1 2015-02-02 04:32:09 <phantomcircuit> im trying to do ibd with loadblock to zfs with the zil on an ssd
  2 2015-02-02 04:33:06 <phantomcircuit> getting terrible performance
  3 2015-02-02 04:33:12 <phantomcircuit> there's seemingly no cpu time being used
  4 2015-02-02 04:33:18 <phantomcircuit> not even in io wait or anything
  5 2015-02-02 04:40:17 <phantomcircuit> ok this is weird
  6 2015-02-02 04:40:47 <phantomcircuit> at 129k and doing maybe 1 block/second
  7 2015-02-02 05:36:45 <phantomcircuit> scratch that i had the path wrong
  8 2015-02-02 08:46:45 <aburan28> if i have a nLockTime tx set to 1000 blocks in the future does that tx essentially sit in the mempool of nodes until 1000 blocks have been found?
  9 2015-02-02 08:56:09 <wumpus> it will only be accepted once it is final, which means that the nLockTime will have passed
 10 2015-02-02 08:56:26 <wumpus> so no, it won't sit there, you'll have to submit it after the nLocktime
 11 2015-02-02 08:56:47 <wumpus> (or someone else, e.g. a timed service)
 12 2015-02-02 09:22:02 <dansmith_> can anyone confirm that replace-by-fee has not yet made it into bitcoind?
 13 2015-02-02 09:22:17 <wumpus> it hasn't
 14 2015-02-02 09:25:05 <xabbix> is there a way to know the status of bitcoind? I've started a rescan and want to know if it completed or not (I don't have a gui available)
 15 2015-02-02 09:28:24 <aburan28> but when i create that transaction and send it to nodes do they add it into their mempool? or does it sit in my client until that block has passed and then send it out to nodes
 16 2015-02-02 09:28:40 <wumpus> through RPC you can query about everything with the get*info commands, and there's debug.log with even more information
 17 2015-02-02 09:29:21 <xabbix> wumpus, was that for me?
 18 2015-02-02 09:29:32 <wumpus> xabbix: I guess?
 19 2015-02-02 09:29:36 <xabbix> wumpus, thanks :)
 20 2015-02-02 09:29:42 <wumpus> aburan28: no, as I said, nodes won't take it into their mempool until nLockTime passed
 21 2015-02-02 09:30:44 <wumpus> aburan28: your node (or someone else) has to send it *after* that to get it mined, for DoS reasons (mempools are already huge) nodes won't mempool what they wouldn't mine within a few blocks
 22 2015-02-02 09:35:03 <aburan28> could a feature be added to validate the tx and if nLockTime is in the future it would then only store the txid for that tx?
 23 2015-02-02 09:35:32 <wumpus> how would that help?
 24 2015-02-02 09:44:15 <aburan28> well as of right now it stores the entire tx in the mempool and the hash is only 64 bytes
 25 2015-02-02 09:45:20 <wumpus> on the other hand, the hash is meaningless for any validation work, so may as well not store it
 26 2015-02-02 13:22:01 <iBog> Luke-Jr I think it was you that had some very strong critism of web wallets the other day.  How popular are they?  I'm curious to know what you think the specific risks are.
 27 2015-02-02 13:27:17 <hearn> they're either quite popular or very popular depending on how you define web wallet
 28 2015-02-02 13:32:48 <iBog> hearn good point.  I suppose it's a third party that stores your private keys and has the ability to sign transactions on your behalf?
 29 2015-02-02 13:33:02 <stonecoldpat> iBog: generally, any website that stores your private key (thus full access to your bitcoins) is unsafe / risky -  once their stolen then their gone, (unlike a bank who can just declare it as fraud/compensate you) although i dont think this is appropriate for bitcoin-dev?
