1 2013-03-09 00:03:59 <jrmithdobbs> HM2: "Beyond NaN"
2 2013-03-09 00:07:49 <xjrn> "EDIT : Also this patch is pointless, you probably spent more CPU cycles browsing this thread than you'll spend verifying SD txes for a full year."
3 2013-03-09 00:08:36 <Diablo-D3> eh
4 2013-03-09 00:08:37 <Diablo-D3> its really the hd churn thats the problem
5 2013-03-09 00:08:51 <HM2> no it's the fees and the confirmation lag
6 2013-03-09 00:09:07 <Diablo-D3> HM2: that doesnt NEED to happen
7 2013-03-09 00:09:22 <Diablo-D3> bitcoin should just disable the no free part of the block by default
8 2013-03-09 00:09:53 <HM2> what about the block size
9 2013-03-09 00:10:58 <OneMiner> Trying to use testnet in Windows 7 64, bitcoind 0.8.0 I've started the .exe but cannot poll it with my bitcoind getinfo or help commands. I get "error conildn't connect to server". Anybody familiar with testnet on Windows?
10 2013-03-09 00:12:03 <K1773R> no, but i know why
11 2013-03-09 00:12:10 <OneMiner> Yay!
12 2013-03-09 00:12:24 <K1773R> bitcoind -testnet -rpcport=18334 getinfo
13 2013-03-09 00:12:30 <K1773R> just replace bitcoind to bitcoind.exe
14 2013-03-09 00:12:51 <K1773R> if you query it somehow dosnt change the rcpport to 18334 if you supply -testnet
15 2013-03-09 00:13:10 <K1773R> therefore u try to connect to the mainnet RPC, which in ur case isnt running
16 2013-03-09 00:13:19 <K1773R> hope this helps ;)
17 2013-03-09 00:13:21 <xjrn> any 2 p2p clients that agree on a trade can form their own validation criterion. what this means is that in theory they can cordon off access to coins sent into thier domain of trade as dead currency to the outside clients.
18 2013-03-09 00:13:43 <warren> apparently that's why testnet bitcoin.conf reads from ~/.bitcoin instead of ~/.bitcoin/testnet3. So testnet in bitcoin.conf would work.
19 2013-03-09 00:13:50 <warren> (not a great reason)
20 2013-03-09 00:14:14 <K1773R> yes, but then if you start bitcoind without options, it would be always testnet
21 2013-03-09 00:14:21 <warren> xjrn: dead currency, except for those already in the block chain?
22 2013-03-09 00:14:35 <warren> xjrn: your position is getting more and more nonsensical
23 2013-03-09 00:14:40 <K1773R> the correct way would be to simply load ~/.bitcoin/testnet3/bitcoin.conf IF -testnet is supplied
24 2013-03-09 00:14:43 <OneMiner> Ding ding ding! K1773R Got it. bitcoind -testnet getinfo returns da info. Thanks! :)
25 2013-03-09 00:14:49 <xjrn> warren: if those already in the blockchain don't control those wallets, how would it be other than dead?
26 2013-03-09 00:15:09 <K1773R> oh, works without -rpcport? nice (didnt try testnet with newer versions)
27 2013-03-09 00:16:13 <K1773R> enjoy ur testnet, remember there is a faucet if you need some coins (send em back when ur done): http://testnet.mojocoin.com/ and http://tpfaucet.appspot.com/
28 2013-03-09 00:16:33 <OneMiner> Thank you. I'll be good. :D
29 2013-03-09 00:17:24 <xjrn> warren: let me put it another way, if any single code base was agreed upon by multiple parties to validate coin ingress as a basis for following a different set of transactions, this would form a different currency traded against the ingress coins.
30 2013-03-09 00:17:52 <warren> xjrn: only if they are mined by incompatible policies.
31 2013-03-09 00:17:56 <xjrn> warren: those coins would not change "the existing blockchain"
32 2013-03-09 00:18:44 <warren> K1773R: would be great to have a testnet <-> BTC exchange, just to give the speculators another toy.
33 2013-03-09 00:19:07 <K1773R> warren: nty, we already had this once...
34 2013-03-09 00:19:16 <warren> really?
35 2013-03-09 00:19:33 <xjrn> namecoin might as well be
36 2013-03-09 00:19:38 <K1773R> warren: testnet is always lowest diff if there is no block found in the last 20 min, if ppl would mine it the diff would stay constantly high and not at 1 (in reality its 0.5)
37 2013-03-09 00:20:24 <warren> I suppose someone keeps on a non-mining seed node.
38 2013-03-09 00:21:30 <warren> K1773R: I know this, I was just making a lame joke.
39 2013-03-09 00:24:15 <K1773R> soon we got TBTC Diff at 100k :D
40 2013-03-09 00:26:39 <jgarzik> random note
41 2013-03-09 00:26:48 <jgarzik> us4.exmulti.net died, so I'm no longer mining testnet full time
42 2013-03-09 00:26:53 <jgarzik> nor providing that testnet node
43 2013-03-09 00:27:04 <xempew> :o
44 2013-03-09 00:27:15 <jgarzik> ACTION needs to get back to BlueMatt, and set up that Foundation-funded VPS
45 2013-03-09 00:28:01 <OneMiner> I've just downloaded the blockchain up to the hight that the testnet explorer states. Everything is good. Just need a block to get mah coins.
46 2013-03-09 00:29:11 <K1773R> testnet explorer gets often stuck for a long time, dont rely on it
47 2013-03-09 00:29:23 <OneMiner> poop
48 2013-03-09 00:29:43 <OneMiner> Hight is 50095 at 121.something diff.
49 2013-03-09 00:30:23 <K1773R> yeah, that was when someone mined with 1GH/s and tryd to sell the TBTC afterwards...
50 2013-03-09 00:30:39 <OneMiner> What a dick! lol that's just supid. :P
51 2013-03-09 00:31:15 <K1773R> i can connect to 2 nodes, i think i should host BTC testnet too on my public altcoin node
52 2013-03-09 00:31:24 <OneMiner> Hahahaha you'd think you could get about the same results with a CPU over time. You know, per watt.
53 2013-03-09 00:31:58 <K1773R> well, you can find alot blocks with CPU, if ur mining alone atleast every 20 min u find a block for sure
54 2013-03-09 00:32:03 <OneMiner> So are we saying that testnet is stuck?
55 2013-03-09 00:32:20 <K1773R> no it isnt, its usual
56 2013-03-09 00:32:54 <OneMiner> Great. I've got my CPU working on it so I'll check back in a bit. Thanks a lot.
57 2013-03-09 00:33:09 <K1773R> im got a public node hosting all altcoins (had problems with some altcoins being stuck because no-one did setup a port forwarding @ home)
58 2013-03-09 00:33:18 <K1773R> which miner?
59 2013-03-09 00:33:26 <OneMiner> Bitcoind CPU.
60 2013-03-09 00:33:30 <K1773R> bad
61 2013-03-09 00:33:35 <OneMiner> :(
62 2013-03-09 00:33:39 <K1773R> https://github.com/pooler/cpuminer/ <--
63 2013-03-09 00:33:48 <OneMiner> Oh snap. I've got that.
64 2013-03-09 00:34:04 <K1773R> its a fork of jgarzik's cpuminer, almost fully implemented in ASM
65 2013-03-09 00:34:08 <OneMiner> Thought that was only for scrypt or something.
66 2013-03-09 00:34:13 <K1773R> its for both
67 2013-03-09 00:35:02 <K1773R> minerd -a sha256d -o http://rpcuser:rcppassword@127.0.0.1:18334
68 2013-03-09 00:35:05 <warren> jgarzik: I think my university wont care about hosting testnet here
69 2013-03-09 00:35:15 <OneMiner> 2.2.3 right? Umm.... Dang. How do I setup my RCP junk for testnet in parallel with regularnet?
70 2013-03-09 00:35:28 <OneMiner> Okey dokey.
71 2013-03-09 00:35:36 <K1773R> you can run bitcoind on main + testnet the same time
72 2013-03-09 00:35:36 <warren> jgarzik: would you folks want ssh access to restart it?
73 2013-03-09 00:36:00 <K1773R> well, its 2 bitcoind instances, but it works
74 2013-03-09 00:36:17 <jgarzik> warren: easier if that task falls on your shoulders ;p
75 2013-03-09 00:36:30 <OneMiner> Gah, I need a damn smoke before I get anything done. bbl
76 2013-03-09 00:36:38 <warren> Vinnie_win: btw, https://en.bitcoin.it/wiki/Incidents#CVE-2010-5140
77 2013-03-09 00:36:39 <jgarzik> warren: just running current codebase on testnet, perhaps one core CPU mining w/ the internal miner
78 2013-03-09 00:37:01 <warren> jgarzik: internal miner is important?
79 2013-03-09 00:37:35 <jgarzik> warren: no, just convenient
80 2013-03-09 00:37:40 <warren> jgarzik: is internal miner a thread within the bitcoind? I would want to limit its CPU usage to like 5%, easier if it is its own process.
81 2013-03-09 00:37:54 <jgarzik> warren: easier than "two moving parts" (bitcoind + cpuminer) to monitor and maintain
82 2013-03-09 00:37:54 <K1773R> jgarzik/warren: how about creating a cron which pulls the git and builds bitcoind every 24h, stops bitcoind, replaces binarys, starts bitcoind again?
83 2013-03-09 00:38:13 <jgarzik> warren: yes, internal miner is its own thread within bitcoind
84 2013-03-09 00:38:15 <K1773R> warren: if you want to limit cpu usage, take a look at the tool cpulimit
85 2013-03-09 00:38:33 <warren> K1773R: unable to build bitcoind here, no ecdsa support, and I refuse to install openssl built by other people. I intend on making a add-on ecdsa-only library at some point for all Fedora and RHEL users.
86 2013-03-09 00:38:46 <K1773R> then build it urself?
87 2013-03-09 00:39:07 <warren> yeah, I will eventually
88 2013-03-09 00:39:39 <K1773R> you can compile openssl in a specific folder (with ./configure --prefix=/some/directory and compile bitcoind with it, if you run bitcoind set the LD_LIBRARAY_PATH to the specific dir
89 2013-03-09 00:39:51 <K1773R> so only bitcoind would use the openssl lib you compiled ;)
90 2013-03-09 00:40:15 <warren> yeah, I could. but if I did that, I might as well make my desired ecdsa only library package
91 2013-03-09 00:40:22 <warren> ecdsa is needed for more than just bitcoin.
92 2013-03-09 00:41:53 <K1773R> sure, up to you. just trying to supply good ideas ;)
93 2013-03-09 00:51:12 <warren> K1773R: cron rebuilding and running testnet every 24H wouldn't ensure that there is at least one testnet seed all the time.
94 2013-03-09 00:51:49 <K1773R> you could be running 2 bitcoinds (1 the release version, and another the git version)
95 2013-03-09 00:51:57 <K1773R> both with a own data directory + different ports
96 2013-03-09 00:52:15 <warren> I'm solving only the "no guaranteed testnet seed" problem.
97 2013-03-09 00:52:27 <warren> because it requires so little effort to do so
98 2013-03-09 00:52:36 <K1773R> no offense ;)
99 2013-03-09 00:54:09 <warren> how often do we restart testnet?
100 2013-03-09 00:55:10 <K1773R> if im correct we are at testnet3 (according to code/directory) which has been created in 0.7 im im correct
101 2013-03-09 00:55:29 <K1773R> gotta ask someone else
102 2013-03-09 01:00:51 <warren> Surprisingly, there are a lot of testnet nodes.
103 2013-03-09 01:01:04 <warren> maybe I shouldn't be concerned about this
104 2013-03-09 01:01:43 <K1773R> i got 6 over here, u?
105 2013-03-09 01:02:18 <warren> what are the chances that everyone will shutdown testnet such that nobody is seeding?
106 2013-03-09 01:02:25 <K1773R> very low
107 2013-03-09 01:02:38 <warren> ok, then I'm killing this
108 2013-03-09 01:02:48 <K1773R> poor testnet :P
109 2013-03-09 01:04:26 <OneMiner> I'm going to run testnet for a bit for one.
110 2013-03-09 01:04:34 <warren> OTOH, I have no limit on bandwidth here. We were dared to saturate the pipe, and we failed with our Linux mirror.
111 2013-03-09 01:09:23 <helo> warren: one of the accepted foundation proposals was for a testnet faucet and node
112 2013-03-09 01:09:28 <warren> OK, I'm leaving it running. Only 64MB RAM it appears.
