1 2014-08-22 00:09:56 <eitan_chatav> Hello, I've a question about payment protocol. How is BIP 70 related to homomorphic payments? Is homomorphic payment supported in any client currently?
2 2014-08-22 00:13:44 <gmaxwell> eitan_chatav: it's not related.
3 2014-08-22 00:14:23 <eitan_chatav> gmaxwell: thanks. is it the subject of a future BIP?
4 2014-08-22 00:15:19 <gmaxwell> Perhaps! The data managment issue for pay to contract seem tricky to solve, as do the privacy ones. I'd like to see it done.
5 2014-08-22 00:15:41 <gmaxwell> I've seen some people doing overly simplistic versions of it which are insecure.
6 2014-08-22 00:16:10 <eitan_chatav> What are the issues involved? Are they specific to homomorphic payment?
7 2014-08-22 00:16:56 <phantomcircuit> bah maths
8 2014-08-22 00:17:03 <eitan_chatav> yay math
9 2014-08-22 00:17:05 <gmaxwell> (E.g. insecure version: if I tell you that my pubkey is X where X is really H(fake_contract)*G and you pay to X+H(real_contract)*G then I can go claim to people that the contract was fake_contract)
10 2014-08-22 00:17:54 <gmaxwell> The other class of issues is where payment is made to a contract and parties don't get the contract written to reliable storage before the payment, and the coins are lost forever.
11 2014-08-22 00:18:28 <phantomcircuit> eitan_chatav, calculate how large the window needs to be such that the actual number of shares submitted at difficulty N for a device expected to run at hashrate H is within X% of the expected Y% of the time
12 2014-08-22 00:19:04 <phantomcircuit> i say again
13 2014-08-22 00:19:06 <phantomcircuit> bah maths
14 2014-08-22 00:19:18 <eitan_chatav> lol, fair enuff phantomcircuit
15 2014-08-22 00:19:41 <eitan_chatav> though I probs wouldn't classify that as math
16 2014-08-22 00:20:41 <gmaxwell> The privacy issue is, say you use pay-to-contract and this is well knownâ then ninjas show up and demand copies of all your contractsâ thus getting cryptographically provable evidence of what the contracts were.
17 2014-08-22 00:21:45 <eitan_chatav> damned ninjas
18 2014-08-22 00:21:48 <gmaxwell> e.g. even if the reciver and transmitter both want to conceal the contract against a hostile third party, they probably can't.
19 2014-08-22 00:23:32 <eitan_chatav> gmaxwell: is there a thread somewhere I can read about this?
20 2014-08-22 00:23:34 <phantomcircuit> eitan_chatav, afaict it's calculating the CDF for the poisson distributon
21 2014-08-22 00:24:02 <eitan_chatav> phantomcircuit: I'm just kidding around but that's applied maths ;-)
22 2014-08-22 00:24:20 <gmaxwell> eitan_chatav: probably not. I made some post with an attempt at improving the privacy, but I'm not super happy with it: https://bitcointalk.org/index.php?topic=318279.0
23 2014-08-22 00:25:14 <eitan_chatav> gmaxwell: I assume that there are problems with current ad-hoc solutions for merchant/customer payment systems, right?
24 2014-08-22 00:28:08 <gmaxwell> eitan_chatav: basically no one is saying "there are contract disputes which pay-to-contract could solve"... perhaps because people are stupid. :)
25 2014-08-22 00:28:26 <eitan_chatav> oh I get it, it's worse because of the provability of the contract, so you can't fool a coercer
26 2014-08-22 00:29:07 <gmaxwell> so pay-to-contract is not so exciting when it solves problems people don't seem to care about ... and then yea, leaves you vulnerable to coercison and makes you more at risk of losing funds. (e.g. if your disk fails right after the deal is done and your last backup was more than a few seconds ago)
27 2014-08-22 00:29:50 <eitan_chatav> I see. Ok :-( Too bad, I was real excited a couple years back when I saw Timo's talk about it.
