1 2015-02-05 00:06:11 <fanquake> ;;blocks
2 2015-02-05 00:06:12 <gribble> 341993
3 2015-02-05 02:16:39 <kryo_> hey i need help (or just a second opinion) with okcoin's websocket API
4 2015-02-05 02:26:39 <kryo_> i guess i will ask okcoin
5 2015-02-05 02:26:47 <kryo_> seems to be on their end..
6 2015-02-05 02:33:39 <jtimon> cfields my rebase of #5681 failed too, I think it may be related wit hthe cache problem you mentioned https://travis-ci.org/bitcoin/bitcoin/jobs/49540263
7 2015-02-05 02:44:45 <cfields> jtimon: yes i saw. i asked wumpus earlier to clear the cache to help take care of the problem. Unfortunately, the cache won't be primed until something's merged in.
8 2015-02-05 02:44:57 <cfields> jtimon: so after that (next merge into master) i think it'll be worked out
9 2015-02-05 02:45:33 <cfields> sorry 'bout that.
10 2015-02-05 02:47:24 <jtimon> ok, thank you for the update
11 2015-02-05 02:47:30 <cfields> jtimon: i'd like to review some of your PRs now, but tbh i don't know where to start. Seems everything's a moving target
12 2015-02-05 02:48:52 <cfields> jtimon: any suggestions?
13 2015-02-05 03:18:40 <roconnor> what version of OpenSSL should bitcoin 0.9.3 be built against?
14 2015-02-05 03:19:22 <phantomcircuit> 1.0.1j
15 2015-02-05 03:19:38 <phantomcircuit> i cant remember the 0.9.x version or the 1.1.x version
16 2015-02-05 03:25:19 <roconnor> ty
17 2015-02-05 05:29:09 <fanquake> wumpus can close #5736
18 2015-02-05 08:49:16 <wumpus> fanquake: done, thanks
19 2015-02-05 09:03:35 <jonasschnelli> is there no way to tell travis not to build pulls when detecting only changes in /docs? I mean #5755 took 2h to build?!
20 2015-02-05 09:05:58 <wumpus> there is something you can put in the pull title to avoid travis testing
21 2015-02-05 09:09:46 <jonasschnelli> wumpus, hmm.. but nobody add this [no-travis] thing... i wonder if there is a something like a -excute-dir option or similar... but low pro...
22 2015-02-05 09:10:17 <jonasschnelli> s/pro/prio
23 2015-02-05 09:16:40 <wumpus> or maybe travis.yml could be made to make a decision whether to test or not, or what to test based on what files changed, I don't know
24 2015-02-05 09:24:45 <jonasschnelli> Travis has no feature for this currently (just checked). Only for branches.
25 2015-02-05 09:24:53 <jonasschnelli> There is only the [skip ci] possibility
26 2015-02-05 09:38:19 <u20024804> 2015-02-05 03:12:09 UpdateTip: new best=000000000086b3a2d1ed70aac5539006bcc5bef060fa65ffaf0053d3656a2622 height=71178 log2_work=52.723193 tx=188690 date=2013-05-06 01:49:24 progress=0.477695 2015-02-05 03:12:09 ProcessBlock: ACCEPTED 2015-02-05 03:12:09 ERROR: CScriptCheck() : 4d0bbf6348726a49600171033e456548a09b246829d649e77b929caf242ae6e7 VerifySignature failed 2015-02-05 03:12:09 InvalidChainFound: invalid block=0000
27 2015-02-05 09:38:46 <u20024804> Hi every one, I got this error and can't sync my node any more.
28 2015-02-05 09:39:11 <u20024804> 2015-02-05 03:12:09 ERROR: ConnectTip() : ConnectBlock 000000000124ea62f4b3af18c2d121fd6055ac47678b858275d55f241badd7c9 failed
29 2015-02-05 09:40:16 <u20024804> I found 4d0bbf6348726a49600171033e456548a09b246829d649e77b929caf242ae6e7 verified script fail because SCRIPT_VERIFY_STRICTENC
30 2015-02-05 09:41:59 <u20024804> It happened in the testnet.
31 2015-02-05 10:01:26 <wumpus> u20024804: what version are you using? do you download executables from bitcoin.org, use a repository package, or build it yourself?