113 2013-03-09 01:09:29 <helo> and dns seed
114 2013-03-09 01:09:58 <warren> helo: well, if that happens, I'll turn it off
115 2013-03-09 01:10:14 <warren> helo: or maybe I'll leave it running so I can extract the 1MB tx's whenever I want.
116 2013-03-09 01:26:54 <Ferroh> I forget, what file do I modify in 0.8 to prevent bitcoind from rechecking the whole chain after importing a new address?
117 2013-03-09 01:28:34 <sipa> you put a false after importprivkey <key>
118 2013-03-09 01:33:29 <Ferroh> Oh, the interface changed?!
119 2013-03-09 01:33:31 <Ferroh> awesome
120 2013-03-09 01:33:34 <Ferroh> thanks sipa
121 2013-03-09 01:55:48 <jaakkos> is anyone using GoxCLI, frontend to MtGox? seems like many commands fail, is this expected?
122 2013-03-09 01:56:12 <jaakkos> is there a python trading library that works well?
123 2013-03-09 02:15:05 <road33> I am developing a WS app for mtgox, today I can not longer connect, is this typical ?
124 2013-03-09 02:32:18 <freewil> road33, it's known to have frequently connectivity issues
125 2013-03-09 03:03:59 <iwilcox> road33: I've had no WS connections work for about 24hrs.
126 2013-03-09 03:07:22 <iwilcox> road33: I've had no WS connections work for about 24hrs.
127 2013-03-09 03:29:52 <freewil> how does bitcoind keep track of the confirmation count for every transaction in the wallet
128 2013-03-09 03:30:23 <freewil> when a new block is received, does it update the confirmation count for each tx?
129 2013-03-09 04:12:44 <BlueMatt> jgarzik: yes you do...
130 2013-03-09 04:12:58 <BlueMatt> anyone know off-hand the lower-bound on a scriptSig assuming it actually signs something
131 2013-03-09 04:13:17 <BlueMatt> so...I suppose the lower-bound on the size of s gi
132 2013-03-09 04:13:44 <BlueMatt> a sig*
133 2013-03-09 04:16:09 <phantomcircuit> BlueMatt, i assume you could shrink it by artificially using a low value private key
134 2013-03-09 04:17:35 <BlueMatt> hmm, I thought there was a lower bound on the sig size
135 2013-03-09 04:17:45 <BlueMatt> or atleast for reasonable values
136 2013-03-09 04:17:48 <BlueMatt> reasonable privkeys
137 2013-03-09 04:20:27 <BlueMatt> phantomcircuit: can you elaborate? that may be the cause of a bug TD saw a few days ago
138 2013-03-09 04:22:05 <phantomcircuit> BlueMatt, er
139 2013-03-09 04:22:06 <phantomcircuit> nvm
140 2013-03-09 04:22:12 <phantomcircuit> well
141 2013-03-09 04:22:26 <phantomcircuit> theoretically the signature size could be 2 bytes
142 2013-03-09 04:22:39 <phantomcircuit> 0, 0 is a valid signature, but exeptionallllly unlikely
143 2013-03-09 04:22:45 <BlueMatt> will openssl/bouncycastle ever encode one as such?
144 2013-03-09 04:22:58 <phantomcircuit> it would be uh 4 bytes?
145 2013-03-09 04:23:22 <phantomcircuit> it's an ASN.1 sequence with 2 ASN.1 integers
146 2013-03-09 04:23:37 <BlueMatt> hmm...I was under the impression I read something about there being a lower bound on what bitcoin will actually generate, but I may be wrong
147 2013-03-09 04:23:41 <BlueMatt> something to dig up tomorrow
148 2013-03-09 04:23:52 <phantomcircuit> so iirc 1 byte to identify the sequence 1 byte for the length, 1 byte to identify the int, 1 byte for the int
149 2013-03-09 04:23:54 <phantomcircuit> so
150 2013-03-09 04:23:55 <phantomcircuit> 6 bytes
151 2013-03-09 04:24:01 <phantomcircuit> but that's exceptionally unlikely
152 2013-03-09 04:24:10 <BlueMatt> yea, well Ill dig it up tomorrow then
153 2013-03-09 04:24:32 <phantomcircuit> there might be something in the ecdsa sign math that im missing
154 2013-03-09 04:26:04 <phantomcircuit> yeah
155 2013-03-09 04:26:26 <phantomcircuit> BlueMatt, both r and s could be 1
156 2013-03-09 04:26:48 <phantomcircuit> there might be something in openssl which invalidated small values for r/s and goes back to picking k
157 2013-03-09 04:27:09 <phantomcircuit> but i sort of doubt it since that would shrink the effective key space
158 2013-03-09 04:28:01 <phantomcircuit> also the valid size for a scriptSig is... infinity
159 2013-03-09 04:29:23 <gmaxwell> phantomcircuit: you can't just get valid tiny r/s without solving the DLP.
160 2013-03-09 04:29:42 <phantomcircuit> gmaxwell, not on purpose
161 2013-03-09 04:29:47 <phantomcircuit> but it could happen by chance
162 2013-03-09 04:30:04 <gmaxwell> Indeed.
163 2013-03-09 04:31:33 <dgriffi> can I get some advice on adding a minor, but rather needed feature?
164 2013-03-09 04:32:01 <dgriffi> I want to add an option to determine which clipboard the "Copy Address" button copies to
165 2013-03-09 04:32:18 <dgriffi> How should I go about doing this so I don't offend the Windows and OSX users?
166 2013-03-09 04:39:06 <Diablo-D3> dgriffi: well, to be fair
167 2013-03-09 04:39:12 <Diablo-D3> dgriffi: qt doesn't work that way
168 2013-03-09 04:39:30 <Diablo-D3> aaaaaand
169 2013-03-09 04:39:45 <Diablo-D3> I'd rather see a patch that treats PRIMARY as an alias for CLIPBOARD
170 2013-03-09 04:39:51 <Diablo-D3> thus solving the whole damned problem to begin with
171 2013-03-09 04:41:21 <dgriffi> Diablo-D3: I'm not that familiar with QT development... what is PRIMARY?
172 2013-03-09 04:41:34 <Diablo-D3> thats not part of QT, its part of X
173 2013-03-09 04:41:42 <Diablo-D3> PRIMARY and CLIPBOARD are the names of the two clipboards in X
174 2013-03-09 04:41:55 <Diablo-D3> select/middle click uses one, standard copy and paste uses the other
175 2013-03-09 04:42:23 <dgriffi> PRIMARY is the middle click one?
176 2013-03-09 04:42:44 <Diablo-D3> I think so
177 2013-03-09 04:42:51 <Diablo-D3> I can never remember which is which
178 2013-03-09 04:43:31 <dgriffi> so should I even bother with an option to control which clipboard is used?
179 2013-03-09 04:43:50 <dgriffi> ...just go in there and change things to write to both?
180 2013-03-09 04:44:13 <Diablo-D3> its not worth it
181 2013-03-09 04:44:25 <Diablo-D3> most people who finally got tired of how fucked up X clipboards are installed parcellite
182 2013-03-09 04:44:39 <Diablo-D3> and turn on synchronize clipboards
183 2013-03-09 04:45:45 <Vinnie_win> So.... :-)
184 2013-03-09 04:46:00 <Vinnie_win> I canceled my date so I could hang out with you guys and lay some blame on peeps
185 2013-03-09 04:46:08 <Vinnie_win> who is WITH ME?
186 2013-03-09 04:46:09 <Diablo-D3> Vinnie_win: right, sure
187 2013-03-09 04:48:28 <dgriffi> Diablo-D3: can I just submit a pull request for shits and giggles and get it accepted?
188 2013-03-09 04:48:51 <Diablo-D3> dgriffi: you could???.
189 2013-03-09 04:48:57 <Diablo-D3> but I can't see it being accepted
190 2013-03-09 04:49:04 <dgriffi> why not?
191 2013-03-09 04:49:12 <Diablo-D3> you're fixing the problem at the wrong level
192 2013-03-09 05:00:17 <warren> It is confusing that Shift-Insert and CTRL-V do different things.
193 2013-03-09 05:00:35 <warren> Fortunately it's mainly us using Linux...
194 2013-03-09 05:03:26 <dgriffi> I just want to get this code written
195 2013-03-09 05:04:47 <warren> Diablo-D3: thanks, I didn't know about parcellite.
196 2013-03-09 05:05:31 <dgriffi> the thing about paracellite is that fewer people will use it.. they'll look at bitcoin-qt and go "why the hell won't the clipboard work right?"
197 2013-03-09 05:05:59 <Diablo-D3> warren: well its because shift-insert is the correct way
198 2013-03-09 05:05:59 <warren> you think "Copy into clipboard 1" and "Copy into clipboard 2" would be better?
199 2013-03-09 05:06:04 <Diablo-D3> its you windows weirdos that do it wrong
200 2013-03-09 05:06:08 <Diablo-D3> warren: no, because qt doesn't work that way
201 2013-03-09 05:06:20 <Diablo-D3> dgriffi: yes, which is why X needs fixed
202 2013-03-09 05:06:25 <Diablo-D3> you think qt is the only broken app? hah!
203 2013-03-09 05:06:34 <dgriffi> Diablo-D3: that's not going to happen very soon
204 2013-03-09 05:06:42 <dgriffi> Diablo-D3: no.. we can start by writing clean apps
205 2013-03-09 05:06:59 <Diablo-D3> how? qt is already broken
206 2013-03-09 05:07:05 <warren> Diablo-D3: parcellite "save history" actually writes your clipboard to disk?
207 2013-03-09 05:07:10 <warren> ACTION scary
208 2013-03-09 05:07:15 <Diablo-D3> warren: no??? turn on persistent history for that
209 2013-03-09 05:07:20 <warren> oh
210 2013-03-09 05:07:47 <warren> Enable scary mode? Yes/No
211 2013-03-09 05:08:52 <warren> Diablo-D3: I'd love a clipboard manager that strips all text formatting, so copying from a web page and pasting in a word processor doesn't attempt to apply its font, size, bold, etc.
212 2013-03-09 05:09:03 <warren> but getting off topic
213 2013-03-09 05:09:05 <Diablo-D3> heh
214 2013-03-09 05:09:06 <Diablo-D3> well
215 2013-03-09 05:09:12 <Diablo-D3> welcome to how fucking stupid computers are
216 2013-03-09 05:09:31 <warren> They aren't stupid. They do EXACTLY what you tell them to do.
217 2013-03-09 05:09:38 <Diablo-D3> dude
218 2013-03-09 05:09:40 <warren> +/- a few stray gamma rays
219 2013-03-09 05:09:43 <Diablo-D3> if my computer did exactly what I told it to
220 2013-03-09 05:09:51 <Diablo-D3> I'd be rich, happy, and have at least three wives
221 2013-03-09 05:10:13 <dgriffi> more digging through the code... why isn't the "Show QR Code" button showing up when I run the application?
222 2013-03-09 05:10:17 <warren> Diablo-D3: If that's the case, I haven't been telling my computer the right things.
223 2013-03-09 05:10:26 <nanotube> warren: it's called notepad passthrough :)
224 2013-03-09 05:10:40 <warren> nanotube: yeah, or gedit. I do it like 100 times a day.
225 2013-03-09 05:10:52 <dgriffi> ACTION hates qedit
226 2013-03-09 05:11:00 <dgriffi> ACTION loves xclip
227 2013-03-09 05:11:05 <warren> there probably exists a solution for this already.
228 2013-03-09 05:11:23 <nanotube> warren: if you use openoffice (and probably msoffice too), doing ctl-shift-v does a 'paste special' and you can choose 'unformatted text'. that's another option.
229 2013-03-09 05:11:26 <dgriffi> so, can anyone tell me why I can't see the "Show QR Code" button?
230 2013-03-09 05:11:43 <warren> nanotube: funny thing is, it only works most of the time. I haven't been able to figure out why.
231 2013-03-09 05:11:57 <warren> nanotube: although sometimes I want to paste in Google Docs, and I have to use notepad/gedit passthrough
232 2013-03-09 05:12:17 <nanotube> yea, when in doubt, use a plain text editor :)
233 2013-03-09 05:12:40 <warren> yeah, Google Docs doesn't print your document the way it looked in the editor anyway. might as well use vi.
234 2013-03-09 05:13:00 <dgriffi> warren: use LaTeX
235 2013-03-09 05:13:10 <warren> haha
236 2013-03-09 05:13:29 <nanotube> latex++ :)
237 2013-03-09 05:13:47 <dgriffi> warren: I mean it. LaTeX is the ultimate when you want your text to look good
238 2013-03-09 05:13:47 <nanotube> though a bit overkill for quickndirty stuff.