28 2014-08-22 02:12:40 <Trind> hi shooter
29 2014-08-22 07:15:08 <arioBarzan> Luke-Jr: probably you're more likely to steal random coins than anyone else, because almost all other nodes don't even bother to relay that kind of transactions, and obviously rest of miners other than Eligius don't confirm them either. So there is a high probability that others don't get to know about the consumed coins with no OP_CHECKSIG in their script, until Eligius put them in a block.
30 2014-08-22 07:16:19 <arioBarzan> (as I mentioned before) like tx/b5c103dae667c4c5067be34510b30bef059090db09355c993cda00517a2b79a6
31 2014-08-22 07:19:42 <poutine> arioBarzan, I made that tx
32 2014-08-22 07:19:47 <arioBarzan> Bytheway I didn't say "you're more likely", I said you "could easily" do that. So I didn't mean that you have any intention of stealing other people's coins :)
33 2014-08-22 07:20:19 <poutine> I don't advocate blind trust, but I don't think it'd be good for anyone's reputation for that to happen, also the value of it was worth the risk
34 2014-08-22 07:20:20 <arioBarzan> poutine: yup, you had your signature in the script as I said yesterday.
35 2014-08-22 07:41:47 <amaclin> arioBarzan, try to send 0.0002 to 3GMLTNSF8ZhuSvysvHvauxP7AbWMfjiu7G and see what happens next :)
36 2014-08-22 07:45:24 <wumpus> cfields: travis is rejecting all pulls :(
37 2014-08-22 07:45:49 <wumpus> Merge with caution!
38 2014-08-22 08:57:30 <elichai2> petertodd, here?
39 2014-08-22 09:17:24 <wumpus> ... -lcrypto -lminiupnpc -lminiupnpc -lminiupnpc -lminiupnpc -lcrypt32 ...
40 2014-08-22 09:17:31 <wumpus> maybe we're slightly over the top there
41 2014-08-22 09:18:14 <gmaxwell> "link, damn you, link!"
42 2014-08-22 09:18:24 <wumpus> and still it gives undefined reference error, maybe one more -miniupnpc will solve it!
43 2014-08-22 09:19:44 <wumpus> --force-link --weld --apply-strong-pressure
44 2014-08-22 09:20:32 <gmaxwell> don't forget -f(ice)pic
45 2014-08-22 09:20:41 <sipa> -fomit-broken-code -Optimize-the-crap-out-of-it -fuse-the-force-luke
46 2014-08-22 09:21:00 <wumpus> hehe
47 2014-08-22 09:21:09 <elichai2> does anyone here ever used petter todd's 'replace-by-fee-tools'?
48 2014-08-22 09:21:31 <gmaxwell> Of probable interest to people here: http://www.reddit.com/r/Bitcoin/comments/2e6q7z/a_fast_relay_network_for_miners/
49 2014-08-22 09:31:31 <wumpus> good writeup
50 2014-08-22 09:34:34 <wumpus> ah, miniupnpc have renamed -DSTATICLIB to -DMINIUPNP_STATICLIB, so thinking it was importing from a DLL library it put damn __imps before the symbols
51 2014-08-22 10:19:00 <elichai2> 0100000001658e70e3860fd59169366e0ee01149c1c484ef0f7ce658c3b3add79ca03aa709000000006b4830450221008dcaf7ae841319ddcaa71e7aa3751c935f7e8070227a084f4b75259e560c4dc5022037442e22666352587834f3d0fd7f86296cee1949b396b17efe5e25612aa701ad01210379d37c226a0deda8b2abbef987cecec9ebeeaba2a63b01cc28910d721e735caeffffffff01a73a0100000000001976a9149824cfa26f1e8b3ff3be47af354f192768eb26a688ac00000000
52 2014-08-22 10:19:00 <elichai2> <elichai2> hey
53 2014-08-22 10:19:00 <elichai2> why do i get 'error: {"code":-25,"message":""}' on this transaction: ?
54 2014-08-22 10:45:31 <josiah> Hi folks
55 2014-08-22 10:45:52 <josiah> I'm trying to test payment processing with bitcoin-qt
56 2014-08-22 10:46:04 <josiah> but bitcoin-qt does not like my uri
57 2014-08-22 10:46:17 <josiah> 'bitcoin:mrzURcDyJsW8wX3t13Hes5ArKAPKvJE8n3?amount=0.5096693602'
58 2014-08-22 10:46:23 <josiah> any ideas?
