1 2011-01-16 00:00:47 <cdecker> Hi guys
2 2011-01-16 00:01:18 <slush> hi
3 2011-01-16 00:01:41 <devon_hillard> What does the -f switch do on Diablo's miner?
4 2011-01-16 00:01:58 <ArtForz> adjust target framerate
5 2011-01-16 00:02:21 <ArtForz> aka inverse of enqueuequernel latency
6 2011-01-16 00:02:47 <ArtForz> low f = lots of workitems per enqueue = laggy like hell but mines faster
7 2011-01-16 00:03:45 <slush> devon_hillard: If you are talking about this; tens or hundreds of pool CPUs are cutting GPUs rewards consistently for every round, but those poor miners have same probability to find a block. So pool have already many blocks from miners which have average in many many months. And those 'unprobably' rewards are rising GPUs rewards back. So everybody is happy
8 2011-01-16 00:08:25 <devon_hillard> ArtForz: what should lag?
9 2011-01-16 00:08:34 <devon_hillard> or is it not a windows issue?
10 2011-01-16 00:08:38 <ArtForz> pretty much everything
11 2011-01-16 00:08:49 <ArtForz> at least my miner does, doesnt matter what OS
12 2011-01-16 00:08:51 <devon_hillard> -f 100 works fine, no difference from before
13 2011-01-16 00:09:01 <[Noodles]> what would be kewl: an option to get a mail when a given worker doesnt submit a hash within a given timeframe (an hour/x hours/a day), default set OFF, then you can set it ON on some workers you won't/can't keep an eye on, fire and forget
14 2011-01-16 00:09:21 <ArtForz> push the # of workitems per call high enough and I can watch windows redrawing in parts
15 2011-01-16 00:09:36 nofuture has quit (Ping timeout: 265 seconds)
16 2011-01-16 00:09:50 <ArtForz> but then I consider "low" -f 2 or 3 ...
17 2011-01-16 00:09:52 <devon_hillard> I disabled flash GPU acceleration and all Windows graphical effects
18 2011-01-16 00:10:04 <slush> [Noodles]: You mean in pool?
19 2011-01-16 00:10:04 <devon_hillard> the UI uses no GPU acceleration that I know
20 2011-01-16 00:10:13 <[Noodles]> slush: yes
21 2011-01-16 00:10:20 <slush> [Noodles]: It's almost done
22 2011-01-16 00:10:27 <ArtForz> yeah, I dont quite get it either, but thats what happens for me
23 2011-01-16 00:10:32 <slush> [Noodles]: 'notify on crash' on profile page
24 2011-01-16 00:10:37 <devon_hillard> so it may cause issues when other apps try to use the GPU acceleration in a background fashion
25 2011-01-16 00:10:46 <ArtForz> might also depend on driver and sdk version and the phase of the moon
26 2011-01-16 00:10:46 <slush> [Noodles]: But still not active, I have to test it more
27 2011-01-16 00:11:00 <[Noodles]> slush : how is "crash" defined?
28 2011-01-16 00:11:12 <slush> [Noodles]: given worker doesnt submit a hash within a given timeframe
29 2011-01-16 00:11:15 <slush> ;)
30 2011-01-16 00:11:23 <[Noodles]> and that's for accounts, means all workers
31 2011-01-16 00:11:30 <[Noodles]> isn't it?
32 2011-01-16 00:11:39 <slush> [Noodles]: Every worker is checked
33 2011-01-16 00:11:51 <slush> [Noodles]: When he was active and now it isn't, mail is sent.
34 2011-01-16 00:12:07 <slush> [Noodles]: Then notification is reset when next share is submitted by worker
35 2011-01-16 00:12:14 <[Noodles]> yeah, but i don't wanna be mailed for any worker, just for those i choose to
36 2011-01-16 00:12:21 <slush> [Noodles]: why?
37 2011-01-16 00:12:40 <slush> [Noodles]: if you have inactive workers, you won't receive emails about them
38 2011-01-16 00:12:54 <[Noodles]> because i often stop one myself (the one on 5850) when playing games
39 2011-01-16 00:13:01 <Raulo> Yesterday, somebody asked about evolution of mining income
40 2011-01-16 00:13:10 <Raulo> I made a graph: http://bitcoin.atspace.com/income.png
41 2011-01-16 00:13:22 * slush sigh
42 2011-01-16 00:13:25 <ArtForz> neat
43 2011-01-16 00:13:33 <ArtForz> what price data is that based on?
44 2011-01-16 00:13:34 <slush> [Noodles]: ok, I'll add per-worker settings
45 2011-01-16 00:13:35 <[Noodles]> so how is "inactivity" defined? ;o)
46 2011-01-16 00:13:42 <Raulo> From mtgox
47 2011-01-16 00:14:01 <Raulo> Long-term income falls
48 2011-01-16 00:14:26 <ArtForz> it actually more looks like it plateaued
49 2011-01-16 00:14:36 <slush> [Noodles]: Currently one hour of inactivity
50 2011-01-16 00:14:49 <[Noodles]> which one?
51 2011-01-16 00:14:50 <Raulo> Yes, but just last month only
52 2011-01-16 00:15:11 <[Noodles]> none of them
53 2011-01-16 00:15:12 <slush> [Noodles]: But I'll add dropdown menu for every worker, so you will be able to select 'never notify', '5 min', and so
54 2011-01-16 00:15:20 <ArtForz> well.. first GPU miners started popping up in late aug/early sept
55 2011-01-16 00:15:26 <[Noodles]> ah, definition, i see
56 2011-01-16 00:15:28 <Raulo> There were a few such periods of plateue
57 2011-01-16 00:15:47 <Raulo> But difficulty long-term rises faster than the bitcoin price
58 2011-01-16 00:15:58 <ArtForz> what long term?
59 2011-01-16 00:16:23 <[Noodles]> slush: yeah, that'd be great, no need to hurry though, take your time, i can handle it a bit longer
60 2011-01-16 00:16:23 kermit has quit (Quit: Leaving.)
61 2011-01-16 00:16:29 <ArtForz> we're still above the btc/kWh you could get with a CPU before GPU mining started
62 2011-01-16 00:16:29 <Raulo> Since the mtgox data is available
63 2011-01-16 00:16:36 kermit has joined
64 2011-01-16 00:16:54 <slush> [Noodles]: currently it has high priority for me, because it is one of the oldest feature request
65 2011-01-16 00:17:13 <Raulo> The graph does not subtract electricity cost
66 2011-01-16 00:17:16 <slush> [Noodles]: then only variable difficulty
67 2011-01-16 00:17:31 <ArtForz> yep
68 2011-01-16 00:17:32 <Raulo> For CPU-miners, the income minus electricity is currently zero
69 2011-01-16 00:17:42 <Raulo> Or less
70 2011-01-16 00:17:43 <ArtForz> sounds aobut right
71 2011-01-16 00:17:44 <slush> Raulo: negative
72 2011-01-16 00:18:39 <ArtForz> so $/hash dropped by a factor of... about 15
73 2011-01-16 00:19:04 <ArtForz> while W/hash dropped by a factor of... about 30
74 2011-01-16 00:19:33 <Raulo> Yep. More or less
75 2011-01-16 00:19:42 <Raulo> But why 30?
76 2011-01-16 00:19:57 <ArtForz> rought guesstimate
77 2011-01-16 00:20:34 <ArtForz> decent CPU does ~6Mh/s using ~100W, 5970 does ~600Mh/s using ~300W
78 2011-01-16 00:20:42 <Raulo> If software did not improve, going from CPU to GPU is a factor of 10
79 2011-01-16 00:21:07 <ArtForz> we also didnt have -4way for CPUs back then
80 2011-01-16 00:22:30 <Raulo> Indeed. 4-way is crucial on newer CPUs
81 2011-01-16 00:23:54 <Raulo> Newer AMD processors have about 3Mh/s/core on 4way with 4way
82 2011-01-16 00:23:58 <ArtForz> yep
83 2011-01-16 00:24:13 <ArtForz> well, not quite it seems
84 2011-01-16 00:24:17 <ArtForz> my Phenom II X6 gets ~15Mh/s @ 4GHz
85 2011-01-16 00:24:48 <Raulo> I ran miner on Opteron Shanghai 6-core 2.6GHz
86 2011-01-16 00:25:23 <Raulo> and got 17 Mh/s
87 2011-01-16 00:27:04 <ArtForz> weird, what miner?
88 2011-01-16 00:27:10 <Raulo> Sorry. Istanbul, not Shanghai. Shanghai is 6-core
89 2011-01-16 00:27:14 <Raulo> jgarzik's
90 2011-01-16 00:27:22 <Raulo> Compiled with Intel Compiler
91 2011-01-16 00:27:28 <ArtForz> I used stock client -4way
92 2011-01-16 00:27:38 <Raulo> (Shanghai is 4-core)
93 2011-01-16 00:27:44 <Raulo> Stock client is slower
94 2011-01-16 00:27:51 <ArtForz> yep
95 2011-01-16 00:28:02 <ArtForz> iirc the cores are pretty much identical
96 2011-01-16 00:28:19 <ArtForz> K10 is K10 for mining, pretty much
97 2011-01-16 00:28:46 <Raulo> Yep. Istanbul is Shanghai with more cores
98 2011-01-16 00:28:50 <ArtForz> aka "who cares about L3 and memory bandwith if everything fits in L1"
99 2011-01-16 00:29:24 <ArtForz> mining seems to scale pretty much 100% with num_cores*clock for a given arch
100 2011-01-16 00:29:29 <Diablo-D3> heh
101 2011-01-16 00:29:44 <ArtForz> perfect [ALU|SIMD]-bound benchmark
102 2011-01-16 00:30:22 <ArtForz> though it kinda makes i7 look worse than it is
103 2011-01-16 00:30:44 <Raulo> 17 Mh/s with 115W is 0.15 Mh/W
104 2011-01-16 00:30:53 <Raulo> Which is better than some of nVidias
105 2011-01-16 00:31:12 <ArtForz> some of the really old ones
106 2011-01-16 00:31:14 <devon_hillard> Raulo: a rough calculations meant my CPU only got back 1/20 of the electricity cost
107 2011-01-16 00:31:37 <ArtForz> recent nvidias get about 0.48Mh/W
108 2011-01-16 00:31:59 <ArtForz> when I have my CAL code finished, HD5970 should get >2Mh/W
109 2011-01-16 00:31:59 <devon_hillard> ArtForz: high-end nvidias?
110 2011-01-16 00:32:04 <ArtForz> yep
111 2011-01-16 00:32:08 <ArtForz> GTX570 and 580
112 2011-01-16 00:32:09 <devon_hillard> ah, yes
113 2011-01-16 00:32:10 <Diablo-D3> what high end nvidias?
114 2011-01-16 00:32:11 <Diablo-D3> there arent any
115 2011-01-16 00:32:14 <Diablo-D3> dohohohoho
116 2011-01-16 00:32:25 <Raulo> CAL code?
117 2011-01-16 00:32:29 <ArtForz> yeah
118 2011-01-16 00:32:39 <ArtForz> renamed CTM
119 2011-01-16 00:32:40 <Raulo> What is CAL code?
120 2011-01-16 00:32:49 <ArtForz> ATIs old GPU framework
121 2011-01-16 00:33:00 <Raulo> Without OpenCL?
122 2011-01-16 00:33:02 <ArtForz> yep
123 2011-01-16 00:33:10 <devon_hillard> nvidia have CUDA, a unique API with a large development community and lots of hardware in place
124 2011-01-16 00:33:13 <ArtForz> kernel code is IL, which is basically portable shader ASM
125 2011-01-16 00:33:16 <Raulo> Sounds very low-level
126 2011-01-16 00:33:20 <ArtForz> it is
127 2011-01-16 00:33:21 <Diablo-D3> devon_hillard: a SMALL development community
128 2011-01-16 00:33:30 <Diablo-D3> devon_hillard: almost no one gives a shit about CUDA
129 2011-01-16 00:33:38 <ArtForz> nvidia has good PR, and thats about it
130 2011-01-16 00:33:41 <devon_hillard> Diablo-D3: practically anyone doing GPGPU work started with CUDA and wound up with legacy hardware
131 2011-01-16 00:33:44 <Diablo-D3> its not open source, it runs on a minority of gpus, and nvidia makes nothing thats fast.
132 2011-01-16 00:33:53 <Diablo-D3> so no one cares about cuda.
133 2011-01-16 00:34:09 <ArtForz> OpenCL is a really great idea
134 2011-01-16 00:34:12 <mrb_> 16:26 < Raulo> 17 Mh/s with 115W is 0.15 Mh/W
135 2011-01-16 00:34:14 <mrb_> 16:27 < devon_hillard> Raulo: a rough calculations meant my CPU only got back 1/20 of the electricity cost
136 2011-01-16 00:34:24 <mrb_> no this shows it is actually slightly profitable
137 2011-01-16 00:34:51 <ArtForz> a well-educated guess, my CAL code should be able to get 610Mh/s with a stock 5970
138 2011-01-16 00:34:54 <devon_hillard> no, for every dollar spent, the CPU only raked back 5 cents
139 2011-01-16 00:34:58 <mrb_> at 17 Mh/s and 115W you spend $14.9 before finding a block
140 2011-01-16 00:35:02 <mrb_> 2**32*18437./(3600*17e6)*115/1000*.10
141 2011-01-16 00:35:02 <mrb_> 14.879780856504054
142 2011-01-16 00:35:04 <ArtForz> or maybe 607
143 2011-01-16 00:35:18 <mrb_> a block that is worth 50 BTC * $.38/BTC = $19
144 2011-01-16 00:35:28 <devon_hillard> mrb_: consider the power use of the whole box, not just the CPU
145 2011-01-16 00:35:28 <ArtForz> so thats 2.02Mh/W
146 2011-01-16 00:35:44 Zarutian has quit (Quit: Zarutian)
147 2011-01-16 00:36:05 <Raulo> devon_hillard: unless you run the box on idle anyway 24/7
148 2011-01-16 00:36:11 <mrb_> devon_hillard: even when considering the power of a regular PC, you are still making money
149 2011-01-16 00:36:30 <mrb_> using my formula if the whole PC consumes less than 150W you are making money
150 2011-01-16 00:36:43 <mrb_> a headless machines with a similar CPU draws less than 150W
151 2011-01-16 00:36:45 <ArtForz> box of 4 5970s @ 830MHz core average, 2.78Gh/s for 1300W, 2.14Mh/W
152 2011-01-16 00:37:45 <Raulo> ATI GPUs are much more effecient but best CPUs are still marginally profitable
153 2011-01-16 00:37:55 <ArtForz> probably not for long
154 2011-01-16 00:38:02 <devon_hillard> mrb_: my core 2 duo E8200 (OC from 2.6 to 3.2GHz) did only 3Mh/s and consumed about 200W, not including the monitors
155 2011-01-16 00:38:18 <ArtForz> stop OCing then
156 2011-01-16 00:38:23 <noagendamarket> someone should design a plug computer
157 2011-01-16 00:38:25 <devon_hillard> for 17Mh/h I don't think even a core i7 could cut it
158 2011-01-16 00:38:27 <ArtForz> if you want to go for efficiency, undervolt
159 2011-01-16 00:38:28 <noagendamarket> for bitcoin
160 2011-01-16 00:38:37 <hacim> a guruplug with a GPU?
161 2011-01-16 00:38:43 <noagendamarket> yeah
162 2011-01-16 00:38:48 <mrb_> devon_hillard: your PSU is probably not very efficient, especially at such a low load
163 2011-01-16 00:38:54 <hacim> how could you possibly plug in the GPU?
164 2011-01-16 00:39:04 <Raulo> Intel Core 2 is much slower than latest Athlons
165 2011-01-16 00:39:08 <tcatm> hacim: eval board of guruplug has a PCI-E slot
166 2011-01-16 00:39:23 <ArtForz> Core2 is on par with older K8 Athlon/Opteron arch
167 2011-01-16 00:39:40 <tcatm> porting the driver to arm could be a little bit of work, though
168 2011-01-16 00:39:41 <mrb_> devon_hillard: I was also assuming a diskless machine
169 2011-01-16 00:39:56 <ArtForz> erm, on you average miner box you have about 40W of CPU+mainboard for 1200W of GPUs, what are you trying to accomplish?
170 2011-01-16 00:40:08 <devon_hillard> the disk consumes some power and so does an optical drive (even idle)
171 2011-01-16 00:40:12 <hacim> tcatm: it does? I have a eval guruplug, but i dont see pci-e
172 2011-01-16 00:40:14 <devon_hillard> and all fans etc...
173 2011-01-16 00:40:31 <tcatm> hacim: the openrd (which the guruplug is based on) has
174 2011-01-16 00:40:40 <hacim> ohh
175 2011-01-16 00:40:40 <mrb_> yep... but trust me. I have diskless headless machines drawing <20W at idle. I know that <150W with a CPU at full load is possible.
176 2011-01-16 00:40:47 <ArtForz> sure is
177 2011-01-16 00:40:58 <ArtForz> depends a lot on the CPU
178 2011-01-16 00:41:22 <ArtForz> it's kidna hard with a modern 6-core
179 2011-01-16 00:41:27 <[Noodles]> mrb_ : and what electircity costs did you assume? i probably won't make any money running 17M for 150, or even 115W, i'll very likely make a loss, and not just a small one
180 2011-01-16 00:41:35 <slush> I'm trying m0mchil on 5970 and have strange problems. When I start only one instance, it gives me ~300mhash, but once I start second device, both are on 600-800k(!)hash.
181 2011-01-16 00:41:35 <mrb_> devon_hillard: just to clarify, I am _not_ an advocate of CPU-mining. I was just found it interesting that some CPUs are unbelievably still slighly profitable
182 2011-01-16 00:41:41 <slush> Any advice?
183 2011-01-16 00:41:48 <ArtForz> slush: what sdk version?
184 2011-01-16 00:41:52 <mrb_> [Noodles]: assuming $.10/kWhr (my location)
185 2011-01-16 00:41:52 <slush> 2.1
186 2011-01-16 00:41:57 <ArtForz> hummm
187 2011-01-16 00:42:05 <ArtForz> windows? linux?
188 2011-01-16 00:42:07 <slush> oooh, m0m isnt compatible with 2.1, right?
189 2011-01-16 00:42:11 <mrb_> [Noodles]: so yeah in many countries it would not be profitable
190 2011-01-16 00:42:13 <slush> I heard something like that
191 2011-01-16 00:42:17 <slush> windows
192 2011-01-16 00:42:18 <ArtForz> does it work if you only start it on the 2nd gpu?
193 2011-01-16 00:42:25 AAA_awright_ has joined
194 2011-01-16 00:42:49 <[Noodles]> even then my calculations make a loss, pay $19 to "earn" $15
195 2011-01-16 00:42:53 <slush> good question. No
196 2011-01-16 00:43:00 <[Noodles]> so how do you do that?
197 2011-01-16 00:43:03 <devon_hillard> [Noodles],mrb_: I would be breaking even at 54Mh/s, assuming $0.2/BTC (with 200W for the entire box), granted, it can be optimized further
198 2011-01-16 00:43:04 <ArtForz> hmmm. I think I heard of that one
199 2011-01-16 00:43:10 <slush> it runs 300mhash only on one core and only when one miner instance is running
200 2011-01-16 00:43:22 AAA_awright has quit (Ping timeout: 240 seconds)
201 2011-01-16 00:43:28 <slush> crossfire?
202 2011-01-16 00:43:32 <ArtForz> Q is, does the 2d core alone work
203 2011-01-16 00:43:33 AAA_awright_ is now known as AAA_awright
204 2011-01-16 00:43:35 <ArtForz> *2nd
205 2011-01-16 00:43:43 <ArtForz> if it does, it's probably crossfire related
206 2011-01-16 00:43:50 <slush> with Di I'm on 600
207 2011-01-16 00:43:55 <ArtForz> if it doesn't, it's the stupid power-saving features
208 2011-01-16 00:44:06 <ArtForz> okay, thats seriously weird
209 2011-01-16 00:44:24 <ArtForz> last time I checked they pretty much used the same OCL calls
210 2011-01-16 00:44:57 <slush> hmm, previous version worked here well. But until then I reinstalled SDK for some reasons
211 2011-01-16 00:45:07 <slush> ...to same version, but... you know...
212 2011-01-16 00:45:10 <slush> reinstalled
213 2011-01-16 00:45:14 <ArtForz> could m0s be picking up wrong sdk version dlls from somewhere?
214 2011-01-16 00:45:33 <slush> well, is there any tool to cleanup old versions?
215 2011-01-16 00:45:34 <niekie> Yikes, 3 hours on the current block, slush :P
216 2011-01-16 00:45:41 <ArtForz> iirc some versions of cat10.12 like installing sdk2.2 dlls
217 2011-01-16 00:45:53 <slush> I definitely need linux here, but I have to live with it at least two weeks
218 2011-01-16 00:46:16 <slush> niekie: We were more than 12 hours on one block, if I remember well
219 2011-01-16 00:46:18 <ArtForz> hmmm
220 2011-01-16 00:46:22 <niekie> Ouch.
221 2011-01-16 00:46:35 <niekie> At least I'm still getting enough shares in it, hehe.
222 2011-01-16 00:46:36 <ArtForz> I can't rememebr the DLL names off the top of my head
223 2011-01-16 00:46:47 <niekie> 50 shares so far.
224 2011-01-16 00:46:51 <ArtForz> it's the dlls in the lib folder of the sdk
225 2011-01-16 00:46:53 <slush> ArtForz: ok, let it be. I will try m0m on freshly installed linux then
226 2011-01-16 00:47:11 <Raulo> Does anybody know which mobo chipsets are Radeons 5xxx compatible? My 5850 did not boot (even no POST) on my old NForce4 mobo which I planned to use for mining
227 2011-01-16 00:47:22 <slush> I don't want to break anything, its on VNC and miner is 50km far away
228 2011-01-16 00:47:26 <ArtForz> I ran dual 5970s on a nforce4
229 2011-01-16 00:47:34 <devon_hillard> so does the CPU need to be minimally powerful to not get in the way of the GPU?
230 2011-01-16 00:47:44 <ArtForz> not really
231 2011-01-16 00:47:59 <niekie> slush: I guess it is statistical punishment for getting the last 2 blocks so fast ;)
232 2011-01-16 00:48:22 <ArtForz> my quad 5970 miners pretty much all have A2-X2s running at 800MHz
233 2011-01-16 00:48:34 <devon_hillard> wow
234 2011-01-16 00:48:50 <ArtForz> avg CPU use is like 2%
235 2011-01-16 00:48:57 <devon_hillard> so CPU undervolting is the key to efficiency
236 2011-01-16 00:49:41 <ArtForz> yep
237 2011-01-16 00:50:06 <slush> I hope singlecore sempron will be enough for 4x5970
238 2011-01-16 00:50:16 <slush> Now on windows, I'm on 100% cpu with single 5970
239 2011-01-16 00:50:21 <slush> weird problems on windows here...
240 2011-01-16 00:50:32 <mrb_> slush: I am running 4x5970 on a Sempron 140 on one of my boxes
241 2011-01-16 00:50:38 <Raulo> My 5850 is definitely OK. Worked on P38 mobo. But no boot on NF4.
242 2011-01-16 00:50:49 <slush> great, this is also 140
243 2011-01-16 00:50:58 <ArtForz> Raulo: weird, might be a BIOS issue
244 2011-01-16 00:50:59 <Raulo> I googled and found several instances of similar problems
245 2011-01-16 00:51:03 <mrb_> but some of my GPGPU tools cause Xorg to spike at 50-60% CPU
246 2011-01-16 00:51:08 <mrb_> haven't had time to investigate yet.
247 2011-01-16 00:51:19 <ArtForz> I had 2 5970s running on a Asus A8N-E
248 2011-01-16 00:51:19 <slush> which tools?
249 2011-01-16 00:51:30 <mrb_> whitepixel (an md5 hash brute forcer)
250 2011-01-16 00:51:36 <slush> oh
251 2011-01-16 00:51:48 <Raulo> ArtForz: maybe. But it's a crappy ASRock and I'm afraid there is no newer BIOS
252 2011-01-16 00:51:54 <ArtForz> ewwww
253 2011-01-16 00:51:59 <ArtForz> yeah, asrock sucks
254 2011-01-16 00:52:13 <Raulo> Before buying a motherboard I'd like to be sure it works
255 2011-01-16 00:52:29 <ArtForz> I have a 880G asrock here that flat out refuses to recognize a 5970
256 2011-01-16 00:52:52 <ArtForz> works fine with any single GPU card, but just doesnt work with dual-GPU cards
257 2011-01-16 00:53:09 <ArtForz> I suspect the BIOS has no clue how to handle PCIe bridges
258 2011-01-16 00:53:16 <lfm> what chipset?
259 2011-01-16 00:53:17 <Raulo> Well. Mine does not work even with 5850. And I planned to use it with 5970
260 2011-01-16 00:53:19 <devon_hillard> can you manually undervolt a CPU further than spec?
261 2011-01-16 00:53:39 <ArtForz> yep
262 2011-01-16 00:53:39 <Raulo> lfm: mine is Nforce4
263 2011-01-16 00:53:58 <lfm> iiieee nforce sucks too
264 2011-01-16 00:54:02 <ArtForz> lfm: Raulos is a nforce4 asrock, my old asus nforce4 works fine even with dual 5970s
265 2011-01-16 00:54:06 <hacim> aren't there a lot of nforce4 boards?
266 2011-01-16 00:54:25 <ArtForz> the fan on the NB died like 3 years ago, still works fine
267 2011-01-16 00:54:32 <slush> bitcoin economy size 2.02 mil USD, nice (bitcoinwatch)
268 2011-01-16 00:54:54 <slush> it was at 1 when I come here
269 2011-01-16 00:54:59 <devon_hillard> are AMD embedded GPUs any good for mining?
270 2011-01-16 00:55:07 <devon_hillard> in parallel with separate GPUs
271 2011-01-16 00:55:20 <devon_hillard> GPUs embedded in the mobo
272 2011-01-16 00:55:20 <ArtForz> yet a modern asrock board is flaky with gpu support -> the chipset is fine, AsRock bios coders are retarded monkeys
273 2011-01-16 00:55:27 <ArtForz> pointless really
274 2011-01-16 00:55:42 <lfm> devon_hillard probably not
275 2011-01-16 00:55:51 <ArtForz> a 5450 runs circles even around Fusion APUs
276 2011-01-16 00:55:51 <Raulo> ArtForz. What mobos you'd recommend?