239 2013-03-09 05:14:03 <dgriffi> for quickanddirty I use straight ascii
240 2013-03-09 05:14:09 <warren> especially math equations
241 2013-03-09 05:14:38 <nanotube> yea if math, latex==good
242 2013-03-09 05:15:11 <warren> During college, a chemistry classmate turned in his labnotes that he wrote in LaTeX. It looked indistinguishable in formatting from the lab manual.
243 2013-03-09 05:15:25 <Diablo-D3> it probably WAS the formatting in the lab manual
244 2013-03-09 05:15:32 <dgriffi> warren: probably because the book was written with latex
245 2013-03-09 05:15:38 <Diablo-D3> TeX is used quite frequently in technical manuals
246 2013-03-09 05:15:56 <warren> The professor laughed when he saw it.
247 2013-03-09 05:17:07 <warren> OK, adding "Implement auto-notepad passthrough hotkey to parcellite" along with the 700 other things I want to do after I finish this stupid thesis.
248 2013-03-09 05:19:24 <dgriffi> warren: what's your thesis on?
249 2013-03-09 05:46:32 <warren> dgriffi: software patent law
250 2013-03-09 05:46:52 <warren> dgriffi: I will likely publish it on groklaw when I'm doe.
251 2013-03-09 05:46:55 <warren> done
252 2013-03-09 05:47:41 <warren> guist of it: I found a potential loophole in the statutes that nobody thought about before.
253 2013-03-09 05:48:44 <dgriffi> oh, nice!
254 2013-03-09 05:49:06 <warren> That could weaken software patent protection.
255 2013-03-09 05:49:09 <dgriffi> warren: are you a law student?
256 2013-03-09 05:49:12 <warren> No.
257 2013-03-09 05:49:23 <warren> Taking a few law courses though.
258 2013-03-09 05:49:32 <nanotube> warren: sounds cool. :)
259 2013-03-09 05:49:44 <warren> I am extremely annoyed at this ecdsa thing.
260 2013-03-09 05:50:16 <iwilcox> Don't *tell* 'em about the loophole. Wait for the next Google v Oracle and hand it to Google.
261 2013-03-09 05:51:20 <dgriffi> ACTION chuckles
262 2013-03-09 05:51:26 <warren> iwilcox: Mark Lemley published a highly influential paper on a different loophole mid-2012, and it's going through the courts in sofware patent cases now.
263 2013-03-09 05:51:47 <nanotube> also... parcellite - looks neat.
264 2013-03-09 05:52:21 <warren> iwilcox: http://www.wired.com/opinion/2012/10/mark-lemley-functional-claiming/
265 2013-03-09 05:52:29 <warren> that's a short summary
266 2013-03-09 05:52:33 <iwilcox> Cool. Well, I hope you get appropriate credit to your name for yours.
267 2013-03-09 05:52:38 <iwilcox> ACTION reads
268 2013-03-09 05:52:39 <warren> I found something similar but entirely different.
269 2013-03-09 05:54:40 <dgriffi> published scholar time
270 2013-03-09 05:57:09 <iwilcox> I must have missed the one where someone claimed to own emoticons.
271 2013-03-09 05:57:47 <dgriffi> I remember reading about that, but I can't remember where and what the outcome was
272 2013-03-09 05:57:53 <warren> I'm very annoyed that Fedora doesn't ship ecdsa, needed by bitcoin and bitmessage among others. I didn't read the ECC patents (and I won't), but reading the analysis by people who did read the patents and understand cryptography, it sounds like ecdsa avoids the patented parts of EC*. However, the lawyers analyzing the situation might not have the skill in the art to distinguish what exactly is in the ecdsa implementation and compare it to the
273 2013-03-09 05:57:53 <warren> patent disclosures. This is all a guess though, since I haven't read the code nor have I read the patents.
274 2013-03-09 05:58:57 <warren> (and I don't know cryptography)
275 2013-03-09 05:59:03 <iwilcox> You won't read the patents...out of the wish to maintain plausible deniability? Out of the wish to avoid accidentally including some element?
276 2013-03-09 05:59:20 <warren> That's theoretically why engineers don't read them.
277 2013-03-09 06:00:07 <warren> I've been considering just saying "screw it" and cutting out openssl's ecdsa into its own library for use on Fedora, and putting that and bitcoin RPMS on my own website.
278 2013-03-09 06:00:42 <warren> Or rather, on my corporate entity's website, which is a legal person with its own bank account.
279 2013-03-09 06:00:51 <warren> I'm just an employee.
280 2013-03-09 06:04:44 <dgriffi> there is a perverse incentive to not read the patents
281 2013-03-09 06:05:09 <iwilcox> We always got told in the strongest terms never to search for any.
282 2013-03-09 06:06:46 <dgriffi> one of my favorite perverse patents is the cat exerciser (laser pointer with a perpendicular handle)
283 2013-03-09 06:06:56 <iwilcox> I never saw how "You knew there was a similar patent, yet you went ahead and implemented" was any worse or better than "You didn't even bother to check whether you were stealing other people's stuff"
284 2013-03-09 06:07:21 <warren> I was issued a software patent in November, and I have another coming soon.
285 2013-03-09 06:07:28 <dgriffi> if you don't bother to check if you're stealing, how is it stealing?
286 2013-03-09 06:07:40 <warren> A few pages before my new patent was a "beach towel with built in pillow"
287 2013-03-09 06:07:46 <warren> Novel invention!
288 2013-03-09 06:07:49 <dgriffi> warren: a friend of mine got one a few months ago (assigned to his employer)
289 2013-03-09 06:08:13 <iwilcox> It's stealing because patent law doesn't seem to distinguish wilful from ignorant copying.
290 2013-03-09 06:08:29 <dgriffi> iwilcox: but how is it copying if you came up with it yourself?
291 2013-03-09 06:08:47 <warren> dgriffi: that's part of the "fiction"
292 2013-03-09 06:08:52 <iwilcox> Never made any sense to me either, but that's how courts seem to see it.
293 2013-03-09 06:09:47 <iwilcox> Does that mean "beach towel with built in pillow" got issued?
294 2013-03-09 06:10:45 <dgriffi> I came across a patent for goggles for chickens so they won't peck each others eyes out
295 2013-03-09 06:12:11 <warren> Patents are not supposed to protect laws of nature, physical phenomena, and abstract ideas. We people skilled in the art recognize mathematics as unpatentable by this rule. The Federal Circuit agrees that math is unpatentable subject matter, ... unless it is HARD.
296 2013-03-09 06:12:29 <warren> As if math you can do with paper and pencil is different from math that a computer can do.
297 2013-03-09 06:13:23 <warren> iwilcox: yes. apparently nobody published prior art or offered for sale ""beach towel with built in pillow" before
298 2013-03-09 06:13:44 <warren> I'm paraphrasing, I don't remember exactly what it said, it was something like that.
299 2013-03-09 06:14:37 <warren> issued November 2012
300 2013-03-09 06:14:46 <warren> I'm seeing a 1980 patent for beach towel with pillow
301 2013-03-09 06:15:52 <iwilcox> Don't patent clerks have pointy-haired bosses?
302 2013-03-09 06:16:11 <warren> I dunno. Let's put it in the IRC log and see if they answer.
303 2013-03-09 06:16:12 <iwilcox> If I were the PHB, why wouldn't I ask for a report per granted patent?
304 2013-03-09 06:17:54 <iwilcox> ACTION shrugs
305 2013-03-09 06:21:44 <dgriffi> okay, I found the place where bitcoin-qt copies stuff to the clipboard...
306 2013-03-09 06:22:10 <dgriffi> it looks like I can pull this off with one or two lines
307 2013-03-09 06:22:53 <warren> dgriffi: you putting the copy into both clipboards?
308 2013-03-09 06:22:58 <warren> I'd like that.
309 2013-03-09 06:25:11 <dgriffi> warren: yeah
310 2013-03-09 06:25:37 <warren> well, your patch goes into MY bitcoin-qt RPM along with ecdsa.
311 2013-03-09 06:25:53 <warren> and the dust tx dropping patch
312 2013-03-09 06:26:35 <dgriffi> I just need to figure out how to extract the text from AddressTableModel::Address and into something that the X11 equivalent will like
313 2013-03-09 06:29:37 <warren> I wonder if I can use the preprocessor to automatically pull out only the ecdsa functions used by bitcoin from openssl source, so it isn't subject to my error...
314 2013-03-09 06:30:02 <Luke-Jr> warren: if you want one clipboard, just use Klipper :P
315 2013-03-09 06:30:43 <warren> Luke-Jr: I keep wishing that with every release GNOME won't become worse again.
316 2013-03-09 06:31:05 <dgriffi> Luke-Jr: I don't want these bolt-on bags-on-the-side
317 2013-03-09 06:31:07 <Luke-Jr> warren: what do you expect, it's GNOME :P
318 2013-03-09 06:31:17 <dgriffi> warren: well, at least the founder jumped ship to Mac
319 2013-03-09 06:31:23 <dgriffi> warren: maybe that'll improve things
320 2013-03-09 06:31:29 <warren> for mac?
321 2013-03-09 06:31:31 <dgriffi> yes
322 2013-03-09 06:31:38 <warren> ACTION shorts Apple
323 2013-03-09 06:31:44 <Luke-Jr> lol
324 2013-03-09 06:31:49 <dgriffi> Miguel deIza I think is his name
325 2013-03-09 06:32:11 <warren> I guess I shouldn't say things that are too mean about people on public IRC logs.
326 2013-03-09 06:37:21 <warren> dgriffi: if you can't figure it out, make it so they can highlight the address =P
327 2013-03-09 06:37:33 <dgriffi> warren: I'll look at doing that too
328 2013-03-09 06:37:39 <dgriffi> maybe do both
329 2013-03-09 06:38:11 <warren> that was a joke, but I guess I'd like it
330 2013-03-09 06:38:36 <dgriffi> oh no. I like being able to highlight whatever I like
331 2013-03-09 06:38:41 <dgriffi> good suggestion
332 2013-03-09 06:39:26 <warren> Another thing on my list of things to do after the thesis, "Port litecoin protocol to bitcoin git master". It's pointless, but good practice to learn the code.
333 2013-03-09 06:39:38 <warren> the litecoin git is a mess
334 2013-03-09 06:43:52 <dgriffi> what is litecoin? bitcoin with different constants?
335 2013-03-09 06:44:59 <warren> dgriffi: 2.5 minute blocks, overly anti-spam fee formula, scrypt instead of sha256
336 2013-03-09 06:45:47 <Luke-Jr> dgriffi: yet another scamcoin
337 2013-03-09 06:45:48 <warren> dgriffi: it makes no sense, but it's somehow traded in a tight range of 6-8 cents over the past year, and a few days ago jumped to ~20 cents
338 2013-03-09 06:46:07 <warren> I doubt it will survive the long-term.
339 2013-03-09 06:46:31 <warren> dgriffi: it's also effectively unmaintained
340 2013-03-09 06:46:33 <Luke-Jr> it's basically being used as a pyramid scheme
341 2013-03-09 06:46:45 <warren> a flat and boring pyramid
342 2013-03-09 06:47:03 <Luke-Jr> lol
343 2013-03-09 06:47:10 <Luke-Jr> nah, they try to sell it to newbies
344 2013-03-09 06:48:52 <Luke-Jr> "look, you missed out on the bitcoin get-rich-quick! litecoin is the next big thing, you should invest in it before the price bursts like bitcoin did!"
345 2013-03-09 06:49:27 <warren> I'm going to use litecoin as an experiment in my "pollution tax" fee formula.
346 2013-03-09 06:49:30 <ThomasV> Luke-Jr: solidcoin!
347 2013-03-09 06:49:52 <ThomasV> errr, no, microcash
348 2013-03-09 06:49:59 <ThomasV> whatever
349 2013-03-09 06:50:44 <warren> Historically litecoin suffered from a massive spam attack of tiny value tx's. In response they jacked up the standard fee to a very high level, which pretty much stopped spam.
350 2013-03-09 06:51:15 <Luke-Jr> warren: I wanted to try out CVE-2012-2459 on it, but they patched it :/
351 2013-03-09 06:51:21 <warren> The gambling there has high fees and no return tx that I can see, too costly. The gamblers pay the high tx because they're addicts.
352 2013-03-09 06:51:56 <warren> This is an experiment, because I think bitcoin actually got its base fee theory wrong.
353 2013-03-09 06:52:27 <warren> The flat per-KB fee doesn't reflect true network costs, allowing the shifting of externalities that we see today.
354 2013-03-09 06:53:35 <warren> litecoin's users want the fee to be reduced by 10x. I'm going to suggest they instead reduce the standard fee far more, but increase the fee on particular behaviors that are costly to the network, like dust tx.
