1 2010-11-24 00:00:27 <anarchy_> it was actually still pretty relevant it seems :)
  2 2010-11-24 00:20:36 <ColonelPanic1> compiling things sucks
  3 2010-11-24 00:21:33 <eureka^> derp
  4 2010-11-24 00:21:45 <ColonelPanic1> mainly because I don't know how to do it on windows
  5 2010-11-24 00:21:53 <eureka^> yeah, that's a real pisser
  6 2010-11-24 00:21:59 <eureka^> cygwin is horrid :P
  7 2010-11-24 00:22:03 <eureka^> mingw isn't much better
  8 2010-11-24 00:22:05 <ColonelPanic1> im using mingw32
  9 2010-11-24 00:22:14 <ColonelPanic1> ive just never had to compile on windows
 10 2010-11-24 00:22:46 <eureka^> i found it easier to use mingw on linux lol
 11 2010-11-24 00:22:58 <ColonelPanic1> I just use gcc/g++
 12 2010-11-24 00:23:12 <ColonelPanic1> okay, where the hell is this home directory
 13 2010-11-24 00:25:36 <ColonelPanic1> and I have to build a bunch of other shit
 14 2010-11-24 00:25:38 <ColonelPanic1> what is this I don't even
 15 2010-11-24 00:56:14 <ColonelPanic1> I fucking hate windows so much
 16 2010-11-24 01:01:42 <tcatm> don't use it
 17 2010-11-24 01:08:25 <ColonelPanic1> I actually solved that problem
 18 2010-11-24 01:08:34 <ColonelPanic1> now I'm just trying to get poclbm.py to work
 19 2010-11-24 01:08:43 <djoot> v0.7 of Mt. Gox Ticker for gdesklets @ http://djoot.se/mtgox/
 20 2010-11-24 01:09:03 <djoot> more configuration options... see screenshots
 21 2010-11-24 01:10:16 <ColonelPanic1> numpy
 22 2010-11-24 01:42:28 <brocktice> Well, question answered
 23 2010-11-24 01:42:42 <brocktice> You can use both cards for mining even if only one is hooked to displays and being used
 24 2010-11-24 01:42:49 <brocktice> the other just has to be enabled as a separate desktop
 25 2010-11-24 01:43:00 <brocktice> the mouse can wander off the edge of the screen there, but no other harm done as far as I can tell.
 26 2010-11-24 01:43:16 <brocktice> AND the eyefinity 5 card is working beautifully. Compiz, finally!
 27 2010-11-24 01:43:24 <ColonelPanic1> congrats
 28 2010-11-24 01:43:27 <brocktice> thanks
 29 2010-11-24 01:43:38 <brocktice> Anyone want some nVidia cards? :)
 30 2010-11-24 01:43:44 <ColonelPanic1> I'm still working on poclbm.py
 31 2010-11-24 01:43:56 <brocktice> "working on"?
 32 2010-11-24 01:44:14 <ColonelPanic1> as in trying to get it to work
 33 2010-11-24 01:44:24 <ColonelPanic1> compiling dependencies and whatnot
 34 2010-11-24 01:46:38 <brocktice> ah, bummer.
 35 2010-11-24 01:47:48 <ColonelPanic1> yeah, I need to start generating again
 36 2010-11-24 01:48:00 <ColonelPanic1> this is a bitch to set up, though
 37 2010-11-24 01:49:39 <brocktice> Windows?
 38 2010-11-24 01:50:51 <ColonelPanic1> yes
 39 2010-11-24 01:51:00 <ColonelPanic1> well, I'm doing it on ubuntu and windows at the same time
 40 2010-11-24 01:51:07 <brocktice> oh
 41 2010-11-24 01:51:16 <brocktice> I can help with ubuntu, windows not so much
 42 2010-11-24 01:52:22 <ColonelPanic1> pyopencl can't find CL/cl.h
 43 2010-11-24 01:52:23 <ColonelPanic1> hmmm
 44 2010-11-24 01:52:31 <brocktice> when building?
 45 2010-11-24 01:52:46 <brocktice> did you set the configure flags appropriately?
 46 2010-11-24 01:53:20 <tylergillies> is it safe to keep my majority of coins in mtgox?
 47 2010-11-24 01:53:26 <tylergillies> or should i store on localhost?
 48 2010-11-24 01:54:02 <brocktice> tylergillies: I'd say he's pretty trustworthy, no idea what his security is though
 49 2010-11-24 01:54:11 <theymos> I don't trust mtgox enough to store all my coins there.
 50 2010-11-24 01:54:21 <brocktice> he sent me $1000 by ACH to my bank account with no issues.
 51 2010-11-24 01:54:43 <ColonelPanic1> apparently I need the ati stream sdk
 52 2010-11-24 01:54:59 <theymos> Considering the number of bugs MtGox has had, I wouldn't be surprised if there's some massive security flaw.
 53 2010-11-24 01:55:17 <brocktice> ColonelPanic1: yep
 54 2010-11-24 01:56:17 <brocktice> wow even at 100% the fans on these 5770s are pretty quiet
 55 2010-11-24 01:56:51 <ArtForz> yeah, compared to 5870/5970 even a reference 5770 is positively quiet
 56 2010-11-24 01:57:30 <brocktice> ArtForz: also with the water cooling I figured out I can get down to a dull roar and only increase core speed on the 5970s by a few degrees C
 57 2010-11-24 01:57:41 <brocktice> Which will be good when my in-laws are visiting and staying in the adjacent room.
 58 2010-11-24 01:58:02 <brocktice> s/speed/temp/
 59 2010-11-24 01:58:10 <brocktice> sorry, I'm three sheets to the wind
 60 2010-11-24 01:58:18 <brocktice> and still managed to successfully swap video cards
 61 2010-11-24 01:58:23 <tylergillies> anyone know how to change timezone in bitcoin forums?
 62 2010-11-24 01:58:25 <ArtForz> hehe
 63 2010-11-24 01:58:29 <ArtForz> yeah
 64 2010-11-24 01:58:33 <brocktice> no bios flashing right now though
 65 2010-11-24 01:58:43 <brocktice> seems one of the 5770s refuses to overclock!
 66 2010-11-24 01:58:46 <tylergillies> brocktice: that takes skill ;)
 67 2010-11-24 01:58:48 <brocktice> I may have to remedy that
 68 2010-11-24 01:59:03 <brocktice> tylergillies: yeah I fucked it up last time and moved a decimal point, lucky I didn't fry anything.
 69 2010-11-24 01:59:08 <brocktice> I wasn't even drinking, just distracted by my family
 70 2010-11-24 01:59:10 <ArtForz> I have the feeling with a big enough rad you could run a set of 5970s pretty near silent
 71 2010-11-24 01:59:24 <brocktice> ArtForz: I can run 'em pretty much silent.
 72 2010-11-24 01:59:55 <brocktice> ArtForz: I don't like the temps, but my perception is skewed
 73 2010-11-24 02:00:02 <brocktice> They're still great compared to stock air cooling.
 74 2010-11-24 02:00:49 <ArtForz> yep
 75 2010-11-24 02:01:04 <ArtForz> for gaming I'd probably go with accelero air coolers
 76 2010-11-24 02:01:21 <brocktice> ArtForz: not sure you saw it in the scroll but the eyefinity 5 is working well
 77 2010-11-24 02:01:30 <ArtForz> yeah, nice
 78 2010-11-24 02:01:36 <brocktice> and the second card works, I just have to enable a dummy x display on it
 79 2010-11-24 02:01:49 <ArtForz> sounds right
 80 2010-11-24 02:02:00 <ArtForz> the acceleros are pretty quiet, at 100% they're about == to stock cooler at 30%
 81 2010-11-24 02:02:45 <brocktice> 5770 should be faster than a GTS 250 right?
 82 2010-11-24 02:02:51 <brocktice> wow
 83 2010-11-24 02:02:55 <brocktice> I saw on overclock.net a lot of people using the acceleros
 84 2010-11-24 02:02:57 <ArtForz> yeah
 85 2010-11-24 02:03:06 <brocktice> don't some xfx cards come with that preinstalled?
 86 2010-11-24 02:03:14 <ArtForz> sapphires I think
 87 2010-11-24 02:03:16 <brocktice> I'm going to have to try starcraft II here on eyefinity
 88 2010-11-24 02:03:18 <brocktice> not my sapphires.
 89 2010-11-24 02:03:20 <brocktice> maybe some other model?
 90 2010-11-24 02:03:22 <ArtForz> yea
 91 2010-11-24 02:03:24 <ColonelPanic1> alright "ImportError: libOpenCL.so: cannot open shared object file: No such file or directory" but pyopencl just installed fine
 92 2010-11-24 02:03:26 <ArtForz> the xfx cooler imo == crap
 93 2010-11-24 02:03:38 <ArtForz> same blower as on ref, but now in the middle
 94 2010-11-24 02:04:01 <brocktice> ArtForz: oh, great, so the hot air goes back into the case
 95 2010-11-24 02:04:02 <brocktice> brilliant
 96 2010-11-24 02:04:11 <Diablo-D3> brocktice
 97 2010-11-24 02:04:14 <ArtForz> aww, it's only on the 4GB sapphire cards
 98 2010-11-24 02:04:14 <brocktice> Diablo-D3:
 99 2010-11-24 02:04:19 <brocktice> yes?
100 2010-11-24 02:04:41 <Diablo-D3> brocktice: did you test
101 2010-11-24 02:04:48 <brocktice> nope, shall I do that now?
102 2010-11-24 02:04:52 <brocktice> Should I just pull from github?
103 2010-11-24 02:04:56 <Diablo-D3> no use the binary
104 2010-11-24 02:04:57 <brocktice> (and edit the miner file)
105 2010-11-24 02:05:02 <brocktice> -test?
106 2010-11-24 02:05:05 <Diablo-D3> yes
107 2010-11-24 02:05:07 <Diablo-D3> brocktice: also
108 2010-11-24 02:05:09 <Diablo-D3> you dont need to edit
109 2010-11-24 02:05:10 <Diablo-D3> just use -a
110 2010-11-24 02:05:13 <brocktice> ok one moment
111 2010-11-24 02:06:09 <brocktice> what has changed now?
112 2010-11-24 02:08:19 <Diablo-D3> brocktice: something big that shouldnt work
113 2010-11-24 02:08:28 <brocktice> oh, brilliant
114 2010-11-24 02:08:39 <brocktice> can you be more specific?
115 2010-11-24 02:08:55 <Diablo-D3> not until I know its faster on 5xxx
116 2010-11-24 02:09:12 <brocktice> faster than what?
117 2010-11-24 02:09:14 <brocktice> last time?
118 2010-11-24 02:09:16 <Diablo-D3> git head
119 2010-11-24 02:10:20 <brocktice> now my miner is going to start complaining to me because minerd is not running
120 2010-11-24 02:10:24 <brocktice> but I will put up with it
121 2010-11-24 02:10:27 <brocktice> just for you
122 2010-11-24 02:10:33 <brocktice> ooh segfault
123 2010-11-24 02:10:36 <brocktice> nice work
124 2010-11-24 02:12:04 <Diablo-D3> what, really?
125 2010-11-24 02:12:23 <brocktice> Added Cypress (20 CU, 1x vector, local work size of 256)
126 2010-11-24 02:12:24 <brocktice> Segmentation fault
127 2010-11-24 02:12:30 <Diablo-D3> hrrrm.
128 2010-11-24 02:12:41 <Diablo-D3> yet it works on my 4850.
129 2010-11-24 02:13:00 <brocktice> seems to be an issue when it tries to initialize on the second card
130 2010-11-24 02:13:05 <brocktice> so you have a 2+ bug
131 2010-11-24 02:13:25 <brocktice> wtf am I going to do with this GTS250
132 2010-11-24 02:13:52 <Diablo-D3> brocktice: do export DISPLAY=:0.0 and try again
133 2010-11-24 02:14:41 <ne0futur> ier2.ww7.be/william@waisse.org.png
134 2010-11-24 02:14:50 <ne0futur> the WoT of my keyring
135 2010-11-24 02:14:59 <brocktice> ok that runs
136 2010-11-24 02:15:31 <ColonelPanic1> can someone tell me how I'm supposed to direct poclbm.py to libOpenCL.so?
