1 2013-10-21 00:00:10 <sipa> why do they even care?
2 2013-10-21 00:00:22 <Urushiol> who knows. it seems they don't trust "scripting" it
3 2013-10-21 00:00:25 <sipa> or what effect would it have for them that they dislike?
4 2013-10-21 00:00:34 <sipa> it has exact
5 2013-10-21 00:00:50 <sipa> it has exactly as much scripting i it as the alternative
6 2013-10-21 00:00:57 <sipa> actually, less
7 2013-10-21 00:01:14 <Urushiol> I know this. You know this. But alas...
8 2013-10-21 00:01:42 <sipa> just do it and they will not even notice
9 2013-10-21 00:02:10 <Urushiol> I don't control the escrow, otherwise I would. And they will notice when one tx is provided to the group.
10 2013-10-21 00:03:01 <Urushiol> I'm mostly venting, really. Sometimes I can't stand stupid luddites.
11 2013-10-21 00:03:28 <Urushiol> sorry, I suppose in this case it's ignorance.
12 2013-10-21 01:09:39 <EasyAt> --rescan should rescan the whole blockchain and find all valid outputs I can use, correct?
13 2013-10-21 01:35:46 <Luke-Jr> EasyAt: -rescan, not --
14 2013-10-21 01:35:56 <Luke-Jr> although -- might be accepted
15 2013-10-21 01:36:04 <EasyAt> Ah, perhaps that's the issue
16 2013-10-21 01:36:32 <Luke-Jr> also, -rescan is deprecated and will probably be removed someday
17 2013-10-21 01:37:36 <EasyAt> Luke-Jr: What will the argument be to rescan then?
18 2013-10-21 01:37:57 <Luke-Jr> none, it's unnecessary
19 2013-10-21 01:38:26 <EasyAt> Ah, hm
20 2013-10-21 01:38:42 <EasyAt> Well, I have an older backup and I'd like to get a current utxo set for my priv keys
21 2013-10-21 01:39:06 <EasyAt> It's showing I have a ton of extra coins that I don't think I have. When I -rescan it shows the same amount of coins
22 2013-10-21 01:41:33 <EasyAt> I did send from 2 nodes using same wallet. There is less than 100 TXs since node 2 started sending. So, I thought they would have the same keypool
23 2013-10-21 01:52:26 <Luke-Jr> EasyAt: loading an old wallet automatically rescans as needed
24 2013-10-21 01:52:59 <EasyAt> Neat
25 2013-10-21 01:53:18 <EasyAt> I'm curious why it thinks I have outputs to spend that I don't after a rescan
26 2013-10-21 01:57:01 <Luke-Jr> EasyAt: unconfirmed transactions are probably missing?
27 2013-10-21 01:57:41 <EasyAt> I havent sent from the address for days
28 2013-10-21 01:58:27 <Luke-Jr> of course not, you never send from addresses
29 2013-10-21 01:58:39 <Luke-Jr> but that doesn't mean the transactions from the wallet were all confirmed
30 2013-10-21 01:58:57 <EasyAt> sorry, I meant the wallet
31 2013-10-21 01:59:03 <EasyAt> I haven't spent an output
32 2013-10-21 01:59:52 <EasyAt> I'm thinking I might just take out all the priv keys and do a manual rescan as a fun project
33 2013-10-21 02:00:36 <Luke-Jr> EasyAt: there's an easy way to test of course
34 2013-10-21 02:00:42 <Luke-Jr> create a new address and send 100% of your balance there
35 2013-10-21 02:01:02 <EasyAt> I did, heh. It appeared to be a double spend as I don't see it on blockchain
36 2013-10-21 02:01:18 <EasyAt> so, I'm curious why the client still thinks I have all those coins
37 2013-10-21 02:01:38 <EasyAt> I don't see it on blockchain.info I mean
38 2013-10-21 02:15:56 <EasyAt> Luke-Jr: Do you happen to have an idea why that would happen?
39 2013-10-21 02:26:50 <Luke-Jr> no
40 2013-10-21 03:00:15 <jgarzik> ASICMINER considering a "blockchain-based exchange" (for trading their shares, I presume): https://bitcointalk.org/index.php?topic=99497.msg3374697#msg3374697
41 2013-10-21 03:00:31 <jgarzik> Lots of projects like that popping up
42 2013-10-21 03:05:28 <EasyAt> What are the circumstances in which a -rescan still can display an improper balance?