59 2014-08-22 10:55:12 <Eliel> josiah: is it running in testnet mode?
60 2014-08-22 10:59:10 <josiah> yep
61 2014-08-22 10:59:19 <josiah> I'm running on ubuntu
62 2014-08-22 10:59:41 <josiah> @Eliel it works on my windows 7 version of qt
63 2014-08-22 11:02:25 <fanquake> josiah Iâm seeing IOUnserialize has detected a string that is not valid UTF-8
64 2014-08-22 11:04:53 <josiah> not sure what it could be
65 2014-08-22 11:05:16 <josiah> the address was generated by bitcoind
66 2014-08-22 11:05:27 <josiah> and it works with Electrum bitcoin Wallet
67 2014-08-22 11:05:42 <fanquake> A further look suggests that my error message is actually some osx issue
68 2014-08-22 11:23:19 <Guest58481> While I wait for my username to be released (robbak), is there any move to switch to a more current BDB?
69 2014-08-22 11:24:04 <hearn> robbak_: afaik no, not if you mean a smooth/automatic migration process
70 2014-08-22 11:24:08 <robbak_> The bdb4 series is being depreciated across the planet.
71 2014-08-22 11:25:17 <hearn> luckily deprecated software does not just magically disappear
72 2014-08-22 11:27:34 <robbak_> As long as it is still supported in some way. With bdb5 a drop-in replacement, I'd like to see us switch with the next release.
73 2014-08-22 11:31:42 <robbak_> FreeBSD has just dropped everything before 4.8, and plans to drop 4.8 within the year. So I'm going to be in the same situation that Ubuntu users are now - either having to go our own way with 5.n, or hacking in a local copy of 4.8.
74 2014-08-22 11:36:25 <hearn> or just use the precompiled binaries
75 2014-08-22 11:36:29 <hearn> only developers have to worry about this
76 2014-08-22 11:36:36 <hearn> the problem with upgrading BDB is that the file formats are incompatible
77 2014-08-22 11:36:40 <hearn> the configure script explains this
78 2014-08-22 11:37:08 <Luke-Jr> hearn: we don't have precompiled BSD binaries O.o
79 2014-08-22 11:37:19 <robbak_> Yeah - well, as I am the person who updates the port that is used to create the precompiled binaries, it is going to be my problem!
80 2014-08-22 11:37:26 <hearn> ah ha :)
81 2014-08-22 11:37:39 <hearn> yeah, so you'll have to include the old bdb i guess. or break wallet compatibility, but that would be bad
82 2014-08-22 11:38:01 <Luke-Jr> robbak_: I don't think we really officially support BSD at all, even with source - need someone interested in testing and maintaining that port ;)
83 2014-08-22 11:38:36 <Luke-Jr> robbak_: (which would indeed make it "your problem" if you choose to be that person)
84 2014-08-22 11:39:02 <Luke-Jr> "some day" hopefully we will switch to a log-based wallet format, but I don't see it happening too soon :/
85 2014-08-22 11:40:43 <robbak_> Yeah. We'll see what happens with bdb. In all likelihood they'll keep 4.8 around longer than they are currently planning.
86 2014-08-22 11:41:10 <gmaxwell> realistically with oracle seemingly trying to kill bdb you'll see more projects doing this.
87 2014-08-22 11:41:40 <gmaxwell> it doesn't make sense to continue on an incompatible upgrade train that introduces bugs and ends in a license incompatibility brick wall.
