1 2011-01-25 00:02:46 <Diablo-D3> lets see if I modified art's kernnel right
   2 2011-01-25 00:04:30 <Diablo-D3> bullshit
   3 2011-01-25 00:04:32 <Diablo-D3> it worked on the first try?
   4 2011-01-25 00:04:36 <Diablo-D3> I must have done something wrong
   5 2011-01-25 00:06:00 masterme120 has quit (Remote host closed the connection)
   6 2011-01-25 00:09:04 <slush> Diablo-D3: what speed diff?
   7 2011-01-25 00:11:39 <Diablo-D3> hrm
   8 2011-01-25 00:27:05 devon_hillard_ has joined
   9 2011-01-25 00:28:06 <lfm> so its java thats slow?
  10 2011-01-25 00:28:26 <Diablo-D3> slush: on a 4850.... none/
  11 2011-01-25 00:28:29 <Diablo-D3> it runs just as fast.
  12 2011-01-25 00:30:33 devon_hillard has quit (Ping timeout: 240 seconds)
  13 2011-01-25 00:30:36 <slush> I'm pretty curious how it will perform on 5xxx
  14 2011-01-25 00:33:14 <CIA-98> bitcoin: Chris Moore master * rd9711a5 / util.cpp : Don't scan the time offsets vector repeatedly once we've displayed the warning that the user's clock may be wrong. - http://bit.ly/h2TSKs
  15 2011-01-25 00:36:22 <Diablo-D3> actually Im wrong
  16 2011-01-25 00:36:34 <Diablo-D3> it performs a lot faster and the hashometer is breaking under the pressure
  17 2011-01-25 00:36:43 <molecular> lol
  18 2011-01-25 00:36:53 <Diablo-D3> -f 1 strikes again
  19 2011-01-25 00:38:27 <cosurgi> what is "-f 1" and "-f 5" ?
  20 2011-01-25 00:38:35 <Diablo-D3> cosurgi: roughly the same.
  21 2011-01-25 00:38:54 <cosurgi> but what is this option - and optimization during compiling?
  22 2011-01-25 00:39:08 sgornick has quit (Read error: Connection reset by peer)
  23 2011-01-25 00:39:52 andrew12 has quit (Ping timeout: 246 seconds)
  24 2011-01-25 00:40:14 <cosurgi> s/and/an/g
  25 2011-01-25 00:40:28 kermit has joined
  26 2011-01-25 00:41:21 sgornick has joined
  27 2011-01-25 00:41:26 <Diablo-D3> cosurgi: no, it just changes how many kernels are ran per second
  28 2011-01-25 00:41:37 <cosurgi> ok.
  29 2011-01-25 00:42:07 <cosurgi> Diablo-D3: so what is the speed improvement? :)
  30 2011-01-25 00:42:29 <Diablo-D3> well, hacked up version of arts doesnt work
  31 2011-01-25 00:42:34 <Diablo-D3> but it seems to be faster
  32 2011-01-25 00:42:40 <cosurgi> ok :)
  33 2011-01-25 00:43:00 <cosurgi> that's a reasonable result after just an hour of hacking :)
  34 2011-01-25 00:43:35 <Diablo-D3> hour? more like 5 minutes
  35 2011-01-25 00:43:52 <cosurgi> cool
  36 2011-01-25 00:45:33 andrew12 has joined
  37 2011-01-25 00:46:14 <Diablo-D3> ArtForz: your kernel, can I get away with just doing fW0 through 2 = block 0 through 2? that doesnt seem right
  38 2011-01-25 00:46:35 m0mchil has joined
  39 2011-01-25 00:46:43 <Diablo-D3> yo m0
  40 2011-01-25 00:46:46 <m0mchil> hi
  41 2011-01-25 00:46:56 <m0mchil> sup?
  42 2011-01-25 00:47:01 <andrew12> hi
  43 2011-01-25 00:47:02 <Diablo-D3> trying art's kernel
  44 2011-01-25 00:47:22 <andrew12> http://www.jwz.org/blog/2011/01/the-day-the-universe-came/ (may or may not be sfw)
  45 2011-01-25 00:48:36 <m0mchil> I am having some problems with two 5970 on a 790g mobo, any ideas? (setting it up for a friend)
  46 2011-01-25 00:48:47 <Diablo-D3> m0mchil: whats the problem?
  47 2011-01-25 00:49:20 <m0mchil> well, it seems the mb doesn't use/see the second 5970
  48 2011-01-25 00:49:35 <Diablo-D3> check if the bios has a 2nd slot switch
  49 2011-01-25 00:49:39 <Diablo-D3> or a physical one
  50 2011-01-25 00:49:59 <m0mchil> hm, physical?
  51 2011-01-25 00:50:05 <Diablo-D3> jumper or switch
  52 2011-01-25 00:50:17 <Diablo-D3> its probably not a x16 x16 board
  53 2011-01-25 00:50:32 <m0mchil> it's x8/x8
  54 2011-01-25 00:50:36 <Diablo-D3> its probably 16/0 <-> 8/8
  55 2011-01-25 00:50:55 <Diablo-D3> older boards required to be switched manaully
  56 2011-01-25 00:50:59 <Diablo-D3> not sure if they need it anymore
  57 2011-01-25 00:51:28 <m0mchil> I'll try upgrading BIOS tomorrow
  58 2011-01-25 00:51:33 <luke-jr> how many proof of works make a block?
  59 2011-01-25 00:52:37 <Diablo-D3> upgrading? no, just look for an option in it
  60 2011-01-25 00:53:47 <m0mchil> there is a gfx/gfx2 peer-to-peer option - enabling or disabling it didn't change anything
  61 2011-01-25 00:54:08 <sipa> luke-jr: one can be enough
  62 2011-01-25 00:54:19 <Diablo-D3> m0mchil: pretty sure that should be enabled
  63 2011-01-25 00:54:27 <luke-jr> sipa: can be?
  64 2011-01-25 00:54:53 <m0mchil> 790x actually, not 790g
  65 2011-01-25 00:54:54 <Diablo-D3> m0mchil: did you check and make sure lspci or device manager didnt see both cards?
  66 2011-01-25 00:54:58 <sipa> or do you means the average number of hashes that is tried by the network per block?
  67 2011-01-25 00:55:02 <Diablo-D3> there is no 790g
  68 2011-01-25 00:55:15 <Diablo-D3> only x/fx
  69 2011-01-25 00:55:44 <Diablo-D3> 790/890 is a 65nm part without a video core
  70 2011-01-25 00:55:56 <m0mchil> aticonfig showed just two GPUs (single card)
  71 2011-01-25 00:56:06 <Diablo-D3> thats not what I asked.