277 2011-01-16 00:55:57 <devon_hillard> I think you can run an AMD card in crossfire with an AMD mobo (hybrid crossfire)
278 2011-01-16 00:56:08 <ArtForz> recommend? MSI 790FX-GD70
279 2011-01-16 00:56:27 <ArtForz> awesomest AMD board evar
280 2011-01-16 00:56:30 <Diablo-D3> ArtForz: except
281 2011-01-16 00:56:41 <Diablo-D3> a 5450 doesnt take 20 watts for the cpu, gpu, and northbridge.
282 2011-01-16 00:56:46 <hacim> A8N-SLI is ok
283 2011-01-16 00:57:41 <tcatm> ArtForz: Which license does your half-client use?
284 2011-01-16 00:57:48 <ArtForz> PD
285 2011-01-16 00:57:54 <tcatm> great :)
286 2011-01-16 00:58:00 <ArtForz> it's intended to be mostly a protocol specification
287 2011-01-16 00:58:05 <ArtForz> for cheap, stay the fuck away from asrock, foxconn and sapphire
288 2011-01-16 00:58:12 <Raulo> MSI 790FX-GD70 has 4 PCI-E slots. Nice but I will fry my PSU and melt my case
289 2011-01-16 00:58:19 <devon_hillard> Artforz: why?
290 2011-01-16 00:58:24 <Raulo> with 4 5970
291 2011-01-16 00:58:34 <slush> ArtForz: Also saphire 5970?
292 2011-01-16 00:58:39 <ArtForz> their bios coders all seem to live under the same rock
293 2011-01-16 00:59:07 <ArtForz> so far havent found a single board from them that could init 2 5970s properly
294 2011-01-16 00:59:57 <ArtForz> either they simply dont work at all with 5970s, or they dont recognize a 5970 in the 2nd slot
295 2011-01-16 01:00:56 <ArtForz> I kinda somewhat got the foxconn board to work by manually poking NB registers from linux over a serial port
296 2011-01-16 01:01:09 <ArtForz> yeah, reconfiguring PCIe lanes live!
297 2011-01-16 01:01:12 <hacim> a lot of people have the MSI 775, i havent seen any of the 790s
298 2011-01-16 01:01:49 <devon_hillard> how is sandy bridge, btw?
299 2011-01-16 01:02:06 <ArtForz> how bout the MSI 870-G45 ?
300 2011-01-16 01:02:07 <devon_hillard> cost-wise is like an amd ii x6
301 2011-01-16 01:02:09 <luke-jr> tcatm: https://en.bitcoin.it/wiki/Tonal_BitCoin#Python
302 2011-01-16 01:02:14 <ArtForz> FAST
303 2011-01-16 01:02:26 <Raulo> Any 2-PCIe mobos recommendations? What about Asus M4N75TD?
304 2011-01-16 01:02:38 <devon_hillard> so would you recommend hexacore amd or sandy bridge?
305 2011-01-16 01:02:43 <ArtForz> 2600K is about == to a $k i7
306 2011-01-16 01:02:49 <Diablo-D3> lol intel fail
307 2011-01-16 01:02:58 <ArtForz> yes, for multithreaded applications
308 2011-01-16 01:03:11 <luke-jr> devon_hillard: Sandy Bridge :P
309 2011-01-16 01:03:13 <ArtForz> single threaded it pretty much smokes anything
310 2011-01-16 01:03:24 <ArtForz> sandy bridge
311 2011-01-16 01:04:00 <ArtForz> 2 PCie, MSI 870-G45
312 2011-01-16 01:04:00 <luke-jr> do note that the good SB mobos aren't out for a month more
313 2011-01-16 01:04:05 <hacim> what are A2-X2s?
314 2011-01-16 01:04:10 <ArtForz> Athlon2 X2
315 2011-01-16 01:04:15 <ArtForz> 2nd slot is only x4 electrical, but it's cheap
316 2011-01-16 01:04:28 <hacim> ah
317 2011-01-16 01:04:28 <Diablo-D3> hacim: phenom IIs without L3, but double the L2.
318 2011-01-16 01:04:28 <ArtForz> and the slots are decently spaced
319 2011-01-16 01:04:29 <Raulo> Artforz: Thanks
320 2011-01-16 01:04:45 <devon_hillard> the GPUs don't really need more than 1x lanes for mining, do they?
321 2011-01-16 01:04:57 <Diablo-D3> hacim: you typically buy dual core as athlon II and quad core as phenom II
322 2011-01-16 01:05:00 <ArtForz> well, they lose a tiny bit of performance, but well < 1%
323 2011-01-16 01:05:13 <ArtForz> theres also quadcore A2s
324 2011-01-16 01:05:15 <devon_hillard> I saw people use a dremel tool on their 1x pci-ex slot to make a graphics card fit
325 2011-01-16 01:05:29 <Diablo-D3> yes, theres AII x4s, and PII x2s
326 2011-01-16 01:05:30 <ArtForz> (they're really PhenomIIs with disabled L3)
327 2011-01-16 01:05:32 <Diablo-D3> both pointless
328 2011-01-16 01:05:47 <Diablo-D3> ArtForz: no, they're AII mobile dies
329 2011-01-16 01:05:52 <Diablo-D3> they still have 2x L2.
330 2011-01-16 01:05:55 <hacim> why do you want that over actually having L3?
331 2011-01-16 01:06:14 <Diablo-D3> hacim: the L3 does nothing on so few cores
332 2011-01-16 01:06:21 <ArtForz> nope
333 2011-01-16 01:06:32 <hacim> ah
334 2011-01-16 01:06:33 <ArtForz> check out 645 for example
335 2011-01-16 01:06:43 <Diablo-D3> hacim: you have slightly better performance per dollar on AII x2 than PII x2
336 2011-01-16 01:06:58 <Diablo-D3> hacim: but AMD's gigantic dick size kicks in on PII x4
337 2011-01-16 01:07:05 <ArtForz> theres really plenty of A2s with 512k L2
338 2011-01-16 01:07:08 <luke-jr> tcatm: hurry up!
339 2011-01-16 01:07:14 <Diablo-D3> man
340 2011-01-16 01:07:18 <Diablo-D3> a 16 core Phenom II
341 2011-01-16 01:07:26 <ArtForz> most of em are normal A2s with half L2 disabled
342 2011-01-16 01:07:26 <Diablo-D3> would probably explode nutsacks
343 2011-01-16 01:07:44 <ArtForz> well, there are 12-core opterons...
344 2011-01-16 01:07:46 <Diablo-D3> you'd just be sitting there and POP
345 2011-01-16 01:07:51 <Diablo-D3> there goes your nutsack
346 2011-01-16 01:08:08 <Diablo-D3> ArtForz: but they cheat and have two memory controllers =P
347 2011-01-16 01:08:36 <ArtForz> anyways, *some* A2-X4s are crippled PhenomIIs
348 2011-01-16 01:09:15 <devon_hillard> when is the next AMD architecture planned? Q1 2011?
349 2011-01-16 01:09:24 <Diablo-D3> it already came out, devon
350 2011-01-16 01:09:30 <ArtForz> shortly after duke nukem forever is released *ducks*
351 2011-01-16 01:09:34 <devon_hillard> there's one with the APUs and then there was AM3+?
352 2011-01-16 01:09:36 <Diablo-D3> ArtForz: thats later this year.
353 2011-01-16 01:09:40 <ArtForz> sorry, Fusion != new CPU arch
354 2011-01-16 01:09:42 <hacim> i wonder what a cheap mining box would be, probably something that can do dual pci-e and run 5850s and above, with enough of a PSU to handle it
355 2011-01-16 01:09:48 <ArtForz> it's still the same crappy K10 cores
356 2011-01-16 01:09:56 <devon_hillard> APUs are for mobile or weak devices, right?
357 2011-01-16 01:09:57 <niekie> Diablo-D3: I'll believe it when I see it.
358 2011-01-16 01:10:03 <Diablo-D3> its a new arch in the sense that the GPU part is on die and a native HTX part.
359 2011-01-16 01:10:15 <Diablo-D3> devon_hillard: no, they'll have opteron fusions too
360 2011-01-16 01:10:22 <ArtForz> yeah, but not a new microarchitecture
361 2011-01-16 01:10:25 <Diablo-D3> devon_hillard: like 16 cores and a 68xx worth of GPU
362 2011-01-16 01:10:40 <ArtForz> = it's still the same CPU cores
363 2011-01-16 01:10:47 <tcatm> luke-jr: code inserted (without testing)
364 2011-01-16 01:10:47 <devon_hillard> but right now, only available for low-power stuff
365 2011-01-16 01:10:48 <hacim> you probably dont need much CPU if you are just GPU mining
366 2011-01-16 01:10:55 <ArtForz> and K10 is really showing it's age compared clock-for-clock with sandy bridge
367 2011-01-16 01:11:04 <Diablo-D3> ArtForz: yeah, but sandy vagina is overpriced shit
368 2011-01-16 01:11:10 <ArtForz> overpriced?
369 2011-01-16 01:11:16 <Diablo-D3> its slow, its overheating, its expensive, and intel is going to go bankrupt
370 2011-01-16 01:11:35 <Diablo-D3> maybe if they cut prices in half they can come close to beating AMD
371 2011-01-16 01:11:38 <devon_hillard> they had enough cash to bribe dell for years :)
372 2011-01-16 01:11:40 <devon_hillard> and still do
373 2011-01-16 01:11:47 <ArtForz> it's as fast as a fucking i7-980X for most workloads FFS
374 2011-01-16 01:11:56 <devon_hillard> and pay those high fines in the US and EU
375 2011-01-16 01:11:57 <Diablo-D3> devon_hillard: you know a company is fucked if they have to bribe people
376 2011-01-16 01:12:03 <luke-jr> [20:07:36] <bitcoincharts> TX 1LVndDxUHXAeH6bQ3j7jdvqmZW31kVLT4z 10 TBC, 1CQWT7zgb5gsLDgRkbkGSACZVBektX37Xm 1951424 uBTCents (b3fa1d5bae3df16b72317c5852694e8cc22f3c8f917ee04cc5b1c17dddd54a83)
377 2011-01-16 01:12:05 <luke-jr> tcatm: looks good there
378 2011-01-16 01:12:20 <devon_hillard> well, I loved the gall of Dell, who treated Intel like a personal ATM :)
379 2011-01-16 01:12:21 <ArtForz> well, AMD is nice if you just want a cheap quadcore
380 2011-01-16 01:12:21 <Diablo-D3> ubtc :D
381 2011-01-16 01:12:33 <devon_hillard> making their targets, quarter after quarter...
382 2011-01-16 01:12:35 <Diablo-D3> Intel is nice if you think your wallet is overweight
383 2011-01-16 01:12:42 <ArtForz> but if you *need* a fast CPU, SB beats anything
384 2011-01-16 01:12:47 <Diablo-D3> I can buy two AMD CPUs for the cost of an Intel
385 2011-01-16 01:12:52 <luke-jr> Diablo-D3: #Bitcoin-monitor
386 2011-01-16 01:12:58 <Diablo-D3> luke-jr: never.
387 2011-01-16 01:12:58 <ArtForz> two *slow* AMD CPUs
388 2011-01-16 01:13:05 <Diablo-D3> ArtForz: two FAST amd cpus
389 2011-01-16 01:13:09 <ArtForz> hexcore PhenomIIs arent exactly cheap
390 2011-01-16 01:13:17 <luke-jr> Diablo-D3: why not
391 2011-01-16 01:13:17 <Diablo-D3> sandy vagina isnt fast.
392 2011-01-16 01:13:24 <Diablo-D3> luke-jr: Im already on three bitcoin channels
393 2011-01-16 01:13:32 <luke-jr> Diablo-D3: but this one is fun
394 2011-01-16 01:13:37 <luke-jr> the game is simple:
395 2011-01-16 01:13:39 <Diablo-D3> thats what these three said
396 2011-01-16 01:13:49 <luke-jr> watch for tx
397 2011-01-16 01:13:55 <luke-jr> guess which address is the receiving end
398 2011-01-16 01:13:57 <ArtForz> it's not fast, it's the fastest per-core on the market
399 2011-01-16 01:14:01 <luke-jr> then send some extra BTC to confuse them
400 2011-01-16 01:14:14 <ArtForz> only thing coming close in total throughput are 12-core opterons and 6-core i7s
401 2011-01-16 01:14:16 <Diablo-D3> ArtForz: straight out? sure... but it uses more watts and costs more dollars
402 2011-01-16 01:14:19 <Diablo-D3> ArtForz: so why bother
403 2011-01-16 01:14:30 <ArtForz> more dollars than a $1k i7 ?
404 2011-01-16 01:14:30 <Diablo-D3> AMD still wins
405 2011-01-16 01:14:34 <ArtForz> I rather think not
406 2011-01-16 01:14:50 <Diablo-D3> more dollars than an AMD that scores better performance/dollar or performance/watt.
407 2011-01-16 01:14:55 <hacim> what does this mean "Military Class/ X6 CPU supported/ CPU CoreUnlock"
408 2011-01-16 01:14:58 <ArtForz> sorry, it doesn't
409 2011-01-16 01:14:59 <hacim> http://www.newegg.com/Product/Product.aspx?Item=N82E16813130290&Tpk=MSI%20870-G45
410 2011-01-16 01:15:06 <Diablo-D3> hacim: just MSI bullshit
411 2011-01-16 01:15:11 <ArtForz> AMD rules, but they really dropped the ball
412 2011-01-16 01:15:15 <tcatm> luke-jr: could be more confusing when you send to the change address ;)
413 2011-01-16 01:15:26 <hacim> ah
414 2011-01-16 01:15:34 <luke-jr> tcatm: you'd have to send more, then
415 2011-01-16 01:15:35 <Diablo-D3> hacim: its their solid state shit brand
416 2011-01-16 01:15:42 <Diablo-D3> hacim: gigabyte otoh calls theirs durable
417 2011-01-16 01:15:48 <ArtForz> a 2600K is nearly twice as fast as a PhenomII X6 1100T for heavily threaded apps
418 2011-01-16 01:15:48 <luke-jr> tcatm: since I'm sending 10 TBC quantities, I'm also screwing up the balance #s
419 2011-01-16 01:15:57 <ArtForz> more than wtwice as fast single threaded
420 2011-01-16 01:15:58 <luke-jr> ArtForz: not a 2600?
421 2011-01-16 01:15:58 <Diablo-D3> ArtForz: I doubt it.
422 2011-01-16 01:16:22 <ArtForz> I benchmarked it
423 2011-01-16 01:16:31 <Diablo-D3> and even if it was, its not enough to save intel from bankruptcy
424 2011-01-16 01:16:40 <Diablo-D3> too little too late
425 2011-01-16 01:16:45 <ArtForz> I like PhenomIIs, but SB is just fucking crazy fast
426 2011-01-16 01:16:46 <Diablo-D3> they should have came out with that 10 years ago.
427 2011-01-16 01:16:49 <ArtForz> and they OC like hell
428 2011-01-16 01:17:05 devon_hillard has quit (Read error: Connection reset by peer)
429 2011-01-16 01:17:12 <ArtForz> my 2600K gets 4.4 on air using the stock cooler
430 2011-01-16 01:17:48 <ArtForz> far from the 5GHz+ reviewers are claiming (handpicked press samples?), but from 3.4 stock it's pretty damn decent
431 2011-01-16 01:18:19 <luke-jr> ArtForz: K cant' do VT-d
432 2011-01-16 01:18:21 <Raulo> Is MSI 870A-G54A just an updated version of 870-G45?
433 2011-01-16 01:18:23 <ArtForz> yep
434 2011-01-16 01:18:32 <ArtForz> ....checks
435 2011-01-16 01:18:55 <Raulo> 870-G45s are hard to find in my country
436 2011-01-16 01:19:32 <ArtForz> sems to be the successor model
437 2011-01-16 01:19:55 <ArtForz> newer southbridge, USB3, 6Gb sata
438 2011-01-16 01:20:33 <hacim> Raulo: buying a mining machine?
439 2011-01-16 01:20:49 <ArtForz> btw, why the FUCK do mainboard mfgs put the 1st x16 slot as slot #2?
440 2011-01-16 01:20:50 <Raulo> It's more expensive but I can't find any reasonable 870-G45s
441 2011-01-16 01:20:51 <LobsterMan> ;;bc,stats
442 2011-01-16 01:20:53 <gribble> Current Blocks: 102889 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 1942 blocks | Next Difficulty In About: 1 week, 4 days, 22 hours, 26 minutes, and 42 seconds | Next Difficulty Estimate: 21028.90627601
443 2011-01-16 01:21:01 <LobsterMan> ;;bc,calc 108000
444 2011-01-16 01:21:02 <gribble> The average time to generate a block at 108000 Khps, given current difficulty of 18437.64439217 , is 1 week, 1 day, 11 hours, 40 minutes, and 32 seconds
445 2011-01-16 01:21:06 <Raulo> Thanks for the help.
446 2011-01-16 01:21:27 <Raulo> I hope I would find an old box that works with 5xxx, though
447 2011-01-16 01:21:31 <ArtForz> it makes no sucking fence
448 2011-01-16 01:21:34 <ArtForz> *fucking sense
449 2011-01-16 01:21:41 <hacim> sucking fence, lol
450 2011-01-16 01:23:39 <ArtForz> seriously, I sometimes want to hit those fuckwits over the head with the fucking ATX spec
451 2011-01-16 01:23:46 <tcatm> luke-jr: UnicodeEncodeError: 'ascii' codec can't encode character u'\ue9dc' in position 65: ordinal not in range(128)
452 2011-01-16 01:23:57 <hacim> Raulo: if you find a cheap configuration that works with the 5xxx cards, i'm interested to know what you find
453 2011-01-16 01:24:05 <ArtForz> preferably engraved in a nice solid tungsten plate
454 2011-01-16 01:25:26 <ArtForz> as I suspect plain steel won't have the neccesary oomph to get the message through
455 2011-01-16 01:25:58 Raulo has quit (Ping timeout: 265 seconds)
456 2011-01-16 01:29:25 <luke-jr> tcatm: do you understand that?
457 2011-01-16 01:29:41 <luke-jr> tcatm: sounds like a problem with the IRC end
458 2011-01-16 01:31:09 <luke-jr> tcatm: what are you using for that?
459 2011-01-16 01:31:39 * kiba ponders how many copies did fabianhjr sold today
460 2011-01-16 01:31:43 <tcatm> luke-jr: supybot
461 2011-01-16 01:32:43 <luke-jr> tcatm: plz paste your irc.reply or whatevr lin
462 2011-01-16 01:32:45 <luke-jr> line
463 2011-01-16 01:32:59 <tcatm> self.irc.queueMsg(ircmsgs.privmsg(self.plugin.channel, msg))
464 2011-01-16 01:33:39 <luke-jr> self.irc.queueMsg(ircmsgs.privmsg(self.plugin.channel, msg.encode('utf-8')))
465 2011-01-16 01:33:40 <luke-jr> try that
466 2011-01-16 01:33:54 acous has quit (Quit: Computer has gone to sleep.)
467 2011-01-16 01:34:12 * molecular is back from having some beers
468 2011-01-16 01:34:14 <tcatm> changed
469 2011-01-16 01:34:21 <luke-jr> tcatm: what's your donation address?
470 2011-01-16 01:34:46 <tcatm> 17asVKkzRGTFvvGH9dMGQaHe78xzfvgSSA
471 2011-01-16 01:35:14 <luke-jr> sent you 1î§ TBC
472 2011-01-16 01:35:39 <tcatm> seems to work
473 2011-01-16 01:35:44 <luke-jr> yep
474 2011-01-16 01:36:00 <luke-jr> [20:31:21] <bitcoincharts> TX 17asVKkzRGTFvvGH9dMGQaHe78xzfvgSSA 1î§ TBC, 1JntVnVu42ip6xPmLke8GnK3CSaQhsquFD 1164992 uBTCents afb99fffbed3f594cd49f7bbed107fb15c06b43f92ffa4a9486dbf5db49a2cc1
475 2011-01-16 01:37:32 <tcatm> which client are you using?
476 2011-01-16 01:37:48 <luke-jr> the original one, with bugs fixed
477 2011-01-16 01:38:01 <tcatm> patches?
478 2011-01-16 01:38:04 <luke-jr> yeah
479 2011-01-16 01:38:17 <tcatm> are they public?
480 2011-01-16 01:38:22 <luke-jr> yep
481 2011-01-16 01:38:34 <luke-jr> lp:~luke-jr/+junk/bitcoin-tonal
482 2011-01-16 01:39:06 <luke-jr> bzr branch lp:~luke-jr/+junk/bitcoin-tonal && cd bitcoin-tonal && qmake && make
483 2011-01-16 01:39:12 <luke-jr> and restart bitcoind with the new one
484 2011-01-16 01:39:34 <luke-jr> tested only on Gentoo so far
485 2011-01-16 01:39:50 <luke-jr> the old Makefiles should still work tho
486 2011-01-16 01:40:14 <luke-jr> it's not really tonal, you still need to specify BTC
487 2011-01-16 01:40:23 <luke-jr> but it allows fractions that have tonal values
488 2011-01-16 01:40:33 <luke-jr> and avoids fees
489 2011-01-16 01:41:15 <luke-jr> fyi, the amount I sent you is properly spoken "ton-vy bitcoins"
490 2011-01-16 01:42:30 <tcatm> mkay ;)
491 2011-01-16 01:43:25 <Slix`> How are bitcoin addresses encoded? Base64?
492 2011-01-16 01:43:33 <tcatm> custom Base58
493 2011-01-16 01:43:58 <Slix`> Why Base58?
494 2011-01-16 01:44:18 <luke-jr> Slix`: I learned by now not to ask "why" questions about bitcoin
495 2011-01-16 01:44:27 <luke-jr> a lot of it is quite frankly retarded XD
496 2011-01-16 01:44:43 <tcatm> Slix`: to avoid confusion with similiar letters like l and I
497 2011-01-16 01:45:11 <Slix`> Ahh. I guess that's a decent reason.
498 2011-01-16 01:45:29 <luke-jr> why is bitcoin protocol mixed with big and little endian?
499 2011-01-16 01:45:39 <ArtForz> it is?
500 2011-01-16 01:45:47 <luke-jr> yes
501 2011-01-16 01:45:57 <ArtForz> last time I checked we used LE everywhere except for ip+port in addr()
502 2011-01-16 01:46:09 <luke-jr> LE ⦠except ⦠is mixing :P
503 2011-01-16 01:46:40 <ArtForz> you'd rather have big-endian *IPs* ?!?
504 2011-01-16 01:46:44 <ArtForz> *liitle.-.endian
505 2011-01-16 01:46:48 <luke-jr> rather than mixed, yes
506 2011-01-16 01:46:55 <ArtForz> thats fucking backwards
507 2011-01-16 01:47:00 <ArtForz> big-endian IP makes sense
508 2011-01-16 01:47:02 <luke-jr> LE is, yes
509 2011-01-16 01:47:16 <ArtForz> because in octets it's the right way around
510 2011-01-16 01:47:29 <luke-jr> can the protocol negotiate the rest?
511 2011-01-16 01:47:31 <ArtForz> having BE port is... weird
512 2011-01-16 01:47:40 <luke-jr> eg, have new versions be pure BE?
513 2011-01-16 01:47:47 <ArtForz> errr... no
514 2011-01-16 01:48:04 <ArtForz> Block and TX data pretty much has to stay LE unless yo uwant to break the chain
515 2011-01-16 01:48:20 <luke-jr> can't convert on-the-fly?
516 2011-01-16 01:48:28 <ArtForz> sure, or byteswap everytime you [de]serialize
517 2011-01-16 01:49:10 <ArtForz> but imo "LE everywhere, except for ip/port pairs" is simple enough to support
518 2011-01-16 01:50:42 <slush> http://img17.allegroimg.pl/photos/400x300/14/06/48/75/1406487515
519 2011-01-16 01:50:42 <slush> from which vendor is this 5970?
520 2011-01-16 01:51:18 <ArtForz> never seen something like that
521 2011-01-16 01:51:31 <ArtForz> this is either some weird prototype or a 'shop
522 2011-01-16 01:51:36 <slush> 12x display port and 3 slots to high
523 2011-01-16 01:52:15 molecular has quit (Ping timeout: 264 seconds)
524 2011-01-16 01:52:34 <ArtForz> looks like a prototype eyefinity 5970
525 2011-01-16 01:53:17 <ArtForz> yep
526 2011-01-16 01:53:22 <slush> http://www.engadget.com/2010/04/30/powercolor-hd5970-eyefinity-12-makes-six-screens-yesterdays-new/
527 2011-01-16 01:53:31 <slush> yes, looks like this is even same picture
528 2011-01-16 01:53:46 <ArtForz> yep, crop from 1st pic
529 2011-01-16 01:54:02 <slush> oh, cool, the third row of display ports can be dismounted
530 2011-01-16 01:54:39 <ArtForz> also looks like it has 2*2GB
531 2011-01-16 01:55:00 <ArtForz> normal 5970 only has 8 ram chips per GPU
532 2011-01-16 01:55:25 <slush> is 360 euro fair price for this?
533 2011-01-16 01:55:57 <Diablo-D3> slush: its a powercolor
534 2011-01-16 01:55:59 <Diablo-D3> fuck. that.
535 2011-01-16 01:56:20 <ArtForz> not to mention it looks like it never went into production
536 2011-01-16 01:56:22 <slush> well, chips are the same, right?
537 2011-01-16 01:56:36 <Diablo-D3> and why the fuck are they selling that shit that way
538 2011-01-16 01:56:51 <slush> I see one in Czech bazaar
539 2011-01-16 01:56:53 molec has joined
540 2011-01-16 01:56:55 <luke-jr> http://blockexplorer.com/tx/a288fec5559c3f73fd3d93db8e8460562ebfe2fcf04a5114e8d0f2920a6270dc
541 2011-01-16 01:56:56 <luke-jr> WTF Is this
542 2011-01-16 01:57:00 <luke-jr> ArtForz?
543 2011-01-16 01:57:03 <slush> so it was probably in production...
544 2011-01-16 01:57:08 <molec> just got kicked
545 2011-01-16 01:57:10 malfy has joined
546 2011-01-16 01:57:14 <Diablo-D3> slush: its dumb.
547 2011-01-16 01:57:24 <ArtForz> "a transaction"
548 2011-01-16 01:57:28 <Diablo-D3> you'd be better off buying two cards with 6
549 2011-01-16 01:57:39 <ArtForz> we have 2 or 3 of these in the block chain
550 2011-01-16 01:57:40 <molec> ArtForz, what's your timezone?
