind
267 2012-08-31 16:27:20 <BlueMatt> ;;seen TD
268 2012-08-31 16:27:21 <gribble> TD was last seen in #bitcoin-dev 3 hours, 37 minutes, and 59 seconds ago: <TD> to see how to convert private to public and then look at the Address/Base58 classes to see how to encode
269 2012-08-31 16:45:49 <gmaxwell> Testnet3 could really use some more peers running current git.
270 2012-08-31 16:46:09 <TD> BlueMatt: yo
271 2012-08-31 16:46:13 <gmaxwell> There are only two connectable testnet nodes on IRC. (jeff and I)
272 2012-08-31 16:46:30 <gmaxwell> ACTION pokes midnightmagic
273 2012-08-31 16:48:01 <TD> ok, there should be another testnet3 node now
274 2012-08-31 16:49:05 <shamoon> gmaxwell: trying: getblockhash / getblock / decoderawtransaction.
275 2012-08-31 16:49:18 <shamoon> but what txid do i pass to decoderawtransaction?
276 2012-08-31 16:50:08 <shamoon> all the tx from getblock?
277 2012-08-31 16:52:16 <TD> trying connection 203.122.247.149:18333 lastseen=0.9hrs
278 2012-08-31 16:52:20 <TD> no route to host
279 2012-08-31 16:52:21 <TD> weird!
280 2012-08-31 16:52:50 <TD> and yet it's pingable?
281 2012-08-31 16:52:55 <TD> smells like a bug
282 2012-08-31 16:53:28 <BlueMatt> possibly the router has no route to the host which the port forward is set to
283 2012-08-31 16:53:35 <BlueMatt> (but handles pings itself0
284 2012-08-31 16:53:36 <BlueMatt> )
285 2012-08-31 16:53:42 <TD> hmm
286 2012-08-31 16:53:49 <TD> right
287 2012-08-31 16:54:08 <TD> that's a good suggestion. i forgot "no route to host" can come from icmp errors anywhere along the path
288 2012-08-31 16:54:26 <TD> looks like it didn't take long for testnet3 to get hosed :)
289 2012-08-31 16:54:44 <gmaxwell> Huh? hosed?
290 2012-08-31 16:54:45 <TD> it's very aggressive about retrying. my node is poking this IP at least once per second
291 2012-08-31 16:55:09 <gmaxwell> TD: its not that it's agressive about retrying; its that there are no other candidates. It's trying the only one it has.
292 2012-08-31 16:55:11 <TD> gmaxwell: not really hosed. seems some peers are being advertised that aren't reachable and it never gives up, so the logs get flooded with errors
293 2012-08-31 16:55:15 <TD> oh
294 2012-08-31 16:55:16 <TD> hah
295 2012-08-31 16:56:04 <gmaxwell> rc1 has a bug that makes it screw up the port number its advertising. Thus there are only two working hosts on IRC.
296 2012-08-31 16:56:29 <TD> oh
297 2012-08-31 16:56:44 <TD> i'm actually using whatever my leveldb branch was last rebased to, not exactly git head or rc1
298 2012-08-31 16:57:03 <TD> i'll switch my testnet3 node off again until i've rebased/rebuilt
299 2012-08-31 16:58:54 <BlueMatt> ACTION needs to buy about 10 mini servers and hide them in various places on campus with ethernet to have a bunch of servers...
300 2012-08-31 17:04:36 <maaku> BlueMatt: the mac mini power usage is probably small enough no one would notice ;)
301 2012-08-31 17:06:15 <BlueMatt> maaku: thats about what I was thinking, though maybe arm would be cheaper...hmm, something to think about
302 2012-08-31 17:06:39 <gmaxwell> BlueMatt: gumstix.
303 2012-08-31 17:06:44 <maaku> does bitcoind run on ARM?
304 2012-08-31 17:06:50 <gmaxwell> or varrious wallplug computers.
305 2012-08-31 17:06:52 <gmaxwell> maaku: yes.
306 2012-08-31 17:07:28 <maaku> is that out of the box, or in someone's branch?
307 2012-08-31 17:09:25 <gmaxwell> Out of the box.
308 2012-08-31 17:09:33 <maaku> cool, thx
309 2012-08-31 17:10:06 <gmaxwell> (technically you didn't provide quite enough information: arm can be BE or LE, but most arm systems today are LE. Bitcoin doesn't currently work on BE systems)
310 2012-08-31 17:11:09 <Luke-Jr> anyone know off-hand if bignum 0 is serialized as 00 or 0100 (because a bit is needed for the sign) ?