  72 2011-01-25 00:56:22 <Diablo-D3> Im specifically not trusting the driver here
  73 2011-01-25 00:56:57 <molecular> m0mchil: does the fan of the "inactive" card go crazy?
  74 2011-01-25 00:57:01 <m0mchil> POST showed only GFX1 at x8
  75 2011-01-25 00:57:09 <Diablo-D3> m0mchil: ahh, its a board problem then
  76 2011-01-25 00:57:10 <Diablo-D3> strange
  77 2011-01-25 00:57:11 <m0mchil> molecular: nope
  78 2011-01-25 00:57:14 <Diablo-D3> sure it was in the slot the whole way?
  79 2011-01-25 00:57:39 <luke-jr> sipa: I mean, what makes a proof-of-work a valid block?
  80 2011-01-25 00:57:40 devon_hillard_ has quit (Quit: Leaving)
  81 2011-01-25 00:57:50 devon_hillard has joined
  82 2011-01-25 00:58:02 <m0mchil> wes, green LED on 5970 was lit, fan working... not sure if this shows it's actually fully armed :)
  83 2011-01-25 00:58:14 <Diablo-D3> it doesnt
  84 2011-01-25 00:58:29 <Diablo-D3> m0mchil: swap the cards and see if the card actually works
  85 2011-01-25 00:59:15 <m0mchil> will hit it hard tomorrow... it'll either work...
  86 2011-01-25 01:00:55 <Diablo-D3> m0mchil: a lone card in the 2nd slot MIGHT not work
  87 2011-01-25 01:00:58 <Diablo-D3> some do some dont
  88 2011-01-25 01:02:50 <m0mchil> I'll try it in first slot
  89 2011-01-25 01:03:02 <m0mchil> pretty sure it's mobo/BIOS problem
  90 2011-01-25 01:04:33 <Diablo-D3> hmm
  91 2011-01-25 01:04:47 <Diablo-D3> I think I figured out what his kernel is doing with fW0 through 2
  92 2011-01-25 01:11:31 <Diablo-D3> [1/24/11 8:05:58 PM] ERROR: Invalid block found on ATI RV770 (#1), possible driver or hardware issue
  93 2011-01-25 01:11:33 <Diablo-D3> an improvement!
  94 2011-01-25 01:12:07 <tcatm> still trying to understand the kernel code?
  95 2011-01-25 01:12:33 <Diablo-D3> not understand as much as make work
  96 2011-01-25 01:13:01 <Diablo-D3> fW* is obviously an optimization
  97 2011-01-25 01:13:37 <tcatm> not really
  98 2011-01-25 01:13:54 <tcatm> he just moved code outside the loop
  99 2011-01-25 01:14:10 <Diablo-D3> yes, which IS an optimization
 100 2011-01-25 01:14:15 <Diablo-D3> since you couldnt do this without the loop
 101 2011-01-25 01:14:50 <Diablo-D3> http://pastebin.com/LFkwNNhw
 102 2011-01-25 01:14:54 <Diablo-D3> but obviously thats not right.
 103 2011-01-25 01:15:24 <tcatm> why don't you use the kernel as is?
 104 2011-01-25 01:15:36 <Diablo-D3> because its wrong
 105 2011-01-25 01:15:47 <tcatm> huh? working fine here
 106 2011-01-25 01:16:04 <Diablo-D3> it doesnt match the args m0's and mine use
 107 2011-01-25 01:16:07 <Diablo-D3> so its wrong
 108 2011-01-25 01:16:14 <tcatm> fix your code
 109 2011-01-25 01:16:37 <tcatm> or even fix the args, doesn't really matter
 110 2011-01-25 01:16:45 <Diablo-D3> fix my code? that'd involve fixing bitcoin and all other miners too
 111 2011-01-25 01:16:53 <luke-jr> lol
 112 2011-01-25 01:17:39 <tcatm> just change ctx->* to your args and it should work
 113 2011-01-25 01:17:53 <Diablo-D3> look at the pastebin
 114 2011-01-25 01:18:20 <tcatm> great
 115 2011-01-25 01:18:27 <m0mchil> tcatm, thank you for the patch BTW
 116 2011-01-25 01:18:40 <Diablo-D3> tcatm: but that doesnt look right
 117 2011-01-25 01:18:40 <tcatm> m0mchil: the fake-push thing?
 118 2011-01-25 01:18:46 <m0mchil> yup
 119 2011-01-25 01:18:51 <tcatm> does it work reliable?
 120 2011-01-25 01:19:19 <andrew12> "does it work reliable"
 121 2011-01-25 01:19:27 <m0mchil> hmhmh, didn't tested it much
 122 2011-01-25 01:19:33 <Diablo-D3> reliably damnit
 123 2011-01-25 01:20:20 <tcatm> it doesn't reconnect to the node when the connection dies so it might end up only sending getworks when it found a solution (= worse than before)
 124 2011-01-25 01:21:34 <m0mchil> am I getting it right that it waits for a new block to be broadcast to request new work from bitcoind / pool?
 125 2011-01-25 01:21:47 <tcatm> yep, block or tx
 126 2011-01-25 01:22:16 <andrew12> hm
 127 2011-01-25 01:22:47 <m0mchil> perhaps block or 60 seconds would be better, not sure
 128 2011-01-25 01:22:50 <andrew12> what was the usd/btc rate in october 2010?
 129 2011-01-25 01:22:56 <Diablo-D3> hey m0mchil
 130 2011-01-25 01:22:58 Cusipzzz has joined
 131 2011-01-25 01:23:03 <Diablo-D3> whats the first three ints of the second block of the header?
 132 2011-01-25 01:23:05 <tcatm> so if you fix the reconnect thing it should reduce getwork/s a lot
 133 2011-01-25 01:23:26 <tcatm> Diablo-D3: merke, ntime, nnonce?
 134 2011-01-25 01:23:35 <tcatm> merkle*
 135 2011-01-25 01:23:43 <Diablo-D3> its not merkle, ntime, and nbits?
 136 2011-01-25 01:23:50 <tcatm> ah yes
 137 2011-01-25 01:23:54 <tcatm> i missed nbits
 138 2011-01-25 01:23:57 <m0mchil> Diablo-D3, yes, end of merkle, time, difficulty
 139 2011-01-25 01:24:00 echelon has joined
 140 2011-01-25 01:24:04 <Diablo-D3> damnit, so Im doing this right
 141 2011-01-25 01:24:15 <tcatm> Diablo-D3: can you paste your complete kernel?
 142 2011-01-25 01:24:28 <Diablo-D3> you've seen his, right?
