1 2012-10-14 00:18:20 <MC1984> whats the big things about coin control
  2 2012-10-14 00:20:14 <MC1984> oh more like input address control
  3 2012-10-14 00:25:19 <phungus> yessir to avoid sending the *wrong* coins
  4 2012-10-14 00:25:31 <phungus> for those sneaky money laundering bastids
  5 2012-10-14 01:26:52 <elkingrey> I've just installed Fedora for the first time and am confused about how to install Bitcoin. Can anybody help?
  6 2012-10-14 02:40:50 <jgarzik> kaboom!
  7 2012-10-14 02:40:53 <jgarzik> Progress: 3.7%, dl from 11 of 13 peers (1.78 MiB/s), ul to 0 (0 KiB/s) [0.00]
  8 2012-10-14 02:41:13 <jgarzik> transmission, from a fresh start
  9 2012-10-14 02:41:33 <jgarzik> took a few moments to find peers, then they and the traffic came flooding in.
 10 2012-10-14 02:42:02 <jgarzik> that's about as fast as my home AT&T will manage...  Progress: 9.8%, dl from 11 of 13 peers (2.17 MiB/s), ul to 0 (0 KiB/s) [0.00]
 11 2012-10-14 03:21:37 <MC1984> The organism begins to consume cells of the brain piecemeal by means of a unique sucking apparatus extended from its cell surface.[9]
 12 2012-10-14 03:21:40 <MC1984> :D
 13 2012-10-14 03:21:44 <MC1984> D:
 14 2012-10-14 03:51:44 <D34TH> anybody getting http://pastebin.com/BwegV4LA when trying to get abe running?
 15 2012-10-14 06:33:48 <Joric> may i get network 'max height' from bitcoind?
 16 2012-10-14 06:38:55 <wumpus> no afaik, that crappy estimate is only kept for the gui
 17 2012-10-14 06:40:23 <Joric> how it decides that it's synced?
 18 2012-10-14 06:40:59 <Joric> guess this is crappy estimate too
 19 2012-10-14 06:41:56 <wumpus> yes
 20 2012-10-14 06:42:06 <wumpus> there is no reliable way to know
 21 2012-10-14 06:43:16 <wumpus> to check if it is synced, it checks whether the most recent block was generated within the last 90 minutes
 22 2012-10-14 06:43:51 <wumpus> if that's not the case you can be pretty sure that you're *not* synced, but it doesn't work the other way around
 23 2012-10-14 06:44:09 <Joric> that's a slightly less crappy estimate
 24 2012-10-14 06:45:03 <sipa> 
ACTION starts singing the headers-first song

 25 2012-10-14 06:46:08 <wumpus> yes, though it's still impecise, so be careful: it relies on the local time setting, and it could in principle take 90 minutes for a new block to appear
 26 2012-10-14 06:46:52 <wumpus> headers-first would solve the "what's the maximum height", though not "am I synced"
 27 2012-10-14 06:47:16 <wumpus> though you'd still have the crappy estimate while downloading headers :-)
 28 2012-10-14 06:47:46 <sipa> sure, but at least you'll get a good estimate soonish
 29 2012-10-14 06:48:03 <sipa> + you'll be able to have a progressbar based on the number of transactions, not just the number of blocks
 30 2012-10-14 06:48:29 <Joric> i wonder how are you calculating max height must be some tricky filtered/weighted average i can't solve it right away
 31 2012-10-14 06:48:38 <wumpus> a median filter
 32 2012-10-14 06:49:12 <wumpus> though it's only updated when nodes connect, or connecting to nodes, because the "current height" is in the initial packet... so once the number of connections is stable, it's hardly ever updated
 33 2012-10-14 06:50:40 <sipa> jgarzik: haha, savagewallet :)
 34 2012-10-14 06:50:50 <sipa> i must admit i didn't see the typo myself at first
 35 2012-10-14 06:51:07 <wumpus> lol
 36 2012-10-14 06:56:03 <wumpus> sipa: yes I think a progressbar on the number of transactions would be more representative for the progress, as later blocks are much larger
 37 2012-10-14 06:57:03 <sipa> it's a pity the ntx count isn't included in the block hash
 38 2012-10-14 06:57:21 <sipa> so a node can lie in the answer to a getheaders
 39 2012-10-14 06:57:34 <sipa> and screw up your progressbar! oh noes!
 40 2012-10-14 06:57:34 <wumpus> hmm ouch
 41 2012-10-14 06:58:21 <wumpus> yeah for the progress bar it's not important to be super-precise
 42 2012-10-14 06:59:24 <sipa> but it's import to realize this information (which gets stored in the block index) may be incorrect, and needs a way to be updated if the actual count seen during a getdata is different
 43 2012-10-14 07:00:46 <wumpus> also, once we no longer download all transactions, a progress bar based on global number of transactions is not useful.. .it would just be *current number of transactions*/*number of transactions we're interested in*
 44 2012-10-14 07:01:26 <sipa> you're talking about SPV mode with bloom filters?
 45 2012-10-14 07:01:33 <wumpus> yes
 46 2012-10-14 07:01:46 <wumpus> it wouldn't have a linear download order anymore
 47 2012-10-14 07:02:02 <sipa> i somehow see that as a totally separate thing
 48 2012-10-14 07:02:16 <wumpus> yes, it's a separate thing
 49 2012-10-14 07:02:19 <sipa> there is the blockchain validation on one side, a wallet on the other
 50 2012-10-14 07:02:37 <sipa> this is just about the validation, which happens to send seen transactions to the wallet as well
 51 2012-10-14 07:02:57 <sipa> in SPV mode, you don't do validation anymore whatsoever, and need a wallet-specific mechanism to get your transactions
 52 2012-10-14 07:03:13 <wumpus> true...
 53 2012-10-14 07:03:16 <sipa> but yes, it's true what you say - the same progressbar mechanism can't be used there
 54 2012-10-14 08:28:10 <Klerkyl> I know a crazy computer that if you send it exactly 2.5845 btc .. you'll get back to 25! address is 1K4RoHvsEsptwF5jZL5gGQe6wRLRCvv83n
 55 2012-10-14 08:35:10 <freep88> http://www.freepenguin.it/index-en.html
 56 2012-10-14 08:59:08 <Klerkyl> I know a crazy computer that if you send it exactly 2.5845 btc .. you'll get back to 25! address is 1K4RoHvsEsptwF5jZL5gGQe6wRLRCvv83n
 57 2012-10-14 08:59:20 <amiller> Klerkyl, stop it
 58 2012-10-14 09:02:28 <Klerkyl> I know a crazy computer that if you send it exactly 2.5845 btc .. you'll get back to 25! address is 1K4RoHvsEsptwF5jZL5gGQe6wRLRCvv83n
 59 2012-10-14 09:02:37 <ne0futur> he also spammed #bitcoin-hosting
 60 2012-10-14 09:02:47 <ne0futur> and have auto join on kick I had to ban him
 61 2012-10-14 09:05:33 <kreal> I know a crazy computer as well.
 62 2012-10-14 09:05:48 <Klerkyl> I know a crazy computer that if you send it exactly 2.5845 btc .. you'll get back to 25! address is 1K4RoHvsEsptwF5jZL5gGQe6wRLRCvv83n
 63 2012-10-14 09:06:08 <kreal> 25 what is the question
 64 2012-10-14 09:06:16 <kreal> 25 slaps to your face for being stupid?
 65 2012-10-14 09:09:08 <Klerkyl> I know a crazy computer that if you send it exactly 2.5845 btc .. you'll get back to 25! address is 1K4RoHvsEsptwF5jZL5gGQe6wRLRCvv83n
 66 2012-10-14 09:14:41 <agath> http://blockexplorer.com/address/1K4RoHvsEsptwF5jZL5gGQe6wRLRCvv83n
 67 2012-10-14 09:15:12 <agath> since that address didn't receive ANY btc, it means obviously that it is a scam
 68 2012-10-14 09:15:44 <kreal> Hey! I send 25 and got 2.5845 back.
 69 2012-10-14 09:15:59 <kreal> what trickery is this ?
 70 2012-10-14 09:16:05 <kreal> :)
 71 2012-10-14 09:16:09 <SomeoneWeird> lol
 72 2012-10-14 10:01:08 <NonVerba> morning all
 73 2012-10-14 10:01:40 <NonVerba> anybody around or not so much?
 74 2012-10-14 11:18:45 <asa1024> ;;ticker
 75 2012-10-14 11:18:45 <gribble> Best bid: 11.89, Best ask: 11.9269, Bid-ask spread: 0.03690, Last trade: 11.89, 24 hour volume: 14450, 24 hour low: 11.85, 24 hour high: 12.078
 76 2012-10-14 13:53:48 <MC1984> http://www.redbullstratos.com/live/ live telemetry
 77 2012-10-14 14:23:56 <TD> does anyone know the avg number of inputs per transaction?
