1 2015-12-11 00:03:16 <desbest> hi i have a question
 2 2015-12-11 00:03:55 <desbest> if bitcoin is decentralised, how is the bitcoin source code updated and pushed out to all the machines?
 3 2015-12-11 00:04:30 <katu> it isnt.
 4 2015-12-11 00:04:31 <moa> it is not pushed it is pulled by the node operaotrs
 5 2015-12-11 00:13:42 <Luke-Jr> desbest: this is why there is no auto-update
 6 2015-12-11 00:15:32 <gmaxwell> desbest: its upgraded by indivigual users on their own terms and schedule.
 7 2015-12-11 00:23:18 <desbest> if i download the windows client that stores parts of the blockchain on my computer, will my computer need to manually update the bitcoin source code?
 8 2015-12-11 00:25:53 <gmaxwell> desbest: you will need to update it from time to time; just like any other program that doesn't automatically update.
 9 2015-12-11 00:31:03 <desbest> for domain names we have root servers, when confirming a transaction happened, how does bitcoin know which computer to look for the transaction on?
10 2015-12-11 00:31:55 <gmaxwell> desbest: bitcoin is a broadcast system; the entire bitcoin state is synchronized on every participating node.
11 2015-12-11 00:32:18 <desbest> sounds complicated
12 2015-12-11 00:32:20 <gmaxwell> So the answer is any system has it because every system must (eventually) have it.
13 2015-12-11 00:32:45 <gmaxwell> desbest: a little, the implementation is simpler than all the implications.
14 2015-12-11 03:13:20 <florianh> Although we all desire some level of credit, we have moved past many of these things...
15 2015-12-11 04:51:21 <Lightsword> oh, another un-upgraded pool https://www.blocktrail.com/BTC/pool/tbdice
16 2015-12-11 04:51:39 <Lightsword> technobits solopool
17 2015-12-11 04:53:20 <Lightsword> ghash still mining v3 blocks as of 6 hours ago https://www.blocktrail.com/BTC/block/00000000000000000902cd3e4ec247c9caa7b8db50bb3f7e5967e6f7ec09db97
18 2015-12-11 04:56:24 <Lightsword> I picked it up from nl1.ghash.io:3333 first on poolbench but uk1.ghash.io:3333 was pretty close as well, so probably one of their EU servers
19 2015-12-11 05:06:00 <btcdrak> Lightsword: I contacted ghash and they thought they had completed their upgrade in November.
20 2015-12-11 05:06:09 <Lightsword> lol
21 2015-12-11 05:06:15 <btcdrak> That was about 6 hours ago, so they should not be producing any more from now
22 2015-12-11 05:06:52 <Lightsword> did they confirm they are fully upgraded?
23 2015-12-11 05:06:55 <btcdrak> At this stage I think if a miner misses a node and that causes an orphan block, well, Natural Selection FTW
24 2015-12-11 05:07:11 <btcdrak> people need to get burned sometimes to get some responsibility...
25 2015-12-11 05:07:26 <Lightsword> I’m worried the chinese miners will pick it up and bounce it around on their SPV links
26 2015-12-11 05:07:35 <btcdrak> Lightsword, ikr?! i mean, half their blocks were v3, it's just ridiculous
27 2015-12-11 05:56:13 <Luke-Jr> Lightsword: that'd be nice. the more losses that gets, the more likely it will get turned off
28 2015-12-11 05:57:01 <Luke-Jr> btcdrak: maybe point out that it's not only bitcoind that needs upgrading
29 2015-12-11 05:57:50 <btcdrak> Luke-Jr: clearly their staff need upgrading too ^.^
30 2015-12-11 05:57:58 <Luke-Jr> btcdrak: and Eloipool
31 2015-12-11 05:58:23 <btcdrak> is that what they are using?
32 2015-12-11 05:58:43 <Luke-Jr> dunno
33 2015-12-11 05:59:23 <Luke-Jr> but it would explain thinking they upgraded when they didn't
34 2015-12-11 06:01:10 <btcdrak> ok I'll pass that on too just in case.
35 2015-12-11 06:02:27 <Lightsword> are you talking about ghash?
36 2015-12-11 06:02:44 <Luke-Jr> yes
37 2015-12-11 06:02:53 <Lightsword> dont’t think they are eloipool
38 2015-12-11 06:03:07 <Lightsword> think they are custom
39 2015-12-11 06:03:59 <Lightsword> they still probably just missed a node, not like all of their blocks are v3
40 2015-12-11 13:50:07 <desbest> when bitcoin confirms whether a transaction happened, which machine does it look for first?