32 2015-02-05 10:02:01 <wumpus> it smells like the openssl1.0.1k compatibility issue, but I don't know the block number where it gets stuck on testnet
33 2015-02-05 10:22:03 <Luke-Jr> wumpus: "+Mining and relay policy enhancements" didn't lengthen the dashes under it (does it need a PR?)
34 2015-02-05 10:23:34 <amaclin_> http://test.webbtc.com/tx/4d0bbf6348726a49600171033e456548a09b246829d649e77b929caf242ae6e7
35 2015-02-05 10:23:50 <wumpus> Luke-Jr: blah, doesn't seem awfully important
36 2015-02-05 10:24:14 <Luke-Jr> nah, not really
37 2015-02-05 10:25:30 <amaclin_> u20024804, yes, this scriptSig has non-der signatures
38 2015-02-05 10:26:54 <wumpus> going to tag rc4
39 2015-02-05 10:26:58 <wumpus> amaclin_: thanks
40 2015-02-05 10:30:00 <wumpus> * [new tag] v0.10.0rc4 -> v0.10.0rc4
41 2015-02-05 10:30:50 <Luke-Jr> in other news, apparently Multibit is charging 0.0001 BTC per second of use soon.. :o
42 2015-02-05 10:30:59 <Luke-Jr> ACTION begins gitian builds
43 2015-02-05 10:31:13 <wumpus> per second of use?! that's kind of weird
44 2015-02-05 10:31:31 <Luke-Jr> yeah
45 2015-02-05 10:31:44 <Luke-Jr> and they compare it to the 0.001 BTC miner fee.. I sure don't pay a miner fee every second
46 2015-02-05 10:32:00 <Luke-Jr> https://beta.multibit.org/blog/2014/04/11/multibit-hd-brit.html
47 2015-02-05 10:32:01 <wumpus> something per transcation would make more sense for monetization of a wallet
48 2015-02-05 10:32:49 <Luke-Jr> oh, per spend, not per second
49 2015-02-05 10:32:55 <wumpus> but having to pay for just having software open... dunno
50 2015-02-05 10:32:56 <Luke-Jr> read it wrong.
51 2015-02-05 10:32:57 <wumpus> oooh :D
52 2015-02-05 10:33:26 <wumpus> I already imagined it like some 80's arcade game, insert coin :p
53 2015-02-05 10:33:31 <Luke-Jr> hehe
54 2015-02-05 10:33:48 <Luke-Jr> if it was a full node, it'd be fun to charge per launch..
55 2015-02-05 10:33:58 <Luke-Jr> that way you discourage people from closing it
56 2015-02-05 10:35:33 <Luke-Jr> if we had micropayments, we could just have everyone pay for blocks.. maybe someday
57 2015-02-05 10:35:36 <wumpus> pay-per-shutdown hehe
58 2015-02-05 10:36:46 <phedny> Luke-Jr: strict pay-per-launch would of course prevent new nodes to be added, because of initial empty wallets ;)
59 2015-02-05 10:37:08 <u20024804> I built it myself from 0.9.3 source code
60 2015-02-05 10:37:22 <Luke-Jr> phedny: heh, true. how about just attaching the fee to the first tx sent per launch?
61 2015-02-05 10:37:32 <Luke-Jr> u20024804: with an incompatible OpenSSL it sounds like
62 2015-02-05 10:38:20 <u20024804> my openssl version is 1.0.1j
63 2015-02-05 10:39:04 <Luke-Jr> u20024804: more likely it's 1.0.1j broken by your distro
64 2015-02-05 10:39:37 <u20024804> So which version should I choose?
65 2015-02-05 10:41:49 <Luke-Jr> just ignore OpenSSL and use BCCore 0.9.4
66 2015-02-05 10:43:04 <u20024804> I have to build bitcoin core from source code to test something.
67 2015-02-05 10:43:36 <wumpus> so rebase your changes to v0.9.4 tag, changes are trivial from v0.9.3
68 2015-02-05 10:44:18 <wumpus> (well, at least trivial to rebase over)
69 2015-02-05 10:45:42 <u20024804> Thank you! Let me try.
70 2015-02-05 10:48:05 <rubensayshi> out of curiousity, why did sipa disable GMP for secp256k1 in bitcoin 0.10 and it's enabled on master?
