1 2011-11-03 00:00:17 <luke-jr> a week ago, it was at "device driver work and final software refinement"
  2 2011-11-03 00:00:20 <midnightmagic> real helpful. no details about it at all. perhaps these are the people art sold his fpga design to.
  3 2011-11-03 00:00:31 <luke-jr> midnightmagic: oh, he did?
  4 2011-11-03 00:00:40 <sipa> art's fpga did 200MH/s
  5 2011-11-03 00:00:47 <gmaxwell> You'd think butterfly would be super eager to prove themselves, since there are a lot of people who'd place order if they believed it.
  6 2011-11-03 00:00:52 <luke-jr> sipa: presumably they'd have improved on it since then :p
  7 2011-11-03 00:01:03 <luke-jr> gmaxwell: I think I'm a beta tester for them ;)
  8 2011-11-03 00:01:10 <luke-jr> gmaxwell: they didn't want to be public yet
  9 2011-11-03 00:01:20 <midnightmagic> luke-jr: so I heard.
 10 2011-11-03 00:01:45 <midnightmagic> gmaxwell: I'm starting to think mining-specific hardware is a bad idea.
 11 2011-11-03 00:02:17 <sipa> why?
 12 2011-11-03 00:02:19 <gmaxwell> midnightmagic: it's not preventable if bitcoin becomes popular (hell, if it just stays as it is!)
 13 2011-11-03 00:03:26 <gmaxwell> plus attackers could be counted on to use it eventually (some potential attackers have access to fabs, e.g. any state level attacker people at some universities with VLSI departments so without it bitcoin would always be very vulnerable to those attackers)
 14 2011-11-03 00:03:29 <midnightmagic> it would suck if I sunk a whole pile of money into some specific hardware that could only do btc-specific sha256 hashing, and suddenly mrb_'s weird double-speed algorithm went out to the public. now everyone else is almost as efficient.
 15 2011-11-03 00:04:41 <gmaxwell> well, on true mining specific hardware you'd have more than a 2x gain. (heck on a work per watt hour, the _FPGAs_ are 7x more efficient than gpus, and they're power pigs!)
 16 2011-11-03 00:04:48 <midnightmagic> or, with this, if only signed firmwares could be uploaded, now butterfly decides they don't like what you're doing to their hardware, and they refuse to release an updated firmware. or the chip they used is slightly too small to efficiently mine with whatever new mining algo goes into the bitcoind core.
 17 2011-11-03 00:05:45 <gmaxwell> midnightmagic: same can be true with anything. if you're going to assume changing the algo (Why?! how?! I call bs!) then any hardware could be written out of it.
 18 2011-11-03 00:05:53 <gmaxwell> (at least written out in terms of efficient execution)
 19 2011-11-03 00:06:15 <midnightmagic> say a sha256 collision is found by someone. now what?
 20 2011-11-03 00:06:23 <gmaxwell> so?
 21 2011-11-03 00:06:31 <genjix> life goes on
 22 2011-11-03 00:06:47 <midnightmagic> the point is, with fpga or other general-purpose, there is .. another purpose.
 23 2011-11-03 00:06:58 <gmaxwell> absolutely, thats the risk you take.
 24 2011-11-03 00:07:03 <gmaxwell> Factor that into the price.
 25 2011-11-03 00:07:20 <gmaxwell> Fortunately a real dedicated piece of hardware could absolutely smoke the more general stuff.
 26 2011-11-03 00:07:37 <midnightmagic> exactly. so mining-specific better be cheap, and I don't see $500-per-GH as cheap with the 7xxx series on the horizon.
 27 2011-11-03 00:08:00 <gmaxwell> midnightmagic: wrong metric.
 28 2011-11-03 00:08:07 <luke-jr> it's per-watt that matters
 29 2011-11-03 00:08:12 <gmaxwell> $500 per gh is not cheap. It's the $ per watt which matters.
 30 2011-11-03 00:08:22 <luke-jr> if everyone's barely breaking even with GPUs, I'll be able to make a nice profit ;)
 31 2011-11-03 00:08:26 <gmaxwell> Thats just cheap enough to make it a complete no brainer.
 32 2011-11-03 00:08:37 <luke-jr> by the time I don't make a profit, GPUs will have died out like CPUs
 33 2011-11-03 00:09:06 <luke-jr> I can mine down to like $0.10/BTC with this
 34 2011-11-03 00:09:25 <genjix> why do they sell them?
 35 2011-11-03 00:09:50 <gmaxwell> midnightmagic: see the figures here: https://bitcointalk.org/index.php?topic=49180.msg587822#msg587822
 36 2011-11-03 00:10:11 <lfm> depends if you have free power too
 37 2011-11-03 00:10:18 <gmaxwell> genjix: hunters vs farmers also, why did the guys with mining hardware sell it during the gold rush  (they made more money than the miners!)
 38 2011-11-03 00:10:38 <genjix> haha ok
 39 2011-11-03 00:10:38 <gmaxwell> lfm: ~almost one has non-trivial amounts of free power.
 40 2011-11-03 00:10:58 <luke-jr> I expect they're planning to sell a bunch at $600 when they know it works
 41 2011-11-03 00:11:11 <luke-jr> and sit back as the difficulty soars
 42 2011-11-03 00:11:18 <lfm> gmaxwell: ya but lots of people have trivial amounts of free power
 43 2011-11-03 00:11:23 <luke-jr> I'm hoping to break even by being an early adopter
 44 2011-11-03 00:11:31 <gmaxwell> lfm: its true.
 45 2011-11-03 00:11:55 <luke-jr> yay for State-subsidized solar panels!
 46 2011-11-03 00:12:19 <gmaxwell> luke-jr: I don't think that board gives you the kind of launch youd need for that. But at least you won't be left cold when a 15gh/s asic for $100 comes out.
 47 2011-11-03 00:13:13 <midnightmagic> 60 days per block, or 0.835 btc per day, at current prices, is $2 per day per unit, assume free electricity and no change in difficulty. 186 days to pay itself off.. and after that, profit hopefully..
 48 2011-11-03 00:13:30 <luke-jr> ;;bc,gen 1000000
 49 2011-11-03 00:13:32 <gribble> The expected generation output, at 1000000 Khps, given current difficulty of 1203461.92638 , is 0.835779145594 BTC per day and 0.0348241310664 BTC per hour.
 50 2011-11-03 00:13:32 <lfm> gmaxwell: there will always be something better coming out real soon now
 51 2011-11-03 00:14:05 <gmaxwell> lfm: well, there is diminishing better once you're unable to improve faster than the underlying silicon process.
 52 2011-11-03 00:14:11 <luke-jr> gmaxwell: well, it works for me with GPU
 53 2011-11-03 00:14:29 <midnightmagic> gmaxwell: that's figures for fpga and gpu. I like fpga, I think that's going to be a fun time when i pick some up in.. I guess another month.
 54 2011-11-03 00:14:35 <luke-jr> my Radeon 5850 paid for itself in under a month
 55 2011-11-03 00:14:55 <gmaxwell> midnightmagic: I was just pointing to it for a point about why power is so important.
 56 2011-11-03 00:15:03 <midnightmagic> all my 5970 paid for themselves in under a month too. that doesn't mean they could do that now.
 57 2011-11-03 00:15:08 <midnightmagic> gmaxwell: ah, yes, I agree.
 58 2011-11-03 00:15:19 <luke-jr> midnightmagic: these things are to GPUs, what GPUs were to CPUs back then
 59 2011-11-03 00:15:43 <lfm> GPUs wont stand still either
 60 2011-11-03 00:15:47 <midnightmagic> only if you buy them in equivalent volume. otherwise it's just a few pennies versus a few more pennies.
 61 2011-11-03 00:15:57 <gmaxwell> There isn't an installed base of these that can be picked up suddenly to crank the difficulty. :)
 62 2011-11-03 00:16:27 <gmaxwell> lfm: yes, but they're always going to be handicapped by being gpus rather than being miners.. lots of useless gates on that card.
 63 2011-11-03 00:16:38 <luke-jr> lfm: GPUs are going backward
 64 2011-11-03 00:16:56 <luke-jr> 6xxx are less efficient than 5xxx for mining
 65 2011-11-03 00:16:58 <luke-jr> etc
 66 2011-11-03 00:17:07 <gmaxwell> of course, they're flexible, but where is the bitcoin miner program that can run password cracking kernels when its more profitable to do so? people aren't bothering to make use of that flexibility.
 67 2011-11-03 00:18:44 <midnightmagic> yeah, flexibility is under-utilized.
 68 2011-11-03 00:19:33 <midnightmagic> anyway, same old argument, same old impossible-to-knows without seeing the future. and nobody can see the future but the people who can manipulate the prices or supply of hardware.
 69 2011-11-03 00:19:54 <midnightmagic> or the people who keep special high-performance algos to themselves..