311 2012-08-31 17:11:16 <Luke-Jr> in scripts
312 2012-08-31 17:22:22 <Luke-Jr> gavinandresen: I don't suppose the ease of subtle bugs in implementing the "canonical" script number format changes your mind? :/
313 2012-08-31 17:27:04 <amiller> i explained satoshidice to someone, they said it reminded them of this: http://www.bash.org/?814243
314 2012-08-31 17:39:25 <Luke-Jr> answers from reading OpenSSL code: zero serializes to 00, everything else has a "sign" bit as bitwise-OR 80 of the last byte in the sequence, with - for both signed or unsigned numbers - an extra padding byte added if the last byte uses 80 itself
315 2012-08-31 17:40:10 <Luke-Jr> and I'm sure that explanation sucked, so I conclude canonical script number format is complicated and vulnerable to subtle implementation bugs :/
316 2012-08-31 17:40:10 <maaku> good, i was worried there might be a "-0" serialization :)
317 2012-08-31 17:40:46 <Luke-Jr> maaku: if bignum allows -0, it would corrupt the OpenSSL serialization (and simply lose the sign in Bitcoin) :/
318 2012-08-31 17:41:59 <jrmithdobbs> Luke-Jr: the fun part is there are calls to bignum that don't have the stupid signed problem
319 2012-08-31 17:42:48 <jrmithdobbs> Luke-Jr: the calls satoshi chose to use to implement weren't the right ones. there's no need to manually parse the bignum like bitcoind does, EVER, there's an openssl function to convert it to an unsigned char * and drop all the encoding crap
320 2012-08-31 17:43:06 <Luke-Jr> jrmithdobbs: I know.
321 2012-08-31 17:43:26 <jrmithdobbs> that -0 thing is annoying
322 2012-08-31 18:43:30 <Evilmax> bitcoin is falling again?
323 2012-08-31 18:44:51 <copumpkin> I don't think -dev cares, unless the protocol is :)
324 2012-08-31 18:45:15 <ciscoftw> whats the deals with bitcoind 'getpeerinfo' ??? not working, running version "version" : 60300
325 2012-08-31 18:46:06 <sipa> ciscoftw: ?
326 2012-08-31 18:46:19 <sipa> 0.6.3 doesn't have getpeerinfo
327 2012-08-31 18:46:28 <sipa> it's a new RPC call for 0.7.0
328 2012-08-31 18:47:01 <ciscoftw> :(
329 2012-08-31 18:47:05 <ciscoftw> where do i get v7?
330 2012-08-31 18:48:09 <maaku> http://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.7.0/test/
331 2012-08-31 18:48:23 <ciscoftw> awesome thanx
332 2012-08-31 18:48:42 <kjj_> 7 is still a release candidate. if you get it, you are testing it before a full release
333 2012-08-31 18:48:52 <maaku> it's release candidate, and might have bugs so beware
334 2012-08-31 18:49:08 <kjj_> been rock solid on my box so far, but just wanted you to know
335 2012-08-31 18:49:30 <ciscoftw> fair enough, not like i use it for anything super major... only serv's for 800Mh/s
336 2012-08-31 18:49:45 <ciscoftw> so not a big deal if its down for couple hrs/days
337 2012-08-31 18:50:05 <maaku> well, backup your wallets is all we're saying
338 2012-08-31 18:50:54 <ciscoftw> [notroot@fc15-btcserv 32]$ ./bitcoind getbalance
339 2012-08-31 18:51:17 <kjj_> ah ha! I think I just figured out why my wallet backup cron wasn't working
340 2012-08-31 18:51:19 <ciscoftw> not to worried about wallet ...sadly
341 2012-08-31 19:01:30 <ciscoftw> 'getpeerinfo' is such a sick addition
342 2012-08-31 19:02:02 <sipa> This is obviously some strange usage of the word 'sick' that I hadn't been previously aware of.
343 2012-08-31 19:03:04 <Diapolo> ^^
344 2012-08-31 19:04:13 <BlueMatt> gmaxwell: next question, would bitcoin run on a 800mhz arm a8...
345 2012-08-31 19:04:50 <Diapolo> sipa: SetLimited(NET_IPV6); means don't use / disable IPv6, right?
346 2012-08-31 19:04:55 <sipa> yup
347 2012-08-31 19:16:34 <sipa> ;;bc,blocks
348 2012-08-31 19:16:35 <gribble> 196604
349 2012-08-31 19:32:59 <Diapolo> sipa: what does SetReachable(NET_TOR) as for NET_IPV4 and _IPV6 this is not used
350 2012-08-31 19:34:00 <sipa> Diapolo: done automatically from AddLocal in net.cpp
351 2012-08-31 19:34:25 <sipa> but you may well be able to reach Tor destinations without reachable on a Tor address yourself
352 2012-08-31 19:34:33 <gmaxwell> BlueMatt: I don't doubt that it would, you might need to do a bit of memory usage optimization. And it may lag a bit...
353 2012-08-31 19:34:41 <Diapolo> sipa: thanks for that hint
354 2012-08-31 19:35:02 <BlueMatt> gmaxwell: or...4 rasp pis in an array, ignoring sig checking for txes relayed from each other
355 2012-08-31 19:35:12 <BlueMatt> +some memory ops as well
356 2012-08-31 19:35:46 <gmaxwell> BlueMatt: pis are less than ideal, there are arm boards with more umph $ but a higher cost per board. Though I guess you want to spread them out.
