1 2013-08-21 00:29:42 <warren> crap. Multibit doesn't have coin control...
2 2013-08-21 00:34:31 <gavinandresen> warren: it has multiple wallet support, which is better than coin control imho
3 2013-08-21 00:34:59 <gavinandresen> warren: I deleted the 0.8.4rc1 tag, will tag rc2 soon
4 2013-08-21 00:40:25 <warren> gavinandresen: setting up live demos for the talk in 1.5 hours
5 2013-08-21 00:40:39 <warren> I guess I'll use multibit because it's fast.
6 2013-08-21 00:40:55 <CodeShark> what are you demoing, warren?
7 2013-08-21 00:41:22 <warren> CodeShark: explaining bitcoin, payment processors, risks and benefits, technology, regulatory environment to a bunch of local business and law people
8 2013-08-21 00:41:39 <CodeShark> ah, good luck :)
9 2013-08-21 00:41:50 <warren> someone willing to donate me like 0.1 BTC? I'll use it from a new wallet to donate to wikipedia during the demo.
10 2013-08-21 00:43:59 <sipa> warren: sure
11 2013-08-21 00:44:07 <warren> sipa: 19YZ9wzsdjUwwF9aEM9cRh661G4VhVvGnc
12 2013-08-21 00:44:12 <warren> sipa: thanks!
13 2013-08-21 00:45:38 <warren> I tried to sync the full blockchain using 0.8.x but my laptop overheats and kills itself each time =)
14 2013-08-21 00:48:32 <warren> hmm, does multibit not see tx's until it hits a block?
15 2013-08-21 00:49:05 <warren> I see it on blockchain.info but not multibit.
16 2013-08-21 00:51:28 <warren> this is worrisome, multibit still doesn't see it.
17 2013-08-21 00:52:33 <warren> does multibit only see it after 6 confirmations?
18 2013-08-21 00:53:55 <CodeShark> that would not make sense
19 2013-08-21 00:54:11 <warren> yeah, first time i'm using this client, and it isn't seeing the tx that sipa just sent
20 2013-08-21 00:54:22 <warren> I rather not demo using blockchain.info (would encourage its use)
21 2013-08-21 00:54:36 <CodeShark> why not just use Bitcoin-Qt?
22 2013-08-21 00:54:50 <warren> I have trouble syncing the blockchain without laptop overheating =)
23 2013-08-21 00:55:01 <lianj> warren: youre on osx? don't show this ugly multibit ui i see on the screenshots
24 2013-08-21 00:55:11 <warren> lianj: linux
25 2013-08-21 00:55:33 <CodeShark> warren: sync it on another machine and then just copy over the datadir
26 2013-08-21 00:55:43 <warren> there's no time before the demo ...
27 2013-08-21 00:56:10 <warren> ok, I had to resync the chain with multibit and now it see's sipa's tx
28 2013-08-21 00:56:19 <warren> hopefully it works when I demo using a payment processor
29 2013-08-21 00:57:20 <warren> heading out
30 2013-08-21 00:57:32 <CodeShark> have fun :)
31 2013-08-21 01:16:22 <gavinandresen> * [new tag] v0.8.4rc2 -> v0.8.4rc2 ??? I'm firing up my gitian builds
32 2013-08-21 01:40:50 <CobaltBlueD> ugh WHy would AMU Mh/s not directly correlate to U/m over an extended period of time?
33 2013-08-21 02:06:58 <warren> well, I hope multibit works
34 2013-08-21 02:07:01 <warren> presenting soon
35 2013-08-21 02:08:25 <Cusipzzz> whatever you do, don't youtube 'presentation disasters' :)
36 2013-08-21 02:24:20 <warren> Hey, someone know where is that explaination of how the repeated random number allows you to find k?
37 2013-08-21 02:24:22 <warren> need URL
38 2013-08-21 02:26:53 <Cusipzzz> http://www.nilsschneider.net/2013/01/28/recovering-bitcoin-private-keys.html ?
39 2013-08-21 02:28:42 <BlueMatt> Cusipzzz: "This transaction was generated by a hardware bitcoin wallet"
40 2013-08-21 02:29:19 <Cusipzzz> oh, looking for the android one?
41 2013-08-21 02:29:38 <BlueMatt> I was just saying that link wasnt about anything people are using in production
42 2013-08-21 02:31:31 <Cusipzzz> that was the one that woke me up to the potential issue if your randomness sucks
43 2013-08-21 02:32:05 <BlueMatt> my randomness?
44 2013-08-21 02:32:15 <Cusipzzz> anyone's*
45 2013-08-21 02:32:37 <BlueMatt> well, ok, yes
46 2013-08-21 02:32:45 <BlueMatt> and android's...
47 2013-08-21 03:19:52 <phantomcircuit> lol
48 2013-08-21 03:20:12 <phantomcircuit> initial block download on windows 8 is slower than on debian in a vm on windows 8
49 2013-08-21 03:20:15 <phantomcircuit> how is that even possible
50 2013-08-21 03:21:42 <weex> sounds like a rhetorical question
51 2013-08-21 03:24:03 <phantomcircuit> weex, im actually seriously interested in how that's possible
52 2013-08-21 03:24:08 <phantomcircuit> but i suspect i wont ever know
53 2013-08-21 03:45:50 <CodeShark> thread contention issues/priorities?
54 2013-08-21 03:46:33 <CodeShark> less efficient context switching?
55 2013-08-21 03:47:21 <CodeShark> less efficient implementations of core libraries?
56 2013-08-21 05:03:49 <dugo> debian downloaded from the bitcoind on host maybe?
57 2013-08-21 05:28:19 <gavinandresen> Review of 0.8.4 release notes welcome: https://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.8.4/test/
58 2013-08-21 05:29:16 <gavinandresen> SHASUMS and announcement will happen after other gitian builders wake up and we make sure the builds match.
59 2013-08-21 05:31:16 <petertodd> looks good to me
60 2013-08-21 06:52:01 <xeroc> since when does bitcoin use gitian?
61 2013-08-21 06:53:11 <moarrr> hasnt it always?
62 2013-08-21 06:53:47 <phantomcircuit> xeroc, for a long time now a number of people have done gitian builds
63 2013-08-21 06:54:03 <phantomcircuit> it's important to be able to trust the binaries
64 2013-08-21 06:54:11 <phantomcircuit> since most people dont build from source
65 2013-08-21 07:25:56 <xeroc> i see .. good to see :-)
66 2013-08-21 07:26:04 <xeroc> yhea .. shure
67 2013-08-21 07:26:16 <warren> btw, used multibit during the live demo. GAH. I hate it.
68 2013-08-21 07:26:33 <warren> It works, just it makes me nervous.
69 2013-08-21 07:26:40 <xeroc> i remember a chat with some guy who claimed that btc does not use gitian .. some while ago .. maybe he was misinformed :-)
70 2013-08-21 07:26:55 <xeroc> ^.^
71 2013-08-21 08:07:40 <sipa> ;;genrate 71
72 2013-08-21 08:07:41 <gribble> The expected generation output, at 71.0 Mhps, given difficulty of 50810339.0483, is 0.000702738130377 BTC per day and 2.92807554324e-05 BTC per hour.
73 2013-08-21 09:56:14 <Luke-Jr> xeroc: gitian was basically created for bitcoin
74 2013-08-21 09:57:27 <michagogo> Luke-Jr: It was? o_O
75 2013-08-21 09:57:50 <Luke-Jr> michagogo: at least, until very recently nobody else had any interest in using it O.o
76 2013-08-21 09:57:50 <michagogo> Or do you mean "for the kind of usage that bitcoin needs"?
77 2013-08-21 09:58:14 <Luke-Jr> michagogo: I don't know if devrandom had some behind-closed-doors other use for it.