41 2015-12-11 13:50:50 <instagibbs> desbest, you might want to try #bitcoin
42 2015-12-11 13:51:23 <desbest> okay
43 2015-12-11 15:04:03 <jl2012> slush also made an empty v3 yesterday
44 2015-12-11 15:16:58 <btcdrak> Lightsword: Luke-Jr: just FYI, TBDice pool will shutdown, so they dont intend upgrading. I guess they'll turn off hashing at enforcement on ~Monday
45 2015-12-11 15:19:12 <btcdrak> Luke-Jr: ghash.io mailed me to say they found their v3 node finally >.>
46 2015-12-11 15:30:39 <jl2012> btcdrak: do you know about slush's recent v3 empty block? https://www.blocktrail.com/BTC/block/0000000000000000037a253d53650c70b85092bf416934ca69fbe043b7f41437
47 2015-12-11 15:31:40 <btcdrak> jl2012 just ask them in their mining channel
48 2015-12-11 15:31:50 <btcdrak> #mining.bitcoin.cz
49 2015-12-11 15:32:05 <btcdrak> maybe they missed a node <.<
50 2015-12-11 19:21:04 <Lightsword> slush still missing blocks?
51 2015-12-11 19:21:16 <Lightsword> or missing v3 nodes I mean
52 2015-12-11 19:22:30 <gmaxwell> no biggy, the world will not end with a few orphaned blocks.
53 2015-12-11 19:43:26 <paulo_> is it possible that SHA2 doesn't generate a low enough hash?
54 2015-12-11 19:45:25 <paulo_> like at some point, it's not possible to have n leading digits as zero?
55 2015-12-11 19:46:12 <sipa> maybe, but that would mean the hash function is broken
56 2015-12-11 19:57:26 <arubi> well what about sha256d?  I guess it's possible that a 'low' enough hash couldn't be generated by trying all inputs that are 32 bytes long, right?
57 2015-12-11 19:57:56 <arubi> considering collisions
58 2015-12-11 20:06:48 <paulo_> well, you can change the hash by changing the timestamp, or changing the transactions you include
59 2015-12-11 20:08:31 <phantomcircuit> arubi, that would basically mean the hash function is broken
60 2015-12-11 20:08:50 <phantomcircuit> also do you have any idea how long it would take to even try all 2^256 inputs? :)
61 2015-12-11 20:08:51 <arubi> phantomcircuit, why though?
62 2015-12-11 20:09:04 <arubi> I do have an idea, I'm not worried :)
63 2015-12-11 20:09:24 <gmaxwell> at least if N is not very close to 256 bits, the hashfunction would have to be quite non-uniform to do that.
64 2015-12-11 20:09:29 <arubi> phantomcircuit, it could be that none of "exactly 32 bytes" inputs can be hashed to "such a low hash"
65 2015-12-11 20:10:41 <arubi> okay, just theories, but fun to think about :)
66 2015-12-11 20:33:11 <phantomcircuit> arubi, yes and in theory my u3 could find ten million blocks in a row
67 2015-12-11 20:33:23 <phantomcircuit> but such a thing is very unlikely
68 2015-12-11 20:39:29 <arubi> right, phantomcircuit, I agree.
69 2015-12-11 22:55:01 <coin_trader> hey all -- humble request..... in a future version of core, may we please split RPC events to its own log file, and segregate that from the catch-all debug.log? it would make much easier post-mortem analysis if relevant events are not included with all the "rejected dust" and mempool operations, finding blocks, new peers, etc...
70 2015-12-11 23:05:28 <sipa> coin_trader: sounds reasonable to me; do you want to file an issue?
71 2015-12-11 23:07:01 <gmaxwell> instagibbs: ^
72 2015-12-11 23:09:46 <moa> event.log?
73 2015-12-11 23:10:52 <coin_trader> sipa, no clue how to do that.... i'm admittedly a noob on git :(
74 2015-12-11 23:11:17 <coin_trader> and i'd rather not muck up the repo with the wrong thing
75 2015-12-11 23:12:19 <coin_trader> i was thinking more like RPCdebug.log or something with "rpc" in the filename, but whatever gets the job done i'm all for :)
76 2015-12-11 23:13:24 <sipa> coin_trader: i mean file an issue (so it can be tracked and implemented by anyone) on https://github.com/bitcoin/bitcoin/issues
77 2015-12-11 23:13:50 <sipa> not write a patch yourself (though you're welcome to try that too, of course)
78 2015-12-11 23:14:26 <coin_trader> ah ha - ok gotcha, yea i'll do the request.... i dont know c, i'm only partially knowledgable on python
79 2015-12-11 23:17:40 <coin_trader> done & done. https://github.com/bitcoin/bitcoin/issues/7199