1 2013-12-13 00:09:47 <jgarzik> LGTM and LGBT are too close to each other
2 2013-12-13 00:09:53 <jgarzik> I keep thinking Mike is gay
3 2013-12-13 01:08:14 <BlueMatt> ACTION starts to write LGBT on pull requests to confuse jgarzik even more
4 2013-12-13 01:08:17 <BlueMatt> awww, he left :(
5 2013-12-13 01:10:59 <saracen> Are the blk0001.dat etc. files still used with the latest bitcoin-qt?
6 2013-12-13 01:11:21 <saracen> The ones in the root of the folder, not under blocks/
7 2013-12-13 01:17:21 <btcdev9991> any objections to use libcurl in my daemon?
8 2013-12-13 01:23:25 <btcdev9991> why doesn't my project compile with libcurl library imports? saying undefined .. i do a -l curl to link it in and include curl.h in the right place...
9 2013-12-13 01:41:14 <__alp__> Got a question about bitcoind, trying to take a txid and figure out what block it was in. Don't see any good way to do that other than brute force
10 2013-12-13 01:41:19 <__alp__> Am I missing something?
11 2013-12-13 01:47:04 <kjj> gettransaction returns the block hash
12 2013-12-13 01:48:36 <__alp__> I swear I couldnt get gettransaction to work from python
13 2013-12-13 01:48:39 <__alp__> let me double check
14 2013-12-13 01:49:05 <kjj> you need to enable transaction indexing if you want it to work on non-wallet transactions
15 2013-12-13 01:50:25 <__alp__> thought I did that and reindexed
16 2013-12-13 01:50:44 <__alp__> probably screwed it up
17 2013-12-13 01:50:57 <__alp__> txindex=1
18 2013-12-13 01:53:47 <__alp__> any way to validate I set up non-wallet tx correctly?
19 2013-12-13 01:53:54 <__alp__> conf file seems to be set up right
20 2013-12-13 01:54:55 <kjj> did you stop, then start with -reindex ?
21 2013-12-13 01:55:13 <__alp__> yes, and it took all day to reindex
22 2013-12-13 01:55:51 <__alp__> .bitcoin/blocks/index directory is 1.5GB now
23 2013-12-13 01:57:04 <kjj> can you gettransaction on 1e19279f6925f12073bdbf48bdc377932320870f3ad1029ac14a1b93a8571ba4 ?
24 2013-12-13 01:57:51 <__alp__> raise JSONRPCException(response['error'])
25 2013-12-13 01:57:52 <__alp__> bitcoinrpc.authproxy.JSONRPCException
26 2013-12-13 01:58:07 <kjj> how about from the command line?
27 2013-12-13 01:58:36 <__alp__> 
28 2013-12-13 01:58:36 <__alp__> 19:58:31
29 2013-12-13 01:58:36 <__alp__> gettransaction '1e19279f6925f12073bdbf48bdc377932320870f3ad1029ac14a1b93a8571ba'
30 2013-12-13 01:58:36 <__alp__> Invalid or non-wallet transaction id (code -5)
31 2013-12-13 01:58:45 <__alp__> guess I need to reindex
32 2013-12-13 02:02:28 <__alp__> give it another shot reindexing I suppose
33 2013-12-13 02:13:07 <trixisowned> anyone know of p2pools code?
34 2013-12-13 02:13:12 <trixisowned> what exactly are peers?
35 2013-12-13 02:13:19 <trixisowned> "0 out, 0 in"
36 2013-12-13 02:13:35 <kjj> peers are other p2pool nodes that you are connected to
37 2013-12-13 02:13:50 <trixisowned> ah
38 2013-12-13 02:13:58 <trixisowned> so if there are no other p2pool nodes for the coin
39 2013-12-13 02:14:04 <trixisowned> its of course gonna say 0 out 0 ib
40 2013-12-13 02:14:05 <trixisowned> in*
41 2013-12-13 02:14:23 <kjj> right
42 2013-12-13 02:22:10 <abishek> for querying bitcoin daemon from a remote server using curl, should we still have to have rpcallowip setup?