355 2013-03-09 06:54:06 <midnightmagic> and effectively kill micropayments! woo
356 2013-03-09 06:54:17 <warren> midnightmagic: micropayments are already dead there
357 2013-03-09 06:54:41 <midnightmagic> well kill them some more, i say
358 2013-03-09 06:54:45 <warren> midnightmagic: not really
359 2013-03-09 06:54:58 <midnightmagic> what? why not
360 2013-03-09 06:55:20 <warren> midnightmagic: this new fee formula would allow 0.01 or 0.001 easily (often without a mandatory fee with sufficient age), but 0.00000001 will have a ridiculously high fee.
361 2013-03-09 06:55:52 <warren> This isn't censorship, it's a pollution tax. Market solution to a network problem.
362 2013-03-09 06:56:31 <warren> midnightmagic: currently it's hard to transfer a zero age 1LTC, you get hit by 0.2 fee, I think.
363 2013-03-09 06:56:43 <warren> ACTION hasn't looked closely enough.
364 2013-03-09 06:58:00 <warren> midnightmagic: this is really similar to Vinnie_win's policy idea of simply disallowing relaying of dust tx's that are smaller than fees. Except I don't suggest it is a blanket ban, just raising the cost of it, and miners like the idea because it increases fees.
365 2013-03-09 06:59:39 <dgriffi> dammit... I can't figure out how to get a plain old string out of AddressTableModel::Address. Any ideas?
366 2013-03-09 06:59:46 <Luke-Jr> warren: ???
367 2013-03-09 07:00:10 <warren> Luke-Jr: I'm open to hearing criticism.
368 2013-03-09 07:00:43 <Luke-Jr> warren: either it's pollution or it isn't. if it is, it should be banned outright; if it isn't, the ridiculous fees hurt someone
369 2013-03-09 07:01:05 <warren> Luke-Jr: ah yes, I agree it *should* be banned, but
370 2013-03-09 07:01:27 <dgriffi> can someone help me understand AddressTableModel::Address?
371 2013-03-09 07:02:08 <Luke-Jr> warren: and if it's so high that it's effectively banned and nobody is ever hurt by it because it really is pollution - then miners have nothing to gain either
372 2013-03-09 07:02:11 <dgriffi> I got the clipboard-writing part done.. can't get at the string I need
373 2013-03-09 07:02:54 <warren> Luke-Jr: Consider Bitcoin. The userbase is screaming that the sky is falling, they like their DP addiction, and they quote the legendary founder as justification for the status quo. I think the discourse has gone in the wrong direction because people suggesting it is pollution are using the wrong arguments.
374 2013-03-09 07:03:10 <Luke-Jr> wtf is DP
375 2013-03-09 07:03:26 <warren> ACTION finds the origin pastebin...
376 2013-03-09 07:03:40 <warren> http://pastebin.com/ng9nF4K3
377 2013-03-09 07:03:57 <Luke-Jr> I don't get it?
378 2013-03-09 07:04:07 <warren> Luke-Jr: that's our new name for SD.
379 2013-03-09 07:04:32 <Luke-Jr> o
380 2013-03-09 07:04:37 <dgriffi> dice parasites?
381 2013-03-09 07:04:46 <Luke-Jr> warren: that's not that many SD addicts
382 2013-03-09 07:04:56 <xjrn> clarity pollution
383 2013-03-09 07:05:01 <warren> Luke-Jr: and that's part of the justification for pollution
384 2013-03-09 07:07:34 <warren> Luke-Jr: 90% of the block for a tiny number of actual users? That's a textbook fact pattern of a negative externality. I thought about this and concluded that our struggles in dealing with this were based on a flawed premise made in the beginning, that every KB has the same cost.
385 2013-03-09 07:08:03 <warren> Luke-Jr: It turns out that some *behaviors* have a higher cost even though they are the same KB size.
386 2013-03-09 07:08:16 <Luke-Jr> warren: how so?
387 2013-03-09 07:08:53 <warren> Luke-Jr: large numbers of dust transactions that are unlikely to be spent. That's fine if it is in proportion to growth of number of actual users on the network.
388 2013-03-09 07:09:11 <Luke-Jr> true
389 2013-03-09 07:09:23 <Luke-Jr> I was thinking the cost of users not adopting bitcoin :P
390 2013-03-09 07:09:30 <xjrn> it seems like there's a protocol denormalization of something with a functionally lower entropy than what is wired
391 2013-03-09 07:09:35 <warren> Luke-Jr: Other examples of behaviors may include chains of 0-conf to 0-conf spending repeatedly.
392 2013-03-09 07:09:47 <iwilcox> Luke-Jr: That's a very difficult one to positively quantify or prove. Not that I doubt it.
393 2013-03-09 07:10:09 <Luke-Jr> iwilcox: yeah, I agree warren's cost is more obvious and direct
394 2013-03-09 07:10:33 <warren> Luke-Jr: I haven't figured out all possible behaviors that are detectable and *always* different from normal behavior.
395 2013-03-09 07:11:32 <warren> Luke-Jr: but my argument here is *nothing* against DP specifically. It is not a value judgement about gambling being good or bad. It is only to examine the fee formula to assign true network costs to behaviors, as per-KB cost isn't cutting it.
396 2013-03-09 07:11:49 <xjrn> warren quanitfy the 'true' network costs
397 2013-03-09 07:11:53 <Luke-Jr> warren: sure, I've never had a problem with DP in terms of it being gambling.
398 2013-03-09 07:12:25 <warren> What I'm suggesting won't stop DP. It might stop the instant lose dust.
399 2013-03-09 07:13:44 <warren> Luke-Jr: Now why I suggest a higher fee instead of a ban ... you balance the magnitude of the higher fee such that it doesn't entirely stop the network cost, it just makes it more expensive. That way miners won't fear losing income and they are more likely to accept the new policy.
400 2013-03-09 07:13:55 <Luke-Jr> most of the problem with DP is short-term; long-term, the fees should take care of it in one way or another
401 2013-03-09 07:14:28 <Luke-Jr> warren: a higher fee and a ban are practically equivalent
402 2013-03-09 07:14:35 <midnightmagic> i'm fairly sure most miners are primarily interested in the blockreward and not the fees.
403 2013-03-09 07:14:40 <warren> Luke-Jr: I keep hearing that, but I don't see how it is a realistic expectation. In a few months we're going to hit against the next soft limit and people will scream again to raise it, with no other changes.
404 2013-03-09 07:14:57 <warren> midnightmagic: the fees are pretty substantial now
405 2013-03-09 07:15:02 <Luke-Jr> warren: short-term as in, a few more years
406 2013-03-09 07:15:20 <Luke-Jr> warren: 5 years from now, I think the fee system alone (perhaps with your modifications) would deal with DP
407 2013-03-09 07:15:35 <midnightmagic> warren: I haven't paid a fee more than once or twice since I started mining in Dec '10.
408 2013-03-09 07:15:52 <dgriffi> midnightmagic: what's your mining rig?
409 2013-03-09 07:16:01 <warren> Luke-Jr: whenever that comes, the "spam" will continue to be disproprotionate to user growth, and our only "solution" will be to just raise the block size.
410 2013-03-09 07:16:21 <warren> Luke-Jr: how will the fee system deal with DP? nobody has explained that.
411 2013-03-09 07:16:23 <xjrn> "only" is a shortsighted classification of your solution warren
412 2013-03-09 07:16:32 <jgarzik> soft limit seems pointless to me
413 2013-03-09 07:16:33 <Luke-Jr> warren: the fees will go up until gamblers are unwilling to flood
414 2013-03-09 07:16:36 <midnightmagic> dgriffi: Essentially it is not possible for me to get any more hashrate without doing something exotic or renting other places.
415 2013-03-09 07:16:54 <jgarzik> unless you like drilling people against the eventuality of raising the hard limit :(
416 2013-03-09 07:16:55 <Luke-Jr> warren: by that time, ordinary users will have to be using off-chain transactions most of the time
417 2013-03-09 07:17:02 <jgarzik> set the soft limit == hard limit immediately
418 2013-03-09 07:17:26 <warren> Luke-Jr: the casino takes in only 1.9% right now. real life casinos are 20-40% more. There's a LONG way to go before fees overcome the addictive drive to stupidly pay them.
419 2013-03-09 07:18:12 <Luke-Jr> seriously? O.O
420 2013-03-09 07:18:25 <Luke-Jr> just how stupid can people get?
421 2013-03-09 07:18:25 <warren> I doubt they'll pay to 20%, maybe 5% easily
422 2013-03-09 07:18:29 <warren> hahahha
423 2013-03-09 07:18:37 <dgriffi> okay, I think I can get at the string I need
424 2013-03-09 07:18:43 <midnightmagic> warren: You have some fee changes proposed somewhere? Do you have a summary or a pull request somewhere I could look at?
425 2013-03-09 07:18:43 <warren> Luke-Jr: the casino tells you up front that you will lose, and people play anyway.
426 2013-03-09 07:18:45 <dgriffi> boy, this code is convoluted.
427 2013-03-09 07:19:11 <warren> midnightmagic: I need to do some statistical analysis before I come up with actual numbers to propose.
428 2013-03-09 07:19:11 <xjrn> reducing the entropy of the block contents, e.g. lossless data compression, and finding a means of relative reference, e.g. hufman tree of wallets and block offsets would smash the wire protocol down a bunch
429 2013-03-09 07:19:12 <Luke-Jr> ACTION facepalms
430 2013-03-09 07:19:55 <warren> xjrn: DP could use compressed keys to make their impact smaller, but they refuse and do anything that the system allows.
431 2013-03-09 07:20:31 <Luke-Jr> I have to admit, after figuring out that there's a 64% chance of "winning" a fair gamble, it was tempting to try it IRL, but .. 20-40% house edge, wtf ..
432 2013-03-09 07:20:35 <kuzetsa> there's STILL a bug where bitcoin-qt / bitcoind likes to claim "blah blah not enough after sending the 0.0???" fee and then when the fee is actually applied, it uses a different value
433 2013-03-09 07:20:47 <warren> Luke-Jr: if we're willing to accept the original design as holy and infallible and forever accept DP's shifting network costs upon others for their private gain, then sure, let's do it.
434 2013-03-09 07:20:52 <xjrn> warren if i understand correctly because the one true blockref.dat lays the blockchain out in its most denormalized form, the complaint has to be about the abusers and not about the one-true blockchain serializer codebase.
435 2013-03-09 07:20:58 <xjrn> that is lame
436 2013-03-09 07:21:06 <dgriffi> is there any way to rudely boot SD out of bitcoindom?
437 2013-03-09 07:21:24 <Luke-Jr> dgriffi: block_dice branch
438 2013-03-09 07:21:26 <warren> dgriffi: no way to do it without the miners agreeing.
439 2013-03-09 07:21:46 <xjrn> if tthe blockchain were represented as a huffman tree of previous tx and blocks, the bits to describe SD would be trivial and self-limitting based on frequency
440 2013-03-09 07:21:49 <dgriffi> warren: that may be the solution to this problem.
441 2013-03-09 07:21:52 <Luke-Jr> warren: we could make a client that punishes miners who don't block SD
442 2013-03-09 07:22:20 <dgriffi> warren: spread the word to the miners that they can speed up the network by shunning SD
443 2013-03-09 07:22:32 <warren> dgriffi: they know. they want the fees.
444 2013-03-09 07:22:35 <Luke-Jr> dgriffi: all the big pools don't care about Bitcoin's long-term viability
445 2013-03-09 07:22:47 <warren> dgriffi: miners are driven by short-term profit
446 2013-03-09 07:23:00 <warren> generally that's good, more hashing, more secure network
447 2013-03-09 07:23:02 <Luke-Jr> warren: some
448 2013-03-09 07:23:32 <kuzetsa> SendCoinsDialog::on_sendButton_clicked() ... on the line with tr("The total exceeds your balance when the %1 transaction fee is included."). it's being passed a different value than the actual fee that is used
449 2013-03-09 07:23:34 <dgriffi> okay, how about something that can be done with bitcoin clients to punish SD?
450 2013-03-09 07:23:40 <iwilcox> dgriffi: Any effort to kick SD that's seen as imposed upon the network instead of evolved by it will fail.
451 2013-03-09 07:24:06 <iwilcox> Kicking SD isn't the goal, anyway.
452 2013-03-09 07:24:21 <kuzetsa> iwilcox: well maybe not YOUR goal
453 2013-03-09 07:24:26 <xjrn> there's no need to punish users if the frequency can be countered with reduced entropy, which seems textbooko here
454 2013-03-09 07:24:27 <iwilcox> Encouraging sustainable behaviour is.