 70 2011-11-03 00:20:15 <gmaxwell> I suggested to ztex (maker of the nicest FPGA board) that he should offer code for e.g. des cracking (which is also quite nice on FPGAs) so that people would feel more comfortable buying them knowing they can be reused if bitcoin goes bust, but he was concerned about them being used for evil. "But I will not implement ready-to-use solutions for this due to moral reasons (could help governments to tyrannize their people)"
 71 2011-11-03 00:20:35 <midnightmagic> yikes
 72 2011-11-03 00:20:48 <midnightmagic> he doesn't sign or encrypt his firmwares does he?
 73 2011-11-03 00:21:08 <midnightmagic> all you need is to feed the onboard microprocessor the new firmware?
 74 2011-11-03 00:21:32 <gmaxwell> well, the xilinx images are always encrypted though the encryption has been cracked. I dunno if he distributes his own sources, but all his products are intended for people to load their own fpga code.
 75 2011-11-03 00:23:03 <gmaxwell> I wonder if he's crazy (After all, who can afford enough FPGAs to cause trouble but can't afford to write code for them??) or if he's just really good at pandering to his audience (wouldn't sound good on bitcointalk to say you don't want to give kiddies cracking code)
 76 2011-11-03 00:23:51 <lfm> gmaxwell: some people prefer to buy open solutions
 77 2011-11-03 00:24:24 <midnightmagic> whoah: http://www.schneier.com/blog/archives/2011/08/breaking_the_xi.html
 78 2011-11-03 00:25:08 <midnightmagic> but regardless of the xilinx images that are encrypted, you can just load your own stuff on his hardware trivially right?
 79 2011-11-03 00:25:14 <gmaxwell> lfm: open has degrees. :)
 80 2011-11-03 00:25:30 <gmaxwell> midnightmagic: sure, if you have the xilinx software.
 81 2011-11-03 00:26:06 <gmaxwell> midnightmagic: his business is much bigger than selling to bitcoin miners.
 82 2011-11-03 00:26:29 <gmaxwell> http://www.ztex.de/
 83 2011-11-03 00:28:42 <CIA-34> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * r6601537 / (4 files in 3 dirs): Split Transaction class and TransactionSchema into separate files. - http://git.io/nxe_EA
 84 2011-11-03 00:28:43 <CIA-34> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * raf79dc9 / lib/util.js : Added some more null constants. - http://git.io/fseEUw
 85 2011-11-03 00:34:19 <midnightmagic> gmaxwell: thanks for the pointers again, sometimes it's hard to keep up
 86 2011-11-03 00:34:44 <gmaxwell> If you're looking at fpgas, also: https://bitcointalk.org/index.php?topic=40058.0
 87 2011-11-03 00:35:33 <gmaxwell> it's slightly vaporware at the moment, but the prior version was completely legit so I'm sure this one will be too, but for mining it will be a bit less costly than ztex's very nice solution.
 88 2011-11-03 00:36:59 <midnightmagic> neat stuff..
 89 2011-11-03 00:38:06 <Diablo-D3> meh
 90 2011-11-03 00:38:08 <Diablo-D3> not that again
 91 2011-11-03 00:38:45 <gmaxwell> midnightmagic: also this thread: https://bitcointalk.org/index.php?topic=44891.0
 92 2011-11-03 00:38:46 <midnightmagic> what?
 93 2011-11-03 00:55:44 <imsaguy> and then there is this: https://bitcointalk.org/index.php?topic=50710.0
 94 2011-11-03 01:30:09 <CIA-34> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * r2f3542b / (lib/binary.js lib/connection.js lib/schema/transaction.js): Renamed Put#var_uint -> Put#varint. (+5 more commits...) - http://git.io/JUWs1g
 95 2011-11-03 03:53:54 <skittixch> Hey, can anyone clarify if mtgox charges fees on both buys and sells?
 96 2011-11-03 04:01:44 <cjdelisle> AFAIK the answer is yes.
 97 2011-11-03 04:02:03 <skittixch> cool, that's what I was thinking
 98 2011-11-03 04:02:14 <skittixch> I think if your transaction is very small, it doesn't take a fee though
 99 2011-11-03 04:02:37 <skittixch> I had a few that were a few cents that rounded off some previous transactions, and they didn't have fees
100 2011-11-03 04:26:20 <copumpkin> can someone please get rid of zhoutong?
101 2011-11-03 04:26:28 <copumpkin> this entire channel log is him joining and parting
102 2011-11-03 04:26:33 <copumpkin> and has been for weeks
103 2011-11-03 04:26:34 <copumpkin> :P
104 2011-11-03 04:26:45 <copumpkin> zhoutong: nice connection
105 2011-11-03 04:38:33 <phantomcircuit> nanotube, UukGoblin nameless| nathan7 please ban zhoutong he join/parts constantly and never talks ever
106 2011-11-03 04:38:38 <phantomcircuit> it's annoying as fuck
107 2011-11-03 06:10:48 <CIA-34> bitcoin: CFSworks * rde17288d5999 Phoenix-Miner/minerutil/RPCProtocol.py: Adjust HTTP timeouts
108 2011-11-03 06:10:49 <CIA-34> bitcoin: CFSworks * rc8db0c300f39 Phoenix-Miner/minerutil/RPCProtocol.py: Use TCP_NODELAY
109 2011-11-03 06:13:08 <CIA-34> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * r0df68ec / lib/schema/transaction.js : Optimization: Direct hashForSignature(), no longer copies tx first. - http://git.io/oQ95sw
110 2011-11-03 06:13:09 <CIA-34> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * r2520a8d / (3 files in 3 dirs): Optimization: Don't instantiate database objects until they're needed. - http://git.io/WzyRhg
111 2011-11-03 06:30:41 <CIA-34> bitcoin: CFSworks * r11f4954675f7 Phoenix-Miner/ (Miner.py minerutil/RPCProtocol.py): Have the connection close upon Phoenix shutdown...
112 2011-11-03 06:30:42 <CIA-34> bitcoin: CFSworks * r085b630f641c Phoenix-Miner/ (16 files in 5 dirs): Remove spaces from empty lines
113 2011-11-03 06:37:05 <CIA-34> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * rdf8c7a4 / (4 files in 2 dirs): Optimization: Added improved receive buffer. - http://git.io/1AVTRQ
114 2011-11-03 07:05:58 <nameless> |phantomcircuit: ignore it?
115 2011-11-03 07:06:19 <phantomcircuit> nameless|, he's filling up my freaking logs
116 2011-11-03 07:06:46 <nameless> |phantomcircuit: /ignore it?
117 2011-11-03 07:06:59 <phantomcircuit> using bouncer
118 2011-11-03 07:07:01 <nameless> |in irssi I can /ignore zhoutong JOIN PART QUIT CRAP
119 2011-11-03 07:07:05 <phantomcircuit> would still fill my logs
120 2011-11-03 07:07:43 <nameless> |shouldn't
121 2011-11-03 07:07:47 <nameless> |it doesn't in irssi
122 2011-11-03 07:07:55 <phantomcircuit> bouncer doesn't work like that
123 2011-11-03 07:08:00 <phantomcircuit> bouncer stores the logs
124 2011-11-03 07:08:08 <phantomcircuit> i'd have to setup ignore on the bouncer
125 2011-11-03 07:12:34 <cjdelisle> -!- zhoutong [~zhoutong@111.221.80.132] has quit [Read error: Connection reset by peer]
126 2011-11-03 07:12:38 <cjdelisle> -!- zhoutong [~zhoutong@111.221.80.132] has joined #bitcoin-dev
127 2011-11-03 07:16:53 <gjs278> # grep zhoutong *|wc -l
128 2011-11-03 07:16:54 <gjs278> 11342
129 2011-11-03 07:16:56 <gjs278> holy fuck
130 2011-11-03 07:20:35 <invisiblemonk> um...
131 2011-11-03 07:21:57 <gjs278> it doesn't work
132 2011-11-03 07:22:01 <gjs278> he penetrates my client ignore
133 2011-11-03 07:22:14 <gjs278> I also don't think he has said anything
134 2011-11-03 07:22:16 <gjs278> ever
135 2011-11-03 07:47:32 <erus`> just turn conference mode on in your client nameless|
136 2011-11-03 07:47:39 <erus`> thn you dont see joins and parts
137 2011-11-03 08:30:01 <lfm> [6~[6~[6~[6~
138 2011-11-03 08:57:03 <lfm> [5~
139 2011-11-03 09:05:46 <CIA-34> bitcoin: CFSworks * re22aadea851e Phoenix-Miner/minerutil/RPCProtocol.py: Lock the HTTP thread
140 2011-11-03 09:15:41 <CIA-34> bitcoin: jedi95 * r1385f7d35667 Phoenix-Miner/.gitignore: Added .elf to .gitignore
141 2011-11-03 09:31:29 <Diablo-D3> man
142 2011-11-03 09:31:34 <Diablo-D3> I love how people join
143 2011-11-03 09:31:36 <Diablo-D3> msg me
144 2011-11-03 09:33:49 <batouzo> Diablo-D3: hey bro I have 50 BTC donation if you want one, but we have to decide on it in next 10 minutes or so, please
145 2011-11-03 09:34:28 <MartianW> Diablo-D3, impatient.