 143 2011-01-25 01:24:32 <tcatm> yep
 144 2011-01-25 01:24:34 <m0mchil> tcatm: I am reluctant to distribute python bitcoin node with miner... for now
 145 2011-01-25 01:24:47 <echelon> can someone add iy6ni3wkqazp4ytu.onion and bitcoinbudtoeks7.onion to the fallback nodes?
 146 2011-01-25 01:25:28 <tcatm> echelon: which fallback nodes? those hardcoded in the source?
 147 2011-01-25 01:25:45 <echelon> on the wiki.. http://www.bitcoin.org/wiki/doku.php?id=fallback_nodes
 148 2011-01-25 01:25:47 <Diablo-D3> http://pastebin.com/iJByrL4e
 149 2011-01-25 01:25:56 <Diablo-D3> tcatm: the rest is the same, I commented out the loop to make debugging easier
 150 2011-01-25 01:26:00 <tcatm> echelon: there's a new wiki at en.bitcoin.it
 151 2011-01-25 01:26:00 <m0mchil> tcatm, also, 5870 would need new hash space every ~12 seconds
 152 2011-01-25 01:26:09 <echelon> oh
 153 2011-01-25 01:26:52 <echelon> why is it on a different domain
 154 2011-01-25 01:27:08 <andrew12> because its a different server
 155 2011-01-25 01:27:41 <tcatm> m0mchil: well you could detect if it's close to nonce overrun and issue a getwork
 156 2011-01-25 01:28:02 <tcatm> or just keep askrate at 5s and additionally getwork on new blocks/tx
 157 2011-01-25 01:28:15 <tcatm> or even better: switch to push
 158 2011-01-25 01:28:45 <m0mchil> yes, and there are many slower GPUs... but yes, push it is
 159 2011-01-25 01:28:49 <tcatm> Diablo-D3: what did you do to W3/tnonce?
 160 2011-01-25 01:29:07 <m0mchil> I now deeply regret my initial getwork...
 161 2011-01-25 01:29:31 <tcatm> getwork was just logical as it was very easy to add to the RPC interface
 162 2011-01-25 01:29:43 <tcatm> push requires a multithreaded TCP server
 163 2011-01-25 01:29:46 devon_hillard has quit (Read error: Connection reset by peer)
 164 2011-01-25 01:30:13 <tcatm> did you start any work on a push interface to bitcoind?
 165 2011-01-25 01:31:02 <m0mchil> no
 166 2011-01-25 01:31:23 <molecular> didn't slush start work on implementing push on his pool using a proxy?
 167 2011-01-25 01:31:35 <Diablo-D3> tcatm: turned loop off
 168 2011-01-25 01:31:46 <tcatm> I'm developing a binary push-based interface (actually porting my push based approach)
 169 2011-01-25 01:32:39 <m0mchil> I am not sure if push should get in official bitcoind... in custom setups it's ok
 170 2011-01-25 01:33:03 <luke-jr> s/official/original
 171 2011-01-25 01:33:22 <tcatm> That's why I made it as a module that can be patched in.
 172 2011-01-25 01:33:29 <m0mchil> great
 173 2011-01-25 01:33:47 <Diablo-D3> tcatm: so Im not sure where it went wrong
 174 2011-01-25 01:34:14 <tcatm> https://github.com/tcatm/bitcoin/tree/FMI early dev state (just a skeleton)
 175 2011-01-25 01:35:02 <m0mchil> and... on a side note... we need more exchanges... moar!
 176 2011-01-25 01:35:15 <m0mchil> my nightmare is mtgox just dissapears
 177 2011-01-25 01:35:15 <tcatm> Diablo-D3: you know, that the code only searches for H==0?
 178 2011-01-25 01:35:22 <tcatm> m0mchil: bitcoin central is opensource
 179 2011-01-25 01:35:25 <Diablo-D3> tcatm: yeah, so does mine
 180 2011-01-25 01:35:40 <tcatm> so what's wrong?
 181 2011-01-25 01:35:52 <Diablo-D3> tcatm: its returning bullshit results
 182 2011-01-25 01:35:59 <tcatm> wait, you're searching for H==0 without the loop?
 183 2011-01-25 01:36:37 <Diablo-D3> yes, the if(H + shit) line just does output[0] = tnonce; like it does in my kernel
 184 2011-01-25 01:36:59 <m0mchil> bitcoin central is nice... but not convenient as mtgox. And mtgox can be much more convenient tha now
 185 2011-01-25 01:37:30 <tcatm> Diablo-D3: what's output[0]? A global variable for all kernels?
 186 2011-01-25 01:38:29 <tcatm> Where's your code?
 187 2011-01-25 01:38:39 noagendamarket has joined
 188 2011-01-25 01:38:53 noagendamarket has quit (Changing host)
 189 2011-01-25 01:38:53 noagendamarket has joined
 190 2011-01-25 01:38:53 * Diablo-D3 facepalms
 191 2011-01-25 01:39:18 <Diablo-D3> https://github.com/Diablo-D3/DiabloMiner
 192 2011-01-25 01:42:03 m0mchil has quit ()
 193 2011-01-25 01:42:19 <tcatm> strange code
 194 2011-01-25 01:42:27 <Diablo-D3> strange in what way?
 195 2011-01-25 01:42:32 <Diablo-D3> its almost identical to m0's
 196 2011-01-25 01:43:20 <tcatm> around line 448 if(buffer.getInt(0)...
 197 2011-01-25 01:43:36 <tcatm> what's that block supposed to do?
 198 2011-01-25 01:43:59 <Diablo-D3> its a bytebuffer, it gets an int.
 199 2011-01-25 01:44:16 <tcatm> the whole if-block?
 200 2011-01-25 01:45:05 <Diablo-D3> that line isnt near 448
 201 2011-01-25 01:45:12 <Diablo-D3> oh wait yes it is, 444
 202 2011-01-25 01:45:17 <Diablo-D3>           if(buffer.getInt(0) > 0) {
 203 2011-01-25 01:45:27 <Diablo-D3> if the returned int is above 0
 204 2011-01-25 01:45:39 <tcatm> oh, github messes up line numbers again, sorry
 205 2011-01-25 01:45:53 <tcatm> yes, but what's the logic?
 206 2011-01-25 01:46:28 <tcatm> is it: if kernel returns nonce != 0 calculate the hash again, this time including G, and check result, then reset buffer to 0?
 207 2011-01-25 01:46:36 <Diablo-D3> yes
 208 2011-01-25 01:46:45 <Diablo-D3> the miner fully calculates the hash to catch gpu bullshit
 209 2011-01-25 01:48:40 <tcatm> What about every miner thread (in GPU) return to output[myid] instead of output[0] and also only return a flag (0 or !0) instead of nonce, then loop (in java) through that array re-calculating every nonce that has the flag set?
