1 2010-12-30 00:01:34 johndrinkwater has quit (Remote host closed the connection)
2 2010-12-30 00:01:50 johndrinkwater has joined
3 2010-12-30 00:08:50 slush has quit (Ping timeout: 240 seconds)
4 2010-12-30 00:12:44 Cusipzzz has joined
5 2010-12-30 00:14:21 <BoBeR> mtgox, around??????????
6 2010-12-30 00:16:28 <afed> i've got 35 bitcoins now f34r
7 2010-12-30 00:16:40 <afed> tryin to bringup this 6950 card
8 2010-12-30 00:16:52 <afed> become a mining baron
9 2010-12-30 00:21:16 <nanotube> BoBeR: might have better luck sending him email.
10 2010-12-30 00:21:21 <nanotube> afed: good going. :)
11 2010-12-30 00:21:31 <BoBeR> ye i guess
12 2010-12-30 00:22:27 devon_hillard has quit (Quit: Leaving)
13 2010-12-30 00:24:26 <afed> hmm
14 2010-12-30 00:24:53 <afed> only 233 megahashes
15 2010-12-30 00:25:02 marioxcc has left ("ERC Version 5.3 (IRC client for Emacs)")
16 2010-12-30 00:25:06 <afed> durrr i didn't flash it to 6970
17 2010-12-30 00:25:42 <ArtForz> still a bit slower than what it should be able to do
18 2010-12-30 00:26:06 <ArtForz> 6950 should be able to get ~260Mh/s
19 2010-12-30 00:27:38 <afed> it's also on an 8x slot
20 2010-12-30 00:27:50 <afed> i didn't think gpu<->system bandwidth had a lot to do with it though
21 2010-12-30 00:27:57 <ArtForz> thats a <0.1% difference
22 2010-12-30 00:28:37 <afed> good to know
23 2010-12-30 00:30:14 <ArtForz> my guess is compiler or driver optimization problem
24 2010-12-30 00:31:09 <afed> could be anything, i haven't done much more than plug the card in yet
25 2010-12-30 00:31:40 grondilu has quit (Quit: leaving)
26 2010-12-30 00:31:51 <ArtForz> 69xx is still very new, and with AMDs internal need-to-know strategy I doubt the driver or opencl dev teams had much time to optimize
27 2010-12-30 00:32:00 davout has quit (Quit: i <3 pork (http://dev.ojnk.net))
28 2010-12-30 00:32:32 BoBeR has quit (Read error: Connection reset by peer)
29 2010-12-30 00:32:47 BoBeR has joined
30 2010-12-30 00:34:51 AAA_awright has joined
31 2010-12-30 00:34:57 RichardG has joined
32 2010-12-30 00:36:39 RichardG has quit (Client Quit)
33 2010-12-30 00:36:52 AAA_awright_ has quit (Ping timeout: 272 seconds)
34 2010-12-30 00:37:37 <da2ce7> yay! Got annother 5970 for $670
35 2010-12-30 00:37:49 <ArtForz> wow, thats expensive
36 2010-12-30 00:38:06 <da2ce7> I know, but they have dryed up.
37 2010-12-30 00:39:09 <da2ce7> it is cheaper than 2x 6950 + new case.
38 2010-12-30 00:39:13 <da2ce7> *
39 2010-12-30 00:39:15 <da2ce7> new mobo.
40 2010-12-30 00:39:24 <afed> i think the reason i picked the 6950 was because of the ability to flash it to 6970
41 2010-12-30 00:39:38 <afed> if i build a new miner it'll be 2x 5870
42 2010-12-30 00:39:44 <afed> newegg is selling them with a free 1 TB disk
43 2010-12-30 00:40:29 <ArtForz> around here they're ~ $570 incl. tax
44 2010-12-30 00:40:39 <ArtForz> and etailers have ~20 in stock
45 2010-12-30 00:41:05 <da2ce7> yeah, I'm in AUS... fucking expencive, even tho our dollar is stronger than yours.
46 2010-12-30 00:41:29 <ArtForz> i'm in europe, we usually get fucked on component prices, too
47 2010-12-30 00:41:37 <afed> did any of you get that ridiculous 5970 that comes with a plastic gun?
48 2010-12-30 00:42:03 <afed> http://www.newegg.com/Product/Product.aspx?Item=N82E16814150500&cm_re=radeon_5970-_-14-150-500-_-Product
49 2010-12-30 00:42:21 <da2ce7> afed, no it is just a stock Gigabyte one.
50 2010-12-30 00:42:36 <da2ce7> hmm, because the 5970 is one pci-e slot, I can me to move arround my graphics cards, and I only retire one 5770
51 2010-12-30 00:43:04 <ArtForz> thats not a 5970, thats a 5870x2 :P
52 2010-12-30 00:43:45 <joe_1> i'm only getting 6 ghash/s what am i doing wrong
53 2010-12-30 00:43:58 <afed> lol
54 2010-12-30 00:43:59 <da2ce7> lol
55 2010-12-30 00:44:01 <ArtForz> or actually it'd be a 5870x2 eyefinity edition
56 2010-12-30 00:44:38 <afed> crazy
57 2010-12-30 00:45:17 <ArtForz> 5870 = 1GB ram 2*DVI+hdmi+DP, 5870 eyefinity = 2GB ram 6*miniDP
58 2010-12-30 00:46:12 <ArtForz> and that "5970" has 5870 clocks, 2*2G ram and 6 miniDP ...
59 2010-12-30 00:48:22 <da2ce7> ArtForz, have you bought any GPU's recently, or working on your next-gen solution to make my new GPU about as usefull as my CPU? lol
60 2010-12-30 00:50:31 <ArtForz> havent added any GPUs for mining
61 2010-12-30 00:50:38 <nanotube> ;;bc,poolstats
62 2010-12-30 00:50:39 <gribble> {"active_workers": 170, "hashes_ps": 5733835232, "shares": 16928, "round_started": "2010-12-29 21:16:38"}
63 2010-12-30 00:50:48 <ArtForz> still 24 5970s and 5 5770s
64 2010-12-30 00:53:13 <nanotube> got your asics in the works? :)
65 2010-12-30 00:53:51 marioxcc has joined
66 2010-12-30 00:54:07 <da2ce7> I love slush's pool. I can have my brothers GPU in my computer, and the coins it generats will be sent to his account.
67 2010-12-30 00:54:14 <da2ce7> :)
68 2010-12-30 00:54:15 Zarutian has quit (Quit: Zarutian)
69 2010-12-30 00:55:24 <nanotube> da2ce7: heh
70 2010-12-30 00:56:53 <da2ce7> nanotube, I manage the hardware, my bro has the money. So I send the coins his account, and he compensates me for the hosting and power.
71 2010-12-30 00:57:08 acous has joined
72 2010-12-30 00:57:08 acous has quit (Changing host)
73 2010-12-30 00:57:08 acous has joined
74 2010-12-30 00:57:24 <nanotube> nice setup :)
75 2010-12-30 00:59:34 gavinandresen has quit (Quit: gavinandresen)
76 2010-12-30 00:59:49 ThomasV has quit (Ping timeout: 240 seconds)
77 2010-12-30 01:02:12 <marioxcc> in my opinion, we need a query to await new block information
78 2010-12-30 01:02:38 <marioxcc> i'm implementing the work-queue
79 2010-12-30 01:03:03 <marioxcc> and it is going to be ineficient to constantly refresh the queue
80 2010-12-30 01:03:06 <ArtForz> huh?
81 2010-12-30 01:03:14 <marioxcc> ArtForz: current cpuminer
82 2010-12-30 01:03:39 <marioxcc> request sends a getwork just when it's need
83 2010-12-30 01:03:46 <marioxcc> that adds a delay
84 2010-12-30 01:03:53 <marioxcc> and it is unused cpu
85 2010-12-30 01:03:55 <ArtForz> my miner does it that way since forever
86 2010-12-30 01:04:33 <marioxcc> ArtForz: my point is:
87 2010-12-30 01:04:45 <marioxcc> to avoid deadtime you need a work queue (because of network latency)
88 2010-12-30 01:04:53 <ArtForz> yes
89 2010-12-30 01:04:56 <ArtForz> well, not really
90 2010-12-30 01:05:05 <marioxcc> and to keep the queue with fresh works (valid ones)
91 2010-12-30 01:05:06 <ArtForz> I use a async push based approach
92 2010-12-30 01:05:08 <marioxcc> you need to refresh its
93 2010-12-30 01:05:24 <marioxcc> becuase you don't know when to renew these
94 2010-12-30 01:05:28 <marioxcc> you need to assume it's always
95 2010-12-30 01:05:33 <marioxcc> and then renew them constantly
96 2010-12-30 01:05:36 <ArtForz> ?
97 2010-12-30 01:05:43 <ArtForz> my node knows pretty damn well when a new block arrives
98 2010-12-30 01:05:47 <marioxcc> well, tellme about your approach
99 2010-12-30 01:06:23 <ArtForz> miner connects to node, node sends "start working on block header X" and "update nTime to Y" messages
100 2010-12-30 01:06:35 <afed> now i'm getting ~280 megahashes :)
101 2010-12-30 01:06:44 <afed> with 6970 bios and clock speeds
102 2010-12-30 01:06:57 <ArtForz> miner responds with #nonces completed since the prev message as soon as it actually does it
103 2010-12-30 01:07:10 <marioxcc> ArtForz: well, i mean cpuminer
104 2010-12-30 01:07:26 <marioxcc> i don't know wheter slush sever support your approach
105 2010-12-30 01:07:30 <ArtForz> well, don't blame me for m0 and satoshi using a bass-ackwards approach for getwork
106 2010-12-30 01:07:50 <marioxcc> ¿bass-ackwards?
107 2010-12-30 01:08:36 <nanotube> ArtForz: maybe you can suggest to them the better approach? :)
108 2010-12-30 01:09:05 <nanotube> ;;ud bass-ackwards
109 2010-12-30 01:09:05 <gribble> http://www.urbandictionary.com/define.php?term=Bass+Ackwards | Ass backwards. The state doing (or having done) something the wrong way.
110 2010-12-30 01:09:08 <nanotube> marioxcc: --^
111 2010-12-30 01:09:21 <marioxcc> ok
112 2010-12-30 01:09:37 <nanotube> (though that said... my ass faces backwards normally... so the expression makes no sense. :) it should be ass forwards.)
113 2010-12-30 01:09:58 <da2ce7> ass-in-mouth kind of way
114 2010-12-30 01:10:09 <marioxcc> haha
115 2010-12-30 01:10:40 <nanotube> fass-orwards? heh
116 2010-12-30 01:12:11 TD has quit (Quit: TD)
117 2010-12-30 01:12:20 <marioxcc> nanotube: LOL
118 2010-12-30 01:27:24 <tcatm> Which video codec is state-of-the-art for videos on websites?
119 2010-12-30 01:27:59 ciuciu has joined
120 2010-12-30 01:28:39 <tcatm> Problem solved, x264 compressed 40 secs of 1280x720 into ~750kb :)
121 2010-12-30 01:28:54 <marioxcc> tcatm: the above phrase is meaningless
122 2010-12-30 01:29:08 <marioxcc> virtually any code can compress any video length to any size
123 2010-12-30 01:29:24 <marioxcc> is the video entropy and output quality what makes the difference
124 2010-12-30 01:30:40 <tcatm> http://bitcoincharts.com/media/stuff/bitcoin.avi bitcoin's commit log visualized using code_swarm
125 2010-12-30 02:08:35 <BoBeR> open codecs ftw
126 2010-12-30 02:10:07 mrb_ has quit (Quit: leaving)
127 2010-12-30 02:21:15 rogutes has joined
128 2010-12-30 02:23:11 <da2ce7> lol, the 5970 is too long! Getting two 6950's insted.
129 2010-12-30 02:24:14 <Diablo-D3> da2ce7: er, thats a bad trade
130 2010-12-30 02:24:23 <Diablo-D3> you'd really want two 5870s
131 2010-12-30 02:25:50 <ArtForz> or just get a $20 case and a dremel :P
132 2010-12-30 02:25:57 <da2ce7> diavlo-d3, afed said that he could get 280mhps on the 6950
133 2010-12-30 02:26:15 <ArtForz> 6950 @ 6970 shaders + clocks
134 2010-12-30 02:26:29 <ArtForz> which is quite a bit short of the ~310Mh/s it should be getting theoretically
135 2010-12-30 02:26:42 <da2ce7> yeah, that is what I'm planning to do.
136 2010-12-30 02:27:24 <ArtForz> and thats not "absolute theoretical peak", thats diablominer on 58xx numbers scaled to 69xx #shader ALUs and clocks
137 2010-12-30 02:27:28 <da2ce7> it works out cheaper, than buying a new case. Just means I need to retire (sell) two cards.
138 2010-12-30 02:32:10 <ArtForz> I dont see how it can be cheaper
139 2010-12-30 02:32:43 <ArtForz> looks like a 5970 is about as fast as 2 6970s
140 2010-12-30 02:33:08 <ArtForz> and 2*6970 should be north of 400W in power
141 2010-12-30 02:33:57 <Diablo-D3> http://www.youtube.com/watch?v=eQ-OcS2Gwvk
142 2010-12-30 02:33:57 <Diablo-D3> ffff
143 2010-12-30 02:34:11 <marioxcc> ArtForz: how do you know how much power do they use?
144 2010-12-30 02:34:21 <ArtForz> magic!
145 2010-12-30 02:34:37 <Diablo-D3> ArtForz: do me a favor
146 2010-12-30 02:34:39 <Diablo-D3> https://code.google.com/p/smhasher/source/browse/trunk/MurmurHash3.cpp?r=75#365
147 2010-12-30 02:34:49 <Diablo-D3> http://pastebin.ca/2033854
148 2010-12-30 02:34:54 <Diablo-D3> are those equiv?
149 2010-12-30 02:36:27 AAA_awright_ has joined
150 2010-12-30 02:36:41 <ArtForz> I assume bmix64 is a macro?
151 2010-12-30 02:36:56 <Diablo-D3> no, its another method I havent copied over yet
152 2010-12-30 02:37:02 <Diablo-D3> in the original C its above that
153 2010-12-30 02:37:15 marioxcc has left ("ERC Version 5.3 (IRC client for Emacs)")
154 2010-12-30 02:37:32 <ArtForz> C with refs?
155 2010-12-30 02:38:00 <ArtForz> I fucking hate implicit pass-by-ref
156 2010-12-30 02:39:14 AAA_awright has quit (Ping timeout: 276 seconds)
157 2010-12-30 02:39:15 <ArtForz> I think your last-partial-bock processing os wrong
158 2010-12-30 02:39:33 <BoBeR> ;;math,60/25
159 2010-12-30 02:39:33 <gribble> Error: "math,60/25" is not a valid command.
160 2010-12-30 02:39:39 <BoBeR> ;;calc,60/25
161 2010-12-30 02:39:39 <gribble> Error: "calc,60/25" is not a valid command.
162 2010-12-30 02:39:41 <ArtForz> tail is a uint8*
163 2010-12-30 02:39:57 <Kiba> ReadWriteWeb interviewed bitcoin!
164 2010-12-30 02:39:59 <Kiba> err
165 2010-12-30 02:40:03 <Kiba> a bitcoin developer!
166 2010-12-30 02:40:11 * Kiba ancipitates traffic surge
167 2010-12-30 02:45:14 <ArtForz> yeah, I think your last-block processing is wrong
168 2010-12-30 02:46:19 <Kiba> man
169 2010-12-30 02:46:30 <Kiba> I am seeing like 30 tweets about bitcoin at once retweeting the same interview
170 2010-12-30 02:46:51 <BoBeR> NEW RECORD
171 2010-12-30 02:48:22 <da2ce7> kiba, do you have a link?
172 2010-12-30 02:48:28 <Cusipzzz> linky
173 2010-12-30 02:48:52 <Kiba> http://www.readwriteweb.com/hack/2010/12/interview-bitcoin.php
174 2010-12-30 02:49:16 <BoBeR> LINKS GALORE
175 2010-12-30 02:51:06 <mrb__> 18:23 < da2ce7> diavlo-d3, afed said that he could get 280mhps on the 6950
176 2010-12-30 02:51:24 <mrb__> I bet that's because the miner he used assumed 20 SIMDs per GPU. the 6950 has 22 SIMDs
177 2010-12-30 02:51:24 <ArtForz> *headdesk*
178 2010-12-30 02:51:26 * Kiba ancipitates price surge
179 2010-12-30 02:51:38 <mrb__> that accounts exactly for the 10% diff in expected perf
180 2010-12-30 02:51:44 <ArtForz> 280MHPS on a 6950 flashed to 6970
181 2010-12-30 02:51:53 <mrb__> still
182 2010-12-30 02:52:03 <mrb__> miner must probably assume 20 SIMDs
183 2010-12-30 02:52:15 <mrb__> and launch a number of threads optimized for 20 SIMDs
184 2010-12-30 02:52:26 <ArtForz> stock 6950 got 233
185 2010-12-30 02:52:40 <mrb__> the increase comes from the MHz increase
186 2010-12-30 02:52:50 <ArtForz> or maybe the OpenCL compiler prduces crappy IL for that kernel for 69xx
187 2010-12-30 02:53:00 <mrb__> basically I think he was benchamrking 20 SIMDs at the 6970 freq
188 2010-12-30 02:53:09 INEEDMONEY has quit (Ping timeout: 264 seconds)
189 2010-12-30 02:53:16 <ArtForz> HE GOT 233 ON A STOCK 5950
190 2010-12-30 02:53:21 <ArtForz> 6950 FUCK
191 2010-12-30 02:53:44 <ArtForz> and a 10% clock increase doesnt account for 233 -> 280
192 2010-12-30 02:54:00 AAA_awright_ is now known as AAA_awright
193 2010-12-30 02:54:20 <ArtForz> 10% clock increase + 9% shader count increase does
194 2010-12-30 02:55:58 <mrb__> I don't think you understand me
195 2010-12-30 02:56:31 <mrb__> a 6950 should be doing 262 Mhash/s with m0mchil's miner
196 2010-12-30 02:56:35 <ArtForz> yep
197 2010-12-30 02:56:46 <ArtForz> the OCL runtime probably uses crappy work dims or procudes horrible IL for 69xx
198 2010-12-30 02:57:24 <mrb__> assuming m0mchil's miner configures the number of thread optimized for 20 SIMDS then: 262*20/22. = 238 Mhash/s and he measured about this number
199 2010-12-30 02:57:38 <mrb__> at 6950 shaders and clocks
200 2010-12-30 02:57:48 <ArtForz> except you dont do anything with the SIMD number manually in OCL
201 2010-12-30 02:58:03 <ArtForz> you just set local and global workqueue size
202 2010-12-30 02:58:21 <mrb__> ...which effectively is mapped to a certain number of threads
203 2010-12-30 02:58:35 <ArtForz> yes, internally, by the OCL runtime
204 2010-12-30 02:58:52 <mrb__> when writing IL code the number of threads has to be fine-tuned to fully exploit all the SIMDs, use full wavefronts, etc
205 2010-12-30 02:59:08 <mrb__> OpenCL just provides abstractions for the same thing
206 2010-12-30 02:59:12 <ArtForz> yep
207 2010-12-30 02:59:20 <ArtForz> iirc ocl local work size corresponds to wavefront multiples
208 2010-12-30 02:59:30 <ArtForz> = 64, 128 and 256 are good bets on 5xxx
209 2010-12-30 02:59:31 <mrb__> in the end I am sure m0mchil's miner can be fine tuned to reach the expected 262 Mhash/s
210 2010-12-30 02:59:42 <Diablo-D3> [09:34:52] <ArtForz> C with refs?
211 2010-12-30 02:59:42 <Diablo-D3> [09:35:20] <ArtForz> I fucking hate implicit pass-by-ref
212 2010-12-30 02:59:44 <Diablo-D3> welcome to C.
213 2010-12-30 03:01:30 <ArtForz> no, welcome to bastard assfuck C-with-fucking-pass-by-ref crap, just use a fucking pointer
214 2010-12-30 03:02:21 <ArtForz> somefunc(a,b,&c) ... hmm... I wonder which parameter might get modified by this function
215 2010-12-30 03:02:55 <OneFixt> ArtForz: ++
216 2010-12-30 03:03:22 <Diablo-D3> ArtForz: so does the java look right?
217 2010-12-30 03:03:28 <ArtForz> no
218 2010-12-30 03:03:39 <Diablo-D3> hrm
219 2010-12-30 03:03:45 <Diablo-D3> well, the tail handling
220 2010-12-30 03:03:46 <Diablo-D3> does
221 2010-12-30 03:03:47 <Diablo-D3> well
222 2010-12-30 03:03:51 <Diablo-D3> see nblocks?
223 2010-12-30 03:03:52 <Diablo-D3> thats 16 bytes
224 2010-12-30 03:03:58 <Kiba> well, it seem that we're going to be FAMOUS :D
225 2010-12-30 03:03:59 <Diablo-D3> I feed it longs
226 2010-12-30 03:04:00 <Diablo-D3> ergo
227 2010-12-30 03:04:06 <Diablo-D3> there can only be two choices
228 2010-12-30 03:04:08 <BoBeR> who is we
229 2010-12-30 03:04:09 <ArtForz> well, what happens if length is a multiple of 16 bytes?
230 2010-12-30 03:04:09 <Kiba> well, at least bitcoin
231 2010-12-30 03:04:24 <Diablo-D3> ArtForz: case 1.
232 2010-12-30 03:04:27 jyaworski has quit (Ping timeout: 240 seconds)
233 2010-12-30 03:04:43 <ArtForz> yours seem to process all blocks then in the main loop AND do another xor
234 2010-12-30 03:05:25 <ArtForz> notice no case 0 in the switch
235 2010-12-30 03:05:33 <Diablo-D3> hrm
236 2010-12-30 03:05:35 <Diablo-D3> rutrow
237 2010-12-30 03:05:39 <Diablo-D3> so theres only one case
238 2010-12-30 03:05:45 <Diablo-D3> if its odd.
239 2010-12-30 03:06:20 <Diablo-D3> uint8_t tail is set to the last set of bytes that is not contained in a block of 16 bytes
240 2010-12-30 03:06:33 <ArtForz> yeah, it's ... weird
241 2010-12-30 03:06:34 <Diablo-D3> it then pads them to the nearest long
242 2010-12-30 03:06:39 <Diablo-D3> and then does its shit
243 2010-12-30 03:06:51 <Diablo-D3> but if Im only using multiples of 8
244 2010-12-30 03:06:54 <Diablo-D3> there can only be two cases
245 2010-12-30 03:07:18 <ArtForz> yeah
246 2010-12-30 03:07:21 <Diablo-D3> and theres no case 0
247 2010-12-30 03:07:36 <ArtForz> yes, because it doesnt do the tail-processing if data is a multiple of 16 bytes
248 2010-12-30 03:07:45 <Diablo-D3> yeah exactly
249 2010-12-30 03:07:54 <ArtForz> wait a second
250 2010-12-30 03:08:44 <ArtForz> is that C for Le or Be system?
251 2010-12-30 03:09:01 <Diablo-D3> uh
252 2010-12-30 03:09:03 <Diablo-D3> art problem
253 2010-12-30 03:09:07 <Diablo-D3> 16 & 15 = 1.
254 2010-12-30 03:09:16 <ArtForz> wtf
255 2010-12-30 03:09:44 <ArtForz> >>> 16 & 15
256 2010-12-30 03:09:44 <ArtForz> 0
257 2010-12-30 03:09:56 <ArtForz> this fucking code is fucking stupid
258 2010-12-30 03:10:02 <BoBeR> 16 & 16 = 32
259 2010-12-30 03:10:08 <ArtForz> it's just "0-pad to next multiple of 16 bytes"
260 2010-12-30 03:10:09 ciuciu has quit (Ping timeout: 240 seconds)
261 2010-12-30 03:10:09 <BoBeR> dunno what you guys are doing wrong
262 2010-12-30 03:10:18 <Diablo-D3> ArtForz: exactly
263 2010-12-30 03:10:25 <tcatm> What core are you talking about?
264 2010-12-30 03:11:08 <ArtForz> so your code is also kinda inefficient
265 2010-12-30 03:11:10 <Diablo-D3> ArtForz: so the ONLY case is if odd set k1 to the left over long, right?
266 2010-12-30 03:11:18 <ArtForz> yep
267 2010-12-30 03:11:23 <Diablo-D3> k1 and not k2, right?
268 2010-12-30 03:11:31 <ArtForz> k2 stays 0 in that case
269 2010-12-30 03:12:13 <ArtForz> so for each pair of longs do a round, if you have a odd number do a k1= last long; k2 = 0; one more bmix round;
270 2010-12-30 03:12:20 larsig has quit (Read error: Operation timed out)
271 2010-12-30 03:12:26 <Diablo-D3> yes
272 2010-12-30 03:12:30 <Diablo-D3> which means
273 2010-12-30 03:12:38 <Diablo-D3> I I had my original code backwards
274 2010-12-30 03:12:50 larsig has joined
275 2010-12-30 03:13:00 <ArtForz> and the last bmix belongs in the if(odd)
276 2010-12-30 03:13:05 <Diablo-D3> yes
277 2010-12-30 03:13:28 <ArtForz> also your code seems to only allow length multiple of 8
278 2010-12-30 03:13:50 <ArtForz> len in the orig code can be anything
279 2010-12-30 03:13:59 <ArtForz> the h2 ^= len part
280 2010-12-30 03:14:09 <ArtForz> while yours does h2 ^= num_blocks * 8
281 2010-12-30 03:14:16 <ArtForz> which isnt quite the same
282 2010-12-30 03:15:28 <ArtForz> if I hash 1 byte, len is 1 in the C code, it ends up 8 in yours
283 2010-12-30 03:16:38 <Diablo-D3> yes, I only allow multiple of 8 for a reason
284 2010-12-30 03:16:44 <Diablo-D3> I cant do the casting tricks
285 2010-12-30 03:16:53 <Diablo-D3> I cant cast 4 bytes into a long
286 2010-12-30 03:16:58 <Diablo-D3> without a temporary copy and shifting
287 2010-12-30 03:16:59 <ArtForz> 8 bytes
288 2010-12-30 03:17:09 <Diablo-D3> er 8
289 2010-12-30 03:17:17 <ArtForz> yeah
290 2010-12-30 03:17:41 <ArtForz> and Ithink your h assigment indexes are wrong at the top
291 2010-12-30 03:18:07 <Diablo-D3> yes should be 0 and 1
292 2010-12-30 03:18:20 <ArtForz> otehrwise yep, shoudl be equiv
293 2010-12-30 03:19:39 mtgox has quit (Ping timeout: 240 seconds)
294 2010-12-30 03:26:34 <da2ce7> woot! http://www.youtube.com/watch?v=HEFMAP0mTvY&feature=player_embedded
295 2010-12-30 03:27:27 <Diablo-D3> da2ce7: you do know what the failure with the ps3 was, right?
296 2010-12-30 03:27:40 <Diablo-D3> ibm signed their root key, the one burned into the chip handling DRM
297 2010-12-30 03:27:42 <Diablo-D3> on
298 2010-12-30 03:27:50 <Diablo-D3> a debian system with the famous broken openssl
299 2010-12-30 03:28:11 <EvanR> lol
300 2010-12-30 03:28:16 <ArtForz> soudns wrong
301 2010-12-30 03:28:28 <Diablo-D3> I know the guy who actually discovered this
302 2010-12-30 03:28:55 <Diablo-D3> evanR: and now, every single ps3 is broken
303 2010-12-30 03:29:07 <Diablo-D3> wanna bypass drm and root your shit? replace the core firmware image.
304 2010-12-30 03:29:14 <ArtForz> errr... the PS3 attack iirc is something different
305 2010-12-30 03:29:25 <Diablo-D3> ArtForz: nope, this is the ps3 attack
306 2010-12-30 03:29:27 <ArtForz> sony fucked up signing the GAMES
307 2010-12-30 03:29:32 <ArtForz> the key would be preferctly fine
308 2010-12-30 03:29:35 <Diablo-D3> none of the news is repeating it right
309 2010-12-30 03:29:44 <Diablo-D3> ArtForz: nope, the _master key_ is now known
310 2010-12-30 03:29:47 <ArtForz> not a weak key, they used the same secret random never-to-be-used-twice nonce ... twice
311 2010-12-30 03:29:49 <EvanR> dont they get these keys with similar technology to star trek the next generation picard simply asking data to scan the device with sensors
312 2010-12-30 03:30:02 <Diablo-D3> evanR: no
313 2010-12-30 03:30:06 <ArtForz> which allows you to recover the privatze key from pubkey, the 2 hashes and signatiures
314 2010-12-30 03:30:24 <Diablo-D3> ArtForz: now anyone can replace the bootloader
315 2010-12-30 03:30:29 <ArtForz> yep
316 2010-12-30 03:30:30 <Diablo-D3> on the ps3, the bootloader loads the firmware.
317 2010-12-30 03:30:31 <ArtForz> read up on how DSA works
318 2010-12-30 03:30:44 <ArtForz> its not a broken private key, they fucking USED IT WRONG
319 2010-12-30 03:30:50 <Diablo-D3> ArtForz: nope
320 2010-12-30 03:31:02 <Diablo-D3> the news is covering it wrong
321 2010-12-30 03:31:11 <Diablo-D3> IBM used a broken debian openssl release.