43 2013-12-13 02:23:24 <BlueMatt> you probably shouldnt do that, period
44 2013-12-13 02:23:44 <BlueMatt> well, at least with strict firewall rules to enforce connection limits (dont rely on bitcoind to do that)
45 2013-12-13 02:36:22 <Morblias> hey guys, first time running a full bitcoin node in vps, and am trying to figure out how to update bitcoind. do i just need to replace /usr/bin/bitcoind with the most recent one or is there more to upgrading it?
46 2013-12-13 02:36:39 <BlueMatt> that should do it
47 2013-12-13 02:36:50 <abishek> ok, where can i find the latest linux binaries of bitcoin daemon?
48 2013-12-13 02:37:56 <Morblias> ahh, sweet. i was going at it the hard way trying to build from source and stuff and it wouldn't work. replacing the file is a lot easier! lol. thanks!
49 2013-12-13 02:37:59 <Morblias> abishek: http://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.8.6/
50 2013-12-13 02:38:13 <abishek> Morblias, thnx
51 2013-12-13 02:38:16 <BlueMatt> PLEASE VERIFY SIGNATURES WHEN DOWNLOADING
52 2013-12-13 02:38:32 <BlueMatt> the software is, after all, holding your money...
53 2013-12-13 02:41:55 <saracen> ACTION wonders if bitcoin-qt will ever auto update
54 2013-12-13 02:42:21 <abishek> BlueMatt, if my bitcoind is secured behind a firewall, what are the secured options to query bitcoind from a remote server?
55 2013-12-13 02:43:09 <BlueMatt> sacrelege: no
56 2013-12-13 02:43:26 <BlueMatt> saracen: ^
57 2013-12-13 02:43:54 <BlueMatt> abishek: I think you can still do it over ssl, but your best bet is really to just do it through an ssh tunnel or so
58 2013-12-13 02:44:04 <BlueMatt> abishek: or, if it doesnt hold money, meh, do whatever
59 2013-12-13 02:44:14 <BlueMatt> though it isnt too hard to dos bitcoind if you can get to the rpc port iirc
60 2013-12-13 02:44:26 <BlueMatt> though maybe thats better now
61 2013-12-13 02:45:28 <sacrelege> BlueMatt kind of funny, switching irc window and reading "NO!" ^^
62 2013-12-13 02:46:01 <BlueMatt> sacrelege: well, bitcoin-qt will hopefully notify users of new releases someday
63 2013-12-13 02:46:04 <BlueMatt> but auto-update, no
64 2013-12-13 02:46:06 <BlueMatt> never
65 2013-12-13 02:48:01 <saracen> BlueMatt: Why? (I'm not arguing that it should feature auto-update, just curious).
66 2013-12-13 02:49:20 <BlueMatt> I think everyone 'round here agrees that some level of user interaction should be required to update
67 2013-12-13 02:49:44 <BlueMatt> if nothing more than "an update is available, click here"
68 2013-12-13 02:49:52 <BlueMatt> bitcoin does, after all, hold your money
69 2013-12-13 02:50:12 <BlueMatt> so users who want to (everyone should, no one does) can go read the update changelog and look at source to make sure its all ok
70 2013-12-13 02:50:43 <saracen> After the "click here" step though, is there any issues from it then automatically installing the update (perhaps showing the changelog, like a Steam restart does)?
71 2013-12-13 02:50:59 <BlueMatt> no, I'd like to see exactly that
72 2013-12-13 02:51:08 <BlueMatt> (assuming click here also does sig verification and all that good stuff)
73 2013-12-13 03:02:23 <go1111111> can I simulate coin control using the Bitcoin-QT UI, by relying on how Bitcoin-QT constructs transactions? For instance if I have unspent outputs of sizes 2, 6, 7, and I try to send 7 BTC, can I rely on the client using the output with 7 coins, because it tries to use as few outputs as possible? In other words, can the heuristic that the client uses to select inputs be described in a couple sentences?
74 2013-12-13 03:04:11 <pigeons> not the answer to the question you asked but git head has a coin control ui
75 2013-12-13 03:06:05 <go1111111> cool.. i'll try to fix my compilation issues then. thanks
76 2013-12-13 04:15:26 <Luke-Jr> go1111111: why not just use coin control?
77 2013-12-13 04:15:46 <Luke-Jr> go1111111: there's a port for 0.8.x, and it's in git master..
78 2013-12-13 04:18:29 <go1111111> Luke-Jr: thats what I plan to do -- wasn't aware of that before. thanks!