71 2015-02-05 10:54:39 <paveljanik> rubensayshi, to not make bitcoind depend on yet another library
72 2015-02-05 10:55:08 <wumpus> rubensayshi: secp256k1 on master needs to be updated
73 2015-02-05 10:55:13 <rubensayshi> secp256k1 does autodetect and use gmp if possible
74 2015-02-05 10:55:19 <wumpus> sipa even has a pull open for that
75 2015-02-05 10:55:21 <rubensayshi> on 0.10 the autodetect is disabled to never use gmp
76 2015-02-05 10:55:33 <wumpus> yes
77 2015-02-05 10:55:49 <wumpus> that was a hot-patch for release-technical reasons
78 2015-02-05 10:57:51 <wumpus> eh we merged #5506, so that should be in master now too
79 2015-02-05 10:58:00 <rubensayshi> and the reason for not wanting gmp at all is to reduce chance of bugs in something that is out of bitcoin core's control I suppose? (which is the reason for switching to the secp256k1 lib in the first place)
80 2015-02-05 10:58:31 <wumpus> right, and it's just not necessary
81 2015-02-05 11:00:33 <wumpus> gmp is a better dependency to have than openssl, sure, but it's indeed better to have it self contained
82 2015-02-05 11:02:53 <rubensayshi> if I run ./configure on the master branch I see "configure: Using bignum implementation: gmp"
83 2015-02-05 11:03:04 <rubensayshi> lemme do a fresh clone just to be usre
84 2015-02-05 11:06:10 <rubensayshi> yea, I'm definitely getting that message when running ./configure on master branch, so if your intentions are to make it never use GMP when people build bitcoin from source then you might want to double check it wumpus
85 2015-02-05 11:09:43 <rubensayshi> made a comment on the PR, so you can take a look at it when you feel like it instead of when I'm nagging on IRC ;-)
86 2015-02-05 11:09:45 <wumpus> rubensayshi: can you try configure with a checkout of secp256k1 from its repository?
87 2015-02-05 11:09:57 <wumpus> (eg outside of bitcoin)
88 2015-02-05 11:10:13 <rubensayshi> latest secp256k1 has autodetect for gmp
89 2015-02-05 11:10:20 <wumpus> if it's fixed there, which I'm quite sure, we just need a bump of secp256k1
90 2015-02-05 11:10:22 <wumpus> hmm right
91 2015-02-05 11:10:52 <rubensayshi> I think sipa might want to keep it an option in secp256k1 while not wanting it to auto detect when used for bitcoin or smt
92 2015-02-05 11:11:11 <wumpus> it could be an option without auto detection
93 2015-02-05 11:11:37 <wumpus> but I thought it was ripped out at some point
94 2015-02-05 11:11:49 <rubensayshi> maybe change the default --with-bignum to NO and then if someone wants to build it with auto/gmp the can explicitly do it
95 2015-02-05 11:12:41 <rubensayshi> I pinged you and sipa on that hotfix PR, I guess he'll notice it at some point
96 2015-02-05 11:13:17 <michagogo> Ooh
97 2015-02-05 11:13:42 <michagogo> ACTION TeamViews into his laptop to boot his VM
98 2015-02-05 11:14:50 <Luke-Jr> why do people use centralised VNC clones like TeamViewer, when standard VNC has been around forever?
99 2015-02-05 11:17:01 <michagogo> Luke-Jr: my modem-router has this really annoying bug
100 2015-02-05 11:17:51 <michagogo> If you assign a static dhcp address, every half hour (or was it an hour?) all connections drop
101 2015-02-05 11:18:20 <michagogo> I tracked it down with wireshark, and for some reason it sends a dhcp NAK
102 2015-02-05 11:19:14 <wumpus> michagogo: hey, I had the same problem with an old router/modem (supplied by the cable company)
103 2015-02-05 11:19:39 <michagogo> wumpus: heh
104 2015-02-05 11:19:48 <wumpus> now I have a spiffy router with openwrt, ISP's router set to bridge mode, and no problems anymore
105 2015-02-05 11:19:58 <michagogo> It's a Netgear VVG2000
106 2015-02-05 11:20:16 <michagogo> Yeah, I wish we had that
107 2015-02-05 11:20:36 <michagogo> What we have works pretty well though, other than that one quirk
108 2015-02-05 11:20:47 <wumpus> it appeared to target specific devices; eg my phone was regularly kicked from the network in that way, and it didn't only happen to static dhcp addresses AFAIK
109 2015-02-05 11:21:01 <wumpus> anyhow, yes lots of routers suck :p
110 2015-02-05 11:21:30 <michagogo> wumpus: in my case, it happened as soon as I reserved an address for the computer, and when I realized that I had done that and the assignment, fixed it immediately.