322 2010-12-30 03:31:20 <EvanR> if you dont want someone to have something, dont give it to them
323 2010-12-30 03:31:29 <ArtForz> I'm not going to watch the 27C3 talk again just to hand you the timestamp
324 2010-12-30 03:31:39 <Diablo-D3> ArtForz: the guys giving the talk didnt even discover the flaw
325 2010-12-30 03:32:03 <ArtForz> still has NOITHING TO DO AT ALL with debian broken OSSL
326 2010-12-30 03:32:21 <ArtForz> they didnt use one-out-of-32768 possible noces to create the sigs, theay all use THE SAME NONCE
327 2010-12-30 03:32:23 <Diablo-D3> you mean except for the fact the key was weak and they brute forced it in almost none of the time?
328 2010-12-30 03:32:55 <Diablo-D3> you can now produce your own keys that any unhacked ps3 will accept.
329 2010-12-30 03:33:03 <ArtForz> yes
330 2010-12-30 03:33:10 <ArtForz> and it's not IBMs or debians fault
331 2010-12-30 03:33:17 <Diablo-D3> its not debian's, no
332 2010-12-30 03:33:19 <Diablo-D3> but it is IBM's
333 2010-12-30 03:33:22 <ArtForz> it's sonys fault for FUCKING USING DSA WRONG
334 2010-12-30 03:33:25 <Diablo-D3> sony didnt create the key.
335 2010-12-30 03:33:45 <ArtForz> THE KEY WAS FINE AND SECRET UNTI LSONY SIGNED 2 HASHES WITH THEN SAME FUCKING NONCE
336 2010-12-30 03:34:00 <Diablo-D3> ArtForz: Im not sure why you keep repeating that
337 2010-12-30 03:34:03 <Diablo-D3> thats not what happened
338 2010-12-30 03:34:15 <Diablo-D3> either that, or they're using it as a cover for the real deal
339 2010-12-30 03:34:49 <ArtForz> remember how [Ec]DSA works?
340 2010-12-30 03:34:50 <ArtForz> Calculate e = HASH(m), where HASH is a cryptographic hash function, such as SHA-1, and let z be the Ln leftmost bits of e.
341 2010-12-30 03:34:51 <ArtForz> Select a random integer k from [1,n ? 1].
342 2010-12-30 03:34:51 <ArtForz> Calculate r = x1(mod n), where (x1,y1) = kG. If r = 0, go back to step 2.
343 2010-12-30 03:34:51 <ArtForz> Calculate s = k ? 1(z + rdA)(mod n). If s = 0, go back to step 2.
344 2010-12-30 03:34:51 <ArtForz> The signature is the pair (r,s).
345 2010-12-30 03:35:11 <ArtForz> guess what happens when you use the same k for 2 different m
346 2010-12-30 03:35:14 <Diablo-D3> ArtForz: either that, or they produced two of the same nonce, randomly.
347 2010-12-30 03:35:25 <Diablo-D3> who knows, maybe they used all of them on the broken openssl already
348 2010-12-30 03:35:26 <ArtForz> for each and every game?
349 2010-12-30 03:35:39 * Diablo-D3 shrugs
350 2010-12-30 03:35:44 <ArtForz> every game exe is signed, all use the *same* k
351 2010-12-30 03:35:51 <Diablo-D3> ArtForz: that sounds wrong
352 2010-12-30 03:35:54 <Diablo-D3> sony isnt that stupid
353 2010-12-30 03:36:09 <ArtForz> thats why they call it EPIC FAIL in the 27C3 presentation
354 2010-12-30 03:36:22 <da2ce7> lol
355 2010-12-30 03:36:54 <Diablo-D3> except the EPIC FAIL is you require zero games to crack the key.
356 2010-12-30 03:37:04 <Kiba> large amount of tweeting about read write web but it seem to fail to have much impact on the forum
357 2010-12-30 03:37:25 <ArtForz> okay, then pray tell the PRNG seed used on debian to generate that key?
358 2010-12-30 03:37:30 <Cusipzzz> full of sound and fury, signifying nothing
359 2010-12-30 03:38:03 <Diablo-D3> ArtForz: yes.
360 2010-12-30 03:39:47 <Diablo-D3> ArtForz: the rmsk in the cell is broken. period.
361 2010-12-30 03:39:52 <Diablo-D3> its an ibm problem, not a sony problem
362 2010-12-30 03:40:04 <Diablo-D3> that means
363 2010-12-30 03:40:20 <ArtForz> whatever dude
364 2010-12-30 03:40:22 <Diablo-D3> I can take the ibm bootrom, and shove my foot up it's ass, wiggle it around, and shit works
365 2010-12-30 03:40:49 <ArtForz> sony disclosing a perfectly safe key by fucking using DSA wrong != IBM generating a unsafe key
366 2010-12-30 03:41:11 <Diablo-D3> yes, and the media needs to learn how to report the issue
367 2010-12-30 03:43:22 sgtstein has joined
368 2010-12-30 03:44:59 <Diablo-D3> [10:36:14] <nenolod> Diablo-D3: sony didn't fucking do it
369 2010-12-30 03:44:59 <Diablo-D3> [10:36:16] <nenolod> Diablo-D3: IBM did it
370 2010-12-30 03:44:59 <Diablo-D3> [10:36:21] <nenolod> Diablo-D3: take it up with IBM
371 2010-12-30 03:44:59 <Diablo-D3> [10:36:43] <nenolod> Diablo-D3: the media are a bunch of retards, it's the RMSK that is compromised
372 2010-12-30 03:45:10 <ArtForz> *headdesk*
373 2010-12-30 03:45:13 <Diablo-D3> [10:36:49] <nenolod> Diablo-D3: sony's keys are completely irrelevant
374 2010-12-30 03:45:13 <Diablo-D3> [10:36:54] <nenolod> Diablo-D3: it's burned into the damn CPU
375 2010-12-30 03:45:13 <Diablo-D3> [10:37:28] <nenolod> Diablo-D3: this has nothing to do with SCEI's keys
376 2010-12-30 03:45:13 <Diablo-D3> [10:37:49] <nenolod> Diablo-D3: SCEI does not even directly sign every game being deployed on PS3
377 2010-12-30 03:45:13 <Diablo-D3> [10:37:59] <nenolod> Diablo-D3: the keychain looks like this
378 2010-12-30 03:45:13 <Diablo-D3> [10:38:50] <nenolod> Diablo-D3: RMSK (root master signing key; cell bootrom) -> bootromverify -> lv0 (SCEI MSK + SCEA/SCEJ/SCEE RSK) -> lv1 (subkey) -> lv2 (subkey)
379 2010-12-30 03:45:24 <Diablo-D3> [10:39:18] <nenolod> Diablo-D3: -> game (Vendor VSK signed by SCEI MSK || SCEA/SCEJ/SCEE RSK)
380 2010-12-30 03:46:15 <Diablo-D3> [10:43:17] <nenolod> the thing is
381 2010-12-30 03:46:15 <Diablo-D3> [10:43:24] <nenolod> you can't replace lv0 without a signature from IBM
382 2010-12-30 03:48:53 <Diablo-D3> ArtForz: so, there.
383 2010-12-30 03:49:36 <ArtForz> nenolod, the guy who added and removed 1Gh/s without the network hashrate changing a single bit?
384 2010-12-30 03:49:51 <ArtForz> and that back when the total network was ~500Mh/s ?
385 2010-12-30 03:50:29 <Diablo-D3> nenolod, the guy whos written more software on the ps3 than you, yes.
386 2010-12-30 03:51:49 <Diablo-D3> ArtForz: this means I can replace sony's firmware
387 2010-12-30 03:51:51 <Diablo-D3> with my own
388 2010-12-30 03:51:56 <Diablo-D3> and they cant stop me
389 2010-12-30 03:52:57 <donpdonp> interesting 27C3 video on that ps3 stuff.
390 2010-12-30 03:53:57 <donpdonp> they said it was two signatures generated with the same nonce and because of that, the private key was revealed. the problem did not seem to be the key itself.
391 2010-12-30 03:54:19 <Diablo-D3> donpdonp: which is bs, sony can revoke master keys afiak
392 2010-12-30 03:54:58 <donpdonp> the talk did mention revokation lists, but it didnt seem to be a problem for them.
393 2010-12-30 03:55:11 <Diablo-D3> if its the cell rmsk, you CANT revoke that
394 2010-12-30 03:55:16 <Diablo-D3> its burned into the cell.
395 2010-12-30 04:07:59 <da2ce7> we need to make 'ps3 bitcoin' :D
396 2010-12-30 04:08:11 <Diablo-D3> not really
397 2010-12-30 04:08:27 <Diablo-D3> ibm has an opencl compiler for cell spes
398 2010-12-30 04:08:35 <ArtForz> even using all SPEs cell is less efficient than a PhenomII
399 2010-12-30 04:08:42 <Diablo-D3> yup
400 2010-12-30 04:09:01 <da2ce7> no, for transaction :)
401 2010-12-30 04:09:18 <da2ce7> you could buy homebrew games with bitcoin.
402 2010-12-30 04:13:27 <da2ce7> it should be easy to extract a proportion of your private bitcoin keys, and make a new wallet. You could use the smaller wallet for every day use.
403 2010-12-30 04:13:54 <da2ce7> but still have a backup in you large wallet.
404 2010-12-30 04:14:43 <Diablo-D3> heh
405 2010-12-30 04:14:52 <Diablo-D3> da2ce7: just send coins to yourself on a different client
406 2010-12-30 04:16:03 Cusipzzz has quit ()
407 2010-12-30 04:21:04 <da2ce7> but I need to keep track and backup two wallets then. It would be better if I could have a one wallet that is small, and on my laptop that I don't need to worry about.
408 2010-12-30 04:21:18 <da2ce7> My big wallet would be a super-set of my smaler ones.
409 2010-12-30 04:23:39 sloob has joined
410 2010-12-30 04:24:29 <sloob> Hello all, I was looking over the bitcoin documentation. Is it correct to say that the specs are derived from the code and not vice versa?
411 2010-12-30 04:24:39 <da2ce7> yep
412 2010-12-30 04:25:02 <sloob> Is this because it's still in flux or just because the original creator hasn't documented it?
413 2010-12-30 04:25:20 <da2ce7> The official specification is the c++ code
414 2010-12-30 04:25:24 <sloob> I guess in a roundabout way, is it expected to change soon
415 2010-12-30 04:25:30 <sloob> is my question
416 2010-12-30 04:25:35 <sloob> okeydoke.
417 2010-12-30 04:25:44 <da2ce7> hopefully oneday it will be the other way arround.
418 2010-12-30 04:25:55 <Diablo-D3> its both
419 2010-12-30 04:25:59 <Diablo-D3> the spec spawned the code
420 2010-12-30 04:26:06 <Diablo-D3> then satoshi realized the spec was bullshit
421 2010-12-30 04:26:11 <Diablo-D3> he needs to update the spec
422 2010-12-30 04:26:14 * Kiba watch another artist draw live
423 2010-12-30 04:26:23 <sloob> ok
424 2010-12-30 04:26:26 <Kiba> a pro artist
425 2010-12-30 04:26:44 <sloob> I was interested in trying to implement a wallet in Java
426 2010-12-30 04:26:56 <sloob> So I'm looking it over
427 2010-12-30 04:27:03 <Diablo-D3> sloob: oh thats easy, its just a bdb 4.7 database
428 2010-12-30 04:27:14 <sloob> OK
429 2010-12-30 04:27:26 <Diablo-D3> so use bdb (4.7 only) in java, and its done
430 2010-12-30 04:27:42 <sloob> I have been messing around with JavaCard, and was like, this would be a cool, secure place to put a wallet
431 2010-12-30 04:27:56 <sloob> I don't think the storage space is large enough tho
432 2010-12-30 04:27:59 <Diablo-D3> you mean except for ... yeah
433 2010-12-30 04:28:03 <Diablo-D3> not enough storage space
434 2010-12-30 04:28:28 <sloob> if they were as cheap as promised, you could hand off hard transactions in one...but they aren't
435 2010-12-30 04:29:53 <sloob> My first goal is just to get a wallet and transactions going in a Java app so it can be used for J2EE, then I want to look at ways to transfer bitcoin outside a computer
436 2010-12-30 04:30:19 <sloob> like a gift certificate or something
437 2010-12-30 04:31:12 <nanotube> sloob: iirc there's a java bitcoin client project out there on sourceforge somewhere.
438 2010-12-30 04:31:29 <sloob> Thanks, I will take a look at that
439 2010-12-30 04:31:33 <nanotube> ;;sl java bitcoin client
440 2010-12-30 04:31:34 <gribble> http://sourceforge.net/projects/bitcoin-client/ | Sep 30, 2010 ... Get Java Bitcoin Client at SourceForge.net. Fast, secure and free downloads from the largest Open Source applications and software ...
441 2010-12-30 04:31:38 <nanotube> yea that looks like it.
442 2010-12-30 04:32:15 <Kiba> with the constant barrage of tweeting traffic, you would think the forum traffic would spike
443 2010-12-30 04:32:16 <Kiba> by now
444 2010-12-30 04:32:39 <sloob> I've known about bitcoin for quite some time, mined a few coins, but have been lazy about getting on board doing actual projects with it
445 2010-12-30 04:32:54 <nanotube> sloob: hope you still have your mined coins. :)
446 2010-12-30 04:33:01 <sloob> I'm just here because I'm unemployed and have free time in the evenings :-/
447 2010-12-30 04:33:35 <nanotube> Kiba: hitting the 'tweet' button is a lot less effort-intensive than signing up to a forum and actually posting something constructive.
448 2010-12-30 04:33:42 <nanotube> so not really much of a surprise. :)
449 2010-12-30 04:33:53 <sloob> makes sense
450 2010-12-30 04:34:05 <nanotube> sloob: if you are unemployed, why don't you have free time during the day, too? :)
451 2010-12-30 04:34:06 <sloob> I would have figured the few times that Slashdot has mentioned it would pull some people in
452 2010-12-30 04:34:15 <sloob> haha, job hunting is a job itself
453 2010-12-30 04:34:19 <nanotube> hehe ic
454 2010-12-30 04:34:20 <Kiba> nanotube: just visiting the forum..due
455 2010-12-30 04:34:24 <Kiba> s/due/dude
456 2010-12-30 04:34:27 <Kiba> no need to register
457 2010-12-30 04:34:33 <nanotube> Kiba: is there a visitor count?
458 2010-12-30 04:35:03 <nanotube> oh yes, 17 guests, 22 users
459 2010-12-30 04:35:14 <nanotube> how does the 17 guests compare to usual?
460 2010-12-30 04:35:23 <Kiba> don't know
461 2010-12-30 04:35:31 <Kiba> but usually quests are more than users I thik
462 2010-12-30 04:35:33 <Kiba> err
463 2010-12-30 04:35:34 <Kiba> guests
464 2010-12-30 04:37:57 <nanotube> mm
465 2010-12-30 04:47:37 <sloob> so, there's an attempt to get bitcoin in the zeitgeist right now?
466 2010-12-30 04:47:59 <Kiba> sometime deliberate
467 2010-12-30 04:48:00 <Kiba> sometime not
468 2010-12-30 04:48:20 <Kiba> like bitcoin meme hooking itself to the wikileak meme
469 2010-12-30 04:48:36 <Kiba> Bitcoin as a meme seem to be quite viral
470 2010-12-30 04:48:50 <sloob> it was never clear to me how that would work since inevitably you need some sort of payment processor or bank to do the conversion
471 2010-12-30 04:49:02 <sloob> I mean whomever is handling it is gonna use a bank
472 2010-12-30 04:49:05 <Kiba> what conversion?
473 2010-12-30 04:49:31 <sloob> like bitcoin to dollar
474 2010-12-30 04:50:09 <sloob> I bought a few bitcoin on paypal once and the place got shut down by paypal
475 2010-12-30 04:51:04 <Kiba> hmm
476 2010-12-30 04:51:05 <Kiba> mtgox?
477 2010-12-30 04:51:07 <Kiba> mtgox isn't dead
478 2010-12-30 04:51:11 <Kiba> it's alive!
479 2010-12-30 04:51:22 <sloob> are they back to doing paypal transactions again?
480 2010-12-30 04:51:25 <Kiba> no
481 2010-12-30 04:51:32 <Kiba> you have to buy LR USD instead
482 2010-12-30 04:51:39 <ArtForz> mtgox switched to LibertyReserve
483 2010-12-30 04:51:41 <sloob> ok
484 2010-12-30 04:51:44 <Kiba> but you can buy with paypal dollars on #bitcoin-otc
485 2010-12-30 04:51:51 <Kiba> to get it transfered to mtgox
486 2010-12-30 04:51:58 <sloob> OK, thanks
487 2010-12-30 04:52:00 <sloob> that's super useful
488 2010-12-30 04:52:16 <Kiba> you have to deal with individual users though
489 2010-12-30 04:53:06 <sloob> no prob
490 2010-12-30 04:53:12 <sloob> how is fraud mitigated
491 2010-12-30 04:53:24 <sloob> or does it just use trust
492 2010-12-30 04:53:50 <nanotube> yes
493 2010-12-30 04:57:48 <Kiba> the bitcoin economy is about 1.5 million dollars
494 2010-12-30 04:57:51 <Kiba> pretty damn small
495 2010-12-30 04:58:48 <nanotube> but also pretty decent, for something so new. :)
496 2010-12-30 04:58:57 <sloob> yeah
497 2010-12-30 04:59:27 <sloob> I have been thinking a lot about how the bitcoin network would be attacked
498 2010-12-30 04:59:39 marioxcc has joined
499 2010-12-30 04:59:40 <sloob> like, if someone wanted to disrupt the economy
500 2010-12-30 05:00:11 <sloob> there are ways around it, but if the IRC channel negotiating connections went down, I Assume that would be catastrophic
501 2010-12-30 05:00:21 <Kiba> no
502 2010-12-30 05:00:31 <Kiba> each bitcoin client have a list of backup node to connect
503 2010-12-30 05:01:09 <sloob> Can you get bootstrapped back into the network via one other client
504 2010-12-30 05:01:22 <Kiba> and each client collect their own addresses to connect to I think
505 2010-12-30 05:01:49 <sloob> like, embed a magnet link in a webpage or tweet or something, people connect to it and it gets up and going again
506 2010-12-30 05:03:24 <Kiba> whatever you come up with, I am sure it's not new
507 2010-12-30 05:03:29 <Kiba> try searching the forum for once
508 2010-12-30 05:03:49 <Kiba> MT`AwAy: did ya answer my interview question yet?
509 2010-12-30 05:03:51 <sloob> I'm not inventing anything, I was just asking for confirmation because it wasn't clear to me
510 2010-12-30 05:04:04 <sloob> I read the section someone wrote on magnet links
511 2010-12-30 05:04:23 <ArtForz> yes, if you get naother doens ip via some other meands you can just -addnode it and your node will bottstrap fro mthere
512 2010-12-30 05:04:34 <nanotube> sloob: yes if you can connect to one client, you'll get addresses of a bunch of others.
513 2010-12-30 05:04:42 <sloob> Thanks
514 2010-12-30 05:04:46 <marioxcc> sloob: what's your question?
515 2010-12-30 05:04:48 <nanotube> bah, beaten by ArtForz ;)
516 2010-12-30 05:04:58 <sloob> I'm cool, it's answered
517 2010-12-30 05:05:10 <marioxcc> ok
518 2010-12-30 05:05:13 <Kiba> No, I beat all of ya
519 2010-12-30 05:05:13 <nanotube> there's a list of fallback nodes on the wiki
520 2010-12-30 05:06:00 <marioxcc> yeah, point to point networks are very difficult to tear down
521 2010-12-30 05:06:04 <marioxcc> that's the point
522 2010-12-30 05:06:12 <marioxcc> that's how bitcoin has been designed
523 2010-12-30 05:06:21 <marioxcc> and that's how it will behave
524 2010-12-30 05:06:22 <marioxcc> :)
525 2010-12-30 05:07:15 <sloob> I've read the specs, gonna read the source code next.
526 2010-12-30 05:07:54 <marioxcc> sloob: what specs?
527 2010-12-30 05:09:02 <sloob> the reverse engineered specs at http://www.bitcoin.org/wiki/doku.php?id=bitcoins_draft_spec_0_0_1
528 2010-12-30 05:09:15 <sloob> and https://github.com/bitcoin/netspec
529 2010-12-30 05:09:39 <Kiba> I thought spec moved to en.bitocin.it the new wiki?
530 2010-12-30 05:09:46 <sloob> wa
531 2010-12-30 05:09:49 <marioxcc> sloob: that's not reverse enginnering
532 2010-12-30 05:10:02 <marioxcc> if we had only the builds or a obfuscated C code
533 2010-12-30 05:10:09 <marioxcc> then it would be reverse enginner to make a specification
534 2010-12-30 05:10:32 <marioxcc> please don't missuse the term "reverse enginnering" it can lead to very bad misunderstanding
535 2010-12-30 05:11:03 <marioxcc> someone which sees a statemen such as "there is a revese enginnering spec of bitcoin in ..." may thought the bitcoin client and protocol are proprietary
536 2010-12-30 05:11:06 <marioxcc> that's not the case
537 2010-12-30 05:11:22 <marioxcc> thanks in advance :9
538 2010-12-30 05:11:25 <marioxcc> oh
539 2010-12-30 05:11:28 <marioxcc> :)
540 2010-12-30 05:11:30 <marioxcc> hehe
541 2010-12-30 05:12:11 <sloob> It's not proprietary, but the code is the authority, right? as opposed to being programmed from a spec first
542 2010-12-30 05:12:26 <ArtForz> kinda
543 2010-12-30 05:12:37 <sloob> I'm not trying to offend or anything, pls don't misunderstand.
544 2010-12-30 05:12:48 <marioxcc> sloob: i feel offended
545 2010-12-30 05:12:50 <sloob> I won't use the term reverse engineer though, point taken
546 2010-12-30 05:13:00 <marioxcc> that's why i gave an explanion :)
547 2010-12-30 05:13:07 <marioxcc> i appreciate your understanding
548 2010-12-30 05:13:12 <sloob> np
549 2010-12-30 05:13:30 <ArtForz> the bitcoin paper pretty much defines how the underlying crypto works, implementation details are "documented" by haveing a reference implementation
550 2010-12-30 05:13:31 * Kiba dreams of doing many thing
551 2010-12-30 05:13:35 <Kiba> writing my own graphic novel
552 2010-12-30 05:13:36 <ArtForz> so yeha, it's mostly spec-by-code
553 2010-12-30 05:13:42 <Kiba> inventing cool stuff
554 2010-12-30 05:13:42 xorrbit has left ()
555 2010-12-30 05:13:46 <Kiba> drawing my own grpahic novel
556 2010-12-30 05:14:01 <sloob> Kiba, do you have samples or anything
557 2010-12-30 05:14:20 <nanotube> ;;bc,wiki protocol spec
558 2010-12-30 05:14:20 <gribble> Google found nothing.
559 2010-12-30 05:14:22 <nanotube> hah
560 2010-12-30 05:14:28 <Kiba> http://www.bitcoinservice.co.uk/files/68
561 2010-12-30 05:14:28 <marioxcc> Kiba: i would love to see a post-quantum public key encryption scheme
562 2010-12-30 05:14:59 <nanotube> https://en.bitcoin.it/wiki/Protocol_Specification
563 2010-12-30 05:15:30 <sloob> nice. I have a number of artist friends, a few professional
564 2010-12-30 05:15:51 <sloob> nanotube: thanks!!!
565 2010-12-30 05:16:00 <marioxcc> it seems a bit like anime
566 2010-12-30 05:16:02 <marioxcc> hehe
567 2010-12-30 05:16:48 <nanotube> sloob: :)
568 2010-12-30 05:17:45 <Kiba> sloob: I am a noob to the whole art thing
569 2010-12-30 05:18:01 <Kiba> my competency as an artist literally have to grow
570 2010-12-30 05:18:15 <sloob> I can give useful advice in this area. I used to be in the fine arts but switched to computers for the money
571 2010-12-30 05:18:25 <sloob> draw all the time. Draw everything. The end.
572 2010-12-30 05:18:43 <Kiba> I draw too few
573 2010-12-30 05:18:52 <Kiba> because it takes forever for me to do an art
574 2010-12-30 05:19:05 <marioxcc> i can say the same of programming and me
575 2010-12-30 05:19:08 <Kiba> but I learned something...
576 2010-12-30 05:19:25 <marioxcc> of course, any art will take practice and time to learn and master
577 2010-12-30 05:20:09 <Kiba> I will be much faster in the next artwork
578 2010-12-30 05:20:17 <Kiba> because I am not scanning and then cleaning up
579 2010-12-30 05:20:22 <Kiba> which take horrific amount of time
580 2010-12-30 05:20:32 <sloob> I still do some illustration every once in a while, and though I am in many ways not as good as I used to be, I can visualize better in my head now.
581 2010-12-30 05:21:01 <sloob> if I could only take the time I put into it and the mental abilities I have now, and put them together
582 2010-12-30 05:21:12 <Kiba> learning computer art technique take time to discover
583 2010-12-30 05:21:38 <Kiba> with my programming skill...
584 2010-12-30 05:21:55 <Kiba> I am sure I can create artistic effects
585 2010-12-30 05:21:57 <Kiba> or template
586 2010-12-30 05:22:14 <sloob> If you haven't read Understanding Comics, I highly recommend it. It's starting to get fashionable to bash it, but it's one of only two books worth a carp on the subject
587 2010-12-30 05:24:30 <sloob> also, draw for yourself but be able to take criticism
588 2010-12-30 05:25:51 <sloob> that's all I have to say about that...I was at comic con a year ago shmoozing, even got invited to a party. Nearly ran into Johnen Vasquez. I mena, physically ran into.
589 2010-12-30 05:26:23 * Kiba wants to do too many cool thing
590 2010-12-30 05:27:12 <sloob> like what
591 2010-12-30 05:27:20 epicurus has joined
592 2010-12-30 05:27:58 <Kiba> write programs, invent self replicating robots, build company, make arts, write novels
593 2010-12-30 05:27:59 <Kiba> etc
594 2010-12-30 05:28:02 <Kiba> well
595 2010-12-30 05:28:19 <marioxcc> anskwer for P=NP?
596 2010-12-30 05:28:22 <marioxcc> *answer
597 2010-12-30 05:28:24 <sloob> lol
598 2010-12-30 05:28:25 <Kiba> I have to figure out how to compose all that into something that build me over time instead of splintering my focus in various area
599 2010-12-30 05:28:35 <sloob> yeah, that's a problem
600 2010-12-30 05:28:39 <sloob> I've not solved it yet
601 2010-12-30 05:28:46 <BoBeR> good cheap beer go
602 2010-12-30 05:28:51 <marioxcc> sloob: sure, not ever the gratest computer scientest have
603 2010-12-30 05:29:26 <sloob> oh, I Was meaning the not splintering
604 2010-12-30 05:29:35 <sloob> focus
605 2010-12-30 05:29:56 <sloob> my life is a graveyard of half-finished projects
606 2010-12-30 05:30:20 <ArtForz> yep, I know that very well
607 2010-12-30 05:30:51 <marioxcc> ah
608 2010-12-30 05:31:42 BCBot has quit (Ping timeout: 240 seconds)
609 2010-12-30 05:32:24 <Kiba> my programming hobby is full of these projects too
610 2010-12-30 05:33:17 <sloob> I wrote a whole direct-connect server in Java one time, but it was way too slow, so it got abandoned
611 2010-12-30 05:33:55 <sloob> for a while was writing a sort of MUD that let you create the world while you were in it.
612 2010-12-30 05:34:31 <sloob> that was pretty cool, but there was no way to plug every possible way for someone to abuse it and crash it
613 2010-12-30 05:35:10 <sloob> someday I'll go back to that, and put evolutionary NPCs in it
614 2010-12-30 05:38:17 BCBot has joined
615 2010-12-30 05:50:32 sgtstein has left ()
616 2010-12-30 05:53:37 remmy_ has joined
617 2010-12-30 05:55:15 AAA_awright_ has joined
618 2010-12-30 05:56:04 paul0 has quit (Quit: paul0)
619 2010-12-30 05:57:50 AAA_awright has quit (Ping timeout: 240 seconds)
620 2010-12-30 06:04:45 AAA_awright_ is now known as AAA_awright
621 2010-12-30 06:17:10 marioxcc is now known as marioxcc-AFK
622 2010-12-30 06:20:44 marioxcc-AFK is now known as marioxcc
623 2010-12-30 06:24:08 marioxcc is now known as marioxcc-AFK
624 2010-12-30 06:28:33 acous has quit (Quit: Computer has gone to sleep.)
625 2010-12-30 06:54:51 sgtstein has joined
626 2010-12-30 06:55:50 RazielZ has joined
627 2010-12-30 06:58:29 epicurus has quit (Quit: This computer has gone to sleep)
628 2010-12-30 07:07:04 larsivi has joined
629 2010-12-30 07:11:37 eureka^ has quit (Ping timeout: 250 seconds)
630 2010-12-30 07:12:22 ThomasV has joined
631 2010-12-30 07:13:40 BitcoinNewb has joined
632 2010-12-30 07:14:10 <BitcoinNewb> hello
633 2010-12-30 07:15:05 <nanotube> hey
634 2010-12-30 07:15:35 <BitcoinNewb> I just found out about bitcoin from Max kieser
635 2010-12-30 07:15:44 <nanotube> cool :)
636 2010-12-30 07:15:53 <BitcoinNewb> I was reading the wiki,and I still dont get it
637 2010-12-30 07:16:14 <BitcoinNewb> Do you spend this money on things?