78 2013-08-21 10:03:04 <gjs278> think it actually was created for bitcoin, nobody else I know uses it
79 2013-08-21 10:03:13 <gjs278> when I google gitian all I get back is bitcoin results
80 2013-08-21 10:03:45 <gjs278> when I first saw it mentioned here I thought it was something other projects were using but nope, just this
81 2013-08-21 10:13:34 <c0rw1n> what's gitian?
82 2013-08-21 10:13:45 <c0rw1n> ive seen that word here before
83 2013-08-21 10:13:55 <c0rw1n> ;;google gitian
84 2013-08-21 10:13:56 <gribble> Gitian: a secure software distribution method: <http://gitian.org/>; devrandom/gitian-builder ?? GitHub: <https://github.com/devrandom/gitian-builder>; bitcoin/gitian.sigs ?? GitHub: <https://github.com/bitcoin/gitian.sigs>
85 2013-08-21 10:13:59 <Luke-Jr> c0rw1n: build framework that produces bit-for-bit identical binaries
86 2013-08-21 10:14:15 <c0rw1n> oh fuck yes
87 2013-08-21 10:14:27 <Luke-Jr> c0rw1n: thus, everyone can prove the downloads on bitcoin.org are accurate to the code, and not with some trojan added
88 2013-08-21 10:14:35 <c0rw1n> again : oh fuck yes
89 2013-08-21 10:16:17 <c0rw1n> it waitwhat oh my fsm
90 2013-08-21 10:16:50 <c0rw1n> hm
91 2013-08-21 10:17:42 <c0rw1n> does it check if the binary is correct? if yes, against what?
92 2013-08-21 10:17:55 <c0rw1n> i mean, where does the check come from?
93 2013-08-21 10:18:00 <Luke-Jr> against PGP signatures of everyone else who built it
94 2013-08-21 10:18:20 <Luke-Jr> https://github.com/bitcoin/gitian.sigs
95 2013-08-21 10:18:40 <c0rw1n> wait
96 2013-08-21 10:18:56 <c0rw1n> just that i finish nerdgasming
97 2013-08-21 10:19:57 <kinlo> :p
98 2013-08-21 10:20:04 <c0rw1n> okay i'm calmed down
99 2013-08-21 10:20:06 <c0rw1n> somewhat
100 2013-08-21 10:21:06 <c0rw1n> i don't wanna get capslocked ascii all over the channel
101 2013-08-21 10:21:23 <c0rw1n> finally someone does it right
102 2013-08-21 10:26:16 <c0rw1n> (well the publishing server is still raidable, and hosted -or at least DNSed- in the US, but then : one piece at a time. SOMEone WILL do the distributed browserver thing some day, and gitian exemplifies a method of distributed signage of binaries, which would be needed in that too, in some form, for trust-free operation of distributed web-apps.)
103 2013-08-21 10:39:32 <marcusw> I don't think it's possible for a trust-free webapp...
104 2013-08-21 10:39:41 <marcusw> unless you hosted it locally yourself
105 2013-08-21 10:41:20 <c0rw1n> i mean a trust-free web-app that serves itself from the browser, and the swarm somehow signs it. With a sybim-resistant DHT maintained by invisible pink unicorns and all.
106 2013-08-21 10:42:56 <marcusw> >sybim-resistant DHT
107 2013-08-21 10:43:05 <marcusw> >invisible pink unicorns
108 2013-08-21 10:43:25 <marcusw> those are of equal feasibility
109 2013-08-21 10:43:48 <c0rw1n> yup
110 2013-08-21 10:43:57 <c0rw1n> there are parts of ideas
111 2013-08-21 10:44:08 <c0rw1n> someone smarter than me will figure it out
112 2013-08-21 10:45:34 <c0rw1n> could get a double-recursive homomorphic thing. would be horrifically slow, but might get towards there
113 2013-08-21 10:46:19 <marcusw> I don't know any of these words
114 2013-08-21 10:48:29 <c0rw1n> http://crypto.stanford.edu/craig/craig-thesis.pdf
115 2013-08-21 10:49:02 <michagogo> ;;google gitian -bitcoin -litecoin
116 2013-08-21 10:49:03 <gribble> devrandom/gitian-builder ?? GitHub: <https://github.com/devrandom/gitian-builder>; Gitian: a source-control oriented distribution: <http://gitian.org/howto.html>; gitian | The Tor Blog: <https://blog.torproject.org/category/tags/gitian>
117 2013-08-21 10:49:20 <gribble> gitian | The Tor Blog: <https://blog.torproject.org/category/tags/gitian>; android-secim in Gitian - Gitorious: <http://gitorious.org/gitian/android-secim>; GITian's Secret Confessions Expressions Crushes | Facebook: <https://www.facebook.com/klsgitconfess>
118 2013-08-21 10:49:20 <michagogo> ;;google gitian -bitcoin -litecoin -site:gitian.org -inurl:devrandom
119 2013-08-21 10:49:38 <c0rw1n> yes i've gone and read what gitian is
120 2013-08-21 10:49:57 <c0rw1n> had no idea it was used by all these alreaday
121 2013-08-21 10:50:24 <michagogo> That was a response to gjs278's "think it actually...nope, just this"
122 2013-08-21 10:50:37 <c0rw1n> ah ok
123 2013-08-21 10:50:42 <michagogo> Hah, that tor blod article was posted yesterday
124 2013-08-21 10:50:45 <michagogo> blog*
125 2013-08-21 10:51:04 <c0rw1n> didn't see it gjs's line
126 2013-08-21 10:51:16 <c0rw1n> (still don't, mebbe it was /msg)
127 2013-08-21 10:52:40 <michagogo> [15:02:37] * Joins: c0rw1n
128 2013-08-21 10:52:40 <michagogo> [15:03:04] <gjs[no ping]278> think it actually was...
129 2013-08-21 10:53:04 <michagogo> c0rw1n: It was the first thing you saw after joining
130 2013-08-21 10:53:08 <michagogo> (it's 3 lines)
131 2013-08-21 10:53:15 <c0rw1n> ah that one sry yes
132 2013-08-21 10:53:20 <michagogo> c0rw1n: Also, I know you say it because you reacted to it :-P
133 2013-08-21 10:53:25 <michagogo> s/say/saw/
134 2013-08-21 10:53:52 <darsie> hi
135 2013-08-21 10:54:20 <c0rw1n> i really need to stop doing caffeine in that kind of amounts ... if it's screwing with my memory and reading, it's clearly counter-roductive
136 2013-08-21 10:54:26 <c0rw1n> gah typos
137 2013-08-21 10:54:34 <darsie> Can you please implement an ETA/time to sync display
138 2013-08-21 10:54:49 <c0rw1n> darsie it can't really be done
139 2013-08-21 10:54:56 <c0rw1n> thee's one, sort of, already
140 2013-08-21 10:55:13 <michagogo> darsie: The client can't know how big the blocks are, nor how fast the peer will send them
141 2013-08-21 10:55:59 <darsie> You can average over the last few blocks and estimate with that. Sure it will vary.
142 2013-08-21 10:56:07 <c0rw1n> technically, you could manage your own mining pool, and ensure you send your tx to that
143 2013-08-21 10:56:26 <c0rw1n> might be just a little expensive
144 2013-08-21 10:56:40 <michagogo> darsie: You can't really, though
145 2013-08-21 10:56:51 <michagogo> Since you might get several small blocks and then several big blocks
146 2013-08-21 10:57:05 <michagogo> Especially when doing IBD
147 2013-08-21 10:57:06 <darsie> I did it manually, just yesterday.
148 2013-08-21 10:57:18 <darsie> or average over the last few minutes.