79 2013-12-13 05:36:49 <Zoop_> http://www.pcworld.com/article/2080041/bitcoin-market-price-app-bitcoin-alarm-is-carefully-cloaked-malware.html
80 2013-12-13 05:37:00 <Zoop_> seems relevant enough to post to the community
81 2013-12-13 05:37:29 <Zoop_> be careful
82 2013-12-13 05:37:50 <Zoop_> your bitcoins are on the line
83 2013-12-13 05:38:19 <kjj> I don't even like going to websites about bitcoin. who downloads and runs random software?
84 2013-12-13 05:44:56 <null> you're our millionth visitor! click here to download 10 bitcoin!
85 2013-12-13 05:45:56 <Plasmastar> lol
86 2013-12-13 05:47:04 <Zoop_> http://xkcd.com/570/
87 2013-12-13 05:47:09 <Zoop_> comes to mind :)
88 2013-12-13 07:24:31 <Alina-malina> can anyone please explain me how that casacious coins worked?
89 2013-12-13 07:25:21 <wumpus> they were basically just paper wallets
90 2013-12-13 07:25:35 <Alina-malina> oh
91 2013-12-13 07:25:49 <Alina-malina> it contains characters like password right and nothing else?
92 2013-12-13 07:26:07 <wumpus> the public key is in the open and the private key is hidden behind some layer
93 2013-12-13 07:26:36 <Alina-malina> you mean and public key and private key are on the same coin?
94 2013-12-13 07:27:31 <wumpus> yes
95 2013-12-13 07:27:36 <Alina-malina> oh
96 2013-12-13 07:28:05 <Alina-malina> so for each coin the public and private key are different right?
97 2013-12-13 07:28:50 <wumpus> yup, each coin is unique
98 2013-12-13 07:29:30 <Alina-malina> are those public keys of those coins available publicly somewhere else rather then printed on coins?
99 2013-12-13 07:29:33 <wumpus> see the photo here http://www.coindesk.com/defcon-hackers-crack-physical-bitcoin-casascius-coins/
100 2013-12-13 07:30:35 <wumpus> yes, http://casascius.uberbills.com/
101 2013-12-13 07:31:50 <Alina-malina> hmm so how many characters can contain maximum the priv key?
102 2013-12-13 07:36:39 <wumpus> Alina-malina: https://en.bitcoin.it/wiki/Mini_private_key_format
103 2013-12-13 07:37:16 <Alina-malina> oh ok thanks!
104 2013-12-13 07:43:25 <Zoop_> Alina trying to crack a casacious coin...
105 2013-12-13 07:45:16 <Apocalyptic> more nothing more than a Malina
106 2013-12-13 07:57:06 <go1111111> what's the process by which the bitcoin ppa gets upgraded to 0.8.6? https://launchpad.net/~bitcoin/+archive/bitcoin
107 2013-12-13 08:03:48 <wumpus> go1111111: ping BlueMatt
108 2013-12-13 09:03:48 <Alina-malina> Zoop_, what you mean trying to crack a casacious coin?
109 2013-12-13 10:23:44 <Guest68621> hey, is there some brief how to how to added fully syn chain to bitcoin dir?
110 2013-12-13 10:30:04 <Guest68621> :)
111 2013-12-13 10:31:47 <jouke> Has there been some changes to the way the ssl certificates are handled in 8.6? Normally I would connect to the node via Ip-addresses, but since 8.6 curl warns about the hostname not belonging to the certificate.
112 2013-12-13 11:59:47 <michagogo> cloud|__alp__, kjj: gettransaction only ever works for transactions in the wallet
113 2013-12-13 12:00:23 <michagogo> cloud|funky3: copy in the blocks/ and chainstate/ directories
114 2013-12-13 12:03:29 <michagogo> cloud|"
115 2013-12-13 12:03:29 <michagogo> cloud|blacklisted it so that future versions are not synced from Debian."
116 2013-12-13 12:03:29 <michagogo> cloud|included in Ubuntu. I've removed the package from trusty now and
117 2013-12-13 12:03:29 <michagogo> cloud|Since this package is in unstable only, I agree that it should not be
118 2013-12-13 12:10:02 <michagogo> cloud|actively harmful, we could accept an SRU that disables the problematic parts
119 2013-12-13 12:10:02 <michagogo> cloud|on upgrade (with a suitable notice)."