638 2010-12-30 07:16:21 <BitcoinNewb> How do you even start
639 2010-12-30 07:16:27 <BitcoinNewb> investing
640 2010-12-30 07:16:37 <nanotube> you can exchange for other currencies
641 2010-12-30 07:16:49 <nanotube> or you can help the network produce blocks and get reward
642 2010-12-30 07:16:51 <BitcoinNewb> ok so I would buy some with dollars
643 2010-12-30 07:16:58 <nanotube> or you can sell stuff for bitcoin
644 2010-12-30 07:17:04 <nanotube> yes you can. see bitcoin.org/trade
645 2010-12-30 07:17:04 <BitcoinNewb> how do you do that?
646 2010-12-30 07:17:08 <nanotube> there is a list of exchanges
647 2010-12-30 07:17:17 <BitcoinNewb> ok
648 2010-12-30 07:17:42 <BitcoinNewb> So I would run a program on my computer and assist with the transactions?
649 2010-12-30 07:17:54 <BoBeR> yes
650 2010-12-30 07:18:00 <BoBeR> the bitcoin application
651 2010-12-30 07:18:03 <BitcoinNewb> like folding
652 2010-12-30 07:18:04 <BitcoinNewb> oh
653 2010-12-30 07:18:13 <nanotube> yes, kinda like folding. :)
654 2010-12-30 07:18:18 <BitcoinNewb> I thought that program was just for transfering coins
655 2010-12-30 07:18:34 <BitcoinNewb> I used to let my ps3 do that, untill it took a shit
656 2010-12-30 07:18:37 genjix has joined
657 2010-12-30 07:18:38 <nanotube> you can turn off block generation, and just use it for transferring.
658 2010-12-30 07:18:40 <BitcoinNewb> go figure
659 2010-12-30 07:18:45 <BitcoinNewb> I C
660 2010-12-30 07:18:58 <genjix> MT`AwAy: hey can we make the font for the sourcecode snippets in the wiki larger?
661 2010-12-30 07:19:23 <BitcoinNewb> So do you reap real rewards for doing that or is it just drops in the bucket?
662 2010-12-30 07:19:28 <genjix> it's tiny... i can hardly see it.
663 2010-12-30 07:20:28 <BitcoinNewb> So to start contributing to I need to buy some bit coins first?
664 2010-12-30 07:20:39 <Kiba> no
665 2010-12-30 07:20:46 <BitcoinNewb> I would love to support this new type of decentralized currency
666 2010-12-30 07:20:46 <Kiba> you can sell stuff or services for bitcoin
667 2010-12-30 07:21:19 <BitcoinNewb> I fix computers. Thats my trade. I could ask for payment in bitcoins then?
668 2010-12-30 07:21:51 <nanotube> yes you could
669 2010-12-30 07:22:05 <genjix> ill sell you 40 btc if u want
670 2010-12-30 07:22:09 <BitcoinNewb> Is there a market place for me to offer up my services for bitcoins
671 2010-12-30 07:22:15 <genjix> foruns
672 2010-12-30 07:22:17 <genjix> m
673 2010-12-30 07:22:17 <genjix> s
674 2010-12-30 07:22:20 <BitcoinNewb> ahh
675 2010-12-30 07:22:29 * genjix pokes MT`AwAy
676 2010-12-30 07:22:29 <BitcoinNewb> Sure how much?
677 2010-12-30 07:22:47 <genjix> $13
678 2010-12-30 07:22:49 <nanotube> BitcoinNewb: see #bitcoin-otc for the trading channel.
679 2010-12-30 07:22:52 <BitcoinNewb> I need to make an account right
680 2010-12-30 07:22:57 <genjix> do you have pokerstars?
681 2010-12-30 07:23:06 <BitcoinNewb> no
682 2010-12-30 07:23:06 <Kiba> BitcoinNewb: it's unlikely that you will get customers who will pay you in bitcoin, but it's a start
683 2010-12-30 07:23:13 <BitcoinNewb> I dont play online poker
684 2010-12-30 07:23:18 <BitcoinNewb> lol
685 2010-12-30 07:23:21 <BitcoinNewb> true
686 2010-12-30 07:23:23 * Kiba pokes MT`AwAy for his own reason
687 2010-12-30 07:23:43 <BitcoinNewb> this stuff is really interesting
688 2010-12-30 07:23:43 <genjix> kk
689 2010-12-30 07:23:45 genjix has left ()
690 2010-12-30 07:23:57 <ThomasV> BitcoinNewb: http://www.bitcoin.org/trade
691 2010-12-30 07:24:00 darrob has quit (Disconnected by services)
692 2010-12-30 07:24:08 darrob has joined
693 2010-12-30 07:24:08 <BitcoinNewb> So would you do the transaction with paypal or something
694 2010-12-30 07:24:27 <BitcoinNewb> how do you convert the dollars into bitcoins
695 2010-12-30 07:24:43 <ThomasV> you just buy some
696 2010-12-30 07:24:52 <BitcoinNewb> where?
697 2010-12-30 07:24:57 <BitcoinNewb> with the program?
698 2010-12-30 07:25:04 <ThomasV> see mtgox.com
699 2010-12-30 07:25:11 <BitcoinNewb> ok
700 2010-12-30 07:25:16 <BitcoinNewb> brb, and thanks
701 2010-12-30 07:26:34 <BitcoinNewb> ohhhh
702 2010-12-30 07:26:44 <BitcoinNewb> duh
703 2010-12-30 07:26:54 <BitcoinNewb> you buy it on the bitcoin exchange like any other currency
704 2010-12-30 07:26:57 <nanotube> as i mentioned before... a bunch of exchange sites listed on /trade
705 2010-12-30 07:27:00 <nanotube> yep
706 2010-12-30 07:27:15 <BitcoinNewb> is this a safe investment yet?
707 2010-12-30 07:27:24 <nanotube> nothing is a safe investment.
708 2010-12-30 07:27:25 <BitcoinNewb> its only 2 years old
709 2010-12-30 07:27:33 <BitcoinNewb> gold and silver lol
710 2010-12-30 07:27:38 <BitcoinNewb> but your right
711 2010-12-30 07:27:48 <nanotube> bitcoin is pretty risky, all things considered.
712 2010-12-30 07:27:55 <BitcoinNewb> Oh yea
713 2010-12-30 07:27:55 <BitcoinNewb> how so
714 2010-12-30 07:28:07 <ThomasV> I would say its is high risk but potentially high reward
715 2010-12-30 07:28:19 <nanotube> on the scale of russian penny stock <-> gold... bitcoin is probably closer to the penny stock. :)
716 2010-12-30 07:29:02 <nanotube> things can happen... serious flaws discovered... govt banning it or ddosing... various things. :)
717 2010-12-30 07:29:10 <BitcoinNewb> Would you say that most people participating in the market are activists?
718 2010-12-30 07:29:30 <BitcoinNewb> I dont really like the current currency markets
719 2010-12-30 07:29:44 <ThomasV> nanotube: how robust is it to ddos ?
720 2010-12-30 07:29:45 <BitcoinNewb> and would love to invest in something diffrent
721 2010-12-30 07:30:00 <nanotube> heh yea, there are a lot of austrian economists and anarchists and libertarians around. :)
722 2010-12-30 07:30:09 <BitcoinNewb> ahh
723 2010-12-30 07:30:15 <BitcoinNewb> good
724 2010-12-30 07:30:25 <nanotube> ThomasV: well, there are only a couple thousand nodes... a well-heeled attacker can probably manage it
725 2010-12-30 07:30:46 <ThomasV> how many nodes active at a given time ?
726 2010-12-30 07:30:50 <nanotube> bump off the nodes, bump off the website and bootstrap irc channel... and down it goes.
727 2010-12-30 07:30:53 <BitcoinNewb> what does that mean? Sorry bout the questions
728 2010-12-30 07:30:56 <nanotube> ThomasV: about a couple thousand.
729 2010-12-30 07:31:07 <nanotube> what does what mean, BitcoinNewb
730 2010-12-30 07:31:09 <BitcoinNewb> so hackers could kill the market
731 2010-12-30 07:31:41 <ThomasV> nodes could protect themselves by imposing a small fee on transactions, couldn't they ?
732 2010-12-30 07:31:45 <BitcoinNewb> with a ddos attack
733 2010-12-30 07:32:11 <BitcoinNewb> to pay for the needed security anti-ddos gear?
734 2010-12-30 07:32:26 <ThomasV> I think that this is the point of the fee option
735 2010-12-30 07:32:51 <nanotube> ThomasV: ddos doesn't have to be from actual bitcoin transactions. just network traffic.
736 2010-12-30 07:32:53 <BitcoinNewb> has the market been attacked?
737 2010-12-30 07:32:56 <Kiba> BitcoinNewb: we're generally not political activist
738 2010-12-30 07:33:06 <Kiba> but a market activist
739 2010-12-30 07:33:19 <BitcoinNewb> yea market activists!
740 2010-12-30 07:33:22 <nanotube> BitcoinNewb: there has been a wave of paypal fraud, a few months ago... now people are a lot more wary about taking paypal.
741 2010-12-30 07:33:39 <BitcoinNewb> I hate paypal anyway
742 2010-12-30 07:33:42 <nanotube> but on a technical level, no i'm not aware of serious attacks.
743 2010-12-30 07:34:31 <ThomasV> nanotube: who was it ?
744 2010-12-30 07:34:39 <BitcoinNewb> but the p2p structure makes it 'vulnerable to attack?
745 2010-12-30 07:34:59 <nanotube> BitcoinNewb: it makes it less vulnerable than non-p2p structure. :)
746 2010-12-30 07:35:06 <ThomasV> well the p2p structure also makes it resilient
747 2010-12-30 07:35:08 <nanotube> ThomasV: nobody knows.
748 2010-12-30 07:35:21 <BitcoinNewb> thats why you guys are trying it
749 2010-12-30 07:35:27 <BitcoinNewb> ok I see
750 2010-12-30 07:35:39 <ThomasV> nanotube: I mean, the paypal addresses must be known, or not ?
751 2010-12-30 07:35:42 <BitcoinNewb> So all I need to do is fire up the program?
752 2010-12-30 07:35:48 <BitcoinNewb> will it kill my bandwith?
753 2010-12-30 07:35:52 <nanotube> ThomasV: i understand that they were using stolen paypal accounts.
754 2010-12-30 07:36:07 <nanotube> ThomasV: so knowing addresses doesn't do much good. that said, the market runners (mtgox, for instance) would know.
755 2010-12-30 07:36:48 <ThomasV> did someone report it to paypal ?
756 2010-12-30 07:37:29 <nanotube> ThomasV: read the related forum threads. in fact, paypal froze mtgox's account, because of too many chargebacks to his account...
757 2010-12-30 07:37:49 <ThomasV> oh I see
758 2010-12-30 07:38:01 <nanotube> BitcoinNewb: no, very light bandwidth use, after your initial blockchain download.
759 2010-12-30 07:38:35 <BitcoinNewb> ok good
760 2010-12-30 07:38:48 <BitcoinNewb> I'm seeding hella torrents lol
761 2010-12-30 07:39:52 sgornick has quit (Quit: Leaving.)
762 2010-12-30 07:40:06 <BitcoinNewb> Cool you can even pay in cash through the mail!
763 2010-12-30 07:40:11 <BitcoinNewb> no paypal though
764 2010-12-30 07:40:28 <ThomasV> nanotube: if Alice gives bitcoins to Bob and decides to pay a transaction fee, how does the network ensure that the recipient node is neither Alices nor Bob's node ?
765 2010-12-30 07:41:28 <ThomasV> is the fee shared among many nodes ?
766 2010-12-30 07:41:46 <edcba> fee is for block generator
767 2010-12-30 07:41:47 <Kiba> today, everyone and their dog in the web world tweeted an interview about bitcoin
768 2010-12-30 07:41:57 <Kiba> with unnoticable traffic level
769 2010-12-30 07:42:09 <Kiba> at the bitcoin forum..
770 2010-12-30 07:42:14 <Kiba> therefore,
771 2010-12-30 07:42:23 <BoBeR> links?
772 2010-12-30 07:42:30 <Kiba> just because everyone tweet aobut you doesn't mean
773 2010-12-30 07:42:33 <ThomasV> edcba: the generator of which block ?
774 2010-12-30 07:42:37 <BoBeR> ;;bc,blocks
775 2010-12-30 07:42:38 <gribble> 100126
776 2010-12-30 07:42:45 <BoBeR> ;;bc,stats
777 2010-12-30 07:42:46 <sloob> I look at a tweet about once a month
778 2010-12-30 07:42:47 <gribble> Current Blocks: 100126 | Current Difficulty: 14484.16236123 | Next Difficulty At Block: 100799 | Next Difficulty In: 673 blocks | Next Difficulty In About: 4 days, 4 hours, 57 minutes, and 0 seconds | Next Difficulty Estimate: 15836.48322474
779 2010-12-30 07:42:49 <Kiba> http://www.readwriteweb.com/hack/2010/12/interview-bitcoin.php
780 2010-12-30 07:42:59 <Kiba> sloob: I only listen to 16 people..
781 2010-12-30 07:43:06 <edcba> the one acknowledging the transactio
782 2010-12-30 07:43:13 <Kiba> and got twice as many followerers
783 2010-12-30 07:43:33 sgornick has joined
784 2010-12-30 07:43:50 <BitcoinNewb> How do I get credit on my transactions when the program doesnt need my account info?
785 2010-12-30 07:43:58 <BitcoinNewb> thanks for all the help btw
786 2010-12-30 07:44:40 <ThomasV> edcba: and how is the block acknowledging the transaction chosen ?
787 2010-12-30 07:45:26 <edcba> maths
788 2010-12-30 07:46:01 <edcba> solving a problem with varying difficulty
789 2010-12-30 07:46:33 <ThomasV> so, if I generate a block, I will not only get 50btc, but I will also get paid for thransactions until the end of the world, no matter if I run a node or not ?
790 2010-12-30 07:46:43 <BitcoinNewb> So to add credit to my balance I need to generate?
791 2010-12-30 07:46:56 <Kiba> add what credit?
792 2010-12-30 07:46:59 <Kiba> bitcoin is not credit
793 2010-12-30 07:47:04 <BitcoinNewb> yes sorry
794 2010-12-30 07:47:05 <Kiba> it's a commodity
795 2010-12-30 07:47:25 <BitcoinNewb> So to add this commodity I need to generate
796 2010-12-30 07:47:34 <Kiba> I propose we rename wallet.dat to key.dat
797 2010-12-30 07:47:36 <edcba> ThomasV: you earn the fees only once for a block
798 2010-12-30 07:47:57 <Kiba> everybody have a hyperspace vault
799 2010-12-30 07:48:07 <Kiba> their wallet.dat simply access these hyperspace wallet
800 2010-12-30 07:48:21 <BitcoinNewb> How big is this download?
801 2010-12-30 07:48:29 <Kiba> s/wallet/vault
802 2010-12-30 07:48:41 <ThomasV> edcba: that's not what you said above, or I did not understand
803 2010-12-30 07:48:56 <sloob> it's in the cloud!
804 2010-12-30 07:49:00 <sloob> the land of wind and ghosts
805 2010-12-30 07:49:09 <BitcoinNewb> lol
806 2010-12-30 07:49:17 <BitcoinNewb> Thats a good fall back
807 2010-12-30 07:50:21 <BitcoinNewb> Im still missing how this program will get bitcoins into the account I just made on Mt Gox
808 2010-12-30 07:50:22 <BitcoinNewb> sorry
809 2010-12-30 07:51:07 <Kiba> you must send bitcoin to an address belonging to Mt Gox
810 2010-12-30 07:51:19 <BitcoinNewb> Ok
811 2010-12-30 07:51:24 <BitcoinNewb> Thats was I was wondering
812 2010-12-30 07:51:36 <ThomasV> BitcoinNewb: try the bitcoin faucet first
813 2010-12-30 07:51:44 <edcba> ThomasV: transaction fee is paid to generator of the first block including that transaction
814 2010-12-30 07:51:46 <ThomasV> it'll give you an idea
815 2010-12-30 07:52:20 <BitcoinNewb> OH! Neet
816 2010-12-30 07:52:25 <BitcoinNewb> free bitcoins!
817 2010-12-30 07:52:30 <ThomasV> edcba: so, not the one who created the bloc in the first place ?
818 2010-12-30 07:53:20 <BitcoinNewb> This is right up my alley
819 2010-12-30 07:53:46 <BitcoinNewb> I hate our current reserve banking system
820 2010-12-30 07:53:49 <edcba> i think you are still confused on how it works :)
821 2010-12-30 07:54:11 <ThomasV> tha's why I ask :-)
822 2010-12-30 07:54:22 <BitcoinNewb> ok
823 2010-12-30 07:54:35 <sloob> there's no reason that bitcoin transactions wouldn't still be taxable, it's just under the radar
824 2010-12-30 07:54:47 <edcba> block have transactions that may have fees
825 2010-12-30 07:54:48 <sloob> but at least it can't get yanked around
826 2010-12-30 07:54:49 <BitcoinNewb> I got a couple of bitcoins on the way, and im generating
827 2010-12-30 07:54:57 <BitcoinNewb> sweet
828 2010-12-30 07:55:15 dduane has quit (Ping timeout: 260 seconds)
829 2010-12-30 07:55:22 <edcba> blocks put in stone transactions
830 2010-12-30 07:55:27 AAA_awright_ has joined
831 2010-12-30 07:55:36 <ThomasV> yeah I got that
832 2010-12-30 07:55:52 <BitcoinNewb> Do you guys know a good program to monitor the bandwith being taken up by bitcoin
833 2010-12-30 07:55:55 <edcba> you need not to gen a block to gen a transaction
834 2010-12-30 07:56:00 <BitcoinNewb> I have other people in the house you see
835 2010-12-30 07:56:09 <sloob> will the number of blocks in an initial download become unmanageable as the number of transactions increase?
836 2010-12-30 07:56:24 <Kiba> sloob: bandwidth will increase
837 2010-12-30 07:56:29 <Kiba> storage will increase
838 2010-12-30 07:56:31 <Kiba> the question
839 2010-12-30 07:56:33 <Kiba> is
840 2010-12-30 07:56:45 <Kiba> will the block chain outpace storage and bandwidth improvement?
841 2010-12-30 07:56:52 <edcba> you need a transaction to be included in a block for receivers to get the money
842 2010-12-30 07:57:23 <ThomasV> edcba: I got that too...
843 2010-12-30 07:57:58 AAA_awright has quit (Ping timeout: 264 seconds)
844 2010-12-30 07:58:36 <ThomasV> so the fee will go to the first node that includes the transaction in the clock ?
845 2010-12-30 07:59:05 <BitcoinNewb> looks like its taking around .50 kb a sec
846 2010-12-30 07:59:22 <BitcoinNewb> I mean MB
847 2010-12-30 07:59:25 <edcba> in the block chain yes
848 2010-12-30 07:59:38 <ThomasV> edcba: then my question remains intact : if Alice gives bitcoins to Bob and decides to pay a transaction fee, how does the network ensure that the recipient node is neither Alices nor Bob's node ?
849 2010-12-30 08:00:07 <nanotube> ThomasV: the recipient is indeed bob's node... since alice gave bitcoins to bob
850 2010-12-30 08:00:26 <nanotube> tx fee doesn't have a preset recipient. person who generates the block containing the tx gets the fee.
851 2010-12-30 08:00:39 <ThomasV> so in that case it is not protecting at all against ddos
852 2010-12-30 08:00:58 <nanotube> and ddos has absolutely no relationship to transactions and fees.
853 2010-12-30 08:01:21 <ThomasV> indeed, that's what I wanted to know
854 2010-12-30 08:01:25 <ThomasV> thanks
855 2010-12-30 08:01:47 <nanotube> read ,,(bc,wiki blocks) and ,,(bc,wiki block chain) and ,,(bc,wiki transaction fees)
856 2010-12-30 08:01:47 <gribble> https://en.bitcoin.it/wiki/Blocks | Dec 19, 2010 ... Data is permanently recorded in the Bitcoin network through blocks. Each block contains all recent transactions, a nonce (random number), ...
857 2010-12-30 08:01:48 <gribble> https://en.bitcoin.it/wiki/Block_chain | Dec 20, 2010 ... Every block contains a hash of the previous block. This has the effect of creating a chain of blocks from the genesis block to the current ...
858 2010-12-30 08:01:49 <gribble> https://en.bitcoin.it/wiki/Transaction_fees | Dec 19, 2010 ... Transaction fees may be included with any transfer of bitcoins from one address to another. At the moment, many transactions are typically ...
859 2010-12-30 08:02:14 dduane has joined
860 2010-12-30 08:03:03 <ThomasV> but this is a flaw imo
861 2010-12-30 08:03:41 <BitcoinNewb> For newbs like me this is a great read
862 2010-12-30 08:03:43 <BitcoinNewb> http://www.bitcoin.org/faq#What_is_a_block
863 2010-12-30 08:04:29 <ThomasV> we should try to ddos bitcoin before someone else does
864 2010-12-30 08:06:16 <Kiba> go do it
865 2010-12-30 08:06:22 <Kiba> and write a security report on it
866 2010-12-30 08:06:58 <ThomasV> if a ddos attack succeeds when btc has become more mainstream, it will be very bad for the btc economy
867 2010-12-30 08:07:37 <edcba> it's still in beta...
868 2010-12-30 08:07:39 davout has joined
869 2010-12-30 08:07:44 <davout> o hai
870 2010-12-30 08:07:48 <sloob> what kind of ddos are you talking about
871 2010-12-30 08:07:54 <sloob> like transaction spamming?
872 2010-12-30 08:07:59 <ThomasV> yes but it's getting more and more attention
873 2010-12-30 08:08:09 <BitcoinNewb> Cool im setting up a cobian backup for my btc wallet!
874 2010-12-30 08:08:16 <edcba> the pb is the btc p2p protocol anyway
875 2010-12-30 08:08:25 <BitcoinNewb> this is way fun! Im so glad I saw this on Kieser Report!
876 2010-12-30 08:08:58 <BitcoinNewb> So instead of a bank I keep my money on my computer in a database file!
877 2010-12-30 08:09:05 <BitcoinNewb> this is the greatest thing ever
878 2010-12-30 08:09:11 <ThomasV> sloob: I was believing that the network could be protected from fake transactions, by imposing transaction fees
879 2010-12-30 08:09:22 <ThomasV> but apparently that's not the case
880 2010-12-30 08:09:56 <sloob> I am not totally convinced that there's no way to clog up the tubes with frivolous transactions
881 2010-12-30 08:10:00 <BitcoinNewb> 4chan would shit all over this if they thought it would ruin somebodys day
882 2010-12-30 08:10:01 <ThomasV> imo the transcation fee should be split and shared among many nodes
883 2010-12-30 08:10:22 <BitcoinNewb> I set mine at .01 is that right?
884 2010-12-30 08:10:31 <edcba> tx fees do noothing againnst ddos
885 2010-12-30 08:10:44 <sloob> I can't quantify what my concern is, so I'm sort of spreading FUD. but I'm thinking about it a lot.
886 2010-12-30 08:11:04 <ArtForz> protection against ddos is quite simple
887 2010-12-30 08:11:26 <ThomasV> ArtForz: yes ?
888 2010-12-30 08:11:28 <ArtForz> yes
889 2010-12-30 08:11:32 Slix` has quit (Read error: Connection reset by peer)
890 2010-12-30 08:12:30 <BitcoinNewb> because its decentrailised?
891 2010-12-30 08:12:32 <edcba> the only pb is to code it and make nodes run that
892 2010-12-30 08:12:58 <ArtForz> especially as satoshi thinks every node caching every transaction is a good idea
893 2010-12-30 08:13:29 <edcba> indeed
894 2010-12-30 08:13:33 <ThomasV> ArtForz: can you be a bit more prolix ?
895 2010-12-30 08:14:00 <ArtForz> I should just waste 0.01 btc in 0.00000001 chunks and prove him wrong
896 2010-12-30 08:14:29 <ArtForz> just because it's small and doesnt include a fee doesnt mean its not a valid transaction, just means it'll pretty much never get into a block
897 2010-12-30 08:15:01 <ArtForz> current nodes will happily cache millions of these until they're restarted
898 2010-12-30 08:15:17 <edcba> also in order to waste less cpu we could make pb difficulty function of public node addr
899 2010-12-30 08:15:35 <ArtForz> huh?
900 2010-12-30 08:16:10 <edcba> not sure about that one because of pregen...
901 2010-12-30 08:17:11 <edcba> ArtForz: solution relative to distance from some public key hash
902 2010-12-30 08:17:40 <ArtForz> solution to what who? where?
903 2010-12-30 08:18:14 <edcba> block gen
904 2010-12-30 08:18:24 <ArtForz> how the fuck would that work
905 2010-12-30 08:20:28 <edcba> instead of h(block) inf to difficulty
906 2010-12-30 08:20:47 <ArtForz> inf to difficulty?
907 2010-12-30 08:20:57 <edcba> inferior to
908 2010-12-30 08:21:37 <edcba> h(block) - h(pubkey) inf to difficulty
909 2010-12-30 08:21:46 marioxcc-AFK has quit (Remote host closed the connection)
910 2010-12-30 08:21:55 <ArtForz> that sounds like a VERY bad idea
911 2010-12-30 08:22:10 <edcba> haha
912 2010-12-30 08:22:45 <ArtForz> so I guess you want to make pubkeys one-time only
913 2010-12-30 08:23:00 <edcba> no of course
914 2010-12-30 08:23:24 <edcba> but yes it needs some tweakinngg
915 2010-12-30 08:23:32 AAA_awright has joined
916 2010-12-30 08:24:29 <edcba> a better idea also is make hashcash instead of fees
917 2010-12-30 08:25:01 <ArtForz> otherwise I just need to build a DB of 2**48 pubkeys or so and have a major advantage
918 2010-12-30 08:25:16 xelister has quit (Remote host closed the connection)
919 2010-12-30 08:25:58 <edcba> yes i see the pb but maybe there is a way to avoid that
920 2010-12-30 08:26:12 AAA_awright_ has quit (Ping timeout: 272 seconds)
921 2010-12-30 08:26:13 <edcba> dunno how yet
922 2010-12-30 08:26:56 <ArtForz> actually in the current system it wouldnt change anything
923 2010-12-30 08:27:11 <ArtForz> pubkey is included in block hash via merkle tree
924 2010-12-30 08:27:13 Diablo-D3 has quit (Ping timeout: 255 seconds)
925 2010-12-30 08:27:30 <ArtForz> so doing h(hdr) - h(pubk) is just a bunch of 32-bit substrations more
926 2010-12-30 08:27:45 <ArtForz> = whats the point?
927 2010-12-30 08:28:09 <edcba> it's general idea
928 2010-12-30 08:28:16 <ArtForz> sha256 already does something like 1500 32-bit adds per bitcoinhash, another 8 32+carry subs wont change anything
929 2010-12-30 08:29:33 <ArtForz> and if you dont include pubkey in hash, see "build DB of fuckton of pubkeys indexed by top hash bytes"
930 2010-12-30 08:30:11 <edcba> you just need to make a distance to make some clients advantaged to some others for a fixed block
931 2010-12-30 08:30:22 <ArtForz> why?
932 2010-12-30 08:30:36 <ArtForz> = you're now limited by amount of storage / access latency
933 2010-12-30 08:31:15 <ArtForz> so instead of clusters of GPUs people will build clusters of SSDs
934 2010-12-30 08:31:28 <edcba> i just tweak difficulty
935 2010-12-30 08:31:48 <ArtForz> no you dont
936 2010-12-30 08:32:23 <ArtForz> for random h, h < target is exactly as likely as (h - pubkey) < target
937 2010-12-30 08:34:04 <edcba> - is more like bitwise distance applied to a mask
938 2010-12-30 08:34:48 <BitcoinNewb> Hey guys!
939 2010-12-30 08:35:12 <ArtForz> errrr... so now count_1_bits(h ^ hash(pubkey)) < target ?
940 2010-12-30 08:35:15 <BitcoinNewb> I just received my first bitcoin transaction from bitcoin faucet
941 2010-12-30 08:35:22 AntonB_ has joined
942 2010-12-30 08:35:23 <edcba> pb of key db remains
943 2010-12-30 08:35:30 <sloob> congrats
944 2010-12-30 08:35:36 <BitcoinNewb> lol
945 2010-12-30 08:35:41 <ArtForz> yep
946 2010-12-30 08:35:43 <AntonB_> Hi all. Where can I read about converting bitcoin to/from USD ? Any ready made web-services ?
947 2010-12-30 08:36:10 <ArtForz> you effectively turned a CPU-bound problem into a time/space tradeoff problem
948 2010-12-30 08:37:04 <edcba> i see some solutions but that would just do what i wanted to prevent
949 2010-12-30 08:37:41 <BitcoinNewb> gotta say this is some really cool shit right here
950 2010-12-30 08:37:53 <BitcoinNewb> is this the first p2p currency?
951 2010-12-30 08:38:16 <sloob> gold is p2p
952 2010-12-30 08:38:23 <BitcoinNewb> lol
953 2010-12-30 08:38:25 <BitcoinNewb> oh yea
954 2010-12-30 08:38:30 <ArtForz> I doubt it's possible to find any one-way problem that doesnt benefit from some kind of special hardware
955 2010-12-30 08:39:08 <BitcoinNewb> it downloaded well over 300 MB
956 2010-12-30 08:39:10 <BitcoinNewb> wow
957 2010-12-30 08:39:32 <BitcoinNewb> my khash is around 950 is that good?