455 2013-03-09 07:24:30 <kuzetsa> some people would like to see SD go away
456 2013-03-09 07:24:30 <warren> Luke-Jr: the cost that we accept by simply accepting bigger block sizes instead of thinking critically, sooner than we wanted to, the average home user don't want to run the full client anymore.
457 2013-03-09 07:25:10 <warren> iwilcox: right, kicking DP isn't the goal. the goal is to better allow the cost of the network to scale with user growth.
458 2013-03-09 07:25:58 <xjrn> dgriffi: yeah changing the clients to handle thier internal representation and wire a normalized signal would do what they're proposing without a cat and mouse game or a mob blocklist
459 2013-03-09 07:26:23 <warren> Luke-Jr: anyway, it appears unlikely that I can convince people here because there is a lot of momentum in the original design, so I will propose it for litecoin as an experiment. Their users want a fee change now, so it is an opportune time.
460 2013-03-09 07:26:46 <iwilcox> Don't tell them it's an experiment :)
461 2013-03-09 07:27:10 <iwilcox> "We thought we'd try this on an inconsequential bullshit currency first..."
462 2013-03-09 07:27:26 <warren> I'm calling it an experiment from bitcoin's perspective, but I'm convinced it will work awesome for litecoin. (Assuming litecoin has any value ... which it doesn't.)
463 2013-03-09 07:27:40 <warren> (but the users seem to think it does, so whatever)
464 2013-03-09 07:28:23 <xjrn> warren try an experiment, gzip your blockchain. does it compress ?
465 2013-03-09 07:28:41 <warren> litecoin is afraid of spam. I will give them a way to drop standard fees by 1000x without the risk of spam.
466 2013-03-09 07:29:31 <warren> xjrn: it isn't the blockchain storage itself, but the unspended coins that must be in memory or fast storage forever that is a problem.
467 2013-03-09 07:33:26 <iwilcox> What baffles me about this whole thing is that nobody seems to object to minimum fees. Minimum fees prevent unsustainable volumes of microtransactions, which might be genuine financial use of Bitcoin. SD's use of Bitcoin as a binary instant messaging system in the lost bet case is clearly not a financial use, yet suggestions to tax it cause uproar.
468 2013-03-09 07:34:17 <iwilcox> Surely if you're against a tax on SD-like behaviour, you want minimum fees abolished entirely. But objectors don't seem to go that far.
469 2013-03-09 07:34:35 <warren> iwilcox: I'd be fine with that too. I am only suggesting a theoretical justification and method of implementation that miners won't object to.
470 2013-03-09 07:35:01 <dgriffi> I am on the verge of getting this clipboard thing working
471 2013-03-09 07:35:09 <warren> iwilcox: problem ... what is the minimum fee?
472 2013-03-09 07:35:57 <dgriffi> IT WORKS!
473 2013-03-09 07:36:12 <warren> iwilcox: we have lots of little miners now collecting 0.002 and shrinking payouts. That seems inconsequential to bitcoin veterans, but that's 8.8 cents now.
474 2013-03-09 07:36:15 <iwilcox> You'd be fine with that; I'm sure you'll do good research too, and the miners will be fine with that. A bunch of folk on bitcointalk will reject it outright as oppression, on some general principle they don't follow through to conclusion.
475 2013-03-09 07:37:01 <warren> iwilcox: ah, users must accept whatever policy miners accept. If reference client accepts, and miners accept, users must.
476 2013-03-09 07:37:28 <Luke-Jr> warren: few miners use the reference client policies as-is
477 2013-03-09 07:37:45 <warren> Luke-Jr: hence my proposal has economic incentive for miners to accept it
478 2013-03-09 07:39:01 <xjrn> warren it approaches infinity asymptotically, but holds to a rate of entropy that's higher than it has to be. consider http://en.wikipedia.org/wiki/Kolmogorov_complexity and the two examples in the beggining.
479 2013-03-09 07:39:27 <warren> xjrn: we're talking about two entirely different things, you entirely ignored my answer
480 2013-03-09 07:39:47 <xjrn> warren: you can solve your own problem by proposing a better normalized wire format and representation which gains consensus via proof. trying to politic a policy change is just going to hurt the poeple you claim to be acting for
481 2013-03-09 07:40:53 <dgriffi> you guys want to check out my modification?
482 2013-03-09 07:40:57 <Luke-Jr> xjrn: you can't do that without a hardfork
483 2013-03-09 07:41:02 <warren> Luke-Jr: I haven't done the math yet so I don't know if this is a good equilibrium point, but for example: currently DP pays 0.001 fee per dust? What if that were forced to be 0.0015 by the dust-specific fee? Miners would love the opportunity to claim a larger portion of DP's ridiculous profit.
484 2013-03-09 07:41:03 <dgriffi> see https://github.com/DavidGriffith/bitcoin
485 2013-03-09 07:41:06 <xjrn> can you prove there's a lower entropy state of either on-0disk, or live, hashed blocks? sure, gzip either one, and see if you can hash them uniquely
486 2013-03-09 07:41:46 <iwilcox> xjrn: Do you consider the SD lost-bet behaviour to be a problem if something (client, protocol, SD itself) doesn't change?
487 2013-03-09 07:41:50 <xjrn> can you add content-encoding to block representation? sure, politic that mod to cooperative p2p clients. e.g. the one-true-bitcoind codebase
488 2013-03-09 07:41:55 <Luke-Jr> warren: you can't force someone to change a fee
489 2013-03-09 07:42:21 <warren> Luke-Jr: the dust fee is not imposed by the reference client per-se, but rather miners realizing "ah ha! I can make bigger profits by colluding with other miners to force more fees for certain behaviors."
490 2013-03-09 07:42:39 <xjrn> iwilcox: i consider it an annoyance that doesn't have enough bitcoins bounty for me to solve myself, a terabyte drive is liek $40 at fry's
491 2013-03-09 07:42:40 <warren> Luke-Jr: it just so happens that the reference client calculates that way too
492 2013-03-09 07:42:48 <xjrn> and my btrfs gzips what i tell it to anyways
493 2013-03-09 07:43:12 <warren> xjrn: You entirely ignored my response. It is NOT about the block chain storage size.
494 2013-03-09 07:43:23 <warren> well, mostly not
495 2013-03-09 07:43:43 <iwilcox> xjrn: What's your opinion on the minimum fee, as currently implemented/enforced by the network?
496 2013-03-09 07:43:45 <xjrn> warren: you said SD could compress, but doesn't, i guess you know something i don't here about the features and options
497 2013-03-09 07:43:58 <xjrn> s/DP/
498 2013-03-09 07:44:22 <warren> xjrn: they could choose to switch to compressed keys if they wanted to
499 2013-03-09 07:44:38 <Luke-Jr> xjrn: everyone else is using compressed keys
500 2013-03-09 07:44:58 <warren> xjrn: their response is "screw you, I'm doing what's allowed by the rules"
501 2013-03-09 07:45:05 <xjrn> iwilcox: so be it. if i patch bitcoind to deflate before hasing, OPTIONALLY, and its adopted, the cost of each tx is a bit smaller. if i make huffman tree backrefs to previous tx, blocks, fractional amounts, the bits go waaay down, the fees can stay as they are.
502 2013-03-09 07:45:33 <xjrn> i guess merkel is already a digest of the tx
503 2013-03-09 07:45:42 <warren> Luke-Jr: oh right, another pollution behavior is un-compressed key use. Increase the fee of that slightly to discourage it.
504 2013-03-09 07:45:59 <warren> That isn't "oppression".
505 2013-03-09 07:46:00 <xjrn> but then they follow as plaintext or blockheaders or whatever suitable for python code.
506 2013-03-09 07:46:56 <iwilcox> xjrn: What do you feel you get for your minimum fee, on those transactions you pay it on?
507 2013-03-09 07:46:59 <dgriffi> well, since you're in this debate, I think I'll depart. I'm filing a pull request for my clipboard change in a couple minutes
508 2013-03-09 07:47:24 <warren> dgriffi: cool, URL please
509 2013-03-09 07:47:34 <dgriffi> warren: https://github.com/DavidGriffith/bitcoin
510 2013-03-09 07:47:55 <dgriffi> warren: the fix was suprisingly simple.. I was getting ready to have to write a bunch of xlib code
511 2013-03-09 07:48:35 <warren> Luke-Jr: If everyone else uses compressed keys by default, we must have decided that's a superior default, so why not make it slightly cheaper than un-compressed keys? Where's the harm?
512 2013-03-09 07:48:50 <dgriffi> how is a key compressible?
513 2013-03-09 07:48:53 <Luke-Jr> warren: it already is, since it's smaller
514 2013-03-09 07:49:02 <xjrn> iwilcox: hypothetical ranking in line is what I "feel i get". i don't see my bitcoin-qt appraoching a fraction of anything else i use my bandwidth for SD or no
515 2013-03-09 07:49:03 <dgriffi> that seems to go against everything I've learned about cryptography
516 2013-03-09 07:49:35 <warren> Luke-Jr: Is the flat fee per KB an infallible doctrine that we must accept?
517 2013-03-09 07:49:48 <iwilcox> dgriffi: I found http://bitcoin.stackexchange.com/questions/3059/what-is-a-compressed-bitcoin-key/3062#3062 digestible
518 2013-03-09 07:49:56 <Luke-Jr> warren: no, just saying it's already cheaper
519 2013-03-09 07:50:39 <warren> Luke-Jr: jack up the price difference by 5%. Ordinary users won't care. Frequent users will change to reduce costs. Everyone wins.
520 2013-03-09 07:51:06 <Luke-Jr> warren: patches welcome (against 0.8.0.eligius branch)
521 2013-03-09 07:51:11 <xjrn> dgriffi: if you do a wireshark dump of bitcoind/qt traffic, and gzip it, that's a sign that there a) is no room for lower entropy over the wire (unlikely), or b) is more normalized forms to be researched
522 2013-03-09 07:51:23 <warren> Luke-Jr: ok, it's on my list
523 2013-03-09 07:52:05 <xjrn> dgriffi: if you assign relative position to previous tx, previous wallets, and previous ratios of tx, you may use fewer bits still
524 2013-03-09 07:52:39 <Luke-Jr> warren: note that GetMinFee is ignored for mining at present
525 2013-03-09 07:52:46 <xjrn> dgriffi: then the vote goes to plausible implementation moreso than agreeing who to ban and who to up-charge
526 2013-03-09 07:52:49 <warren> What's new in Bitcoin 0.6.0: ... New wallets created with this version will use 33-byte 'compressed' public keys instead of 65-byte public keys, resulting in smaller transactions and less traffic on the bitcoin network.
527 2013-03-09 07:53:37 <warren> Luke-Jr: none of this will be possible overnight. It will be part of a basket of cost analysis proposals.
528 2013-03-09 07:54:08 <Luke-Jr> warren: a basket isn't likely to get adopted.
529 2013-03-09 07:54:12 <warren> Luke-Jr: and I expect miners to adjust these parameters on their own to compete for whatever they think will maximize profit
530 2013-03-09 07:54:29 <warren> Luke-Jr: ok, then separate but related proposals
531 2013-03-09 07:54:44 <Luke-Jr> IMO the punishment for uncompressed keys is a no-brainer
532 2013-03-09 07:54:50 <warren> yeah
533 2013-03-09 07:54:59 <Luke-Jr> I'd deploy that overnight
534 2013-03-09 07:55:02 <warren> can you think of other no brainers?
535 2013-03-09 07:55:17 <xjrn> isn't punishment for uncompressed keys implicit already? you increase your tx sizes?
536 2013-03-09 07:55:38 <Luke-Jr> xjrn: already said that. "punishment" means an ADDITIONAL penalty
537 2013-03-09 07:55:46 <warren> all for it
538 2013-03-09 07:56:07 <warren> xjrn: there's no switching costs, so not crying for the "victims"
539 2013-03-09 07:56:09 <xjrn> i had no idea, btw warren that you were referring to an implementation detail, so i really did overlook it as nonsequitor
540 2013-03-09 07:56:29 <Luke-Jr> warren: actually, there is a small switching cost
541 2013-03-09 07:56:33 <lianj> Luke-Jr: isnt that just rude for already existing uncompressed keys?