120 2013-12-13 12:10:02 <michagogo> cloud|releases. If there are versions of the package in stable releases that are
121 2013-12-13 12:10:02 <michagogo> cloud|"Unfortunately, it is not feasible to remove the package from stable
122 2013-12-13 12:17:29 <Blaster> Hmm, my Ubuntu testnet node experiences a delay before it receives payment notifications, of sometimes up to 5 minutes. I am trying to build a payment service that allows someone access to a resource immediately upon payment and it doesn't work very well with that delay. Is there any way to resolve it? Will it resolve when I move from testnet to normal?
123 2013-12-13 12:21:31 <michagogo> cloud|Blaster: if the transaction isn't being relayed to your node at all, there's not much you can do
124 2013-12-13 12:21:56 <Blaster> michagogo|cloud: why wouldn't it be relayed?
125 2013-12-13 12:22:04 <michagogo> cloud|(Besides wait for either the transaction, or a block containing it, to arrive at your node
126 2013-12-13 12:22:07 <michagogo> cloud|)
127 2013-12-13 12:22:20 <michagogo> cloud|There are many possible reasons.
128 2013-12-13 12:22:40 <Blaster> that's dumb, this is supposed to programmable money, instantaneous.
129 2013-12-13 12:23:33 <michagogo> cloud|Blaster: many things can affect the speed of a transaction's propagation
130 2013-12-13 12:23:54 <michagogo> cloud|The size, the transaction fee, and many others
131 2013-12-13 12:24:21 <Blaster> but the node you're sending payment to should be aware of the payment immediately shouldn't it?
132 2013-12-13 12:24:44 <michagogo> cloud|Blaster: if the transaction gets relayed, yes
133 2013-12-13 12:25:14 <michagogo> cloud|Is there anything unusual about the transaction?
134 2013-12-13 12:25:19 <Blaster> michagogo|cloud: no
135 2013-12-13 12:25:23 <michagogo> cloud|Does it pay a transaction fee?
136 2013-12-13 12:25:25 <Blaster> except that it's on testnet
137 2013-12-13 12:25:30 <Blaster> yes .0001
138 2013-12-13 12:26:51 <michagogo> cloud|Are the sending and receiving nodes both well-connected?
139 2013-12-13 12:27:26 <Blaster> they are both on the same network
140 2013-12-13 12:27:47 <Blaster> The receiving node my Vbox VM
141 2013-12-13 12:28:58 <michagogo> cloud|ACTION shrugs
142 2013-12-13 12:29:34 <michagogo> cloud|I assume they're fully synced?
143 2013-12-13 12:30:17 <Blaster> michagogo|cloud: yup
144 2013-12-13 12:31:07 <tholenst> I don't think i'll be able to help you, but... what does it mean when you say "it receives payment notifications"? Is that the time you wait until the payment is confirmed?
145 2013-12-13 12:32:22 <Blaster> the point when walletnotify is called, it shows a notification bubble, and it shows the payment in recent transactions with 0 confirmations
146 2013-12-13 12:33:28 <Blaster> I was thinking that should take place immediately when a payment is sent to one of my addresses
147 2013-12-13 12:34:25 <tholenst> I would think so too, but mind you, I know almost nothing :)
148 2013-12-13 12:39:21 <tholenst> If you like, I will try running walletnotify on my testnet node and see if I observe the same, but -- I don't actually see how this would help
149 2013-12-13 13:16:26 <__alp__> Ugh, just spent all night running reindexing (-reindex) and still gettransaction doesnt work
150 2013-12-13 13:23:20 <darsie> hi
151 2013-12-13 13:24:34 <darsie> Is it possible to use private keys that are to large or too small? 0 or >0xfffffffffffffff3623462346342 (or whatever the limit is)
152 2013-12-13 13:25:45 <darsie> The idea is to create a bitcoin address of PK 0 and use that to reliably and trustworthily distroy btc.
153 2013-12-13 13:26:05 <darsie> destroy
154 2013-12-13 13:27:47 <wumpus> in that case you should use public key 0, private key 0 would be trivial for everyone to claim
155 2013-12-13 13:28:19 <tholenst> See address 1111111111111111111114oLvT2, but I think the proper way to destroy bitcoins is OP_RETURN
156 2013-12-13 13:29:01 <darsie> wumpus: So even PK 0 could be used to sign a tx?