 30 2015-02-02 13:33:28 <Luke-Jr> the topic is being discussed in #bitcoin more already
 31 2015-02-02 13:33:38 <iBog> stonecoldpat you're correct... I took it to a different channel
 32 2015-02-02 14:37:03 <Shaan> anydevelopers in here looking to do a bit of freelance work?, i need a plugin for handling payments for prestashop using BIP44
 33 2015-02-02 15:00:57 <unicodesnowman> Shaan, post it on bitcointalk
 34 2015-02-02 15:01:03 <unicodesnowman> i'm sure ppl will fund a bounty if it's made open source.
 35 2015-02-02 15:06:17 <Shaan> unicodesnowman: ive never done anything like it before, mind pointing me in the right direction
 36 2015-02-02 15:46:00 <freename> Using the RPC interface for bitcoin-qt, is it possible to get a block number from a transaction ID?  To say it another way, is it possible to figure out what block a transaction is a part of given a transaction ID (without downloading and indexing all the blocks and transactions first)?
 37 2015-02-02 15:46:39 <sipa> no
 38 2015-02-02 15:47:11 <sipa> well you can run with -txindex, which lets bitcoin core fo the indexing of transactions for you
 39 2015-02-02 15:47:27 <justanotheruser> freename: how do you expect to query a database you don't have?
 40 2015-02-02 15:49:03 <freename> justanotheruser:  Thanks for the question,  The RPC interface for bitcoin-qt has a database behind it.  I've turned on some additional indexing but there appears to be no RPC call to get a block from a transaction ID.
 41 2015-02-02 15:49:41 <freename> sipa:  Yea,  I've got -txindex on, but that doesn't seem to be enough.  I just wanted to make sure I've not missed something before I throw in the towel on the RPC approach.
 42 2015-02-02 15:49:42 <justanotheruser> freename: yes, but you qualified it with "without downloading and indexing all the blocks and transactions first".
 43 2015-02-02 15:49:42 <Luke-Jr> freename: with -txindex, use getrawtransaction
 44 2015-02-02 15:51:10 <freename> justanotheruser:  gott-cha... yea,  I meant something else there.  Sorry about that confusion.  I meant downloading all the transactions out of the RPC interface.  I've downloaded all of the data into bitcoin-qt and indexed it where possible.
 45 2015-02-02 15:51:31 <justanotheruser> freename: well like they said, you can do that with -txindex and a -reindex
 46 2015-02-02 15:51:38 <justanotheruser> gl;hf
 47 2015-02-02 15:51:44 <michagogo> freename: then yes, if txindex is on getrawtransaction should work
 48 2015-02-02 15:51:55 <michagogo> You'll want to tell it to be verbose
 49 2015-02-02 15:52:20 <michagogo> (I can never remember if you need to give it true, false, 1, or 0 to do that)
 50 2015-02-02 15:52:33 <freename> michagogo: Luke-Jr:  I'll take a look at that part of the documentation again and see what I missed.
 51 2015-02-02 15:52:55 <michagogo> freename: check `bitcoin-cli help getrawtransaction`
 52 2015-02-02 15:53:23 <michagogo> (Dropping the first part if you're using the debug window or a different RPC client)
 53 2015-02-02 15:53:40 <michagogo> Pretty sure that command does it as long as you tell it to be verbose
 54 2015-02-02 16:04:26 <rubensayshi> 64: non-mandatory-script-verify-flag (No error)
 55 2015-02-02 16:04:28 <rubensayshi> :(
 56 2015-02-02 16:05:49 <rubensayshi> I guess it's still something wrong with the signature, but what could it be if bitcoind fails to give a proper error message? if it's really a non canonical DER it seems to give me a error message stating exactly that
 57 2015-02-02 16:07:25 <freename> michagogo: sipa: Luke-Jr: justanotheruser:  Setting the verbose flag to 1 solved my problem.  Thanks for the help!
 58 2015-02-02 16:08:10 <michagogo> rubensayshi: I think that means it's non-standard
 59 2015-02-02 16:08:42 <michagogo> I'd say, go over the non-mandatory verification flags and see which one you fail
 60 2015-02-02 16:09:01 <rubensayshi> michagogo, I'm having this issue trying to sign the 2nd signature of a 2of3 using bitcoinlib-js (or sipa's secp256k1) and it's 'random'
 61 2015-02-02 16:09:13 <rubensayshi> with some privatekeys it fails, with others it passes
 62 2015-02-02 16:09:18 <michagogo> rubensayshi: what software are you running?
