1 2012-01-25 00:05:38 eoss has joined
2 2012-01-25 00:05:41 eoss has quit (Changing host)
3 2012-01-25 00:05:41 eoss has joined
4 2012-01-25 00:05:45 <dwon> Hrm. Got a corrupted wallet.dat when I was mucking around yesterday. That sort of thing seems to happen with every app I ever use that uses Berkeley DB (OpenLDAP being a good other example).
5 2012-01-25 00:06:13 <dwon> Has anyone thought about replacing wallet.dat with a plaintext or JSON file?
6 2012-01-25 00:06:38 <gmaxwell> dwon: what does 'mucking around' mean?
7 2012-01-25 00:07:06 <roconnor> dwon: I think Amory is like that
8 2012-01-25 00:07:06 <gmaxwell> And are you using binaries or did you compile it yourself? and if the latter, what version of libdb are you linked against?
9 2012-01-25 00:07:39 <dwon> I tried importing a casascius private key using pywallet, but it ignored my --wallet= option and wrote to ~/.bitcoin/wallet.dat anyway.
10 2012-01-25 00:07:48 <dwon> I had an encrypted wallet, so the whole thing broke.
11 2012-01-25 00:07:56 <gmaxwell> ...
12 2012-01-25 00:07:59 <upb> LOL
13 2012-01-25 00:08:02 <dwon> (also annoying: There's no option to decrypt an encrypted wallet)
14 2012-01-25 00:08:23 <dwon> With a text file, I would have just stopped bitcoind, added a line of text, and started it again.
15 2012-01-25 00:08:27 <gmaxwell> You can't really blame libdb for some crazy tool smashing the file while it was in use.
16 2012-01-25 00:08:52 <gmaxwell> We have an rpc call to import keys.
17 2012-01-25 00:09:09 <dwon> There's no bitcoind command to import keys, as far as I can tell.
18 2012-01-25 00:09:16 <gmaxwell> There is.
19 2012-01-25 00:09:29 theorb has joined
20 2012-01-25 00:09:30 theorbtwo has quit (Ping timeout: 248 seconds)
21 2012-01-25 00:09:40 theorb is now known as theorbtwo
22 2012-01-25 00:09:50 <gmaxwell> importprivkey <bitcoinprivkey> [label]
23 2012-01-25 00:09:59 <gmaxwell> (from help)
24 2012-01-25 00:10:02 <dwon> error: {"code":-32601,"message":"Method not found"}
25 2012-01-25 00:10:09 <dwon> I don't think that patch was ever merged into mainline
26 2012-01-25 00:10:18 <sipa> it was
27 2012-01-25 00:10:20 <gmaxwell> dwon: yes, it was.
28 2012-01-25 00:10:21 <sipa> but not yet released
29 2012-01-25 00:10:25 <sipa> it will be in 0.6
30 2012-01-25 00:10:58 <josephcp> if you compile from github it should be there
31 2012-01-25 00:11:13 <gmaxwell> dwon: what you described is actually part of an argument against having just a plain text file for the wallet unfortunately. :(
32 2012-01-25 00:11:38 <gmaxwell> dwon: if you did what you suggested, it wouldn't know to rescan, and so it would not know about the bitcoin waiting for it at that address.
33 2012-01-25 00:11:49 <dwon> gmaxwell: So what? I'd restart bitcoind and then it would work.
34 2012-01-25 00:11:53 <gmaxwell> No it wouldn't.
35 2012-01-25 00:11:59 <dwon> It would if it were implemented that way.
36 2012-01-25 00:12:09 <sipa> no there is no way it could now to rescan
37 2012-01-25 00:12:15 <sipa> know
38 2012-01-25 00:12:49 <dwon> sipa: The wallet consists of, what, a few hundred keys? It could easily be compared with a binary cached list at startup
39 2012-01-25 00:12:55 <gmaxwell> (not that the benefits of simple text file wouldn't outweigh those costsâ I think they doâ but the particular example you gave was unfortunate)
40 2012-01-25 00:12:58 copumpkin has joined
41 2012-01-25 00:13:05 <sipa> sure, but then you have more than just a text file
42 2012-01-25 00:13:07 <gmaxwell> dwon: so there would need to be a wallet to monitor the wallet? :(
43 2012-01-25 00:13:10 <sipa> of course that is possible
44 2012-01-25 00:13:15 <sipa> but what is the advantage then?
45 2012-01-25 00:13:53 <sipa> i'm much more in favor of an import/export option to convert the wallet to plaintain, and be able to import that
46 2012-01-25 00:13:54 <gmaxwell> In any case, the way to do the import would be via an import call like we have nowâ even if the wallet were just a text file.
47 2012-01-25 00:13:58 <dwon> The advantage is that I can just add/remove/edit/encrypt/transfer my wallet at will.
48 2012-01-25 00:14:07 <dwon> backups, etc
49 2012-01-25 00:14:22 <dwon> I'm mostly asking to see if people would be interested in a patch that ripped out wallet.dat and replaced it with a text format.
50 2012-01-25 00:14:25 <gmaxwell> But ... you can't. Because you will screw things up editing it. Badly. And confuse yourself and potentially lose bitcoins.
51 2012-01-25 00:14:33 <josephcp> you're probably having it in plaintext on your HDD slack space if you do it though
52 2012-01-25 00:14:35 Zarutian has quit (Quit: Zarutian)
53 2012-01-25 00:14:36 <sipa> if you'd remove keys from your wallet, especially chance keys, your wallet will look very very weird
54 2012-01-25 00:14:39 <dwon> gmaxwell: Have you ever looked at how gnucash saves its data?
55 2012-01-25 00:14:42 <roconnor> dwon: AFAIU amory has a text-format wallet
56 2012-01-25 00:15:00 <sipa> dwon: bitcoin is not gnucash, it needs to track individual coins
57 2012-01-25 00:15:01 JFK911 has quit (Ping timeout: 240 seconds)
58 2012-01-25 00:15:05 <onelineproof> I have a program that dumps your address & priv keys in text format from wallet.dat
59 2012-01-25 00:15:19 <roconnor> *armory
60 2012-01-25 00:15:24 <sipa> onelineproof: me too, my showwallet branch, and it allows importing too
61 2012-01-25 00:15:25 wirehead` has joined
62 2012-01-25 00:15:30 rdponticelli_ has joined
63 2012-01-25 00:15:33 <gmaxwell> Having a textfile is fine and good, having it for _editing_ is not good.
64 2012-01-25 00:15:42 <sipa> exactly
65 2012-01-25 00:15:49 <onelineproof> ya well mines better :)
66 2012-01-25 00:16:00 <sipa> excuse me?
67 2012-01-25 00:16:04 <onelineproof> no jk
68 2012-01-25 00:16:09 rdponticelli has quit (Ping timeout: 240 seconds)
69 2012-01-25 00:16:13 <onelineproof> its all good
70 2012-01-25 00:16:15 <dwon> sipa: There's no reason why the "wallet" needs to be one file. The most important thing is the list of private keys. Really, those should be in their own file, with nothing else. The accounting could be done separately.
71 2012-01-25 00:16:16 <sipa> :D
72 2012-01-25 00:16:20 <onelineproof> ill check it out when its released
73 2012-01-25 00:16:24 <josephcp> (selfish low blow: privkey imports wont work with p2sh)
74 2012-01-25 00:16:29 <sipa> dwon: yes, there's talk about that
75 2012-01-25 00:16:38 <luke-jr> josephcp: why not?
76 2012-01-25 00:16:39 <dwon> gmaxwell: I'm pretty sure it could be made to work with editing.
77 2012-01-25 00:16:40 <sipa> josephcp: why not?
78 2012-01-25 00:16:45 <josephcp> beuase you need to store the transactions
79 2012-01-25 00:16:49 <luke-jr> â¦
80 2012-01-25 00:16:50 <gmaxwell> dwon: ... the accounting depends on relational consistency with the private keys... unless you're going to do the very expensive relating operations over the blockchain at every startup.
81 2012-01-25 00:16:55 <josephcp> outputs
82 2012-01-25 00:16:59 <luke-jr> you need to store the script, you mean
83 2012-01-25 00:17:00 <sipa> josephcp: you always need to store the transactions
84 2012-01-25 00:17:04 <josephcp> yeah
85 2012-01-25 00:17:05 <sipa> josephcp: how will you know what to spend?
86 2012-01-25 00:17:09 <josephcp> exactly
87 2012-01-25 00:17:11 <luke-jr> new import format :D
88 2012-01-25 00:17:24 <josephcp> :-P
89 2012-01-25 00:17:28 <gmaxwell> josephcp: you import the "private key" which includes all the required data.
90 2012-01-25 00:17:38 <sipa> what is your point, josephcp?
91 2012-01-25 00:17:39 <josephcp> you need to import individual transactions
92 2012-01-25 00:17:47 <gmaxwell> No you don't.
93 2012-01-25 00:17:50 <dwon> gmaxwell: Here's my view: The private keys are what are important. Everything else is cache, and can be reconstructed if you have the private keys.
94 2012-01-25 00:17:59 <sipa> dwon: in a simply wallet, yes
95 2012-01-25 00:18:09 <sipa> dwon: but if you have comments to the transactions, and accounts, and ...
96 2012-01-25 00:18:17 <sipa> you cannot reconstruct things from the block chain
97 2012-01-25 00:18:30 <gmaxwell> dwon: Not quiteâ labels, comments, accounts? but even still only 'recovered' with a very expensive scan of all transactions in the history of bitcoin.
98 2012-01-25 00:18:47 <sipa> the idea that your entire wallet is reconstructable from the block chain will prevent us from moving to a real payment protocol
99 2012-01-25 00:19:12 <onelineproof> does the bitcoin do a check to make sure that the private keys correctly correspond to the addresses?
100 2012-01-25 00:19:20 <onelineproof> *bitcoin client
101 2012-01-25 00:19:36 <gmaxwell> josephcp: All you need to import with P2SH is the "private key" where the private key includes the actual ECDSA private key, and also the hash input used to create the p2sh. It's just more keying data from a data management perspective.
102 2012-01-25 00:20:15 <josephcp> yeah, if it's a standard private key AND you know the escrow keys
103 2012-01-25 00:20:27 <josephcp> err standard transaction
104 2012-01-25 00:20:46 <gmaxwell> (this view is adopted by bitcoin itself, look at the names of the input and output scriptsâ the script providing the satisfaction instructions is the signature, the script with the requirements is the pubkey)
105 2012-01-25 00:20:46 <sipa> onelineproof: the bitcoin client does not store addresses
106 2012-01-25 00:20:48 <josephcp> if it's a standard transaction script AND you know the escrow keys, rather
107 2012-01-25 00:20:54 <sipa> onelineproof: it only stores keypairs
108 2012-01-25 00:21:07 <onelineproof> because my program does a multiplication in elliptic curve space to make sure...
109 2012-01-25 00:21:20 <dwon> In my view, wallet.dat mixes concerns that could be separated, and I'm toying with the idea of making a patch to bitcoind that actually does that.
110 2012-01-25 00:21:30 <josephcp> if you forget you made an escrow trasaction you're SOL, it's impossible for you to know you have coins waiting for you :-P
111 2012-01-25 00:21:42 <onelineproof> ya but to get the addresses, it looks at the public key from the wallet, or the private key, and then coverts it to public and then to address?
112 2012-01-25 00:21:47 <sipa> dwon: they are more intimately linked than you think
113 2012-01-25 00:21:49 <dwon> I also don't think labels, comments, accounts, etc. belong in the same program that actually does the crypto.
114 2012-01-25 00:21:49 <gmaxwell> josephcp: there isn't any 'if', all the associated required data (the keys themselves, the template, the places to go to get the keys if they aren't there) is just part of the private key.
115 2012-01-25 00:22:16 <josephcp> gmaxwell: yeah, just saying importprivatekey wouldn't work the same, it'd be understood differently yes
116 2012-01-25 00:22:32 <sipa> onelineproof: it only loads the private keys, actually
117 2012-01-25 00:22:58 <dwon> sipa: From the protocol's perspective, you have private keys that are authorized to move bitcoins around, and you have the block chain. That's it. All the other wallet-related stuff is gravy that belongs in a different program.
118 2012-01-25 00:22:59 <gmaxwell> josephcp: no, it really would work exactly the same. It would take a blob, and stick the result into the private key field... invoke the code to figure out the address matching that private key, etc.
119 2012-01-25 00:23:02 <onelineproof> ok good, so it seems fine...
120 2012-01-25 00:23:16 <gmaxwell> josephcp: the blob is just a bit bigger.
121 2012-01-25 00:23:26 <dwon> The Satoshi bitcoind does way too much already.
122 2012-01-25 00:23:49 <gmaxwell> dwon: there really aren't "private keys that are authorized to move bitcoins around" in the protocol.
123 2012-01-25 00:23:55 <josephcp> yeah, good point, it'd just be a minor change from taht perspective, you're right
124 2012-01-25 00:24:29 <sipa> dwon: i disagree
125 2012-01-25 00:24:36 <sipa> dwon: the block handling and wallet handling should be separated
126 2012-01-25 00:24:55 <dwon> sipa: I actually agree with that.
127 2012-01-25 00:25:00 <sipa> but the list of private keys, and the transaction outputs that can be spent using them are very tightly connected
128 2012-01-25 00:25:28 <gmaxwell> dwon: there are transactions, which have requirements which can be satisfied by data in new transactions. There are a few sets of rules about what kind of pairings are allowed (e.g. to hold bitcoin count constant). But the whole notion of addresses and such isn't part of the underlying bitcoin protocol at all.
129 2012-01-25 00:26:10 <gmaxwell> But without the structure built on top of it it's not a usable system.
130 2012-01-25 00:26:23 <dwon> gmaxwell: Yes, I know. My point is that the accounting, comments, etc. have nothing to do with the protocol, and really don't belong in bitcoind at all.
131 2012-01-25 00:26:24 baz has quit (Ping timeout: 245 seconds)
132 2012-01-25 00:26:39 <dwon> and certainly not in the same file as the private keys
133 2012-01-25 00:27:17 erle- has quit (Quit: erle-)
134 2012-01-25 00:27:39 <gmaxwell> There must be relational integrety between transactions and addresses, it's hard to provide that when they're split between files. (and accounts/labels are a property of addresses)
135 2012-01-25 00:28:23 <gmaxwell> Also, I propose we put dwon and etotheipi in a room. Only one may exit. etotheipi has been saying that the wallet software doesn't do enough. dwon says it does too much.
136 2012-01-25 00:28:48 JFK911 has joined
137 2012-01-25 00:28:59 devrandom has quit (Ping timeout: 276 seconds)
138 2012-01-25 00:29:20 <gmaxwell> Since etotheipi has been backing his position with code (in the form of a quasiclient), I'm betting on him being the survivor in that battle. I'll give 2:1 odds, who's placing bets in the dwon/etotheipi death match?
139 2012-01-25 00:31:07 <dwon> The real question was, if I actually put the work into replacing wallet.dat with a more stable, human-readable format---in bitcoind---would such a patch be likely to be considered, or should I just forget about bitcoind and work on one of the alternative implementations.
140 2012-01-25 00:31:13 <luke-jr> etotheipi_: does your client support BIP 17 yet? :D
141 2012-01-25 00:32:07 <sipa> dwon: you'll at least have to convince some people that such a separation is beneficial
142 2012-01-25 00:32:50 <luke-jr> dwon: probably forget about bitcoind. they don't merge much
143 2012-01-25 00:33:03 <luke-jr> dwon: I've got a stable, well-tested patch that's 8 months old, and it's STILL not merged.
144 2012-01-25 00:33:04 traviscj has quit (Remote host closed the connection)
145 2012-01-25 00:33:24 <gmaxwell> luke-jr: A patch which _no one but you_ will step up and say they will use it!
146 2012-01-25 00:33:25 <sipa> because there is hardly evidence that your feature is wanted
147 2012-01-25 00:33:34 <luke-jr> (ignore gmaxwell's lie)
148 2012-01-25 00:33:44 <luke-jr> sipa: there's lots
149 2012-01-25 00:33:52 <sipa> then why don't they say so?
150 2012-01-25 00:34:05 <luke-jr> sipa: because most of them gave up on bitcoind by now. and one did, at least.
151 2012-01-25 00:34:22 <gmaxwell> I used to want it, and had said so, but I don't need it anymore.
152 2012-01-25 00:35:05 <luke-jr> sipa: bitcoind is basically useless for mining on mainnet nowadays, so people have been forced to build alternatives since bitcoind maintainers refuse to merge the stuff miners need
153 2012-01-25 00:35:15 <gmaxwell> luke-jr: You're kind of a poor teammate. You can't get your way all the time.
154 2012-01-25 00:35:16 devrandom has joined
155 2012-01-25 00:35:36 <luke-jr> gmaxwell: teammates work together, not constantly find excuses to interfere with stuff
156 2012-01-25 00:35:50 <luke-jr> and I think BIP16 is the first change I'm actually opposing.
157 2012-01-25 00:35:57 <gmaxwell> luke-jr: one patch of yours not going in isn't "constantly find excuses to interfere with stuff"
158 2012-01-25 00:36:14 <luke-jr> gmaxwell: it's not unique. just one of the best examples
159 2012-01-25 00:36:26 <gmaxwell> and what you say about mining is simply misleading. Show me the actually mergable async rpc pull?
160 2012-01-25 00:36:48 <gmaxwell> (which is _the_ reason that you can't solomine with stock bitcoind on mainnet)
161 2012-01-25 00:37:17 <luke-jr> gmaxwell: https://github.com/bitcoin/bitcoin/pull/568
162 2012-01-25 00:37:35 <luke-jr> over 3 months old now
163 2012-01-25 00:37:52 jjjrmy has joined
164 2012-01-25 00:38:01 <gmaxwell> Thats the patch that async it all and is trivial to crash.. just call getnewaddress in a loop at the same time as getinfo IIRC.
165 2012-01-25 00:38:06 jjjrmy has left ()
166 2012-01-25 00:38:06 <dwon> sipa: It just seems really odd to me that bitcoin-stable didn't have private key import/export from day one, and still doesn't have it (although I'm glad it's coming). I get the sense that bitcoind's lack of modularity makes it a bottleneck in bitcoin-related feature development.
167 2012-01-25 00:38:32 <luke-jr> gmaxwell: no such report.
168 2012-01-25 00:38:33 <dwon> sipa: a plaintext wallet is one of a few things that would help alleviate that
169 2012-01-25 00:38:39 <luke-jr> gmaxwell: try it and see.
170 2012-01-25 00:38:52 <sipa> dwon: the source code is definitely very non-modular, and improvements are necessary and being made (though slowly)
171 2012-01-25 00:39:00 <luke-jr> gmaxwell: an early revision (May 2011) didn't work with SSL, but that was fixed
172 2012-01-25 00:39:12 <gmaxwell> I did try it, months agoâ and I was running some version of it. It was stable so long as I only called getwork on it as I recall.
173 2012-01-25 00:39:15 <sipa> dwon: but i am not convinced that a text-file wallet would help there
174 2012-01-25 00:39:34 <dwon> sipa: Curious, what operating system are you running?
175 2012-01-25 00:39:37 <luke-jr> gmaxwell: I'm sure if you did, you would have posted to that effect.
176 2012-01-25 00:39:42 <sipa> dwon: ubuntu
177 2012-01-25 00:39:52 <luke-jr> gmaxwell: I haven't seen it crash.
178 2012-01-25 00:40:03 <gmaxwell> luke-jr: I don't think I was aware of the pull request, I had it as a seperate patch.
179 2012-01-25 00:40:20 <luke-jr> gmaxwell: this one only ever existed as a pullreq
180 2012-01-25 00:41:06 <luke-jr> (in public)
181 2012-01-25 00:41:32 <luke-jr> I just set a getinfo loop and listtransactions \* 1000 loop in the bg
182 2012-01-25 00:41:52 <luke-jr> getnewaddress would be troublesome to clean up
183 2012-01-25 00:42:03 <luke-jr> suppose I could use a dummy walletâ¦
184 2012-01-25 00:42:13 iocor has quit (Quit: Computer has gone to sleep.)
185 2012-01-25 00:42:24 <gmaxwell> I'll test it. If you do too make sure you compile with the deadlock detection enabled.
186 2012-01-25 00:42:30 <gmaxwell> Any chance this applies against head?
187 2012-01-25 00:43:29 marf_away has quit (Ping timeout: 245 seconds)
188 2012-01-25 00:43:56 <gmaxwell> luke-jr: also, that pull doesn't appear mergable as is, https://github.com/luke-jr/bitcoin/commit/c3ec6e260886f309bfc47a15040ec5fcbe603b31
189 2012-01-25 00:44:03 <gmaxwell> (shows how much actual attention it's had. :( )
190 2012-01-25 00:45:11 <gmaxwell> luke-jr: okay, this is not the same as what I was runningâ it looks like it _only_ makes getwork async?
191 2012-01-25 00:46:17 Incitatus has quit (Quit: Leaving)
192 2012-01-25 00:46:24 Turingi has quit (Read error: Connection reset by peer)
193 2012-01-25 00:46:36 cronopio has quit (Quit: leaving)
194 2012-01-25 00:49:21 <luke-jr> gmaxwell: it makes every RPC call a new thread.
195 2012-01-25 00:49:36 <luke-jr> gmaxwell: by default, they lock a mutex
196 2012-01-25 00:49:45 <luke-jr> gmaxwell: they can be flagged threadsafe as they're audited
197 2012-01-25 00:50:07 <gmaxwell> hmmm.
198 2012-01-25 00:51:07 <gmaxwell> Thats prudent, but .. can I OOM a system by just calling getinfo (which is kinda slow) 10,000 times at once causing about 80 gigs of thread stacks? (not that I really worry about RPC attacks as much as existing cases that might use a lot more ram)
199 2012-01-25 00:51:50 <luke-jr> I'm pretty sure you'd use up the socket limits first
200 2012-01-25 00:52:09 <gmaxwell> 10,000 sockets isn't a problem.
201 2012-01-25 00:52:17 <luke-jr> in a single process it is
202 2012-01-25 00:52:19 <luke-jr> with select
203 2012-01-25 00:52:20 <gmaxwell> and the thread stack in linux is 8mbytes by default IIRC.
204 2012-01-25 00:52:33 <luke-jr> thread stack size is irrelevant if it never gets used ;)
205 2012-01-25 00:52:40 <gmaxwell> ah. I guess that should be checked. yea.. limiting concurrent connections is probably fine.
206 2012-01-25 00:52:44 wasabi3 has joined
207 2012-01-25 00:52:55 <luke-jr> well, not sure if there's an intentional limit in there
208 2012-01-25 00:53:08 <luke-jr> but bitcoind isn't designed in a way that it can handle over 1024 connections
209 2012-01-25 00:53:18 <gmaxwell> sure it can, if you up the ulimit.
210 2012-01-25 00:53:23 <luke-jr> not with select
211 2012-01-25 00:53:27 <luke-jr> select() maxes out at 1024
212 2012-01-25 00:53:31 <sipa> indeed
213 2012-01-25 00:53:34 <luke-jr> after that, it crashes
214 2012-01-25 00:53:44 <luke-jr> (or worse)
215 2012-01-25 00:54:09 <gmaxwell> Mm. The only time I've seen bitcoin crash with lots of connections is when it was clearly out of FDs.
216 2012-01-25 00:54:33 <luke-jr> well, you'll get random memory corruption
217 2012-01-25 00:54:42 <sipa> what is worse than crashing?
218 2012-01-25 00:54:43 <luke-jr> usually the fd limit is 1024 anyway, so it's never a problem
219 2012-01-25 00:54:48 <sipa> launching nuclear missiles?
220 2012-01-25 00:54:50 <luke-jr> sipa: buffer overflow exploits? :p
221 2012-01-25 00:54:53 <luke-jr> stealing money
222 2012-01-25 00:54:53 <gmaxwell> (I didn't remember select being limited thusly, though it sucks because of the wait when you have fewer things then even that)
223 2012-01-25 00:54:55 JFK911 has quit (Read error: Operation timed out)
224 2012-01-25 00:54:59 <sipa> luke-jr: agree
225 2012-01-25 00:55:29 <gmaxwell> Why does the select man page not mention this limit?
226 2012-01-25 00:56:08 <sipa> FD_SETSIZE is mentioned at the bottom
227 2012-01-25 00:56:34 <gmaxwell> Indeed.
228 2012-01-25 00:56:50 JFK911 has joined
229 2012-01-25 00:57:22 <luke-jr> I don't care if it's me or someone else, but SOMEONE should be given the opportunity to maintain getwork-based mining IMO
230 2012-01-25 00:57:45 <BlueMatt> can I ask why getwork-based mining is being killed? what are we replacing it with?
231 2012-01-25 00:58:03 <sipa> BlueMatt: it is not being killed, but it has been ignored more or less in bitcoind
232 2012-01-25 00:58:11 <doublec> I got bit by the select 1024 limit in my first pool software - the language runtime I used used select underneath and it died when the pool got big
233 2012-01-25 00:58:14 <gmaxwell> should the fact that FD_SETSIZE doesn't show up in the bitcoin source concern me? :)
234 2012-01-25 00:58:23 <sipa> BlueMatt: at least, from a pool perspective
235 2012-01-25 00:58:41 <gmaxwell> It's been a long time since I wrote anything that used select for more than 3 FDs. :)
236 2012-01-25 00:58:49 <BlueMatt> sipa: I was under the impression even the highly-modified pool bitcoinds still use getwork to return work?
237 2012-01-25 00:58:50 <doublec> BlueMatt: getmemorypool and generating work externally
238 2012-01-25 00:59:06 <BlueMatt> afaik only p2pool is capable of doing that?
239 2012-01-25 00:59:08 <gmaxwell> BlueMatt: nah, now everything worth talking about uses getmemorypool.
240 2012-01-25 00:59:14 <doublec> BlueMatt: so does poolserverj
241 2012-01-25 00:59:15 <BlueMatt> mmm, oh well then yea
242 2012-01-25 00:59:18 <doublec> BlueMatt: and luke's new one
243 2012-01-25 00:59:19 <BlueMatt> kill getwork
244 2012-01-25 00:59:35 <BlueMatt> then we can kill all the make new block crap
245 2012-01-25 00:59:35 <gmaxwell> BlueMatt: nah, bitcoinj, too. and the erlang thing.
246 2012-01-25 00:59:39 SomeoneWeirdzzzz is now known as SomeoneWeird
247 2012-01-25 00:59:41 <luke-jr> BlueMatt: apparently the plan is to replace it with p2pool
248 2012-01-25 00:59:47 <doublec> mine is in testing to use it too
249 2012-01-25 00:59:50 <BlueMatt> luke-jr: sounds good to me
250 2012-01-25 00:59:59 <gmaxwell> er poolserverj
251 2012-01-25 01:00:02 <sipa> not yet
252 2012-01-25 01:00:11 <gmaxwell> yea, no way â not yet.
253 2012-01-25 01:00:18 <BlueMatt> no ofc not
254 2012-01-25 01:00:23 <sipa> but i'm certainly watching p2pool
255 2012-01-25 01:00:28 <luke-jr> BlueMatt: maybe long-run, but I don't think getwork should just be neglected either
256 2012-01-25 01:00:33 <BlueMatt> I do
257 2012-01-25 01:00:39 <BlueMatt> it would mean we can rip out more code
258 2012-01-25 01:00:44 <BlueMatt> (which Im always in favor of)
259 2012-01-25 01:00:50 <BlueMatt> esp the new block code
260 2012-01-25 01:00:53 <luke-jr> maybe in 0.7 or 0.8
261 2012-01-25 01:01:01 <gmaxwell> But yea, getwork based mining is unmaintained but the biggest reason it sucks bad (though not only) is the rpc blocking... and that matters for more than just mining.
262 2012-01-25 01:01:03 <BlueMatt> more crap that goes around CBlockStore that doesnt need to
263 2012-01-25 01:01:12 <luke-jr> but until p2pool is mature enough, bitcoind mining should be maintained
264 2012-01-25 01:01:14 <gmaxwell> The fact that bitcoin will do ... nothing else while you run getinfo is pretty annoying.
265 2012-01-25 01:01:34 <BlueMatt> luke-jr: so make a mining proxy that makes getwork requests from getmemorypool
266 2012-01-25 01:01:39 <sipa> wish i had the time to redo the locking in bitcoin properly :)
267 2012-01-25 01:01:44 <BlueMatt> (or isnt there one)
268 2012-01-25 01:01:56 <sipa> with abstraction layers before wallet and block code that lock the necessary things
269 2012-01-25 01:02:03 <BlueMatt> sipa: hey atleast cblockstore gets rid of some of cs_main
270 2012-01-25 01:02:04 <gmaxwell> BlueMatt: p2pool can just have the peer to peer stuff commented out and then you have that.
271 2012-01-25 01:02:09 <BlueMatt> (though probably wont by merge...)
272 2012-01-25 01:02:10 <luke-jr> BlueMatt: p2pool can do that fine in the long run
273 2012-01-25 01:02:14 <sipa> and no rpc or gui code accessing internal data strucutres
274 2012-01-25 01:02:21 <gmaxwell> BlueMatt: and plus it does snazzy crap like generate cool graphs of your hash rat.e
275 2012-01-25 01:02:24 <BlueMatt> gmaxwell: mmm, fair enough
276 2012-01-25 01:03:06 <sipa> in bitcoin-seeder i use readwrite locks
277 2012-01-25 01:03:16 <BlueMatt> readwrite locks would be nice
278 2012-01-25 01:03:26 <sipa> so the dns code can still do requests while the database is being dumped, for example
279 2012-01-25 01:03:32 <BlueMatt> (and not hard to write support for them in the current CriticalSection stuff)
280 2012-01-25 01:03:43 <luke-jr> I see no good reason to avoid merging the well-tested pulls for getwork into 0.6
281 2012-01-25 01:03:54 <luke-jr> who cares if they'll be yanked out later, when getwork goes away?
282 2012-01-25 01:03:57 <sipa> no, i adapted bitcoin's criticalsection code to do readwrite locks
283 2012-01-25 01:04:20 <BlueMatt> oh bitcoin-seeder uses bitcoin code?
284 2012-01-25 01:04:22 marf_away has joined
285 2012-01-25 01:04:25 <BlueMatt> or am I misreading you?
286 2012-01-25 01:04:25 <sipa> yes
287 2012-01-25 01:04:28 <BlueMatt> oh
288 2012-01-25 01:04:35 <sipa> heavily modified
289 2012-01-25 01:04:39 <k9quaint> RELEASE THE DMCA KRAKEN!
290 2012-01-25 01:04:39 <BlueMatt> ofc
291 2012-01-25 01:04:42 <sipa> but some network code is copied
292 2012-01-25 01:05:45 <sipa> BlueMatt: https://github.com/sipa/bitcoin-seeder/blob/master/bitcoin.cpp e.g.
293 2012-01-25 01:05:55 <BlueMatt> mmm
294 2012-01-25 01:06:36 * BlueMatt ->
295 2012-01-25 01:06:36 sacarlson has quit (Read error: Connection reset by peer)
296 2012-01-25 01:06:54 <sipa> BlueMatt -> space?
297 2012-01-25 01:07:12 <luke-jr> BlueMatt: you know there's no air out there?
298 2012-01-25 01:07:23 <sipa> and there is no sound either
299 2012-01-25 01:07:32 <sipa> unless you're called George Lucas
300 2012-01-25 01:08:07 <roconnor> luke-jr gmaxwell: I don't suppose I can get you guys to induldge me in another attempt at BIP 17 in testnet?
301 2012-01-25 01:08:30 <roconnor> luke-jr: you could connect directly to gmaxwell and even send the second transaction first.
302 2012-01-25 01:09:49 JRWR has quit (Quit: "If we desire respect for the law, we must first make the law respectable." ~Louis D. Brandeis)
303 2012-01-25 01:10:06 <BlueMatt> nvm
304 2012-01-25 01:19:16 <gmaxwell> luke-jr: hey, p2pool with a --solo option and a minor tweak to move worker graphs to seperate pages would be a drop in replacement for eligius. 0_o
305 2012-01-25 01:20:31 <gmaxwell> roconnor: I'm game now if luke is game.. I had to run out before... I don't have a listening testnet node, but I could fix that.
306 2012-01-25 01:21:12 <gmaxwell> or he could open an RPC port for me. and I could mine against his node.
307 2012-01-25 01:21:17 <roconnor> gmaxwell: how do you get transactions?
308 2012-01-25 01:21:25 <gmaxwell> roconnor: outbound connections.
309 2012-01-25 01:21:53 <roconnor> also, sending out the second transaction first would help
310 2012-01-25 01:22:04 <roconnor> though I don't know how easy that is for luke-jr to do
311 2012-01-25 01:22:53 <luke-jr> gmaxwell: um, no?
312 2012-01-25 01:23:12 <roconnor> gmaxwell: you could start a regular testnet peer
313 2012-01-25 01:23:24 <roconnor> luke-jr: can connect to it
314 2012-01-25 01:23:27 <gmaxwell> luke-jr: to what? replacing eligius? I didn't expect you to do itâ but it's an interesting observation.
315 2012-01-25 01:23:42 <luke-jr> gmaxwell: except that it wouldn't do anything like Eligius
316 2012-01-25 01:23:45 <roconnor> and gmaxwell can connect his miner to his peer
317 2012-01-25 01:23:46 Barracuda555 has joined
318 2012-01-25 01:23:55 <roconnor> maybe I'm making this overly complex
319 2012-01-25 01:24:09 <luke-jr> roconnor: problem is, gmaxwell would relay my BIP 17 txnâ¦
320 2012-01-25 01:24:12 sacarlson has joined
321 2012-01-25 01:24:13 <dwon> gmaxwell: To answer your previous question, I'm building against libdb5.1-dev. Does that sound fine to you?
322 2012-01-25 01:24:15 <doublec> gmaxwell: the 10 second longpolls would be a slight difference
323 2012-01-25 01:24:22 <gmaxwell> luke-jr: p2pool supports payout address in username.. and it's pplns. and it generates per worker graphs. ... thats a lot like eligius.
324 2012-01-25 01:24:35 <roconnor> luke-jr: is it easy for you to send the second transaction first?
325 2012-01-25 01:24:36 <luke-jr> gmaxwell: Eligius isn't PPLNS.
326 2012-01-25 01:24:37 <gmaxwell> doublec: turn 'em off.. not needed if solo.
327 2012-01-25 01:24:41 <luke-jr> roconnor: no
328 2012-01-25 01:24:47 <roconnor> heh
329 2012-01-25 01:24:48 <gmaxwell> luke-jr: oh I thought you switched when you chaged servers.
330 2012-01-25 01:24:49 <roconnor> what a pain
331 2012-01-25 01:24:51 <doublec> oh right, I see - I thought you meant luke running it for others to use
332 2012-01-25 01:25:11 <luke-jr> gmaxwell: I see little benefit to PPLNS at this point. ;)
333 2012-01-25 01:25:16 <gmaxwell> doublec: yea, the 10 sec longpolls would result in poor performance.
334 2012-01-25 01:26:13 <gmaxwell> luke-jr: okay, save for the fact that it would be PPLNS instead of SMPPS, and would have lots of fun dos vulnerabilties. ... it would be a pretty much drop in replacement.