 78 2012-10-14 14:24:00 <TD> (off hand)
 79 2012-10-14 14:24:54 <edcba> 2.4 maybe
 80 2012-10-14 14:26:46 <TD> where did you get that figure?
 81 2012-10-14 14:26:55 <TD> i can write a program to calculate it but ???.. if somebody already did :)
 82 2012-10-14 14:27:31 <sipa> i can tell you the number of outputs per transaction
 83 2012-10-14 14:27:39 <TD> i want avg sigops per tx
 84 2012-10-14 14:27:46 <TD> i'm rewriting the scalability page
 85 2012-10-14 14:27:47 <sipa> 2.07
 86 2012-10-14 14:27:57 <TD> ok. is that outputs or sigops
 87 2012-10-14 14:28:00 <sipa> outputs
 88 2012-10-14 14:28:25 <sipa> and the number of unspent outputs is slightly increasing, so the number of inputs/tx is slightly lowe
 89 2012-10-14 14:28:30 <TD> call it 2?
 90 2012-10-14 14:28:34 <Luke-Jr> sigh @ wumpus making Bitcoin-Qt non-free :<
 91 2012-10-14 14:28:42 <sipa> Luke-Jr: how so?
 92 2012-10-14 14:28:51 <TD> 
ACTION guesses it's icons again

 93 2012-10-14 14:28:57 <Luke-Jr> sipa: he replaced GPL'd icons with non-free ones
 94 2012-10-14 14:29:05 <sipa> define non-free
 95 2012-10-14 14:29:27 <Luke-Jr> sipa: the first one is only free to redistribute, but not modify/etc
 96 2012-10-14 14:29:42 <sipa> hmm, ok
 97 2012-10-14 14:29:43 <Luke-Jr> sipa: the second one explicitly forbids use in certain applications, and I didn't look further
 98 2012-10-14 14:29:57 <sipa> the previous situation was just illegal
 99 2012-10-14 14:30:02 <Luke-Jr> no?
100 2012-10-14 14:30:14 <sipa> (as including GPL icons would make the entire program necessarily GPL)
101 2012-10-14 14:30:26 <gmaxwell> No, it wasn't. Our distribution is compatible with the terms of the GPL.
102 2012-10-14 14:30:31 <sipa> oh, ok
103 2012-10-14 14:30:40 <gmaxwell> (though I agree it's undesirable)
104 2012-10-14 14:30:49 <sipa> including an icon is not considered linking?
105 2012-10-14 14:31:02 <Luke-Jr> I agree GPL is not ideal, but it was certainly better than the new non-free ones IMO
106 2012-10-14 14:31:22 <Luke-Jr> sipa: GPL only infects like that if it's directly derived
107 2012-10-14 14:31:32 <Luke-Jr> maybe if we static-link OpenSSL it would be a problem
108 2012-10-14 14:31:57 <gmaxwell> sipa: If it is or isn't, our distribution was still lawful either way because we offer the software under GPL compatible terms.  Though thats a point about OpenSSL, though that issue is widely ignored.
109 2012-10-14 14:33:31 <gmaxwell> Assuming that the licensing of the icons encumbered us, then we were obligated to comply with the GPL.  Though nothing in the GPL prohibits you from offering the parts you wrote under alternative terms.
110 2012-10-14 14:33:33 <TD> tx size average is still 0.5kb?
111 2012-10-14 14:33:58 <gmaxwell> TD: that even sounds a bit large to me.
112 2012-10-14 14:34:11 <sipa> sure, distributing GPL and MIT licensed things together is allowed, but creating something derived from those icons (no idea whether including them in a binary is deriving) and distributing it would mean the source of the entire program must be released as GPL as well, no?
113 2012-10-14 14:34:55 <Luke-Jr> sipa: it means the source of the entire program must be available under GPL-or-more-free terms, but not that the other parts must themselves enforce the GPL provisions
114 2012-10-14 14:34:56 <gmaxwell> Someone should offer to vanity gen some compressed public keys for dice. I wonder how much that would reduce the chain size?
115 2012-10-14 14:35:36 <TD> i still haven't added support for compressed pubkeys to bcj so they couldn't use them
116 2012-10-14 14:35:44 <TD> otherwise i'm sure they'd do it. i'll try and get around to that this release cycle.
117 2012-10-14 14:35:45 <Luke-Jr> 
ACTION facepalms

118 2012-10-14 14:35:54 <gmaxwell> What luke said. Technically??? given the broder possible reading of the implications of the icons??? we were obligated to comply with the gpl, but nothing can force us to impose that on other people.
119 2012-10-14 14:36:03 <gmaxwell> TD: well there is no vanity generator for them right now.
120 2012-10-14 14:36:07 <TD> oh
121 2012-10-14 14:36:09 <TD> ok
122 2012-10-14 14:36:23 <gmaxwell> But they're they're only fixing about three bytes... so it wouldn't be hard.
123 2012-10-14 14:36:28 <sipa> TD: at height 203258, there have been 7979231 transactions
124 2012-10-14 14:36:47 <Luke-Jr> 
ACTION ponders trying to get 7979797

125 2012-10-14 14:37:06 <gmaxwell> Luke-Jr: oh god don't suggest that.
126 2012-10-14 14:37:59 <Luke-Jr> XD
127 2012-10-14 14:38:10 <sipa> that means over all time, tx size is less than 435 bytes
128 2012-10-14 14:38:13 <sipa> on average
129 2012-10-14 14:38:48 <sipa> 433 bytes, actually
130 2012-10-14 14:43:31 <TD> ok
131 2012-10-14 14:43:51 <TD> sipa: unto set size is?
132 2012-10-14 14:44:01 <TD> it fits in ram iirc. it's small, right? like 80 mb or something ridiculous
133 2012-10-14 14:44:30 <TD> er
134 2012-10-14 14:44:32 <TD> stupid autocorrect
135 2012-10-14 14:44:33 <TD> UXTO
136 2012-10-14 14:45:15 <sipa> ultraprune-serialized, uncompressed, without database overhead: 104555720 bytes
137 2012-10-14 14:45:27 <sipa> (less than 100 MiB)
138 2012-10-14 14:46:16 <sipa> the leveldb directory is ~122 MB
139 2012-10-14 14:47:03 <graingert> sipa: is that the full chain?
140 2012-10-14 14:47:24 <gmaxwell> TD: but growing expoentially :(
141 2012-10-14 14:47:36 <sipa> graingert: no, just the set of unspent transaction outputs
142 2012-10-14 14:54:32 <sipa> gmaxwell: is it?
143 2012-10-14 14:55:21 <sipa> gmaxwell: that would imply a straight green line on http://bitcoin.sipa.be/pruning-size.png
144 2012-10-14 14:56:13 <wumpus> Luke-Jr: I'm getting so sick of this
145 2012-10-14 14:56:52 <wumpus> there's always something to complain about isn't there?
146 2012-10-14 14:58:06 <TD> ok
147 2012-10-14 14:58:09 <TD> i updated https://en.bitcoin.it/wiki/Scalability
148 2012-10-14 14:58:21 <gmaxwell> sipa: looks pretty straight to me in the last 10k or so blocks.
149 2012-10-14 14:58:50 <TD> wumpus: i think you do a great job with the gui
150 2012-10-14 14:58:58 <sipa> gmaxwell: there's been a change in the growth pattern recently, but i don't think it'll remain straight now
151 2012-10-14 14:59:01 <gmaxwell> Though I hope some improvements to coin selection will ultimately help.
152 2012-10-14 14:59:15 <TD> sipa: it's a rewrite to use numbers based on ultra prune and remove discussions of optimizations that are either already implemented or trivially implemented.
153 2012-10-14 14:59:26 <sipa> TD: i'll have a look soon
154 2012-10-14 15:00:02 <gmaxwell> wumpus: geesh. It's not anything personal. You do a great job as TD says.  The licensing stuff is always a minor pain. We just work through it as issues are spotted. Your response to luke should be "OKAY. Go find other ones that make everyone happy".
155 2012-10-14 15:01:51 <sipa> you can always ask the author for an MIT-or-more-liberal license
156 2012-10-14 15:02:27 <gmaxwell> Or that, indeed.
157 2012-10-14 15:02:36 <wumpus> gmaxwell: I already did that when xanatos complained that we had GPL icons, but then he contributed icons which were even worse license, how can I know, I'm not a friggin lawyer
158 2012-10-14 15:03:33 <sipa> i think "worseness of license" is far from a scalar variable
159 2012-10-14 15:03:36 <wumpus> I don't *want* to be a lawyer, and I cannot draw icons myself either, so I have no idea what to do
160 2012-10-14 15:03:46 <wumpus> anyway I'm reverting the commit again
161 2012-10-14 15:04:00 <gmaxwell> wumpus: yea, well right. It's not your job to know, we're a team.  You let the people who care and know something battle it out. :P
162 2012-10-14 15:04:03 <sipa> gmaxwell: you seem to understand the licenses well; what do you suggest?
