1 2012-07-26 00:00:06 <Diablo-D3> you should never create another function that just exists to be the internals of another
  2 2012-07-26 00:00:16 <jgarzik> Diablo-D3: rofl
  3 2012-07-26 00:00:30 <Diablo-D3> it should focus on a well defined concept.
  4 2012-07-26 00:00:53 <jgarzik> that means to God objects and MEGAMOTHs
  5 2012-07-26 00:00:57 <jgarzik> *leads to
  6 2012-07-26 00:01:05 <Diablo-D3> nope
  7 2012-07-26 00:01:10 <Diablo-D3> because thats the other end of it
  8 2012-07-26 00:01:13 <Diablo-D3> code should be able to be read.
  9 2012-07-26 00:01:21 <Joric> not a pythonista either mostly wrote c/c++ games
 10 2012-07-26 00:01:36 <Diablo-D3> Ive written more C in my life than any other language
 11 2012-07-26 00:01:36 <jgarzik> small, readable, debuggable.  break up code into smaller functions, best for human consumption.  compiler will inline what it needs.
 12 2012-07-26 00:01:38 <Diablo-D3> its nearly perfect.
 13 2012-07-26 00:01:53 <jgarzik> of course, if you are programming in Java, your code is broken by design anyway, so that doesn't apply
 14 2012-07-26 00:02:43 <Diablo-D3> I hate how I had to write one part of lugh, though
 15 2012-07-26 00:03:13 <Diablo-D3> due to how setjump is so goddamned braindead
 16 2012-07-26 00:03:19 <Diablo-D3> I had to add internal functions.
 17 2012-07-26 00:04:29 <Diablo-D3> setjump _must_ be called inside the function you're calling it from
 18 2012-07-26 00:04:36 <Diablo-D3> you cant call it from a wrapper function
 19 2012-07-26 00:04:52 <Diablo-D3> bastards.
 20 2012-07-26 00:10:32 <gribble> New news from bitcoinrss: fanquake opened pull request 1630 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1630>
 21 2012-07-26 00:25:52 <gribble> New news from bitcoinrss: laanwj opened pull request 1631 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1631>
 22 2012-07-26 01:32:38 <gribble> New news from bitcoinrss: luke-jr opened pull request 1632 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1632>
 23 2012-07-26 01:36:11 <Bitxk> ?
 24 2012-07-26 01:50:13 <midnightmagic> has anyone written scripting that can do coincontrol with the raw tx api yet?
 25 2012-07-26 01:50:24 <midnightmagic> (on an address level)
 26 2012-07-26 01:52:19 <t3a> !rank
 27 2012-07-26 01:52:20 <gribble> #1 pool (88), #2 gribble (7), #3 faq (5), #4 wiki (4), #5 blockchainsnapshot (4), #6 website (2), #7 googlegroup (1), #8 bbe (1), #9 trade (1), #10 freebitcoins (0), #11 faucet (0)
 28 2012-07-26 02:17:25 <midnightmagic> luke-jr: would adding an 's' to named structures might be miconstrued as a differently-named structure? Why did you quote 'return 0;' ? Finally, if you are settling on American spelling convention (canceled), why did you change zeroes to zeros?
 29 2012-07-26 02:17:53 <Bitxk> what kind of "format" is the 51 character lenght prv key?
 30 2012-07-26 02:19:17 <luke-jr> midnightmagic: maybe; proper grammar; English spelling, though "zeros" is correct either way
 31 2012-07-26 02:19:30 <luke-jr> also, I didn't change American spelling to English spelling or vice-versa
 32 2012-07-26 02:22:25 <midnightmagic> luke-jr: I notice you are splitting up doublewords in some cases but not others. "ciphertext" for example you changed to "cipher text" but you didn't change "loopback" to "loop back", even though "ciphertext" is more common. Come to think of it, I don't recall ever seeing "cipher text" in the context of crypto papers like in iacr archive for example. That might make some people grin. :)
 33 2012-07-26 02:23:19 <midnightmagic> also you left "plaintext".
 34 2012-07-26 02:24:31 <midnightmagic> luke-jr: "cancelled" is English spelling, and you changed it to "canceled", which is the American spelling.
 35 2012-07-26 02:25:04 <midnightmagic> anyway, don't care in the end, but that kind of churn might make merging harder-ish for people.
 36 2012-07-26 02:27:33 <midnightmagic> lol where's vrag anyway sheesh
 37 2012-07-26 02:28:44 <copumpkin> hiding
 38 2012-07-26 02:28:47 <copumpkin> he'll be back soon
 39 2012-07-26 02:31:25 <luke-jr> midnightmagic: pretty sure it was canceled -> cancelled O.o
 40 2012-07-26 02:31:39 <midnightmagic> luke-jr: er..  either way. :)
 41 2012-07-26 02:32:44 <luke-jr> midnightmagic: is canceled actually valid American ?
 42 2012-07-26 02:33:32 <phantomcircuit> cancelled
 43 2012-07-26 02:33:43 <midnightmagic> luke-jr: American spelling has switched to the one-'l' versions but it's a 1:5 ratio of ll to l for those words it's changed for.
 44 2012-07-26 02:33:53 <phantomcircuit> canceled is indeed the american spelling
 45 2012-07-26 02:33:58 <midnightmagic> luke-jr: So not everybody got the memo.
 46 2012-07-26 02:33:59 <phantomcircuit> i've always used cancelled
 47 2012-07-26 02:34:36 <luke-jr> same here
 48 2012-07-26 02:34:48 <luke-jr> canceled seems more like a typo people are tolerating IMO
 49 2012-07-26 02:34:48 <midnightmagic> you guys are in the group of people who didn't get the memo. :)
 50 2012-07-26 02:34:58 <midnightmagic> latent canadians..
 51 2012-07-26 02:35:48 <Bitxk> :D i succed importing a key into electrum :D
 52 2012-07-26 02:35:59 <Bitxk> i know no one cares, but i just say it
 53 2012-07-26 02:36:19 <midnightmagic> nooo, SuprTiggr come back
 54 2012-07-26 02:48:01 <MC-Eeepc> is it true that forest fixed p2pools strange inefficiency
 55 2012-07-26 02:49:32 <Bitxk> no idea
 56 2012-07-26 02:49:48 <Bitxk> anyone use electrum? i have a strange behaviour
 57 2012-07-26 02:50:07 <gmaxwell> MC-Eeepc: it's old news.
 58 2012-07-26 02:50:28 <Joric> Bitxk, like what
 59 2012-07-26 02:50:31 <MC-Eeepc> so its true?
 60 2012-07-26 02:50:47 <Bitxk> i succed to import a wallet into electrum, i mean, an address
 61 2012-07-26 02:50:57 <gmaxwell> MC-Eeepc: Yes. like over a month ago.
 62 2012-07-26 02:50:58 <Bitxk> now i close electrum, and i open it and address list is empty
 63 2012-07-26 02:51:04 <Bitxk> i close it and open it again and its full
 64 2012-07-26 02:51:09 <Bitxk> i repeat several times
 65 2012-07-26 02:51:13 <MC-Eeepc> jolly good
 66 2012-07-26 02:51:15 <Bitxk> and it continously alternate
 67 2012-07-26 02:51:19 <Bitxk> between nothing and all
 68 2012-07-26 02:51:30 <MC-Eeepc> now there is no reason not to use it
 69 2012-07-26 02:55:48 <gribble> 190804
 70 2012-07-26 02:55:48 <phantomcircuit> ;;bc,blocks
 71 2012-07-26 02:59:00 <MC-Eeepc> zou did it
 72 2012-07-26 02:59:16 <MC-Eeepc> https://bitcointalk.org/index.php?topic=95738.0 shit just got real
 73 2012-07-26 03:00:12 <Bitxk> one again?