137 2010-11-24 02:15:44 <Diablo-D3> brocktice: is it faster?
138 2010-11-24 02:15:53 <ColonelPanic1> python is moonspeak to me
139 2010-11-24 02:16:00 <brocktice> Diablo-D3: dunno I didn't kill my other miner
140 2010-11-24 02:16:02 <brocktice> hang on
141 2010-11-24 02:16:08 <brocktice> ColonelPanic1: just think of it as pseudocode
142 2010-11-24 02:16:10 <ArtForz> shouldnt it look for OpenCL.dll or something on windows?
143 2010-11-24 02:16:12 <Diablo-D3> ColonelPanic1: you dont direct it
144 2010-11-24 02:16:13 <brocktice> it basically is
145 2010-11-24 02:16:17 <ColonelPanic1> not doing it on windows
146 2010-11-24 02:16:22 <Diablo-D3> ColonelPanic1: its like every other program on windows
147 2010-11-24 02:16:28 <Diablo-D3> ColonelPanic1: are you on ATI?
148 2010-11-24 02:16:31 <ColonelPanic1> yes
149 2010-11-24 02:16:36 <brocktice> ooh that's interestgin
150 2010-11-24 02:16:36 <ne0futur> oups :
151 2010-11-24 02:16:38 <Diablo-D3> install the catalyst driver WITH the opencl runtime
152 2010-11-24 02:16:39 <brocktice> segfault again
153 2010-11-24 02:16:42 <brocktice> even with :0.0
154 2010-11-24 02:16:42 <ColonelPanic1> ubuntu 10.10, radeon 5850
155 2010-11-24 02:16:51 <Diablo-D3> oh
156 2010-11-24 02:16:52 <Diablo-D3> ubuntu?
157 2010-11-24 02:17:01 <brocktice> aand this time it seems to not be segfaulting
158 2010-11-24 02:17:02 <ArtForz> oh, then, set LD_LIBRARY_PATH to include the lib/x86 or x86_64 dir of your ati sdk
159 2010-11-24 02:17:03 <brocktice> that's odd
160 2010-11-24 02:17:03 <ColonelPanic1> im running the catalyst driver and already installed pyopencl
161 2010-11-24 02:17:06 <Diablo-D3> ColonelPanic1: you forgot to download the ati stream sdk
162 2010-11-24 02:17:16 <ColonelPanic1> I have it, just don't know where I set the path
163 2010-11-24 02:17:16 <tylergillies> Bitcoin Miners, because only the elite can make money from smoking hash
164 2010-11-24 02:17:24 <Diablo-D3> ColonelPanic1: [10:17:02] <ArtForz> oh, then, set LD_LIBRARY_PATH to include the lib/x86 or x86_64 dir of your ati sdk
165 2010-11-24 02:17:28 <brocktice> blargh this sucks
166 2010-11-24 02:17:35 <brocktice> 248000 khash/sec
167 2010-11-24 02:17:42 <Diablo-D3> brocktice: so its randomly segfaulting?
168 2010-11-24 02:17:47 <brocktice> yep
169 2010-11-24 02:17:53 <Diablo-D3> brocktice: is it faster than head is?
170 2010-11-24 02:18:02 <brocktice> no I don't think so
171 2010-11-24 02:18:10 <Diablo-D3> is it slower?
172 2010-11-24 02:18:12 <brocktice> before I was getting around 300,000 khash/s each
173 2010-11-24 02:18:16 <brocktice> yes slower
174 2010-11-24 02:18:16 <ne0futur> sorry bad chan
175 2010-11-24 02:18:18 <ColonelPanic1> oh, I need to set an environment variable
176 2010-11-24 02:18:30 <brocktice> (on your miner 300,000)
177 2010-11-24 02:18:37 <Diablo-D3> brocktice: frown.
178 2010-11-24 02:18:50 <brocktice> sorry, them's the breaks
179 2010-11-24 02:19:37 <Diablo-D3> I wonder if its faster on geforce
180 2010-11-24 02:19:51 <ArtForz> well, at least it segfaults faster
181 2010-11-24 02:20:29 <brocktice> heh
182 2010-11-24 02:20:31 <Diablo-D3> heh
183 2010-11-24 02:20:45 <ColonelPanic1> that didn't work, actually
184 2010-11-24 02:20:48 <ColonelPanic1> fff
185 2010-11-24 02:20:59 <Diablo-D3> ColonelPanic1: you're on x86? or x86-64?
186 2010-11-24 02:21:08 <ColonelPanic1> x86, I set it to the right directory
187 2010-11-24 02:21:20 <Diablo-D3> export LD_LIBRARY_PATH=/path/to/ati-stream/ati-stream-sdk-v2.2-lnx64/lib/x86/:$LD_LIBRARY_PATH
188 2010-11-24 02:21:27 <Diablo-D3> (or 2.1 if you're on that)
189 2010-11-24 02:21:43 <Diablo-D3> we recommend 2.1
190 2010-11-24 02:21:55 <brocktice> you're running x86?
191 2010-11-24 02:22:04 <Diablo-D3> ColonelPanic1: and remember, that only sets it in the terminal you're on
192 2010-11-24 02:22:10 <hippich> hey, kiba, check out http://ubitio.us !! =)
193 2010-11-24 02:22:32 <ColonelPanic1> ah, that helped
194 2010-11-24 02:22:38 <ColonelPanic1> now to get jsonrpc installed
195 2010-11-24 02:23:15 <Diablo-D3> ColonelPanic1: http://json-rpc.org/wiki/python-json-rpc
196 2010-11-24 02:23:26 <Diablo-D3> ubuntu still doesnt fucking package it
197 2010-11-24 02:24:59 <Kiba> hippich: saw it already
198 2010-11-24 02:25:03 <ColonelPanic1> this one sounds fun "pyopencl.LogicError: clGetPlatformIDs failed: invalid/unknown error code"
199 2010-11-24 02:25:24 <brocktice> oh
200 2010-11-24 02:25:28 <brocktice> register your sdk
201 2010-11-24 02:25:34 <ArtForz> you forget that tgz that goes into /etc
202 2010-11-24 02:25:38 <brocktice> I just encountered this one an hour ago on my workstation
203 2010-11-24 02:25:49 <ColonelPanic1> wait, wat
204 2010-11-24 02:25:59 <brocktice> djoot: Nice, you added newlines!
205 2010-11-24 02:26:01 <ArtForz> the registration one
206 2010-11-24 02:26:52 <brocktice> I hate to be like this but people really need to RTFM
207 2010-11-24 02:26:58 <ArtForz> icd-registration.tgz or so
208 2010-11-24 02:27:20 <brocktice> and then fgi
209 2010-11-24 02:27:20 <djoot> brocktice: yep, any string as separator really, except for newline, then only newline
210 2010-11-24 02:27:25 <brocktice> and then ask for help
211 2010-11-24 02:27:37 <djoot> my python-fu not very good :)
212 2010-11-24 02:27:43 <Kiba> hippich: now, we need a freakishly easy to use browser extension
213 2010-11-24 02:27:51 <djoot> I prefer to have it on one line though
214 2010-11-24 02:27:58 <Kiba> so that we can pay for stuff
215 2010-11-24 02:28:04 <hippich> yea. like 1-click payments.
216 2010-11-24 02:28:12 <brocktice> djoot: Ah, I have a hard time finding the right bit of info with it all on one line
217 2010-11-24 02:28:13 <Kiba> chromium extension is a good start
218 2010-11-24 02:28:16 <Kiba> but develoopment is slow
219 2010-11-24 02:28:19 <brocktice> all jumbles together
220 2010-11-24 02:29:29 <djoot> brocktice: you can also choose which tags you want to display, if you find high/low/vol superfluous..
221 2010-11-24 02:30:14 <brocktice> ooh nice
222 2010-11-24 02:30:36 <brocktice> djoot: perfect, all I want is buy, sell, and last
223 2010-11-24 02:30:41 <brocktice> nice work
224 2010-11-24 02:30:45 <djoot> thanks
225 2010-11-24 02:30:52 <brocktice> (donated again a minute ago btw)
226 2010-11-24 02:31:13 <djoot> cool, thank you, very generous :)
227 2010-11-24 02:31:16 <brocktice> Surprisingly that's my favorite feature of bitcoin so far
228 2010-11-24 02:31:28 <brocktice> Very easy to donate quickly to things I like\n57007
229 2010-11-24 02:31:36 <Diablo-D3> no
230 2010-11-24 02:31:39 <Diablo-D3> it goes in /
231 2010-11-24 02:31:42 <Diablo-D3> it just makes files in /etc
232 2010-11-24 02:31:46 <brocktice> well, it pust stuff in /et.... yeah
233 2010-11-24 02:32:22 <brocktice> djoot: not so generous, I'm just glad you made it and want to support this and future updates
234 2010-11-24 02:32:38 <Diablo-D3> ColonelPanic1: http://developer.amd.com/Downloads/icd-registration.tgz
235 2010-11-24 02:32:39 <brocktice> People's time is worth a lot and I think that's often disregarded.
236 2010-11-24 02:32:58 <Kiba> ubitio.us is MADE OF AWESOME WIN
237 2010-11-24 02:33:05 <Diablo-D3> ColonelPanic1: download that, cd /; sudo tar zxvf /path/to/that.tgz
238 2010-11-24 02:33:46 <brocktice> niiiice
239 2010-11-24 02:33:55 <brocktice> much better than that other domain name
240 2010-11-24 02:33:59 <brocktice> it was kind of scary
241 2010-11-24 02:34:06 <brocktice> and this one actually loads in my firefox
242 2010-11-24 02:34:12 <hippich> well. it was development server )
243 2010-11-24 02:34:21 <hippich> and now - it found it's name )
244 2010-11-24 02:34:26 <brocktice> sure, fair enough
245 2010-11-24 02:34:27 <brocktice> cute name
246 2010-11-24 02:36:02 <Diablo-D3> so close!
247 2010-11-24 02:36:06 <Kiba> Very nice and direct to the point, hippich
248 2010-11-24 02:36:08 <Kiba> keep it
249 2010-11-24 02:36:25 <hippich> well. at least for one year i will keep it for sure ))))
250 2010-11-24 02:36:31 <Kiba> idea: don't distract the user until he is done downloading.
251 2010-11-24 02:36:33 <hippich> jsut registered domain )
252 2010-11-24 02:36:47 <Kiba> hippich: the front page is direct to the point, no frill
253 2010-11-24 02:36:49 <Kiba> me like.
254 2010-11-24 02:36:55 <hippich> aa)
255 2010-11-24 02:36:56 <hippich> i see )
256 2010-11-24 02:37:18 <hippich> yeah. this service is just to upload file. link to download should send Uploaders.
257 2010-11-24 02:38:04 <xelister> so now getwork is part of officiall bitcoind and no need to manually patch anymore?
258 2010-11-24 02:38:18 <Diablo-D3> xelister: since when?
259 2010-11-24 02:38:37 <Diablo-D3> so close =/
260 2010-11-24 02:38:45 <xelister> r189 | s_nakamoto | 2010-11-23 20:16:36 +0100 (Tue, 23 Nov 2010) | 1 line
261 2010-11-24 02:38:46 <xelister> new getwork
262 2010-11-24 02:39:05 <Diablo-D3> I hate 5xxx now
263 2010-11-24 02:39:08 <Diablo-D3> how dare it be slower
264 2010-11-24 02:39:14 <xelister> how much slower then m0m's
265 2010-11-24 02:39:31 <Diablo-D3> you mean vs my git head
266 2010-11-24 02:39:43 <xelister> yes
267 2010-11-24 02:40:17 <Diablo-D3> brocktice tried it, per 5970 gpu, it gets 243 where head gets 300
268 2010-11-24 02:40:21 <Diablo-D3> le sigj
269 2010-11-24 02:40:24 <Diablo-D3> *sigh
270 2010-11-24 02:41:02 <xelister> newer version of yours miner is slower then previous version of it?