149 2013-08-21 10:58:50 <darsie> I did:
150 2013-08-21 10:58:52 <darsie> echo >>x `date +%s` 252531
151 2013-08-21 10:59:00 <darsie> cat x
152 2013-08-21 10:59:09 <c0rw1n> i think you'd want to average over the period of difficulty adjustment, so it follows the network : faster when a lot of hash is deployed, slower if large hashpower disappear
153 2013-08-21 10:59:10 <darsie> echo 'scale=3;(1481-0829)/(2531-2394)*(3269-2531)/60'|bc
154 2013-08-21 11:00:26 <darsie> What is IBD?
155 2013-08-21 11:02:40 <michagogo> Initial Block Download
156 2013-08-21 11:03:04 <darsie> You mean the first part of the block chain loads much faster?
157 2013-08-21 11:03:30 <michagogo> (BTW, does anyone know if that term also applies to the catching up when you start a node, or if it's only referring to the very first download for a new node?)
158 2013-08-21 11:05:37 <Luke-Jr> darsie: it does
159 2013-08-21 11:06:24 <darsie> Yes, it does. That would give a very off ETA, indeed.
160 2013-08-21 11:08:26 <darsie> The ETA display could be an option to activate in the preferences, off by default and with a note about its inaccuracy, especially for IBD.
161 2013-08-21 11:09:24 <darsie> But every ETA is just an *estimate*. When you d/l a torrent, the speed varies, too. Still they all show an ETA. Initially the ETA shows days, then goes down to hours.
162 2013-08-21 11:09:57 <darsie> It gives you a ballpark figure, which is useful.
163 2013-08-21 11:13:38 <darsie> And btw, your display of downloaded block count informs the user about the progress. They then estimate the TA in their head. Show computed ETA, too. It's what you intended by your display, just more useful.
164 2013-08-21 11:14:09 <michagogo> "It's what you intended by your display" <-- Why do you say that?
165 2013-08-21 11:14:32 <darsie> It appears conclusive to me.
166 2013-08-21 11:14:46 <michagogo> For example, telling the user how far behind they are can be good for the user to know what they should expect to see in terms of older transactions
167 2013-08-21 11:14:52 <marcusw> users will be much less mad about an ETA they come up with being wrong than one bitcoin gave them being wrong
168 2013-08-21 11:15:46 <marcusw> that being said, network transfer time estimation (especially on p2p networks) is expected to be a crappy estimate at best
169 2013-08-21 11:16:04 <marcusw> think of it like a torrent client...that's what the users expect
170 2013-08-21 11:29:34 <runeks> darsie: The problem with comparing it to a torrent download is that the first 1000 blocks will take like 10 seconds while the last 1000 blocks might take 1000 seconds. So your estimate would be off by a factor of 100. That's not useful in any way. Quite the opposite.
171 2013-08-21 11:30:17 <marcusw> runeks: so turn that into an equation and scale it appropriately
172 2013-08-21 11:30:57 <darsie> runeks: That can be improved by computing the ETA based on the last few blocks/minutes. And printing a note about the inaccuracy of the ETA in the preferences where the ETA option is off by default.
173 2013-08-21 11:31:27 <runeks> marcusw: You mean benchmark an IBD and bake the constants into bitcoin-qt?
174 2013-08-21 11:31:59 <pigeons> cool darsie, make a pull request when you're done coding it
175 2013-08-21 11:32:10 <marcusw> runeks: why not?
176 2013-08-21 11:34:26 <runeks> marcusw: I guess I think it seems clumsy and unnecessary. At least we should wait until it's possible to do an IBD from multiple nodes. When only using a single node it becomes completely impossible to calculate an ETA. Torrents are fairly constant in speed when the node count is over 100. But with Bitcoin, you basically only have a single connection.
177 2013-08-21 11:35:07 <michagogo> Hopefully sipa's headers-first will be ready soon
178 2013-08-21 11:36:38 <runeks> michagogo: Does it get all headers from a single node and then tries to sync full blocks from multiple nodes?
179 2013-08-21 11:36:47 <michagogo> runeks: Ask sipa
180 2013-08-21 11:36:54 <michagogo> (or look at his github fork)
181 2013-08-21 11:37:45 <arioBarzan> I am running bitcoind 0.2.0 on two locally connected nodes, but couldn't mine. In version 0.3.15 Satoshi made sure that generation doesn't start before block 74000 downloaded. Is there also any similar limitation on 0.2.0 ?
182 2013-08-21 11:37:50 <runeks> Looks like that's the plan: https://github.com/sipa/bitcoin/commit/81899f54fd26505a60cb84f51b4ff40c045f4be6
183 2013-08-21 11:37:58 <michagogo> 0.2.0? o_O
184 2013-08-21 11:38:00 <michagogo> Why?
185 2013-08-21 11:38:02 <c0rw1n> what the fuck are you doing with those versions?
186 2013-08-21 11:38:16 <c0rw1n> forking from old code?
187 2013-08-21 11:38:30 <arioBarzan> I need just some fake coins for doing some tests, just for learning.
188 2013-08-21 11:38:40 <michagogo> arioBarzan: What's wrong with testnet?
189 2013-08-21 11:38:46 <c0rw1n> just get testnet-in-a-box
190 2013-08-21 11:38:52 <michagogo> Just start bitcoind or bitcoin-qt with -testnet
191 2013-08-21 11:38:57 <c0rw1n> easier and will probably work
192 2013-08-21 11:38:57 <michagogo> Or what c0rw1n said
193 2013-08-21 11:39:34 <c0rw1n> ;;google bitcoin "testnet in a box"
194 2013-08-21 11:39:34 <gribble> freewil/bitcoin-testnet-box ?? GitHub: <https://github.com/freewil/bitcoin-testnet-box>; Testnet - Bitcoin: <https://bitcoin.it/wiki/Testnet>; Bitcoin - Browse /Bitcoin/testnet-in-a-box at SourceForge.net: <http://sourceforge.net/projects/bitcoin/files/Bitcoin/testnet-in-a-box/>
195 2013-08-21 11:39:50 <arioBarzan> could I compile new versions (e.g.) and strip them of their mining limitation rules ?
196 2013-08-21 11:40:33 <darsie> arioBarzan: I can give you some testnet coins. Which address?
197 2013-08-21 11:40:36 <arioBarzan> I wonder how Satoshi could mine with 0.2.0 but I couldn't.
198 2013-08-21 11:41:26 <Zoop_> he mined more than most of us together
199 2013-08-21 11:41:33 <marcusw> plot twist: he was actually premining on unshipped BFL asics
200 2013-08-21 11:41:36 <runeks> arioBarzan: If you just want to test something then use testnet.
201 2013-08-21 11:41:38 <arioBarzan> I Have used testnet coins. They are good for learning. But I assume one should be able to fork his local blockchain
202 2013-08-21 11:41:55 <runeks> marcusw: At 100KH/s
203 2013-08-21 11:42:28 <marcusw> what a high roller
204 2013-08-21 11:42:49 <Zoop_> ahah
205 2013-08-21 11:43:34 <CobaltBlueD> I've been narrowing down the cause on an interesting symptom here. eruptors running @ 330Mh/s, cgminer reporting that. The reason I am seeing poor rates with this particular setup is long delays cause extended idle time betweeeeen the submissions.
206 2013-08-21 11:44:43 <CobaltBlueD> anyone have ideas on what would cause that? Hardware looks fine, voltage is good, same version of cgminer works fine with these chips on other machines.
207 2013-08-21 11:45:10 <runeks> Actually, scratch that. Looks like the average network speed for the first year was a whopping 4 MH/s. Not 100KH/s.
208 2013-08-21 11:45:11 <arioBarzan> could I disable IsInitialBlockDownload() in the code and compile bitcoind again to be able to mine some coins in a local network?
209 2013-08-21 11:45:40 <runeks> arioBarzan: What is it you want to do exactly?