335 2012-01-25 01:26:50 <gmaxwell> doublec: running it for others to use, yesâ but as a central pool, not as a p2pool node.
336 2012-01-25 01:26:58 <gmaxwell> (a p2pool network of one, if you will)
337 2012-01-25 01:27:00 <luke-jr> gmaxwell: I disagree.
338 2012-01-25 01:27:28 <luke-jr> gmaxwell: btw, that bitcoind hasn't crashed yet - 1 thread of getinfo, 2 threads of getnewaddress, plus initial blockchain download
339 2012-01-25 01:27:31 <gmaxwell> luke-jr: okay, you'd have to add a bunch of tonal to it too. :)
340 2012-01-25 01:27:38 Barracuda555 has left ()
341 2012-01-25 01:27:47 <luke-jr> gmaxwell: would you like my TBC branch of Bitcoin-Qt?
342 2012-01-25 01:28:24 <gmaxwell> luke-jr: I'll replace 100% of my bitcoin-qt usage with it!
343 2012-01-25 01:28:29 <doublec> it's proving very difficult to get in touch with the namecoin creator so he can respond to the current attack going on
344 2012-01-25 01:28:38 <doublec> definitely a negative point for namecoin uptake there
345 2012-01-25 01:28:44 <luke-jr> gmaxwell: so none? :P
346 2012-01-25 01:28:45 b4epoche_ has joined
347 2012-01-25 01:28:51 <gmaxwell> doublec: you're the obvious next in line, I think.
348 2012-01-25 01:28:57 <doublec> I don't suppose anyone here knows a contact?
349 2012-01-25 01:28:58 <roconnor> doublec: oh ya, try contacting the bitcoin creator
350 2012-01-25 01:28:59 * luke-jr concurs with gmaxwell
351 2012-01-25 01:29:02 <luke-jr> LOL
352 2012-01-25 01:29:05 <doublec> haha
353 2012-01-25 01:29:21 <gmaxwell> oh my.. its contagious.
354 2012-01-25 01:29:33 <sipa> soon RealSolid will go missing, damn :(
355 2012-01-25 01:29:34 <gmaxwell> makomk: you dodged a bullet, you should thank luke.
356 2012-01-25 01:29:41 b4epoche has quit (Ping timeout: 240 seconds)
357 2012-01-25 01:29:41 b4epoche_ is now known as b4epoche
358 2012-01-25 01:31:31 <gmaxwell> luke-jr: soâ care to connect a testnet node to me to test this for roconnor? (or open an RPC port to a testnet node for me?)
359 2012-01-25 01:31:58 <luke-jr> gmaxwell: can you avoid relaying non-std txns? ;)
360 2012-01-25 01:32:18 <gmaxwell> oh.. hah.. it'll get spent on your first...
361 2012-01-25 01:32:29 <luke-jr> or actually
362 2012-01-25 01:32:32 <gmaxwell> just disconnect -connect=127.0.0.1 and write both.
363 2012-01-25 01:32:34 <luke-jr> I think you can just -connect=0.0.0.0
364 2012-01-25 01:32:44 <luke-jr> and I -connect=you
365 2012-01-25 01:32:49 <gmaxwell> yes, that'll work.
366 2012-01-25 01:32:55 <luke-jr> OK
367 2012-01-25 01:33:03 <gmaxwell> okay.. lemme get it listening. what port is testnet? 18333 or something?
368 2012-01-25 01:33:09 <roconnor> yes
369 2012-01-25 01:33:21 wasabi3 has quit (Ping timeout: 245 seconds)
370 2012-01-25 01:33:39 <gmaxwell> maybe I'll mine a few and reorg testnet a bit to.
371 2012-01-25 01:33:45 <luke-jr> :p
372 2012-01-25 01:33:59 coblee has quit (Ping timeout: 276 seconds)
373 2012-01-25 01:34:06 <gmaxwell> less fun to do that without the threaded rpc patches...
374 2012-01-25 01:34:27 <luke-jr> gmaxwell: you could take the opportunity to test them ;)
375 2012-01-25 01:34:41 <luke-jr> gmaxwell: maybe even test all of my next-test branch
376 2012-01-25 01:34:56 <gmaxwell> I completely trust that the getwork part works, since you've been running it
377 2012-01-25 01:35:23 wasabi1 has joined
378 2012-01-25 01:35:33 <luke-jr> http://pastebin.com/XGavcpD7 <-- everything merged into next-test presently
379 2012-01-25 01:35:51 <luke-jr> gmaxwell: if it didn't, I would be shocked. I don't run next-test on Eligius though
380 2012-01-25 01:37:34 Diablo-D3 has joined
381 2012-01-25 01:37:56 <gmaxwell> luke-jr: okay, sent address, tell me when you've relayed txn and I'll start mining.
382 2012-01-25 01:38:07 da2ce7 has joined
383 2012-01-25 01:38:11 <gmaxwell> stay connected or it won't mine.
384 2012-01-25 01:40:38 <gmaxwell> wee... connected.
385 2012-01-25 01:40:56 <gmaxwell> hopefully thats luke.
386 2012-01-25 01:41:30 <gmaxwell> hm.. perhaps not.
387 2012-01-25 01:42:26 <luke-jr> wtf
388 2012-01-25 01:42:29 * luke-jr ponders
389 2012-01-25 01:42:51 <gmaxwell> It's working, because someone is getblocksing me.
390 2012-01-25 01:43:14 <luke-jr> my bitcoind is refusing to send :/
391 2012-01-25 01:43:19 <luke-jr> OH
392 2012-01-25 01:43:24 <luke-jr> maybe I need to unlock wallet
393 2012-01-25 01:43:26 <luke-jr> <.<
394 2012-01-25 01:43:27 <gmaxwell> thought they are behaving really weird.
395 2012-01-25 01:43:43 <sipa> y00 r bein h4xx0rd
396 2012-01-25 01:43:54 <gmaxwell> oh no, I'm just being dsylexic.
397 2012-01-25 01:44:48 <gmaxwell> askfor tx 22505d90cd1cca139578 0
398 2012-01-25 01:44:48 <gmaxwell> sending getdata: tx 22a837d2d78b4e592b3d
399 2012-01-25 01:44:48 <gmaxwell> sending getdata: tx 22505d90cd1cca139578
400 2012-01-25 01:44:48 <gmaxwell> ERROR: AcceptToMemoryPool() : transaction with out-of-bounds SigOpCount
401 2012-01-25 01:44:48 <gmaxwell> ERROR: AcceptToMemoryPool() : transaction with out-of-bounds SigOpCount
402 2012-01-25 01:45:32 <luke-jr> O.o
403 2012-01-25 01:45:54 <gmaxwell> I did accept this: AcceptToMemoryPool(): accepted 942c593a7b
404 2012-01-25 01:46:15 <gmaxwell> whats your transactions?
405 2012-01-25 01:46:32 <luke-jr> doh
406 2012-01-25 01:46:54 <gmaxwell> sending getdata: tx a439e34e87740475ecf0
407 2012-01-25 01:46:54 <gmaxwell> ERROR: ConnectInputs() : a439e34e87 mapTransactions prev not found 67db4e950b
408 2012-01-25 01:46:57 <gmaxwell> ERROR: AcceptToMemoryPool() : ConnectInputs failed a439e34e87
409 2012-01-25 01:47:06 <luke-jr> if (nSigOps > nSize / 65 || nSize < 100)
410 2012-01-25 01:47:07 <luke-jr> return error("AcceptToMemoryPool() : transaction with out-of-bounds SigOpCount");
411 2012-01-25 01:47:09 <luke-jr> no check for testnet
412 2012-01-25 01:47:46 <roconnor> gmaxwell: you enabled --testnet?
413 2012-01-25 01:47:46 <luke-jr> it's a multisig 1-of-1
414 2012-01-25 01:47:51 <roconnor> or just the testnet prot?
415 2012-01-25 01:47:54 <roconnor> *port
416 2012-01-25 01:47:55 <gmaxwell> roconnor: yes.
417 2012-01-25 01:48:03 <luke-jr> roconnor: -testnet only skips the "IsStandard" check, not the sigop-standard check :/
418 2012-01-25 01:48:12 <roconnor> oh right
419 2012-01-25 01:48:13 <sipa> luke-jr: multisig == 20 sigops
420 2012-01-25 01:48:18 <luke-jr> sipa: I know
421 2012-01-25 01:48:38 <sipa> testnet has the same network rules, i suppose, so it also enforces the sigop check?
422 2012-01-25 01:48:38 <gmaxwell> If only there was a proposal which fixed that counting⦠;)
423 2012-01-25 01:49:01 <roconnor> still, it shouldn't be exceeded
424 2012-01-25 01:49:12 <sipa> wait, this means that multisig is effectively unusable?
425 2012-01-25 01:49:24 <gmaxwell> the data has to be big enough.
426 2012-01-25 01:49:26 <sipa> as its sigopcount immediately exceeds the nSize/65 check?
427 2012-01-25 01:49:33 <luke-jr> sipa: this isn't a network rule
428 2012-01-25 01:49:47 <sipa> oh, right
429 2012-01-25 01:49:54 <sipa> indeed, it's just a memory pool check
430 2012-01-25 01:50:04 <gmaxwell> I can knock out that check if you like.
431 2012-01-25 01:50:05 <luke-jr> it should probably be || fTestnet there too
432 2012-01-25 01:50:07 <luke-jr> gmaxwell: please
433 2012-01-25 01:50:08 <sipa> indeed
434 2012-01-25 01:50:18 <luke-jr> gmaxwell: do I need to hack my end to retransmit? :P
435 2012-01-25 01:50:36 <gmaxwell> probably. :(
436 2012-01-25 01:50:43 <roconnor> sorry, this is all much more difficult than I expected.
437 2012-01-25 01:50:46 <gmaxwell> or we can just wait.
438 2012-01-25 01:51:21 <roconnor> though it is useful to know that I need to knock out that check to test multisig
439 2012-01-25 01:51:26 <roconnor> etotheipi_: might care about this info
440 2012-01-25 01:51:38 <luke-jr> gmaxwell: hacked mine to resend
441 2012-01-25 01:51:48 <gmaxwell> yea, no need to apologize. ... this is useful learning.
442 2012-01-25 01:52:03 <gmaxwell> lots of slow cores in this machine.. linking..
443 2012-01-25 01:52:04 <onelineproof> so, sipa, the thing I was talking about, im testing now and I think I found one problem. I opened a wallet.dat, changed one character from the private key, then reloaded the client, and it still showed that I have the same address as with the original private key.
444 2012-01-25 01:52:10 <luke-jr> sipa: care to quickly skip that check on testnet in mainline? ;)
445 2012-01-25 01:52:24 <onelineproof> now testing to see what happens when i do a transaction
446 2012-01-25 01:52:34 JFK911 has quit (Ping timeout: 240 seconds)
447 2012-01-25 01:52:37 <luke-jr> gmaxwell: ready when you are
448 2012-01-25 01:52:46 <gmaxwell> luke-jr: up again.
449 2012-01-25 01:53:00 graingert1 has quit (Read error: Connection reset by peer)
450 2012-01-25 01:53:20 gjs278 has quit (Ping timeout: 252 seconds)
451 2012-01-25 01:53:24 <luke-jr> gmaxwell: ECONNREFUSED
452 2012-01-25 01:53:40 <gmaxwell> hm? it's connected. or is that not you?
453 2012-01-25 01:53:56 <gmaxwell> someone is connected.
454 2012-01-25 01:54:04 <luke-jr> oh well
455 2012-01-25 01:54:08 <luke-jr> debug.log lies then
456 2012-01-25 01:54:14 <gmaxwell> not relaying yet.
457 2012-01-25 01:54:53 <gmaxwell> I'll brb.. going to drive home while this thinks about it.
458 2012-01-25 01:55:03 <sipa> onelineproof: nothing will happen, it effectively ignores the pubkey in the db key
459 2012-01-25 01:55:49 <onelineproof> but the bitcoin-qt client doesnt notify you that you no longer have access to that address
460 2012-01-25 01:57:28 <onelineproof> and the address that it displays obviously comes from the pubkey in the db, since its unchanged, so it does read the pubkey for displaying the address
461 2012-01-25 01:57:44 <onelineproof> maybe when it does a transaction it uses the priv key, but not for everything
462 2012-01-25 01:58:58 JFK911 has joined
463 2012-01-25 01:59:12 <sipa> no it does not use the pubkey at all
464 2012-01-25 01:59:25 <sipa> oh wait
465 2012-01-25 01:59:30 <sipa> maybe it does, never mind
466 2012-01-25 01:59:56 <gmaxwell> luke-jr: no relay yet. :(
467 2012-01-25 02:00:17 <sipa> onelineproof: it does for encrypted wallets, but for normal wallets, the pubkey is effectively not read
468 2012-01-25 02:00:21 <onelineproof> so my main point is that it should somehow notify the user if the priv key does not correspond to the address (which comes from the pubkey)
469 2012-01-25 02:00:25 <sipa> well, read but ignored
470 2012-01-25 02:00:34 <sipa> that would be a good safety check, yes
471 2012-01-25 02:00:34 <onelineproof> o ok, i didnt try encrypted wallets...
472 2012-01-25 02:01:17 BTC_Bear is now known as hbrntng!~BTC_Bear@unaffiliated/btc-bear/x-5233302|BTC_Bear
473 2012-01-25 02:01:18 <onelineproof> well my cwallet does that check...all it does is do a simple multiplication in elliptic curve space using the crypto library
474 2012-01-25 02:01:28 <sipa> yes, it is not hard
475 2012-01-25 02:04:40 Kolky has quit (Quit: Bye bye!)
476 2012-01-25 02:05:22 <onelineproof> something like EC_POINT_mul(group, pub_key, privbn, 0, 0, ctx);
477 2012-01-25 02:06:02 <onelineproof> ok then maybe ill add it as a suggestion on the github thingie
478 2012-01-25 02:06:54 <luke-jr> gmaxwell: got them yet?
479 2012-01-25 02:07:16 <gmaxwell> nope.
480 2012-01-25 02:07:26 JRWR has joined
481 2012-01-25 02:07:38 <gmaxwell> the normal retransmit won't trigger without a block I think!
482 2012-01-25 02:08:00 <gmaxwell> I guess I could mine one.
483 2012-01-25 02:08:47 <gmaxwell> (that was fast.
484 2012-01-25 02:09:33 <gmaxwell> oh there it goes...
485 2012-01-25 02:09:35 <gmaxwell> hmph. orphans.
486 2012-01-25 02:09:47 <gmaxwell> luke-jr: whats your txn id?
487 2012-01-25 02:09:55 <sipa> onelineproof: already implementing
488 2012-01-25 02:10:17 <onelineproof> cool
489 2012-01-25 02:10:50 <luke-jr> gmaxwell: 942c593a7bb5499c2b44fb36e2456285b03e857138131309c2053ff10ff54b80, 22505d90cd1cca13957894ec80c6e97702be09453f834f9116091e975e5ae19e, 6e44f73750745cb623f3bf58cd899b33f38157eabaf254f4851d4177aedbbb0e, 1204863bfbe719eb543341e62b6dbb2ed0a5b0e0fd34c2e9e4d2f4adab265efe, 6501e74021dd6c3838495d6a133491e070a19f198f023ae670418415745a6230, c484754441efc7999ed1521f926654b279711997ae045d24ad6b842a03e0caec,
490 2012-01-25 02:10:51 <luke-jr> 4b8056d8145500bce11504becb4fd6cc6dc6702218b97c1ec312637b8eae9841, 67db4e950b895db8e682610d251a50963e8c43bef42c6c6855c8212f713f494c, a439e34e87740475ecf004652fd4d8b6393400756ccad683fdd39840a922de16
491 2012-01-25 02:10:59 <gmaxwell> I suspect it's here.
492 2012-01-25 02:11:08 <k9quaint> altogether too many notes
493 2012-01-25 02:11:49 <gmaxwell> time for a bit of a reorg.
494 2012-01-25 02:12:01 <sipa> onelineproof: https://github.com/bitcoin/bitcoin/pull/782
495 2012-01-25 02:13:07 <luke-jr> gmaxwell: oh?
496 2012-01-25 02:13:20 <gribble> New news from bitcoinrss: sipa opened pull request 782 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/782>
497 2012-01-25 02:13:37 <gmaxwell> luke-jr: I think testnet gained a block while we were off in space.
498 2012-01-25 02:13:43 <luke-jr> gmaxwell: i c
499 2012-01-25 02:13:44 <gmaxwell> luke-jr: in any case, I've reconnected to the real network.
500 2012-01-25 02:14:24 <gmaxwell> (and dropped like eight blocks on it)
501 2012-01-25 02:14:55 <luke-jr> BBE hasn't got it yet
502 2012-01-25 02:15:17 <gmaxwell> weird.. I don't appear to be telling nodes about it!
503 2012-01-25 02:15:29 <roconnor> gmaxwell: mine another :D
504 2012-01-25 02:15:34 coblee has joined
505 2012-01-25 02:15:35 <gmaxwell> thats a bug.
506 2012-01-25 02:15:50 <gmaxwell> roconnor: yea.. that will do it.
507 2012-01-25 02:16:05 <roconnor> there it goes
508 2012-01-25 02:16:30 <gmaxwell> ERROR: ConnectInputs() : 22a837d2d7 prev tx already used at (nFile=1, nBlockPos=31538368, nTxPos=31538583)
509 2012-01-25 02:16:52 <gmaxwell> the idiot coin grabber doesn't care if they're already used or not I guess?
510 2012-01-25 02:17:03 <luke-jr> lol
511 2012-01-25 02:17:12 <gmaxwell> man, I wish this node had IP logging â I'd know who did it, because they must be connected to me.
512 2012-01-25 02:17:38 <roconnor> gmaxwell: I don't see the redemption transaction
513 2012-01-25 02:17:51 <gmaxwell> well, darn it. luke-jr?
514 2012-01-25 02:18:02 <luke-jr> I don't see anyâ¦
515 2012-01-25 02:18:08 <gmaxwell> :-/
516 2012-01-25 02:18:10 <gmaxwell> whats the txn id?
517 2012-01-25 02:18:16 <luke-jr> gmaxwell: 942c593a7bb5499c2b44fb36e2456285b03e857138131309c2053ff10ff54b80, 22505d90cd1cca13957894ec80c6e97702be09453f834f9116091e975e5ae19e, 6e44f73750745cb623f3bf58cd899b33f38157eabaf254f4851d4177aedbbb0e, 1204863bfbe719eb543341e62b6dbb2ed0a5b0e0fd34c2e9e4d2f4adab265efe, 6501e74021dd6c3838495d6a133491e070a19f198f023ae670418415745a6230, c484754441efc7999ed1521f926654b279711997ae045d24ad6b842a03e0caec,
518 2012-01-25 02:18:18 <luke-jr> 4b8056d8145500bce11504becb4fd6cc6dc6702218b97c1ec312637b8eae9841, 67db4e950b895db8e682610d251a50963e8c43bef42c6c6855c8212f713f494c, a439e34e87740475ecf004652fd4d8b6393400756ccad683fdd39840a922de16
519 2012-01-25 02:18:34 <roconnor> http://blockexplorer.com/testnet/t/6cgChXgz1E is one end
520 2012-01-25 02:18:37 <gmaxwell> I'd checked to see I'd accepted the first one.
521 2012-01-25 02:18:42 <roconnor> that the 942
522 2012-01-25 02:19:00 <gmaxwell> luke-jr: don't just give me all of them whats missing?
523 2012-01-25 02:19:04 <luke-jr> aha, I missed /tx
524 2012-01-25 02:19:20 <roconnor> it's totally gonna get stolen :D
525 2012-01-25 02:19:30 <roconnor> it's a lot this time :D
526 2012-01-25 02:19:31 <gmaxwell> hah. if it hasn't been already.
527 2012-01-25 02:19:36 <luke-jr> roconnor: until gmaxwell orphans it hopefully
528 2012-01-25 02:20:01 <luke-jr> gmaxwell: mine another block and see if that helps?
529 2012-01-25 02:20:33 Joric has joined
530 2012-01-25 02:20:34 Joric has quit (Changing host)
531 2012-01-25 02:20:34 Joric has joined
532 2012-01-25 02:20:38 <gmaxwell> lemme shut down... so I shut down and went into connect=0.0.0 mode again, to wipe my memrory pool of any theft.
533 2012-01-25 02:20:48 <luke-jr> hmm
534 2012-01-25 02:20:55 <luke-jr> gmaxwell: you don't seem to have mined them, even if you accepted them :x
535 2012-01-25 02:21:20 <roconnor> another hidden opsigcount check?
536 2012-01-25 02:21:29 <gmaxwell> well, give me an ID of something missing?
537 2012-01-25 02:21:33 <luke-jr> 22505d90cd1cca13957894ec80c6e97702be09453f834f9116091e975e5ae19e
538 2012-01-25 02:22:13 <gmaxwell> hm.
539 2012-01-25 02:22:23 <gmaxwell> ERROR: FetchInputs() : 22505d90cd mapTransactions prev not found 22a837d2d7
540 2012-01-25 02:22:23 <gmaxwell> ERROR: AcceptToMemoryPool() : FetchInputs failed 22505d90cd
541 2012-01-25 02:22:23 <gmaxwell> storing orphan tx 22505d90cd
542 2012-01-25 02:22:38 <gmaxwell> ERROR: ConnectInputs() : 22a837d2d7 prev tx already used at (nFile=1, nBlockPos=31538368, nTxPos=31538583)
543 2012-01-25 02:22:42 <gmaxwell> ERROR: AcceptToMemoryPool() : ConnectInputs failed 22a837d2d7
544 2012-01-25 02:24:20 <roconnor> something appears to be wrong with luke-jr transaction series
545 2012-01-25 02:24:23 <luke-jr> I don't see anything redeeming it in BBE
546 2012-01-25 02:24:43 <gmaxwell> did you manage to use the same input in two transactions?
547 2012-01-25 02:24:49 <roconnor> luke-jr: 22505d90cd depends on 22a837d2d7 which doesn't appear to be in your list
548 2012-01-25 02:25:42 <gmaxwell> dependency on an already stolen coin and luke and I are on mutually exclusive forks? luke-jr whats your height?
549 2012-01-25 02:25:51 <gmaxwell> I'm at "blocks" : 44795,
550 2012-01-25 02:26:06 vsrinivas has joined
551 2012-01-25 02:26:30 <luke-jr> 44795
552 2012-01-25 02:26:44 <gmaxwell> hm. then you should have the same view of the network as me.
553 2012-01-25 02:27:05 <roconnor> right, and the redeeming transaction isn't there
554 2012-01-25 02:27:27 <roconnor> a transaction is missing from luke-jr's list of transactions
555 2012-01-25 02:27:44 JZavala has joined
556 2012-01-25 02:27:44 <luke-jr> gmaxwell: try orphaning the theft from before? :/
557 2012-01-25 02:27:51 <luke-jr> bitcoind can't handle thefts I guess
558 2012-01-25 02:28:44 <roconnor> I don't think this latest one is stolen
559 2012-01-25 02:28:56 <gmaxwell> roconnor: I think he respent the old theft or something like that.
560 2012-01-25 02:29:03 <luke-jr> roconnor: no, but bitcoind won't give up the other txns
561 2012-01-25 02:29:06 <roconnor> oh
562 2012-01-25 02:29:31 <roconnor> that is unfortuante
563 2012-01-25 02:29:36 <gmaxwell> luke-jr: if you have bip17 code why aren't you rejecting this chain?
564 2012-01-25 02:29:43 <luke-jr> gmaxwell: ⦠good question
565 2012-01-25 02:29:49 <luke-jr> maybe because I already verified the blocks?
566 2012-01-25 02:29:57 * luke-jr deletes blk* and tries over
567 2012-01-25 02:30:04 <gmaxwell> right okay so you turned that on after?
568 2012-01-25 02:30:06 <luke-jr> oh
569 2012-01-25 02:30:08 <luke-jr> because the timestamp
570 2012-01-25 02:30:14 * luke-jr sets that to 0 too
571 2012-01-25 02:30:21 <gmaxwell> okay, so set the timestamp, delete your blocks..
572 2012-01-25 02:30:29 <gmaxwell> then when you sync I'll delete mine and sync from you.
573 2012-01-25 02:30:47 <gmaxwell> then I'll reorg the #@$#@ out of testnet, I guess.
574 2012-01-25 02:31:01 <gmaxwell> ah, thats why it's so easy now.. diff of 8.8
575 2012-01-25 02:31:13 <gmaxwell> yea, I'll totally unsteal those coins for you.
576 2012-01-25 02:31:30 <roconnor> heh, I'll have to rerun my experiments but I"m okay with that. My mistakes will be covered up. :P
577 2012-01-25 02:31:50 <gmaxwell> "Watson, prepare the time machine!"
578 2012-01-25 02:32:31 <gmaxwell> I'm going to adjust my firewall settings so only luke can connect..
579 2012-01-25 02:32:56 <luke-jr> found a trivial bug in BIP17 impl :P
580 2012-01-25 02:33:06 <roconnor> luke-jr: is it related to core?
581 2012-01-25 02:34:07 <luke-jr> no
582 2012-01-25 02:34:11 <luke-jr> the GetArg is missing the -
583 2012-01-25 02:34:14 <luke-jr> so changing the time is impossible
584 2012-01-25 02:34:16 <luke-jr> <.<
585 2012-01-25 02:34:17 <gmaxwell> hah
586 2012-01-25 02:34:49 <luke-jr> come to think of it, I bet that bug is in git master too
587 2012-01-25 02:35:36 <luke-jr> 8k blocks so far
588 2012-01-25 02:36:14 <luke-jr> 11k
589 2012-01-25 02:36:23 <gmaxwell> slowpoke.
590 2012-01-25 02:36:33 <roconnor> thats almost as slow as my client
591 2012-01-25 02:37:17 <gmaxwell> once you've resynced. I'll pop my chain. and learn it from you... then I remain a mostly unmodified node.. but I'll be cut back to the point I need to reorg from.
592 2012-01-25 02:38:26 <gmaxwell> hm anyone have handy timewarp code? I could push testnet down to 1... :) then it would be easy to maintain.
593 2012-01-25 02:38:36 <gribble> New news from bitcoinrss: sp1d3rx opened issue 783 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/783>
594 2012-01-25 02:38:37 <gmaxwell> "testnet difficulty problem solved!"
595 2012-01-25 02:38:57 <roconnor> see, it is a feature afterall
596 2012-01-25 02:39:02 vsrinivas has left ()
597 2012-01-25 02:39:21 <sipa> gmaxwell: ArtForz
598 2012-01-25 02:39:58 <gmaxwell> sipa: would actually be useful for testing your fix, I suppose.
599 2012-01-25 02:40:57 pickett_ has quit (Ping timeout: 276 seconds)
600 2012-01-25 02:42:08 <luke-jr> 25k
601 2012-01-25 02:42:08 slush has quit (Read error: Connection reset by peer)
602 2012-01-25 02:42:31 <gmaxwell> Anyone know a faster way to reach art than email?
603 2012-01-25 02:42:52 <doublec> he's often on irc
604 2012-01-25 02:42:54 <gmaxwell> I don't feel like waiting but if I can get the patch now I'll go ahead and timewarp testnet in the process.
605 2012-01-25 02:43:13 <roconnor> gmaxwell: you can shine the ArtForz signal in the sky
606 2012-01-25 02:43:25 <sipa> what is testnet's difficulty:
607 2012-01-25 02:43:26 <sipa> ?
608 2012-01-25 02:43:28 <roconnor> if it is cloudy
609 2012-01-25 02:43:29 <gmaxwell> sipa: 8.
610 2012-01-25 02:43:39 <gmaxwell> å
611 2012-01-25 02:43:48 <sipa> and your hashrate? :p
612 2012-01-25 02:44:03 <gmaxwell> sipa: I've been throwing 2gh/s at it.
613 2012-01-25 02:44:15 <sipa> ;;bc,calcd 2000000 8
614 2012-01-25 02:44:15 <gmaxwell> I can though a lot more, except the getwork rpc falls down.
615 2012-01-25 02:44:15 <gribble> The average time to generate a block at 2000000 Khps, given the supplied difficulty of 8, is 17 seconds
616 2012-01-25 02:44:30 <roconnor> is that 8 mainnet difficulty or 8 testnet difficulty
617 2012-01-25 02:44:36 <roconnor> stupid different scales
618 2012-01-25 02:44:43 <gmaxwell> what? they're different scales?!
619 2012-01-25 02:44:48 <sipa> yes, factor 2
620 2012-01-25 02:45:05 <gmaxwell> 0_o well getinfo returns "difficulty" : 8.88353262,
621 2012-01-25 02:45:07 <sipa> worst idea ever, imho
622 2012-01-25 02:45:07 Cablesaurus has quit (Quit: Man who run behind car get exhausted)
623 2012-01-25 02:45:31 <roconnor> My client measures everything in mainnet difficulty IIRC
624 2012-01-25 02:45:42 <gmaxwell> hopefully we sent gavin to rehab after that one?
625 2012-01-25 02:46:00 <gmaxwell> factor of 2 in which direction?
626 2012-01-25 02:46:06 <roconnor> sipa: OP_RETURN was a pretty bad idea
627 2012-01-25 02:46:11 <sipa> 1 testnet = 0.5 realnet
628 2012-01-25 02:46:14 <luke-jr> 30k
629 2012-01-25 02:46:16 <roconnor> I thought it was a factor of 4
630 2012-01-25 02:47:32 <gmaxwell> so at 2gh/s it would take me about 9 hours or so to knock it all the way down..
631 2012-01-25 02:47:45 <roconnor> gmaxwell: down to 1?
632 2012-01-25 02:47:47 <gmaxwell> but if 1 is 0.5 the miner will lose efficiency .. no point in going below 2.
633 2012-01-25 02:48:14 <sipa> actually, i can't find that in the code anymore :s
634 2012-01-25 02:48:18 <gmaxwell> roconnor: assming it takes me 2016*2 blocks to do it.
635 2012-01-25 02:48:26 <roconnor> ah
636 2012-01-25 02:49:34 <luke-jr> 38k
637 2012-01-25 02:49:51 gjs278 has joined
638 2012-01-25 02:52:03 <sipa> gmaxwell: can you verify based on how frequently you find a block, whether there actually is a different difficulty scale?
639 2012-01-25 02:52:12 <sipa> i can't find any trace of it in the code
640 2012-01-25 02:52:13 <gmaxwell> sipa: it's so fast it's hard to tell.
641 2012-01-25 02:52:17 theymos has joined
642 2012-01-25 02:52:28 <gmaxwell> It sure didn't seem like 17 seconds on average.
643 2012-01-25 02:52:30 <roconnor> sipa: the difficulty is the ratio to the difficulty of the genessis block
644 2012-01-25 02:53:12 <roconnor> er
645 2012-01-25 02:53:20 pickett_ has joined
646 2012-01-25 02:53:21 <roconnor> ratio of the target to the genessis block
647 2012-01-25 02:54:00 <sipa> that's not how it is calculated
648 2012-01-25 02:54:07 <gmaxwell> Vanitygen users are gonna love compressed public keys.. they'll get a great big speedup by searching both.
649 2012-01-25 02:54:36 <roconnor> gmaxwell: is the EC operation the slow bit?
650 2012-01-25 02:54:42 <sipa> definitely
651 2012-01-25 02:55:02 <gmaxwell> They've made it quite fast now.. but yes.
652 2012-01-25 02:55:19 <gmaxwell> ~18Maddr/s on $random ati gpu now.
653 2012-01-25 02:55:48 <roconnor> gmaxwell: shh, don't tell them and just take thier buisness
654 2012-01-25 02:55:49 <gmaxwell> amusingly â this suggests that we could potentially speed up chain validation by doing ecdsa on the gpu.
655 2012-01-25 02:56:26 Cablesaurus has joined
656 2012-01-25 02:56:26 Cablesaurus has quit (Changing host)
657 2012-01-25 02:56:26 Cablesaurus has joined
658 2012-01-25 02:56:53 <roconnor> gives 50:50 odds on luke-jr getting all the way upto block 44795 :P
659 2012-01-25 02:57:13 <gmaxwell> roconnor: yea.. he's been quiet... :)
660 2012-01-25 02:58:08 <gmaxwell> doesn't look like it.
661 2012-01-25 02:58:15 <sipa> i'm quote sure the unly thing different is that difficulty starts are 0.5 for testnet
662 2012-01-25 02:58:20 <sipa> but there is no different scale
663 2012-01-25 02:58:43 <gmaxwell> roconnor: he didn't getdata for it.. so I think he didn't get that far.
664 2012-01-25 02:58:47 <gmaxwell> luke-jr: you should be done now.
665 2012-01-25 02:59:04 <roconnor> gmaxwell: he was pulling from you?
666 2012-01-25 02:59:08 <gmaxwell> yes.
667 2012-01-25 02:59:13 <roconnor> ah
668 2012-01-25 02:59:13 <doublec> with these namecoin spam transactions, I should be expecting to see them in my memory pool on a node, right?
669 2012-01-25 02:59:36 <sipa> not necessarily
670 2012-01-25 03:00:04 <doublec> I'm logging all transactions relayed to me but don't seem them on a node with lots of connections
671 2012-01-25 03:00:06 <gmaxwell> roconnor: hm. the last block he pulled from me was 44794 it looks like so I guess you were right.
672 2012-01-25 03:00:15 <roconnor> heh
673 2012-01-25 03:00:21 <roconnor> well at least we get to debug some BIP 17 code
674 2012-01-25 03:00:41 slush has joined
675 2012-01-25 03:01:36 <gmaxwell> doublec: maybe no new ones are being emitted right now?
676 2012-01-25 03:02:07 <sipa> gmaxwell: actually, you could just look at getwork()'s target, and compare it with difficulty to know
677 2012-01-25 03:02:33 <doublec> gmaxwell: the blocks are still full of them
678 2012-01-25 03:02:47 <sipa> doublec: unless they are being added by a miner
679 2012-01-25 03:03:03 <doublec> right, that was my thought - or sent directly to the miner
680 2012-01-25 03:03:25 <gmaxwell> doublec: someone in #bitcoin was talking about doing this the other day, they didn't sound that sophicated.
681 2012-01-25 03:03:42 <doublec> maybe my logging is wrong
682 2012-01-25 03:03:44 * doublec double checks
683 2012-01-25 03:03:57 <doublec> I'm seeing other peoples transactions just not the spam ones
684 2012-01-25 03:04:00 <roconnor> that's how doublec gets his nic
685 2012-01-25 03:04:01 <gmaxwell> Do they violate some normal fee rule? e.g. might they only be relayed by modified nodes?
686 2012-01-25 03:04:07 <doublec> haha
687 2012-01-25 03:04:25 <doublec> ah, good point
688 2012-01-25 03:04:40 <sipa> i'm sure he got his NIC from a regular vendor
689 2012-01-25 03:05:06 <gmaxwell> if so â then (1) you might be dropping them yourself, and (2) they may only be getting forwarded in some free-relay subset of the network.
690 2012-01-25 03:05:26 <luke-jr> InvalidChainFound: invalid block=00000000036eb1b913c6 height=44758 work=13480243478959218
691 2012-01-25 03:05:28 <luke-jr> InvalidChainFound: current best=0000000019680aaa3920 height=44757 work=13480205323894932
692 2012-01-25 03:05:37 <gmaxwell> luke-jr: oh, good.