157 2013-12-13 13:29:46 <wumpus> darsie: sure
158 2013-12-13 13:30:03 <wumpus> I guess the first key here is key 0 http://directory.io/
159 2013-12-13 13:30:19 <darsie> Why is it exluded then from key generators?
160 2013-12-13 13:30:41 <wumpus> is it?
161 2013-12-13 13:31:10 <darsie> Specifically, any 256-bit number between 0x1 and 0xFFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFE BAAE DCE6 AF48 A03B BFD2 5E8C D036 4141 is a valid private key.
162 2013-12-13 13:31:56 <darsie> https://en.bitcoin.it/wiki/Private_key
163 2013-12-13 13:32:19 <darsie> Well, it doesn't say the keys outside this range are invalid ...
164 2013-12-13 13:32:52 <wumpus> oh right, I think I remember, private key 0 wouldn't provide any security because it will stay 0 under the ECDSA manipulations
165 2013-12-13 13:33:15 <wumpus> after all it's a multiply
166 2013-12-13 13:33:36 <__alp__> It's not 0 in that case, right? It's just G?
167 2013-12-13 13:33:39 <darsie> ic. So there isn't really a PK that can't be used for signing?
168 2013-12-13 13:33:52 <darsie> sorry, secret key
169 2013-12-13 13:33:55 <darsie> SK
170 2013-12-13 13:34:44 <darsie> ahh, both private and public abbreveviate to P ... inconvenient.
171 2013-12-13 13:35:17 <Tyupi> hi
172 2013-12-13 13:42:33 <darsie> How about 0xFFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF 1234? Can you make valid signatures with that?
173 2013-12-13 13:43:07 <michagogo> cloud|darsie: some software will simply wrap around
174 2013-12-13 13:43:22 <notthemessiah> So, running the qt client on my 2005 laptop seems to crash the computer while indexing the block chain.
175 2013-12-13 13:44:02 <michagogo> cloud|darsie: if you want to destroy bitcoins, send them to a scriptPubKey of OP_RETURN
176 2013-12-13 13:44:28 <darsie> michagogo|cloud: Ok. I don't understand that, though.
177 2013-12-13 13:44:36 <darsie> thx
178 2013-12-13 13:44:54 <michagogo> cloud|OP_RETURN makes a transaction's validation fail
179 2013-12-13 13:45:24 <michagogo> cloud|Any output with a scriptPubKey that is, or begins with, OP_RETURN is provably unspendable
180 2013-12-13 13:45:43 <michagogo> cloud|And therefore can be pruned from the utxo set
181 2013-12-13 13:46:01 <michagogo> cloud|__alp__: gettransaction is a wallet rpc
182 2013-12-13 13:46:12 <michagogo> cloud|It will only ever work for transactions in your wallet
183 2013-12-13 14:31:52 <Tyupi> are there some server side php payment processor for accepting bitcoin instead of bitfreak's client-side processing?
184 2013-12-13 14:35:32 <abishek> what is the suitable datatype to store the amount in bitcoin?
185 2013-12-13 14:37:33 <kjj> decimal, arbitrary, string or 64-bit integer (the least good)
186 2013-12-13 14:38:28 <abishek> kjj, how many decimal points ?
187 2013-12-13 14:38:38 <kjj> can also use floats, but need to be very careful about width, rounding, and error accumulation
188 2013-12-13 14:38:44 <wumpus> as long as you don't use floating point it should be ok
189 2013-12-13 14:39:28 <abishek> kjj, varchar is much better if you want stay away from width, rounding and error accumulation, isn't it
190 2013-12-13 14:39:33 <abishek> what you store is what you get
191 2013-12-13 14:39:33 <kjj> currently, the protocol works to a precision of 8 places past the decimal
192 2013-12-13 14:39:55 <abishek> ok
193 2013-12-13 14:40:25 <kjj> if you understand how floats work and how to use them (you probably don't, and I don't mean that as a personal insult), they work just fine
194 2013-12-13 14:40:57 <wumpus> using floats is fine, but not for monetary amounts. just don't. you'll regret it.