357 2012-08-31 19:36:18 <D34TH> bluematt why not a 1.4 ghz arm 10?
358 2012-08-31 19:36:21 <BlueMatt> like/
359 2012-08-31 19:36:22 <BlueMatt> ?
360 2012-08-31 19:36:25 <gmaxwell> (the pi's cpu is a seriously underpowered arm, only has one execution engine, no l2 cache, etc.. that SOC was really designed to spend most of its power budget in the DSP that you're not going to use at all)
361 2012-08-31 19:36:35 <BlueMatt> yea, the pi's cpu is really, really cheap
362 2012-08-31 19:37:01 <gmaxwell> not just cheap, but most of its gates are in the dsp. More applicable to your workload would be things like http://armdevices.net/2012/07/12/129-exynos4412-quad-core-arm-cortex-a9-development-board/
363 2012-08-31 19:37:04 <BlueMatt> reading up on gumstix dont look like a great deal either, there are better ones...
364 2012-08-31 19:37:40 <gmaxwell> yea, the gumstix are a bit overpriced, they've been around forever and for a long time were basically all there was in the really-really smal linux pc market.
365 2012-08-31 19:37:49 <D34TH> APC or CottonCandy
366 2012-08-31 19:37:51 <gmaxwell> though there isn't much as small as them even now.
367 2012-08-31 19:38:34 <D34TH> cottoncandy has an A9@1.2GHz
368 2012-08-31 19:39:06 <Luke-Jr> BlueMatt: gmaxwell: FWIW, RPi is having some USB/Ethernet issues
369 2012-08-31 19:39:17 <BlueMatt> Luke-Jr: yea, I heard about that, havent read up on it though
370 2012-08-31 19:39:18 <D34TH> ^
371 2012-08-31 19:39:29 <Luke-Jr> it's a real practical problem for miners
372 2012-08-31 19:39:57 <gmaxwell> Luke-Jr: WHY ARE PEOPLE MINING USING @#$#@@#$#@$@ rasberry-pi ::stab:: ::stab::
373 2012-08-31 19:40:05 <Luke-Jr> gmaxwell: cheap and low power
374 2012-08-31 19:40:15 <Luke-Jr> fits well with the various USB FPGAs
375 2012-08-31 19:40:19 <gmaxwell> ::stab::
376 2012-08-31 19:40:21 <BlueMatt> I have to say this 4 core a9 for 129 is looking damn good
377 2012-08-31 19:40:44 <Luke-Jr> gmaxwell: ? problem?
378 2012-08-31 19:40:51 <gmaxwell> Luke-Jr: a panda board or the alike would have been a much better choice. :P
379 2012-08-31 19:41:03 <Luke-Jr> meh, it's all proprietary ARM crap
380 2012-08-31 19:41:11 <Luke-Jr> I'd like to get it working on my MIPS router :P
381 2012-08-31 19:41:20 <gmaxwell> Luke-Jr: well the pi is more propritary than most.
382 2012-08-31 19:41:52 <gmaxwell> 9/10 of the gates on the chip are in the videocore dsp, which doesn't even have a publically documented instruction set.
383 2012-08-31 19:43:45 <copumpkin> sounds like a job for hardcore reverse engineers
384 2012-08-31 19:44:54 <Luke-Jr> copumpkin: somehow I don't know how practical that is
385 2012-08-31 19:45:08 <copumpkin> reverse engineering is good for its own sake
386 2012-08-31 19:45:08 <Luke-Jr> it's been years and nvidia drivers are still mediocre afaik
387 2012-08-31 19:45:16 <copumpkin> :)
388 2012-08-31 19:45:18 <jrmithdobbs> gmaxwell: ya i'm less than thrilled by how the pi has turned out
389 2012-08-31 19:45:22 <jrmithdobbs> as far as openness
390 2012-08-31 19:45:30 <copumpkin> Luke-Jr: clearly the reverse engineers there aren't hardcore enough
391 2012-08-31 19:45:33 <copumpkin> I said we need hardcore ones
392 2012-08-31 19:45:41 <copumpkin> no true hardcore reverse engineer...
393 2012-08-31 19:46:00 <gmaxwell> They're off with the scottsmen.
394 2012-08-31 19:46:10 <Luke-Jr> there's no such thing as a scotsman
395 2012-08-31 19:46:17 <copumpkin> I'm in england now, so I might travel up north to meet a true scotsman
396 2012-08-31 19:49:56 <Diapolo> sipa: can you try to reach my hidden service (testnet) via 36zog7xdlcb2yeog.onion
397 2012-08-31 19:51:20 <sipa> let's see
398 2012-08-31 19:51:39 <gmaxwell> If only I could addnode without shutting down!
