1 2013-11-25 00:07:10 <ers35> gmaxwell: I spun up a testnet node at 142.4.212.125:18333 or [2607:5300:60:117d::1]:18333 if you still need one.
 2 2013-11-25 00:09:15 <gmaxwell> ers35: thanks. :)
 3 2013-11-25 00:21:42 <warren> Ryan52: ping
 4 2013-11-25 00:41:23 <Ryan52> warren: pong. progress?
 5 2013-11-25 00:42:45 <Ryan52> warren: my mac sits here happily syncing from the network, with no corruption yet. I plan to throw some more force at it again tomorrow.
 6 2013-11-25 00:45:01 <diki> I just opened blockchain.info and was greeted by a Windows XP loading bar for a few seconds including the Windows XP logo, I was like wtf?
 7 2013-11-25 00:45:41 <Ryan52> 
ACTION heard the domain expired without anybody picking it up, people said it was broken anyways or something.

 8 2013-11-25 00:46:06 <diki> Ryan52:That was for blockexplorer.com
 9 2013-11-25 00:47:36 <Ryan52> diki: oh, right, sorry, my brain stopped at block...
10 2013-11-25 00:47:42 <Ryan52> 
ACTION feels silly now

11 2013-11-25 00:51:57 <Burritoh> diki: that happens when their site crashes and comes back.
12 2013-11-25 00:52:17 <diki> so it's intentional and not say added by a malicious person?
13 2013-11-25 00:52:48 <Burritoh> diki: yeah, intentional, if you are talking about the possibility of defacement
14 2013-11-25 00:53:00 <Burritoh> diki: slightly better than Bitstamp's "And it's gone" sparking panic all round.
15 2013-11-25 01:09:02 <diki> Anyway, pardon me for not searching but how do I read the blockchain by writing my own custom program?
16 2013-11-25 01:15:16 <gulli> Was wondering about extra security regarding HD wallets. Start using a node, say, 150 nodes down the tree as the starting point where the ecternal and internal chains start
17 2013-11-25 01:15:28 <gulli> Even if an attacked would figure out the seed
18 2013-11-25 01:15:40 <gulli> he would then need to find out what node to start at
19 2013-11-25 01:15:55 <sipa> that's security through obscurity
20 2013-11-25 01:16:14 <sipa> you've essemtially just made the secret be the seed + the starting poimt
21 2013-11-25 01:16:23 <sipa> without adding any security
22 2013-11-25 01:16:25 <gulli> yup
23 2013-11-25 01:16:49 <sipa> if you think your seed is not secure enough, make it longer
24 2013-11-25 02:22:47 <BW^-> BW^-: *,gmaxwell: what is meant in the topic with "There is no from address"?
25 2013-11-25 02:22:52 <BW^-> *,gmaxwell: what is meant in the topic with "There is no from address"?
26 2013-11-25 02:23:55 <gmaxwell> BW^-: why are you directing that at me?
27 2013-11-25 02:24:25 <gmaxwell> I didn't put that there, I just changed the topic last ... though I can explain it:
28 2013-11-25 02:25:35 <gmaxwell> Transactions don't contain any from address in them. Attempts to infer a 'from address' must use the historical transaction data and are risky to use for things like refunds because they sometimes result in sending funds to the wrong party or into a black hole, and people have lost money from doing that.
29 2013-11-25 02:25:51 <BW^-> how do you define transaction radar?
30 2013-11-25 02:26:18 <BW^-> mhm
31 2013-11-25 02:26:21 <gmaxwell> It's sometimes a FAQ in here where people show up asking where in a transaction they can read the 'from' from— to which the answer is: you can't, if you want to know who paid you the correct way to do that in bitcoin is to distinguish them based on what address they paid.
32 2013-11-25 02:26:55 <BW^-> gmaxwell: is there any annotation or alike tha tcould be added to a TX script to communicate to the receiving party which the refund-to address is *safely*?
33 2013-11-25 02:27:31 <BW^-> ] "...is to distinguish them based on what address they paid."
34 2013-11-25 02:27:36 <BW^-> gmaxwell: mhm, you mean, what address they paid *to*
35 2013-11-25 02:27:37 <BW^-> mhm
36 2013-11-25 02:28:45 <gmaxwell> BW^-: things like the payment protocol were created to provide that information out of band.  (though even without them the best practice is to simply require the sending party provide a refund address before you give them an address to pay to)
37 2013-11-25 03:08:19 <warren>  https://bitcointalk.org/index.php?topic=337294.0;all  MacOS X corruption fix bounty now increased to 10 BTC + 200 LTC thanks to new a pledge from BitcoinTalk.
38 2013-11-25 03:13:03 <gmaxwell> warren: can you update the original post
39 2013-11-25 03:13:23 <BW^-> what do we mean by "height" here?
