1 2010-10-31 01:02:04 <sneak> i lied before
  2 2010-10-31 01:02:09 <sneak> it appears poclbm is cpu-bound on my machine
  3 2010-10-31 01:02:16 <Diablo-D3> sneak: its not
  4 2010-10-31 01:02:26 <sneak> then why is it using 100% cpu?
  5 2010-10-31 01:02:29 <sneak> on one core?
  6 2010-10-31 01:02:30 <Keefe> oh, it's only 26 blocks since the last adjust
  7 2010-10-31 01:02:41 <Diablo-D3> its chewing cpu time due to a really idiotic design in ati's impl
  8 2010-10-31 01:02:52 <Keefe> bitbot's new diff estimate would be pretty worthless at this point
  9 2010-10-31 01:03:28 <Diablo-D3> sneak: it event waits on the buffer read for the kernel output
 10 2010-10-31 01:03:41 <Diablo-D3> sneak: but instead of ati using, say, libevent, its being dickheaded
 11 2010-10-31 01:04:11 <Diablo-D3> sneak: on my box it uses 60% no matter what the hash rate is, effectively
 12 2010-10-31 01:04:42 <Diablo-D3> sneak: otoh, stream sdk 2.1 doesnt have this problem
 13 2010-10-31 01:05:01 <Diablo-D3> but the compiler isnt nearly as good
 14 2010-10-31 01:05:13 <sneak> oh, this is a problem specific to 2.2?
 15 2010-10-31 01:05:16 <Diablo-D3> yeah
 16 2010-10-31 01:05:18 <Diablo-D3> just ignore it
 17 2010-10-31 01:05:22 <Diablo-D3> it doesnt actually cause any problems
 18 2010-10-31 01:06:08 <Diablo-D3> sneak: what cpu do you have?
 19 2010-10-31 01:06:09 <sneak> hrm
 20 2010-10-31 01:06:21 <sneak> c2d E8200 2.66
 21 2010-10-31 01:06:28 <Diablo-D3> I have a e8500 @ 3.16
 22 2010-10-31 01:06:33 <Diablo-D3> so there really is no issue
 23 2010-10-31 01:08:00 <Diablo-D3> sneak: btw, check after doing cpufreq-set -g performance -c 0, and again with -c1
 24 2010-10-31 01:09:37 <Diablo-D3> sneak: btw, my host client is far better engineered, and it still has the same bug
 25 2010-10-31 01:09:48 <Diablo-D3> its clearly in ATI's code
 26 2010-10-31 01:09:49 <sneak> you said it's a bug in the lib, not the client
 27 2010-10-31 01:10:07 <Diablo-D3> see above.
 28 2010-10-31 01:11:44 <sneak> Keefe: even with crossfire disabled, still not seeing above 53% usage on either gpu core
 29 2010-10-31 01:12:35 <sneak> seeing appx 190M/s on one, 180M/sec on the other
 30 2010-10-31 01:15:07 <Diablo-D3> but does ANYONE here have a 5870?
 31 2010-10-31 01:23:35 <XxMalouinxX> guess not
 32 2010-10-31 01:24:29 <Diablo-D3> well, if we knew the speed of that
 33 2010-10-31 01:24:36 <Diablo-D3> we'd know how much a 5970 is actually supposed to be
 34 2010-10-31 01:24:42 <Diablo-D3> and I dont remember art syaing his 5970s did 6xx
 35 2010-10-31 01:27:18 <sneak> he said 300M
 36 2010-10-31 01:27:23 <sneak> but he could have had them misconfigured
 37 2010-10-31 01:28:33 <Diablo-D3> but he has a 5770
 38 2010-10-31 01:28:40 <Diablo-D3> which performs about 1/6th of his 5970
 39 2010-10-31 01:28:44 <Diablo-D3> which is current
 40 2010-10-31 01:28:56 <Diablo-D3> 5770 == 1/3rd of 5870
 41 2010-10-31 01:31:05 <Diablo-D3> s/current/correct/
 42 2010-10-31 01:37:02 <Diablo-D3> fuck
 43 2010-10-31 01:37:06 <Diablo-D3> theres something wrong with my client
 44 2010-10-31 01:37:13 <Diablo-D3> its been mining for the past 20 minutes
 45 2010-10-31 01:37:15 <Diablo-D3> and found nothing
 46 2010-10-31 01:38:43 <MacRohard> try it on the test net maybe
 47 2010-10-31 01:39:06 <Diablo-D3> Im running on my own private test net =P
 48 2010-10-31 01:39:08 <Diablo-D3> diff 1
 49 2010-10-31 01:39:11 <MacRohard> ahh
 50 2010-10-31 01:39:17 <Diablo-D3> fuck.
 51 2010-10-31 01:39:53 <Diablo-D3> pym0 found two just sitting here
 52 2010-10-31 01:41:15 <Diablo-D3> three
 53 2010-10-31 01:41:22 <Diablo-D3> I bet my sha routine is wrong
 54 2010-10-31 01:44:11 <Diablo-D3> found: 61925223, 30/10/2010 22:38
 55 2010-10-31 01:44:13 <Diablo-D3> found: 243648564, 30/10/2010 22:42
 56 2010-10-31 01:44:15 <Diablo-D3> found: 136833986, 30/10/2010 22:43
 57 2010-10-31 01:44:19 <Diablo-D3> well we know m0's works >_>
 58 2010-10-31 02:08:50 <XxMalouinxX> Diablo-D3:  lol
 59 2010-10-31 02:09:55 <XxMalouinxX> Diablo-D3: yes Art already said his radeon 5970 are getting 6xx
 60 2010-10-31 02:10:12 <XxMalouinxX> Diablo-D3: He was saying that each core/heart got 320mhash/s
 61 2010-10-31 02:10:24 <Diablo-D3> goddamnit
 62 2010-10-31 02:18:19 <Diablo-D3> Ive gotten two matches and... bitcoin didnt think they passed
 63 2010-10-31 02:20:56 <Diablo-D3> four, and still no valid matches
 64 2010-10-31 02:20:57 <Diablo-D3> ftw
 65 2010-10-31 02:53:54 <XxMalouinxX> ArtForz: here?
 66 2010-10-31 02:56:32 <Diablo-D3> goddamnit
 67 2010-10-31 02:56:34 <Diablo-D3> I hate python
 68 2010-10-31 02:56:57 <Keefe> is java your favorite?
 69 2010-10-31 02:57:26 <Diablo-D3> that isnt it
 70 2010-10-31 02:57:32 <Diablo-D3> 
 71 2010-10-31 02:57:38 <Diablo-D3> what the fuck does that line _do_
 72 2010-10-31 02:59:35 <Diablo-D3> I thought it took ints 0 through 18, converted output[1] to hex string, and then put in the rest of the ints, 19 through 31
 73 2010-10-31 03:01:02 <Diablo-D3> er, 20 through 31
 74 2010-10-31 03:02:55 <Keefe> i don't really know python
 75 2010-10-31 03:03:26 <nanotube> Diablo-D3: seems like work['block'] is a list. it takes the first 152 elements of the list (0-151), adds the hex-encoded output[1], then adds the rest of the elements in the work['block'] list, starting from element 160 and going until the end.
 76 2010-10-31 03:03:53 <Diablo-D3> yeah
 77 2010-10-31 03:05:42 <Keefe> if generation continues at the average rate of the last 288 blocks, the next diff will be 3896
 78 2010-10-31 03:05:54 <nanotube> mmm
 79 2010-10-31 03:08:43 <nanotube> Diablo-D3: actually, work['block'] is a string, since the pack().encode produces a string. so we're concatenating strings.
 80 2010-10-31 03:08:56 <nanotube> of course, with string being simply a list of chars... the basic idea still stands. :)
 81 2010-10-31 03:09:35 <Diablo-D3> nanotube: yeah, its just hex representation shit
 82 2010-10-31 03:09:41 <nanotube> yea
 83 2010-10-31 03:09:45 <Diablo-D3> "block" : "00000001279301853905419ecece893fb63efff909234992afb1b8ca6c9c6c9a0000000067001fffc88d270d9857372aa65d49ea0df500c4bd99818850f07bf4eae3ba7d4cccec031d00ffff00000000000000800000000000000000000000000000000000000000000000000000000000000000000000000000000080020000",
 84 2010-10-31 03:09:55 <Diablo-D3> which, as you can notice, is little endian
 85 2010-10-31 03:10:06 <Diablo-D3> that last number is 280 not 20000
 86 2010-10-31 03:11:46 <Diablo-D3> er, not 80020000
 87 2010-10-31 03:11:52 <Keefe> when i gen'd my first block, the diff was 713. when i first discovered bitcoin, diff was 352
 88 2010-10-31 03:23:18 <Diablo-D3> 
 89 2010-10-31 03:23:20 <Diablo-D3> hrm
 90 2010-10-31 03:23:22 <Diablo-D3> only the first 16
 91 2010-10-31 03:23:44 <Diablo-D3> er sorry
 92 2010-10-31 03:23:49 <Diablo-D3> skips the first 16
 93 2010-10-31 03:23:51 <Diablo-D3> no wait
 94 2010-10-31 03:23:52 <Diablo-D3> I was right
 95 2010-10-31 03:23:54 <Diablo-D3> only first 16
 96 2010-10-31 03:26:04 <Diablo-D3> meh doesnt matter
 97 2010-10-31 03:27:26 <Diablo-D3> wait no
 98 2010-10-31 03:27:34 <Diablo-D3> it does skip the first 16
 99 2010-10-31 03:31:15 <Diablo-D3> goddamnit, its still not working right!
100 2010-10-31 03:37:02 <Diablo-D3> AHGAHAHKJDAESDTkladst
101 2010-10-31 03:37:03 <Diablo-D3> SATajtawklrtaw
102 2010-10-31 03:37:08 <Diablo-D3> IT WOOOOOOORKS
103 2010-10-31 03:37:11 <Diablo-D3> AHAAAAAAAAHAHAHAHAHHAHAHA
104 2010-10-31 03:37:14 <Diablo-D3> TAKE THAT MOTHERFUCKERS
105 2010-10-31 03:37:15 <Diablo-D3> TAKE
106 2010-10-31 03:37:16 <Diablo-D3> THAT
107 2010-10-31 03:37:17 <Diablo-D3> MOTHER
108 2010-10-31 03:37:18 <Diablo-D3> FUCKERS
109 2010-10-31 03:37:20 <Diablo-D3> UNF
110 2010-10-31 03:38:02 <XxMalouinxX> Diablo-D3:  =)
111 2010-10-31 03:40:01 <Diablo-D3> lawls
112 2010-10-31 03:40:16 <Diablo-D3> it dont
113 2010-10-31 03:41:35 <XxMalouinxX> haha
114 2010-10-31 03:41:53 <Diablo-D3> doesnt fucking surprise me
115 2010-10-31 03:42:19 <XxMalouinxX> one day it will work, and then you gonna say YES ! It work ! FUCKERS !
116 2010-10-31 03:42:25 <XxMalouinxX> and then, 24h later
117 2010-10-31 03:42:40 <XxMalouinxX> "Art, what the fucking X command do, i don't understand why it give me an error"
118 2010-10-31 03:42:54 <Diablo-D3> what I wanna know is
119 2010-10-31 03:42:55 <XxMalouinxX> ;)
120 2010-10-31 03:42:58 <Diablo-D3> why the motherfucking fuck does
121 2010-10-31 03:43:03 <Diablo-D3> numpy endian swap
122 2010-10-31 03:43:07 <Diablo-D3> fucking fucker
123 2010-10-31 03:43:17 <XxMalouinxX> Art didn't come today?