399 2012-08-31 19:52:26 <sipa> hmm, Proxy error: connection not allowed
400 2012-08-31 19:53:03 <Diapolo> I used: -tor=127.0.0.1:9050 -externalip=36zog7xdlcb2yeog.onion -listen
401 2012-08-31 19:54:25 <sipa> did you configure a service on tor?
402 2012-08-31 19:54:30 <Diapolo> yes
403 2012-08-31 19:54:53 <gmaxwell> hidden service don't always work instantly. How long has your node been up?
404 2012-08-31 19:54:57 <Diapolo> LOL I configured main net ^^ 8333
405 2012-08-31 19:55:01 <Diapolo> just a sec
406 2012-08-31 19:55:39 <Luke-Jr> gmaxwell: BlueMatt's pullreq for that didn't get in? :<
407 2012-08-31 19:56:02 <Diapolo> changed to 18333
408 2012-08-31 19:56:15 <gmaxwell> Luke-Jr: we were too busy shedpainting it.
409 2012-08-31 19:56:21 <Luke-Jr> :/
410 2012-08-31 19:56:31 <sipa> Diapolo: it connected, and disconnected...
411 2012-08-31 19:56:33 <BlueMatt> I had a pull for network-something?
412 2012-08-31 19:56:39 <gmaxwell> BlueMatt: addnode.
413 2012-08-31 19:56:39 <Luke-Jr> ACTION blames Diapolo
414 2012-08-31 19:56:43 <BlueMatt> ahh
415 2012-08-31 19:56:43 <Luke-Jr> j/k :p
416 2012-08-31 19:56:53 <Diapolo> Luke-Jr: for what? ^^
417 2012-08-31 19:57:13 <Luke-Jr> Diapolo: not sure!
418 2012-08-31 19:57:20 <gmaxwell> well, even I'm not totally happy with the behavior of it, though no fault of bluematt the whole addnode dns interaction is weird.
419 2012-08-31 19:57:26 <Diapolo> sipa: I restarted Tor after changing the port just to be sure, can you try again?
420 2012-08-31 19:57:32 <sipa> Diapolo: already connected
421 2012-08-31 19:57:56 <Diapolo> 127.0.0.1:50085 is via Tor I guess?
422 2012-08-31 19:58:00 <sipa> yes
423 2012-08-31 19:58:16 <Diapolo> then it seems I did setup my first hidden service ^^
424 2012-08-31 19:58:23 <gmaxwell> hurrah!
425 2012-08-31 19:58:26 <sipa> \\o/
426 2012-08-31 19:58:35 <Diapolo> thanks dear sirs ^^
427 2012-08-31 19:58:43 <gmaxwell> "If Diapolo can do it then anyone can!"
428 2012-08-31 19:58:45 <gmaxwell> ACTION ducks
429 2012-08-31 19:58:46 <gmaxwell> ;)
430 2012-08-31 19:58:48 <Luke-Jr> ouch
431 2012-08-31 19:58:53 <Diapolo> 8-)
432 2012-08-31 19:58:56 <gmaxwell> (I am kidding, of course)
433 2012-08-31 19:59:02 <BlueMatt> gmaxwell: Id say that is my fault...I wrote addnode...
434 2012-08-31 19:59:19 <Diapolo> gmaxwell: I feel a little truth, but no offense ^^. You know I'm that Windows guy :-P.
435 2012-08-31 19:59:46 <BlueMatt> gmaxwell: thanks for the tip, two $150 quad-core arms on order :)
436 2012-08-31 19:59:52 <gmaxwell> well, no, you are a good test subject. jrmithdobbs, sipa, gavin, and I are the only people I know that have tested it... and we all knew a bit more about tor than you did.
437 2012-08-31 20:00:40 <Luke-Jr> ACTION knows nothing about Tor hidden services ??? but probably still too much to be a good tester XD
438 2012-08-31 20:00:42 <Diapolo> that Vidalia package is rather easy to use and the Tor.txt is fine ... I'm not sure about the novice user, but it can be done
439 2012-08-31 20:00:50 <gmaxwell> BlueMatt: Great, I look forward to hearing how they work.
440 2012-08-31 20:01:26 <gmaxwell> Diapolo: yea, I don't expect joe-average to set it up. Someday tor is supposted to get a way for applications to request hidden service setup.. and when that happens it will be much easier.
441 2012-08-31 20:01:50 <Luke-Jr> gmaxwell: I wonder if UPnP would work for that
442 2012-08-31 20:01:56 <Diapolo> sipa: as I used just -tor: it means all connection I have (incoming ones) are other users that use Tor proxies?
443 2012-08-31 20:02:14 <sipa> Diapolo: no
444 2012-08-31 20:02:45 <sipa> -tor is just "which proxy to use for outgoing tor connections"
445 2012-08-31 20:03:22 <sipa> it doesn't prevent connections to or fro; normal IPv4
446 2012-08-31 20:03:49 <Diapolo> so this is not the optimum in terms of privacy, right?