43 2013-10-21 03:09:17 <EasyAt> Other than generating a new keypool with a different node using the same wallet. Which, I would think could only mean you could have BTC that the older backup doesn't know about
44 2013-10-21 03:10:00 <olalonde> is there a standard for brainwallets?
45 2013-10-21 03:10:15 <maaku> olalonde: bip 32
46 2013-10-21 03:10:17 <olalonde> how are they commonly implemented? just sha256(passphrase) ?
47 2013-10-21 03:10:18 <olalonde> ok
48 2013-10-21 03:10:27 <maaku> that is not how they are commonly implemented
49 2013-10-21 03:10:43 <maaku> /that/bip 32/
50 2013-10-21 03:10:46 <maaku> bip 32 is how they should be implemented
51 2013-10-21 03:11:58 <olalonde> is there a flaw in sha256(passphrase)
52 2013-10-21 03:16:25 <olalonde> oh got it
53 2013-10-21 03:24:29 <olalonde> there is nothing about brainwallets in bip 32 though
54 2013-10-21 03:25:39 <olalonde> oh I guess the passphrase would be the seed in this case
55 2013-10-21 03:47:50 <olalonde> does bitcoin-qt implement HD-wallet?
56 2013-10-21 03:57:34 <Luke-Jr> olalonde: not yet
57 2013-10-21 04:59:10 <olalonde> is there any software that imports the blockchain into a (relational) database?
58 2013-10-21 05:00:09 <Luke-Jr> yes.
59 2013-10-21 05:00:11 <sipa> EasyAt: -rescan will still result in an incorrect (too low) balance if your wallet contains unconfirmed sends; it can result in a too low balance if you have non-confirming receives in it
60 2013-10-21 05:00:18 <sipa> olalonde: abe
61 2013-10-21 05:00:46 <olalonde> thanks
62 2013-10-21 05:01:02 <sipa> EasyAt: -rescan only adds missing transactions from the chain, it doesn't delete anything
63 2013-10-21 05:43:57 <EasyAt> sipa: Is there a way to remove all and rescan all transactions relating to my keys
64 2013-10-21 05:44:34 <EasyAt> Remove all TXs from the cache in the wallet
65 2013-10-21 05:45:09 <weex> EasyAt: you could use pywallet to dump keys, reimport to a new wallet, and then rescan the new wallet
66 2013-10-21 05:45:45 <weex> shut down bitcoind and backup your existing wallet first of course
67 2013-10-21 05:46:42 <EasyAt> weex: Indeed, I'm thinking a fork of pywallet that can do that on multiple wallets and recombine into a wallet useable by the reference client
68 2013-10-21 05:46:51 <EasyAt> would be neat
69 2013-10-21 05:47:27 <EasyAt> I have a directory of backups that might have different keypools but starting from the same initial wallet I'd like to comebine
70 2013-10-21 05:52:14 <weex> EasyAt: sounds interesting but dangerous, do you think many would want to do that?
71 2013-10-21 05:53:15 <sipa> EasyAt: yes, -salvagewallet
72 2013-10-21 05:53:36 <sipa> it throws everything away, and the salvages private keys from the wallet file
73 2013-10-21 05:53:45 <sipa> and rescans using those
74 2013-10-21 05:54:03 <Polyatomic> where can I get the coincontrol for Bitcoin-QT
75 2013-10-21 05:56:01 <EasyAt> weex: Why would that be dagnerous, given a backup beforehand?
76 2013-10-21 06:03:20 <weex> EasyAt: if you have a backup then it's not especially dangerous, i just mean to be cautious anytime some software is accessing a wallet file
77 2013-10-21 06:05:45 <EasyAt> Ah, okay
78 2013-10-21 06:06:43 <olalonde> does the standard client always use 1 or 2 outputs?
79 2013-10-21 06:07:14 <EasyAt> sipa weex, thank you
80 2013-10-21 06:08:23 <olalonde> 1 destination address + 1 change address?
81 2013-10-21 06:11:54 <sipa> olalonde: if necessary, yes
82 2013-10-21 06:12:09 <olalonde> no more then 2 right?