88 2014-08-22 11:54:19 <wumpus> I don't see it as a big issue, right now bdb48 is easy enough to build from source
89 2014-08-22 11:55:29 <wumpus> of course, at some point in the future code rot will set in, as the existing archive will be incompatible with new compilers and libraries, and it doesn't sound very appealing either to take maintenance upon us
90 2014-08-22 11:56:58 <wumpus> then again, maybe still more appealing than the nastiness that 5.x+ may bring
91 2014-08-22 12:05:51 <wumpus> I hope that someone will pick it up at some point, introduce a new wallet format that doesn't have dependencies on any db lib
92 2014-08-22 12:20:23 <wumpus> going to tag 0.9.3rc1 in a bit, can anyone have a last look https://github.com/bitcoin/bitcoin/blob/0.9.3/doc/release-notes.md https://github.com/bitcoin/bitcoin/commits/0.9.3
93 2014-08-22 12:22:58 <Luke-Jr> wumpus: I looked over the release notes, didn't see anything to change.
94 2014-08-22 12:37:23 <wumpus> thanks
95 2014-08-22 13:02:54 <wumpus> * [new tag] v0.9.3rc1 -> v0.9.3rc1
96 2014-08-22 13:04:39 <wumpus> gitian builds can start
97 2014-08-22 13:13:03 <michagogo> ACTION just got his SMS
98 2014-08-22 13:29:47 <michagogo> wumpus: was the qt win dep version intentionally not bumped?
99 2014-08-22 13:30:22 <michagogo> (I ask because in the past, openssl update has usually bumped the deps version)
100 2014-08-22 13:34:38 <wumpus> michagogo: yes
101 2014-08-22 13:34:59 <wumpus> the windows qt dep doesn't change when the openssl version changes
102 2014-08-22 13:35:27 <michagogo> Really? Maybe I'm misremembering...
103 2014-08-22 13:35:29 <michagogo> ACTION looks
104 2014-08-22 13:35:36 <wumpus> I checked.
105 2014-08-22 13:35:52 <michagogo> How come OS X does?
106 2014-08-22 13:36:03 <wumpus> because it's not deterministic
107 2014-08-22 13:36:55 <michagogo> erm
108 2014-08-22 13:36:56 <michagogo> https://github.com/bitcoin/bitcoin/commit/4a811b00534cb1ba39f8d31c4912ba995a4ac268
109 2014-08-22 13:37:03 <michagogo> was immediately followed by https://github.com/bitcoin/bitcoin/commit/178825dec338cb0840aef1d23415981012a5593f
110 2014-08-22 13:37:28 <michagogo> Were we wrong back then?
111 2014-08-22 13:37:44 <wumpus> yes, we were wrong
112 2014-08-22 13:38:34 <michagogo> Ah. So what does Qt use OpenSSL for?
113 2014-08-22 13:38:52 <michagogo> (and why does it not need updating?)
114 2014-08-22 13:38:53 <wumpus> it's easy to understand: openssl is a c library without implementation in the header files, qt is a library that *uses* openssl, the link-time interface doesn't change between the openssl versions
115 2014-08-22 13:39:01 <michagogo> Ahh.
116 2014-08-22 13:39:21 <wumpus> so there is no reason at all that you'd expect qt to change
117 2014-08-22 13:39:47 <michagogo> (I'm sure that's obvious/intuitive to you, but I'm not really familiar with the whole c world and how stuff like that works...)
118 2014-08-22 13:39:53 <wumpus> and it doesn't... I always build *everything* on a new release and verify the hashes
119 2014-08-22 13:41:08 <wumpus> you're free of course to verify if you want to check
120 2014-08-22 13:41:21 <michagogo> Um.
121 2014-08-22 13:41:23 <michagogo> ./bin/gbuild:83:in `build_one_configuration': undefined method `exists?' for Dir:Class (NoMethodError)
122 2014-08-22 13:41:23 <michagogo> Preparing build environment
123 2014-08-22 13:42:44 <michagogo> ACTION checks the gitian-builder history
124 2014-08-22 13:43:03 <wumpus> strange
125 2014-08-22 13:43:24 <wumpus> (works for me, but I my last update of gitian was a few days ago)
126 2014-08-22 13:44:35 <michagogo> I just pulled
127 2014-08-22 13:45:39 <michagogo> Looks like that line was added in https://github.com/devrandom/gitian-builder/commit/ea24af1043b964394e7406a3e36e863620939efb
128 2014-08-22 13:45:52 <michagogo> ACTION looks up the ruby Dir class
129 2014-08-22 13:46:23 <michagogo> ACTION also looks up what version of ruby is in precise
130 2014-08-22 13:48:40 <michagogo> Aha
131 2014-08-22 13:49:08 <michagogo> Dir.exist[s]? exists in 2.1.2 and not in 1.8.7
132 2014-08-22 13:49:37 <michagogo> It's also in 1.9.3
133 2014-08-22 13:50:22 <wumpus> hm
134 2014-08-22 13:56:19 <michagogo> Looks like my pull was ae83ac0..8e907ad
135 2014-08-22 13:56:31 <michagogo> What's the git command to revert again?