146 2011-11-03 09:36:17 <Diablo-D3> ...
147 2011-11-03 09:36:20 <Diablo-D3> SEE WHAT I MEAN
148 2011-11-03 10:10:38 <CIA-34> bitcoin: CFSworks * rfe36698ca6d0 Phoenix-Miner/minerutil/RPCProtocol.py: Remove typo yield
149 2011-11-03 10:10:44 <CIA-34> bitcoin: CFSworks * r3be3e9935775 Phoenix-Miner/minerutil/RPCProtocol.py: Remove extra space
150 2011-11-03 10:10:47 <CIA-34> bitcoin: CFSworks * r8d89717715ed Phoenix-Miner/minerutil/RPCProtocol.py: Support https connections
151 2011-11-03 10:19:50 <UukGoblin> interesting
152 2011-11-03 10:19:58 <UukGoblin> every time I try to get a ticker from mtgox
153 2011-11-03 10:20:00 <UukGoblin> at first I get
154 2011-11-03 10:20:08 <UukGoblin> curl: (35) Unknown SSL protocol error in connection to mtgox.com:443
155 2011-11-03 10:20:16 <UukGoblin> and then the next attempt shortly after works fine
156 2011-11-03 10:20:40 <UukGoblin> guess it's some kind of smart gray-listing for ddos protection
157 2011-11-03 10:25:50 <CIA-34> bitcoin: CFSworks * rbf3ebecdb890 Phoenix-Miner/minerutil/RPCProtocol.py: RPCProtocol should catch timeouts too
158 2011-11-03 10:45:40 <CIA-34> bitcoin: CFSworks * ra949a20a6ce4 Phoenix-Miner/phoenix.py: Make phoenix.py executable by default
159 2011-11-03 11:08:19 <UukGoblin> ;;help bc,diff
160 2011-11-03 11:08:20 <gribble> (bc,diff <an alias, 0 arguments>) -- Alias for "web fetch http://blockexplorer.com/q/getdifficulty".
161 2011-11-03 11:09:09 <lfm> kinda roundabout way of doing it it seems
162 2011-11-03 11:11:10 <UukGoblin> and why the hell doesn't google calculator API return valid json
163 2011-11-03 11:12:00 <lfm> hehe, how would they be able to feed you the adds then?
164 2011-11-03 11:14:36 <invisiblemonk> fuck aws.
165 2011-11-03 11:15:20 <lfm> invisiblemonk: <- I see you (whats aws?)
166 2011-11-03 11:20:41 <SomeoneWeird> Anyone here used BitcoinSharp?
167 2011-11-03 11:21:42 <lfm> I havnt, what is i
168 2011-11-03 11:21:45 <lfm> it?
169 2011-11-03 11:21:53 <SomeoneWeird> C# bitcoin library
170 2011-11-03 11:22:16 <lfm> oh just a lib, not a complete app
171 2011-11-03 11:22:24 <SomeoneWeird> mhm
172 2011-11-03 11:22:28 <kinlo> a library should be enough :)
173 2011-11-03 11:22:37 <SomeoneWeird> can't figure out how to get it working though haha
174 2011-11-03 11:22:40 <kinlo> but why do people use .net
175 2011-11-03 11:22:50 <kinlo> that's for me the confusing part :)
176 2011-11-03 11:22:50 <lfm> well some libs are not exactly complete
177 2011-11-03 11:25:19 <SomeoneWeird> its pretty complete if you look at it lfm
178 2011-11-03 11:25:29 <SomeoneWeird> i do kinlo :P
179 2011-11-03 11:25:38 <CIA-34> bitcoin: CFSworks * r977f08547d91 Phoenix-Miner/minerutil/RPCProtocol.py: Intercept HTTP requests to close them directly as well.
180 2011-11-03 11:25:40 <CIA-34> bitcoin: CFSworks * r429e0f145d34 Phoenix-Miner/minerutil/RPCProtocol.py: Shut down socket when closing connections.
181 2011-11-03 11:25:42 <lfm> whats kinlo?
182 2011-11-03 11:25:52 <kinlo> I think he was just replying to me
183 2011-11-03 11:25:59 <kinlo> :)
184 2011-11-03 11:26:07 <lfm> oh ok
185 2011-11-03 11:26:38 <kinlo> perhaps a comma would have been in place, you might even read that the wrong way :p
186 2011-11-03 11:26:55 <lfm> not much point me looking at it tho cuz I have not inclination to learn C# or .net
187 2011-11-03 11:27:28 <kinlo> well, imho it would be better to write a wrapper and to extend bitcoins rpc functionality then just dumb-rewriting the code
188 2011-11-03 11:27:44 <kinlo> it is already in a far superiour language, why rewrite it ? :)
189 2011-11-03 11:28:42 <lfm> my reasons for not wanting to learn C# have noting to do with its suitability for any particular purpose, I just dont trust microssoft
190 2011-11-03 11:29:28 <SomeoneWeird> i do mono stuff too
191 2011-11-03 11:29:41 <kinlo> my reasons is because microsoft has so many customers, so many users, but no measurable community
192 2011-11-03 11:29:48 <SomeoneWeird> kinlo, bitcoinsharp has alot of stuff, wallet reading/writing, blockchain searching etc.etc.
193 2011-11-03 11:30:07 <kinlo> SomeoneWeird: stuff which should be done via rpc calls imho
194 2011-11-03 11:30:17 <batouzo> why not use C++
195 2011-11-03 11:30:25 <SomeoneWeird> well it can be full on stand alone kinlo
196 2011-11-03 11:30:29 <SomeoneWeird> not have to rely on the client
197 2011-11-03 11:30:51 <SomeoneWeird> batouzo, because i dont know C++
198 2011-11-03 11:30:59 <SomeoneWeird> dont plan on learning it either.
199 2011-11-03 11:31:47 <kinlo> why not?:)
200 2011-11-03 11:31:52 <kinlo> c++ is a cool language
201 2011-11-03 11:31:57 <kinlo> very difficult, but cool
202 2011-11-03 11:32:20 <cjdelisle> The world is written in C++
203 2011-11-03 11:32:20 <SomeoneWeird> my point exactly
204 2011-11-03 11:32:24 <SomeoneWeird> too busy at the moment
205 2011-11-03 11:32:28 <SomeoneWeird> -.- cjdelisle
206 2011-11-03 11:32:43 <batouzo> C# is like C++ but microsoftied
207 2011-11-03 11:32:51 <lfm> I dont think c++ would be primarily charcterized as "cool".
208 2011-11-03 11:32:55 <batouzo> and "easier". Such as visual basic is "easy"
209 2011-11-03 11:33:03 <SomeoneWeird> visual basic sucks lol
210 2011-11-03 11:33:05 <Ycros> batouzo: not at all, C# is more like java
211 2011-11-03 11:33:10 <SomeoneWeird> ^
212 2011-11-03 11:33:14 <Ycros> batouzo: managed C++ is microsoftied C++
213 2011-11-03 11:33:24 <batouzo> yeah something in between
214 2011-11-03 11:33:42 <batouzo> my friends learns visual basic 6 at uni.
215 2011-11-03 11:33:52 <lfm> well ya C# is ms's anser to java for people who dont want to learn java
216 2011-11-03 11:33:55 <batouzo> I think she was trolled by her uni.
217 2011-11-03 11:34:05 <cjdelisle> ouch
218 2011-11-03 11:34:22 <SomeoneWeird> damn
219 2011-11-03 11:34:24 <batouzo> vb6 was deprecated old shit 10 years ago
220 2011-11-03 11:34:24 <cjdelisle> c# is vbasic with squigley brackets
221 2011-11-03 11:34:41 <SomeoneWeird> ok guys stop the language war.
222 2011-11-03 11:34:43 <SomeoneWeird> -.-
223 2011-11-03 11:34:51 <batouzo> I think [Turbo] Pascal even is better
224 2011-11-03 11:35:14 <lfm> batouzo: well duh! Pascal is perfect.
225 2011-11-03 11:35:23 <batouzo> :}
226 2011-11-03 11:35:24 <SomeoneWeird> lawl
227 2011-11-03 11:37:41 <kinlo> pascal is indeed a good language
228 2011-11-03 11:37:59 <kinlo> and anyone comparing c# with C++ has no clue, they are fundamentally different
229 2011-11-03 11:38:27 <cjdelisle> c# is from the java strand of trying to save the programmer from himself
230 2011-11-03 11:38:32 <lfm> well there wouldnt be much point in comparing them if they were identical.
231 2011-11-03 11:38:51 <cjdelisle> but the folly in sun's logic was to assume that they wouldn't make a better idiot
232 2011-11-03 11:39:39 <SomeoneWeird> im not comparing them kinlo
233 2011-11-03 11:39:54 <kinlo> you werent indeed
234 2011-11-03 11:40:03 <da2ce7> cjdelisle, C# and VB are very similar, just differnt sintax... However new features get added to C# first.