124 2010-10-31 03:44:19 <Diablo-D3> hes probably too busy with h&b
125 2010-10-31 03:44:33 <XxMalouinxX> h&b ?
126 2010-10-31 03:44:39 <Diablo-D3> hookers and blow
127 2010-10-31 03:44:44 <XxMalouinxX> lawl
128 2010-10-31 03:45:27 <XxMalouinxX> btw, you're from USA right ?
129 2010-10-31 03:46:10 <XxMalouinxX> how many day it take to receive a package from : United States Postal Service (1 x 0.30lbs) (First-Class Mail International Package
130 2010-10-31 03:47:34 <Diablo-D3> several
131 2010-10-31 03:47:35 <Diablo-D3> USPS sucks
132 2010-10-31 03:48:03 <XxMalouinxX> fuck
133 2010-10-31 03:48:16 <Diablo-D3> probably 3-5
134 2010-10-31 03:48:25 <XxMalouinxX> ok
135 2010-10-31 03:48:38 <XxMalouinxX> The Dual PSU thing I ordered come from florida with usps
136 2010-10-31 03:50:32 <nanotube> XxMalouinxX: where are you located?
137 2010-10-31 03:50:54 <XxMalouinxX> Canada / Quebec
138 2010-10-31 03:50:59 <XxMalouinxX> ;)
139 2010-10-31 03:51:11 <nanotube> ah well... a few days sounds about right then
140 2010-10-31 03:51:33 <XxMalouinxX> It would be nice, I would hate to receive the radeon 5970 before the dual PSU :p
141 2010-10-31 03:53:43 <Keefe> don't be surprised if it takes weeks to get it across the border
142 2010-10-31 03:54:06 <XxMalouinxX> that would really suck :/
143 2010-10-31 03:54:13 <XxMalouinxX> it's only a cable =(
144 2010-10-31 03:54:16 <Keefe> within the USA, USPS is great, almost always 2 or 3 days
145 2010-10-31 03:54:35 <Keefe> but hit the border and everything changes :(
146 2010-10-31 03:54:58 <Diablo-D3> usps sucks
147 2010-10-31 03:55:04 <Diablo-D3> ups and fedex > usps
148 2010-10-31 03:55:10 <Keefe> i discourage my canadian customers from using standard mail (usps) and recommend fedex ground instead
149 2010-10-31 03:55:16 <XxMalouinxX> performance pcs charged too much for other option
150 2010-10-31 03:55:24 <XxMalouinxX> by too much, I mean 40$ and higher for a cable
151 2010-10-31 03:55:34 <Keefe> yes, good cross-border shipping is expensive
152 2010-10-31 03:55:50 <Diablo-D3> wait, shipping to do canada?
153 2010-10-31 03:56:03 <Keefe> but $40 is a ripoff for just a cable
154 2010-10-31 03:56:04 <XxMalouinxX> If I had bought something of 200$, I would have paid 30-40, but for 9.99$...
155 2010-10-31 03:56:08 <Diablo-D3> canada will just rip open your packages and charge you $50 for having that done
156 2010-10-31 03:56:14 <Keefe> 1 lb by fedex ground to canada should cost about $15
157 2010-10-31 03:56:36 <XxMalouinxX> Wait i'm gonna copy paste the price, they were insane
158 2010-10-31 03:56:47 <Diablo-D3> I fucking hate canada's postal shit
159 2010-10-31 03:56:47 <Keefe> takes 1 week
160 2010-10-31 03:56:52 <Diablo-D3> they can go fuck themselves
161 2010-10-31 03:57:08 <Keefe> from southern california to canada at least
162 2010-10-31 03:58:02 <XxMalouinxX> Its come from florida
163 2010-10-31 03:58:03 <Keefe> maybe that was $40 for fedex express?
164 2010-10-31 03:58:14 <Keefe> makes more sense
165 2010-10-31 03:58:22 <XxMalouinxX> 4.27$ for the usps first class mail inter
166 2010-10-31 03:58:27 <XxMalouinxX> after that come
167 2010-10-31 03:58:32 <XxMalouinxX> 22$ usps priority mail
168 2010-10-31 03:58:34 <Keefe> and that should take just 1 day if you're in a major city
169 2010-10-31 03:58:51 <Keefe> priority mail is a ripoff. no faster than first class
170 2010-10-31 03:58:54 <XxMalouinxX> they don't have fedex, only UPS, and that go from 25$ to 60$
171 2010-10-31 03:59:07 <XxMalouinxX> for something of 0.30lbs
172 2010-10-31 03:59:36 <Keefe> i'm not familiar with ups. it's either usps for cheap lightweight shipping or fedex for guaranteed speed or heavier stuff
173 2010-10-31 03:59:54 <XxMalouinxX> ok
174 2010-10-31 04:00:05 <XxMalouinxX> They shipped it oct. 28
175 2010-10-31 04:00:10 <Diablo-D3> ups and fedex are baiscally identical
176 2010-10-31 04:00:20 <Keefe> ups is usually more expensive
177 2010-10-31 04:00:32 <XxMalouinxX> I like purolator
178 2010-10-31 04:00:36 <Keefe> at least for low-volume shippers like myself
179 2010-10-31 04:02:15 <Keefe> in many cases, usps first-class within the usa is both faster and much cheaper than fedex ground
180 2010-10-31 04:03:06 <XxMalouinxX> I didn't know about the border, I will try to avoid usps when I can, thx :p
181 2010-10-31 04:03:39 <Keefe> for example, i can send a little electronic component by usps first-class from california on a saturday to new york for $1.41 with tracking, and they'll probably get it monday
182 2010-10-31 04:04:06 <Keefe> fedex ground would cost like $6 and not deliver til a whole week later
183 2010-10-31 04:04:12 <XxMalouinxX> well, you can call that fast :p
184 2010-10-31 04:04:53 <Diablo-D3> I suspect in m0's code, numpy isnt swapping
185 2010-10-31 04:04:59 <Diablo-D3> pyopencl is
186 2010-10-31 04:06:12 <Diablo-D3> now, first three rounds should work
187 2010-10-31 04:08:08 <Diablo-D3> meh
188 2010-10-31 04:17:16 <kiba> hmm, somebody brought bitcoins for 0.141 on bcm
189 2010-10-31 04:17:27 <kiba> mtgox's last price is .1933
190 2010-10-31 04:21:06 <XxMalouinxX> Diablo-D3: do the "meh" mean "It failed" ?
191 2010-10-31 04:21:33 <Diablo-D3> yes
192 2010-10-31 04:21:55 <Diablo-D3> its not a big issue
193 2010-10-31 04:22:13 <XxMalouinxX> ok
194 2010-10-31 05:07:02 <Diablo-D3> fuck
195 2010-10-31 05:07:05 <Diablo-D3> it is numpy doing it I think
196 2010-10-31 05:07:16 <Diablo-D3> because I just genned a block using opencl
197 2010-10-31 05:07:19 <Diablo-D3> on x86 I mean
198 2010-10-31 05:07:25 <Diablo-D3> and it worked
199 2010-10-31 05:08:55 <Diablo-D3> radeon == big, x86 == little endian
200 2010-10-31 05:10:49 <XxMalouinxX> lol
201 2010-10-31 05:11:23 <Diablo-D3> so
202 2010-10-31 05:12:01 <Diablo-D3> and I technically should if() that
203 2010-10-31 05:12:13 <Diablo-D3> bitcoind DOESNT byteswap on it's side
204 2010-10-31 05:12:25 <Diablo-D3> but java only parses shit like this as network endian (ie, big)
205 2010-10-31 05:12:29 <XxMalouinxX> Don't give up, COME fucking bastard
206 2010-10-31 05:12:40 <XxMalouinxX> damn, I forget an ON
207 2010-10-31 05:13:00 <XxMalouinxX> thats better, and yeah
208 2010-10-31 05:13:02 <XxMalouinxX> come on me
209 2010-10-31 05:13:10 <Diablo-D3> anyhow, yeah, fucking endian shit
210 2010-10-31 05:13:15 <Diablo-D3> and this isnt even bitcoind's fault
211 2010-10-31 05:13:17 <Diablo-D3> its m0's fault
212 2010-10-31 05:13:22 <Diablo-D3> since he wrote the getwork patch
213 2010-10-31 05:13:22 <XxMalouinxX> lolll
214 2010-10-31 05:13:50 <XxMalouinxX> btw, do you think m0's give wrong hash/s ?
215 2010-10-31 05:14:01 <Diablo-D3> dont think so
216 2010-10-31 05:14:10 <XxMalouinxX> ok
217 2010-10-31 05:14:16 <Diablo-D3> I use a different way of calculating it, and it gives mostly the same numbers
218 2010-10-31 05:14:21 <Diablo-D3> at least, they're in the same ballpark
219 2010-10-31 05:14:24 <Diablo-D3> they're not 2x different
220 2010-10-31 05:14:24 <XxMalouinxX> thats good to know
221 2010-10-31 05:14:55 <XxMalouinxX> and what about usps, did you alread ship to canada with that? is it fast? I WANT MY CABLE !! NOW !
222 2010-10-31 05:15:11 <Diablo-D3> Ive never shipped anything to canada
223 2010-10-31 05:15:17 <Diablo-D3> I just know a few canadians
224 2010-10-31 05:15:22 <XxMalouinxX> ok
225 2010-10-31 06:07:42 <Diablo-D3> http://forums.amd.com/devforum/messageview.cfm?catid=390&threadid=138994&enterthread=y
226 2010-10-31 06:07:44 <Diablo-D3> hmm interesting
227 2010-10-31 06:20:32 <Diablo-D3> goddamnit
228 2010-10-31 06:29:25 <Diablo-D3> hmm
229 2010-10-31 06:29:34 <Diablo-D3> if I set the worksize very low it doesnt go poof
230 2010-10-31 06:37:36 <Diablo-D3> gets a little choppy there
231 2010-10-31 06:37:37 <Diablo-D3> HRM
232 2010-10-31 06:38:16 <Diablo-D3> thats what I was using by default and shit blew up
233 2010-10-31 06:38:42 <joe_1> oh
234 2010-10-31 06:38:46 <Diablo-D3> I suspect Im running out of registers somewhere
235 2010-10-31 06:49:43 <joe_1> what are you working on. i was ctually kidding about trying a local work size of 256 not really sure what you are doing.
236 2010-10-31 06:49:55 <Diablo-D3> gpu miner
237 2010-10-31 06:50:39 <joe_1> yea, i want to deploy miners on random machines somehow and have them send me the bitcoins they generate
238 2010-10-31 06:54:25 <Diablo-D3> heh
239 2010-10-31 07:06:33 <Diablo-D3> well this is just fucking absurd
240 2010-10-31 07:07:04 <joe_1> have you made any coins yet?
241 2010-10-31 07:07:13 <Diablo-D3> fictional ones, yes
242 2010-10-31 07:07:18 <joe_1> o
243 2010-10-31 07:07:27 <joe_1> why not jump on the main network?