136 2014-08-22 13:56:56 <michagogo> Not git checkout ae83ac0, that would put me in detached HEAD...
137 2014-08-22 13:58:19 <michagogo> Aha
138 2014-08-22 13:58:20 <michagogo> git reset HEAD@{1}
139 2014-08-22 13:59:30 <michagogo> Okay, that should fix it for now (after I update the VM and reinstall the guest additions...)
140 2014-08-22 14:03:53 <wumpus> gitian should really not require a recent ruby version :/
141 2014-08-22 14:10:07 <michagogo> I agree
142 2014-08-22 14:12:18 <wumpus> I'd assume that older ruby has another way to do Dir.exist
143 2014-08-22 14:13:13 <wumpus> hah, that German sniffer node still hasn't given up connecting to me
144 2014-08-22 14:13:41 <gavinandresen> brew install ruby and then /usr/local/bin/ruby bin/gbuild ../bitcoin/contrib/gitian-descriptors/deps-linux.yml ⦠is working so far for me on OSX
145 2014-08-22 14:15:52 <wumpus> gavinandresen: good to hear
146 2014-08-22 14:24:08 <michagogo> gavinandresen: well, I'd assume Homebrew installs new ruby, so there shouldn't be a problem
147 2014-08-22 14:25:00 <gavinandresen> michagogo: right, what I done did is a workaround, not a fix. gitian shouldn't require a recent ruby version
148 2014-08-22 14:25:16 <michagogo> Oh, you were on system ruby?
149 2014-08-22 14:25:26 <gavinandresen> michagogo: yes
150 2014-08-22 14:25:30 <michagogo> What ruby does os x ship?
151 2014-08-22 14:25:47 <gavinandresen> OSX 10.8.5 has ruby 1.8.7 (2012-02-08 patchlevel 358) [universal-darwin12.0]
152 2014-08-22 14:25:59 <gavinandresen> ⦠don't know what 10.9 ships
153 2014-08-22 14:26:27 <gavinandresen> brew is ruby 2.1.2p95 (2014-05-08 revision 45877) [x86_64-darwin12.0]
154 2014-08-22 14:27:00 <michagogo> Ah, so in the same boat as Ubuntu
155 2014-08-22 14:27:24 <michagogo> I could probably install rvm or something to get a newer ruby, but for now reverting those changes should be fine
156 2014-08-22 14:28:15 <michagogo> Looks like none of it affects building 0.9.3, since it's not using the cache stuff yet
157 2014-08-22 14:28:37 <michagogo> (and I don't need to use @ in a URL)
158 2014-08-22 14:32:14 <jgarzik> wumpus, gavinandresen, sipa, cfields: ping RE Add more bitcoin-tx tests #4733 https://github.com/bitcoin/bitcoin/pull/4733
159 2014-08-22 14:33:07 <wumpus> michagogo: right, 0.9.3 doesn't need a recent gitian
160 2014-08-22 14:33:18 <wumpus> jgarzik: ACK
161 2014-08-22 14:35:33 <wumpus> michagogo: replacing Dir.exists? with File.directory? should solve the issue too
162 2014-08-22 14:44:01 <michagogo> I may not get my build in today... I g2g in ~an hour
163 2014-08-22 14:44:11 <michagogo> and the VM is still updating
164 2014-08-22 14:44:39 <michagogo> It seems like every time I use it there's a linux update, and those seem to take a long time... :-/