 210 2011-01-25 01:49:04 <Diablo-D3> why would it?
 211 2011-01-25 01:49:19 <Diablo-D3> the only way that'd make sense if I was likely to get more than one valid nonce per session
 212 2011-01-25 01:49:34 <Diablo-D3> s/session/kernel/
 213 2011-01-25 01:50:02 <tcatm> That's a lot more likely than solving the block at current difficulties
 214 2011-01-25 01:50:11 <echelon> does "#" work as a comment in bitcoin.conf?
 215 2011-01-25 01:50:17 <Diablo-D3> its not as likely as you think
 216 2011-01-25 01:50:27 <Diablo-D3> when I last tested that, it had zero change on my rate of H==0
 217 2011-01-25 01:50:45 <Diablo-D3> not only that, my rate of H==0 matches the estimated output.
 218 2011-01-25 01:50:50 <tcatm> My miner find H==0 quite often, even multiple times per kernel run
 219 2011-01-25 01:51:16 <Diablo-D3> anyhow, this doesnt solve my problem
 220 2011-01-25 01:52:36 <Diablo-D3> tcatm: btw
 221 2011-01-25 01:52:39 <Diablo-D3> setting a flag makes no sense
 222 2011-01-25 01:52:44 <Diablo-D3> != 0 IS the flag
 223 2011-01-25 01:52:54 <Diablo-D3> so it'd only change a single output to a few thousand outputs.
 224 2011-01-25 01:53:02 <Diablo-D3> and art's kernel still suffers from the same problem
 225 2011-01-25 01:53:17 <Diablo-D3> a loop of 1024 nonces could still, theoretically, have multiple valid nonces
 226 2011-01-25 01:53:25 <tcatm> yep
 227 2011-01-25 01:53:48 <tcatm> that's why he calculates all those 1024 hashes again to find any hash<=target
 228 2011-01-25 01:54:16 <Diablo-D3> heh
 229 2011-01-25 01:54:35 <tcatm> doing 1024 or even 4096 hashes in CPU costs almost no time, you can do that while waiting for getwork response
 230 2011-01-25 01:54:41 <Diablo-D3> he only gave the kernel, not the miner source
 231 2011-01-25 01:55:07 <tcatm> he described his miner pretty well over the last months
 232 2011-01-25 01:55:15 <Diablo-D3> not as well as you think
 233 2011-01-25 01:55:15 dukeleto has quit (Excess Flood)
 234 2011-01-25 01:55:19 <Diablo-D3> anyhow, this doesnt solve my problem
 235 2011-01-25 01:55:45 dukeleto has joined
 236 2011-01-25 01:56:12 <tcatm> hm hm
 237 2011-01-25 01:56:34 <molecular> I might be talking crap: but what about the 3 rounds at the beginning and at the end. aren't these included in art's loop?
 238 2011-01-25 01:56:47 <Diablo-D3> molecular: dont think so
 239 2011-01-25 01:56:50 <echelon> what's the timeout for each connection on tor?
 240 2011-01-25 01:56:52 <Diablo-D3> at least I know the last three arent
 241 2011-01-25 01:56:52 <echelon> err
 242 2011-01-25 01:56:54 <echelon> on bitcoin
 243 2011-01-25 01:57:03 <comboy> Diablo-D3: thanks for your miner, great work, very stable and my cpu stays cool
 244 2011-01-25 01:58:00 <Diablo-D3> molecular: notice it stops at W12 and then does if(H + shit)... thats the hallmark of one that doesnt do the last 3
 245 2011-01-25 01:58:06 <Diablo-D3> comboy: welcome
 246 2011-01-25 01:58:23 Slix` has joined
 247 2011-01-25 01:58:30 <molecular> Diablo-D3, ok
 248 2011-01-25 01:58:35 <tcatm> Diablo-D3: http://pastebin.com/Bghjj3Lr this code calculates the ctx (here called blk)
 249 2011-01-25 01:58:52 <Diablo-D3> tcatm: hah
 250 2011-01-25 01:59:11 <Diablo-D3> tcatm: its identical to my code
 251 2011-01-25 01:59:21 <comboy> anybody know if and can I find somewhere history of difficulty over time?
 252 2011-01-25 01:59:26 <tcatm> Diablo-D3: great
 253 2011-01-25 01:59:57 <Diablo-D3> tcatm: but he misses out on the obvious win
 254 2011-01-25 02:00:03 <tcatm> comboy: http://nullvoid.org/bitcoin/difficultiez.php
 255 2011-01-25 02:00:06 <Diablo-D3> you only need 6 of the 8 parts for cty
 256 2011-01-25 02:00:24 <tcatm> saving what? 2 mov's?
 257 2011-01-25 02:00:31 <Diablo-D3> because A and E dont change with only 2 rounds
 258 2011-01-25 02:00:32 <comboy> sweeet, thanks! :)
 259 2011-01-25 02:00:34 <Diablo-D3> er 3 rounds
 260 2011-01-25 02:00:55 <tcatm> also that precalc_hash is from my miner
 261 2011-01-25 02:01:14 <Diablo-D3> both me and m0 do it this way
 262 2011-01-25 02:01:37 <tcatm> sure, the compiler optimizes that pretty good
 263 2011-01-25 02:02:11 <tcatm> + 64 more bits as a constant don't impact performance at all as the DMA will probably transfer a larger chunk of memory anyway
 264 2011-01-25 02:02:24 * Diablo-D3 shrugs
 265 2011-01-25 02:02:26 <Diablo-D3> took it out anyhow
 266 2011-01-25 02:06:13 <Diablo-D3> hrrrrrrrm
 267 2011-01-25 02:06:14 <Diablo-D3> tcatm: hey you
 268 2011-01-25 02:06:20 <tcatm> ?
 269 2011-01-25 02:06:24 <Diablo-D3> show me the source for R
 270 2011-01-25 02:06:28 <tcatm> r?
 271 2011-01-25 02:06:30 <Diablo-D3> the macro
 272 2011-01-25 02:06:42 <tcatm> Just a SHA round
 273 2011-01-25 02:06:51 <Diablo-D3> yeah, but does it match mine/
 274 2011-01-25 02:07:12 <tcatm> maybe?