958 2010-12-30 08:39:53 <ArtForz> ;;bc,calc 950
959 2010-12-30 08:39:54 <gribble> The average time to generate a block at 950 Khps, given current difficulty of 14484.16236123 , is 2 years, 3 weeks, 6 days, 21 hours, 46 minutes, and 1 second
960 2010-12-30 08:40:21 <BitcoinNewb> you used that tool didnt you? I will use that next time
961 2010-12-30 08:40:24 <BitcoinNewb> thats really shitty
962 2010-12-30 08:40:29 <BitcoinNewb> I need a better computer
963 2010-12-30 08:40:34 <ArtForz> might want to check out ,bc,,pool
964 2010-12-30 08:40:35 <gribble> No fancy GPU farm, and don't want to wait for months for a block gen? Join the mining pool! http://mining.bitcoin.cz/
965 2010-12-30 08:41:09 <BitcoinNewb> man you guys think of everything!
966 2010-12-30 08:41:12 <BitcoinNewb> Sweet!
967 2010-12-30 08:41:24 <BitcoinNewb> pool users and share the benifits!
968 2010-12-30 08:41:36 <ArtForz> pooled mining still provides the same average payout as normal mining, just less... chunky
969 2010-12-30 08:43:19 <ArtForz> instead of randomly 50 btc every 2 years average, pooled mining gives more like 0.06 or so btc/day
970 2010-12-30 08:44:45 <BitcoinNewb> what miner do you suggest
971 2010-12-30 08:44:59 <ArtForz> dunno, I gave up on CPU mining long ago
972 2010-12-30 08:45:03 <sloob> thanks for that. i've got a celeron crapbox that coud be mining coins
973 2010-12-30 08:45:27 <ArtForz> running 24 HD5970 + 5 HD5770 GPUs
974 2010-12-30 08:45:42 <sloob> but i didn't want to wait 12 years for a payout
975 2010-12-30 08:45:54 <BitcoinNewb> mybitcoin is the same thing just on a cloud
976 2010-12-30 08:46:02 <ArtForz> = about 16Gh/s
977 2010-12-30 08:46:10 RazielZ has quit (Ping timeout: 240 seconds)
978 2010-12-30 08:47:52 <edcba> now i wonder if it's more beneficial to consume lot of hdd or lot of cpu...
979 2010-12-30 08:48:04 <BoBeR> lots of CPU
980 2010-12-30 08:48:13 <BoBeR> more heat
981 2010-12-30 08:48:15 <BoBeR> you can heat your home
982 2010-12-30 08:48:45 <ArtForz> and yes, that's 16000000 khash/s
983 2010-12-30 08:50:00 <BoBeR> um
984 2010-12-30 08:50:09 <BoBeR> 16 000 000 000 hash
985 2010-12-30 08:50:14 <BoBeR> looks bigger like that
986 2010-12-30 08:50:19 <ArtForz> or about the equivalent of 10000 low-end CPUs or 1000 high end 6-core monsters
987 2010-12-30 08:50:42 <BoBeR> dont ever go offline
988 2010-12-30 08:50:50 <BoBeR> bitcoin will slow to a crawl
989 2010-12-30 08:50:54 <BoBeR> can i ask
990 2010-12-30 08:51:00 <BoBeR> how many bitcoins do you have
991 2010-12-30 08:51:07 Bananaphone has joined
992 2010-12-30 08:51:11 <BoBeR> or are you selling them on mtgox
993 2010-12-30 08:51:36 <ArtForz> lots of hdd is actually worse I think
994 2010-12-30 08:51:45 <davout> ;;bc,calc 16000000
995 2010-12-30 08:51:46 RazielZ has joined
996 2010-12-30 08:51:46 <gribble> The average time to generate a block at 16000000 Khps, given current difficulty of 14484.16236123 , is 1 hour, 4 minutes, and 48 seconds
997 2010-12-30 08:51:50 <ArtForz> because it scales superlinear with #HDDs
998 2010-12-30 08:51:55 <davout> :D
999 2010-12-30 08:52:22 <BitcoinNewb> so I mine or generate
1000 2010-12-30 08:52:27 <BitcoinNewb> I cant do both on one machine
1001 2010-12-30 08:52:29 <ArtForz> same thing
1002 2010-12-30 08:52:47 <BoBeR> generate is mining
1003 2010-12-30 08:52:53 <BoBeR> using the bitcoin client
1004 2010-12-30 08:53:03 <BoBeR> which is not even 100% efficient
1005 2010-12-30 08:53:12 <BitcoinNewb> I understand that
1006 2010-12-30 08:53:37 <BitcoinNewb> but if I used pooled mining I get a trickle instead of a lump sum
1007 2010-12-30 08:53:43 <ArtForz> yep
1008 2010-12-30 08:53:59 <BitcoinNewb> but really it is the same thing
1009 2010-12-30 08:54:07 <ArtForz> yep
1010 2010-12-30 08:54:18 sloob has quit (Ping timeout: 276 seconds)
1011 2010-12-30 08:55:26 <BitcoinNewb> So I should get more computers generating for me then
1012 2010-12-30 08:56:56 <ArtForz> yeah, if you're paying for power CPU is pretty much a non-starter
1013 2010-12-30 08:57:40 <ArtForz> still kinda useful if it's winter and you have electric heating anyways
1014 2010-12-30 08:58:01 <BitcoinNewb> well at this rate it will take 765 days for me to get 50 bitcoins
1015 2010-12-30 08:58:16 <BitcoinNewb> Looks like I have to mine just to see new coins this year
1016 2010-12-30 08:58:27 <ArtForz> or just buy some
1017 2010-12-30 08:58:37 <BitcoinNewb> yes or earn them
1018 2010-12-30 08:58:47 <ArtForz> or that
1019 2010-12-30 08:58:51 <BitcoinNewb> I guess I will take a look at the forum, and put up an add
1020 2010-12-30 08:58:57 elryry has quit (Quit: DAMN!)
1021 2010-12-30 08:58:58 <BitcoinNewb> ad
1022 2010-12-30 08:59:17 <Bananaphone> i'm gonna throw a bunch of dvds up there sooner or later
1023 2010-12-30 09:05:15 <davout> bidding pond is best suited for that
1024 2010-12-30 09:06:32 <ThomasV> hey I have a bunch of stamps to sell
1025 2010-12-30 09:06:51 <ThomasV> but no idea if it's worth the effort to put ads
1026 2010-12-30 09:07:44 <ThomasV> ops sorry, wrong channel
1027 2010-12-30 09:08:35 elryry has joined
1028 2010-12-30 09:10:25 <BoBeR> mtgox alive
1029 2010-12-30 09:11:38 <ThomasV> bearish
1030 2010-12-30 09:26:39 BitcoinNewb has quit (Quit: Page closed)
1031 2010-12-30 09:29:45 ciuciu has joined
1032 2010-12-30 09:50:32 AAA_awright_ has joined
1033 2010-12-30 09:51:15 johndrinkwater has quit (Read error: Operation timed out)
1034 2010-12-30 09:52:23 AAA_awright has quit (Ping timeout: 240 seconds)
1035 2010-12-30 09:53:12 johndrinkwater has joined
1036 2010-12-30 09:54:52 ciuciu has quit (Ping timeout: 272 seconds)
1037 2010-12-30 09:56:24 Xunie has quit (Remote host closed the connection)
1038 2010-12-30 09:58:38 <mizerydearia> `bitcoind backupwallet ./wallet3.dat` => error: {"code":-1,"message":"boost::filesystem::copy_file: Permission denied: \"/var/lib/bitcoin/.bitcoin/wallet.dat\", \"./wallet3.dat\""}
1039 2010-12-30 09:59:07 <mizerydearia> Am I providing incorrect input?
1040 2010-12-30 09:59:13 Xunie has joined
1041 2010-12-30 09:59:13 Xunie has quit (Changing host)
1042 2010-12-30 09:59:13 Xunie has joined
1043 2010-12-30 09:59:30 <ArtForz> dont use relative paths
1044 2010-12-30 10:00:16 <mizerydearia> Aha, thanks.
1045 2010-12-30 10:01:46 <mizerydearia> Is it possible relative paths can be used in a future release?
1046 2010-12-30 10:02:00 <ArtForz> relative to what?
1047 2010-12-30 10:02:02 <mizerydearia> Or, configuration of path in bitcoin.conf
1048 2010-12-30 10:02:21 <mizerydearia> Relative to some designated backup path.
1049 2010-12-30 10:03:00 <ArtForz> should be possible
1050 2010-12-30 10:03:05 slush has joined
1051 2010-12-30 10:13:11 slush has quit (Read error: Connection reset by peer)
1052 2010-12-30 10:13:14 slush1 has joined
1053 2010-12-30 10:13:54 TheAncientGoat has joined
1054 2010-12-30 10:16:38 TD has joined
1055 2010-12-30 10:22:29 slush1 is now known as slush
1056 2010-12-30 10:22:37 slush has quit (Changing host)
1057 2010-12-30 10:22:37 slush has joined
1058 2010-12-30 10:25:56 kikea has joined
1059 2010-12-30 10:26:42 <slush> ;;getrating slush
1060 2010-12-30 10:26:42 <gribble> User slush was created on Thu Nov 11 07:34:28 2010, and has a cumulative rating of 4, from a total of 4 ratings. Of these, 4 are positive and 0 are negative. This user has also sent 5 positive ratings, and 0 negative ratings to others.
1061 2010-12-30 10:28:46 TD has quit (Quit: TD)
1062 2010-12-30 10:29:50 kikea has quit (Client Quit)
1063 2010-12-30 10:43:48 <davout> fuck yea, just generated a block
1064 2010-12-30 10:43:56 <davout> fuck yea, it has a trading fee
1065 2010-12-30 10:44:18 <davout> hey slush, does your pool manage to split trading fees too ?
1066 2010-12-30 10:44:31 <davout> *transaction fees
1067 2010-12-30 10:47:23 <UukGoblin> ;;getrating UukGoblin
1068 2010-12-30 10:47:23 <gribble> User UukGoblin was created on Mon Nov 8 11:05:40 2010, and has a cumulative rating of 1, from a total of 1 ratings. Of these, 1 are positive and 0 are negative. This user has also sent 1 positive ratings, and 0 negative ratings to others.
1069 2010-12-30 10:47:27 <UukGoblin> mrhm
1070 2010-12-30 10:47:32 <UukGoblin> I should work on my rating I guess
1071 2010-12-30 10:51:03 <slush> davout: Not yet; Because pool earned 0.02 bitcoins at first 30 blocks, I did not complete logic for it
1072 2010-12-30 10:51:36 <slush> davout: once it became significant amount, I will do it, of course
1073 2010-12-30 10:51:59 Bananaphone has quit (Ping timeout: 260 seconds)
1074 2010-12-30 10:52:57 <davout> yea
1075 2010-12-30 10:53:12 <davout> so that's what i thought
1076 2010-12-30 10:53:20 <davout> you *are* a scammer
1077 2010-12-30 10:53:25 <slush> lol :)
1078 2010-12-30 10:53:30 <davout> i'm sending the logs to gene
1079 2010-12-30 10:53:45 <slush> yeah, I do micro-scamming
1080 2010-12-30 10:53:46 <davout> haven't seen him around much :)
1081 2010-12-30 10:53:58 <slush> 0.01 bitcoins per 20 blocks, I become rich soon
1082 2010-12-30 10:55:07 <davout> yup
1083 2010-12-30 11:04:45 <slush> davout: If I count 0.001btc as my daily income from this scamming and wage 2$ per hour of my work (which is something like minimal wage in my country), and say I spent month of common mandays, my return from pool investment will be back in 1166000 hours (~6600 months). yay!
1084 2010-12-30 11:05:28 slush has quit (Quit: Leaving.)
1085 2010-12-30 11:05:29 slush1 has joined
1086 2010-12-30 11:05:31 slush1 is now known as slush
1087 2010-12-30 11:19:03 <ThomasV> has someone computed the average number of transactions per block ?
1088 2010-12-30 11:19:26 <davout> what would be the use of such a stat ?
1089 2010-12-30 11:19:28 <ThomasV> I polled a few random blocks at blockexplorer
1090 2010-12-30 11:19:58 <ThomasV> most of them have zero transaction, except the inintial generation of 50 ntc
1091 2010-12-30 11:20:03 <ThomasV> *btc*
1092 2010-12-30 11:20:31 <ThomasV> it suggests that most bitcoins are not being used
1093 2010-12-30 11:21:23 noagendamarket has joined
1094 2010-12-30 11:23:59 <ThomasV> it would be interesting to see how transactions are distributed
1095 2010-12-30 11:36:01 <davout> i'm not sure how you'd reach to a figure of x% bitcoins being used
1096 2010-12-30 11:36:25 <davout> bitcoins can be in transactions without being actually traded between people
1097 2010-12-30 11:36:37 <davout> even if you discard generation transactions
1098 2010-12-30 11:36:49 <sipa> ThomasV: i made a graph of that once
1099 2010-12-30 11:36:56 <sipa> ThomasV: but it's very irregular
1100 2010-12-30 11:37:07 <ThomasV> a histogram ?
1101 2010-12-30 11:37:27 <sipa> an average number of transactions/day
1102 2010-12-30 11:37:48 <sipa> oh you just want per block? let me see
1103 2010-12-30 11:46:16 eureka^ has joined
1104 2010-12-30 12:02:31 slush has quit (Ping timeout: 240 seconds)
1105 2010-12-30 12:06:46 TheAncientGoat_ has joined
1106 2010-12-30 12:07:52 TheAncientGoat has quit (Ping timeout: 276 seconds)
1107 2010-12-30 12:14:22 TheAncientGoat_ has quit (Ping timeout: 276 seconds)
1108 2010-12-30 12:18:05 remmy_ has quit (Ping timeout: 240 seconds)
1109 2010-12-30 12:20:11 TheAncientGoat has joined
1110 2010-12-30 12:23:02 slush has joined
1111 2010-12-30 12:28:38 mtgox has joined
1112 2010-12-30 12:31:08 <ThomasV> sipa: that would be great
1113 2010-12-30 12:33:45 TheAncientGoat has quit (Ping timeout: 276 seconds)
1114 2010-12-30 12:34:33 TheAncientGoat has joined
1115 2010-12-30 12:48:05 TheAncientGoat_ has joined
1116 2010-12-30 12:48:49 TheAncientGoat has quit (Ping timeout: 276 seconds)
1117 2010-12-30 12:53:23 slush has quit (Ping timeout: 240 seconds)
1118 2010-12-30 12:53:34 noagendamarket_ has joined
1119 2010-12-30 12:53:41 noagendamarket has quit (Ping timeout: 240 seconds)
1120 2010-12-30 12:55:58 TheAncientGoat_ has quit (Ping timeout: 276 seconds)
1121 2010-12-30 12:58:00 devon_hillard has joined
1122 2010-12-30 12:58:13 TheAncientGoat has joined
1123 2010-12-30 12:58:22 <devon_hillard> Why are NVidia cards so much slower than ATI at bitcoin crunching?
1124 2010-12-30 12:58:57 <devon_hillard> or what are applications where an nvidia card would be better, price-per-price than an equivalent ati card?
1125 2010-12-30 13:00:41 noagendamarket__ has joined
1126 2010-12-30 13:00:42 <sipa> the architecture is different
1127 2010-12-30 13:01:11 <sipa> from what i heard, nvidia cards would be better for floating point
1128 2010-12-30 13:01:23 <sipa> than for integer operations
1129 2010-12-30 13:01:39 <sipa> and they also have an advantage when more branching is involved
1130 2010-12-30 13:04:29 noagendamarket_ has quit (Ping timeout: 240 seconds)
1131 2010-12-30 13:04:30 noagendamarket__ is now known as noagendamarket_
1132 2010-12-30 13:05:01 <sipa> ThomasV: http://sipa.ulyssis.org/static/bitcoin/bc-tx-hist.txt
1133 2010-12-30 13:05:16 noagendamarket_ has quit (Client Quit)
1134 2010-12-30 13:05:26 <ThomasV> nice
1135 2010-12-30 13:06:02 <wumpus> devon_hillard: well, ati cards have much less control logic, they are really CPU-driven number crunching machines.. NV cards are becoming more and more like independent CPUs and have a lot of control logic
1136 2010-12-30 13:06:35 <wumpus> devon_hillard: apart from that, ATI have instructions like bit rotate left/right, and generally faster integer handling
1137 2010-12-30 13:06:56 <devon_hillard> so it really depends on the application
1138 2010-12-30 13:08:26 <devon_hillard> so nvidia would want to muscle in on IBM Cell turf, while ATI is more content with having powerful vector processors
1139 2010-12-30 13:09:30 <wumpus> yep, it depends on the application
1140 2010-12-30 13:11:49 mndrix has quit (Ping timeout: 265 seconds)
1141 2010-12-30 13:12:25 genjix has joined
1142 2010-12-30 13:13:36 <devon_hillard> what's a motherboard that can accomodate 4 graphics cards?
1143 2010-12-30 13:13:37 <wumpus> for bitcoin generation ATI is by far the best, though I'm not sure how long it will remain profitable to buy them for mining, as the difficulty keeps increasing
1144 2010-12-30 13:13:53 <wumpus> especially if you have to pay electricity :)
1145 2010-12-30 13:14:03 TheAncientGoat has quit (Ping timeout: 276 seconds)
1146 2010-12-30 13:14:13 TheAncientGoat has joined
1147 2010-12-30 13:14:14 <devon_hillard> it would be nice to live in Iceland from that perspective
1148 2010-12-30 13:14:22 <devon_hillard> 4 cents per kWh :)
1149 2010-12-30 13:16:43 RichardG has joined
1150 2010-12-30 13:18:53 johndrinkwater has quit (Remote host closed the connection)
1151 2010-12-30 13:23:06 johndrinkwater has joined
1152 2010-12-30 13:25:27 <devon_hillard> ;;bc,calc 1000
1153 2010-12-30 13:25:29 <gribble> The average time to generate a block at 1000 Khps, given current difficulty of 14484.16236123 , is 1 year, 50 weeks, 5 days, 0 hours, 16 minutes, and 43 seconds
1154 2010-12-30 13:25:36 <devon_hillard> ;;bc,calc 1000000
1155 2010-12-30 13:25:37 <gribble> The average time to generate a block at 1000000 Khps, given current difficulty of 14484.16236123 , is 17 hours, 16 minutes, and 49 seconds
1156 2010-12-30 13:25:45 RichardG has quit (Ping timeout: 240 seconds)
1157 2010-12-30 13:25:52 TheAncientGoat has quit (Ping timeout: 276 seconds)
1158 2010-12-30 13:26:47 * sipa wonders if he can put a machine with some 5970's in the computer lab
1159 2010-12-30 13:27:46 TheAncientGoat has joined
1160 2010-12-30 13:29:35 <devon_hillard> ;;bc,calc 4650000
1161 2010-12-30 13:29:37 <gribble> The average time to generate a block at 4650000 Khps, given current difficulty of 14484.16236123 , is 3 hours, 42 minutes, and 58 seconds
1162 2010-12-30 13:29:46 RichardG has joined
1163 2010-12-30 13:30:11 kermit has quit (Ping timeout: 265 seconds)
1164 2010-12-30 13:30:12 <ThomasV> sipa: similarly, is it possible to see the number of transactions *per bitcoin* ?
1165 2010-12-30 13:30:25 <sipa> how do you mean?
1166 2010-12-30 13:30:39 <ThomasV> the number of times a given bitcoin has been exchanged
1167 2010-12-30 13:30:43 acous has joined
1168 2010-12-30 13:30:44 acous has quit (Changing host)
1169 2010-12-30 13:30:44 acous has joined
1170 2010-12-30 13:30:58 <sipa> define 'a given bitcoin' ?
1171 2010-12-30 13:31:07 <ThomasV> heh
1172 2010-12-30 13:31:37 <ThomasV> I would like to see this data for all of them
1173 2010-12-30 13:31:40 <ThomasV> in a histogram
1174 2010-12-30 13:31:49 <sipa> yes, but what is a given bitcoin?
1175 2010-12-30 13:32:07 <sipa> account balances are results of transactions
1176 2010-12-30 13:32:32 <sipa> and ok, you can probably trace every transaction back to the block that generated it
1177 2010-12-30 13:32:46 <ThomasV> and eight by the number of bitcoins
1178 2010-12-30 13:32:50 <ThomasV> *weight*
1179 2010-12-30 13:32:54 TheAncientGoat has quit (Ping timeout: 276 seconds)
1180 2010-12-30 13:34:27 RG has joined
1181 2010-12-30 13:35:09 <ThomasV> sipa: which tools do you use to do this ?
1182 2010-12-30 13:35:30 <sipa> bitcoin + getblock patch
1183 2010-12-30 13:35:41 <sipa> perl script that dumps these in a text file
1184 2010-12-30 13:35:58 <ThomasV> where can I get it ?
1185 2010-12-30 13:36:21 <sipa> and some unix tools, like sort / grep / uniq / ... to create the histogram
1186 2010-12-30 13:36:30 RichardG has quit (Ping timeout: 260 seconds)
1187 2010-12-30 13:36:35 <sipa> eh i can put it online, but it's not much
1188 2010-12-30 13:38:17 <ThomasV> that would be great
1189 2010-12-30 13:38:22 RichardG has joined
1190 2010-12-30 13:38:39 <ThomasV> although I have not even managed to compile it atm
1191 2010-12-30 13:39:02 RG has quit (Ping timeout: 250 seconds)
1192 2010-12-30 13:40:28 <tcatm> http://blk.bitcoinwatch.com/ has the same information as getblock
1193 2010-12-30 13:41:47 <ThomasV> what doesthe getblock patch do ?
1194 2010-12-30 13:42:44 <tcatm> output something like http://blk.bitcoinwatch.com/b?h=99737
1195 2010-12-30 13:42:53 AntonB_ has quit (Quit: AntonB_)
1196 2010-12-30 13:43:18 <devon_hillard> how large is the BTC trade, daily? $1K?
1197 2010-12-30 13:44:28 <tcatm> devon_hillard: http://bitcoincharts.com/charts/mtgoxUSD#rg90zigDailyzvzcvzm1g10zm2g25 last 3 months, daily mtgox volume in USD
1198 2010-12-30 13:44:31 <genjix> bitcoinwatch.com
1199 2010-12-30 13:53:47 RichardG has quit (Quit: The number you have dialed, 911 is no longer in service, or has been temporarily been disconnected. if you feel you have reaches this recording in error, please check the number and dial again or dial an operator.)
1200 2010-12-30 13:59:23 <ThomasV> sipa: bc-tx-hist.txt means that at least 78% of the bitcoins never circulated
1201 2010-12-30 13:59:44 <sipa> how so?
1202 2010-12-30 13:59:59 <sipa> it doesn't say anything about amounts
1203 2010-12-30 14:00:22 <ThomasV> no, but blocks generated so far have the same amount
1204 2010-12-30 14:00:50 <sipa> so?
1205 2010-12-30 14:01:20 <ThomasV> I mean, on creation
1206 2010-12-30 14:01:31 <davout> you forget tx fees
1207 2010-12-30 14:01:44 <ThomasV> those 78703 blocks have 50 bitcoins
1208 2010-12-30 14:01:49 <sipa> yes
1209 2010-12-30 14:01:52 <ThomasV> tx fees are much smaller than that
1210 2010-12-30 14:02:04 <ThomasV> they can be neglected
1211 2010-12-30 14:02:07 <sipa> but what says they never circulated?
1212 2010-12-30 14:02:30 <sipa> it simply means there are no other transactions done at the time they were generated
1213 2010-12-30 14:02:33 <ThomasV> they were never in a transcation, so they never circulated
1214 2010-12-30 14:02:38 <sipa> no no
1215 2010-12-30 14:02:45 <ThomasV> no ?
1216 2010-12-30 14:03:21 <sipa> i think you think the histogram means something else
1217 2010-12-30 14:03:41 <ThomasV> well, you made it so you know better than me what you put in it :-)
1218 2010-12-30 14:03:50 <ThomasV> please explain
1219 2010-12-30 14:04:01 <sipa> those 78703 blocks with no transactions
1220 2010-12-30 14:04:24 <sipa> mean that there are 78703 blocks with no transactions in them
1221 2010-12-30 14:04:43 <sipa> it doesn't say anything about what later can have happened with the bitcoins generated in those blocks
1222 2010-12-30 14:04:52 <ThomasV> oh, sorry I forgot that
1223 2010-12-30 14:04:55 <ThomasV> indeed
1224 2010-12-30 14:05:05 <sipa> what you ask is a lot harder
1225 2010-12-30 14:05:20 <ThomasV> yes, but now you understand what I ask :-D
1226 2010-12-30 14:05:55 <tcatm> Are you asking for the number of BTC that were generated but never sent?
1227 2010-12-30 14:06:19 <sipa> i think so
1228 2010-12-30 14:06:21 <ThomasV> tcatm: I would like to know how often bitcoins circulate
1229 2010-12-30 14:06:31 <ThomasV> ideally, with a histogram
1230 2010-12-30 14:06:54 <sipa> you know that a lot of bitcoins circulate within accounts belonging to the same person/wallet?
1231 2010-12-30 14:07:05 <ThomasV> it would show the proportion of bitcoins that were exchanged x times
1232 2010-12-30 14:07:08 <tcatm> ThomasV: How should that look like?
1233 2010-12-30 14:07:28 <ThomasV> sipa: yes I know, it would not include transactions at mtgox
1234 2010-12-30 14:07:37 <sipa> that too
1235 2010-12-30 14:07:46 <ThomasV> and the same person can have several wallets too
1236 2010-12-30 14:08:13 <sipa> but if i have 50 BTC in my wallet, and i send you 10 BTC, the client may send 40 BTC to another account within my own wallet, just to obscure things
1237 2010-12-30 14:08:26 <sipa> within the same transaction
1238 2010-12-30 14:08:39 <ThomasV> sipa : that's what it always does, afaik
1239 2010-12-30 14:08:47 <sipa> if possible, yes
1240 2010-12-30 14:08:59 <tcatm> Yep, the client has to split coins most of the time
1241 2010-12-30 14:09:15 <ThomasV> but it would still give us an indication
1242 2010-12-30 14:10:11 <ThomasV> basically, we need to know if the economy is paralyzed by deflation
1243 2010-12-30 14:10:31 <ThomasV> that's why I would like to see this kind of figures
1244 2010-12-30 14:10:31 <sipa> ThomasV: there is also no tracing of actual bitcoins
1245 2010-12-30 14:10:50 <sipa> let's say three accounts exist, A, B and C
1246 2010-12-30 14:11:09 <sipa> B and C generate a block, they get 50 BTC each
1247 2010-12-30 14:11:35 <sipa> so the balances are (50,50,0)
1248 2010-12-30 14:11:52 <sipa> oh (0,50,50)
1249 2010-12-30 14:11:59 <sipa> B sends 10 to A, we get (10,40,50)
1250 2010-12-30 14:12:15 <sipa> C then sends 20 BTC to A, we get (30,40,30)
1251 2010-12-30 14:13:32 <sipa> now A sends everything to another person D, we get (0,40,30,30)
1252 2010-12-30 14:14:40 <sipa> oh wait, never mind!
1253 2010-12-30 14:14:42 <lfm> ya its more like bank accounts than wallets. only the balance is important. it doesnt matter what size transactions were
1254 2010-12-30 14:20:10 <sipa> ThomasV: so you basically want to trace each fraction of a BTC, and see how many times it was transferred?
1255 2010-12-30 14:20:30 <ThomasV> yes
1256 2010-12-30 14:21:22 <lfm> use the block browser
1257 2010-12-30 14:21:24 <ThomasV> it would give a similar histogram
1258 2010-12-30 14:21:43 <sipa> what if a transaction has 2 inputs A and B, and one output C, and another transaction that takes C, and sends it to D en E
1259 2010-12-30 14:22:13 <ThomasV> then all BTC have 2 transactions
1260 2010-12-30 14:22:22 <sipa> ok
1261 2010-12-30 14:22:42 <sipa> but imagine the BTC at A has already 2 transactions
1262 2010-12-30 14:22:49 <sipa> and the one at B only has 1
1263 2010-12-30 14:23:17 <sipa> then C's money has 2.5 transactions?
1264 2010-12-30 14:23:18 <lfm> all transactions can have many inputs and 1 or 2 or occactionaly more outputs
1265 2010-12-30 14:23:29 <ThomasV> then you need to weight
1266 2010-12-30 14:23:43 <sipa> yes, assuming the amount is equal
1267 2010-12-30 14:23:55 <sipa> and D and E's money eventually has 3.5
1268 2010-12-30 14:23:56 <ThomasV> weight by the amount
1269 2010-12-30 14:24:05 <sipa> ok, sounds interesting
1270 2010-12-30 14:24:25 <sipa> but you would get a strange histogram
1271 2010-12-30 14:24:35 <sipa> with a lot of different fractional amounts
1272 2010-12-30 14:25:16 <lfm> what do you mean "paralysed by deflation?
1273 2010-12-30 14:26:05 <ThomasV> sipa: it is not as informative as if we knew which output belongs with which input, but I guess it is the best we can do
1274 2010-12-30 14:26:24 <ThomasV> lfm: I happen to believe that deflation is a crappy idea
1275 2010-12-30 14:27:08 <ThomasV> if bitcoin is successful, there will be bitcoin variants that have built-in inflation rates
1276 2010-12-30 14:27:58 <lfm> well I guess first of all you need to really determine IF deflation is happening.