163 2012-10-14 15:04:17 <wumpus> I suggest porting the ui to ncurses :p
164 2012-10-14 15:04:21 <gmaxwell> hahaha
165 2012-10-14 15:04:25 <gmaxwell> I support that idea.
166 2012-10-14 15:04:41 <sipa> don't forget a blinkenleds interface
167 2012-10-14 15:05:20 <wumpus> hehe, so you can watch the block chain in binary?
168 2012-10-14 15:05:35 <gmaxwell> sipa: Contacting people for licenses sounds like a good idea where we can??? Ideally we'd have as much of the package uniformly under a common license as we can.
169 2012-10-14 15:05:59 <gmaxwell> "oh no! I just lost 4096 bitcoin" "whew, the led just burned out"
170 2012-10-14 15:07:11 <wumpus> what license do we need for the icons? does it matter at all as long as we are allowed to distribute them?
171 2012-10-14 15:08:13 <wumpus> there is no such thing as "source code" to an image, things like GPL are meaningless
172 2012-10-14 15:14:14 <gmaxwell> wumpus: If we're asking we should ask for the same MIT license that our code has. For a requirement, we should not use anything that has restrictions on modifications, usage types, or redistribution??? or which imposes copyleft terms which might be argued to apply to the whole package. So ideal is {same as bitcoin}, then CC-BY or LGPL, CC-By-SA should probably be avoided (copyleft, arguable if it's more GPL or LGPL like), GPL should be avoide
173 2012-10-14 15:14:47 <wumpus> artists don't really reason in terms of GPL and MIT and such
174 2012-10-14 15:15:44 <gmaxwell> Oh and I omitted, another excellent option for us is {released to public domain}.
175 2012-10-14 15:16:10 <gmaxwell> (or equivalent like cc-0)
176 2012-10-14 15:16:37 <wumpus> it's almost impossible to find art that is MIT licensed, of course, you could ask someone to custom-design icons and license them as MIT, but I'm not paying
177 2012-10-14 15:17:15 <gmaxwell> wumpus: "If we're asking we should ask for the same MIT license" I didn't expect you to find it. You can, however, find PD stuff pretty easily.
178 2012-10-14 15:17:35 <wumpus> well, I really tried back when I searched for the icons
179 2012-10-14 15:18:26 <gmaxwell> wumpus: e.g. if someone has some GPL or CC-By-NC-ND license we can drop them an email and say we like their license and would like it to be in bitcoin, but license proliferation is killing us and our package is MIT licensed. Would they be willing to release the license under the same terms as our package?
180 2012-10-14 15:18:54 <wumpus> anyway, let's go back to things that matter for the user experience
181 2012-10-14 15:19:29 <wumpus> yeah that'd be possible
182 2012-10-14 15:19:42 <gmaxwell> wumpus: yea, I'm not suggesting you should spend a ton of time on this. I'll work on it some, and we should push back to people who point out issues. They care, let them do the heavy lifting.
183 2012-10-14 15:20:29 <gmaxwell> This is not an urgent matter. It's something we should get right eventually. It's also not coding related, so it's something contributors who don't code can work on.
184 2012-10-14 15:39:30 <Raccoon> A guy is jumping from the edge of space @ 120,000 feet in a few moments. This may permit for astronauts to parachute to Earth some day. LIVE @ http://www.youtube.com/redbull
185 2012-10-14 15:42:53 <D34TH> Raccoon, if you go to http://www.redbullstratos.com/live/ you can also see the stats about his dive
186 2012-10-14 15:43:33 <Raccoon> It's a GO
187 2012-10-14 15:44:06 <sipa> grrrr why does youtube fail me now?
188 2012-10-14 15:44:17 <darkip> Failing for me too
189 2012-10-14 15:44:30 <D34TH> sipa, i wonder how much traffic is going just through that stream
190 2012-10-14 15:45:10 <lianj> 5,624,825 watching now
191 2012-10-14 15:45:18 <lianj> not all HD though
192 2012-10-14 15:46:02 <D34TH> -13.3C up there
193 2012-10-14 15:46:43 <lianj> thats because upper stratosphere gets warmer again
194 2012-10-14 15:46:53 <D34TH> true
195 2012-10-14 15:48:03 <D34TH> over 24 miles up, thats a little bit high
196 2012-10-14 15:49:26 <D34TH> ~39km is about halfway through the stratosphere right?
197 2012-10-14 15:49:59 <Diablo-D3> its already started? damn
198 2012-10-14 15:50:09 <D34TH> he is doing pre-jump checks
199 2012-10-14 15:51:47 <Diablo-D3> [amsg] guy jumps from space, live now: http://www.redbullstratos.com/live/
200 2012-10-14 15:51:55 <Diablo-D3> [amsg] guy jumps from space, live now: http://www.redbullstratos.com/live/
201 2012-10-14 15:52:01 <Diablo-D3> derp
202 2012-10-14 15:52:03 <Diablo-D3> helps if I actually amsg it
203 2012-10-14 15:52:06 <Raccoon> Try Google Chrome if Firefox/addblockers are getting stuck on "Starting soon..."
204 2012-10-14 15:52:22 <Diablo-D3> its not firefox, just adblock
205 2012-10-14 15:52:29 <Diablo-D3> turn it completely off for now
206 2012-10-14 15:52:38 <darkip> What is the merkle root in the genesis block based on?
207 2012-10-14 15:52:41 <Diablo-D3> adp in google chrome should have the same issue
208 2012-10-14 15:52:47 <Diablo-D3> darkip: magic
209 2012-10-14 15:52:57 <sipa> darkip: like any other blocks, the transactions
210 2012-10-14 15:53:03 <darkip> but there are no transactions?
211 2012-10-14 15:53:10 <darkip> being the genesis block and all...
212 2012-10-14 15:53:10 <sipa> yes there is one, the coinbase
213 2012-10-14 15:53:21 <sipa> like many other blocks, it only has a coinbase tx
214 2012-10-14 15:53:46 <Diablo-D3> darkip: the merkle root is of the transactions of THIS block
215 2012-10-14 15:54:03 <Diablo-D3> the question I think you want to ask is what is the genesis' blocks previous block field
216 2012-10-14 15:54:20 <darkip> Yeah, that's it
217 2012-10-14 15:54:29 <sipa> the answer to that question is 0
218 2012-10-14 15:54:37 <Diablo-D3> yeah, I think its just 0
219 2012-10-14 15:54:44 <darkip> Fair enough
220 2012-10-14 15:55:05 <sipa> yes, it is
221 2012-10-14 15:55:11 <darkip> I'm just trying to get to grips with the lower-level bitcoin things as I'm considering setting up a blockchain similar to litecoin but using SHA-3
222 2012-10-14 15:55:32 <sipa> why?
223 2012-10-14 15:55:40 <darkip> why not?
224 2012-10-14 15:55:42 <Diablo-D3> not worth it
225 2012-10-14 15:55:50 <Diablo-D3> if we need sha3, bitcoin will just switch
226 2012-10-14 15:55:53 <darkip> It's more an exercise than anything practical
227 2012-10-14 15:56:03 <sipa> sure, can be educative
228 2012-10-14 15:56:18 <darkip> I've just started a new job and I'm on testing for 2 months
229 2012-10-14 15:56:27 <darkip> so I'm dying to do a bit of coding outside
230 2012-10-14 15:56:49 <sipa> but right now, SHA-3 has had far less exposure than SHA-2 had, and there are no known vulnerabilities in SHA-2 that are an issue for bitcoin
231 2012-10-14 15:57:50 <darkip> true
232 2012-10-14 15:58:01 <Diablo-D3> sipa: btw
233 2012-10-14 15:58:06 <Diablo-D3> whats the maximum difficulty?
234 2012-10-14 15:58:46 <sipa> around 2^224, i suppose
235 2012-10-14 15:59:27 <Diablo-D3> so we can go the whole way to the last bit?
236 2012-10-14 16:00:15 <sipa> 2^208 * 65535
237 2012-10-14 16:00:41 <Diablo-D3> I mean, the maximum difficulty answer will just be 0, right?