271 2010-11-24 02:41:09 <Diablo-D3> yes.
272 2010-11-24 02:41:10 <Diablo-D3> but
273 2010-11-24 02:41:12 <Diablo-D3> its faster on 4xxx
274 2010-11-24 02:41:18 <Diablo-D3> ;_;
275 2010-11-24 02:41:29 <djoot> brocktice: well, I'm glad you like it, and thanks for the support :)
276 2010-11-24 02:41:49 <brocktice> np
277 2010-11-24 02:42:01 <Diablo-D3> xelister: like I said, on my 4850, it does 77.9 vs 74.5, where the theoretical max is 78.6
278 2010-11-24 02:44:37 <ColonelPanic1> IT WORKS!
279 2010-11-24 02:44:46 <Diablo-D3> xelister: so bleh
280 2010-11-24 02:44:49 <ColonelPanic1> 10800 khash/s
281 2010-11-24 02:44:58 <ColonelPanic1> sorry, 108000*
282 2010-11-24 02:44:58 <Diablo-D3> ColonelPanic1: wow thats slow, wtf are you on?
283 2010-11-24 02:45:04 <Diablo-D3> oh thats a little more like it
284 2010-11-24 02:45:07 <xelister> ColonelPanic1: what card/miner/os ?
285 2010-11-24 02:45:17 <Diablo-D3> sounds like lower end 5xxx hardware
286 2010-11-24 02:45:24 <xelister> is it...
287 2010-11-24 02:45:31 <Diablo-D3> nothing geforce goes that fast
288 2010-11-24 02:45:33 <ColonelPanic1> it's a 5850 mobile with poclbm.py on ubuntu 10.10
289 2010-11-24 02:45:57 <ColonelPanic1> oh shi- I can hear it firing up now
290 2010-11-24 02:46:11 <Diablo-D3> ColonelPanic1: I do not recommend running a miner on a laptop
291 2010-11-24 02:46:19 <Diablo-D3> laptops are not designed to actually be used
292 2010-11-24 02:46:22 <Diablo-D3> they usually burn up
293 2010-11-24 02:46:24 <xelister> is it a bit slower then Radeon HD 5750 ?
294 2010-11-24 02:46:29 <xelister> ah, mobile
295 2010-11-24 02:46:41 <Kiba> so, I seem to have received additional bitcoins
296 2010-11-24 02:46:44 <xelister> yeah I advice against miningon latptops
297 2010-11-24 02:46:53 <ColonelPanic1> really now?
298 2010-11-24 02:46:57 <Kiba> 0.02
299 2010-11-24 02:47:04 <ColonelPanic1> but it's a sweet laptop
300 2010-11-24 02:47:07 <ArtForz> yep, really bad idea imo
301 2010-11-24 02:47:13 <xelister> ColonelPanic1: well you can try it.. just watch it to not overheat
302 2010-11-24 02:47:32 <ArtForz> laptop GPU cooling is mostly a fucking joke
303 2010-11-24 02:47:44 <ColonelPanic1> it's an alienware m15x, figured that would be okay
304 2010-11-24 02:48:00 <ArtForz> which also happens to be the reason we had nvidia bumpgate first on mobile GPUs
305 2010-11-24 02:48:10 <ColonelPanic1> fans were roaring like a motherfucker a second ago, just slowed down though
306 2010-11-24 02:49:00 <Diablo-D3> laptops are shit
307 2010-11-24 02:50:07 <ColonelPanic1> so can I run one instance of this using my GPU, and another with my CPU on the same machine?
308 2010-11-24 02:50:16 <Diablo-D3> dont bother
309 2010-11-24 02:50:23 <Diablo-D3> ColonelPanic1: and no
310 2010-11-24 02:50:24 <ColonelPanic1> :(
311 2010-11-24 02:50:25 <Diablo-D3> thats not how it works
312 2010-11-24 02:50:28 <xelister> Diablo-D3: your miner seems to throw exception with 189 bitcoind
313 2010-11-24 02:50:33 <xelister> at parsing getwork
314 2010-11-24 02:50:36 <Diablo-D3> xelister: yeah screw 189
315 2010-11-24 02:50:47 <Diablo-D3> m0's should fuck up too
316 2010-11-24 02:51:00 <ColonelPanic1> just asking, because I have a choice of which device to use
317 2010-11-24 02:51:07 <ColonelPanic1> so why not both?
318 2010-11-24 02:52:40 <Diablo-D3> dont
319 2010-11-24 02:52:41 <Diablo-D3> only use one
320 2010-11-24 02:52:47 <Diablo-D3> ColonelPanic1: and opencl is slower on the cpu
321 2010-11-24 02:52:55 <Diablo-D3> if you're going to mine on the cpu, just use bitcoin itself
322 2010-11-24 02:52:57 <Diablo-D3> which is a waste of time
323 2010-11-24 02:52:59 <ColonelPanic1> yes, I noticed that
324 2010-11-24 02:53:01 <Diablo-D3> so dont mine on the cpu at all
325 2010-11-24 02:53:12 <xelister> diablooooooooooooooooooooo
326 2010-11-24 02:53:20 <xelister> can you fix your miner for 189
327 2010-11-24 02:53:47 <Diablo-D3> ColonelPanic1: as a comparison, my c2d e8500 uses about 65 watts, it does 2.9 mhash... my radeon 4850 uses about twice that and does 75.
328 2010-11-24 02:53:52 <Diablo-D3> ColonelPanic1: cpu mining is dead
329 2010-11-24 02:54:01 <Diablo-D3> xelister: no, bitcoin needs to fix their getwork
330 2010-11-24 02:54:08 <xelister> lololololol
331 2010-11-24 02:54:17 <xelister> satoshi will fix bitcoind to fit your miner?
332 2010-11-24 02:54:27 <Diablo-D3> xelister: no, he will fix it to fit m0's too
333 2010-11-24 02:54:35 <xelister> String blocks = responseMessage.get("block").getValueAsText();
334 2010-11-24 02:54:37 <Diablo-D3> my miner is only compatible with m0's getwork
335 2010-11-24 02:54:44 <xelister> why is satoshi not in this channel
336 2010-11-24 02:54:59 <xelister> ArtForz: are you satoshi?
337 2010-11-24 02:55:02 <Diablo-D3> xelister: so if satoshi commited a broken getwork that isnt compatible with my or m0's at all
338 2010-11-24 02:55:04 <Diablo-D3> xelister: fuck him
339 2010-11-24 02:55:36 <xelister> bad satoshi.
340 2010-11-24 02:56:02 <Diablo-D3> xelister: I mean, you do realize who wrote the getwork patch, right?
341 2010-11-24 02:56:04 <Diablo-D3> m0 did
342 2010-11-24 02:56:16 <Diablo-D3> so not only is r189 broken with mine, its broken with m0's too
343 2010-11-24 02:56:30 <xelister> I know
344 2010-11-24 02:56:45 <Diablo-D3> xelister: pastebin the getwork output from ./bitcoind getwork
345 2010-11-24 02:57:33 <xelister> nah I will just use 186 then
346 2010-11-24 02:57:38 <xelister> perhaps in fact satoshi will fix
347 2010-11-24 02:57:49 <Diablo-D3> xelister: pastebin the fucking output
348 2010-11-24 02:57:55 <Diablo-D3> I wanna see what he fucked up
349 2010-11-24 02:59:45 <Diablo-D3> xelister: dooo eeet
350 2010-11-24 03:00:45 <xelister> k
351 2010-11-24 03:02:41 <ColonelPanic1> anyone have suggestions for how I should monitor GPU temp on linux?
352 2010-11-24 03:03:04 <ArtForz> aticonfig --odgt ?
353 2010-11-24 03:03:29 <ColonelPanic1> that fails for some reason
354 2010-11-24 03:03:37 <ArtForz> uhhh
355 2010-11-24 03:11:09 <jgarzik> blah.  lost in a maze of ByteReverse().  for example, I cannot figure out why upstream 'getwork' does a ByteReserve on 128-byte pdata, prior to passing to CheckWork()
356 2010-11-24 03:11:40 <xelister> $ ./bitcoind getwork
357 2010-11-24 03:11:42 <xelister> {
358 2010-11-24 03:11:43 <xelister> "midstate" : "fbea2bafccc3fa9c68a1a2663c0d90050c09bd63b2d2bbed43b9d8a0258bb195",
359 2010-11-24 03:11:45 <xelister> "data" : "000000010b0e80982b65b2796ce2285fe7434059c82931f3364c2bea0002e52d00000000544ca9c0120b8955224b2985d9eba010f32367b0a520a31a5d9b6f3705ede5544cec906f1b098b2a00000000000000800000000000000000000000000000000000000000000000000000000000000000000000000000000080020000",
360 2010-11-24 03:11:46 <xelister> "hash1" : "00000000000000000000000000000000000000000000000000000000000000000000008000000000000000000000000000000000000000000000000000010000",
361 2010-11-24 03:11:48 <xelister> "target" : "0000000000000000000000000000000000000000000000002a8b090000000000"
362 2010-11-24 03:11:49 <xelister> }
363 2010-11-24 03:11:51 <xelister> ^- 189
364 2010-11-24 03:12:06 <xelister> jgarzik: yeah 189's getwork by satoshi is some incompatible pile of.. it seems
365 2010-11-24 03:12:39 <jgarzik> I am writing a miner from scratch
366 2010-11-24 03:13:05 <jgarzik> so hopefully no incompatibility problems exist
367 2010-11-24 03:13:07 <xelister> what language?
368 2010-11-24 03:15:31 <jgarzik> C
369 2010-11-24 03:15:50 <ColonelPanic1> ah, something I might understand
370 2010-11-24 03:16:13 <xelister> pure C O_o ?
371 2010-11-24 03:16:18 <jgarzik> yes
372 2010-11-24 03:16:27 <xelister> its sadomaso
373 2010-11-24 03:16:36 <xelister> :)
374 2010-11-24 03:16:49 <ColonelPanic1> I would like to see this
375 2010-11-24 03:16:51 <xelister> why not just c++
376 2010-11-24 03:17:00 <ColonelPanic1> because he wants bare metal
377 2010-11-24 03:17:00 <xelister> ColonelPanic1: btw the bitcoind itself is in c++
378 2010-11-24 03:17:02 <MT`AwAy> why one would want to use C++ ?
379 2010-11-24 03:17:10 <ColonelPanic1> xelister: I knew that?
380 2010-11-24 03:17:13 <gavinandresen> the // comments are nice
381 2010-11-24 03:17:27 <MT`AwAy> gavinandresen: use C99 and not $1?????? crappy ANSI standard
382 2010-11-24 03:17:39 <xelister> MT`AwAy: tons of applications are written in c++ because it is popular, extreamly established/stable/known, well-tooled, and fast-compilable
383 2010-11-24 03:18:03 <xelister> why use C, it will not give any speed advantages
384 2010-11-24 03:18:13 <MT`AwAy> xelister: and tons of stuff that must be fast and in which memory allocations/etc must be closely monitored are written in C
385 2010-11-24 03:18:18 <gavinandresen> I haven't written code in pure C since... umm... 1992?