244 2010-10-31 07:07:33 <Diablo-D3> Im running this on a diff=1 client
245 2010-10-31 07:07:39 <Diablo-D3> so I can verify it actually does coins
246 2010-10-31 07:07:47 <joe_1> what is a diff=1 client
247 2010-10-31 07:07:57 <Diablo-D3> a client set to do difficulty of 1
248 2010-10-31 07:08:04 <joe_1> oh
249 2010-10-31 07:19:00 <Diablo-D3> boned
250 2010-10-31 07:20:27 <Diablo-D3> well
251 2010-10-31 07:20:29 <Diablo-D3> art should be happy
252 2010-10-31 07:20:41 <Diablo-D3> I broke his 75 mhash estimate
253 2010-10-31 07:22:32 <joe_1> how many blocks a week can you expect with that
254 2010-10-31 07:22:33 <Diablo-D3> Im just not sure by how much
255 2010-10-31 07:22:41 <Diablo-D3> joe_1: not sure
256 2010-10-31 07:22:43 <Diablo-D3> use the calculator
257 2010-10-31 07:24:32 <joe_1> about 10 bucks a month
258 2010-10-31 07:24:43 <Diablo-D3> no, thats not right.
259 2010-10-31 07:25:04 <joe_1> sorry forgot a zero
260 2010-10-31 07:25:07 <joe_1> 100 bucks a month
261 2010-10-31 07:25:13 <Diablo-D3> thats more like it
262 2010-10-31 07:25:28 <joe_1> what are you running
263 2010-10-31 07:25:34 <Diablo-D3> a radeon 4850
264 2010-10-31 07:25:37 <joe_1> im about to go shopping for a new computer tomorrow if this is true
265 2010-10-31 07:25:48 <Diablo-D3> joe_1: now see
266 2010-10-31 07:25:52 <Diablo-D3> if everyone goes and does this
267 2010-10-31 07:25:55 <Diablo-D3> difficulty will shoot up
268 2010-10-31 07:27:28 <joe_1> u should jump on the main network before people do it
269 2010-10-31 07:27:40 <Diablo-D3> PEOPLE ALREADY ARE
270 2010-10-31 07:29:23 <joe_1> how does the 4850 compare with just running it on your regular processor?
271 2010-10-31 07:29:47 <Diablo-D3> my c2d e8500 @ 3.16ghz does 3 mhash
272 2010-10-31 07:29:56 <Diablo-D3> art's estimate for my 4850 was 75
273 2010-10-31 07:30:07 <joe_1> this is insane.... why do gpu's do it so much faster than cpu's?
274 2010-10-31 07:30:25 <Diablo-D3> because they're designed for massively parallel mathematics
275 2010-10-31 07:31:05 <joe_1> im not a big gamer... i might just plug my GPU where my CPU should go, and move my CPU to where my GPU should go.
276 2010-10-31 07:31:27 <Diablo-D3> that dont work that way.
277 2010-10-31 07:36:14 <joe_1> will calculating coins blow out my gpu?
278 2010-10-31 07:36:41 <Diablo-D3> its no worse than playing games
279 2010-10-31 07:57:26 <edcba> but you may mine longer than play
280 2010-10-31 07:57:53 <edcba> and maybe gaming is less intensive for gpu
281 2010-10-31 07:58:05 <Diablo-D3> its not.
282 2010-10-31 07:58:12 <Diablo-D3> gaming uses every square inch of the chip
283 2010-10-31 07:58:21 <edcba> doubt that
284 2010-10-31 07:58:29 <Diablo-D3> remember, its just math
285 2010-10-31 07:58:33 <Diablo-D3> math is math is math
286 2010-10-31 07:58:49 <edcba> games need not running more than 60fps
287 2010-10-31 07:59:02 <Diablo-D3> gpus dont know what fps are.
288 2010-10-31 07:59:31 <edcba> they will heat more if they are fully utilized
289 2010-10-31 07:59:41 <Diablo-D3> that isnt true
290 2010-10-31 07:59:55 <edcba> why is it hotter when i use it then ? :)
291 2010-10-31 08:00:11 <Diablo-D3> if you're using 3D accel functions, it cranks up anyhow
292 2010-10-31 08:00:20 <ArtForz> sorry, wrong
293 2010-10-31 08:00:20 <Diablo-D3> it doesnt measure how much you're using, just that you're using it
294 2010-10-31 08:00:49 <Diablo-D3> ArtForz: btw, you said 75 right?
295 2010-10-31 08:00:50 <edcba> submitting one job every second isn't same than submitting same job 5000 times a second
296 2010-10-31 08:01:03 <Diablo-D3> edcba: except thats not how it works
297 2010-10-31 08:01:07 <ArtForz> 75.3
298 2010-10-31 08:01:16 <Diablo-D3> games submit thousands of jobs a second
299 2010-10-31 08:01:26 <Diablo-D3> I submit at least 60 a second
300 2010-10-31 08:01:40 <edcba> yes but they not necessarily use it at 100%
301 2010-10-31 08:01:44 <Diablo-D3> edcba: and it depends how you define job too; I use instancing in my 3D engine, for example
302 2010-10-31 08:01:58 <Diablo-D3> which requires very little GL->driver->hardware interaction for the bulk of the rendering time spent
303 2010-10-31 08:02:06 <ArtForz> no game stresses a GPU anywhere close to 100%
304 2010-10-31 08:02:09 <Diablo-D3> ArtForz: well, I hit a peak of 135.
305 2010-10-31 08:02:18 <Diablo-D3> no game? lol.
306 2010-10-31 08:02:23 <ArtForz> 135? no way in hell
307 2010-10-31 08:02:53 <Diablo-D3> at 100, diff=1, how long does it take to gen 50 blocks?
308 2010-10-31 08:03:09 <ArtForz> yes, no game. 0. zero. zilch.
309 2010-10-31 08:08:15 <ArtForz> mining stresses a GPU about as much as furmark
310 2010-10-31 08:08:26 <Diablo-D3> ArtForz: ^ math plz
311 2010-10-31 08:08:54 <ArtForz> what do you think furmark does?
312 2010-10-31 08:09:16 <Diablo-D3> [05:02:53] <Diablo-D3> at 100, diff=1, how long does it take to gen 50 blocks?
313 2010-10-31 08:09:49 <ArtForz> 2147s
314 2010-10-31 08:11:03 <Diablo-D3> 36 minutes
315 2010-10-31 08:11:35 <ArtForz> yup
316 2010-10-31 08:15:19 <Diablo-D3> ArtForz: the fun part is, if I run this newly hacked up kernel on a worksize of 256, it basically locks the card up
317 2010-10-31 08:16:13 <Diablo-D3> er, local worksize
318 2010-10-31 08:24:21 <Diablo-D3> ArtForz: maybe I screwed up
319 2010-10-31 08:55:39 <lfm> bitcoin doesnt use the floating point tho
320 2010-10-31 08:57:39 <Diablo-D3> lfm: that doesnt quite matter
321 2010-10-31 08:57:47 <Diablo-D3> the alus are designed optimally
322 2010-10-31 09:00:37 <lfm> but surely there must be some FP/DP circuitry that is unused and thus could be used to stress the gpu further than bitcoin does
323 2010-10-31 09:00:56 <Diablo-D3> there is one part we're not using
324 2010-10-31 09:00:58 <Diablo-D3> texture units
325 2010-10-31 09:01:11 <Diablo-D3> or anything to do with global memory beyond what the hardware itself does
326 2010-10-31 09:02:52 <lfm> so concievably there could be other algorithms that stress it more than bitcoin hashes
327 2010-10-31 09:03:10 <Diablo-D3> yes, such as playing games.
328 2010-10-31 09:03:27 <Diablo-D3> lfm: btw, no hardware out there has DP circuitry
329 2010-10-31 09:03:49 <Diablo-D3> all the GPUs that do DP have ALUs designed to take twice as long
330 2010-10-31 09:04:34 <lfm> ok so they dont have single cycle DP but they may still have circuitry dedicated to DP in some way
331 2010-10-31 09:04:40 <Diablo-D3> slightly yes
332 2010-10-31 09:04:56 <Diablo-D3> both gtx 4xx and radeon 5xxx have medium and low end models that cant do it
333 2010-10-31 09:05:04 <Diablo-D3> but its not full blown circuitry
334 2010-10-31 09:05:08 <Diablo-D3> its like an extra 10% or some shit
335 2010-10-31 09:05:13 <Diablo-D3> I saw it quoted in some AMD shit
336 2010-10-31 09:05:31 <Diablo-D3> its enough to drop it on hardware that doesnt need it, but not enough to really consider it useful
337 2010-10-31 09:05:42 <Diablo-D3> lfm: on radeons, it ends up with shitty ALU performance
338 2010-10-31 09:05:53 <Diablo-D3> since you either get 5x SP, or 1x DP
339 2010-10-31 09:05:53 <lfm> twice as long is still way better than trying to handle it in software Id think
340 2010-10-31 09:05:59 <Diablo-D3> see above.
341 2010-10-31 09:06:23 <Diablo-D3> 4 units of the ALU (since its a 5 wide WLIW arch) are not filled when using DP
342 2010-10-31 09:06:38 <Diablo-D3> and on top of that, it STILL takes twice as long
343 2010-10-31 09:06:48 <lfm> 5x as long is still better than trying to do it with software
344 2010-10-31 09:07:06 <Diablo-D3> I think what Im saying is, no one is using DP in anything that requires performance.
345 2010-10-31 09:07:40 <lfm> well when you need that accuracy there isn't really any substitute
346 2010-10-31 09:07:52 <Diablo-D3> if you need that accuracy, you should be considering other hardware
347 2010-10-31 09:09:01 <Diablo-D3> ArtForz: in 36 minutes it did 32
348 2010-10-31 09:09:35 <lfm> I think the tesla's have something for that if I understood the presentation I went to the other day
349 2010-10-31 09:10:15 <Diablo-D3> er sorry, 33
350 2010-10-31 09:10:18 <lfm> 32 could be within a random sample of 50
351 2010-10-31 09:10:38 <Diablo-D3> well, that makes it closer to 66 mhash =/
352 2010-10-31 09:10:43 <Diablo-D3> I wonder what Im doing wrong
353 2010-10-31 09:11:07 <Diablo-D3> lfm: teslas just have the double time DP hardware
354 2010-10-31 09:11:29 <Diablo-D3> lfm: but since they use a straight forward 4x simd arch, they dont have the penalty AMD does
355 2010-10-31 09:11:51 <Diablo-D3> and its quite possible AMD doesnt have the penalty anymore either
356 2010-10-31 09:12:00 <Diablo-D3> Im very interested in the other arch they have
357 2010-10-31 11:06:49 <Diablo-D3> ArtForz: you around?
358 2010-10-31 13:11:30 <Diablo-D3> alright
359 2010-10-31 13:11:34 <Diablo-D3> I cant get it any faster
360 2010-10-31 13:11:41 <Diablo-D3> it just doesnt. get. any. faster.
361 2010-10-31 13:17:24 <Diablo-D3> its going around 68 khash/sec give or take
362 2010-10-31 13:18:42 <genjix> hi
363 2010-10-31 13:19:06 <genjix> looks like bitcoin finally tailed off for a bit
364 2010-10-31 13:42:45 <genjix> bye
365 2010-10-31 13:47:04 <yebyen> tailed off?
366 2010-10-31 13:49:19 <yebyen> come onnnn cdebootstrap
367 2010-10-31 13:49:29 <yebyen> i need dhcp/dns at my house
368 2010-10-31 13:49:52 <yebyen> there's no outside network connection, and my bitcrapsd-TEST nodes are not talking to each other on the IRC server I gave them in /etc/hosts
369 2010-10-31 13:50:07 <yebyen> can anyone think of another reason why it wouldn't work?
370 2010-10-31 13:50:20 <ne0futur> firewall ?