 74 2012-07-26 03:00:18 <Bitxk> once again*
 75 2012-07-26 03:00:32 <Bitxk> mtgox get hacked continously
 76 2012-07-26 03:01:27 <MC-Eeepc> no zou reused an old email
 77 2012-07-26 03:02:56 <Joric> Erik Voorhees? His brother must be Jason.
 78 2012-07-26 03:05:51 <gmaxwell> This is OT for this channel. move it it #bitcoin or elsewhere please.
 79 2012-07-26 03:07:24 <MC-Eeepc> ok dad
 80 2012-07-26 03:48:31 <jgarzik> just run a query over the entire blockchain, looking at script opcode use
 81 2012-07-26 03:48:47 <jgarzik> I counted the number of times each opcode was used, in total: https://gist.github.com/3180470
 82 2012-07-26 03:50:35 <jgarzik> p2sh -probably- not included
 83 2012-07-26 03:52:35 <gmaxwell> jgarzik: can you do the same for testnet3
 84 2012-07-26 03:52:46 <jgarzik> gmaxwell: sure, one sec
 85 2012-07-26 03:52:55 <gmaxwell> (I want to see which ones are missing tests)
 86 2012-07-26 03:54:27 <jgarzik> gmaxwell: https://gist.github.com/3180486
 87 2012-07-26 03:56:04 <jgarzik> gmaxwell: you're going to make my life harder, by injecting tests into mainnet, aren't you?  ;-)
 88 2012-07-26 03:56:30 <gmaxwell> No. no. Thats what testnet3 is for.
 89 2012-07-26 03:56:52 <gmaxwell> I mean I wanted to see if there were any missing in testnet3.
 90 2012-07-26 03:56:58 <jgarzik> gotcha :)
 91 2012-07-26 03:59:25 <jgarzik> on mainnet, there are exactly 182 instances of each of these opcodes: OP_2OVER, OP_2SWAP, OP_DEPTH, OP_IF, OP_IFDUP, OP_VERIFY
 92 2012-07-26 03:59:41 <jgarzik> a new transaction pattern?
 93 2012-07-26 04:07:18 <gmaxwell> jgarzik: it's the word "SCRIPT" or something like that.
 94 2012-07-26 04:07:44 <gmaxwell> That was a p2pool bug.
 95 2012-07-26 04:09:52 <gmaxwell> sorry, it's lower case "script"
 96 2012-07-26 04:11:57 <jgarzik> BN_bn2mpi(this, &vch[0])
 97 2012-07-26 04:12:08 <jgarzik> anybody know where MPI format might be found?  my google fu is weak
 98 2012-07-26 04:12:36 <jgarzik> maybe its something simple like 32-bit big endian ints, cat'd together?
 99 2012-07-26 04:15:48 <theymos> Described here: http://www.openssl.org/docs/crypto/BN_bn2bin.html
100 2012-07-26 04:16:06 <copumpkin> "the representation of a"
101 2012-07-26 04:16:39 <copumpkin> oh
102 2012-07-26 04:16:44 <copumpkin> I see, right above that
103 2012-07-26 04:19:43 <leotreasure> hello,
104 2012-07-26 04:20:03 <leotreasure> i really need some help with Abe i think something is going horribly wrong
105 2012-07-26 04:20:33 <leotreasure> can anyone tell me how I can identify the progress of Abe syncing with bitcoind?
106 2012-07-26 04:20:44 <gmaxwell> how the @#$#@$#@$@# do I get a decode of a non-wallet transaction now?
107 2012-07-26 04:21:15 <leotreasure> specifically i don't know if this is a transaction hash, # or what ---> block_tx 185378 4317408
108 2012-07-26 04:22:05 <leotreasure> the syncing process has taken over three days and the space on the vps is using 14gigs
109 2012-07-26 04:23:36 <gmaxwell> sounds like abe sucks.
110 2012-07-26 04:24:57 <copumpkin> gmaxwell: what?
111 2012-07-26 04:26:02 <jgarzik> meh.  i already require openssl for ecdsa, might as well use their bignums too
112 2012-07-26 04:32:05 <leotreasure> gmaxwell: do you know what format (hash/tx #&) this block_tx number represents?
113 2012-07-26 04:50:00 <leotreasure> gmaxwell: ah nm i found it - format is block number, transaction number
114 2012-07-26 05:31:04 <luke-jr> jgarzik: sorry if it's obvious and I'm just too tired to realize it, but& what's the purpose in counting opcodes?
115 2012-07-26 05:42:03 <weex> in testing bitcoind seems to parse <amount> if given in exponential notation
116 2012-07-26 05:42:21 <weex> it even rounds a sub-satoshi properly up to one
117 2012-07-26 05:42:47 <weex> is this behavior something that can be counted on longterm?
118 2012-07-26 05:49:42 <midnightmagic> is there some sneaky way of translating from my .bitcoin/debug.log reference of orphan block 7013fe34aa2de9f1 to that orphaned block's full hash?
119 2012-07-26 05:50:33 <theymos> Hexdump blkxxxx.dat and search for that hex, maybe? (I haven't tried this.)
120 2012-07-26 05:50:55 <midnightmagic> hrm..
121 2012-07-26 05:52:19 <theymos> -printblock=7013fe34aa2de9f1 would print details, but not the full hash IIRC.
122 2012-07-26 05:52:33 <midnightmagic> ah good idea
123 2012-07-26 05:54:59 <midnightmagic> doh segfault
124 2012-07-26 07:10:11 <Enas> why silent?
125 2012-07-26 08:57:39 <Anoosa> hello I am trying to build bitcoin on windows 7 and get this error : cannot find -lboost_filesystem-mgw44-mt-s-1_49
126 2012-07-26 08:57:48 <Anoosa> I am using Qt Creator
127 2012-07-26 08:58:32 <Anoosa> and downloaded boost library and complied it
128 2012-07-26 09:00:04 <Anoosa> I am building bitcoin under windows 7
129 2012-07-26 09:00:15 <Anoosa> I got this error : :-1: error: cannot find -lboost_filesystem-mgw44-mt-s-1_49
130 2012-07-26 09:00:25 <ersi> Yeah thanks, we saw your message three minutes ago.
131 2012-07-26 09:00:30 <Anoosa> I downloaded boost library and compiled it
132 2012-07-26 09:01:17 <ersi> Well, it's probably not in the right place or you havn't told the compiler where the boost library is located.
133 2012-07-26 09:01:27 <Anoosa> and got the lib file but it has this name : :libboost_system-mgw47-mt-s-1_43.a
134 2012-07-26 09:01:46 <Anoosa> where is  the right place?
135 2012-07-26 09:02:02 <ersi> I have no idea. It's your system. :-P
136 2012-07-26 09:02:08 <Anoosa> qtsdkmingwlib?
137 2012-07-26 09:04:17 <Anoosa> I need real help
138 2012-07-26 09:06:22 <ersi> I'd recommend either sticking around, plenty of people have what you wrote in their IRC history and might be able to help - or you could try finding a guide/tutorial/README that'd assist you - or sum everything up about your environment and do a comprehensive forum post
139 2012-07-26 09:13:00 <Anoosa> ok
140 2012-07-26 09:23:11 <Diablo-D3> gmaxwell: I feel funny
141 2012-07-26 09:23:27 <Diablo-D3> I just said, out loud, why the fuck are function calls so slow
142 2012-07-26 09:23:56 <Diablo-D3> I feel weird.