 63 2015-02-02 16:09:20 <rubensayshi> and I think the R and S might be negative
 64 2015-02-02 16:12:03 <rubensayshi> guess I gotta isolate it more so I can share a code snippet ...
 65 2015-02-02 16:12:22 <sipa> give an example signature?
 66 2015-02-02 16:13:59 <rubensayshi> https://gist.github.com/rubensayshi/bec320d12345d98c15da
 67 2015-02-02 16:14:13 <sipa> rubensayshi: if libsecp256k1 fails to sign with correct DER that would be a very unexpected bug...
 68 2015-02-02 16:14:40 <sipa> rubensayshi: that's not DER
 69 2015-02-02 16:14:59 <sipa> you need a padding byte for R to avoid it being interpreted as anegative number
 70 2015-02-02 16:15:03 <rubensayshi> yea, well I'm using nodejs bindings and just replaced some stuff in bitcoinlib-js to use the bindings
 71 2015-02-02 16:15:10 <rubensayshi> oh sec
 72 2015-02-02 16:15:30 <rubensayshi> that's from trying to debug what went wrong :/
 73 2015-02-02 16:16:52 <rubensayshi> ugh without that debug stuff this one works, gimme a sec to try a different one ...
 74 2015-02-02 16:20:12 <rubensayshi> now I'm getting "mandatory-script-verify-flag-failed (Script evaluated without error but finished with a false\/empty top stack element)" - I'll ping you if I have something you can work with
 75 2015-02-02 16:22:48 <sipa> rubensayshi: that means it's just invalid
 76 2015-02-02 16:24:20 <rubensayshi> yea, the sucky part is that I'm verifying them by doing sendrawtransaction and for the ones that failed in the past I've also tried them using a PHP Lib which uses a random K and that seemed to succeed
 77 2015-02-02 16:24:26 <rubensayshi> so I can't reuse my old testcases :/
 78 2015-02-02 16:24:48 <rubensayshi> I guess I could run it against secp256k1 verify method instead
 79 2015-02-02 16:32:57 <cfields> rubensayshi: you can use libbitcoinconsensus for that
 80 2015-02-02 17:20:52 <michagogo> Hm, apparently there's an
 81 2015-02-02 17:21:01 <michagogo> A RPi 2 now
 82 2015-02-02 17:21:25 <cfields> michagogo: runs on windows, even
 83 2015-02-02 17:21:35 <michagogo> Quad core ARMv7, 1GB RAM, same price
 84 2015-02-02 17:21:48 <michagogo> I wonder if one of those could run a node
 85 2015-02-02 17:22:08 <michagogo> cfields: oh? Didn't see that
 86 2015-02-02 17:22:38 <cfields> michagogo: http://dev.windows.com/en-us/featured/raspberrypi2support
 87 2015-02-02 17:23:11 <michagogo> ...I wonder if there's any other computer where you pay quadruple what the hardware costs for the OS license
 88 2015-02-02 17:23:36 <michagogo> Ah, cool.
 89 2015-02-02 17:23:40 <michagogo> Nvm, I guess.
 90 2015-02-02 17:23:46 <cfields> embrace...
 91 2015-02-02 20:20:34 <warren> michagogo: I suspect that thing doesn't have enough RAM to safely run bitcoind
 92 2015-02-02 20:21:20 <wumpus> it can, just make sure to dial down the dbcache
 93 2015-02-02 20:22:36 <wumpus> but yes, 2GB would be have been better for a bitcoin node
 94 2015-02-02 20:22:39 <sipa> i'm running a bitcoind with 100 connections with less than 500 MiB of ram
 95 2015-02-02 20:23:00 <sipa> with some tuned settings (lower rpcthreads, no parallellism, lower dbcache)
 96 2015-02-02 20:27:55 <WolfGoethe> wow. sipa. could u share those settings? we wanna get 3 bitcoin cores on a dedicated box with 8 ram
 97 2015-02-02 20:28:08 <sipa> i just told you
 98 2015-02-02 20:28:09 <gmaxwell> 8 ram?