235 2011-11-03 11:40:07 <batouzo> kinlo: I said it was microsoftized.  Same as windows is fundamentally different to real OS
236 2011-11-03 11:40:11 <da2ce7> *syntax
237 2011-11-03 11:40:12 <SomeoneWeird> mhm
238 2011-11-03 11:40:32 <batouzo> wait, we should be working
239 2011-11-03 11:40:39 <kinlo> anyway
240 2011-11-03 11:40:41 <CIA-34> bitcoin: CFSworks * r216323cfff48 Phoenix-Miner/minerutil/RPCProtocol.py: Shouldn't try to shutdown the RPC socket if it doesn't exist
241 2011-11-03 11:40:49 <kinlo> anyone ever worked with google go?
242 2011-11-03 11:40:59 <lfm> da2ce7: no I think you were right the first time, C# has the ms sintax
243 2011-11-03 11:41:34 <da2ce7> lfm, ofcoures, it was made by MS.
244 2011-11-03 11:41:43 <SomeoneWeird> mhm
245 2011-11-03 12:40:40 <CIA-34> bitcoin: jedi95 * r9fb667c07c7f Phoenix-Miner/kernels/phatk2/__init__.py: Removed time.clock import from phatk2 (not used)
246 2011-11-03 12:40:41 <CIA-34> bitcoin: jedi95 * re04cdb19305b Phoenix-Miner/minerutil/RPCProtocol.py: Merge branch 'master' of github.com:jedi95/Phoenix-Miner
247 2011-11-03 13:00:43 <CIA-34> bitcoin: jedi95 * r83ec5e1bef87 Phoenix-Miner/kernels/ (6 files in 3 dirs): Remove ATI-only restriction on phatk/phatk2 as these no longer give errors on Nvidia hardware.
248 2011-11-03 13:10:24 <graingert> :D
249 2011-11-03 13:20:46 <CIA-34> libbitcoin: genjix * r38d18d646c0b / (Makefile examples/subvertx/mktx.cpp): Accept [HOST[:PORT]] for mktx.
250 2011-11-03 13:49:00 <CIA-34> poolserverj: shadders * f11b17698972 r183 / (15 files in 8 dirs): Some optimisazion, removal of debug code and tidy of workmaker threading to take advantage of multiple cores. Improved work output by about 300%.
251 2011-11-03 13:50:48 <CIA-34> bitcoin: CFSworks * ra4475dde8752 Phoenix-Miner/minerutil/RPCProtocol.py: Correctly close the RPC HTTP connection socket
252 2011-11-03 14:10:02 <CIA-34> bitcoin: Gavin Andresen master * r434a483 / (doc/build-msw.txt doc/release-process.txt): Update documentation to reflect 0.5 reality. And removed leading $ from shell commands, so it is easier to copy and paste from release-process.txt. - http://git.io/mr2XbA
253 2011-11-03 14:10:03 <CIA-34> bitcoin: Mac Deployment Script
254 2011-11-03 14:10:03 <CIA-34> bitcoin: p2k master * r6eaa1b3 / (6 files in 2 dirs):
255 2011-11-03 15:27:59 <d33tah> hi guys
256 2011-11-03 15:28:09 <d33tah> i've got an evil idea i'd like to discuss with you
257 2011-11-03 15:28:34 <d33tah> first, is it true that bitcoin miners calculate pure SHA hashes all the time?
258 2011-11-03 15:28:45 <da2ce7> d33tah
259 2011-11-03 15:28:51 <da2ce7> welcome, what do you want to decuss?
260 2011-11-03 15:29:10 <da2ce7> d33tah, they calcualte doubble sha256
261 2011-11-03 15:29:12 <d33tah> well, i thought of an idea that could make mining more profitable
262 2011-11-03 15:29:18 <luke-jr> d33tah: no
263 2011-11-03 15:29:38 <d33tah> so, these are hashes of hashes...
264 2011-11-03 15:29:49 <luke-jr> they do a double-sha256, and only of a specific fixed-length input
265 2011-11-03 15:29:51 <cocktopus> its something liek sha256(sha256()) right
266 2011-11-03 15:30:09 <da2ce7> *well technicaly it isn't sha256, as the miners can get away with skiping a few rounds.
267 2011-11-03 15:30:12 <d33tah> but the first run of the hasing, is it exactly the same as php's sha256?
268 2011-11-03 15:30:46 <roconnor> skiping a few rounds?
269 2011-11-03 15:30:49 <d33tah> could the data from the first hashing be used as legit sha256?
270 2011-11-03 15:31:03 <da2ce7> sure, the last few rounds of the 2nd hash is not needed.
271 2011-11-03 15:31:15 <da2ce7> as it only effects the higher numbers.
272 2011-11-03 15:31:53 <luke-jr> da2ce7: no, you can't skip anymore
273 2011-11-03 15:32:07 <roconnor> da2ce7: and those higher numbers correspond to the lower bits when comparing with the target.
274 2011-11-03 15:32:17 <d33tah> does anybody know the answer to this question? are the hashes calculated legit sha256 hashes?
275 2011-11-03 15:32:20 <da2ce7> luke-jr, why not?
276 2011-11-03 15:32:26 <luke-jr> d33tah: no, they're double
277 2011-11-03 15:32:26 <roconnor> da2ce7: interesting
278 2011-11-03 15:32:39 <d33tah> luke-jr: let's say we interrupt after the first pass. would it?
279 2011-11-03 15:32:41 <luke-jr> da2ce7: as roconnor said, the bits you need to be 0 are last, and midstate is not part of work anymore
280 2011-11-03 15:32:46 <luke-jr> d33tah: yes
281 2011-11-03 15:32:49 <d33tah> okay
282 2011-11-03 15:33:05 <da2ce7> ok
283 2011-11-03 15:33:25 <roconnor> luke-jr: midstate?
284 2011-11-03 15:33:43 <luke-jr> roconnor: midstate = first part of first round of sha256 which is the same regardless of nonce
285 2011-11-03 15:34:20 <roconnor> luke-jr: has the definition of work changed?
286 2011-11-03 15:34:28 <d33tah> so, now suppose we introduce a new kind of transaction - one side withdraws some amount of its money as "bounty" assigned to a specific hash, the other side checks his first passes everytime if it's not something we're looking for - if it is, we announce to the network what data collides with the hash and then claim the bounty. if hash isn't cracked in the specific time, the money returns to its original owner
287 2011-11-03 15:34:34 <d33tah> is it possible?
288 2011-11-03 15:34:35 <luke-jr> roconnor: getwork JSON-RPC used to provide midstate
289 2011-11-03 15:34:40 <luke-jr> roconnor: now you need to do that yourself in the miner
290 2011-11-03 15:35:17 <luke-jr> d33tah: no, there's no support for time-based transactions
291 2011-11-03 15:35:26 <luke-jr> d33tah: furthermore, you would "never" find such a collision
292 2011-11-03 15:35:39 <sipa> d33tah: and a miner who sees such a claim transaction will just change it to one that sends it to himself
293 2011-11-03 15:36:07 <d33tah> sipa: right, anything wrong in that?
294 2011-11-03 15:36:22 <sipa> eh
295 2011-11-03 15:36:27 <d33tah> i mean, not the one that sees - the one that receives
296 2011-11-03 15:36:27 <sipa> it makes it worthless
297 2011-11-03 15:36:35 <d33tah> it'd be like "first to claim"
298 2011-11-03 15:36:42 <luke-jr> d33tah: there is no first to claim
299 2011-11-03 15:36:53 <d33tah> why not?
300 2011-11-03 15:36:55 <sipa> i managed to find a collision, i create a transaction that claims the bounty, and sends it to me
301 2011-11-03 15:37:00 <sipa> i broadcast that transaction
302 2011-11-03 15:37:15 <graingert> once it's in a block
303 2011-11-03 15:37:16 <sipa> smart miner sees the transaction, changes it to payment to himself, and puts it in a block
304 2011-11-03 15:37:17 <graingert> release the hash
305 2011-11-03 15:37:24 <d33tah> oh fuck
306 2011-11-03 15:37:47 <d33tah> so it'd have to be secret between the donor and the receiver to work?
307 2011-11-03 15:37:49 <luke-jr> anyhow, if you could find collisions, Bitcoin wouldn't work
308 2011-11-03 15:37:54 <luke-jr> the point is that you *can't* find collisions
309 2011-11-03 15:38:00 <graingert> yeah so it doesn't help :p
310 2011-11-03 15:38:16 <d33tah> so it's basically impossible to trade hashes this way?
311 2011-11-03 15:38:22 <graingert> the bounty would have to be huge
312 2011-11-03 15:38:23 <sipa> why would you want to?