143 2012-07-26 10:19:14 <Anoosa> https://github.com/bitcoin/bitcoin/issues/1401
144 2012-07-26 10:21:06 <Anoosa> I did the same as they mentioned in the link
145 2012-07-26 10:21:15 <Anoosa> but I got an error
146 2012-07-26 10:22:08 <edcba> so ?
147 2012-07-26 10:22:24 <Anoosa> http://penpaste.com/6kyv9f4
148 2012-07-26 10:22:47 <Anoosa> what to do?
149 2012-07-26 10:23:16 <edcba> wsock32 lib .
150 2012-07-26 10:23:17 <edcba> ?
151 2012-07-26 10:30:04 <Anoosa> what do u mean?
152 2012-07-26 10:31:42 <Anoosa> should I add it this lib?
153 2012-07-26 10:31:52 <Anoosa> from where to get it?
154 2012-07-26 10:35:49 <Anoosa> !!!
155 2012-07-26 10:35:50 <gribble> Error: "!!" is not a valid command.
156 2012-07-26 10:36:05 <Anoosa> ?
157 2012-07-26 10:38:51 <edcba> !*
158 2012-07-26 10:38:52 <gribble> Error: "*" is not a valid command.
159 2012-07-26 10:46:08 <Anoosa> :) I got u now u meant to add -lwsock32?
160 2012-07-26 10:46:32 <Anoosa> but till now its building widh it will solve the problem
161 2012-07-26 10:48:27 <Anoosa> it worked .....
162 2012-07-26 10:48:33 <Anoosa> yaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaay
163 2012-07-26 10:48:38 <Anoosa> horrrrrrrray
164 2012-07-26 10:49:05 <Anoosa> but did not execute .. :((
165 2012-07-26 10:49:23 <Anoosa> it gives me this mesage mingw10.dll is missing
166 2012-07-26 11:39:54 <Anoosa> ??
167 2012-07-26 12:34:15 <lianj> jgarzik: OP_NOP2 15 is more likely OP_CHECKHASHVERIFY
168 2012-07-26 13:46:20 <luke-jr> weex: JSON defines the format of Number to include exponent notation, and bitcoind has defined rounding as a protocol rule to help out float-based clients
169 2012-07-26 13:49:57 <BlueMatt> <luke-jr> BlueMatt: did you find a way to get debug info in Win32 builds? --> I never looked into that, the only thing I know about debug symbols on win32 is, if we dont strip, we get gcc debug symbols that are entirely useless unless you get gdb for windows, which essentially means installing msys, which is quite a package just to debug...
170 2012-07-26 13:53:55 <BlueMatt> s/quite a package to debug/quite a package just to debug bitcoin/
171 2012-07-26 14:55:25 <BlueMatt> anyone know off-hand the total number of addresses used in the chain or a recent number?
172 2012-07-26 14:56:12 <copumpkin> used?
173 2012-07-26 14:56:29 <copumpkin> distinct inputs?
174 2012-07-26 14:56:31 <copumpkin> or what?
175 2012-07-26 14:57:16 <luke-jr> distinct outputs, I imagine
176 2012-07-26 14:57:24 <BlueMatt> distinct public keys
177 2012-07-26 14:57:27 <BlueMatt> or addresses
178 2012-07-26 14:57:41 <copumpkin> hmm, my code doesn't speak addresses yet, otherwise I could just run a quick query over it :/
179 2012-07-26 14:58:48 <riush> i got about 4682712 addresses (pubkey, hash160, multisig, p2sh) around block 187100
180 2012-07-26 14:59:06 <BlueMatt> thanks
181 2012-07-26 14:59:28 <sipa> how many address with nonzero unspent outputs?
182 2012-07-26 15:01:40 <sipa> +es
183 2012-07-26 15:09:02 <copumpkin> sipa: too complicated, didn't test!
184 2012-07-26 15:09:05 <copumpkin> tc;dt?
185 2012-07-26 15:10:09 <BlueMatt> sipa: I may do that later, it would be a nice api to have unspent outputs in bitcoinj indexed by address/pubkey/target, so I may do that later
186 2012-07-26 15:12:02 <BlueMatt> anyone have an opinion on suggested false-positive rates for bloom filters?
187 2012-07-26 15:12:14 <riush> hm yea that's gonna take a while :)
188 2012-07-26 15:12:38 <BlueMatt> ?
189 2012-07-26 15:13:06 <riush> querying for all unspent outputs instead of only number of used addresses
190 2012-07-26 15:16:41 <copumpkin> BlueMatt: as small as posislbe within your memory constraints :P
191 2012-07-26 15:17:26 <BlueMatt> real question: how many false-positive addresses would be considered "anonymous" ie 5k addresses which are in the chain which are fp, is that "anonymous"?
192 2012-07-26 15:18:32 <copumpkin> oh, so the goal is to have high false positives
193 2012-07-26 15:18:39 <copumpkin> have you come across k-anonymity?
194 2012-07-26 15:18:45 <copumpkin> seems like what you're trying to capture there
195 2012-07-26 15:18:53 <copumpkin> http://epic.org/privacy/reidentification/Sweeney_Article.pdf
196 2012-07-26 15:20:02 <copumpkin> seems hard to quantify address anonymity that way though
197 2012-07-26 15:21:30 <BlueMatt> yea...
198 2012-07-26 15:21:52 <copumpkin> meaning that false positive rate in your bloom filter only tells part of the picture
199 2012-07-26 15:22:04 <copumpkin> but I dunno, depends what you're doing with it I guess :)
200 2012-07-26 15:22:35 <copumpkin> everything else being equal, I'd probably feel safe with singificantly fewer than 5k people confused with me
201 2012-07-26 15:23:04 <BlueMatt> yep, it really depends on the use-case, but you still have to write docs with suggested values...
202 2012-07-26 15:23:14 <BlueMatt> "give me a %" is kinda poor docs
203 2012-07-26 15:23:36 <copumpkin> what is this for?
204 2012-07-26 15:23:38 <copumpkin> I missed the context
205 2012-07-26 15:23:41 <BlueMatt> bitcoinj
206 2012-07-26 15:23:44 <BlueMatt> just the docs for it
207 2012-07-26 15:24:09 <sipa> i wonder if a filter of the form "match all addresses that are bip30 derived from ..." is useful
208 2012-07-26 15:24:17 <BlueMatt> (writing bloom filter which filters transactions sent to a client to reduce bw and derserialization costs)
209 2012-07-26 15:24:25 <sipa> for those who do not want anonymity
210 2012-07-26 15:24:50 <sipa> but also don't know how many addresses to precalculate
211 2012-07-26 15:25:18 <BlueMatt> bip30?
212 2012-07-26 15:25:33 <BlueMatt> I dont follow...
213 2012-07-26 15:25:38 <sipa> eh
214 2012-07-26 15:25:40 <sipa> bip32
215 2012-07-26 15:25:56 <BlueMatt> ahh
216 2012-07-26 15:26:24 <BlueMatt> I thought you could derive infinitely many addresses from bip32?
217 2012-07-26 15:26:30 <BlueMatt> would need a n param
218 2012-07-26 15:26:37 <sipa> no
219 2012-07-26 15:26:44 <sipa> that's the point
220 2012-07-26 15:27:11 <sipa> if the client does not know how many of the addresses are used already
221 2012-07-26 15:27:41 <BlueMatt> given an arbitrary address, matching a bip32 base would require a lot of derivation work, no?