551 2011-01-16 01:57:44 <luke-jr> ArtForz: a blackhole transaction?
552 2011-01-16 01:57:50 <ArtForz> some joker tried to consume a lot of CPU with duplicate OP_CHECKSIGS
553 2011-01-16 01:57:57 <ArtForz> GMT+1
554 2011-01-16 01:58:08 <molec> ArtForz, ah, hello!
555 2011-01-16 01:58:20 <slush> everybody in GMT+1 should be in bed now
556 2011-01-16 01:58:23 <slush> including me
557 2011-01-16 01:58:30 <molec> slush, no!
558 2011-01-16 01:58:49 <ArtForz> who cares, it's saturday... err... sunday
559 2011-01-16 01:58:56 <slush> :-D
560 2011-01-16 01:59:03 <molec> I tried my girlfriends yoga-lesson last week. since then, I can't sleep before 5:30
561 2011-01-16 01:59:14 <Diablo-D3> http://www.engadget.com/2011/01/15/man-discovers-glasses-free-3d-tech-in-the-blink-of-an-eye-video/
562 2011-01-16 01:59:15 <slush> Diablo-D3: well, is there any specific reason why reject this card? Except you don't like manufacturer? :)
563 2011-01-16 01:59:18 <Diablo-D3> this is disturbing
564 2011-01-16 01:59:24 <Diablo-D3> slush: cant trust the driver suport
565 2011-01-16 01:59:50 <ArtForz> shrug, I'd get one for the novelty value, but BIOs and driver support are probably flaky if existing at all
566 2011-01-16 01:59:51 <Diablo-D3> they're using the 6 display port controllers off the 2nd gpu
567 2011-01-16 01:59:56 <ArtForz> yep
568 2011-01-16 02:00:03 <Diablo-D3> it was never meant to be used that way
569 2011-01-16 02:00:05 <luke-jr> ArtForz: like you?
570 2011-01-16 02:00:15 <ArtForz> does 12-screen eyefinity with 2 5870s work?
571 2011-01-16 02:00:22 <Diablo-D3> also they're fucks
572 2011-01-16 02:00:27 <luke-jr> ArtForz: cmon, we know you're the only one capable of generating such tx
573 2011-01-16 02:00:31 <Diablo-D3> they should have put it on the vent slot
574 2011-01-16 02:00:33 <ArtForz> nope, wasnt me
575 2011-01-16 02:00:39 <Diablo-D3> and then put vents between each display port plug
576 2011-01-16 02:00:50 <ArtForz> I created a bunch of weird tx on testnet though
577 2011-01-16 02:01:08 <slush> Diablo-D3: ok, I accept it, quite good reasons
578 2011-01-16 02:01:19 <slush> too bad there are only few 5970 on market
579 2011-01-16 02:01:40 <Diablo-D3> slush: well, they can do the same trick with a 6990 anyhow
580 2011-01-16 02:01:40 <ArtForz> ?
581 2011-01-16 02:01:49 <Diablo-D3> they just connected the displayport pins
582 2011-01-16 02:01:52 <ArtForz> german etailers still have fucktons of em
583 2011-01-16 02:02:35 <slush> ArtForz: but for what price? In czech new cards are fucking expensive and old card are... who knows where?
584 2011-01-16 02:02:45 <ArtForz> 415EUR incl. tax
585 2011-01-16 02:02:53 <slush> ArtForz: new?
586 2011-01-16 02:03:19 <ArtForz> http://www.hoh.de/PC-Komponenten/Grafikkarten/ATI/PCIe/HD5970/Sapphire-HD-5970-2048MB-GDDR5-PCIe-Full-Retail_i9199_124724.htm
587 2011-01-16 02:03:28 <slush> in czech 550 euro and more
588 2011-01-16 02:03:30 Cusipzzz has quit ()
589 2011-01-16 02:04:20 <ArtForz> wow, that sucks
590 2011-01-16 02:04:37 cdecker has quit (Remote host closed the connection)
591 2011-01-16 02:04:42 <ArtForz> but yeah, seems .cz gets raped for tech prices even more than .de
592 2011-01-16 02:04:50 <ArtForz> still better than australia though ;)
593 2011-01-16 02:04:52 <slush> well, just find somebody who knows german and its done
594 2011-01-16 02:04:54 cdecker has joined
595 2011-01-16 02:05:01 <slush> mostly because we don't have euro, I think
596 2011-01-16 02:05:24 <ArtForz> not really, iirc I saw a bunch in france for 380EUR
597 2011-01-16 02:05:42 <slush> ohh, nice price
598 2011-01-16 02:06:20 <ArtForz> hmm... looks like HOH is actually the last shop with the 400EUR 5970s in stock
599 2011-01-16 02:06:35 <ArtForz> there were about half a dozen shops that still had em in stock like a week ago
600 2011-01-16 02:06:49 <slush> lol, just found eshop with 850euro for 5970
601 2011-01-16 02:07:19 <slush> (in czech, of course)
602 2011-01-16 02:07:26 <ArtForz> wow.
603 2011-01-16 02:07:42 <ArtForz> "normal" 5970s are ~500EUR in .de
604 2011-01-16 02:07:49 <ArtForz> well, 500-550
605 2011-01-16 02:08:01 <ArtForz> only the sapphires were that cheap for a while
606 2011-01-16 02:08:32 <ArtForz> as sapphire is AMDs fav. OEM, I guess they got the last remaining stock of cypress chips
607 2011-01-16 02:08:49 <Diablo-D3> s/AMD/ATI/ s/ATI/AMD/ *cough*
608 2011-01-16 02:09:08 <ArtForz> s/AMD/DAAMIT/
609 2011-01-16 02:09:19 <Diablo-D3> hee
610 2011-01-16 02:09:48 <ArtForz> we're red! we're green! we're in between!
611 2011-01-16 02:09:59 <slush> http://www.anacomp.cz/gigabyte-hd5870-eyefinityx6-edition-gv-r597d5-2gd-b-2gb-gddr5-256bit-pci-e-6xmini-display-port_d44148.html
612 2011-01-16 02:09:59 <slush> This is the same shit as those 12x displayport box, right?
613 2011-01-16 02:10:09 <slush> I ask because it is cheap and new and in czech :)
614 2011-01-16 02:10:39 <ArtForz> anyways, sapphire builds all of AMD/ATI/he-who-shall-not-be-nameds reference cards
615 2011-01-16 02:10:59 <ArtForz> thats a normal x6 eyefinity
616 2011-01-16 02:11:03 <ArtForz> 5870
617 2011-01-16 02:11:15 <slush> oh, its 5870, soooryyy
618 2011-01-16 02:11:33 <slush> That explains a lot (at least price)
619 2011-01-16 02:13:33 <ArtForz> hmmm... looks like 5850s might actually be a better deal
620 2011-01-16 02:13:43 <ArtForz> 160EUR in .de
621 2011-01-16 02:14:13 <ArtForz> not bad considering a 5970 is only 2.23x as fast
622 2011-01-16 02:15:42 <ArtForz> so you need about twice the mainboards+CPU+ram, same amount of PSUs (4*5850 on one 800W shouldnt be a problem)
623 2011-01-16 02:16:53 <slush> maybe cheaper, but what about power efficiency?
624 2011-01-16 02:17:08 <ArtForz> 5850 is nearly as good as 5970
625 2011-01-16 02:17:12 <ArtForz> better than 5870
626 2011-01-16 02:18:00 <ArtForz> the official 151W figure for 5850 is too high, they're closer to 130-135W
627 2011-01-16 02:18:32 <ArtForz> guess AMD wanted some headroom if they had bad chips that needed higher voltage to be stable or something
628 2011-01-16 02:19:31 <ArtForz> cheap dual x16 board + cpu + ram is ~120 EUR
629 2011-01-16 02:19:43 <ArtForz> so 60 EUR per card
630 2011-01-16 02:20:13 <molec> ArtForz, wohnst du in DE?
631 2011-01-16 02:20:23 <ArtForz> yep, nordbayern
632 2011-01-16 02:20:35 <molec> würz burg?
633 2011-01-16 02:20:36 <ArtForz> yeah, 5850 should actually end up cheaper
634 2011-01-16 02:20:51 <ArtForz> n stück südlich von regensburg
635 2011-01-16 02:21:05 <cdecker> Nice I see we have a lot of German guys here :D
636 2011-01-16 02:21:18 * cdecker is aus Hamburg
637 2011-01-16 02:21:36 <ArtForz> yeah, kinda crazy how many germans like bitcoin
638 2011-01-16 02:21:44 <cdecker> hehe
639 2011-01-16 02:22:03 <cdecker> Euro's not doing so fine
640 2011-01-16 02:22:17 <cdecker> I shouldn't complain
641 2011-01-16 02:22:24 * molec wohnt in Hamburg
642 2011-01-16 02:22:24 <cdecker> I live in CHF land :D
643 2011-01-16 02:22:43 <molec> cdecker, where in Hamburg?
644 2011-01-16 02:23:00 <cdecker> Long time since I lived in Hamburg
645 2011-01-16 02:23:06 <molec> am in eimsbüttel
646 2011-01-16 02:23:15 <cdecker> But I still like to call myself a hamburger :D
647 2011-01-16 02:23:15 <molec> allright
648 2011-01-16 02:23:18 * tcatm ~100km nördlich von Hamburg
649 2011-01-16 02:23:46 <molec> tcatm: not much choice... 100km north of hamburg.
650 2011-01-16 02:23:50 <ArtForz> anyways, from US PoV we're supposed to be liberal hippie commies or something ;)
651 2011-01-16 02:23:55 <cdecker> Nice should meet some day :D
652 2011-01-16 02:24:02 <cdecker> Anyway got to go
653 2011-01-16 02:24:08 altamic has quit (Ping timeout: 250 seconds)
654 2011-01-16 02:24:09 <cdecker> See you soon ^^
655 2011-01-16 02:24:33 <molec> when I was in the stats in 1992, some people seriously asked me, whether we had (electrical) power in germany.
656 2011-01-16 02:24:42 <molec> cu, cdecker
657 2011-01-16 02:24:55 <molec> s/stats/states
658 2011-01-16 02:24:55 <ArtForz> don't forget about the medical death boards ;)
659 2011-01-16 02:25:38 <ArtForz> maybe something with the hacker (counter)culture in .de ?
660 2011-01-16 02:26:02 <molec> not sure, something in their past about avoiding authority??
661 2011-01-16 02:26:11 <ArtForz> yep
662 2011-01-16 02:26:28 <molec> ArtForz, would love to meet you some time in person
663 2011-01-16 02:26:30 <ArtForz> overall german hackers seem to like decentralized crypto-anything ;)
664 2011-01-16 02:26:37 <molec> yeah
665 2011-01-16 02:26:37 <slush> hmm, 5850 is nowhere in bazaars and new one is more than 50% of 5970 here. So when I count in also more needed slots and CPUs, it perform worse than 5970 for me
666 2011-01-16 02:26:50 <molec> also: you should demo your ASICs next year at 28c3
667 2011-01-16 02:26:53 <ArtForz> freenet, tor, I2P, now bitcoin
668 2011-01-16 02:26:55 <slush> yes, also tor is very popular in german
669 2011-01-16 02:27:04 <molec> also popular in iran
670 2011-01-16 02:27:08 <ArtForz> I maybe will
671 2011-01-16 02:27:26 <ArtForz> also, the guys presentation on DES on FPGAs @ 27C3 was ... bad
672 2011-01-16 02:27:30 <molec> if you want any help, I'm here!
673 2011-01-16 02:27:35 <ArtForz> several HUGE factual errors
674 2011-01-16 02:27:50 <molec> didn't see that (yet)
675 2011-01-16 02:28:08 <ArtForz> his claimed numbers for DES on GPU were 1/20 of what I could get on a 5970 without even trying
676 2011-01-16 02:29:09 <slush> but yes, 5850 looks like better deal than 5870, definitely
677 2011-01-16 02:29:45 <molec> can you point me to the vid of the talk?
678 2011-01-16 02:30:01 <ArtForz> I only watched the live stream
679 2011-01-16 02:30:09 <molec> allright... they guy's name?
680 2011-01-16 02:30:17 <luke-jr> ;;bc,calc 4205
681 2011-01-16 02:30:19 <gribble> The average time to generate a block at 4205 Khps, given current difficulty of 18437.64439217 , is 31 weeks, 0 days, 23 hours, 8 minutes, and 43 seconds
682 2011-01-16 02:30:20 <ArtForz> http://events.ccc.de/congress/2010/Fahrplan/events/4203.en.html
683 2011-01-16 02:30:58 <molec> thx
684 2011-01-16 02:31:34 <ArtForz> I kinda liked his idea of using JTAG for communication
685 2011-01-16 02:32:04 <molec> nice idea to harvest old electronics for fpgas
686 2011-01-16 02:32:12 <ArtForz> yep
687 2011-01-16 02:32:32 <ArtForz> it's pretty slow, but you pretty much can't get lower pincount, and finding JTAG on old boards is usually pretty easy
688 2011-01-16 02:32:44 <molec> but less effective than buying asics if you have the $
689 2011-01-16 02:32:50 <ArtForz> depends
690 2011-01-16 02:33:01 <ArtForz> if you get a good price, it could be on par
691 2011-01-16 02:33:08 <molec> u think?
692 2011-01-16 02:33:13 <ArtForz> yep
693 2011-01-16 02:33:22 <ArtForz> problem is you really need a quite big FPGA to fit a unrolled sha256 engine, let alone 2
694 2011-01-16 02:34:11 <molec> if u use a small fpga @(say)200mhz, say, too small to fit sha-256 pipeline, what will you get?
695 2011-01-16 02:34:30 * kiba landed a 100 dollars worth of bitcoin job
696 2011-01-16 02:34:30 <ArtForz> one bitcoinhash is 2 sha256 blocks
697 2011-01-16 02:34:50 <ArtForz> so if you can fit 1/2 of a block pipelined, you get 1 hash / 4 clocks
698 2011-01-16 02:35:06 <molec> 1 has / 4 clocks sounds good
699 2011-01-16 02:35:29 <ArtForz> getting >120MHz out of a spartan6 is a small miracle :/
700 2011-01-16 02:35:39 <ArtForz> and spartan3 is a lot slower still
701 2011-01-16 02:36:26 <molec> does the max clock somehow depend on the logic implemented on the fpga?
702 2011-01-16 02:36:35 <ArtForz> yes
703 2011-01-16 02:37:06 <ArtForz> a single layer of logic+flipflop can get 350MHz+ on a S6
704 2011-01-16 02:37:18 * molec has to play with a fpga
705 2011-01-16 02:37:18 <molec> do you have some old fpga dev-board laying around that I could useÃ
706 2011-01-16 02:37:18 <molec> s/Ã/?
707 2011-01-16 02:37:21 <molec> ... just to get the hang=
708 2011-01-16 02:37:25 <ArtForz> nope
709 2011-01-16 02:37:29 <molec> s/=/.
710 2011-01-16 02:37:32 <molec> ok
711 2011-01-16 02:37:34 <ArtForz> I only have 3
712 2011-01-16 02:37:45 <molec> can you recommend how I could search on ebay?
713 2011-01-16 02:38:04 <ArtForz> thats kinda tricky, the talk explains it pretty well
714 2011-01-16 02:38:26 <molec> I really enjoiyed making that 8bit cpu in vhdl and seeing that run at UNI-stgt
715 2011-01-16 02:38:47 <ArtForz> yep, seeing a VHDL design work in hardware is pretty amazing
716 2011-01-16 02:39:05 <ArtForz> at least for the first few times
717 2011-01-16 02:39:10 <molec> only ran @1Mhz, but still, it was like a baby born and living
718 2011-01-16 02:39:26 <molec> 1khz?
719 2011-01-16 02:39:40 <molec> dunno, it was really slow
720 2011-01-16 02:39:51 <molec> some bug in the c-code that setup the shit, not my fault ;I)
721 2011-01-16 02:39:59 <ArtForz> kinda like building your first electric vehicle, or completely overhauling your first engine
722 2011-01-16 02:40:13 <molec> never did one of these
723 2011-01-16 02:40:22 Cusipzzz has joined
724 2011-01-16 02:40:25 <ArtForz> did all of them
725 2011-01-16 02:40:30 <molec> might be easier to just fuck some girl ;)
726 2011-01-16 02:41:04 <molec> ArtForz, if you ever around Hamburg, and looking for place to sleep, call 0176 6126597
727 2011-01-16 02:41:07 <ArtForz> completely rebuilt the diesel engine in my '87 VW T3
728 2011-01-16 02:41:32 <molec> sorry, got the nubmer wrong
729 2011-01-16 02:41:34 <ArtForz> ex-BW sani-bus
730 2011-01-16 02:41:36 <molec> 0176 61265697
731 2011-01-16 02:43:34 <ArtForz> my daily driver is a '89 ford transit ;)
732 2011-01-16 02:44:13 <ArtForz> old car = 1.5kEUR/yr+ tax
733 2011-01-16 02:44:34 <ArtForz> old "small truck" ~ 120EUR/yr tax
734 2011-01-16 02:46:30 <necrodearia> http://www.youtube.com/user/stefbot#p/u/14/7R4ppEADvm4
735 2011-01-16 02:47:23 <molec> what's your work to require a daily drive?
736 2011-01-16 02:47:34 <molec> you seem to be making enough BTC to live from it
737 2011-01-16 02:48:15 <necrodearia> and http://www.youtube.com/user/stefbot#p/u/12/Cu9HUBs5mp0
738 2011-01-16 02:48:33 <ArtForz> I'm doing electronic/control design for printing/binding machines
739 2011-01-16 02:48:54 <ArtForz> = on-demand one-off book runs
740 2011-01-16 02:49:31 <ArtForz> pretty cool tech, and I have pretty much the coolest boss ever
741 2011-01-16 02:50:43 <ArtForz> no fixed work hours *at all*
742 2011-01-16 02:51:09 Cusipzzz has quit ()
743 2011-01-16 02:51:26 <ArtForz> work at home, come in 2 days a week at 2-7AM, who cares as long as the job gets done
744 2011-01-16 02:52:40 <molec> you come in at 02:00-0700 ?
745 2011-01-16 02:52:57 <ArtForz> if I want to, yes
746 2011-01-16 02:53:16 <molec> you have this insanely nice job
747 2011-01-16 02:53:21 <ArtForz> or work 9-5 too or just work from home if I want to
748 2011-01-16 02:53:23 <molec> envy!
749 2011-01-16 02:53:32 <molec> arsch!
750 2011-01-16 02:53:34 <molec> :)
751 2011-01-16 02:53:41 <ArtForz> like I said. coolest. boss. ever.
752 2011-01-16 02:54:01 <molec> nice
753 2011-01-16 02:54:14 <molec> what more can you ask for?
754 2011-01-16 02:54:25 <ArtForz> nothing really, I'm pretty happy
755 2011-01-16 02:54:55 <molec> we were not put in this world to be happy
756 2011-01-16 02:55:13 <ArtForz> well, if you're not happy, you're doing it wrong ;)
757 2011-01-16 02:55:21 <molec> there must be something to solve/do/fuck
758 2011-01-16 02:55:45 <ArtForz> aka strive for greatness, be content with mediocrity
759 2011-01-16 02:55:46 <molec> I'm happy.... just not happy enough
760 2011-01-16 03:05:17 <luke-jr> [21:44:28] <bitcoincharts> TX 1NHeFo9B5VUsAWw42P6f2Q7nhdh6CJbQy 1239744 uBTCents, 1KpLXJqQLSx87sjns33AKgZDP9cSDuKBXx î§î§ TBC bdbea3ba47130140214620ae1775fd334ac07478e2abe58d91e8350351d97877
761 2011-01-16 03:05:47 <tcatm> Is there anyone else using TBC?
762 2011-01-16 03:05:54 <ArtForz> I like this guy
763 2011-01-16 03:06:16 <luke-jr> tcatm: with time âº
764 2011-01-16 03:06:29 <luke-jr> tcatm: right now, lack of clients makes it difficult
765 2011-01-16 03:06:37 <tcatm> luke-jr: So you're the only one? :P
766 2011-01-16 03:06:48 <luke-jr> tcatm: I've been giving prizes in TBC, and such âº
767 2011-01-16 03:07:16 <tcatm> => lots of subcent change that'll end up in fees. Good for miners :)
768 2011-01-16 03:07:30 <luke-jr> yeah, until people get fixed clients :P
769 2011-01-16 03:11:09 <luke-jr> I must say, marioxcc's idea would have been a really bad idea :P
770 2011-01-16 03:11:22 <tcatm> What is his idea?
771 2011-01-16 03:11:24 <luke-jr> doing it the way he wanted, would have made an algorithm to detect BTC vs TBC impossible
772 2011-01-16 03:11:28 <luke-jr> he wanted 1 BTC = 1 TBC
773 2011-01-16 03:14:05 <niekie> What on earth is TBC.
774 2011-01-16 03:14:15 <luke-jr> niekie: Tonal BitCoin
775 2011-01-16 03:14:19 <luke-jr> as opposed to Decimal BitCoin
776 2011-01-16 03:14:27 <nameless> !~root@weowntheinter.net|I want 1NBC
777 2011-01-16 03:14:30 <niekie> So using another numbering base?
778 2011-01-16 03:14:35 <luke-jr> niekie: yes
779 2011-01-16 03:14:37 <nameless> !~root@weowntheinter.net|1NBC = 9999999999999999999999999999999999999999999BTC
780 2011-01-16 03:14:44 <niekie> What exactly are the advantages in that?
781 2011-01-16 03:14:45 <luke-jr> nameless|: impossible then
782 2011-01-16 03:14:49 <luke-jr> niekie: decimal sucks
783 2011-01-16 03:14:51 <nameless> !~root@weowntheinter.net|luke-jr: nonsense
784 2011-01-16 03:15:00 <nameless> !~root@weowntheinter.net|I just need more processing power than all of you combined
785 2011-01-16 03:15:07 * tcatm hopes luke-jr will send him 1 TBTC by accident
786 2011-01-16 03:15:10 <luke-jr> nameless|: BTC is capped at 21mil
787 2011-01-16 03:15:16 <nameless> !~root@weowntheinter.net|luke-jr: Like I said
788 2011-01-16 03:15:21 <luke-jr> tcatm: I thought I did
789 2011-01-16 03:15:34 <tcatm> (Tera Bitcoin)
790 2011-01-16 03:15:47 <luke-jr> lame SI/Decimal
791 2011-01-16 03:33:25 * kiba finally trumps over a Khan Academy problem set
792 2011-01-16 03:48:03 Diablo-D3 has quit (Ping timeout: 264 seconds)
793 2011-01-16 04:12:12 eureka^ is now known as Diet_Dr_Thunder
794 2011-01-16 04:18:38 <newsham> *raises hands in air and yells*KHANNNNNN!!!!
795 2011-01-16 05:26:29 jgarzik_ is now known as jgarzik
796 2011-01-16 05:26:32 sabalaba has joined
797 2011-01-16 05:26:37 jgarzik has quit (Changing host)
798 2011-01-16 05:26:37 jgarzik has joined
799 2011-01-16 05:27:26 ArtForz has quit (Read error: Connection timed out)
800 2011-01-16 05:28:03 ArtForz has joined
801 2011-01-16 05:36:32 <malfy> why is there a connection to an irc network/channel from the bitcoin daemon?
802 2011-01-16 05:37:15 <nanotube> it's used for bootstrapping.
803 2011-01-16 05:38:18 <malfy> i dont understand
804 2011-01-16 05:38:29 <malfy> isn't it supposed to be a p2p network?
805 2011-01-16 05:38:36 <malfy> there should be no need for a huge channel
806 2011-01-16 05:38:41 <nanotube> yes, but any p2p network needs a method to bootstrap, to find other peers.
807 2011-01-16 05:38:56 <nanotube> you don't need irc if you don't want it, you can run with -noirc. there's a hardcoded list of seednodes.
808 2011-01-16 05:39:10 <malfy> ok
809 2011-01-16 05:54:51 noagendamarket has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
810 2011-01-16 06:06:30 presence has joined
811 2011-01-16 06:06:45 <presence> any mining.bitcoin.cz guys on?
812 2011-01-16 06:10:24 <presence> it seems that your account pw and your worker pws have to align
813 2011-01-16 06:10:41 <presence> yet it displays the worker pw on the account url...which I would prefer not to reveal.
814 2011-01-16 06:11:30 presence has left ()
815 2011-01-16 06:23:38 <andrew12> aww
816 2011-01-16 06:23:47 <andrew12> I wanted to say something witty like 'people can mine for you ;)'
817 2011-01-16 06:24:01 pr0wler has quit (Ping timeout: 240 seconds)
818 2011-01-16 06:33:18 sabalaba has quit (Ping timeout: 240 seconds)
819 2011-01-16 06:34:07 afed_ has joined
820 2011-01-16 06:44:44 RazielZ has joined
821 2011-01-16 06:59:07 Bth8 has quit (Ping timeout: 255 seconds)
822 2011-01-16 07:00:40 Bth8 has joined
823 2011-01-16 07:06:58 joe_1 has joined
824 2011-01-16 07:07:11 <hacim> what kind of wattage is needed in a PSU to fire 5xxx cards?
825 2011-01-16 07:10:06 <RazielZ> One 5xxx card?
826 2011-01-16 07:10:09 <RazielZ> And which one?
827 2011-01-16 07:10:10 <RazielZ> ._.
828 2011-01-16 07:10:11 <mrb_> hacim: depends. a 5450 only needs 19W for itself. a 5970 needs 300W for itself. a power-efficient machine may draw no more than 80W under full load so in theory, even with a high-end 5970, a 380W PSU is sufficient.
829 2011-01-16 07:10:27 <RazielZ> 380W?
830 2011-01-16 07:10:31 <RazielZ> Yeah, go for atleast 800
831 2011-01-16 07:10:34 <RazielZ> :D
832 2011-01-16 07:10:50 <mrb_> I have done that myself (400W PSU for a 5970)
833 2011-01-16 07:11:04 <RazielZ> What the nts
834 2011-01-16 07:11:06 <RazielZ> nuts even
835 2011-01-16 07:11:35 <RazielZ> I'd be too scareed to power my PC with anything less than 750W
836 2011-01-16 07:11:36 <mrb_> only possible to do safely... if you have a clamp-meter and knows exactly how many Amps you are drawing :)
837 2011-01-16 07:11:36 <RazielZ> ._.
838 2011-01-16 07:12:05 <mrb_> I have also run 4 x 5970 at full load with only two 560W PSU
839 2011-01-16 07:12:23 <RazielZ> 4 5970
840 2011-01-16 07:12:28 <RazielZ> want.