542 2013-03-09 07:56:56 <Luke-Jr> but since people aren't supposed to use addresses more than once anyway, I don't care :p
543 2013-03-09 07:57:11 <xjrn> i would stick with the implicits, personally
544 2013-03-09 07:57:21 <warren> lianj: make the additional fee small enough that ordinary users won't care, as coin age makes the fee zero anyway
545 2013-03-09 07:57:30 <lianj> Luke-Jr: pfft, lots people use addresses more than once
546 2013-03-09 07:57:40 <Luke-Jr> lianj: they're not supposed to, it's bad practice, etc
547 2013-03-09 07:57:53 <lianj> Luke-Jr: yep, still happens
548 2013-03-09 07:57:53 <Luke-Jr> and 0.6.0 was released nearly a year ago
549 2013-03-09 07:57:54 <iwilcox> lianj: There's nothing implicit in an existing address that make it uncompressible.
550 2013-03-09 07:57:59 <Luke-Jr> plenty of time to use up your keypool
551 2013-03-09 07:58:05 <Luke-Jr> iwilcox: yes there is
552 2013-03-09 07:58:25 <Luke-Jr> iwilcox: you cannot compress an existing address
553 2013-03-09 07:58:53 <warren> lianj: ordinary users won't notice the change
554 2013-03-09 07:58:54 <lianj> iwilcox: every privkey has two addresses, the compressed and the uncompressed
555 2013-03-09 07:59:17 <midnightmagic> wonder if the vanitygen does compressed keys.. hrm..
556 2013-03-09 07:59:32 <Luke-Jr> lianj: worst case, the sends just take a little longer to make it up in priority
557 2013-03-09 07:59:37 <Luke-Jr> midnightmagic: good question
558 2013-03-09 07:59:38 <lianj> warren: just let them pay the increased tx size like it is now. why isnt that enough?
559 2013-03-09 08:00:19 <warren> lianj: read the last hour. Per-KB size flat fee is a design flaw. Tiny adjustments are needed to reflect true network costs.
560 2013-03-09 08:00:52 <warren> lianj: I'm not suggesting a 1000% punishment. like 5-10% that won't hurt anyone with coin age.
561 2013-03-09 08:01:17 <Luke-Jr> warren: well, technically, the only cost of an uncompressed key IS the size
562 2013-03-09 08:01:26 <lianj> warren: better kill 1dice then
563 2013-03-09 08:01:42 <warren> Luke-Jr: doesn't stop us from deciding to make it bigger
564 2013-03-09 08:02:09 <Luke-Jr> lianj: exactlyu
565 2013-03-09 08:02:29 <warren> There's no drawback to this.
566 2013-03-09 08:02:47 <warren> well, bitcoin-0.6.0 decided the drawback wasn't worthwhile to think about.
567 2013-03-09 08:03:02 <xjrn> shaving 20 bytes is one thing. resolving a spammer to less than 32 bits wallet reference is doable if they keep the spam alive over time
568 2013-03-09 08:03:44 <iwilcox> warren: The motivations for proposals in your basket are going to have to be meaningful to people with fat pipes, big disks and loads of RAM why the status quo is bad, in very pragmatic terms. I don't think "true network costs" will appeal to non-miners.
569 2013-03-09 08:04:03 <iwilcox> You'll need "why it's bad for your gran" arguments.
570 2013-03-09 08:05:21 <Luke-Jr> yo gran so old she still uses uncompressed keys
571 2013-03-09 08:05:34 <xjrn> http://en.wikipedia.org/wiki/Huffman_coding a coordinate system for blockchains could affect every claim of spam with a frequency based feedback loop reflected in lookup size
572 2013-03-09 08:06:11 <xjrn> it would "reward" the spammer and the victim of spam with smaller tokens to describe SD
573 2013-03-09 08:06:14 <iwilcox> xjrn: Patches welcome :)
574 2013-03-09 08:06:29 <warren> iwilcox: the bloat of externalized true network costs make it more unpleasant for more users to join. If transaction volume scales with user adoption, then that is efficient cost acceptance. If when new users hear about the full client and find that it uses way too much resources because there's 10000x preventable dust tx's, then we have potentially lost a new user.
575 2013-03-09 08:06:45 <warren> Luke-Jr: lol
576 2013-03-09 08:06:56 <dgriffi> I have another bitcoin-qt question... why doesn't the program have an "install" make target?
577 2013-03-09 08:07:00 <xjrn> iwilcox: my dayjob is my favorite pastime
578 2013-03-09 08:07:14 <Luke-Jr> dgriffi: because you didn't write one yet
579 2013-03-09 08:07:42 <warren> dgriffi: most makefiles get it wrong anyway, the distos and windows differ substantially
580 2013-03-09 08:08:00 <xjrn> bitcoin rates anywhere from 1/200th to 1/2000th as important per day to code for until BFL decides to ship
581 2013-03-09 08:08:08 <dgriffi> so, maybe a little shell script that does the job for the unix people?
582 2013-03-09 08:08:10 <warren> dgriffi: packaging software into an rpm, you do not "make install" something directly into /usr/bin.
583 2013-03-09 08:08:21 <warren> dgriffi: where is the "correct" location?
584 2013-03-09 08:08:31 <iwilcox> xjrn: More seriously, I don't think anyone's arguing that a more efficient Bitcoin would be a bad thing; AIUI the argument is that the resources aren't available to reduce the effect on Grandma today by that route.
585 2013-03-09 08:08:40 <dgriffi> warren: if you build something yourself, it gets installed to /usr/local
586 2013-03-09 08:08:56 <warren> dgriffi: yeah, that's not desired for packagers
587 2013-03-09 08:09:13 <warren> is that where mac puts it?
588 2013-03-09 08:09:40 <xjrn> iwilcox: a bitcoin client with 2 flavors is reasonable, but not trivial. it's what happens to any python codebase in the long run
589 2013-03-09 08:09:43 <dgriffi> warren: I know. the packaging schemes all maintain and control stuff in /usr/.. but /usr/local is intended to be manually manipulated
590 2013-03-09 08:10:15 <dgriffi> warren: /opt is another place to put things, with its own pros and cons
591 2013-03-09 08:10:28 <warren> dgriffi: the usual way to do what you want is with autoconf to auto-generate the target system's desired paths, but qt generally has its own thing
592 2013-03-09 08:10:33 <midnightmagic> /opt/sbin/bitcoind !
593 2013-03-09 08:10:44 <iwilcox> No, /opt/bitcoin/bitcoind :)
594 2013-03-09 08:10:55 <iwilcox> (Hi, Sun!)
595 2013-03-09 08:11:03 <midnightmagic> /opt/bitcoin/sbin/bitcoind
596 2013-03-09 08:11:14 <midnightmagic> /usr/pkgsrc/sbin/bitcoind ?
597 2013-03-09 08:11:29 <xjrn> pools (glaring at Luke-Jr ) have nothing to gain by not aggressively normalizing the wire protocol.
598 2013-03-09 08:11:37 <dgriffi> /opt/bitcoin/bin/bitcoin-qt and /opt/bitcoin/share
599 2013-03-09 08:11:47 <Luke-Jr> /Software/net-p2p/bitcoind-0.9.0-r1/bin/bitcoind
600 2013-03-09 08:11:47 <warren> xjrn: you're good at ignoring responses
601 2013-03-09 08:11:51 <xjrn> normalizing the shitlist will only be a cat and mouse game band-aid
602 2013-03-09 08:12:17 <midnightmagic> I wonder if xjrn is a dadabot
603 2013-03-09 08:12:36 <midnightmagic> xjrn: Hey! Are you a dadabot?
604 2013-03-09 08:12:54 <warren> midnightmagic: hahaha
605 2013-03-09 08:12:55 <xjrn> warren: don't get so touchy I'm a troll fer god sakes
606 2013-03-09 08:13:12 <iwilcox> If eric publicly coughed up the funds he pledged in that thread, perhaps the devs could take a two-week holiday to work on Bitcoin wire/disk format optimisations ;)
607 2013-03-09 08:13:15 <warren> midnightmagic: that would make SO much sense if true.
608 2013-03-09 08:13:47 <midnightmagic> Just regurgitating ancient #bitcoin-dev from way back that is ~ relevant.
609 2013-03-09 08:13:58 <Luke-Jr> iwilcox: huh?
610 2013-03-09 08:14:24 <midnightmagic> warren: See? No answer.
611 2013-03-09 08:15:09 <warren> midnightmagic: thank you.
612 2013-03-09 08:16:08 <iwilcox> Luke-Jr: https://bitcointalk.org/index.php?topic=150493.msg1601093#msg1601093
613 2013-03-09 08:19:01 <warren> "We cannot advance Bitcoin by asking anyone who is putting stress on the system to "please refrain." I do not think Bitcoin is so delicate and poorly constructed as you. It can handle SatoshiDice, it can handle far more than SatoshiDice, and I congratulate and encourage everyone else out there who is building infrastructure and putting stress on this system."
614 2013-03-09 08:19:15 <dgriffi> about installing bitcoin-qt... is share/ used at all after compilation is finished?
615 2013-03-09 08:19:19 <warren> He's correct. And making the block size bigger isn't our best solution either.
616 2013-03-09 08:20:18 <warren> dgriffi: I'm surprised that we don't ship some scripts like the wallet cleanup script
617 2013-03-09 08:20:24 <warren> I've been wondering why.
618 2013-03-09 08:20:42 <dgriffi> bitinstant was robbed
619 2013-03-09 08:20:47 <midnightmagic> Except there's a reason if the stress is greater than it can sustain; or if the stress impedes longterm growth. Satoshi specifically told people NOT to encourage Wikileaks to start using bitcoin as early as was being contemplated.
620 2013-03-09 08:20:52 <dgriffi> http://www.wired.com/wiredenterprise/2013/03/digital-thieves-pull-off-12000-bitcoin-heist/
621 2013-03-09 08:21:07 <xjrn> in order for bitcoin to reduce representational sizes, a coordinate system needs to be devised so that absolute pointers are not absolutely necessary to relay new blocks
622 2013-03-09 08:21:59 <midnightmagic> I wonder if I'm op here yet..
623 2013-03-09 08:22:01 <xjrn> moving to offsets from absolute addressing has its tradeoffs, but huffman coding and datacompression is in the win column
624 2013-03-09 08:22:10 <midnightmagic> nope.
625 2013-03-09 08:24:19 <Luke-Jr> wow
626 2013-03-09 08:24:19 <xjrn> or markov chains, just to be extra geek fashionable
627 2013-03-09 08:24:25 <Luke-Jr> Site5 is trying to pass the blame to BitInstant
628 2013-03-09 08:27:31 <MC1984> could it be said that in hindsight it was a mistake for bitcoin to have access to the full 8 decimal places this early
629 2013-03-09 08:27:41 <MC1984> given the dust issue
630 2013-03-09 08:29:14 <grau> MC1984: It would still not be late to modify relay rules, so transactions with output below the fee are not relayed.
631 2013-03-09 08:29:15 <dgriffi> warren: what script are you referring to?
632 2013-03-09 08:30:02 <MC1984> hmm yeah
633 2013-03-09 08:30:14 <MC1984> i wonder why satoshi didnt do that
634 2013-03-09 08:30:48 <grau> he was a human
635 2013-03-09 08:31:02 <warren> MC1984: +
636 2013-03-09 08:31:34 <warren> grau: If miners accept the "output below the fee are not relayed" then I'm perfectly happy.
637 2013-03-09 08:32:17 <grau> warren: you cant get them accept that, but one can lower the chances of dust being forwarded to them
638 2013-03-09 08:32:44 <warren> grau: that isn't realistic. You can connect directly to miners that do, to make sure that they do.
639 2013-03-09 08:33:11 <grau> warren: you can not connect to all miners directly
640 2013-03-09 08:33:14 <warren> grau: my proposal is far more complicated, but it actually stands a chance of miners agreeing to it in order to maximize profit
641 2013-03-09 08:33:22 <MC1984> miners are pretty happy ith dice though
642 2013-03-09 08:33:24 <MC1984> they get fees
643 2013-03-09 08:34:12 <midnightmagic> I am personally unaware of a single miner who's happy with DP.
644 2013-03-09 08:34:18 <wumpus> dgriffi: no, share is never used by the program itself
645 2013-03-09 08:34:27 <grau> sure, they only care of the fee/KB and until there is no competition to fill up the block they are happy with SD
646 2013-03-09 08:35:26 <grau> i mean no fee paying competition.
647 2013-03-09 08:35:29 <MC1984> so why dont allmost miners do that drop below minimum fee patch
648 2013-03-09 08:35:55 <Luke-Jr> MC1984: Satoshi's clients never used more than 2 decimal places
649 2013-03-09 08:36:27 <MC1984> Luke-Jr in the GUI or protocol?