693 2012-01-25 03:05:52 <gmaxwell> so your height is 44757?
694 2012-01-25 03:05:54 <etotheipi_> so what's up with multi-sig not working?
695 2012-01-25 03:06:00 <luke-jr> gmaxwell: yes
696 2012-01-25 03:06:10 <doublec> gmaxwell: thanks for the tip, I'll check
697 2012-01-25 03:06:11 <roconnor> etotheipi_: something about relay rules and opsigcounts
698 2012-01-25 03:06:27 <etotheipi_> I seem to remember Gavin mentioning in a post somewhere, that he had to change some constant to 200
699 2012-01-25 03:06:39 <gmaxwell> okaydokie.
700 2012-01-25 03:06:42 <etotheipi_> from... 65
701 2012-01-25 03:06:49 <gmaxwell> luke-jr: I'm going to mindwipe my node. one minute.
702 2012-01-25 03:06:49 <etotheipi_> ?
703 2012-01-25 03:07:09 rdponticelli_ has quit (Read error: Connection reset by peer)
704 2012-01-25 03:07:19 <roconnor> height should be 44774?
705 2012-01-25 03:07:22 rdponticelli has joined
706 2012-01-25 03:07:30 <roconnor> oops
707 2012-01-25 03:07:32 <roconnor> nevermind
708 2012-01-25 03:07:59 <gmaxwell> luke-jr: can you kick yours to make it reconnect to me?
709 2012-01-25 03:09:22 <roconnor> awesome; this will totally hide my mistakes :D
710 2012-01-25 03:09:30 <roconnor> I approve of this reorg.
711 2012-01-25 03:09:47 <roconnor> ... unless the transactions are reinserted :(
712 2012-01-25 03:09:55 eldentyrell has joined
713 2012-01-25 03:10:16 <sipa> they should be
714 2012-01-25 03:10:40 <gmaxwell> roconnor: better come up with some double spends of that coin.
715 2012-01-25 03:10:51 <roconnor> heh
716 2012-01-25 03:10:52 <gmaxwell> you could wipe your nodes mind and then connect to me too. :)
717 2012-01-25 03:11:05 <luke-jr> gmaxwell: done I think
718 2012-01-25 03:11:14 <gmaxwell> luke-jr: yea, it's syncing now.
719 2012-01-25 03:11:52 <roconnor> does this rely on the source client rebroadcasting?
720 2012-01-25 03:11:56 <gmaxwell> stupid code is prior to the mlock fix on my end.. I might need to apply that.
721 2012-01-25 03:12:03 <sipa> roconnor: no
722 2012-01-25 03:12:34 <etotheipi_> so does this mean it may not be possible to enable multisig without forking?
723 2012-01-25 03:12:45 <sipa> etotheipi_: it's not a network rule
724 2012-01-25 03:12:52 <roconnor> etotheipi_: no, it is strickly a client relay rule
725 2012-01-25 03:12:55 <roconnor> *strictly
726 2012-01-25 03:13:04 <roconnor> the multisig transactions are valid
727 2012-01-25 03:13:08 <etotheipi_> ahh
728 2012-01-25 03:13:11 <roconnor> they just are not relayed at the moment
729 2012-01-25 03:13:13 <roconnor> IIUC
730 2012-01-25 03:13:19 <etotheipi_> so propagation may be tough, but you can still mine them
731 2012-01-25 03:13:24 <roconnor> possibly not accepted by miners
732 2012-01-25 03:13:29 <roconnor> you can mine them yourself.
733 2012-01-25 03:13:32 <roconnor> IIUC
734 2012-01-25 03:13:43 <roconnor> but really this ought to be changed
735 2012-01-25 03:13:46 <roconnor> they should be relayed
736 2012-01-25 03:13:54 <etotheipi_> understood...
737 2012-01-25 03:14:10 <roconnor> at least on testnet
738 2012-01-25 03:14:22 <etotheipi_> so next question... about 4 days ago, I tried to empty my Satoshi wallet
739 2012-01-25 03:14:36 <gmaxwell> luke-jr: I'm restarting here with the mlock fixes.. should sync faster now.. (had only made it to 4000 after a couple minutes)
740 2012-01-25 03:14:39 <roconnor> sipa: can etotheipi_ add "filler" to make it pass the check?
741 2012-01-25 03:14:42 <etotheipi_> the tx was not accepted by the network, but my client believes it was valid
742 2012-01-25 03:15:14 <sipa> roconnor: i suppose, an extra drop in the scriptSig should work
743 2012-01-25 03:15:14 <etotheipi_> (my client is the satoshi client, not Armory)
744 2012-01-25 03:15:26 <sipa> *push
745 2012-01-25 03:15:28 <sipa> not drop
746 2012-01-25 03:15:46 <etotheipi_> so my Satoshi client is rebroadcasting this tx every 30 min or so...
747 2012-01-25 03:15:50 <roconnor> sipa: if that works then this check is retarded
748 2012-01-25 03:16:01 <etotheipi_> the outputs are locked, so I can't do anything with it
749 2012-01-25 03:16:09 <gmaxwell> roconnor: it's only a little retarded.
750 2012-01-25 03:16:10 <roconnor> etotheipi_: so I think if you add filler to your scripts they will be relayed
751 2012-01-25 03:16:13 <theymos> etotheipi_: Is it actually invalid?
752 2012-01-25 03:16:15 <gmaxwell> roconnor: it's a size to computation check.
753 2012-01-25 03:16:26 <etotheipi_> theymos, it must not be
754 2012-01-25 03:16:29 <roconnor> etotheipi_: though I guess they become more expensive
755 2012-01-25 03:16:31 <gmaxwell> roconnor: e.g. it prevents txn which are very small but do lots of computation.
756 2012-01-25 03:16:36 <etotheipi_> but I have the raw tx if anyone wants to check it
757 2012-01-25 03:16:46 <etotheipi_> I'm not sure why my client produced an invalid tx
758 2012-01-25 03:17:20 <gmaxwell> etotheipi_: do you have a second node that you can see rejecting it?
759 2012-01-25 03:17:36 <gmaxwell> etotheipi_: do you know the txn id? (maybe I've seen it and can see why I was rejecting it)
760 2012-01-25 03:17:48 <etotheipi_> well Armory is connected to the Satoshi node... and every time I clear Armory's memory pool, it comes back
761 2012-01-25 03:18:02 <etotheipi_> 6acfd032e78a94d7611b5a2b00f190bc0c970224b39817562f3b1f1a7de6c9ee
762 2012-01-25 03:18:22 <etotheipi_> so it seems the Satoshi node is doing what it's supposed to... keeps broadcasting
763 2012-01-25 03:18:27 <theymos> etotheipi_: The Satoshi client somtimes produces (and keeps) invalid transactions that are double-spending.
764 2012-01-25 03:18:48 <gmaxwell> Doesn't appear to have appeared for me.
765 2012-01-25 03:18:49 <luke-jr> gmaxwell: how goes it?
766 2012-01-25 03:19:06 <gmaxwell> luke-jr: height 8000.
767 2012-01-25 03:19:11 <roconnor> theymos: there is a bug in BBE that you might want to see before gmaxwell bulldozes it
768 2012-01-25 03:19:25 <theymos> roconnor: What is it?
769 2012-01-25 03:19:36 <gmaxwell> Has bbe been tested against deepish reorgs?
770 2012-01-25 03:19:55 <sipa> artforz caused one once
771 2012-01-25 03:20:00 <etotheipi_> theymos, gmaxwell, whoever is interested in why this might've happened: http://pastebin.com/xaQ4hGiH
772 2012-01-25 03:20:04 <sipa> iirc BBE didn't deal with it cleanly
773 2012-01-25 03:20:10 <roconnor> theymos: http://blockexplorer.com/testnet/tx/6c010586ba139c5ef64a0ab8352f16125497cc742fe10f41535b0bc99124beed#o0 says not redeemed
774 2012-01-25 03:20:10 marf_away has quit (Ping timeout: 245 seconds)
775 2012-01-25 03:20:15 <roconnor> theymos: but it is redeemed by
776 2012-01-25 03:20:21 <roconnor> http://blockexplorer.com/testnet/tx/087e66d6f7869d5a3c73b77ad44bb435e582067cbf2c78415403c32778b16096#o0
777 2012-01-25 03:20:43 <gmaxwell> etotheipi_: is that tx id in the same order bitcoin would log it in?
778 2012-01-25 03:20:58 <etotheipi_> it's exactly how it's serialized in blk0001.dat
779 2012-01-25 03:21:01 <etotheipi_> and on the network
780 2012-01-25 03:21:20 <roconnor> gmaxwell: there was a big reorg on testnet some time ago
781 2012-01-25 03:21:25 <roconnor> gmaxwell: probably done by you
782 2012-01-25 03:21:28 <gmaxwell> etotheipi_: super useful! https://blockchain.info/rejected
783 2012-01-25 03:21:55 <roconnor> oh it was artzford
784 2012-01-25 03:21:55 <etotheipi_> gmaxwell, thanks! I didn't even know about it
785 2012-01-25 03:21:57 <gmaxwell> roconnor: hm? no. I've never previously intentionally reorged testnet.
786 2012-01-25 03:22:04 <theymos> roconnor: It shows up as redeemed for me.
787 2012-01-25 03:22:07 <gmaxwell> etotheipi_: it has your txn
788 2012-01-25 03:22:11 <roconnor> artforz
789 2012-01-25 03:22:13 <etotheipi_> gmaxwell, (oh, I misunderstood your question... I belive that ID is LE)
790 2012-01-25 03:22:41 <gmaxwell> etotheipi_: or rather, it did.. google cache shows it there.
791 2012-01-25 03:23:00 <roconnor> theymos: damn cache
792 2012-01-25 03:23:04 <roconnor> theymos: sorry about that
793 2012-01-25 03:23:20 <theymos> roconnor: np. Thanks for letting me know of a possible problem.
794 2012-01-25 03:23:32 <theymos> BBE testnet should handle reorgs OK, though I haven't tested it much. Mainnet shuts down when there's more than 5 blocks replaced.
795 2012-01-25 03:23:39 <gmaxwell> etotheipi_: 6acfd032e78a94d7611b5a2b00f190bc0c970224b39817562f3b1f1a7de6c9ee ConnectInputs failed 6acfd032e7
796 2012-01-25 03:23:41 <etotheipi_> gmaxwell, so most likely "connectInputs failed"
797 2012-01-25 03:23:57 <gmaxwell> etotheipi_: useless site! it doesn't show the imporant part of that log entry.
798 2012-01-25 03:24:14 <gmaxwell> luke-jr: 13500.
799 2012-01-25 03:24:34 <etotheipi_> I guess I just gotta check the inputs manually...
800 2012-01-25 03:24:35 <gmaxwell> etotheipi_: that log entry usually shows which input it can't connect.
801 2012-01-25 03:24:44 <gmaxwell> but the site isn't showing it.
802 2012-01-25 03:24:54 <etotheipi_> or actually, I'll just wait until I dump private keys and import into Armory
803 2012-01-25 03:25:06 <etotheipi_> it does a fresh-rescan of the blockchain and I can construct a new tx
804 2012-01-25 03:25:23 <gmaxwell> etotheipi_: would be nice to know what happened...
805 2012-01-25 03:25:59 <etotheipi_> gmaxwell, I'm not sure how helpful it will be... I don't even remember where the inputs are from or how I might've ended up with this
806 2012-01-25 03:26:16 <etotheipi_> the only thign I know is I tried to send the exact amt of the wallet, with 0 tx fee
807 2012-01-25 03:26:25 <roconnor> etotheipi_: they are in your transaction
808 2012-01-25 03:26:32 <roconnor> you should be able to trace them by hand
809 2012-01-25 03:26:48 <roconnor> at the very least
810 2012-01-25 03:27:14 <Joric> didn't know there are so many rejected blocks, is it normal?
811 2012-01-25 03:27:29 <etotheipi_> roconnor, I knwo how to find them... I'm just not sure what to do with the information...
812 2012-01-25 03:27:51 <roconnor> etotheipi_: you can plug the inputs into blockexplorer and see which ones are redeemed
813 2012-01-25 03:28:10 ferroh has quit (Ping timeout: 240 seconds)
814 2012-01-25 03:28:38 <roconnor> then hand craft a new transaction ommiting those inputs
815 2012-01-25 03:28:50 <etotheipi_> roconnor, I know exactly HOW to do it
816 2012-01-25 03:29:06 <etotheipi_> I'm just not seeing the priority when in a couple weeks I can do it in like 2 sec with Armory
817 2012-01-25 03:29:14 <roconnor> sure
818 2012-01-25 03:29:19 <etotheipi_> (once the dump -priv key update is available in satoshi client)
819 2012-01-25 03:29:37 <etotheipi_> I've hand-crafted waaaay too many tx... I just don't feel like digging up my old scripts and doing it again
820 2012-01-25 03:30:11 <roconnor> heh
821 2012-01-25 03:30:34 <roconnor> etotheipi_: send me your private keys and I'll do it for you. ^_^
822 2012-01-25 03:31:18 TheSeven has quit (Read error: Operation timed out)
823 2012-01-25 03:31:30 <Joric> etotheipi_, found out bsddb ships with python 2.7 exe, so the only dep is pycrypto that could be easily replaced with aes.py
824 2012-01-25 03:31:49 theymos has quit (Remote host closed the connection)
825 2012-01-25 03:32:14 <etotheipi_> Joric, if I use any of your code, I"ll probably just modify to use my existing crypto library
826 2012-01-25 03:32:23 TheSeven has joined
827 2012-01-25 03:32:33 <etotheipi_> do you know if bsddb is built into 2.6?
828 2012-01-25 03:32:49 <Joric> it doesn't really affect speed there's a few thousands sha512 iterations but it uses aes just once
829 2012-01-25 03:33:04 <etotheipi_> well, I've got both avail in Armory already
830 2012-01-25 03:33:05 <Joric> idk, i checked 2.7 it's there
831 2012-01-25 03:33:10 <etotheipi_> I use something similar in Armory
832 2012-01-25 03:35:17 <Joric> also i should use xrange here i think ) // for i in range(nDerivIterations)
833 2012-01-25 03:35:18 <etotheipi_> I don't know why I asked... I just imported bsddb in python and it worked (and I don't think I"ve installed it)
834 2012-01-25 03:35:56 <etotheipi_> so Joric, what do I have to run to pull my own private keys out of my encrypted wallet?
835 2012-01-25 03:36:26 <etotheipi_> roconnor, I just realized that since Armory goes through localhost, and localhost is the one with the invalid tx... it won't broadcast my new tx
836 2012-01-25 03:36:29 <Joric> etotheipi_, https://github.com/joric/pywallet it's pretty straightforward
837 2012-01-25 03:37:09 <roconnor> etotheipi_: you can run another bitcoin client under a new user
838 2012-01-25 03:37:26 <roconnor> or the same user in a new home bitcoin directory
839 2012-01-25 03:37:44 <etotheipi_> roconnor, yeah, or i can use that website
840 2012-01-25 03:37:52 <roconnor> oh right
841 2012-01-25 03:37:57 <roconnor> bitsend or whatever
842 2012-01-25 03:38:02 <etotheipi_> I'm just amused that the one node on the entire network that won't take my tx is the only one I'm connected to
843 2012-01-25 03:40:06 <luke-jr> [19:25:48] <luke-jr> etotheipi_: does your client support BIP 17 yet? :D
844 2012-01-25 03:41:04 <roconnor> luke-jr: etotheipi_ doesn't do validation
845 2012-01-25 03:41:46 ferroh has joined
846 2012-01-25 03:42:02 barmstrong has joined
847 2012-01-25 03:42:11 <etotheipi_> luke-jr, no it doesn't... I gotta iron out the regular client stuff first...
848 2012-01-25 03:42:19 <luke-jr> roconnor: oh
849 2012-01-25 03:42:34 <etotheipi_> but I will have to add, at least a big chunk of it, validation when I cut the umbilical cord to the Satoshi client
850 2012-01-25 03:42:42 <Joric> oddly reference client thinks the presense of 'key' and 'ckey' simultaneously is an error
851 2012-01-25 03:43:32 <Joric> it should have better error handling imo ) why not have both encrypted and unencrypted keys
852 2012-01-25 03:43:54 <roconnor> I'll probably make a BIP 17 branch for my code sometime, since it is so easy to implement.
853 2012-01-25 03:43:57 eoss has quit (Quit: Leaving)
854 2012-01-25 03:44:03 <gmaxwell> Joric: because people who want to cause you trouble could slip a bunch of unencrypted keys in your wallet and try to trick you into using them.
855 2012-01-25 03:44:44 <roconnor> gmaxwell: encryption is not integrety
856 2012-01-25 03:44:46 <gmaxwell> (you'll notice if they delete your encrypted ones in the process)
857 2012-01-25 03:44:50 <sipa> roconnor: easier than BIP16?
858 2012-01-25 03:45:01 <roconnor> sipa: much easier IMHO
859 2012-01-25 03:45:07 <roconnor> at least changes to the core
860 2012-01-25 03:45:21 <roconnor> for BIP16 I'll have to change my script parser.
861 2012-01-25 03:45:32 BLZNGPNGN has quit (Quit: ~qq)
862 2012-01-25 03:45:33 <etotheipi_> I guess I should start looking at BIP 0017
863 2012-01-25 03:45:37 <sipa> roconnor: how so?
864 2012-01-25 03:45:48 BLZNGPNGN has joined
865 2012-01-25 03:46:08 <roconnor> sipa: well, you want to parse first using the template, and if it suceeds then use that, otherwise parse using the old method
866 2012-01-25 03:46:17 <roconnor> then you need to branch on script eval depending on the parse
867 2012-01-25 03:46:34 <roconnor> I mean, maybe I haven't though it through well enough yet.
868 2012-01-25 03:47:03 <etotheipi_> oh sweet, BIP 0017 uses OP_CS
869 2012-01-25 03:47:12 <roconnor> BIP17 is a small change to the scriptMonad and implementing a new opcode
870 2012-01-25 03:47:38 <sipa> just parse scriptPubKey and scriptSig, then pattern match on scriptPubKey, if so, parse scriptSig's last instruction's data, and run that, otherwise run both scriptSig and scriptPubKey
871 2012-01-25 03:48:03 <sipa> i haven't seen that part of your code though, so i could be wrong :)
872 2012-01-25 03:48:23 <roconnor> sipa: you have to disassemble the sigscript into the code and stack pushes
873 2012-01-25 03:48:50 <roconnor> rather than executing the stack pushes I'd just set up the stack literally
874 2012-01-25 03:48:57 <sipa> right, true
875 2012-01-25 03:49:21 <roconnor> there is relatively a lot more code to write
876 2012-01-25 03:49:28 <roconnor> to do it properly
877 2012-01-25 03:49:43 <roconnor> I don't mean this is a bad thing
878 2012-01-25 03:49:51 <sipa> i guess it depends a lot on how you've structured your implementation
879 2012-01-25 03:50:11 <roconnor> BIP 16 is probably easier to implement if you do a half-assed job :)
880 2012-01-25 03:50:33 <roconnor> I mean I could actually execute the stack pushes
881 2012-01-25 03:50:45 <roconnor> so I don't have to write the stack serialization code
882 2012-01-25 03:51:18 <roconnor> but that would be bad code
883 2012-01-25 03:51:44 <roconnor> If you are going to have an invarient that all the code are stack pushes, then you might as well take advantage of it.
884 2012-01-25 03:51:51 num1 has quit (Quit: Textual IRC Client: http://www.textualapp.com/)
885 2012-01-25 03:52:04 num1 has joined
886 2012-01-25 03:52:25 <roconnor> The flip side of BIP16 is you get lots of guarentees
887 2012-01-25 03:53:43 <roconnor> and no more garbage in scriptsigs like in http://blockexplorer.com/testnet/tx/087e66d6f7869d5a3c73b77ad44bb435e582067cbf2c78415403c32778b16096#o0
888 2012-01-25 03:57:07 <gmaxwell> luke-jr: okay, I've been synced a bit..
889 2012-01-25 03:57:55 <gmaxwell> commencing unthievery.
890 2012-01-25 03:59:02 <roconnor> heh
891 2012-01-25 03:59:12 <roconnor> gmaxwell is the testnet police
892 2012-01-25 03:59:38 <gmaxwell> zomg 50% attack!
893 2012-01-25 04:00:08 <roconnor> I'm on the phone with the New York Times right now.
894 2012-01-25 04:00:20 <roconnor> telling them that bitcoin is breaking right NOW!
895 2012-01-25 04:00:22 <Joric> i'm the testnet-in-a-box police it's handy to have 0.25 difficulty
896 2012-01-25 04:00:30 <gmaxwell> [2012-01-24 22:54:53] Rejected 00000000.18cabadd.25b08253 BLOCK! GPU 4 thread 4
897 2012-01-25 04:00:35 <gmaxwell> @#!?
898 2012-01-25 04:00:53 <gmaxwell> How the heck does that happen?
899 2012-01-25 04:01:06 <gmaxwell> IM THE ONLY MINER IN THE UNIVERSE
900 2012-01-25 04:01:17 <roconnor> gmaxwell: you are double spending yourself!
901 2012-01-25 04:01:52 <gmaxwell> roconnor: nah, block solutions rejected because they're stale. solving so fast I'm colliding myself. :-/
902 2012-01-25 04:01:56 <gmaxwell> But it's not even that fast!
903 2012-01-25 04:02:22 <dwon> I just mined a testnet block a few minutes ago. Did I break something?
904 2012-01-25 04:02:33 <gmaxwell> dwon: no you didn't.
905 2012-01-25 04:03:01 <roconnor> dwon: but spend it while you can
906 2012-01-25 04:03:06 <gmaxwell> you may think you did... :)
907 2012-01-25 04:03:09 <gmaxwell> haha.
908 2012-01-25 04:03:33 <dwon> heh. what's your hash rate?
909 2012-01-25 04:03:34 <Diablo-D3> ?
910 2012-01-25 04:03:55 <gmaxwell> wtf man.. this cgminer stuff... kinda scarry.
911 2012-01-25 04:04:01 <gmaxwell> (5s):1991.2 (avg):1969.5 Mh/s | Q:328 A:14 R:8 HW:0 E:4% U:2.14/m
912 2012-01-25 04:04:51 <Diablo-D3> did you bork it?
913 2012-01-25 04:04:56 <gmaxwell> buncha stales .. and I'm the only miner in the universe.
914 2012-01-25 04:05:05 <Diablo-D3> wtf are you doing?
915 2012-01-25 04:05:11 <gmaxwell> Diablo-D3: reorging testnet.
916 2012-01-25 04:05:42 <roconnor> Diablo-D3: gmaxwell is doing opening a can of justice on testnet's ass
917 2012-01-25 04:06:08 <sipa> He is smiting it
918 2012-01-25 04:06:15 <gmaxwell> I'm half wayish.
919 2012-01-25 04:06:38 <gmaxwell> cutting down the max scan time solved my stupid orphans.
920 2012-01-25 04:07:21 MobiusL has quit (Remote host closed the connection)
921 2012-01-25 04:07:23 wirehead has quit (Ping timeout: 248 seconds)
922 2012-01-25 04:08:07 MobiusL has joined
923 2012-01-25 04:09:38 * gmaxwell drops another 2gh/s on it.
924 2012-01-25 04:11:39 <gmaxwell> "blocks" : 44792,
925 2012-01-25 04:11:48 <gmaxwell> everyone ready to watch your testnet logs soon?
926 2012-01-25 04:12:02 <gmaxwell> "blocks" : 44795,
927 2012-01-25 04:12:02 <gmaxwell> "blocks" : 44795,
928 2012-01-25 04:12:15 <roconnor> watching
929 2012-01-25 04:12:34 <gmaxwell> Okay, I'm tied.
930 2012-01-25 04:12:41 <gmaxwell> one ahead...
931 2012-01-25 04:12:47 <gmaxwell> "blocks" : 44799,
932 2012-01-25 04:13:07 <gmaxwell> k.. time for fireworks.
933 2012-01-25 04:13:22 <roconnor> dwon: quick mine 4 blocks
934 2012-01-25 04:13:27 * sipa starts testnet client
935 2012-01-25 04:13:39 <roconnor> IRC got join
936 2012-01-25 04:13:52 <roconnor> DelayedRepaint
937 2012-01-25 04:13:58 <gmaxwell> Yall ready?
938 2012-01-25 04:14:09 <roconnor> ooh
939 2012-01-25 04:14:19 <sipa> REORGANIZE
940 2012-01-25 04:14:20 <sipa> done
941 2012-01-25 04:14:23 <sipa> 44800
942 2012-01-25 04:14:48 RobinPKR has quit (Read error: Connection reset by peer)
943 2012-01-25 04:15:11 <gmaxwell> overshot a bit there...
944 2012-01-25 04:16:01 RobinPKR has joined
945 2012-01-25 04:16:02 * roconnor waits for BBE
946 2012-01-25 04:16:02 <gmaxwell> is bbe broken?
947 2012-01-25 04:16:29 <Mad7Scientist> I'm against encrypting the wallet
948 2012-01-25 04:16:35 <Mad7Scientist> I just think everyone should have a good backup
949 2012-01-25 04:16:41 <gmaxwell> Mad7Scientist: ... then don't?
950 2012-01-25 04:16:49 <Mad7Scientist> and if your laptop is stolen then immediately move all coins to a new wallet
951 2012-01-25 04:17:18 <gmaxwell> luke-jr: okay... so are your transactions all hunky dunky now?
952 2012-01-25 04:17:18 <roconnor> Mad7Scientist: and if your laptop is stolen and returned?
953 2012-01-25 04:17:46 <roconnor> gmaxwell: 50:50 that the transactions were not spent and how the stealer is at work
954 2012-01-25 04:17:53 <Mad7Scientist> it makes no differenc if the wallet is encrypted or not in that case
955 2012-01-25 04:18:01 <Mad7Scientist> they could just as easily install spyware
956 2012-01-25 04:18:12 <roconnor> Mad7Scientist: ya, you are probably right
957 2012-01-25 04:18:41 <gmaxwell> Mad7Scientist: is your laptop chained to your wrist?
958 2012-01-25 04:18:54 <gmaxwell> roconnor: I got transactions from him when the mining was going on.
959 2012-01-25 04:19:30 paraipan has quit (Remote host closed the connection)
960 2012-01-25 04:20:36 <gmaxwell> anyone have a use for 2350 TNBTC?
961 2012-01-25 04:21:27 * roconnor wishes BBE would update
962 2012-01-25 04:21:59 <gmaxwell> yea, I think it's busted.
963 2012-01-25 04:22:08 <sipa> oh no, you killed bbe
964 2012-01-25 04:22:37 <gmaxwell> oh there it goes!
965 2012-01-25 04:22:45 <gmaxwell> hahaha!
966 2012-01-25 04:22:50 <gmaxwell> 1 1 1 1 1 ..
967 2012-01-25 04:22:56 <Mad7Scientist> so I got bitcoin-qt compiled with -g3 and everything and now the I/O lockup when mining thing isn't happening now
968 2012-01-25 04:23:33 <gmaxwell> I guess I started at 44757
969 2012-01-25 04:23:55 <sipa> anyone know what the reasoning is behind deprioritizing connections to non-standard ports?
970 2012-01-25 04:24:12 paraipan has joined
971 2012-01-25 04:24:36 <gmaxwell> sipa: DOS attack.
972 2012-01-25 04:24:42 <sipa> explain?
973 2012-01-25 04:25:07 <gmaxwell> I announce yourhost:port and then a bunch of bitcoin nodes are connecting to it and yabbering some unwelcome noise before hanging up.
974 2012-01-25 04:25:19 <gmaxwell> I think BBE is kinda broken?
975 2012-01-25 04:26:55 <gmaxwell> I think its showing the old block content, not the ones I overwrote.
976 2012-01-25 04:27:03 <roconnor> I think so too
977 2012-01-25 04:27:11 <k9quaint> I think encrypting the wallet makes a difference in the stolen laptop scenario
978 2012-01-25 04:28:18 <gmaxwell> roconnor: yea, it revised the last 19 but none before that
979 2012-01-25 04:28:29 <gmaxwell> This is pretty awesome: https://blockexplorer.com/testnet/block/000000001c1dc1509cb24319ad048bc7089c9da4c3401050579c0044d176c20b
980 2012-01-25 04:28:30 <etotheipi_> as long as there is key-stretching, and/or a ton of entropy in the passphrase
981 2012-01-25 04:28:35 <gmaxwell> click previous block.
982 2012-01-25 04:28:50 <etotheipi_> unfortunately, so many users are going to pick ridiculously simple encryption passphrases,
983 2012-01-25 04:28:55 <BlueMatt> gmaxwell: wtf?
984 2012-01-25 04:29:20 <gmaxwell> etotheipi_: alas. they will. reference client uses 100ms of whatever the user's cpu is...
985 2012-01-25 04:29:37 <k9quaint> etotheipi_: users can also be fooled into transferring BTC directly to scammers and thieves with no need to steal the laptop
986 2012-01-25 04:29:46 <gmaxwell> (minimum of 25k iterations, which is 100ms of some fairly slow machine)
987 2012-01-25 04:30:08 <etotheipi_> I figured I would customize it in Armory, so if feel like using such a ludicrous passphrase, you can make it take an hour to unlock :)
988 2012-01-25 04:30:27 <gmaxwell> 100ms is enough that an okay password should survive. It's not enough to save "fuckyou" or "Password!" or "$bitcoin$"
989 2012-01-25 04:30:39 <gmaxwell> BlueMatt: BBE doesn't handle reorgs right at all.
990 2012-01-25 04:30:59 <roconnor> etotheipi_: you use scrypt?
991 2012-01-25 04:31:00 <k9quaint> I use the full text of the first 7 books of the Illiad as my passphrase
992 2012-01-25 04:31:19 <etotheipi_> roconnor, I use an algorithm in the same paper as scrypt
993 2012-01-25 04:31:25 <roconnor> good enough
994 2012-01-25 04:31:27 <etotheipi_> it has the same benefits, but it's a bit simpler to implement
995 2012-01-25 04:31:41 <k9quaint> in fact, I can give it out and still be secure as no human on earth can successfully string together that many asterisks without a typo
996 2012-01-25 04:31:42 <etotheipi_> it's just not quite as flexible
997 2012-01-25 04:31:43 <gmaxwell> etotheipi_: scrypt itself I feel pretty so so about.
998 2012-01-25 04:32:02 <etotheipi_> (ROMix is what it's called)
999 2012-01-25 04:32:03 <BlueMatt> gmaxwell: odd, I remember the author spending a ton of time reworking reorgs to work up to a certain number of blocks (like 8 or something)
1000 2012-01-25 04:32:05 <sipa> gmaxwell: how do you mean?
1001 2012-01-25 04:32:35 <etotheipi_> I just want to get unlocking above 1 MB/thread, effectively, completely disarming any chance of GPUs helping out
1002 2012-01-25 04:32:47 <sipa> scrypt looks like an extremely well-researched solution to the problem they try to tackle
1003 2012-01-25 04:33:08 <gmaxwell> sipa: I'm a big fan of the memory hard idea. the implementation choices in scrypt seemes a little random.
1004 2012-01-25 04:33:22 <etotheipi_> gmaxwell, that's why I liked ROMix... it's very pure
1005 2012-01-25 04:33:34 <etotheipi_> just sha512 and some XOR ops
1006 2012-01-25 04:33:44 <gmaxwell> the ltc people have now managed something like 3x speedup from their initial version, about half from algebraic simplifications to the hash.
1007 2012-01-25 04:34:13 <gmaxwell> (the other half from good old ASM elbow greese)
1008 2012-01-25 04:34:54 <gmaxwell> etotheipi_: you're using 1MB? Good. some of the upcoming gpu like manycore devices have a lot more cache then the gpus.
1009 2012-01-25 04:35:08 <etotheipi_> gmaxwell, it actually uses 32 MB on my system
1010 2012-01-25 04:35:22 <etotheipi_> it uses the highest value it can (up to the 32 MB) in the time you specify
1011 2012-01-25 04:35:25 <gmaxwell> though.. you do know what happens when you roll your own crypto. :)
1012 2012-01-25 04:35:32 <etotheipi_> it's not rolled
1013 2012-01-25 04:35:38 <etotheipi_> it's in published literature
1014 2012-01-25 04:35:43 <gmaxwell> "I use -funroll-loops!"
1015 2012-01-25 04:35:53 <etotheipi_> :)
1016 2012-01-25 04:36:07 <phantomcircuit> FUN ROLL LOOPS ARE DELICIOUS
1017 2012-01-25 04:36:27 <phantomcircuit> everytime i read that
1018 2012-01-25 04:36:29 <etotheipi_> and I read the paper... apparently ROMix is provably memory-hard, where scrypt "has nice properties, but they can't prove it"
1019 2012-01-25 04:36:41 <gmaxwell> phantomcircuit: I found a chinese company that will make you custom cereal boxes with whatever yo uwant.
1020 2012-01-25 04:37:00 <phantomcircuit> gmaxwell, omg yes
1021 2012-01-25 04:37:01 <gmaxwell> phantomcircuit: I wanted to get a box of -funroll-loops but they had a rather large minimum order
1022 2012-01-25 04:37:01 <etotheipi_> (I don't mean to discredit scrypt... it's great... I just didn't feel it necessary to integrate all these extra crypto modules)
1023 2012-01-25 04:37:02 <luke-jr> gmaxwell: BBE sees none
1024 2012-01-25 04:37:12 <gmaxwell> luke-jr: bbe is totally busted by this
1025 2012-01-25 04:37:24 <gmaxwell> luke-jr: https://blockexplorer.com/testnet/block/000000001c1dc1509cb24319ad048bc7089c9da4c3401050579c0044d176c20b < click previous block
1026 2012-01-25 04:38:07 <sipa> etotheipi_: scrypt uses ROMix
1027 2012-01-25 04:38:29 <gmaxwell> I'd use scrypt over $homegrown but ISTM they got the right core idea, and a good implementation of that.. then they waved their arms randomly around the rest. But at least it's published.
1028 2012-01-25 04:38:30 <etotheipi_> scrypt I thought was an extension of ROMix
1029 2012-01-25 04:38:30 <sipa> it extends it with other operation to compensate for the fact that memory-operations take finite time
1030 2012-01-25 04:39:25 <etotheipi_> right... ROMix is much more flexible in terms of balancing compute and mem-ops... but I didn't need that, and ROMix is very simple, and proven to be memory-hard... I'm not too concerned about it
1031 2012-01-25 04:39:37 <gmaxwell> I'm kinda surprised that they didn't manage to wedge ECOH in there too. ;)
1032 2012-01-25 04:39:46 <etotheipi_> err... scrypt is much more flexible
1033 2012-01-25 04:41:36 <luke-jr> gmaxwell: :/
1034 2012-01-25 04:42:04 <luke-jr> I'm stuck in safe mode -.-
1035 2012-01-25 04:43:13 <phantomcircuit> luke-jr, LOL
1036 2012-01-25 04:44:06 <luke-jr> how do I get out
1037 2012-01-25 04:44:14 <Joric> https://bitcointalk.org/index.php?topic=61044.0 a beautiful mind
1038 2012-01-25 04:44:47 <Joric> also lier there's no moneys :( https://bitcointools.appspot.com/?k=deadbeef
1039 2012-01-25 04:45:31 <phantomcircuit> Joric, stupidity is not something i believe anybody is particularly worried about protecting...