 275 2011-01-25 02:07:28 <tcatm> http://pastebin.com/LEsKW5cB
 276 2011-01-25 02:08:08 <Diablo-D3> rut row
 277 2011-01-25 02:08:21 <Diablo-D3> those rotates seem to be the opposite direction of mine
 278 2011-01-25 02:08:41 <tcatm> #define rotate(x,y) ((x<<y) | (x>>(sizeof(x)*8-y)))
 279 2011-01-25 02:09:06 <Diablo-D3>   static int rot(int x, int y) {
 280 2011-01-25 02:09:06 <Diablo-D3>     return (x >>> y) | (x << (32 - y));
 281 2011-01-25 02:09:06 <Diablo-D3>   }
 282 2011-01-25 02:09:07 <Diablo-D3> faaawk
 283 2011-01-25 02:09:20 <Diablo-D3> no wonder this shits broke
 284 2011-01-25 02:10:09 lfm has quit (Read error: Connection reset by peer)
 285 2011-01-25 02:10:25 <Diablo-D3> meh nm
 286 2011-01-25 02:10:30 <Diablo-D3> it should end up producing the same answer
 287 2011-01-25 02:10:40 lfm has joined
 288 2011-01-25 02:11:16 <Diablo-D3> 32 - 6 = 26, 32 - 11 = 21, 32 - 25 = 7
 289 2011-01-25 02:11:17 <tcatm> yep
 290 2011-01-25 02:11:26 <Diablo-D3> so its not that
 291 2011-01-25 02:11:28 <Diablo-D3> damnit
 292 2011-01-25 02:11:28 <tcatm> I'm just rotating right
 293 2011-01-25 02:11:58 <tcatm> what's x >>> y (triple >)?
 294 2011-01-25 02:12:11 <Diablo-D3> C >>
 295 2011-01-25 02:12:20 <tcatm> what's >> in java?
 296 2011-01-25 02:12:28 <Diablo-D3> same thing, but it respects the sign
 297 2011-01-25 02:12:42 <tcatm> strange language
 298 2011-01-25 02:13:02 <Diablo-D3> well, Java has no unsigned variables
 299 2011-01-25 02:13:07 <dirtyfilthy> yeh it sucks
 300 2011-01-25 02:13:13 <Diablo-D3> C >> on signed will do the same thing as Java >>
 301 2011-01-25 02:13:19 <Diablo-D3> it wont shift the sign bit
 302 2011-01-25 02:13:27 <Diablo-D3> so you need >>> to do full shifts
 303 2011-01-25 02:13:46 <Diablo-D3> its just one of those things
 304 2011-01-25 02:14:18 <tcatm> can you paste your whole kernel.cl?
 305 2011-01-25 02:15:02 <Diablo-D3> http://pastebin.com/CG1CbRGu
 306 2011-01-25 02:16:00 <tcatm> hm hm
 307 2011-01-25 02:16:05 <tcatm> this is funny
 308 2011-01-25 02:16:46 <Diablo-D3> cant figure out why it fails?
 309 2011-01-25 02:16:52 <Diablo-D3> it should be identical if his kernel works in the first place
 310 2011-01-25 02:17:14 <tcatm> did you verify his kernel?
 311 2011-01-25 02:18:07 <lfm> is it some byteswapping thing? he sed he doesnt but getwork does
 312 2011-01-25 02:18:23 <tcatm> shouldn't matter in this case
 313 2011-01-25 02:19:00 <luke-jr> should 500 W PSU be enough for two HD 5770?
 314 2011-01-25 02:23:21 <Diablo-D3> doesnt matter, I byteswap to make mine work
 315 2011-01-25 02:23:29 <Diablo-D3> tcatm: havent verified it
 316 2011-01-25 02:23:31 <Diablo-D3> but I assume it works
 317 2011-01-25 02:23:38 kermit has quit (Quit: Leaving.)
 318 2011-01-25 02:23:41 <lfm> luke ati recomends 600
 319 2011-01-25 02:24:49 tcatm has quit (Read error: Operation timed out)
 320 2011-01-25 02:25:17 <luke-jr> lfm: for one, or for two?
 321 2011-01-25 02:25:27 <lfm> for ywo
 322 2011-01-25 02:25:32 <lfm> two
 323 2011-01-25 02:25:48 <luke-jr> hmm
 324 2011-01-25 02:25:51 <luke-jr> think I could resell one?
 325 2011-01-25 02:28:05 <Keefe> luke-jr: how much 12V watts can your psu do?
 326 2011-01-25 02:28:14 <luke-jr> ?
 327 2011-01-25 02:28:19 <luke-jr> dunno?
 328 2011-01-25 02:28:28 <Keefe> tell me the model # and i'll look it up
 329 2011-01-25 02:28:35 <luke-jr> http://www.newegg.com/Product/Product.aspx?Item=N82E16817371007
 330 2011-01-25 02:29:28 <luke-jr> auction ends in 5 min XD
 331 2011-01-25 02:29:36 <Keefe> 408W 12V
 332 2011-01-25 02:29:39 <Keefe> auction for what?
 333 2011-01-25 02:29:46 <luke-jr> $155 for two 5770
 334 2011-01-25 02:30:11 <Keefe> if you have a low-end system, 408W 12V should be plenty for two 5770's
 335 2011-01-25 02:30:42 <luke-jr> I think it uses about 150 W right now
 336 2011-01-25 02:30:51 <Keefe> my mining boxes have sempron 140 and 1gb ram and one drive, and use about 50W without graphics
 337 2011-01-25 02:30:59 <luke-jr> (measured at AC level and calculated)
 338 2011-01-25 02:31:06 <Keefe> still should be ok
 339 2011-01-25 02:31:19 <Keefe> 5770 uses ~110W iirc
 340 2011-01-25 02:31:21 <luke-jr> $155 sound good deal for that?
 341 2011-01-25 02:31:24 <Keefe> yep
 342 2011-01-25 02:31:36 <Keefe> but i bet the bidding goes much higher :)
 343 2011-01-25 02:31:50 <luke-jr> well, what should I max?
 344 2011-01-25 02:31:59 <Keefe> $200
 345 2011-01-25 02:32:04 <luke-jr> looks like one 5770 goes $146 retail
 346 2011-01-25 02:32:13 <Keefe> i'm selling mine for $120
 347 2011-01-25 02:32:26 <Keefe> if you are really patient, you could probably get one for $100 on ebay
 348 2011-01-25 02:32:28 <luke-jr> $195.01 + $5 shipping then? :P
 349 2011-01-25 02:32:32 <luke-jr> hmm
 350 2011-01-25 02:32:35 <luke-jr> ;;bc,calc
 351 2011-01-25 02:32:35 <gribble> (bc,calc <an alias, 1 argument>) -- Alias for "echo The average time to generate a block at $1 Khps, given current difficulty of [bc,diff], is [time elapsed [math calc 1/((2**224-1)/[bc,diff]*$1*1000/2**256)]]".