650 2013-03-09 08:36:31 <Luke-Jr> GUI
651 2013-03-09 08:36:35 <Luke-Jr> and bitcoind
652 2013-03-09 08:36:46 <MC1984> i meant protocol
653 2013-03-09 08:36:46 <warren> grau: if you convince 40% of miners to not accept "output below the fee are not relayed", it is reasonably easy for DP to connect directly to friendly nodes that do to get it to the miners. If 70% of miners accept the "pollution tax" and demand higher fees or dust tx lower priority, then DP can still connect directly to friendly node miners, but their delay to confirmation becomes a lot more variable. As other payments are batched with the DP
654 2013-03-09 08:36:47 <warren> dust, someone else's payout becomes delayed, and DP users begin freaking out about confirmation delays.
655 2013-03-09 08:36:54 <Luke-Jr> MC1984: that's stupid
656 2013-03-09 08:37:03 <MC1984> so is dust
657 2013-03-09 08:37:10 <Luke-Jr> MC1984: it won't be dust in the future
658 2013-03-09 08:37:26 <MC1984> yeah so its bumped up eventually
659 2013-03-09 08:37:29 <Luke-Jr> MC1984: protocol changes take years to deploy, it makes sense to have it flexible
660 2013-03-09 08:37:32 <grau> SD dust problem will be solved as soon as there are fee paying other services competing for blockspace. Just matter of time.
661 2013-03-09 08:37:45 <Luke-Jr> MC1984: you can't just "bump up" it
662 2013-03-09 08:37:58 <warren> grau: that will happen slower if costs are higher than expected when new users join.
663 2013-03-09 08:38:00 <MC1984> i know its a fork
664 2013-03-09 08:38:30 <grau> warren: costs are low, what os the worry?
665 2013-03-09 08:38:39 <warren> grau: my proposal is also good in that it requires no soft or hard fork, and miners can change the policy at any time.
666 2013-03-09 08:39:16 <warren> wow, someone is DoS attacking my server
667 2013-03-09 08:39:20 <warren> this never happened before
668 2013-03-09 08:39:33 <grau> warren: point me to a summary of your proposal please
669 2013-03-09 08:39:34 <MC1984> ive read the agumnt that dic is burning up bitcoins 'startup capital' and it makes a lot of sense
670 2013-03-09 08:39:50 <warren> grau: I'll write it out in detail, not sure when
671 2013-03-09 08:39:58 <warren> grau: this is a long-term problem
672 2013-03-09 08:41:29 <kuzetsa> so I've been trying to ask about this bug for nearly a year now... I just tried again about an hour ago or something
673 2013-03-09 08:41:37 <grau> MC1984: yes, it burns on "startup capital" of the system, just like any early adopter of a technology does that
674 2013-03-09 08:41:40 <kuzetsa> the value in these two lines of code (for the fee does not match:
675 2013-03-09 08:41:52 <iwilcox> MC1984: Unfortunately it's an appeal to sentiment that seems to fall on rather unsympathetic ears.
676 2013-03-09 08:42:21 <MC1984> its not an appeal to sentiment
677 2013-03-09 08:43:05 <kuzetsa> tr("This transaction is over the size limit. You can still send it for a fee of %1, which goes to the nodes that process your transaction and helps to support the network. Do you want to pay the fee?").arg(BitcoinUnits::formatWithUnit(BitcoinUnits::BTC, nFeeRequired));
678 2013-03-09 08:43:09 <kuzetsa> tr("The total exceeds your balance when the %1 transaction fee is included.").
679 2013-03-09 08:43:09 <MC1984> dice seems to be burning capital in the 20 lines of coke on a 6ft mirror and greasy chinese food way
680 2013-03-09 08:43:18 <MC1984> not the way that advances the enterprise
681 2013-03-09 08:43:50 <warren> grau: " will be solved as soon as there are fee paying other services competing for blockspace" is an argument to just allow the startup capital to be burned prematurely while ONE participant profits from it.
682 2013-03-09 08:44:44 <kuzetsa> one was in sendcoinsdialog, the other was in the bitcoingui.cpp file
683 2013-03-09 08:44:53 <iwilcox> MC1984: It's an appeal to sentiment in that it's an appeal to our sense of fairness. If you're coldly rational or a diehard free market economist, you don't care for those arguments, was my point.
684 2013-03-09 08:45:14 <warren> grau: we wouldn't be caring about this if they only agreed to simple things like compressed keys and no more dust response. but Eric is right in asking nicely for an abuser to stop isn't a solution. This is forcing us to think creatively about the problem. I've identified the actual source of the problem is our flawed original theory of fee calculation.
685 2013-03-09 08:45:41 <MC1984> ok
686 2013-03-09 08:45:58 <kuzetsa> they appear to be using an inconsistent method to calculate the fee ammount to use and it's really annoying when you're trying to sweep your wallet (roll all the inputs into a single, larger transaction ID so you can better predict your fees at a later date)
687 2013-03-09 08:46:14 <warren> they?
688 2013-03-09 08:46:15 <MC1984> though destabalising the system at an arly stage and killing your cash cow should appeal to dice own sense of rationality one would hope
689 2013-03-09 08:46:38 <kuzetsa> warren: the two lines in question (one from sendcoinsdialog, the other was in the bitcoingui.cpp)
690 2013-03-09 08:47:00 <warren> MC1984: dice participants are willing to pay a much higher usage fee than new, scared users. THAT is the startup cost we're burning now because the original design is most holy.
691 2013-03-09 08:47:06 <MC1984> another theory is that the guys behind dice are purposefully trying to 'temper' bitcoin with heat
692 2013-03-09 08:47:17 <MC1984> especially some of them are with the founation i read
693 2013-03-09 08:47:19 <kuzetsa> one of the lines in the bitcoin-qt gui source code calculates the fee one way, the other from a different file of the source uses a different (inconsistent) method
694 2013-03-09 08:47:28 <warren> MC1984: anyone with money can join the foundation
695 2013-03-09 08:47:42 <MC1984> yeah but the eric guy is on the board
696 2013-03-09 08:48:33 <warren> I might be wrong, but the board doesn't decide what goes into the protocol, isn't the Foundation more of a promotion arm?
697 2013-03-09 08:48:51 <MC1984> i didnt say they do
698 2013-03-09 08:49:27 <xjrn> every venture starts out with denormalized data, and faces a crunch between throwing hardware at each milestone's solution, or normalizing what's relevant.
699 2013-03-09 08:49:42 <warren> xjrn: you actually made sense there.
700 2013-03-09 08:49:48 <MC1984> implied maybe someone with nough interest to be on the board of the foundation is dlibeately trying to temper bitcoin for what they see as the greater good
701 2013-03-09 08:49:50 <xjrn> bitcoin is pretty unremarkable in this sense
702 2013-03-09 08:50:11 <xjrn> warren: im responding to you now
703 2013-03-09 08:50:33 <warren> Right. So let's throw more hardware at the problem!
704 2013-03-09 08:50:48 <warren> Sounds great?
705 2013-03-09 08:51:15 <xjrn> afaict blockchains exist as denormalized data at every level
706 2013-03-09 08:51:46 <xjrn> it might as well be a fat mysql database with no indexes trying to update a million rows per hour
707 2013-03-09 08:51:54 <xjrn> that's what startups do
708 2013-03-09 08:52:09 <grau> You seem to claim SD uses an "unfair" share of the "startup capital". I think they simply do what is their interest is. If no one else finds a profitable business model for the chain, then their share is fair as is.
709 2013-03-09 08:52:54 <grau> I hope someone will soon find some better use of the chain and the problem is solved.
710 2013-03-09 08:53:05 <warren> grau: They're doing what is allowed under the current rules. That's fine. The network can decide to change the rules to adjust to their own interests that do not align with DP's interests.
711 2013-03-09 08:53:43 <warren> grau: " I hope someone will soon find some better use of the chain and the problem is solved." By competing on fees?
712 2013-03-09 08:54:40 <grau> warren: even if fees are equal, miner has at least to choose from.
713 2013-03-09 08:55:04 <grau> currently they would forgo profit if they would not mine SD
714 2013-03-09 08:55:09 <warren> grau: This is exactly the flawed reasoning I'm talking about.
715 2013-03-09 08:55:18 <warren> grau: I'm not suggesting they forego DP.
716 2013-03-09 08:55:19 <warren> at all
717 2013-03-09 08:55:28 <warren> grau: I'm suggesting they can demand more and DP can't stop them.
718 2013-03-09 08:55:47 <warren> And DP doesn't have a moral high horse to complain from.
719 2013-03-09 08:57:07 <warren> grau: My suggestion is not the most extreme proposal here. The leading proposal from yesterday seems to be to just block outputs that are smaller than fees.
720 2013-03-09 08:57:35 <xjrn> that creates a more anachronistic system in the long run
721 2013-03-09 08:58:02 <iwilcox> No, it buys us time to optimise stuff.
722 2013-03-09 08:58:14 <iwilcox> Without putting Grandma off for the rest of her life.
723 2013-03-09 09:00:51 <grau> warren: I support suppressing the relay of transactions without fee or with an output less than the fee of the transaction.
724 2013-03-09 09:01:53 <grau> My motivation is purely technical to avoid spam and UTXO fragmentation.
725 2013-03-09 09:02:04 <warren> grau: Right, both that proposal and mine would suppress it, either by ban or a tax.
726 2013-03-09 09:02:17 <Luke-Jr> grau: that's everyone's motivation who wants to block DP
727 2013-03-09 09:02:37 <warren> grau: I'm at least glad we're calling it DP.
728 2013-03-09 09:02:46 <warren> ACTION hint.
729 2013-03-09 09:05:36 <xjrn> iwilcox: that sounds as reactionary as warren and Luke-Jr. i'm only barely cognizant of the technical details but i can already tell it looks like amateur hour in the datacenter
730 2013-03-09 09:06:09 <warren> "i'm only barely cognizant of the technical details" describes all of your comments.
731 2013-03-09 09:06:21 <Luke-Jr> ^
732 2013-03-09 09:07:15 <xjrn> warren: I'm not ashamed of admitting what I haven't picked up yet. most of it looks like it's not worth perpetuating, it leads to the panic.
733 2013-03-09 09:08:43 <warren> You're lucky I never bothered to learn how to ignore someone in IRC.
734 2013-03-09 09:08:46 <xjrn> i had a mild interest in whether i should worry about SD a short while ago. I'm more worried about those who are taking quick shortcuts in here
735 2013-03-09 09:08:51 <Luke-Jr> xjrn: criticising solutions without having any better ideas is useless
736 2013-03-09 09:09:24 <warren> xjrn: Quick shortcut is throwing more hardware at the problem
737 2013-03-09 09:09:31 <xjrn> Luke-Jr: i don't have a reason to sit down and bang out an alternate client at this time, but for those of you who think you do, I can lend some perspective I'm not seeing in here
738 2013-03-09 09:10:23 <Luke-Jr> xjrn: looking forward to it
739 2013-03-09 09:10:27 <xjrn> Luke-Jr: you think you know C code, have you ever used relative addressing in asm code? is there a TIME AND PLACE for aboslute addressing, and relative addressing?
740 2013-03-09 09:11:02 <Luke-Jr> xjrn: I hold the position that there is no excuse to be using assembly except in bootstrapping C
741 2013-03-09 09:11:50 <xjrn> Luke-Jr: then go on about your politics with your pitch fork, you can't debate information thoery by stating compiler preferences
742 2013-03-09 09:12:08 <Luke-Jr> it's not compiler-related.
743 2013-03-09 09:13:15 <xjrn> then answer the question about relative and absolute addressing, or just write me off as a troll, either way I'm definitely not worried about dust transactions or SD or what follows when you plug the dyke with your finger and stand there
744 2013-03-09 09:13:22 <grau> Luke-Jr: The only excluse for using C is writing a virual machine ...
745 2013-03-09 09:13:48 <warren> Why are we debating with this guy?
746 2013-03-09 09:14:43 <xjrn> warren: I don't think you can
747 2013-03-09 09:14:44 <Luke-Jr> xjrn: http://www.youtube.com/watch?v=oHg5SJYRHA0 proves my point
748 2013-03-09 09:14:54 <Luke-Jr> warren: dunno, bored?
749 2013-03-09 09:21:41 <xjrn> Luke-Jr: i agreed with you at one point about banning SD tx before i gave any thought at all to the expense, which is repeating meaningless extra bits at every possible transaction from mining to p2p to on-disk indexing. it helps to know a 64 byte hash index the first time you publish it, but thereafter you can refer to 4 million pervious blocks with 4 bytes, to put it in language a c...
750 2013-03-09 09:21:43 <xjrn> ...programmer might recognize
751 2013-03-09 09:23:06 <Luke-Jr> xjrn: nobody cares about the disk space cost
752 2013-03-09 09:23:20 <warren> xjrn: I told you this multiple times.