313 2011-11-03 15:38:26 <graingert> to compete with bitcoin mining
314 2011-11-03 15:38:26 <luke-jr> d33tah: it's basically impossible to find hashes
315 2011-11-03 15:38:39 <graingert> more than galactic huge :p
316 2011-11-03 15:38:48 <d33tah> hm, why that huge?
317 2011-11-03 15:38:55 <d33tah> the miner code would have to be polymorphic
318 2011-11-03 15:39:00 <graingert> because it would be much mroe profitable to mine
319 2011-11-03 15:39:10 <sipa> it costs 2^128 SHA256 operations to find a collision, if no weakness is found
320 2011-11-03 15:39:32 <sipa> and 2^256 SHA256 operations to find a preimage, which is probably what you're aiming for
321 2011-11-03 15:39:50 <d33tah> if the bounty is over the limit, add the code checking if we have a collision to the main hashing routine
322 2011-11-03 15:40:07 <d33tah> hm, damn, i was so enthusiastic about it :P
323 2011-11-03 15:40:14 <sipa> i still don't see the purpose
324 2011-11-03 15:40:18 <d33tah> well
325 2011-11-03 15:40:18 <sipa> what problem are you trying to solve?
326 2011-11-03 15:40:32 <luke-jr> d33tah: you *can't* find a hash.
327 2011-11-03 15:40:46 <luke-jr> if you can, you've broken Bitcoin
328 2011-11-03 15:41:00 <graingert> d33tah: there is not enough energy in the universe to brute force it
329 2011-11-03 15:41:11 <d33tah> finding a preimage of a single hash would break the network?
330 2011-11-03 15:41:35 <sipa> no, but being able to do so cheaply would break the usefulness
331 2011-11-03 15:42:19 <d33tah> i just was thinking
332 2011-11-03 15:42:26 <graingert> d33tah: no you wern't
333 2011-11-03 15:42:56 <d33tah> if thousands of users keep mining milions of hashes per seconds for hours... the possibility should be within reach
334 2011-11-03 15:43:06 <sipa> no
335 2011-11-03 15:43:09 <d33tah> though i guess i should believe you guys it's not ;p
336 2011-11-03 15:43:18 <d33tah> graingert: it wasn't too nice, man.
337 2011-11-03 15:44:39 <graingert> It's a truefact
338 2011-11-03 15:45:05 <d33tah> the fact is that i was thinking indeed, just didn't have enough knowledge of the topic
339 2011-11-03 15:45:17 <graingert> the amount of energy it takes to increment a number 2^256 times
340 2011-11-03 15:46:01 <graingert> not counting the sha operation
341 2011-11-03 15:46:16 <graingert> is more than it is physically possible to obtain
342 2011-11-03 15:46:30 <d33tah> alright
343 2011-11-03 15:46:45 <d33tah> it's not what offended me, it's you implying i wasn't thinking
344 2011-11-03 15:47:03 <sipa> finding a weakness in ECDSA or SHA256 is more likely that it being brute-forced
345 2011-11-03 15:47:32 <graingert> compromising the algorithm is possible
346 2011-11-03 15:47:33 <sipa> and a bug in the software implementation is even much more likely than that, imho
347 2011-11-03 15:47:55 <d33tah> that's pretty much how I imagine the death of Bitcoin
348 2011-11-03 15:47:56 <graingert> there was a very odd bug in bitcoin
349 2011-11-03 15:48:04 <d33tah> a single bug would screw it all up, wouldn't it?
350 2011-11-03 15:48:15 <graingert> nope it's happened before
351 2011-11-03 15:48:26 <sipa> the integer overflow issue?
352 2011-11-03 15:48:27 <d33tah> i mean, a bug in the client, not the protocol
353 2011-11-03 15:48:35 <graingert> yep integer overflow
354 2011-11-03 15:49:02 <sipa> that's the only actual software bug in bitcoin itself with security consequences i know of
355 2011-11-03 15:50:10 <d33tah> though, if there was some bug allowing the offender to force the client to send any kinds of transactions, wouldn't it be the end of Bitcoin?
356 2011-11-03 15:50:32 <d33tah> what would be done? fast-rewind to the first uncompromised block? doesn't sound too rational
357 2011-11-03 15:51:10 <sipa> with an encrypted wallet, that would be impossible
358 2011-11-03 15:51:26 <graingert> an offline wallet it would be impossible
359 2011-11-03 15:51:31 <graingert> as the client isn't running
360 2011-11-03 15:51:34 <sipa> as the client doesn't have the keys necessary to create the transaction, unless the passphrase is given
361 2011-11-03 15:51:44 <d33tah> keyloggers? code injection?
362 2011-11-03 15:51:51 <sipa> yes, those are threats
363 2011-11-03 15:51:55 <graingert> yeah but that would only be a few
364 2011-11-03 15:51:57 <sipa> but beyond the control of bitcoin
365 2011-11-03 15:51:59 <graingert> not everyone all at once
366 2011-11-03 15:52:26 <d33tah> but in a pretty short time though
367 2011-11-03 15:58:34 <graingert> nope
368 2011-11-03 15:58:45 <nameless> |erus`: The ban wasn't for me. It was requested, and the last thing I want to do when waking up at 4am for work is hear IRC people bitch about join/parts
369 2011-11-03 15:58:52 <nameless> |I can /ignore it
370 2011-11-03 15:59:00 <nameless> |other' seem to have problems with that
371 2011-11-03 15:59:02 <BlueMatt> what does laanwj go by on here, again?
372 2011-11-03 16:00:22 <BlueMatt> also, does someone want to explain https://github.com/bitcoin/bitcoin/commit/fbea7eca656bf5797a2c5cf748715a1ae332835b#diff-1
373 2011-11-03 16:08:39 <sipa> BlueMatt: wumpus
374 2011-11-03 16:09:55 <BlueMatt> wumpus: ping
375 2011-11-03 16:20:05 <ciscoftw> "Set gen=1 to attempt to generate bitcoins" (bitcoin.conf) ...what purpose does this serve? seems like its not neccessary to set this flag...
376 2011-11-03 16:29:56 <invisiblemonk> god, aws being slow again.
377 2011-11-03 16:33:23 <nanotube> ciscoftw: not necessary, that's only if you want to use the built-in miner
378 2011-11-03 16:33:27 <nanotube> which you probably don't
379 2011-11-03 16:41:49 <ciscoftw> thanx for reply nontube: so if i had rpc clients connected to the bitcoind server, i would need that flag set?
380 2011-11-03 16:58:37 <sipa> ciscoftw: you don't need that flag, and i suggest to turn it off
381 2011-11-03 16:58:49 <sipa> unless for testing purposes, and not on mainnet
382 2011-11-03 17:36:08 <luke-jr> ciscoftw: where do you see that, even?
383 2011-11-03 18:06:29 <ciscoftw> luke-jr: within the bitcoin.conf file...
384 2011-11-03 18:07:38 <luke-jr> ciscoftw: what bitcoin.conf file?
385 2011-11-03 18:08:31 <ciscoftw> just seems like i cannot find any info regarding why i would/wouldnt want it.. still dl'ing the blockchain, so i cant test yet... but im thinking without it set gen=0 (the client will not attempt to mine) ...regardless how many clients are connected to it
386 2011-11-03 18:09:06 <luke-jr> you always want it off
387 2011-11-03 18:09:17 <luke-jr> it'll probably be removed entirely soon
388 2011-11-03 18:09:33 <cocktopus> it is less efficient than standalone alternatives
389 2011-11-03 18:09:41 <cocktopus> by like, a lot
390 2011-11-03 18:11:18 <ciscoftw> less efficient in what respects? ...so the answer to my question would be, using the bitcoind client to mine does NOT matter how the "gen" flag is set.
391 2011-11-03 18:11:43 <cocktopus> correct, the gen flag uses the client as a miner
392 2011-11-03 18:11:55 <cocktopus> you need the server flag to just host a solo mining setup
393 2011-11-03 18:12:51 <cocktopus> using server allows you to solo mine with an optimized miner, instead of what's built in to the client
394 2011-11-03 18:13:12 <luke-jr> 'gen'
395 2011-11-03 18:13:17 <luke-jr> 'gen' controls the bitcoind miner
396 2011-11-03 18:13:26 <luke-jr> bitcoind provides work via JSON-RPC with or without it
397 2011-11-03 18:13:48 <luke-jr> ciscoftw: FWIW, you're better off not CPU mining at all, and if you have a GPU using an optimized pool like #Eligius
398 2011-11-03 18:13:49 <ciscoftw> ohhhh, i never even knew the client could mine... thanks make much more sense. many thanx. all my clients connect over rpc to the bitcoind service
399 2011-11-03 18:14:23 <cocktopus> luke-jr, do you know when it will be removed completely?
400 2011-11-03 18:14:32 <cocktopus> and how much space is estimated to save?
401 2011-11-03 18:14:47 <luke-jr> cocktopus: who needs space?