 352 2011-01-25 02:32:37 <lfm> i got one 5770 new for 105 on sale
 353 2011-01-25 02:32:52 <luke-jr> but one for $100 is > 2 for $190
 354 2011-01-25 02:32:55 <luke-jr> ;;bc,stats
 355 2011-01-25 02:32:57 <gribble> Current Blocks: 104441 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 390 blocks | Next Difficulty In About: 2 days, 5 hours, 50 minutes, and 30 seconds | Next Difficulty Estimate: 21920.36910764
 356 2011-01-25 02:33:42 <luke-jr> I bit 165.01
 357 2011-01-25 02:34:10 <luke-jr> outbid :p
 358 2011-01-25 02:35:56 <Keefe> auction over?
 359 2011-01-25 02:36:33 <luke-jr> yeah
 360 2011-01-25 02:36:39 <luke-jr> it went up past $200 total
 361 2011-01-25 02:37:00 <Keefe> link?
 362 2011-01-25 02:37:23 <Keefe> nvm, found it
 363 2011-01-25 02:38:37 <Keefe> i can sell you mine for $130 incl shipping and can ship it tomorrow
 364 2011-01-25 02:39:30 <Keefe> not a great price though. but i'm patient
 365 2011-01-25 02:39:54 <Keefe> fwiw, it OCs to 960 at stock V just fine
 366 2011-01-25 02:40:15 <Keefe> gigabyte super overclock edition
 367 2011-01-25 02:41:25 <Keefe> i ran a 5970 and 5770 together on a psu with 450W 12V for a month before upgrading
 368 2011-01-25 02:41:41 <Keefe> but my base system only pulls 50W
 369 2011-01-25 02:41:44 <Diablo-D3> damnit
 370 2011-01-25 02:41:48 <Diablo-D3> where the fuck is art when you need him,
 371 2011-01-25 02:42:00 <Kiba> hey
 372 2011-01-25 02:42:07 <Kiba> you want to draw art?
 373 2011-01-25 02:42:21 <Diablo-D3> no, I want art to figure out why his kernel doesnt work
 374 2011-01-25 02:42:28 <Keefe> Diablo-D3: can't you figure out that opencl stuff on your own?
 375 2011-01-25 02:42:42 <luke-jr> Keefe: I'm not in a rush
 376 2011-01-25 02:42:50 <luke-jr> I passed by a 5770 for $100 earlier
 377 2011-01-25 02:42:59 <luke-jr> I figure 5770 is only worth $80 at most :p
 378 2011-01-25 02:43:07 <Keefe> ok :)
 379 2011-01-25 02:43:40 tcatm has joined
 380 2011-01-25 02:43:41 tcatm has quit (Changing host)
 381 2011-01-25 02:43:41 tcatm has joined
 382 2011-01-25 02:43:48 <tcatm> re
 383 2011-01-25 02:43:56 <tcatm> Diablo-D3: does it work now?
 384 2011-01-25 02:44:18 <Diablo-D3> tcatm: no?
 385 2011-01-25 02:44:24 <Keefe> luke-jr: but keep in mind that if it takes you a week more to find the deal you want, you might as well have paid $20 and got it right away
 386 2011-01-25 02:45:13 <Keefe> more like $28/wk actually
 387 2011-01-25 02:45:28 <luke-jr> for a 5770?
 388 2011-01-25 02:45:31 <Keefe> yep
 389 2011-01-25 02:45:36 <tcatm> Diablo-D3: you made a mistake while copy&pasting the code
 390 2011-01-25 02:45:49 <luke-jr> ;;bc,calc 156000
 391 2011-01-25 02:45:49 <gribble> The average time to generate a block at 156000 Khps, given current difficulty of 18437.64439217 , is 5 days, 21 hours, and 22 seconds
 392 2011-01-25 02:46:01 <Keefe> ;;bc,calc 177000
 393 2011-01-25 02:46:02 <gribble> The average time to generate a block at 177000 Khps, given current difficulty of 18437.64439217 , is 5 days, 4 hours, 16 minutes, and 35 seconds
 394 2011-01-25 02:46:09 <luke-jr> $17/wk :p
 395 2011-01-25 02:46:46 <luke-jr> ;;bc,calc 240000
 396 2011-01-25 02:46:46 <gribble> The average time to generate a block at 240000 Khps, given current difficulty of 18437.64439217 , is 3 days, 19 hours, 39 minutes, and 14 seconds
 397 2011-01-25 02:47:25 <luke-jr> $181 max bid on single 5850 sound good?
 398 2011-01-25 02:47:56 <Diablo-D3> tcatm: er, I did?
 399 2011-01-25 02:48:20 <luke-jr> actually, for $184 I could BuyItNow
 400 2011-01-25 02:48:31 <Keefe> luke-jr: $28/wk at $0.42/btc and 177 mhps (960mhz OC)
 401 2011-01-25 02:48:36 <tcatm> Diablo-D3: yep
 402 2011-01-25 02:48:47 <Diablo-D3> which line?
 403 2011-01-25 02:48:49 <luke-jr> Keefe: I don't OC
 404 2011-01-25 02:49:11 <luke-jr> and this is all assuming I manage to get a GPU miner working
 405 2011-01-25 02:49:21 <tcatm> Diablo-D3: fW0 and the following few lines
 406 2011-01-25 02:49:22 <Keefe> so $25/wk then
 407 2011-01-25 02:49:25 <luke-jr> which isn't looking too good from my experience with the Mac
 408 2011-01-25 02:49:43 <luke-jr> although nevermind… I just remembered, the Mac OS ver is too old for OpenCL XD
 409 2011-01-25 02:50:56 <Keefe> $180 for 5850 sounds fair
 410 2011-01-25 02:50:57 kermit has joined
 411 2011-01-25 02:51:53 <luke-jr> but $184 too much? :P
 412 2011-01-25 02:52:13 <Keefe> why do you ask me
 413 2011-01-25 02:52:17 <hacim> whats the channel with live block stats?