 99 2015-02-02 20:28:13 <sipa> -rpcthreads=1 -par=1 -dbcache=10
100 2015-02-02 20:28:41 <wumpus> 3 bitcoin core? bitcoin multi-core?
101 2015-02-02 20:29:29 <WolfGoethe> 8 cores, 8 gigs of ram dedicated box
102 2015-02-02 20:29:33 <wumpus> see also https://gist.github.com/laanwj/efe29c7661ce9b6620a7
103 2015-02-02 20:29:43 <gmaxwell> uh you shouldn't need any special anything for that.
104 2015-02-02 20:29:45 <WolfGoethe> 3 seperate instances of bitcoincore 9.3
105 2015-02-02 20:29:51 <WolfGoethe> running fine now
106 2015-02-02 20:29:55 <gmaxwell> 8/3 = 2 2/3.
107 2015-02-02 20:30:00 <WolfGoethe> always looking to optimize
108 2015-02-02 20:30:24 <wumpus> well it's not really opimization, it uses less memory at the expense of performance
109 2015-02-02 20:31:09 <jikoz> I really need the ability to go `bitcoin-cli importprivkey <> ""  $RESCAN_FROM_THIS_BLOCK`
110 2015-02-02 20:31:22 <jikoz> Since I know exactly which bock the address got its first transaction
111 2015-02-02 20:31:26 <jikoz> and full scans are too slow
112 2015-02-02 20:31:36 <wumpus> e.g. less dbcache means more i/o, less parallelism means slower verification, etc
113 2015-02-02 20:32:07 <jikoz> There's no easy way to rescan from a particular block, I assume?
114 2015-02-02 20:32:09 <harding> jikoz: use importwallet instead.  The file format is human-editable
115 2015-02-02 20:32:34 <jikoz> I need to dynamically add new addresses
116 2015-02-02 20:32:39 <wumpus> yes, importwallet does support adding key birthdates
117 2015-02-02 20:32:44 <jikoz> oh!
118 2015-02-02 20:32:47 <jikoz> Let me take a look
119 2015-02-02 20:32:59 <sipa> if you want to know the file format, try fdumpwallet first
120 2015-02-02 20:33:33 <harding> jikoz: what sipa said, or there's an example of the file format here: https://bitcoin.org/en/developer-reference#dumpwallet
121 2015-02-02 20:33:58 <Eliel_> sipa: which of those parameters would be most useful to increase in case you do have a little more memory available?
122 2015-02-02 20:34:14 <wumpus> dbcache
123 2015-02-02 20:34:18 <phantomcircuit> wumpus, orphan blocks thing is gone already i think
124 2015-02-02 20:34:32 <sipa> Eliel_: don't do anything, the defaults are fine for like 800 MiB
125 2015-02-02 20:34:39 <sipa> in 0.10
126 2015-02-02 20:35:05 <wumpus> phantomcircuit: yes, good catch
127 2015-02-02 20:35:38 <jikoz> What's  `reserver=1` mean in the wallet dump?
128 2015-02-02 20:35:45 <jikoz> *reserve=1
129 2015-02-02 20:36:07 <sipa> that it's a reserve key
130 2015-02-02 20:36:18 <sipa> which is not to be exposed in the address book
131 2015-02-02 20:36:20 <jikoz> Oh, like an address that hasn't yet been used?
132 2015-02-02 20:36:22 <sipa> yes
133 2015-02-02 20:37:25 <jikoz> And if I call importwallet, and its already got that address -- it will just ignore it?