222 2012-07-26 15:27:42 <sipa> so the server can automatically add an address to the filter if it encounters an earlier one already used
223 2012-07-26 15:28:02 <sipa> oh, you need some lookahead parameter, yes
224 2012-07-26 15:28:45 <BlueMatt> mmm...Im not a big fan of making nodes do that much work for a remote node downloading the chain
225 2012-07-26 15:28:51 <BlueMatt> seems like waaay too easy a dos target
226 2012-07-26 15:28:52 <sipa> like address >N+k shouldn't be looked for until N is used
227 2012-07-26 15:28:57 <BlueMatt> yea
228 2012-07-26 15:29:26 <sipa> it's much lighter than bloom filter matching with continuous updating
229 2012-07-26 15:29:50 <sipa> anyway, gotta go visit Reyljavik
230 2012-07-26 15:29:55 <sipa> Reykjavik
231 2012-07-26 15:33:41 <jgarzik> MagicalTux, on the forum:  "As of today, we have received no contact from any involved party stating a police investigation is ongoing, nor any contact from any law enforcement agency regarding anything related to Bitcoinica."
232 2012-07-26 15:34:28 <jgarzik> so glad I had zero to do with Bitcoinica...  and so glad we have MT and Charlie on bitcoin's side
233 2012-07-26 15:39:28 <lianj> so much to read. i just want to know how genjix explains him leaking the source (only after the hack happened)
234 2012-07-26 15:40:20 <luke-jr> lianj: is there real evidence of that?
235 2012-07-26 15:41:15 <lianj> yes
236 2012-07-26 15:41:41 <luke-jr> just because the leaked git repo was cloned by genjix != real evidence genjix leaked it
237 2012-07-26 15:41:58 <gavinandresen> jgarzik gmaxwell TD and anybody else who might be interested in transaction fees: can I get some code review / ACKs on https://github.com/bitcoin/bitcoin/pull/1590
238 2012-07-26 15:42:31 <jgarzik> could have sworn I already ack'd that, but apparently not.  will re-review
239 2012-07-26 15:43:06 <lianj> luke-jr: cloned and packed. maybe not real evidence but worth explaining, no?
240 2012-07-26 15:43:24 <luke-jr> lianj: many people use git clones on their servers
241 2012-07-26 15:43:41 <luke-jr> I think it most likely the attacker just tar'd up one of those
242 2012-07-26 15:43:56 <luke-jr> the fact that the files were owned by genjix is not surprising
243 2012-07-26 15:44:49 <jgarzik> gavinandresen: what is the use case for -blockminsize ?
244 2012-07-26 15:45:09 <jgarzik> gavinandresen: I understand -blockmaxsize and -blockprioritysize, those look OK
245 2012-07-26 15:45:39 <gavinandresen> jgarzik: luke suggested it; it is for the "I want to sort by fee, but include free transactions if there is room at the end of the block"
246 2012-07-26 15:45:47 <lianj> luke-jr: then he could still state that his box (one that he has an account on) was hacked. also besides one reddit post, his forum post was the first that made the link truely public
247 2012-07-26 15:46:27 <gavinandresen> jgarzik: I suggested some possible fee policies here: https://bitcointalk.org/index.php?topic=95837.0
248 2012-07-26 15:47:47 <jgarzik> gavinandresen: thanks, that post helps understand the context
249 2012-07-26 15:48:01 <lianj> luke-jr: if the hacker tar'd it he would prolly be root. or got it (non-root) with genjix's account
250 2012-07-26 15:49:50 <luke-jr> gavinandresen: perhaps CheckNewBlock should be merged first?
251 2012-07-26 15:51:17 <luke-jr> (unit tests)
252 2012-07-26 15:51:21 <gavinandresen> luke-jr: I'll re-test 1246 (I was ready to ACK until it didn't compile with my gcc last time around)
253 2012-07-26 15:55:08 <jgarzik> lianj: OP_CHECKHASHVERIFY?  you mean luke-jr's proposal?
254 2012-07-26 15:55:19 <lianj> yes
255 2012-07-26 15:56:09 <luke-jr> jgarzik: yes, gavinandresen was running a trollbot on testnet, so I had to test it on mainnet
256 2012-07-26 15:56:26 <jgarzik> trollbot/
257 2012-07-26 15:56:27 <jgarzik> ?
258 2012-07-26 15:56:51 <luke-jr> jgarzik: it automatically stole CHV transactions in the context of non-CHV clients/miners
259 2012-07-26 15:57:10 <luke-jr> (like BBE)
260 2012-07-26 15:59:55 <lianj> have OP_NOP2 aliased to OP_CHECKHASHVERIFY in my code
261 2012-07-26 16:05:29 <riush> sipa: 2127798 addresses with unspent txouts at block 187299
262 2012-07-26 16:09:07 <sipa> eh?
263 2012-07-26 16:09:39 <sipa> i cery recently counted only 1.5M unspent txouts
264 2012-07-26 16:09:42 <sipa> very
265 2012-07-26 16:11:47 <riush> oh, hmm.. might have messed something up then :p
266 2012-07-26 16:11:55 <midnightmagic> luke-jr: do you run something that auto-spends free bitcoins? someone said not to spend open tx into the blockchain because you autospend them on detection.
267 2012-07-26 16:12:18 <sipa> or SD messed things up very quickly, riush?
268 2012-07-26 16:12:38 <sipa> my data is maybe a month old
269 2012-07-26 16:13:06 <riush> hehe also possible. i'll rescan the chain and check again - could also be that i used different versions of the code during one sync or something :)
270 2012-07-26 16:19:30 <gavinandresen> midnightmagic: if Luke isn't running something like that, it is extremely likely somebody else is.  If your transaction doesn't include a signature check, you're asking to get it stolen.
271 2012-07-26 16:41:11 <luke-jr> midnightmagic: yes
272 2012-07-26 16:41:43 <luke-jr> but that doesn't mean you shouldn't do it
273 2012-07-26 16:44:53 <midnightmagic> gavinandresen: Putting small amounts out there might be an interesting experiment.
274 2012-07-26 16:45:36 <midnightmagic> luke-jr: thanks, just curious. it would be interesting to see a tx autospend race. :)
275 2012-07-26 16:46:33 <luke-jr> midnightmagic: mine is just a patch for Bitcoin-Qt that I run on my desktop PC
276 2012-07-26 16:48:05 <midnightmagic> luke-jr: Oh, you don't automatically do it on a well-connected relay? that's interesting.
277 2012-07-26 16:48:51 <luke-jr> midnightmagic: it's not especially poorly connected
278 2012-07-26 16:49:06 <gavinandresen> luke-jr : miner_tests includes a "won't ever happen" test -- an orphan (missing input) transaction in the memory pool.  It's causing problems with the sort-by-fee code.
279 2012-07-26 16:50:53 <luke-jr> gavinandresen: hmm, better to handle that possibility just in case some bug adds an orphan?
280 2012-07-26 16:51:18 <luke-jr> what if it wasn't an orphan when it got into the mempool, but became one later by a double-spend?
281 2012-07-26 16:51:32 <jgarzik> gavinandresen: actually.....
282 2012-07-26 16:51:35 <gavinandresen> double-spends aren't orphans
283 2012-07-26 16:51:45 <jgarzik> gavinandresen: that can happen during a reorg, IIUC
284 2012-07-26 16:51:52 <jgarzik> gavinandresen: otherwise, you are correct
285 2012-07-26 16:52:11 <luke-jr> gavinandresen: but the transaction that depends on the reversed one is
286 2012-07-26 16:52:29 <jgarzik> gavinandresen: reorg dumps into memory pool, because otherwise, dumping into orphan pool might result in transactions being zapped
287 2012-07-26 16:53:26 <luke-jr> IMO, the possibility of a bug getting an orphan into the mempool is significant enough that it should be tested and handled
288 2012-07-26 16:53:34 <luke-jr> even if they shouldn't
289 2012-07-26 16:55:56 <sipa> any progress towards 0.7, by the way?