753 2013-03-09 09:23:28 <xjrn> so then you can debate whetehr a 56 byte or a 33 byte wallet is better, or add anachronistic penalties to the simple system that already works, but you can refer to most previously seen wallets of any size with 9 bytes, given the previous 32 bit offset
754 2013-03-09 09:23:41 <warren> good night
755 2013-03-09 09:23:53 <Luke-Jr> night
756 2013-03-09 09:24:29 <xjrn> Luke-Jr: it doesn't matter if its disk or wire, the same thing applies. the same complaints are about absolute representations.
757 2013-03-09 09:24:38 <MC1984> what the fuck am i reading
758 2013-03-09 09:24:54 <Luke-Jr> xjrn: nobody cares about on wire either
759 2013-03-09 09:25:49 <grau> !ticker
760 2013-03-09 09:25:50 <gribble> BTCUSD ticker | Best bid: 46.06400, Best ask: 46.22998, Bid-ask spread: 0.16598, Last trade: 46.06400, 24 hour volume: 46746.73118119, 24 hour low: 42.50000, 24 hour high: 46.23000, 24 hour vwap: 44.03643
761 2013-03-09 09:25:51 <xjrn> Luke-Jr: i gave you more credit than i should have, my apoligies
762 2013-03-09 09:26:27 <xjrn> Luke-Jr: you're looking to ban tx based on niether a savings of disk or bandwidth, just out of pure spite
763 2013-03-09 09:26:45 <xjrn> good luck with that :)
764 2013-03-09 09:28:13 <Luke-Jr> nope, you're just too short sighted to see other limited resources
765 2013-03-09 09:28:30 <xjrn> Luke-Jr: like your intelligence?
766 2013-03-09 09:29:11 <Luke-Jr> guess I'll take after warren and stick you on ignore
767 2013-03-09 09:29:19 <xjrn> MC1984: see http://en.wikipedia.org/wiki/Kolmogorov_complexity
768 2013-03-09 09:29:44 <warren> Luke-Jr: I'm supposed to be sleeping. I'm almost tempted to learn how to ignore someone.
769 2013-03-09 09:30:18 <Luke-Jr> warren: it's easy, just don't read what they write, don't respond to it under any circumstances, or type /ignore <nick>
770 2013-03-09 09:30:41 <warren> Luke-Jr: so much effort
771 2013-03-09 09:30:43 <warren> but ok
772 2013-03-09 09:53:40 <anddam> warren: or, you know, /ignore
773 2013-03-09 09:57:00 <Luke-Jr> anddam: I did say that
774 2013-03-09 10:00:06 <anddam> Luke-Jr: I missed that, I was catching with the scroll history
775 2013-03-09 10:00:28 <anddam> Luke-Jr: I just saw a thread of yours on bitcointalk, you wrote a ASIC/FPGA miner
776 2013-03-09 10:00:29 <anddam> in C
777 2013-03-09 10:00:40 <anddam> is that correct?
778 2013-03-09 10:00:58 <Luke-Jr> anddam: I didn't write all of it, but I maintain one, yes
779 2013-03-09 10:13:30 <TheButterZone> https://bitcointalk.org/index.php?topic=150341.0 Generate own random numbers then incorporate into bitaddress.org script (.1 btc)
780 2013-03-09 11:03:59 <ciphermonk> what happened on testnet? My block count went back from 56118 to 56112
781 2013-03-09 11:06:09 <lianj> reorg?
782 2013-03-09 11:10:52 <Luke-Jr> lianj: reorgs should never go backward
783 2013-03-09 11:11:12 <Luke-Jr> ciphermonk: are you sure?
784 2013-03-09 11:11:28 <lianj> Luke-Jr: oh, true.
785 2013-03-09 11:11:30 <richweskus> Hey folks, is there any rss feed where i can get the ticker price?
786 2013-03-09 11:11:55 <richweskus> not necessarily rss any service to consume will do
787 2013-03-09 11:12:07 <Luke-Jr> richweskus: various exchanges might have them, but there is no "the price"
788 2013-03-09 11:12:26 <richweskus> mtgox?
789 2013-03-09 11:12:29 <Luke-Jr> #mtgox
790 2013-03-09 11:12:45 <ciphermonk> Luke-Jr: Yeah I've got blocks up to 56118 processed in my CouchDB ;) and bitcoind is advertising 56113 now
791 2013-03-09 11:13:04 <Luke-Jr> ciphermonk: did you restart bitcoind between the change?
792 2013-03-09 11:13:33 <ciphermonk> nope, it's on my VPS, I don't restart it
793 2013-03-09 11:13:56 <ciphermonk> I also have 3 orphaned generated blocks on there
794 2013-03-09 11:13:58 <Luke-Jr> ciphermonk: can you upload the debug.log somewhere? please be careful not to lose it (restarting bitcoind often truncates ti)
795 2013-03-09 11:14:11 <lianj> richweskus: https://api.twitter.com/1/statuses/user_timeline.rss?screen_name=bitcoinprice
796 2013-03-09 11:14:14 <ciphermonk> ok
797 2013-03-09 11:14:25 <lianj> and hope its the avg since last tweet/rss post
798 2013-03-09 11:14:34 <Luke-Jr> lianj: actually, I did think of a scenario where a reorg would move backward
799 2013-03-09 11:14:42 <richweskus> perfect lianj
800 2013-03-09 11:14:59 <richweskus> straight to the chase i like you :D
801 2013-03-09 11:19:15 <ciphermonk> I have this data for block 56118 on testnet
802 2013-03-09 11:19:20 <ciphermonk> {
803 2013-03-09 11:19:20 <ciphermonk> "hash": "00000000e9fbde2ae0f658a33b631b3e123df5a8eb1f393707210d03df4631c0", "time": 1362817248, "height": 56118, "transaction_count": 3}
804 2013-03-09 11:19:40 <Luke-Jr> ciphermonk: are you mining on this bitcoind?
805 2013-03-09 11:19:47 <ciphermonk> yes
806 2013-03-09 11:19:58 <Luke-Jr> ciphermonk: do you have the difficulties of each block?
807 2013-03-09 11:20:46 <ciphermonk> application-wise I just store hash, time, height and transaction count. I'm not sure how I would find the difficulties
808 2013-03-09 11:21:52 <ciphermonk> if they are not in the debug.log then probably not
809 2013-03-09 11:23:03 <Luke-Jr> ciphermonk: bitcoind getblock <hash>
810 2013-03-09 11:23:18 <Luke-Jr> ciphermonk: if you could pastebin the difficulty of every block in question, that would help
811 2013-03-09 11:23:26 <Luke-Jr> REORGANIZE: Disconnect 17 blocks; 0000000000fda51f20b0feb20e8eaf27c3e829ed0d4380995c8b247436b5d592..00000000fb95794f259f3b43040023d938a1988deb9822ca613cd4d231fdd906
812 2013-03-09 11:23:27 <Luke-Jr> REORGANIZE: Connect 2 blocks; 0000000000fda51f20b0feb20e8eaf27c3e829ed0d4380995c8b247436b5d592..0000000000c8c22a7e7f3dc5bef949fe856a4bd7538915b492b09eff02cde408
813 2013-03-09 11:23:31 <Luke-Jr> those are the two ranges in question, fwiw
814 2013-03-09 11:23:35 <ciphermonk> ok
815 2013-03-09 11:28:01 <ciphermonk> I'll pastebin those 17 blocks
816 2013-03-09 11:32:44 <ciphermonk> I'm still pasting but from what I see they have difficulty 1
817 2013-03-09 11:41:45 <gmaxwell> 04:10 < Luke-Jr> lianj: reorgs should never go backward
818 2013-03-09 11:41:48 <gmaxwell> sure, they can.
819 2013-03-09 11:42:13 <gmaxwell> E.g. you would prefer a short high difficulty chain to a longer diff=1 chain.
820 2013-03-09 11:42:16 <Luke-Jr> [12:14:07] <Luke-Jr> lianj: actually, I did think of a scenario where a reorg would move backward
821 2013-03-09 11:44:02 <ciphermonk> ok here are the blocks in the range: http://pastebin.com/U9tnePjD
822 2013-03-09 11:44:48 <Luke-Jr> ciphermonk: 0000000000c8c22a7e7f3dc5bef949fe856a4bd7538915b492b09eff02cde408 is misisng
823 2013-03-09 11:45:19 <ciphermonk> ah crap
824 2013-03-09 11:51:25 <Luke-Jr> ???
825 2013-03-09 11:51:59 <ciphermonk> http://pastebin.com/PWgpjXa8
826 2013-03-09 11:52:48 <Luke-Jr> yep, that's it
827 2013-03-09 11:52:58 <ciphermonk> I'm confused, I'm not sure I've got them all right. I have blocks 56112 to 56118 in by application database. The rest I worked my way through the previousblockhash
828 2013-03-09 11:53:17 <ciphermonk> I have 2 blocks 56110 for instance
829 2013-03-09 11:53:17 <Luke-Jr> ciphermonk: 0000000000c8c22a7e7f3dc5bef949fe856a4bd7538915b492b09eff02cde408 is worth 121 of the other blocks combined
830 2013-03-09 11:53:24 <Luke-Jr> "difficulty" : 121.01183650,
831 2013-03-09 11:54:14 <ciphermonk> ah
832 2013-03-09 11:54:51 <Luke-Jr> so it sees that, vs a pile of up to 121 "difficulty 1" blocks, it throws the diff1 blocks away and prefers the diff121 block
833 2013-03-09 11:56:49 <ciphermonk> which part of the code deals with this? I thought the rule "longest chain wins" was pretty immutable. Are the difficulty 1 blocks invalid in any way?
834 2013-03-09 11:57:15 <Luke-Jr> a difficulty 121 block is "longer" than 121 difficulty 1 blocks
835 2013-03-09 11:57:23 <Luke-Jr> no, they're not invalid
836 2013-03-09 11:57:27 <Luke-Jr> (on testnet, that is)
837 2013-03-09 11:57:38 <ciphermonk> oh length is defined as the difficulty!
838 2013-03-09 11:57:56 <ciphermonk> euh sorry
839 2013-03-09 11:58:10 <ciphermonk> ok
840 2013-03-09 11:58:24 <Luke-Jr> testnet has a special rule to allow diff1 blocks every 20 minutes
841 2013-03-09 11:58:35 <Luke-Jr> on real bitcoin, the opportunity for something like this is very small
842 2013-03-09 11:58:49 <Luke-Jr> only once every 2 weeks, during a major network upgrade
843 2013-03-09 11:58:54 <Luke-Jr> like the one for ASICs possibly
844 2013-03-09 11:59:00 <Luke-Jr> (which has only barely just started)
845 2013-03-09 11:59:10 <ciphermonk> ok fair enough
846 2013-03-09 11:59:38 <ciphermonk> The only impact I see is miners having some coinbases invalidated
847 2013-03-09 11:59:57 <Luke-Jr> pretty much
848 2013-03-09 12:00:11 <ciphermonk> cool, learned something new
849 2013-03-09 12:00:16 <Luke-Jr> note that the "height stays the same, but the top block changed" scenario is (I think?) more likely on mainnet
850 2013-03-09 12:00:34 <Luke-Jr> but height moving backward is the "only during major network upgrades" unliklihood
851 2013-03-09 12:00:48 <Luke-Jr> gmaxwell: hmm, you here still?
852 2013-03-09 12:01:15 <Luke-Jr> gmaxwell: is it possible that timejacking-related "bug" was intentional to avoid "height stays the same" reorgs? :/
853 2013-03-09 12:07:28 <gmaxwell> Luke-Jr: you can eliminate it without creating those.
854 2013-03-09 12:07:54 <Luke-Jr> gmaxwell: right, but does the current bug also avoid those?
855 2013-03-09 12:08:01 <Luke-Jr> (I added a note to the Hardfork Wishlist on this matter)
856 2013-03-09 12:08:35 <gmaxwell> timejacking bug is created by a lack of overlap, if you add an overlap on the more recent side you can create the issue you're thinking of, I think.
857 2013-03-09 12:22:07 <ciphermonk> is a reorg of, say, 6 block or more realistically possible on prodnet? (going back 6 blocks)
858 2013-03-09 12:22:47 <grau> blockchain.info now lags by about 10 blocks.
859 2013-03-09 12:23:51 <grau> ... just that I said. it is catching up
860 2013-03-09 12:26:59 <_dr> do you guys plan to enter gsoc?
861 2013-03-09 12:30:11 <MrKain> Does anyone know of a service that can deposit funds into someones bank account., and that it be available instantly -
862 2013-03-09 12:34:04 <ciphermonk> bitcoin? :D