1277 2010-12-30 14:28:14 <ThomasV> no you do not
1278 2010-12-30 14:28:50 <x6763> price deflation or monetary deflation?
1279 2010-12-30 14:29:01 <ThomasV> monetary mass
1280 2010-12-30 14:29:26 <ThomasV> well, with btw monetary mass is constant, but this leads to price deflation
1281 2010-12-30 14:29:28 <lfm> ok you lost me
1282 2010-12-30 14:29:44 <ThomasV> monetary mass is 21 million
1283 2010-12-30 14:29:48 <ThomasV> it is fixed
1284 2010-12-30 14:29:50 <lfm> not yet
1285 2010-12-30 14:29:51 <x6763> so?
1286 2010-12-30 14:30:04 <ThomasV> it is not already mined
1287 2010-12-30 14:30:07 <ThomasV> but it is there
1288 2010-12-30 14:30:17 <lfm> only have 6 million or so so far
1289 2010-12-30 14:30:48 <ThomasV> yeah, but you understand what I mean
1290 2010-12-30 14:31:11 <ThomasV> the total monetary mass is programmed to tend towards 21 million
1291 2010-12-30 14:31:15 <lfm> the potential is fixed at 21e6 ya
1292 2010-12-30 14:32:02 <ThomasV> in the same way, we could program another currency, a bitcoin variant, where the total monetary mass increases every year by 2%
1293 2010-12-30 14:32:14 <ThomasV> that would be "BTC2"
1294 2010-12-30 14:32:24 <ThomasV> ..or by 5%, btc5
1295 2010-12-30 14:32:28 <x6763> solving what problem?
1296 2010-12-30 14:32:45 <ThomasV> solving the problem that deflation is very bad for an economy
1297 2010-12-30 14:32:50 <x6763> how so?
1298 2010-12-30 14:33:06 MJD has quit (Ping timeout: 255 seconds)
1299 2010-12-30 14:33:10 <ThomasV> well, some people believe it is
1300 2010-12-30 14:33:22 <x6763> if demand for bitcoins increases, prices in terms of bitcoins will decrease...if demand for bitcoins decreases, prices in terms of bitcoins will increase
1301 2010-12-30 14:33:31 <x6763> where's the problem?
1302 2010-12-30 14:33:41 <lfm> well if you start with the premis of "deflation is bad" then you can say bitcoin is bad but we dont understand why you think defaltion is bad in the first place
1303 2010-12-30 14:33:43 <ThomasV> the problem is the real world my dear
1304 2010-12-30 14:34:03 <lfm> no need for the condecention
1305 2010-12-30 14:34:04 <x6763> more valuable money is a problem?
1306 2010-12-30 14:34:05 <ThomasV> deflation is an incentive to keep your coins, not to spend them
1307 2010-12-30 14:34:36 <ThomasV> deflation makes it impossible to hire someone for a fixed salary
1308 2010-12-30 14:34:51 <x6763> then maybe hiring people at a fixed salary is a bad idea
1309 2010-12-30 14:35:00 <lfm> so that makes people want to acquire more driving up the demand
1310 2010-12-30 14:35:15 grondilu has joined
1311 2010-12-30 14:35:26 <x6763> a money that increases in value rather than decreases is a different paradigm from government money
1312 2010-12-30 14:35:30 <lfm> not impossible, I bet I could do it
1313 2010-12-30 14:35:38 sgtstein has left ()
1314 2010-12-30 14:36:06 <EvanR-work> ThomasV: bitcoins will be lost unrecoverably as time goes on, leading to inflation
1315 2010-12-30 14:36:12 <EvanR-work> er
1316 2010-12-30 14:36:28 <ThomasV> no, it leads to more deflation
1317 2010-12-30 14:36:34 <EvanR-work> ok
1318 2010-12-30 14:36:40 <lfm> in fact employees wont need to demand "indexed" salery if there is deflation
1319 2010-12-30 14:36:48 <EvanR-work> as coins get generated it leads to inflation as your coins youre holding lose value
1320 2010-12-30 14:37:16 <ThomasV> EvanR-work: only for the moment, because coins are generated
1321 2010-12-30 14:37:59 <EvanR-work> itll be like that for a long time
1322 2010-12-30 14:38:02 <lfm> it makes it easier to hire people for fixed salery cuz deflation gives a built in increase
1323 2010-12-30 14:38:17 <ThomasV> normally, the role of a central bank is to increase the monetary mass, because more physical goods are created each year
1324 2010-12-30 14:38:30 <EvanR-work> thats bull
1325 2010-12-30 14:39:00 <sipa> ThomasV: how do you mean deflation makes it impossible to hire someone for a fixed salary?
1326 2010-12-30 14:39:40 <lfm> thomasv more physical good wear out each year too, thats irrelevant
1327 2010-12-30 14:39:40 <ThomasV> sipa: because, if you are a company and sell goods whose price decrease, your revenues will decrease
1328 2010-12-30 14:39:57 <x6763> yet your spending power remains the same or increases
1329 2010-12-30 14:39:59 <EvanR-work> theres a real reason for the fed
1330 2010-12-30 14:40:11 <lfm> thomasv so dont decrease the price
1331 2010-12-30 14:40:32 <sipa> you may have to decrease the price
1332 2010-12-30 14:40:50 <ThomasV> you will have to decrease prices if your competitors do
1333 2010-12-30 14:40:54 <sipa> indeed
1334 2010-12-30 14:41:07 <ThomasV> tha's what happens when there is deflation
1335 2010-12-30 14:41:13 <x6763> *price* deflation
1336 2010-12-30 14:41:24 <ThomasV> yes, price deflation
1337 2010-12-30 14:41:46 <x6763> so you want your goods to become more and more expensive, making it harder for you to buy things?
1338 2010-12-30 14:41:51 <lfm> thomasv revenue can decrease with deflation and you could still have more buying power with less revenue if you have defaltion
1339 2010-12-30 14:41:54 <EvanR-work> thats what inflation / deflation is referring to, prices
1340 2010-12-30 14:43:00 <ThomasV> lfm: yes, if people were accepting to be hired for a salary that decreases each year. but in general they do not like this idea.
1341 2010-12-30 14:43:08 <devon_hillard> inflation is a system where the government gets to derive value without presenting a proof-of-work chain
1342 2010-12-30 14:43:22 <devon_hillard> it can double-spend at leisure
1343 2010-12-30 14:43:28 <x6763> *price* inflation/deflation makes more sense when you just call them price increases/decreases, since the changes in price are brought about the exact same way, regardless of the change (or lack thereof) in money supply...inflation/deflation makes more sense when you're talking about the amount of currency in the economy
1344 2010-12-30 14:43:41 <larsivi> EvanR-work: only in popular media, infation/deflation _really_ refers to the amount of money available in a given currency
1345 2010-12-30 14:43:54 <ThomasV> x6763: that's not the actual definition
1346 2010-12-30 14:44:01 <lfm> oh by "fixed" salery you mean you want to cut saleries every year indexed to deflation
1347 2010-12-30 14:44:05 <ThomasV> inflation/deflation refers to prices
1348 2010-12-30 14:44:05 <x6763> when gas goes up/down one day, you don't say, "oh, gas inflated today" or "gas deflated today"
1349 2010-12-30 14:44:29 <ThomasV> the amount of money is called monetary mass
1350 2010-12-30 14:44:37 <EvanR-work> larsivi: ok, if theres more money, thats going to drive prices down? and thats deflation?
1351 2010-12-30 14:44:40 <larsivi> since inflation/deflation has an _effect_ on prices, these words are typically used
1352 2010-12-30 14:44:40 <devon_hillard> ThomasV: even when we consider inflation as a price phenomenon, it is at its base driven by money quantity factors
1353 2010-12-30 14:44:44 <ThomasV> that's how it happens to be called
1354 2010-12-30 14:44:47 genjix has quit (Ping timeout: 260 seconds)
1355 2010-12-30 14:44:55 <ThomasV> devon_hillard: not only
1356 2010-12-30 14:44:57 <EvanR-work> larsivi: so how does it refer to the money
1357 2010-12-30 14:45:31 <larsivi> EvanR-work: more money means higher prices (you have more of it and can thus pay more)
1358 2010-12-30 14:45:32 <devon_hillard> ThomasV: only money supply variations can explain long term value variations
1359 2010-12-30 14:45:39 <ThomasV> it is driven by money quantity AND physical goods that are available AND psychological factors
1360 2010-12-30 14:45:58 <larsivi> so inflate the amount of money, and prices rise (assuming no other relevant effects)
1361 2010-12-30 14:46:11 <ThomasV> no
1362 2010-12-30 14:46:14 <EvanR-work> larsivi: i dont have more money, but the country does
1363 2010-12-30 14:46:32 <larsivi> EvanR-work: yes, I didn't say that your amount was relevant
1364 2010-12-30 14:46:43 <EvanR-work> well prices going up in that case would suck for me
1365 2010-12-30 14:46:45 <larsivi> it is the total of some currency that is
1366 2010-12-30 14:47:07 <lfm> thomasv seems to me if an employers insist on indexing saleries to deflation then employees would just have to get used to it
1367 2010-12-30 14:47:30 <larsivi> EvanR-work: sure, but that is typically what has happened in inflated economies (30's germany, to some degree the USA of today)
1368 2010-12-30 14:47:33 <ThomasV> larsivi: if your economy produces 5% more goods each year, and if you increase the amount of money by only 3%, you will see price deflation, not inflation
1369 2010-12-30 14:47:47 <ThomasV> even though you increased the amount of money
1370 2010-12-30 14:48:05 <larsivi> ThomasV: sure, I said assuming no other effects
1371 2010-12-30 14:48:16 <ThomasV> "no other effects" ?
1372 2010-12-30 14:48:35 <ThomasV> let me remind you that the goal of the economy is to produce goods
1373 2010-12-30 14:48:40 <larsivi> an "ideal" world with no economical growth, no psychological factors, etc
1374 2010-12-30 14:48:41 <ThomasV> not to produce money
1375 2010-12-30 14:48:42 ciuciu has joined
1376 2010-12-30 14:48:52 <larsivi> which economy?
1377 2010-12-30 14:48:56 <EvanR-work> ThomasV: i thought it was to satisfy peoples wants
1378 2010-12-30 14:49:17 <larsivi> I never heard of that particular goal (except possibly from an industry owner)
1379 2010-12-30 14:49:23 <x6763> economies don't have goals, last i knew
1380 2010-12-30 14:49:41 <x6763> only people did, and each person's goals are different
1381 2010-12-30 14:49:50 <lfm> thomasv on the other hand if you see that your money buys more every year instaed of less that is a positive psycological effect if you ask me
1382 2010-12-30 14:49:56 <ThomasV> well, let's say the goal of industry owners and customers who buy things
1383 2010-12-30 14:50:30 <ThomasV> in the end they are the people who use the currency
1384 2010-12-30 14:50:40 <larsivi> the only thing that really matters for people is buying power, which improves also with deflation (you can buy more with the same money)
1385 2010-12-30 14:50:58 <larsivi> the only reason inflation is considered good, is that it makes a somewhat false pretense of growth
1386 2010-12-30 14:51:13 <larsivi> ThomasV: there's a lot of information on this topic in the bitcoin wiki
1387 2010-12-30 14:51:28 <x6763> there's even more info on this topic at mises.org
1388 2010-12-30 14:51:37 <devon_hillard> You have A. >demand-for-goods or B. <demand-for-money or C. <supply-of-goods or D. >supply-of-money. Spending more money on goods (>demand-for-goods) can't go up forever because there is a limit to how much people can spend. Demand for money can't go down forever because people still need to buy goods to live. The supply of goods can't go down forever unless humans are actually wiped out. That leaves only D: only the supply of money can go up for
1389 2010-12-30 14:51:37 <devon_hillard> ever (at least in principle).
1390 2010-12-30 14:51:38 <ThomasV> larsivi: I think economy books are perhaps more relevant than the wiki ?
1391 2010-12-30 14:51:58 <larsivi> there are historically long periods of time with economic growth at the same time as there was deflation
1392 2010-12-30 14:52:23 <lfm> thomasv in the end bitcoin will not stand alone. it will compete with all the other currencies old and new out there. if people use it fine, if they dont well then maybe defaltion doesnt work. I think its worth a try
1393 2010-12-30 14:52:25 <larsivi> ThomasV: well, you don't need to read a full book to understand the simpler mechanics, and the wiki has some good summaries
1394 2010-12-30 14:52:49 <ThomasV> anyway...
1395 2010-12-30 14:53:26 <ThomasV> I am not trying to convince you guys ; I just want to know if some people share my silly opinion
1396 2010-12-30 14:53:30 <larsivi> theory has it that deflation is better for society in general, as it favours smart and not so risky investments
1397 2010-12-30 14:53:58 <larsivi> I guess I didn't get your silly opinion :P
1398 2010-12-30 14:53:58 <ThomasV> theory...
1399 2010-12-30 14:54:14 <sipa> ThomasV: i find it reasonable that people do not want salaries to go down each year, psychologically
1400 2010-12-30 14:54:24 <ThomasV> my silly opinion is that there should be a variant of BTC with built-in inflation
1401 2010-12-30 14:54:25 <devon_hillard> it actually has to do with the government's ability to convert pieces of paper into actual goods
1402 2010-12-30 14:54:27 <larsivi> ThomasV: sure (and history verifies that theory to some degree)
1403 2010-12-30 14:54:31 <lfm> thomasv there are probably some people who share your point of view yes. they may be rare here cuz bitcoin probably attract people who dislike inflation
1404 2010-12-30 14:54:42 <sipa> ThomasV: with interest?
1405 2010-12-30 14:55:15 <ThomasV> sipa: with a predefined increase of the monetary mass, for example 2% per year
1406 2010-12-30 14:55:23 <sipa> yes
1407 2010-12-30 14:55:29 <sipa> but where do you increase it?
1408 2010-12-30 14:55:32 <larsivi> BTC with inflation would mean that there wouldn't be a roof on the total amount of bitcoins
1409 2010-12-30 14:55:32 <ThomasV> miners
1410 2010-12-30 14:55:42 <lfm> devon the peices of paper are pretty irrelevant. if they mattered a $100 bill would be worth the same as a $1 bill
1411 2010-12-30 14:55:46 <ThomasV> no there would be no roof
1412 2010-12-30 14:56:03 <devon_hillard> instead of doubling difficulty on every turn, you could increase it by a suitable factor to favor low inflation
1413 2010-12-30 14:56:06 <ThomasV> (except that the speed is controlled)
1414 2010-12-30 14:56:16 <lfm> sipa it would be a small change to make bitcoin inflationary
1415 2010-12-30 14:56:18 Zarutian has joined
1416 2010-12-30 14:57:23 <devon_hillard> the advantage is that bitcoin growth can be controlled and limited algorithmically
1417 2010-12-30 14:57:31 <devon_hillard> a government won't be able to restrain itself
1418 2010-12-30 14:57:38 <ThomasV> sure
1419 2010-12-30 14:58:06 <ThomasV> but you can use this to increase the monetary mass too, algorithmically
1420 2010-12-30 14:58:11 <lfm> and totally predictably, unlike fiat currencies which get manipulated for political purposes
1421 2010-12-30 14:58:18 <ThomasV> exactly
1422 2010-12-30 14:58:37 <sipa> ThomasV: but currently, there is a 69%/year inflation in bitcoin
1423 2010-12-30 14:58:47 <lfm> so its irrelevant if its deflation or inflation really, its just what you're used to
1424 2010-12-30 14:58:53 <sipa> which decreases of course continually
1425 2010-12-30 14:59:30 <ThomasV> sipa: currently the monetary mass increases, but also the number of users
1426 2010-12-30 14:59:32 <larsivi> there is no point in bitcoin being inflationary, since it wouldn't really really benefit anyone but miners (which could possibly be smart to have miners keep generating blocks, but transaction costs are meant to have that role)
1427 2010-12-30 14:59:37 <lfm> sipa whered that figure come from?
1428 2010-12-30 14:59:47 <devon_hillard> OK, speaking of CUDA vs. OpenCL is it simply NVidia vs. ATI, or would NVidia be planning OpenCL-optimized GPUs in the future at some point? Effectively leading to two production branches.
1429 2010-12-30 15:00:06 riush has joined
1430 2010-12-30 15:00:29 <sipa> lfm: 50*86400/600 BTC per day, 5000000 BTC in total
1431 2010-12-30 15:00:31 <larsivi> devon_hillard: CUDA vs OpenCL isn't really a hardware issue
1432 2010-12-30 15:00:49 <sipa> lfm: (1+50*86400/600/5000000)^365 yearly
1433 2010-12-30 15:00:51 <lfm> devon_hillard nvidia supports opencl now
1434 2010-12-30 15:01:00 <sipa> which is 1.69
1435 2010-12-30 15:01:05 <devon_hillard> lfm: yes, but apparently ATI is better in opencl
1436 2010-12-30 15:01:23 <lfm> devon_hilliard, yes it is a hardware thing
1437 2010-12-30 15:01:33 <sipa> no ATI is better in opencl for the purpose of very parallel integer code
1438 2010-12-30 15:02:25 <lfm> devon_hilliard ati just happens to run bitcoin mining faster is all, nothing mysterious
1439 2010-12-30 15:03:22 <lfm> miners written for cuda arnt significantly faster
1440 2010-12-30 15:03:37 <ThomasV> lfm: when you buy VPS hosting, you generally want to pay a fixed amount per month.
1441 2010-12-30 15:03:59 <lfm> if you can, sure. depends what is available
1442 2010-12-30 15:04:09 <ThomasV> when you sell it too
1443 2010-12-30 15:04:19 <lfm> thomasv in fact Id rather pay LESS each month
1444 2010-12-30 15:04:20 akem has joined
1445 2010-12-30 15:04:55 <lfm> if I was selling Id rather they pay more each month, sure, so what?
1446 2010-12-30 15:05:22 <x6763> each party agrees on a price for a period of time, and both are compromising...who cares?
1447 2010-12-30 15:05:29 <sipa> by the way
1448 2010-12-30 15:05:37 <EvanR-work> is that 92 billion coin transfer in the blockchain somewhere?
1449 2010-12-30 15:05:38 <sipa> in the assumption that mining remains profitable
1450 2010-12-30 15:06:04 <sipa> either by decreasing number of miners, or through transaction fees
1451 2010-12-30 15:06:06 <lfm> evanr-work not any more
1452 2010-12-30 15:06:10 <EvanR-work> oh
1453 2010-12-30 15:06:39 <EvanR-work> can i see a record somewhere?
1454 2010-12-30 15:06:40 <sipa> i assume that eventually people will invest in mining
1455 2010-12-30 15:06:55 <devon_hillard> has anyone played with DirectCompute so far?
1456 2010-12-30 15:06:59 <lfm> evenr-work that was the point of the fix to unwind the bad transaction out of the block chain
1457 2010-12-30 15:07:44 <sipa> if it was done at much lower gains than now, people will need to invest a lot to do it efficiently
1458 2010-12-30 15:07:53 <sipa> eg. design specialized hardware for it
1459 2010-12-30 15:08:05 <lfm> sipa people ARE investing in mining every day, thats why the difficulty keeps increasing
1460 2010-12-30 15:08:19 <sipa> lfm: yes of course, but not on the scale i mean
1461 2010-12-30 15:08:44 <sipa> so, you could have 'banks' where you put money (BTC), and tell you they pay you back after X months
1462 2010-12-30 15:08:53 <sipa> including an interest
1463 2010-12-30 15:09:02 <lfm> sipa custom hardware may never be profitable.
1464 2010-12-30 15:09:21 <sipa> maybe, maybe not
1465 2010-12-30 15:09:43 <EvanR-work> i heard there is a bitcoin bank in tor
1466 2010-12-30 15:09:46 <lfm> sipa yes, bitcoin banks and loansharks are possible
1467 2010-12-30 15:11:17 <devon_hillard> loanshark is a loaded word
1468 2010-12-30 15:11:20 <EvanR-work> isnt there more 'financial' money out there than actual dollars?
1469 2010-12-30 15:11:28 <devon_hillard> call it "lender of last resort for the common people"
1470 2010-12-30 15:11:29 <EvanR-work> how is that possible with bitcoin?
1471 2010-12-30 15:11:30 <sipa> anyone know that the 'coinbase' field is in the getblock json output?
1472 2010-12-30 15:11:47 <sipa> *what
1473 2010-12-30 15:12:10 <devon_hillard> the threat of broken kneecaps is what's keeping interest rates low, relatively speaking :)
1474 2010-12-30 15:12:13 larsivi has quit (Read error: Operation timed out)
1475 2010-12-30 15:12:20 <lfm> evan-work yes, paper money is no doubt the vast minority of money at least in developed nations
1476 2010-12-30 15:13:21 <EvanR-work> is that possible with bitcoins?
1477 2010-12-30 15:13:39 <EvanR-work> banks accounting has so much coins, the sum total of such is more than 21M?
1478 2010-12-30 15:14:18 <EvanR-work> seems like an error
1479 2010-12-30 15:14:27 <lfm> yes I suspect fractional reserve banking would be possible with btc too
1480 2010-12-30 15:14:42 <EvanR-work> but i do know they put the expected future value of a house on their accounting
1481 2010-12-30 15:15:37 <lfm> yes net future value is standard accounting procedure
1482 2010-12-30 15:16:25 <EvanR-work> so somehow total bitcoin wealth will go past 21M
1483 2010-12-30 15:16:52 <sipa> the sum of bank accounts in BTC may go above 21M
1484 2010-12-30 15:17:07 <sipa> but i don't think it can grow exponentially
1485 2010-12-30 15:17:10 <lfm> well the number of coins in an economy doesnt really relate to the gnp or whatever
1486 2010-12-30 15:17:35 <lfm> the circulation rate comes into effect
1487 2010-12-30 15:18:07 <lfm> a coin can be spent and baught many times every year
1488 2010-12-30 15:19:30 <lfm> there is no real limit to a circulation rate thus no limit to the size of an economy set by the amoun t of money in circulkation
1489 2010-12-30 15:20:15 xelister has joined
1490 2010-12-30 15:20:45 EvanR-work has quit (Quit: leaving)
1491 2010-12-30 15:21:13 <lfm> ok well in btc I guess there is a circulation limit, the max is 21 million coins spent every 10 minutes
1492 2010-12-30 15:21:25 <hundfred> e
1493 2010-12-30 15:21:42 <xelister> [2010-12-30_14:37:42] 346 Mhash/sec (cycle=21960 attempts=490351)
1494 2010-12-30 15:21:43 <xelister> [2010-12-30_14:37:42] error: [calc] 1-Cypress-core2: INVALID BLOCK found. nonce=16777215 G=168361 H=3958514400
1495 2010-12-30 15:21:45 <xelister> woot?
1496 2010-12-30 15:21:53 AntonB_ has joined
1497 2010-12-30 15:22:18 <lfm> but if you add in things like mybitcoin.com and mtgox accoutns they can spend and accept btc at much faster rates~
1498 2010-12-30 15:22:27 <x6763> the fact that prices can change is what removes any limits (not in monetary terms, but in the amount of goods exchanged for that money)...it's the goods being traded that matters, not the amount of money or the rate at which that money is exchanged
1499 2010-12-30 15:23:47 <lfm> yes, i am assuming the transactions are btc for some services or products. the number of coins in the economy does not limit the number of transactions or their total value
1500 2010-12-30 15:28:10 <lfm> of course its silly to expect a total circulation of 21 million coins every 10 min. some money doesnt circulate while some can actually circulate faster than the block chain in the side markets like mtgox
1501 2010-12-30 15:30:39 <lfm> so it would be possible to have 2100 million btc in transactions on say mygox in less than the time the main block chain issues one block
1502 2010-12-30 15:31:19 <lfm> mybe not too likely but possible.
1503 2010-12-30 15:33:46 larsig has quit (Ping timeout: 240 seconds)
1504 2010-12-30 15:38:52 <ThomasV> mtgox really quiet today
1505 2010-12-30 15:41:05 paul0 has joined
1506 2010-12-30 15:47:13 larsig has joined
1507 2010-12-30 15:49:32 ThomasV has quit (Ping timeout: 240 seconds)
1508 2010-12-30 15:50:43 cdecker has joined
1509 2010-12-30 15:58:09 ThomasV has joined
1510 2010-12-30 16:02:27 johndrinkwater has quit (Ping timeout: 265 seconds)
1511 2010-12-30 16:13:35 <tcatm> I made "real" bitcoin: http://bitcoincharts.com/media/stuff/bitcoin.jpg
1512 2010-12-30 16:13:36 jyaworski has joined
1513 2010-12-30 16:16:11 <cdecker> Cool :D
1514 2010-12-30 16:16:17 <cdecker> Are you going to sell it?
1515 2010-12-30 16:16:20 <cdecker> xD
1516 2010-12-30 16:17:05 <tcatm> Would someone buy it? :)
1517 2010-12-30 16:17:34 <ThomasV> is it backed by a real one ?
1518 2010-12-30 16:17:52 <ThomasV> I mean, a virtual one, of course
1519 2010-12-30 16:17:55 <cdecker> That would be cool
1520 2010-12-30 16:18:04 <tcatm> It's backed by nothing except half an hour of time :D
1521 2010-12-30 16:18:17 <cdecker> Microdots with the private key of an address :D
1522 2010-12-30 16:18:29 <tcatm> Though I'm thinking engraving a hash on the back
1523 2010-12-30 16:18:53 <cdecker> Oh even better embed an RFID with the private key :D
1524 2010-12-30 16:19:28 <tcatm> Yep, had that thought, too.
1525 2010-12-30 16:19:36 <cdecker> Nice
1526 2010-12-30 16:19:49 <ThomasV> write the private key on the edge
1527 2010-12-30 16:19:56 <tcatm> Next step will be a prettier surface
1528 2010-12-30 16:20:23 <ThomasV> what is it made of ?
1529 2010-12-30 16:20:53 larsivi has joined
1530 2010-12-30 16:21:00 <cdecker> Would be even better if you had to destroy the coin to get to the private key
1531 2010-12-30 16:21:02 <tcatm> Copper + glass epoxy
1532 2010-12-30 16:21:24 <cdecker> But use RFID to sign challenges verifying the authenticity
1533 2010-12-30 16:21:27 <cdecker> :D
1534 2010-12-30 16:21:32 <tcatm> I.e. printed circuit board material :P
1535 2010-12-30 16:21:39 EvanR-work has joined
1536 2010-12-30 16:22:22 <ThomasV> what size is it ?
1537 2010-12-30 16:22:35 <wumpus> devon_hillard: opencl is roughly based on cuda, both languages are very similar and target the same kinds of hardware
1538 2010-12-30 16:22:38 <tcatm> 24x1.5mm
1539 2010-12-30 16:35:47 davout has quit (Quit: i <3 pork (http://dev.ojnk.net))
1540 2010-12-30 16:36:17 grondilu has quit (Quit: leaving)
1541 2010-12-30 16:37:16 <EvanR-work> lol the pool just found a block after 83 shares
1542 2010-12-30 16:37:28 <EvanR-work> if you have at least 1 share, you now have the mother load ;)
1543 2010-12-30 16:38:34 <tcatm> http://a59.img-up.net/bitcoin2d08rj.jpg better picture + size comparision
1544 2010-12-30 16:40:19 <ThomasV> tcatm: nice !
1545 2010-12-30 16:43:58 <devon_hillard> EvanR-work: sweet!
1546 2010-12-30 16:44:11 <EvanR-work> and i wasnt running the miner :(
1547 2010-12-30 16:44:13 <EvanR-work> lol
1548 2010-12-30 16:50:35 saltyballs has joined
1549 2010-12-30 17:08:10 TD has joined
1550 2010-12-30 17:09:50 shortcircuit has quit (Remote host closed the connection)
1551 2010-12-30 17:10:12 cdecker has left ()
1552 2010-12-30 17:12:07 <mizerydearia> $ bitcoind gettransaction 0x36c86ca499275945a3f1805d6e070ac00b2d32e5e16a4c9cf2c83b189a347d88 => error: {"code":-5,"message":"Invalid transaction id"}
1553 2010-12-30 17:12:20 <mizerydearia> I obtained the txid from http://www.bitcoin.org/smf/index.php?topic=2343.msg31365#msg31365 - How come it doesn't work?
1554 2010-12-30 17:12:23 <bitbot> [RFC] reporting addresses, listtransactions/gettransaction : gavinandresen: We're pretty close to being able to duplicate the bitcoin GUI using the RPC. I think the only thing missing is reporting bitcoin addresses in listtransactions. That gets a little sticky with multisends (one transactions that sends coins to multiple recipients); here's what I'm thinking of implem...
1555 2010-12-30 17:14:52 <EvanR-work> got the new cpuminer, cryptopp_asm32 gives me 2x speed boost
1556 2010-12-30 17:16:40 <EvanR-work> 1.8M total, using 2 threads
1557 2010-12-30 17:16:55 <mizerydearia> Ah, I think the id must exist from my wallet.
1558 2010-12-30 17:17:17 <mizerydearia> yep
1559 2010-12-30 17:18:40 <Kiba> mizerydearia: how the pizza project going?
1560 2010-12-30 17:18:53 <mizerydearia> Delayed, working on other projects atm.
1561 2010-12-30 17:19:11 shortcircuit has joined
1562 2010-12-30 17:19:14 <mizerydearia> I will soon release a project I am working. It shall be quite exciting I think.