402 2011-11-03 18:14:51 <luke-jr> software is tiny
403 2011-11-03 18:15:00 <cocktopus> i know but meh
404 2011-11-03 18:15:04 <luke-jr> I'd guess 0.6 at the earliest
405 2011-11-03 18:15:20 <luke-jr> probably later than that tho
406 2011-11-03 18:15:44 <cocktopus> if i compile it, can i disable a flag to leave it out, or is it not possible that way
407 2011-11-03 18:16:00 <luke-jr> no
408 2011-11-03 18:16:04 <luke-jr> patches welcome, I'm sure
409 2011-11-03 18:16:25 <cocktopus> yah ok
410 2011-11-03 18:16:27 <luke-jr> but really, I don't think anyone *cares* to do the work to #ifdef or remove it
411 2011-11-03 18:16:47 <ciscoftw> i do about 2,000 Mh/s (between 4 machines), been in a pool for last 2months (got 25btc) i'm just gonna go solo now, hence all these questions regarding the gen flag. didnt know that gen flag was to have the bitcoin service itself mine... not really what a bitcoin client is for imo, then again i suppose when cpu mining was still possible it had its purpoes
412 2011-11-03 18:17:49 <Raccoon> Google Chrome users:  http://www.google.com/search?q=Do+a+Barrel+Roll
413 2011-11-03 18:18:05 <cocktopus> works in other browsers too
414 2011-11-03 18:18:23 <Raccoon> like what browsers
415 2011-11-03 18:18:39 <luke-jr> ciscoftw: better to stick with a good pool
416 2011-11-03 18:18:41 <cocktopus> firefox and equivalent
417 2011-11-03 18:19:05 <Raccoon> not any finished version of Firefox
418 2011-11-03 18:19:38 <luke-jr> also, note that the built-in miner has been slowed down to simplify the code, recently
419 2011-11-03 18:20:00 <luke-jr> since its new "purpose" is to make it easy to testnet-in-a-box mine
420 2011-11-03 18:21:12 <cocktopus> !bc,calcd 2000000 [bc,diff]
421 2011-11-03 18:21:13 <gribble> The average time to generate a block at 2000000 Khps, given the supplied difficulty of 1203461.92638, is 4 weeks, 1 day, 21 hours, 53 minutes, and 34 seconds
422 2011-11-03 18:21:29 <ciscoftw> luke-jr: dont really care about btc at this point, all my hardware is nvidia anyway... used to run pyrit (preshare master key generation) ...just runs bitcoin when not gen'ing psks
423 2011-11-03 18:21:56 <luke-jr> ciscoftw: just sayin
424 2011-11-03 18:22:25 <cocktopus> not running constantly reduces your probability of finding a block anyway
425 2011-11-03 18:22:39 <ciscoftw> si si, i'm pretty sure i'll never solve a block, took me a month to submit 1mil shares..
426 2011-11-03 18:22:40 <luke-jr> not really, no
427 2011-11-03 18:23:22 <luke-jr> ciscoftw: you know the nvidias will use less power if you leave them idle instead?
428 2011-11-03 18:23:34 <ciscoftw> :)
429 2011-11-03 18:23:34 <cocktopus> power == money :D
430 2011-11-03 18:23:57 <ciscoftw> this runs at work, power = free (kinda) ...free for me
431 2011-11-03 18:24:16 <luke-jr> pretty sure most companies would consider that an abuse
432 2011-11-03 18:24:39 <ciscoftw> should see this rig running when pyrit is cooking
433 2011-11-03 18:25:27 <cocktopus> ciscoftw: what do you charge for key cracking
434 2011-11-03 18:26:08 <ciscoftw> nodda, its part of the audit(s)
435 2011-11-03 18:26:29 <ciscoftw> cracking isnt all that intense,  its the table generation
436 2011-11-03 18:26:32 <cocktopus> ah ic
437 2011-11-03 18:27:18 <ciscoftw> would be nice to have some ati hardware for btc though... better off just buying. nvidia is just horrible at sha
438 2011-11-03 18:27:46 <ciscoftw> but buying ati hardware would surely be an abuse :)
439 2011-11-03 18:28:04 <cocktopus> lol yes since it is slower at floating point you couldn't justify it
440 2011-11-03 18:28:04 <luke-jr> indeed
441 2011-11-03 18:29:14 <ciscoftw> and i'm not gonna pony up 500bucks for coulpe cards, when i could just buy it, wanna say two/three weeks ago btw was less than 2bucks
442 2011-11-03 18:32:08 <luke-jr> lol
443 2011-11-03 18:32:17 <luke-jr> hasn't been that low since like February :P
444 2011-11-03 19:37:50 <CIA-34> bitcoin: Wladimir J. van der Laan master * rc589f09 / contrib/gitian-descriptors/qt-win32.yml : Make qt-win32 gitian build deterministic - http://git.io/GVFmWw
445 2011-11-03 19:37:51 <CIA-34> bitcoin: Merge pull request #604 from laanwj/winqt_deterministic
446 2011-11-03 19:55:19 <dgores> CTransaction.nLockTime is always 0?  I do not see anywhere its set to anything else.
447 2011-11-03 19:56:04 <gmaxwell> dgores: Yes, it's a feature that we currently do not use.
448 2011-11-03 19:56:24 <gmaxwell> Its validated by the network, but the official client doesn't make use of it.
449 2011-11-03 19:56:40 <dgores> Thanks.  How about... in general, what does IsFinal mean?
450 2011-11-03 19:59:14 <gmaxwell> Part of the idea behind nlocktime was that locked transactions could be replaced... IsFinal is just the check that the replacement has passed (e.g. because the transaction is now unlocked)
451 2011-11-03 19:59:46 <gmaxwell> If you're reading the code, keep in mind that just because e.g. nothing in the official bitcoin client sets nlocktime!=0 that doesn't mean that something else on the network couldn't.
452 2011-11-03 19:59:57 <gmaxwell> So there is code to deal with that, and it's not dead code.
453 2011-11-03 20:00:32 <dgores> you mean... a transaction could be replaced before the block has been generated?
454 2011-11-03 20:01:35 <gmaxwell> dgores: Locked transactions (transactions with nlocktime!=0) could be replaced before their lock expires.
455 2011-11-03 20:02:01 <sipa> locked transactions are never included in blocks
456 2011-11-03 20:02:08 <gmaxwell> (though the actual _replacement_ code doesn't exist IIRC, but nodes still can't mine non-final transactions)
457 2011-11-03 20:02:21 <sipa> indeed
458 2011-11-03 20:03:47 <b4epoche> wtf?  ask a question and the network immediate hiccups
459 2011-11-03 20:04:07 <copumpkin> we saw nothing
460 2011-11-03 20:04:26 <gmaxwell> We never saw your question. Perhaps your question was not harmonious with the views of the powers that be?
461 2011-11-03 20:04:29 <gmaxwell> ;)
462 2011-11-03 20:04:42 <sipa> peer intervened
463 2011-11-03 20:04:44 <b4epoche> does it seem reasonable that the current network hashing power could crack a 27-lowercase-alphabetic character password in 10 minutes?
464 2011-11-03 20:04:57 <b4epoche> maybe the network hiccuped just before asking...
465 2011-11-03 20:05:16 <sipa> that's comparable to a 127 bit keyspace
466 2011-11-03 20:05:20 <sipa> *equivalent
467 2011-11-03 20:05:24 <sipa> so i'd say no :)
468 2011-11-03 20:05:46 <gmaxwell> log2(26^27)=126.91
469 2011-11-03 20:05:50 <gmaxwell> yea. _No_.
470 2011-11-03 20:09:03 <dgores> Thanks for the clarification.  So miners never include non-final transactions in their blocks. :)
471 2011-11-03 20:09:29 <gmaxwell> Or more importantly, the block would be invalid if it contained any.
472 2011-11-03 20:14:24 <b4epoche> hmm&  what's wrong with my calculation of 27 k-multicombos of 27 being about 970 trillion?
473 2011-11-03 20:15:13 <b4epoche> and at roughly 1 trillion hashes per second...
474 2011-11-03 20:15:18 <sipa> no idea, how did you come to that result?
475 2011-11-03 20:15:37 <sipa> 26^27 = 160059109085386090080713531498405298176
476 2011-11-03 20:16:24 <gmaxwell> b4epoche: are you couting only selection and not permutation?
477 2011-11-03 20:16:30 <b4epoche> 27 combinations of 27 objects with repeats
478 2011-11-03 20:17:04 <sipa> 26 objects, no?
479 2011-11-03 20:17:52 <b4epoche> eh, duh, yea...
480 2011-11-03 20:17:59 <sipa> otherwise that's correct
481 2011-11-03 20:18:21 <sipa> but selecting n elements from m possibilities, with order important and repeats allowed: m^n
482 2011-11-03 20:19:18 <b4epoche> ah, yea, I was doing order unimportant&
483 2011-11-03 20:19:29 <gmaxwell> That why I asked about not permutation.