210 2013-08-21 11:45:57 <c0rw1n> if he does that it's basically altcoin-in-a-box no?
211 2013-08-21 11:46:10 <arioBarzan> I want to mine a valid block at the height of for e.g. 200
212 2013-08-21 11:46:23 <arioBarzan> but on my pair of local nodes
213 2013-08-21 11:47:16 <arioBarzan> I tried bitcoin 0.2.0 after one hour of mining bitcoin found NO new block.
214 2013-08-21 11:47:31 <c0rw1n> ooh you could feed it a known set of tx
215 2013-08-21 11:47:47 <c0rw1n> good luck with doing it, but might work
216 2013-08-21 11:48:00 <arioBarzan> unrotunately 0.2.0 did not have rpc
217 2013-08-21 11:48:10 <c0rw1n> it must get the tx from somewhere
218 2013-08-21 11:49:05 <c0rw1n> if you're willing to go that far, you can take apart the c++ code and call it from a kludged buttonpad GUI
219 2013-08-21 11:49:18 <arioBarzan> c0rw1n: could I disable IsInitialBlockDownload() in current version to be able to mine at the height of 200?
220 2013-08-21 11:49:25 <c0rw1n> no idea
221 2013-08-21 11:50:01 <CobaltBlueD> so my question to any devs floating around, what has to run after a submission and before the new work is sent to the ASIC. it's using stratum of course.
222 2013-08-21 11:51:17 <arioBarzan> Satoshi was mining in first 120 blocks, but he couldn't possibly spend them. so it wouldn't be necessary to have a tx to be able to mine at that height.
223 2013-08-21 11:52:10 <c0rw1n> why couldn't he spend them? he could have made addresses and sent the coinbases between those
224 2013-08-21 11:52:33 <arioBarzan> either 0.2.0 was a fake ( so only satoshi could mine possibly with another version) or I am doing something wrong that I couldn't mine by 0.2.0
225 2013-08-21 11:52:35 <c0rw1n> not that i've gone and traced the first block mind you
226 2013-08-21 11:53:05 <arioBarzan> c0rw1n: he had to wait 120 blocks after generating new coin.
227 2013-08-21 11:53:17 <c0rw1n> ooh *click* i'm an idioty
228 2013-08-21 11:54:26 <c0rw1n> ah but then do you get the same blockchain every time? I mean the same hashes from the same "first 120, no-tx" blocks
229 2013-08-21 11:54:50 <c0rw1n> or is it time-sensitive? i really have no idea
230 2013-08-21 11:54:58 <c0rw1n> i mean date-sensitive
231 2013-08-21 11:56:27 <arioBarzan> it shouldn't be date-sensitive. The two-connected nodes have no more blocks after 175, and their mining difficulty is 1
232 2013-08-21 11:56:49 <CobaltBlueD> I have a bad feeling I'm going to be looking at code today
233 2013-08-21 12:00:04 <jgarzik> mornin'
234 2013-08-21 12:01:12 <CobaltBlueD> morning. definitely time for coffee.
235 2013-08-21 12:01:54 <arioBarzan> I am running bitcoind 0.2.0 on two locally connected nodes, but couldn't mine. In version 0.3.15 Satoshi made sure that generation doesn't start before block 74000 downloaded (https://en.bitcoin.it/wiki/Changelog#0.3.15.5B24.5D). Is there also any similar limitation on 0.2.0 ?
236 2013-08-21 12:08:20 <marcusw> man...add a bit about heaven and 2nd coming and satoshi = jesus
237 2013-08-21 12:08:55 <marcusw> japanese Jesus Chrypto
238 2013-08-21 12:08:59 <c0rw1n> Satoshi is the Cyberchrist
239 2013-08-21 12:09:04 <c0rw1n> or crypstomessiah
240 2013-08-21 12:09:15 <c0rw1n> (gah typos grr)
241 2013-08-21 12:09:49 <marcusw> >mfw satoshi was actually jesus
242 2013-08-21 12:16:40 <michagogo> How long ago has there been a 0.8.4rc2?
243 2013-08-21 12:17:23 <c0rw1n> that's not listed right next to the release?
244 2013-08-21 12:18:16 <michagogo> c0rw1n: I just happened to notice "v0.8.4rc2" in github's tags list when looking for 0.2.0 to check for checkpoints
245 2013-08-21 12:18:26 <c0rw1n> ah ok
246 2013-08-21 12:18:34 <michagogo> Anyway, looks like it was tagged yesterday
247 2013-08-21 12:18:40 <michagogo> Ooh, with a new checkpoint
248 2013-08-21 12:18:51 <gribble> 3412
249 2013-08-21 12:18:51 <michagogo> ;;calc [blocks] - 250000
250 2013-08-21 12:18:57 <michagogo> 3412 blocks ago
251 2013-08-21 12:19:43 <vrs> has anybody in here successfully built bitcoin on freebsd with clang?
252 2013-08-21 12:19:55 <vrs> I'm getting gmake[1]: *** No rule to make target `clang'. Stop.
253 2013-08-21 12:21:44 <Luke-Jr> CobaltBlueD: upgrade to BFGMiner
254 2013-08-21 12:22:18 <Luke-Jr> arioBarzan: 0.2.0 isn't compatible with the current Bitcoin protocol, note
255 2013-08-21 12:23:08 <arioBarzan> Luke-Jr: I am running my own set of nodes
256 2013-08-21 12:23:17 <Luke-Jr> vrs: BSD isn't really supported.
257 2013-08-21 12:23:21 <arioBarzan> Luke-Jr: but I can not generate blocks
258 2013-08-21 12:23:49 <vrs> Luke-Jr: ...aw
259 2013-08-21 12:24:12 <Luke-Jr> vrs: would you like to be the BSD team? :P
260 2013-08-21 12:24:19 <vrs> uh
261 2013-08-21 12:24:27 <vrs> let me gain a few levels in C and then perhaps
262 2013-08-21 12:24:30 <Luke-Jr> :p
263 2013-08-21 12:24:36 <vrs> err C/C++
264 2013-08-21 12:24:48 <Luke-Jr> vrs: the error suggests you ran 'make clang' or something
265 2013-08-21 12:24:52 <Luke-Jr> which is wrong in any case
266 2013-08-21 12:24:59 <vrs> yes, and that's confusing me
267 2013-08-21 12:25:14 <vrs> because it was a simple portmaster invocation and nowhere was making clang involved
268 2013-08-21 12:25:16 <Luke-Jr> try the documented build steps for Linux
269 2013-08-21 12:25:22 <vrs> might just be a problem of the port
270 2013-08-21 12:27:55 <arioBarzan> Luke-Jr: If I disable IsInitialBlockDownload() could I generate coins on locally-connected-nodes at height for e.g. 200 ?
271 2013-08-21 12:28:03 <michagogo> Is 0.8.4 mostly a bugfix release?
272 2013-08-21 12:28:25 <Luke-Jr> arioBarzan: perhaps
273 2013-08-21 12:28:34 <Luke-Jr> arioBarzan: note that 0.2.0 was likely long after block 200
274 2013-08-21 12:28:42 <Luke-Jr> michagogo: yeah, just trivial things
275 2013-08-21 12:30:40 <michagogo> Brb, going to find my external hard drive with Ubuntu on it so I can gitian-build
276 2013-08-21 12:31:07 <vrs> btw regarding the bootstrap.dat torrent, I seeded 371690.7MB of it
277 2013-08-21 12:31:29 <michagogo> I've seeded 34.6 GB
278 2013-08-21 12:31:33 <vrs> which is a ratio of roughly 46
279 2013-08-21 12:31:59 <michagogo> (is jgarzik going to update bootstrap for 250k?)
280 2013-08-21 12:33:03 <michagogo> Also: How come gitian.sigs has sigs for 0.8.4rc1, but that tag doesn't seem to exist?