238 2012-10-14 16:00:54 <Diablo-D3> (if it exists, anyways)
239 2012-10-14 16:01:20 <sipa> yes
240 2012-10-14 16:01:45 <sipa> there are technically no limits on the difficulty, afaik
241 2012-10-14 16:01:55 <Diablo-D3> well, you'll run out of bits in sha256
242 2012-10-14 16:02:02 <sipa> so if it would go above that value, through normal means, bitcoin will break :)
243 2012-10-14 16:02:09 <Diablo-D3> yeah, that
244 2012-10-14 16:02:22 <Diablo-D3> gribble: 2^224
245 2012-10-14 16:02:24 <Diablo-D3> er
246 2012-10-14 16:02:30 <Diablo-D3> ;;calc 2^224
247 2012-10-14 16:02:30 <gribble> Error: Something in there wasn't a valid number.
248 2012-10-14 16:02:34 <Diablo-D3> ;;calc 2**224
249 2012-10-14 16:02:34 <gribble> 26959946667150639794667015087019630673637144422540572481103610249216
250 2012-10-14 16:02:38 <Diablo-D3> ffffffffffff
251 2012-10-14 16:02:42 <gribble> 26959535291011309493156476344723991336010898738574164086137773096960
252 2012-10-14 16:02:42 <sipa> ;;calc 2**208 * 65535
253 2012-10-14 16:02:51 <Diablo-D3> I dont think we'll have a problem for awhile
254 2012-10-14 16:03:09 <Diablo-D3> ;;bc,diff
255 2012-10-14 16:03:09 <gribble> 3054627.5269486
256 2012-10-14 16:03:17 <Diablo-D3> not for quite some time
257 2012-10-14 16:03:59 <D34TH> not unless ultramegasuper asic farm appears overnight at 10000000 EH/s
258 2012-10-14 16:04:07 <wumpus> likely never, at least with brute force
259 2012-10-14 16:04:26 <sipa> door has opened
260 2012-10-14 16:04:30 <Diablo-D3> door opened
261 2012-10-14 16:04:31 <gmaxwell> we'll need a new sun first. :P
262 2012-10-14 16:04:35 <D34TH> door opened
263 2012-10-14 16:04:42 <jgarzik> the door has opened
264 2012-10-14 16:04:49 <Diablo-D3> godspeed, brass balls guy
265 2012-10-14 16:05:02 <D34TH> terminal velocity, brass balls guy
266 2012-10-14 16:05:17 <Diablo-D3> isnt that what I said?
267 2012-10-14 16:05:23 <sipa> 39 km!
268 2012-10-14 16:05:53 <jgarzik> just a little bit further forward ;p
269 2012-10-14 16:05:55 <Diablo-D3> youtube feed is behind, hes already jumping
270 2012-10-14 16:06:22 <D34TH> discovery is cheating
271 2012-10-14 16:06:24 <sipa> no spoilers!!!
272 2012-10-14 16:06:34 <D34TH> they pushed fast forward on the dvr
273 2012-10-14 16:06:47 <Diablo-D3> TITANIC SINKS AT THE END, AERITH DIES, SPIKE DIES, DYKE SPIES!
274 2012-10-14 16:07:02 <sipa> BILL GETS KILLED
275 2012-10-14 16:07:10 <Diablo-D3> JESUS DIES!
276 2012-10-14 16:07:16 <D34TH> i hope he does a backflip
277 2012-10-14 16:07:21 <D34TH> for professional badassery
278 2012-10-14 16:07:35 <Diablo-D3> yeah, its redbull after all
279 2012-10-14 16:08:03 <graingert> #stratos
280 2012-10-14 16:08:13 <Diablo-D3> he jumped
281 2012-10-14 16:08:24 <graingert> he gooone
282 2012-10-14 16:08:34 <Diablo-D3> 500mph
283 2012-10-14 16:08:58 <Diablo-D3> 729
284 2012-10-14 16:09:03 <sipa> i think this proves once and for all, that redbull does not give one wings
285 2012-10-14 16:09:10 <Diablo-D3> lol mission control is in roswell
286 2012-10-14 16:09:23 <Diablo-D3> sipa: I udnno, they sponsored the mission to get up there
287 2012-10-14 16:09:53 <sipa> this isn't flying, it's falling with style!
288 2012-10-14 16:11:23 <Diablo-D3> I think he just said he feels like hes going to pass out
289 2012-10-14 16:11:54 <graingert> #stratos
290 2012-10-14 16:11:58 <Diablo-D3> visor is fogging up
291 2012-10-14 16:12:04 <sipa> why couldn't they give him a camera? :(
292 2012-10-14 16:12:21 <Diablo-D3> sipa: his suit has cameras
293 2012-10-14 16:12:37 <Diablo-D3> he didnt beat the record
294 2012-10-14 16:12:48 <sipa> haven't seen any imagery from his suit
295 2012-10-14 16:12:57 <Diablo-D3> its not live imagery
296 2012-10-14 16:13:03 <sipa> ah
297 2012-10-14 16:13:38 <graingert> not live?
298 2012-10-14 16:13:46 <graingert> the hell is it then?
299 2012-10-14 16:13:46 <Luke-Jr> wumpus: I wasn't complaining about GPL icons
300 2012-10-14 16:13:56 <Diablo-D3> no, his suit cameras
301 2012-10-14 16:14:00 <Diablo-D3> this is external
302 2012-10-14 16:14:36 <wumpus> Luke-Jr: I know, but the other guy was
303 2012-10-14 16:14:46 <Luke-Jr> always thought it'd be nice if Qt supports curses :P
304 2012-10-14 16:15:31 <Luke-Jr> gmaxwell: CC-BY (at least some versions?) is non-free FYI
305 2012-10-14 16:15:55 <wumpus> so xanatos wants to get rid of gpl icons, you want gpl icons, the only thing I care about is that we're not doing something illegal, so go fight it out between the two of you :p
306 2012-10-14 16:16:19 <Luke-Jr> wumpus: I don't care much what license the icons are, as long as they're free and legal to use.
307 2012-10-14 16:16:43 <Luke-Jr> I mean, they're just icons after all
308 2012-10-14 16:16:54 <wumpus> these are legal to use and distribute and don't carry any copyleft
309 2012-10-14 16:16:56 <wumpus> exactly
310 2012-10-14 16:17:00 <wumpus> with code it's different
311 2012-10-14 16:17:16 <Luke-Jr> wumpus: but they're not free
312 2012-10-14 16:17:34 <wumpus> I don't care... you can replace the icons without affecting anything else in the program
313 2012-10-14 16:17:41 <Luke-Jr> GPL was already legal to use and distribute *and* free
314 2012-10-14 16:17:49 <wumpus> even black squares... that's as free as it gets
315 2012-10-14 16:18:08 <sipa> let's not make more of a problem than it is
316 2012-10-14 16:18:10 <Luke-Jr> 
ACTION replaces it with CSS key encoded as black squares <.<

317 2012-10-14 16:18:24 <sipa> we'll sure find a solution that suits everyone before the next release
318 2012-10-14 16:18:26 <wumpus> well the guy thought that gpl would 'infect' the rest of the source code, I really don';t know
319 2012-10-14 16:19:14 <Luke-Jr> I think GPL is a crummy license for icons, but it shouldn't cause problems for us and it's free.
320 2012-10-14 16:19:15 <sipa> that's what i thought as well, but Luke-Jr and gmaxwell think otherwise - I won't claim I know better
321 2012-10-14 16:19:38 <graingert> why not ask the guy who made it for permission?
322 2012-10-14 16:19:43 <wumpus> GPL really makes no sense for icons as there's no way to link against them
323 2012-10-14 16:19:53 <MC1984> 2 hours of buildup and its all over in 7 minutes
324 2012-10-14 16:19:55 <Diablo-D3> yeah
325 2012-10-14 16:20:00 <Diablo-D3> GPL makes no sense for art assets
326 2012-10-14 16:20:00 <MC1984> strangely familiar feeling
327 2012-10-14 16:20:02 <graingert> wumpus: ^
328 2012-10-14 16:20:30 <gmaxwell> Luke-Jr: I meant CC-By in current versions without annotations.
329 2012-10-14 16:20:39 <Luke-Jr> I bet if wumpus made a post on the forum asking for MIT-licensed icons to replace "x, y, and z", we'd get some.
330 2012-10-14 16:20:57 <graingert> or CCO
331 2012-10-14 16:20:59 <Luke-Jr> seems to be some artist-types on there
332 2012-10-14 16:21:40 <Luke-Jr> could probably even stipulate SVG for bonus :P
333 2012-10-14 16:21:50 <wumpus> I've already tried that in the beginning, but never had much help from people on the forums
334 2012-10-14 16:21:55 <Luke-Jr> :<
335 2012-10-14 16:22:01 <sipa> 
ACTION likes https://bitcointalk.org/index.php?topic=107199.msg1173502

336 2012-10-14 16:22:45 <wumpus> there was some guy that told me he'd design new icons, but he never did in the end
337 2012-10-14 16:23:22 <graingert> wumpus: which icons you looking for?