1040 2012-01-25 04:46:25 <phantomcircuit> gmaxwell, what's the minimum order :)
1041 2012-01-25 04:47:28 ferroh has quit (Ping timeout: 252 seconds)
1042 2012-01-25 04:47:35 user__ has quit (Quit: Leaving)
1043 2012-01-25 04:49:02 ferroh has joined
1044 2012-01-25 04:49:18 <dwon> Can we also add code to /bin/rm to prevent you from deleting ~/.ssh/id_rsa ?
1045 2012-01-25 04:50:34 <roconnor> sipa: my experimental transactions are gone
1046 2012-01-25 04:52:20 <Joric> 0xfacefeed doesn't have any btc on it aswell
1047 2012-01-25 04:54:45 JRWR has quit (Quit: "If we desire respect for the law, we must first make the law respectable." ~Louis D. Brandeis)
1048 2012-01-25 04:58:15 da2ce7 has joined
1049 2012-01-25 04:58:51 <k9quaint> I thought ROMix was only proven to be sequential memory hard with access to a true random oracle
1050 2012-01-25 04:58:52 <luke-jr> dwon: I alias'd 'rm' to 'echo NO U'
1051 2012-01-25 04:59:04 <dwon> luke-jr: heh
1052 2012-01-25 04:59:10 <luke-jr> so every time I habitually try to rm something without thinking, it stops me
1053 2012-01-25 04:59:15 <luke-jr> and I have to remember to type out the /bin/
1054 2012-01-25 04:59:24 <etotheipi_> k9quaint, where does the random oracle come in? it's a deterministic algorithm
1055 2012-01-25 05:00:19 <luke-jr> gmaxwell: sure you didn't re-mine some invalid txn? :P
1056 2012-01-25 05:00:30 <dwon> etotheipi_: hash functions are often referred to as "random oracles". They're not quite the same thing, but almost.
1057 2012-01-25 05:00:43 <CIA-2> bitcoin: ckolivas * r312724cea1c0 cgminer/adl.c: Adjust fan speed gently while in the optimal range when temperature is drifting to minimise overshoot in either direction. http://tinyurl.com/7aoadso
1058 2012-01-25 05:00:54 <dwon> etotheipi_: http://en.wikipedia.org/wiki/Random_oracle
1059 2012-01-25 05:01:06 <etotheipi_> oh
1060 2012-01-25 05:01:36 <etotheipi_> well the entirety of Bitcoin relies on SHA having that property (or at least being "close enough")
1061 2012-01-25 05:02:46 <k9quaint> um, SHA is not a random oracle
1062 2012-01-25 05:03:03 <k9quaint> if you give me a second, I can dig up the attack as proof
1063 2012-01-25 05:03:06 <luke-jr> k9quaint: it is when you give it random input
1064 2012-01-25 05:03:27 <gmaxwell> luke-jr: I couldn't have, I restarted and nuked my chainâ if I did I'd have to have gotten them from you.
1065 2012-01-25 05:03:39 <k9quaint> luke-jr: of course :P
1066 2012-01-25 05:03:40 <etotheipi_> k9quaint, the point is, it's "randomness" is good enough for the entire security model of bitcoin
1067 2012-01-25 05:03:43 <luke-jr> gmaxwell: hmm
1068 2012-01-25 05:03:55 <dwon> luke-jr: Nope. All MerkleÂDamgÃ¥rd hash functions are vulnerable to length-extension attacks.
1069 2012-01-25 05:03:57 <gmaxwell> k9quaint: No MD hash is.. but it's kind of boring.
1070 2012-01-25 05:04:09 <luke-jr> gmaxwell: sure you don't want to try next-test with BIP17 on your end? :P
1071 2012-01-25 05:04:36 da2ce7 has quit (2!~da2ce7@gateway/tor-sasl/da2ce7|Ping timeout: 276 seconds)
1072 2012-01-25 05:04:38 <gmaxwell> (what dwon said)
1073 2012-01-25 05:04:45 <dwon> http://en.wikipedia.org/wiki/Merkle%E2%80%93Damg%C3%A5rd_construction#Security_characteristics
1074 2012-01-25 05:04:47 <gmaxwell> dwon: do your miss your testnet btc yet?
1075 2012-01-25 05:05:08 wasabi2 has quit (Read error: Connection reset by peer)
1076 2012-01-25 05:05:19 <dwon> gmaxwell: "block 1b0b884af584bbbd, accepted" <-- I got some more
1077 2012-01-25 05:05:31 <gmaxwell> dwon: I can fix that too!
1078 2012-01-25 05:05:34 <roconnor> wtf
1079 2012-01-25 05:05:35 <roconnor> http://blockexplorer.com/testnet/block/000000000b36050cf688bf13a934eec2eb7d385c19cedeb8783deeaeb2092ea7
1080 2012-01-25 05:05:38 <dwon> gmaxwell: :P
1081 2012-01-25 05:05:43 <dwon> gmaxwell: you'll break BBE again
1082 2012-01-25 05:06:00 <gmaxwell> roconnor: it's all busted.
1083 2012-01-25 05:06:16 <luke-jr> so how do I get out of safe mode?
1084 2012-01-25 05:06:19 <roconnor> gmaxwell: is it legal for dupicate transactions to be in blocks?
1085 2012-01-25 05:06:22 <gmaxwell> Soâ good point here, we can't count on BBE to help us troubleshoot attacks on bitcoin, it can't even handle boring reorgs right.
1086 2012-01-25 05:06:30 JRWR has joined
1087 2012-01-25 05:06:39 <luke-jr> gmaxwell: neither can my client apparently
1088 2012-01-25 05:06:39 <roconnor> duplicate
1089 2012-01-25 05:06:45 <gmaxwell> roconnor: it's not a duplicate.. bbe is confused.
1090 2012-01-25 05:06:50 <k9quaint> dwon: so if you take a true random oracle and hash its output using SHA, its vulnerable to an attack?
1091 2012-01-25 05:06:51 <roconnor> okay
1092 2012-01-25 05:07:03 <doublec> does abe handle it better?
1093 2012-01-25 05:07:38 <luke-jr> I'm still at block 44760
1094 2012-01-25 05:07:54 <gmaxwell> luke-jr: hah. uh oh!
1095 2012-01-25 05:08:19 <gmaxwell> luke-jr: I _started_ at 44758 .. so you got some of mine but then got stuck?
1096 2012-01-25 05:08:22 * roconnor really really hopes no one mounts a 50% attack on mainnet
1097 2012-01-25 05:08:37 <gmaxwell> roconnor: mainnet sees 1/2 deep splits from time to time.
1098 2012-01-25 05:08:58 <roconnor> gmaxwell: maybe you minded a BIP17 illegal transaction
1099 2012-01-25 05:08:59 <gmaxwell> so I wonder why they haven't broken BBE â perhaps no conflicting txn in them.
1100 2012-01-25 05:09:02 <roconnor> *mined
1101 2012-01-25 05:09:07 <luke-jr> InvalidChainFound: invalid block=000000001b0b884af584 height=44805 work=13482036766980660
1102 2012-01-25 05:09:08 <luke-jr> InvalidChainFound: current best=00000000047ded1a1b6d height=44760 work=13480319789087790
1103 2012-01-25 05:09:11 <gmaxwell> roconnor: I don't know how I could have. I was only connected to luke.
1104 2012-01-25 05:09:27 barmstro_ has joined
1105 2012-01-25 05:09:31 <luke-jr> gmaxwell: can you dump your 44761?
1106 2012-01-25 05:09:32 <gmaxwell> roconnor: and I'd restarted, so I had no memory.
1107 2012-01-25 05:09:44 <roconnor> luke-jr had memory
1108 2012-01-25 05:09:53 <roconnor> maybe he relayed BIP 17 illegal blocks
1109 2012-01-25 05:09:54 <gmaxwell> roconnor: but he shouldn't have relayed bad txn..
1110 2012-01-25 05:10:00 <luke-jr> :/
1111 2012-01-25 05:10:21 <gmaxwell> he didn't have any illegal blocks.
1112 2012-01-25 05:10:21 <roconnor> explains the symptoms
1113 2012-01-25 05:10:35 <roconnor> gmaxwell: he may have had orphan transactions
1114 2012-01-25 05:10:42 <gmaxwell> the sequence was, he erased his chain and connected to me. (restarting in the process)
1115 2012-01-25 05:10:53 <gmaxwell> he resynced until the invalid point and stopped.
1116 2012-01-25 05:10:57 <roconnor> after all, the statoshi client keeps illegal transactions
1117 2012-01-25 05:11:03 <gmaxwell> Then I shut down and erased my chain.
1118 2012-01-25 05:11:04 <luke-jr> gmaxwell: you used -connect?
1119 2012-01-25 05:11:04 barmstrong has quit (Ping timeout: 240 seconds)
1120 2012-01-25 05:11:20 <gmaxwell> luke-jr: yes, and I checked I was only connected to you. my listen only accepts connections from you too
1121 2012-01-25 05:11:23 <gmaxwell> roconnor: in memory
1122 2012-01-25 05:11:27 <roconnor> gmaxwell: right you sent the illegal transactions to him
1123 2012-01-25 05:11:30 <luke-jr> gmaxwell: can you dump your 44761?
1124 2012-01-25 05:11:42 <gmaxwell> roconnor: yes, he should have not accepted them.
1125 2012-01-25 05:11:43 <roconnor> gmaxwell: he stashed them because the satoshi client is retarded
1126 2012-01-25 05:11:51 <roconnor> and then sent them back to you
1127 2012-01-25 05:11:53 <roconnor> which you accepted
1128 2012-01-25 05:12:12 <gmaxwell> roconnor: it only stashes its own txn in the wallet, non-self txn are only stored in memory.
1129 2012-01-25 05:12:16 <gmaxwell> we need a whiteboard.
1130 2012-01-25 05:12:21 <gmaxwell> luke-jr: how?
1131 2012-01-25 05:12:27 <roconnor> gmaxwell: the statoshi cleient stashes illegal transactions
1132 2012-01-25 05:12:30 <etotheipi_> out of random curiosity, I wonder what percentage of nodes need to fix multi-sig relay logic in order for the network to handle them "normally"...
1133 2012-01-25 05:12:33 <gmaxwell> roconnor: in memory!
1134 2012-01-25 05:12:43 <roconnor> gmaxwell: did luke-jr shut down before reconnecting to you?
1135 2012-01-25 05:12:44 <luke-jr> gmaxwell: getblockhash 44761
1136 2012-01-25 05:12:47 <luke-jr> gmaxwell: getblock <hash>
1137 2012-01-25 05:12:59 <luke-jr> roconnor: I think so
1138 2012-01-25 05:13:02 <gmaxwell> roconnor: yes! he had to to wipe the blocks
1139 2012-01-25 05:13:06 <roconnor> oh
1140 2012-01-25 05:13:08 <etotheipi_> I guess it depends mostly on the avg number of connections per node.... though I would guess you'd only need like 10% to reach am iner
1141 2012-01-25 05:13:11 <roconnor> well there goes that theory
1142 2012-01-25 05:13:15 <gmaxwell> 00000000140dc9cc6c849eab9652c3ad64b3e8c609f3bf69098a1ed4c15509b1
1143 2012-01-25 05:13:22 <roconnor> luke-jr: maybe you are the one stealing your coins?
1144 2012-01-25 05:13:24 <roconnor> :P
1145 2012-01-25 05:13:27 <luke-jr> gmaxwell: getblock 00000000140dc9cc6c849eab9652c3ad64b3e8c609f3bf69098a1ed4c15509b1
1146 2012-01-25 05:13:34 <gmaxwell> "tx" : [
1147 2012-01-25 05:13:34 <gmaxwell> "33cc930003d413e5e3dade604dd6b79428fdf89ba768194f48c39fd24e513841",
1148 2012-01-25 05:13:34 <gmaxwell> "cb2de044ea7001d5d87ed05816e9d0e5b4bab539ffed9d88b8727f93242eaf20",
1149 2012-01-25 05:13:34 <gmaxwell> "b5d0a285e5d3587bb1c7ec9239ba77b8097636e34169205b9d1de568b9cbc3f4",
1150 2012-01-25 05:13:34 <gmaxwell> "f1e0cadf564cd56d9705ede1f259d84bc472a81d34061c7a81a78d159fc80711",
1151 2012-01-25 05:13:36 <gmaxwell> "22a837d2d78b4e592b3d68d4144ce7b0290aa83cb7906ef6f15fb2caae439e3e"
1152 2012-01-25 05:13:37 <luke-jr> â¦
1153 2012-01-25 05:13:45 <gmaxwell> (thats it)
1154 2012-01-25 05:14:02 <luke-jr> gettransaction <all of those>
1155 2012-01-25 05:14:03 <gmaxwell> looking
1156 2012-01-25 05:14:14 <gmaxwell> first is the coinbase.
1157 2012-01-25 05:14:23 da2ce7 has joined
1158 2012-01-25 05:14:32 wasabi2 has joined
1159 2012-01-25 05:14:37 <gmaxwell> looks like gettransaction only works on my transactions. :(
1160 2012-01-25 05:14:49 <luke-jr> x.x
1161 2012-01-25 05:15:06 <gmaxwell> luke-jr: feel free to fix that and send me a patch.
1162 2012-01-25 05:15:12 * luke-jr grumbles at gavin reinventing a neutered dumpblock and merging it instead of the original well-tested dumpblock
1163 2012-01-25 05:16:38 <gmaxwell> you might try bbe ..
1164 2012-01-25 05:17:10 <gmaxwell> yea... they are there.
1165 2012-01-25 05:17:39 <roconnor> luke-jr: does your log say which transaction you rejected in the block?
1166 2012-01-25 05:17:47 <luke-jr> gmaxwell: not that one
1167 2012-01-25 05:17:51 <gmaxwell> doesn't have f1e
1168 2012-01-25 05:18:07 <gmaxwell> or 22a
1169 2012-01-25 05:18:11 Guest52283 has joined
1170 2012-01-25 05:18:21 <gmaxwell> luke-jr: also, those stupid error messages should log which txn triggered them.
1171 2012-01-25 05:18:21 <luke-jr> ERROR: ConnectInputs() : f1e0cadf56 VerifySignature failed
1172 2012-01-25 05:18:40 <sipa> gmaxwell: a gettransaction that works for all transactions it quite easy...
1173 2012-01-25 05:18:51 <sipa> i wonder why it doesn't exist
1174 2012-01-25 05:19:15 <gmaxwell> luke-jr: (and also what peer it got them from, I think all of our "sign of bad deeds" log entries should log ips.)
1175 2012-01-25 05:19:19 <roconnor> AcceptToMemoryPool(): accepted f1e0cadf56
1176 2012-01-25 05:19:20 has quit (Clown|!~clown@static-87-79-93-140.netcologne.de|Ping timeout: 245 seconds)
1177 2012-01-25 05:19:20 <roconnor> SetBestChain: new best=000000000348fa7591af height=44797 work=13481731526466372
1178 2012-01-25 05:19:22 <roconnor> MainFrameRepaint
1179 2012-01-25 05:20:00 <roconnor> oh that was the reorg
1180 2012-01-25 05:20:31 <gmaxwell> hmph.
1181 2012-01-25 05:20:39 <gmaxwell> I think luke might have given me it in the form of a bad block!
1182 2012-01-25 05:20:50 <luke-jr> ?
1183 2012-01-25 05:20:50 <gmaxwell> lemme post my debug.log
1184 2012-01-25 05:20:51 da2ce7 has quit (2!~da2ce7@gateway/tor-sasl/da2ce7|Ping timeout: 276 seconds)
1185 2012-01-25 05:22:01 * roconnor wonders what is in this mysterious transaction
1186 2012-01-25 05:22:09 <gmaxwell> http://people.xiph.org/~greg/testnet.debug.txt
1187 2012-01-25 05:22:10 * luke-jr wishes there was some way to export blocks/txns as hexdumps from bitcoind
1188 2012-01-25 05:23:01 Guest52283 has quit (Ping timeout: 252 seconds)
1189 2012-01-25 05:24:57 <gmaxwell> so... I boggle. Why did I askfor tx f1e0cadf564cd56d9705 0
1190 2012-01-25 05:25:02 <gmaxwell> I was only connected to luke at the time.
1191 2012-01-25 05:25:11 <gmaxwell> He shouldn't have forwarded that txn because it's invalid to him.
1192 2012-01-25 05:25:29 <roconnor> gmaxwell: maybe you are the coin stealer
1193 2012-01-25 05:25:41 <luke-jr> lol
1194 2012-01-25 05:26:13 <gmaxwell> It didn't happen until I'd solved a couple blocks...
1195 2012-01-25 05:26:22 <k9quaint> and all this time we thought it was Eleutheria that was stealing our coins
1196 2012-01-25 05:26:32 <k9quaint> turns out it was gmaxwell the whole time
1197 2012-01-25 05:26:33 MrTiggr has joined
1198 2012-01-25 05:26:37 <luke-jr> f1e0cadf56 is one of my earlier attempts :O
1199 2012-01-25 05:26:39 <roconnor> gmaxwell: but you asked for it from luke-jr, and he gave it to you
1200 2012-01-25 05:26:39 <gmaxwell> so luke and I had done those restarts.. I synced up.. sat for many minutes.. mined two blocks.. and the evil txn just ..emerged. :)
1201 2012-01-25 05:27:01 <gmaxwell> luke-jr: well shit! thats why, it was in your wallet, and it WILL relay it even if it thinks its invalid.
1202 2012-01-25 05:27:03 <luke-jr> maybe my wallet is corrupt -.-
1203 2012-01-25 05:27:24 <luke-jr> any way to purge txn history in my wallet?
1204 2012-01-25 05:27:26 <luke-jr> will -rescan help
1205 2012-01-25 05:27:28 <luke-jr> ?
1206 2012-01-25 05:27:34 <gmaxwell> roconnor: it doesn't log the advertisement.
1207 2012-01-25 05:27:44 <roconnor> ya, strange
1208 2012-01-25 05:27:52 <roconnor> though your -debug flag is off
1209 2012-01-25 05:27:57 <sipa> luke-jr: -rescan will only add/update things in your wallet, not delete
1210 2012-01-25 05:28:02 <gmaxwell> luke-jr: so YOU were the stealer all along!
1211 2012-01-25 05:28:18 <luke-jr> it's not a theft tho
1212 2012-01-25 05:28:41 <gmaxwell> roconnor: git diff tells me I'm not the stealer.
1213 2012-01-25 05:28:53 <luke-jr> sipa: how can I delete?
1214 2012-01-25 05:28:57 <roconnor> gmaxwell: clearly someone patched git as well
1215 2012-01-25 05:29:01 <sipa> luke-jr: not...
1216 2012-01-25 05:29:06 <luke-jr> -.-
1217 2012-01-25 05:29:21 * luke-jr ponders seeing if he can validate the txns he did on mainnetâ¦
1218 2012-01-25 05:29:31 <sipa> once i update rejectedtx, you'll be able to delete transactions from your wallet
1219 2012-01-25 05:30:06 <sipa> what is this "coin stealer", actually?
1220 2012-01-25 05:30:36 <CIA-2> bitcoin: ckolivas * rc009c6cd7403 cgminer/adl.c: Don't try to explicitly drift to exactly the target temperature, aim for just below it. http://tinyurl.com/7n5alnr
1221 2012-01-25 05:30:39 <roconnor> sipa: there appears to be a bot that takes BIP 17 transactions and steals them
1222 2012-01-25 05:30:41 <gmaxwell> sipa: I don't think I can show you now that BBE is a hunk of slag.
1223 2012-01-25 05:30:58 <gmaxwell> sipa: or probably any TXN that requires nothing to satisfy it.
1224 2012-01-25 05:30:59 <roconnor> sipa: by using the sigScript OP_1
1225 2012-01-25 05:31:30 MrTiggr has quit (Ping timeout: 240 seconds)
1226 2012-01-25 05:31:50 <luke-jr> there's no stray OP_NOP2s on mainnet, right?
1227 2012-01-25 05:32:01 MrTiggr has joined
1228 2012-01-25 05:33:27 * roconnor heads to bed
1229 2012-01-25 05:33:59 <dwon> So why wouldn't I see the 50 generated BTC when I run listaccounts 0 ?
1230 2012-01-25 05:35:05 <gmaxwell> dwon: because its not mature yet, which isn't the same as not confirmed.
1231 2012-01-25 05:35:24 <dwon> even when I set minconf=0 ?
1232 2012-01-25 05:35:37 <gmaxwell> Correct. minconf isn't relevant here.
1233 2012-01-25 05:35:46 <roconnor> dwon: mined transactions always require at least 100 confirmations before they are spendable
1234 2012-01-25 05:35:51 <gmaxwell> There is a _network rule_ that generated coin is unspendable for 100 blocks.
1235 2012-01-25 05:36:10 <gmaxwell> (the client imposes a somewhat higher standard than the network rule)
1236 2012-01-25 05:36:20 <gmaxwell> (of 120)
1237 2012-01-25 05:36:27 <dwon> oh, wow
1238 2012-01-25 05:36:47 <sipa> and the client completely hides it before it reaches depth 1
1239 2012-01-25 05:37:13 <sipa> because otherwise people complain too much about disappearing blocks, in the case of a stale ;)
1240 2012-01-25 05:37:35 <gmaxwell> the maturity check stuff is important. Normal transactions can get remined if they get orphaned, so long as someone wasn't being evil and respending.
1241 2012-01-25 05:37:47 <dwon> yeah, makes sense. ^^^ I was just going to speculate that.
1242 2012-01-25 05:38:01 <gmaxwell> But block can't be replaced.. and anything based on them vanishes forever to never return even if there was no evil.
1243 2012-01-25 05:38:52 <dwon> Also, a related question: When I mined those coins (using poclbm pointed directly at my bitcoind rpc port) the *address* still only showed up in my keypool, rather than in any account. I had to use setaccount. Is that normal?
1244 2012-01-25 05:39:22 <gmaxwell> generated txn don't show up in accountsâ limitation in the software.
1245 2012-01-25 05:39:44 <etotheipi_> gmaxwell, what's the easiest way to set up one of my miners to mine on testnet?
1246 2012-01-25 05:39:50 b4epoche_ has joined
1247 2012-01-25 05:39:59 <etotheipi_> do you have a server I can connect one of my GPUs to?
1248 2012-01-25 05:40:35 <sipa> etotheipi_: just point them to a bitcoind?
1249 2012-01-25 05:40:35 b4epoche has quit (Ping timeout: 245 seconds)
1250 2012-01-25 05:40:35 b4epoche_ is now known as b4epoche
1251 2012-01-25 05:41:07 <etotheipi_> do I have to set up RPC stuff?
1252 2012-01-25 05:41:22 <sipa> well, you need an rpc user and password
1253 2012-01-25 05:41:37 <sipa> but without those you can't run bitcoind
1254 2012-01-25 05:42:17 <etotheipi_> andi t's gotta be bitcoind?
1255 2012-01-25 05:42:22 <sipa> no
1256 2012-01-25 05:42:33 <sipa> bitcoin-qt in -server mode will work as well
1257 2012-01-25 05:42:37 <sipa> you never used an RPC call?
1258 2012-01-25 05:42:56 <etotheipi_> nope... I never needed it
1259 2012-01-25 05:43:05 <gmaxwell> wow, you're missing out. bitcoin cli is nice.
1260 2012-01-25 05:43:30 <etotheipi_> once I start doing more networking stuff, I'm sure I'll figure it out
1261 2012-01-25 05:43:32 <gmaxwell> my testnet setup is ~/src/bitcoin/src/bitcoind -testnet -rpcport=9992 --datadir=/tmp/ -rpcallowip=192.168.16.* -rpcuser=abcabc -rpcpassword=32413123421 -gen=0 -daemon
1262 2012-01-25 05:43:50 <gmaxwell> (you can through those params in the conf of course)
1263 2012-01-25 05:44:19 <gmaxwell> to use the rpc interface even if you don't have a conf... ~/src/bitcoin/src/bitcoind -testnet -rpcport=9992 --datadir=/tmp/ -rpcallowip=192.168.16.* -rpcuser=abcabc -rpcpassword=32413123421 help
1264 2012-01-25 05:44:31 * sipa doesn't dare to compile addrman... 550 lines of never-compiled code :S
1265 2012-01-25 05:45:18 <gmaxwell> sipa: the day I first wrote a 2000 line C program and it compiled and ran on the first time â I felt I finally knew C.
1266 2012-01-25 05:45:43 <gmaxwell> (I didn't. But hey)
1267 2012-01-25 05:46:01 <k9quaint> i was gonna say, that just makes you pedantic :P
1268 2012-01-25 05:46:52 <gmaxwell> I think I can't got more than 10 lines without pissing off the C++ compiler.
1269 2012-01-25 05:46:56 <gmaxwell> s/got/go/
1270 2012-01-25 05:47:23 <sipa> i don't think i've ever written more than 20 lines of C code without getting a compiler error
1271 2012-01-25 05:49:30 <gmaxwell> I should setup some wrapper to log the ones I get.
1272 2012-01-25 05:49:36 <etotheipi_> gmaxwell, is there anything I need to do with the miner itself? do I just point it to that IP and port, and use the same rpc pass/user? any testnet flags or anything?
1273 2012-01-25 05:49:59 <gmaxwell> etotheipi_: thats it, aim and fire. no flags or anything.
1274 2012-01-25 05:50:37 <CIA-2> bitcoin: ckolivas * r4995b7c241fa cgminer/adl.c: On dual GPUs, only autotune the shared fan once per gpu autotune cycle. http://tinyurl.com/799z3qc
1275 2012-01-25 05:54:43 <etotheipi_> wtf, I get "Could not locate specified kernel" ... which is weird
1276 2012-01-25 05:54:47 <k9quaint> gmaxwell: C is tolerant of line breaks between arithmetic operators, just linebreak after every scalar and operand to pad your results ;P
1277 2012-01-25 05:55:27 <etotheipi_> because when I run it from my run_miner script, I have no problem, but when I run the exact same command from the CLI, I get that error, even when I copy what I think is the identical command
1278 2012-01-25 05:56:18 <gmaxwell> etotheipi_: path setup for the opengl stuff?
1279 2012-01-25 05:56:40 <etotheipi_> I've been running 4 GPUs on this rig for 6 months straight
1280 2012-01-25 05:57:06 <etotheipi_> and my run_miner script works... but when I pull the CLI line out of it, it doesn't seem to work
1281 2012-01-25 05:57:48 <etotheipi_> ehh... doesn't matter now, I just made a copy of the script and hardcoded the change
1282 2012-01-25 05:58:45 <etotheipi_> looks like it works now... thanks!
1283 2012-01-25 06:01:16 <etotheipi_> oh... my bad... I just noticed that the python script uses the "cwd" options... so you were right
1284 2012-01-25 06:08:21 AAA_awright has quit (Read error: Connection reset by peer)
1285 2012-01-25 06:08:26 da2ce7 has quit (Quit: KVIrc 4.1.3 Equilibrium http://www.kvirc.net/)
1286 2012-01-25 06:08:36 AAA_awright has joined
1287 2012-01-25 06:12:48 BTC_Bear is now known as BTC_Bear|hbrntng
1288 2012-01-25 06:13:10 cosurgi has quit (Ping timeout: 240 seconds)
1289 2012-01-25 06:17:37 wirehead has joined
1290 2012-01-25 06:25:51 BGL has quit (Ping timeout: 276 seconds)
1291 2012-01-25 06:29:01 <Mad7Scientist> Your paste can be seen here: http://paste.pocoo.org/show/540379/
1292 2012-01-25 06:29:43 Tiggr has joined
1293 2012-01-25 06:30:09 Tiggr is now known as Guest65076
1294 2012-01-25 06:30:52 MrTiggr has quit (Ping timeout: 252 seconds)
1295 2012-01-25 06:31:17 da2ce7 has joined
1296 2012-01-25 06:32:08 <Mad7Scientist> #0 0xb6384871 in open () from /lib/libc.so.6
1297 2012-01-25 06:32:09 <Mad7Scientist> #1 0xb63288bf in _IO_file_open () from /lib/libc.so.6
1298 2012-01-25 06:32:09 <Mad7Scientist> #2 0xb6328a83 in _IO_file_fopen () from /lib/libc.so.6
1299 2012-01-25 06:32:09 <Mad7Scientist> #3 0xb631c300 in ?? () from /lib/libc.so.6
1300 2012-01-25 06:32:09 <Mad7Scientist> #4 0xb631c36d in fopen () from /lib/libc.so.6
1301 2012-01-25 06:32:09 <Mad7Scientist> #5 0x0809dbd1 in OpenBlockFile (nFile=1, nBlockPos=923920348,
1302 2012-01-25 06:32:41 <Mad7Scientist> pszMode=0x821dee9 "rb") at src/main.cpp:1455
1303 2012-01-25 06:33:01 MrTiggr has joined
1304 2012-01-25 06:33:55 <Mad7Scientist> (gdb) backtrace
1305 2012-01-25 06:33:55 <Mad7Scientist> #0 0xb6558e5d in pread64 () from /lib/libpthread.so.0
1306 2012-01-25 06:33:55 <Mad7Scientist> #1 0xb760d1bb in __os_io () from /usr/lib/libdb_cxx-4.8.so
1307 2012-01-25 06:34:02 <Mad7Scientist> ...
1308 2012-01-25 06:34:06 <Mad7Scientist> Well that's my I/O issue
1309 2012-01-25 06:34:22 <gmaxwell> thats different from the pastebin? whats with the pastebin?
1310 2012-01-25 06:34:23 <Mad7Scientist> 3 backtraces (the middle one cut off at #5
1311 2012-01-25 06:34:33 <Mad7Scientist> I interrupted it 3 times
1312 2012-01-25 06:34:36 <Mad7Scientist> 3 different ones
1313 2012-01-25 06:34:38 Guest65076 has quit (Ping timeout: 245 seconds)
1314 2012-01-25 06:36:04 hexTech has joined
1315 2012-01-25 06:36:10 <BlueMatt> ;;later tell devrandom feature request for gitian-builder/downloader: side output (ie debug symbols) that gets signed but that can be downloaded separately
1316 2012-01-25 06:36:10 <gribble> The operation succeeded.
1317 2012-01-25 06:36:14 <gmaxwell> can you put the three complete backtraces in a pastebin?
1318 2012-01-25 06:36:26 <Mad7Scientist> I'll get the full middle one
1319 2012-01-25 06:36:29 <Mad7Scientist> the last one only had two
1320 2012-01-25 06:37:54 <Mad7Scientist> Your paste can be seen here: http://paste.pocoo.org/show/540380/ (the one that goes to #5 in this chat)
1321 2012-01-25 06:40:13 <Mad7Scientist> 4th backtrace for tonight: http://paste.pocoo.org/show/540381/
1322 2012-01-25 06:41:58 <Mad7Scientist> backtrace #5 http://paste.pocoo.org/show/540382/
1323 2012-01-25 06:42:00 <k9quaint> thats on an NFS mount right?
1324 2012-01-25 06:42:05 <Mad7Scientist> yes
1325 2012-01-25 06:42:22 <k9quaint> they all hang during blocking file operations right?
1326 2012-01-25 06:42:48 <Mad7Scientist> I think the program can't receive signals during NFS file I/O
1327 2012-01-25 06:43:10 <Mad7Scientist> not sure what you mean
1328 2012-01-25 06:43:14 BlueMatt has quit (Quit: Ex-Chat)
1329 2012-01-25 06:43:29 <Mad7Scientist> The 0.3.whatever bitcoin didn't have this problem
1330 2012-01-25 06:43:29 <phantomcircuit> Mad7Scientist, nfs has a bad tendency to lockup randomly
1331 2012-01-25 06:43:44 <Mad7Scientist> what do you mean by lock up
1332 2012-01-25 06:44:26 <k9quaint> is this a linux system?
1333 2012-01-25 06:45:45 <Mad7Scientist> the thread that is hanging up is doing 8MB/sec read 50KB/sec write during this time
1334 2012-01-25 06:45:49 <Mad7Scientist> yes it is
1335 2012-01-25 06:46:25 <k9quaint> does the thread hang with status D?
1336 2012-01-25 06:46:53 <Mad7Scientist> yes
1337 2012-01-25 06:47:03 <k9quaint> did you set the mounts to be soft mounts in fstab?
1338 2012-01-25 06:48:08 <Mad7Scientist> users,intr,exec are my NFS flags
1339 2012-01-25 06:48:11 <Mad7Scientist> so I guess not
1340 2012-01-25 06:48:18 <k9quaint> try soft mounting
1341 2012-01-25 06:48:24 has joined
1342 2012-01-25 06:48:30 <k9quaint> if that doesnt work, try using sync writes
1343 2012-01-25 06:48:47 <k9quaint> and if that doesnt work, take that abortion of a protocol out back and shoot it in the head
1344 2012-01-25 06:48:55 <k9quaint> bury the corpse somewhere where the RFCs can't find it
1345 2012-01-25 06:50:09 <Mad7Scientist> hand/soft is just for a network problem I thought where with hard your programs just hang forever
1346 2012-01-25 06:50:41 sgstair has quit (Ping timeout: 252 seconds)
1347 2012-01-25 06:53:07 <Mad7Scientist> It started working fine new
1348 2012-01-25 06:53:45 <Mad7Scientist> from time to time these bursts of I/O lockups which last 20 - 40 seconds come around and the program is responsive for like 4 seconds in between those times
1349 2012-01-25 06:54:17 barmstro_ has quit (Remote host closed the connection)
1350 2012-01-25 06:54:21 lianj has quit (Ping timeout: 245 seconds)
1351 2012-01-25 06:54:49 lianj has joined
1352 2012-01-25 06:54:49 lianj has quit (Changing host)
1353 2012-01-25 06:54:49 lianj has joined
1354 2012-01-25 06:55:25 <Mad7Scientist> I guess on NFS without sync a network file can be modified but accessing it from another computer will get you the old file
1355 2012-01-25 06:55:38 <Mad7Scientist> unless you run sync on the computer that wrote to it
1356 2012-01-25 06:55:52 <Mad7Scientist> or some similar system calls are used
1357 2012-01-25 07:02:47 larsivi has quit (Ping timeout: 252 seconds)
1358 2012-01-25 07:07:08 MrTiggr has quit (Ping timeout: 248 seconds)
1359 2012-01-25 07:18:35 JRWR has quit (Remote host closed the connection)
1360 2012-01-25 07:30:24 sgstair has joined
1361 2012-01-25 07:40:49 bagit has joined
1362 2012-01-25 07:45:03 booo has joined
1363 2012-01-25 07:59:28 hexTech has quit (Ping timeout: 276 seconds)
1364 2012-01-25 08:00:11 auronandace has left ("Leaving")
1365 2012-01-25 08:02:16 booo has quit (Ping timeout: 245 seconds)
1366 2012-01-25 08:03:39 m00p has quit (Ping timeout: 260 seconds)
1367 2012-01-25 08:15:03 m00p has joined
1368 2012-01-25 08:15:33 MrTiggr has joined
1369 2012-01-25 08:19:30 larsivi has joined
1370 2012-01-25 08:22:52 sacarlson has quit (Ping timeout: 276 seconds)
1371 2012-01-25 08:30:39 larsivi has quit (Read error: Connection reset by peer)
1372 2012-01-25 08:35:24 abbe has quit (Quit: Heroes die once, Cowards live longer!)