841 2011-01-16 07:13:34 <RazielZ> Really, if you have any spares you can send em here, I won't mind
842 2011-01-16 07:17:54 CyanDynamo has quit (Quit: Leaving.)
843 2011-01-16 07:18:41 ElectRo` has left ()
844 2011-01-16 07:18:49 <hacim> ok, so two 5870s would be fine on a 500w or so, just shopping around for a PSU
845 2011-01-16 07:19:51 <mrb_> hacim: watch the max Amp on the 12V rail. two 5870s = 2*188W = 31 Amp at 12V
846 2011-01-16 07:20:11 <mrb_> I bet some crappy 500W PSU may not even be capable of supplying 31A@12V
847 2011-01-16 07:20:27 <mrb_> on the other hand some high-end 400W ones may be able to do it
848 2011-01-16 07:20:33 <mrb_> read the specs.
849 2011-01-16 07:20:59 <hacim> mrb_: what is the max amp on the 12v?
850 2011-01-16 07:22:15 <hacim> so, eg. this corsair 500w is +12V@34A, so it could handle the two 5870s
851 2011-01-16 07:22:32 <mrb_> yep this one seems good
852 2011-01-16 07:22:56 <mrb_> although that leaves only 3 Amps for the rest of the system
853 2011-01-16 07:23:07 <mrb_> 3*12 = 36 Watt
854 2011-01-16 07:23:30 <mrb_> if you have an energy efficient CPU, that may be ok
855 2011-01-16 07:24:08 <RazielZ> also if you don't plan on overclocking
856 2011-01-16 07:24:21 <mrb_> on the other hand you are unlikely to hit the maximum amp on two 5870, even when mining 24/7
857 2011-01-16 07:24:26 * kiba set to work on his first bitcoin freelance programming job
858 2011-01-16 07:24:49 <mrb_> maximum *theoretical* amp
859 2011-01-16 07:25:07 <hacim> hmm, so maybe that is cutting it too close?
860 2011-01-16 07:25:25 <mrb_> I dunno, what's your hardware?
861 2011-01-16 07:26:36 <mrb_> if you want to be comfortable, leave 10 Amps to the rest of the system. 120W will account for a 95W CPU at full load plus a ~30W on the mobo/ram/hdd
862 2011-01-16 07:27:01 <hacim> mrb: haven't bought anything yet, but this is what I was looking at: https://micah.riseup.net/shots/2011.01.16-02.21.1295162515.xTjPuKdczM.png
863 2011-01-16 07:27:30 <hacim> i welcome any suggestions for improvements
864 2011-01-16 07:27:37 <mrb_> 95W CPU.
865 2011-01-16 07:28:16 <mrb_> yeah look for a PSU capable of at least 40A@12V
866 2011-01-16 07:28:27 <hacim> perhaps I can get a cheaper/lower power cpu
867 2011-01-16 07:28:27 <RazielZ> that RAM is cheap D:
868 2011-01-16 07:28:41 <hacim> RazielZ: yeah, I'm worried it might be shit
869 2011-01-16 07:28:51 <RazielZ> I don't know...
870 2011-01-16 07:28:55 <mrb_> your config at CPU and GPU full load would draw (95+188*2)/12 = 39.25 Amp
871 2011-01-16 07:29:06 <RazielZ> crucial is usually decent I think
872 2011-01-16 07:29:11 <RazielZ> we don't really have it around here
873 2011-01-16 07:29:30 <RazielZ> but the whole PC is cheap so sure, why not
874 2011-01-16 07:29:42 <RazielZ> I'm not really familiar with the lower end of components >.>
875 2011-01-16 07:30:04 <hacim> i could drop that CPU down to 65W
876 2011-01-16 07:30:26 <mrb_> good idea
877 2011-01-16 07:30:37 * mrb_ off to watch a movie
878 2011-01-16 07:31:09 <RazielZ> I hate your prices so much. Things are way more expensive here
879 2011-01-16 07:38:33 <kiba> price hating, RazielZ?
880 2011-01-16 07:48:19 xelister_ has quit (Quit: Konversation terminated!)
881 2011-01-16 07:48:51 xelister has joined
882 2011-01-16 07:48:52 xelister has quit (Changing host)
883 2011-01-16 07:48:52 xelister has joined
884 2011-01-16 07:56:14 <RazielZ> Yeah.
885 2011-01-16 08:07:48 Myckel has joined
886 2011-01-16 08:07:49 Myckel has quit (Changing host)
887 2011-01-16 08:07:49 Myckel has joined
888 2011-01-16 08:16:27 Slix` has quit (Read error: Connection reset by peer)
889 2011-01-16 08:19:52 TheAncientGoat has joined
890 2011-01-16 08:32:38 ArtForz has quit (Read error: Connection timed out)
891 2011-01-16 08:33:06 ArtForz has joined
892 2011-01-16 09:03:04 SirDidi has joined
893 2011-01-16 09:13:57 <lfm> ;;bc,calcd 150000 373.70862152
894 2011-01-16 09:13:57 <gribble> The average time to generate a block at 150000 Khps, given the supplied difficulty of 373.70862152, is 2 hours, 58 minutes, and 20 seconds
895 2011-01-16 09:16:02 <nanotube> lfm: why 373?
896 2011-01-16 09:16:31 <lfm> testnet
897 2011-01-16 09:16:55 <nanotube> ah
898 2011-01-16 09:17:03 <nanotube> testnet diff is based on 1==0.4
899 2011-01-16 09:17:06 <nanotube> 0.5 that is
900 2011-01-16 09:17:12 <nanotube> so mainnet equivalent difficulty is 186
901 2011-01-16 09:17:20 <nanotube> point something
902 2011-01-16 09:18:59 <lfm> well i dont understand that
903 2011-01-16 09:20:10 <lfm> does that mean it averages to 5 min per block then?
904 2011-01-16 09:20:10 <nanotube> so... in mainnet, when difficulty is 1.0, the hextarget is 0x00000000FFFF0000000000000000000000000000000000000000000000000000
905 2011-01-16 09:20:32 <nanotube> on testnet, difficulty of 1.0, hextarget is twice that. (lshift by one more digit)
906 2011-01-16 09:21:30 <nanotube> it still 'aims' at 10min per block, just that 'minimum diff' is essentially 0.5, rather than 1
907 2011-01-16 09:22:07 <nanotube> so as far as actual 'probability to find hash', tn diff of 373 (meaning, 373 times more difficult than at 1.0), is actually only 186 times more difficult than mainnet 1.0
908 2011-01-16 09:22:38 <nanotube> because mainnet 1.0 is twice as hard as testnet 1.0
909 2011-01-16 09:22:51 <nanotube> ... hope that makes sense. :)
910 2011-01-16 09:23:54 <nanotube> but anyway, the final takeaway, is that if you want to know the actual avg time between blocks... use "bc,calcd 150000 186.x" :)
911 2011-01-16 09:25:28 <lfm> seems like it would make it less useful as a test [platform to have different rules
912 2011-01-16 09:27:53 <nanotube> not sure why min diff was chosen to be different.. probably just to make it easier to generate blocks for testing. (which, now that artforz has driven the difficulty up, is now a moot point :)
913 2011-01-16 09:28:01 <nanotube> )
914 2011-01-16 09:28:26 <lfm> well it is still easiuer
915 2011-01-16 09:28:48 <nanotube> yes
916 2011-01-16 09:28:57 <nanotube> twice as easy, in fact. ;)
917 2011-01-16 09:29:39 <lfm> no I mean even 373 is easier than 18000
918 2011-01-16 09:30:21 <nanotube> ah yes indeed. and in terms of 'real difficulty', testnet is about 100 times easier than mainnet currently.
919 2011-01-16 09:30:34 <nanotube> which is still plenty difficult, for a cpu. :)
920 2011-01-16 09:32:50 <joe_1> theymos is your bitcoin address checker at http://www.bitcoin.org/smf/index.php?topic=1844.0 valid and working?
921 2011-01-16 09:34:00 <lfm> I think testnet could just be thrown out and reset and restarted
922 2011-01-16 09:38:09 altamic has joined
923 2011-01-16 09:41:29 altamic has quit (Client Quit)
924 2011-01-16 09:43:12 <mrb_> who accidentally left his GPUs running on testnet again?
925 2011-01-16 09:48:56 <joe_1> does a bitcoin address contain 160 bits?
926 2011-01-16 09:50:43 <lfm> mrb_ why you think it was accident?
927 2011-01-16 09:54:52 ducki2p has joined
928 2011-01-16 09:56:28 james_ has quit (Ping timeout: 265 seconds)
929 2011-01-16 09:56:56 <joe_1> the i2p community wants a couple minor changes to the network protocol code to get it to talk to i2p
930 2011-01-16 09:57:46 <ducki2p> thats a bit of a harsh statement :)
931 2011-01-16 09:58:35 <ducki2p> at this point I am exploring what would need to be done to get BC to work on I2P
932 2011-01-16 09:59:13 <lfm> joe_1 so would it still be able to talk to old bitcoin versions?
933 2011-01-16 09:59:54 <joe_1> no --- good point. the i2p version would only talk to other i2p users
934 2011-01-16 10:00:08 <joe_1> so there would need to be an option to talk to both i2p and non-i2p users, so that the network will have these "bridges" to the real network
935 2011-01-16 10:00:20 <xelister> what kind of latences can bitcoin sustain to work properly?
936 2011-01-16 10:00:38 <xelister> is 30 minute lag in posting the transaction order ok?
937 2011-01-16 10:00:40 <lfm> joe_1 so you would also need Id think some sort of gaeway then to bridge the two types
938 2011-01-16 10:01:09 <xelister> are transaction orders tied to anything, like time or last block, or are they just untied messages (untill incorportated in some block)?
939 2011-01-16 10:01:42 <lfm> xelister the latter I think. no timestamp in txns
940 2011-01-16 10:02:23 <lfm> xelister so long as there are no delays propogating "solved" blocks
941 2011-01-16 10:02:56 <joe_1> 30 minute lag is probably unacceptable for miners, but users willing to wait 2 hours for 1 confirmation on sent and received transactions will be OK
942 2011-01-16 10:03:43 <lfm> well users would still not like extra delays but the system would still work I think
943 2011-01-16 10:04:08 <joe_1> of course i2p does not have 30 minute lag, it has something like 45 second lag
944 2011-01-16 10:04:41 <lfm> even 45 second lag would be bad for blcok propogation
945 2011-01-16 10:04:56 <lfm> block
946 2011-01-16 10:05:01 <xelister> actually i2p can be the fastest afair
947 2011-01-16 10:05:08 <xelister> you can configure chain length
948 2011-01-16 10:06:30 <xelister> tunnel
949 2011-01-16 10:06:36 <lfm> ya I think youd want to keep chains quite short for bitcoin
950 2011-01-16 10:06:40 <ducki2p> joe_1: most of my connections have a RTT of 1-2 seconds
951 2011-01-16 10:06:43 <ducki2p> (that is with 2 hops)
952 2011-01-16 10:07:20 <lfm> 2 seconds should be ok for bitcoin
953 2011-01-16 10:07:31 <joe_1> it would be bad for block propagation but another question is, would the addition of high latency peers decrease the efficiency of existing mining on the low latency network that already exists?
954 2011-01-16 10:08:27 <xelister> higher latency for miners just mean that more of the generated blocks would turn out to be chain branches that are rejected later
955 2011-01-16 10:08:31 <lfm> if a lot of the node were on high l;atency you would have more often races and blocks rejected
956 2011-01-16 10:09:35 <xelister> ArtForz: would you please put say 4, then 2, then 1 5970 on testnet to "fix it" (get diff back to 20 usable for testing ;) ?
957 2011-01-16 10:09:38 <joe_1> also, the high-latency net could build chains fast and reject chains that the low latency net had been working on, if the bridge went out for a few minutes then came back online to deliver a longer chain.
958 2011-01-16 10:09:50 <lfm> miners would probably prefer to be on direct connections to minimize chances of their blocks getting rejkected
959 2011-01-16 10:12:00 <lfm> new mined blocks are already quite annonymous on the existing bitcoin net so I doubt there would be much call for extra measues
960 2011-01-16 10:12:43 <lfm> measures
961 2011-01-16 10:14:24 <xelister> lfm: bitcoin is quite tracable
962 2011-01-16 10:14:59 <lfm> ok tell me who won the last three blocks?
963 2011-01-16 10:15:14 <nanotube> xelister: it only takes 22 min to gen a block on testnet with a 5970, at current difficulty. not a big deal for anyone with a decent gpu.
964 2011-01-16 10:15:26 <xelister> lfm: give me 10.000 BTC and I will
965 2011-01-16 10:15:41 <xelister> .. and I will e ready to respond to such question for next blocks
966 2011-01-16 10:15:44 <xelister> *will be
967 2011-01-16 10:16:07 <xelister> remember that 10.000 usd is not even 1% of LANUNCH MONEY for irs etc
968 2011-01-16 10:16:22 <xelister> or [other] criminals that would like to trace your activity
969 2011-01-16 10:16:28 <lfm> the info is not worth 10 btc to me
970 2011-01-16 10:16:35 <xelister> yeah for me neither
971 2011-01-16 10:16:40 <xelister> but it may be for someone in the future
972 2011-01-16 10:16:56 <xelister> eg what if IRS says - every bitcoiner, pay income tax for btcs earned/mined
973 2011-01-16 10:17:01 <xelister> or like "bitcoin is now illegal"
974 2011-01-16 10:17:35 <lfm> so how you gonna tell if a btc is mined in usa or not?
975 2011-01-16 10:17:57 <xelister> for sums like 10k USD bitcoin can be traced
976 2011-01-16 10:18:09 <xelister> its not something impossible
977 2011-01-16 10:18:22 <xelister> even average small time hacker/badguy could do it
978 2011-01-16 10:18:55 noagendamarket has joined
979 2011-01-16 10:19:28 <lfm> easy to say
980 2011-01-16 10:19:44 noagendamarket has quit (Changing host)
981 2011-01-16 10:19:44 noagendamarket has joined
982 2011-01-16 10:19:54 <xelister> you say this is hard to do for 10 kbtc?
983 2011-01-16 10:24:58 <lfm> i yhink youd be lucky to figure out where the next block came from and if you were not prepared you could not figure out where the last block came from no matter how skilled you were
984 2011-01-16 10:41:30 <niekie> Hm. Amazon offers GPU instances. This is new to me.
985 2011-01-16 10:42:13 <niekie> I guess paying $2.10/hour for them is not currently worth it though.
986 2011-01-16 10:43:34 <niekie> Specs are 'Cluster GPU Quadruple Extra Large 22 GB memory, 33.5 EC2 Compute Units, 2 x NVIDIA Tesla âFermiâ M2050 GPUs, 1690 GB of local instance storage, 64-bit platform, 10 Gigabit Ethernet'
987 2011-01-16 10:47:45 <mrb_> lfm: accidental, because I fail to see why anybody would want to put so much resources mining on testnet
988 2011-01-16 10:48:24 <mrb_> niekie: this has been discussed in the past. if you do the math, ec2 instances are not worth it.
989 2011-01-16 10:48:48 <mrb_> both gpu and cpu instances.
990 2011-01-16 10:49:22 <lfm> mrb_ maybe it was a test to see how badly they could break testnet
991 2011-01-16 10:50:32 <mrb_> I must admit I once mined 20 blocks in < 5 minutes on testnet
992 2011-01-16 10:50:49 <nevezen> is testnet a simulation?
993 2011-01-16 10:51:14 <niekie> Yes and no.
994 2011-01-16 10:51:35 <niekie> It's a real network, and real clients are looking for blocks on it. But it's used for testing purposes.
995 2011-01-16 10:51:36 <lfm> ya, I think it only took 20 min for art to mine 2016 blocks
996 2011-01-16 10:51:48 <mrb_> it's another instance of the peer-to-peer bitcoin network that operate on another port
997 2011-01-16 10:52:58 <mrb_> by convention people declare it as the "test network" but it's really no different than the real network other than people does not recognize coins on testnet as having any value
998 2011-01-16 10:55:46 <lfm> and since testnet is not valued it could theoreticlly be reset back to zero
999 2011-01-16 10:56:48 <mrb_> yeah, a software update would have to be published so that all bitcoin nodes cooperate and agree to reset it at a determined time.
1000 2011-01-16 10:57:43 <mrb_> but really, just not doing anything would cause it to reset itself anyway. since it's gonna take a while to mine the next 2016 blocks
1001 2011-01-16 10:58:56 <lfm> yes once the 2016 blocks are found at the hi difficulty then the difficulty will drop back to a more reasonable level but it could take a long time
1002 2011-01-16 10:59:37 <mrb_> if you try to mine the blocks somewhate quickly, then the difficulty will not decrease much
1003 2011-01-16 11:00:01 <mrb_> the longer it takes to mine the block, the bigger the difficulty drop will be
1004 2011-01-16 11:00:13 <lfm> yup
1005 2011-01-16 11:02:16 cdecker has quit (Remote host closed the connection)
1006 2011-01-16 11:03:00 <slush> current testnet difficulty is pretty high. Is there any plan to reset it?
1007 2011-01-16 11:03:15 cdecker has joined
1008 2011-01-16 11:03:33 <lfm> no real plan but there are a few suggestions to reset it
1009 2011-01-16 11:03:38 <slush> I mean, testnet is mainly for testing, but currently it takes too long to test finding a block, even on decent gpu
1010 2011-01-16 11:03:46 <slush> I definitely vote for it
1011 2011-01-16 11:04:20 <lfm> the question is, is it a design flaw revealed that it needs to be reset?
1012 2011-01-16 11:04:23 <xelister> we could ask ArtForz to put ther 4, then 2, then 1 gpu, and it would reset it reasonably quick
1013 2011-01-16 11:04:25 <slush> For example, I need tens or even hundreds of solved blocks for full pool testing. With current difficulty, it will take ages on 5970
1014 2011-01-16 11:04:27 <xelister> lfm: yes
1015 2011-01-16 11:04:46 <xelister> lfm: the difficulty should be reevaluated not only each 2016 blocks, but also each 48 hours
1016 2011-01-16 11:05:35 <slush> I can kindly ask ArtForz to not test cartel mode too often ;), but I doubt it will have an effect
1017 2011-01-16 11:05:50 <slush> + some people began to speculate on testnet value
1018 2011-01-16 11:05:52 <xelister> well we wrecked testnet :|
1019 2011-01-16 11:05:57 <slush> which put more miners here
1020 2011-01-16 11:07:38 cdecker has quit (Read error: Connection reset by peer)
1021 2011-01-16 11:10:02 <lfm> so if its a design flaw will we have the same problem on the main net if a big miner drops out and the block rate slows to a crawl?
1022 2011-01-16 11:11:49 johnyh has joined
1023 2011-01-16 11:12:39 <johnyh> btw, anyone have testnet bitcoins lying around? over 10,000 best
1024 2011-01-16 11:13:03 <lfm> I think art has some
1025 2011-01-16 11:14:11 <joe_1> im getting scared
1026 2011-01-16 11:14:35 <lfm> why joe?
1027 2011-01-16 11:14:41 <joe_1> what if the test net takes over
1028 2011-01-16 11:14:52 <joe_1> all my wealth is in main net coins
1029 2011-01-16 11:15:40 <lfm> I think people are silly to value testnet. it coulkd be reset tommorow and all the old balances would be dead
1030 2011-01-16 11:16:06 <joe_1> only if people update...
1031 2011-01-16 11:17:38 <slush> maybe automatic chain reset after...20 000 blocks... can be good idea for testnet?
1032 2011-01-16 11:18:34 <joe_1> actually i dont think it could ever take over,
1033 2011-01-16 11:18:47 <joe_1> bitcoin has value by virtue that it is the first
1034 2011-01-16 11:19:05 scibotic has joined
1035 2011-01-16 11:19:05 scibotic has quit (Changing host)
1036 2011-01-16 11:19:05 scibotic has joined
1037 2011-01-16 11:19:17 <lfm> i think its best to be manual reset. if testnet gets too many special cases it will lose value as a testing ground
1038 2011-01-16 11:19:49 <slush> who can decide testnet reset?
1039 2011-01-16 11:21:54 <lfm> I think satoshi or someone just has to make new block zero and then get people to stop uing the old testnet chain. might need new version, not sure really
1040 2011-01-16 11:21:55 <joe_1> theoretically, nobody. practically, the bitcoin client developers
1041 2011-01-16 11:23:15 <slush> I hope they don't release version with block reset with forgotten condition to testnet ;)
1042 2011-01-16 11:24:32 <lfm> "forgotten condition"?
1043 2011-01-16 11:24:32 <joe_1> well it's resilient enough that that wouldn't be too big of a problem. it would only take one person who didn't upgrade, to re-populate the network once they realized the mistake and released the fixed version
1044 2011-01-16 11:25:20 <lfm> joe if its a different block zero then the old chain wouldnt work any more
1045 2011-01-16 11:25:43 <joe_1> i'm saying after they fixed it back.
1046 2011-01-16 11:26:28 <lfm> huh?
1047 2011-01-16 11:26:42 <joe_1> slush was saying i hope they don't accidentally reset the main chain.
1048 2011-01-16 11:26:46 <slush> lfm: "forget to make condition which perform block reset only on testnet"
1049 2011-01-16 11:27:29 <lfm> oh, ya I guess we dont really wanna reset the main net now
1050 2011-01-16 11:27:44 <lfm> even if it is still beta
1051 2011-01-16 11:28:25 <joe_1> is it a coincidence that satoshi used the biblical term "genesis" to refer to the first block?
1052 2011-01-16 11:29:32 Zarutian has joined
1053 2011-01-16 11:29:50 <lfm> um, no, it is the beginning, thats what genisis means
1054 2011-01-16 11:29:56 <joe_1> meaning that God himself handed down the hash to the first block, so the main bitcoin chain is the "in god we trust" bitcoin currency.
1055 2011-01-16 11:30:28 <lfm> not neccesarily anything to do with god
1056 2011-01-16 11:30:50 <joe_1> yeah, but i think that's what we should tell other people, so nobody gets any wise ideas about starting new chains.
1057 2011-01-16 11:31:23 <lfm> genisis is from the same root as generate
1058 2011-01-16 11:43:06 Tritonio_GR has quit (Read error: Connection reset by peer)
1059 2011-01-16 11:46:13 <noagendamarket> joe satoshi himself says people should start new block chains to setrup competing currencies
1060 2011-01-16 11:47:06 <joe_1> i dont think that's right
1061 2011-01-16 11:47:37 <joe_1> first, it splits up valuable CPU power needed to fend off attacks from governments
1062 2011-01-16 11:48:11 <joe_1> second, if people see that currencies are coming and going then none of them will ever have value, since value is based on expectation of future value.
1063 2011-01-16 11:51:38 <joe_1> when an ordinary person looks at a candidate currency and asks themselves if it's valueable, they're asking what makes it forever and what makes it self evident, to the rest of the world, that it's a currency?
1064 2011-01-16 11:52:30 <joe_1> for bitcoin, it's the first application of the proof-of-work chain cryptocurrency.
1065 2011-01-16 11:54:35 <joe_1> if i wake up tomorrow and start my own chain, someone asks me why is it a currency? "it's a proof-of-work chain based cryptocurrency that I created yesterday"
1066 2011-01-16 11:57:10 <joe_1> i could theoretically imagine 2 bitcoin chains, but nothing more than 2. The reason i say that, is that we already accept gold and silver-- 2 metals-- as currency, when we should really only have 1 metal currency.
1067 2011-01-16 11:59:10 <joe_1> maybe i should start a thread asking satoshi about this. or look for any threads already on there about this subject
1068 2011-01-16 12:11:18 <lfm> if you stand behind your currency it will have value. if you make standing offer to trade it for another currency of anything of value, gold, cookies, your labor, if you back it up then it will have value
1069 2011-01-16 12:19:06 hundfred has quit (Ping timeout: 260 seconds)
1070 2011-01-16 12:21:06 EvanR has quit (Ping timeout: 246 seconds)
1071 2011-01-16 12:21:55 <T_X> slush: thanks for the awesome extra statstics :). that somehow gives a good feeling for having selected a %-donation value :)
1072 2011-01-16 12:22:20 <slush> thanks :)
1073 2011-01-16 12:23:55 <slush> that's my first graphs in javascript, it was real pain
1074 2011-01-16 12:26:12 <T_X> :D. but looks nice :)
1075 2011-01-16 12:26:58 altamic has joined
1076 2011-01-16 12:31:16 [Noodles] has quit (Ping timeout: 246 seconds)
1077 2011-01-16 12:31:47 Diablo-D3 has joined
1078 2011-01-16 12:37:45 [Noodles] has joined
1079 2011-01-16 12:40:55 <T_X> on mtgox, only trades on the market cost 0.65%, right? I mean, sending btc to the mtgox account and withdrawing them again does not have a transaction fee, right?
1080 2011-01-16 12:41:07 <slush> correct
1081 2011-01-16 12:45:36 <T_X> kk. actually what I would find useful is the possiblity to share an mtgox wallet over multiple accounts for one thing and having the possibility for reviewing transactions from and to that wallet publically.
1082 2011-01-16 12:47:22 <slush> is it related to trading somehow? This feature would be more useful on mybitcoin, which is more 'bank' thank exchange
1083 2011-01-16 12:47:37 <T_X> hmm, okay
1084 2011-01-16 12:47:54 <sipa> T_X: sending them euros does incur a 2% fee as well
1085 2011-01-16 12:48:17 <T_X> no, there's no plan to send EUR
1086 2011-01-16 12:48:33 <T_X> just to have a deposit which is not controlled by a single person
1087 2011-01-16 12:48:52 <sipa> ?
1088 2011-01-16 12:49:04 <slush> both are talking about different things :)
1089 2011-01-16 12:49:22 <T_X> looks like :D
1090 2011-01-16 12:49:26 <slush> sipa talks about mtgox fees
1091 2011-01-16 12:49:34 <slush> not about shared wallets
1092 2011-01-16 12:49:45 <sipa> oh, i hadn't read everything afterwards yet :)
1093 2011-01-16 12:51:01 <T_X> and maybe a certain amount of elected people within this community were having the right to make transactions out of this account. however, it shall be visible for everyone else in the community who did any of those transactions
1094 2011-01-16 13:00:32 devon_hillard has joined
1095 2011-01-16 13:03:50 <T_X> mybitcoin does not have such nice donation buttons as mtgox has, has it?