386 2010-11-24 03:18:18 <xelister> anyway here where 99.99% of time comes from the miner CL program speed
387 2010-11-24 03:18:31 <ColonelPanic1> I write code in C errday
388 2010-11-24 03:18:40 <xelister> MT`AwAy: bullshit, except for kernels and drivers
389 2010-11-24 03:19:01 <jgarzik> and python plugins and perl plugins and core libraries and
390 2010-11-24 03:19:24 <MT`AwAy> xelister: good for me, I'm writing kernels. However whatever language you use should be your own liberty, and you should be able to say "I'll write that in C" without getting bashed :D
391 2010-11-24 03:19:35 <gavinandresen> jgarzik: plain C, open source?  I've been tempted to get GPU mining on my Mac, but haven't bothered to get all the right Java/Python pieces put together...
392 2010-11-24 03:19:38 <xelister> MT`AwAy: actually the oposite, c++ have tools that can track precise allocation of resources etc, c lacks them making such memory managment harder
393 2010-11-24 03:19:45 <ByteCoin> Are you writing a client too?
394 2010-11-24 03:20:01 <xelister> MT`AwAy: Im not bashing anyone :)
395 2010-11-24 03:20:05 <ByteCoin> Or just the mining portion?
396 2010-11-24 03:20:07 <jgarzik> ArtForz has already written a client in Python
397 2010-11-24 03:20:11 <MT`AwAy> xelister: C++ needs those because it does stuff implicitly
398 2010-11-24 03:20:17 <MT`AwAy> in C, if you don't do something, it won't happen
399 2010-11-24 03:20:22 <ByteCoin> I think it's half a client
400 2010-11-24 03:20:22 <jgarzik> not GPU
401 2010-11-24 03:20:25 <xelister> jgarzik: cpu miner?
402 2010-11-24 03:20:32 <jgarzik> yes
403 2010-11-24 03:20:38 <ColonelPanic1> jgarzik: can I get in on this?
404 2010-11-24 03:20:41 <xelister> but cpu miners can never catch up to gpu / opencomputing/opencl ones it seems
405 2010-11-24 03:20:57 <MT`AwAy> xelister: with my CPU miner I do 30~35Mhash/sec, and I'm quite happy with that
406 2010-11-24 03:20:58 <MT`AwAy> :D
407 2010-11-24 03:21:09 <ByteCoin> Perhaps I'm missing something... what's the point of writing another miner?
408 2010-11-24 03:21:13 <MT`AwAy> (not going to reach some people's 6Ghash/sec)
409 2010-11-24 03:21:20 <MT`AwAy> ByteCoin: fun
410 2010-11-24 03:21:26 <xelister> MT`AwAy: 50 usd gfx card does that, while not blocking your PC and eating less W ;)
411 2010-11-24 03:21:45 <xelister> MT`AwAy: what cpus? :)
412 2010-11-24 03:21:55 <MT`AwAy> xelister: I'm using idle cpu power from my servers, which does not block anything and I don't pay for the W anyway
413 2010-11-24 03:22:03 <xelister> ah
414 2010-11-24 03:22:19 <ByteCoin> Oooookkaaaayyyy.... seriously, you're writing a CPU miner for fun? Is that right jgarzik
415 2010-11-24 03:22:32 <MT`AwAy> ByteCoin: you can also write one to understand how the whole thing works
416 2010-11-24 03:22:38 <MT`AwAy> reasons are not important
417 2010-11-24 03:22:48 <jgarzik> ByteCoin: correct
418 2010-11-24 03:23:18 <Kiba> http://ub.yepcorp.com/file/download/4
419 2010-11-24 03:23:26 <Kiba> hmm
420 2010-11-24 03:23:27 <Kiba> wrong link
421 2010-11-24 03:23:29 <ByteCoin> Ok. Thx
422 2010-11-24 03:23:46 <MT`AwAy> I'm writing a bitcoin client for fun, too :p
423 2010-11-24 03:24:15 <MT`AwAy> (and one that won't need wx2.9 or any weird libs, and won't have mining at all, mining will have to be done externally)
424 2010-11-24 03:24:34 <ByteCoin> MT AwAy: Which language?
425 2010-11-24 03:25:17 <MT`AwAy> ByteCoin: C++
426 2010-11-24 03:25:27 <MT`AwAy> (no, I don't do UI stuff in C)
427 2010-11-24 03:25:49 <ByteCoin> Is the wallet handling going to be a separate process to the peer to peer stuff?
428 2010-11-24 03:25:51 <MT`AwAy> basically I want to do something easier to read that current bitcoin client where all the shit is in main.cpp
429 2010-11-24 03:26:02 <MT`AwAy> ByteCoin: yep
430 2010-11-24 03:26:15 <MT`AwAy> if the OS allows it
431 2010-11-24 03:26:30 <Kiba> I would write the bitcoin client in python
432 2010-11-24 03:26:36 <gavinandresen> Mt`Away: design it as a libbitcoin and then an app on top that uses it and you'll have my eternal gratitude....
433 2010-11-24 03:26:37 <ByteCoin> I suggest that the interface between the wallet handler and the p2p portion be TCP
434 2010-11-24 03:26:44 <MT`AwAy> (ie on linux I'm positive I can do it, dunno if windows supports threads without memory sharing
435 2010-11-24 03:26:53 <MT`AwAy> gavinandresen: that's my current app design
436 2010-11-24 03:26:58 <Kiba> can somebody work on chromium extension?
437 2010-11-24 03:27:02 <ArtForz> threads without memoy sharing? like... processes?
438 2010-11-24 03:27:04 <Kiba> I guess not.
439 2010-11-24 03:27:07 <MT`AwAy> it uses a common bitcoin lib, and two frontends (cli & ui)
440 2010-11-24 03:27:17 <MT`AwAy> ArtForz: there's no fork() on win32
441 2010-11-24 03:27:21 <theymos> nanotube: I will shortly move BTC real-time stats to http://blockexplorer.com/q/ . If gribble can follow 301 redirects, no changes should be required.
442 2010-11-24 03:31:04 <xelister> Diablo-D3: pasted
443 2010-11-24 03:31:21 <ColonelPanic1> so is it really not a good idea to GPU mine on my laptop?
444 2010-11-24 03:31:28 <ColonelPanic1> as in, dangerous to my hardware's health
445 2010-11-24 03:32:03 <Diablo-D3> xelister: url?
446 2010-11-24 03:32:44 <Diablo-D3> oh wait
447 2010-11-24 03:32:45 <Diablo-D3> I see it
448 2010-11-24 03:32:46 <Diablo-D3> yeah
449 2010-11-24 03:32:47 <Diablo-D3> wtf
450 2010-11-24 03:32:53 <Diablo-D3> that patch isnt even close to compatible with m0's
451 2010-11-24 03:35:27 <ColonelPanic1> 100Mhash/s seems worth it to me
452 2010-11-24 03:35:56 <xelister> ArtForz: crossair is good PSU?
453 2010-11-24 03:37:38 <Diablo-D3> xelister: you mean corsair
454 2010-11-24 03:37:41 <Diablo-D3> depends on which one
455 2010-11-24 03:38:53 <Diablo-D3> xelister: how many watts you need?
456 2010-11-24 03:39:04 <xelister> that is a question to
457 2010-11-24 03:39:06 <xelister> too
458 2010-11-24 03:39:17 <Diablo-D3> what do you have?
459 2010-11-24 03:42:17 <ArtForz> so far it seems every PSU is more or less crap
460 2010-11-24 03:42:19 <Diablo-D3> because I doubt you need more than 650 watts unless you're going with quite a few GPUs
461 2010-11-24 03:42:21 <Diablo-D3> ArtForz: they are
462 2010-11-24 03:42:31 <Diablo-D3> fucking pc power might as well be out of business
463 2010-11-24 03:42:36 <Diablo-D3> the new silencer IIs are shit
464 2010-11-24 03:43:06 <Diablo-D3> perfect psu == single 12v rail, 20 pounds of heatsink, a 80mm fan, giant caps
465 2010-11-24 03:43:11 <ArtForz> doesnt matter if it's $100 or $250, continous load >80% or so = fail
466 2010-11-24 03:43:20 <Diablo-D3> ergo, the original silencer series
467 2010-11-24 03:43:23 <Diablo-D3> absolute perfect
468 2010-11-24 03:43:39 <ArtForz> the problem is internal temps
469 2010-11-24 03:43:39 <Diablo-D3> but you cant buy them anymore
470 2010-11-24 03:43:45 <Diablo-D3> ArtForz: yes
471 2010-11-24 03:43:57 <Diablo-D3> those fucking giant fans fuck everything up
472 2010-11-24 03:44:01 <ArtForz> yep
473 2010-11-24 03:44:16 <Diablo-D3> you cant increase airflow enough to deal with the fact your heatsinks are now tiny and shit
474 2010-11-24 03:44:28 <Diablo-D3> the AT/ATX PSU design was meant for 80mm fans
475 2010-11-24 03:44:31 <ArtForz> actually the heatsinks are big enough
476 2010-11-24 03:44:40 <ArtForz> but how the f* is the airflow supposed to make a 90??? bend?
477 2010-11-24 03:44:41 <Diablo-D3> yeah but your airflow is now shot
478 2010-11-24 03:44:44 <Diablo-D3> yeah exactly
479 2010-11-24 03:44:58 <Diablo-D3> ArtForz: I think if I have to replace my psu
480 2010-11-24 03:45:04 <Diablo-D3> Im going to turn the psu fan off
481 2010-11-24 03:45:12 <jgarzik> ColonelPanic1: git://github.com/jgarzik/cpuminer.git
482 2010-11-24 03:45:14 <Diablo-D3> my case has a 120 fan thats between the bottom drive cage and the psu
483 2010-11-24 03:45:22 <Diablo-D3> so it blows the air right into the psu and out the case
484 2010-11-24 03:45:24 <jgarzik> ColonelPanic1: it doesn't work yet, so don't get excited.  something is broken
485 2010-11-24 03:45:33 <jgarzik> and I bet it's fscking ByteReverse related
486 2010-11-24 03:45:42 <Diablo-D3> either that, or figure out how to externally mount a 80mm fan
487 2010-11-24 03:45:54 <Diablo-D3> the only way around this
488 2010-11-24 03:45:59 <Diablo-D3> is buy PSUs rated absurdly high
489 2010-11-24 03:46:24 <Diablo-D3> like, 850w when you only use about half that
490 2010-11-24 03:46:56 <Diablo-D3> and make sure its a thermally driven fan
491 2010-11-24 03:47:04 <ArtForz> or something like this: http://www.trcelectronics.com/Meanwell/se-1500-12.shtml ;)
492 2010-11-24 03:47:07 <Diablo-D3> http://www.newegg.com/Product/Product.aspx?Item=N82E16817139009
493 2010-11-24 03:47:20 <Diablo-D3> ArtForz: hah
494 2010-11-24 03:48:36 <ArtForz> yeah, in my current design I went with a ultra kaze supplying the PSU section
495 2010-11-24 03:48:55 <Diablo-D3> you've seen the inside of a p180b right?
496 2010-11-24 03:49:07 <Diablo-D3> psu is in an isolated air chamber with a 4 drive cage
497 2010-11-24 03:49:11 <ArtForz> yep
498 2010-11-24 03:49:15 <Diablo-D3> so that works for me
499 2010-11-24 03:49:19 <Diablo-D3> I hope my psu never dies
500 2010-11-24 03:49:27 <ArtForz> but the fan is comparatively puny
501 2010-11-24 03:50:09 <ColonelPanic1> noticed something very disturbing about my laptop, the fans turn off when I close the lid
502 2010-11-24 03:50:23 <Diablo-D3> ColonelPanic1: thats normalish
503 2010-11-24 03:50:24 <ArtForz> sure, a 5W kaze aint a 30W delta, but it seems to help a lot
504 2010-11-24 03:50:31 <Diablo-D3> ColonelPanic1: fans turn off when you sleep
505 2010-11-24 03:50:36 <ColonelPanic1> Diablo-D3: not when it's still pumping out 100Mhash/s
506 2010-11-24 03:50:36 <Diablo-D3> hrm
507 2010-11-24 03:50:41 <Diablo-D3> http://www.newegg.com/Product/Product.aspx?Item=N82E16817139011
508 2010-11-24 03:50:53 <Diablo-D3> ColonelPanic1: your laptop should never be still awake when the lid is closed
509 2010-11-24 03:51:00 <ColonelPanic1> why not?