1373 2012-01-25 08:36:44 sacarlson has joined
1374 2012-01-25 08:36:45 larsivi has joined
1375 2012-01-25 08:38:53 blomqvist has joined
1376 2012-01-25 08:39:00 blomqvist has quit (Changing host)
1377 2012-01-25 08:39:01 blomqvist has joined
1378 2012-01-25 08:40:40 mpr has quit (Ping timeout: 252 seconds)
1379 2012-01-25 08:52:01 mpr has joined
1380 2012-01-25 08:56:25 marf_away has joined
1381 2012-01-25 08:57:43 coblee_ has joined
1382 2012-01-25 09:01:01 coblee has quit (Ping timeout: 245 seconds)
1383 2012-01-25 09:01:01 coblee_ is now known as coblee
1384 2012-01-25 09:17:57 _sgstair has joined
1385 2012-01-25 09:18:00 sgstair has quit (Disconnected by services)
1386 2012-01-25 09:18:02 _sgstair is now known as sgstair
1387 2012-01-25 09:22:05 molecular has quit (Ping timeout: 240 seconds)
1388 2012-01-25 09:30:23 topi` has joined
1389 2012-01-25 09:34:25 dr_win has quit (Remote host closed the connection)
1390 2012-01-25 09:34:54 molecular has joined
1391 2012-01-25 09:40:36 marf_away has quit (Read error: Connection reset by peer)
1392 2012-01-25 09:41:16 marf_away has joined
1393 2012-01-25 09:41:55 da2ce7 has joined
1394 2012-01-25 09:42:41 MC1984 has quit (Ping timeout: 245 seconds)
1395 2012-01-25 09:42:42 Ken` has quit (Ping timeout: 245 seconds)
1396 2012-01-25 09:42:49 da2ce7 has quit (Ping timeout: 276 seconds)
1397 2012-01-25 09:43:06 MC1984 has joined
1398 2012-01-25 09:43:31 onelineproof has quit (Ping timeout: 245 seconds)
1399 2012-01-25 09:43:38 Ken` has joined
1400 2012-01-25 09:44:19 sacredch1o has joined
1401 2012-01-25 09:44:28 erle- has joined
1402 2012-01-25 09:46:43 sacredchao has quit (Ping timeout: 276 seconds)
1403 2012-01-25 09:49:16 Lexa has joined
1404 2012-01-25 09:49:27 onelineproof has joined
1405 2012-01-25 09:51:57 b4epoche_ has joined
1406 2012-01-25 09:52:51 b4epoche has quit (Ping timeout: 240 seconds)
1407 2012-01-25 09:52:51 b4epoche_ is now known as b4epoche
1408 2012-01-25 10:04:49 wasabi2 has quit (Read error: Connection reset by peer)
1409 2012-01-25 10:07:13 larsivi has quit (Ping timeout: 252 seconds)
1410 2012-01-25 10:08:30 sacarlson has quit (Read error: Connection reset by peer)
1411 2012-01-25 10:12:30 hexTech has joined
1412 2012-01-25 10:12:30 knotwork has quit (Read error: Connection reset by peer)
1413 2012-01-25 10:12:31 NickelBot has quit (Read error: Connection reset by peer)
1414 2012-01-25 10:13:35 knotwork has joined
1415 2012-01-25 10:16:07 larsivi has joined
1416 2012-01-25 10:21:52 NickelBot has joined
1417 2012-01-25 10:26:11 sacarlson has joined
1418 2012-01-25 10:38:56 cosurgi has joined
1419 2012-01-25 10:41:44 marf_away has quit (Read error: Connection reset by peer)
1420 2012-01-25 10:42:56 marf_away has joined
1421 2012-01-25 10:43:55 Karmaon has quit (Ping timeout: 276 seconds)
1422 2012-01-25 10:46:39 Karmaon has joined
1423 2012-01-25 10:58:34 slush1 has joined
1424 2012-01-25 11:00:49 slush has quit (Ping timeout: 276 seconds)
1425 2012-01-25 11:01:38 iocor has joined
1426 2012-01-25 11:08:13 slush has joined
1427 2012-01-25 11:08:13 sacarlson has quit (Read error: Connection reset by peer)
1428 2012-01-25 11:09:45 slush1 has quit (Ping timeout: 260 seconds)
1429 2012-01-25 11:10:47 booo has joined
1430 2012-01-25 11:11:40 RazielZ has joined
1431 2012-01-25 11:11:57 <Joric> where's 0.5.2 on github? which tag or branch?
1432 2012-01-25 11:16:36 slush has quit (Ping timeout: 240 seconds)
1433 2012-01-25 11:20:06 booo has quit (Ping timeout: 240 seconds)
1434 2012-01-25 11:20:45 <tcatm> Joric: looks like it is not tagged. maybe luke-jr knows which commit is 0.5.2?
1435 2012-01-25 11:23:09 wirehead has quit (Ping timeout: 255 seconds)
1436 2012-01-25 11:24:14 <Joric> looks like you were really in a hurry
1437 2012-01-25 11:26:51 sacarlson has joined
1438 2012-01-25 11:31:48 <UukGoblin> probably someone forgot to push the tag
1439 2012-01-25 11:32:01 <UukGoblin> it's sometimes tricky to see which tags are pushed
1440 2012-01-25 11:33:06 larsivi has quit (Quit: No Ping reply in 180 seconds.)
1441 2012-01-25 11:33:28 larsivi has joined
1442 2012-01-25 11:36:43 rdponticelli has quit (Ping timeout: 240 seconds)
1443 2012-01-25 11:46:23 <UukGoblin> (well, git ls-remote does it, but you can't easily see it in gitk)
1444 2012-01-25 12:06:09 rdponticelli has joined
1445 2012-01-25 12:07:06 iocor has quit (Quit: Computer has gone to sleep.)
1446 2012-01-25 12:10:21 <diki> was wondering
1447 2012-01-25 12:10:40 <diki> would gememorypool or something affect the database so that abe fails to scan it
1448 2012-01-25 12:10:41 <diki> ?
1449 2012-01-25 12:11:24 user__ has joined
1450 2012-01-25 12:14:17 paraipan has quit (Remote host closed the connection)
1451 2012-01-25 12:14:56 devrandom has quit (Ping timeout: 276 seconds)
1452 2012-01-25 12:15:52 paraipan has joined
1453 2012-01-25 12:19:53 iocor has joined
1454 2012-01-25 12:20:45 wirehead` has quit (Ping timeout: 255 seconds)
1455 2012-01-25 12:22:40 eldentyrell has quit (Ping timeout: 260 seconds)
1456 2012-01-25 12:27:46 <roconnor> anyone awake know the details of how repeated transaction IDs are handled in the core network rules?
1457 2012-01-25 12:28:35 <roconnor> e.g. if I mine TxID1 and spend it to TxID2, and then remine TxID1 again, can I still spend it to TxID3?
1458 2012-01-25 12:33:19 devrandom has joined
1459 2012-01-25 12:45:40 <UukGoblin> 'remine' on an alternative block, or one building up on the old one?
1460 2012-01-25 12:45:49 danbri has quit (Ping timeout: 248 seconds)
1461 2012-01-25 12:48:42 danbri has joined
1462 2012-01-25 12:49:05 danbri has quit (Remote host closed the connection)
1463 2012-01-25 12:49:16 danbri has joined
1464 2012-01-25 12:51:01 <roconnor> UukGoblin: just mine a block with the exact same output and producing a new transaction with an identical ID.
1465 2012-01-25 12:51:42 <roconnor> in a new block occuring after the block containing TxID2
1466 2012-01-25 12:53:40 <UukGoblin> ID of a transaction is a has of it, isn't it?
1467 2012-01-25 12:53:55 <UukGoblin> are you asking about a collision there?
1468 2012-01-25 12:54:08 <UukGoblin> s/has /hash /
1469 2012-01-25 12:54:22 <roconnor> well, it isn't a collision if exactly the same data is being hashed
1470 2012-01-25 12:54:36 <etotheipi_> geez roconnor, how much (little) do you sleep?
1471 2012-01-25 12:55:03 <roconnor> UukGoblin: if you mine to the same public key and use exactly same extranonce you get exactly the same transaction ID.
1472 2012-01-25 12:55:05 <etotheipi_> I thought you were going to bed like 6 hours ago
1473 2012-01-25 12:55:09 <roconnor> etotheipi_: I got up early today
1474 2012-01-25 12:55:14 <roconnor> 7 hours ago
1475 2012-01-25 12:55:33 <UukGoblin> if exactly the same data is re-sent it'd be a double spend, so I guess it'd get rejected
1476 2012-01-25 12:55:50 <UukGoblin> but I'm sure I'm missing something there
1477 2012-01-25 12:56:13 <roconnor> but it isn't really a double spend if the first use of the TxID1 is already spent
1478 2012-01-25 12:56:54 <roconnor> etotheipi_: building a fully validating client is so hard
1479 2012-01-25 12:57:12 <etotheipi_> roconnor, that's why I didn't do it :)
1480 2012-01-25 12:58:31 <UukGoblin> oh... that'd just be a doubled transaction?
1481 2012-01-25 12:59:02 <UukGoblin> meh, I'm lost
1482 2012-01-25 12:59:12 <roconnor> step 1 mine a coin
1483 2012-01-25 12:59:40 <roconnor> lets say this coin is the output of TxID1234
1484 2012-01-25 13:00:13 <UukGoblin> if you "mine" it it'd have no previous output
1485 2012-01-25 13:00:20 <roconnor> step 2 wait for 100 confirmation and then spend this coin. Let say that TxID5678 spends this coin.
1486 2012-01-25 13:00:24 <roconnor> UukGoblin: right
1487 2012-01-25 13:00:46 <UukGoblin> ah, right
1488 2012-01-25 13:00:55 <roconnor> so now TxID1234 is all spent and can be "freed" in priniciple.
1489 2012-01-25 13:00:56 <UukGoblin> right
1490 2012-01-25 13:00:57 <Joric> dammit how to upload tag to github? git push says everything is up to date
1491 2012-01-25 13:01:11 booo has joined
1492 2012-01-25 13:01:21 <UukGoblin> Joric, git push tag v0.5.2
1493 2012-01-25 13:01:25 <roconnor> step 3 mine again using exactly the same output publick key and exactly the same extranonce.
1494 2012-01-25 13:01:29 <UukGoblin> nah
1495 2012-01-25 13:01:36 <roconnor> this produces a new coin
1496 2012-01-25 13:01:50 <UukGoblin> Joric, or push origin tag v0.5.2
1497 2012-01-25 13:01:52 <roconnor> but when you hash it you get again TxID1234, because exactly the same data is being hashed.
1498 2012-01-25 13:02:22 <Joric> UukGoblin, my hero!
1499 2012-01-25 13:02:31 <roconnor> can you, step 4, spend this coin? or will bitcoin complain that it is already spent by TxID5678
1500 2012-01-25 13:03:56 <UukGoblin> Joric, there was also push --tags or sth to push all of them
1501 2012-01-25 13:04:26 <UukGoblin> roconnor, ah, I get it...
1502 2012-01-25 13:04:47 <UukGoblin> coinbase also has a timestamp, so that'd have to be identical too
1503 2012-01-25 13:04:56 <roconnor> ya
1504 2012-01-25 13:04:57 <UukGoblin> unless you screwed with it in the first place
1505 2012-01-25 13:05:21 <roconnor> or just use an entirely empty coinbase
1506 2012-01-25 13:07:05 <UukGoblin> nice question :)
1507 2012-01-25 13:07:37 d4de has quit (Ping timeout: 252 seconds)
1508 2012-01-25 13:08:53 <UukGoblin> I'd hope the block from step 3 would be rejected
1509 2012-01-25 13:09:05 <roconnor> I'm 80% sure it won't be
1510 2012-01-25 13:10:06 <roconnor> thought I think that would be a reasonable patch ... unless dupilicate coinbase coins already exist
1511 2012-01-25 13:11:19 <UukGoblin> well is the block's timestamp somewhere else than coinbase?
1512 2012-01-25 13:11:35 <UukGoblin> yeah IU think it would be
1513 2012-01-25 13:11:44 <roconnor> there is a timestamp in block
1514 2012-01-25 13:11:50 <roconnor> but that won't affect the transaction ID
1515 2012-01-25 13:12:25 agricocb has quit (Quit: Leaving.)
1516 2012-01-25 13:12:28 <UukGoblin> yeah
1517 2012-01-25 13:14:44 paraipan has quit (Quit: Saliendo)
1518 2012-01-25 13:17:54 ski_ has quit (Ping timeout: 255 seconds)
1519 2012-01-25 13:21:10 ski_ has joined
1520 2012-01-25 13:21:13 d4de has joined
1521 2012-01-25 13:21:30 <etotheipi_> roconnor, btw, I updated Armory to use C++ ripemd160... so you shouldn't have that problem anymore
1522 2012-01-25 13:21:45 <etotheipi_> (it's compiled into Armory now, so there can't be any problem)
1523 2012-01-25 13:21:59 <roconnor> etotheipi_: I think that is wise
1524 2012-01-25 13:22:08 <roconnor> I haven't gotten aroudn to updating Armory yet
1525 2012-01-25 13:22:11 <etotheipi_> roconnor, I had no idea that some people would have a problem with that
1526 2012-01-25 13:22:21 <etotheipi_> roconnor, you can wait... I still gotta figure out this last bug
1527 2012-01-25 13:22:40 <UukGoblin> https://raw.github.com/goblin/gbl-btc-doc/master/btc_txn.txt - I guess neither some_hash() nor simplified_version_of() (from At1's txin1 use anything else than the transaction's data?
1528 2012-01-25 13:23:05 <etotheipi_> it's going to be *amazing* once I do --- every tx-related issue will be resolved... right now the ledger/balances don't update quite right without restarting Armory
1529 2012-01-25 13:23:19 <UukGoblin> you guys writing a new client?
1530 2012-01-25 13:23:57 <etotheipi_> Uuk, I am: http://bitcoinarmory.com/
1531 2012-01-25 13:24:21 <etotheipi_> actually: http://bitcoinarmory.com/index.php/start-page/sample-page/screenshots
1532 2012-01-25 13:25:42 <UukGoblin> cool
1533 2012-01-25 13:26:16 gavinandresen has joined
1534 2012-01-25 13:26:52 <CIA-2> bitcoin: Gavin Andresen master * rb6a35d2 / (4 files in 2 dirs): Merge pull request #773 from gavinandresen/p2shSigOpCount ... https://github.com/bitcoin/bitcoin/commit/b6a35d2d5235585b728c4373d6c0e69e563caec6
1535 2012-01-25 13:27:10 <UukGoblin> hi gavin :) roconnor just asked an interesting question
1536 2012-01-25 13:27:26 <gavinandresen> Easy to summarize or should I go read the logs?
1537 2012-01-25 13:27:52 paraipan has joined
1538 2012-01-25 13:28:13 <UukGoblin> I think last hour's logs would be best
1539 2012-01-25 13:29:37 <gavinandresen> About creating a "duplicate" transaction? The transaction hash depends on the hashes of all the transaction's inputs, so you can only create same-txid-but-different-inputs in a coinbase.
1540 2012-01-25 13:30:00 <gavinandresen> And there are a couple of those in the main chain (whoever screwed up their mining software can't spend them)
1541 2012-01-25 13:31:14 <gavinandresen> Unless I'm misinterpreting roconnor's question.... muck with the scriptSig and you can create "same transaction, different txids", but I know roconnor knows that....
1542 2012-01-25 13:31:18 <UukGoblin> same-txid-but-different-inputs? wouldn't different inputs imply a different txid (short of hash collision)?
1543 2012-01-25 13:31:33 <gavinandresen> afk for a while, back later
1544 2012-01-25 13:36:48 ski_ has quit (Ping timeout: 255 seconds)
1545 2012-01-25 13:39:47 Guest61378 has joined
1546 2012-01-25 13:41:06 sje has joined
1547 2012-01-25 13:41:06 sje has quit (Changing host)
1548 2012-01-25 13:41:07 sje has joined
1549 2012-01-25 13:42:53 sneak has quit (Ping timeout: 252 seconds)
1550 2012-01-25 13:43:06 ovidiusoft has joined
1551 2012-01-25 13:44:51 sneak has joined
1552 2012-01-25 13:47:28 Guest61378 has quit (Quit: kthxbai)
1553 2012-01-25 13:47:35 <gavinandresen> UukGoblin: I should have said same-txid-appearing-twice-in-the-blockchain
1554 2012-01-25 13:47:47 datagutt_ has joined
1555 2012-01-25 13:49:16 <CIA-2> bitcoin: Gavin Andresen master * r4c932cc / src/db.cpp : Merge pull request #782 from sipa/checkkeys ... https://github.com/bitcoin/bitcoin/commit/4c932cca6f579bfb11e0144415c3004d382b8e42
1556 2012-01-25 13:50:52 etotheipi_ has quit (Ping timeout: 240 seconds)
1557 2012-01-25 13:51:27 <UukGoblin> gavinandresen, that'd be "... so you can only create same-txid-appearing-twice-in-the-blockchain in a coinbase."
1558 2012-01-25 13:55:30 ski_ has joined
1559 2012-01-25 13:56:09 <gmaxwell> Testnet BBE is still molten slag.
1560 2012-01-25 13:58:48 iocor has quit (Quit: Computer has gone to sleep.)
1561 2012-01-25 13:58:49 <gmaxwell> ;;later tell theymos BBE doesn't handle (deep) reorgs and the testnet results are now pasta.
1562 2012-01-25 13:58:49 <gribble> The operation succeeded.
1563 2012-01-25 13:59:50 pickett_ has quit (Remote host closed the connection)
1564 2012-01-25 14:00:30 pickett_ has joined
1565 2012-01-25 14:00:57 <UukGoblin> gmaxwell, oh and putting stuff in the search box (on mainnet) is quite slow
1566 2012-01-25 14:01:09 <UukGoblin> like, very slow
1567 2012-01-25 14:01:18 <UukGoblin> I'm not sure if it actually comes back with results
1568 2012-01-25 14:01:43 iocor has joined
1569 2012-01-25 14:01:51 iocor has quit (Client Quit)
1570 2012-01-25 14:01:59 <gmaxwell> UukGoblin: yea, it doesâ (unless it just stopped working) but it's been getting slower.
1571 2012-01-25 14:02:24 graingert has joined
1572 2012-01-25 14:02:28 <gmaxwell> UukGoblin: I reorged testnet last night and the testnet results are wrong and psycho now.
1573 2012-01-25 14:02:57 b4epoche_ has joined
1574 2012-01-25 14:03:07 etotheipi_ has joined
1575 2012-01-25 14:03:08 <UukGoblin> gmaxwell, try searching for b33145e9910fb6842c03c9b353017e03183e84ce63fc5eae973
1576 2012-01-25 14:03:24 agricocb has joined
1577 2012-01-25 14:03:37 b4epoche has quit (Ping timeout: 244 seconds)
1578 2012-01-25 14:03:37 b4epoche_ is now known as b4epoche
1579 2012-01-25 14:03:59 <gavinandresen> Can I get a quick reality check? Does anybody besides Luke prefer BIP 17 to BIP 16 ?
1580 2012-01-25 14:04:34 <gavinandresen> (well, Luke and apparently Steve from BitPay)
1581 2012-01-25 14:05:12 <UukGoblin> gavinandresen, I think I do, but I'm not 100% sure yet
1582 2012-01-25 14:05:29 <gmaxwell> gavinandresen: less obvious than I thought it was 24 hours ago. roconnor went throught the patch and implemented it and was making purring sounds about it.
1583 2012-01-25 14:06:01 <gavinandresen> Has roconnor implemented BIP 16 ?
1584 2012-01-25 14:06:33 JRWR has joined
1585 2012-01-25 14:08:05 <lianj> due to mining power he forced bip17 txs into the mainnet :D
1586 2012-01-25 14:08:25 <gavinandresen> lianj: did he manage to redeem them?
1587 2012-01-25 14:08:40 <gmaxwell> he there is luke
1588 2012-01-25 14:08:55 <lianj> yes
1589 2012-01-25 14:09:24 <UukGoblin> gmaxwell, that search timed out after ~6 minutes or so
1590 2012-01-25 14:09:24 <etotheipi_> dumb question: how can such scripts be redeemed on main-net without other nodes supporting it?
1591 2012-01-25 14:09:41 <gmaxwell> gavinandresen: it didn't sound like roconnor had, but it sounded like he started to. search the backscroll here for "roconnor> I'll probably make a BIP 17 branch for my code sometime"
1592 2012-01-25 14:09:43 <gavinandresen> Mine a block containing both the funding and redeeming tx
1593 2012-01-25 14:10:01 <gmaxwell> gavinandresen: are you running the BIP17 stealer on testnet?
1594 2012-01-25 14:10:07 <gmaxwell> :)
1595 2012-01-25 14:10:10 <etotheipi_> gavinandresen, wouldn't other nodes reject it because they don't understand BIP 0017 and consider it invalid?
1596 2012-01-25 14:10:27 <gavinandresen> etotheipi_: no, backwards compatible....
1597 2012-01-25 14:10:27 <UukGoblin> ;;later tell theymos oh and searches on BBE time out - try searching for b33145e9910fb6842c03c9b353017e03183e84ce63fc5eae973
1598 2012-01-25 14:10:28 <gribble> The operation succeeded.
1599 2012-01-25 14:10:34 <gmaxwell> etotheipi_: no. Thats not how it works, if it did it would be completely non deployable.
1600 2012-01-25 14:10:50 <etotheipi_> gmaxwell, I know... that's why I'm confused and wish I had paid more attention
1601 2012-01-25 14:11:00 <gmaxwell> etotheipi_: it only makes some valid things invalid, not the other way around.
1602 2012-01-25 14:11:37 <gavinandresen> If BIP 17 had been proposed back in October, then I might have been for it. But I would have proposed a different BIP 11, to get around the CHECKMULTISIG==20 sigops rules
1603 2012-01-25 14:11:56 <etotheipi_> so non-BIP0017 nodes are willing to accept OP_NOP2 scripts as long as they are already in a block?
1604 2012-01-25 14:11:58 <gmaxwell> On testnet there is some agent that steals BIP17 transactions, though, it appears.
1605 2012-01-25 14:12:20 <gmaxwell> etotheipi_: Yes, thats what the NOPs are for.
1606 2012-01-25 14:12:29 <lianj> etotheipi_: yes, true for every nonstandard but valid tx
1607 2012-01-25 14:12:57 <etotheipi_> it just seems weird that a full-validation node would allow a tx to be considered valid if it can't fully evaluate it
1608 2012-01-25 14:13:11 slush has joined
1609 2012-01-25 14:13:30 <UukGoblin> gavinandresen, so in short, roconnor was asking about how is same-txid-same-data-in-generation-transaction handled for multiple generations
1610 2012-01-25 14:13:31 <lianj> my script runner said 'meh, raise error' though because i had not implemeneted NOP2 ^^
1611 2012-01-25 14:13:32 <etotheipi_> or rather, it would allow the block into it's own blockchain if it can't evaluate the script
1612 2012-01-25 14:13:39 <gmaxwell> gavinandresen: the checkmultisig thing is still really unfortunate. Otherwise BIP17 seems okay dokey.
1613 2012-01-25 14:13:45 <UukGoblin> i.e. you generate/mine 2 blocks with identical generation transaction
1614 2012-01-25 14:14:24 <gmaxwell> UukGoblin: he was? are you looking at where he was confused by the totally insane and busted testnet bbe results?
1615 2012-01-25 14:14:38 <UukGoblin> gmaxwell, no, I'm looking at a question he asked about 2 hours ago in here
1616 2012-01-25 14:14:47 <gavinandresen> gmaxwell: I still worry about exactly what the OP_CHECKSIGS are signing. It looked to me like Luke's code had a bug there, but I didn't spend time thinking deeply about it
1617 2012-01-25 14:14:49 <UukGoblin> more like 1h50
1618 2012-01-25 14:16:15 <gavinandresen> UukGoblin: if you have a majority of hashing power you can confuse clients using duplication coinbase transactions. That's a known issue in the category of "if you have majority hashing power you can do much worth things so not worth fixing"
1619 2012-01-25 14:16:30 <gmaxwell> gavinandresen: yea, I have no clue if the code is busted junk or not. We were trying to produce some example transactions last night on testnet but broke bbe in the process.
1620 2012-01-25 14:17:16 <gavinandresen> gmaxwell: that's why I want the reality check: why spend all the time banging on BIP 17 if nobody actually thinks it is better?
1621 2012-01-25 14:17:20 <UukGoblin> gavinandresen, nono, /without/ majority of hashpower
1622 2012-01-25 14:17:24 <gmaxwell> gavinandresen: roconnor asked a more subtle question: Say you currently produce dupe coinbases, you can only spend one. Fine. Say you produce one, spend it, then produce it again.
1623 2012-01-25 14:17:27 <UukGoblin> you just mine 2 regular blocks
1624 2012-01-25 14:17:28 <gavinandresen> (except that banging on things in general is good)
1625 2012-01-25 14:17:51 <UukGoblin> with identical coinbase and identical pubkeys in the generation transaction, so that there's actually 2 identical transactions with the same ID
1626 2012-01-25 14:17:52 <gmaxwell> gavinandresen: because banging on things is good in general, we found out bbe is broken by some reorgs.
1627 2012-01-25 14:18:34 <gavinandresen> gmaxwell: the rest of the network will see the second as a double spend, so it can never be spent
1628 2012-01-25 14:19:11 <UukGoblin> gavinandresen, but the first transaction, after it's spent, can theoretically be pruned
1629 2012-01-25 14:19:15 <gmaxwell> UukGoblin: we have blocks like that already.
1630 2012-01-25 14:19:41 <gmaxwell> UukGoblin: you can't accept third party pruned blocks without making yourself greatly vulnerable already.
1631 2012-01-25 14:20:06 <UukGoblin> I guess miner nodes can't prune at all
1632 2012-01-25 14:21:00 <gmaxwell> UukGoblin: they can. oh. hm.
1633 2012-01-25 14:21:03 <gavinandresen> RE: pruning: pruning nodes could just store all the coinbase transactions.
1634 2012-01-25 14:21:20 <gmaxwell> Yea, thats kind of ugly but seems correct.
1635 2012-01-25 14:21:21 <makomk> gmaxwell: by the way, I have a feeling that there are probably *still* an awful lot of pushpoold-based pools out there that use getwork.
1636 2012-01-25 14:21:22 <gavinandresen> ... and prune the rest. It is a good edge case
1637 2012-01-25 14:21:59 <gavinandresen> If I could go back in time I'd make duplicate coinbases 100% illegal. Going forward, probably a good idea to "discourage" them, maybe eventually ban....
1638 2012-01-25 14:22:31 <UukGoblin> so the blocks would get accepted, but people couldn't spend them because they'd be seen as duplicate
1639 2012-01-25 14:22:43 cryptoxchange has joined
1640 2012-01-25 14:22:43 cryptoxchange has quit (Changing host)
1641 2012-01-25 14:22:43 cryptoxchange has joined
1642 2012-01-25 14:23:54 larsivi has quit (Ping timeout: 252 seconds)
1643 2012-01-25 14:24:29 <gmaxwell> gavinandresen: I don't think there would be any grave risk of putting in a check that makes them 100% illegal after, say, height 210000... if it was introduced nowish.
1644 2012-01-25 14:25:39 <gmaxwell> I guess the problem is that the check required to make them illegal is _also_ pruning incompatible.
1645 2012-01-25 14:25:54 <gavinandresen> gmaxwell: very true!
1646 2012-01-25 14:26:42 <UukGoblin> also, would it make just the dupe-coinbase illegal, or coinbase+pubkey combo?
1647 2012-01-25 14:27:13 <gmaxwell> UukGoblin: by coinbase we mean the coinbase txn hash which covers all of it including the pubkey.
1648 2012-01-25 14:27:23 <gavinandresen> yup
1649 2012-01-25 14:28:06 <UukGoblin> right
1650 2012-01-25 14:29:15 <UukGoblin> well, coming back in time, I'd perhaps have the generation txn contain the hash of the previous block in the 'from_addr' field...
1651 2012-01-25 14:29:25 <UukGoblin> that'd teach them
1652 2012-01-25 14:30:11 <gmaxwell> that would have made the problem impossible. yup.
1653 2012-01-25 14:30:34 <gavinandresen> from_addr field?
1654 2012-01-25 14:30:42 <gmaxwell> the input..
1655 2012-01-25 14:30:47 <UukGoblin> from. whatever it's called
1656 2012-01-25 14:30:58 <gmaxwell> make the generation txn have an "input" field which is just the _block_ hash of the previous block.
1657 2012-01-25 14:31:16 <gmaxwell> (and of course the network rule would be to check that differently than normal inputs)
1658 2012-01-25 14:31:19 <gavinandresen> vin[0].scriptSig[0...32] == previous block hash. Good Idea.
1659 2012-01-25 14:31:42 <UukGoblin> scriptSig, that's it
1660 2012-01-25 14:31:58 <gavinandresen> Wait, no, you don't know "the" previous block
1661 2012-01-25 14:32:05 <gavinandresen> Wait, yes you do...
1662 2012-01-25 14:32:18 <UukGoblin> you perhaps don't know current, but previous you should know
1663 2012-01-25 14:32:23 <gavinandresen> (gotta get some caffeine in me...)
1664 2012-01-25 14:32:42 <UukGoblin> no hang on...
1665 2012-01-25 14:32:47 <UukGoblin> hrm.
1666 2012-01-25 14:32:55 <gmaxwell> well, it would mean you'd have to toss and rebuild the coinbase txn every block, but thats no big deal.
1667 2012-01-25 14:33:31 <UukGoblin> https://github.com/goblin/gbl-btc-doc/blob/master/btc_txn.txt <- I called it "prev_tx" in At1::txin1 here
1668 2012-01-25 14:33:57 <UukGoblin> coinbase/scriptSig would stay... but nvmd, it's not changeable now anyway
1669 2012-01-25 14:34:18 <UukGoblin> (in Ag1 it was just "no previous output"
1670 2012-01-25 14:34:19 <UukGoblin> )
1671 2012-01-25 14:34:32 <gmaxwell> it doesn't even matter where you put it.
1672 2012-01-25 14:34:36 erle- has quit (Quit: erle-)
1673 2012-01-25 14:34:47 <UukGoblin> https://en.bitcoin.it/wiki/Transactions calls it "Previous tx"
1674 2012-01-25 14:35:00 <gmaxwell> so long as its there, and its required to be there.. you can't get dupes (except via astronomical odds)
1675 2012-01-25 14:35:02 <UukGoblin> yeah, doesn't matter that much
1676 2012-01-25 14:36:25 wasabi2 has joined
1677 2012-01-25 14:37:22 iocor has joined
1678 2012-01-25 14:40:01 <gmaxwell> UukGoblin: not that it'll ever matter but I added your idea to https://en.bitcoin.it/wiki/Hardfork_Wishlist in the bugfixes section.
1679 2012-01-25 14:41:08 <UukGoblin> cool thanks ;-)
1680 2012-01-25 14:44:24 <UukGoblin> hm, "Miners/relays should not be able to inject extra arbitrary data into transactions?" <- I recall that they are, but it's somehow limited?
1681 2012-01-25 14:45:12 etotheipi_ has quit (Ping timeout: 240 seconds)
1682 2012-01-25 14:45:47 <gmaxwell> we have some protection now by a new behavior that requires txn to not have excess data stuffed.. but it's not a network rule, and it's stupid that its possible in the first place.
1683 2012-01-25 14:45:51 <UukGoblin> again, boils down to me not knowing what the simplified_version_of() is (quoted from https://en.bitcoin.it/wiki/Transactions#Input)
1684 2012-01-25 14:46:15 JRWR has quit (Remote host closed the connection)
1685 2012-01-25 14:48:20 JRWR has joined
1686 2012-01-25 14:50:22 <UukGoblin> is SHA "post-quantum" safe? :-)
1687 2012-01-25 14:50:29 <UukGoblin> I guess not
1688 2012-01-25 14:50:57 datagutt_ is now known as datagutt
1689 2012-01-25 14:52:31 vhsjon has joined
1690 2012-01-25 14:53:34 iocor has quit (Quit: Computer has gone to sleep.)
1691 2012-01-25 14:54:31 <UukGoblin> oh it is :-O
1692 2012-01-25 14:55:00 inlikeflynn has quit ()
1693 2012-01-25 14:55:37 <diki> So I did some math, though I suck at it
1694 2012-01-25 14:55:52 <diki> and tried to figure out the max possible bitcoin addresses
1695 2012-01-25 14:56:06 <diki> so 34^58?
1696 2012-01-25 14:57:46 <UukGoblin> diki, what's the limit on length of scriptPubKey?
1697 2012-01-25 14:58:13 <diki> I am not familiar with the bitcoin source code
1698 2012-01-25 14:58:23 <UukGoblin> diki, cause addresses can be longer than the usual
1699 2012-01-25 14:58:40 <diki> well when I say address, I mean a bitcoin base58 encoded address
1700 2012-01-25 14:59:21 <UukGoblin> diki, BIP 11 addresses are longer than your typical address
1701 2012-01-25 15:00:28 <diki> sample?
1702 2012-01-25 15:00:37 danbri has quit (Ping timeout: 240 seconds)
1703 2012-01-25 15:01:02 <UukGoblin> but yeah, you can calculate max number of typical addresses that way... however at the bottom of https://en.bitcoin.it/wiki/Address you see they can be 33 character-long too
1704 2012-01-25 15:01:27 <gmaxwell> UukGoblin: SHA256 obviously post-quantum unsafe.
1705 2012-01-25 15:01:35 <gmaxwell> er damnit
1706 2012-01-25 15:01:39 <gmaxwell> ISN'T
1707 2012-01-25 15:01:55 <UukGoblin> it wasn't that obvious to me...
1708 2012-01-25 15:02:15 <UukGoblin> but I don't get exactly how quantum physics works. It's too "magical" for me.
1709 2012-01-25 15:02:30 iocor has joined
1710 2012-01-25 15:04:20 <gmaxwell> UukGoblin: It's just 'strange'. In any case, the best speedup a quantum computer can give you against the inversion of generic non-linear functions is SQRT(N) (though the bound it tight, a sufficiently large QC also gives at least that). So a 256 bit hash gets the security against collision the same as a 128 bit hash, which is still okay.
1711 2012-01-25 15:05:09 <gmaxwell> UukGoblin: some special things QC could do even faster, but there isn't any reason to think any of the hash functions anyone uses fall into that class.