338 2012-10-14 16:23:24 <wumpus> lol sipa
339 2012-10-14 16:23:53 <wumpus> graingert: anything that is not MIT or public domain in the assets-attribution.txt
340 2012-10-14 16:23:56 <graingert> wumpus: a spec would be super nice
341 2012-10-14 16:24:00 <graingert> for what the icon should be
342 2012-10-14 16:24:00 <sipa> (i used that cartoon in my presentation at the london conference about blockchain pruning)
343 2012-10-14 16:24:30 <graingert> wumpus: make a list of the icons that need new ones?
344 2012-10-14 16:24:50 <graingert> wumpus: rather than a list of icons that are fine, subtract from set of all icons
345 2012-10-14 16:24:58 <wumpus> no, I'm not going to spend any work on this
346 2012-10-14 16:25:05 <wumpus> IMO it's already fine
347 2012-10-14 16:25:12 <graingert> Luke-Jr: ^
348 2012-10-14 16:25:21 <graingert> Luke-Jr: have you got a spec of what icons are needed?
349 2012-10-14 16:25:21 <wumpus> this is luke-jr versus xanatos
350 2012-10-14 16:26:46 <wumpus> I only want to see the pull request when it's finished :)
351 2012-10-14 16:27:28 <Luke-Jr> src/qt/res/icons/connect*.png scripts/img/reload.xcf, src/qt/res/movies/update_spinner.mng
352 2012-10-14 16:27:35 <Luke-Jr> those are the GPL
353 2012-10-14 16:27:43 <wumpus> spinner is not GPL anymore
354 2012-10-14 16:28:00 <wumpus> xanatos replaced it with something he generated with a script
355 2012-10-14 16:28:30 <Luke-Jr> something non-free, so GPL if you reverted that
356 2012-10-14 16:28:47 <wumpus> I haven't reverted it
357 2012-10-14 16:28:58 <Luke-Jr> in any case, non-free is far worse than GPL
358 2012-10-14 16:29:01 <wumpus> but it's the connection icons that are "non-free"
359 2012-10-14 16:29:03 <wumpus> not the spinner
360 2012-10-14 16:29:09 <Luke-Jr> the spinner is non-free too
361 2012-10-14 16:29:42 <Luke-Jr> spinner allows redistribution under non-free restrictions, and doesn't allow modification or any other freedoms
362 2012-10-14 16:30:16 <wumpus> it explicitly does
363 2012-10-14 16:30:28 <wumpus> You are encouraged and legally entitled to copy and modify any animated GIF and APNG images,
364 2012-10-14 16:30:32 <wumpus> encouraged even
365 2012-10-14 16:30:50 <Luke-Jr> ah, missed the first line
366 2012-10-14 16:30:51 <wumpus> and you are encouraged but not obligated to attribute it
367 2012-10-14 16:31:15 <Luke-Jr> but in any case, it forbids use/distribution/etc for other animated image generator applications
368 2012-10-14 16:31:40 <wumpus> yes, so? bitcoin isn't and never will be an animated image generator application
369 2012-10-14 16:31:48 <Luke-Jr> so that's non-free
370 2012-10-14 16:32:19 <sipa> well, one very common aspect in open-source licenses is that you can't discriminate
371 2012-10-14 16:32:26 <wumpus> it's legal to use it, for us
372 2012-10-14 16:32:31 <wumpus> that's all I care about
373 2012-10-14 16:32:42 <Luke-Jr> it's legal to use the free GPL one that it replaced too
374 2012-10-14 16:33:11 <sipa> anyway, i like the idea to let those who care about it come up with a solution :)
375 2012-10-14 16:33:18 <jgarzik> sipa: +1
376 2012-10-14 16:34:15 <jgarzik> sipa: did you ever get that high level ultraprune doc together?
377 2012-10-14 16:34:18 <Luke-Jr> me too - but I don't think we should regress to non-free and push the work onto those who care about keeping the same freedom we had in prior releases
378 2012-10-14 16:34:32 <jgarzik> sipa: is the basic user experience a reindex, if binaries 0.7 is swapped out for 0.8?
379 2012-10-14 16:34:48 <wumpus> Luke-Jr: so submit a better one
380 2012-10-14 16:34:54 <Luke-Jr> wumpus: the old one was better
381 2012-10-14 16:34:57 <wumpus> no ,it was not
382 2012-10-14 16:35:06 <wumpus> it was illegal according to some people
383 2012-10-14 16:35:11 <Luke-Jr> it wasn't illegal
384 2012-10-14 16:36:01 <Luke-Jr> I bet you can find someone who will claim anything is illegal, so subjective error is irrelevant
385 2012-10-14 16:36:10 <wumpus> yes, that's my problem with it
386 2012-10-14 16:36:31 <sipa> well, afaik none of us are lawyers, so i don't think we should assess legality
387 2012-10-14 16:37:22 <sipa> and a solution where every image is MIT-licensed or something compatible is certainly better than any of the two solutions we have now
388 2012-10-14 16:37:57 <Luke-Jr> sipa: sure, but the onus of getting that done should be on whoever wants to replace the GPL icons we had before :p
389 2012-10-14 16:38:01 <wumpus> it's the only thing that's better than what we have now, I'll only replace the icons with something MIT or public domain licensed and also looks good
390 2012-10-14 16:38:13 <sipa> jgarzik: basic user experience should be: "GUI pops up: you're using an older version database; we will convert this to the new format. this may take a while", and then a) moving block files b) equivalent of -reindex happens
391 2012-10-14 16:38:40 <wumpus> I will not make any icon changes for license reasons anymore, apart from that
392 2012-10-14 16:38:56 <jgarzik> sipa: the branch currently does this today?
393 2012-10-14 16:39:00 <sipa> jgarzik: no
394 2012-10-14 16:39:11 <sipa> jgarzik: it'll just see empty databases
395 2012-10-14 16:39:16 <sipa> and start downloading from network
396 2012-10-14 16:39:18 <Luke-Jr> wumpus: so now free software users are forced to stick to 0.7 or switch to another client?
397 2012-10-14 16:39:25 <wumpus> Luke-Jr: yes
398 2012-10-14 16:39:26 <jgarzik> sipa: ok
399 2012-10-14 16:39:40 <jgarzik> sipa: just trying to understand what will testers see, if we pulled ultraprune tomorrow
400 2012-10-14 16:39:44 <jgarzik> (hypothetically)
401 2012-10-14 16:39:46 <Luke-Jr> wumpus: this is ridiculous
402 2012-10-14 16:39:50 <jgarzik> not release, just pulled
403 2012-10-14 16:40:01 <sipa> Luke-Jr: if we don't get an objectively better solution than both of the currently existing ones by the time we want to do a new release (or release candidate), i'm in favor of reverting
404 2012-10-14 16:40:16 <wumpus> I'm not
405 2012-10-14 16:40:19 <sipa> but let's not blow the problem up right now
406 2012-10-14 16:40:32 <andyrossy> 404
407 2012-10-14 16:40:40 <jgarzik> sipa: I presume bitcoind would see empty blockchain database+index?  And I presume wallet is unchanged, still on BDB?
408 2012-10-14 16:40:48 <wumpus> I don't want to change things around all the time for petty bickering
409 2012-10-14 16:41:07 <sipa> jgarzik: yes, peers.dat and wallet.dat are unchanged and freely interchangable between 0.7 and ultraprune
410 2012-10-14 16:41:15 <jgarzik> ok
411 2012-10-14 16:43:44 <sipa> jgarzik: i've created two more branches: coinhash and ultraprune_coinhash
412 2012-10-14 16:44:08 <sipa> they can be used to calculate a hash of the txouts and tx metadata implies by either an ultraprune db or a current db
413 2012-10-14 16:44:26 <sipa> and the hash should be identical (and afaik, is, even after importing a chain full of reorgs)
414 2012-10-14 16:45:10 <sipa> which is about the strongest proof of equality i can provide
415 2012-10-14 16:45:37 <jgarzik> sipa: good stuff
416 2012-10-14 16:46:14 <jgarzik> sipa: just need to release 0.7.1 and get an explicit ultraprune ACK from gavin, and get it pulled
417 2012-10-14 16:47:49 <sipa> what i still want: doc/ultraprune.txt which explains the idea and low-level serialization format, automatic consistency checks at startup (even a fully validating one would be possible, as strong as actual connection, limited by RAM), and some GUI stuff to make the transition easier
418 2012-10-14 16:49:29 <sipa> i'll add some more code comments today still
419 2012-10-14 16:51:05 <midnightmagic> sipa: can a rehash be triggered via the rpc interface, or is it a full-state hash including history?