371 2010-10-31 13:50:29 <yebyen> there's no outside network connection
372 2010-10-31 13:50:40 <yebyen> the nodes work if i put them on the real internet, where they can get to irc.lfnet.org
373 2010-10-31 13:50:55 <yebyen> firewall will do it though, they don't work at my work/office connection where they can't reach irc.lfnet.org
374 2010-10-31 13:51:24 <yebyen> i have an inspircd and i listed it in /etc/hosts
375 2010-10-31 13:51:26 <recon> How long does a transaction take to propagate? I've got 23 connections and all the current blocks, and at least overnight I didn't get the 0.05 from the faucet.
376 2010-10-31 13:51:39 <yebyen> but i'll tell you tomorrow, if it was only DNS that i needed to add
377 2010-10-31 13:52:31 <yebyen> new newport in non-menthol?
378 2010-10-31 13:52:32 <yebyen> wtf
379 2010-10-31 14:22:50 <MacRohard> yebyen, there's some command line switch you can set to tell them an IP to connect too
380 2010-10-31 15:52:43 <nanotube> recon: shouldn't take long at all to propagate. most of the time you see the 0/unconfirmed transaction in your client within seconds.
381 2010-10-31 15:52:58 <brocktice> and the rally continues, for now
382 2010-10-31 15:53:11 <recon> nanotube: Yeah, turns out the problem was that I had noscript on and hadn't filled out the captcha
383 2010-10-31 15:53:17 <recon> because I am an idiot, apparently
384 2010-10-31 15:53:27 <nanotube> brocktice: prices seem to be pretty stable at the moment....
385 2010-10-31 15:53:34 <nanotube> recon: haha
386 2010-10-31 15:53:51 <brocktice> nanotube: many have been saying they expected it to tank soon
387 2010-10-31 15:53:55 <brocktice> and it was starting to head back down
388 2010-10-31 15:53:58 <recon> i actually find the whole thing in theory quite interesting, and I'm reading the technical paper now
389 2010-10-31 15:54:05 <nanotube> recon: cool :)
390 2010-10-31 15:54:06 <brocktice> but it does seem to be leveling out about 0.18
391 2010-10-31 15:54:07 <recon> but I'm kinda wondering how the client bootstraps itself
392 2010-10-31 15:54:40 <ArtForz> initial bootstrap via irc and a hardcoded list of seednodes
393 2010-10-31 15:55:06 <nanotube> brocktice: yea, my personal expectation is a bit of stability at under .2 ... some people expect a bit of a pullback... probably the short sellers, mostly. :)
394 2010-10-31 15:55:15 <ArtForz> once it connects to a few nodes exchange of node addresses via the p2p network
395 2010-10-31 15:55:20 <nanotube> brocktice: only way to find out is to wait and see hehe
396 2010-10-31 15:55:37 <recon> huh.
397 2010-10-31 15:55:40 <brocktice> nanotube: well, I'm doing more of a buy and hold so it's fine if it fluctuates some.
398 2010-10-31 15:55:52 <brocktice> As long as it doesn't collapse and disband, but I don't think that's looking likely soon.
399 2010-10-31 15:56:47 <brocktice> So, I'm a key offender here, but should someone start a #bitcoin-mining if they haven't already?
400 2010-10-31 15:56:57 <brocktice> I understand that it's closely tied with development and all
401 2010-10-31 16:06:22 <brocktice> ArtForz: if running m0's miner, each GPU is sort of competing against the others in a machine with multiple GPUs, right?
402 2010-10-31 16:06:51 <ArtForz> erm, kinda
403 2010-10-31 16:07:50 <ArtForz> his latest getwork patch is buggy, multiple miners on the same node effectively do duplicate work
404 2010-10-31 16:08:30 <brocktice> which is why some people seem to be reporting finding the same block on multiple processes
405 2010-10-31 16:08:39 <ArtForz> yep
406 2010-10-31 16:09:06 <ArtForz> I think instead of each miner getting it's own part of the nonce space to search, they all get assigned the same range
407 2010-10-31 16:09:13 <brocktice> oh
408 2010-10-31 16:09:13 <recon> whichever one publishes first gets the credit, then?
409 2010-10-31 16:09:18 <brocktice> I see he updated the thread
410 2010-10-31 16:09:25 <ArtForz> yep
411 2010-10-31 16:09:39 <brocktice> I guess when I build for the new machine I'll use 166
412 2010-10-31 16:09:51 <brocktice> I should probably learn how this stuff works, though.
413 2010-10-31 16:10:16 <brocktice> Also, I guess I don't understand Crossfire, but why not crossfire the cards and just run on them as one big card? OpenCL just doesn't support it?
414 2010-10-31 16:11:25 <ArtForz> yep
415 2010-10-31 16:12:15 <ArtForz> amds opencl implementation doesn't support crossfire at all
416 2010-10-31 16:12:31 <brocktice> hm
417 2010-10-31 16:13:32 <ArtForz> which seems to be a slight problem with 5970s under windows, no way to disable the internal crossfire
418 2010-10-31 16:13:38 <brocktice> Also, do you know the actual max power draw of a 5970?
419 2010-10-31 16:13:49 <brocktice> The 'guidelines' given for the card indicate 650w or such to start with
420 2010-10-31 16:13:55 <brocktice> 850 or so for two
421 2010-10-31 16:13:59 <brocktice> so that tells me 200ish?
422 2010-10-31 16:14:05 <ArtForz> erm... thats for the whole system
423 2010-10-31 16:14:09 <brocktice> right
424 2010-10-31 16:14:12 <ArtForz> pretty damn close to 300W stock
425 2010-10-31 16:14:17 <brocktice> huh, ok
426 2010-10-31 16:14:26 <ArtForz> OCed, quite a bit more
427 2010-10-31 16:14:38 <brocktice> I looked into combined heat and power devices yesterday
428 2010-10-31 16:14:48 <brocktice> because kiba or someone mentioned stirling engines
429 2010-10-31 16:14:52 <ArtForz> about 370W @ 1.15Vcore and 910Mhz core
430 2010-10-31 16:14:53 <brocktice> Is that what you have for heat/power?
431 2010-10-31 16:14:59 <ArtForz> pretty much
432 2010-10-31 16:15:01 <brocktice> Have you considered water cooling?
433 2010-10-31 16:15:20 <brocktice> I heard there was a test ongoing in the UK and Germany.
434 2010-10-31 16:15:24 <ArtForz> yes, cost of water blocks makes it pretty much not worth it
435 2010-10-31 16:15:31 <brocktice> ah, ok
436 2010-10-31 16:15:36 <brocktice> Yeah it's like $150/card.
437 2010-10-31 16:16:03 <brocktice> I wonder how much more you could eke out though, I'm sure it's published somewhere.
438 2010-10-31 16:17:35 <ArtForz> my combined heat + power system is pretty much a small industrial 3cyl diesel with a water-water heat exchanger and exhaust heat exchanger + a watercooled 3-phase generator
439 2010-10-31 16:17:52 <brocktice> Is that normal where you are or your own thing?
440 2010-10-31 16:18:10 <ArtForz> the control logic is my own design
441 2010-10-31 16:18:33 <ArtForz> the actual engine+generator are a commercial heat+power unit
442 2010-10-31 16:19:26 <brocktice> hm here's a water block for $80?
443 2010-10-31 16:19:33 <brocktice> Guy is saying 85C on full load with air
444 2010-10-31 16:19:38 <brocktice> 42 or 43 with water
445 2010-10-31 16:19:52 <brocktice> Is the hash rate / clock cycle linear?
446 2010-10-31 16:19:59 <ArtForz> yes
447 2010-10-31 16:20:26 <ArtForz> from my experience on 5970s core temp isn't an issue, VRM temps are
448 2010-10-31 16:20:33 <brocktice> ah
449 2010-10-31 16:20:35 <brocktice> well, that's covered too
450 2010-10-31 16:20:59 <ArtForz> and yeah, with a decent water block you should be able to run a 5970 at 1.20V core and ~1GHz
451 2010-10-31 16:26:54 <brocktice> Hm, yeah that's not too much of a benefit over air cooling.
452 2010-10-31 16:27:06 <brocktice> That VRAM does get hot. Shuts off at 125C??
453 2010-10-31 16:27:24 <brocktice> I should hook up a phase change system and use it to boil water for pasta or something.
454 2010-10-31 16:27:46 <ArtForz> VRM, Voltage Regulator Module
455 2010-10-31 16:27:52 <brocktice> oh
456 2010-10-31 16:28:17 <brocktice> misread, that makes more sense
457 2010-10-31 16:28:40 <ArtForz> and even with a good waterblock VRM temps hit >80???C, those ICs are just too damn small
458 2010-10-31 16:34:51 <genjix> ;bcm
459 2010-10-31 16:36:01 <nanotube> brocktice: looks like -mining already exists. :)
460 2010-10-31 16:36:04 <nanotube> genjix: bitbot's down
461 2010-10-31 16:36:09 <genjix> figured
462 2010-10-31 16:36:21 <brocktice> so it does
463 2010-10-31 16:37:41 <nanotube> brocktice: also, my strategy re:bitcoin is the same, but i try to facilitate some market activity as well.
464 2010-10-31 16:37:43 <recon> what's bitbot do?
465 2010-10-31 16:37:53 <nanotube> recon: carries various useful information. :)
466 2010-10-31 16:38:00 <brocktice> nanotube: sure, ditto
467 2010-10-31 16:38:00 <nanotube> gribble has some too... but not as much.
468 2010-10-31 16:38:00 <recon> like blocks and difficulty and such?
469 2010-10-31 16:38:11 <nanotube> recon: right. like ,,(bc,blocks) e.g.
470 2010-10-31 16:38:12 <gribble> 88842
471 2010-10-31 16:38:32 <nanotube> but bitbot also had stats on bcm/mtg market info, next difficulty estimate, etc.
472 2010-10-31 16:38:39 <recon> that's... pretty cool
473 2010-10-31 16:38:51 <recon> not that I have the computing power to effectively mine (still leave it on overnight), but cool
474 2010-10-31 16:38:54 <nanotube> http://www.bitcoin.org/wiki/doku.php?id=bitbot
475 2010-10-31 16:42:49 <brocktice> oh, supybot
476 2010-10-31 16:42:53 <brocktice> I used to run one of those
477 2010-10-31 16:43:08 <recon> as did I
478 2010-10-31 16:49:36 <XxMalouinxX> ArtForz: Keefe doubted my number yesterday, he said to me it was near to impossible to get 360mhash/s witha radeon 5970 OC'ed to 900Mhz with m0 kernel
479 2010-10-31 16:50:52 <ArtForz> 360 soudns a bit high
480 2010-10-31 16:51:45 <XxMalouinxX> Well, it vary from 318 to 335mhash
481 2010-10-31 16:51:49 <ArtForz> I'd say 330-335 is more in the ballpark
482 2010-10-31 16:52:37 <XxMalouinxX> http://img714.imageshack.us/img714/52/keefe.jpg
483 2010-10-31 16:52:57 <XxMalouinxX> oh btw, I did my rig yesterday : http://img827.imageshack.us/img827/3483/gedc0043.jpg
484 2010-10-31 16:53:09 <brocktice> Mine's coming Tuesday
485 2010-10-31 16:53:13 <brocktice> I'll post it then
486 2010-10-31 16:53:19 <brocktice> the second card is coming Weds.
487 2010-10-31 16:53:31 <brocktice> What're the black boxes in back?