1563 2010-12-30 17:19:14 saltyballs has quit (Ping timeout: 265 seconds)
1564 2010-12-30 17:20:37 * tcatm just polished his bitcoin. The glass epoxy almost looks like a hologram now.
1565 2010-12-30 17:20:58 <ThomasV> holograms are good
1566 2010-12-30 17:21:09 <ThomasV> it means that they cannot be fake
1567 2010-12-30 17:21:25 <EvanR-work> lol
1568 2010-12-30 17:21:28 <tcatm> It's not a hologram ;)
1569 2010-12-30 17:22:08 <tcatm> It's just the tiny glass fibers within the epoxy that confuse the eyes so it looks like the B is floating in the air.
1570 2010-12-30 17:27:32 <Kiba> tcatm: nice way to imbube bitcoin intristic value
1571 2010-12-30 17:27:36 <Kiba> with intristic value!
1572 2010-12-30 17:27:42 <Kiba> make it with copper!
1573 2010-12-30 17:27:46 <BoBeR> http://www.theregister.co.uk/2010/12/30/ps3_jailbreak_hack/
1574 2010-12-30 17:27:59 <EvanR-work> intrinsic value >_<
1575 2010-12-30 17:28:02 BoBeR has quit (Quit: (.)(.)
1576 2010-12-30 17:28:24 <tcatm> Kiba: Brass would be a better choice I think.
1577 2010-12-30 17:29:17 ThomasV has quit (Ping timeout: 260 seconds)
1578 2010-12-30 17:29:29 <EvanR-work> cpuminer failed, and reconnected! \o/
1579 2010-12-30 17:29:31 <EvanR-work> success
1580 2010-12-30 17:34:24 ThomasV has joined
1581 2010-12-30 17:41:15 marioxcc has joined
1582 2010-12-30 17:41:26 <marioxcc> hi
1583 2010-12-30 17:41:33 <marioxcc> do someone run a ASIC miner?
1584 2010-12-30 17:46:13 <xelister> well... I run an ASCII miner
1585 2010-12-30 17:46:21 <xelister> O _
1586 2010-12-30 17:46:23 <xelister> \|/ '|`
1587 2010-12-30 17:46:24 <xelister> V |
1588 2010-12-30 17:46:25 <xelister> / \ |
1589 2010-12-30 17:46:49 <xelister> on my C64, he does 0.5 hps
1590 2010-12-30 17:46:51 <tcatm> http://y72.img-up.net/bitcoin3d6phb.jpg high-res picture (Need to clean it better next time)
1591 2010-12-30 17:47:24 <xelister> tcatm: watch out for FBI you counterfeti terrorist
1592 2010-12-30 17:49:12 AAA_awright has joined
1593 2010-12-30 17:49:50 <marioxcc> xelister: nice
1594 2010-12-30 17:50:04 <marioxcc> but ASIC=application specific IC
1595 2010-12-30 17:50:30 <xelister> yeah. ask ArtForz he ordered fabrication in some factory afair
1596 2010-12-30 17:50:36 <marioxcc> tcatm: sweet, it's an etched PCB?
1597 2010-12-30 17:50:57 <tcatm> marioxcc: It's milled, but yep, it's a PCB.
1598 2010-12-30 17:51:01 <marioxcc> ok
1599 2010-12-30 17:51:25 AAA_awright_ has quit (Ping timeout: 264 seconds)
1600 2010-12-30 17:52:14 <mizerydearia> It appears the number parameter for "listaccounts" method does not function properly. Is it just me or do others notice the same?
1601 2010-12-30 17:57:54 <EvanR-work> correct
1602 2010-12-30 17:58:09 <EvanR-work> ive been having several difficulties with bitcoind
1603 2010-12-30 17:59:17 <nanotube> tcatm: nice coin :)
1604 2010-12-30 18:00:26 <EvanR-work> tcatm: how big is that ?
1605 2010-12-30 18:01:18 <nanotube> EvanR-work: about the size of a 1euro coin. (see earlier image)
1606 2010-12-30 18:01:49 <EvanR-work> oh
1607 2010-12-30 18:02:55 <EvanR-work> its probably worth a lot more than 1 coin ;)
1608 2010-12-30 18:02:55 <tcatm> EvanR-work: 24x1.5mm
1609 2010-12-30 18:03:44 dwdollar1 has left ()
1610 2010-12-30 18:05:28 <tcatm> I don't think it's worth more than 1 BTC ;)
1611 2010-12-30 18:07:01 <marioxcc> well, it's trivial to do if you have a way to mask the PCB
1612 2010-12-30 18:07:07 <marioxcc> and you have an etching agent
1613 2010-12-30 18:07:10 <marioxcc> but not everyone does
1614 2010-12-30 18:07:26 <tcatm> And have the gerber files
1615 2010-12-30 18:08:43 <EvanR-work> tcatm: if you have an auction, i think youll get more
1616 2010-12-30 18:09:24 <ThomasV> I'm bidding 1BTC
1617 2010-12-30 18:09:40 <tcatm> ThomasV: That doesn't even cover shipping :P
1618 2010-12-30 18:10:00 <ThomasV> sure, but there has to be a first bid
1619 2010-12-30 18:10:50 <marioxcc> ThomasV: there is a minimum bid too
1620 2010-12-30 18:11:35 <tcatm> min bid: 40 BTC :)
1621 2010-12-30 18:12:12 Diablo-D3 has joined
1622 2010-12-30 18:12:39 <tcatm> end of auction: block 101000
1623 2010-12-30 18:12:43 AAA_awright_ has joined
1624 2010-12-30 18:14:05 <EvanR-work> youre actually selling it?
1625 2010-12-30 18:14:42 <ThomasV> will you put it on biddingpond.com ?
1626 2010-12-30 18:14:54 AAA_awright has quit (Ping timeout: 265 seconds)
1627 2010-12-30 18:15:28 <tcatm> EvanR-work: yep
1628 2010-12-30 18:15:41 <tcatm> it has a lot of scratches on the back, though
1629 2010-12-30 18:16:05 <ThomasV> that makes it look authentic
1630 2010-12-30 18:16:18 <EvanR-work> counterfeit bitcoin!
1631 2010-12-30 18:16:42 <EvanR-work> the community will have to raid your place and confiscate it
1632 2010-12-30 18:16:47 <tcatm> ThomasV: does biddingpond support blocks as end times?
1633 2010-12-30 18:16:58 <ThomasV> no idea
1634 2010-12-30 18:17:07 <tcatm> They'd rather confiscate my GPU miners than the coin :P
1635 2010-12-30 18:21:18 <omglolbbq> i bought something on bidding pond once, i won the auction but nothing happened after that...
1636 2010-12-30 18:21:25 <omglolbbq> that was a few months ago though, might have been a bug
1637 2010-12-30 18:21:37 <ThomasV> who runs the site ?
1638 2010-12-30 18:23:28 <sipa> is anyone familiar with the details of how transaction fees are done in transactions/blocks?
1639 2010-12-30 18:23:32 <nanotube> tcatm: see #bitcoin-auction.
1640 2010-12-30 18:23:53 <ThomasV> join #bitcoin-auction
1641 2010-12-30 18:23:56 <sipa> see http://blockexplorer.com/block/00000000560af7bd66afc13012ec3209bf659aa5b305c4eeecb8884d4fa35afb eg
1642 2010-12-30 18:24:04 <nanotube> tcatm: grondilu has had success starting auctions on forum, then taking it over to irc in the last 10-20 blocks.
1643 2010-12-30 18:24:13 <sipa> oh wait
1644 2010-12-30 18:30:37 AAA_awright_ is now known as AAA_awright
1645 2010-12-30 18:33:41 xelister has quit (Remote host closed the connection)
1646 2010-12-30 18:34:18 TD has quit (Quit: TD)
1647 2010-12-30 18:42:26 mndrix has joined
1648 2010-12-30 18:42:26 mndrix has quit (Changing host)
1649 2010-12-30 18:42:26 mndrix has joined
1650 2010-12-30 18:48:11 <afed> getting 157 mhash from each 5770, and 255 mhash from the modded 6950
1651 2010-12-30 18:48:31 <afed> contrary to popular reports, the 6950 is not stable at 6970 clocks
1652 2010-12-30 18:48:35 <afed> perhaps for gaming but not for our purposes
1653 2010-12-30 18:48:56 <afed> or it might be because i'm using a five year old machine and a 500 watt PSU
1654 2010-12-30 18:49:12 <afed> but the shader unlock seems to work flawlessly
1655 2010-12-30 18:55:06 midnightmagic has quit (Quit: midnightmagic)
1656 2010-12-30 18:59:50 <ArtForz> tried clocking only the memory down to 5950 levels?
1657 2010-12-30 19:01:09 <ArtForz> 5950 has 5GHz memory chips, 5970 has 6GHz chips, 5970 stock mem clock is 1350 QDR = 5400MHz
1658 2010-12-30 19:01:30 <marioxcc> ArtForz: you said 5 GHz?
1659 2010-12-30 19:01:37 <ArtForz> yep
1660 2010-12-30 19:02:13 <marioxcc> i think that's marketing BS
1661 2010-12-30 19:02:21 <ArtForz> well, 1.25GHz QDR
1662 2010-12-30 19:02:40 <Kiba> libertarians don't compartmenize. Ever.
1663 2010-12-30 19:02:48 <marioxcc> that sounds more reasonable, but still too high
1664 2010-12-30 19:02:52 <ArtForz> nope
1665 2010-12-30 19:03:21 <Kiba> They are nut to everybody, but that's only because everybody else never bothers to figure out how nuts their own belief system is.
1666 2010-12-30 19:04:01 <Diablo-D3> afed: btw, a lot of 6970s were not binned as 6950 because of broken pipes... they did because they leaked too much power
1667 2010-12-30 19:04:10 <ArtForz> PC memory is very limited by signal integrity issues from having multiple modules ona bus and the reflectiosn caused by module connectors
1668 2010-12-30 19:04:18 <afed> Diablo-D3: ah, that makes sense
1669 2010-12-30 19:04:22 <Diablo-D3> afed: so the 6950s that work with unlocked pipes may STILL fail at full speed
1670 2010-12-30 19:04:40 <afed> so modding the power connect wouldn't do me any good
1671 2010-12-30 19:04:45 <marioxcc> ArtForz: i think it's impractical to run the RAM chips at such high frecuencies
1672 2010-12-30 19:04:46 <Diablo-D3> afed: yield quality for 69xx parts have been very good
1673 2010-12-30 19:04:48 <ArtForz> if you know you only have a single chip on a channel and can properly terminate the bus, 800MHz base clock with DDR3 is possible
1674 2010-12-30 19:04:55 <marioxcc> instead they run they much lower and use multipliers
1675 2010-12-30 19:05:05 <ArtForz> no... they really run that fast
1676 2010-12-30 19:05:13 <Diablo-D3> afed: you could unlock the 6950 and underclock... you might still get a win here
1677 2010-12-30 19:05:17 <marioxcc> the chips don't
1678 2010-12-30 19:05:22 <ArtForz> yes they do
1679 2010-12-30 19:05:25 <marioxcc> at least not in DDR
1680 2010-12-30 19:05:29 <marioxcc> http://lwn.net/Articles/250967/
1681 2010-12-30 19:05:37 <afed> Diablo-D3: it's running at standard 6950 speed now
1682 2010-12-30 19:06:03 <Diablo-D3> btw, in ddr3, 800mhz base clock under normal voltage IS possible
1683 2010-12-30 19:06:09 <ArtForz> <ArtForz> PC memory is very limited by signal integrity issues from having multiple modules ona bus and the reflectiosn caused by module connectors
1684 2010-12-30 19:06:15 <marioxcc> 800 MHz bus clock
1685 2010-12-30 19:06:16 <Diablo-D3> afed: that might be as far as you're getting without switching to third party cooling
1686 2010-12-30 19:06:19 <afed> actually the memory IS at 6970 standard speed
1687 2010-12-30 19:06:24 <marioxcc> but the ram chips run much lower
1688 2010-12-30 19:06:33 <afed> i'm satisfied with the way it's running
1689 2010-12-30 19:06:41 <marioxcc> for DDR3 it's 1/4 the bus frecuency
1690 2010-12-30 19:06:44 <Diablo-D3> 800 mhz bus clock ddr3 == ddr3-1600
1691 2010-12-30 19:06:49 <ArtForz> yep
1692 2010-12-30 19:06:54 <ArtForz> DDR3@800Mhz base clock, data lanes do 1.6Gbits each
1693 2010-12-30 19:06:55 <afed> wouldn't buy another one though
1694 2010-12-30 19:06:57 <Diablo-D3> ddr3 still only doubles, it doesnt quadruple
1695 2010-12-30 19:07:09 <afed> perhaps i will return it
1696 2010-12-30 19:07:13 <ArtForz> GDDR5 does
1697 2010-12-30 19:07:21 <afed> 5870 is better bang for buck
1698 2010-12-30 19:07:23 <marioxcc> Diablo-D3: I undersand the bus is quad pumped
1699 2010-12-30 19:07:28 <Diablo-D3> marioxcc: not on ddr3.
1700 2010-12-30 19:07:32 <marioxcc> and transmits on both rising and falling edge
1701 2010-12-30 19:07:46 <ArtForz> ddr3 is DDR, Double Data Rate, = double pumped
1702 2010-12-30 19:07:54 <Diablo-D3> marioxcc: ddr3 is the third generation ddr spec
1703 2010-12-30 19:08:12 <Diablo-D3> its double pumped, it uses a 4x length bit buffer, and retains the 240 pin dimm design of ddr2
1704 2010-12-30 19:08:18 <ArtForz> why the fuck GPU mfgs call QDR dram GDDR5 is beyond me
1705 2010-12-30 19:08:23 xelister has joined
1706 2010-12-30 19:08:38 <ArtForz> *sdram
1707 2010-12-30 19:08:40 <Diablo-D3> the gddr family of specs is NOT related to the ddr family
1708 2010-12-30 19:08:51 <marioxcc> i guess
1709 2010-12-30 19:08:55 <marioxcc> i'm talking about the DDR
1710 2010-12-30 19:09:00 paul0 has quit (Quit: paul0)
1711 2010-12-30 19:09:06 <ArtForz> yeah, still having a qdr spec called ddr is annoying
1712 2010-12-30 19:09:09 <Diablo-D3> there _is_ a conventional qdr design
1713 2010-12-30 19:09:26 <Diablo-D3> btw, gddr5 is not quad
1714 2010-12-30 19:09:29 <marioxcc> so are you saying GDDR5 runs at 1.25 GHz?
1715 2010-12-30 19:09:31 <Diablo-D3> its double and double width
1716 2010-12-30 19:10:11 <Diablo-D3> so it has a 8 bit wide prefetch buffer, that has double data lines and is double pumped
1717 2010-12-30 19:10:23 <Diablo-D3> its no more qdr than dual channel ddr is.
1718 2010-12-30 19:10:37 <xelister> Diablo-D3: look at -mining
1719 2010-12-30 19:10:45 <Diablo-D3> if anything, its on-chip dual channel
1720 2010-12-30 19:10:45 <marioxcc> with double pumped you mean it works on falling and rising edge?
1721 2010-12-30 19:10:49 <Diablo-D3> marioxcc: yes
1722 2010-12-30 19:11:05 <marioxcc> so what's the real difference to DDR3?
1723 2010-12-30 19:11:06 <Diablo-D3> true qdr is bottom, rising, top, falling
1724 2010-12-30 19:11:06 <ArtForz> errr. no
1725 2010-12-30 19:11:18 <ArtForz> weird part about GDDR5 is that it effectively doubles clock twice
1726 2010-12-30 19:11:22 <Diablo-D3> marioxcc: different specs
1727 2010-12-30 19:11:25 <Diablo-D3> ArtForz: except it doesnt
1728 2010-12-30 19:11:30 <Diablo-D3> it uses the same core clock and doubles it
1729 2010-12-30 19:11:42 <Diablo-D3> it has 2x data pins.
1730 2010-12-30 19:11:49 <ArtForz> you have 1.25Ghz base clock, 2.5GHz RCK/WCK, and the data lanes transfer on rising and falling edge of [WR]CK
1731 2010-12-30 19:12:03 <marioxcc> ArtForz: ¿1.25 GHz BASE clock?
1732 2010-12-30 19:12:09 <Diablo-D3> yes
1733 2010-12-30 19:12:17 <Diablo-D3> 4.5gbit chips run at 1.25ghz base
1734 2010-12-30 19:12:26 <Diablo-D3> it doubles it, and then sends two things at once due to double data lines
1735 2010-12-30 19:12:49 <ArtForz> more-or-less
1736 2010-12-30 19:12:58 <Diablo-D3> its not qdr
1737 2010-12-30 19:13:03 <marioxcc> so which frecuency runs the DRAM aray?
1738 2010-12-30 19:13:06 <ArtForz> you have double the data lanes, but they're always used alternatingly
1739 2010-12-30 19:13:07 <marioxcc> *array
1740 2010-12-30 19:13:10 <Diablo-D3> marioxcc: the base clock
1741 2010-12-30 19:13:21 <ArtForz> and it's internally wide-as-fuck
1742 2010-12-30 19:13:51 <marioxcc> so the state-of-art DDR3 runs at 200 MHz
1743 2010-12-30 19:14:00 <marioxcc> and you're saying GDDR5 runs at 1.25 GHz
1744 2010-12-30 19:14:01 <Diablo-D3> marioxcc: no
1745 2010-12-30 19:14:09 <Diablo-D3> ddr3-1333 runs at 666
1746 2010-12-30 19:14:19 <ArtForz> state of the art DDR3 runs at 666+
1747 2010-12-30 19:14:26 <marioxcc> 666 bus
1748 2010-12-30 19:14:38 <marioxcc> 166 to DRAM chips
1749 2010-12-30 19:14:39 <ArtForz> iirc seeing DDR3-1600 = 800 base clk and above isnt too uncommon
1750 2010-12-30 19:14:42 <Diablo-D3> (you arent going to find ddr3-1600 that compiles with the 1.5v spec requirement)
1751 2010-12-30 19:14:52 <Diablo-D3> er, complies
1752 2010-12-30 19:14:57 <ArtForz> yep
1753 2010-12-30 19:15:06 <marioxcc> ok
1754 2010-12-30 19:15:08 <Diablo-D3> crank the voltage up and you can get anything you want
1755 2010-12-30 19:15:16 <ArtForz> that isnt really a problem for GPU mfgs though
1756 2010-12-30 19:15:22 <marioxcc> what about propagation delay?
1757 2010-12-30 19:15:30 <marioxcc> and power dicipation
1758 2010-12-30 19:15:33 <Diablo-D3> marioxcc: btw, the main hack betwween the different generations is they keep increasing the prefetch buffer
1759 2010-12-30 19:15:37 <ArtForz> yep
1760 2010-12-30 19:15:46 <Diablo-D3> marioxcc: ddr2 has 2x the prefetch buffer, ddr3 has 2x the prefetch buffer of ddr2
1761 2010-12-30 19:15:52 <ArtForz> access latency keeps more-or-less the same
1762 2010-12-30 19:16:01 <Diablo-D3> marioxcc: which means its 2x easier to synchronize the transfers, but it doubles the latency
1763 2010-12-30 19:16:06 <ArtForz> while streaming bandwidth keeps going up
1764 2010-12-30 19:16:08 <marioxcc> sure
1765 2010-12-30 19:16:17 <Diablo-D3> marioxcc: but if you run the chips 2x faster, the actual latency stays the same
1766 2010-12-30 19:16:39 <marioxcc> but the bandwidth rises, don't it?
1767 2010-12-30 19:16:43 <ArtForz> yep
1768 2010-12-30 19:17:03 <marioxcc> I undestand latency is less of an issues because the wide bus allows to fill complete cache lines
1769 2010-12-30 19:17:09 <ArtForz> and iirc GDDR5 doubles the prefect buffer over DDR3
1770 2010-12-30 19:17:14 <ArtForz> *prefetch
1771 2010-12-30 19:17:30 <Diablo-D3> ddr2 and ddr3, both same fab size, both 800mhz, due to reduced communication noise, you get the same bandwidth, double the latency, but use like 2/3rds the power
1772 2010-12-30 19:18:14 <Diablo-D3> or double the ddr3 speed, double the bandwidth, have about the same time to settle, keep the latency the same, and use about the same power
1773 2010-12-30 19:18:31 <Diablo-D3> ArtForz: gddr4 has 2x the prefetch buffer over ddr3
1774 2010-12-30 19:18:38 <Diablo-D3> except gddr4 didnt import the changes from ddr3
1775 2010-12-30 19:18:54 <Diablo-D3> gddr5 retained that 8 bit wide buffer
1776 2010-12-30 19:19:17 <ArtForz> yeah, it's still double DDR3
1777 2010-12-30 19:19:31 sweatyballs has joined
1778 2010-12-30 19:19:45 <Diablo-D3> increasing the prefetch buffer has side effects... its only useful if they NEED to cut down settle time
1779 2010-12-30 19:20:12 <ArtForz> yep
1780 2010-12-30 19:20:57 <marioxcc> so then what's the reason to double the prefetch buffer?
1781 2010-12-30 19:21:35 <marioxcc> for the manufacturers
1782 2010-12-30 19:21:40 <Diablo-D3> marioxcc: you tick the clock every x bits + overhead time
1783 2010-12-30 19:21:54 <Diablo-D3> marioxcc: doubling it means you need to address the overhead half as much
1784 2010-12-30 19:22:34 <marioxcc> ok
1785 2010-12-30 19:22:43 <Diablo-D3> gddr5 is just a fucking hack
1786 2010-12-30 19:23:02 <ArtForz> but a damn effective one for getting a lot of bandwidth
1787 2010-12-30 19:23:06 <Diablo-D3> yeah
1788 2010-12-30 19:23:14 <Diablo-D3> I bet ddr4 will import that change in the next few years
1789 2010-12-30 19:23:49 <Diablo-D3> marioxcc: basically, and Im really doing gddr5 a disservice here, while its waiting for one buffer to settle, its operating another
1790 2010-12-30 19:24:10 <marioxcc> ok
1791 2010-12-30 19:24:58 <Diablo-D3> so the write clock internally ticks at 2x the doubled base clock (but this doesnt make it qdr)
1792 2010-12-30 19:25:41 <Diablo-D3> because they dont use the same pins, they just have one write completing when another is beginning
1793 2010-12-30 19:25:51 <ArtForz> errr... no, they use the same pins
1794 2010-12-30 19:26:15 <marioxcc> interesting
1795 2010-12-30 19:26:19 <marioxcc> where did you get that information from?
1796 2010-12-30 19:26:31 <ArtForz> DQ0..31 is sampled on both edges of WCK
1797 2010-12-30 19:26:39 <ArtForz> (ina 32-bit chip)
1798 2010-12-30 19:27:01 <Diablo-D3> ArtForz: how many data pins does a gddr4 chip have?
1799 2010-12-30 19:27:08 <Diablo-D3> or pins total on the package?
1800 2010-12-30 19:27:55 <ArtForz> I'd have to check for GDDR4
1801 2010-12-30 19:28:25 <ArtForz> GDDR5 has something like 60 data/control pins
1802 2010-12-30 19:28:30 <Diablo-D3> 67
1803 2010-12-30 19:28:32 <Diablo-D3> with 170 total
1804 2010-12-30 19:29:42 <ArtForz> yep
1805 2010-12-30 19:30:01 <marioxcc> have you considered building a FPGA miner?
1806 2010-12-30 19:30:04 <ArtForz> how the fuck do you want to fit 2 32 bit data buses, clock, addr, control, ... in 67 pins
1807 2010-12-30 19:30:09 <ArtForz> yes, not worth it
1808 2010-12-30 19:30:13 <Diablo-D3> gddr4 is 136 total
1809 2010-12-30 19:30:27 <marioxcc> ArtForz: why not?
1810 2010-12-30 19:30:29 <ArtForz> http://www.bitcoin.org/smf/index.php?topic=2362.msg33869#msg33869
1811 2010-12-30 19:30:33 <bitbot> An estimate of fpga performance : ArtForz: The real issue on FPGA isnt the logic ops(cheap) or the rotates(pretty much free), but the 32-bit adds. A_out = H + s0 + s1 + maj + ch + K + W -> at least 3 level adder tree ((H + s0) + (s1 + maj)) + ((ch + K) + W) Carry chain delay in a single 32-bit adder on a -3 speed grade Spartan6 is ~2ns, so without ANY routing delays...
1812 2010-12-30 19:31:29 <marioxcc> what a pity
1813 2010-12-30 19:33:22 oleaa has joined
1814 2010-12-30 19:33:26 <ArtForz> DES is fast on FPGA because it was DESIGNED to be fast in hardware
1815 2010-12-30 19:34:02 <marioxcc> of course
1816 2010-12-30 19:34:16 <marioxcc> but why is then SHA256 so fast in GPUs?
1817 2010-12-30 19:34:16 <oleaa> If someone offered food in exchange for a digital currency, it might work.
1818 2010-12-30 19:34:28 oleaa has left ()
1819 2010-12-30 19:34:33 <ArtForz> sha2 was designed to be fast on CPUs
1820 2010-12-30 19:34:54 <Diablo-D3> marioxcc: many gpus provide everything sha2 needs in low cost ops
1821 2010-12-30 19:35:02 <Diablo-D3> like radeon 5xxx does rot in I think a single cycle
1822 2010-12-30 19:35:04 <ArtForz> GPUs have a crapload of CPU-like ALUs
1823 2010-12-30 19:35:10 <ArtForz> yep
1824 2010-12-30 19:35:36 <ArtForz> but then rots are pretty much free in hardware
1825 2010-12-30 19:35:49 <Diablo-D3> sha2 is all easy instructions
1826 2010-12-30 19:35:57 <ArtForz> except for the 32-bit adders
1827 2010-12-30 19:36:02 sweatyballs has quit (Ping timeout: 265 seconds)
1828 2010-12-30 19:36:11 <Diablo-D3> ArtForz: well
1829 2010-12-30 19:36:13 <Diablo-D3> thats due to a gpu quirk
1830 2010-12-30 19:36:19 <marioxcc> what's the problem with adding?
1831 2010-12-30 19:36:26 <ArtForz> well, except when you have a crapton of hardIP adders
1832 2010-12-30 19:36:32 <ArtForz> carry logic
1833 2010-12-30 19:36:41 <Diablo-D3> gpus can do 24bit MADs and shit in no time
1834 2010-12-30 19:36:53 <ArtForz> you either have a 1-gate delay per bit or you need carry-lookahead ROMs
1835 2010-12-30 19:37:15 <ArtForz> no, I mean 32-bit adds are what makes sha2 so unimpressive on FPGA
1836 2010-12-30 19:37:28 <marioxcc> but I don't understand
1837 2010-12-30 19:37:36 <marioxcc> why don't this problem hits GPUs too?
1838 2010-12-30 19:37:41 <marioxcc> how do they work around?
1839 2010-12-30 19:38:01 <ArtForz> errr... they have fast hardware adders in the ALU
1840 2010-12-30 19:38:16 darrob has quit (Remote host closed the connection)
1841 2010-12-30 19:38:40 <ArtForz> a single gate on a modern process is really damn fast
1842 2010-12-30 19:38:56 <marioxcc> ArtForz: but can't the same idea of the GPU/CPU adder be implemented on the FPGA?
1843 2010-12-30 19:39:25 <ArtForz> yes, if someone were to create a FPGA with dedicated hardware adders, it would be pretty damn fast
1844 2010-12-30 19:39:52 <ArtForz> you *can* use the DSP MULADD hardware most modern FPGAs have
1845 2010-12-30 19:39:59 <sneak> hi guys
1846 2010-12-30 19:40:05 <marioxcc> hi
1847 2010-12-30 19:40:13 <marioxcc> ArtForz: hmm
1848 2010-12-30 19:40:17 <sneak> ArtForz: wie gehts?
1849 2010-12-30 19:40:29 <ArtForz> but then you only have something like 150 adders on a S6LX150
1850 2010-12-30 19:40:52 <sneak> i borrowed your logarithmic-scale graph of bitcoin aggregate p2p hash rate for a slide yesterday
1851 2010-12-30 19:40:56 <sneak> hope you don't mind, gave you credit :)
1852 2010-12-30 19:41:12 <ArtForz> whoops, 180 DSP hadr-adders
1853 2010-12-30 19:41:49 <Diablo-D3> marioxcc: see, DSPs have gotten to the point where they're just extremely fast common instructions
1854 2010-12-30 19:41:59 <Diablo-D3> marioxcc: they dont think, they dont optimize, they barely have anything but the ALUs
1855 2010-12-30 19:42:08 <ArtForz> yep
1856 2010-12-30 19:42:15 <Diablo-D3> marioxcc: GPU and DSP design has merged
1857 2010-12-30 19:42:23 <marioxcc> oh
1858 2010-12-30 19:42:27 <ArtForz> btw, Distributed FPGA Number Crunching For The Masses sucked
1859 2010-12-30 19:42:34 <Diablo-D3> marioxcc: true GPGPUs are just extremely complex, extremely wide, huge DSPs
1860 2010-12-30 19:42:37 <xelister> ArtForz: why
1861 2010-12-30 19:42:42 <ArtForz> his numbers for DEs on GPUs were off by over an order of magnitude
1862 2010-12-30 19:42:55 <Diablo-D3> marioxcc: radeon 5xxx are basically _perfect_ for what we need
1863 2010-12-30 19:43:06 <ArtForz> bitslicing on a 5970 gets close to 12Gkey/s
1864 2010-12-30 19:43:24 <Diablo-D3> marioxcc: it just happens to implement all the instructions we need, cheaply, and in a way that makes it easy for us to do it
1865 2010-12-30 19:43:48 <xelister> marioxcc: apart from that that ATI sucks donkey cocks about software / drivers
1866 2010-12-30 19:43:52 <marioxcc> DSP?