290 2012-07-26 16:56:16 <sipa> (w00t hotel with free wifi)
291 2012-07-26 16:57:08 <gavinandresen> The only thing on my "must have for 0.7" list is what I'm working on right now-- the sort-by-fee change for miners.
292 2012-07-26 16:57:27 <gavinandresen> Although what's the status of block propagation concerns?
293 2012-07-26 16:57:48 <gavinandresen> (and Luke's suggested patch for propagate-before-fully-validated)
294 2012-07-26 16:58:32 <gmaxwell> BlueMatt: going to rebase your addnode stuff? it's very helpful when working with the new networking code.
295 2012-07-26 16:58:34 <midnightmagic> Does anyone have orphaned block 000000000000023d4293f434a240fe325933842ecf3112c70b48e1102d5bf216 in their blk*? and if so, can you getrawtransaction on its coinbase tx, 5e70cc8945b7c37f7d176ac26cb2072e1ceaccb9297b36a8cbe19c082dd89b57 ?
296 2012-07-26 16:58:52 <gavinandresen> luke-jr jgarzik : y'all are right, better safe than sorry, I'll make the CreateBlock code do the right thing even if there is an orphan txn in the mempool
297 2012-07-26 16:58:57 <BlueMatt> gmaxwell: just got back from vacation today, Ill look into it later
298 2012-07-26 16:59:02 <gmaxwell> Fking getrawtransaction is not working on non-wallet transactions.
299 2012-07-26 16:59:27 <luke-jr> gavinandresen: I think we need to have a group gettogether to test it
300 2012-07-26 16:59:29 <gavinandresen> gmaxwell: really?  should....
301 2012-07-26 16:59:34 <midnightmagic> gmaxwell: You mean gettransaction?   getrawtransaction seems to work okay for me so far..?
302 2012-07-26 16:59:38 <luke-jr> (fastblockrelay that is)
303 2012-07-26 17:00:10 <gavinandresen> luke-jr: did you test fastblockrelay with invalid blocks?
304 2012-07-26 17:00:16 <gmaxwell> midnightmagic: I have that block, and already tried doing what you're doing.
305 2012-07-26 17:00:20 <gavinandresen> (invalid-but-valid-pow)
306 2012-07-26 17:00:20 <gmaxwell> [bitcoin01@carbide80 ~]$ ~/bitcoin/src/bitcoind -datadir=/tmp/bitcoin01/ getblock 000000000000023d4293f434a240fe325933842ecf3112c70b48e1102d5bf216 | grep -A 1 tx
307 2012-07-26 17:00:21 <midnightmagic> gettransaction was switched back to wallet-only not so long ago i think because Gavin was worried people depended on it being wallet-only (so, past-consistent)
308 2012-07-26 17:00:22 <BlueMatt> re: fastblockrelay, can we do the "blocks are headers + vector<tx hash>" stuff first?
309 2012-07-26 17:00:23 <gmaxwell> "tx" : [
310 2012-07-26 17:00:25 <gmaxwell> "5e70cc8945b7c37f7d176ac26cb2072e1ceaccb9297b36a8cbe19c082dd89b57",
311 2012-07-26 17:00:28 <gmaxwell> [bitcoin01@carbide80 ~]$ ~/bitcoin/src/bitcoind -datadir=/tmp/bitcoin01/ getrawtransaction 5e70cc8945b7c37f7d176ac26cb2072e1ceaccb9297b36a8cbe19c082dd89b57
312 2012-07-26 17:00:31 <gmaxwell> error: {"code":-5,"message":"No information available about transaction"}
313 2012-07-26 17:00:33 <gmaxwell> [bitcoin01@carbide80 ~]$
314 2012-07-26 17:00:34 <luke-jr> gavinandresen: no, I haven't tested it at all - it needs a group IMO
315 2012-07-26 17:01:03 <midnightmagic> gmaxwell: Yeah, but try the next ones, all the ones I tested so far work fine. Like 9fc16d59c11503e89d15f4063048eab1aa9d9631cb1eae0bfdabec6956fd6dd2
316 2012-07-26 17:01:22 <gmaxwell> 0_o
317 2012-07-26 17:01:27 <gmaxwell> it's just that coinbase it doesn't work on?!@
318 2012-07-26 17:01:32 <midnightmagic> gmaxwell: I know you have. I became curious about it specifically because you were complaining about it in #p2pool
319 2012-07-26 17:01:33 <gmaxwell> I see.
320 2012-07-26 17:01:44 <midnightmagic> gmaxwell: Try the coinbase of another orphan.
321 2012-07-26 17:01:55 <sipa> gmaxwell: you're not accidentally running ultraprune code, as that removes the ability to getrawtransaction
322 2012-07-26 17:02:00 <gmaxwell> I had just ass_u_me_d that the api calls dropped non-wallet txn.
323 2012-07-26 17:02:00 <sipa> ?
324 2012-07-26 17:02:07 <gmaxwell> sipa: no, that node isn't ultraprune.
325 2012-07-26 17:02:10 <sipa> ok
326 2012-07-26 17:02:13 <midnightmagic> gmaxwell: I was trying to translate one of the reorg block hash to a real hash so I could getblock on it last night when I had to go sleep..
327 2012-07-26 17:02:21 <gmaxwell> it's git master as of 12e5881c
328 2012-07-26 17:02:45 <luke-jr> gmaxwell: my blk0001 index also breaks getrawtxn
329 2012-07-26 17:02:52 <sipa> haven't follow git master developments the past two weeks
330 2012-07-26 17:02:58 <gmaxwell> luke-jr: right, this node isn't pruned.
331 2012-07-26 17:03:06 <gmaxwell> And this block is also new.
332 2012-07-26 17:03:33 <midnightmagic> lol wth does this mean: "07/26/12 18:51:10 REORGANIZE: Disconnect 0 blocks; 81ea778dcb50081c..81ea778dcb50081c"
333 2012-07-26 17:03:40 <luke-jr> BlueMatt: fastblockrelay probably needs to eliminate the bandwidth impact too
334 2012-07-26 17:03:58 <midnightmagic> cool: InvalidChainFound: invalid block=7307e584d6eaa80d  height=190906
335 2012-07-26 17:04:04 <sipa> midnightmagic: it means you did a reorg without any disconnects
336 2012-07-26 17:04:05 <jgarzik> gavinandresen: CreateNewBlock() always did a final ConnectInputs, so it should always DTRT WRT orphans
337 2012-07-26 17:04:16 <BlueMatt> I just hate the forward-block-before-checking stuff, even if its hard and you mark it as such, it just seems a waste...
338 2012-07-26 17:04:18 <gmaxwell> midnightmagic: I guess we just can't getrawtransaction orphaned coinbases, because they're not in the index.
339 2012-07-26 17:04:24 <gmaxwell> kinda stinks for troubleshooting.
340 2012-07-26 17:04:47 <jgarzik> gmaxwell: really, we need a snapshotting filesystem that stores the entire state for each tree (chain) fork
341 2012-07-26 17:04:49 <sipa> gmaxwell: i intend to add an optional txindex to ultraprune
342 2012-07-26 17:05:03 <sipa> gmaxwell: independent of the connected chain
343 2012-07-26 17:05:27 <midnightmagic> gmaxwell: 7013fe34aa2de9f1 is another orphan but I've been having trouble retrieving the full block hash. does -printblocktree work for you?
344 2012-07-26 17:06:38 <gmaxwell> midnightmagic: I can't getblock a partial hash...
345 2012-07-26 17:07:54 <midnightmagic> gmaxwell: me neither. I thought if i printblocktree I could just grep for it, but since the last time I used it, it's been gummed up and makes no visual sense (and doesn't bring full blockhashes in any event) anymore.