488 2010-10-31 16:53:36 <XxMalouinxX> My rig without the radeon 5970* :P mine should arrive next week too
489 2010-10-31 16:53:38 <brocktice> power?
490 2010-10-31 16:53:41 <XxMalouinxX> psu
491 2010-10-31 16:53:43 <XxMalouinxX> yeah
492 2010-10-31 16:53:53 <brocktice> no behind the PSU
493 2010-10-31 16:54:11 <brocktice> either side of the CD-ROM and drives
494 2010-10-31 16:54:17 <XxMalouinxX> the first black box is the cpu fan, coolmaster V8
495 2010-10-31 16:54:30 <brocktice> that's a cpu fan?
496 2010-10-31 16:54:30 <XxMalouinxX> the 2 black boxes behind are 2 psu, one of 1000w and one other of 1200w
497 2010-10-31 16:54:38 <ArtForz> PSU mounting, hole tape FTW!
498 2010-10-31 16:55:10 <XxMalouinxX> What ? xD lol
499 2010-10-31 16:56:13 <brocktice> eh, wow
500 2010-10-31 16:56:19 <brocktice> just looked it up
501 2010-10-31 16:57:43 <XxMalouinxX> with that I can OC the i7 920 to 4ghz, That is the main reason why i bought it
502 2010-10-31 16:58:04 <ArtForz> I just use boxed coolers for my mining rigs
503 2010-10-31 16:58:26 <XxMalouinxX> like i said earlier, this was my gaming rig :p that why I got that one
504 2010-10-31 16:58:32 <ArtForz> yeah
505 2010-10-31 16:58:54 <ArtForz> remember, my first rig also had a tower cooler :P
506 2010-10-31 16:59:13 <XxMalouinxX> you can certainly call that a tower hehe
507 2010-10-31 16:59:23 <XxMalouinxX> when it was in a case, it was hard to close it without touching it
508 2010-10-31 16:59:35 <brocktice> So, the case I ordered has a long spot for internal 3.5" drives.
509 2010-10-31 16:59:50 <brocktice> 80mm fans should fit in there nicely.
510 2010-10-31 16:59:58 <brocktice> Might mount a bunch of them up front
511 2010-10-31 17:00:16 <brocktice> Come to think of it, I should probably swap cases, which doesn't excite me.
512 2010-10-31 17:01:23 <ArtForz> for my boxes that actually have a case, sharkoon rebel9
513 2010-10-31 17:01:29 <XxMalouinxX> hehe up to you. but I can say for sure, that I fan before and after your graphic card, help a LOT for cooling
514 2010-10-31 17:01:39 <XxMalouinxX> I dropped of 7-10c
515 2010-10-31 17:01:43 <ArtForz> yep
516 2010-10-31 17:02:20 <brocktice> really?
517 2010-10-31 17:02:29 <brocktice> You'd think the GPU fan could pull enough air on its own.
518 2010-10-31 17:02:34 <ArtForz> had a similar push-pull setup for my quad 5770s before I got the industrial fans
519 2010-10-31 17:02:58 <brocktice> not sure if this link will work
520 2010-10-31 17:03:02 <brocktice> http://www.newegg.com/Product/ImageGallery.aspx?CurImage=11-128-020-S01&ISList=11-128-020-S01,11-128-020-S02,11-128-020-S03,11-128-020-S04,11-128-020-S05,11-128-020-S06,11-128-020-S07,11-128-020-S08,11-128-020-S09,11-128-020-S10&S7ImageFlag=1&Item=N82E16811128020&Depa=0&WaterMark=1&Description=ARK%20IPC-4U600%20Black%20Steel%204U%20Rackmount%20Server%20Case
521 2010-10-31 17:03:22 <brocktice> Oh they've already got a 120mm in there
522 2010-10-31 17:03:23 <brocktice> hahaha
523 2010-10-31 17:03:31 <brocktice> I already knew this was basically the case I have
524 2010-10-31 17:03:35 <brocktice> but I just saw what they did
525 2010-10-31 17:03:47 <brocktice> the one I have now has two of those 5.25" drive cages
526 2010-10-31 17:03:52 <brocktice> and another little cage between them.
527 2010-10-31 17:04:07 <brocktice> For this one they just took out the right two cages and put in the horizontal one.
528 2010-10-31 17:04:46 <brocktice> But if I take that cage out it'd be the perfect place for a radiator...
529 2010-10-31 17:21:07 <brocktice> lol, reading a review of the 5970: "The card itself is longer than a fourth-grade piano recital."
530 2010-10-31 17:24:28 <ArtForz> even a 5970 is still a bit shorter than what the PCIe spec allows
531 2010-10-31 17:24:41 <brocktice> well in a deep 4u rackmount case I'm not going to have an issue
532 2010-10-31 17:24:52 <brocktice> I just thought it was a funny expression
533 2010-10-31 17:25:49 <brocktice> 70.5 db at 100% fan?
534 2010-10-31 17:26:03 <ArtForz> sounds about right
535 2010-10-31 17:26:16 <XxMalouinxX> I confirm that
536 2010-10-31 17:26:34 <ArtForz> PCIe spec says full length card is 312mm + bracket + card retainer
537 2010-10-31 17:27:27 <brocktice> not so bad I guess
538 2010-10-31 17:27:37 <ArtForz> 5970 is 305mm I think
539 2010-10-31 17:27:46 <brocktice> apparently it's between standing next to a freeway and runing the garbage disposal
540 2010-10-31 17:27:51 <brocktice> *running
541 2010-10-31 17:28:13 <ArtForz> hehehe
542 2010-10-31 17:28:36 <brocktice> so AMD underclocked it to keep to the PCI-e spec?
543 2010-10-31 17:28:47 <ArtForz> yep
544 2010-10-31 17:29:41 <brocktice> "
545 2010-10-31 17:29:44 <brocktice> oops
546 2010-10-31 17:29:54 <brocktice> "plans are afoot to enable dual 8-pin power connectors on future 5970 cards, but don't expect to see such an animal until next year."
547 2010-10-31 17:30:14 <ArtForz> yep, stock 5970 PCB has space for dual 8-pins
548 2010-10-31 17:30:35 <brocktice> I was cursing myself for not getting a bigger power supply
549 2010-10-31 17:30:40 <ArtForz> also has pads for a 4th Vcore VRM phase
550 2010-10-31 17:30:49 <brocktice> but it sounds like I couldn't get 3 cards on one PSU regardless.
551 2010-10-31 17:30:56 <brocktice> at least not with overclocking.
552 2010-10-31 17:31:05 <ArtForz> yep
553 2010-10-31 17:31:23 <ArtForz> btw, I think thats pretty much what sapphire did for the 4GB 5970
554 2010-10-31 17:31:33 <brocktice> what's that?
555 2010-10-31 17:31:46 <brocktice> I mean, what did they do?
556 2010-10-31 17:31:59 <ArtForz> a 5970 with 2*2GB ram, running 850MHz stock and 900 in the factory OCed version
557 2010-10-31 17:32:06 <brocktice> ahh
558 2010-10-31 17:32:43 <ArtForz> chip layout directly around the GPUs is a bit different to accomodate the extra ram chips, but the power supply section is == 5970, just with dual 8-pins and all VRM phases populated
559 2010-10-31 17:33:12 <brocktice> gotcha
560 2010-10-31 17:33:19 <ArtForz> of course it's also expensive as hell
561 2010-10-31 17:33:27 <brocktice> so 1 GHz gets an extra 30-40 mhash/s?
562 2010-10-31 17:33:38 <brocktice> and it seems that's the upper limit.
563 2010-10-31 17:33:42 <brocktice> (vs. 900)
564 2010-10-31 17:33:47 <ArtForz> yea
565 2010-10-31 17:33:51 <brocktice> hrm
566 2010-10-31 17:34:16 <ArtForz> 5870/5970 doesn't seem to go much past 1Ghz
567 2010-10-31 17:34:37 <ArtForz> even with >1.3V core most cards wont reach 1.1
568 2010-10-31 17:34:57 <ArtForz> and to be honest I wouldnt want to try that on a 5970
569 2010-10-31 17:35:28 <brocktice> so at current rates of difficulty and prices
570 2010-10-31 17:35:43 <brocktice> pushing each card to 1 GHz is worth about $3-4/day.
571 2010-10-31 17:35:57 <brocktice> I can see why you're not bothering with water cooling then
572 2010-10-31 17:36:20 <ArtForz> yep
573 2010-10-31 17:37:23 <ArtForz> I lowered the OC on my cards a bit
574 2010-10-31 17:38:02 <brocktice> why?
575 2010-10-31 17:38:56 <ArtForz> 900 @ 1.14V
576 2010-10-31 17:39:35 <brocktice> what was it?
577 2010-10-31 17:39:41 <ArtForz> 910 @ 1.15V
578 2010-10-31 17:40:15 <ArtForz> lowered core temps by ~1???C and VRM by 4-5???C
579 2010-10-31 17:40:25 <brocktice> were they making you nervous?
580 2010-10-31 17:40:42 <ArtForz> yeah, VRM temps were still a bit too high
581 2010-10-31 17:41:01 <brocktice> fans at 90%?
582 2010-10-31 17:41:32 <ArtForz> GPU fans @ 70%
583 2010-10-31 17:41:49 <ArtForz> 100% on the cards w/ accelero coolers
584 2010-10-31 17:41:53 <brocktice> right
585 2010-10-31 17:42:03 <brocktice> nm
586 2010-10-31 17:42:10 <brocktice> plus you have that wind tunnel box
587 2010-10-31 17:42:13 <ArtForz> yep
588 2010-10-31 17:42:50 <ArtForz> without it VRM temps hit 110???C even at 100% fan
589 2010-10-31 17:43:33 <ArtForz> afair spec for the VRM ICs says says long-term max is 110???C, short term 125???C
590 2010-10-31 17:47:00 <brocktice> apt-get upgrade on a production server always makes me nervous.
591 2010-10-31 17:47:03 <brocktice> :(
592 2010-10-31 17:47:53 <brocktice> I wonder whether the recent increase in difficulty will drive up prices.'
593 2010-10-31 17:48:07 <ArtForz> probably wont
594 2010-10-31 17:48:29 <brocktice> they're moving up now but it's hard to say whether that's just a momentary thing.
595 2010-10-31 17:50:27 <Kiba> increase in price for bitcoin incentivize mining
596 2010-10-31 17:50:46 <brocktice> increase in difficulty disincentivizes it
597 2010-10-31 17:51:05 <brocktice> but I'm thinking that the increase in difficulty may make miners more reluctant to part with their btcs at lower prices.
598 2010-10-31 17:51:21 <ArtForz> that's a perfectly cromulent word.
599 2010-10-31 17:51:44 <Kiba> increase in difficulty mean there are less miners
600 2010-10-31 17:51:47 <brocktice> I just hope the increase in mining doesn't drive away the people who like bitcoin because their idle cpu slices make free money.
601 2010-10-31 17:52:35 <Kiba> spectulation is good for the economy
602 2010-10-31 17:54:22 <Kiba> however, bitcoin have value only because there are people who believes in bitcoin
603 2010-10-31 17:54:47 <Kiba> so I hope these people realize that their return is based in part on how many goods & services there are in the btc economy
604 2010-10-31 17:54:50 <brocktice> More than they believe in fiat currencies anyway
605 2010-10-31 17:55:18 <brocktice> bitcoin has value on its own as long as people believe it is more reliable than other currencies.