510 2010-11-24 03:51:07 <ColonelPanic1> sometimes I just have to close it to temporarily move it
511 2010-11-24 03:51:16 <Diablo-D3> ColonelPanic1: because the keyboard is an important way of venting heat
512 2010-11-24 03:51:19 <ArtForz> = compared to open air, PSU exhaust temps dropped like a rock when I added a closed shroud over the 2 PSUs fed by a ultra kaze 3000
513 2010-11-24 03:51:26 <Diablo-D3> ArtForz: lol
514 2010-11-24 03:51:36 <ColonelPanic1> oh, I thought the giant ports in the front and back were for venting heat
515 2010-11-24 03:52:33 <Diablo-D3> ColonelPanic1: I swear to god they're just for look
516 2010-11-24 03:52:38 <Diablo-D3> laptops are so fucking badly designed
517 2010-11-24 03:52:44 <Diablo-D3> doubly so if its hp/compaq or dell
518 2010-11-24 03:52:56 <ColonelPanic1> I can feel the airflow going through it, man
519 2010-11-24 03:53:04 <Diablo-D3> yeah because of the shitty fans
520 2010-11-24 03:53:06 <Diablo-D3> ColonelPanic1: trust me
521 2010-11-24 03:53:16 <Diablo-D3> do not close the lid while in operation if you dont have lid close->sleep on
522 2010-11-24 03:53:21 <ColonelPanic1> I do, I don't think I'm going to leave this on unattended
523 2010-11-24 03:53:46 <ColonelPanic1> wireless just got disconnected for no apparent reason
524 2010-11-24 03:53:50 <ColonelPanic1> heat related?
525 2010-11-24 03:54:30 <Diablo-D3> could be
526 2010-11-24 03:54:44 <Diablo-D3> http://www.newegg.com/Product/Product.aspx?Item=N82E16817139015
527 2010-11-24 03:54:47 <Diablo-D3> ooh
528 2010-11-24 03:54:56 <ColonelPanic1> thanks for the help, by the way
529 2010-11-24 03:54:56 <Diablo-D3> http://www.newegg.com/Product/Product.aspx?Item=N82E16817139014
530 2010-11-24 03:54:57 <Diablo-D3> ooh!
531 2010-11-24 03:54:59 <Diablo-D3> there
532 2010-11-24 03:55:02 <Diablo-D3> the ultimate PSU
533 2010-11-24 03:55:30 <Diablo-D3> at 400 watts it should last like 20 years
534 2010-11-24 03:56:41 <ArtForz> meh
535 2010-11-24 03:56:52 <Diablo-D3> ArtForz: well
536 2010-11-24 03:57:04 <Diablo-D3> with such overkill
537 2010-11-24 03:57:09 <Diablo-D3> with a thermally driven fan
538 2010-11-24 03:57:13 <Diablo-D3> the fan should never have to turn on
539 2010-11-24 03:58:11 <Diablo-D3> my pc power's 80mm is thermally driven, rarely turns on
540 2010-11-24 03:58:33 <Diablo-D3> 650 actual rated (not max), I use 350 max
541 2010-11-24 03:58:44 <ArtForz> pfff
542 2010-11-24 03:58:56 <Diablo-D3> you know whats depressing?
543 2010-11-24 03:59:05 <Diablo-D3> coursair's ax series is their best
544 2010-11-24 03:59:11 <ArtForz> my homebrew passive can do that
545 2010-11-24 03:59:12 <Diablo-D3> their ax 650 is cheaper than mine
546 2010-11-24 03:59:27 <Diablo-D3> actually, lets do a real comparison
547 2010-11-24 03:59:31 <Diablo-D3> their ax 750 is cheaper than mine
548 2010-11-24 03:59:38 <Diablo-D3> since my max is like 750
549 2010-11-24 04:00:41 <ArtForz> $20 "650W" PSU, replaced heatsinks with beefy alu Ls, bolted the whole thing to a 5" x 10" heatsink from a dead plate amp
550 2010-11-24 04:00:51 <Diablo-D3> hah
551 2010-11-24 04:00:57 <Diablo-D3> ArtForz: well
552 2010-11-24 04:01:00 <Diablo-D3> you know what I wanna see?
553 2010-11-24 04:01:07 <Diablo-D3> radical new psu designs
554 2010-11-24 04:01:11 <ArtForz> it's been working in my media center PC for about 10 years
555 2010-11-24 04:01:34 <Diablo-D3> say
556 2010-11-24 04:01:43 <Diablo-D3> like
557 2010-11-24 04:02:08 <Diablo-D3> a 92mm fan mounted like the 80 is
558 2010-11-24 04:02:23 <ArtForz> can't do it, ATX PSU height is 86mm
559 2010-11-24 04:02:31 <Diablo-D3> you can
560 2010-11-24 04:02:38 <Diablo-D3> take the frame apart
561 2010-11-24 04:02:50 <Diablo-D3> they have 92mm fans that JUST BARELY fit
562 2010-11-24 04:03:15 <Diablo-D3> either that, or use two 60s
563 2010-11-24 04:03:21 <Diablo-D3> and put the power plug in the middle not on the end
564 2010-11-24 04:03:29 <ArtForz> or just use a 80x38 or 80x51 :P
565 2010-11-24 04:03:39 <Diablo-D3> well I wanna maximize airflow
566 2010-11-24 04:04:11 <ArtForz> erm... those are the kind of 80mms you usally see in 2Us
567 2010-11-24 04:04:55 <Diablo-D3> yeah, but the slight problem with 80mm is you cant cover the entire board
568 2010-11-24 04:05:01 <Diablo-D3> you have a 80mm wide channel
569 2010-11-24 04:05:12 <ArtForz> which should be plenty
570 2010-11-24 04:05:18 <Diablo-D3> this is why they mistakenly chose 120/140 top fans
571 2010-11-24 04:05:52 <ArtForz> the only parts that really heat up are the power semiconductors and the main transformer
572 2010-11-24 04:06:05 <Diablo-D3> yes BUT
573 2010-11-24 04:06:09 <Diablo-D3> you now have more room
574 2010-11-24 04:06:34 <ArtForz> btw, if you shave em a bit 2x80 should also fit
575 2010-11-24 04:06:45 <Diablo-D3> eeeh Im not seeing it
576 2010-11-24 04:06:52 <ArtForz> ATX = 150mm width
577 2010-11-24 04:07:00 <Diablo-D3> 80 + 80 = 160
578 2010-11-24 04:07:06 <ArtForz> 92 > 86
579 2010-11-24 04:07:07 <Diablo-D3> about 140 if you take them out of the frames
580 2010-11-24 04:07:12 <ArtForz> yep
581 2010-11-24 04:07:24 <Diablo-D3> oh wait
582 2010-11-24 04:07:25 <ArtForz> so you could put 2 80s in a figure-8-ish common frame
583 2010-11-24 04:07:29 <Diablo-D3> if I put them on the intake side
584 2010-11-24 04:07:35 <ArtForz> yeah
585 2010-11-24 04:07:39 <Diablo-D3> huh
586 2010-11-24 04:07:40 <Diablo-D3> that could work
587 2010-11-24 04:07:48 <Diablo-D3> it'd also suck all the fucking air out of the case
588 2010-11-24 04:07:51 <Diablo-D3> if they go on high
589 2010-11-24 04:08:18 <Diablo-D3> ArtForz: you know what else I wanna see?
590 2010-11-24 04:08:22 <Diablo-D3> a bias screw
591 2010-11-24 04:08:31 <ArtForz> a what?
592 2010-11-24 04:08:38 <Diablo-D3> a knob you turn to make the fans go on earlier
593 2010-11-24 04:08:45 <ArtForz> ahhh...yeah, that would rule
594 2010-11-24 04:08:56 <Diablo-D3> so just turn it to 11, and the fans go full speed 24/7
595 2010-11-24 04:09:12 <ArtForz> because most current designs seem happy running their fan at 60% or so while exhaust air temp is > 70???C
596 2010-11-24 04:09:18 <Diablo-D3> well
597 2010-11-24 04:09:20 <Diablo-D3> its thermally driven
598 2010-11-24 04:09:21 <Diablo-D3> ie
599 2010-11-24 04:09:23 <Diablo-D3> THEIR thermals
600 2010-11-24 04:09:45 <Diablo-D3> a better design would be measure intake air temp and the temp of your psu
601 2010-11-24 04:10:02 <ArtForz> yeah, they're going "power semiconductors are oversized and should be fine at < 100???C heatsink temp"
602 2010-11-24 04:10:14 <Diablo-D3> yeah pretty much
603 2010-11-24 04:10:20 <Diablo-D3> which is why I said overkill the wattage
604 2010-11-24 04:10:31 <Diablo-D3> if your psu design sucks, a 1200 psu is going to last forever at 400 watts
605 2010-11-24 04:10:39 <ArtForz> they are, but you're slow-cooking the damn electrolytics when you keep interior temp at a constant 70???C+
606 2010-11-24 04:11:10 <Diablo-D3> yeah
607 2010-11-24 04:11:21 <Diablo-D3> which is double fucked when you realize you're doing that to shit that actually matters
608 2010-11-24 04:11:31 <Diablo-D3> since if the intake air is hot, its not being vented right
609 2010-11-24 04:11:34 <ArtForz> yep
610 2010-11-24 04:11:42 <Diablo-D3> ergo
611 2010-11-24 04:11:45 <Diablo-D3> 2x 80mm oriented right
612 2010-11-24 04:11:48 <ArtForz> well, ATX case airflow = major WTf in itself
613 2010-11-24 04:11:51 <Diablo-D3> measure intake temp
614 2010-11-24 04:11:53 <Diablo-D3> ArtForz: not quite
615 2010-11-24 04:11:57 <Diablo-D3> well
616 2010-11-24 04:11:59 <Diablo-D3> MINE is
617 2010-11-24 04:12:07 <Diablo-D3> since I'm now venting the two hottest components in the box out the top
618 2010-11-24 04:12:50 <Diablo-D3> if I designed a computer
619 2010-11-24 04:12:55 <Diablo-D3> I think, for one, I'd make it skinnier
620 2010-11-24 04:13:08 <Diablo-D3> make it like 2u width
621 2010-11-24 04:13:09 <ArtForz> I much prefer defined airflow channels
622 2010-11-24 04:13:14 <Diablo-D3> ArtForz: well wait
623 2010-11-24 04:13:18 <Diablo-D3> 2u width
624 2010-11-24 04:13:20 <Diablo-D3> put PSU sideways
625 2010-11-24 04:13:29 <Diablo-D3> take fans off of everything
626 2010-11-24 04:13:32 <Diablo-D3> make everything flat
627 2010-11-24 04:13:38 <Diablo-D3> and make everything one giant air channel
628 2010-11-24 04:13:43 <Diablo-D3> and line the back with fans
629 2010-11-24 04:13:46 <ArtForz> that would also work
630 2010-11-24 04:14:06 <ArtForz> it's pretty much how a lot of industrial power electronics is designed
631 2010-11-24 04:14:36 <Diablo-D3> thats how the first AT basically was
632 2010-11-24 04:14:43 <Diablo-D3> the cpu was infront of the psu
633 2010-11-24 04:14:47 <ArtForz> make the whole power section one big air channel and slap a fan (or half a dozen) at each end
634 2010-11-24 04:14:51 <Diablo-D3> and it had a plain finned heatsink
635 2010-11-24 04:14:58 <Diablo-D3> ArtForz: well
636 2010-11-24 04:15:00 <Diablo-D3> look at 1Us
637 2010-11-24 04:15:07 <Diablo-D3> thats exactly how the psu is
638 2010-11-24 04:15:13 <Diablo-D3> it air channels the whole way from the front
639 2010-11-24 04:15:21 <ArtForz> btw, most P3-era slim desktops were cooled like this
640 2010-11-24 04:15:22 <Diablo-D3> and nothing goes infront of it except maybe a drive
641 2010-11-24 04:15:32 <Diablo-D3> this is how a wii is cooled
642 2010-11-24 04:15:40 <Diablo-D3> the wii is kind of the design you'd want
643 2010-11-24 04:15:46 <Diablo-D3> except its a tad too skinny
644 2010-11-24 04:16:06 <Diablo-D3> needs to be a bit fatter to put 80mm fans and bigger heatsink fins
645 2010-11-24 04:16:13 <Diablo-D3> ArtForz: also
646 2010-11-24 04:16:20 <Diablo-D3> heatpipes could work well here
647 2010-11-24 04:16:44 <ArtForz> yep
648 2010-11-24 04:16:45 <Diablo-D3> heatpipe the hottest components away from everything to the physically top channel
649 2010-11-24 04:17:57 <ArtForz> yep
650 2010-11-24 04:18:07 <Diablo-D3> ArtForz: see
651 2010-11-24 04:18:13 <Diablo-D3> this is why I dont work for the computer indutsry
652 2010-11-24 04:18:21 <Diablo-D3> I'd cure computer cancer and computer aids
653 2010-11-24 04:19:39 <ColonelPanic1> how well will m0mchil's miner work over the internet, not on a LAN?