281 2013-08-21 12:33:34 <Luke-Jr> was it deleted? :o
282 2013-08-21 12:34:59 <petertodd> it was, rc1 has a show-stopper bug
283 2013-08-21 12:37:01 <michagogo> What bug?
284 2013-08-21 12:37:31 <petertodd> just a thinko related to the bloom patch that made android wallets not sync up
285 2013-08-21 12:38:06 <petertodd> problem was multibit synced, and that's all greg had available to test with, where as myself all I had was an android wallet...
286 2013-08-21 12:38:26 <petertodd> next time I'll screw up and it'll be gmaxwell finding that I broke multibit :/
287 2013-08-21 13:32:56 <michagogo> Gitian build done.
288 2013-08-21 13:32:59 <michagogo> PR'd.
289 2013-08-21 13:39:01 <Ry4an> I hope that's a homestarrunner reference :)
290 2013-08-21 13:41:21 <michagogo> Hmm?
291 2013-08-21 13:41:30 <Ry4an> I guess not. :)
292 2013-08-21 13:41:53 <Ry4an> http://www.hrwiki.org/wiki/Arrow'd_Guy
293 2013-08-21 13:42:18 <michagogo> ACTION has never heard of Homestar Runner
294 2013-08-21 13:42:48 <Luke-Jr> it's .com
295 2013-08-21 13:43:04 <michagogo> (wouldn't http://www.hrwiki.org/wiki/%27d be better?)
296 2013-08-21 13:43:31 <Ry4an> ah, it was an early 2000s online funny thing. There was a recurring character that VERB'd people to death for various types of verb. Was popular amongst geeks at the time, so I thought you were invoking it. Not a big deal.
297 2013-08-21 13:43:42 <michagogo> Yeah, nope :-P
298 2013-08-21 13:44:06 <michagogo> Anyone have any idea on release estimate for 0.8.4? (assuming no show-stoppers)
299 2013-08-21 14:20:41 <runeks> For anyone interested: I've updated my text-only version of historic Bitcoin difficulties to include an estimate of the next difficulty adjustment (estimations are marked with an asterisk): http://runeks.dk/bitcoin/diff.txt
300 2013-08-21 14:23:42 <darsie> Does the bitcoin address include the public key?
301 2013-08-21 14:24:07 <edcba> hash of it
302 2013-08-21 14:24:10 <edcba> iirc :)
303 2013-08-21 14:24:57 <vrs> darsie: it does not and I think it's a ripemd-160 hash
304 2013-08-21 14:26:01 <vrs> https://en.bitcoin.it/wiki/Protocol_specification#Addresses
305 2013-08-21 14:35:35 <handle> darsie: no
306 2013-08-21 14:35:43 <handle> it is a hash of the public key
307 2013-08-21 14:35:47 <darsie> thx
308 2013-08-21 14:36:06 <handle> so if you have the address, you can verify the correct public key, however you cannot get the actual public key from it
309 2013-08-21 14:36:25 <gmaxwell> kinda sad that darsie is getting worse advice in here than they got in #bitcoin ...
310 2013-08-21 14:36:58 <darsie> well, I didn't pursue the purpose here, since I already got it.
311 2013-08-21 14:38:33 <handle> gmaxwell: is my advice bad?
312 2013-08-21 14:38:43 <gmaxwell> Yea, kinda.
313 2013-08-21 14:39:08 <handle> 00:25 < darsie> Does the bitcoin address include the public key?
314 2013-08-21 14:39:16 <arioBarzan> handle: no
315 2013-08-21 14:39:19 <handle> I think my answer accurately answers that question - it doesn't include the public key
316 2013-08-21 14:39:28 <gmaxwell> It's not pedantically correct, depending on whats meant by public key, and it didn't solve darsie's actual problem (actually it would have given the opposite from correct behavior.
317 2013-08-21 14:39:34 <gmaxwell> 09:25 < gmaxwell> darsie: You should probably ask more of your actual question.
318 2013-08-21 14:39:37 <gmaxwell> 09:26 < gmaxwell> darsie: Because I can't actually give you a good answer with what you gave.
319 2013-08-21 14:39:40 <gmaxwell> 09:27 < gmaxwell> (A typical 1 style address is a hash160 of a ecdsa public key. But the hash160 itself is effectively a public key for signing validation)
320 2013-08-21 14:39:43 <gmaxwell> 09:27 < darsie> Well, I want to make a backup of my wallet on paper, but the wallet is too big. So I wondered if printing address + private key suffices.
321 2013-08-21 14:39:46 <gmaxwell> 09:27 < gmaxwell> darsie: saving the private keys suffice, all the rest can be derrived from that.
322 2013-08-21 14:40:03 <handle> ah
323 2013-08-21 14:42:58 <arioBarzan> I'm running bitcoin 0.2.0 on two locally connected nodes, siolated from internet. I couldn't generate any new block. Satoshi blocked generating blocks before 74000 in version 0.3.15 (https://en.bitcoin.it/wiki/Changelog#0.3.15.5B24.5D) is there any similar checkpoint on 0.2.0 ?
324 2013-08-21 14:43:08 <gmaxwell> handle: The address itself _is_ a public key. It's not an ecc public key, but it's public data which you can use to verify knoweldge of private data. One completely valid way (and the way the blockchain works for 1 type addresses) is that the address is the public key (thing you identify payees with), and the ecdsa-public key is just part of the signature (data the payee provides to prove their identity)
325 2013-08-21 14:44:50 <edcba> darsie about backup remember each time a transaction is issued new key is issued
326 2013-08-21 14:45:28 <darsie> edcba: I doubt that.
327 2013-08-21 14:45:34 <edcba> i mean backuping bitcoin keys suck
328 2013-08-21 14:46:03 <handle> actualy, there's a pool of 100 keys when you first start up bitcoin, IIRC
329 2013-08-21 14:46:08 <runeks> Armory works really well for backing up a wallet on paper IMO.
330 2013-08-21 14:46:12 <handle> and when you exhaust those 100 keys, it generates nore
331 2013-08-21 14:46:15 <handle> more*
332 2013-08-21 14:46:50 <arioBarzan> From where could I download bitcoin 0.1.0 ?
333 2013-08-21 14:47:05 <edcba> (internet)
334 2013-08-21 14:50:57 <edcba> didn't find the forum link about the guy having lost all his bitcoins still having a backup of his "old style" wallet :)
335 2013-08-21 14:54:12 <runeks> arioBarzan: http://www.zorinaq.com/pub/bitcoin-0.1.0.tgz
336 2013-08-21 14:54:29 <arioBarzan> runeks: tnx
337 2013-08-21 14:54:49 <edcba> why that version btw ?
338 2013-08-21 14:55:12 <runeks> arioBarzan: Bear in mind that it won't communicate with other nodes on the network. And it's Windows-only.
339 2013-08-21 14:55:35 <runeks> I mean, it will communicate with instances of itself, but not with the general Bitcoin network.
340 2013-08-21 14:56:11 <runeks> Because: http://bitcoin.stackexchange.com/questions/2788/what-are-the-february-20-changes-to-the-bitcoin-protocol
341 2013-08-21 14:56:19 <arioBarzan> runeks: Hopefully it would generate a block at difficulty 1 which is the main thing I am looking for
342 2013-08-21 14:56:44 <arioBarzan> I will run it on isolated nodes
343 2013-08-21 14:57:59 <gmaxwell> arioBarzan: I'm not aware of any explicit inhibition, though the oldest mining code is really slow.
344 2013-08-21 14:59:49 <arioBarzan> gmaxwell: is there any not-so-difficult way that I could disable checkpoints for generating blocks at height for e.g. 200 at the current bitcoin version and then compile it again?
345 2013-08-21 15:01:08 <gmaxwell> arioBarzan: what exactly are you trying to do?