195 2013-12-13 14:41:15 <BlueMatt> it is illegal to store money as floating point in some countries
196 2013-12-13 14:41:17 <BlueMatt> (seriously)
197 2013-12-13 14:41:35 <wumpus> BlueMatt: that makes sense
198 2013-12-13 14:42:10 <BlueMatt> ok, maybe I'm wrong, but it should be
199 2013-12-13 14:42:19 <kjj> most of the time, you'll want to handle amounts in a way that matches how you will be interfacing with bitcoin. if you are writing transactions in binary, using an int for satoshis may be your best bet
200 2013-12-13 14:42:51 <kjj> if you are working with RPC, you'll probably want a string (arbitrary) representation
201 2013-12-13 14:44:40 <kjj> I don't recommend the integer satoshi method except when you are writing your own transactions from scratch
202 2013-12-13 14:53:02 <abishek> i think i will go with varchar
203 2013-12-13 15:06:30 <BlueMatt> lol @king james programming
204 2013-12-13 15:17:05 <jgarzik> mornin'
205 2013-12-13 15:18:09 <BlueMatt> hi jgarzik
206 2013-12-13 15:49:12 <Temper> hello
207 2013-12-13 15:49:22 <Temper> does anyone want to talk theory?
208 2013-12-13 16:05:03 <jgarzik> Temper, you are welcome to ask or discuss anything on-topic
209 2013-12-13 16:05:26 <Temper> well i want to talk about designs in code not to loose bitcoins
210 2013-12-13 16:05:37 <Temper> since they have suddenly become so valuable
211 2013-12-13 16:05:41 <Temper> :)
212 2013-12-13 16:06:16 <BlueMatt> Temper: generally, if you start talking you['re more likely to get a response than if you ask to ask :)
213 2013-12-13 16:06:28 <BlueMatt> (though it is still rather early in the us...maybe not so many people around)
214 2013-12-13 16:11:18 <kjj> HD wallets will remove the last structural opportunity for coin loss that I'm aware of
215 2013-12-13 16:11:29 <Diablo-D3> high definition wallets
216 2013-12-13 16:11:34 <kjj> well, reduce, I suppose
217 2013-12-13 16:12:24 <kjj> yup. oversampled 192 kHz wallets
218 2013-12-13 16:12:41 <BlueMatt> ACTION still thinks sipa chose a terrible name
219 2013-12-13 16:13:08 <kjj> meh. he's earned the right to a half dozen terrible names before I get worried
220 2013-12-13 16:13:18 <BlueMatt> heh, thats true
221 2013-12-13 16:13:28 <BlueMatt> though Im not sure there is an amount to put on that...
222 2013-12-13 16:13:55 <kjj> I'll draw the line when he starts naming things for members of One Direction
223 2013-12-13 16:14:05 <BlueMatt> heh
224 2013-12-13 16:14:38 <kjj> Temper: what's your mind?
225 2013-12-13 16:15:12 <Temper> ok so storage is not 100%
226 2013-12-13 16:15:17 <Temper> you make a write..
227 2013-12-13 16:15:22 <Temper> but it can fail
228 2013-12-13 16:15:42 <Temper> however, a network like bitcoin is pretty much fail proof
229 2013-12-13 16:16:08 <Temper> so write to db, transact bitcoin, publish record, fail
230 2013-12-13 16:16:15 <Temper> you just lost it all
231 2013-12-13 16:16:37 <BlueMatt> hd wallets help a ton with this
232 2013-12-13 16:16:44 <Temper> hd?
233 2013-12-13 16:16:46 <BlueMatt> once you get your one key safely stored, the rest is easy
234 2013-12-13 16:16:53 <BlueMatt> heirachical deterministic
235 2013-12-13 16:17:04 <Temper> hrmm
236 2013-12-13 16:17:10 <BlueMatt> hierarchical*
237 2013-12-13 16:17:14 <kjj> it means you only need to save one secret, and all other keys can be derived from that secret
238 2013-12-13 16:17:24 <Temper> what i guessed
239 2013-12-13 16:17:36 <kjj> they have more capabilities, but in this discussion the important part is the single secret
240 2013-12-13 16:17:37 <BlueMatt> its coming down the pipe in wallets, but its still some time off
241 2013-12-13 16:17:41 <Temper> doesn't that mean that the keys are not that.. unique
242 2013-12-13 16:17:43 <Temper> ?