447 2012-08-31 20:04:02 <sipa> not at all
448 2012-08-31 20:04:07 <sipa> use -proxy for that
449 2012-08-31 20:04:19 <gmaxwell> perhaps in 0.8 we should add some tor icon that lights up if you have _only_ tor connections?
450 2012-08-31 20:05:03 <gmaxwell> (or rather, have at least one connection and are only able to have tor connections)
451 2012-08-31 20:05:20 <Diapolo> I'm not sure if I understood the difference here sipa, can you explain a little more detailed.
452 2012-08-31 20:05:34 <Diapolo> -tor vs. -proxy -tor vs. -proxy
453 2012-08-31 20:05:40 <sipa> Diapolo: -proxy means "use this proxy for everything"
454 2012-08-31 20:05:52 <sipa> -tor means "use this proxy for connections to .onion addresses"
455 2012-08-31 20:06:23 <sipa> -proxy -tor means "use this proxy for everything, but that one for .onion addresses"
456 2012-08-31 20:06:49 <Diapolo> so -proxy is the general thing in between in/outgoing connections, while -tor just for outgoing ones to tor
457 2012-08-31 20:06:51 <sipa> plus, -proxy disables listening by default
458 2012-08-31 20:07:21 <Luke-Jr> 507 testnet blocks found without a problem..
459 2012-08-31 20:08:06 <Luke-Jr> is there a good way to copy an address from testnet2 to testnet3? :P
460 2012-08-31 20:08:17 <Diapolo> What would you define as best in terms of privacy but allow others to allow hs connections then? -proxy -tor -listen
461 2012-08-31 20:09:00 <sipa> Diapolo: best for privacy is only outgoing connections through Tor i guess, and if you're paranoid, only to other hidden services
462 2012-08-31 20:09:14 <sipa> -proxy=<tor> -onlynet=tor
463 2012-08-31 20:09:29 <Diapolo> I'm not but was interested :).
464 2012-08-31 20:09:57 <sipa> Diapolo: oh, i've downloaded 21000 testnet blocks through tor from you
465 2012-08-31 20:10:00 <gmaxwell> you can't bootstrap that way however.
466 2012-08-31 20:10:15 <sipa> yeah you'll need some known onion peers
467 2012-08-31 20:11:14 <Diapolo> sipa: so it works good :)
468 2012-08-31 20:12:03 <Diapolo> Sipa: To extend my questions, a seperate IPv6 proxy would be for outgoing connections to IPv6 peers / addresses then right?
469 2012-08-31 20:12:13 <sipa> yes
470 2012-08-31 20:25:56 <Diapolo> I love getpeerinfo
471 2012-08-31 20:29:32 <gmaxwell> yea, gettpeerinfo is goodness.
472 2012-08-31 20:30:52 <Diapolo> Blizzard would nerf it IMHO.
473 2012-08-31 20:35:52 <sipa> ...?
474 2012-08-31 20:38:18 <Diapolo> "Nerf (video gaming), a change to a computer game that makes the "nerfed" game element less desirable"
475 2012-08-31 20:38:31 <sipa> i know that
476 2012-08-31 20:38:41 <sipa> but i don't see what blizzard has to do with getpeerinfo
477 2012-08-31 20:39:12 <Diapolo> as getpeerinfo == goodness == imbalanced in terms of gameplay ... well it was just a random though related to my old WoW phase
478 2012-08-31 20:39:18 <gmaxwell> its seems that blizzard nerfing things is a meme
479 2012-08-31 20:40:04 <Diapolo> gmaxwell: sipa: Can you take a look https://github.com/Diapolo/bitcoin/commit/ff10bde440e9a7837b1df9b338741dac126b6ade
480 2012-08-31 20:41:36 <sipa> Diapolo: i don't think just -proxy6 should imply -nolisten; but i'm not sure in what practical setting that would be used anyway, so it's hard to guess best defaults
481 2012-08-31 20:43:01 <Diapolo> well as your wish was to have a Qt setting for reaching IPv6 peers there needs to be code in the core :) what -proxy6 implies can be changed, no problem with that
482 2012-08-31 20:43:13 <Diapolo> Qt proxy setting
483 2012-08-31 20:43:25 <sipa> sure; no problem with adding -proxy6
484 2012-08-31 20:43:44 <sipa> but the Qt settings thing will directly call SetProxy, no?
485 2012-08-31 20:44:09 <sipa> so it doesn't really need -proxy6 to exist, or am I missing something?
486 2012-08-31 20:44:28 <Diapolo> it would work like it does now for Qt, just the switch should be in the core too IMO
487 2012-08-31 20:44:56 <sipa> ok, agree
488 2012-08-31 20:45:00 <Diapolo> or I missunderstand, why we need it in Qt but not the core ^^
489 2012-08-31 20:45:23 <sipa> no i agree it's good for consistency that both interfaces have the same functionality
490 2012-08-31 20:45:45 <sipa> i was just wondering whether something like -proxy6 was required to have it in Qt
491 2012-08-31 20:46:24 <Diapolo> gmaxwell or you suggested such an option a few weeks ago
492 2012-08-31 20:47:02 <Diapolo> did I remember correctly, that IPv6 requires SOCKS5?