346 2012-07-26 17:08:07 <midnightmagic> s/bring/print/
347 2012-07-26 17:08:29 <jgarzik> for pynode, I'm liking a dual-tx index approach:
348 2012-07-26 17:08:37 <jgarzik> one full index with all id's, and one with just unspent stuff
349 2012-07-26 17:08:45 <midnightmagic> so.. as far as I can tell, there's no way to use normal bitcoind calls to go from the stuff in debug.log to real data.
350 2012-07-26 17:08:55 <jgarzik> the unspent-index is what gets hit hard
351 2012-07-26 17:09:18 <midnightmagic> gavinandresen: Should I open an issue?
352 2012-07-26 17:09:20 <jgarzik> disk space is not a concern, only working set size
353 2012-07-26 17:09:50 <gmaxwell> midnightmagic: see the discussion on https://github.com/bitcoin/bitcoin/pull/1426
354 2012-07-26 17:09:55 <midnightmagic> ok
355 2012-07-26 17:12:57 <gmaxwell> gavinandresen: did you and jeff come to any meeting of the minds on testnet protocol version stuff?
356 2012-07-26 17:13:26 <midnightmagic> gmaxwell: ???_???
357 2012-07-26 17:17:13 <gmaxwell> midnightmagic: yea, it seems obvious enough to me. :) But also not worth arguing over.
358 2012-07-26 17:17:38 <midnightmagic> gmaxwell: So was there some outcome from that that I'm just not aware of?   bah. one more patch i have to maintain myself. :(
359 2012-07-26 17:18:06 <gmaxwell> Well, I'm the only person commenting on that pull request saying that it should log the whole hash.
360 2012-07-26 17:18:50 <gmaxwell> And I'm unwilling to bludgeon people over the parameters to a blinking subtring operation.
361 2012-07-26 17:22:54 <midnightmagic> #define BLOCKSUBSTR 48
362 2012-07-26 17:24:08 <gmaxwell> Yea, though this won't help me much when I'm trying to help some poor sucker who's sent me a debug.log.
363 2012-07-26 17:28:18 <Icoin> hi, can someone tell me why my bitcoind does have like 10 connections but dnt get any blocks ? http://pastebin.com/gnGZJWax
364 2012-07-26 17:44:39 <gmaxwell> Icoin: how long has it been running?
365 2012-07-26 17:45:03 <Icoin> ok its running now
366 2012-07-26 17:45:18 <Icoin> i deleted everything and recompiled
367 2012-07-26 17:45:24 <Icoin> thanks
368 2012-07-26 17:45:45 <Icoin> gmaxwell it was running like 10 mins
369 2012-07-26 17:59:12 <jgarzik> This is the 'bitcoin' sub-directory of pynode: https://github.com/jgarzik/pynode/tree/master/bitcoin
370 2012-07-26 17:59:29 <jgarzik> hopefully that becomes a "python-bitcoin" package, perhaps reuseable by others
371 2012-07-26 18:00:04 <jgarzik> luke-jr: I'd be interested in a "what needs to change for python3" review, as I've little p3 knowledge
372 2012-07-26 18:00:38 <jgarzik> well, I'm of very little python skill in general... :) ... but even less with python3
373 2012-07-26 18:01:00 <jgarzik> general python code suggestions welcome too
374 2012-07-26 18:06:00 <helo> i'm glad you released that before i started trying to get caesure up to speed
375 2012-07-26 18:08:50 <helo> would pyopencl speed up pynode syncing??
376 2012-07-26 18:10:23 <gmaxwell> Like a rocket motor helps a cow chew its cud.
377 2012-07-26 18:11:42 <helo> was afraid of that :)
378 2012-07-26 18:15:09 <luke-jr> jgarzik: Eloipool's "bitcoin" subdir would fit well in a python-bitcoin pkg
379 2012-07-26 18:15:24 <jgarzik> note the massive caveats in 'TODO'...  it will sync a blockchain, but script verf and chain reorg (major, needed features for security) remain incomplete
380 2012-07-26 18:15:36 <jgarzik> still a dev-only release
381 2012-07-26 18:22:09 <luke-jr> bitcoind is using ~/.bitcoin/testnet/ again? O.o
382 2012-07-26 18:24:23 <gmaxwell> 0_o
383 2012-07-26 18:25:08 <gmaxwell> util.cpp:        path /= "testnet3";
384 2012-07-26 18:27:47 <jgarzik> luke-jr: just looked at Eloipool's bitcoin subdir...  I think I cover most of that already
385 2012-07-26 18:28:06 <luke-jr> jgarzik: ah, ok
386 2012-07-26 18:28:13 <jgarzik> luke-jr: TX assembling and disassembling code is absent perhaps, but most other stuff is covered
387 2012-07-26 18:29:01 <luke-jr> jgarzik: well, an important part of designing it was to not do anything unnecessary, since Eloipool is usually doing txn stuff in the busy sections
388 2012-07-26 18:29:23 <gribble> New news from bitcoinrss: gavinandresen opened pull request 1633 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1633>
389 2012-07-26 18:29:26 <luke-jr> so maybe not the best fit anyway
390 2012-07-26 18:30:20 <luke-jr> anyhow, something is broken with my testnet - it's trying to use /testnet/ (not /testnet3/) and dies as soon as I getwork it (not today-new - I'm testing with an old master)
391 2012-07-26 18:30:31 <luke-jr> but I gotta run, will debug more later
392 2012-07-26 18:31:01 <BlueMatt> why are we still spending time on testnet3 network connections? we didnt bother to solve this problem on testnet2, and it seems like a waste of effort, no?
393 2012-07-26 18:39:08 <jgarzik> BlueMatt: given that it makes testnet useless, answering that question in the negative sort of implies giving up on testnet completely
394 2012-07-26 18:39:35 <BlueMatt> I was under the impression we use #testnet3 on irc for testnet3 seeding again?
395 2012-07-26 18:39:46 <gavinandresen> yes
396 2012-07-26 18:40:07 <BlueMatt> worst case `alias bitcoind=rm ~/.bitcoin/peers.dat; bitcoind`
397 2012-07-26 18:40:59 <gavinandresen> ~/.bitcoin/testnet3/peers.dat
398 2012-07-26 18:41:03 <BlueMatt> sorry
399 2012-07-26 18:41:21 <gavinandresen> (I just did that to test the patch....)
400 2012-07-26 18:45:25 <jgarzik> BlueMatt: (re IRC)  yes, but then we start to connect, and download addresses from testnet 1/2 clients
401 2012-07-26 18:45:34 <jgarzik> BlueMatt: ...and then start connecting to them all, etc...
402 2012-07-26 18:45:56 <jgarzik> spreading their addresses, even if you're testnet3 and they're not
403 2012-07-26 18:46:31 <BlueMatt> so we patch testnet3 stuff and debate about it, so that we can add a patch which can then be removed when 0.7 is released?
404 2012-07-26 18:48:28 <gavinandresen> huh?  the patch will stay
405 2012-07-26 18:49:13 <BlueMatt> s/then be removed/then be useless and could have been skipped at that point/
406 2012-07-26 18:49:41 <gavinandresen> we have lots of code that is "should be removed at some point"
407 2012-07-26 18:49:49 <gmaxwell> BlueMatt: what should be removed?
408 2012-07-26 18:49:52 <BlueMatt> so lets add more?
409 2012-07-26 18:49:52 <gavinandresen> ... and we eventually remove it
410 2012-07-26 18:50:05 <gmaxwell> oh the testnet protocol version workaround? "meh"
411 2012-07-26 18:50:20 <jgarzik> gavinandresen: commit e0e54740b10e28dac2fe7aa9b461e956452e9649 is confusing
412 2012-07-26 18:50:21 <gmaxwell> Testnet usability is quite poor now, something must be done about it.