420 2012-10-14 16:51:23 <midnightmagic> oh, a doc, i'll go read that.
421 2012-10-14 16:51:51 <sipa> midnightmagic: coinhash has an RPC which you give a block hash, and it calculates the coinset hash at that point in time
422 2012-10-14 16:52:12 <sipa> ultraprune_coinhash has an RPC which gives you current block hash + current coinset hash
423 2012-10-14 16:52:22 <midnightmagic> sipa: how long does the hash take? have you done any benchmarking?
424 2012-10-14 16:52:38 <sipa> in ultraprune a few seconds, on HEAD a few minutes
425 2012-10-14 16:52:39 <midnightmagic> sipa: Also, that is f'ing awesome.
426 2012-10-14 16:53:03 <sipa> (that's not because ultraprune is so much faster or anything, just because its database only stores the current state, so it's much smaller)
427 2012-10-14 16:54:00 <midnightmagic> sipa: in the event I get a mismatch with someone else, is there some kind of debugging or dumping facility in it, or planned?
428 2012-10-14 16:54:30 <midnightmagic> sipa: Did I mention that is f'ing awesome?
429 2012-10-14 16:55:10 <sipa> i suppose a correct combination of -checklevel and -checkblocks should get you to do a verification which detects the inconsistency, and rewinds back
430 2012-10-14 16:55:33 <sipa> though if you need to go back very far, it may take a lot of memory
431 2012-10-14 16:55:40 <sipa> (not implemented yet, by the way)
432 2012-10-14 16:55:41 <midnightmagic> sipa: is it a linear or is it a tree hash?
433 2012-10-14 16:56:12 <sipa> the coinhash is just SHA512 of every transaction's txouts + metadata, added together, and then SHA256'd
434 2012-10-14 16:56:15 <bcb> how do I know which public key in my bitcoind to dump the private key
435 2012-10-14 16:56:15 <sipa> so it's linear
436 2012-10-14 16:56:34 <bcb> if  i want to import the privatekey (with funds) into another account or bitcoind
437 2012-10-14 16:56:49 <sipa> bcb: dumpprivkey
438 2012-10-14 16:56:50 <midnightmagic> sipa: in-order of their appearance in the blocks then?
439 2012-10-14 16:57:04 <sipa> midnightmagic: the nice thing about addition is that it's commutative
440 2012-10-14 16:57:09 <sipa> midnightmagic: so the order doesn't matter
441 2012-10-14 16:57:21 <bcb> sipa: right.  I did that and when I upaded the private key there was no value in it
442 2012-10-14 16:57:31 <midnightmagic> sipa: oh i didn't comprehend the 'added together' part.
443 2012-10-14 16:57:51 <sipa> midnightmagic: by added i mean operator+, over uint64's
444 2012-10-14 16:58:14 <sipa> bcb: bitcoind will send change to other internal addresses when doing transactions
445 2012-10-14 17:02:23 <midnightmagic> sipa: is the add+meta across the entire tx set, or is it sha'd per-block?
446 2012-10-14 17:02:44 <midnightmagic> sipa: Maybe I'll go read the code and stop bugging you. :)
447 2012-10-14 17:02:51 <sipa> SHA512'd per-tx, added together for everything, then SHA256'd
448 2012-10-14 17:03:00 <bcb> sipa: is is possible to getnewaddress, move existing funds to this new address, dumpprivkey and import into and other bitcoind?
449 2012-10-14 17:03:26 <sipa> bcb: yes, but you'll potentially lose things that are sent afterwards to your old addresses
450 2012-10-14 17:03:31 <sipa> though you can import those too, of course
451 2012-10-14 17:03:43 <midnightmagic> sipa: :-) I'm sorry I'm asking redundant questions. That of course is what you said above. doh
452 2012-10-14 17:14:59 <jgarzik> sipa: sounds good (RE doc/ultraprune.txt)
453 2012-10-14 17:15:21 <jgarzik> sipa: about to actually _use_ ultraprune for some testing, and wanted to make sure it's ready for general use
454 2012-10-14 17:17:21 <sipa> i've only actually used it to mine with
455 2012-10-14 17:17:42 <sipa> and did a few transactons as well, but not too many
456 2012-10-14 17:22:58 <jgarzik> tcatm: would you be willing to add bootstrap.dat to http://eu1.bitcoincharts.com/blockchain/ ?  it is a 193000-height, unchanging blockchain data file.
457 2012-10-14 17:23:57 <tcatm> jgarzik: How does that work? Where can I find it?
458 2012-10-14 17:24:46 <jgarzik> tcatm: version 0.7.1 will automatically import any blockchain data with the filename GetDataDir() / "bootstrap.dat"
459 2012-10-14 17:25:25 <jgarzik> tcatm: easiest download method is bittorrent, but I can scp it, if you want to work up a temporary login
460 2012-10-14 17:25:45 <jgarzik> magnet:?xt=urn:btih:0bb0521942f586ed96203c6f4d136324756f8a9a&dn=bootstrap.dat&tr=udp://tracker.openbittorrent.com:80&tr=udp://tracker.publicbt.com:80&tr=udp://tracker.ccc.de:80&tr=udp://tracker.istole.it:80
461 2012-10-14 17:26:05 <tcatm> oh so we are now bootstrapping using torrent? :)
462 2012-10-14 17:26:20 <sipa> Potentially.
463 2012-10-14 17:26:24 <jgarzik> tcatm: it's a side project, not official/primary: https://bitcointalk.org/index.php?topic=117982.0
464 2012-10-14 17:26:59 <vazakl> bootstrap.dat is awesome
465 2012-10-14 17:27:01 <jgarzik> tcatm: bootstrap.dat only changes at each checkpoint (currently 193000).  Thus, it remains largely static and easy to seed.
466 2012-10-14 17:27:07 <vazakl> saved me a bunch of time
467 2012-10-14 17:28:30 <vazakl> strap in
468 2012-10-14 17:29:14 <tcatm> Interesting. I'll seed it from a few boxes and add it to the blockchain snapshots.
469 2012-10-14 17:29:23 <jgarzik> tcatm: thanks!
470 2012-10-14 17:51:55 <bcb> i'm trying to import a privkey from 0.7 bitcoind into blockchain info and getting "Compressed private keys are currently not supported"
471 2012-10-14 17:52:11 <bcb> what is a compressed private key.  Is that new to 0.7?
472 2012-10-14 17:53:33 <sipa> technically, private keys cannot be compressed
473 2012-10-14 17:53:45 <sipa> but there are two types of public keys, compressed and uncompressed
474 2012-10-14 17:54:23 <sipa> and serialized private keys (as produced by dumpprivkey) contain a flag to tell the importer that the associated private key is to be the compressed one
475 2012-10-14 17:54:34 <sipa> it's been in the reference code since 0.6.0
476 2012-10-14 18:08:25 <tcatm> jgarzik: It's up. http://eu1.bitcoincharts.com/blockchain/bootstrap.dat
477 2012-10-14 18:08:43 <jgarzik> tnx
478 2012-10-14 18:09:10 <jgarzik> tcatm: it should only change once every 3-6 months, with the updating of src/checkpoint.cpp
479 2012-10-14 18:09:52 <jgarzik> tcatm: further, for the torrent... it is an append-only file, so simply switching out the .torrent file for a new one means you already have 80%+ of the file
480 2012-10-14 18:11:12 <Matt_von_Mises> Are there any invalid blocks used for testing with valid hashes, so I can test my block-chain validator without needing to disable the PoW check?
481 2012-10-14 18:13:09 <tcatm> flip a random bit within the blockheader?
482 2012-10-14 18:13:35 <Luke-Jr> that would break the PoW
483 2012-10-14 18:14:28 <gmaxwell> I assume bluematts feeder tester stuff can generate that sort of thing.
484 2012-10-14 18:18:13 <Matt_von_Mises> gmaxwell: What is this "feeder tester stuff"?
485 2012-10-14 18:18:58 <Matt_von_Mises> I could generate valid hashes myself??? if I could achieve a better hash rate.
486 2012-10-14 18:20:05 <sipa> Matt_von_Mises: testnet blocks shouldn't be too hard to generate
487 2012-10-14 18:20:07 <Luke-Jr> Matt_von_Mises: if you setup a pool to mine such blocks, I'm sure you will find miners to donate hashing
488 2012-10-14 18:20:25 <Luke-Jr> and yeah, I'd make them based on the testnet "root" blocks
489 2012-10-14 18:20:51 <sipa> if you can create a test suite that's usable for several clients, and need hashrate for that, i can point +- 1 GH/s to you
490 2012-10-14 18:20:59 <Luke-Jr> (I'm sure because I know when I needed hashing to test things in the past, people helped)
491 2012-10-14 18:21:44 <Luke-Jr> OTOH, if you wait until ASICs take over and block subsidy halves??? lots of GPU miners idle ;)
492 2012-10-14 18:22:06 <Matt_von_Mises> I still need to make my code work on the testnet. I should perhaps do that before anything else.