493 2012-08-31 20:47:08 <sipa> yes
494 2012-08-31 20:49:28 <Diapolo> I'm going to remove that -proxy6 implies -listen to be disabled. Are you fine with opening that as pull or are there big bugs jumping on you?
495 2012-08-31 20:49:51 <sipa> nothing obvious, looks good
496 2012-08-31 20:50:28 <BlueMatt> bug bugs jumping on you...oh shit that would not be good, call an exterminator or something
497 2012-08-31 20:50:58 <Diapolo> I think the funny thing will be to bring that bunch of different possibilities into Qt ^^.
498 2012-08-31 20:51:03 <Diapolo> BlueMatt: :-D
499 2012-08-31 20:53:27 <sipa> Diapolo: and I hope you don't need my approval to open a pull request :)
500 2012-08-31 20:54:21 <Diapolo> Gavin scared me a few times ^^ and as he will get informed I wanted to lower my fear when opening a core-changing pull.
501 2012-08-31 20:54:59 <sipa> haha
502 2012-08-31 20:55:11 <gmaxwell> Diapolo: at worst your pull request gets closed and you get to open a new one later. No biggie.
503 2012-08-31 20:55:18 <sipa> ACTION 's first pull request took half a year to get merged
504 2012-08-31 20:57:05 <Diapolo> It was a rather big one I guess?
505 2012-08-31 20:57:32 <sipa> import/export of private keys
506 2012-08-31 20:59:01 <BlueMatt> holy shit...wtf? I wrote -logtimestamps
507 2012-08-31 20:59:08 <sipa> really?
508 2012-08-31 20:59:16 <BlueMatt> apparently
509 2012-08-31 20:59:17 <BlueMatt> https://github.com/bitcoin/bitcoin/pull/110
510 2012-08-31 20:59:18 <Diapolo> I changed a comment in the translation process ^^.
511 2012-08-31 20:59:30 <BlueMatt> going back through my own pulls is kinda funny...I dont remember doing >1/2 of them
512 2012-08-31 21:00:06 <Diapolo> So, what's that favorite alcoholic drink while coding?
513 2012-08-31 21:00:09 <kjj_> you can get timestamps in your logs?! brb
514 2012-08-31 21:00:15 <gmaxwell> hahah
515 2012-08-31 21:00:18 <BlueMatt> heh
516 2012-08-31 21:00:23 <Diapolo> hrhr that was a good one
517 2012-08-31 21:00:59 <sipa> Diapolo: don't drink and code :)
518 2012-08-31 21:01:10 <BlueMatt> sipa: awww...thats no fun :(
519 2012-08-31 21:01:17 <sipa> Diapolo: (or at least, don't push to bitcoin/bitcoin.git when drunk(
520 2012-08-31 21:01:26 <BlueMatt> well, ok yea, never push while drunk
521 2012-08-31 21:01:33 <Diapolo> I guess there would be some really great commit messages ^^.
522 2012-08-31 21:02:17 <sipa> i have IRC logs of things I said while being drunk :S
523 2012-08-31 21:02:47 <BlueMatt> sipa: heh, yea, those ones are always...fun
524 2012-08-31 21:04:21 <Diapolo> I have to be here when that happens ... things will get less techie and more philosophic.
525 2012-08-31 21:08:12 <sipa> Diapolo: it's been a few years ago, thankfully :)
526 2012-08-31 21:12:00 <Diapolo> the internet doesn't forget ;) I'm outa here for now...
527 2012-08-31 21:46:09 <jrmithdobbs> has anyone else been seeing this kind of stuff? TCP: Peer 188.222.86.47:64587/8333 unexpectedly shrunk window 1764239751:1764242671 (repaired)
528 2012-08-31 21:57:24 <jrmithdobbs> more exactly: http://pastebin.com/pkWrSppA
529 2012-08-31 21:58:25 <kjj_> harmless. google it
530 2012-08-31 22:12:56 <unclemantis> I am having trouble running the bitcoin-qt PPA http://pastebin.com/raw.php?i=zkAN7TDP
531 2012-08-31 22:13:24 <kjj_> platform?
532 2012-08-31 22:14:14 <unclemantis> Ubuntu 10.11
533 2012-08-31 22:14:27 <unclemantis> sorry 11.10
534 2012-08-31 22:14:50 <kjj_> can you pop open a shell and head to your bitcoin data directory?
535 2012-08-31 22:15:59 <unclemantis> to be honest this is the first time I am running bitcoin-qt in linux
536 2012-08-31 22:16:48 <kjj_> heh. I don't use it either, but the answer to your problem is quite possibly hiding in a log file
537 2012-08-31 22:16:59 <unclemantis> wonderful!