484 2011-11-03 20:20:15 <b4epoche> for some stupid reason I was thinking order was unimportant which is just dumb
485 2011-11-03 20:20:48 <gmaxwell> the n^m form is pretty intutive though ... N ways ... times N ways.. and so on.
486 2011-11-03 20:21:08 <b4epoche> yea, I was making it harder than it is
487 2011-11-03 20:23:06 <sipa> @seen BlueMatt
488 2011-11-03 20:23:14 <gribble> BlueMatt was last seen in #bitcoin-dev 4 hours, 13 minutes, and 17 seconds ago: <BlueMatt> wumpus: ping
489 2011-11-03 20:23:14 <sipa> ;;seen BlueMatt
490 2011-11-03 20:23:41 <b4epoche> so, about an 11 character (lowercase alphabetic) password
491 2011-11-03 20:23:52 <sipa> 51 bits
492 2011-11-03 20:24:50 <gribble> Error: "bc,spotest" is not a valid command.
493 2011-11-03 20:24:50 <sipa> ;;bc,spotest
494 2011-11-03 20:24:52 <sipa> ;;bc,spotestimate
495 2011-11-03 20:24:53 <gribble> 1179311.65836
496 2011-11-03 20:25:14 <b4epoche> 51 bits?
497 2011-11-03 20:25:44 <gribble> Alias bc,24hprc, Alias bc,altprofit, Alias bc,avgprc, Alias bc,bcm, Alias bc,bitpenny, Alias bc,blocks, Alias bc,bounty, Alias bc,btceur, Alias bc,btcgbp, Alias bc,btcguild, Alias bc,btcrub, Alias bc,btcto, Alias bc,calc, Alias bc,calcd, Alias bc,channels, Alias bc,convert, Alias bc,deepbit, Alias bc,diff, Alias bc,diffchange, Alias bc,eligius, Alias bc,estimate, Alias bc,exchb, Alias bc,fx, (2 more messages)
498 2011-11-03 20:25:44 <sipa> ;;bc,help
499 2011-11-03 20:25:49 <sipa> ;;bc,hashspeed
500 2011-11-03 20:25:50 <gribble> Error: "bc,hashspeed" is not a valid command.
501 2011-11-03 20:25:52 <sipa> ;;bc,hashrate
502 2011-11-03 20:25:53 <gribble> Error: "bc,hashrate" is not a valid command.
503 2011-11-03 20:25:58 <b4epoche> password length?  I'm trying to come up with examples lay-people can relate to
504 2011-11-03 20:25:59 <gribble> 1190018.35695522
505 2011-11-03 20:25:59 <sipa> ;;bc,estimate
506 2011-11-03 20:26:29 <midnightmagic> permutations and combinations isn't statistics, it's discrete math :-)
507 2011-11-03 20:26:34 <midnightmagic> :-D
508 2011-11-03 20:26:36 <sipa> it's combinatorics
509 2011-11-03 20:26:48 <b4epoche> and I still hate statistics
510 2011-11-03 20:27:16 <sipa> b4epoche: i say "51 bits" because it corresponds to searching a 2^51 keyspace
511 2011-11-03 20:27:54 <midnightmagic> commonly lumped together with discrete math, in a course usually called discrete math.
512 2011-11-03 20:28:50 <midnightmagic> but i guess technically is combinatorics.
513 2011-11-03 20:28:59 <sipa> b4epoche: so if each password attempt costs as much as a bitcoin block hash attempt, the current bitcoin network would do it in 450s
514 2011-11-03 20:29:23 <sipa> however, if e.g. key strengthening is applied, it costs a lot more per attempt
515 2011-11-03 20:29:39 <gmaxwell> Or if, e.g. it was just regular 1x md5 .. it would go _much_ faster.
516 2011-11-03 20:30:24 <sipa> @later tell BlueMatt ok, i got gitian to build bitcoin again!
517 2011-11-03 20:30:35 <gribble> The operation succeeded.
518 2011-11-03 20:30:35 <sipa> ;;later tell BlueMatt ok, i got gitian to build bitcoin again!
519 2011-11-03 20:31:19 <midnightmagic> it would be interesting if the bitcoin network itself included a way to request, offer to pay, and distribute the results of, cracking problems.
520 2011-11-03 20:31:21 <b4epoche> I'm just looking for a few tidbits to get people oriented with hashing, 'big numbers', and 'hard problems'
521 2011-11-03 20:31:48 <b4epoche> this damn TEDx talk is fast approaching...
522 2011-11-03 20:32:02 <midnightmagic> b4epoche: are you doing one?
523 2011-11-03 20:32:09 <b4epoche> yes
524 2011-11-03 20:32:39 <b4epoche> and I need to explain to the graphic designer putting the slides together how to illustrate all this for a general audience
525 2011-11-03 20:33:20 <midnightmagic> b4epoche: sweet, someone I know is doing one too.. he's doing biohacking. do you have any links that will eventually be a way to watch it?
526 2011-11-03 20:33:43 <b4epoche> I was thinking of trying to explain the proof-of-work in terms of password cracking.  with difficulty adjustments being changes in the password length
527 2011-11-03 20:34:09 <b4epoche> probably here:  http://www.tedxpsu.com/
528 2011-11-03 20:34:15 <gmaxwell> b4epoche: it's comprehensible but it fosters a view that bitcoin is somehow illigimate or unlawful.
529 2011-11-03 20:34:28 <midnightmagic> cool beans man, i hope you rock it.
530 2011-11-03 20:35:00 <gmaxwell> b4epoche: There is some youtube video with an idiot blabering on about that "Oh? So what do you think bitcoin is doing with all that computing power?!?! It's cracking passwords!"
531 2011-11-03 20:36:45 <midnightmagic> b4epoche: You're Eric? That is an AWESOME last name man. :)
532 2011-11-03 20:37:29 <b4epoche> Yea, I'm Eric
533 2011-11-03 20:38:14 <b4epoche> gmaxwell:  that's true&
534 2011-11-03 20:38:56 <b4epoche> I'm trying to come up with a some way of explaining 'adjustably hard problems'
535 2011-11-03 20:38:57 <gmaxwell> If you just want to wow them count up the add operations used in 2xsha256
536 2011-11-03 20:39:19 <gmaxwell> Then estimate how long it would take to do a 32,32 add on a set of abacuses.
537 2011-11-03 20:39:30 <gmaxwell> Then divide by the worlds population.
538 2011-11-03 20:39:47 <midnightmagic> the best way i've found so far to explain it to people is a dartboard analogy, with shrinking pie-slices to aim for..
539 2011-11-03 20:40:13 <edcba> b4epoche: what about problem of varying difficulty ? :)
540 2011-11-03 20:40:32 <gmaxwell> I've explained it as rolling many-sided dice with accutable solutions being increasingly small values.
541 2011-11-03 20:40:43 <b4epoche> the problem with problems is that most people think that the smarter you are the faster you'll solve them
542 2011-11-03 20:40:44 <edcba> a sudoku less and less filled :)
543 2011-11-03 20:40:48 <gmaxwell> But people tend not to believe that the mining process is like dice.
544 2011-11-03 20:41:22 <gmaxwell> edcba: sudoku have superliner solvers, not really the same kind of problem.
545 2011-11-03 20:41:51 <b4epoche> dart board also suggests that if you're better at darts you'll have better odds
546 2011-11-03 20:42:08 <midnightmagic> the leap of faith happens the moment they accept that putting data into SHA256 gives you a mystery number back out of it.
547 2011-11-03 20:42:15 <b4epoche> maybe it's like being able to predict the weather accurately N days out
548 2011-11-03 20:42:17 <gmaxwell> Well, back to my dice.  .. except the dice have 2^256 sides.. a bit hard to visualize.
549 2011-11-03 20:42:22 <b4epoche> well, for N straight days
550 2011-11-03 20:42:27 <edcba> finding something at random with some properties
551 2011-11-03 20:42:35 <gmaxwell> midnightmagic: yes, well, we're only hoping that SHA256 is a random oracle!
552 2011-11-03 20:43:08 <gmaxwell> b4epoche: the weather still has an enormous gain from getting the basic behavior right.
553 2011-11-03 20:43:33 <b4epoche> true
554 2011-11-03 20:43:39 <edcba> phone calling ppl at random and hoping they are called julia
555 2011-11-03 20:44:19 <edcba> or you could just use mining with rock having particular shape
556 2011-11-03 20:44:29 <b4epoche> thx all&  I gotta get home to eat some dinner
557 2011-11-03 20:44:49 <midnightmagic> i call it a mystery number.. i avoid calling it random in case someone thinks a calculation can't be random, which i encounter surprisingly often.
558 2011-11-03 20:45:02 <edcba> ok it's like doing jenga with rocks
559 2011-11-03 20:45:02 <midnightmagic> hoping they're called julia. awesome. :)
560 2011-11-03 20:45:09 <gmaxwell> midnightmagic: yea, I have that problem with the dice analogy, they think the process can't be like throwing dice.