606 2010-10-31 17:56:24 <nevezen> I don't know what to believe in
607 2010-10-31 17:56:56 <Kiba> bitcoin's reliablity is pretty much proven when paypal decided to suspend MtGox's account
608 2010-10-31 17:57:19 <brocktice> ?
609 2010-10-31 17:57:26 <brocktice> ah, meaning vs. paypal
610 2010-10-31 17:57:32 <brocktice> as a transfer medium
611 2010-10-31 17:57:59 <nevezen> I've been running bitcoin for a few months now. And it hasn't generated one coin yet.
612 2010-10-31 17:58:06 <Kiba> most of other ways have chargeback ability too
613 2010-10-31 17:58:13 <Kiba> nevezen: you're unlucky
614 2010-10-31 17:58:19 <brocktice> yeah that's bad luck
615 2010-10-31 17:58:27 <nevezen> yeah
616 2010-10-31 17:58:28 <brocktice> even rather slow machines of mine have now and then got a block.
617 2010-10-31 17:58:50 <nevezen> which is why I don't think bitcoin works.
618 2010-10-31 17:58:51 <brocktice> although it's getting less and less frequent
619 2010-10-31 17:58:56 <brocktice> Oh it works.
620 2010-10-31 17:59:07 <brocktice> its primary purpose is not to generate coins for users though ;)
621 2010-10-31 17:59:08 <Kiba> if bitcoin is reliant on mining, than it might as well be useless
622 2010-10-31 17:59:23 <nevezen> I'll accept that it works when it generates one coin
623 2010-10-31 17:59:32 <brocktice> Tell you what
624 2010-10-31 17:59:35 <brocktice> you post your address
625 2010-10-31 17:59:40 <brocktice> I'll send you a coin
626 2010-10-31 17:59:43 <brocktice> then you'll see that it works
627 2010-10-31 17:59:59 <brocktice> or there's always the faucet
628 2010-10-31 17:59:59 <Kiba> nevezen: you're missing the point of bitcoin
629 2010-10-31 18:00:04 <nevezen> meh, like bribing someone to stfu? :)
630 2010-10-31 18:00:07 <recon> brocktice: is that an open offer? :B
631 2010-10-31 18:00:10 <nevezen> kiba: perhaps.
632 2010-10-31 18:00:14 <brocktice> recon: no
633 2010-10-31 18:00:17 <recon> damn
634 2010-10-31 18:00:34 <brocktice> https://freebitcoins.appspot.com/
635 2010-10-31 18:00:49 <brocktice> there, that's what the faucet is for.
636 2010-10-31 18:00:58 <nevezen> thanks for the offer though :)
637 2010-10-31 18:01:20 <brocktice> wow the faucet has a decent amount of money there.
638 2010-10-31 18:01:31 <Kiba> don't abuse the faucet
639 2010-10-31 18:01:57 <Kiba> ya know, people always miss the point of bitcoins
640 2010-10-31 18:02:17 <nevezen> so what is the point of bitcoins? really?
641 2010-10-31 18:02:17 <recon> what, to have a decentralized yet secure commodity that won't inflate?
642 2010-10-31 18:02:20 <Kiba> and mining is somehow an "improtant aspect"
643 2010-10-31 18:02:26 <Kiba> nevezen: a monentary system
644 2010-10-31 18:02:43 <Kiba> to coordinate the production of the economy
645 2010-10-31 18:03:08 <recon> maybe I should organize a microstakes poker game with it and my friends, like with maybe a 5 BTC -> 500 chips conversion
646 2010-10-31 18:03:15 <Kiba> to coordinate resource of an economy
647 2010-10-31 18:03:22 <recon> use pokerth and pay out after the fact
648 2010-10-31 18:03:25 <recon> hm....
649 2010-10-31 18:03:26 <Kiba> mining is just an incentive
650 2010-10-31 18:03:37 <jrabbit> bitcoin because fuck paypal
651 2010-10-31 18:03:53 <brocktice> Not just paypal.
652 2010-10-31 18:03:56 <Kiba> and a poor one at that
653 2010-10-31 18:04:00 <brocktice> I wish my company could pay me in bitcoins.
654 2010-10-31 18:04:00 <Kiba> brocktice: credit cards
655 2010-10-31 18:04:02 <brocktice> hey wait
656 2010-10-31 18:04:04 <Kiba> anything with a chargeback
657 2010-10-31 18:04:08 <brocktice> I think it can
658 2010-10-31 18:04:16 <brocktice> because the damn bank charges for every paycheck
659 2010-10-31 18:04:27 <brocktice> but if the company buys bitcoins and then pays me... hmmm..
660 2010-10-31 18:04:45 <Kiba> no payroll taxes
661 2010-10-31 18:04:51 <brocktice> well, that's dicier
662 2010-10-31 18:04:55 <Kiba> no other fucking silly taxes
663 2010-10-31 18:04:59 <brocktice> I'm not looking to evade that explicitly
664 2010-10-31 18:05:06 <nevezen> not very tangible like the feel of paper cash though :P
665 2010-10-31 18:05:16 <brocktice> but legitimately evading the bank's fees would be nice
666 2010-10-31 18:05:30 <Kiba> legitimately?
667 2010-10-31 18:05:43 <Kiba> there's nothing legimitate about hoisting a boatload of taxes on you
668 2010-10-31 18:05:53 <Kiba> and on services that you probably won't ever have a chance to use
669 2010-10-31 18:05:58 <brocktice> I mean, there's nothing remotely illegal about using a different method of paying me
670 2010-10-31 18:06:09 <brocktice> instead of their ACH transfers with attached fees.
671 2010-10-31 18:07:24 <recon> your place of employment might raise a bit of protest over having to convert to bitcoins, assuming they can do that and don't have to go cash -> western union -> LRUSD -> bitcoins or some other equally absurd chain
672 2010-10-31 18:07:37 <brocktice> I am part owner of my place of employment. :)
673 2010-10-31 18:08:03 <brocktice> should be able to buy them on the company credit card or something.
674 2010-10-31 18:08:15 <Kiba> hmm
675 2010-10-31 18:08:17 <recon> it's... difficult to buy bitcoins on a credit card
676 2010-10-31 18:08:25 <Kiba> business opportunity: buy a bunch of walmart credit cards
677 2010-10-31 18:08:26 <recon> because chargebacks and non-chargebackable currency
678 2010-10-31 18:08:32 <Kiba> sell them at a bitcoin premium
679 2010-10-31 18:08:36 <brocktice> hmm
680 2010-10-31 18:08:40 <recon> I should know, I spent a few hours researching that
681 2010-10-31 18:08:49 <brocktice> recon: good to know.
682 2010-10-31 18:08:51 <Kiba> like 5.10 for a $5 credit card
683 2010-10-31 18:09:05 <brocktice> there's some fee
684 2010-10-31 18:09:07 <brocktice> it'd be more than that
685 2010-10-31 18:09:07 <recon> you can put that into paypal or liqpay and maybe throught that through exchangezone
686 2010-10-31 18:09:10 <recon> but it's still a hassle
687 2010-10-31 18:10:02 <Kiba> sell people prepaid credit cards
688 2010-10-31 18:10:03 <nanotube> brocktice: once bcm or mtg get ach going, then it will be a whole lot easier
689 2010-10-31 18:10:06 <recon> and of course I wouldn't even want to have to deal with the IRS on income taxes with that
690 2010-10-31 18:10:16 <Kiba> nanotube: I thought mtgox stop?
691 2010-10-31 18:10:30 <brocktice> "Oh yeah, they paid me 5000 bitcoins, which was worth X USD at the time, but now they're worth Y USD."
692 2010-10-31 18:10:36 <nanotube> Kiba: he got denied by one ach provider... but is still working with another.
693 2010-10-31 18:10:39 <brocktice> but this must be a common thing with people paid in foreign currency.
694 2010-10-31 18:10:43 <nanotube> brocktice: no different than getting paid in gold.
695 2010-10-31 18:10:49 <nanotube> brocktice: right also fx
696 2010-10-31 18:11:00 <recon> brocktice: it's decided by the exchange rate the Treasury Department sets when you pay your income taxes, IIRC
697 2010-10-31 18:11:17 <recon> the problem here being that the treasury department probably doesn't recognize an exchange rate for bitcoins atm
698 2010-10-31 18:11:19 <nanotube> you pay taxes on 'income' based on the value of commodity as it was when you received it. any further increase is a capital gain, which only gets taxed once you realize it by selling.
699 2010-10-31 18:11:29 <recon> or that
700 2010-10-31 18:11:59 <Kiba> the government can't set a simple tax
701 2010-10-31 18:12:11 <Kiba> and decided to throughly piss off people
702 2010-10-31 18:12:18 <nanotube> Kiba: heh, well, that's a whole different problem. :)
703 2010-10-31 18:12:31 <Kiba> well, taxes are immoral/wrong anyway
704 2010-10-31 18:12:35 <Kiba> but we got no choice
705 2010-10-31 18:12:47 <jrabbit> Don't evade taxes with bitcoin. :(
706 2010-10-31 18:12:54 <recon> my head is spinning now. no wonder people are paid to figure this stuff out
707 2010-10-31 18:13:10 <brocktice> yeah I did the accounting for my two companies the first year
708 2010-10-31 18:13:22 <brocktice> The taxes are a big clusterfuck
709 2010-10-31 18:13:28 <nanotube> haha yea
710 2010-10-31 18:13:55 <Kiba> I am sure a lot of people who will use bitcoin will use it to evade taxes
711 2010-10-31 18:14:05 <Kiba> just like a large portion of Americans don't pay their income taxes each year
712 2010-10-31 18:14:25 <recon> I don't have any income worth evading taxes for anyway, so there's that
713 2010-10-31 18:14:29 <brocktice> something like half of americans don't *owe* income taxes each year.
714 2010-10-31 18:15:06 <Kiba> clusterfuck incentives
715 2010-10-31 18:15:23 <recon> i'm pretty sure I fall below the poverty line insofar as taxes go, I'm not even full-time employed at the moment
716 2010-10-31 18:15:40 <recon> and social security, medicare, etc all just get taken out of my paycheck anyway
717 2010-10-31 18:15:43 <Kiba> Rules for Ruling an Evil Empire:
718 2010-10-31 18:16:14 <nanotube> rule #1: there are no rules.
719 2010-10-31 18:16:15 <nanotube> haha
720 2010-10-31 18:16:18 <Kiba> 1. Tax code should be simple and should be low. Budget accordingly so that you have to cancel tax collection because the imperial treasury is overflowing
721 2010-10-31 18:16:40 <brocktice> I thought you said an evil empire?
722 2010-10-31 18:16:49 <genjix> yeah sounds pretty good to me
723 2010-10-31 18:16:51 <Kiba> all empire ARE evil
724 2010-10-31 18:16:56 <nanotube> where do i join?
725 2010-10-31 18:17:20 <Kiba> just that all empires ALWAY FALL to inflation, wars, or war debt
726 2010-10-31 18:17:27 <nanotube> but some empires are more evil than others. </animal farm.