83 2013-10-21 06:12:21 <sipa> if the output matches thw input exactly, no change is necessary
84 2013-10-21 06:12:35 <sipa> wiuth sendmany there can be more
85 2013-10-21 06:12:53 <olalonde> ah ok
86 2013-10-21 06:13:01 <olalonde> thx
87 2013-10-21 06:17:34 <EasyAt> sipa: I'm curious, why doesn't rescan remove all TXs from the wallet and rescan everything?
88 2013-10-21 07:09:19 <olalonde> what happened with https://en.bitcoin.it/wiki/BIP_0016?
89 2013-10-21 07:09:25 <olalonde> what was the result of the vote
90 2013-10-21 07:11:37 <Luke-Jr> it's not a vote, it's an upgrade
91 2013-10-21 07:11:48 <Luke-Jr> it went live over a year ago
92 2013-10-21 07:34:54 <sipa> EasyAt: because that would mean throwing away unconfirmed transactions in your wallet
93 2013-10-21 07:55:01 <sturles> bitcoind -debug
94 2013-10-21 07:55:01 <sturles> Error: wallet.dat corrupt, salvage failed
95 2013-10-21 07:55:01 <sturles> I've been struggling with this all morning:
96 2013-10-21 07:55:17 <sturles> Same for all my backups.
97 2013-10-21 07:55:36 <tgs3> sturles: first make a backup, next, there are some scripts and tools to try wallet recovery. strange about the backups, are you sure there is free space on the drive?
98 2013-10-21 07:55:40 <K1773R> sturles: different DBD libs?
99 2013-10-21 07:56:05 <K1773R> sturles: ppl tend to build bitcoind with libdb5 instead of libdb4.8
100 2013-10-21 07:56:19 <sturles> Could be. I did a dist-upgrade while bitcoind was running. It had been running for a couple of months when I rebooted yesterday.
101 2013-10-21 07:56:42 <K1773R> sturles: older wallets are mostly with libdb4.8 and therefore are incompatible with libdb5
102 2013-10-21 07:56:45 <sturles> libdb is libdb4.8.
103 2013-10-21 07:57:09 <sturles> libdb_cxx-4.8.so => /usr/lib/libdb_cxx-4.8.so (0x00007f511fb59000)
104 2013-10-21 07:57:14 <K1773R> check with ldd
105 2013-10-21 07:57:16 <K1773R> ok
106 2013-10-21 07:57:20 <tgs3> K1773R: *database* library that is not backward compatible?
107 2013-10-21 07:57:40 <K1773R> tgs3: yes, blame oracle :P
108 2013-10-21 07:57:57 <sturles> tgs3: Do you have some pointers to scripts?
109 2013-10-21 07:57:59 <tgs3> now we are moving away from this shit, right?
110 2013-10-21 07:58:04 <K1773R> sturles: pywallet will do the trick then (ie, recover wallet or extract privkeys)
111 2013-10-21 07:58:24 <K1773R> tgs3: AFAIK no
112 2013-10-21 07:58:35 <tgs3> sturles: not now but if you google you will find some.. yes possibly it was pywallet. there were at least 2 set of tools old and new. But if it's the lib then just fix that
113 2013-10-21 07:58:44 <K1773R> sturles: https://bitcointalk.org/index.php?topic=34028.0
114 2013-10-21 07:59:16 <midnightmagic> olalonde: The "vote" as in, the entries of /p2sh/ in coinbase txn which were meant to indicate direct miner support for p2sh, "went through". We don't have to push /p2sh/ into our coinbase anymore.
115 2013-10-21 07:59:26 <sturles> I don't think it is just the lib.
116 2013-10-21 08:00:15 <K1773R> sturles: check the wallet with file
117 2013-10-21 08:00:52 <K1773R> ie, 4.8 looks like this
118 2013-10-21 08:00:53 <K1773R> .bitcoin/wallet.dat: Berkeley DB (Btree, version 9, native byte-order)
119 2013-10-21 08:00:53 <K1773R> $ file .bitcoin/wallet.dat
120 2013-10-21 08:01:41 <sturles> .bitcoin/wallet.dat: Berkeley DB (Btree, version 9, native byte-order)
121 2013-10-21 08:01:50 <sturles> Yep. Same.