 414 2011-01-25 02:52:18 * luke-jr is including "$15" shipping in all these btw
 415 2011-01-25 02:52:19 <Keefe> i prefer 5970's
 416 2011-01-25 02:52:25 <luke-jr> hacim: #bitcoin-monitor
 417 2011-01-25 02:52:30 <luke-jr> Keefe: sure, but that's $600
 418 2011-01-25 02:52:36 <hacim> ah thanks
 419 2011-01-25 02:52:42 <Keefe> i'll sell you a new 5970 for $460
 420 2011-01-25 02:53:13 <Keefe> $478 incl sh
 421 2011-01-25 02:53:20 acous has quit (Remote host closed the connection)
 422 2011-01-25 02:53:28 <luke-jr> 5850 still a better deal :P
 423 2011-01-25 02:53:31 <Keefe> (within usa)
 424 2011-01-25 02:53:48 <hacim> if you spend a lot of time looking, you can get a 5970 for cheaper than that
 425 2011-01-25 02:53:52 <tcatm> Diablo-D3: also there's another difference in your code and art's
 426 2011-01-25 02:53:53 <molecular> tcatm you mean it should read
 427 2011-01-25 02:53:59 <Keefe> time is money ;)
 428 2011-01-25 02:54:02 <molecular>   const uint fW0 = block0 + (rotr(block1, 7) ^ rotr(block1, 18) ^ (block1 >> 3));
 429 2011-01-25 02:54:05 <molecular> const uint fW1 = block1 + (rotr(block2, 7) ^ rotr(block2, 18) ^ (block2 >> 3)) + 0x01100000;
 430 2011-01-25 02:54:11 <luke-jr> 5850 = 1.33 MH/s/$ vs 5970 = 1.12 MH/s/$
 431 2011-01-25 02:54:30 <Diablo-D3> argh
 432 2011-01-25 02:54:46 * Diablo-D3 smacks art
 433 2011-01-25 02:54:55 <tcatm> better smack yourself
 434 2011-01-25 02:54:58 <luke-jr> Keefe: $405 total or no sale! :P
 435 2011-01-25 02:55:02 <Keefe> no deal
 436 2011-01-25 02:55:03 <Kiba> Diablo-D3: you want to pay me to draw some art for ya?
 437 2011-01-25 02:55:22 <noagendamarket> so who bought all the 5970's in the Us ?
 438 2011-01-25 02:55:24 <noagendamarket> lol
 439 2011-01-25 02:55:28 <Diablo-D3> noagendamarket: art did
 440 2011-01-25 02:55:35 <noagendamarket> heh
 441 2011-01-25 02:55:36 <Keefe> i bought 9
 442 2011-01-25 02:55:38 <Diablo-D3> tcatm: whats the other difference?
 443 2011-01-25 02:55:39 <luke-jr> Keefe: in reality, AMD says I'd need to buy a new PSU to get it anyway
 444 2011-01-25 02:55:43 <Diablo-D3> art bought... what, 26?
 445 2011-01-25 02:55:55 <noagendamarket> I can only find them in canada
 446 2011-01-25 02:56:01 kermit has quit (Quit: Leaving.)
 447 2011-01-25 02:56:05 <noagendamarket> http://www.a-power.com/product-14264
 448 2011-01-25 02:56:12 kermit has joined
 449 2011-01-25 02:56:24 <noagendamarket> yeah art bought that many
 450 2011-01-25 02:56:37 <Keefe> for a 5970, ya 408W 12V isn't enough when you're starting at 150W
 451 2011-01-25 02:57:00 <noagendamarket> we need to source some for #bitcoind
 452 2011-01-25 02:57:49 <tcatm> Diablo-D3: you don't want to know
 453 2011-01-25 02:58:19 <Diablo-D3> theres none
 454 2011-01-25 02:58:22 <Diablo-D3> kernel works now
 455 2011-01-25 02:59:46 * molecular can verify that
 456 2011-01-25 02:59:52 <Diablo-D3> okay, what is it then?
 457 2011-01-25 02:59:58 <molecular> now what has to be done to put the loop in?
 458 2011-01-25 03:00:16 <Diablo-D3> yes obviously the loop is turned off
 459 2011-01-25 03:00:23 <molecular> I know
 460 2011-01-25 03:00:27 <tcatm> turning the loop off is stupid
 461 2011-01-25 03:00:35 <Diablo-D3> tcatm: its for debugging reasons
 462 2011-01-25 03:01:31 <tcatm> so turn it on
 463 2011-01-25 03:01:44 <molecular> any change to the java-code necessary when loop is on?
 464 2011-01-25 03:02:03 <Diablo-D3> tcatm: what is art feeding the kernel for the nonce base?
 465 2011-01-25 03:02:13 <Diablo-D3> is he skipping 1024 at a time in the host code?
 466 2011-01-25 03:03:56 <tcatm> what would you say? RTFC?
 467 2011-01-25 03:04:32 <Diablo-D3> RTSL
 468 2011-01-25 03:04:44 <Diablo-D3> but neither your code nor arts is open source
 469 2011-01-25 03:05:17 <tcatm> err.. the .cl contains all information?
 470 2011-01-25 03:05:55 ArtForzZz has joined
 471 2011-01-25 03:06:14 <Diablo-D3> tcatm: yes, but I'd still like to know, for sure, what its being fed
 472 2011-01-25 03:07:05 <tcatm> well if you're lazy you could do tnonce = (ctx->nonce + myid)<<10
 473 2011-01-25 03:07:49 <tcatm> then loop for(ctx->nonce = 0; ctx->nonce < ~0>>10; ctx->nonce++)
 474 2011-01-25 03:08:06 <Diablo-D3> well, if I was REALLY lazy
 475 2011-01-25 03:08:17 <Diablo-D3> ... yeah, it'd be that
 476 2011-01-25 03:09:03 <Diablo-D3> (base + myid) << 10 seems to be right for my use case
 477 2011-01-25 03:09:16 ArtForz has quit (Ping timeout: 272 seconds)
 478 2011-01-25 03:09:19 <tcatm> it avoids nasty byteswaps :P
 479 2011-01-25 03:09:20 bk128 has quit (Quit: bk128)
 480 2011-01-25 03:09:58 <tcatm> while you're at it, why not move f* to java and add args to the kernel?
 481 2011-01-25 03:10:42 <Diablo-D3> tcatm: I might at some point
 482 2011-01-25 03:12:28 <tcatm> also (E & F) | (G & (E & F)) might be optimized a little
 483 2011-01-25 03:12:44 <Diablo-D3> probably
 484 2011-01-25 03:13:03 <Diablo-D3> /AMD doesn't do BFI_INT via bitselect yet; when it does change (g^(e&(f^g))) to (bitselect(e, g, f))
 485 2011-01-25 03:13:12 <tcatm> saving you 1..2 instructions per round
 486 2011-01-25 03:13:13 <Diablo-D3> thats a note from my kernel
 487 2011-01-25 03:13:23 <Diablo-D3> the REAL optimization is using the hardware for it
 488 2011-01-25 03:13:31 <tcatm> even without bitselect you can cache half of the calculation
 489 2011-01-25 03:13:45 <Diablo-D3> yeah, but jesus, fucking AMD
 490 2011-01-25 03:14:56 <Diablo-D3> I'll have to limit the nonce to 2**22, btw
 491 2011-01-25 03:15:18 <tcatm> why?