111 2015-02-05 11:25:56 <Luke-Jr> wumpus: rc4 sigs pushed
112 2015-02-05 11:26:41 <Luke-Jr> michagogo: eh, most ISPs don't let you do static DHCP.. usually you fix this with dynamic DNS :p
113 2015-02-05 11:26:57 <Luke-Jr> wumpus: I get lots of problems with OpenWrt :/
114 2015-02-05 11:28:28 <michagogo> Luke-Jr: I mean NAT-internal
115 2015-02-05 11:28:57 <Luke-Jr> ACTION ponders whether to preemptively do an autoprune rc4 for the people who are no doubt going to ask in a few hours
116 2015-02-05 11:29:01 <Luke-Jr> michagogo: ah, weird
117 2015-02-05 11:29:08 <michagogo> Yeah, I'd use dynamic DNS, but the problem is I can't do dynamic port forwarding
118 2015-02-05 11:30:06 <Luke-Jr> michagogo: could set a static IP on the host :P
119 2015-02-05 11:33:03 <michagogo> Luke-Jr: I don't think windows lets you do that per network. It would be set on the adapter, and I would need to go back and change it every time I take the computer elsewhere, which is annoying. I could, but it's easier to use something like TeamViewer.
120 2015-02-05 11:33:29 <michagogo> It's also nice to have installed when my grandmother calls me to ask for help...
121 2015-02-05 11:33:42 <Luke-Jr> XD
122 2015-02-05 11:34:31 <Luke-Jr> anyhow, rc4 bins uploading to http://luke.dashjr.org/programs/bitcoin/files/bitcoind/0.10.0/test/rc4/
123 2015-02-05 11:34:34 <Luke-Jr> night
124 2015-02-05 11:34:51 <michagogo> And their TeamViewer QS applet works really well, so when other people (who don't have TV installed on their computers) need help, I can just point them to it
125 2015-02-05 11:34:56 <Luke-Jr> michagogo: are you doing a reddit post again? :p
126 2015-02-05 11:35:20 <michagogo> Luke-Jr: if nobody has gotten around to it when I get home
127 2015-02-05 11:35:33 <michagogo> Which may not be until 6:30-7 UTC
128 2015-02-05 11:37:56 <Luke-Jr> well, my wife is yelling at me to go to bed, so it won't be me. XD
129 2015-02-05 12:58:43 <gdm85> just got this when gitian-building 0.10.0rc4: http://pastebin.com/raw.php?i=qgfN5EUu
130 2015-02-05 13:00:59 <gdm85> FYI: rating C: https://www.ssllabs.com/ssltest/analyze.html?d=bitcoincore.org
131 2015-02-05 13:02:00 <gdm85> although my issue is prolly due to one of the CA certs missing from debian
132 2015-02-05 13:03:04 <jonasschnelli> gdm85, i just builded the linux part without problems... https://builds.jonasschnelli.ch/releasebuilds/v0.10.0rc4/build-linux.log
133 2015-02-05 13:03:53 <gdm85> jonasschnelli: on Debian some CA certs are not installed by default.
134 2015-02-05 13:04:15 <jonasschnelli> gdm85, i built on wheezy
135 2015-02-05 13:04:47 <fanquake> gdm85 Yes the cert problem is a known issue
136 2015-02-05 13:05:15 <gdm85> jonasschnelli: you have extra certs installed, I assume.
137 2015-02-05 13:05:35 <gdm85> https://www.ssllabs.com/ssltest/analyze.html?d=bitcoincore.org
138 2015-02-05 13:05:35 <gdm85> on a freshly debootstrapped debian wheezy you'll get:
139 2015-02-05 13:05:42 <gdm85> ERROR: The certificate of `bitcoincore.org' hasn't got a known issuer.
140 2015-02-05 13:05:42 <gdm85> ops..wrong paste: ERROR: The certificate of `bitcoincore.org' is not trusted.
141 2015-02-05 13:08:40 <gdm85> fanquake: I think it's really the root CA missing from Debian, otherwise you'd be able to fix this by adding the intermediates.
142 2015-02-05 13:08:52 <gdm85> (directly appended on the SSL cert)
143 2015-02-05 13:16:56 <u20024804> I tried upgrade bitcoin source code to 0.9.4 but still can't work; and then I downgrade my openssl to 1.0.1g but still can't work
144 2015-02-05 13:21:35 <wumpus> u20024804: did you restart the sync / reindex?