122 2013-10-21 08:02:40 <K1773R> then you can recover your privkeys with pywallet (linked above)
123 2013-10-21 08:03:33 <sturles> ERROR:root:Couldn't open wallet.dat/main. Try quitting Bitcoin and running this again.
124 2013-10-21 08:03:40 <sturles> From pywallet.
125 2013-10-21 08:03:52 <sturles> (bitcoin is not running.)
126 2013-10-21 08:03:56 <K1773R> latest version?
127 2013-10-21 08:04:01 <sturles> Yes.
128 2013-10-21 08:04:04 <sturles> Cloned from git.
129 2013-10-21 08:05:40 <K1773R> you specified the wallet's location correctly? there is also a WUI which makes things easier -> --web AFAIK
130 2013-10-21 08:07:53 <sturles> $ python src/pywallet/pywallet.py --dumpwallet .bitcoin/wallet.dat
131 2013-10-21 08:10:00 <sturles> $ python src/pywallet/pywallet.py --dumpwallet --walletfile=.bitcoin/wallet.dat
132 2013-10-21 08:10:03 <sturles> fails as well
133 2013-10-21 08:10:24 <sturles> And just pywallet.py --dumpwallet inside .bitcoin
134 2013-10-21 08:10:45 <sturles> The wallet.dat file looks sane.
135 2013-10-21 08:11:19 <sturles> Mostly galble but recogniceable strings, i.e. accounts.
136 2013-10-21 08:15:56 <sturles> More ideas?
137 2013-10-21 08:17:06 <Polyatomic> sturles is pywallet working for u
138 2013-10-21 08:17:19 <sturles> No, see error message above.
139 2013-10-21 08:17:27 <sturles> ERROR:root:Couldn't open wallet.dat/main. Try quitting Bitcoin and running this again.
140 2013-10-21 08:17:44 <K1773R> is wallet encrypted?
141 2013-10-21 08:17:48 <sturles> No.
142 2013-10-21 08:18:01 <K1773R> well, post in the thread then, jackjack should help you
143 2013-10-21 08:18:07 <Polyatomic> try opening a command prompt
144 2013-10-21 08:18:23 <K1773R> Polyatomic: hes not on winblows
145 2013-10-21 08:18:24 <sturles> Polyatomic: Huh? I only work from the command prompt.
146 2013-10-21 08:18:39 <Polyatomic> ah shit sorry sturles
147 2013-10-21 08:19:51 <Polyatomic> I got it to work like K1773R mentioned with --web
148 2013-10-21 08:20:11 <sturles> Is your wallet corrupt..?
149 2013-10-21 08:20:29 <sturles> 09:54 < sturles> Error: wallet.dat corrupt, salvage failed
150 2013-10-21 08:20:31 <Polyatomic> n
151 2013-10-21 08:20:46 <Polyatomic> fark man
152 2013-10-21 08:21:31 <sturles> And my backups have the same corruption. :-(
153 2013-10-21 08:23:52 <Polyatomic> could u try importing one to Blockc.info
154 2013-10-21 08:30:04 <sturles> Hmm. Worth a try perhaps.
155 2013-10-21 08:31:16 <sturles> Actually it looks like pywallet.py is able to recover keys with --recover. I placed the wallet backup on a small device created for the purpose, and it looks like it is recovering keys.
156 2013-10-21 08:31:40 <sturles> Will try that first.
157 2013-10-21 08:31:44 <Polyatomic> sweet
158 2013-10-21 08:32:18 <sturles> Will be some work to restore the diffrerent accounts, but at least I will have my BTC. I hope..
159 2013-10-21 08:33:20 <sturles> Should be about 100 in that wallet. Which is why I have been taking backups often. But when at least a month of backups fail in the same way.. :-(
160 2013-10-21 08:34:04 <sturles> Found a working backup from July. It is only half as big and cointains ~0.001 BTC.
161 2013-10-21 08:40:58 <K1773R> sturles: no need to recreate accounts and assingn addresses to it, all you need is importing privkeys
162 2013-10-21 08:44:16 <sturles> Old wallet.dat: 9613312 bytes Recovered wallet.dat: 1212416 bytes
163 2013-10-21 08:44:25 <sturles> Doesn't look that good..