538 2012-08-31 22:17:06 <kjj_> can you open a shell and type "cd .bitcoin/" ?
539 2012-08-31 22:17:17 <kjj_> without the quotes
540 2012-08-31 22:17:27 <unclemantis> yu[
541 2012-08-31 22:17:36 <unclemantis> yup... got stuff that looks like bitcoin stuff
542 2012-08-31 22:17:57 <kjj_> ok, that's probably it. anything that claims to be a database error log?
543 2012-08-31 22:18:08 <kjj_> I don't remember exactly what it would be called.
544 2012-08-31 22:18:08 <unclemantis> yup
545 2012-08-31 22:18:11 <unclemantis> debug.log
546 2012-08-31 22:18:20 <kjj_> well, that's different, but it might do
547 2012-08-31 22:18:26 <kjj_> type "tail debug.log"
548 2012-08-31 22:19:23 <unclemantis> you want the contents?
549 2012-08-31 22:19:44 <kjj_> hopefully not. does it show something like that exception?
550 2012-08-31 22:19:59 <kjj_> or does it mention trying to open a database?
551 2012-08-31 22:20:00 <unclemantis> yes it does
552 2012-08-31 22:20:03 <unclemantis> yes it does
553 2012-08-31 22:20:08 <kjj_> which database?
554 2012-08-31 22:20:16 <unclemantis> Loading addresses...
555 2012-08-31 22:20:22 <unclemantis> dbenv.open LogDir=/home/msheldon/.bitcoin/database ErrorFile=/home/msheldon/.bitcoin/db.log
556 2012-08-31 22:20:45 <unclemantis> hmmm
557 2012-08-31 22:20:47 <kjj_> ahh, addr.dat. it'll be gone soon
558 2012-08-31 22:20:57 <kjj_> just delete addr.dat and try starting it again
559 2012-08-31 22:21:35 <unclemantis> ahh same shit
560 2012-08-31 22:21:55 <kjj_> does that db.log file exist?
561 2012-08-31 22:22:05 <unclemantis> yes
562 2012-08-31 22:22:13 <kjj_> tail db.log
563 2012-08-31 22:22:19 <unclemantis> http://pastebin.com/raw.php?i=drqaDqqz
564 2012-08-31 22:22:46 <kjj_> you change versions recently?
565 2012-08-31 22:22:52 <unclemantis> nope
566 2012-08-31 22:22:59 <unclemantis> just installed this just a few moments ago
567 2012-08-31 22:23:07 <unclemantis> spanking new install, etc..
568 2012-08-31 22:23:20 <kjj_> did it run once?
569 2012-08-31 22:23:24 <unclemantis> nope
570 2012-08-31 22:23:28 <kjj_> or are you getting this on the first attempt?
571 2012-08-31 22:23:28 <unclemantis> crashed at once
572 2012-08-31 22:23:32 <unclemantis> yup
573 2012-08-31 22:23:48 <unclemantis> should i remove and install and try again?
574 2012-08-31 22:24:02 <kjj_> nah, let's see if we can get it going anyway.
575 2012-08-31 22:24:09 <kjj_> which bitcoin version, by the way?
576 2012-08-31 22:24:22 <unclemantis> https://launchpad.net/~bitcoin/+archive/bitcoin
577 2012-08-31 22:25:25 <kjj_> I'm not familiar with PPA. Does it install the deps automatically? or did you install db4.8 ?
578 2012-08-31 22:26:08 <unclemantis> i installed the PPA and then it showed up in my USC and i installed it from there
579 2012-08-31 22:26:47 <unclemantis> i don't know about db4.8. Should have installed that too right?
580 2012-08-31 22:27:11 <unclemantis> i was just following the directions on the home page of www.bitcoin.org
581 2012-08-31 22:27:29 <unclemantis> if there is ANOTHER way to do this, please inlighten me
582 2012-08-31 22:27:44 <kjj_> heh. I grab the source and compile it
583 2012-08-31 22:27:51 <unclemantis> i guess i can do that
584 2012-08-31 22:27:52 <unclemantis> brb
585 2012-08-31 22:28:00 <kjj_> hang on, that shouldn't be necessary
586 2012-08-31 22:28:18 <unclemantis> removing it anyway
587 2012-08-31 22:28:50 <unclemantis> back to square one
588 2012-08-31 22:29:25 <unclemantis> downloading the tar.gz
589 2012-08-31 22:30:00 <kjj_> ooh. my CNC controller is a Ubuntu box. let me see if it works for me
590 2012-08-31 22:30:19 <gmaxwell> 17:23 < unclemantis> spanking new install, etc..
591 2012-08-31 22:30:30 <gmaxwell> you mean a totally fresh install with no preexisting wallet or blockchain?