1712 2012-01-25 15:05:24 <gmaxwell> (but ECC and RSA do)
1713 2012-01-25 15:06:06 danbri has joined
1714 2012-01-25 15:06:07 MrTiggr has quit (Read error: Connection reset by peer)
1715 2012-01-25 15:06:18 <lianj> http://michaelnielsen.org/blog/quantum-computing-for-the-determined/
1716 2012-01-25 15:06:24 danbri has quit (Remote host closed the connection)
1717 2012-01-25 15:06:30 MrTiggr has joined
1718 2012-01-25 15:08:41 Nicksasa has joined
1719 2012-01-25 15:10:33 Tiggr has joined
1720 2012-01-25 15:10:40 <UukGoblin> thanks
1721 2012-01-25 15:10:59 Tiggr is now known as Guest15264
1722 2012-01-25 15:12:23 MrTiggr has quit (Ping timeout: 276 seconds)
1723 2012-01-25 15:18:57 TD has joined
1724 2012-01-25 15:19:21 copumpkin has quit (Quit: Computer has gone to sleep.)
1725 2012-01-25 15:22:15 danbri has joined
1726 2012-01-25 15:24:07 Guest15264 is now known as MrTiggr
1727 2012-01-25 15:29:29 <luke-jr> Joric: git://gitorious.org/+bitcoin-stable-developers/bitcoin/bitcoind-stable.git
1728 2012-01-25 15:30:02 <gmaxwell> vitevite~.
1729 2012-01-25 15:30:54 Tuxavant_ has joined
1730 2012-01-25 15:32:48 rcorreia_ has joined
1731 2012-01-25 15:32:51 dikidera has joined
1732 2012-01-25 15:32:58 james has joined
1733 2012-01-25 15:33:01 james has quit (Remote host closed the connection)
1734 2012-01-25 15:33:34 slush has quit (Ping timeout: 252 seconds)
1735 2012-01-25 15:36:32 JRWR has quit (Quit: "If we desire respect for the law, we must first make the law respectable." ~Louis D. Brandeis)
1736 2012-01-25 15:37:06 <luke-jr> gavinandresen: was my objection to pull #773 wrong or inaccurate in some way?
1737 2012-01-25 15:37:52 Tycale_ has joined
1738 2012-01-25 15:38:41 pasky_ has joined
1739 2012-01-25 15:41:51 nathan7 has joined
1740 2012-01-25 15:41:51 iz has joined
1741 2012-01-25 15:44:15 upb has joined
1742 2012-01-25 15:44:15 bagit has joined
1743 2012-01-25 15:44:15 JZavala has joined
1744 2012-01-25 15:44:15 Internet13 has joined
1745 2012-01-25 15:44:15 JStoker has joined
1746 2012-01-25 15:44:15 RAWRwins254 has joined
1747 2012-01-25 15:44:31 p0s has joined
1748 2012-01-25 15:44:56 riush has quit (Ping timeout: 252 seconds)
1749 2012-01-25 15:45:08 <luke-jr> no really mquin?
1750 2012-01-25 15:45:11 copumpkin has joined
1751 2012-01-25 15:45:45 iocor has quit (Quit: Computer has gone to sleep.)
1752 2012-01-25 15:47:02 schilly has joined
1753 2012-01-25 15:47:02 bushing has joined
1754 2012-01-25 15:47:02 Bwild has joined
1755 2012-01-25 15:47:16 copumpkin has quit (Client Quit)
1756 2012-01-25 15:47:34 copumpkin has joined
1757 2012-01-25 15:48:37 ThomasV has joined
1758 2012-01-25 15:48:37 BCBot has joined
1759 2012-01-25 15:49:10 Graet has joined
1760 2012-01-25 15:49:19 Internet13 has quit (Read error: Connection reset by peer)
1761 2012-01-25 15:49:40 Sedra has joined
1762 2012-01-25 15:49:57 Internet13 has joined
1763 2012-01-25 15:50:49 twobitcoins has joined
1764 2012-01-25 15:54:15 Cherothald has joined
1765 2012-01-25 15:54:55 gwillen has joined
1766 2012-01-25 15:56:28 tomat has quit (Ping timeout: 255 seconds)
1767 2012-01-25 15:57:22 forests has joined
1768 2012-01-25 15:58:36 topace_ has joined
1769 2012-01-25 15:59:52 marf_away has joined
1770 2012-01-25 15:59:52 sacarlson has joined
1771 2012-01-25 15:59:52 onelineproof has joined
1772 2012-01-25 15:59:52 mpr has joined
1773 2012-01-25 15:59:52 coderrr has joined
1774 2012-01-25 15:59:52 Cryo has joined
1775 2012-01-25 16:00:23 iocor has joined
1776 2012-01-25 16:00:23 therealnanotube has joined
1777 2012-01-25 16:01:15 maqr has joined
1778 2012-01-25 16:01:15 wizkid057_ has joined
1779 2012-01-25 16:01:15 ski_ has joined
1780 2012-01-25 16:01:15 booo has joined
1781 2012-01-25 16:01:15 coblee has joined
1782 2012-01-25 16:01:15 TheSeven has joined
1783 2012-01-25 16:01:15 Nesetalis has joined
1784 2012-01-25 16:01:15 ahbritto has joined
1785 2012-01-25 16:01:15 _W_ has joined
1786 2012-01-25 16:01:15 osmosis has joined
1787 2012-01-25 16:01:15 bobke has joined
1788 2012-01-25 16:01:15 MBS has joined
1789 2012-01-25 16:01:15 spaola has joined
1790 2012-01-25 16:01:15 Hunner has joined
1791 2012-01-25 16:01:15 welterde has joined
1792 2012-01-25 16:01:15 jercos has joined
1793 2012-01-25 16:01:15 phantomcircuit has joined
1794 2012-01-25 16:01:15 cjdelisle has joined
1795 2012-01-25 16:01:15 nullrouten has joined
1796 2012-01-25 16:01:15 doublec has joined
1797 2012-01-25 16:01:15 toggles_ has joined
1798 2012-01-25 16:01:15 nameless has joined
1799 2012-01-25 16:01:34 MrTiggr has quit (Ping timeout: 248 seconds)
1800 2012-01-25 16:01:46 maqr has quit (Max SendQ exceeded)
1801 2012-01-25 16:02:44 BCBot has quit (Read error: Operation timed out)
1802 2012-01-25 16:04:40 SuprTiggr has joined
1803 2012-01-25 16:05:01 cdecker has quit (Ping timeout: 255 seconds)
1804 2012-01-25 16:06:01 maqr has joined
1805 2012-01-25 16:06:29 cdecker has joined
1806 2012-01-25 16:06:42 vhsjon has joined
1807 2012-01-25 16:08:30 <roconnor> gmaxwell: did we get an answer to my question?
1808 2012-01-25 16:09:45 bob12323 has quit (Quit: Lost terminal)
1809 2012-01-25 16:10:27 <luke-jr> or mine?
1810 2012-01-25 16:11:42 <roconnor> luke-jr: what's your question? :D
1811 2012-01-25 16:11:50 <luke-jr> [10:31:37] <luke-jr> gavinandresen: was my objection to pull #773 wrong or inaccurate in some way?
1812 2012-01-25 16:11:54 <gmaxwell> roconnor: about coinbases? kinda. We already have duplicate ones. Only one can be spent. If you were to reintroduce one, you'd still only have one that could be spent.
1813 2012-01-25 16:12:09 <gmaxwell> at least thats what gavin said. I didn't step through the logic myself.
1814 2012-01-25 16:12:23 therealnanotube has quit (Changing host)
1815 2012-01-25 16:12:23 therealnanotube has joined
1816 2012-01-25 16:12:23 iocor has quit (Changing host)
1817 2012-01-25 16:12:23 iocor has joined
1818 2012-01-25 16:12:23 <luke-jr> it's starting to seem to me, the primary factor to whether something is merged, is who wrote it :/
1819 2012-01-25 16:12:23 JStoker has quit (Changing host)
1820 2012-01-25 16:12:23 JStoker has joined
1821 2012-01-25 16:12:24 copumpkin has quit (Changing host)
1822 2012-01-25 16:12:24 copumpkin has joined
1823 2012-01-25 16:12:41 <gmaxwell> UukGoblin pointed out this has negative implications for pruning.
1824 2012-01-25 16:12:41 BCBot has joined
1825 2012-01-25 16:12:47 Graet has quit (Changing host)
1826 2012-01-25 16:12:47 Graet has joined
1827 2012-01-25 16:12:58 <luke-jr> #773 appears to break the ability to accept non-standard transactions, yet Gavin just merged it anyway
1828 2012-01-25 16:13:16 <luke-jr> ignoring my objection
1829 2012-01-25 16:13:32 <roconnor> gmaxwell: heh I didn't quite understand what you mean
1830 2012-01-25 16:13:50 <roconnor> mine, spend, mine again, can spend or not?
1831 2012-01-25 16:14:15 forests has quit (Quit: forests)
1832 2012-01-25 16:14:33 <gmaxwell> roconnor: you can mine it all you want. that ID can only ever be spent once.
1833 2012-01-25 16:14:34 forests has joined
1834 2012-01-25 16:14:42 forests has quit (Remote host closed the connection)
1835 2012-01-25 16:14:55 <roconnor> gmaxwell: even if you mine it after it as been "freed"?
1836 2012-01-25 16:14:57 <gmaxwell> luke-jr: I don't see where you're coming from there.
1837 2012-01-25 16:15:28 <gmaxwell> roconnor: thats what gavin said. The input has been used. Once it's used it can't be used again. (which makes sense to me)
1838 2012-01-25 16:15:29 <luke-jr> gmaxwell: I want to accept non-std rtxns
1839 2012-01-25 16:15:38 <gmaxwell> luke-jr: yep. Great. Do so.
1840 2012-01-25 16:15:44 <roconnor> gmaxwell: it's fair ... though I have a bug in my code.
1841 2012-01-25 16:15:46 <luke-jr> gmaxwell: pull #773 breaks that
1842 2012-01-25 16:15:47 <roconnor> apparently
1843 2012-01-25 16:15:49 <roconnor> :D
1844 2012-01-25 16:15:50 <gmaxwell> luke-jr: no it doesn't.
1845 2012-01-25 16:15:54 <luke-jr> https://github.com/bitcoin/bitcoin/pull/773/files#L0R506
1846 2012-01-25 16:16:32 weather has joined
1847 2012-01-25 16:16:35 <gmaxwell> luke-jr: Yes. And? There isn't a protcol rule being enforced.
1848 2012-01-25 16:16:56 <luke-jr> gmaxwell: there isn't a clear code path to accept non-standard transactions anymore.
1849 2012-01-25 16:17:23 <gmaxwell> luke-jr: It's just saying that if you're accepting non-standard stuff it would be prudent to still have some sanity checking (I know your non-standard acceptance does)
1850 2012-01-25 16:17:38 <gmaxwell> luke-jr: HUH?
1851 2012-01-25 16:17:55 <luke-jr> gmaxwell: it's removing the existing sanity checking, and forcing anyone who wants to support non-std txns implement their own
1852 2012-01-25 16:18:25 <gmaxwell> luke-jr: they're not forced. That was non-standard checking leakage.
1853 2012-01-25 16:19:11 <gmaxwell> luke-jr: if all the non-standard checking code were duplicated there, would you also complain the same thing about its removal?
1854 2012-01-25 16:19:14 helo has joined
1855 2012-01-25 16:19:35 <gmaxwell> I'd agree with you if that was enforcing some protocol rule, but it's not.
1856 2012-01-25 16:19:45 booo has quit (Ping timeout: 255 seconds)
1857 2012-01-25 16:19:50 upb has quit (Changing host)
1858 2012-01-25 16:19:50 upb has joined
1859 2012-01-25 16:20:20 <gmaxwell> I'd also agree that someone should publish some good sanity checking for non-standard txn which would include something like that (and a lot more).
1860 2012-01-25 16:20:21 weather is now known as BeTep
1861 2012-01-25 16:20:25 <luke-jr> gmaxwell: OK, so the answer to my question is "No, the code you're supposed to add here is in fact code you bypass if you truly want to accept non-standard transactions"?
1862 2012-01-25 16:21:00 <gmaxwell> luke-jr: Yes. If you truly want to accept non-standard txn with no protection against crazy rubbish, spamming, etc.
1863 2012-01-25 16:21:21 <luke-jr> would have been nice if Gavin had clarified that
1864 2012-01-25 16:21:22 <luke-jr> oh well
1865 2012-01-25 16:21:43 <gmaxwell> (perhaps you might want to protect against those things instead with changed fee rules)
1866 2012-01-25 16:22:12 _Fireball has joined
1867 2012-01-25 16:22:51 <gmaxwell> luke-jr: to be fair, it says should not must. :)
1868 2012-01-25 16:23:29 <gmaxwell> luke-jr: but it should instead say something like "if you do this, you really should add basic sanity checking or improved fee rules in its place"
1869 2012-01-25 16:24:14 <gmaxwell> Cause I think, for example that you might accept some single TXN that uses up most of your checksigs in the block. You won't produce an invalid block, but a block with one or two obnoxious txn is probably not what you wanted.
1870 2012-01-25 16:24:15 <luke-jr> ok, I think I understand it now.
1871 2012-01-25 16:24:33 <nameless> !~root@mindjail.subluminal.net|SuprTiggr: Do me a favor and exclude me from your "Don't IRC as root" script please
1872 2012-01-25 16:24:43 <nameless> !~root@mindjail.subluminal.net|I'll IRC as whoever I damn well please
1873 2012-01-25 16:24:43 <UukGoblin> I'm guessing value (amount) is part of the txid hash?
1874 2012-01-25 16:24:45 <gmaxwell> nameless|: don't IRC as root.
1875 2012-01-25 16:24:51 <roconnor> UukGoblin: yes
1876 2012-01-25 16:24:53 <luke-jr> hmm
1877 2012-01-25 16:24:57 * nameless !~root@mindjail.subluminal.net|beats gmaxwell with root
1878 2012-01-25 16:25:00 <roconnor> UukGoblin: the output amounts
1879 2012-01-25 16:25:02 <luke-jr> gmaxwell: I wonder if that could be the DoS attack Eligius is experiencingâ¦
1880 2012-01-25 16:25:07 <roconnor> UukGoblin: the input amounts only indirectly.
1881 2012-01-25 16:25:17 <luke-jr> gmaxwell: will bitcoind block RPC while it verifies txns? :\
1882 2012-01-25 16:25:29 <gmaxwell> luke-jr: heck yea. it blocks for just about everything.
1883 2012-01-25 16:25:38 <UukGoblin> thanks roconnor
1884 2012-01-25 16:25:38 <nameless> !~root@mindjail.subluminal.net|SuprTiggr: See above message for SuprTiggr
1885 2012-01-25 16:25:56 <nameless> !~root@mindjail.subluminal.net|err
1886 2012-01-25 16:26:07 * nameless !~root@mindjail.subluminal.net|palms face
1887 2012-01-25 16:26:09 <UukGoblin> actually I'd love to know how the transaction is actually hashed, as well as what simplified version of it is signed in scriptSig
1888 2012-01-25 16:26:30 BTC_Bear has joined
1889 2012-01-25 16:26:38 <gmaxwell> luke-jr: you should get oprofile setup. It's very useful for troubleshooting this kind of stuff.
1890 2012-01-25 16:26:43 <luke-jr> hmm
1891 2012-01-25 16:26:53 <gmaxwell> (it's a little it of a pain, but its worth it)
1892 2012-01-25 16:26:54 <luke-jr> gmaxwell: I don't think CPU usage is a problem tho
1893 2012-01-25 16:26:57 <roconnor> UukGoblin: transactions are serizalied according to the wiki page format and that is hashed
1894 2012-01-25 16:27:09 <nameless> !~root@mindjail.subluminal.net|gmaxwell: Based on my client version, do you really think I'm IRCing as root?
1895 2012-01-25 16:27:17 <gmaxwell> luke-jr: could be some other dos attack... though most would cause interesting logs.
1896 2012-01-25 16:27:32 <gmaxwell> nameless|: I haven't looked!
1897 2012-01-25 16:28:08 <roconnor> UukGoblin: https://en.bitcoin.it/wiki/Protocol_specification#tx
1898 2012-01-25 16:28:16 <gmaxwell> nameless|: though it's stupid to indicate thatâ it just gives you an excuse to whine at people who are doing helpful things for the vast majority of people out there who are actually idiots.
1899 2012-01-25 16:28:48 <roconnor> UukGoblin: Hmm, actually I can be more specific than than if I cut and paste my code
1900 2012-01-25 16:29:08 <nameless> !~root@mindjail.subluminal.net|gmaxwell: If I want to IRC as motherfuckin root, I want to do it in peace!
1901 2012-01-25 16:29:31 <nameless> !~root@mindjail.subluminal.net|gmaxwell: And, while yes, I appreciate helping n00bs out, I've been kickbanned from channels by bots saying "don't IRC as root"
1902 2012-01-25 16:29:37 <roconnor> hmm, my code is actually splayed out over a few modules
1903 2012-01-25 16:29:41 <nameless> !~root@mindjail.subluminal.net|So excuse me for going on a vendeta
1904 2012-01-25 16:29:56 <roconnor> UukGoblin: how about I answer any questions you have :D
1905 2012-01-25 16:30:19 <UukGoblin> roconnor, thanks... reading that protocol specification
1906 2012-01-25 16:31:24 <roconnor> UukGoblin: there is a serialized tx packet example below
1907 2012-01-25 16:31:43 <roconnor> UukGoblin: but the "Message header" is not part of the transaction and isn't used in the hash.
1908 2012-01-25 16:32:29 <UukGoblin> roconnor, so the hash as used in prev_tx in inputs is the hash of the entire transaction, serialized as for the protocol
1909 2012-01-25 16:32:58 <roconnor> yes
1910 2012-01-25 16:33:18 <BTC_Bear> The rest of nameless|'s host mask, kind of takes care of the ~root part. :)
1911 2012-01-25 16:33:18 <UukGoblin> what's in the signature in scriptSig? it can't be signing the hash, right? Is it signing the same serialized version that's being hashed?
1912 2012-01-25 16:33:32 <UukGoblin> or some subset of it?
1913 2012-01-25 16:34:00 sje has quit (Remote host closed the connection)
1914 2012-01-25 16:34:03 TheDealer is now known as TheFakeNanotube
1915 2012-01-25 16:34:15 <roconnor> It basically signs the transaction with all the scriptSigs removed. There are some variations possible using different hashkinds.
1916 2012-01-25 16:35:18 <UukGoblin> ah... so that's why someone can fuck around with scriptSigs?
1917 2012-01-25 16:35:27 <UukGoblin> as in, append data to them or sth?
1918 2012-01-25 16:35:28 <roconnor> yes
1919 2012-01-25 16:35:52 <roconnor> or, more positively, miners and relayers could filter crap out
1920 2012-01-25 16:36:43 <UukGoblin> weird.
1921 2012-01-25 16:36:51 <UukGoblin> but I guess only satoshi could answer, "why?"
1922 2012-01-25 16:38:42 <roconnor> Honestly, bitcoin isn't all that well designed.
1923 2012-01-25 16:39:12 <roconnor> Not that any one person could have gotten it right by him/herself.
1924 2012-01-25 16:39:31 gavinandresen has joined
1925 2012-01-25 16:39:46 <roconnor> Satoshi seems to have been a bit too enamoured by concatinative langauges.
1926 2012-01-25 16:39:48 <Joric> satoshi isn't one person haven't you see the movie
1927 2012-01-25 16:39:58 <nameless> !~root@mindjail.subluminal.net|BTC_Bear: lol
1928 2012-01-25 16:40:01 SomeoneWeird is now known as SomeoneWeirdzzzz
1929 2012-01-25 16:41:18 Joric has quit ()
1930 2012-01-25 16:42:17 <UukGoblin> "concatinative languages"? :-)
1931 2012-01-25 16:42:26 TheFakeNanotube is now known as TheDealer
1932 2012-01-25 16:43:11 <UukGoblin> ah, got a wikipage for it
1933 2012-01-25 16:44:12 <devrandom> ;;later tell BlueMatt perhaps mark it as optional for download purposes? otherwise it should already work, since each file is hashed separately
1934 2012-01-25 16:44:12 <gribble> The operation succeeded.
1935 2012-01-25 16:45:55 <gmaxwell> roconnor: Be nice, as I pointed out before the list of things that could easily have been gotten wrong but were done right is very, very long. And most of the unfortunate stuff is in functionality that could have easily been missed. E.g. the malleability is unfortunateâ but it's still better than a system with no script at all.
1936 2012-01-25 16:46:04 ThomasV has quit (Quit: Leaving)
1937 2012-01-25 16:46:20 ThomasV has joined
1938 2012-01-25 16:46:22 <roconnor> gmaxwell: sorry, you are absolutely right.
1939 2012-01-25 16:46:45 etotheipi_ has joined
1940 2012-01-25 16:47:48 Cablesaurus has joined
1941 2012-01-25 16:47:48 Cablesaurus has quit (Changing host)
1942 2012-01-25 16:47:48 Cablesaurus has joined
1943 2012-01-25 16:47:57 <roconnor> I'd like to double check the code about this prevesion of spending duplicate coinbases.
1944 2012-01-25 16:48:04 <gmaxwell> roconnor: heh. well it's not my pride you're woundingâ I'm pointing it out because someone with less expirence with the system might take your very narrow statement as something more broad. :)
1945 2012-01-25 16:48:06 <roconnor> it is part of the client that I've never looked at myself.
1946 2012-01-25 16:48:20 <gavinandresen> gmaxwell: I didn't answer your question before as to whether or not I wrote the BIP-17-coin-stealing bot... I have to confess (confession is good for the soul, right?) I did.
1947 2012-01-25 16:48:32 <roconnor> hah
1948 2012-01-25 16:48:42 <gmaxwell> hahah
1949 2012-01-25 16:48:46 <gmaxwell> gavinandresen++
1950 2012-01-25 16:49:09 <gavinandresen> I'm a little surprised Luke hasn't warned people that testing BIP 17 on testnet won't work.
1951 2012-01-25 16:49:13 <gmaxwell> Though I'm kinda sad that there aren't more people who are comfortable enough with the internals to have done that. :(
1952 2012-01-25 16:49:38 <roconnor> Now I feel more motived to finish my BIP stealing relayer.
1953 2012-01-25 16:49:45 <roconnor> *BIP16
1954 2012-01-25 16:49:51 <roconnor> *motivated
1955 2012-01-25 16:50:08 <Eliel> :D
1956 2012-01-25 16:50:17 <BTC_Bear> gavinandresen: there is a #bitcoin-church that would hear your confessions. :))
1957 2012-01-25 16:50:20 ovidiusoft has quit (Ping timeout: 252 seconds)
1958 2012-01-25 16:50:33 <etotheipi_> whoa, I came in at an interesting time... who stole what? is BIP0017 screwed?
1959 2012-01-25 16:50:46 <gavinandresen> roconnor: stealing a BIP 16 redemption means you get lucky (or can 50%) and solve a block before the rest of the mining nodes
1960 2012-01-25 16:50:49 <Eliel> etotheipi_: only works when there's a majority of old nodes.
1961 2012-01-25 16:51:08 <etotheipi_> ahh
1962 2012-01-25 16:51:13 <gavinandresen> roconnor: ... because BIP 16 redemptions look like IsStandard() transactions to old clients (assuming the scriptSig is < 200 bytes)
1963 2012-01-25 16:51:33 <etotheipi_> so the fact that you would only need like 10% new nodes to propagate doesn't matter... you need 50% for security
1964 2012-01-25 16:51:34 <roconnor> gavinandresen: not it just means that whoever I relay to solves a block.
1965 2012-01-25 16:51:36 <UukGoblin> gavinandresen, or that you can spread your version of the transaction faster than the originator
1966 2012-01-25 16:51:52 <gavinandresen> roconnor: but the people you're relaying to have probably already got the original transaction
1967 2012-01-25 16:52:00 <UukGoblin> etotheipi_, both BIPs won't work until majority of hashpower supports it ;-]
1968 2012-01-25 16:52:12 <Eliel> etotheipi_: I think coding the new version so that there's a subnetwork with full connectivity composed entirely of new nodes will help.
1969 2012-01-25 16:52:16 <gavinandresen> roconnor: ... because transaction percolation through the network is a race, and the original redeemer will always have a head start
1970 2012-01-25 16:52:18 <etotheipi_> is it 50% of hashpower, or 50% of all nodes?
1971 2012-01-25 16:52:30 <UukGoblin> hashpower
1972 2012-01-25 16:52:35 <roconnor> gavinandresen: What fraction of the peers will I get on average?
1973 2012-01-25 16:53:02 <gavinandresen> roconnor: depends on how directly connected the original redeemer is to you, and whether they send the transaction to you before telling their other peers, etc.....
1974 2012-01-25 16:53:18 <etotheipi_> roconnor, I know you're busy... but I just committed a 99% fix to the qtdev branch of Armory... it has the ripemd160 update, and zero-conf/mem-pool seems to work well
1975 2012-01-25 16:53:34 <roconnor> gavinandresen: assume the network is a random 8-incident graph
1976 2012-01-25 16:53:44 <roconnor> etotheipi_: thanks for the tip
1977 2012-01-25 16:53:44 <etotheipi_> roconnor, I think it's going to be ready soon, and you're definitely a person I'd love to help test :)
1978 2012-01-25 16:53:53 <etotheipi_> err.. ready for alpha soon
1979 2012-01-25 16:54:10 <gavinandresen> roconnor: I dunno, my networking-fu is weak.
1980 2012-01-25 16:54:24 <Eliel> gavinandresen: do you think it would be difficult to modify the networking of the new version so that it makes a subnetwork of new nodes to help with propagation?
1981 2012-01-25 16:54:53 <gavinandresen> Eliel: I think that would be more trouble than it is worth.
1982 2012-01-25 16:55:10 <etotheipi_> eliel, it seems that if everyone is connected to 20 or more peers... you'd only need a very small percentage of upgraded nodes in order for propagation to get to miners
1983 2012-01-25 16:55:25 <roconnor> I like how your bot was so BIP 17 specific, it left my transaction http://blockexplorer.com/testnet/tx/6c010586ba139c5ef64a0ab8352f16125497cc742fe10f41535b0bc99124beed#o0 free :D
1984 2012-01-25 16:55:45 <UukGoblin> heh
1985 2012-01-25 16:55:47 <UukGoblin> nasty ;-]
1986 2012-01-25 16:56:00 <UukGoblin> BIP-bot war ;-]
1987 2012-01-25 16:56:18 <gavinandresen> roconnor: yes, I was tempted to create a generic "if the scriptPubKey evaluates to true with any input then steal" but it was easier to just target CHV transactions
1988 2012-01-25 16:56:48 cdecker has quit (Ping timeout: 252 seconds)
1989 2012-01-25 16:56:49 <Eliel> gavinandresen: even just code that would try to keep a minimum of 2 of the connections to new nodes would probably help a lot I think.
1990 2012-01-25 16:57:05 BCBot has quit (Ping timeout: 240 seconds)
1991 2012-01-25 16:58:12 ciscoftw has joined
1992 2012-01-25 16:58:14 <gavinandresen> Eliel: good idea. Probably should do that when (if?) p2sh features start to appear in clients
1993 2012-01-25 16:58:30 iocor has quit (Quit: Computer has gone to sleep.)
1994 2012-01-25 16:59:16 <roconnor> gavinandresen: oh right, I wanted to tell you that if you ignore all of luke's anti-BIP16 rhetoric, and look at the BIP 17 proposal by itself, it is actually pretty good proposal. It is less conservative than BIP 16, but this is balanced by being much simpler to implement IMHO. I think you would do well by giving it serious consideration.
1995 2012-01-25 16:59:41 <gavinandresen> Eliel: I actually think somebody aught to take a good stiff brush to the networking code, and implement phantomcircuit's idea of rotating peers (to help prevent Sybil attacks), be smarter about what nodes you keep, what IP ranges you're connected to, what peer versions, ....
1996 2012-01-25 17:00:05 <phantomcircuit> gavinandresen, hi
1997 2012-01-25 17:00:23 <gavinandresen> roconnor: I thought I did give it serious consideration in this post: https://bitcointalk.org/index.php?topic=60433.msg703923#msg703923
1998 2012-01-25 17:01:02 <gavinandresen> roconnor: I tried to set aside my personal biases and explain why I prefer BIP 16.
1999 2012-01-25 17:01:32 <roconnor> gavinandresen: oh good. As long as you have given it serious consideration, then I'm happy.
2000 2012-01-25 17:02:34 <roconnor> gavinandresen: you were around during the OP_RETURN incident?
2001 2012-01-25 17:02:42 <gavinandresen> roconnor: yes
2002 2012-01-25 17:03:00 TD has quit (Quit: TD)
2003 2012-01-25 17:03:06 BCBot has joined
2004 2012-01-25 17:03:53 <dikidera> gavinandresen:I was looking at the index page of bitcoin.org...and somehow I disagree with the whole "instant payments" since a transactions needs to have 6 confirms before it is spendable
2005 2012-01-25 17:03:57 <roconnor> gavinandresen: do you know why the script executions were separated and interfraced by the main stack?
2006 2012-01-25 17:04:13 Clipse has joined
2007 2012-01-25 17:04:14 <Eliel> phantomcircuit: Would you consider writing a BIP of your idea of how to improve the networking code?
2008 2012-01-25 17:04:31 <roconnor> gavinandresen: I mean, turning OP_RETURN into something that just fails was a good move, but the code changes made then were quite substatial.
2009 2012-01-25 17:04:38 <phantomcircuit> Eliel, i would have to take a bit more of a look but yeah
2010 2012-01-25 17:04:43 <roconnor> gavinandresen: I called it heavy-handed.
2011 2012-01-25 17:04:45 <dikidera> whereas paypal for instance, if you send money from one account to another, now that is instant
2012 2012-01-25 17:04:52 <gavinandresen> roconnor: no, Satoshi never took the time to explain things like that in detail.
2013 2012-01-25 17:04:56 <roconnor> though I respect the urgency of the situation.
2014 2012-01-25 17:05:18 <sipa> many changes in the code look to me like they were never intended to stay there for enternity
2015 2012-01-25 17:05:29 <sipa> but we're kinda stuck with them now
2016 2012-01-25 17:05:32 <Eliel> phantomcircuit: great :)
2017 2012-01-25 17:05:34 <gavinandresen> That was just after the first slashdotting, and I think Satoshi felt like he was under seige-- there were several "oops!" moments right around that time
2018 2012-01-25 17:05:53 <Eliel> dikidera: that's up to the recipient.
2019 2012-01-25 17:06:03 <Eliel> dikidera: it's instant in the sense that it's immediately visible.
2020 2012-01-25 17:06:05 MrTiggr has joined
2021 2012-01-25 17:06:13 <gavinandresen> I was pushing for whitelisting "known good" transactions instead of trying to blacklist "might be bad"... (which I still think is the right approach)
2022 2012-01-25 17:06:34 <roconnor> gavinandresen: BIP16 undoes some of the "heavy-handed" changes of the OP_RETURN fix, it restores the situation where the signature is separted by be verifyer by an OP_CODESEPARATOR. If anything it restores the original design, not subverts it.
2023 2012-01-25 17:06:39 <phantomcircuit> dikidera, it's also immediately spendable so long as the recipient is willing to accept it after 0 or 1 confirmas as well
2024 2012-01-25 17:06:46 <sipa> roconnor: you mean BIP17
2025 2012-01-25 17:06:50 <roconnor> gavinandresen: well restores the
2026 2012-01-25 17:06:53 <gavinandresen> roconnor: that's why it makes me nervous
2027 2012-01-25 17:06:54 <roconnor> yes BIP17
2028 2012-01-25 17:07:26 <roconnor> Which is perfectly fair
2029 2012-01-25 17:07:40 <roconnor> and it should make you nervous
2030 2012-01-25 17:07:43 <gavinandresen> roconnor: weren't you the one arguing that we don't understand Script well enough to be reversing decisions like that?
2031 2012-01-25 17:07:52 <roconnor> yes
2032 2012-01-25 17:08:05 <roconnor> like I said BIP 17 is less conservative than BIP 16
2033 2012-01-25 17:08:28 <roconnor> I'm just responding to your "just makes me feel uncomfortable" remarks
2034 2012-01-25 17:08:33 <roconnor> which, I sort of agree with
2035 2012-01-25 17:08:35 TD has joined
2036 2012-01-25 17:08:40 <UukGoblin> was OP_RETURN the 0.3.9 overflow bug?
2037 2012-01-25 17:08:43 <gavinandresen> roconnor: ok
2038 2012-01-25 17:08:53 <gavinandresen> UukGoblin: no, overflow was different bug
2039 2012-01-25 17:09:12 <roconnor> but you seem to be painting the change as BIP 17 chaning the way things were intendened, which in fact they restore thing as they originally originaly were to some extent.
2040 2012-01-25 17:09:28 <gavinandresen> roconnor: ok, fair criticism.
2041 2012-01-25 17:09:55 <gavinandresen> roconnor: ... but Arguments from Antiquity is a well-known logical fallacy
2042 2012-01-25 17:09:56 <roconnor> but you are right, the things as they originally orignally were was not neccesarilly so great.
2043 2012-01-25 17:10:05 <roconnor> yep
2044 2012-01-25 17:12:12 <UukGoblin> for what's that worth, a feature I'd like to see is handling transaction fees properly :->
2045 2012-01-25 17:12:31 <UukGoblin> and double-spends to increase a tx fee
2046 2012-01-25 17:12:47 <UukGoblin> and for miners to be able to communicate to users in a standardised way what fees they expect
2047 2012-01-25 17:13:52 <UukGoblin> but I'm just a selfish miner ;-]
2048 2012-01-25 17:14:06 iocor has joined
2049 2012-01-25 17:15:26 <phantomcircuit> UukGoblin, i actually had a plan for the communicating fees part
2050 2012-01-25 17:15:34 <phantomcircuit> i guess i have 2 BIPs to write
2051 2012-01-25 17:15:45 <gavinandresen> phantomcircuit: how do you prevent miners from lying about their fee policies?
2052 2012-01-25 17:16:08 <gavinandresen> (they have a strong incentive to claim that they'll only accept high fees, when they'll really accept lower...)
2053 2012-01-25 17:16:10 <phantomcircuit> you guess based on a combination of fee policy and actual fees in the previous block
2054 2012-01-25 17:16:23 <UukGoblin> yeah
2055 2012-01-25 17:16:40 <phantomcircuit> or rather you display both to users and ask them what to do
2056 2012-01-25 17:16:51 <phantomcircuit> that seems the only sensible way to handle it
2057 2012-01-25 17:16:54 BGL has joined
2058 2012-01-25 17:17:07 <UukGoblin> yup, users should have the choice in the end
2059 2012-01-25 17:17:12 <cjdelisle> average fees over previous blocks is an interesting metric since it would theoretically be more stable than bitcoin itself.
2060 2012-01-25 17:17:13 <UukGoblin> phantomcircuit, yup, I was thinking about similar
2061 2012-01-25 17:17:20 <phantomcircuit> the main problem with the current system is that it seems like black magic too... pretty much everybody
2062 2012-01-25 17:17:29 <gavinandresen> phantomcircuit: I was planning on having the client watch transactions and infer the network-wide fee policy from what gets into blocks and what doesn't....