346 2013-08-21 15:02:01 <arioBarzan> gmaxwell: I want to generate a new block on two isolated nodes, locally-connected to each other
347 2013-08-21 15:02:34 <arioBarzan> gmaxwell: I tried 0.2.0 and after several hours I was unsuccessful
348 2013-08-21 15:02:58 <arioBarzan> although the difficulty was just 1
349 2013-08-21 15:04:09 <edcba> arioBarzan: i think you can use any client and 2 instances of them in an isolated network
350 2013-08-21 15:04:25 <edcba> or hmm maybe not because of some hardcoded checks...
351 2013-08-21 15:05:03 <arioBarzan> edcba: exactly, not possible because of checkpoints https://en.bitcoin.it/wiki/Changelog#0.3.15.5B24.5D
352 2013-08-21 15:05:19 <edcba> thanks gavinandresen...
353 2013-08-21 15:05:47 <edcba> just adds complexity for nothing imo
354 2013-08-21 15:06:00 <gmaxwell> arioBarzan: just use regular testnet and isolate the nodes.
355 2013-08-21 15:06:18 <gmaxwell> arioBarzan: you'll get a block after 20 minutes plus whatever it takes you to mine a diff 1 block.
356 2013-08-21 15:09:25 <arioBarzan> gmaxwell: everybody today just said the sam that I use testnet. I wonder how Satoshi mined those blocks.
357 2013-08-21 15:10:19 <edcba> using his cpu ?
358 2013-08-21 15:11:02 <gmaxwell> arioBarzan: huh? _lots_ of people were using bitcoin by 0.2.
359 2013-08-21 15:12:20 <arioBarzan> gmaxwell: did 0.2 have checkpoints to prevent one from generating blocks at low height?
360 2013-08-21 15:14:29 <gmaxwell> 09:57 < gmaxwell> arioBarzan: I'm not aware of any explicit inhibition, though the oldest mining code is really slow.
361 2013-08-21 15:14:56 <gmaxwell> I'm also not about to go digging through the 0.2 code to find out, especially because what you're trying to do has nothing to do with 0.2.
362 2013-08-21 15:15:05 <gmaxwell> If you mine on testnet, I know that that works.
363 2013-08-21 15:15:58 <arioBarzan> gmaxwell: thanks
364 2013-08-21 15:44:53 <jgarzik> michagogo, yes
365 2013-08-21 15:48:06 <michagogo> ACTION goes to find what he asked jgarzik
366 2013-08-21 15:48:14 <michagogo> Ah, right.
367 2013-08-21 15:48:47 <gaantr2> Anyone here know anything about bitcoin-j?
368 2013-08-21 15:51:06 <sipa> gaantr2: talk to TD or BlueMatt i guess
369 2013-08-21 15:51:08 <sipa> or Goonie
370 2013-08-21 15:51:39 <Goonie> heard my name
371 2013-08-21 15:58:53 <gaantr2> sipa: thanks
372 2013-08-21 15:59:09 <gaantr2> Goonie: do you know how to recover 0.7 BTC from a bitcoin-j wallet?
373 2013-08-21 15:59:32 <sipa> gaantr2: android wallet, multibit, or something else?
374 2013-08-21 15:59:41 <Goonie> gaantr2: load it into wallet-tool, create a spend
375 2013-08-21 16:00:00 <Goonie> gaantr2: you can probably use multibit as well
376 2013-08-21 16:00:07 <gaantr2> okay I will try multibit
377 2013-08-21 16:00:41 <Goonie> gaantr2: how did you create the wallet?
378 2013-08-21 16:01:28 <gaantr2> I created it with bitcoinj
379 2013-08-21 16:02:05 <gaantr2> I had someone send 0.7 BTC to an address on the wallet, but for whatever reason it never seemed to confirm in the wallet
380 2013-08-21 16:02:27 <gaantr2> the transaction is confirmed: https://blockchain.info/address/1Ee9KBg1VtC8yHds8F8YLcxJDktppHUvnt
381 2013-08-21 16:03:12 <Goonie> gaantr2: then you have a good incentive to fix your app (-:
382 2013-08-21 16:08:03 <michagogo> How long does it usually take from a new release version being tagged until it goes live for end-user download?
383 2013-08-21 16:22:00 <gaantr2> Goonie: Yep. The issue is difficult to pin point. Seems like after it has been running for a while it "freezes" and no longer processes coin transactions sent to the program. It is based on the ping app
384 2013-08-21 16:22:17 <gaantr2> I have not been able to fix this particular bug yet.
385 2013-08-21 16:24:44 <michagogo> Question: I found a utility that allows me to close (a) TCP connection(s). Wireshark tells me that this is done with a RST,ACK packet.
386 2013-08-21 16:25:28 <michagogo> If IBD is stuck, will using getpeerinfo to locate the syncnode and closing that connection unstick it?
387 2013-08-21 16:26:38 <marcusw> should just be RST, not RSTACK, but yeah
388 2013-08-21 16:28:35 <gaantr2> yay! I got my coins back lol
389 2013-08-21 16:28:59 <gaantr2> Goonie++ (bitcoin-dev++)
390 2013-08-21 16:29:35 <gaantr2> I guess I should also give a multibit++
391 2013-08-21 16:29:51 <michagogo> It would be cool if there were somewhere where hex dumps of blocks were easily accessible to be pasted into the console following "submitblock "
392 2013-08-21 16:31:38 <helo> michagogo: ./bitcoind getblock <block hash> false
393 2013-08-21 16:32:35 <michagogo> helo: Is there an always-on bitcoind node with its RPC port open somewhere that I could pull that from? :P
394 2013-08-21 16:32:51 <helo> prawlie not
395 2013-08-21 16:32:55 <michagogo> (I was kidding)
396 2013-08-21 16:35:50 <marcusw> why doesn't anyone run one of those?
397 2013-08-21 16:36:08 <helo> they want people to run their own to halp the network
398 2013-08-21 16:36:17 <michagogo> marcusw: Actually, I think it could be done now that there's a nowallet patch
399 2013-08-21 16:36:46 <marcusw> empty wallet would be fine anyway
400 2013-08-21 16:38:08 <michagogo> Really looking forward to sipa's headers-first, so that one peer won't stall IBD
401 2013-08-21 17:05:17 <arioBarzan> is there any way to circumvent mining checkpoints ? I want to mine a new block after genesis block in a network of locally connected nodes, isolated from internet.
402 2013-08-21 17:05:44 <Krellan> Luke-Jr: I glanced at bfgminer and somehow it printed (.100%) in the "R" column, thus throwing it off of alignment. Looked through percentf2() and don't see how that would be possible, it's correct. Maybe weird floating-point rounding up?
403 2013-08-21 17:07:10 <Krellan> Really nice column alignment btw, the page looks rather elegant in a large terminal window.
404 2013-08-21 17:07:44 <gmaxwell> arioBarzan: Checkpoints aren't whats stopping you in any case.
405 2013-08-21 17:08:37 <gmaxwell> arioBarzan: though as an aside, you can turn them off with the @#$@# commandline parameter which is right there for it. And I advised you above to use testnet.
406 2013-08-21 17:10:39 <runeks> helo: Does that command work for you? All I get is an error. I'm guessing you're running master and not a stable release?
407 2013-08-21 17:11:22 <michagogo> runeks: I think you mean HEAD?
408 2013-08-21 17:11:56 <michagogo> runeks: And yeah, it doesn
409 2013-08-21 17:12:04 <runeks> michagogo: Not sure. Doesn't the master branch contain the newest patches?
410 2013-08-21 17:12:05 <michagogo> runeks: And yeah, it doesn't seem to be working for me on 0.8.4rc2
411 2013-08-21 17:12:23 <michagogo> runeks: In general, more or less (though only those that have been merged)
412 2013-08-21 17:12:58 <helo> yeah, i'm on 8fa9b5cc45c1826b...