1096 2011-01-16 13:13:49 <Diablo-D3> http://www.youtube.com/watch?v=cTQNwMxqM3E
1097 2011-01-16 13:13:51 <Diablo-D3> what
1098 2011-01-16 13:50:08 <npouillard> I'm looking for info about the use of testnet, any pointers?
1099 2011-01-16 13:51:31 * npouillard found the page on the wiki
1100 2011-01-16 13:52:11 <sipa> what do you need to know about it?
1101 2011-01-16 13:55:09 <sipa> how often does https://smsz.net/btcStats/bitcoin.kml get updated?
1102 2011-01-16 13:57:49 <npouillard> http://www.bitcoin.org/smf/index.php?PHPSESSID=fe6b3d0c1739bb2914976ffe1209fbb9&topic=363.0;all <== this thread looks quite outdated no?
1103 2011-01-16 14:00:29 hundfred has joined
1104 2011-01-16 14:10:46 RazielZ has quit ()
1105 2011-01-16 14:11:36 noagendamarket has quit (Ping timeout: 255 seconds)
1106 2011-01-16 14:19:51 altamic has quit (Read error: Connection reset by peer)
1107 2011-01-16 14:20:12 altamic has joined
1108 2011-01-16 14:29:28 scibotic has quit (Quit: Leaving)
1109 2011-01-16 14:33:04 rain has joined
1110 2011-01-16 14:33:31 rain is now known as Guest59002
1111 2011-01-16 14:40:19 Guest59002 has quit (Quit: Page closed)
1112 2011-01-16 14:43:55 altamic has quit (Ping timeout: 240 seconds)
1113 2011-01-16 14:45:16 EvanR has joined
1114 2011-01-16 14:51:40 scibotic has joined
1115 2011-01-16 14:52:13 EvanR has quit (Ping timeout: 240 seconds)
1116 2011-01-16 14:55:42 EvanR has joined
1117 2011-01-16 15:01:54 AAA_awright has quit (Ping timeout: 264 seconds)
1118 2011-01-16 15:04:12 AAA_awright has joined
1119 2011-01-16 15:07:17 AAA_awright_ has joined
1120 2011-01-16 15:08:51 AAA_awright has quit (Ping timeout: 240 seconds)
1121 2011-01-16 15:09:41 AAA_awright has joined
1122 2011-01-16 15:12:06 AAA_awright_ has quit (Ping timeout: 264 seconds)
1123 2011-01-16 15:12:37 EvanR has quit (Ping timeout: 246 seconds)
1124 2011-01-16 15:18:09 <sipa> ;;bc,calc 600000
1125 2011-01-16 15:18:10 <gribble> The average time to generate a block at 600000 Khps, given current difficulty of 18437.64439217 , is 1 day, 12 hours, 39 minutes, and 41 seconds
1126 2011-01-16 15:18:12 <sipa> ;;bc,calc 1200000
1127 2011-01-16 15:18:13 <gribble> The average time to generate a block at 1200000 Khps, given current difficulty of 18437.64439217 , is 18 hours, 19 minutes, and 50 seconds
1128 2011-01-16 15:22:05 EvanR has joined
1129 2011-01-16 15:27:15 EvanR has quit (Ping timeout: 255 seconds)
1130 2011-01-16 15:28:58 Tritonio has joined
1131 2011-01-16 15:29:11 theymos has joined
1132 2011-01-16 15:29:28 <theymos> joe_1: My address checker is working.
1133 2011-01-16 15:33:59 <niekie> ;;bc
1134 2011-01-16 15:33:59 <gribble> Error: "bc" is not a valid command.
1135 2011-01-16 15:34:02 <niekie> ;;help
1136 2011-01-16 15:34:03 <gribble> The bot responds when you start a line with the ! character. A good starting point for exploring the bot is the !facts command. You can also visit the bot's website for a list of help topics and documentation: http://gribble.sourceforge.net/
1137 2011-01-16 15:35:16 <niekie> ;;bc,blocks
1138 2011-01-16 15:35:17 <gribble> 102998
1139 2011-01-16 15:35:27 <niekie> ;;bc,stats
1140 2011-01-16 15:35:28 <gribble> Current Blocks: 102998 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 1833 blocks | Next Difficulty In About: 1 week, 3 days, 10 hours, 30 minutes, and 36 seconds | Next Difficulty Estimate: 22596.89665695
1141 2011-01-16 15:35:32 <tcatm> hey theymos. did the coin arrive?
1142 2011-01-16 15:35:48 <theymos> Yes. It looks pretty cool! Thanks.
1143 2011-01-16 15:36:04 <tcatm> cool :)
1144 2011-01-16 15:39:16 skeledrew1 has quit (Quit: Instantbird 0.3a1pre)
1145 2011-01-16 15:40:13 <tcatm> theymos: can blockexplorer be modified to include TX that aren't in blocks yet? I have a python script that'll connect to a bitcoin node and output them as soon as they arrive
1146 2011-01-16 15:40:24 skeledrew has joined
1147 2011-01-16 15:44:08 <theymos> Yes. Can you post the script? I might not implement it for a while, though -- there are some other features I want to do first.
1148 2011-01-16 15:45:24 altamic has joined
1149 2011-01-16 15:46:05 <tcatm> Might as well use it to have realtime update for blocks ;)
1150 2011-01-16 15:46:45 RazielZ has joined
1151 2011-01-16 15:47:28 ducki2p has quit (Remote host closed the connection)
1152 2011-01-16 15:47:38 <tcatm> save this as bitcoin.py http://pastebin.com/JZM6Hemt
1153 2011-01-16 15:48:01 ducki2p has joined
1154 2011-01-16 15:48:51 <tcatm> this will dump blocks and TX as soon as it sees them http://pastebin.com/LsSebetx
1155 2011-01-16 15:50:16 Zarutian has quit (Quit: Zarutian)
1156 2011-01-16 15:51:24 <theymos> Thanks. Probably I'll just use block messages from this to trigger block updates from getblock, since I already have getblock parsing coded.
1157 2011-01-16 15:51:49 Zarutian has joined
1158 2011-01-16 15:53:30 EvanR has joined
1159 2011-01-16 15:55:42 Cusipzzz has joined
1160 2011-01-16 16:08:35 slush has quit (Ping timeout: 240 seconds)
1161 2011-01-16 16:12:54 sabalaba has joined
1162 2011-01-16 16:14:42 EvanR has quit (Ping timeout: 260 seconds)
1163 2011-01-16 16:17:44 sabalaba has quit (Read error: Operation timed out)
1164 2011-01-16 16:19:17 AAA_awright_ has joined
1165 2011-01-16 16:19:35 EvanR has joined
1166 2011-01-16 16:22:02 AAA_awright has quit (Ping timeout: 276 seconds)
1167 2011-01-16 16:24:36 EvanR has quit (Ping timeout: 272 seconds)
1168 2011-01-16 16:26:38 <nanotube> <lfm> [02:51:34] and since testnet is not valued it could theoreticlly be reset back to zero <-- i beg to differ about the 'not valued' bit. there are several outstanding bid offers for testnet btc on OTC, so clearly testnet does have value.
1169 2011-01-16 16:26:51 <nanotube> resetting it willy-nilly is not a good idea, i think.
1170 2011-01-16 16:27:29 <sipa> it has value if someone wants to pay for it
1171 2011-01-16 16:27:41 <sipa> the question is whether it is useful that it has value
1172 2011-01-16 16:28:14 <ArtForz> oh, resetting testnet is easy
1173 2011-01-16 16:28:39 hundfred has quit (Read error: Operation timed out)
1174 2011-01-16 16:28:47 <ArtForz> <12h with something like my cluster
1175 2011-01-16 16:29:03 <tcatm> how would you reset it? fake nTime?
1176 2011-01-16 16:29:07 <ArtForz> yep
1177 2011-01-16 16:29:23 <sipa> any sentence which contains "my cluster", said by ArtForz, is by definition not "easy"
1178 2011-01-16 16:29:51 <ArtForz> basically restart from genesis, play with block times to first drive difficulty way up, then let is settle back down
1179 2011-01-16 16:30:09 <ArtForz> new chain has higher total work than current chain = takes over
1180 2011-01-16 16:30:35 <tcatm> sounds like something worth to do
1181 2011-01-16 16:30:54 <ArtForz> and total testnet work only represents about 12h @ 16Gh/s
1182 2011-01-16 16:31:12 <ArtForz> unless my total work calculation is off for testnet
1183 2011-01-16 16:31:30 * sipa wonders whether a tool which generates a custom genesis block, port, magic numbers, and dumps those into a config file, with a patch for bitcoind to load settings from that config file is useful
1184 2011-01-16 16:31:39 <sipa> essentially allowing everyone to create their own testnet
1185 2011-01-16 16:32:08 <sipa> -test
1186 2011-01-16 16:32:44 <ArtForz> lets see if the math works out
1187 2011-01-16 16:33:44 <tcatm> would even be interesting to see how clients react to all coinbases becoming invalid
1188 2011-01-16 16:34:05 <sipa> ?
1189 2011-01-16 16:34:33 <tcatm> sipa: to create your own testnet only connect to your nodes
1190 2011-01-16 16:34:42 <ArtForz> we have about 16M seconds to work with
1191 2011-01-16 16:36:15 <ArtForz> we can stuff about 5 blocks in a second thanks to the median-of-prev-11-plus-1 rule
1192 2011-01-16 16:37:21 <ArtForz> = you can easily put 2016 blocks in 600 seconds to drive difficulty up without wasting too much time
1193 2011-01-16 16:38:07 INEEDMONEY has joined
1194 2011-01-16 16:38:47 <tcatm> => difficulty of 1008?
1195 2011-01-16 16:38:54 <ArtForz> ?
1196 2011-01-16 16:39:12 <ArtForz> testnet is also limited by /4 *4 per period
1197 2011-01-16 16:39:26 <tcatm> right, forgot about that
1198 2011-01-16 16:40:25 slush has joined
1199 2011-01-16 16:47:39 AAA_awright has joined
1200 2011-01-16 16:48:25 <INEEDMONEY> so how's the number-crunching going?
1201 2011-01-16 16:48:54 <ArtForz> yep, should work
1202 2011-01-16 16:49:28 AAA_awright_ has quit (Ping timeout: 240 seconds)
1203 2011-01-16 16:49:29 devon_hillard has quit (Read error: Connection reset by peer)
1204 2011-01-16 16:51:18 <ArtForz> 4 * 2016 blocks in 2400s -> get difficulty up to 32, 2016 block in 1209600s, 3* 2016 block in 4838400s each -> final difficulty is 0.5
1205 2011-01-16 16:52:03 <ArtForz> total work in that chain is 732738600566784, testnet chain is 674195433361632 or so
1206 2011-01-16 16:52:30 <ArtForz> that chain needs 15727200s after genesis
1207 2011-01-16 16:53:45 <ArtForz> and yeah, it's < 2/3 the # of blocks of the current testnet chain, yet has higher sum-of-work
1208 2011-01-16 16:54:40 <ArtForz> 16128 blocks
1209 2011-01-16 16:55:11 <nanotube> wouldn't it be cheaper/faster to just let the diff on the current testnet come down?
1210 2011-01-16 16:55:32 <ArtForz> well... that will take a while
1211 2011-01-16 16:55:51 <molec> there's something I don't understand about cooperative mining: what keeps a miner from keeping the block for himself in case he finds one instead of submitting it to the cooperative?
1212 2011-01-16 16:56:11 <nanotube> ;;bc,wiki pooled mining
1213 2011-01-16 16:56:12 <gribble> https://en.bitcoin.it/wiki/Pooled_mining | Dec 30, 2010 ... Pooled mining is an approach where multiple generating clients contribute to the generation of a block, and then split the block reward ...
1214 2011-01-16 16:56:15 <nanotube> molec: --^
1215 2011-01-16 16:56:26 <molec> nanotube, thx
1216 2011-01-16 16:56:27 <ArtForz> well... easier... yes
1217 2011-01-16 16:56:43 <ArtForz> we just need a bit more hashpower on testnet
1218 2011-01-16 16:56:53 <nanotube> molec: in short, blocks that clients are working on contain the tx giving btc to the pool. 'keeping it' is useless.
1219 2011-01-16 16:56:56 <ArtForz> so we get ~2400s/block for maximum slowing
1220 2011-01-16 16:57:23 <tcatm> how long would it take?
1221 2011-01-16 16:57:43 <tcatm> ;;bc,calc 300000 42
1222 2011-01-16 16:57:43 <gribble> Error: invalid syntax (<string>, line 1)
1223 2011-01-16 16:57:49 <tcatm> ;;bc,calcd 300000 42
1224 2011-01-16 16:57:49 <gribble> The average time to generate a block at 300000 Khps, given the supplied difficulty of 42, is 10 minutes and 1 second
1225 2011-01-16 16:57:51 <ArtForz> 8 weeks to get to the next quartering
1226 2011-01-16 16:58:36 <tcatm> instead of 1-2 days when restarting the chain
1227 2011-01-16 16:58:42 <ArtForz> yep
1228 2011-01-16 16:58:53 <ArtForz> and then we're still at 42 diff ...
1229 2011-01-16 16:59:06 <ArtForz> = another 8 weeks to get to 10.something
1230 2011-01-16 16:59:36 <ArtForz> restarting the chain somehow seems *easier*
1231 2011-01-16 16:59:40 <tcatm> so only nTime needs to be carefully adjusted (i.e. ++ every 5s)?
1232 2011-01-16 16:59:49 <ArtForz> well... kinda
1233 2011-01-16 16:59:53 <ArtForz> but yeah
1234 2011-01-16 17:00:12 <ArtForz> basically ntime +=1 every 5 blocks for the first 4*2016 blocks
1235 2011-01-16 17:00:25 <ArtForz> then ntime += 600 for 2016 blocks
1236 2011-01-16 17:00:33 <ArtForz> and ntime += 2400 for 3*2016 blocks
1237 2011-01-16 17:00:51 * tcatm hacks client
1238 2011-01-16 17:01:31 redMBA has joined
1239 2011-01-16 17:01:40 <tcatm> ,,bc,calcd 300000 0.5
1240 2011-01-16 17:01:40 <gribble> (bc,calcd <an alias, 2 arguments>) -- Alias for "echo The average time to generate a block at $1 Khps, given the supplied difficulty of $2, is [time elapsed [math calc 1/((2**224-1)/$2*$1*1000/2**256)]]".
1241 2011-01-16 17:01:44 <tcatm> ;;bc,calcd 300000 0.5
1242 2011-01-16 17:01:44 <gribble> The average time to generate a block at 300000 Khps, given the supplied difficulty of 0.5, is 7 seconds
1243 2011-01-16 17:02:00 <ArtForz> would certainly be fun to see what happens when everyones balance goes *poof*
1244 2011-01-16 17:04:01 <nanotube> mm yea.. fun.... O_.
1245 2011-01-16 17:04:21 <ArtForz> well, it's called TESTnet for a reason
1246 2011-01-16 17:04:55 <nanotube> heh true
1247 2011-01-16 17:05:32 <nanotube> so how long would it take for someone with a few Thps to do the same to the mainnet?
1248 2011-01-16 17:06:18 <ArtForz> mainnet work is ... 657980749827397368
1249 2011-01-16 17:06:19 scibotic has quit (Read error: Connection reset by peer)
1250 2011-01-16 17:06:30 scibotic has joined
1251 2011-01-16 17:06:42 <ArtForz> 182h @ 1Thps
1252 2011-01-16 17:07:17 <ArtForz> + a few more h for the work mainnet is doing in those 182h
1253 2011-01-16 17:07:31 <ArtForz> so... about 200h @ 1Thps
1254 2011-01-16 17:07:38 cdecker has joined
1255 2011-01-16 17:07:44 <tcatm> what's the easiert way to calculate network work?
1256 2011-01-16 17:08:22 <ArtForz> errr
1257 2011-01-16 17:08:31 <nanotube> ArtForz: mmm not so encouraging. let's hope that it'll be a while before adversaries with Thps's to throw around takes note. :)
1258 2011-01-16 17:09:02 <ArtForz> mainline uses sum(2**256 / (block[i].target + 1))
1259 2011-01-16 17:09:06 <kiba> gene guy seem to have a strong hatred against multinational corporations
1260 2011-01-16 17:09:32 <ArtForz> and my code does, too
1261 2011-01-16 17:09:58 <ArtForz> it's about the same as sum(block[i].difficulty * 2**32)
1262 2011-01-16 17:10:19 <ArtForz> or simpler: sum(block[i].difficulty) * 2**32
1263 2011-01-16 17:11:05 <ArtForz> that should come out to roughly the same
1264 2011-01-16 17:13:19 <tcatm> so sum... target+1 is the accurate work?
1265 2011-01-16 17:13:27 <ArtForz> yes
1266 2011-01-16 17:13:50 <tcatm> what was the formula to calculate target from nbits again?
1267 2011-01-16 17:14:41 <ArtForz> off the top of my head, (nbits & 0xFFFFFF) << (8 * (((nbits >> 24) & 0xFF) - 3))
1268 2011-01-16 17:15:16 <tcatm> oh wait, getblock outputs diff instead of nbits :/
1269 2011-01-16 17:16:29 <ArtForz> submit a feature request for BBE? ;)
1270 2011-01-16 17:16:50 <theymos> What feature?
1271 2011-01-16 17:16:54 <tcatm> bitcoincharts keeps the blockchain in a sqltable
1272 2011-01-16 17:17:00 <tcatm> mysql> select sum(difficulty)*pow(2,32) from bitcoin_block;
1273 2011-01-16 17:17:05 <tcatm> 6.58050765980394e+17
1274 2011-01-16 17:17:06 <ArtForz> total chain work for blocks
1275 2011-01-16 17:17:38 <ArtForz> @ block 103010, it should be 658059940115424900
1276 2011-01-16 17:18:19 <ArtForz> error comes from 1.0 diff != 2**224
1277 2011-01-16 17:18:41 <tcatm> can that be compensated?
1278 2011-01-16 17:18:46 <ArtForz> not easily
1279 2011-01-16 17:19:11 <ArtForz> sec
1280 2011-01-16 17:20:26 <tcatm> theymos: can you do pow(2,256)/sum((nbits & 0xFFFFFF) << (8 * (((nbits >> 24) & 0xFF) - 3)) + 1) in your block database?
1281 2011-01-16 17:20:27 <ArtForz> try using 4295032833 instead of 2**32
1282 2011-01-16 17:20:47 <tcatm> 6.58060807191439e+17
1283 2011-01-16 17:20:59 <ArtForz> thats a lot closer
1284 2011-01-16 17:21:02 <tcatm> yep
1285 2011-01-16 17:21:03 <tcatm> thanks
1286 2011-01-16 17:21:27 <ArtForz> it's 4295032833.0000153... really (0x100010001.000100010001...)
1287 2011-01-16 17:22:07 <ArtForz> = 2**256 / ((0xFFFF << 208) + 1)
1288 2011-01-16 17:22:52 <ArtForz> you're within 0.001% or something anyways
1289 2011-01-16 17:22:56 <theymos> tcatm: Not easily.
1290 2011-01-16 17:22:59 <ArtForz> good enough for a rough estimate ;)
1291 2011-01-16 17:23:33 sabalaba has joined
1292 2011-01-16 17:23:34 <hacim> is memory or CPU resources ever a concern WRT to running a GPU miner?
1293 2011-01-16 17:23:46 <ArtForz> nope
1294 2011-01-16 17:25:07 altamic has quit (Quit: altamic)
1295 2011-01-16 17:25:28 <ArtForz> tcatm: your calculation is 0.0001318% high ... using floating point... so who cares ;)
1296 2011-01-16 17:25:49 <nanotube> so... i guess this basically means that if we want to "prevent blockchain competition" from any other competing chains (and thereby 'preserve the mainnet chain value'), we could just 'hunt down' any startup new chains by throwing a bunch of our gen power at them, and 'resetting' their net. heh...
1297 2011-01-16 17:26:01 <ArtForz> you can't do it with mainnet
1298 2011-01-16 17:26:07 <nanotube> yes, hardcoded checkpoints
1299 2011-01-16 17:26:12 <ArtForz> yep
1300 2011-01-16 17:26:15 <nanotube> but you can do it starting from latest checkpoint
1301 2011-01-16 17:26:20 <ArtForz> yep
1302 2011-01-16 17:26:23 <tcatm> ArtForz: and the unit is? hashes?
1303 2011-01-16 17:26:28 <ArtForz> yes
1304 2011-01-16 17:26:30 <nanotube> e.g., if you did this since block 70k-something... that'd still be pretty havoc-y
1305 2011-01-16 17:26:44 <ArtForz> unit for work is bitcoin block hashes
1306 2011-01-16 17:26:52 <ArtForz> = about 2*sha256
1307 2011-01-16 17:27:13 <ArtForz> or roughly 1500 core cycles using SSE2
1308 2011-01-16 17:27:15 <hacim> cool, so I can get just the bare minimum for my mining machine
1309 2011-01-16 17:28:18 <ArtForz> = mainnet chain is equivalent to about 987090 peta sse2 ops
1310 2011-01-16 17:28:28 <theymos> If a "reset" happened on the real network, a new version would be released that locked in the real main chain. You'll still have the blocks -- they're not deleted when they become orphans.
1311 2011-01-16 17:28:38 <ArtForz> or about 987 exaops
1312 2011-01-16 17:28:49 <ArtForz> yep
1313 2011-01-16 17:29:51 <ArtForz> so in a while we'll have the equivalent of 1 zetta-sse-ops in the main chain
1314 2011-01-16 17:30:09 TheAncientGoat has quit (Ping timeout: 276 seconds)
1315 2011-01-16 17:30:35 <ArtForz> another 6 orders of magnitude and we run out of SI prefixes :P
1316 2011-01-16 17:30:59 <UukGoblin> hmm, what prevents LRUSD<->USD exchange rate from drifting far off 1.0?
1317 2011-01-16 17:31:59 <ArtForz> wild guess: LR offering premium exchangers close to 1.0 exchange rate for LRUSD<->USD
1318 2011-01-16 17:32:01 <nanotube> UukGoblin: 'official lr exchangers'
1319 2011-01-16 17:32:10 <nanotube> right
1320 2011-01-16 17:32:43 <nanotube> they can buy lr at close to 1.0 (if not exactly 1.0) from lr.
1321 2011-01-16 17:33:18 <nanotube> due to competition, their premiums to others are generally single-digit percents.
1322 2011-01-16 17:33:28 <UukGoblin> ah
1323 2011-01-16 17:33:48 <UukGoblin> so LR /does/ sell LRUSD but only to '(un)official exchangers'
1324 2011-01-16 17:34:07 <nanotube> s/(un)//
1325 2011-01-16 17:34:19 <tcatm> ()official?
1326 2011-01-16 17:34:27 <nanotube> you have to go through an official registration process, to be able to buy lr from them.
1327 2011-01-16 17:34:35 <UukGoblin> "In order to add or remove funds from your LR account you should use one of the independent exchange providers from the list below. Please note that businesses listed on this page are independent and in no way related to Liberty Reserve, and Liberty Reserve does not guarantee any outcome that you may experience with these companies."
1328 2011-01-16 17:34:39 <nanotube> tcatm: haha, ok, s/\(un\)//
1329 2011-01-16 17:35:03 <nanotube> UukGoblin: yes, they're not legally related, but they're 'official' in that they have to register, and lr has to approve
1330 2011-01-16 17:35:24 <UukGoblin> heh
1331 2011-01-16 17:35:26 <UukGoblin> thanks niekie
1332 2011-01-16 17:35:33 <niekie> No problem :)
1333 2011-01-16 17:35:39 <nanotube> woo more ops. :)
1334 2011-01-16 17:35:48 <niekie> Exactly.
1335 2011-01-16 17:35:58 devon_hillard has joined
1336 2011-01-16 17:36:03 * nanotube already had ops, niekie :)
1337 2011-01-16 17:36:14 <nanotube> or ... wait i thought i did. heh
1338 2011-01-16 17:36:18 <niekie> Nope :)
1339 2011-01-16 17:36:25 <nanotube> in that case, tx. :)
1340 2011-01-16 17:41:29 redMBA has quit (Ping timeout: 240 seconds)
1341 2011-01-16 17:42:14 <kiba> the bitcoin job market sucks
1342 2011-01-16 17:42:30 akem has quit (Read error: Connection reset by peer)
1343 2011-01-16 17:43:06 <UukGoblin> kiba, no wonder. paying income tax off bitcoin must be a pain...
1344 2011-01-16 17:44:08 <nanotube> UukGoblin: haha i doubt that that is the reason. :)
1345 2011-01-16 17:44:21 <kiba> so we got the base economy
1346 2011-01-16 17:44:24 <kiba> Hosting
1347 2011-01-16 17:44:32 <kiba> that's like the first order capital goods
1348 2011-01-16 17:44:45 <kiba> and then we got a few sites
1349 2011-01-16 17:45:01 <UukGoblin> yup, I guess it's mostly due to lack of popularity :-)
1350 2011-01-16 17:45:13 <kiba> well
1351 2011-01-16 17:45:25 <kiba> we have a thriving "bounty" system
1352 2011-01-16 17:45:39 <kiba> but nobody cares about getting bitcoin from doing work
1353 2011-01-16 17:45:55 <kiba> I think everybody just buy bitcoin rather than work for it
1354 2011-01-16 17:46:10 doublec has joined
1355 2011-01-16 17:46:14 <tcatm> I got about 3000 BTC from programming/webdesign jobs
1356 2011-01-16 17:46:35 <UukGoblin> "bounty" isn't good enough for me... I need a job that'll provide me a constant income for a fixed number of hours of work per day
1357 2011-01-16 17:47:14 <theymos> Bitcoin Block Explorer is finally back to normal! HTTPS is enabled again, and updates happen at the normal 1-minute frequency. Page loading should also be faster. I changed some things, so let me know if there are any bugs.
1358 2011-01-16 17:47:17 <kiba> 1,200 dollars for tcatm
1359 2011-01-16 17:47:32 <kiba> everyone is a programmer so they don't want to hire other programmers
1360 2011-01-16 17:47:34 <Cusipzzz> theymos: nice, I sent a contrib yesterday.
1361 2011-01-16 17:47:45 <theymos> Cusipzzz: Thanks!
1362 2011-01-16 17:48:25 <Cusipzzz> been doing a lot of testing and BBE has been invaluable
1363 2011-01-16 17:49:20 cdecker has left ()
1364 2011-01-16 17:49:46 <tcatm> ArtForz: What rpm would you recommened for milling solid copper (instead of FR4)?
1365 2011-01-16 17:50:43 <slush> whats mode +ooo ?