164 2013-10-21 08:44:41 <sturles> bitcoind: Error: Cannot initialize keypool
165 2013-10-21 08:45:00 <sturles> Trying again with -rescan
166 2013-10-21 08:45:08 <michagogo> cloud|sturles: did you try db_dump or whatever it's called?
167 2013-10-21 08:45:23 <sturles> No.
168 2013-10-21 08:45:27 <michagogo> cloud|Since the wallet is just a Berkeley DB
169 2013-10-21 08:46:44 <sturles> Installing db-util..
170 2013-10-21 08:46:59 <sturles> Hmm. Getting db5.1-util. Hope it is compatible..
171 2013-10-21 08:48:42 <sturles> db_dump: file wallet.dat has LSN 9/198824, past end of log at 1/28
172 2013-10-21 08:50:44 <sturles> Ah! Tried it in a new copy of my old .bitcoin directory. Worked there!
173 2013-10-21 08:52:43 <sturles> ACTION runs db_load and hopes for the best..
174 2013-10-21 08:55:42 <sturles> Loaded wallet.dat is 600k smaller than the one I dumped. Making another copy of .bitcoin before testing..
175 2013-10-21 08:59:38 <michagogo> cloud|sturles: well, you can open the output of db_dump and see what it looks like
176 2013-10-21 09:00:31 <sturles> What I saw didn't make much sense.
177 2013-10-21 09:00:38 <michagogo> cloud|But if you're using the 5.1 version of db-util, I'm assuming db_load won't give you something you can use with a bitcoin[d|-qt] compiled with 4.8
178 2013-10-21 09:03:06 <sturles> My fear as well. :-/
179 2013-10-21 09:08:45 <sipa> michagogo|cloud: db_dump on a 5.1 wallet, and db_load'ing it in a 4.8 one should work afaik
180 2013-10-21 09:15:58 <sturles> How can I fix this?: Error: Cannot initialize keypool
181 2013-10-21 09:21:24 <sturles> $ db4.8_load wallet.dat < wallet.dump
182 2013-10-21 09:21:33 <sturles> $ bitcoind -debug
183 2013-10-21 09:21:33 <sturles> Error: Error loading wallet.dat: Wallet corrupted
184 2013-10-21 09:21:36 <sturles> :-(
185 2013-10-21 09:22:42 <Polyatomic> if you create a new wallet.dat is that corrupted also
186 2013-10-21 09:23:15 <sturles> This was a new wallet.dat.
187 2013-10-21 09:23:21 <sturles> Old one moved away.
188 2013-10-21 09:23:31 <sipa> new as in emoty
189 2013-10-21 09:23:34 <sipa> empty
190 2013-10-21 09:24:00 <sturles> Ah. Haven't tried that.
191 2013-10-21 09:24:22 <sturles> Running -salvagewallet with the dumped/loaded wallet first.
192 2013-10-21 09:33:01 <sturles> Salvaging a 9 GiB wallet.dat takes time..
193 2013-10-21 09:38:20 <tgs3> sturles: I can help you with the salvage operatoin, upload your wallet to my ftp
194 2013-10-21 09:39:48 <sturles> Can you do more than -salvagewallet?
195 2013-10-21 09:40:31 <sipa> 9 GiB? wtf :o
196 2013-10-21 09:40:54 <sturles> I've been using it since 2010..
197 2013-10-21 09:40:58 <tgs3> sturles: satoshi how is life in japan?
198 2013-10-21 09:41:02 <sturles> :-)
199 2013-10-21 09:41:31 <sturles> 4 GiB salvaged..
200 2013-10-21 09:41:32 <tgs3> serves you right for choosing such db format
201 2013-10-21 09:42:27 <sturles> I'm sure satoshi would have fixed a wallet.dat by editing it in emacs.
202 2013-10-21 09:42:50 <sipa> satoshi used vim, of course
203 2013-10-21 09:43:13 <sturles> No. :-(
204 2013-10-21 09:46:31 <sturles> Hmm. The db_loaded wallet is dumpable by pywallet.
205 2013-10-21 09:46:50 <sturles> A lot of errors like this: Wallet data not recognized: {'__type__': 'keymeta', '__value__': '\x01\...
206 2013-10-21 09:48:05 <sturles> Is it possible to import this data dumped by pywallet?