592 2012-08-31 22:31:03 <unclemantis> correct
593 2012-08-31 22:31:05 <unclemantis> hi max
594 2012-08-31 22:31:21 <unclemantis> ok, well i have a PPA and i have the tar.gz
595 2012-08-31 22:31:28 <unclemantis> i have nothing installed right now
596 2012-08-31 22:31:41 <unclemantis> so... do i take the red pill or the blue pill?
597 2012-08-31 22:33:16 <gmaxwell> well, try deleting the .bitcoin directory then. Though this is really really weird.
598 2012-08-31 22:34:11 <unclemantis> i did
599 2012-08-31 22:34:30 <kjj_> installed just fine on my CNC machine. heh. how long did it take to crash the first time? instant?
600 2012-08-31 22:34:48 <gmaxwell> also, you're not out of space (I know silly question, but I had to ask)
601 2012-08-31 22:34:51 <unclemantis> instaqnt
602 2012-08-31 22:34:58 <unclemantis> max i was just thinking that
603 2012-08-31 22:35:00 <unclemantis> brb
604 2012-08-31 22:35:22 <kjj_> ooh. the QT gui is pretty
605 2012-08-31 22:35:44 <unclemantis> hmmm i should be ok
606 2012-08-31 22:36:06 <kjj_> how much free space?
607 2012-08-31 22:36:52 <unclemantis> 61GB
608 2012-08-31 22:38:06 <unclemantis> did it again, first try
609 2012-08-31 22:39:14 <kjj_> did it throw the DB exception on when it crashed?
610 2012-08-31 22:39:52 <unclemantis> http://pastebin.com/T4whChbq
611 2012-08-31 22:40:28 <kjj_> oh, did you delete the .bitcoin folder before you start it?
612 2012-08-31 22:40:34 <unclemantis> no
613 2012-08-31 22:40:39 <kjj_> do that
614 2012-08-31 22:40:48 <unclemantis> why?
615 2012-08-31 22:40:57 <unclemantis> the install created it
616 2012-08-31 22:41:07 <kjj_> no, running it for the first time creates it
617 2012-08-31 22:41:12 <unclemantis> hmmm
618 2012-08-31 22:41:13 <unclemantis> ok
619 2012-08-31 22:41:45 <unclemantis> hmmm weird
620 2012-08-31 22:41:53 <unclemantis> worked this time
621 2012-08-31 22:41:55 <unclemantis> bug?
622 2012-08-31 22:42:12 <kjj_> good question
623 2012-08-31 22:42:38 <gmaxwell> might have crashed for an unrelated reason on the first startup, leaving a corrupted database behind.
624 2012-08-31 22:42:43 <unclemantis> weird
625 2012-08-31 22:42:47 <unclemantis> cause i did it TWICE
626 2012-08-31 22:42:50 <unclemantis> clean install that is
627 2012-08-31 22:43:04 <unclemantis> either way, i got it working....well. WE got it working
628 2012-08-31 22:43:13 <unclemantis> it is downloading the chain as we speak
629 2012-08-31 22:43:16 <kjj_> I'll nuke mine and see what state the PPA install leaves
630 2012-08-31 22:43:34 <unclemantis> :-)
631 2012-08-31 22:46:11 <kjj_> when you say you did a clean install twice, do you mean you wiped out the box and completely reinstalled Ubuntu between attempts?
632 2012-08-31 22:46:42 <unclemantis> no
633 2012-08-31 22:46:54 <unclemantis> cleaned out the previous install of bitcoin-qt and reinstalled
634 2012-08-31 22:47:01 <kjj_> ah! mystery solved
635 2012-08-31 22:47:19 <kjj_> when you ran it the first time, it created the .bitcoin directory for your user, and then crashed for some unknown reason
636 2012-08-31 22:47:36 <kjj_> when you removed it, it didn't remove your user data, so when you reinstalled, you still had your corrupt database
637 2012-08-31 22:47:43 <unclemantis> it did
638 2012-08-31 22:47:53 <unclemantis> it removed the .bitcoin folder
639 2012-08-31 22:48:19 <kjj_> I just did "sudo apt-get remove bitcoin-qt" on my box, and it didn't remove my .bitcoin/ folder
640 2012-08-31 22:48:34 <unclemantis> either way
641 2012-08-31 22:48:41 <unclemantis> got it kickstarted and I am happy :)
642 2012-08-31 22:48:48 <kjj_> cool. happy bitcoining
643 2012-08-31 22:48:53 <unclemantis> :-D
644 2012-08-31 22:49:05 <unclemantis> moving from Windows 7 to Ubuntu
645 2012-08-31 22:49:16 <unclemantis> i have been told I am moving from Windows to Windows :-P
646 2012-08-31 23:12:04 <helo> should bitcoin-qt respond to SIGINT?
647 2012-08-31 23:22:55 <kjj_> helo: I just took a quick look at the code and I don't see any particular signal handling. so either there isn't any, or it isn't something that I'd recognize