654 2010-11-24 04:19:47 <ColonelPanic1> or other situations with high latency
655 2010-11-24 04:19:59 <Diablo-D3> ColonelPanic1: not worth it
656 2010-11-24 04:20:07 <theymos> Bitcoin Block Explorer has moved: http://blockexplorer.com/
657 2010-11-24 04:20:08 <ColonelPanic1> but why
658 2010-11-24 04:20:15 <Diablo-D3> ColonelPanic1: the miner goes dead if it cant contact bitcoind
659 2010-11-24 04:20:22 <ArtForz> theymos: great
660 2010-11-24 04:20:54 <nanotube> theymos: 301 redirects shouldn't be a problem... but let's see.
661 2010-11-24 04:20:55 <Kiba> So, I have a crappy website that I attempted to theme.
662 2010-11-24 04:20:55 <nanotube> ;;bc,stats
663 2010-11-24 04:20:59 <gribble> Current Blocks: 93532 | Current Difficulty: 6866.89864897 | Next Difficulty At Block: 94752 | Next Difficulty In: 1220 blocks | Next Difficulty In About: 1 week, 1 day, 8 hours, 17 minutes, and 41 seconds | Next Difficulty Estimate: 6971.07111698
664 2010-11-24 04:22:14 <nanotube> but i guess to reduce latency, i can update the links... everything stays the same except for the domain, theymos ?
665 2010-11-24 04:22:26 <theymos> Yes.
666 2010-11-24 04:22:38 <theymos> Different port, too.
667 2010-11-24 04:23:56 <nanotube> right. :)
668 2010-11-24 04:25:42 <ColonelPanic1> Diablo-D3: hmm, but it does auto-reconnect
669 2010-11-24 04:26:50 <Diablo-D3> ColonelPanic1: yes
670 2010-11-24 04:27:05 <Diablo-D3> ColonelPanic1: but why bother mining if you're going to screw it up
671 2010-11-24 04:28:24 <ColonelPanic1> now it sends all of the completed work back to bitcoind immediately, right?
672 2010-11-24 04:28:28 <nanotube> theymos: thanks for the heads-up, all done. :)
673 2010-11-24 04:28:52 <ColonelPanic1> I was just wondering, because my laptop often leaves home
674 2010-11-24 04:29:20 <Diablo-D3> ColonelPanic1: erm
675 2010-11-24 04:29:25 <Diablo-D3> just run bitcoind on your laptop
676 2010-11-24 04:51:40 <Kiba> economic reasoning is horribly easy to get wrong
677 2010-11-24 04:52:10 <Kiba> yet any fool notice that deflation ain't all that bad once you look at computers
678 2010-11-24 04:53:30 <bd_> the drop in computer prices isn't deflation
679 2010-11-24 04:53:44 <bd_> deflation is when the price for _EVERYTHING_ drops
680 2010-11-24 04:54:16 <bd_> which means people avoid investing, and simply hang on to their money
681 2010-11-24 04:54:22 <bd_> which slows down economic growth
682 2010-11-24 04:55:19 <Kiba> The value of bitcoin bounty rises, increasing the incentives for people to spend their time
683 2010-11-24 04:55:52 <Kiba> but people will offer less money for new bounties.
684 2010-11-24 04:59:50 <Kiba> bd_: I decide that your reasoning is correct.
685 2010-11-24 05:04:21 <ColonelPanic1> wow man
686 2010-11-24 05:06:16 <Kiba> wow, what?
687 2010-11-24 05:06:44 <Kiba> Hmm,
688 2010-11-24 05:06:54 <Kiba> I got this big tome called Man Economy and the State.
689 2010-11-24 05:07:06 <ColonelPanic1> can't get on here through tor...
690 2010-11-24 05:08:47 <nanotube> bd_: by your reasoning, then, people would avoid investing in computers, since they expect the price to drop ?
691 2010-11-24 05:09:50 <Kiba> TIME PREFERENCE THEORY STRIKE
692 2010-11-24 05:15:35 <bd_> nanotube: people expect computers to drop in price, but computers have a high utility anyway
693 2010-11-24 05:15:51 <bd_> alternately, you could say there's a high opportunity cost for not buying them
694 2010-11-24 05:16:01 <bd_> the thing about investments, though, is with price deflation, the value of your investment drops
695 2010-11-24 05:16:18 <bd_> and the only point to making an investment is because you expect the price to go up (and/or interest)
696 2010-11-24 05:16:27 <nanotube> bd_: yes... so what you are saying that in the case of general deflation... people will only buy things with high utility? oh the horror, what will become of the useless crappy trinkets industry!!?
697 2010-11-24 05:16:35 <bd_> by having the price and interest drop on your investment, you might not get a return
698 2010-11-24 05:16:45 <bd_> nanotube: No, I mean that people won't invest in _new_ businesses
699 2010-11-24 05:17:01 <bd_> because they expect the price of that business to drop faster than it makes profit (= a loss for the investor)
700 2010-11-24 05:17:01 <nanotube> they will, if they expect the new business to generate a return in excess of the rate of deflation.
701 2010-11-24 05:17:18 <bd_> nanotube: Right. So the more deflation, the higher that barrier to investment.
702 2010-11-24 05:17:50 <nanotube> yes... but things reach an equilibrium. if there's no investment, fewer goods get created, prices go up
703 2010-11-24 05:17:55 <Kiba> doesn't sound horrible.
704 2010-11-24 05:18:08 <bd_> Sure, eventually. But until it reaches equilibrium, you can have a recession
705 2010-11-24 05:18:09 <Kiba> I mean, inflations would encourage crappy investment
706 2010-11-24 05:18:26 <Kiba> and more risk taking
707 2010-11-24 05:18:38 <Kiba> but deflation would encourage more conservative investment and less investment
708 2010-11-24 05:18:39 <nanotube> bd_: sure you could... and of course, if we have inflation instead of deflation, we never have those, right?
709 2010-11-24 05:19:09 <bd_> nanotube: Sure you can. :) I'm just saying too much of anything can be a bad thing.
710 2010-11-24 05:19:23 <bd_> A small amount of inflation can help encourage investment
711 2010-11-24 05:19:26 <Kiba> when the economy grows, it's not a bad thing.
712 2010-11-24 05:19:38 <Kiba> deflationary economic growth.
713 2010-11-24 05:19:44 <Kiba> which is happening right now.
714 2010-11-24 05:19:49 <nanotube> in general, inflation encourages people to invest, at the margin, in shitty projects, just to try to make up for inflation. then when things go kaboom because there has been overinvestment, you have a crash.
715 2010-11-24 05:19:50 <bd_> A large amount of inflation... well, it can destroy your currency.
716 2010-11-24 05:19:59 <bd_> nanotube: That's when there's too much inflation :)
717 2010-11-24 05:20:10 <Kiba> people are investing right now
718 2010-11-24 05:20:19 <Kiba> but they're not doing capital intensive projects
719 2010-11-24 05:20:25 <Kiba> they're doing things like web services, etc
720 2010-11-24 05:20:30 <nanotube> in deflation, there's a larger barrier to investment, so you shouldn't get that much irrational overinvestment.
721 2010-11-24 05:20:39 <nanotube> so arguably, deflation would be conducive to more stability.
722 2010-11-24 05:20:44 <bd_> But too little inflation can be a bad thing too. Investment groups build up too much money, can't find anywhere to put it, and start investing in crappy projects.
723 2010-11-24 05:21:20 <bd_> Assuming there's still enough inflation for investment to make sense at all
724 2010-11-24 05:21:41 <nanotube> i'd argue that any investment that is caused solely by inflation is superfluous investment.
725 2010-11-24 05:21:49 <Kiba> with deflation, it might make more sense to make smaller investment in less intensive capital projects
726 2010-11-24 05:21:56 <bd_> nanotube: Nothing is caused _solely_ by inflation :)
727 2010-11-24 05:22:08 <Kiba> and you could spread your investment over several businesss
728 2010-11-24 05:22:16 <nanotube> bd_: at the margin, sure there is.
729 2010-11-24 05:22:18 <bd_> The point I'm trying to make is too much of either is a bad thing
730 2010-11-24 05:22:55 <nanotube> bd_: yes, 'too much' of anything, by definition, is bad :) the question is, is an inflationary regime or a deflationary regime more conducive to economic stability
731 2010-11-24 05:23:29 <bd_> Perhaps alternating between the two would be a good idea :)
732 2010-11-24 05:23:41 <nanotube> it is far from a foregone conclusion, imho, that inflation creates more stability than deflation, assuming 'moderate' amounts of both.
733 2010-11-24 05:23:49 <Kiba> let the market work it out
734 2010-11-24 05:24:04 <bd_> Kiba: when you do that bubbles happen :)
735 2010-11-24 05:24:21 <nanotube> i'd argue that with 'moderate' deflation, with a higher threshold return for investment, the economy is more stable.
736 2010-11-24 05:24:43 <Kiba> economic intervention make recovery slower, not faster
737 2010-11-24 05:24:50 <Kiba> plus, they don't know shit.
738 2010-11-24 05:25:01 <bd_> Kiba: People like to say that, but what do you think would happen if they just let all the banks collapse? :/
739 2010-11-24 05:25:17 <Kiba> banks collapses because they practice fractional reserve.
740 2010-11-24 05:25:19 <Kiba> no shit.
741 2010-11-24 05:25:27 <nanotube> bd_: some banks would survive, and take over the niches that the failed banks left.
742 2010-11-24 05:25:42 <bd_> nanotube: But in the meantime, you'd be left with very little lending for a very long time
743 2010-11-24 05:25:58 <bd_> Which is exactly what causes a depression
744 2010-11-24 05:26:00 <Kiba> we have no saving to back it up anyway. It's an illusion.
745 2010-11-24 05:26:10 <bd_> Kiba: All money is an illusion
746 2010-11-24 05:26:12 <nanotube> bd_: well, one might argue that there was /too much/ lending before. :)
747 2010-11-24 05:26:21 <bd_> nanotube: Sure, but too little is quite bad too
748 2010-11-24 05:26:22 <Kiba> I call it reality.
749 2010-11-24 05:26:43 <nanotube> as an analogy, once you are used to breathing pure oxygen, it sucks to go back to normal air. :)
750 2010-11-24 05:27:08 <nanotube> bd_: yes... once the pit is dug, it's best to get out of it slowly.
751 2010-11-24 05:27:22 <Kiba> high saving rate = cheaper interest = greater economic investment
752 2010-11-24 05:27:26 <nanotube> but it would have been better if you didn't dig the pit in the first place.