40 2013-11-25 03:13:46 <BW^-> in the context of retrieving the history of all transactions that regard one single bitcoin address
41 2013-11-25 03:14:31 <warren> gmaxwell: yea
42 2013-11-25 03:14:33 <BW^-> i mean, are we talking about an integer index counter relative to now
43 2013-11-25 03:14:39 <gmaxwell> warren: and maybe submit to hacker news?  "Solve a stupid database corruption bug on OSX, recieve ~$10k in Bitcoin and Litcoin"
44 2013-11-25 03:14:40 <BW^-> or are we talking about a globally unique ID??
45 2013-11-25 03:14:52 <warren> gmaxwell: URL?
46 2013-11-25 03:15:24 <gmaxwell> BW^-: it's not clear enough to me what you're talking about, but usually when we talk about height we're talking about the absolute blockchain height (in blocks) starting from the genesis block.
47 2013-11-25 03:15:36 <gmaxwell> warren: https://news.ycombinator.com/submit  ?
48 2013-11-25 03:16:12 <BW^-> aha!!!
49 2013-11-25 03:16:20 <BW^-> gmaxwell: awesomeness
50 2013-11-25 03:16:57 <BW^-> gmaxwell: so this way, if you have made a safe retrieval of all transactions regarding a particular address to a particular height X
51 2013-11-25 03:17:23 <BW^-> then you know that afterwards you will ever only need to make new checks from that height X to whatever new height new blocks get to
52 2013-11-25 03:17:25 <BW^-> mhm
53 2013-11-25 03:17:56 <BW^-> gmaxwell: what about if there's blockchain forks that then heal themselves, that could cause lots of undefined qualities to any counting of heights couldn't it?
54 2013-11-25 03:18:12 <gmaxwell> BW^-: we rescan a bit before the recorded height in such cases.
55 2013-11-25 03:18:37 <BW^-> gmaxwell: when does that happen?
56 2013-11-25 03:18:52 <gmaxwell> though thats only an issue if the wallet and blockchain on a node have been seperated.
57 2013-11-25 03:19:06 <gmaxwell> otherwise the wallet would have witnessed any reorg regardless.
58 2013-11-25 03:20:57 <BW^-> gmaxwell: wait, can you please re-express that as an example?
59 2013-11-25 03:21:04 <BW^-> just to be sure i understood what you meant
60 2013-11-25 03:24:33 <diki> Anyway, pardon me for not searching but how do I read the blockchain by writing my own custom program?
61 2013-11-25 03:32:56 <warren> gmaxwell: could you please submit it there?
62 2013-11-25 03:33:53 <gmaxwell> warren: sure
63 2013-11-25 03:34:59 <kjj> diki: fopen()
64 2013-11-25 03:35:06 <diki> very funny
65 2013-11-25 03:35:15 <diki> I do know that blocks are concatenated
66 2013-11-25 03:35:22 <diki> but I'd need a bit more than that
67 2013-11-25 03:36:33 <gmaxwell> warren: https://news.ycombinator.com/newest
68 2013-11-25 03:36:33 <kjj> https://bitcointalk.org/index.php?topic=101514.0
69 2013-11-25 03:36:50 <gmaxwell> oh someone else did.. /me deletes mine
70 2013-11-25 03:37:10 <gmaxwell> haha there is a cypherpunks account on hackernews! :P
71 2013-11-25 03:38:33 <gmaxwell> https://news.ycombinator.com/item?id=6792325
72 2013-11-25 03:46:31 <phantomcircuit> gmaxwell, would anybody be opposed to adding some paranoid checks to leveldb?
73 2013-11-25 03:46:52 <phantomcircuit> i *might* be able to make them backwards compatible
74 2013-11-25 03:48:17 <phantomcircuit> mostly i want to add sequence numbers to the journal
75 2013-11-25 03:48:26 <phantomcircuit> which should also make a bunch of stuff safer
76 2013-11-25 03:48:37 <phantomcircuit> im actually kind of surprised they didn't already have them
77 2013-11-25 03:48:38 <gmaxwell> phantomcircuit: if they were compatible... well, (1) we could at least use them for debugging. (2) though I'd really like them to go upstream.
78 2013-11-25 03:49:03 <phantomcircuit> gmaxwell, i have not a clue how to get a patch accepted upstream
79 2013-11-25 03:49:09 <phantomcircuit> or rather even where to submit such a patch
80 2013-11-25 03:49:52 <gmaxwell> maybe first you have to get a job at google, as in some google open source projects. :P  (well, I think I've seen them take patches submitted on their bugtracker)
81 2013-11-25 03:50:24 <phantomcircuit> lolol
82 2013-11-25 03:52:45 <kjj> gmaxwell: do you have any scripts for BIP10?
83 2013-11-25 03:53:31 <gmaxwell> kjj: no. alas.
84 2013-11-25 03:53:52 <kjj> ok.  if all goes well tomorrow, I'm going to start on some
85 2013-11-25 03:55:10 <kjj> I just wish his example code wasn't in python
86 2013-11-25 03:56:12 <kjj> heh, turns out not to matter.  his example links are all 404 now