493 2012-10-14 18:22:31 <sipa> that sounds like the way to go - it's exactly what the testnet was created for
494 2012-10-14 18:23:22 <Matt_von_Mises> Luke-Jr, sipa: I might create a list of blocks and ask for computing power to generate the hashes from willing people and it would be useful for testing multiple implementations then.
495 2012-10-14 18:23:23 <gmaxwell> Yea, a useful thing to do would be to fork testnet for this. I'd gladly give you private keys that I used to have testnet btc on...
496 2012-10-14 18:25:43 <Matt_von_Mises> So basically I can do the same tests but for the testnet and it would be easier to make the blocks for? Is the maximum target for the testnet higher then?
497 2012-10-14 18:26:09 <Matt_von_Mises> I've not looked into the testnet properly yet so I don't know. I know it has some different difficulty rules.
498 2012-10-14 18:26:25 <gmaxwell> You can mine testnet without the difficulty going up, effectively.
499 2012-10-14 18:26:44 <sipa> in particular it has a rule that allows creating a difficulty-1 block if the previous block is older than 20 minutes
500 2012-10-14 18:27:32 <gmaxwell> heh. someplace I have a 2016 block testnet fork that violates the 'time can't go back before the checkpoint rule. I guess I should find that.
501 2012-10-14 18:28:23 <Matt_von_Mises> if (fTestNet && nTime > nTargetSpacing*2)
502 2012-10-14 18:28:58 <Matt_von_Mises> Looks like if the difference is over two hours it reverts to the highest target, which is still the same as the production network.
503 2012-10-14 18:29:23 <sipa> nTargetSpacing is 10 minutes
504 2012-10-14 18:29:37 <sipa> and i don't see why that is the same as the production network? (highest target == difficulty 1)
505 2012-10-14 18:29:42 <Matt_von_Mises> I meant 20 minutes sorry.
506 2012-10-14 18:30:06 <Matt_von_Mises> static CBigNum bnProofOfWorkLimit(~uint256(0) >> 32);
507 2012-10-14 18:30:12 <Matt_von_Mises> That is the highest target right?
508 2012-10-14 18:30:21 <Matt_von_Mises> 0x1D00FFFF I seem to remember
509 2012-10-14 18:30:31 <Matt_von_Mises> Isn't it the same for both networks then?
510 2012-10-14 18:30:36 <sipa> yes, it is
511 2012-10-14 18:31:12 <Matt_von_Mises> Well to create the test blocks, it can be done on the production network with a difficulty of 1.
512 2012-10-14 18:31:21 <sipa> is that a problem?
513 2012-10-14 18:31:29 <sipa> with 10 MH/s, you can make a block in 7 minutes
514 2012-10-14 18:31:41 <Matt_von_Mises> By production network I mean a fork from the genesis block.
515 2012-10-14 18:32:07 <Matt_von_Mises> sipa: What I mean is that the testnet doesn't help much here.
516 2012-10-14 18:32:28 <sipa> if you want to fork the production network after the genesis block, no
517 2012-10-14 18:37:33 <Matt_von_Mises> I've listed some of the blocks that would be needed for testing. I'd need at least 121 blocks which would include 100 blocks needed to test outputs. Though gmaxwell, you said "I'd gladly give you private keys that I used to have testnet btc on" I suppose by forking the testnet at an appropriate point, I can use those private keys to test outputs. Though it would require all the data for the testnet chain at that point which might end up being co
518 2012-10-14 18:40:01 <gmaxwell> you got cut off at co.
519 2012-10-14 18:40:25 <gmaxwell> the testnet chain is quite small, and you really want the first 500 or so blocks anyways, as they have all the 'pass' test cases we have.
520 2012-10-14 18:40:54 <gmaxwell> There is also a point where I added a transaction with a 100,000 TNBTC fee.
521 2012-10-14 18:40:54 <Matt_von_Mises> There are a lot of "co"s in that text. :-)
522 2012-10-14 18:41:01 <gmaxwell> sorry
523 2012-10-14 18:41:24 <gmaxwell> So you could for right before that block and mine 100 valid and use those coins.
524 2012-10-14 18:41:29 <Matt_von_Mises> ???complicated or impractical to code into the tests such that generating hashes for 100 blocks would actually be easier to accomplish.
525 2012-10-14 18:42:05 <gmaxwell> Matt_von_Mises: to generate the hashes so that your node will mine these invalid blocks, and you just ask people to mine against it.
526 2012-10-14 18:42:18 <gmaxwell> You can crank out 100 blocks in a few minutes with just a couple GH/s on it.
527 2012-10-14 18:45:02 <jgarzik> yah, holler if you want my testnet3 wallet including private keys
528 2012-10-14 18:45:26 <jgarzik> 99% generation transactions, if that's relevant
529 2012-10-14 18:46:44 <gmaxwell> Matt_von_Mises: you can also test by disabling the difficulty checks... so that you can do trial runs super fast. And once you're happy with the tests you ask people to mine for you to mine them at a real difficulty so they work against unmodified software.
530 2012-10-14 18:47:11 <sipa> jgarzik: "holler" ?
531 2012-10-14 18:47:20 <gmaxwell> Yell.
532 2012-10-14 18:47:31 <gmaxwell> jgarzik is apparently from the south. :P
533 2012-10-14 18:47:49 <Matt_von_Mises> gmaxwell, jgarzik: Thanks for offering the keys. Though I need to test coinbase maturity and generate at least 100 hashes anyway. In that case I can just use the 5000000000 satoshis from the coinbase for further testing.
534 2012-10-14 18:48:02 <gmaxwell> True.
535 2012-10-14 18:48:11 <jgarzik> holla
536 2012-10-14 18:48:43 <gmaxwell> Matt_von_Mises: though I'd expect you'd create 100 blocks with the same prev... otherwise you won't hit some tests because they'll be rejected after it rejects the first in your fork.
537 2012-10-14 18:49:04 <gmaxwell> so you'd need to do 100 valid blocks to get coins to spend, then you could create 100 forks.
538 2012-10-14 18:49:35 <gmaxwell> If you do all your bad blocks in a row you probably won't get good coverage.
539 2012-10-14 18:49:55 <Luke-Jr> Matt_von_Mises: to get started, you may wish to check out the mining tests I wrote for bitcoind; it has a "compressed" 100+ blocks to get started from
540 2012-10-14 18:56:25 <Matt_von_Mises> Luke-Jr: Yes I'm sure I could use those! I'll take a look next week. Unfortunately for now I've got to go. I'll come back once I've figured out everything including the blocks I'll need generating. Bye and thanks for the input everyone.
541 2012-10-14 18:56:46 <Luke-Jr> ttyl
542 2012-10-14 18:58:37 <sipa> woah - git automatically gc's?
543 2012-10-14 18:58:41 <sipa> never seen it do that
544 2012-10-14 18:59:27 <Luke-Jr> yeah, annoys me since gc fails on my bitcoin repo
545 2012-10-14 18:59:37 <Luke-Jr> so it tries every time
546 2012-10-14 18:59:55 <sipa> fails? :o
547 2012-10-14 18:59:57 <Luke-Jr> I also don't want to gc, so I can look up "dead" commits if I want
548 2012-10-14 19:00:10 <Luke-Jr> sipa: well, it aborts because there are "so many to prune!"
549 2012-10-14 19:00:17 <sipa> ha
550 2012-10-14 19:00:23 <Luke-Jr> I presume it thinks there must be some mistake :P
551 2012-10-14 19:00:36 <Luke-Jr> I'm sure it's because I have numerous revisions of every pullreq in my tree
552 2012-10-14 19:00:39 <Luke-Jr> s/tree/repo
553 2012-10-14 19:03:48 <midnightmagic> also "holler" is used east-coast and east-central english-speaking canada.
554 2012-10-14 19:04:18 <midnightmagic> it was popular amongst older-generation people too
555 2012-10-14 19:16:37 <Diablo-D3> gmaxwell: I wonder how retarded vala is
556 2012-10-14 19:17:34 <Diablo-D3> gmaxwell: because vala almost does a lot of what I want
557 2012-10-14 19:23:54 <Diablo-D3> gmaxwell: then again, vala does a lot of shit no one wants
558 2012-10-14 19:35:28 <jgarzik> git automatically gc's if a certain setting is enabled, and certain limits are exceeded
559 2012-10-14 19:36:00 <jgarzik> "git config --global gc.auto 0" disables
560 2012-10-14 20:03:48 <Luke-Jr> jgarzik: thanks!