413 2012-07-26 18:50:36 <BlueMatt> release 0.7, done?
414 2012-07-26 18:50:45 <BlueMatt> why bother fixing it on master?
415 2012-07-26 18:50:53 <gmaxwell> BlueMatt: No, testnet will be unusable in 0.7 release too.
416 2012-07-26 18:50:58 <jgarzik> gavinandresen: #1, comment is incorrect.  It is known that orphans in mempool may happen during reorg.  TX mempool code also knows this.
417 2012-07-26 18:51:11 <jgarzik> gavinandresen: #2, we call ConnectInputs, so the code appears redundant?
418 2012-07-26 18:51:13 <BlueMatt> gmaxwell: after a reasonable upgrade time, it would be perfectly useable, no?
419 2012-07-26 18:51:18 <gmaxwell> BlueMatt: so long as there are a lot of older nodes out there it is _very_ likely that all 8 of your outbound slots will be to useless nodes.
420 2012-07-26 18:51:44 <gmaxwell> BlueMatt: this hasn't turned out to be true for the other testnet rule change.
421 2012-07-26 18:51:58 <BlueMatt> are the testnet node version counts that old?
422 2012-07-26 18:52:18 <gmaxwell> People are a little less agressive in upgrading testnet, I think...
423 2012-07-26 18:52:33 <BlueMatt> oh, I would have guessed more
424 2012-07-26 18:52:46 <jgarzik> I'm surprised at how many testnet nodes exist, at all
425 2012-07-26 18:52:47 <gavinandresen> jgarzik: I suspect we're talking about two different kinds of 'orphans'
426 2012-07-26 18:53:35 <jgarzik> gavinandresen: you added a check for missing inputs...  ConnectInputs does that too
427 2012-07-26 18:54:06 <gavinandresen> Right. And ConnectInputs is always called before adding to the memory pool
428 2012-07-26 18:54:29 <jgarzik> gavinandresen: Right.  And ConnectInputs is called _again_ when building a block, on purpose, for reasons like this.
429 2012-07-26 18:54:33 <gavinandresen> ... therefore, the "This should never happen" comment is correct.  I should have used a different word for the commit message "missing parent
430 2012-07-26 18:54:41 <jgarzik> gavinandresen: look down in the same function you modified...
431 2012-07-26 18:55:21 <gavinandresen> Right.  Luke added a test case that calls mempool.AddUnchecked with a transaction that has a missing parent
432 2012-07-26 18:55:52 <gavinandresen> ... which messed up the transaction-fee-calculation code and prompted that commit
433 2012-07-26 18:56:18 <jgarzik> gavinandresen: ok?  Even without your change, that means ConnectInputs will fail, and that TX will not get included
434 2012-07-26 18:56:45 <gmaxwell> sipa: did you ever finish that peer rotation code? It would be helpful for the problems testnet3 has.
435 2012-07-26 18:57:37 <gavinandresen> jgarzik: I see, you're suggesting the code could be simpler....
436 2012-07-26 18:59:26 <gavinandresen> jgarzik: there's a difference between missing a parent (which should never happen) and having a parent that got re-organized away from you, though (which would cause ConnectInputs to fail)
437 2012-07-26 19:11:03 <gmaxwell> gavinandresen: " What's the migration plan"  just break it. I'll fix my node to skip the checks, and feed the new nodes.
438 2012-07-26 19:11:10 <gmaxwell> There are only 9 nodes in #bitcoinTEST3.
439 2012-07-26 19:11:33 <gavinandresen> ok, then go for it
440 2012-07-26 19:11:50 <gmaxwell> K.
441 2012-07-26 19:12:18 <gavinandresen> just make sure eleven is one of the new magic bytes and I'll be happy.
442 2012-07-26 19:13:22 <gavinandresen> actually, make them 11 and 0x11 and 011 and 111 and I'll be really happy
443 2012-07-26 19:31:49 <gmaxwell> weird.
444 2012-07-26 19:32:16 <gmaxwell> I have three connections from freaky nodes that don't care about the protocol version.
445 2012-07-26 19:32:43 <gmaxwell> oh I guess they're just connections that aren't up yet.
446 2012-07-26 19:37:16 <gmaxwell> gavinandresen: https://github.com/bitcoin/bitcoin/pull/1634 ?  I've got two nodes running that and successfully migrated the blockchain.
447 2012-07-26 19:37:35 <gavinandresen> gmaxwell: cool.  IP address that I can connect to?
448 2012-07-26 19:40:20 <gribble> New news from bitcoinrss: gmaxwell opened pull request 1634 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1634>
449 2012-07-26 19:41:43 <MC-Eeepc> i cant believe the txn pol is about 3000 now
450 2012-07-26 19:43:29 <D34TH> wait a minute, 11 fetish?
451 2012-07-26 19:43:34 <D34TH> meh
452 2012-07-26 19:43:39 <D34TH> who am i to judge
453 2012-07-26 19:48:17 <jgarzik> gavinandresen gmaxwell: pchMessageStart ACK.
454 2012-07-26 19:48:37 <jgarzik> gavinandresen gmaxwell: change IRC to #testnet4, or leave as is?
455 2012-07-26 19:48:53 <gmaxwell> with only 9 nodes in #testnet3 it hardly matters.
456 2012-07-26 19:48:56 <jgarzik> ok
457 2012-07-26 19:49:08 <gmaxwell> it happily tries connecting and fails.
458 2012-07-26 19:49:22 <gavinandresen> ... over and over and over again....
459 2012-07-26 19:49:43 <gmaxwell> The retries back off though.
460 2012-07-26 19:56:01 <gavinandresen> testing to see what happens if I don't erase testnet3/blk* ....
461 2012-07-26 19:57:40 <gmaxwell> I tested not deleting the peers.dat, and it rejected it and replaced it. But indeed, I didn't test the blk/index.
462 2012-07-26 19:58:12 <gavinandresen> Starting with 500-something blocks, it's happily appending to the old blk0001.dat
463 2012-07-26 19:58:33 <gavinandresen> blkindex.dat doesn't have magic numbers in it
464 2012-07-26 19:59:26 <gavinandresen> ... so it might Just Work.  I suspect bad things might happen if it tried to validate a transaction spending from one of those first 500 blocks, though
465 2012-07-26 19:59:28 <jgarzik> should have a testnet3 v2 node up in a few minutes, with public incoming
466 2012-07-26 20:05:32 <gavinandresen> so far so good, I spent block1's 50BTC and my mixed testnet node accepted it into it's wallet.
467 2012-07-26 20:15:03 <jgarzik> apropos of nothing...  figuring out how to properly index multiple chains sure is a headache.  I definitely understand why Satoshi bailed, and decided to only index the main chain.
468 2012-07-26 20:16:31 <gmaxwell> I wonder if anything will blow up if a fewer-blocks chain has higher sum difficulty?
469 2012-07-26 20:17:24 <jgarzik> I don't think so
470 2012-07-26 20:18:25 <gmaxwell> well, obviously but I don't think I've ever tested that. Should be easy to do on testnet.
471 2012-07-26 20:19:47 <jgarzik> once a chain is long enough, you're not looking at connecting to confirmed (lower height from fork point in question) TXs, but TXs inside weaker-chain blocks connecting to other TXs inside weaker-chain blocks.
472 2012-07-26 20:31:43 <jgarzik> current testnet3 v2 height?
473 2012-07-26 20:32:17 <gmaxwell> "blocks" : 8716,
474 2012-07-26 20:32:41 <gmaxwell> it syncs really slow around 4000 some because I mined some jumbo blocks there.