727 2010-10-31 18:17:36 <recon> four legs good, two legs bad
728 2010-10-31 18:17:52 <genjix> :) i love that book
729 2010-10-31 18:18:02 <nanotube> yea classic
730 2010-10-31 18:18:04 <ArtForz> yup, great book
731 2010-10-31 18:18:15 <genjix> re-watched both the movies the other day
732 2010-10-31 18:18:22 <nanotube> there are movies? O_O
733 2010-10-31 18:18:45 <genjix> yeah the newer one is quite good but spoilt by the hollywood-ified ending
734 2010-10-31 18:18:52 <genjix> (they give it a happy ending)
735 2010-10-31 18:18:53 <ArtForz> yup
736 2010-10-31 18:18:55 <nanotube> hmm
737 2010-10-31 18:19:00 <genjix> which ruins the whole point of the book
738 2010-10-31 18:19:01 <Kiba> 2. Wars should be avoided as a mean of income. Wars cause ruin to many empires. Wars should be done fast and quick, if neccesary.
739 2010-10-31 18:19:14 <ArtForz> old one is a bit more primitive on animation but truer to the book
740 2010-10-31 18:19:24 <Kiba> 3. Cutt off heads of impercial beaucracts who are too conservative.
741 2010-10-31 18:19:33 <genjix> nanotube: they're on youtube
742 2010-10-31 18:19:49 <genjix> newer one uses real actors... watch the ending for true lolz
743 2010-10-31 18:19:56 <nanotube> genjix: heh, i'll put it on my todo list for when i have some time. :)
744 2010-10-31 18:19:57 <Kiba> Conservatism kill empires because they cause stagnation in civilization advancement
745 2010-10-31 18:20:08 <genjix> i was going WHYYYYYY??? hollywood why?
746 2010-10-31 18:20:13 <nanotube> hehe
747 2010-10-31 18:20:30 <genjix> nanotube: http://www.youtube.com/watch?v=EL_QlCxukng
748 2010-10-31 18:20:33 <genjix> only 10 mins
749 2010-10-31 18:24:16 <nanotube> Kiba: 'conservatism' is such a broad label that it has little meaning. am i 'conservative' when i want to 'conserve natural resources' ? am i conservative when i vote 'no' for a new tax on $something? the word has no meaning without context.
750 2010-10-31 18:24:40 <Kiba> nanotube: those who do not embrace new ideas
751 2010-10-31 18:24:45 <Kiba> hmm
752 2010-10-31 18:24:48 <Kiba> that's a bad rule
753 2010-10-31 18:24:51 <nanotube> Kiba: some new ideas are... shitty
754 2010-10-31 18:24:57 <recon> Kiba: New idea! We should kill all the asian people
755 2010-10-31 18:25:00 <recon> oh wait
756 2010-10-31 18:25:03 <recon> that's not a good idea
757 2010-10-31 18:25:07 <nanotube> hehe
758 2010-10-31 18:25:14 <recon> forget I said that
759 2010-10-31 18:25:25 <genjix> conservatives are defined as believing things are better in the past unchanged
760 2010-10-31 18:25:27 <Kiba> nanotube: let say imperial beaucracts who do not embrace new innovations
761 2010-10-31 18:25:44 <Kiba> like important military technologies
762 2010-10-31 18:25:46 <nanotube> an innovation is 'new' by definition. not all of them are good, either.
763 2010-10-31 18:26:00 <recon> <science> I have created a new machine that allows us to use asian people as a consumable power source!
764 2010-10-31 18:26:08 <recon> <recon> I'm not powering my car with asians.
765 2010-10-31 18:26:30 <recon> <science> But it has a 84% efficiency ra--
766 2010-10-31 18:26:33 <recon> <recon> no.
767 2010-10-31 18:26:36 <nanotube> e.g.... the shift of the lcd market from the old 16x10 widescreeen to the 16x9 hdtv-matching widescreen is an 'innovation' but it sure is shitty, cuz it steals pixels from you.
768 2010-10-31 18:26:53 <nanotube> recon: haha good one
769 2010-10-31 18:27:04 <genjix> are you anti-science?
770 2010-10-31 18:27:11 <recon> listen
771 2010-10-31 18:27:17 <recon> no matter what you believe the mathematical tolerances to be
772 2010-10-31 18:27:21 <recon> it's still probably a bad idea to shake a baby
773 2010-10-31 18:27:31 <Kiba> shakababy!
774 2010-10-31 18:27:38 <recon> "Damnit that joint was supposed to be good to fifteen joules"
775 2010-10-31 18:27:41 <nanotube> Kiba: i think it's better to focus on people who refuse to consider things on their merits, rather than some idea of 'conservatism' or 'innovations'
776 2010-10-31 18:27:45 <genjix> <science> I have invented a way to build colonies on Mars
777 2010-10-31 18:27:53 <genjix> <recon> No.
778 2010-10-31 18:27:59 <genjix> n1
779 2010-10-31 18:28:08 <recon> genjix: does the mars colony involve incinerating asian people?
780 2010-10-31 18:28:16 <nanotube> <science> the new way involves building walls out of asians
781 2010-10-31 18:28:22 <nanotube> <recon> zomg no
782 2010-10-31 18:28:29 <nanotube> haha
783 2010-10-31 18:28:44 <recon> well
784 2010-10-31 18:28:50 <recon> if they're not being burned, I *guess* it's okay
785 2010-10-31 18:29:09 <Kiba> how about DEAD asian people?
786 2010-10-31 18:29:11 <nanotube> yea, just... compressed a bit.
787 2010-10-31 18:29:30 <ArtForz> nah, just freeze dried, ground and compacted
788 2010-10-31 18:30:10 <recon> but grinding them compromises the structural integrity of the bones!
789 2010-10-31 18:30:23 <recon> they're like lincoln logs
790 2010-10-31 18:30:36 <nanotube> recon: after freeze drying... i think structural integrity of the bones is going to be the least of your concerns. :)
791 2010-10-31 18:30:38 <recon> really macabre lincoln logs
792 2010-10-31 18:32:28 <recon> Pictured: the new colony on Mars: http://i.crackedcdn.com/phpimages/article/1/9/5/34195.jpg?v=1
793 2010-10-31 18:42:06 <nanotube> genjix: oh wow... such a reversal, that new ending. let's just go over here, and wait until they die. great.
794 2010-10-31 18:42:29 <genjix> yeah its terrible, isn't it.
795 2010-10-31 18:42:47 <nanotube> heh yea, i think so.
796 2010-10-31 18:42:50 <genjix> if george orwell wanted a happy ending, he'd have made one
797 2010-10-31 18:43:04 <Kiba> Orwell is a....SOCIALIST
798 2010-10-31 18:43:11 <Kiba> a naive SOCIALIST
799 2010-10-31 18:43:30 <nanotube> genjix: key error: the farmers usually don't allow the farmed ones to just leave.
800 2010-10-31 18:44:07 <genjix> the whole ending is just... derp.
801 2010-10-31 18:44:10 <nanotube> Kiba: i don't know what his political views were... but his books are a powerful statement against authoritarian governments.
802 2010-10-31 18:44:17 <nanotube> genjix: yea
803 2010-10-31 18:44:26 <genjix> he was anarchist/socialist
804 2010-10-31 18:44:33 <genjix> he fought in catalonia
805 2010-10-31 18:45:06 <genjix> i'm surprised tbh that they didn't also try to fit a romance in there too
806 2010-10-31 18:46:18 <nanotube> haha yea, a dog-horse romance... at least they'd then be able to count on the audience of furries. :)
807 2010-10-31 18:46:29 <ArtForz> lol
808 2010-10-31 18:46:45 <nanotube> though i guess they're probably not numerous enough to be a market segment anyone actually cares about...
809 2010-10-31 18:46:54 <nanotube> too bad. the world needs more furries. :)
810 2010-10-31 18:49:38 <Kiba> we need moar goods and services!
811 2010-10-31 18:51:38 <nanotube> that too :)
812 2010-10-31 18:52:22 <Kiba> so that I can short!
813 2010-10-31 18:57:19 <brocktice> Kiba: wouldn't you want to short *first*?
814 2010-10-31 18:57:47 <Kiba> I am alreay shorting, just waiting for the price to come down
815 2010-10-31 18:57:56 <brocktice> ah, courtesy of nanotube?
816 2010-10-31 18:57:58 <Kiba> I notice a dispranency in bcm and mtgox pricing
817 2010-10-31 18:58:24 <brocktice> Kiba: I don't know if it's large enough to consider worth noticing.
818 2010-10-31 18:59:14 <recon> when the bots syas "400 @ $0.1750" does that mean "400 BTC got sold at $0.1750 a piece" or...
819 2010-10-31 18:59:37 <ArtForz> yep
820 2010-10-31 18:59:49 <Kiba> so the disparency is resolved
821 2010-10-31 18:59:58 <Kiba> price is surging for bitcoinmarket
822 2010-10-31 19:00:31 <brocktice> 0.18 is surging?
823 2010-10-31 19:00:35 <brocktice> It's behind mtgox
824 2010-10-31 19:00:52 <Kiba> behind or whatever
825 2010-10-31 19:00:59 <Kiba> the thing is, it's catching up to mtgox's price
826 2010-10-31 19:01:06 <ArtForz> well, bcm used to trade a lot lower than mtgox
827 2010-10-31 19:01:08 <brocktice> mtgox had been lagging bcm
828 2010-10-31 19:01:52 <ArtForz> over 30% lower, now it's about 10%
829 2010-10-31 19:02:42 <ArtForz> seems to have started shortly after mtgox switched to LR$... coincidence?
830 2010-10-31 19:03:26 <Kiba> started shortly?
831 2010-10-31 19:03:35 <Kiba> it was several days when mtgox already switch to LR?
832 2010-10-31 19:03:43 <recon> when did mtgox switch to LR anyway?
833 2010-10-31 19:05:15 <Keefe> well, i first withdrew usd from mtg to lr on oct 27 at 18:15 utc
834 2010-10-31 19:05:38 <ArtForz> oct 26 I think
835 2010-10-31 19:05:40 <Keefe> i was checking often, so it was probably less than a day before that
836 2010-10-31 19:06:01 <Keefe> then mtgox ran out of lr funds for a short time
837 2010-10-31 19:06:22 <Kiba> fractional reserve banking?
838 2010-10-31 19:06:26 <Keefe> he has funds in other forms, but just didn't have enough moved to lr
839 2010-10-31 19:06:28 <Kiba> mtgox should be a full reserve...
840 2010-10-31 19:06:37 <recon> oh wow, I just missed it
841 2010-10-31 19:06:52 <lfm> it?
842 2010-10-31 19:07:03 <recon> the point where I could have used paypal on mtgox
843 2010-10-31 19:07:14 <recon> as opposed to something with insane fees like western union
844 2010-10-31 19:07:23 <recon> i'm kinda disappointed I missed my chance to cash in on $10 worth of bitcoins with my account in just so many days
845 2010-10-31 19:08:05 <Keefe> paypal withdrawal from mtg stopped around oct 10
846 2010-10-31 19:08:25 <recon> i meant 'buy' by 'cash in'
847 2010-10-31 19:08:27 <recon> stupid english
848 2010-10-31 19:08:30 <Keefe> from then til oct 26, the only way to take funds out of mtg was in btc
849 2010-10-31 19:09:04 <ArtForz> yup, 26 oct mtgox switched to lr
850 2010-10-31 19:09:11 <Kiba> so how many days left  before paypal might unfreeze his account?
851 2010-10-31 19:09:13 <Keefe> except for possibly private arrangements with some people
852 2010-10-31 19:09:24 <Keefe> i have no idea
853 2010-10-31 19:09:26 <lfm> kiba might never
854 2010-10-31 19:09:37 <brocktice> hmm, too bad havenco shut down
855 2010-10-31 19:10:00 <ArtForz> on bcm 26 oct 1 btc == lr$0.10, 27 oct 1 btc=lr$0.17
856 2010-10-31 19:10:47 <Keefe> has anyone checked out the idea of using green dot cards to transfer money from cash to someone else's paypal account?