243 2013-12-13 16:17:47 <BlueMatt> no
244 2013-12-13 16:17:59 <BlueMatt> from the one secret you can derive infinite (essentially) keys
245 2013-12-13 16:18:05 <Temper> the the addr is basically the pub key right?
246 2013-12-13 16:18:18 <BlueMatt> well, more unique keys than you can calculate before the universe ends
247 2013-12-13 16:18:19 <kjj> if you assume that ECDSA is safe, then HD wallets are also safe.
248 2013-12-13 16:18:26 <BlueMatt> addr is hash(pubkey)
249 2013-12-13 16:18:36 <Temper> ECDSA?
250 2013-12-13 16:18:46 <kjj> the crypto we use
251 2013-12-13 16:18:47 <BlueMatt> the signature algorithm bitcoin uses
252 2013-12-13 16:19:08 <Temper> well how does this one key allow you to recover transactions?
253 2013-12-13 16:19:17 <BlueMatt> magic :)
254 2013-12-13 16:19:21 <Temper> lost transactions
255 2013-12-13 16:19:26 <kjj> transactions are never lost
256 2013-12-13 16:19:26 <Temper> oh well
257 2013-12-13 16:19:30 <Temper> i know
258 2013-12-13 16:19:34 <BlueMatt> https://github.com/bitcoin/bips/blob/master/bip-0032.mediawiki
259 2013-12-13 16:19:37 <Temper> not for the btc network
260 2013-12-13 16:19:39 <kjj> or rather, a lost transaction is the same as a transaction that never happened
261 2013-12-13 16:19:43 <Temper> but maybe for me
262 2013-12-13 16:19:59 <Temper> not if it has been published
263 2013-12-13 16:20:03 <null> Temper: think of a seed x_0 you input to a formula like x_i+1 = a*x_i + b mod c
264 2013-12-13 16:20:17 <BlueMatt> if a tx has been published and ended up in the chain, its easy to find with a bloom filter and a peer
265 2013-12-13 16:20:22 <Temper> yeah but none of this helps me do accounting
266 2013-12-13 16:20:23 <BlueMatt> if its not in the chain, its lost
267 2013-12-13 16:20:31 <null> Temper: you can generate a lot of random numers with it, but the sequence is uniquely defined by your seed, yet deterministic
268 2013-12-13 16:20:36 <Temper> yeah the transactions happened
269 2013-12-13 16:20:45 <Temper> but i mean i don't remember what i did
270 2013-12-13 16:20:50 <Temper> who i sent the money to
271 2013-12-13 16:20:57 <Temper> because my hd got amnesia
272 2013-12-13 16:21:05 <Temper> they got it..
273 2013-12-13 16:21:08 <kjj> that's not a problem that bitcoin can solve (nor should it try)
274 2013-12-13 16:21:10 <Temper> sure
275 2013-12-13 16:21:15 <Temper> oh no
276 2013-12-13 16:21:16 <Temper> i know
277 2013-12-13 16:21:22 <BlueMatt> you could use a function to convert from order id -> addr
278 2013-12-13 16:21:25 <BlueMatt> (deterministically)
279 2013-12-13 16:21:27 <Temper> i am talking btc interface
280 2013-12-13 16:21:34 <Temper> ahh
281 2013-12-13 16:21:46 <Temper> using like 2 factor auth?
282 2013-12-13 16:22:05 <Temper> well i don't have the order id either
283 2013-12-13 16:22:09 <BlueMatt> esp if your order ids are monotonically increasing
284 2013-12-13 16:22:44 <Temper> like take btc-e
285 2013-12-13 16:22:53 <Temper> they do hundreds of transactions?