2063 2012-01-25 17:17:56 <phantomcircuit> well and it can never be perfect
2064 2012-01-25 17:18:12 <phantomcircuit> a miner can mine a transaction only they see with a huge fee distorting the whole thing
2065 2012-01-25 17:18:25 <phantomcircuit> which is why ultimately you need the user to have control
2066 2012-01-25 17:18:57 <phantomcircuit> proper way to do it is, build optimal transactions, guess fee, ask user to adjust or accept
2067 2012-01-25 17:19:01 <josephcp> an interesting way to do it would be for pools to publish their fees policy in the coinbase in a machine readable fashion
2068 2012-01-25 17:19:13 <phantomcircuit> josephcp, stealin mah ideas :(
2069 2012-01-25 17:19:16 <gavinandresen> phantomcircuit: sure. But you should be able to give the user a pretty good idea of how long their transaction will take to get into a block with no fee or a fee of X...
2070 2012-01-25 17:19:20 <josephcp> oh sorry phantomcircuit :-(
2071 2012-01-25 17:19:41 <phantomcircuit> gavinandresen, well you can try but that's a system i dont think there is a winner for
2072 2012-01-25 17:19:41 <gavinandresen> josephcp: again, why wouldn't a pool just claim that they require big fees?
2073 2012-01-25 17:19:42 <phantomcircuit> :)
2074 2012-01-25 17:19:56 <phantomcircuit> which is why ultimately need to ask the user
2075 2012-01-25 17:20:17 <gavinandresen> How would the user know what a reasonable fee is?
2076 2012-01-25 17:20:24 <phantomcircuit> they dont
2077 2012-01-25 17:20:27 <phantomcircuit> it's a guess
2078 2012-01-25 17:20:31 <josephcp> i assume some kind of UI that gives an estimate for processing time?
2079 2012-01-25 17:20:32 <phantomcircuit> but at least it's their guess :)
2080 2012-01-25 17:20:40 <gavinandresen> Good UI design says don't ask users for information that they don't know.
2081 2012-01-25 17:20:46 <sipa> a nice slider to select the fee
2082 2012-01-25 17:20:51 <roconnor> miners.com -- 5-star miners at 2-star prices.
2083 2012-01-25 17:20:52 <phantomcircuit> yeah thus you ask
2084 2012-01-25 17:20:58 <phantomcircuit> "accept or adjust?"
2085 2012-01-25 17:21:01 <sipa> with information about how long they can expect the transaction to get confir,ed
2086 2012-01-25 17:21:09 <gavinandresen> sipa: ++
2087 2012-01-25 17:21:14 <phantomcircuit> thus dont ask them something they dont know
2088 2012-01-25 17:21:21 <phantomcircuit> and give them the option to do so if they do
2089 2012-01-25 17:21:28 <sipa> but ultimately, it will be the receiver who cares about this, and should pay the fee, imho
2090 2012-01-25 17:21:32 <gavinandresen> sipa: ... although a slider is probably overkill, I think most users care about only two cases: "Free" and "Fast"
2091 2012-01-25 17:21:41 <roconnor> sometimes miners don't get enough transactions to fill their rooms, that where we come in.
2092 2012-01-25 17:23:09 <josephcp> yeah then you can artificially create two markets (which IMO is a bit controvertial but i think it's a good thing) i.e. pools artifically waiting 12 hours before confirming a free transaction, as the users had sufficient UI warning
2093 2012-01-25 17:23:26 * jgarzik recalls the much-mocked debacle of the MandrakeSoft Linux installer: they had a slider for "how much linux do you want?" left side (ostensibly 0%) was "minimal boot" and right side (100%) was "all packages" ... but the positions in the middle were far too vague to be of any practical use.
2094 2012-01-25 17:23:40 <cjdelisle> haha
2095 2012-01-25 17:24:05 <gavinandresen> I want 82% Linux, please
2096 2012-01-25 17:24:28 <Eliel> :D
2097 2012-01-25 17:24:39 booo has joined
2098 2012-01-25 17:24:53 <josephcp> well i think this siutation is a lot easier assuming as sufficeintly accurate estimate, you're only dealing with one thing, time, with linux it's packages (is apache at 47% or is vim at 18% or tuxracer at 85%)
2099 2012-01-25 17:25:00 <josephcp> one variable i mean
2100 2012-01-25 17:25:07 <gavinandresen> The graph of transaction fee versus time-to-include-in-block is radically non-linear, so slider is almost certainly wrong.
2101 2012-01-25 17:25:42 <gavinandresen> (and it's actually a complicated higher-dimensional graph that includes transaction priority and size....)
2102 2012-01-25 17:25:52 <josephcp> ooh right
2103 2012-01-25 17:26:06 <UukGoblin> well, if "as cheap as possible" is the goal, the client should transmit the transaction with 0 fee, waith ~3 blocks, if it's not included, increase the fee slightly and resubmit
2104 2012-01-25 17:26:13 <josephcp> that does make it way harder
2105 2012-01-25 17:26:26 <sipa> gavinandresen: i was once working on an idea to assign a time-dependent score to each transaction group (based on sigops, size, age of inputs, ...), and modify the memory pool to choose a cutoff point and include all transaction below the cutoff
2106 2012-01-25 17:26:33 <sipa> gavinandresen: then it becomes one-dimensional
2107 2012-01-25 17:26:38 storrgie has joined
2108 2012-01-25 17:26:57 <Eliel> probably best to keep the number of possible choices down to simply things (at least for the novice view, if we want such separately). So, I think the choices should be "Next Block", "2 hours", "one day" and then display fees next to them that ought to get the transaction included that fast.
2109 2012-01-25 17:26:58 <sipa> assuming miners follow the same function (they'd still be free to choose their own cutoff-point-coosing algorithm)
2110 2012-01-25 17:27:07 <gavinandresen> sipa: I was thinking of creating buckets in the multidimensional space and keeping track of average time to confirm for each bucket....
2111 2012-01-25 17:27:28 <sipa> gavinandresen: ambitious :)
2112 2012-01-25 17:27:37 <gavinandresen> (and interpolating if you find the transaction you're sending was empty)
2113 2012-01-25 17:27:53 <gavinandresen> sipa: the ambitious projects are the most fun ones!
2114 2012-01-25 17:28:06 <sipa> that's why we're here :)
2115 2012-01-25 17:28:32 <gavinandresen> bitcoin? ambitious? what do you mean? :)
2116 2012-01-25 17:28:39 <josephcp> haha
2117 2012-01-25 17:28:57 <Eliel> is it possible to have a transaction output that's 0.00 bitcoins (yes, empty). There was an article that people might want to use such outputs to timestamp something so they can later prove they already had it way back.
2118 2012-01-25 17:29:05 dikidera is now known as diki
2119 2012-01-25 17:30:00 <roconnor> Eliel: yes that is legal
2120 2012-01-25 17:30:07 <roconnor> Eliel: you can spend it too
2121 2012-01-25 17:30:24 <josephcp> you'll probably have to pay fees though
2122 2012-01-25 17:30:40 <gavinandresen> Yes, it'll match the definition of "dust spam"
2123 2012-01-25 17:31:13 <roconnor> Eliel: you can mine it yourself :P
2124 2012-01-25 17:32:06 <Eliel> that kind of transaction would need enforced minimum fee of course.
2125 2012-01-25 17:33:43 <josephcp> i actually have a bitcoin timestamping service written out, but didn't release it because it seemed like a dick thing to do :-P (it did aggregate all signatures into 1 address though)
2126 2012-01-25 17:34:05 <josephcp> webservice, it did a daily transaction
2127 2012-01-25 17:38:45 <midnightmagic> timestamping is an excellent, excellent service. the pgp timestamper is still running strong @ itconsult but it would be much less reproachable to skip the dying usenet posts as proof and go straight to the bitcoin blockchain
2128 2012-01-25 17:42:38 Visalleras has quit (Read error: Connection reset by peer)
2129 2012-01-25 17:44:27 <helo> where is the line drawn between "excellent services" and "blockchain bloat/spam"?
2130 2012-01-25 17:45:19 <[eval]> why not create a blockchain specifically for timestamping and merge mine it. or use namecoin instead of bitcoin, since it's purpose-built for data storage in the chain.
2131 2012-01-25 17:45:30 <josephcp> [eval]: because there's not enough money in merge mining it for the pools
2132 2012-01-25 17:45:47 <[eval]> josephcp: namecoin is being merge-mined
2133 2012-01-25 17:45:51 MrTiggr has quit (Ping timeout: 260 seconds)
2134 2012-01-25 17:45:59 <josephcp> there's enough of an economy for NMC
2135 2012-01-25 17:46:25 <[eval]> so use namecoin instead of a separate chain; namecoin is already made for storing data in the chain
2136 2012-01-25 17:46:44 <[eval]> bitcoin is made primarily for financial transactions
2137 2012-01-25 17:47:11 p0s has quit (Remote host closed the connection)
2138 2012-01-25 17:47:48 blomqvist has quit (Quit: Heroes die once, Cowards live longer!)
2139 2012-01-25 17:47:59 abbe has joined
2140 2012-01-25 17:48:02 abbe has quit (Changing host)
2141 2012-01-25 17:48:02 abbe has joined
2142 2012-01-25 17:48:45 <diki> hey...since I have the blockchain almost imported, anyone know how one can determine the "final balance" of an address?
2143 2012-01-25 17:49:20 <[eval]> blockexplorer is an easy way :P
2144 2012-01-25 17:49:35 <josephcp> yeah you can't precisely do it with bitcoind
2145 2012-01-25 17:49:40 mtve has joined
2146 2012-01-25 17:49:49 <diki> I have the blockchain imported into mysql
2147 2012-01-25 17:49:54 <diki> all txes everything
2148 2012-01-25 17:50:44 <josephcp> get all unredeemed output scripts that match what you want... (is that too obvious to be a tautology?)
2149 2012-01-25 17:51:08 <Eliel> curious, I got paid in the coinbase transaction in both of these blocks http://blockchain.info/block-index/852772/000000000000068c6ddbae5bec15c975671d9ef40a0b49fac106f28c990d326d http://blockchain.info/block-index/852769/00000000000002b87f51a4e2476582ad8d764c3a731e3e0831998fe4a5be8c5e and if I do "bitcoind listtransactions" both show up. Only one of them shows up in bitcoin-qt UI though.
2150 2012-01-25 17:51:29 <Eliel> (they're p2pool blocks)
2151 2012-01-25 17:52:12 <Eliel> running 0.5.2-beta
2152 2012-01-25 17:52:36 <diki> define "output scripts"
2153 2012-01-25 17:52:49 <diki> As I said before, I am rather uneducated in the bitcoin protocol stuff
2154 2012-01-25 17:54:54 user__ has quit (Ping timeout: 276 seconds)
2155 2012-01-25 17:55:07 <josephcp> diki: you should observe how blockexplorer does it (i think they just watch for pubkey outputs and standard transactions)
2156 2012-01-25 17:55:35 <josephcp> so if you convert an address to a <pubkeyhash>, just look for OP_DUP OP_HASH160 <pubKeyHash> OP_EQUALVERIFY OP_CHECKSIG
2157 2012-01-25 17:55:47 <josephcp> as well as your <pubkey> OP_CHECKSIG
2158 2012-01-25 17:56:12 <josephcp> i think this is how blockexplorer does it at least..
2159 2012-01-25 17:59:42 <diki> Eliel:usually, if it has 0 confirmations, or just 1 it wouldnt show uo
2160 2012-01-25 17:59:44 <diki> *up
2161 2012-01-25 17:59:57 <diki> But if it's more than 1, it is weird
2162 2012-01-25 18:00:58 <roconnor> Eliel: I haven't looked, but coinbase output require at least 100 confirmations before they can be spent
2163 2012-01-25 18:01:15 rasengan has joined
2164 2012-01-25 18:01:19 Kolky has joined
2165 2012-01-25 18:01:27 nathan7 has quit (Changing host)
2166 2012-01-25 18:01:27 nathan7 has joined
2167 2012-01-25 18:01:28 <josephcp> 120?
2168 2012-01-25 18:01:40 <roconnor> 120 by the standard client, 100 by official rules
2169 2012-01-25 18:01:46 <josephcp> ohok
2170 2012-01-25 18:02:19 <roconnor> Eliel: that is 16 hours and 40 minutes on average IIRC
2171 2012-01-25 18:02:22 riush has joined
2172 2012-01-25 18:02:23 riush has quit (Changing host)
2173 2012-01-25 18:02:23 riush has joined
2174 2012-01-25 18:05:02 <Eliel> roconnor: I'm not talking about spending. Simply whether they show up in the transactions list or not.
2175 2012-01-25 18:05:19 <Eliel> and one of those shows, the other not. Both have less than 120 confirmations
2176 2012-01-25 18:05:31 <roconnor> oh, strange
2177 2012-01-25 18:05:39 <josephcp> are they both the same version?
2178 2012-01-25 18:06:02 <Eliel> only in bitcoin-qt UI though, "bitcoind listtransactions" shows it fine. It's the same process giving both lists
2179 2012-01-25 18:06:08 <Eliel> so yes, same version.
2180 2012-01-25 18:06:11 <josephcp> oh
2181 2012-01-25 18:06:12 <josephcp> weird
2182 2012-01-25 18:06:13 Diablo-D3 has joined
2183 2012-01-25 18:07:40 user__ has joined
2184 2012-01-25 18:07:55 TD has quit (Quit: TD)
2185 2012-01-25 18:08:00 _Fireball has quit (Read error: No route to host)
2186 2012-01-25 18:08:16 <gavinandresen> The GUI used to have some code to hide mined blocks before they had X confirmations, I have no idea if the QT code does the same.
2187 2012-01-25 18:08:27 <gavinandresen> (the wx-GUI code used to...)
2188 2012-01-25 18:08:43 <gavinandresen> (and I don't remember what X was)
2189 2012-01-25 18:08:53 _Fireball has joined
2190 2012-01-25 18:09:14 Turingi has joined
2191 2012-01-25 18:09:29 <Eliel> I first noticed this when they had 1 and 2 respectively. and it's still only showing only one of them.
2192 2012-01-25 18:10:08 <Eliel> and they have 4 and 5 now
2193 2012-01-25 18:11:08 <gavinandresen> Eliel: definitely sounds like a bug. github.com/bitcoin/bitcoin/issues is the bug tracker
2194 2012-01-25 18:12:11 <gavinandresen> Eliel: if you could put the output of gettransaction <txid> for both transactions in the bug report it might help
2195 2012-01-25 18:12:45 <Eliel> I was thinking to include those blockchain.info links but I can do that too.
2196 2012-01-25 18:13:36 <Eliel> I'm also not the only one seeing this behauviour, quite a few in #p2pool noticed this.
2197 2012-01-25 18:14:12 b4epoche has quit (Read error: Operation timed out)
2198 2012-01-25 18:14:20 <gmaxwell> dikidera: (why am I responding to diki and how did he get unignored) "needs to have 6 confirms before it is spendable" is not true.
2199 2012-01-25 18:14:44 b4epoche has joined
2200 2012-01-25 18:14:52 <Eliel> I'll add this to the issue tracker.
2201 2012-01-25 18:16:49 <gmaxwell> midnightmagic: it would be fairly easy to make a MM timestamper that you just give a set of hashes to and it just emits a set of messages with connecting tree fragments back to a bitcoin header.
2202 2012-01-25 18:18:09 <Eliel> gmaxwell: the only problem with such that is that it'd need to be included in a regular bitcoin transaction, otherwise there's no way to pay miners for the hashpower.
2203 2012-01-25 18:18:43 <gmaxwell> Eliel: No payment is needed, frankly. It's a service we should run just to discourage less efficient ways of doing the same.
2204 2012-01-25 18:18:57 <gmaxwell> The marginal cost of commiting a hash is damn near zero.
2205 2012-01-25 18:19:03 <josephcp> i have like 90% of code written for Ruby on Rails already
2206 2012-01-25 18:19:17 <josephcp> it's so small because it's easy stuff
2207 2012-01-25 18:19:22 <midnightmagic> gmaxwell: capital idea. :) actually, namecoin might be the easiest place to cram it into since the data blocks can be updated, and updated only by their owner.
2208 2012-01-25 18:19:26 <Eliel> then all it'd take is to convince pools to add support.
2209 2012-01-25 18:19:38 traviscj has joined
2210 2012-01-25 18:19:57 <josephcp> i just didn't release it because it seemed like it'd be controvertial
2211 2012-01-25 18:19:59 <gmaxwell> Yes, though since people don't need instant timestamping usually (if they do bitcoin is a bad fit) it doesn't matter if only a fraction run it.
2212 2012-01-25 18:20:17 <midnightmagic> josephcp: Bah. Nothing is controversial.
2213 2012-01-25 18:20:17 <gmaxwell> josephcp: make it just work with merged mining. There isn't anything controversial there.
2214 2012-01-25 18:20:27 <midnightmagic> josephcp: Well.. no technical solutions anyway.
2215 2012-01-25 18:20:28 <helo> namecoin seems like an obvious solution... already merge mined, etc
2216 2012-01-25 18:20:30 <josephcp> well it's currently written to use addresses
2217 2012-01-25 18:20:43 <Eliel> midnightmagic: the only problem being that namecoin only has a fraction of bitcoin hashpower behind it... well, not that it matters much, this application can make do with much less hashpower given the timescales involved.
2218 2012-01-25 18:20:52 <josephcp> i wrote it months ago and didn't realize you can have zero coin outputs so there were some testnet transactions that had 1 satoshi outputs :-P
2219 2012-01-25 18:21:19 <gmaxwell> midnightmagic: for straight timestamping you don't want anything fancy like that. You just want tractability to a header. not even a seperate chain.
2220 2012-01-25 18:21:25 <josephcp> how about i clean it up a little bit and show you guys a prototype/example tomorrow?
2221 2012-01-25 18:21:54 <gmaxwell> josephcp: please make it work by MM before releasing it. :)
2222 2012-01-25 18:22:08 <midnightmagic> Eliel: difficulty is 500k. That's enough. Combine it with cumulative usenet posts a la itconsulting.co.uk does, and voila.
2223 2012-01-25 18:22:16 <josephcp> doesn't that require acceptance from pools?
2224 2012-01-25 18:22:39 <midnightmagic> josephcp: Who gives a crap what the pools think.
2225 2012-01-25 18:22:49 <gmaxwell> josephcp: Noâ I mean actually making it useful requires miners to run it, but there isn't any 'acceptance' part. I'll run it.
2226 2012-01-25 18:22:59 <midnightmagic> I'll run it too.
2227 2012-01-25 18:23:10 <josephcp> that's like 10x more work -_-;
2228 2012-01-25 18:23:13 pusle has joined
2229 2012-01-25 18:23:19 <midnightmagic> I'll run it anyway! :)
2230 2012-01-25 18:23:43 <gmaxwell> josephcp: yes, but it's good workâ without doing that the code is a liability, with doing that its an asset.
2231 2012-01-25 18:23:43 BlueMatt has joined
2232 2012-01-25 18:24:02 <josephcp> nah i wouldn't be interesting in fixing/releasing it via merged mining :-P doesn't seem useful enough for me, i'd rather just leave it ignored haha
2233 2012-01-25 18:24:16 <gmaxwell> :(
2234 2012-01-25 18:24:45 <BlueMatt> devrandom: ping
2235 2012-01-25 18:25:03 <josephcp> address outputs on the bitcoin testnet blockchain were way easy to do and was a fun side project
2236 2012-01-25 18:25:51 ovidiusoft has joined
2237 2012-01-25 18:27:30 TheDealer is now known as imsaguy2
2238 2012-01-25 18:27:32 <gmaxwell> midnightmagic: an neat point is that you could give a hash for notarizing to a notary node.. he builds his notary tree... sends the root to all the other notary nodes he knows of.
2239 2012-01-25 18:27:42 <devrandom> BlueMatt: pong
2240 2012-01-25 18:27:48 <BlueMatt> devrandom: see gribble
2241 2012-01-25 18:28:04 <gmaxwell> midnightmagic: then if one of them solves a block, that notary gets their result from the solver, and just tacks on the tree fragment.
2242 2012-01-25 18:28:08 <devrandom> okay, sounds good
2243 2012-01-25 18:28:18 <midnightmagic> gmaxwell: I was thinking about cooperative stamping services who could as a unit provide some redundancy.
2244 2012-01-25 18:28:22 <gmaxwell> midnightmagic: so it would have simply awesome scalability... since you don't need to flood any notary requests.
2245 2012-01-25 18:28:38 <josephcp> my problem is that the economic incentives aren't there for a timestamping for an alternate blockchain :-/
2246 2012-01-25 18:28:40 <gmaxwell> you just send one hash.
2247 2012-01-25 18:28:50 <midnightmagic> I should email that itconsulting guy, he seems like a stand-up fellow.
2248 2012-01-25 18:28:52 <josephcp> i know it's merged mining, i'm counting manpower/effort/cognitive
2249 2012-01-25 18:28:59 <gmaxwell> josephcp: We're not talking about an alternative chain. There is absolutely no need for a chain.
2250 2012-01-25 18:29:09 <josephcp> oh you mean dump it in the coinbase?
2251 2012-01-25 18:30:01 <gmaxwell> ... ::cries::
2252 2012-01-25 18:30:16 <gmaxwell> The coinbase has a single 32 byte merged mining commitment.
2253 2012-01-25 18:30:19 <Eliel> yep, merkletree root of all the hashes that are to be verified somewhere in a block, can be anywhere in any transaction really.
2254 2012-01-25 18:30:52 <Eliel> but the space defined for merged mining seems like an obvious fit
2255 2012-01-25 18:30:54 <josephcp> oh right
2256 2012-01-25 18:31:01 <midnightmagic> gmaxwell: wait.. how do you prove it existed in the *past* and wasn't created five minutes ago without piling on a blockchain of effort on top of it?
2257 2012-01-25 18:31:23 <midnightmagic> gmaxwell: You mean in a txn?
2258 2012-01-25 18:31:29 <gmaxwell> midnightmagic: you use the bitcoin chain for thatâ you only count trees that made it into the bitcoin chain.
2259 2012-01-25 18:31:41 <gmaxwell> No, not in a txn. That horiffic and shouldn't be done.
2260 2012-01-25 18:32:12 <midnightmagic> gmaxwell: But what's stopping me from hashing to a block from last week?
2261 2012-01-25 18:32:41 <luke-jr> gmaxwell: could you help test Eloipool maybe?
2262 2012-01-25 18:32:47 <gmaxwell> midnightmagic: people can look at the chain and see what was actually there
2263 2012-01-25 18:32:49 Stove has joined
2264 2012-01-25 18:32:51 danbri_ has joined
2265 2012-01-25 18:33:02 <gmaxwell> luke-jr: I don't really have the cycles for it right now. :(
2266 2012-01-25 18:33:03 <Eliel> also, when it's included in the block, the recipient ought to save all data needed to verify that it indeed was part of the block as long as you have the block headers.
2267 2012-01-25 18:33:04 danbri has quit (Read error: Connection reset by peer)
2268 2012-01-25 18:33:29 <Eliel> umm s/recipient/the person who wants to timestamp the hash/
2269 2012-01-25 18:33:39 <gmaxwell> midnightmagic: what the service gives you back is enough that if you have the bitcoin headers you can verify the hash with at least as much security as bitcoin.
2270 2012-01-25 18:34:11 <gmaxwell> (and actually more, because the notary services should also add a conventional signature to the messages)
2271 2012-01-25 18:35:05 JimRogers has joined
2272 2012-01-25 18:35:05 JimRogers has quit (Client Quit)
2273 2012-01-25 18:35:13 BlueMatt has quit (Quit: Ex-Chat)
2274 2012-01-25 18:35:22 JimRogers has joined
2275 2012-01-25 18:35:43 <midnightmagic> Right. But the idea is to prove past chronological existence. I have last week's header, today. Past immutability can only be achieved by traditional pgp stamping services by posting digests to usenet groups and relying on the unlikelihood of a mass-hack. We would have to rely on people having a live feed just trusting that last week's stamp did arrive last week, and not five minutes ago..? I must be missing something fundamental
2276 2012-01-25 18:35:49 <midnightmagic> to what you're saying.
2277 2012-01-25 18:36:49 <gmaxwell> midnightmagic: You are.
2278 2012-01-25 18:36:54 <midnightmagic> thought so.
2279 2012-01-25 18:36:55 BlueMatt has joined
2280 2012-01-25 18:37:16 <gmaxwell> midnightmagic: You know that it couldn't be only five minutes old because you trust that someone didn't just attack bitcoin and rewrite the last week of blocks in order to insert the commitment.
2281 2012-01-25 18:37:24 <Eliel> midnightmagic: bitcoin couldn't work if we couldn't trust the past transactions are very expensive to change.
2282 2012-01-25 18:37:34 <gmaxwell> You're thinking too much about how MM is used for alt chains.
2283 2012-01-25 18:38:06 slush has joined
2284 2012-01-25 18:38:21 <midnightmagic> Eliel: But that's data which is embedded in the blockchain via a merkle hash. The actual merkle tree is perturbed by last week's data. gmaxwell: are you saying *any* data is perturbing the bitcoin blockchain? like in coinbase or something?
2285 2012-01-25 18:38:27 robblesz has joined
2286 2012-01-25 18:38:29 robblesz has quit (Excess Flood)
2287 2012-01-25 18:38:40 <gmaxwell> midnightmagic: the MM hash root.
2288 2012-01-25 18:38:58 robblesz has joined
2289 2012-01-25 18:39:07 <midnightmagic> ah, so there is an altchain, and that requires miner participation.
2290 2012-01-25 18:39:30 <gmaxwell> no.
2291 2012-01-25 18:39:31 Stove has quit (Ping timeout: 260 seconds)
2292 2012-01-25 18:39:31 <midnightmagic> I thought you were just saying having a hashchain outside of mining effort.. :)
2293 2012-01-25 18:39:31 <Eliel> it needs no altchain, really.
2294 2012-01-25 18:39:32 * gmaxwell cries
2295 2012-01-25 18:39:39 <midnightmagic> wow brutal.
2296 2012-01-25 18:39:45 <gmaxwell> H(your message)=1 H(1+2+3) = Notary1 H(Note1 + Namecoin) = AUX H( coinbase+aux) = Bitcoin block.
2297 2012-01-25 18:39:57 <midnightmagic> *ah* namecoin!
2298 2012-01-25 18:40:01 <gmaxwell> no.
2299 2012-01-25 18:40:03 <gmaxwell> die.
2300 2012-01-25 18:40:04 <sipa> it is an altchain with everything except the fact that the 'blocks' of the altchain are not linked to eachother
2301 2012-01-25 18:40:05 <midnightmagic> LOL
2302 2012-01-25 18:40:12 <gmaxwell> sipa: righ. You get it.
2303 2012-01-25 18:40:13 larsivi has joined
2304 2012-01-25 18:40:22 <sipa> you could do so
2305 2012-01-25 18:40:26 <sipa> but there is no reason to
2306 2012-01-25 18:40:49 Dagger3 is now known as Dagger2
2307 2012-01-25 18:41:08 * BlueMatt has an official time for a google phone interview, and is now really freaking out
2308 2012-01-25 18:41:30 <sipa> BlueMatt: job interview?
2309 2012-01-25 18:41:37 hahuang65 has joined
2310 2012-01-25 18:41:38 <BlueMatt> internship
2311 2012-01-25 18:41:40 <gmaxwell> midnightmagic: I'm a notary. You give me a hash. I combine it into a tree. And use that to perturb the bitcoin chain (via a miner). I give you back all the info needed to prove that pertubation to a third party.
2312 2012-01-25 18:41:42 <sipa> cool
2313 2012-01-25 18:41:59 <midnightmagic> "via a miner".
2314 2012-01-25 18:42:00 <gmaxwell> midnightmagic: someone who wants to validate it, gets the bitcoin chain headers.. and the proof data from you.. and sees that it checks out.
2315 2012-01-25 18:42:00 <BlueMatt> no way in hell Ill get it, but meh
2316 2012-01-25 18:42:00 BTC_Bear is now known as BTC_Bear|hbrntng
2317 2012-01-25 18:42:27 <Eliel> midnightmagic: when you want to check if a bitcoin transaction was really included in a past block, you take hash of the transaction and check to see if it's hash exists in the merkle tree for transactions in that block. If it's there, it existed at the time that block was mined.
2318 2012-01-25 18:42:28 <gmaxwell> midnightmagic: yes, the miner puts a tree of notary hashes inside the tree they use for merged mining.. which eventually goes into the coinbase txn.
2319 2012-01-25 18:42:40 cande has quit (Quit: Lämnar)
2320 2012-01-25 18:42:51 <midnightmagic> Right! So that requires miner participation..?
2321 2012-01-25 18:42:55 <sipa> yes
2322 2012-01-25 18:43:18 <gmaxwell> Eliel: it can't be in the merkle tree of transactions but otherwise, right. It would be in the mm tree. (because thats not validated by bitcoin)
2323 2012-01-25 18:43:52 <Eliel> midnightmagic: the same way, here we have a random file someone wants to prove exists, they hash it and have the hash included in a merkle tree inserted into the merged mining area in coinbase.
2324 2012-01-25 18:44:06 <josephcp> the problem is realistic, you're not going to get enough mining power from people because there's insufficient incentive
2325 2012-01-25 18:44:20 <gmaxwell> midnightmagic: an intersting point is that anyone could be a notary service, and any notary service only must give one hash per block to a miner. So it has log(n) scaling. E.g. even if there was only one or two miners doing this we could confirm googleplexes of hashes.
2326 2012-01-25 18:44:57 <Eliel> josephcp: a small pool doing this is sufficient.
2327 2012-01-25 18:44:59 <gmaxwell> josephcp: There is plenty of incentive. It keeps people from runining the decenteralization of bitcoin by creating junk transactions which waste O(N) space for all full validating nodes forever.
2328 2012-01-25 18:45:12 <midnightmagic> okay, yeah. I do and did understand. It definitely requires perturbation of the bitcoin blockchain via participating miners. That's what I've been trying to ask. It sounded like you were saying "no, no perturbation of the bitcoin blockchain required"
2329 2012-01-25 18:45:20 <Eliel> josephcp: because once it's part of a bitcoin block, every other miner is piling confirmations for it
2330 2012-01-25 18:45:23 <midnightmagic> So POW effort is piled on top.
2331 2012-01-25 18:45:28 <gmaxwell> The actual miner himself need only do one hash operation and return one result, in in the process commit to a googleplex of hashes.
2332 2012-01-25 18:45:29 <josephcp> it was way easier to just dump all the hashes into a single file as a simple tree, then RIPEMD160 it, you get full hashing power of bitcoin.
2333 2012-01-25 18:45:35 <midnightmagic> Whichallows us to prove it happened last week.
2334 2012-01-25 18:45:47 <gmaxwell> midnightmagic: no additional _storage_. But yes, perturbation.
2335 2012-01-25 18:45:48 <josephcp> making alternate chain => you're not getting enough security is the problem
2336 2012-01-25 18:45:59 <gmaxwell> josephcp: No...
2337 2012-01-25 18:46:06 <gmaxwell> Thats not a problem here.
2338 2012-01-25 18:46:11 <sipa> josephcp: how is it less secure?
2339 2012-01-25 18:46:13 <josephcp> ohhhhhhh
2340 2012-01-25 18:46:16 <josephcp> i see what you mean
2341 2012-01-25 18:46:21 <josephcp> because it's still going into bitcoin chain
2342 2012-01-25 18:46:22 <Eliel> josephcp: it's the same here, just the place you put the hash is different
2343 2012-01-25 18:46:23 <josephcp> never mind!
2344 2012-01-25 18:46:24 <midnightmagic> Yes, but it's as difficult to change as it would be to re-mine the effort back to that time. Yes, same side-benefits that the altchains like namecoin get.
2345 2012-01-25 18:46:31 <gmaxwell> Halleula!
2346 2012-01-25 18:47:01 <midnightmagic> lol dude I asked like four times and got blanket "no"s.
2347 2012-01-25 18:47:03 <josephcp> but doesn't this still add noise to the coinbase?
2348 2012-01-25 18:47:04 <gmaxwell> midnightmagic: it's different than what namecoin getsâ because it's not totally independant of bitcoin. If bitcoin goes away, your notary proof isn't that useful.
2349 2012-01-25 18:47:13 <gmaxwell> josephcp: no!
2350 2012-01-25 18:47:15 <Eliel> but realistically, the simplest place for a person without connections to bitcoin pools to put such a thing into is a bitcoin transaction.
2351 2012-01-25 18:47:26 <Eliel> so I expect it will end up done that way, a lot.
2352 2012-01-25 18:47:26 <gmaxwell> josephcp: we have 32 bytes of noise already for MM. And this doesn't increase it.
2353 2012-01-25 18:47:34 <midnightmagic> josephcp: only as much noise as it already in there..
2354 2012-01-25 18:47:55 <josephcp> oh i see
2355 2012-01-25 18:47:55 <gmaxwell> Eliel: Noâ they don't have to! they just give it to a service.. who can give it to a service.. who can give it to a service.. who finally has a connection.
2356 2012-01-25 18:48:04 <gmaxwell> (no need for so much indirection, but it's not precluded either)
2357 2012-01-25 18:48:17 <Eliel> gmaxwell: yes, therefore the service needs to get widely known so people know not to use transactions for it
2358 2012-01-25 18:48:22 <gmaxwell> Yes.
2359 2012-01-25 18:48:26 <midnightmagic> gmaxwell: Namecoin isn't effectively independent from bitcoin at this point. I'm strongly doubting the existence of any namecoin-only hashing going on right now.
2360 2012-01-25 18:48:29 <gmaxwell> and have easy to use tools.
2361 2012-01-25 18:48:54 <gmaxwell> midnightmagic: we could all decide that namecoin rocks and bitcoin sucks.. and bitcoin could stop getting mining and namecoin could get it.. and life would be just ducky.
2362 2012-01-25 18:49:04 <gmaxwell> midnightmagic: not that this is likely... but the design allows it.
2363 2012-01-25 18:49:17 <Eliel> gmaxwell: this service in question could face the problem of being spammed with simply random hashes though.
2364 2012-01-25 18:49:33 <gmaxwell> midnightmagic: you don't have to have any bitcoin data to mine namecoin. (you do if you also want to mine bitcoin of course)
2365 2012-01-25 18:49:42 <gmaxwell> Eliel: yes but it scales fantastically.
2366 2012-01-25 18:49:45 storrgie has quit (Quit: Leaving)
2367 2012-01-25 18:49:50 <gavinandresen> Eliel: ain't no such thing as a free lunch....
2368 2012-01-25 18:49:52 Clown- has joined
2369 2012-01-25 18:49:59 <Eliel> yep, the spam would just increase the size of the merkle tree.
2370 2012-01-25 18:50:02 <gavinandresen> Eliel: ... says the person who runs the bitcoin free lunch faucet....
2371 2012-01-25 18:50:11 <gmaxwell> Eliel: no, because you build trees inside trees.
2372 2012-01-25 18:50:13 <Eliel> but no-one needs to store it completely so it's workable
2373 2012-01-25 18:50:18 <gmaxwell> Eight.
2374 2012-01-25 18:50:19 <gmaxwell> er right.
2375 2012-01-25 18:51:04 <gmaxwell> Eliel: E.g. I run could run an accept all notary.. and just rate limit my input to whatever bandwidth I can afford.. no matter how much I get, I still output 1 hash to an upstream notary or miner.
2376 2012-01-25 18:51:36 <gmaxwell> Though perhaps my notarized results are a little big (though the size there has log2(n) scaling so even thats not terrible).