1867 2010-12-30 19:44:00 <xelister> marioxcc: and disabling xfire shit is pain in the ass
1868 2010-12-30 19:44:00 <Diablo-D3> xelister: bullshit
1869 2010-12-30 19:44:04 <ArtForz> Digital Signal Processing
1870 2010-12-30 19:44:04 <Diablo-D3> ATI's software is fine
1871 2010-12-30 19:44:12 <xelister> Diablo-D3: Ati's drivers are legendary for being shit
1872 2010-12-30 19:44:15 <xelister> pure shit
1873 2010-12-30 19:44:18 <Diablo-D3> ati's, yes
1874 2010-12-30 19:44:19 <Diablo-D3> amd's, no
1875 2010-12-30 19:44:39 <xelister> Diablo-D3: say wat? read again what I written
1876 2010-12-30 19:44:41 <Diablo-D3> and since amd fired half of ati when they bought them
1877 2010-12-30 19:44:51 <Diablo-D3> you havent used a truly ati driver since... oh... 2005
1878 2010-12-30 19:45:01 <Diablo-D3> so go troll elsewhere
1879 2010-12-30 19:45:01 <xelister> well 5970's recent drivers still suck cock
1880 2010-12-30 19:45:08 <Diablo-D3> 10.12 works fine, douchenuts
1881 2010-12-30 19:45:20 <xelister> except ati shit hangs each 1-2 days and nvidia each 1-2 months
1882 2010-12-30 19:45:29 <Diablo-D3> it hangs because your shits overheating
1883 2010-12-30 19:45:43 <Diablo-D3> we already confirmed your card is somewhat retarded
1884 2010-12-30 19:45:53 <xelister> and switching virtual terminals / screen on ati 5970 every 1 in 10 times garble screen (glitches) or textures or just hangs x
1885 2010-12-30 19:46:02 <Diablo-D3> yes; don't do that.
1886 2010-12-30 19:46:07 <Diablo-D3> theres no reason to switch terminals.
1887 2010-12-30 19:46:08 <xelister> Diablo-D3: no bitch, when miner is off and nothing is overheating
1888 2010-12-30 19:46:12 <ArtForz> "doctor, it hurts when I do that"
1889 2010-12-30 19:46:18 <Diablo-D3> ESPECIALLY if you havent turned off KMS
1890 2010-12-30 19:46:18 <xelister> lol. <Diablo-D3> theres no reason to switch terminals.
1891 2010-12-30 19:46:26 <Diablo-D3> fglrx and kms do not handle each other properly
1892 2010-12-30 19:46:53 <xelister> - doctor Diablo, my legs hurt when I walk. - just do not walk! use cars, duh. Simple. Your leg bone cancer is bullshit
1893 2010-12-30 19:47:21 <ArtForz> seriously, why use VTs when you have X up?
1894 2010-12-30 19:47:40 <xelister> to have other users with other X server
1895 2010-12-30 19:47:49 <xelister> to switch between people on one computer
1896 2010-12-30 19:47:52 <xelister> or roles
1897 2010-12-30 19:48:06 <xelister> or access root more securly a bit (only login from secure plain terminal)
1898 2010-12-30 19:48:12 <xelister> (avoid X based troyan)
1899 2010-12-30 19:48:19 <xelister> (if one of users is infected)
1900 2010-12-30 19:48:20 <Diablo-D3> xelister: I never thought that was a good solution
1901 2010-12-30 19:48:20 <xelister> etc
1902 2010-12-30 19:48:37 <Diablo-D3> multiple X servers like that papers over the problem
1903 2010-12-30 19:48:44 <xelister> papers?
1904 2010-12-30 19:48:45 <Kiba> hmm
1905 2010-12-30 19:49:03 <ArtForz> *shrug*
1906 2010-12-30 19:49:09 <xelister> anyway, proposing to not use one of basic linux feauture that is here since ever is plain trolling
1907 2010-12-30 19:49:11 <Diablo-D3> you have to reinitalize the hardware every time you switch users
1908 2010-12-30 19:49:19 <xelister> because of ati's fuckery whne using this feauture
1909 2010-12-30 19:49:23 <Diablo-D3> (or switch between X and VT)
1910 2010-12-30 19:49:34 <Diablo-D3> xelister: then turn KMS off
1911 2010-12-30 19:49:37 <xelister> yeah that is the point. more isolation = increasing security a bit
1912 2010-12-30 19:49:51 <xelister> Diablo-D3: probably I have KMS off if I use binary driver
1913 2010-12-30 19:50:08 <xelister> but it garbles when switching between X users too - between two X sessions
1914 2010-12-30 19:50:27 <xelister> and that is fault of Ati's suckery
1915 2010-12-30 19:50:30 <ArtForz> yes
1916 2010-12-30 19:50:37 <xelister> because it says so IN THEIR OWN DOCUMENTATION NOTES FOR THE DRIVER
1917 2010-12-30 19:50:47 <xelister> that card can fuck up when switching terminals/server
1918 2010-12-30 19:51:00 <ArtForz> actually driver fucks up, not card
1919 2010-12-30 19:51:17 <xelister> so to conclude, ati's drivers suck cock for the above mentioned reason (and because of stupid xfire disabling handling in config)
1920 2010-12-30 19:51:21 <ArtForz> radeon driver handles it just fine
1921 2010-12-30 19:51:43 <ArtForz> but then radeon driver sucks in other interesting and fun ways
1922 2010-12-30 19:51:50 <Diablo-D3> blah blah blah
1923 2010-12-30 19:51:54 <Diablo-D3> xelister: troll elsewhere
1924 2010-12-30 19:51:55 <xelister> ArtForz: what other ways you know uncle Art?
1925 2010-12-30 19:52:00 <Diablo-D3> want it to stop sucking? turn kms off.
1926 2010-12-30 19:52:15 <xelister> Diablo-D3: how to check if it is even off? I probably have it off
1927 2010-12-30 19:52:23 <Diablo-D3> have you turned it off? if not, its on.
1928 2010-12-30 19:52:30 <xelister> then how to?
1929 2010-12-30 19:52:37 <Diablo-D3> add the kernel argument for it.
1930 2010-12-30 19:53:14 <ArtForz> actually I think it's AMDs strategy of supporting open source, privde documentation and make the propertiary drivers suck badly
1931 2010-12-30 19:53:53 TD has joined
1932 2010-12-30 19:54:19 <Diablo-D3> ArtForz: well
1933 2010-12-30 19:54:21 <Diablo-D3> I will admit
1934 2010-12-30 19:54:23 <Diablo-D3> they want fglrx to die
1935 2010-12-30 19:54:27 <Diablo-D3> so does everyone else
1936 2010-12-30 19:54:33 <Diablo-D3> but xelister is just a troll
1937 2010-12-30 19:54:35 <Diablo-D3> so
1938 2010-12-30 19:54:36 <Diablo-D3> ymmv
1939 2010-12-30 19:54:36 <zygf> too bad there's nobody around to make the open drivers not suck :P
1940 2010-12-30 19:54:40 <ArtForz> yep
1941 2010-12-30 19:54:52 <xelister> Diablo-D3 is just a cocksucker
1942 2010-12-30 19:54:55 <Diablo-D3> zygf: you mean the ~$2 million AMD has spent isnt enough?
1943 2010-12-30 19:54:56 <ArtForz> well, radeon works surprisingly damn well for 2D
1944 2010-12-30 19:55:03 * xelister wellcome back real discussion
1945 2010-12-30 19:55:28 <Diablo-D3> hell, mesa even has a prototype opencl impl for gallium
1946 2010-12-30 19:55:32 <ArtForz> yep
1947 2010-12-30 19:55:38 * EvanR-work just finished reading the log pertaining to bitDNS. the afterparty included Diablo-D3 and xelister verbally abusing each other about market strategy, if i remember correctly
1948 2010-12-30 19:55:40 <ArtForz> wonder how they pull that
1949 2010-12-30 19:56:10 <ArtForz> hey, trollfights are fun ;)
1950 2010-12-30 19:56:11 <zygf> I recall seeing discussions on the gallium list about how many features gallium doesn't and might not support, so I don't have much faith in them
1951 2010-12-30 19:56:30 <ArtForz> well, it's all still pretty early in development
1952 2010-12-30 19:56:45 <zygf> s/might not/might never/
1953 2010-12-30 19:56:53 <Diablo-D3> zygf: gallium is just a framework to use the hardware
1954 2010-12-30 19:57:10 <Diablo-D3> zygf: it'll do whatever the underlying hardware supports as long as someone codes that for the gallium backend
1955 2010-12-30 19:57:23 <zygf> precisely, and it needs to expose all the hardware's features >_>
1956 2010-12-30 19:57:29 <Diablo-D3> yes, thats what a driver does.
1957 2010-12-30 19:57:46 <xelister> what is alternative for fglrx?
1958 2010-12-30 19:58:21 <Diablo-D3> xelister: selling your computer and becoming a sherpa somewhere in the andes
1959 2010-12-30 19:58:36 <xelister> how this will change fglrx involvment in linux world?
1960 2010-12-30 19:58:46 <xelister> ok unless I take my father Linus with me to andes
1961 2010-12-30 19:58:49 <xelister> but lets asume not
1962 2010-12-30 20:01:04 <xelister> you've got around 3 seconds. Since, know this: first person to speak below this line gets 1 btc.
1963 2010-12-30 20:01:10 <xelister> Well ok I will transfer to myself ;)
1964 2010-12-30 20:01:38 <EvanR-work> ._.
1965 2010-12-30 20:02:16 <xelister> first person to say when they made out with Diablo-D3's sister for less then 10 usd gets 2 btc; offer stands for 15 seconds
1966 2010-12-30 20:02:19 <xelister> and Diablo-D3 doesnt count
1967 2010-12-30 20:02:29 <xelister> (can't participate)
1968 2010-12-30 20:02:37 <zygf> eww
1969 2010-12-30 20:02:57 <xelister> zygf: what?? Some people are into big bones
1970 2010-12-30 20:02:58 <Diablo-D3> just put him on ignore, I have
1971 2010-12-30 20:03:04 <xelister> he I win ;)
1972 2010-12-30 20:04:03 xelister is now known as retsilex
1973 2010-12-30 20:04:17 <retsilex> ok so I did, yesterday. Did I win anything?
1974 2010-12-30 20:04:45 <retsilex> Diablo-D3: dont be so serious Im sure this xelister lad is just kidding, mate
1975 2010-12-30 20:04:48 <Kiba> man
1976 2010-12-30 20:04:50 <Kiba> I am so hungry
1977 2010-12-30 20:04:59 <Kiba> that I cannot function intellectually
1978 2010-12-30 20:05:17 <EvanR-work> retsilex of course you would say that
1979 2010-12-30 20:05:21 <Diablo-D3> retsilex: my client still lists people changing nicks.
1980 2010-12-30 20:05:31 <retsilex> huh >_>
1981 2010-12-30 20:07:25 Xanie has joined
1982 2010-12-30 20:07:28 ApertureScience has quit (Read error: Connection reset by peer)
1983 2010-12-30 20:09:37 <EvanR-work> oh
1984 2010-12-30 20:09:49 <EvanR-work> i totally thought retselix was a third party ;)
1985 2010-12-30 20:12:32 <retsilex> EvanR-work: want to buy 10000 BTC at just 25.00 USD? it's the standard price nowdays, really
1986 2010-12-30 20:12:48 <EvanR-work> yes
1987 2010-12-30 20:12:52 <retsilex> @ 25
1988 2010-12-30 20:12:58 <retsilex> as in 25 usd/btc
1989 2010-12-30 20:13:07 <EvanR-work> then no
1990 2010-12-30 20:14:00 <retsilex> hm what to do, what to do... better sell bitcoins or hold to them for few more months? (just 800 btc but always something)
1991 2010-12-30 20:15:30 <EvanR-work> maybe sell them at a slow rate, to partially insure yourself against complete collapse of bitcoin
1992 2010-12-30 20:16:06 AAA_awright_ has joined
1993 2010-12-30 20:16:41 gavinandresen has joined
1994 2010-12-30 20:17:13 RazielZ has quit (Ping timeout: 264 seconds)
1995 2010-12-30 20:18:11 AAA_awright has quit (Ping timeout: 255 seconds)
1996 2010-12-30 20:24:02 darsk1ez has quit (Read error: Operation timed out)
1997 2010-12-30 20:24:22 AntonB_ has quit (Quit: AntonB_)
1998 2010-12-30 20:25:52 eureka^ has quit (Read error: Connection reset by peer)
1999 2010-12-30 20:25:54 newsham has quit (Read error: Connection reset by peer)
2000 2010-12-30 20:26:31 AAA_awright_ is now known as AAA_awright
2001 2010-12-30 20:26:38 HarryS has quit (Ping timeout: 260 seconds)
2002 2010-12-30 20:27:07 darkskiez has quit (Read error: Operation timed out)
2003 2010-12-30 20:28:29 <devon_hillard> Diablo-D3: what documentation would you recommend for starting out with OpenCL and Java?
2004 2010-12-30 20:29:50 <edcba> programming for dummies
2005 2010-12-30 20:32:39 eureka^ has joined
2006 2010-12-30 20:36:16 MJD has joined
2007 2010-12-30 20:38:43 newsham has joined
2008 2010-12-30 20:38:55 jyaworski has quit (Ping timeout: 240 seconds)
2009 2010-12-30 20:38:58 mtgox has quit (Ping timeout: 255 seconds)
2010 2010-12-30 20:40:40 darrob has joined
2011 2010-12-30 20:40:56 darsk1ez has joined
2012 2010-12-30 20:42:34 darkskiez has joined
2013 2010-12-30 20:49:55 noagendamarket has joined
2014 2010-12-30 20:50:14 noagendamarket has quit (Changing host)
2015 2010-12-30 20:50:14 noagendamarket has joined
2016 2010-12-30 20:50:36 <Diablo-D3> devon_hillard: I wouldnt.
2017 2010-12-30 20:50:48 <Diablo-D3> devon_hillard: lwjgl's opencl wrapper is not yet compplete
2018 2010-12-30 20:51:26 <Kiba> now that my stomach is filled, I can expend intellectual energy!
2019 2010-12-30 20:55:36 Bth8 has joined
2020 2010-12-30 20:56:25 Bananaphone has joined
2021 2010-12-30 21:07:13 retsilex has quit (Ping timeout: 240 seconds)
2022 2010-12-30 21:23:15 slush_cz has joined
2023 2010-12-30 21:23:48 slush_cz has quit (Client Quit)
2024 2010-12-30 21:24:35 slush has joined
2025 2010-12-30 21:26:54 <EvanR-work> Kiba: stop harming the environment
2026 2010-12-30 21:26:59 <EvanR-work> conserve energy
2027 2010-12-30 21:27:34 theadmin_ has joined
2028 2010-12-30 21:27:35 <lfm> the earth will survive, its the people who are in trouble
2029 2010-12-30 21:28:34 <EvanR-work> as long as we can drive our soccer tanks until the sun goes down forever
2030 2010-12-30 21:28:44 ciuciu has quit (Ping timeout: 276 seconds)
2031 2010-12-30 21:30:11 <EvanR-work> just making sure, bitcoin network right now consists of dedicated miners who are making a serious profit?
2032 2010-12-30 21:31:18 <tcatm> Most don't make serious profit.
2033 2010-12-30 21:31:22 Kiba has quit (Ping timeout: 260 seconds)
2034 2010-12-30 21:32:40 <lfm> evanr-work most are probably losing money (if they pay for their own power)
2035 2010-12-30 21:33:05 <EvanR-work> because the array of ati cards take a lot of power?
2036 2010-12-30 21:33:36 <tcatm> Because those cards have to be bought first..
2037 2010-12-30 21:33:44 <EvanR-work> well yeah
2038 2010-12-30 21:33:46 <lfm> evanr-work huh? if they have array of ati cards they can make a profit. most dont
2039 2010-12-30 21:33:51 <EvanR-work> oh ok
2040 2010-12-30 21:33:51 HarryS has joined
2041 2010-12-30 21:34:18 <EvanR-work> so theres a break even point on power vs coin rate
2042 2010-12-30 21:34:38 <lfm> evanr-work isnt that obvious?
2043 2010-12-30 21:34:44 <EvanR-work> no. thats why im asking
2044 2010-12-30 21:34:58 <EvanR-work> you just confirmed that there are people who make a profit
2045 2010-12-30 21:35:22 <lfm> evanr-work many run miner without considering the costs, just cuz it is interesting
2046 2010-12-30 21:35:28 <EvanR-work> if coin rate was proportional to power, then youd never make a profit
2047 2010-12-30 21:35:53 <lfm> why not?
2048 2010-12-30 21:36:21 <lfm> maybe there are more than one factor
2049 2010-12-30 21:36:36 <EvanR-work> assuming theres an operating point where you make coins that dont cover electricity... then proportionality means you cant do it at any operating point
2050 2010-12-30 21:36:47 <EvanR-work> thats true
2051 2010-12-30 21:37:20 <lfm> the fact is you can make profit mining with just one gpu
2052 2010-12-30 21:37:54 <EvanR-work> oh really
2053 2010-12-30 21:37:56 <zygf> can you?
2054 2010-12-30 21:38:10 <lfm> if it is a good one yes
2055 2010-12-30 21:38:29 <sipa> for now
2056 2010-12-30 21:38:31 <zygf> gribble's calculator doesn't take into account the exponential difficulty growth
2057 2010-12-30 21:38:44 <nanotube> i'm making a profit with cpu mining! (generating on a vps that i pay for anyway, with no extra charge for electricity used).
2058 2010-12-30 21:39:06 <sipa> it's hard to predict that
2059 2010-12-30 21:39:32 <zygf> sure, but it's better to try than assume a constant difficulty :P
2060 2010-12-30 21:39:39 <EvanR-work> nanotube: and i can decrease entropy of part of a system ;)
2061 2010-12-30 21:39:47 <lfm> well ya if you get free power you can make profit at the expense who pays the power bill
2062 2010-12-30 21:40:07 <nanotube> hey, externalities ftw. ;)
2063 2010-12-30 21:40:12 <EvanR-work> lol
2064 2010-12-30 21:40:31 <lfm> zygf you cant assume constant growth either tho
2065 2010-12-30 21:40:53 devon_hillard has quit (Read error: Connection reset by peer)
2066 2010-12-30 21:40:55 <EvanR-work> the calculator is more useful for shorter term
2067 2010-12-30 21:41:02 <EvanR-work> like the weather forecast
2068 2010-12-30 21:41:07 theadmin_ has quit (Remote host closed the connection)
2069 2010-12-30 21:41:08 devon_hillard has joined
2070 2010-12-30 21:41:11 <zygf> but you can't go around telling people their GPUs will generate profit if it's based on flawed calculations with constant difficulty :P
2071 2010-12-30 21:41:14 <EvanR-work> and with constant its very good for short term
2072 2010-12-30 21:41:37 <zygf> yeah, but no GPU will cover its cost in 10 days
2073 2010-12-30 21:41:55 <lfm> zygf based on current difficulty and current btc prices yes you can, just tell them what its based on
2074 2010-12-30 21:42:00 <EvanR-work> you know whats missing, choosing a difficulty
2075 2010-12-30 21:42:00 <nanotube> zygf: if you are thinking of buying a gpu, you make your own calculations. (and don't forget resale value of the gpu)
2076 2010-12-30 21:42:06 <zygf> at constant difficulty it takes 3-4 months, in that time you have 9-12 difficulty bumps
2077 2010-12-30 21:42:46 <lfm> zygf of course it also depends on your electricity prices which I dont know
2078 2010-12-30 21:43:08 <EvanR-work> and water prices for hydro cooling!
2079 2010-12-30 21:43:16 <nanotube> and depends on your assumptions about btc prices on the morket
2080 2010-12-30 21:43:21 <nanotube> market
2081 2010-12-30 21:43:47 <zygf> now those are harder to predict than difficulty is!
2082 2010-12-30 21:43:49 <lfm> zygf and we dont know if or when the difficulty will level off
2083 2010-12-30 21:44:12 <EvanR-work> moores law
2084 2010-12-30 21:44:12 <nanotube> ;;bc,stats
2085 2010-12-30 21:44:14 <gribble> Current Blocks: 100235 | Current Difficulty: 14484.16236123 | Next Difficulty At Block: 100799 | Next Difficulty In: 564 blocks | Next Difficulty In About: 3 days, 10 hours, 52 minutes, and 36 seconds | Next Difficulty Estimate: 15997.81635055
2086 2010-12-30 21:44:27 <nanotube> lfm: hey, almost at 16k.
2087 2010-12-30 21:44:47 <lfm> its been up and down all week
2088 2010-12-30 21:45:39 <nanotube> of course.. but on average, it's been climbing steadily toward 16k through the last 1000 blocks.
2089 2010-12-30 21:50:08 <ArtForz> which is still a lot less growth than the previous difficulties
2090 2010-12-30 21:50:42 <Keefe> ArtForz: have a link to an updated chart of total power?
2091 2010-12-30 21:50:50 <ArtForz> nope
2092 2010-12-30 21:51:01 <lfm> i could update mine if you like
2093 2010-12-30 21:51:11 <Keefe> i need to get myself a better charting app than excel :)
2094 2010-12-30 21:51:19 <Keefe> Art's looked awesome
2095 2010-12-30 21:51:31 <ArtForz> if this rate of growth continues, we'll only be seeing a ~14% diffincrease
2096 2010-12-30 21:51:39 <nanotube> indeed
2097 2010-12-30 21:51:41 <Keefe> i'll just update my excel chart for now
2098 2010-12-30 21:51:45 <nanotube> seems to be doing some leveling off
2099 2010-12-30 21:52:00 <ArtForz> mine was OO calc
2100 2010-12-30 21:52:01 <nanotube> until the gamers catch on and everyone starts running a gpu miner. :)
2101 2010-12-30 21:52:20 <nanotube> and/or until art gets his batch of asic
2102 2010-12-30 21:52:40 <ArtForz> first batch is in feb, and it's only 20Gh/s
2103 2010-12-30 21:52:56 <nanotube> still, pretty decent. :)
2104 2010-12-30 21:53:13 <nanotube> and then you'll be selling off your gpus to finance further asic purchases... :)
2105 2010-12-30 21:53:32 <ArtForz> probably
2106 2010-12-30 21:53:37 <Keefe> if Art keeps to ~25% of total power, it won't necessarily bury us gpu miners
2107 2010-12-30 21:55:08 <ArtForz> well, I have to do *something* because when outside temps go >10°C or so I'll have to switch my miners to utility power @ $0.27/kWh
2108 2010-12-30 21:55:25 <Keefe> ouch
2109 2010-12-30 21:55:56 <TD> ArtForz: where do you get your power today?
2110 2010-12-30 21:56:10 <ArtForz> micro combined heat+power unit
2111 2010-12-30 21:56:35 <ArtForz> = central heating provided by diesel genset running on heating oil
2112 2010-12-30 22:00:43 OneFixt has quit (Ping timeout: 272 seconds)
2113 2010-12-30 22:02:00 Slix` has joined
2114 2010-12-30 22:02:01 <EvanR-work> haha bitcoins or coal powerwed
2115 2010-12-30 22:02:03 <EvanR-work> are*
2116 2010-12-30 22:04:17 <lfm> http://www3.telus.net/millerlf/hashes.png
2117 2010-12-30 22:04:49 <ArtForz> old
2118 2010-12-30 22:05:03 <ArtForz> data ends end of november
2119 2010-12-30 22:05:24 <lfm> doh it didnt add the latest yet ...
2120 2010-12-30 22:13:07 <lfm> http://www3.telus.net/millerlf/hashes.png ok now, does kinda show leveling off the last bit
2121 2010-12-30 22:18:04 kermit has joined
2122 2010-12-30 22:21:59 tg has quit (Read error: Connection reset by peer)
2123 2010-12-30 22:22:24 tg has joined
2124 2010-12-30 22:25:10 <Keefe> looks like until 2010, there were just a few cpus
2125 2010-12-30 22:25:24 <lfm> right
2126 2010-12-30 22:25:46 <nanotube> lfm: so what are we at now, about 110ghps?
2127 2010-12-30 22:26:25 <Keefe> sounds about right
2128 2010-12-30 22:26:27 <tcatm> 114.39
2129 2010-12-30 22:26:27 <newsham> big farm came online in aug 2010
2130 2010-12-30 22:26:50 <lfm> naw, many small ones, that was the slashdot effect
2131 2010-12-30 22:26:51 <newsham> or maybe july
2132 2010-12-30 22:27:00 <Keefe> mid july, /.
2133 2010-12-30 22:27:13 <nanotube> yea, like 10 mhps until 2010. ... /me so sad he wasn't around then.
2134 2010-12-30 22:27:34 <newsham> i wish ibought msft in 1980
2135 2010-12-30 22:27:54 <nanotube> yea
2136 2010-12-30 22:27:55 <nanotube> heh
2137 2010-12-30 22:28:41 akem has quit (Ping timeout: 255 seconds)
2138 2010-12-30 22:29:23 OneFixt has joined
2139 2010-12-30 22:29:24 <Keefe> i almost gave my family and relatives each a self-made bitcoin gift card for christmas this year, but hadn't finished the research to prove it would work to just print on a card the 256-bit secret from a keypair, as a QR code or maybe just hex
2140 2010-12-30 22:29:30 OneFixt has quit (Changing host)
2141 2010-12-30 22:29:30 OneFixt has joined
2142 2010-12-30 22:29:30 <newsham> if only i put my life savings into gold in 2004
2143 2010-12-30 22:29:43 <Keefe> was going to give each 20 btc and tell them to hold onto it for a year
2144 2010-12-30 22:30:07 xelister has joined
2145 2010-12-30 22:30:16 <lfm> coulda given them thumb drives
2146 2010-12-30 22:30:22 <nanotube> Keefe: yea, may be better to give them a usb stick with a wallet. more immediately usable.
2147 2010-12-30 22:30:31 <nanotube> lfm: hehe right
2148 2010-12-30 22:30:49 <nanotube> (and of course, keep backups, cuz you know half of them will just lose it.
2149 2010-12-30 22:30:51 <nanotube> )
2150 2010-12-30 22:30:52 <newsham> that way if btc doesnt pan out, they still have a (year old) usb stick!
2151 2010-12-30 22:30:54 <devon_hillard> ;;bc,help
2152 2010-12-30 22:30:54 <gribble> Alias bc,bcm, Alias bc,blocks, Alias bc,btcex, Alias bc,calc, Alias bc,diff, Alias bc,estimate, Alias bc,help, Alias bc,hextarget, Alias bc,markets, Alias bc,mtgox, Alias bc,nexttarget, Alias bc,poolstats, Alias bc,stats, Alias bc,timetonext, Alias bc,totalbc, and Alias bc,wiki
2153 2010-12-30 22:31:01 <TD> it'd be nice if there was an interchange format for keypairs that is not wallet.dat
2154 2010-12-30 22:31:05 <devon_hillard> ;;bc,poolstats
2155 2010-12-30 22:31:06 <gribble> {"active_workers": 133, "hashes_ps": 5398184064, "shares": 3846, "round_started": "2010-12-30 21:37:21"}
2156 2010-12-30 22:31:27 <lfm> you could put the whole bitcoin runnable and block chain on a cheap thumb drive ready to go
2157 2010-12-30 22:31:45 <marioxcc> nanotube: if he gives a usb stick with a wallet, wouldn't they have to replace they current wallet if they have one?
2158 2010-12-30 22:31:46 <devon_hillard> ;;bc,diff
2159 2010-12-30 22:31:46 <gribble> 14484.16236123
2160 2010-12-30 22:32:02 <devon_hillard> what is the current total crunching speed of the network?
2161 2010-12-30 22:32:02 <newsham> lfm: you could put tinycore linux with bitcoin and wallet, bootable.
2162 2010-12-30 22:32:10 <newsham> and have it boot right into the program
2163 2010-12-30 22:32:18 <lfm> mario just transfer the balance to a new address
2164 2010-12-30 22:32:33 <EvanR-work> devon_hillard: 110G
2165 2010-12-30 22:32:34 <Keefe> the QR code idea attracted me because i hear some smartphones have an app for reading them with the cam
2166 2010-12-30 22:32:43 <marioxcc> lfm: why not to transfer it directly, as a gift
2167 2010-12-30 22:33:01 <marioxcc> instead of giving a "gift card" give the bitcoins to the familariy address
2168 2010-12-30 22:33:12 <lfm> mario the assumption was most of them didnt know about btc already
2169 2010-12-30 22:33:16 <Keefe> was thinking either i or someone else would eventually make an addon/patch for bitcoin to import/export keys
2170 2010-12-30 22:33:35 <marioxcc> lfm: ok
2171 2010-12-30 22:33:54 <devon_hillard> so the network is producing around $50K at current exchange rates
2172 2010-12-30 22:34:00 <devon_hillard> every month
2173 2010-12-30 22:34:12 neptunepink has quit (Read error: Connection reset by peer)
2174 2010-12-30 22:34:32 <Keefe> $65K
2175 2010-12-30 22:34:41 <nanotube> devon_hillard: network always aims for 1 block per 10 min... regardless of total diff.