207 2013-10-21 09:49:51 <sturles> Reminds me of the following from the perlref manual:
208 2013-10-21 09:50:02 <sturles> That's it for creating references. By now you're probably dying to know how to use references to get back to your long-lost data.
209 2013-10-21 09:50:11 <sturles> Somewhere in the middle. :-)
210 2013-10-21 10:07:46 <sturles> No luch this time either. I salvaged the wallet on another computer, which hadn't synched up with the blockchain for a while.
211 2013-10-21 10:08:13 <sturles> Nothing there now. Just some very old dust.
212 2013-10-21 10:39:37 <sturles> If I remove everything from .bitcoin, except wallet.dat and bitcoin.conf, I get the following error:
213 2013-10-21 10:39:40 <sturles> Error: An error occurred while setting up the RPC port 8332 for listening on IPv4: Cannot assign requested address
214 2013-10-21 10:39:43 <sturles> Segmentation fault
215 2013-10-21 10:39:49 <sturles> port 8332 is not taken.
216 2013-10-21 10:40:47 <sturles> This is obviously a bug.
217 2013-10-21 10:43:52 <sturles> And this is very interresting, because this time it seems to be able to load my wallet:
218 2013-10-21 10:43:55 <sturles> init message: Loading wallet...
219 2013-10-21 10:43:58 <sturles> nFileVersion = 89900
220 2013-10-21 10:44:00 <sturles> Keys: 2012 plaintext, 0 encrypted, 383 w/ metadata, 2012 total wallet 15708ms
221 2013-10-21 10:44:03 <sturles> init message: Loading addresses...
222 2013-10-21 10:48:20 <sturles> Specifying other rpcports doesn't help.
223 2013-10-21 11:05:37 <sturles> Doh! Please don't laugh.. Computer was in single user mode, and I had forgot to bring up the loopback interface. Not my day today..
224 2013-10-21 11:14:39 <wumpus> FYI I added a "build system" category under bitcoin issues on github
225 2013-10-21 11:20:43 <sturles> It is starting to look promising now. I re-installed some old versions of boost libraries to make my old bitcoind, the one which was running (self compiled) working again. This version accepts the backed up wallet, but b0rks on something else. Some database issue. Started again from an empty .bitcoin, except for wallet.dat and bitcoin.conf, and it is downloading blocks now.
226 2013-10-21 11:20:53 <sturles> ACTION crosses fingers
227 2013-10-21 11:23:29 <Polyatomic> all the best .
228 2013-10-21 12:39:22 <sturles> This is going to take a week.. Can I speed it up somehow using the old blk*.dat files?
229 2013-10-21 12:40:59 <sipa> a week? :o
230 2013-10-21 12:41:14 <sipa> just put the old blocks and chainstate directory back
231 2013-10-21 12:41:29 <sipa> if you only have the blk*.dat files, move those back and start with -reindex
232 2013-10-21 12:42:27 <sturles> Very slow net, old CPU and a very slow 5400 RPM disk. Not a good combination when downloading blocks.
233 2013-10-21 12:42:52 <sturles> Will try -reindex.
234 2013-10-21 12:43:01 <sipa> do you have the chainstate dir?
235 2013-10-21 12:44:37 <sturles> Yes.
236 2013-10-21 12:45:04 <sipa> then use that
237 2013-10-21 12:45:10 <sipa> reindex still requires processing all blocks
238 2013-10-21 12:45:19 <sipa> just put the chainstate and blocks directory back
239 2013-10-21 12:45:45 <sturles> I may have to delete some recent blk files. It crashes when starting in the current state. Not sure what's wrong.
240 2013-10-21 12:46:17 <sturles> Crashes during verification.
241 2013-10-21 12:47:03 <sturles> I'll give you the error message in a minute, when I'm done copying..
242 2013-10-21 12:48:07 <sturles> Copying is done buy a snail walking across the platters, taking one bit at a time..
243 2013-10-21 12:50:46 <sturles> bitcoind.bak: ./db/dbformat.h:96: leveldb::Slice leveldb::ExtractUserKey(const leveldb::Slice&): Assertion `internal_key.size() >= 8' failed.
244 2013-10-21 12:50:49 <sturles> Aborted
245 2013-10-21 12:51:54 <sturles> Trying again after removing index..
246 2013-10-21 12:52:02 <sturles> Same error.