561 2011-11-03 20:45:25 <edcba> the higher the block chain the most secure it is
562 2011-11-03 20:45:35 <b4epoche> well, isn't it the inverse problem that's random
563 2011-11-03 20:45:39 <edcba> and the difficulty change according previous piece
564 2011-11-03 20:46:04 <midnightmagic> just wave my hands and say, "Nobody knows how to predict the number or go backwards.. YET.." dun dun DUNNNN
565 2011-11-03 20:48:00 <gmaxwell> b4epoche: The hash is believed to be a random oracle, upto computational intractability.
566 2011-11-03 20:48:21 <gmaxwell> You put a number in .. a number comes out. You can't predict anything about the number that comes out, or go backwards.
567 2011-11-03 20:48:28 <midnightmagic> hey Tycho! did the ddos people stop yet? also, your IRC client joins a channel before it authenticates with nickserv, fyi.
568 2011-11-03 20:48:54 <RuzzlePuzzle> hey guys
569 2011-11-03 20:48:57 <RuzzlePuzzle> i got a massive problem
570 2011-11-03 20:48:58 <midnightmagic> except maybe a good distribution..
571 2011-11-03 20:49:14 <[Tycho]> midnightmagic, yes, I know it.
572 2011-11-03 20:49:17 <gavinandresen> I like the throwing dice analogy-- you're throwing 200 dice and hoping 20 of them come up ones.  Difficulty goes up... hoping that 21 of them come up ones....
573 2011-11-03 20:49:30 <[Tycho]> I think that we have no DDoS at this moment.
574 2011-11-03 20:49:33 <gmaxwell> midnightmagic: No, thats not promised either. There may be subspaces where it's all zeros that come out or something. :)
575 2011-11-03 20:49:49 <[Tycho]> It usualy doesn't last for a long time because it's expensive.
576 2011-11-03 20:50:16 <midnightmagic> gmaxwell: hrm. I thought one of the properties was a random-like distribution? like we're confident that it's not going to shift towards the upper-end of the possible result space?
577 2011-11-03 20:50:40 <gmaxwell> gavinandresen: the best thing about the dice is that people instantly get the linearity of it.  If you and I are both playing this game, and you throw 2x faster than me, you'll get 2x the solutions.. but I'll still get solutions too.
578 2011-11-03 20:50:48 <[Tycho]> Sometimes some home users try to flood us with just http requests, but this doesn't affect my servers, I only see it in my logs :)
579 2011-11-03 20:50:51 <gavinandresen> gmaxwell: yup
580 2011-11-03 20:51:26 <gmaxwell> midnightmagic: If you can reliably show that its not random then its not a random oracle, but thats not the same as being able to show that it meets any perticular constraint. Perhaps this is a bit too pedantic.
581 2011-11-03 20:51:36 <midnightmagic> [Tycho]: well that's good. :) hey I don't suppose you have namecoin merged-mining in deepbit's eventual future?
582 2011-11-03 20:52:14 <RuzzlePuzzle> I have a bunch of servers which dont allow me to open connections, but they allow me to create connections (cheaper) but... I want to use them as proxies so... I register them all on a centralised (i guess) DNS type server which I use to proxy through into the machines. The problem is, either I leave a socket open for each reverse connection to tell them when they are needed and to establish
583 2011-11-03 20:53:02 <RuzzlePuzzle> if I leave a bunch of connections open, thats sucks, and if I make them spam a resource on the server that sucks even more
584 2011-11-03 20:53:07 <RuzzlePuzzle> what can I do?
585 2011-11-03 20:53:18 <gmaxwell> Pay for access to a better botnet?
586 2011-11-03 20:53:18 <midnightmagic> gmaxwell: hrm. i'm pretty sure SHA has as one of its qualities uniform distribution.. ? well now I can't find any reference for that.
587 2011-11-03 20:53:39 <RuzzlePuzzle> these are cheap webservers
588 2011-11-03 20:54:06 <RuzzlePuzzle> im using php to create the connections
589 2011-11-03 20:54:14 <midnightmagic> ah, I'm mixing it up. Uniform distribution != random-like distribution.
590 2011-11-03 20:54:17 <RuzzlePuzzle> open ports is not allowed
591 2011-11-03 20:56:46 <RuzzlePuzzle> programs like skype, teamview and bitcoin must do this already?
592 2011-11-03 20:56:51 <RuzzlePuzzle> how do they deal with it?
593 2011-11-03 20:57:26 <gmaxwell> RuzzlePuzzle: clients that listen, listen. Clients that don't don't.  UPNP is used to try to increase the number of listening clients.
594 2011-11-03 20:58:15 <RuzzlePuzzle> where can I find an upnp example in the bitcoin source?
595 2011-11-03 21:00:36 <DrHaribo> Bah. I created some broken generation transactions on testnet. Anyone know how different transactions must be to not be considered duplicates?
596 2011-11-03 21:00:40 <gmaxwell> RuzzlePuzzle: bitcoin uses a upnp library, but thats not what you wan't if you're not permitted to listen.
597 2011-11-03 21:00:55 <gmaxwell> DrHaribo: they need to not have the same hash.
598 2011-11-03 21:01:00 <DrHaribo> Is it enough to have 1 byte be different in the coinbase?
599 2011-11-03 21:01:06 <DrHaribo> Ah great - then that should do it
600 2011-11-03 21:01:35 <DrHaribo> Thanks, gmaxwell !
601 2011-11-03 21:04:35 <RuzzlePuzzle> gmaxwell: true. im not sure what other options I have
602 2011-11-03 21:05:16 <RuzzlePuzzle> maybe I just need to pay a bit more for vpn's
603 2011-11-03 21:14:45 <[Tycho]> midnightmagic, no.
604 2011-11-03 21:14:59 <midnightmagic> [Tycho]: okie doke! I'll stop bugging you.
605 2011-11-03 21:15:44 <CIA-34> bitcoin: jedi95 * r2a3a6d2a9ee6 Phoenix-Miner/ (Miner.py minerutil/RPCProtocol.py): Set RPC LP timeout to 10 minutes. Version bump to 1.7.0
606 2011-11-03 21:28:33 <gmaxwell> a 1m buckball dice.
607 2011-11-03 21:29:32 <gmaxwell> "The nucleus to nucleus diameter of a C60 molecule is about 0.71 nm"
608 2011-11-03 21:31:17 <b4epoche_> yea, that too&  but a buckball die would not have equi-probability sides
609 2011-11-03 21:31:44 <gmaxwell> nor would any spherical die.
610 2011-11-03 21:31:51 <b4epoche_> I think you'd need a crystal with a tetragonal lattice
611 2011-11-03 21:32:06 <b4epoche_> well, it obviously wouldn't be perfectly spherical
612 2011-11-03 21:32:30 <gmaxwell> only the platonic solids let you make equalprobable quasispherical dice.
613 2011-11-03 21:33:07 <gmaxwell> but you can make cylinerical ones, since you can construct a n-equal-sided polygon for any n.
614 2011-11-03 21:33:52 <invisiblemonk> damn it.
615 2011-11-03 21:34:08 <gmaxwell> if you admit more more dimensions you do get more regular polytopes though. Not exactly helpful for visualization.
616 2011-11-03 21:34:52 <gmaxwell> (well, one more)
617 2011-11-03 21:35:25 <b4epoche_> yea, I was thinking of a triangulated sphere
618 2011-11-03 21:36:40 <midnightmagic> someone used to play.. d&d..?
619 2011-11-03 21:37:16 <b4epoche_> long long ago
620 2011-11-03 22:36:45 <BlueMatt> sipa: re: gitian YAY!
621 2011-11-03 22:37:38 <BlueMatt> sipa: we need gavin to release a qt-win32 build as its build is not yet deterministic, then we can get everyone building the release and even do gitian releases :)
622 2011-11-03 23:06:04 <BlueMatt> ;;later tell sipa re: gitian YAY!
623 2011-11-03 23:06:04 <gribble> The operation succeeded.
624 2011-11-03 23:06:10 <BlueMatt> ;;later tell sipa we need gavin to release a qt-win32 build as its build is not yet deterministic, then we can get everyone building the release and even do gitian releases :)
625 2011-11-03 23:06:10 <gribble> The operation succeeded.
626 2011-11-03 23:25:20 <jjjrmy> Anyone hungry for some pizza?!
627 2011-11-03 23:25:37 <sipa> i'll get you one for 10k BTC
628 2011-11-03 23:26:20 <jjjrmy> sipa: www.bitpizza.net
629 2011-11-03 23:27:41 <sipa> haha
630 2011-11-03 23:28:42 <jjjrmy> DO IT!
631 2011-11-03 23:28:48 <sipa> https://bitcointalk.org/index.php?topic=137.0
632 2011-11-03 23:29:13 <jjjrmy> sipa: Yeah, I know. :p old news
633 2011-11-03 23:29:25 <sipa> :)
634 2011-11-03 23:29:33 <sipa> very old news