286 2013-12-13 16:22:59 <Temper> an hour/min
287 2013-12-13 16:23:00 <Temper> sec
288 2013-12-13 16:23:02 <BlueMatt> bitcoin doesnt have a way to solve account tracking, no
289 2013-12-13 16:23:07 <Temper> i know
290 2013-12-13 16:23:13 <Temper> i am trying to solve it
291 2013-12-13 16:23:16 <BlueMatt> you could use hd wallet chains to map user id -> address chain
292 2013-12-13 16:23:23 <BlueMatt> (ie set of addresses)
293 2013-12-13 16:23:32 <Temper> hrmm
294 2013-12-13 16:23:40 <Temper> so when i sign a transaction
295 2013-12-13 16:23:46 <Temper> is it immediately sent
296 2013-12-13 16:23:53 <Temper> outwards
297 2013-12-13 16:24:00 <Temper> can i said a transaction
298 2013-12-13 16:24:04 <Temper> err sign
299 2013-12-13 16:24:18 <Temper> and then hold on to it
300 2013-12-13 16:24:22 <kjj> if you use sendtoaddress, etc, then yes, it is sent immediately
301 2013-12-13 16:24:24 <Temper> secretly?
302 2013-12-13 16:24:38 <kjj> if you use createrawtransaction/sign../send.. it isn't sent until you sendrawtransaction
303 2013-12-13 16:24:51 <Temper> ahh
304 2013-12-13 16:25:03 <Temper> and bitcoind
305 2013-12-13 16:25:12 <Temper> how many "wallets" can it have open?
306 2013-12-13 16:25:24 <kjj> because of coinjoin, I'm thinking about adding rpc calls similar to sendtoaddress, sendmany, etc that return the raw TX rather than sending
307 2013-12-13 16:25:37 <kjj> the reference client only uses one wallet
308 2013-12-13 16:25:51 <BlueMatt> Temper: if youre looking at running a large set of transactions/addrs, bitcoind may not be the best bet...
309 2013-12-13 16:26:01 <BlueMatt> not that there is one I can truly recommend, but its not all that great
310 2013-12-13 16:26:03 <Temper> why not?
311 2013-12-13 16:26:15 <Temper> and what is?
312 2013-12-13 16:26:17 <BlueMatt> it doesnt scale particularly well
313 2013-12-13 16:26:20 <kjj> you have to shut the node down to switch, but that is almost certainly not the right way to solve whatever actual problem you are trying to solve
314 2013-12-13 16:26:24 <BlueMatt> its wallet, that is
315 2013-12-13 16:26:53 <Temper> well like coinbase must have 600k wallets
316 2013-12-13 16:27:03 <BlueMatt> I doubt they use bitcoind
317 2013-12-13 16:27:15 <kjj> why must coinbase have so many wallets?
318 2013-12-13 16:27:23 <BlueMatt> or even a client that keeps user wallet distinct beyond a db to track accounts
319 2013-12-13 16:27:25 <Temper> they advertise that many customers
320 2013-12-13 16:27:45 <kjj> doesn't mean they each need their own wallet
321 2013-12-13 16:27:48 <Temper> but then 1 key opens them all?
322 2013-12-13 16:27:59 <kjj> in fact, it is better for privacy if all users are pooled
323 2013-12-13 16:28:15 <BlueMatt> a shitton of keys open them all
324 2013-12-13 16:28:24 <atian> kjj: its not pooled anymore
325 2013-12-13 16:28:24 <BlueMatt> but they're all in one place (short cold storage wallet(s))
326 2013-12-13 16:28:41 <Temper> how is it more private to keep all that info in one place?
327 2013-12-13 16:28:53 <Temper> seems like a horrible idea
328 2013-12-13 16:29:03 <kjj> atian: that's crazy then.
329 2013-12-13 16:29:07 <BlueMatt> Temper: its gonna be in one place whether you have 100 wallets on one server or 1 wallet with 100 accounts on one server
330 2013-12-13 16:29:27 <BlueMatt> atian: ummm...recent withdraws of mine do not show a consistent set of coins being withdrawn from?
331 2013-12-13 16:29:34 <kjj> if all of the accounts are pooled, the coins you withdraw were almost certainly someone else's deposit, not your own
332 2013-12-13 16:29:56 <Temper> yeah that seems bad too
333 2013-12-13 16:29:57 <BlueMatt> to be fair they were purchased with usd, but still
334 2013-12-13 16:30:13 <Temper> because then the network does not enforce account balances
335 2013-12-13 16:30:18 <BlueMatt> Temper: coinbase does (as with any sane large bitcoinsite) keep most of their funds offline
336 2013-12-13 16:30:34 <kjj> most large operations have offline ("cold") wallets for most storage, and only keep enough hot to settle short term flows
337 2013-12-13 16:30:39 <Temper> they they have a slush fund