134 2015-02-02 20:37:31 <sipa> i believe so
135 2015-02-02 20:37:33 <jikoz> And only rescan for the new addresses it hasn't seen
136 2015-02-02 20:37:47 <wumpus> better to just not add anything to the file that you don't want imported
137 2015-02-02 20:38:10 <sipa> jikoz: it rescans for everything, but only goes back as far as it needs
138 2015-02-02 20:38:22 <jikoz> Ok, perfect. I'll only add the new addresses
139 2015-02-02 20:38:36 <jikoz> Ha, this is easier than I thought. I was expecting to need to patch my bitcoind
140 2015-02-02 21:04:10 <n01d3x> im playing around with a pos alt coin and want it to only stake wallets where nBalance > 99999 and the vary the rate based on the year. I have the varying rate set I'm just not sure how to get the code to check the nBalance then move on to the varying rate code or return 0
141 2015-02-02 21:04:46 <n01d3x> of anyone can help please pm me and ill show you what ive got so far
142 2015-02-02 21:04:48 <teward> "alt coin" - wrong channel
143 2015-02-02 21:05:24 <n01d3x> figured as much but didn't know of any other channels to look for help
144 2015-02-02 21:05:29 <n01d3x> google just kept returning here
145 2015-02-02 22:46:48 <jikoz> What date should I use, for the address in `bitcoin-cli importwallet`? I know exactly which block the address was first seen
146 2015-02-02 22:47:02 <jikoz> Should I get the block it was first seen's timestamp, and subtract 1 second ?
147 2015-02-02 22:47:33 <jikoz> Or do I need to make it in local time? (like the time my bitcoind saw it?)
148 2015-02-02 22:52:36 <jikoz> Ah, never mind. Found it in the code
149 2015-02-02 22:52:36 <jikoz> thanks
150 2015-02-02 22:59:21 <earlz> So, what is a decent way to store and query blockchain data?
151 2015-02-02 22:59:25 <earlz> A faster way than just scanning through the blocks
152 2015-02-02 22:59:42 <earlz> The data isn't really well structured for SQL..
153 2015-02-02 23:00:23 <earlz> like tracking origin of transactions through multiple levels, counting the amount of coins an address "owns" (total of vouts) at a height/time.. etc
154 2015-02-02 23:01:02 <kanzure> what's wrong with sql here?
155 2015-02-02 23:01:26 <kanzure> https://github.com/monetizeio/sqlalchemy-bitcoin
156 2015-02-02 23:01:39 <earlz> If you want to query the origin of a transaction you eithe rhave to do super deep joins or multiple round trips..
157 2015-02-02 23:02:21 <kanzure> hardly, you can just track origins
158 2015-02-02 23:02:43 <kanzure> also, you can use things like postgresql scripting (at the database level) which is very speedy
159 2015-02-02 23:02:51 <kanzure> but i doubt that is necessary here
160 2015-02-02 23:03:46 <earlz> I'm just wondering if there is some other DB that might be better suited for the data than SQL..
161 2015-02-02 23:04:11 <kanzure> well there's always berkeleydb
162 2015-02-02 23:04:19 <kanzure> </snark>
163 2015-02-02 23:04:21 <earlz> I have a working script that uses a similar structure actually.. but querying it was painfully difficult, and indexing it wasn't trivial
164 2015-02-02 23:04:27 <earlz> bleh
165 2015-02-02 23:04:56 <earlz> I'm wondering if it keyvalue with mapreduce might be able to do anything for it
166 2015-02-02 23:32:37 <sipa> Luke-Jr: do you think the 0.10 release notes should include something about block proposals?
167 2015-02-02 23:33:33 <Luke-Jr> hm, dunno. it's mainly only useful to pools right now.
168 2015-02-02 23:33:52 <sipa> public awareness may not be bad
169 2015-02-02 23:34:22 <Luke-Jr> perhaps
170 2015-02-02 23:34:38 <Luke-Jr> 
ACTION checks out the current relnotes

171 2015-02-02 23:35:00 <Luke-Jr> does git have the latest?
172 2015-02-02 23:36:02 <sipa> yes, afaik, in the 0.10 branch
173 2015-02-02 23:51:23 <Luke-Jr> sipa: https://github.com/bitcoin/bitcoin/pull/5740/files