2176 2010-12-30 22:34:50 <nanotube> total hash power, i mean
2177 2010-12-30 22:34:50 <lfm> ;;bc,calc 144*50*30
2178 2010-12-30 22:34:53 <gribble> The average time to generate a block at 144*50*30 Khps, given current difficulty of 14484.16236123 , is 3 days, 8 hours, and 4 seconds
2179 2010-12-30 22:35:05 <lfm> ;;google,calc 144*50*30
2180 2010-12-30 22:35:05 <gribble> Error: "google,calc" is not a valid command.
2181 2010-12-30 22:35:15 <lfm> ;;calc 144*50*30
2182 2010-12-30 22:35:15 <gribble> Error: The command "calc" is available in the Google and Math plugins. Please specify the plugin whose command you wish to call by using its name as a command before "calc".
2183 2010-12-30 22:35:24 <EvanR-work> ;;bc,calc 110000000
2184 2010-12-30 22:35:24 <lfm> ;;calc,google 144*50*30
2185 2010-12-30 22:35:25 <gribble> The average time to generate a block at 110000000 Khps, given current difficulty of 14484.16236123 , is 9 minutes and 25 seconds
2186 2010-12-30 22:35:26 <gribble> Error: "calc,google" is not a valid command.
2187 2010-12-30 22:35:26 <marioxcc> lfm: just use your shell/REPL..
2188 2010-12-30 22:36:16 <lfm> ;;bah
2189 2010-12-30 22:36:17 <gribble> I do not know about 'bah', but I do know about these similar topics: 'bbe'
2190 2010-12-30 22:37:04 <marioxcc> LOL
2191 2010-12-30 22:37:05 <ArtForz> ;;math
2192 2010-12-30 22:37:05 <gribble> Error: "math" is not a valid command.
2193 2010-12-30 22:37:15 <ArtForz> I know it fucking has calc capabilities
2194 2010-12-30 22:37:31 <newsham> [12:24] <@nanotube> devon_hillard: network always aims for 1 block per 10 min... regardless of total diff.
2195 2010-12-30 22:37:39 <newsham> thats only adjusted ever week or two though, right?
2196 2010-12-30 22:37:46 <marioxcc> every N blocks
2197 2010-12-30 22:37:48 <newsham> so cant it be considerably off if there's a massive change in hps?
2198 2010-12-30 22:37:54 <ArtForz> every 2016 blocks (two weeks nominal)
2199 2010-12-30 22:38:08 <lfm> ;;calc,math 144*50*30
2200 2010-12-30 22:38:08 <gribble> Error: "calc,math" is not a valid command.
2201 2010-12-30 22:38:15 <devon_hillard> nanotube: perhaps, but I was talking about exchange rates
2202 2010-12-30 22:38:15 <marioxcc> what if there is a massive injection of computing power just after the adjustment?
2203 2010-12-30 22:38:31 <ArtForz> then the next adjustment period will still be 2016 blocks
2204 2010-12-30 22:38:39 <devon_hillard> you would be like a central bank suddenly printing money
2205 2010-12-30 22:38:45 <ArtForz> which will be a lot < 2 weeks
2206 2010-12-30 22:38:48 <xelister> ;;calc 2+@
2207 2010-12-30 22:38:48 <gribble> Error: The command "calc" is available in the Google and Math plugins. Please specify the plugin whose command you wish to call by using its name as a command before "calc".
2208 2010-12-30 22:38:49 <xelister> ;;calc 2+2
2209 2010-12-30 22:38:50 <gribble> Error: The command "calc" is available in the Google and Math plugins. Please specify the plugin whose command you wish to call by using its name as a command before "calc".
2210 2010-12-30 22:38:51 <devon_hillard> the market doesn't know it immediately
2211 2010-12-30 22:38:55 <xelister> ;;math calc 2+2
2212 2010-12-30 22:38:57 <gribble> 4
2213 2010-12-30 22:38:59 <newsham> devon: the point is that it will be about 6 * 50 * (exchange rate) per hour
2214 2010-12-30 22:38:59 <ArtForz> ahhh!
2215 2010-12-30 22:39:01 <xelister> I rule
2216 2010-12-30 22:39:08 <xelister> ArtForz: that will be 2 blocks please
2217 2010-12-30 22:39:12 <lfm> ;;google calc 144*50*30
2218 2010-12-30 22:39:13 <gribble> 144 * 50 * 30 = 216,000
2219 2010-12-30 22:39:18 <lfm> ah!
2220 2010-12-30 22:39:22 <newsham> so like $75/hr at 25c a pop
2221 2010-12-30 22:39:29 <xelister> 18QmFkHw2Ms6XgfB2dxZyNLt2UMAKWRTZV
2222 2010-12-30 22:39:31 <xelister> >_>
2223 2010-12-30 22:39:34 <lfm> ;;google calc 144*50*30*0.30
2224 2010-12-30 22:39:35 <gribble> 144 * 50 * 30 * 0.30 = 64,800
2225 2010-12-30 22:40:16 <lfm> so I think its like $64k/month
2226 2010-12-30 22:40:49 <ArtForz> yep, looks about right
2227 2010-12-30 22:41:02 <newsham> inflation!
2228 2010-12-30 22:41:32 osmosis has joined
2229 2010-12-30 22:41:33 <EvanR-work> thats the $64k/month question
2230 2010-12-30 22:41:55 <newsham> thats a lot of coal
2231 2010-12-30 22:42:00 purpleposeidon has joined
2232 2010-12-30 22:42:24 <EvanR-work> were on a bitcoin train to somewhere
2233 2010-12-30 22:42:32 <EvanR-work> and ArtForz is shovelling
2234 2010-12-30 22:43:08 <newsham> has anyone started a bitcoin lotto yet?
2235 2010-12-30 22:43:27 <ArtForz> I'm down to only 14% of total network
2236 2010-12-30 22:43:31 <lfm> newsham the whole bitcoin mining system is a lottery
2237 2010-12-30 22:43:36 <devon_hillard> bitcoin bingo
2238 2010-12-30 22:43:45 <newsham> lfm: that one is no fun.. the payout is too high
2239 2010-12-30 22:43:45 <EvanR-work> lol
2240 2010-12-30 22:43:53 <newsham> and its spread too evenly
2241 2010-12-30 22:43:55 <EvanR-work> bingo strategy, pay attention
2242 2010-12-30 22:44:06 <newsham> need something where one person can make lots of bitcoins and the person running it can make even more
2243 2010-12-30 22:44:31 <devon_hillard> BTC casino pool
2244 2010-12-30 22:44:38 <devon_hillard> only one miner gets all the coins
2245 2010-12-30 22:44:55 <newsham> nah, thats cpu for coin.
2246 2010-12-30 22:44:56 <devon_hillard> based on his share weight
2247 2010-12-30 22:45:01 <newsham> i think coin for coin is better
2248 2010-12-30 22:45:14 <newsham> i dont want your useless cpu grinding
2249 2010-12-30 22:45:32 <marioxcc> newsham: those useless thigs make 50 BTC
2250 2010-12-30 22:45:32 <marioxcc> :)
2251 2010-12-30 22:45:39 <lfm> devon_hilliard that would be exactly like running the standard client for yourself then.
2252 2010-12-30 22:46:00 <devon_hillard> lfm: yes, but with an economy of electricity
2253 2010-12-30 22:46:08 <lfm> without the overhead
2254 2010-12-30 22:46:09 <marioxcc> devon_hillard: ???
2255 2010-12-30 22:46:27 <devon_hillard> a pool will economize electricity
2256 2010-12-30 22:46:27 <marioxcc> how would you pool casino compare with individual minign?
2257 2010-12-30 22:46:38 <newsham> i know.. a BTC change service. give me 50BTC and I'll give you 10 5BTC transactions, or 50 1BTC transactions
2258 2010-12-30 22:46:41 <marioxcc> a pool don't increases performance, nor efficiency
2259 2010-12-30 22:46:42 <newsham> or 2 25BTC's
2260 2010-12-30 22:46:44 <lfm> devon how do you figure that?
2261 2010-12-30 22:46:45 <marioxcc> it just reduces granularity
2262 2010-12-30 22:46:47 <devon_hillard> since there is no search duplication
2263 2010-12-30 22:46:57 <marioxcc> devon_hillard: there is NO duplication
2264 2010-12-30 22:47:05 <marioxcc> all miners work on different bloks by default
2265 2010-12-30 22:47:16 <lfm> devon there is never search duplication unless its a bug
2266 2010-12-30 22:47:17 <marioxcc> same nonce range but i think merkle root varies
2267 2010-12-30 22:47:20 <newsham> or how about a BTC sorting service?
2268 2010-12-30 22:47:29 <newsham> give us your BTC and we'll sort and band them into one large BTC
2269 2010-12-30 22:47:34 <marioxcc> newsham: sorting?
2270 2010-12-30 22:47:38 <marioxcc> oh
2271 2010-12-30 22:47:56 <lfm> newsham haha
2272 2010-12-30 22:48:12 <newsham> lfm: https://www.cufatcats.org/Images/CoinMachine.jpg
2273 2010-12-30 22:48:16 <newsham> err i mean mario
2274 2010-12-30 22:48:18 <afed> when i connect a m0mchil cl miner to bitcoin.cz i get block found output, but when i connect it to the bitcoin client on my own machine i get nothing except X khashes per second, is that correct?
2275 2010-12-30 22:48:29 <devon_hillard> so there is a space of unclaimed blocks, everybody picks a block and attempts to break it
2276 2010-12-30 22:48:40 <Diablo-D3> afed: yes
2277 2010-12-30 22:48:42 <afed> i wish i'd see an aggregate hashes per second on the client or some other indication it had miners connected and working
2278 2010-12-30 22:48:43 <lfm> afed yes
2279 2010-12-30 22:48:44 <devon_hillard> well, with a pool there is less duplication in searching for the same block
2280 2010-12-30 22:48:46 <Diablo-D3> afed: pool shares are very low difficulty
2281 2010-12-30 22:48:54 <marioxcc> devon_hillard: there is no duplication
2282 2010-12-30 22:49:04 <marioxcc> a pool actually reduces eficiency a bit
2283 2010-12-30 22:49:11 <marioxcc> but increases granularity
2284 2010-12-30 22:49:13 <Diablo-D3> the pool does NOT reduce efficiency
2285 2010-12-30 22:49:17 <Diablo-D3> not in any useful terms
2286 2010-12-30 22:49:20 <afed> Diablo-D3: oh i get it, higher difficulty, not finding things nearly as often
2287 2010-12-30 22:49:34 <marioxcc> Diablo-D3: i mean a bit, think bandwidth and so
2288 2010-12-30 22:49:36 <afed> Diablo-D3: so when a miner finds a block for my client it'll appear as normal?
2289 2010-12-30 22:49:41 <lfm> devon no there is no duplication no one is searching for the "same" block. they all search for different blocks
2290 2010-12-30 22:49:50 <joe_1> we already discussed that pools are broken because people won't turn in the winning hashes
2291 2010-12-30 22:49:58 <marioxcc> someone should put that on the FAQ
2292 2010-12-30 22:50:01 <devon_hillard> marioxcc: so the network guarantees that only one client will search for one block at any one time?
2293 2010-12-30 22:50:04 <Diablo-D3> joe_1: but they cant USE the winning hashes
2294 2010-12-30 22:50:10 <joe_1> they still won't turn them in
2295 2010-12-30 22:50:11 <Diablo-D3> so there is no attack
2296 2010-12-30 22:50:15 <marioxcc> devon_hillard: well, the network don't ensures that
2297 2010-12-30 22:50:16 <Diablo-D3> why wont they turn them in?
2298 2010-12-30 22:50:19 <marioxcc> is just probability
2299 2010-12-30 22:50:19 <Diablo-D3> their effort is wasted then
2300 2010-12-30 22:50:21 <joe_1> for fun
2301 2010-12-30 22:50:23 <ArtForz> it's a completely pointless attack
2302 2010-12-30 22:50:44 <marioxcc> but the probability of two clients serching the same space is almost 0
2303 2010-12-30 22:50:48 <marioxcc> for all practical purposes, it's ok
2304 2010-12-30 22:50:49 <afed> if they can't use the winning hash to claim the coins for themselves, they don't benefit, they only hurt the pool
2305 2010-12-30 22:50:53 <afed> they'll get fewer bitcoins that way
2306 2010-12-30 22:50:53 <devon_hillard> marioxcc: so then, with a large enough pool you would get less duplicate searches
2307 2010-12-30 22:51:06 <lfm> devon pretty much, there is lots of room in 256 bit hashes for different blocks both good and bad
2308 2010-12-30 22:51:11 <Diablo-D3> yes, if they DONT turn the winning block in, THEY lose coins
2309 2010-12-30 22:51:12 <Diablo-D3> period
2310 2010-12-30 22:51:12 <marioxcc> devon_hillard: no, there are no duplicate searches
2311 2010-12-30 22:51:12 <devon_hillard> less searches in the same space
2312 2010-12-30 22:51:16 <Diablo-D3> there is no way to win in this attack
2313 2010-12-30 22:51:23 <ArtForz> Diablo-D3: and so does everyone else, thats the point
2314 2010-12-30 22:51:30 <Diablo-D3> ArtForz: so why mine at all?
2315 2010-12-30 22:51:33 <ArtForz> exactly
2316 2010-12-30 22:51:43 <ArtForz> like I said, pointless attack
2317 2010-12-30 22:52:24 <lfm> devon_hilliard IT IS NOT THE SAME SPACE. the block includes hashes of the txn and the txn are all different to different addreses and with different timestamps
2318 2010-12-30 22:52:32 <ArtForz> yep
2319 2010-12-30 22:52:37 <marioxcc> devon_hillard: what I mean is that there may be duplicate seraches, but there is plenty of space for variants, which are random so probability is so low than you don't really need to account for that
2320 2010-12-30 22:52:42 <ArtForz> everyone works on their own merle root
2321 2010-12-30 22:52:45 <ArtForz> *merkle root
2322 2010-12-30 22:52:51 <marioxcc> 1 in 2^256, i don't see the point
2323 2010-12-30 22:52:53 <Diablo-D3> marioxcc: its impossible to get the same space twice randomly
2324 2010-12-30 22:52:59 <marioxcc> Diablo-D3: ¿why?
2325 2010-12-30 22:53:01 <newsham> is there a page listing the mining pools?
2326 2010-12-30 22:53:05 <Diablo-D3> its easier to win powerball
2327 2010-12-30 22:53:12 <marioxcc> newsham: https://en.bitcoin.it/wiki/Pooled_Mining
2328 2010-12-30 22:53:14 <ArtForz> not impossible, but so damn unlikely it'll probably never ahppen
2329 2010-12-30 22:53:21 <Diablo-D3> milisecond time, nonce, and merkle root has to match
2330 2010-12-30 22:53:24 <Diablo-D3> all has to match
2331 2010-12-30 22:53:26 <ArtForz> 1 in 2**128 is a REALLY small chance
2332 2010-12-30 22:53:37 <lfm> like i said, there is lots of room in 256 bits
2333 2010-12-30 22:53:42 <ArtForz> yep
2334 2010-12-30 22:53:57 <marioxcc> Diablo-D3: it's not impossible if i understand well, but probability is too low to even consider it
2335 2010-12-30 22:54:04 <ArtForz> I think effectively sqrt that, birthday paradox
2336 2010-12-30 22:54:11 <marioxcc> ArtForz: oh, yeah
2337 2010-12-30 22:54:26 <marioxcc> still very low
2338 2010-12-30 22:54:29 <ArtForz> yep
2339 2010-12-30 22:54:43 <ArtForz> and we have nonce, too ...
2340 2010-12-30 22:54:57 <ArtForz> so 1 in 2**144
2341 2010-12-30 22:54:58 <nanotube> ArtForz: lfm: "math calc" or "google calc"
2342 2010-12-30 22:55:25 <lfm> it is very low chance of lightning hitting you but it is incredibly more likely you get hit by lightning every day for a year that we get duplicate blocks by chance
2343 2010-12-30 22:55:39 <ArtForz> yep
2344 2010-12-30 22:56:00 <ArtForz> chance per block is something like 5e-44
2345 2010-12-30 22:56:31 <lfm> not to mention we arnt reuing the dates yet
2346 2010-12-30 22:56:46 <ArtForz> well, I assume everyone is working on the same nTime
2347 2010-12-30 22:57:10 <ArtForz> which isnt true, but people should be generally within 10 seconds of each other, so add another order of magnitude
2348 2010-12-30 22:57:26 <lfm> ya but we only need to consider two blocks the same if they are created the same second
2349 2010-12-30 22:58:03 <marioxcc> what happens if SHA256 is brooken or a method is found to resolve PoW orderes of magnitude faster than brute force?
2350 2010-12-30 22:58:16 <ArtForz> anyways miners randomly working on the same block should happen about once every 3e39 years
2351 2010-12-30 22:58:22 <lfm> well then we start using sha512
2352 2010-12-30 22:58:34 <TD> anyone know who this was
2353 2010-12-30 22:58:34 <ArtForz> well, if sha256 is borken, so probably is sha512
2354 2010-12-30 22:58:35 <TD> http://blockexplorer.com/tx/73037ad07148f367b783338ee1eb056fabefdc0e3c9c9422354ac9410013ae97
2355 2010-12-30 22:58:41 <TD> some big miner consolidating his generations ?
2356 2010-12-30 22:59:10 <marioxcc> lfm: SHA 2 hash familiy share the same structure
2357 2010-12-30 22:59:26 <marioxcc> just different word widths, initialization vectors and constants
2358 2010-12-30 22:59:27 <lfm> ok so sha3x1024 or whatever. just as quick almost as we can spread a new version of the software
2359 2010-12-30 22:59:50 <marioxcc> lfm: but is that plausible?
2360 2010-12-30 22:59:55 <ArtForz> and I doubt any normal attack is applicable to bitcoin
2361 2010-12-30 23:00:04 <marioxcc> we know not everyone is going to install the new version in the same day it's released
2362 2010-12-30 23:00:11 <ArtForz> we use double-sha256 everywhere
2363 2010-12-30 23:00:20 <lfm> I dont think it will happen but you never know what thos eggheads might do
2364 2010-12-30 23:00:27 <marioxcc> ArtForz: anyway, in the hypotetical case
2365 2010-12-30 23:00:43 <ArtForz> in that hypothetical case, bitcoin is the smallest problem
2366 2010-12-30 23:00:50 <marioxcc> what do you mean?
2367 2010-12-30 23:01:00 <devon_hillard> ok, so if a client is given a block to break and the client is rebooted, all the work would have been wasted?
2368 2010-12-30 23:01:14 <marioxcc> devon_hillard: uh?
2369 2010-12-30 23:01:17 <marioxcc> there is no work wasted
2370 2010-12-30 23:01:21 <devon_hillard> i.e. if you are running your miner for 12 hours a day instead of continuously
2371 2010-12-30 23:01:27 <lfm> marioxcc other currency exchange systems use some of the same encryption algorthms
2372 2010-12-30 23:01:32 <marioxcc> devon_hillard: yo don't make progress towards solving the block
2373 2010-12-30 23:01:40 <marioxcc> the chance to solve it for every try is the same
2374 2010-12-30 23:01:42 <marioxcc> read the wiki
2375 2010-12-30 23:01:50 <lfm> marioxcc like banks you know
2376 2010-12-30 23:02:23 <marioxcc> lfm: ok
2377 2010-12-30 23:02:30 purpleposeidon has quit (Quit: you know stuff)
2378 2010-12-30 23:04:10 <marioxcc> so do you think double hasning was choosen because it is more difficult to break (maybe like double the rounds)?
2379 2010-12-30 23:04:54 <joe_1> satoshi would know.
2380 2010-12-30 23:05:43 <marioxcc> i don't ask if you know, becuase i guess only satoshi knows the real reason, but whether you think so
2381 2010-12-30 23:07:11 <joe_1> in what way do we use double hashing?
2382 2010-12-30 23:07:33 <TD> when bitcoin hashes something, it does sha256(sha256(data))
2383 2010-12-30 23:07:45 <joe_1> oh, that's horrible
2384 2010-12-30 23:07:50 <lfm> sha256(sha256(block header))
2385 2010-12-30 23:08:02 <joe_1> it's probably weaker than one hash
2386 2010-12-30 23:08:10 <ArtForz> proof?
2387 2010-12-30 23:08:11 <lfm> not
2388 2010-12-30 23:08:44 <joe_1> proof is simple: if the square of the sha256 function were stronger, then they would define sha256 as such, and it would be common knowledge.
2389 2010-12-30 23:09:01 <ArtForz> yes, because speed/security dradeoffs dont exist
2390 2010-12-30 23:09:08 <ArtForz> *tradeoffs
2391 2010-12-30 23:09:08 <joe_1> exactly
2392 2010-12-30 23:09:29 <xelister> joe_1: what you say is idiotic
2393 2010-12-30 23:09:32 <ArtForz> so you just proved yourself wrong
2394 2010-12-30 23:09:43 <lfm> and des3 would be weaker than des
2395 2010-12-30 23:09:46 purpleposeidon has joined
2396 2010-12-30 23:10:00 <ArtForz> any compentent hash function doesnt throw entropy away, adding more rounds only makes it stronger
2397 2010-12-30 23:10:18 <xelister> joe_1: http://en.wikipedia.org/wiki/PBKDF2
2398 2010-12-30 23:10:24 <newsham> joe: not necessarily.
2399 2010-12-30 23:10:55 <xelister> joe_1: its well known that PBKDF2 is the stronger (or at least the harder to brute force) the longer it is, and yet noone uses 'infinity' long PBKDF, as a speed tradeoff
2400 2010-12-30 23:11:05 <xelister> although, actually, my example with PBKDF2 is a bit other
2401 2010-12-30 23:13:11 <newsham> how many people are heating their houses with bitcoin generators?
2402 2010-12-30 23:13:40 <joe_1> maybe your right
2403 2010-12-30 23:14:42 <newsham> how much easier is it to find a slush block vs. a real block?
2404 2010-12-30 23:16:00 <joe_1> a lot easier
2405 2010-12-30 23:16:15 <ArtForz> ;;bc,diff
2406 2010-12-30 23:16:15 <gribble> 14484.16236123
2407 2010-12-30 23:16:23 <slush> newsham: 14000x easier
2408 2010-12-30 23:16:26 <ArtForz> about 14484 times easier
2409 2010-12-30 23:16:47 <slush> lol, 'slush block'
2410 2010-12-30 23:17:33 <joe_1> will using the sha256 function twice also get rid of the threat of the u.s. government's "backdoor" key into the sha256 hash?
2411 2010-12-30 23:17:44 <TD> there is no "backdoor" in sha
2412 2010-12-30 23:19:03 <newsham> so a reasonably fast cpu is gonna find one every 30min or so?
2413 2010-12-30 23:19:06 <newsham> does that sound about right?
2414 2010-12-30 23:19:46 <marioxcc> newsham: it depeneds what "reasonably fast" is
2415 2010-12-30 23:19:59 <marioxcc> but you're going to find several per day, unless your CPU is really crappy
2416 2010-12-30 23:20:02 <newsham> somewhere in the 3k khps range
2417 2010-12-30 23:20:24 <joe_1> ;;bc,estimate 3k
2418 2010-12-30 23:20:25 <gribble> 15984.36049930
2419 2010-12-30 23:20:49 <joe_1> it'll take 15984 seconds to find a real block, so about 1 second to find each of the slush blocks
2420 2010-12-30 23:20:57 <newsham> so 4.5hrs
2421 2010-12-30 23:21:04 <newsham> err wait..
2422 2010-12-30 23:21:19 <newsham> i thought its 160 days or so to find a real block
2423 2010-12-30 23:21:54 <newsham> oops, i meant 3 khps i guess :)
2424 2010-12-30 23:21:54 <devon_hillard> bitcoin.com should just link directly to one of the GPU miners
2425 2010-12-30 23:22:01 <xelister> what is slush block?
2426 2010-12-30 23:22:05 <joe_1> ;;bc,estimate 3000000
2427 2010-12-30 23:22:06 <gribble> 15988.82342648
2428 2010-12-30 23:22:11 <joe_1> stupid bot
2429 2010-12-30 23:22:11 <slush> >>> 2**32 / 3000000
2430 2010-12-30 23:22:11 <slush> 1431L
2431 2010-12-30 23:22:15 <slush> ~ half hour
2432 2010-12-30 23:22:15 <xelister> yey, btc @ 3.00 again
2433 2010-12-30 23:22:20 <devon_hillard> I suppose lots of users come in, see 3K khashes/s, then realize it takes more than a year of work, then give up
2434 2010-12-30 23:22:39 <xelister> devon_hillard: bitcoin is not just about mining you know.
2435 2010-12-30 23:23:08 <marioxcc> please don't say 3K khashes/s
2436 2010-12-30 23:23:15 <marioxcc> that's a barbarism
2437 2010-12-30 23:23:21 <marioxcc> say 3 Mhash/s
2438 2010-12-30 23:23:34 <newsham> thats what you get for putting the "k" next to "hash"!
2439 2010-12-30 23:23:57 <marioxcc> newsham: i didn't invented the name
2440 2010-12-30 23:24:19 <marioxcc> i would use maybe the existing SI unit Bq
2441 2010-12-30 23:24:24 <marioxcc> 3 MBq
2442 2010-12-30 23:24:37 <marioxcc> or 3 MHz in its deffect
2443 2010-12-30 23:26:27 <xelister> BTH bitcoin hash?
2444 2010-12-30 23:26:42 <marioxcc> xelister: what?
2445 2010-12-30 23:26:50 <xelister> thinming out loud
2446 2010-12-30 23:30:28 kiba has joined
2447 2010-12-30 23:35:34 noagendamarket has quit (Ping timeout: 240 seconds)
2448 2010-12-30 23:42:27 <kiba> bencoder: yo yo yo yo?
2449 2010-12-30 23:42:37 <kiba> I did not see any feature emerging from your download service
2450 2010-12-30 23:45:04 kiba has quit (Remote host closed the connection)
2451 2010-12-30 23:48:08 <afed> lol bequerels?
2452 2010-12-30 23:49:23 <marioxcc> afed: yeah
2453 2010-12-30 23:49:25 <ArtForz> btw, same problem with MIPS
2454 2010-12-30 23:49:46 <ArtForz> M Instructions Per Second
2455 2010-12-30 23:50:01 <marioxcc> i don't see a problem with MIPS or khash/s
2456 2010-12-30 23:50:20 <marioxcc> but I do with X K MIPS or Y M khash/s
2457 2010-12-30 23:50:25 <ArtForz> exactly
2458 2010-12-30 23:50:35 <ArtForz> nobody uses GIPS or TIPS though
2459 2010-12-30 23:50:43 <marioxcc> yes :(
2460 2010-12-30 23:50:55 <afed> i think we should start using it
2461 2010-12-30 23:51:00 <afed> gflops and tflops are already acceptable
2462 2010-12-30 23:51:09 <ArtForz> yep
2463 2010-12-30 23:51:17 <afed> maybe intstructions per second isn't meaningful
2464 2010-12-30 23:51:19 <afed> integer ops per second
2465 2010-12-30 23:51:48 <afed> or do we just consider those the same?
2466 2010-12-30 23:52:08 <ArtForz> usually, yes
2467 2010-12-30 23:52:22 <marioxcc> well, with today pipelined processors
2468 2010-12-30 23:52:48 <marioxcc> there is hardly a single number to give a objetive performance rating
2469 2010-12-30 23:53:00 <ArtForz> same thing you do for peak flops, just ass-u-me the thing can keep all units busy
2470 2010-12-30 23:53:14 Kiba has joined
2471 2010-12-30 23:53:55 <afed> http://www.dell.com/us/business/p/poweredge-c410x/pd?refid=poweredge-c410x&baynote_bnrank=0&baynote_irrank=3&~ck=dellSearch
2472 2010-12-30 23:54:26 <ArtForz> and for peak flops multiply by 2 if you have muladd
2473 2010-12-30 23:55:01 <ArtForz> which is kinda BSy
2474 2010-12-30 23:55:44 <lfm> why can more systems do the zero overhead loops like some signal processors
2475 2010-12-30 23:55:51 <lfm> cant
2476 2010-12-30 23:56:10 <ArtForz> zero overhead loops ?
2477 2010-12-30 23:56:36 <lfm> they can do a for loop as fast as if it was unrolled
2478 2010-12-30 23:56:56 <lfm> stuff with a predetermined number of passes
2479 2010-12-30 23:56:56 <tcatm> one of my miners behaves strangely. it disappears (on the network) for a few minutes and then comes back without errors...
2480 2010-12-30 23:57:19 <ArtForz> you can do pretty much the same on a CPU with branch predicates
2481 2010-12-30 23:57:45 <ArtForz> not 0-cycle, but 1-cycle
2482 2010-12-30 23:58:33 <ArtForz> (or just knowing how your branch predictor works)
2483 2010-12-30 23:58:45 gavinandresen has quit (Quit: gavinandresen)
2484 2010-12-30 23:59:13 <ArtForz> lemme guess 16 Tesla 2050 or 2070s
2485 2010-12-30 23:59:20 <lfm> loop instruction specifies the whole loop range of instructions and the loop count so no need for branching at all
2486 2010-12-30 23:59:53 <afed> dell says you can install anything you want
2487 2010-12-30 23:59:58 <afed> ati cards