413 2013-08-21 17:13:44 <michagogo> But the current state of the master branch (actually, I think it's whatever you happen to have checked out at the moment, but it usually is used in this context to refer to the master branch's most recent commit) is known as HEAD
414 2013-08-21 17:14:17 <runeks> I see. So 'master' is a branch and HEAD is a certain point (commit) in that branch?
415 2013-08-21 17:16:18 <michagogo> runeks: http://en.wikipedia.org/wiki/Revision_control
416 2013-08-21 17:16:36 <michagogo> Also sometimes called tip, this refers to the most recent commit, either to the trunk or to a branch. The trunk and each branch have their own head, though HEAD is sometimes loosely used to refer to the trunk.[7]
417 2013-08-21 17:16:36 <michagogo> Head
418 2013-08-21 17:16:58 <runeks> Well thank you. TIL.
419 2013-08-21 17:17:02 <warren> The talk went extremely well. ~36 people who never used Bitcoin before learned a lot.
420 2013-08-21 17:17:14 <gmaxwell> warren: was it recorded?
421 2013-08-21 17:17:16 <warren> Someone else brought a Bitcoin ATM to demo.
422 2013-08-21 17:17:18 <michagogo> [7] == http://garygregory.wordpress.com/2011/02/03/trunk-vs-head-in-version-control-systems/
423 2013-08-21 17:17:25 <warren> gmaxwell: only photos I think
424 2013-08-21 17:17:29 <michagogo> warren: Which talk?
425 2013-08-21 17:17:50 <warren> michagogo: small local talk in the middle of nowhere
426 2013-08-21 17:18:01 <michagogo> I see.
427 2013-08-21 17:18:07 <michagogo> What was the content of the talk?
428 2013-08-21 17:18:20 <warren> I donated sipa's money to Wikipedia during the demo to demonstrate payment processors.
429 2013-08-21 17:18:30 <gmaxwell> warren: as an aside??? I almost always record myself (just using my laptop) when I give a talk. I think doing so helped me improve my speaking style quite a bit.
430 2013-08-21 17:18:31 <michagogo> Lol
431 2013-08-21 17:18:35 <michagogo> (with sipa
432 2013-08-21 17:18:40 <michagogo> (with sipa's consent, I assume?)
433 2013-08-21 17:18:45 <warren> yes
434 2013-08-21 17:18:54 <michagogo> Also: didn't know they take bitcoins
435 2013-08-21 17:18:59 <gmaxwell> haha "and now I demostrate wallet hacking"
436 2013-08-21 17:19:05 <warren> hehehe
437 2013-08-21 17:19:16 <warren> "Here's a flaw in secp256k1 ..."
438 2013-08-21 17:19:18 <gmaxwell> michagogo: they don't, but bitpay will accept bitcoins and pay to wikipedia.
439 2013-08-21 17:19:28 <michagogo> Interesting.
440 2013-08-21 17:20:02 <tgs3> how is land of the free progressing on making some legal trouble @ NY bitcoiners?
441 2013-08-21 17:20:10 <michagogo> Is that in collaboration with wikipedia, or just you giving it to them to give to wikipedia?
442 2013-08-21 17:20:12 <warren> I had no time to prep a slide deck, so it was a conversation that went mostly smoothly
443 2013-08-21 17:20:15 <gmaxwell> warren: alas, if only I'd been giving a talk on any of the days when I triggered amusing bc.i behavior. "And now??? I have all the bitcoins."
444 2013-08-21 17:20:31 <michagogo> gmaxwell: LOL. What behavior was that?
445 2013-08-21 17:20:57 <warren> Maybe 4 people in the audience had GPG keys but never tried Bitcoin.
446 2013-08-21 17:20:58 <gmaxwell> michagogo: there have been a couple ones, e.g. https://people.xiph.org/~greg/21mbtc.png (a screenshot someone else took, I didn't think to take one)
447 2013-08-21 17:21:06 <michagogo> And how do bitpay compare to other sites, like coinbase?
448 2013-08-21 17:21:44 <gmaxwell> michagogo: or the XSS in a transaction one, https://blockchain.info/tx/59bd7b2cff5da929581fc9fef31a2fba14508f1477e366befb1eb42a8810a000?show_adv=true
449 2013-08-21 17:22:20 <michagogo> jL_<script>window.alert("If this were an actual exploit, your mywallet would be empty.")</script>
450 2013-08-21 17:22:21 <michagogo> lol
451 2013-08-21 17:22:25 <gmaxwell> (which I don't have a screenshot of, alas)
452 2013-08-21 17:22:38 <michagogo> It's simple enough to figure out what it does :-P
453 2013-08-21 17:23:17 <michagogo> gmaxwell: How come https://bitcointalk.org/index.php?action=profile;u=11425 lists two different addresses for you?
454 2013-08-21 17:23:59 <gmaxwell> michagogo: because I rotate the profile one ~every time it gets a payment.
455 2013-08-21 17:24:06 <michagogo> Interesting.
456 2013-08-21 17:24:11 <michagogo> But not the signature one?
457 2013-08-21 17:24:16 <warren> I showed this video in response to the scary parts of Bitcoin: video.cnbc.com/gallery/?video=3000163361&play=1
458 2013-08-21 17:24:24 <gmaxwell> michagogo: you really shouldn't be reusing addresses at all...
459 2013-08-21 17:24:33 <warren> the first few minutes up until Jared Cohen's answer about Bitcoin is decent
460 2013-08-21 17:24:43 <michagogo> gmaxwell: So why do you? :-P
461 2013-08-21 17:25:02 <gmaxwell> michagogo: I generally don't, the 1GMaxwell one is hardly ever used.
462 2013-08-21 17:25:28 <gmaxwell> I've just used it for those few things that should be conspiciously involving me mostly.
463 2013-08-21 17:25:28 <warren> multibit was ever affected by the PRNG problem?
464 2013-08-21 17:27:05 <gmaxwell> michagogo: the profile one mostly only gets used because thats where forum moderation payments go.
465 2013-08-21 17:27:14 <michagogo> Ah.
466 2013-08-21 17:27:18 <michagogo> Forum moderation payments?
467 2013-08-21 17:28:13 <gmaxwell> michagogo: yea, the forum pays moderators, not a ton, but enough to make it feel like a little extra time checking it is well spent.
468 2013-08-21 17:28:24 <michagogo> Cool. How much is it?
469 2013-08-21 17:28:37 <michagogo> (on what order of magnitude?)
470 2013-08-21 17:28:44 <warren> greater than 1 satoshi
471 2013-08-21 17:29:43 <gmaxwell> depends on activity level and some known only to theymos magic, last month I got 3 BTC which is by far the most I've gotten from it.
472 2013-08-21 17:30:14 <gmaxwell> it's mostly been about 1 BTC a month or so, for me. Some of the global moderators get quite a bit more, but they're also spending a whole lot of time on it.
473 2013-08-21 17:30:41 <michagogo> Wow, nice
474 2013-08-21 17:31:40 <Luke-Jr> Krellan: probably, thanks for the report
475 2013-08-21 17:31:56 <marcusw> gmaxwell: and that's all from donations?
476 2013-08-21 17:32:27 <gmaxwell> marcusw: the forum has non-trivial income from the advertising on it.
477 2013-08-21 17:33:31 <marcusw> the fun question is, how much does theymos make?
478 2013-08-21 17:34:26 <gmaxwell> marcusw: a kazillion. No doubt. But ... like .. running the bitcoin forum, you have to be some kind of madman to run that place anyways.
479 2013-08-21 17:35:34 <marcusw> lol
480 2013-08-21 17:35:46 <marcusw> it's nice that he pays in BTC
481 2013-08-21 17:53:07 <gmaxwell> OMG
482 2013-08-21 17:53:16 <gmaxwell> I think I just invented something mildly awesome.