1366 2011-01-16 17:50:54 <hacim> slush: three individual +o
1367 2011-01-16 17:50:55 <niekie> Mass-opping.
1368 2011-01-16 17:51:03 <hacim> slush: ie. ops for those three people
1369 2011-01-16 17:51:04 <slush> wow, thanks :)
1370 2011-01-16 17:52:00 <slush> is any help for irc admin commands lying around?
1371 2011-01-16 17:52:02 <UukGoblin> hmm, rpm... as long as it spins... ;-)
1372 2011-01-16 17:52:20 omglolbbq has joined
1373 2011-01-16 17:52:31 <niekie> slush: basic commands are /kick which takes 2 arguments, username and reason.
1374 2011-01-16 17:53:03 slush has joined
1375 2011-01-16 17:53:03 <omglolbbq> i just started bitcooin and i get "WARNING: Displayed transactions may not be correct! You may need to upgrade, or other nodes may need to upgrade"... whats going on?
1376 2011-01-16 17:53:14 <UukGoblin> slush, it's unlikely you'll have to use any commands... freenode is such a friendly place
1377 2011-01-16 17:53:15 <slush> wow
1378 2011-01-16 17:53:17 <niekie> slush: yeah, it works like that.
1379 2011-01-16 17:53:26 <niekie> Now you lost op though ;)
1380 2011-01-16 17:53:36 <slush> :-D
1381 2011-01-16 17:53:42 <theymos> omglolbbq: How many blocks do you have?
1382 2011-01-16 17:54:19 <omglolbbq> 102882
1383 2011-01-16 17:55:02 <ArtForz> ;;bc,blocks
1384 2011-01-16 17:55:02 <gribble> 103013
1385 2011-01-16 17:55:25 <omglolbbq> i have 8 connections but it's not downloading new blocks
1386 2011-01-16 17:55:51 <ArtForz> weird
1387 2011-01-16 17:56:01 <ArtForz> what client version?
1388 2011-01-16 17:56:05 <omglolbbq> .19
1389 2011-01-16 17:56:23 <theymos> How strange. That's the error for "longer invalid chain".
1390 2011-01-16 17:56:33 <omglolbbq> i had 102881 blocks when i started the client, it downloaded 1 block, then gave that message and stopped downloading
1391 2011-01-16 17:56:49 <theymos> Try deleting your block database files. Maybe they're corrupted.
1392 2011-01-16 17:57:09 <omglolbbq> ok that will take some time :)
1393 2011-01-16 17:57:46 <ArtForz> yep, prob corrupted block chain
1394 2011-01-16 17:58:02 <niekie> Make a backup though ;)
1395 2011-01-16 17:58:27 <nevezen> where does it download newer blocks from?
1396 2011-01-16 17:58:38 <sipa> the network
1397 2011-01-16 17:58:40 <omglolbbq> nevezen, other clients
1398 2011-01-16 17:58:55 <omglolbbq> blkindex and blk0001 is all?
1399 2011-01-16 17:59:05 <nevezen> no central server? ic..
1400 2011-01-16 17:59:21 <omglolbbq> nevezen, thats one of bitcoins strenghts
1401 2011-01-16 17:59:30 <sipa> nevezen: that's what peer-to-peer means
1402 2011-01-16 17:59:38 <sipa> everyone is server
1403 2011-01-16 17:59:47 <theymos> omglolbbq: All of the __db files, too, I think.
1404 2011-01-16 18:01:10 <omglolbbq> ok done
1405 2011-01-16 18:01:20 <omglolbbq> ohh thats much faster downloading than it used to
1406 2011-01-16 18:02:55 riX2000 has joined
1407 2011-01-16 18:03:19 <omglolbbq> now lets hope the network will agree on my previous balance :)
1408 2011-01-16 18:04:48 <niekie> Doesn't Bitcoin still rely on IRC for peer-introducing though?
1409 2011-01-16 18:05:00 <niekie> (or was that changed during my prolonged absence from Bitcoin stuff)
1410 2011-01-16 18:05:03 <omglolbbq> can it swarm the blockchain from multiple nodes or does it download block by block chronologicaly from one node?
1411 2011-01-16 18:05:18 <theymos> niekie: There's also a decentralized seednode method, though IRC is used first.
1412 2011-01-16 18:05:40 <theymos> omglolbbq: The initial download is from one node.
1413 2011-01-16 18:05:52 <niekie> What network does it currently use again?
1414 2011-01-16 18:05:58 <niekie> Still LFnet?
1415 2011-01-16 18:06:06 <theymos> Yes.
1416 2011-01-16 18:06:52 davout has quit (Read error: Operation timed out)
1417 2011-01-16 18:08:12 <hacim> damn, you can get dual 5870's in a laptop?
1418 2011-01-16 18:08:49 <omglolbbq> :o
1419 2011-01-16 18:09:29 <ArtForz> I doubt it
1420 2011-01-16 18:09:59 <hacim> this seems to suggest this guy has it: http://newyork.craigslist.org/brk/sys/2162737411.html
1421 2011-01-16 18:10:07 sabalaba has quit (Ping timeout: 255 seconds)
1422 2011-01-16 18:10:35 <hacim> alienware m17x
1423 2011-01-16 18:10:38 <ArtForz> M5870 = HD5770 core @ 700MHz
1424 2011-01-16 18:10:43 sabalaba has joined
1425 2011-01-16 18:11:39 <ArtForz> = HD5770 at HD5750 clocks
1426 2011-01-16 18:12:37 <hacim> huh, so you can overclock 5770s to achieve the same Mhash/sec as a 5870?
1427 2011-01-16 18:12:46 <ArtForz> errr... no?
1428 2011-01-16 18:12:53 <ArtForz> 5770 = half a 5870
1429 2011-01-16 18:13:30 <ArtForz> 2* M5870 = about 1* desktop HD5850
1430 2011-01-16 18:13:31 <hacim> right
1431 2011-01-16 18:14:03 <hacim> err
1432 2011-01-16 18:14:21 <scibotic> Yay for choosing passwords. :/
1433 2011-01-16 18:14:39 <hacim> a 5870 does ~313Mhash/s, and a 5850 does ~240
1434 2011-01-16 18:14:44 <ArtForz> yes?
1435 2011-01-16 18:14:56 <hacim> so how does 2 5870's = 1 5850?
1436 2011-01-16 18:15:07 <ArtForz> M5870, not the M
1437 2011-01-16 18:15:09 <ArtForz> *note
1438 2011-01-16 18:15:18 <ArtForz> M5870 = 800 ALUs @ 700MHz
1439 2011-01-16 18:15:20 <hacim> ohh i didn't realize there was a mobile edition
1440 2011-01-16 18:15:27 <ArtForz> HD5870 = 1600 ALUs @ 850MHz
1441 2011-01-16 18:15:42 <ArtForz> HD5770 = 800 ALUs @ 850MHz
1442 2011-01-16 18:15:46 <scibotic> Just chose a password that was 69 characters long, holy hell that's going to suck.
1443 2011-01-16 18:15:57 <ArtForz> HD5750 = 720 ALUs @ 700MHz
1444 2011-01-16 18:16:16 <ArtForz> so M5870 == HD5770 @ HD5750 clock
1445 2011-01-16 18:16:36 <ArtForz> = 560Gop/s
1446 2011-01-16 18:16:36 <hacim> i have been wondering why there are so many different brands of these radeon cards (ASUS, SAPPHIRE, DIAMOND, HIS, XFX, GIGABYTE, Velocity Micro Edge...)
1447 2011-01-16 18:16:49 <ArtForz> HD5850 = 1044 Gop/s
1448 2011-01-16 18:17:00 <hacim> err, scratch that velocity one
1449 2011-01-16 18:17:02 <ArtForz> HD5870 = 1360Gop/s
1450 2011-01-16 18:17:26 <omglolbbq> hacim, why are there so many different brands of cars, or peanutbutter?
1451 2011-01-16 18:17:33 <ArtForz> so 2* M5870 is a small bit faster than a HD5850 for raw shader throughput
1452 2011-01-16 18:18:44 <hacim> omglolbbq: sure, but the price differences are big
1453 2011-01-16 18:19:12 <omglolbbq> hacim, some brands guarantee certain performance or OC capabilities, and ofcourse, you pay for the name... like a benz :)
1454 2011-01-16 18:19:18 <ArtForz> for reference cards, you pretty much pay for the name on the sticker
1455 2011-01-16 18:19:18 <hacim> eg. I can get a sapphire for $269, or a visontek for $549
1456 2011-01-16 18:19:26 redMBA has joined
1457 2011-01-16 18:19:39 <hacim> i thought people were saying that sapphire sucks
1458 2011-01-16 18:19:55 <omglolbbq> sapphire always worked for me
1459 2011-01-16 18:19:57 <ArtForz> sapphire builds all AMD/ATI referencer cards
1460 2011-01-16 18:20:00 <ArtForz> *reference
1461 2011-01-16 18:20:18 <scibotic> I'd be more concerned about their warranty policy. :V
1462 2011-01-16 18:20:33 <ArtForz> so for example every 5970 except for the 4GB crazy editions are really sapphires
1463 2011-01-16 18:21:19 <ArtForz> other vendors just speed bin em to make "OC editions"
1464 2011-01-16 18:22:11 <hacim> they dont add something different on some cards, like Cyprus XT or so?
1465 2011-01-16 18:22:21 <sipa> "speed bin" ?
1466 2011-01-16 18:22:22 <omglolbbq> ArtForz, i had a saphire that sad guaranteed OCable to xxxMHz
1467 2011-01-16 18:22:27 <ArtForz> yes
1468 2011-01-16 18:22:29 <omglolbbq> *said
1469 2011-01-16 18:22:35 <ArtForz> put card in test box, see how far it OCs @ stock V
1470 2011-01-16 18:22:44 <sipa> oh right
1471 2011-01-16 18:22:47 <ArtForz> if it's > 800 or so for both cores, you have a possible OC edition
1472 2011-01-16 18:22:50 <hacim> or have openGL on some and not others
1473 2011-01-16 18:23:02 <ArtForz> if it doesnt, it goes into the "stock cards" bin
1474 2011-01-16 18:23:25 <omglolbbq> my 5850 runs 900MHz on stock cooler and voltage
1475 2011-01-16 18:23:32 <ArtForz> thats pretty decent
1476 2011-01-16 18:23:42 <omglolbbq> 300mhash :)
1477 2011-01-16 18:23:50 <ArtForz> well, okay, 5850s are a bit of a special case
1478 2011-01-16 18:24:30 <ArtForz> difference between a plain sapphire 5970 and a asus "voltmod" 5970? sticker, box, and different manufacturer ID in BIOS
1479 2011-01-16 18:25:06 <ArtForz> flash a sapphire 5970 with the asus bios and ASUSs voltmod util happily accepts it
1480 2011-01-16 18:25:27 <ArtForz> same thing with MSI and afterburner
1481 2011-01-16 18:26:18 <omglolbbq> ArtForz, afterburner won't let me pick anything higher then 900, and rivatuner doesn't see the card
1482 2011-01-16 18:26:26 <ArtForz> so real difference between a $500 sapphire and a $600 msi? $100 ;)
1483 2011-01-16 18:27:17 <ArtForz> iirc afterburner has a secret ini setting or registry key to allow higher-than-allowed OCs
1484 2011-01-16 18:27:36 <ArtForz> normally it limits max clock/voltage/... to "safe" levels
1485 2011-01-16 18:28:05 <necrodearia> Woooo, I look forward to Stefan Molyneux eventually supporting and promoting Bitcoin!
1486 2011-01-16 18:28:07 <omglolbbq> well i'll go find that settingg :)
1487 2011-01-16 18:28:20 <omglolbbq> Who's Stefan Molyneux?
1488 2011-01-16 18:28:26 <necrodearia> It will be very interesting to hear his take on Bitcoin
1489 2011-01-16 18:28:30 <necrodearia> omglolbbq, Freedomain Radio
1490 2011-01-16 18:28:32 <ArtForz> I like to call it "just do what I say" mode ;)
1491 2011-01-16 18:29:27 <omglolbbq> ArtForz, you have any idea what the performance loss/win is on windows vs linux?
1492 2011-01-16 18:29:31 <omglolbbq> on same gpu
1493 2011-01-16 18:29:40 <ArtForz> from my benchmarks, negligible
1494 2011-01-16 18:30:07 <ArtForz> makes well < 1% difference
1495 2011-01-16 18:30:10 <luke-jr> â¦
1496 2011-01-16 18:30:16 <omglolbbq> ahh thats not worth switching right now :)
1497 2011-01-16 18:30:19 <luke-jr> omglolbbq: Linux doesn't support OpenCL
1498 2011-01-16 18:30:30 <niekie> ... it doesn't?
1499 2011-01-16 18:30:31 <ArtForz> no, ATIs driver does
1500 2011-01-16 18:30:41 <niekie> nVidia's does too.
1501 2011-01-16 18:30:45 <luke-jr> but that's not Linux
1502 2011-01-16 18:30:50 <luke-jr> it's a proprietary fork of Linux
1503 2011-01-16 18:30:53 <ArtForz> not for ATI cards ;)
1504 2011-01-16 18:30:54 <luke-jr> no better than Windows
1505 2011-01-16 18:30:59 <luke-jr> ArtForz: no?
1506 2011-01-16 18:31:07 <luke-jr> I admit, I only have witnessed the condition of nVidia
1507 2011-01-16 18:31:12 <ArtForz> no, it's a propertiary driver on top of linux
1508 2011-01-16 18:31:19 <luke-jr> so it's userland?
1509 2011-01-16 18:31:22 <ArtForz> last time I checked ATI doesn't provide the whole kernel tree
1510 2011-01-16 18:31:22 <hacim> "Turbo Edition"?
1511 2011-01-16 18:31:26 <ArtForz> OpenCL is
1512 2011-01-16 18:31:28 <luke-jr> doesn't matter
1513 2011-01-16 18:31:32 <luke-jr> a patch is still a derived work
1514 2011-01-16 18:31:33 <ArtForz> yes it does
1515 2011-01-16 18:31:42 <ArtForz> a patch is not a fork
1516 2011-01-16 18:32:01 <sipa> and it is not a patch
1517 2011-01-16 18:32:04 <luke-jr> a patch is a fork if it doesn't get merged
1518 2011-01-16 18:32:06 <sipa> the source is still linux
1519 2011-01-16 18:32:07 <ArtForz> which parts of core linux functionality does it change?
1520 2011-01-16 18:32:11 <ArtForz> oh, NONE?
1521 2011-01-16 18:32:14 <luke-jr> ArtForz: video
1522 2011-01-16 18:32:17 <sipa> it's a non-free plugin loaded into linux
1523 2011-01-16 18:32:17 <ArtForz> nope
1524 2011-01-16 18:32:24 <luke-jr> sipa: Linux isn't a microkernel.
1525 2011-01-16 18:32:25 <ArtForz> console is still plain ole linux VESA driver
1526 2011-01-16 18:32:29 <sipa> luke-jr: so?
1527 2011-01-16 18:32:38 <luke-jr> sipa: so only microkernels have drivers/plugins
1528 2011-01-16 18:32:43 <sipa> ok
1529 2011-01-16 18:32:50 <sipa> it's a non-free module loaded into linux
1530 2011-01-16 18:32:57 <luke-jr> it is impossible to run something in kernel-mode in Linux that isn't derived from Linux
1531 2011-01-16 18:33:04 <ArtForz> bullshit
1532 2011-01-16 18:33:46 <luke-jr> per the legal definition of derived anyhow
1533 2011-01-16 18:33:54 AAA_awright has quit (Ping timeout: 240 seconds)
1534 2011-01-16 18:34:03 <luke-jr> which, keep in mind, includes books set in a Star Wars universe
1535 2011-01-16 18:34:12 <ArtForz> the final running result is a derviced work
1536 2011-01-16 18:34:12 <luke-jr> even if they don't use any of the same characters, places
1537 2011-01-16 18:34:21 <ArtForz> the driver itself doesn't have to be
1538 2011-01-16 18:34:24 <luke-jr> ArtForz: law doesn't care about technicalities
1539 2011-01-16 18:34:29 <ArtForz> yes it fucking does
1540 2011-01-16 18:34:47 <ArtForz> so all windows NDIS drivers are derived works of linux?
1541 2011-01-16 18:34:53 <luke-jr> NDIS isn't Windows
1542 2011-01-16 18:35:04 <luke-jr> NDIS is a standard network driver plugin API
1543 2011-01-16 18:35:07 <ArtForz> yep
1544 2011-01-16 18:35:13 <ArtForz> so, are they derived works of linux?
1545 2011-01-16 18:35:16 <luke-jr> ndiswrapper is
1546 2011-01-16 18:35:30 <ArtForz> I asked if the DRIVER is a derived work, not ndiswrapper
1547 2011-01-16 18:35:35 <LobsterMan> ;;bc,stats
1548 2011-01-16 18:35:37 <luke-jr> nope
1549 2011-01-16 18:35:37 <gribble> Current Blocks: 103019 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 1812 blocks | Next Difficulty In About: 1 week, 3 days, 9 hours, 39 minutes, and 12 seconds | Next Difficulty Estimate: 22404.40357018
1550 2011-01-16 18:35:43 AAA_awright has joined
1551 2011-01-16 18:36:12 <ArtForz> while loading the same driver with ndiswrapper makes the whole thing a derived work
1552 2011-01-16 18:36:35 <luke-jr> that'd probably be a grey area
1553 2011-01-16 18:36:58 <luke-jr> more likely mere aggregation
1554 2011-01-16 18:37:17 <ArtForz> distributing the combined work would violate GPL and probably also the propertiary license of the original driver
1555 2011-01-16 18:37:51 <luke-jr> IANAL. doesn't seem clear-cut like that.
1556 2011-01-16 18:37:58 <ArtForz> but distributing the GPL parts and the propertiary driver is a-ok
1557 2011-01-16 18:38:37 <luke-jr> anyhow, the problem isn't the blobs generally
1558 2011-01-16 18:38:49 <luke-jr> it's the code that is derived from both Linux and the blob
1559 2011-01-16 18:39:01 <ArtForz> which is under GPL2 last time I checked
1560 2011-01-16 18:39:04 <luke-jr> per the GPL, that code cannot be distributed
1561 2011-01-16 18:39:10 <ArtForz> huh?
1562 2011-01-16 18:39:12 <luke-jr> GPL2 cannot link to blobs
1563 2011-01-16 18:39:19 <ArtForz> sure can
1564 2011-01-16 18:39:21 <luke-jr> nope
1565 2011-01-16 18:39:25 <luke-jr> not even LGPL can
1566 2011-01-16 18:39:29 <ArtForz> GPL2 doesnt limit use last time I checked
1567 2011-01-16 18:39:39 <sipa> it limits distribution
1568 2011-01-16 18:39:40 <luke-jr> linking isn't use.
1569 2011-01-16 18:39:50 <ArtForz> I am free to patch the binary resulting from GPL2 to hell and back as long as I dont distribute it
1570 2011-01-16 18:39:54 <sipa> you're allowed to link with whatever you want if you don't distribute
1571 2011-01-16 18:39:58 <luke-jr> you cannot distribute GPL code which links to blob
1572 2011-01-16 18:40:08 <Diablo-D3> luke-jr: not exactly
1573 2011-01-16 18:40:18 <ArtForz> sure can
1574 2011-01-16 18:40:18 <Diablo-D3> you cant use it as a way of violating the GPL.
1575 2011-01-16 18:40:44 <luke-jr> ArtForz: nope, go read GPL and consult a lawyer
1576 2011-01-16 18:40:56 <Diablo-D3> also, GPL can link to closed source code that is considered "the OS"
1577 2011-01-16 18:40:59 <ArtForz> you do not distribute the linked product
1578 2011-01-16 18:41:04 <Diablo-D3> yeah what art said
1579 2011-01-16 18:41:07 <luke-jr> Diablo-D3: true
1580 2011-01-16 18:41:07 <Diablo-D3> the big issue is DISTRIBUTING it
1581 2011-01-16 18:41:12 <ArtForz> because that would violate the gpl
1582 2011-01-16 18:41:17 <Diablo-D3> I can write GPL that 100% depends on an outside binary blob
1583 2011-01-16 18:41:19 <Diablo-D3> and Im fine
1584 2011-01-16 18:41:28 <Diablo-D3> its not my blob, and Im not distributing the two together
1585 2011-01-16 18:41:54 <sipa> and even with the blob, it is still your program
1586 2011-01-16 18:41:59 <luke-jr> ArtForz: Copyright does not define "link" as what C libraries do when they load things
1587 2011-01-16 18:42:15 <luke-jr> Legal "link" applies neutral to mechanisms in place, or languages used
1588 2011-01-16 18:42:18 <Diablo-D3> sipa: yes, but you cant use it as a way to violate the GPL
1589 2011-01-16 18:42:22 <ArtForz> no, but copyright defines "distribution" and "mere aggregation"
1590 2011-01-16 18:42:55 <Diablo-D3> sipa: you cant put your entire program into a binary blob, wrap it in a think GPL layer, and then claim its GPL and open source
1591 2011-01-16 18:43:02 <luke-jr> Diablo-D3: you can't *modify* someone else's GPL software such that it uses your blob, and distribute that modification
1592 2011-01-16 18:43:23 <Diablo-D3> luke-jr: I can modify their GPL code and distribute it.
1593 2011-01-16 18:43:34 <sipa> Diablo-D3: it won't be GPL, and it won't be open source, but it will be your program
1594 2011-01-16 18:43:38 <luke-jr> Diablo-D3: you can't, if your modification links to something proprietary
1595 2011-01-16 18:43:48 <Diablo-D3> actually, yes I can.
1596 2011-01-16 18:43:51 <luke-jr> nope
1597 2011-01-16 18:43:53 <Diablo-D3> go read the GPL
1598 2011-01-16 18:43:57 <Diablo-D3> even RMS agrees with me on this
1599 2011-01-16 18:44:04 <Diablo-D3> so go license troll elsewhee, noobs.
1600 2011-01-16 18:44:04 <luke-jr> I've read the GPL
1601 2011-01-16 18:44:19 <luke-jr> IP lawyers agree that closed source Linux modules are illegal
1602 2011-01-16 18:44:21 <luke-jr> :p
1603 2011-01-16 18:44:22 <ArtForz> yet you horribly misunderstand it
1604 2011-01-16 18:44:28 <ArtForz> which ones?
1605 2011-01-16 18:44:47 <ArtForz> "my imaginary lawyer friend" doesn't count
1606 2011-01-16 18:44:48 <Diablo-D3> IP lawyers do _not_ agree
1607 2011-01-16 18:45:01 <Diablo-D3> a GPL program dynamically loading a binary blob does not constitute anything in the realm of licensing
1608 2011-01-16 18:45:06 <Diablo-D3> this includes the kernel.
1609 2011-01-16 18:45:18 <Diablo-D3> you simply cant statically link it into your kernel.
1610 2011-01-16 18:45:34 <Diablo-D3> (nor, re: the kernel, is that even a good idea)
1611 2011-01-16 18:45:57 <luke-jr> Diablo-D3: the law doesn't see a difference between "dynamic linking" and "static linking"
1612 2011-01-16 18:46:04 <sipa> These requirements apply to the modified work as a whole. If
1613 2011-01-16 18:46:05 <sipa> identifiable sections of that work are not derived from the Program,
1614 2011-01-16 18:46:05 <sipa> and can be reasonably considered independent and separate works in
1615 2011-01-16 18:46:05 <sipa> themselves, then this License, and its terms, do not apply to those
1616 2011-01-16 18:46:05 <sipa> sections when you distribute them as separate works. But when you
1617 2011-01-16 18:46:07 <sipa> distribute the same sections as part of a whole which is a work based
1618 2011-01-16 18:46:10 <sipa> on the Program, the distribution of the whole must be on the terms of
1619 2011-01-16 18:46:12 <sipa> this License, whose permissions for other licensees extend to the
1620 2011-01-16 18:46:13 <Diablo-D3> luke-jr: no, but the GPL does.
1621 2011-01-16 18:46:13 <kiba> REAL LIFE FPS!
1622 2011-01-16 18:46:15 <sipa> entire whole, and thus to each and every part regardless of who wrote it.
1623 2011-01-16 18:46:17 <luke-jr> ArtForz: http://www.kroah.com/log/linux/ols_2006_keynote.htm
1624 2011-01-16 18:46:24 <Diablo-D3> luke-jr: ergo, the law does by extension.
1625 2011-01-16 18:46:27 <luke-jr> Diablo-D3: nope
1626 2011-01-16 18:46:27 <kiba> remote controlling bots to shoot paintball at each other
1627 2011-01-16 18:46:34 <Diablo-D3> Anyhow, Im not going to bother arguing this
1628 2011-01-16 18:47:00 <Diablo-D3> RMS already put his foot down on the definitive interpretation of the law.
1629 2011-01-16 18:47:08 <sipa> file not found
1630 2011-01-16 18:47:10 <Diablo-D3> You're a dick if you do it, but as long as you dont violate the GPL, you're fine.
1631 2011-01-16 18:47:39 <Diablo-D3> loading plugins, modules, or anything else that is not distributed with a GPL program is legal even if they are binary blobs
1632 2011-01-16 18:48:05 <Diablo-D3> linking GPL programs to binary libraries that are required to use OS or hardware functionality is also legal within the GPL.
1633 2011-01-16 18:48:06 <luke-jr> Diablo-D3: that might be true, but Linux doesn't have loadable plugins like that
1634 2011-01-16 18:48:11 <sipa> luke-jr: so, you think that loading fglrx into the linux kernel (being separately distributed, and linux not requiring fglrx to work), is a GPL infringement?
1635 2011-01-16 18:48:21 <Diablo-D3> luke-jr: actually, both RMS and Linus said it does
1636 2011-01-16 18:48:31 <Diablo-D3> all dynamically loaded modules are.
1637 2011-01-16 18:49:02 <Diablo-D3> what do you think modprobe does? masturbate using ground unicorn dust?
1638 2011-01-16 18:49:05 <luke-jr> sipa: I don't know the technicalities of fglrx. For it to work, you need to load an infringing piece of code. That might or might not be part of the 'fglrx.ko' file
1639 2011-01-16 18:49:11 theymos has quit (Remote host closed the connection)
1640 2011-01-16 18:49:22 <luke-jr> Diablo-D3: modprobe loads kernel internals that have been split off
1641 2011-01-16 18:49:39 <Diablo-D3> luke-jr: modules, which may or may not be sourced from the kernel source.