145 2015-02-05 13:26:53 <u20024804> Yes, I runed with -reindex -par=1
146 2015-02-05 13:29:23 <fanquake> jonasschnelli Do you have a linux sig I can compare too?
147 2015-02-05 13:30:02 <jonasschnelli> fanquake, sure: https://builds.jonasschnelli.ch/releasebuilds/v0.10.0rc4/bitcoin-linux-0.10-build.assert
148 2015-02-05 13:30:13 <jonasschnelli> fanquake, whole build is here: https://builds.jonasschnelli.ch/releasebuilds/v0.10.0rc4/
149 2015-02-05 13:30:36 <fanquake> jonasschnelli nice, sigs match :)
150 2015-02-05 13:31:22 <jonasschnelli> fanquake, see also https://github.com/bitcoin/gitian.sigs/pull/86/files
151 2015-02-05 13:33:26 <jonasschnelli> I'm trying to understand the net level better: when having a fresh main net sync. Why only the header at the top gets loaded? "initial getheaders (0) to peer=1 (startheight:342065)"?
152 2015-02-05 13:33:37 <jonasschnelli> Is startheight somehow related to key generation time?
153 2015-02-05 13:35:25 <michagogo> jonasschnelli: I think that's how many blocks that node had when you connected to it
154 2015-02-05 13:36:53 <jonasschnelli> michagogo, i thought header-first will first download all headers. But i just see one or two getheaders, ..the it starts getting blocks.
155 2015-02-05 13:39:47 <michagogo> jonasschnelli: pretty sure getheaders is "give me all the headers you have starting from x"
156 2015-02-05 13:41:13 <michagogo> So you're asking that peer for headers, it gives you all of them at once (I think), then you start getting blocks
157 2015-02-05 13:43:37 <jonasschnelli> michagogo, Yes. But why does it not start with header height 0? Instead it will start at the height received from the version message from the remote peer.
158 2015-02-05 13:44:13 <jonasschnelli> i always thought bitcoin will first form a complete chain of headers
159 2015-02-05 13:51:07 <michagogo> 15:33:42 <jonasschnelli> I'm trying to understand the net level better: when having a fresh main net sync. Why only the header at the top gets loaded? "initial getheaders (0) to peer=1 (startheight:342065)"?
160 2015-02-05 13:51:33 <michagogo> My interpretation of that message is that it's sending a getheaders starting with height 0
161 2015-02-05 13:52:02 <michagogo> To peer #1, and when we connected to that peer it had 342065 blocks in the best chain (in the version message)
162 2015-02-05 13:52:45 <jonasschnelli> michagogo, no. I had e empty data dir and this log message came at the very beginning.
163 2015-02-05 13:53:05 <michagogo> jonasschnelli: right...
164 2015-02-05 13:53:12 <michagogo> So it's asking for headers starting from 0
165 2015-02-05 13:53:41 <jonasschnelli> michagogo, i also thought that. But it start by loading header at height 342065
166 2015-02-05 13:53:53 <michagogo> Pretty sure the 342065 is how many blocks *peer=1* claimed to have in the version message when we connected to it
167 2015-02-05 13:56:05 <jonasschnelli> michagogo, Let me analyze this more deep. But there is no complete header-chain before loading blocks... because it takes 2,3 seconds, than it starts loading blocks (fresh main net sync).
168 2015-02-05 13:56:28 <michagogo> jonasschnelli: Hm, interesting
169 2015-02-05 13:56:36 <michagogo> Maybe try a fresh sync with -debug?
170 2015-02-05 13:56:44 <jonasschnelli> i did.
171 2015-02-05 14:00:49 <michagogo> Hm, I think this might be the first gbuild that I've done almost entirely from a bus
172 2015-02-05 14:01:04 <michagogo> (Or, a series of buses)
173 2015-02-05 14:17:28 <michagogo> Uh, what? o_O https://i.imgur.com/xwIaIFz.jpg
174 2015-02-05 14:20:35 <LeMiner> anyone have any idea why bitcoind doesn't start/crashes when running some low intensity applications in the background? (this is on a regulair-hdd-pc)
175 2015-02-05 14:21:01 <LeMiner> or more specifically, armory doesn't get an connection to bitcoin core when running these applications
176 2015-02-05 14:28:56 <michagogo> Okay, sigs up, PR #88