475 2012-07-26 20:33:07 <jgarzik> heh, I was wondering what was going on there
476 2012-07-26 20:33:29 <gmaxwell> there is a block with something like 4000 transactions in it.
477 2012-07-26 20:33:40 <jgarzik> ok, node up at us4.exmulti.net:18333
478 2012-07-26 20:36:57 <gavinandresen> I'm 97.107.133.172:18311
479 2012-07-26 20:37:57 <gmaxwell> I'm up at 6hgmaxwellgpv2oe.onion:18333
480 2012-07-26 20:38:53 <gavinandresen> I'll switch the testnet faucet over tomorrow... time for dinner now.
481 2012-07-26 20:41:01 <jgarzik> gmaxwell: cute, there is a vanitygen for onion addresses?  :)
482 2012-07-26 20:43:05 <jgarzik> connection timeout
483 2012-07-26 20:43:05 <jgarzik> trying connection 31.221.2.51:18333 lastseen=1.1hrs
484 2012-07-26 20:43:06 <jgarzik> connection timeout
485 2012-07-26 20:43:10 <jgarzik> trying connection 31.221.2.51:18333 lastseen=1.1hrs
486 2012-07-26 20:43:18 <jgarzik> Sure does like to try the same addresses over and over
487 2012-07-26 20:43:35 <jgarzik> probably hitting the same empty addrman bucket over and over
488 2012-07-26 20:45:07 <jgarzik> seems like somebody is mining with far more :)
489 2012-07-26 20:47:52 <Matt_von_Mises> Can someone tell me the purpose of this line when looking at address times? https://github.com/bitcoin/bitcoin/blob/master/src/main.cpp#L2497
490 2012-07-26 20:49:13 <jgarzik> Matt_von_Mises: CAddress is a versioned binary structure, and only recently gained 'nTime'
491 2012-07-26 20:50:12 <gmaxwell> jgarzik: yes.
492 2012-07-26 20:51:02 <gmaxwell> jgarzik: it only knows the couple on IRC.
493 2012-07-26 20:52:09 <gmaxwell> jgarzik: testnet3 isn't mined much, but previously someone revved the difficulty up to 16 in a "I'll get all to testnetcoins effort", and I told them if they didn't stop I'd just rewrite all their work. :)
494 2012-07-26 20:52:39 <jgarzik> gmaxwell: I figure 1 cpu core will keep it going, if everybody else drops out
495 2012-07-26 20:53:00 <gmaxwell> I should unlazy and put in a proper patch to only start mining at N minutes.
496 2012-07-26 20:54:59 <Matt_von_Mises> jgarzik: Yes but why does it set the times in these strange ways?
497 2012-07-26 20:59:24 <Matt_von_Mises> Also where does the IsReachable function come from and what does it do?
498 2012-07-26 20:59:56 <sipa> Matt_von_Mises: for example, you cannot reach an ipv6 address from an ipv4 one
499 2012-07-26 21:00:14 <sipa> it's part of the logic that decides which public address to advertize
500 2012-07-26 21:01:37 <sipa> gmaxwell: no, i didn't finish peer rotation
501 2012-07-26 21:02:37 <gmaxwell> sipa: What do you think of having hidden service seednodes that work by just explicitly prepopulating the database with a few static HS nodes?
502 2012-07-26 21:03:08 <gmaxwell> holy crap. why is this wallet.dat 481MBytes 0_o
503 2012-07-26 21:03:23 <edcba> because there is a lot of bitcoins in it ? :)
504 2012-07-26 21:03:25 <sipa> so, you connect, and they feed you some addr packets?
505 2012-07-26 21:03:36 <gmaxwell> sipa: right.
506 2012-07-26 21:04:16 <sipa> sure, that's what i intended to do
507 2012-07-26 21:04:36 <sipa> if i ever get to rewrite a large part of bitcoin-seeder
508 2012-07-26 21:04:43 <gmaxwell> edcba: The testnet3 blockchain is only about 6mbytes in total.
509 2012-07-26 21:05:04 <gmaxwell> But my testnet3 wallet with lots of stuff in it is 481MBytes.
510 2012-07-26 21:06:06 <gavinandresen> gmaxwell: weird. My 24,000 BTC testnet3 wallet is only 1M
511 2012-07-26 21:07:59 <edcba> what is wallet storing ?
512 2012-07-26 21:08:05 <gmaxwell> Well, I do have like ... 2000 orphaned coinbases in it.
513 2012-07-26 21:08:07 <edcba> all tx ?
514 2012-07-26 21:08:20 <edcba> no i mean generally
515 2012-07-26 21:08:41 <gmaxwell> I wasn't responding to you.
516 2012-07-26 21:08:49 <gmaxwell> It stores your private keys and your own txn.
517 2012-07-26 21:08:56 <edcba> maybe some ppl spammed the network with something and you end up with those
518 2012-07-26 21:08:57 <edcba> hmm
519 2012-07-26 21:09:09 <edcba> noting more ?
520 2012-07-26 21:15:33 <gmaxwell> the bitcointools wallet stuff is really unhappy with a file this big.
521 2012-07-26 21:19:17 <Eliel> gmaxwell: the problem apparently wasn't due to the blk000x.dat file data. bitcoind has synced up fine.
522 2012-07-26 21:19:55 <gmaxwell> $ ./dbdump.py --wallet-tx | wc -l
523 2012-07-26 21:20:05 <midnightmagic> fuck you, whoever wrote satoshi dice
524 2012-07-26 21:20:19 <gmaxwell> midnightmagic: hm?
525 2012-07-26 21:22:36 <midnightmagic> gmaxwell: Block ACCEPTED           69435ms (block 188357)
526 2012-07-26 21:23:41 <Matt_von_Mises> sipa: THanks for mentioning the IPv6 addresses. I still need my code to detect it's own IP for that. Though I'm more worried about the addr messages. I'm so confused I made a topic about it: https://bitcointalk.org/index.php?topic=95907.0
527 2012-07-26 21:30:25 <gmaxwell> Matt_von_Mises: You should have asked the question you asked there here.
528 2012-07-26 21:31:03 <Matt_von_Mises> I did... Well maybe not as precisely.
529 2012-07-26 21:31:16 <Matt_von_Mises> But it is something that needs in-depth explaination
530 2012-07-26 21:31:26 <Matt_von_Mises> I need to go to bed soon as well.
531 2012-07-26 21:34:07 <gmaxwell> Matt_von_Mises: I can only speak to what I expirenced. What you asked on IRC didn't give me anything to answer, what you asked on the forum did.
532 2012-07-26 21:35:34 <Matt_von_Mises> Well the line I was looking at that I did not understand was the setting times to 5 days ago. But anyway, I've got to go to bed sorry. Bye.
533 2012-07-26 22:20:03 <jgarzik> getblocks -1 to 00000000000000000000 limit 500
534 2012-07-26 22:20:11 <jgarzik> on an outgoing-only, no-listen box
535 2012-07-26 22:20:18 <jgarzik> I wonder if that is somebody probing
536 2012-07-26 22:20:59 <jgarzik> run a public node,  attract connections like flies, then probe their block database?
537 2012-07-26 22:23:23 <jgarzik> sipa: bitcoin connecting to the same addresses, over and over and over again: http://gtf.org/garzik/bitcoin/debug.log.bz2
538 2012-07-26 22:23:34 <jgarzik> testnet3, not mainnet
539 2012-07-26 22:28:05 <sipa> i'm on my phone here; i'll check it when i get back (juli 30)
540 2012-07-26 22:29:08 <jgarzik> sipa: apologies... I thought you were back.  Enjoy the cold beer!
541 2012-07-26 22:31:34 <sipa> no prob!