1642 2011-01-16 18:49:56 <luke-jr> Diablo-D3: if they don't derive from Linux, they won't work
1643 2011-01-16 18:50:00 <luke-jr> Linux has no driver APIs
1644 2011-01-16 18:50:02 <ArtForz> sure do
1645 2011-01-16 18:50:06 <ArtForz> then you MAKE ONE
1646 2011-01-16 18:50:14 <ArtForz> which is exactly what the GPL wrapper does
1647 2011-01-16 18:50:21 <Diablo-D3> luke-jr: yes it does
1648 2011-01-16 18:50:26 <luke-jr> ArtForz: I'm not aware of anyone doing that besides ndiswrapper
1649 2011-01-16 18:50:32 <Diablo-D3> fglrx and nvidia both use the existing APIs.
1650 2011-01-16 18:50:39 <luke-jr> Diablo-D3: there are no APIs
1651 2011-01-16 18:50:45 <Diablo-D3> yet, it uses them.
1652 2011-01-16 18:50:47 <luke-jr> they use INTERNAL PRIVATE SYMBOLS of Linux
1653 2011-01-16 18:50:51 <ArtForz> what do you think the GPLed parts disributed with ati/nvidia/... do?
1654 2011-01-16 18:50:53 <ArtForz> no they dont
1655 2011-01-16 18:50:54 <sipa> the GPL part of the driver does
1656 2011-01-16 18:51:02 <ArtForz> the call the wrapper API, which calls internal linux code
1657 2011-01-16 18:51:02 <luke-jr> sipa: GPL cannot link to blob
1658 2011-01-16 18:51:21 <Diablo-D3> luke-jr: C has no concept for private symbols
1659 2011-01-16 18:51:21 <luke-jr> ArtForz: where is this API defined?
1660 2011-01-16 18:51:34 <Diablo-D3> luke-jr: the kernel contains headers full of signatures of said symbols
1661 2011-01-16 18:51:44 <luke-jr> Diablo-D3: all symbols are internal and private unless otherwise specified in a document.
1662 2011-01-16 18:51:45 <ArtForz> mainly, by the wrapper code
1663 2011-01-16 18:51:50 <Diablo-D3> luke-jr: the kernel loads the modules from fglrx, and nvidia, and then those modules call the symbols.
1664 2011-01-16 18:52:04 <Diablo-D3> luke-jr: according to _BOTH_ RMS _AND_ LINUS, THIS IS OKAY
1665 2011-01-16 18:52:06 <ArtForz> but feel free to write a spec and distribute it to your "binary blob" and "linux wrapper" teams
1666 2011-01-16 18:52:23 <Diablo-D3> the guy who wrote the first fucking version of GPL, and the guy who wrote the kernel, agree with me
1667 2011-01-16 18:52:24 <luke-jr> yawn
1668 2011-01-16 18:52:30 <Diablo-D3> so where the fuck do you think you're going with this?
1669 2011-01-16 18:52:39 <luke-jr> Diablo-D3: Linus is an idiot. I doubt your claim that RMS agrees.
1670 2011-01-16 18:52:39 <Diablo-D3> somehow you know more than those two than the legal aspects of this?
1671 2011-01-16 18:52:42 <Diablo-D3> get a clue.
1672 2011-01-16 18:52:43 <luke-jr> Finally, neither are IP lawyers.
1673 2011-01-16 18:52:51 <luke-jr> and the IP lawyers say it's illegal
1674 2011-01-16 18:52:57 <Diablo-D3> IP lawyers who do work for the FSF says its okay.
1675 2011-01-16 18:53:04 <Diablo-D3> so blow it out your ass.
1676 2011-01-16 18:53:05 <luke-jr> Diablo-D3: prove it
1677 2011-01-16 18:53:05 <sipa> can you provide a source for that, luke-jr?
1678 2011-01-16 18:53:05 <davex__> i say fuck ip and lawyers.
1679 2011-01-16 18:53:09 <luke-jr> sipa: I already did
1680 2011-01-16 18:53:17 <luke-jr> http://www.kroah.com/log/linux/ols_2006_keynote.html
1681 2011-01-16 18:53:18 <sipa> 19:42:53 < sipa> file not found
1682 2011-01-16 18:53:24 <luke-jr> sipa: fix your browser
1683 2011-01-16 18:53:35 <Diablo-D3> luke-jr: and btw, if you claim Linus is an idiot, whos going to sue? hmmm?
1684 2011-01-16 18:53:36 <sipa> apologies, now it works
1685 2011-01-16 18:53:38 <ArtForz> thats ONE IP lawyer, working for novell, saying somethign totally unrelated
1686 2011-01-16 18:53:42 <Diablo-D3> LOL
1687 2011-01-16 18:53:44 <Diablo-D3> NOVELL
1688 2011-01-16 18:53:46 <luke-jr> Diablo-D3: hopefully someone with balls
1689 2011-01-16 18:53:48 <Diablo-D3> the ultimate closed source company
1690 2011-01-16 18:53:49 <Diablo-D3> fuck them
1691 2011-01-16 18:53:50 <ArtForz> "you can't have your program include linux headers wihtout it being a derived work"
1692 2011-01-16 18:53:51 <johnyh> heh you guys are so easly trolled by luke-jr :)
1693 2011-01-16 18:53:55 <Diablo-D3> its just more microsoft oriented fud
1694 2011-01-16 18:53:59 <ArtForz> which EVERYONE FUCKING AGREES ON
1695 2011-01-16 18:54:13 <Diablo-D3> johnyh: not really, I called him a troll about 200 lines up
1696 2011-01-16 18:54:19 <luke-jr> Diablo-D3: Linus doesn't hold copyright on Linux.
1697 2011-01-16 18:54:27 <ArtForz> I called him a troll about 2 weeks ago
1698 2011-01-16 18:54:34 <ArtForz> and thats relevant to anything how?
1699 2011-01-16 18:54:40 <luke-jr> [13:49:21] <Diablo-D3> luke-jr: and btw, if you claim Linus is an idiot, whos going to sue? hmmm?
1700 2011-01-16 18:54:51 Xanie has quit (Quit: if assholes could fly,this place would be a airport!)
1701 2011-01-16 18:54:55 <luke-jr> sole copyright*
1702 2011-01-16 18:54:59 <Diablo-D3> Uh-huh.
1703 2011-01-16 18:55:07 <ArtForz> again, thats relevant to anything how?
1704 2011-01-16 18:55:14 <luke-jr> *any* Linux contributor that infringe on can sue
1705 2011-01-16 18:55:17 <ArtForz> again, thats relevant to anything how?
1706 2011-01-16 18:55:21 <Diablo-D3> okay, I have code in the kernel
1707 2011-01-16 18:55:23 * Diablo-D3 sues luke-jr
1708 2011-01-16 18:55:30 <luke-jr> unfortunately, my own kernel patches are too trivial for this
1709 2011-01-16 18:55:30 <ArtForz> I do, too actually
1710 2011-01-16 18:55:35 <luke-jr> Diablo-D3: I don't infringe âº
1711 2011-01-16 18:55:38 <Diablo-D3> ArtForz: who doesnt at this point
1712 2011-01-16 18:55:53 <luke-jr> Diablo-D3: trivial code isn't copyrightable :p
1713 2011-01-16 18:56:12 <Diablo-D3> welp, linux cant be copyrighted then
1714 2011-01-16 18:56:15 <luke-jr> not saying yours is trivial, but lots of people "doesnt" have non-trivial code in there
1715 2011-01-16 18:56:22 <Diablo-D3> since its ONLY like a million lines
1716 2011-01-16 18:57:01 * Diablo-D3 goes ahead and takes luke-jr's advice and steals all the code
1717 2011-01-16 18:57:04 <Diablo-D3> Im glad I know an IP lawyer
1718 2011-01-16 18:57:05 <ArtForz> and again, how is that fucking relevant to anything
1719 2011-01-16 18:57:11 <Diablo-D3> oh whats that luke? you're NOT one?
1720 2011-01-16 18:57:13 <luke-jr> Diablo-D3: you still fail to provide a citation for RMS endorsing binary Linux modules
1721 2011-01-16 18:57:14 <ArtForz> yeah, his name is 127.0.0.1
1722 2011-01-16 18:57:15 <Diablo-D3> go blow it out your ass.
1723 2011-01-16 18:57:48 <luke-jr> at least I can cite IP lawyers who agree.
1724 2011-01-16 18:58:03 <ArtForz> ... in your fantasy world
1725 2011-01-16 18:58:24 <Diablo-D3> Im not sure how an IP lawyer working for Microsoft/Novell has any right to weigh in on the issue
1726 2011-01-16 18:58:31 sabalaba has quit (Ping timeout: 272 seconds)
1727 2011-01-16 18:58:40 <ArtForz> not to mention he talks aboput INCLUDING LINUX HEADER FILES
1728 2011-01-16 18:59:12 <luke-jr> Diablo-D3: Novell owns more of Linux than you do I bet
1729 2011-01-16 18:59:25 <ArtForz> if that .h contains a 50-line macro, thats sure as hell including GPL2 code
1730 2011-01-16 18:59:31 <ArtForz> and again, how is that fucking relevant to anything
1731 2011-01-16 18:59:47 <ArtForz> yay for cursor-up
1732 2011-01-16 18:59:53 * Diablo-D3 just puts luke-jr on ignore
1733 2011-01-16 18:59:57 <ArtForz> makes arguing with trolls much more fun
1734 2011-01-16 19:00:19 sipa has quit (Quit: zie u op de omdraai zijde)
1735 2011-01-16 19:00:32 <Diablo-D3> I have more important things to deal with arm chair lawyers on the Internet
1736 2011-01-16 19:00:42 <Diablo-D3> luke has never written a day of published code in his life
1737 2011-01-16 19:00:48 <Diablo-D3> his opinion means squat shit
1738 2011-01-16 19:00:50 <luke-jr> lol
1739 2011-01-16 19:01:15 <ArtForz> but he can cite totally unrelated stuff that popped up in a google search
1740 2011-01-16 19:01:33 <luke-jr> actually I have, but writing code is kinda irrelevant to competency in law anyway
1741 2011-01-16 19:02:19 <Diablo-D3> ArtForz: yeah so?
1742 2011-01-16 19:02:23 <Diablo-D3> if I wanted to go argue with google, I would
1743 2011-01-16 19:02:59 <ArtForz> crap, forgot the <sarcasm> tag
1744 2011-01-16 19:06:18 CyanDynamo has joined
1745 2011-01-16 19:08:54 akem has joined
1746 2011-01-16 19:08:54 akem has quit (Changing host)
1747 2011-01-16 19:08:54 akem has joined
1748 2011-01-16 19:09:03 <omglolbbq> ;;bc,blocks
1749 2011-01-16 19:09:04 <gribble> 103023
1750 2011-01-16 19:09:25 <omglolbbq> got all blocks, no errors, balance is back :)
1751 2011-01-16 19:09:45 <ArtForz> so you had a corrupted block chain.. weird
1752 2011-01-16 19:09:57 <omglolbbq> how could that happen?
1753 2011-01-16 19:10:04 <ArtForz> exactly, should figure out howTF this can happen, and how to prevent it
1754 2011-01-16 19:10:05 <omglolbbq> downloaded something from a malicious node?
1755 2011-01-16 19:10:10 <ArtForz> nah
1756 2011-01-16 19:10:11 sabalaba has joined
1757 2011-01-16 19:10:31 <ArtForz> did your client/box crash before it started doing that?
1758 2011-01-16 19:10:50 <Cusipzzz> bad disk/write error ros omething
1759 2011-01-16 19:10:56 <ArtForz> I suspect it might've crashed mid-write to blkXXXX.dat or something
1760 2011-01-16 19:11:08 <omglolbbq> don't recall any crashes :)
1761 2011-01-16 19:11:12 <ArtForz> = half-a-block at the end of blkXXXX.dat
1762 2011-01-16 19:11:20 <omglolbbq> but i had some weird behaviour today though
1763 2011-01-16 19:11:24 <omglolbbq> that caused me to reboot
1764 2011-01-16 19:11:37 <omglolbbq> every application i started popped the message 'access denied'
1765 2011-01-16 19:11:53 <omglolbbq> i could only close programs, not start any, reboot solved it... :S
1766 2011-01-16 19:12:01 <ArtForz> I hink Cusipzzz might be right
1767 2011-01-16 19:12:19 <ArtForz> disk write error can cause crazy symptoms like that
1768 2011-01-16 19:12:23 Slix` has joined
1769 2011-01-16 19:12:54 <omglolbbq> i guess i can expect more special actions to come then :)
1770 2011-01-16 19:13:15 <ArtForz> also fun: running stuff on a box with flaky RAM
1771 2011-01-16 19:14:39 <omglolbbq> im glad to have my 15 btc back! :D
1772 2011-01-16 19:14:49 <ArtForz> aka 1+1=2147483650
1773 2011-01-16 19:15:25 <omglolbbq> my ram is fine i think, but disk may be problem
1774 2011-01-16 19:15:37 <omglolbbq> oh well im running windows, so it could realy be anything i suppose
1775 2011-01-16 19:15:55 <ArtForz> when in doubt, blame cosmic rays
1776 2011-01-16 19:16:03 <nevezen> right
1777 2011-01-16 19:16:08 <omglolbbq> :
1778 2011-01-16 19:16:08 <omglolbbq> :)
1779 2011-01-16 19:24:09 johnyh has quit (Remote host closed the connection)
1780 2011-01-16 19:24:43 james has joined
1781 2011-01-16 19:25:09 james is now known as Guest12529
1782 2011-01-16 19:25:16 redMBA has quit (Ping timeout: 240 seconds)
1783 2011-01-16 19:25:26 johnyh has joined
1784 2011-01-16 19:38:39 devon_hillard has quit (Read error: Connection reset by peer)
1785 2011-01-16 19:40:03 <necrodearia> philosophy type of question (I'm not sure if it's written so well): Hmm, is it possible to produce exploits (not software, but perhaps related to social engineering) in which individuals around the world are led to continue their support of ideas as they normally do, but in a way in which their support is directed towards those that have exploited and instead redirect the support into other ideas (particularly those expressed clearly by
1786 2011-01-16 19:40:03 <necrodearia> Stefan Molyneux)? For example, individuals participative in the collective "powers that be" (similar to the swarm of anonymous) can simply pose as such and establish support from and control over others, in which the support is funding not continued enslavement of humanity, but instead funding help to reduce and minimize such control and power.
1787 2011-01-16 19:40:37 <necrodearia> mm, I shall post on forum instead
1788 2011-01-16 19:42:38 AAA_awright_ has joined
1789 2011-01-16 19:44:44 <tcatm> slush: does your pool run a standard bitcoind on port 8333?
1790 2011-01-16 19:45:00 <slush> tcatm: yes
1791 2011-01-16 19:45:13 AAA_awright_ has quit (Client Quit)
1792 2011-01-16 19:45:17 * tcatm connects miner
1793 2011-01-16 19:45:30 <andrew12> heh
1794 2011-01-16 19:45:42 <slush> you mean rpc api? It is on 8332
1795 2011-01-16 19:45:47 <tcatm> no, 8333
1796 2011-01-16 19:45:52 AAA_awright_ has joined
1797 2011-01-16 19:45:55 AAA_awright has quit (Ping timeout: 240 seconds)
1798 2011-01-16 19:46:32 AAA_awright_ has quit (Client Quit)
1799 2011-01-16 19:46:42 <slush> tcatm: why do you need connect miner to 8333?
1800 2011-01-16 19:46:52 <tcatm> fixing efficiency issues
1801 2011-01-16 19:47:27 <slush> what is default bitcoin port for p2p?
1802 2011-01-16 19:47:34 <tcatm> 8333
1803 2011-01-16 19:48:01 <slush> omg
1804 2011-01-16 19:48:04 <slush> i'm on bad server
1805 2011-01-16 19:49:02 <andrew12> o.o
1806 2011-01-16 19:49:09 <slush> ok, everything is ok
1807 2011-01-16 19:49:17 <slush> 8333 is open on pool, too
1808 2011-01-16 19:49:22 <tcatm> great
1809 2011-01-16 19:49:35 AAA_awright has joined
1810 2011-01-16 19:49:38 <andrew12> hmm
1811 2011-01-16 19:49:43 <tcatm> grrr... aptitude install sucks over NFS
1812 2011-01-16 19:49:46 <andrew12> my miner isnt running
1813 2011-01-16 19:49:49 <andrew12> though it is now
1814 2011-01-16 19:49:56 <andrew12> because i just started it :p
1815 2011-01-16 19:50:52 <andrew12> should probably run a miner on my mac too
1816 2011-01-16 19:55:17 * andrew12 gets all excited whenever he sees that his miner found a hash
1817 2011-01-16 19:58:29 <tcatm> slush: works much better this way :)
1818 2011-01-16 19:58:48 <slush> heh, 'which' way? :)
1819 2011-01-16 19:59:07 <tcatm> fake-push
1820 2011-01-16 19:59:41 RazielZ has quit ()
1821 2011-01-16 19:59:41 <tcatm> doing getwork only when your node has new TX/blocks (or all nonces are tried)
1822 2011-01-16 19:59:49 <andrew12> hmm
1823 2011-01-16 20:00:01 <tcatm> i.e. not every 5 seconds ;)
1824 2011-01-16 20:00:10 <slush> I don't see your mners connected
1825 2011-01-16 20:00:20 <tcatm> only connected one
1826 2011-01-16 20:00:23 <andrew12> jgarzik's miner is giving me a 400 error on my mac
1827 2011-01-16 20:00:29 <andrew12> with slush's pool
1828 2011-01-16 20:00:33 <slush> I'm thinking heavily about push based improvements. I think I'm very close...
1829 2011-01-16 20:00:50 <slush> andrew12: 400 are with bad parameters
1830 2011-01-16 20:01:01 <andrew12> then i'll check the stuff..
1831 2011-01-16 20:01:08 <slush> it indicate that pool is working, but you are giving bad arguments
1832 2011-01-16 20:01:42 <andrew12> ahh its cause i spelled the password wrong
1833 2011-01-16 20:01:50 <andrew12> now that that's running, i'll be off :P
1834 2011-01-16 20:02:36 <tcatm> slush: what's your push approach?
1835 2011-01-16 20:03:38 <slush> comet based + unique merkle for every worker + ability to change ntime on workers
1836 2011-01-16 20:03:56 <tcatm> comet?
1837 2011-01-16 20:04:26 <slush> extranonce based on worker login to avoid cross-worker spending of shares
1838 2011-01-16 20:04:30 andrew12_ has joined
1839 2011-01-16 20:04:39 <andrew12_> On my phone ;)
1840 2011-01-16 20:04:44 <slush> https://secure.wikimedia.org/wikipedia/en/wiki/Comet_%28programming%29
1841 2011-01-16 20:04:47 andrew12_ is now known as andrew12^droid
1842 2011-01-16 20:06:37 omglolbbq has left ()
1843 2011-01-16 20:08:55 <slush> tcatm: after share is submitted, server just check if merkle root correspond with worker login, if share is valid and ntime for this merkle is used first time
1844 2011-01-16 20:09:03 <tcatm> slush: http://pastebin.com/J4qZRgXT (+ adds logging support)
1845 2011-01-16 20:09:09 riX2000_ has joined
1846 2011-01-16 20:09:35 riX2000_ is now known as Guest12689
1847 2011-01-16 20:11:32 <slush> tcatm: nice. I have studying of poclbm in my todo
1848 2011-01-16 20:12:47 riX2000 has quit (Ping timeout: 260 seconds)
1849 2011-01-16 20:12:52 SirDidi has quit (Ping timeout: 240 seconds)
1850 2011-01-16 20:19:28 <andrew12^droid> Should try running a miner on my phone :P
1851 2011-01-16 20:19:59 andrew12^droid has quit (Quit: parting is such sweet sorrow...)
1852 2011-01-16 20:20:13 <tcatm> Use two phone and combine them with gloves
1853 2011-01-16 20:21:16 hundfred has joined
1854 2011-01-16 20:21:46 darrob has joined
1855 2011-01-16 20:25:21 SirDidi has joined
1856 2011-01-16 20:28:02 AAA_awright_ has joined
1857 2011-01-16 20:30:17 AAA_awright has quit (Ping timeout: 260 seconds)
1858 2011-01-16 20:31:38 <Cusipzzz> lol 8 rails
1859 2011-01-16 20:32:26 <ArtForz> ?
1860 2011-01-16 20:32:30 <Cusipzzz> mt
1861 2011-01-16 20:38:36 AAA_awright_ is now known as AAA_awright
1862 2011-01-16 20:42:20 lyspooner has joined
1863 2011-01-16 20:43:36 * tcatm needs to improve cooling of his cluster
1864 2011-01-16 20:48:21 <kiba> /me offers avatar art for 100 BTC
1865 2011-01-16 20:48:25 <kiba> err
1866 2011-01-16 20:48:38 * kiba offers forum avatar for 25 BTC. 100x100 pixels
1867 2011-01-16 20:50:04 darrob has quit (Ping timeout: 240 seconds)
1868 2011-01-16 20:51:43 darrob has joined
1869 2011-01-16 20:52:30 AAA_awright_ has joined
1870 2011-01-16 20:52:58 AAA_awright has quit (Ping timeout: 260 seconds)
1871 2011-01-16 20:54:12 <tcatm> selling hot air at -100 BTC :)
1872 2011-01-16 20:55:08 andrew12_ has joined
1873 2011-01-16 20:58:30 AAA_awright_ is now known as AAA_awright
1874 2011-01-16 20:58:44 joe_1 has quit (Ping timeout: 250 seconds)
1875 2011-01-16 20:59:05 <tcatm> ArtForz: How do you keep your 5970s cool?
1876 2011-01-16 21:01:18 <ArtForz> magic.
1877 2011-01-16 21:01:58 <ArtForz> "lots of fans" might also work decently
1878 2011-01-16 21:02:53 <ArtForz> my hottest core right now is at... 74.50 C
1879 2011-01-16 21:03:06 <ArtForz> at 60% card fan speed
1880 2011-01-16 21:03:07 <tcatm> I'm using the "lots of fans" approach which worked perfect until outside temperatures reached 9°C
1881 2011-01-16 21:03:24 <tcatm> What was the command to force fanspeed?
1882 2011-01-16 21:03:42 <ArtForz> aticonfig --pplib-cmd 'set fanspeed 0 X'
1883 2011-01-16 21:03:55 <ArtForz> it uses DISPLAY, not --adapter
1884 2011-01-16 21:04:26 * kiba got two jobs lined up
1885 2011-01-16 21:04:38 <tcatm> i.e. :0.[0..X]?
1886 2011-01-16 21:04:53 <ArtForz> .0 .2 .4 .6 for 4 5970s
1887 2011-01-16 21:05:54 <ArtForz> fan controller is connected to primary GPU on 5970
1888 2011-01-16 21:06:53 <tcatm> looks like some default to 43% fan
1889 2011-01-16 21:07:04 <ArtForz> 'get fanspeed 0' to read current fan speed
1890 2011-01-16 21:07:30 <ArtForz> iirc auto-fan maxes at like 60%
1891 2011-01-16 21:07:38 <tcatm> Much better! :) -11 °C
1892 2011-01-16 21:08:01 <ArtForz> for some wheird reason they decided to never let the cards hit max fan on auto
1893 2011-01-16 21:08:12 <ArtForz> iirc max RPM with auto is 4500RPM
1894 2011-01-16 21:08:20 <ArtForz> 100% is >6kRPM
1895 2011-01-16 21:08:26 <tcatm> Can tehy safely run at 100%?
1896 2011-01-16 21:08:42 <ArtForz> yep
1897 2011-01-16 21:08:54 <ArtForz> well, I never ran em at 100% for long, pointless
1898 2011-01-16 21:09:16 <ArtForz> anything >80% doesnt seem to help at all
1899 2011-01-16 21:09:41 <ArtForz> except for creating a really annying whine
1900 2011-01-16 21:09:53 <tcatm> Noise doesn't matter.
1901 2011-01-16 21:10:20 sipa1024 has joined
1902 2011-01-16 21:10:26 <ArtForz> I've run one of my 5970s @ 80% for months, no problems so far
1903 2011-01-16 21:11:10 <tcatm> Hm. My totalhashrate changed too, with fan speed.
1904 2011-01-16 21:11:20 <ArtForz> up?
1905 2011-01-16 21:11:23 <tcatm> yep
1906 2011-01-16 21:11:28 <ArtForz> your VRMs were probably overheating
1907 2011-01-16 21:11:34 <tcatm> +109 MHash/s
1908 2011-01-16 21:11:45 <ArtForz> what was core temp before?
1909 2011-01-16 21:12:07 <tcatm> 94°C (was 78°C a few weeks ago)
1910 2011-01-16 21:12:24 <ArtForz> yep, >90°C you can pretty much bet the VRMs are hitting 120°C
1911 2011-01-16 21:12:51 <tcatm> one 5870 still causes troubles jumping around 230...350 Mhash/s
1912 2011-01-16 21:13:16 <ArtForz> sounds like it's still throttling :/
1913 2011-01-16 21:13:32 <tcatm> The one closest to CPU (3x 5870 on that board)
1914 2011-01-16 21:13:43 <ArtForz> a few runs of aticonfig --odgc should show if the card is clocking down
1915 2011-01-16 21:13:53 <tcatm> Even though it has a fan blowing at ~45° from to its back
1916 2011-01-16 21:14:14 <ArtForz> if it jumps between ~500MHz and max, it's trottling
1917 2011-01-16 21:14:17 <ArtForz> +h
1918 2011-01-16 21:14:42 <tcatm> yep. 600/900
1919 2011-01-16 21:14:51 <ArtForz> yep :/
1920 2011-01-16 21:15:01 <ArtForz> needs more cowb... fans
1921 2011-01-16 21:15:23 <tcatm> it already has two (+ onboard fan)
1922 2011-01-16 21:15:34 <tcatm> Guess I need to change airflow
1923 2011-01-16 21:15:47 <ArtForz> probably
1924 2011-01-16 21:20:52 <slush> is possibility to check VRM temperature in linux?
1925 2011-01-16 21:21:24 <ArtForz> nope
1926 2011-01-16 21:21:33 <tcatm> Looks like the hot 5870 has a bad fan
1927 2011-01-16 21:21:53 <ArtForz> ewww
1928 2011-01-16 21:21:58 <tcatm> Or rather feels like. Air pressure is a lot lower than on the other cards
1929 2011-01-16 21:22:25 <ArtForz> intake unobstructed?