753 2010-11-24 05:27:44 <bd_> nanotube: Of course. but bubble happen :/
754 2010-11-24 05:27:46 <bd_> bubbles*
755 2010-11-24 05:27:55 <bd_> they've been happening for hundreds of years, and they're not going to stop anytime soon
756 2010-11-24 05:28:21 <nanotube> yes... so what... such is life.
757 2010-11-24 05:28:42 <bd_> I can see why people don't like bank bailouts though
758 2010-11-24 05:29:05 <bd_> it's the bank's fault for making those bad loans, why should we be bailing them out, right?
759 2010-11-24 05:29:23 <Kiba> it's some people's money that was put in bank.
760 2010-11-24 05:29:27 <bd_> But losing too many banks all at once is a huge shock to the economy
761 2010-11-24 05:29:30 <Kiba> but the bank was irresponsible in the first place.
762 2010-11-24 05:29:32 <Kiba> what to do?
763 2010-11-24 05:29:56 <Kiba> but if you bailout the banks, it won't teach any lesson to the savers not to put their money into banks
764 2010-11-24 05:29:59 <bd_> Kiba: Save the banks. It's more important to fix the economy than get revenge.
765 2010-11-24 05:30:15 <bd_> Then put in regulations to fix it
766 2010-11-24 05:30:15 <Kiba> bd_: moral hazard.
767 2010-11-24 05:30:22 <bd_> Or punish the execs directly
768 2010-11-24 05:30:27 <Kiba> regulations will get twisted
769 2010-11-24 05:30:40 <Kiba> people are lazy to check if their bank can be trusted
770 2010-11-24 05:30:42 <bd_> But don't let the banks fail outright, unless you want to see a depression occuring
771 2010-11-24 05:30:51 <Kiba> you fail them, just like any other business.
772 2010-11-24 05:31:06 <bd_> Kiba: People don't check whether their bank can be trusted, because the gov't will repay the individuals
773 2010-11-24 05:32:46 <bd_> it's just the sudden reduction in lending _will_ kill the economy
774 2010-11-24 05:32:46 <Kiba> if your business is based on borrowing money you're dead.
775 2010-11-24 05:32:46 <nanotube> bd_: well, bank bailouts are a rather different discussion from inflationary/deflationary regimes. :) fwiw, i agree that letting a whole bunch of banks fail outright would have been not very wise.
776 2010-11-24 05:32:51 <Kiba> but the problem: structural deficency in employment market.
777 2010-11-24 05:32:52 <bd_> Kiba: Lots of business are based on borrowing money safely, you know?
778 2010-11-24 05:32:59 <nanotube> bd_: punishing investors/managers, possibly by way of letting creditors take over, as in bankruptcy, would probably be a good happy medium.
779 2010-11-24 05:33:13 <bd_> Businesses take out lines of credit to buy inventory, pay them back with the proceeds, etc.
780 2010-11-24 05:34:35 <Kiba> I am a cashflow kind of guy. I don't get loans unless I can pay them off.
781 2010-11-24 05:35:02 <nanotube> right... technically businesses should do the same. :)
782 2010-11-24 05:35:23 <nanotube> but in a free market, there's nothing to stop a business from taking on too much debt - only the due diligence from the creditors themselves.
783 2010-11-24 05:35:50 <Kiba> nanotube: sure, if you think you can grow big.
784 2010-11-24 05:35:57 <Kiba> but me, I don't have that kind of risk tolerance.
785 2010-11-24 05:35:59 <bd_> _some_ debt isn't necessarily a bad idea - the question is if the business can make money faster than the bank could on its own
786 2010-11-24 05:37:14 <Kiba> my process is to generate positive cashflow, save some of them for raining day, and generate some more
787 2010-11-24 05:37:58 <bd_> Sure, it's nice if you can do that :)
788 2010-11-24 05:38:03 <bd_> Some businesses can indeed operate that way
789 2010-11-24 05:38:12 <bd_> Lots of web applications, for example.
790 2010-11-24 05:38:20 <bd_> Capital-intensive businesses have a harder time
791 2010-11-24 05:38:20 <Kiba> bd_: not saying that people should operate my way
792 2010-11-24 05:38:41 <bd_> fair enough, then
793 2010-11-24 05:39:52 <Kiba> what we have here though, is a market totally unregulated.
794 2010-11-24 05:40:15 <bd_> actually, the banking market is very heavily regulated
795 2010-11-24 05:40:23 <bd_> the problem is the banks are good at finding loopholes
796 2010-11-24 05:40:25 <Kiba> bd_: not the bitcoin economy
797 2010-11-24 05:40:32 <bd_> oh, the bitcoin market
798 2010-11-24 05:40:38 <bd_> Actually ,some of the same regulations may still apply
799 2010-11-24 05:40:47 <Kiba> the problem of regulations is unintended consequence and finding a way for devils to run the regulatory agency
800 2010-11-24 05:40:56 <bd_> The bitcoin market's not yet big enough for enforcement to take place
801 2010-11-24 05:41:19 <Kiba> there should be no enforcement. It would be a natural experiment.
802 2010-11-24 05:41:45 <bd_> Kiba: Incidentally, there have been examples of bank crashes in certain online games
803 2010-11-24 05:41:56 <bd_> google for the ginko bank incident in second life
804 2010-11-24 05:42:31 <bd_> basically, since there was no regulation, ginko bank became way too overleveraged, there was a bank run, and, well, you can probably guess the rest
805 2010-11-24 05:43:17 <nanotube> Kiba: some regulations are necessary for the 'free market' to operate well. most notably, those regarding disclosure. free market fails when there's information asymmetry (or plain fraud)
806 2010-11-24 05:43:25 <Kiba> fractional reserve banking
807 2010-11-24 05:43:52 <Diablo-D3> s/fractional/fictional/
808 2010-11-24 05:43:53 <Diablo-D3> dohohoho
809 2010-11-24 05:44:13 <nanotube> fractional reserve is fine - as long as there's full disclosure, and healthy competition.
810 2010-11-24 05:44:30 <bd_> nanotube: and regulations to protect the depositors
811 2010-11-24 05:44:48 <nanotube> as iirc i mentioned before, people can choose between full-reserve, and actually /paying/ the bank for the safekeeping of the currency, or fractional reserve, and getting interest on deposits... or anything in between.
812 2010-11-24 05:45:19 <nanotube> bd_: what kind of regulations to protect depositors, other than full disclosure of what the bank is doing with depositors' money, is needed?
813 2010-11-24 05:45:31 <bd_> nanotube: limitations on leverage, for one
814 2010-11-24 05:45:33 <nanotube> (assuming contract law is in effect)
815 2010-11-24 05:45:41 <bd_> A bank can't lend out _all_ of its deposits
816 2010-11-24 05:46:07 <Kiba> if people are dumb enough to do business with banks that lend out all of its deposits, let them do it.
817 2010-11-24 05:46:15 <bd_> Kiba: people _are_ that dumb
818 2010-11-24 05:46:15 <nanotube> bd_: i'd say no to that. i as a depositor should be able to decide if i want to deposit in bankA, which lends out 80 %, and pays say, 3% interest, and bankB, which lends out 99%, and pays 5%
819 2010-11-24 05:46:23 <bd_> Kiba: This is one of the things that caused the great depression
820 2010-11-24 05:46:33 <nanotube> bd_: as long as depositors are aware of the terms ex-ante, there's no need to regulate reserve requirement.
821 2010-11-24 05:46:43 <bd_> nanotube: People _don't read_ the terms :)
822 2010-11-24 05:47:04 <nanotube> yes... whose fault is that? :)
823 2010-11-24 05:47:08 <bd_> nanotube: Human nature
824 2010-11-24 05:47:17 <bd_> If a single person makes a mistake, it's their own damn fault
825 2010-11-24 05:47:39 <jgarzik> grumble.  we need an easynet
826 2010-11-24 05:47:41 <nanotube> well, 'full disclosure' should be such that the most salient and important bits of information are indeed made salient, rather than buried.
827 2010-11-24 05:47:45 <bd_> If _everyone_ makes the same mistake, it's a flaw in the design of the human brain, and we should probably prevent people from being ruined by it.
828 2010-11-24 05:47:55 <jgarzik> a genesis block that permits easy proof of work for a miner
829 2010-11-24 05:47:56 <bd_> nanotube: People still won't read it
830 2010-11-24 05:48:17 <bd_> Even if you have 99.99% leveraged - THIS IS REALLY DANGEROUS... BUT YOU GET A 200% INTEREST RATE!!!!!!!1!
831 2010-11-24 05:48:22 <bd_> people will buy into it
832 2010-11-24 05:48:33 <Kiba> a sucker is born every minute
833 2010-11-24 05:48:46 <bd_> Kiba: This is why we have laws against fraud
834 2010-11-24 05:48:53 <bd_> And regulations for banking
835 2010-11-24 05:49:00 <Kiba> bd_: but it's not fraud if you state that you're going to do risky things liek that
836 2010-11-24 05:49:18 <Kiba> bd_: but regulations are easy to skirts and the agents are the devil, and regulation cause unintended consequence.
837 2010-11-24 05:49:22 <bd_> Anyway. The point is, the average person makes stupid mistakes like that
838 2010-11-24 05:49:24 <nanotube> bd_: well, it is somewhat a subjective place where you draw the line at "protecting people from their own stupidity" and "letting nature take its course"
839 2010-11-24 05:49:49 <bd_> The regulations are put into place to try to protect people from their stupidity. You can argue whether that's a good thing or not, I suppose.
840 2010-11-24 05:50:23 <nanotube> yes i can. :)
841 2010-11-24 05:50:46 <bd_> Kiba: Regulation isn't easy to skirt. It's really hard to skirt. But banks have a strong motivation to get around them, and the resources to hire the very best lawyers and accountants.
842 2010-11-24 05:50:54 <bd_> It looks easy in retrospect, of course.
843 2010-11-24 05:51:23 <nanotube> bd_: i suppose my pov would be that you protect people from their own stupidity, when that stupidity would harm /someone else/ as well.
844 2010-11-24 05:51:48 <bd_> nanotube: When a large fraction of the population loses their entire savings, it's bad for EVERYONE
845 2010-11-24 05:52:02 <bd_> When loaning stops, and capital-heavy businesses have to cut back sharply, it's bad for EVERYONE
846 2010-11-24 05:52:42 <Kiba> I heard capital markets are hard to do right
847 2010-11-24 05:53:08 <bd_> And when 51% of your voters are hurt, you have a strong incentive to protect them.
848 2010-11-24 05:53:34 <Kiba> President Hoover and Rooselvets spend a lot of money...
849 2010-11-24 05:55:05 <nanotube> bd_: well, that's the current argument. i'm a bit on the fence about it, though. i might go for something like requiring a person to demonstrate sound logical reasoning and financial education by means of taking a test, before allowing them to invest in anything above a certain riskiness threshold.
850 2010-11-24 05:55:27 <bd_> nanotube: Actually, there are some restrictions on high-risk investments, such as startups
851 2010-11-24 05:55:29 <nanotube> that would of course bring with it some logistics problems.
852 2010-11-24 05:55:37 <bd_> google for "institutional investors"
853 2010-11-24 05:55:38 <nanotube> so what they do instead is measure by net worth. :)
854 2010-11-24 05:55:53 <bd_> yep. The idea being it won't hurt you too much if you screw up
855 2010-11-24 05:55:55 <nanotube> yea i know all about that.
856 2010-11-24 05:56:02 <nanotube> s/all/some/ :P
857 2010-11-24 05:56:53 <nanotube> the problem that bugs me is that i, as a relatively-low-net-worth individual, should be able to invest some 'discretionary' portion of my wealth into the riskier investments.