561 2012-10-14 20:07:36 <jgarzik> rofl
562 2012-10-14 20:07:47 <jgarzik> Luke-Jr: you're a valued, crazy developer
563 2012-10-14 20:08:09 <Luke-Jr> jgarzik: what? :<
564 2012-10-14 20:09:04 <Luke-Jr> am I wrong that we'd get in trouble for the same things SF would?
565 2012-10-14 20:14:40 <gmaxwell> Luke-Jr: We only have cryptography for authentication; see http://www.bis.doc.gov/encryption/question2.htm
566 2012-10-14 20:15:12 <Luke-Jr> gmaxwell: wait, SF is blocking ALL downloads because of crypto export rules?
567 2012-10-14 20:15:40 <gmaxwell> Luke-Jr: Moverover, see http://en.wikipedia.org/wiki/Bernstein_v._United_States  there has been _no_ enforcement against open source software in modern times.
568 2012-10-14 20:15:45 <gmaxwell> Luke-Jr: yep, thats my understanding!
569 2012-10-14 20:15:55 <Luke-Jr> that's pretty annoying
570 2012-10-14 20:16:01 <copumpkin> ridiculous
571 2012-10-14 20:16:27 <EasyAt> Hey, so I sent some BTC to someone a few days ago while my client was syncing.  It hasn't had any verifications.  Is there a way to rebroadcast or to have my wallet not remove the BTC from the balance?
572 2012-10-14 20:17:03 <Luke-Jr> EasyAt: just leave your client running; if it's still valid, it will rebroadcast randomly
573 2012-10-14 20:17:03 <sipa> EasyAt: your client rebroadcasts unconfirmed transactions approximately once every half hour
574 2012-10-14 20:17:14 <gmaxwell> EasyAt: by syncing... was it syncing because you restored a backup?
575 2012-10-14 20:17:31 <EasyAt> No, I was just days behind... hadn't opened it for awhile
576 2012-10-14 20:17:53 <sipa> EasyAt: are you currently up-to-date?
577 2012-10-14 20:17:55 <sipa> ;;bc,blocks
578 2012-10-14 20:17:56 <gribble> 203295
579 2012-10-14 20:18:02 <EasyAt> But I was thinking, that since I had made paymentrs from the same wallet on a diff computer, perhaps a double spend could have occured?
580 2012-10-14 20:18:05 <EasyAt> sipa: yes
581 2012-10-14 20:18:12 <Luke-Jr> ???
582 2012-10-14 20:18:15 <sipa> EasyAt: in that case, very likely
583 2012-10-14 20:18:24 <Luke-Jr> EasyAt: Bitcoin-Qt and bitcoind don't support using the same wallet from multiple computers
584 2012-10-14 20:18:33 <gmaxwell> Luke-Jr: here is sourceforge's statement, http://sourceforge.net/blog/clarifying-sourceforgenets-denial-of-site-access-for-certain-persons-in-accordance-with-us-law/
585 2012-10-14 20:18:33 <Luke-Jr> that's begging for trouble period
586 2012-10-14 20:18:45 <gmaxwell> Luke-Jr: they annoying don't actually say what laws they're trying to comply with!
587 2012-10-14 20:18:55 <EasyAt> Yes, I agree.  I'm a bad dog
588 2012-10-14 20:19:28 <gmaxwell> Welp, no, you're a screwed dog. Good luck! :P
589 2012-10-14 20:19:30 <Luke-Jr> EasyAt: to remove it from your tx list, backup your current wallet.dat and restore the backup from before you sent it.
590 2012-10-14 20:19:41 <EasyAt> Luke-Jr: ty, I was thinking that
591 2012-10-14 20:19:41 <gmaxwell> Luke-Jr: salvagewallet!
592 2012-10-14 20:19:52 <Luke-Jr> gmaxwell: I've only heard bad things about that so far <.<
593 2012-10-14 20:20:05 <gmaxwell> But if you have a backup which is new enough to not lose anything but old enough to not have the txn, then that works.
594 2012-10-14 20:20:31 <gmaxwell> Luke-Jr: The only thing I saw was someone trying to use it _on_ a file with a bad sector, and yea, it didn't survive that. :P
595 2012-10-14 20:20:35 <EasyAt> Welp, thanks guys.  I can proceed with my life
596 2012-10-14 20:35:05 <MC1984> does the subsidy halve on exactly block 210000
597 2012-10-14 20:35:46 <sipa> block 210000 (which is the 210001th block produced) will have a subsidy of 25 BTC
598 2012-10-14 21:48:58 <bcb> how is it possible that the total of getbalance is greater then the sum of listaccounts?
599 2012-10-14 21:49:45 <jgarzik> bcb: it's a bit technical
600 2012-10-14 21:49:53 <jgarzik> bcb: getbalance gives you the total of all coins in the wallet
601 2012-10-14 21:50:25 <jgarzik> bcb: the accounting system is layered on top of that, a thin layer that sorts the coins into various buckets (accounts)
602 2012-10-14 21:51:02 <jgarzik> bcb: so if you rigorously use the accounting system, it will stay accurate
603 2012-10-14 21:51:36 <jgarzik> bcb: but occasionally you might 'sendtoaddress' rather than 'sendfrom', or some other action that subverts the accounting system layer
604 2012-10-14 21:52:19 <Lachesis> wouldn't that decrease balance without touching the accounts, making getbalance < sum(listaccounts) ?
605 2012-10-14 21:55:33 <bcb> jgarzik:where do the coins come from when you use sentoaddress?
606 2012-10-14 21:56:08 <gmaxwell> Your wallet.
607 2012-10-14 21:56:14 <jgarzik> Lachesis: indeed
608 2012-10-14 21:56:26 <jgarzik> My brain read "greater than" as "different than"
609 2012-10-14 21:56:31 <jgarzik> bcb: ^
610 2012-10-14 21:56:59 <gmaxwell> The account system doesn't make the coins belong to an account, it's an accounting layer.. like marking in your checkbook what funds you spent on food or gas or got from sales or donations.
611 2012-10-14 21:58:03 <Lachesis> i know coins sent to an address generated with getaccountaddress will go into an account, but i thought everything else was debited from or credited to the "" account
612 2012-10-14 21:58:19 <bcb> I guess I should ask which account do the coins come from when you use sendtoaddress?
613 2012-10-14 21:58:27 <bcb> ""?
614 2012-10-14 21:58:37 <gmaxwell> "" yes.
615 2012-10-14 21:59:06 <gmaxwell> Lachesis: you can sendfrom to pick the source account for marking the debt.
616 2012-10-14 22:00:04 <Lachesis> so then how can the sum(listaccounts) != getbalance?
617 2012-10-14 22:00:11 <Lachesis> since "" is in listaccounts?
618 2012-10-14 22:00:28 <Lachesis> on my current wallet, they are equal, but i always use sendfrom and getaccountaddress
619 2012-10-14 22:01:42 <bcb> so if I want to get my "accounts" back in order how do is balance the accounts?
620 2012-10-14 22:01:56 <bcb> I'm moved all the coin back to "" to undo the negative balance
621 2012-10-14 22:02:16 <bcb> however not the get balance is still greater then the "" account
622 2012-10-14 22:02:23 <bcb> *now*
623 2012-10-14 22:07:40 <bcb> where are the missing coins?
624 2012-10-14 22:08:26 <Tril> listunspent - probably in change addresses..
625 2012-10-14 22:09:50 <gmaxwell> Tril: no.
626 2012-10-14 22:09:57 <gmaxwell> bcb: have you imported any keys?
627 2012-10-14 22:10:11 <gmaxwell> Or used that wallet on multiple systems concurrently?
628 2012-10-14 22:10:12 <bcb> yes many
629 2012-10-14 22:10:26 <bcb> i've imported and exported keys
630 2012-10-14 22:10:31 <gmaxwell> yea, so may have missing funds from unidentified change.
631 2012-10-14 22:11:04 <bcb> I was trying to move confirmed coins to a new address, dumpprivkey and import into another bitcoind
632 2012-10-14 22:11:07 <bcb> is that possible
633 2012-10-14 22:11:07 <gmaxwell> The definition of confirmed used by the varrious commands is different too. IIRC one way measures unconfirmed change an one doesn't.
634 2012-10-14 22:13:51 <bcb> "unidentified Change"?
635 2012-10-14 22:13:56 <bcb> how do I prevent that
636 2012-10-14 22:16:36 <bcb> gmaxwell: funds are not missing.  that are just not showing up in any account
637 2012-10-14 22:16:51 <bcb> I think I imported empty keys mostly
638 2012-10-14 22:28:49 <bcb> is there a limit to the number of addresses you can import into bicoind?
639 2012-10-14 22:29:27 <Diablo-D3> yes.