1930 2011-01-16 21:22:33 <tcatm> yep
1931 2011-01-16 21:22:49 <ArtForz> yeah, sounds like a bad fan, unless it's a non-ref cooler
1932 2011-01-16 21:22:49 <tcatm> There's even a fan blowing from the side of the case on the cards
1933 2011-01-16 21:23:10 <ArtForz> reference 5870 and 5970 have the exact same fan
1934 2011-01-16 21:23:23 <tcatm> I think it's non-ref (some XFX)
1935 2011-01-16 21:23:33 <tcatm> Just as the other two 5870 next to it
1936 2011-01-16 21:23:38 <ArtForz> well, that could explain it
1937 2011-01-16 21:23:44 <ArtForz> reference has a really quite oversized fan
1938 2011-01-16 21:24:05 <ArtForz> iirc a 10.2W w/ double ball bearing
1939 2011-01-16 21:25:14 <tcatm> Guess I should replace those 5870 with two 5970. A lot easier to cool because of space between them
1940 2011-01-16 21:26:12 <ArtForz> and as I said, stock fan curve never runs it > 4.5kRPM, it can hit >6k at 100%
1941 2011-01-16 21:26:18 ApertureScience has joined
1942 2011-01-16 21:26:20 <davex__> hey ArtForz, what was the new radeon you said is coming out next month?
1943 2011-01-16 21:26:25 <ArtForz> 6990
1944 2011-01-16 21:26:36 darrob has quit (Remote host closed the connection)
1945 2011-01-16 21:26:41 <ArtForz> probably a dual 6970 with more aggressive pwoer limiting
1946 2011-01-16 21:27:01 <davex__> ok
1947 2011-01-16 21:27:04 <ArtForz> so I guess a few 3rd party makers decided that a 10W fan never running >60% was a waste of money
1948 2011-01-16 21:27:37 darrob has joined
1949 2011-01-16 21:27:45 <ArtForz> = a 6W fan probably still wont go >80% PWM or so on the stock RPM profile
1950 2011-01-16 21:28:13 <ArtForz> nice way to shave off quite a few cents per unit
1951 2011-01-16 21:28:21 <tcatm> Yup.
1952 2011-01-16 21:28:59 <tcatm> Then again those XFX have a nice cooler design so I might be able to fit better fan on it.
1953 2011-01-16 21:29:29 <ArtForz> is it one of those center-fan cards?
1954 2011-01-16 21:29:32 <tcatm> yep
1955 2011-01-16 21:30:00 <ArtForz> try NOT directing a case fan at the ass-end, instead blow in from the case side = card top
1956 2011-01-16 21:30:53 <tcatm> that's what I'm doing
1957 2011-01-16 21:31:14 <tcatm> + one from cpu to back of that card
1958 2011-01-16 21:31:29 <ArtForz> weird
1959 2011-01-16 21:31:47 <nevezen> is the 6690 a single or double slot card?
1960 2011-01-16 21:32:19 <ArtForz> 6690?
1961 2011-01-16 21:32:33 <nevezen> 6990
1962 2011-01-16 21:32:38 <ArtForz> double slot
1963 2011-01-16 21:32:44 <ArtForz> it's 300W ffs
1964 2011-01-16 21:33:18 <nevezen> are there motherboards that can hold them without occupying an unused pcie slot..
1965 2011-01-16 21:33:26 <ArtForz> nope
1966 2011-01-16 21:34:07 <nevezen> so if a mobo had 4x pcie, you can only stick two of those in there?
1967 2011-01-16 21:34:16 <ArtForz> depends on the slot spacing
1968 2011-01-16 21:35:13 <sipa1024> usually x16 PCIe slots are separated by at least one other slot
1969 2011-01-16 21:35:24 <ArtForz> more than 2 cards per board is pretty hard without flexible risers as you pretty much NEED to leave 1 slot space next to a card to not block its intake
1970 2011-01-16 21:35:29 <nevezen> probably an x1 right
1971 2011-01-16 21:35:41 <ArtForz> except when the board designer is braindead (which most seem to be)
1972 2011-01-16 21:36:03 <nevezen> what was that motherboard that had 4x pcie?
1973 2011-01-16 21:36:13 <nevezen> I saw another picture of 4x cards in a system..
1974 2011-01-16 21:36:20 <ArtForz> pretty much every 790FX / 890FX board
1975 2011-01-16 21:36:26 <ArtForz> for example a lot of 790FX boards have something like x16 - x16 - x1 - x16 - x16 - PCI - PCI layout
1976 2011-01-16 21:36:29 <ArtForz> completely stupid
1977 2011-01-16 21:37:19 <ArtForz> one is even x1 - x16 - x16 - x16 - x1 - x16 - PCI
1978 2011-01-16 21:37:52 <tcatm> Yay. Temperature back to 81°C on that card (from 94)
1979 2011-01-16 21:37:59 <ArtForz> with slot #2 and #4 being the true x16 slots... who the fuck came up with that stupid idea...
1980 2011-01-16 21:38:14 <ArtForz> MSI 790FX has x16 - x1 - x16 - x1 - x16 - PCI - x16
1981 2011-01-16 21:38:33 <ArtForz> with slots #1 and #5 being the true x16 slots
1982 2011-01-16 21:38:59 <ArtForz> on that board more than 2 cards without PCIe risers to get at least 2 out of the way = bad idea
1983 2011-01-16 21:39:23 <tcatm> added a little bit of foam to widen the gap between the cards
1984 2011-01-16 21:39:25 <nevezen> what do you mean true x16?
1985 2011-01-16 21:39:48 <ArtForz> 790FX only has 32 lanes for GPU PCIe
1986 2011-01-16 21:40:13 <ArtForz> runs either x16 - 0 - x16 - 0, x16 - 0 - x8 - x8 or , x8 - x8 - x8 - x8
1987 2011-01-16 21:40:29 <ArtForz> but it can't do 0 - x16 - ...
1988 2011-01-16 21:40:42 <ArtForz> instead it runs (unused x8) - x8 - ...
1989 2011-01-16 21:41:07 <ArtForz> so you have 2 slots that can do x16 or x8 and 2 slots that can do 0 or x8
1990 2011-01-16 21:41:35 <ArtForz> = theres 2 true x16 slots and 2 slots that max out at x8
1991 2011-01-16 21:41:44 <nevezen> oh that sucks
1992 2011-01-16 21:41:50 <ArtForz> doesnt matter for mining anyways, but it kinda does matter for 3D
1993 2011-01-16 21:42:18 <nevezen> is that a pcie limitation though?
1994 2011-01-16 21:42:25 <ArtForz> nope, chipset limitation
1995 2011-01-16 21:42:46 <ArtForz> there are boards that have 4 full x16 slots, but the prices are crazy
1996 2011-01-16 21:43:02 <ArtForz> usually $400+ boards
1997 2011-01-16 21:43:15 <ArtForz> and most cheat by using onboard PCIe bridges
1998 2011-01-16 21:43:57 <ArtForz> = you still only have 32 lanes to chipset, connected to 2 bridges doing x16 -> x16 + x16
1999 2011-01-16 21:44:20 <tcatm> btw, with fans at 100% on 5970 I can't hear any whining
2000 2011-01-16 21:44:28 <ArtForz> weird
2001 2011-01-16 21:44:43 <ArtForz> might have something to do with me pretty much force-feeding the cards with case fans
2002 2011-01-16 21:45:23 <ArtForz> = I have 2 scythe ultra kazes blowing into a pretty much airtight compartment containing the GPUs
2003 2011-01-16 21:45:24 <tcatm> My setup is similiar with one fan for 2 cards
2004 2011-01-16 21:45:40 <tcatm> + cardboard air channel
2005 2011-01-16 21:45:55 <ArtForz> yep, cardboard = teh awesum
2006 2011-01-16 21:46:06 <ArtForz> cheap, easy to customize, sturdy enough
2007 2011-01-16 21:46:16 <ArtForz> and you usually don't get your computers wet :P
2008 2011-01-16 21:46:38 <nevezen> more combustible though?
2009 2011-01-16 21:46:40 <nevezen> cardboard..
2010 2011-01-16 21:47:08 <ArtForz> dude, if your cards are getting hot enough to ignite cardboard, you have other problems...
2011 2011-01-16 21:47:59 <ArtForz> not to mention the whole thing is usually still inside a steel chassis
2012 2011-01-16 21:48:11 <tcatm> mhm there's a faint sine-like noise at 2.7 kHz
2013 2011-01-16 21:48:29 <ArtForz> if it's only at 2.7kHz, it *is* a sine ;)
2014 2011-01-16 21:48:54 <ArtForz> if it's anything else, it's at 2.7k + harmonics
2015 2011-01-16 21:49:04 <ArtForz> *ducks*
2016 2011-01-16 21:49:30 <tcatm> that's why I wrote -like
2017 2011-01-16 21:49:53 <tcatm> there might be more (and probably is) but I can't see it
2018 2011-01-16 21:50:31 <ArtForz> probably fan noise
2019 2011-01-16 21:50:59 <ArtForz> blower fan + fins = the makings of a siren
2020 2011-01-16 21:51:00 <tcatm> http://bitcoincharts.com/media/stuff/noise.png
2021 2011-01-16 21:51:34 <ArtForz> thats pretty damn faint I'd say
2022 2011-01-16 21:51:48 <ArtForz> < 60Hz hum
2023 2011-01-16 21:52:05 <ArtForz> and you have something at ~ 500Hz
2024 2011-01-16 21:52:13 <sipa1024> ArtForz: to be even more formal, if you're interpreting his sentence as "looking *only* at 2.7 kHz, ...", there can't be anything but a (shifted) sine :p
2025 2011-01-16 21:52:22 <tcatm> probably resonance
2026 2011-01-16 21:52:37 <tcatm> or the big fans infront of the cards
2027 2011-01-16 21:52:53 <ArtForz> it's probably not fan motor noise
2028 2011-01-16 21:53:39 <ArtForz> that tends to be at RPM/(some multiple of 4)
2029 2011-01-16 21:54:22 <ArtForz> err... the inverse
2030 2011-01-16 21:54:40 <ArtForz> rpm / 60 * some multiple of 4
2031 2011-01-16 21:55:04 <tcatm> http://bitcoincharts.com/media/stuff/noise2.png
2032 2011-01-16 21:55:22 <ArtForz> while blade noise tends to be at rpm / 60 * some multiple of #blades
2033 2011-01-16 21:55:41 <ArtForz> unless you have weird stuff like helmholtz resonance in an air channel
2034 2011-01-16 21:57:41 <ArtForz> the ~350Hz is probably fan motor
2035 2011-01-16 21:57:42 <tcatm> the noise around 7500ms is strange. it's periodic
2036 2011-01-16 21:57:59 <ArtForz> weird
2037 2011-01-16 21:58:25 <ArtForz> fan controller?
2038 2011-01-16 21:59:19 <tcatm> no, just plain 12V from PSU
2039 2011-01-16 21:59:42 <ArtForz> still no clue, but I think I saw virgin mary in there...
2040 2011-01-16 22:00:23 <ArtForz> and whats with the ~2.8kHz noise periodically dropping out
2041 2011-01-16 22:00:42 <ArtForz> and shifting in pitch
2042 2011-01-16 22:01:07 <tcatm> maybe PSU having problems keeping 12V stable?
2043 2011-01-16 22:01:34 <ArtForz> what would a PSU be doing at 2.8kHz ?
2044 2011-01-16 22:02:43 InitBrain has joined
2045 2011-01-16 22:02:46 <ArtForz> could be a subharmonic
2046 2011-01-16 22:03:42 <InitBrain> plop all ^^
2047 2011-01-16 22:03:54 <ArtForz> hmmm
2048 2011-01-16 22:04:08 <ArtForz> or a harmonic of whats at ~1375
2049 2011-01-16 22:04:10 <tcatm> mhm I should get better equipment. some sounds could be from my laptop (used onboard mic)
2050 2011-01-16 22:05:03 <ArtForz> shrug
2051 2011-01-16 22:05:08 <ArtForz> doesnt look too bad really
2052 2011-01-16 22:05:23 <ArtForz> but it *could* be the laptop itself
2053 2011-01-16 22:05:37 <ArtForz> CFL inverter anyone?
2054 2011-01-16 22:05:55 <tcatm> maybe
2055 2011-01-16 22:06:43 <tcatm> at least there are no signs of dying fans in the spectrum
2056 2011-01-16 22:06:47 <ArtForz> nope
2057 2011-01-16 22:07:05 <ArtForz> but looking at it, that could really be a CFL inverter
2058 2011-01-16 22:07:56 <ArtForz> they're generally resonant topologies, regulation loooks a lot like frequency shifting
2059 2011-01-16 22:08:19 lyspooner has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.6/20100625231939])
2060 2011-01-16 22:08:29 <ArtForz> at least it does with my guinea pig
2061 2011-01-16 22:10:06 <lfm> does laptop use ccfl backlight?
2062 2011-01-16 22:10:14 <ArtForz> normally, yes
2063 2011-01-16 22:10:43 <tcatm> Anyway. this http://pastebin.com/rTFENpXf looks way better than 70% of them at >92C
2064 2011-01-16 22:10:49 <ArtForz> ones with LED backlights usually say so on a Huge Ugly Sticker(tm)
2065 2011-01-16 22:10:57 <lfm> hehe
2066 2011-01-16 22:11:14 <ArtForz> yep
2067 2011-01-16 22:11:50 <ArtForz> I should record my papst fans spinning up to 100% compared to a 5970 :P
2068 2011-01-16 22:12:23 <andrew12_> It is way too cold outside
2069 2011-01-16 22:12:34 <andrew12_> My toes are freezing
2070 2011-01-16 22:12:51 <ArtForz> needs more cowbe... GPUs
2071 2011-01-16 22:13:18 <lfm> nice toes warmers
2072 2011-01-16 22:13:22 <ArtForz> it's a nice cozy 26°C in here now
2073 2011-01-16 22:13:36 <andrew12_> I need some :P
2074 2011-01-16 22:13:48 <ArtForz> luckly it went well below freezing outside, so hopefully temps will go down a bit overnight
2075 2011-01-16 22:14:12 <andrew12_> I just finished my boy scouts meeting... we were outside for some reason.
2076 2011-01-16 22:14:30 <andrew12_> I can't feel my toes
2077 2011-01-16 22:14:36 <andrew12_> And my dad doesn't care
2078 2011-01-16 22:16:37 <andrew12_> But whatever. We're finally leaving
2079 2011-01-16 22:17:53 <tcatm> slush: ping?
2080 2011-01-16 22:18:35 <slush> tcatm: pong
2081 2011-01-16 22:18:52 <tcatm> your stats are wrong/lagging behind
2082 2011-01-16 22:19:05 <slush> how so?
2083 2011-01-16 22:19:30 <tcatm> My miner reports accepted shares that don't show in my profile
2084 2011-01-16 22:20:01 <slush> oh, you don't think CDF, right?
2085 2011-01-16 22:20:27 <tcatm> oh wait there might be something wrong with my miner
2086 2011-01-16 22:20:28 <slush> I don't see anything wrong
2087 2011-01-16 22:20:53 <slush> ok, I have console open. Tell me when miner find a share
2088 2011-01-16 22:23:26 <andrew12_> Miners gonna mine?
2089 2011-01-16 22:23:42 <andrew12_> :p
2090 2011-01-16 22:24:18 <slush> tcatm: well, you are not using newest m0mchil miner, right?
2091 2011-01-16 22:24:22 <tcatm> now they only report invalid/stale
2092 2011-01-16 22:24:23 <slush> tcatm: you have fork...
2093 2011-01-16 22:24:26 <slush> yes, I see
2094 2011-01-16 22:24:29 <tcatm> slush: yep
2095 2011-01-16 22:24:35 Zarutian has quit (Quit: Zarutian)
2096 2011-01-16 22:24:55 <slush> m0mchil removed feature with increasing ntime
2097 2011-01-16 22:25:03 <slush> which was problematic with pool
2098 2011-01-16 22:25:12 <slush> see diff at 5.January
2099 2011-01-16 22:25:30 devon_hillard has joined
2100 2011-01-16 22:25:42 <Diablo-D3> slush: I never added it for that reason
2101 2011-01-16 22:25:52 <Diablo-D3> it'd fuck with clients that serve a lot of requests AND
2102 2011-01-16 22:25:57 <Diablo-D3> it costs nothing to make another getwork
2103 2011-01-16 22:26:07 <Diablo-D3> (of course, via pool, now it does, and I dont care)
2104 2011-01-16 22:26:08 <slush> yep
2105 2011-01-16 22:26:22 andrew12_ has quit (Quit: parting is such sweet sorrow...)
2106 2011-01-16 22:26:29 <Diablo-D3> btw, did me adding zlib support help much?
2107 2011-01-16 22:26:30 <slush> I now accept only shares which don't touch ntime
2108 2011-01-16 22:26:50 <slush> I don't think so, because most miners are still on old version
2109 2011-01-16 22:27:00 <slush> I plan to announce change when I fix m0mchil's for gzip, too
2110 2011-01-16 22:27:17 <slush> and jgarzik's supports gzip already
2111 2011-01-16 22:27:45 <Diablo-D3> yeah
2112 2011-01-16 22:27:53 <Diablo-D3> I mentioned in my thread, but meh
2113 2011-01-16 22:27:59 <lfm> slush why not allow ntime touch
2114 2011-01-16 22:29:13 <slush> lfm: Potential security issue and possibility to crunching same range by more miners
2115 2011-01-16 22:29:40 <slush> getwork is unique only when central instance distribute them. When miner can change ntime, it makes duplicities
2116 2011-01-16 22:29:59 <ArtForz> huh?
2117 2011-01-16 22:30:21 <slush> well, we worked on this with m0m many many hours
2118 2011-01-16 22:30:30 <ArtForz> you hand out work with same bnExtraNonce and different nTime to different miners?
2119 2011-01-16 22:30:51 <sipa1024> bnExtraNonce = coinbase?
2120 2011-01-16 22:31:01 <ArtForz> yep
2121 2011-01-16 22:31:15 <ArtForz> yeah, *same coinbase
2122 2011-01-16 22:31:16 <slush> ArtForz: yes, sometimes
2123 2011-01-16 22:31:22 <sipa1024> why?
2124 2011-01-16 22:31:31 <slush> its how getwork works
2125 2011-01-16 22:31:41 <sipa1024> include the username in the coinbase
2126 2011-01-16 22:31:45 <sipa1024> and it's all unique
2127 2011-01-16 22:31:55 <slush> sipa1024: wow, great idea!
2128 2011-01-16 22:32:01 <slush> but I need fork bitcoin ;) ;)
2129 2011-01-16 22:32:22 <sipa1024> yes, you'll need to patch it for that, but it doesn't sound that hard
2130 2011-01-16 22:32:24 <slush> (how I can express sarcasm)
2131 2011-01-16 22:32:47 <slush> sipa1024: Today I described this idea here already
2132 2011-01-16 22:32:56 <sipa1024> oh ok
2133 2011-01-16 22:33:04 <lfm> getwork is broken then
2134 2011-01-16 22:33:08 <tcatm> slush: backported changes
2135 2011-01-16 22:34:09 <slush> When miner receive getwork (with given merkle) and modify it's time, there is some possibility that the same job was given to another worker
2136 2011-01-16 22:34:22 <slush> When is whole thing driven by bitcoin, it is not possible
2137 2011-01-16 22:34:33 <lfm> each worker should have unique extranonce
2138 2011-01-16 22:34:41 <slush> it's because extranonce is resetting after crossing 128
2139 2011-01-16 22:34:47 <slush> I kno
2140 2011-01-16 22:34:49 <slush> I know
2141 2011-01-16 22:35:10 <lfm> so getwork in bitcoin is broke, fix it
2142 2011-01-16 22:35:27 <tcatm> getwork is the wrong approach anyway
2143 2011-01-16 22:35:43 <slush> I plan to hack bitcoind to get source data and calculate merkle hash in my app
2144 2011-01-16 22:36:09 <tcatm> 1. add multithreaded RPC server, 2. rewrite getwork to use raw TCP connection for push after first getwork
2145 2011-01-16 22:36:10 <slush> then I can encode worker login to extranonce and use push based protocol instead of getwork
2146 2011-01-16 22:36:38 <slush> every miner can crunch on same job by increasing ntime and nonce, until he receive merkle update from pool
2147 2011-01-16 22:37:57 <slush> tcatm: I will go around C code and bitcoind as much as possible.
2148 2011-01-16 22:39:16 <tcatm> so you prefer complicated workarounds?
2149 2011-01-16 22:39:50 <slush> workarounds? Is pushing worker login to bitcoind more clear way?
2150 2011-01-16 22:40:00 <slush> ...to calculate merkle hash in bitcoin
2151 2011-01-16 22:40:19 <slush> btw I think there is already patch which get merkle from bitcoin
2152 2011-01-16 22:40:39 <tcatm> you need to get work pushed from bitcoind before you can start pushing to miners
2153 2011-01-16 22:41:16 <slush> afaik I only need merkle to construct all data for miners
2154 2011-01-16 22:41:49 <tcatm> yes but how do you know when to push work?
2155 2011-01-16 22:41:52 <slush> so combining active notification about new block/tx + retrieving actual merkle is good way, imho
2156 2011-01-16 22:42:50 <slush> tcatm: Now I'm calling getwork 110x per second. In the worst case I can poll bitcoind 10x per second to check if there is new merkle. Uglier than active notifications, but will work
2157 2011-01-16 22:43:16 <sipa1024> the getwork() implementation keeps a static int with the extra nonce in it
2158 2011-01-16 22:43:44 <slush> yes and I replace extra nonce with unique number for worker
2159 2011-01-16 22:44:32 <slush> sipa1024: pool workers are not talking directly with bitcoind, of course. I have app in the middle
2160 2011-01-16 22:45:24 <sipa1024> slush: i assumed so :)
2161 2011-01-16 22:46:43 <tcatm> Problems communicating with bitcoin RPC
2162 2011-01-16 22:46:56 <tcatm> + some invalid/stales
2163 2011-01-16 22:47:23 <slush> tcatm: with upgraded miner?
2164 2011-01-16 22:47:48 <tcatm> yep
2165 2011-01-16 22:48:15 <slush> oh, i see few connection failures also in my miner
2166 2011-01-16 22:48:18 <slush> but now it is oko
2167 2011-01-16 22:48:20 <slush> ok
2168 2011-01-16 22:48:48 <slush> problem on network? Looks like, server is idle
2169 2011-01-16 22:52:23 <luke-jr> https://en.bitcoin.it/wiki/Units <-- useful?
2170 2011-01-16 22:53:47 * kiba is gleeful today
2171 2011-01-16 22:54:51 * sipa1024 thinks the value for TBC of dBTC is wrong
2172 2011-01-16 22:54:56 <luke-jr> kiba: like the reindeer when Rudolph lit the way?
2173 2011-01-16 22:55:30 <luke-jr> sipa1024: wd
2174 2011-01-16 22:55:51 <kiba> I am merely freelancing
2175 2011-01-16 22:55:58 <sipa1024> obviously this was a test
2176 2011-01-16 22:57:10 <luke-jr> fixed
2177 2011-01-16 22:58:31 <tcatm> slush: do you pass all getworks to bitcoind or only those that match the target?
2178 2011-01-16 23:02:35 <slush> now - all. Just because I don't care, it's few requests. I'll make full check with implementing dynamic worker targets
2179 2011-01-16 23:02:39 <slush> I have to go, bye
2180 2011-01-16 23:03:09 <alhazred> is the difficulty on blocks checked to make sure it's the currently agreed on difficulty?
2181 2011-01-16 23:03:38 <alhazred> on new blocks
2182 2011-01-16 23:05:28 <alhazred> just wondering what's stopping me from generating a block with a low difficulty, then it gets accepted into the chain, then all the miners start working on the new chain cos it's now the longest, and it gets confirmed
2183 2011-01-16 23:05:46 <lfm> its cgecked
2184 2011-01-16 23:05:53 <lfm> checked
2185 2011-01-16 23:05:58 <alhazred> right, chur
2186 2011-01-16 23:06:04 <sipa1024> others won't build further on your block
2187 2011-01-16 23:12:15 Hal9001 has joined
2188 2011-01-16 23:16:42 <sipa1024> i wonder what would happen if someone managed to create a transaction with a constant true script
2189 2011-01-16 23:16:48 <sipa1024> and got it into a block
2190 2011-01-16 23:16:55 <sipa1024> anyone could claim it :p
2191 2011-01-16 23:17:03 <tcatm> nobody would
2192 2011-01-16 23:18:14 hippich has joined
2193 2011-01-16 23:18:22 <edcba> why ?
2194 2011-01-16 23:18:50 Hal9001 has quit (Quit: Hal9001)
2195 2011-01-16 23:18:53 <tcatm> 1) you need to notice that TX 2) you need to write a custom TX to claim it
2196 2011-01-16 23:18:53 <lfm> there is lots of easier ways to give away byc
2197 2011-01-16 23:18:57 <lfm> btc
2198 2011-01-16 23:19:06 Hal9001 has joined
2199 2011-01-16 23:19:14 <edcba> that would encourage development :)
2200 2011-01-16 23:19:15 <sipa1024> if the amount was large enough, people would surely do the effort
2201 2011-01-16 23:24:37 Bth8 has quit (Remote host closed the connection)
2202 2011-01-16 23:26:53 Bth8 has joined
2203 2011-01-16 23:26:58 chuck251 has joined
2204 2011-01-16 23:34:12 Hal9001 has quit (Ping timeout: 246 seconds)
2205 2011-01-16 23:34:57 dukeleto has joined
2206 2011-01-16 23:35:37 <dukeleto> hola
2207 2011-01-16 23:36:36 Myckel has quit (Quit: Ik ga weg)
2208 2011-01-16 23:40:00 devon_hillard has quit (Read error: Connection reset by peer)
2209 2011-01-16 23:49:17 darrob has quit (Ping timeout: 240 seconds)
2210 2011-01-16 23:53:28 darrob has joined
2211 2011-01-16 23:55:31 <dukeleto> is bitcoin still trying to transition to git?
2212 2011-01-16 23:55:38 doublec has quit (Quit: Leaving)
2213 2011-01-16 23:55:52 <tcatm> dukeleto: it's mostly finished I think
2214 2011-01-16 23:57:05 INEEDMONEY1 has joined
2215 2011-01-16 23:58:36 <dukeleto> tcatm: where does the git repo live?
2216 2011-01-16 23:59:43 INEEDMONEY has quit (Ping timeout: 260 seconds)
2217 2011-01-16 23:59:59 <tcatm> https://github.com/bitcoin/