483 2013-08-21 17:53:30 <gmaxwell> I have a way to _provably_ timelock bitcoins.
484 2013-08-21 17:53:34 <michagogo> gmaxwell: Ooh?
485 2013-08-21 17:53:42 <michagogo> How?
486 2013-08-21 17:53:59 <gmaxwell> You have 10 BTC you want to be unspendable until 2014
487 2013-08-21 17:54:38 <gmaxwell> You write a transaction with no inputs which pays 10 BTC to a new destination. The nlocktime is set to 2014.
488 2013-08-21 17:55:10 <michagogo> Erm...
489 2013-08-21 17:55:18 <gmaxwell> In the scriptsig for its input you set the sighash type to ANYONECANPAY. You set the signature to some random data.
490 2013-08-21 17:55:25 <gmaxwell> Then you solve for the public key.
491 2013-08-21 17:55:43 <gmaxwell> You have now computed your redemption transaction without the payment transaction existing.
492 2013-08-21 17:55:56 <gmaxwell> then you pay your coins you want to lock to that address.
493 2013-08-21 17:57:13 <marcusw> you can solve for pubkey with sig+hash? 0.o
494 2013-08-21 17:57:19 <gmaxwell> marcusw: yes.
495 2013-08-21 17:58:03 <marcusw> what crypto is that?
496 2013-08-21 17:58:30 <gmaxwell> Works fine for ECDSA.
497 2013-08-21 17:59:07 <michagogo> ACTION doesn't understand gmaxwell's plan because he doesn't know all the very-base-level details, like what a sighash is
498 2013-08-21 17:59:51 <gmaxwell> oh darn, it won't work because of course ANYONECANPAY still binds the single input.
499 2013-08-21 17:59:55 <marcusw> lol ECDSA
500 2013-08-21 18:00:01 <Scrat> it's a salute of the dev army
501 2013-08-21 18:00:09 <Scrat> ACTION sig hashes michagogo
502 2013-08-21 18:00:22 <marcusw> fuck one thing up and you can solve algebraically for the privkey
503 2013-08-21 18:01:56 <gmaxwell> marcusw: A lot of signature systems have that property, in fact depending on how you define "one thing" they all do.
504 2013-08-21 18:02:33 <marcusw> gmaxwell: but dupe IVs causing privkey leaks with trivial effort?
505 2013-08-21 18:03:09 <gmaxwell> marcusw: sure, any DSA has that property.
506 2013-08-21 18:03:16 <gmaxwell> marcusw: though it's trivially solved.
507 2013-08-21 18:03:41 <marcusw> kinda seems a little risky, hmm?
508 2013-08-21 18:04:13 <michagogo> Also that's not a problem if have a good source of entropy OR you don't reuse addresses (though the second one's not entirely up to you, of course)
509 2013-08-21 18:04:27 <michagogo> BTW, how large can the R value be>
510 2013-08-21 18:04:30 <michagogo> be?*
511 2013-08-21 18:11:17 <gmaxwell> michagogo: R is a point on our curve.
512 2013-08-21 18:11:26 <gmaxwell> marcusw: whats risky?
513 2013-08-21 18:12:02 <gmaxwell> michagogo: 'good source of entropy' just has certificaiton problems. Someone gives you a black box signing tool. How can you tell if its evil?
514 2013-08-21 18:12:24 <michagogo> [23:11:17] <gmaxwell> michagogo: R is a point on our curve.
515 2013-08-21 18:12:24 <michagogo> Hmm?
516 2013-08-21 18:12:36 <gmaxwell> michagogo: in any case, derandomized dsa solves this, and I'd like the bitcoin ecosystem to switch to derandomized dsa (and have been whining about that for a long time)
517 2013-08-21 18:13:04 <michagogo> Does a switch to derandomized dsa allow for backwards compatibility, or is it a hardfork?
518 2013-08-21 18:13:18 <gmaxwell> It's fully compatible.
519 2013-08-21 18:13:27 <michagogo> How does it differ
520 2013-08-21 18:13:28 <michagogo> ?
521 2013-08-21 18:14:36 <gmaxwell> The details of any particular scheme are slightly more complicated, but effectively you set K = H(hashed-message || private key).
522 2013-08-21 18:16:27 <marcusw> so why haven't we switched?
523 2013-08-21 18:20:37 <gmaxwell> marcusw: because no one that needs to switch thinks they need to by definition. ... Also, until recently, there wasn't a published standard for doing derandomized DSA, so it was easy to be cryptoparanoid about any particular approach someone might have suggested.
524 2013-08-21 18:34:50 <sipa> warren: are you saying i never did a gitian build?
525 2013-08-21 18:36:35 <michagogo> ...what?
526 2013-08-21 18:36:42 <michagogo> ACTION checks his email
527 2013-08-21 18:38:41 <michagogo> Ah
528 2013-08-21 18:40:01 <michagogo> sipa: Er, he didn't say "s"...
529 2013-08-21 18:43:00 <[7]> ACTION slaps michagogo :P
530 2013-08-21 18:43:00 <michagogo> ACTION slaps [7] around a bit with a large trout
531 2013-08-21 18:43:09 <michagogo> .w [7]
532 2013-08-21 18:43:10 <michagogo> bah
533 2013-08-21 18:43:26 <michagogo> [7]: Why?
534 2013-08-21 18:43:41 <[7]> because these stupid quotations keep highlighting me :P
535 2013-08-21 18:44:00 <michagogo> erm
536 2013-08-21 18:44:17 <michagogo> Wait, what?
537 2013-08-21 18:44:24 <michagogo> You highlight on the numeral 7?
538 2013-08-21 18:44:30 <[7]> I highlight [7]
539 2013-08-21 18:44:43 <michagogo> Oh, LOL
540 2013-08-21 18:44:49 <michagogo> The wikipedia article?
541 2013-08-21 18:45:15 <[7]> yep. keeps happening once a week or so
542 2013-08-21 18:45:31 <[7]> the number 7 seems to be used fairly often in references :/
543 2013-08-21 19:02:54 <sipa> why we need to split bitcoind from the wallet: http://www.reddit.com/r/Bitcoin/comments/1kt970/bitcoin_developers_need_api_to_analyze_the/cbse0k8
544 2013-08-21 19:04:00 <warren> sipa: no
545 2013-08-21 19:04:05 <gmaxwell> sipa: I don't actually understand whats being said there. They're worried about people trying to get their private keys so they are using blockchain.info's api?
546 2013-08-21 19:04:17 <gmaxwell> (not that I disagree with your conclusions in the slightest!!)
547 2013-08-21 19:04:45 <warren> sipa: I'm saying that j, p and g, prominent devs have never contributed to gitian.sigs.
548 2013-08-21 19:04:48 <SomeoneWeird> they don't want to use the API on a node that has keys
549 2013-08-21 19:05:36 <sipa> gmaxwell: you may miss some context in the post they're replying to
550 2013-08-21 19:06:02 <sipa> gmaxwell: the reason for not using bitcoind as an API is because that bitcoind also manages their wallet, and they can't run because of hardware constraints
551 2013-08-21 19:06:10 <sipa> can't run 2
552 2013-08-21 19:06:33 <sipa> warren: i have no clue what 'p' you're referring to then :)
553 2013-08-21 19:06:41 <warren> sipa: he knows who he is. =)
554 2013-08-21 19:06:45 <sipa> ok
555 2013-08-21 19:07:06 <sipa> Ah.
556 2013-08-21 19:09:05 <Krellan> Luke-Jr: Thanks. I experimented with this patch to perhaps work around floating-point rounding that would cause that percentage cosmetic display glitch: http://pastebin.com/YfaYwGDH