858 2010-11-24 05:57:12 <Kiba> interseting data, bd_
859 2010-11-24 05:57:20 <nanotube> i'm smart enough not to throw it all out there. but i might want to do it with some amount of 'play' money. :)
860 2010-11-24 05:57:43 <bd_> nanotube: Startups aren't interested in your $500 :)
861 2010-11-24 05:57:45 <Kiba> it didn't say it cause collapse of second life economy
862 2010-11-24 05:59:06 <nanotube> bd_: yes, but they may be intrested in $10m, coming from an 'aggregator' of many $500s :)
863 2010-11-24 05:59:15 <nanotube> which is currently impossible, due to regulations
864 2010-11-24 05:59:20 <bd_> nanotube: Those are called mutual funds, I think :)
865 2010-11-24 05:59:45 <Kiba> spectulation is...essentially gambling.
866 2010-11-24 05:59:48 <nanotube> mutual funds can't invest in startups. those are called hedge funds. and hedgefunds are limited to high-net-worth individuals.
867 2010-11-24 06:00:01 <bd_> ah.
868 2010-11-24 06:00:06 <nanotube> or venture capital funds
869 2010-11-24 06:00:07 <bd_> Can mutual funds invest in hedge funds? >.>
870 2010-11-24 06:00:11 <nanotube> same restriction.
871 2010-11-24 06:00:18 <bd_> interesting
872 2010-11-24 06:00:34 <nanotube> bd_: i'm not sure about that last question...
873 2010-11-24 06:00:41 <nanotube> but i suspect the answer is no.
874 2010-11-24 06:00:46 <Kiba> Banks in the 19th century knew that when they lent out their customer's money, a run could and at times was orchestrated against them by rivals.
875 2010-11-24 06:00:52 <Kiba> an interesting comment at the bottom
876 2010-11-24 06:00:56 <Kiba> but mostly crap though
877 2010-11-24 06:01:04 <Kiba> that bit is intersting, the rest was crap
878 2010-11-24 06:01:34 <Kiba> he does make a point about virtual world scarcity doesn't exist though
879 2010-11-24 06:01:52 <Kiba> bitcoin is actually scarce too, but's it's a different kind of scarcity too
880 2010-11-24 06:02:02 <bd_> There's artificial scarcity, and it seems to be enough that some people in SL can make money
881 2010-11-24 06:02:03 <Kiba> faux-scarcity.
882 2010-11-24 06:02:16 <bd_> but I'm not sure if lending makes sense, since SL businesses aren't capital intensive
883 2010-11-24 06:02:21 <bd_> it makes more sense in something like eve
884 2010-11-24 06:02:24 <Kiba> bd_: it make sense to rent lands...but maybe not goods and objects
885 2010-11-24 06:02:53 <Kiba> some players have the time to climb the ranks and then destroy company assets
886 2010-11-24 06:03:20 <Kiba> an example..of...a very good scammers..but we're all vulnerable to that.
887 2010-11-24 06:03:21 <bd_> eve is actually a pretty good example of a mostly unregulated free-market economy, actually
888 2010-11-24 06:03:39 <bd_> Of course ... there is vigilante regulation, of a sort
889 2010-11-24 06:07:33 <Kiba> hmm
890 2010-11-24 06:07:53 <Kiba> it would be very hard to get legislators aligned with the effort to make an economy grows
891 2010-11-24 06:08:11 <Kiba> human sucks.
892 2010-11-24 06:08:32 <bd_> You could try giving them a bonus based on the GDP.... but that just encourages short-term growth spurts, followed by a crash after they get out of office
893 2010-11-24 06:08:43 <nanotube> unfortunately, legislators tend to only be in it for their own profit.
894 2010-11-24 06:08:45 <bd_> So maybe a bonus 10 years later based on the GDP growth over that period
895 2010-11-24 06:09:26 <nanotube> profit == bribes, under the table deals, book deals after the fact, networking in business to get cushy positions afterwards, etc.
896 2010-11-24 06:09:27 <Kiba> sound goods, but there's no such thing as a free lunch
897 2010-11-24 06:10:07 <nanotube> bd_: guess what i'll do if i'm a politician and i get compensated in a bonus based on gdp growth? i'll be printing some serious money. :)
898 2010-11-24 06:10:25 <bd_> GDP growth, adjusted for inflation :)
899 2010-11-24 06:10:25 <nanotube> gdp grew from 10 trillion to 12 trillion. 20 % growth o/ profit
900 2010-11-24 06:10:42 <nanotube> bd_: yea, and how do we measure inflation? it is some... gasp ... government agency, isn't it? :)
901 2010-11-24 06:10:55 <Kiba> oooooooooh. shit.
902 2010-11-24 06:11:03 <bd_> actually, if there's inflation... that makes their bonus worth less, right? :)
903 2010-11-24 06:11:12 <bd_> sounds self-balancing to me
904 2010-11-24 06:11:16 <nanotube> i mean, inflation is already a highly shady figure at best. guess how much shadier it'll be.
905 2010-11-24 06:11:20 <Kiba> if they are close to the printing press, they'll benefit.
906 2010-11-24 06:11:47 <theymos> BBE supports encryption now: https://blockexplorer.com
907 2010-11-24 06:11:54 <bd_> Or you could measure it based on some other currency (let them choose which currency to use when they're elected in)
908 2010-11-24 06:12:09 <bd_> as long as everyone picks different currencies you should be good
909 2010-11-24 06:12:15 <Kiba> ?
910 2010-11-24 06:12:17 <nanotube> bd_: to a point. say you have 0 gdp growth, you get nothing, right? now, say you have 10% growth, just because you inflated. now you get "something". though it may be less valuable now that there's been inflation, it's still bigger than nothing.
911 2010-11-24 06:12:50 <bd_> nanotube: You'd have no growth if you measured based on a foreign currency
912 2010-11-24 06:12:50 <nanotube> theymos: nice. :)
913 2010-11-24 06:12:54 <Kiba> if you decentralize money production, like bitcoin does
914 2010-11-24 06:12:58 <Kiba> then they can't do that
915 2010-11-24 06:13:09 <nanotube> bd_: unless you got in a devaluation war with said foreign currency. :)
916 2010-11-24 06:13:25 <nanotube> bd_: i'm the legislator, i'll be choosing nigerian currency as my measuring stick
917 2010-11-24 06:13:30 <Kiba> then of course, they will manipulate the currency to benefit them
918 2010-11-24 06:13:33 <nanotube> since i expect them to be inflating at 50% per year.
919 2010-11-24 06:13:36 <bd_> hmm, point :)
920 2010-11-24 06:13:42 <Kiba> err
921 2010-11-24 06:13:43 <Kiba> manipulate
922 2010-11-24 06:13:45 <Kiba> the statistics
923 2010-11-24 06:13:49 <Kiba> incentives matter
924 2010-11-24 06:13:54 <Kiba> but incentives are hard to get right
925 2010-11-24 06:14:13 <nanotube> bd_: it's an interesting idea, though :) but there are some implementation problems. :)
926 2010-11-24 06:14:32 <nanotube> bd_: also, considering the fact that most compensation to a politician is not of direct monetary kind, anyway...
927 2010-11-24 06:14:47 <jgarzik> poor testnet.  many hours between blocks.
928 2010-11-24 06:16:00 <bd_> One option would be to forbid politicians from working anywhere else after retirement, and just pay them a fixed percentage of GDP
929 2010-11-24 06:16:08 <bd_> a bit extreme, admittedly
930 2010-11-24 06:16:15 <Kiba> they'll find sneaky way around that
931 2010-11-24 06:16:39 <bd_> Kiba: Nah, they'll just never vote it in in the first place :)
932 2010-11-24 06:17:02 <Kiba> yup
933 2010-11-24 06:17:04 <Kiba> human sucks
934 2010-11-24 06:17:14 <Kiba> we are the devil
935 2010-11-24 06:17:48 <bd_> nah, we're just not built to be able to deal with a society of more than a hundred or so people
936 2010-11-24 06:17:59 <Kiba> true
937 2010-11-24 06:18:05 <bd_> see http://en.wikipedia.org/wiki/Dunbar's_number
938 2010-11-24 06:19:31 <Kiba> well
939 2010-11-24 06:19:33 <Kiba> it's more than that
940 2010-11-24 06:19:43 <Kiba> we don't think inutitively about division of labor
941 2010-11-24 06:19:45 <Kiba> errr
942 2010-11-24 06:19:50 <Kiba> correctly
943 2010-11-24 06:20:06 <Kiba> we got this whole irrational behaviors
944 2010-11-24 06:20:39 <bd_> Right.
945 2010-11-24 06:20:51 <bd_> and then neoclassical economics assumes rational behavior and perfect information :P
946 2010-11-24 06:21:15 <Kiba> if we know everything about the future, we wouldn't need a market
947 2010-11-24 06:23:20 <nanotube> Kiba: we'd still need a market. since people's preferences would still differ, and there would be need for trade.
948 2010-11-24 06:23:37 <nanotube> it just wouldn't be a very volatile market. :)
949 2010-11-24 06:23:54 <Kiba> no volatilty: no fun
950 2010-11-24 06:23:56 <Kiba> no emotions
951 2010-11-24 06:24:04 <Kiba> no strong emotions.
952 2010-11-24 06:24:10 <Kiba> but also no CRUSHING feeling
953 2010-11-24 06:25:52 <nanotube> hehe you can always get your emotions elsewhere.
954 2010-11-24 06:27:08 <Kiba> http://www.youtube.com/watch?v=d0nERTFo-Sk
955 2010-11-24 06:27:11 <Kiba> let rap!
956 2010-11-24 06:59:57 <kermit> how long does it take for a transaction to be confirmed?
957 2010-11-24 07:00:21 <bd_> on average, one confirmation per 10 minutes.
958 2010-11-24 07:00:40 <bd_> different bitcoin acceptors may have different thresholds for how many confirmations they need
959 2010-11-24 08:48:19 <LobsterMan> http://imgur.com/gR9RU
960 2010-11-24 08:49:47 <LobsterMan> http://i.imgur.com/7laLD.jpg
961 2010-11-24 10:12:17 <CIA-106> DiabloMiner: Patrick McFarland master * rf18385e / (2 files in 2 dirs):
962 2010-11-24 10:12:41 <LobsterMan> :]
963 2010-11-24 10:14:35 <Diablo-D3> so meh
964 2010-11-24 10:16:09 <DELTA9> ul#list-nav li {
965 2010-11-24 10:16:10 <DELTA9> display:inline
966 2010-11-24 10:17:09 <Diablo-D3> does this look like vim to you?
967 2010-11-24 10:17:22 <Diablo-D3> LobsterMan: I really need to quit hacking on this
968 2010-11-24 10:17:26 <Diablo-D3> LobsterMan: its not like anyone uses it
969 2010-11-24 10:17:44 <LobsterMan> well you use it
970 2010-11-24 10:17:47 <LobsterMan> isn't that reason enough?
971 2010-11-24 10:17:49 <LobsterMan> or will when it works
972 2010-11-24 10:17:55 <LobsterMan> btw....have you determined if it does work yet>?
973 2010-11-24 10:18:03 <Diablo-D3> yes, I determined that awhile ago
974 2010-11-24 10:18:05 <Diablo-D3> it works fine
975 2010-11-24 10:18:15 <Diablo-D3> its now the fastest miner EVAR
976 2010-11-24 10:18:50 <Diablo-D3> LobsterMan: art says theoretical maximum is 78.6 mhash.... I get 77.9 with the radeon 4xxx hack
977 2010-11-24 10:19:00 <Diablo-D3> LobsterMan: 74.6 without
978 2010-11-24 10:19:36 <LobsterMan> i should give yours another try ^_________^
979 2010-11-24 10:22:06 <joe_1> has bitcoin reached 100$ a coin yet?
980 2010-11-24 10:22:34 <Diablo-D3> joe_1: no
981 2010-11-24 10:22:40 <Diablo-D3> LobsterMan: you know whats hilarious?