2377 2012-01-25 18:51:43 has quit (Clown|!Clown@static-87-79-93-140.netcologne.de|Ping timeout: 252 seconds)
2378 2012-01-25 18:52:40 <Eliel> gmaxwell: the system also needs some way of sending each sender of a hash the merkle tree path to their hash
2379 2012-01-25 18:52:58 <gmaxwell> Yes, you would poll the place you got it from to get the result.
2380 2012-01-25 18:53:16 <gmaxwell> And they'd poll the place they had notarizing their tree.. and so on, up to the block.
2381 2012-01-25 18:53:59 <Eliel> timestamp certificates :)
2382 2012-01-25 18:54:13 <gmaxwell> Notary services could also earn income not from the bitcoin notary part (because thats so damn cheap that it really will be free) but from additional services, like a signature by a honest to god expert you could bring into court, along with a more precise timestamp.
2383 2012-01-25 18:54:42 paraipan has quit (Ping timeout: 276 seconds)
2384 2012-01-25 18:54:42 sacredch1o has quit (Ping timeout: 276 seconds)
2385 2012-01-25 18:54:57 <gmaxwell> So there is some business room there, even though baseline "evidence" level is basically impossible to make not-free.
2386 2012-01-25 18:55:03 <josephcp> i still don't see the financial incentives for a pool to do merged mining though
2387 2012-01-25 18:55:21 <Eliel> there's advertisement value in it
2388 2012-01-25 18:55:23 JRWR has joined
2389 2012-01-25 18:55:32 <Eliel> you can tell people you do this charitably
2390 2012-01-25 18:55:39 <Eliel> and it's dead simple to implement
2391 2012-01-25 18:56:10 <gmaxwell> josephcp: Why do people pick up litter on the sidewalk in front of their house? ... but moreover, because its so stupidly cheap past the inital implementation.
2392 2012-01-25 18:56:33 sacredchao has joined
2393 2012-01-25 18:57:14 <Eliel> yep, one small pool doing it is more than enough too.
2394 2012-01-25 18:57:22 BTC_Bear is now known as hbrntng!~BTC_Bear@unaffiliated/btc-bear/x-5233302|BTC_Bear
2395 2012-01-25 18:57:24 <gmaxwell> (or alternatively because we'll just get the miner part of the functionality built into pool/p2pool software and no one will even think about it)
2396 2012-01-25 18:58:32 <Eliel> that might also be useful to use with intrusion detection systems. Have the system build a merkle tree out of all files in the system and put the root into the blockchain.
2397 2012-01-25 18:59:33 paraipan has joined
2398 2012-01-25 19:00:04 <Mad7Scientist> What do you think the issue with the I/O lockup is?
2399 2012-01-25 19:00:10 <josephcp> it just doesn't make sense for something as simple as timestamping IMO, the amount of blockchain spam from a service that outputs a single transaction per day is minimal, even with 3-4 competing services
2400 2012-01-25 19:00:16 <Mad7Scientist> I would switch back to wx bitcoin if it didn't crash so often
2401 2012-01-25 19:00:40 <josephcp> (but i still didn't release it for a reason ;-) just comparing it to a full blown altchain
2402 2012-01-25 19:00:44 <gmaxwell> josephcp: 6 billion transactions a day if everyone uses it.
2403 2012-01-25 19:00:49 <josephcp> no
2404 2012-01-25 19:00:59 <josephcp> the service distributes a single file (equivalent to a merle root)
2405 2012-01-25 19:01:08 <josephcp> per day
2406 2012-01-25 19:01:25 <josephcp> merkle
2407 2012-01-25 19:01:56 <josephcp> you need the file to validate, the hash of the file is what's broadcasted on the bitcoin network
2408 2012-01-25 19:03:28 <josephcp> (the social benefit is if you make this service free, it disincentivizes others from rolling their own timestamps)
2409 2012-01-25 19:03:33 <josephcp> which actually does create spam
2410 2012-01-25 19:05:26 <gmaxwell> josephcp: but you can have all that without a single transaction at allâ and without the risk of a socially non-productive competion between peer altruists services creating the very spam you sought to prevent.
2411 2012-01-25 19:05:33 <Eliel> so, because it's existence will reduce transaction spam, better get it working ASAP :)
2412 2012-01-25 19:06:55 <gmaxwell> http://pastebin.com/5cXeT64Q < this is what I was thinking basically.
2413 2012-01-25 19:08:38 traviscj has quit (Ping timeout: 240 seconds)
2414 2012-01-25 19:09:01 traviscj has joined
2415 2012-01-25 19:10:16 <josephcp> hrm you do have a point, you don't need to bother constructing a full blown altchain or whatever
2416 2012-01-25 19:10:29 <gmaxwell> Eliel: to submit you should just be able to go to https://service.host/notary/hashIwanttoCommit ... and URL will say "rejected" "pending since {date}" or return a message with the stuff in that pastebin.
2417 2012-01-25 19:10:42 <gmaxwell> Eliel: and after a while it will just forget about your request.
2418 2012-01-25 19:10:44 <josephcp> since the txns don't need to inter-relate
2419 2012-01-25 19:11:22 <gmaxwell> and the service itself may comput up its root.. then simply contact another service and so on.
2420 2012-01-25 19:11:50 <gmaxwell> (though I'd expect there to only be two layers normallyâ ones run by miners, and ones run facing the general public that use the ones run by miners)
2421 2012-01-25 19:12:25 <josephcp> i haven't taken a look at merged mining much at all, i think i get it but i need to double check in the source code to see exactly -_-;
2422 2012-01-25 19:12:36 [Tycho] has joined
2423 2012-01-25 19:13:15 iocor has quit (Quit: Computer has gone to sleep.)
2424 2012-01-25 19:13:24 <Eliel> gmaxwell: well, nothing preventing the ones run by miners facing the general public :P
2425 2012-01-25 19:13:35 <gmaxwell> Eliel: Correct. And thats fine.
2426 2012-01-25 19:14:05 <gmaxwell> my only point is that the miners need to invest pratically nothing in thisâ no risk of running a high traffic service, they could keep their url private and just tell a few downstream services.
2427 2012-01-25 19:14:36 <Eliel> josephcp: I don't think any code in the merged mining thing will help clear this up.
2428 2012-01-25 19:16:01 <diki> gmaxwell:fail, but meh
2429 2012-01-25 19:16:17 <diki> I was talking about something entirely different
2430 2012-01-25 19:19:55 <gmaxwell> Eliel: I revised the pastbin a bit to allow each stage to add some additional data, http://pastebin.com/1fVRZEBk I don't know what value X should have. Needs to be big enough to include a hash, a short message/url, a timestamp, and a cryptographic signature.
2431 2012-01-25 19:20:31 <gmaxwell> (the reason X should be limited is so that long chains become stupid)
2432 2012-01-25 19:21:18 slush has quit (Ping timeout: 240 seconds)
2433 2012-01-25 19:24:00 <Eliel> I don't think the data needs to separate where different hash tree parts begin and end... well, unless the hash algo is different.
2434 2012-01-25 19:24:14 Zarutian has joined
2435 2012-01-25 19:26:31 <midnightmagic> gmaxwell: indeed we could. here's an awesome video. you in? http://vimeo.com/15479617
2436 2012-01-25 19:27:18 <gmaxwell> Eliel: ah I wanted each notary service to be able to add some data.
2437 2012-01-25 19:27:23 JRWR has quit (Remote host closed the connection)
2438 2012-01-25 19:27:49 <gmaxwell> Eliel: the reason for that is so that they could e.g. add their own more accurate timestamps and contact info if you wanted to get them to testify to their accuracy.
2439 2012-01-25 19:28:10 wirehead` has joined
2440 2012-01-25 19:28:44 <gmaxwell> Really that only needs to happen at the edge, but I wanted the hierarchy to be invisible. So even if EvilService had an exclusive deal with all the miners, Free service could still get in committments by just using EvilService themselves.
2441 2012-01-25 19:28:45 Diablo-D3 has quit (Ping timeout: 255 seconds)
2442 2012-01-25 19:35:40 <CIA-2> bitcoin: p2k * r76a1c4f8185e ecoinpool/apps/ecoinpool/src/ecoinpool_rpc.erl: Support for older versions of cgminer http://tinyurl.com/89mnzaq
2443 2012-01-25 19:36:02 <Eliel> gmaxwell: hmm, true, it would be useful if the services can add information.
2444 2012-01-25 19:37:14 datagutt has quit (Quit: kthxbai)
2445 2012-01-25 19:37:49 erle- has joined
2446 2012-01-25 19:39:20 baz has joined
2447 2012-01-25 19:47:22 devrandom has quit (Ping timeout: 276 seconds)
2448 2012-01-25 19:47:51 inlikeflynn has joined
2449 2012-01-25 19:48:08 iocor has joined
2450 2012-01-25 19:48:08 devrandom has joined
2451 2012-01-25 19:52:51 tomat has joined
2452 2012-01-25 19:53:41 Cablesaurus has quit (Quit: We be chillin - IceChat style)
2453 2012-01-25 19:56:14 traviscj has quit (Remote host closed the connection)
2454 2012-01-25 19:58:11 danbri_ has quit (Read error: Connection reset by peer)
2455 2012-01-25 19:58:15 danbri has joined
2456 2012-01-25 20:01:00 <Eliel> Can I somehow include images with the bugreport?
2457 2012-01-25 20:01:04 <diki> How does one extract the scripts? I was browsing through the ABE db, and am only seeing http://pastebin.com/N4kFwuLK
2458 2012-01-25 20:03:15 <BlueMatt> Eliel: just put it in an image upload site and link it
2459 2012-01-25 20:06:00 <Eliel> hmm.. I don't usually use those, which one has least annoying features?
2460 2012-01-25 20:06:17 <BlueMatt> I use them so rarely I have no idea
2461 2012-01-25 20:06:51 vite has joined
2462 2012-01-25 20:07:03 vite has left ("Leaving")
2463 2012-01-25 20:07:32 <Eliel> oh well, I guess I'll just go with the first one that pops into my mind then
2464 2012-01-25 20:07:49 <BlueMatt> thats what I usually do...
2465 2012-01-25 20:07:58 <josephcp> diki: i think ABE has some builtin deserializer written in python from what I remember
2466 2012-01-25 20:08:15 <josephcp> you can call a method
2467 2012-01-25 20:08:44 lyspooner has joined
2468 2012-01-25 20:08:46 <josephcp> yeah looking at their repo what you're looking for should be in deserialize.py
2469 2012-01-25 20:10:09 <diki> I am still unclear on how to actually find the final balance of an address
2470 2012-01-25 20:10:27 <diki> As in I have the data, just don't know how to use it
2471 2012-01-25 20:13:02 <josephcp> read the source code of ABE
2472 2012-01-25 20:13:28 <josephcp> 1 minute i found in abe.py q_getreceivedbyaddress shows the SQL query
2473 2012-01-25 20:14:24 <diki> I bet it's a massive left join
2474 2012-01-25 20:14:33 <diki> not sure if it works on views though
2475 2012-01-25 20:15:42 danbri has quit (Ping timeout: 252 seconds)
2476 2012-01-25 20:15:45 <josephcp> oh looks like you need to just subtract the sends though (or construct a better query yourself, it should be enough information)
2477 2012-01-25 20:16:21 <diki> I am still unsure of how blockexplorer.com does such complex queries and is still fast
2478 2012-01-25 20:16:38 <josephcp> it's not complex
2479 2012-01-25 20:16:57 <diki> just look at them joins
2480 2012-01-25 20:17:04 danbri has joined
2481 2012-01-25 20:17:49 <gribble> New news from bitcoinrss: jojkaart opened issue 784 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/784>
2482 2012-01-25 20:18:09 BlueMatt has quit (Ping timeout: 245 seconds)
2483 2012-01-25 20:18:49 krysits has joined
2484 2012-01-25 20:19:11 <krysits> a do not want code todau
2485 2012-01-25 20:19:16 <josephcp> you don't a lot of joins if you want to find the balance
2486 2012-01-25 20:22:10 <josephcp> oh looks like the way ABE is constructed you do because of the way it handles blocks, oh well
2487 2012-01-25 20:22:58 <Eliel> there, issue 784 is my bugreport :)
2488 2012-01-25 20:24:55 Stove has joined
2489 2012-01-25 20:25:24 <Eliel> roconnor, do you have thoughts on this? https://bitcointalk.org/index.php?topic=60433.msg712674#msg712674
2490 2012-01-25 20:25:43 <CIA-2> bitcoin: Luke Dashjr * ra60ae4b2a8cf eloipool/jsonrpcserver.py: Refactor JSONRPCServer using a (mostly) asynchronous model http://tinyurl.com/7bz3er4
2491 2012-01-25 20:26:18 RazielZ has quit (Ping timeout: 240 seconds)
2492 2012-01-25 20:26:19 <diki> Eliel:is this your wallet in the image?
2493 2012-01-25 20:26:25 danbri_ has joined
2494 2012-01-25 20:26:25 danbri has quit (Read error: Connection reset by peer)
2495 2012-01-25 20:29:14 user__ has quit (Quit: Leaving)
2496 2012-01-25 20:29:21 <roconnor> Eliel: in priciple you can link C++ against Haskell ... though you have to bring in the whole haskell runtime :D
2497 2012-01-25 20:30:12 <roconnor> Eliel: maybe one could compile the pure fragment of Haskell to C
2498 2012-01-25 20:30:23 <roconnor> Eliel: but more practically one would get Haskell to generate C
2499 2012-01-25 20:30:52 danbri_ has quit (Ping timeout: 248 seconds)
2500 2012-01-25 20:31:39 <Eliel> roconnor: precompiling it would throw away some of the benefits I think.
2501 2012-01-25 20:32:01 <roconnor> yes
2502 2012-01-25 20:32:08 <roconnor> but not all of them.
2503 2012-01-25 20:32:53 <Eliel> a subset of full haskell runtime should be enough for this purpose I think.
2504 2012-01-25 20:34:06 danbri has joined
2505 2012-01-25 20:34:51 <Eliel> if we could get that done, it'd make it a much easier task to get network wide upgrades done.
2506 2012-01-25 20:35:09 <CIA-2> bitcoin: Luke Dashjr * ra60ae4b2a8cf eloipool/jsonrpcserver.py: Refactor JSONRPCServer using a (mostly) asynchronous model http://tinyurl.com/7bz3er4
2507 2012-01-25 20:36:22 <diki> I am still unsure of what this bip 17 will do actually
2508 2012-01-25 20:36:49 <diki> What I got is that it will make it so that two clients need to be online to spend coins
2509 2012-01-25 20:36:57 <diki> of course, I could've gotten it wrong
2510 2012-01-25 20:37:35 <Eliel> roconnor: what other languages do you think could be considered for this?
2511 2012-01-25 20:37:58 <roconnor> Eliel: Agda, coq, epigram
2512 2012-01-25 20:38:04 <roconnor> but of course I'd think that
2513 2012-01-25 20:38:06 <roconnor> PVS
2514 2012-01-25 20:38:35 <Eliel> would one of those be significantly lighter weight?
2515 2012-01-25 20:39:01 <roconnor> coq can extract to haskell, ocaml, and scheme, ... though not perfrectly :(
2516 2012-01-25 20:39:41 <roconnor> PVS was used to verify the seL4 kernel IIRC
2517 2012-01-25 20:39:59 <gmaxwell> ocaml is pretty easy runtime wise.
2518 2012-01-25 20:40:42 <roconnor> ocaml extraction is probably well supported, but scheme has no type system to satify.
2519 2012-01-25 20:41:28 <CIA-2> bitcoin: Luke Dashjr * r478d9a7ddafd eloipool/eloipool.py: Move JSONRPCServer binding to config file http://tinyurl.com/7l7yzjb
2520 2012-01-25 20:41:37 slush has joined
2521 2012-01-25 20:41:57 <Mad7Scientist> (gdb) backtrace
2522 2012-01-25 20:41:57 <Mad7Scientist> #0 0xb6558e5d in pread64 () from /lib/libpthread.so.0
2523 2012-01-25 20:41:57 <Mad7Scientist> #1 0xb760d1bb in __os_io () from /usr/lib/libdb_cxx-4.8.so
2524 2012-01-25 20:42:06 <Mad7Scientist> It's locking up for 30 seconds at a time like that
2525 2012-01-25 20:42:14 <Mad7Scientist> i'm going to have to switch back to the older version
2526 2012-01-25 20:42:29 <gmaxwell> Mad7Scientist: I'd bet money that its no different.
2527 2012-01-25 20:42:45 <gmaxwell> Nothing you've shown so far looks like anything that would be different between different versions.
2528 2012-01-25 20:42:49 <gmaxwell> :(
2529 2012-01-25 20:43:13 RazielZ has joined
2530 2012-01-25 20:43:27 <Mad7Scientist> http://paste.pocoo.org/show/540723/
2531 2012-01-25 20:43:33 <Mad7Scientist> Well I'll go back to the old one anyway
2532 2012-01-25 20:43:42 <Mad7Scientist> This problem with the qt version comes and goes
2533 2012-01-25 20:43:46 <Mad7Scientist> It'll work for a half hour okay
2534 2012-01-25 20:43:50 <Mad7Scientist> then this happens for 10 minutes
2535 2012-01-25 20:44:42 userhg has joined
2536 2012-01-25 20:45:13 <gmaxwell> it's probably happening when it gets a block and your IO problems make it take forever to verify it.
2537 2012-01-25 20:45:45 slush has quit (Ping timeout: 240 seconds)
2538 2012-01-25 20:49:22 <Mad7Scientist> nah
2539 2012-01-25 20:49:35 <Mad7Scientist> It's happened when the block was 13 minutes old according to the indicator
2540 2012-01-25 20:49:47 bob12321 has joined
2541 2012-01-25 20:49:49 <Mad7Scientist> It's related to mining only
2542 2012-01-25 20:50:17 <Mad7Scientist> If I stop this one thread my miners disconnect but the bitcoin program interface still works
2543 2012-01-25 20:50:43 <CIA-2> libbitcoin: genjix * r29b6f622ccec / (9 files in 4 dirs): fetch_outputs(pubkey_hash, completion_handler) http://tinyurl.com/85uy9jc
2544 2012-01-25 20:51:28 danbri has quit (Read error: Connection reset by peer)
2545 2012-01-25 20:54:10 <Mad7Scientist> !!!!!
2546 2012-01-25 20:54:11 <gribble> Error: "!!!!" is not a valid command.
2547 2012-01-25 20:54:17 <Mad7Scientist> Now the old version says "run database recovery"
2548 2012-01-25 20:54:20 <Mad7Scientist> oh no :(
2549 2012-01-25 20:54:23 <Mad7Scientist> what happened now
2550 2012-01-25 20:55:16 <gmaxwell> where did the old version you're running come from?
2551 2012-01-25 20:55:34 <gmaxwell> You _cannot_ downgrade libdb versions. The bitcoin.org binaries use 4.x.
2552 2012-01-25 20:55:47 <sipa> 4.8, currently
2553 2012-01-25 20:56:23 <Mad7Scientist> I want back to 0.3.21
2554 2012-01-25 20:56:37 <luke-jr> nobody maintains that
2555 2012-01-25 20:56:40 <Mad7Scientist> OH NO
2556 2012-01-25 20:56:49 <Mad7Scientist> DID THE NEW version update the databases without asking?
2557 2012-01-25 20:56:53 <luke-jr> of course
2558 2012-01-25 20:56:58 <Mad7Scientist> ****
2559 2012-01-25 20:57:09 <Mad7Scientist> There goes a whole day of downloading the block chain again
2560 2012-01-25 20:57:27 <Mad7Scientist> WHY.
2561 2012-01-25 20:57:34 <Mad7Scientist> This is awful
2562 2012-01-25 20:57:42 <Mad7Scientist> it might take 2 days
2563 2012-01-25 20:57:44 <gmaxwell> Wasn't .21 _also_ 4.8?
2564 2012-01-25 20:57:46 [eval] has quit (Ping timeout: 240 seconds)
2565 2012-01-25 20:57:47 <luke-jr> gmaxwell: no?
2566 2012-01-25 20:57:50 <luke-jr> 4.7 back then
2567 2012-01-25 20:57:56 <luke-jr> except on Gentoo
2568 2012-01-25 20:58:12 <Mad7Scientist> I could have compiled the bitcoin-qt version with the older libdb couldn't I
2569 2012-01-25 20:58:34 <sipa> yes
2570 2012-01-25 20:58:44 <Mad7Scientist> Is there a program I can use to go back a version
2571 2012-01-25 20:59:17 <Mad7Scientist> Why did you guys make it update the databases without asking
2572 2012-01-25 20:59:25 <Mad7Scientist> I backed up wallet.dat but that's it
2573 2012-01-25 20:59:53 <sipa> because 4.7 was not available anymore in recent distro's
2574 2012-01-25 21:00:01 <Mad7Scientist> There is no performance advantage to using the latest DB anyway is there
2575 2012-01-25 21:00:11 <sipa> no
2576 2012-01-25 21:00:33 <luke-jr> Mad7Scientist: it's not like the switch didn't have HUGE WARNINGS
2577 2012-01-25 21:00:45 <luke-jr> Mad7Scientist: also, there is no excuse to be using an unmodified 0.3.21 anymore
2578 2012-01-25 21:00:50 <Mad7Scientist> I didn't know about the warnings
2579 2012-01-25 21:01:04 <luke-jr> 0.4.x is the minimum supported version
2580 2012-01-25 21:01:14 <Mad7Scientist> Well 0.3.21 is the only one with the I/O block mining issue that I know of and I don't want to try 10 other versions
2581 2012-01-25 21:01:45 <Mad7Scientist> Will 0.4 work with db 4.8?
2582 2012-01-25 21:01:50 danbri has joined
2583 2012-01-25 21:01:55 <gmaxwell> Mad7Scientist: I'm doubtful that it'll be any better. I expect your expirence is just goofed up by remembering the performance from a smaller network.
2584 2012-01-25 21:02:15 <gmaxwell> Mad7Scientist: yes. the .4 binaries use 4.8 too. IIRC.
2585 2012-01-25 21:02:26 <Mad7Scientist> I'll try the oldest 4.x
2586 2012-01-25 21:05:19 traviscj has joined
2587 2012-01-25 21:06:04 <Mad7Scientist> Where were these HUGE warnings?
2588 2012-01-25 21:08:19 <luke-jr> I/O block mining issue ?
2589 2012-01-25 21:10:11 <Mad7Scientist> When miners are connected to bitcoin-qt -server the thread that does the mining hangs on I/O to ~/.bitcoin/ for 30 seconds at a time then frees up for 5 seconds
2590 2012-01-25 21:10:16 <Mad7Scientist> and keeps doing that for 10 minutes
2591 2012-01-25 21:10:21 <Mad7Scientist> then it works okay for a half hour or so
2592 2012-01-25 21:14:04 <Mad7Scientist> 4.0 doesn't work.
2593 2012-01-25 21:17:03 <Mad7Scientist> 0.4.1 says bad database too
2594 2012-01-25 21:18:22 BCBot has quit (Ping timeout: 276 seconds)
2595 2012-01-25 21:19:46 cdecker has joined
2596 2012-01-25 21:20:38 BCBot has joined
2597 2012-01-25 21:22:44 <etotheipi_> ll
2598 2012-01-25 21:22:47 <etotheipi_> gah!
2599 2012-01-25 21:22:57 <Mad7Scientist> I don't know what's going on
2600 2012-01-25 21:23:08 <Mad7Scientist> db4.5_dump is able to read all the files in ~/.bitcoin/
2601 2012-01-25 21:23:36 danbri_ has joined
2602 2012-01-25 21:23:50 <Mad7Scientist> but in database/ there is a log.something file which is "version 16" according to file
2603 2012-01-25 21:24:05 slush has joined
2604 2012-01-25 21:25:03 danbri has quit (Read error: Connection reset by peer)
2605 2012-01-25 21:26:29 lyspooner has quit (Quit: ChatZilla 0.9.88 [Firefox 8.0.1/20111120135848])
2606 2012-01-25 21:26:39 <Mad7Scientist> hey it's working
2607 2012-01-25 21:26:45 <Mad7Scientist> I just had to delete the database/log file
2608 2012-01-25 21:27:07 <gmaxwell> Mad7Scientist: ugh. you should have shut down bitcoin cleanly before changing versions. :(
2609 2012-01-25 21:28:03 num1 has joined
2610 2012-01-25 21:28:06 <gmaxwell> Mad7Scientist: if your node was in the process of validating a block when you shut down then deleted the logs then its possible when the transaction it was processing is eventually spent you'll reject that block and just get stuck.
2611 2012-01-25 21:28:56 TD has joined
2612 2012-01-25 21:30:36 JZavala has quit (Ping timeout: 248 seconds)
2613 2012-01-25 21:40:09 ne0futur has joined
2614 2012-01-25 21:40:09 ne0futur has quit (Changing host)
2615 2012-01-25 21:40:09 ne0futur has joined
2616 2012-01-25 21:40:40 egecko_ has quit (Quit: ~ Trillian Astra - www.trillian.im ~)
2617 2012-01-25 21:42:47 egecko has joined
2618 2012-01-25 21:43:37 <Mad7Scientist> gmaxwell, I did
2619 2012-01-25 21:44:13 <Mad7Scientist> I shut down .5 then the other versions wouldn't start because of the database. I deleted the log file and now .4.2 started up just fine
2620 2012-01-25 21:44:15 <gmaxwell> Mad7Scientist: if you were cleanly shut down there shouldn't have been any database logs to delete. IIRC.
2621 2012-01-25 21:44:36 <sipa> gmaxwell: bdb always keeps one log file
2622 2012-01-25 21:44:58 <gmaxwell> ah.
2623 2012-01-25 21:45:09 <Mad7Scientist> log file "version 16"
2624 2012-01-25 21:45:14 <Mad7Scientist> now there is a new one version 15
2625 2012-01-25 21:45:43 BlueMatt has joined
2626 2012-01-25 21:49:40 egecko has quit (Quit: ~ Trillian Astra - www.trillian.im ~)
2627 2012-01-25 21:50:06 <gmaxwell> sipa: is this pull request no good anymore with the network stack refactor that happened recently? https://github.com/bitcoin/bitcoin/pull/427
2628 2012-01-25 21:50:37 <sipa> gmaxwell: netbase contains most of that pull request actually
2629 2012-01-25 21:50:41 <CIA-2> bitcoin: p2k * r8223e4273bcc ecoinpool/apps/ecoinpool/src/mysql_sharelogger.erl: Circumvent bug in MySQL driver http://tinyurl.com/7c8cabo
2630 2012-01-25 21:50:42 <CIA-2> bitcoin: p2k * r67c06e69287b ecoinpool/apps/ecoinpool/src/ecoinpool_db.erl: Bugfix http://tinyurl.com/7dot9ke
2631 2012-01-25 21:50:54 <gmaxwell> sipa: yea I though so.
2632 2012-01-25 21:51:02 <sipa> ipv6 support (at least the obvious part) is trivial now, but i'm waiting until addrman is done and working
2633 2012-01-25 21:51:06 <gmaxwell> sipa: did your address code compile? :)
2634 2012-01-25 21:51:40 <sipa> it did!
2635 2012-01-25 21:51:48 <sipa> but there is still some integration work to do
2636 2012-01-25 21:51:49 egecko has joined
2637 2012-01-25 21:51:58 <sipa> after fixing about 20 compile errors
2638 2012-01-25 22:02:26 ovidiusoft2 has joined
2639 2012-01-25 22:03:57 ThomasV_ has joined
2640 2012-01-25 22:05:23 ovidiusoft has quit (Ping timeout: 252 seconds)
2641 2012-01-25 22:06:17 traviscj has quit (Remote host closed the connection)
2642 2012-01-25 22:12:33 traviscj has joined
2643 2012-01-25 22:14:44 ovidiusoft2 has quit (Ping timeout: 245 seconds)
2644 2012-01-25 22:17:42 gavinandresen has quit (Quit: gavinandresen)
2645 2012-01-25 22:26:10 b4epoche_ has joined
2646 2012-01-25 22:26:43 b4epoche has quit (Read error: Operation timed out)
2647 2012-01-25 22:26:44 b4epoche_ is now known as b4epoche
2648 2012-01-25 22:27:55 paraipan has quit (Ping timeout: 276 seconds)
2649 2012-01-25 22:28:38 traviscj has quit (Ping timeout: 240 seconds)
2650 2012-01-25 22:29:59 Nicksasa has quit (Ping timeout: 240 seconds)
2651 2012-01-25 22:33:36 ThomasV_ has quit (Ping timeout: 244 seconds)
2652 2012-01-25 22:33:55 Zarutian has quit (Quit: Zarutian)
2653 2012-01-25 22:35:59 dstien has quit (Ping timeout: 245 seconds)
2654 2012-01-25 22:36:28 dstien has joined
2655 2012-01-25 22:36:53 userhg has quit (Quit: Leaving)
2656 2012-01-25 22:39:19 <Cryo> regarding the topic, isn't there a way for the network to tell the client to update?
2657 2012-01-25 22:40:13 <sipa> no-- that's pretty hard to do if you do not want to rely on a centralized service
2658 2012-01-25 22:40:28 paraipan has joined
2659 2012-01-25 22:40:35 <BlueMatt> actually, yes
2660 2012-01-25 22:40:37 <BlueMatt> gitian
2661 2012-01-25 22:40:43 <BlueMatt> it just hasnt been implemented yet
2662 2012-01-25 22:40:53 <sipa> i didn't say "impossible" :)
2663 2012-01-25 22:41:15 <Cryo> because tor does it :)
2664 2012-01-25 22:41:34 <BlueMatt> do they not rely on a signed central service?
2665 2012-01-25 22:41:48 <Cryo> I don't know the details, just that it works
2666 2012-01-25 22:42:00 <BlueMatt> they probably just use a central service that is eg pgp signed
2667 2012-01-25 22:42:19 stalled has quit (Ping timeout: 240 seconds)
2668 2012-01-25 22:42:24 <BlueMatt> (if I ever get around to implementing it) were gonna do one better ;)
2669 2012-01-25 22:42:30 pusle has quit ()
2670 2012-01-25 22:42:39 <Cryo> cool
2671 2012-01-25 22:42:52 <BlueMatt> or someone else feels like going it (but I dont think anyone else feels like writing platform-specific gitian code)...
2672 2012-01-25 22:43:06 <BlueMatt> anyway, if you are on linux use the ppa
2673 2012-01-25 22:43:23 <BlueMatt> s/linux/ubuntu/
2674 2012-01-25 22:43:30 <BlueMatt> (maybe debian if you feel like hacking around)
2675 2012-01-25 22:46:18 gp5st has quit (Quit: Leaving.)
2676 2012-01-25 22:47:52 pickett_ has quit (Remote host closed the connection)
2677 2012-01-25 22:48:24 traviscj has joined
2678 2012-01-25 22:51:50 pickett has joined
2679 2012-01-25 22:53:03 TD has quit (Quit: TD)
2680 2012-01-25 22:56:51 graingert has quit (Quit: Leaving.)
2681 2012-01-25 22:56:51 RazielZ has quit (Read error: Connection reset by peer)
2682 2012-01-25 22:57:10 stalled has joined
2683 2012-01-25 22:57:12 RazielZ has joined
2684 2012-01-25 22:57:42 topi` has quit (Read error: Operation timed out)
2685 2012-01-25 22:57:47 topi` has joined
2686 2012-01-25 22:57:57 graingert has joined
2687 2012-01-25 22:58:01 smtmnyz has quit (Read error: Operation timed out)
2688 2012-01-25 22:58:03 graingert has quit (Changing host)
2689 2012-01-25 22:58:03 graingert has joined
2690 2012-01-25 22:58:20 smtmnyz has joined
2691 2012-01-25 22:58:32 booo has quit (Ping timeout: 260 seconds)
2692 2012-01-25 22:59:46 JFK911 has quit (Ping timeout: 240 seconds)
2693 2012-01-25 22:59:59 [Tycho] has quit (Remote host closed the connection)
2694 2012-01-25 23:00:44 iocor has quit (Quit: Computer has gone to sleep.)
2695 2012-01-25 23:01:19 JFK911 has joined
2696 2012-01-25 23:02:23 _Fireball has quit (Quit: I love my HydraIRC -> http://www.hydrairc.com <-)
2697 2012-01-25 23:02:39 <diki> How can, after years of existence of bitcoin, only 2.8 million addresses exist in the blockchain(not counting the default keypool size of 100 pre-generated addresses)
2698 2012-01-25 23:03:22 agricocb has quit (Quit: Leaving.)
2699 2012-01-25 23:03:48 <diki> 2.8 is what I "counted" without searching for unique oness
2700 2012-01-25 23:03:54 <doublec> addresses don't appear in the blockchain unless they're involved in a transaction
2701 2012-01-25 23:04:08 <diki> doublec:every single day there are so many transactions
2702 2012-01-25 23:13:47 vhsjon has quit (Quit: vhsjon)
2703 2012-01-25 23:14:17 <helo> maybe 10k transactions/day?
2704 2012-01-25 23:15:50 graingert has quit (Read error: Connection reset by peer)
2705 2012-01-25 23:19:07 davex__ has joined
2706 2012-01-25 23:19:40 hexTech has quit (Remote host closed the connection)
2707 2012-01-25 23:20:26 wirehead` has quit (Ping timeout: 240 seconds)
2708 2012-01-25 23:20:38 hexTech has joined
2709 2012-01-25 23:21:43 pasky_ is now known as pasky
2710 2012-01-25 23:22:03 graingert has joined
2711 2012-01-25 23:22:08 davex__ has quit (Client Quit)
2712 2012-01-25 23:22:47 RazielZ has quit (Quit: Leaving)
2713 2012-01-25 23:23:20 Prattler has joined
2714 2012-01-25 23:24:30 agricocb has joined
2715 2012-01-25 23:26:14 Turingi has quit (Read error: Connection reset by peer)
2716 2012-01-25 23:27:28 krysits has quit (Remote host closed the connection)
2717 2012-01-25 23:27:36 krysits has joined
2718 2012-01-25 23:32:29 Disposition has quit (Read error: Connection reset by peer)
2719 2012-01-25 23:32:47 Clipse has quit (Read error: Connection reset by peer)
2720 2012-01-25 23:33:03 iocor has joined
2721 2012-01-25 23:33:19 Disposition has joined
2722 2012-01-25 23:34:09 hexTech has quit (Remote host closed the connection)
2723 2012-01-25 23:35:03 hexTech has joined
2724 2012-01-25 23:38:58 Cory has joined
2725 2012-01-25 23:40:44 hexTech has quit (Remote host closed the connection)
2726 2012-01-25 23:41:54 hexTech has joined
2727 2012-01-25 23:45:57 larsivi_ has joined
2728 2012-01-25 23:49:46 copumpkin has quit (Quit: Computer has gone to sleep.)
2729 2012-01-25 23:56:32 bob12321 has joined
2730 2012-01-25 23:57:04 BurtyB has quit (Read error: Connection reset by peer)
2731 2012-01-25 23:57:40 dikidera has joined
2732 2012-01-25 23:58:26 BurtyB has joined
2733 2012-01-25 23:59:36 wtfman[away] has quit (Ping timeout: 252 seconds)