857 2010-10-31 19:11:07 <recon> that'd be interesting...
858 2010-10-31 19:11:31 <recon> but the card can't be verified without using it making escrow an iffy proposition
859 2010-10-31 19:11:37 <Keefe> it appears that it may be possible to buy a green dot card at walmarts with no or little id, then give the numbers to someone to deposit into their pp acct
860 2010-10-31 19:12:26 <brocktice> recon: actually a trusted escrow person could deposit the funds into their own account
861 2010-10-31 19:12:31 <brocktice> then personal-send them to the recipient
862 2010-10-31 19:12:47 <brocktice> I think if I could determine that was legal I would do it for little to no fee.
863 2010-10-31 19:12:57 <recon> point
864 2010-10-31 19:13:09 <recon> Keefe: i can buy a green dot moneypak at the rite aid down the street without any ID
865 2010-10-31 19:13:10 <recon> just cash
866 2010-10-31 19:13:21 <Keefe> that's what i was thinking
867 2010-10-31 19:13:23 <brocktice> AFACT it's legal to be an escrow without license in all but 3 states in the USA.
868 2010-10-31 19:13:30 <brocktice> AFAICT I mean
869 2010-10-31 19:13:50 <nanotube> brocktice: i think so. see what escrow.com has for their licensing.
870 2010-10-31 19:13:59 <nanotube> they have lic in california, and... a couple other states.
871 2010-10-31 19:14:01 <brocktice> nanotube: yeah they're in California
872 2010-10-31 19:14:08 <brocktice> that's what I based that on, and a little other reading
873 2010-10-31 19:14:08 <Keefe> if i knew that there was no chargeback risk, i'd gladly sell my btc for moneypak numbers
874 2010-10-31 19:14:12 <brocktice> They are licensed in 3 states.
875 2010-10-31 19:14:31 <brocktice> Keefe: there's no chargeback risk with personal payments on paypal, is what I heard.
876 2010-10-31 19:14:34 <recon> Keefe: we could try an experiment between ourselves sometimes but we'd need to find an escrow of some sort
877 2010-10-31 19:14:41 <brocktice> I will escrow.
878 2010-10-31 19:14:44 <recon> and yeah, if it's sent by gift you can't chareback
879 2010-10-31 19:14:46 <Keefe> could be an attractive way for people who want to buy btc anonymously
880 2010-10-31 19:14:52 <brocktice> I use my real name and it can go through my business and PO box.
881 2010-10-31 19:15:14 <Keefe> i'm not thinking of escrow, just direct buyer-to-seller, like how many otc txns are done
882 2010-10-31 19:15:18 <brocktice> ah
883 2010-10-31 19:15:20 <recon> we could try that too
884 2010-10-31 19:15:23 <nanotube> there have been quite a few tx on -otc like that already.
885 2010-10-31 19:15:24 <recon> i'm willing to risk the $10
886 2010-10-31 19:15:54 <recon> Keefe: with a number, assuming the guy doesn't redeem it before you, there's no chargeback risk
887 2010-10-31 19:16:11 <Keefe> can't somehow claim the moneypak card was stolen from you?
888 2010-10-31 19:16:12 <recon> so you'd redeem it first and then send out the BTC payment, but I'd be pretty much at your mercy if you screwed me over
889 2010-10-31 19:16:20 <recon> nope, no liability there
890 2010-10-31 19:16:34 <recon> their answer would be "well, shoulda redeemed it earlier"
891 2010-10-31 19:16:43 <Keefe> sweet. and yes, there's the usual trust issue as with any non-chargebackable payment method
892 2010-10-31 19:17:05 <brocktice> we just need a way to buy bitcoin cards at wal-mart
893 2010-10-31 19:17:07 <recon> well, I'm willing to try that for the venture. Might help with the whole web of trust thing if i ever want to do more otc stuff
894 2010-10-31 19:17:26 <nanotube> well, if you have two non-chargeback-able methods... the problem is completely solved with an escrow.
895 2010-10-31 19:17:28 <brocktice> which comes back to my wallet idea
896 2010-10-31 19:17:37 <brocktice> hm, wallet on a USB stick?
897 2010-10-31 19:18:13 <recon> nanotube: the thing is, the moneypak can't be verified without using it, and as soon as you use it it becomes chargebackable
898 2010-10-31 19:18:23 <recon> of course if the escrow is trusted that's not a problem
899 2010-10-31 19:18:36 <Keefe> i'm willing to risk $10 worth of btc, and give btc first before getting the moneypak numbers, if the buyer has been at least around here chatting for a few days
900 2010-10-31 19:18:44 <nanotube> p1 sends btc to escrow. escrow confirms receipt of funds to p2. p2 sends card number to p1. p1 redeems card... confirms to escrow, escrow sends along the btc.
901 2010-10-31 19:19:04 <recon> Keefe: well, I've been here for one day, but I'd throw over the moneypak numbers first because you seem to be trusted
902 2010-10-31 19:19:12 <nanotube> Keefe: recon seems pretty new here... so you might want to get the moneypak first.
903 2010-10-31 19:19:29 <nanotube> recon: yes, i throw my vote of trust for Keefe
904 2010-10-31 19:19:46 <recon> on top of that the rite aid is closed for halloween so I'd actually have to pick up the moneypak tomorrow
905 2010-10-31 19:19:51 <nanotube> heh
906 2010-10-31 19:20:09 <brocktice> Why bother with the moneypak?
907 2010-10-31 19:20:10 <recon> 'closing for halloween', what with the last-minute candy runs and such, doesn't seem like the best decision anyway
908 2010-10-31 19:20:11 <Kiba> enterprising geeks!
909 2010-10-31 19:20:15 <recon> but *shrug*
910 2010-10-31 19:20:21 <brocktice> If personal payments are not chargeback-able?
911 2010-10-31 19:20:42 <Keefe> brocktice: if recon doesn't have a pp acct or wants to use cash, or stay anonymous
912 2010-10-31 19:21:07 <Keefe> but what about fees...
913 2010-10-31 19:21:27 <brocktice> no fees
914 2010-10-31 19:21:36 <brocktice> you just say it's a personal payment owed
915 2010-10-31 19:21:51 <Keefe> i was wondering about moneypak fees
916 2010-10-31 19:21:59 <brocktice> oh, there might be
917 2010-10-31 19:22:09 <brocktice> whoa
918 2010-10-31 19:22:13 <brocktice> precipitous drop on mtgox
919 2010-10-31 19:22:36 <brocktice> but then it came back up I think?
920 2010-10-31 19:22:37 <Keefe> someone taking advantage of the bcm-mtg arbitrage opportunity?
921 2010-10-31 19:22:40 <brocktice> yes
922 2010-10-31 19:22:46 <brocktice> I mean, yes it came back up
923 2010-10-31 19:23:02 <recon> Keefe: the retailer charges a flat service fee for the moneypak
924 2010-10-31 19:23:10 <Keefe> about $5
925 2010-10-31 19:23:26 <recon> Keefe: so all fees would be on my end instead of yours
926 2010-10-31 19:23:44 <brocktice> The problem with all of this stuff is it takes too long
927 2010-10-31 19:23:44 <Keefe> i think $4.95 is the max fee to get the card
928 2010-10-31 19:23:50 <brocktice> we all want automated exchanges, yes?
929 2010-10-31 19:24:01 <brocktice> or at least digital
930 2010-10-31 19:24:31 <recon> for the purposes of the exercise I'm assuming I want to use cash and stay anonymous
931 2010-10-31 19:24:42 <Kiba> hehe
932 2010-10-31 19:24:45 <Keefe> there's room for various trade methods. yes, automated all-digital exchanges are great, but there's a niche for other methods, especially involving cash
933 2010-10-31 19:24:49 <Kiba> I just missed my opportunity to short sell
934 2010-10-31 19:24:50 <brocktice> then the best thing to do is probably to mail cash
935 2010-10-31 19:25:00 <brocktice> which is already being done
936 2010-10-31 19:25:00 <recon> $4.95 is a hell of a lot less than a $50 wire fee
937 2010-10-31 19:25:20 <Keefe> mailing cash involves the risk of mail loss or delay
938 2010-10-31 19:25:25 <recon> brocktice: mailing has the means for a possible man-in-the-middle attack of "postal service jacks your shit"
939 2010-10-31 19:25:37 <recon> i know I wouldn't trust the USPS 'round these parts with mine
940 2010-10-31 19:25:38 <brocktice> doesn't mailing a prepaid credit card or moneypak have the same risk?
941 2010-10-31 19:25:39 <Keefe> and both parties suspecting the other of cheating when that happens
942 2010-10-31 19:25:52 <recon> brocktice: we're not mailing the moneypak, we're giving the number to the other person digitally
943 2010-10-31 19:25:57 <brocktice> ah I see
944 2010-10-31 19:26:16 <brocktice> that's not so bad then
945 2010-10-31 19:26:49 <Kiba> opportunity for courier services to develop?
946 2010-10-31 19:26:50 <Keefe> and the $5 cost of getting the card can be one-time only, if trading with a single counterparty multiple times
947 2010-10-31 19:27:23 <recon> Keefe: a moneypak isn't reloadable
948 2010-10-31 19:27:25 <Kiba> distributed courier service
949 2010-10-31 19:27:29 <Keefe> oh
950 2010-10-31 19:27:32 <recon> Keefe: you're confusing it with the actual green card
951 2010-10-31 19:27:45 <recon> a green card is like a prepaid credit card that you use moneypaks to load up
952 2010-10-31 19:27:56 <recon> you can bypass the green card entirely and use moneypaks to load a paypal account
953 2010-10-31 19:28:01 <recon> both incur the same fee, paid to the retailer
954 2010-10-31 19:29:21 <recon> Also, unless you 'verify' with GreenDot which involves forking over your name, address, date of birth, and SSN, there's a max of $250 per year of 'moneypak -> paypal'
955 2010-10-31 19:29:48 <recon> doing so raises the max to $2,000 per month, although if you're working with that volume of money I'm pretty sure you can afford a wire or western union or moneygram or something
956 2010-10-31 19:30:08 <recon> </lecture>
957 2010-10-31 19:30:32 <Kiba> the government really hate "money laundering"
958 2010-10-31 19:30:33 <Keefe> who needs to provide that id? the person doing moneypak-to-pp or the person buying the moneypak?
959 2010-10-31 19:31:20 <recon> the person buying the moneypak needs no ID-- it's assumed that's going to a PP or a GreenCard and both require self-identification
960 2010-10-31 19:31:35 <recon> the person doing moneypak->PP needs no other sort of ID other than what he's already given to PP
961 2010-10-31 19:31:37 <recon> to get an account
962 2010-10-31 19:33:14 <recon> moneypaks are pretty much just like cash, and there is absolutely no way to get a chargeback or anything once it's redeemed, even if you can prove it's via fraud
963 2010-10-31 19:33:32 <recon> making it useful for this kind of thing
964 2010-10-31 19:33:43 <Teppy1> OneFixt: You around?
965 2010-10-31 19:34:55 <Teppy1> Wow, what's with the market crazyness?
966 2010-10-31 19:35:13 <recon> ...oh, huh, moneypak also has a $20 minimum
967 2010-10-31 19:35:16 <recon> I thought it was a $10
968 2010-10-31 19:35:52 <recon> still no matter, assuming Keefe has $20 worth of BTC under the future exchange rate