 492 2011-01-25 03:15:25 <tcatm> wait, that's what I told you?
 493 2011-01-25 03:15:29 <Diablo-D3>   const uint tnonce = (base + myid) << 10;
 494 2011-01-25 03:15:32 <Diablo-D3> thats why
 495 2011-01-25 03:15:51 <Diablo-D3> since Im consuming nonces 2**10 times faster, the max isnt 2**32 anymore
 496 2011-01-25 03:16:00 <tcatm> yep
 497 2011-01-25 03:16:13 <tcatm> guess what's ~0>>10+1 is?
 498 2011-01-25 03:16:30 <Diablo-D3> 0?
 499 2011-01-25 03:16:59 <Diablo-D3> oh wait you said ~
 500 2011-01-25 03:17:12 <tcatm> yep
 501 2011-01-25 03:17:13 <Diablo-D3> its, what, 2**21?
 502 2011-01-25 03:18:18 <tcatm> 2**22
 503 2011-01-25 03:20:15 <Diablo-D3> heh
 504 2011-01-25 03:20:17 <molecular> I have it running with a loop-count of 16 here. of course the hashmeter is wrong. but when I multiply with 16 it gives me 528mhash/s (compared to 535mhash/s before).
 505 2011-01-25 03:20:38 <molecular> damn
 506 2011-01-25 03:20:51 <Diablo-D3> what hardware?
 507 2011-01-25 03:20:59 <tcatm> 5970
 508 2011-01-25 03:21:00 <molecular> 1 5970 stock mhz
 509 2011-01-25 03:21:07 <luke-jr> ;;bc,calc 53000
 510 2011-01-25 03:21:08 <gribble> The average time to generate a block at 53000 Khps, given current difficulty of 18437.64439217 , is 2 weeks, 3 days, 7 hours, 2 minutes, and 13 seconds
 511 2011-01-25 03:21:25 <Diablo-D3> lets see what it does on my 4850
 512 2011-01-25 03:21:37 <tcatm> My 5970 do 556 Mhash/s stock
 513 2011-01-25 03:21:48 <molecular> I think I saw some higher value before, when the loop was disabled.. will have to verify that, though
 514 2011-01-25 03:22:02 <molecular> tcatm using which miner?
 515 2011-01-25 03:22:08 <tcatm> m0's
 516 2011-01-25 03:22:22 <Diablo-D3> yeah but you got some weird shit going on
 517 2011-01-25 03:22:31 <molecular> weird, couldn't get that kind of performance with m0s
 518 2011-01-25 03:22:37 <tcatm> Diablo-D3: not really
 519 2011-01-25 03:22:48 <tcatm> you could easily port it to your miner
 520 2011-01-25 03:24:10 <tcatm> Actually I'm currently using m0's kernel. unmodified
 521 2011-01-25 03:24:18 <Diablo-D3> dont you have that mismatched stream sdk set?
 522 2011-01-25 03:24:27 <tcatm> yep
 523 2011-01-25 03:24:32 <tcatm> 2.3 compiler backported to 2.1
 524 2011-01-25 03:25:01 <molecular> that could well make a difference
 525 2011-01-25 03:25:26 <tcatm> where backported = cp clc and llc and some libs to 2.1 directory :P
 526 2011-01-25 03:25:32 sabalaba has joined
 527 2011-01-25 03:26:39 <molecular> what did you not copy back? I mean: what's wrong with using sdk 2.3 fully?
 528 2011-01-25 03:26:57 <tcatm> the opencl lib is broken
 529 2011-01-25 03:27:09 sabalaba has quit (Read error: Connection reset by peer)
 530 2011-01-25 03:27:13 <Diablo-D3> molecular: basically, the actual compiler is in the driver not steam sdk
 531 2011-01-25 03:27:42 <Diablo-D3> molecular: also, it may not matter with this kernel
 532 2011-01-25 03:27:43 <tcatm> btw, you can patch llc to produce bfi instead of bytealign
 533 2011-01-25 03:29:31 <Diablo-D3> yeah probably
 534 2011-01-25 03:29:32 <molecular> just hung my system using 256 loops
 535 2011-01-25 03:29:37 <Diablo-D3> molecular: not hung
 536 2011-01-25 03:29:42 <Diablo-D3> loop is just going to take a long while
 537 2011-01-25 03:29:49 <molecular> I know
 538 2011-01-25 03:30:03 <molecular> it's been running for a while, finding blocks and reporting hashrate
 539 2011-01-25 03:30:11 <molecular> then I hit ctrl-c and it froze
 540 2011-01-25 03:30:19 <Diablo-D3> yeah it'll take awhile to come back
 541 2011-01-25 03:30:21 <molecular> ping worked, no ssh, htough
 542 2011-01-25 03:30:30 <molecular> ok, maybe I wasn't patient enough
 543 2011-01-25 03:30:30 <Diablo-D3> it WILL come back
 544 2011-01-25 03:30:42 <Diablo-D3> I once waited an hour for mine to come back after I had kernel time runaway
 545 2011-01-25 03:31:21 <molecular> the hashmeter gets screwed badly when a block is found (show too high)
 546 2011-01-25 03:31:50 <tcatm> solution: don't find blocks :P
 547 2011-01-25 03:31:58 <luke-jr> LOL
 548 2011-01-25 03:32:12 <Diablo-D3> thats not true at all
 549 2011-01-25 03:32:17 <Diablo-D3> its largely variable anyhow
 550 2011-01-25 03:32:25 <tcatm> variable?
 551 2011-01-25 03:32:29 <Diablo-D3> it goes up and down
 552 2011-01-25 03:32:38 <tcatm> hashrate is pretty much constant
 553 2011-01-25 03:32:51 <Diablo-D3> not when you have other shit running on the machine
 554 2011-01-25 03:33:11 <luke-jr> ok, I bid on a 5850
 555 2011-01-25 03:34:17 <tcatm> Diablo-D3: can you paste your working kernel.cl?
 556 2011-01-25 03:34:28 <molecular> should it be "const uint tnonce = (base + myid) << 10;" or "const uint tnonce = base + myid << 8;" ?
 557 2011-01-25 03:34:37 <molecular> s/8/10
 558 2011-01-25 03:34:45 <Diablo-D3> first one
 559 2011-01-25 03:34:50 <tcatm> molecular: when you don't do 1024 nonces at one, yep
 560 2011-01-25 03:36:02 DoomDumas has left ()
 561 2011-01-25 03:36:31 <molecular> btw: Diablo, you could change "debug("Block found, but rejected by Bitcoin, on " + deviceName);" to an info. might save some people trouble.
 562 2011-01-25 03:37:13 <Diablo-D3> molecular: nope, because its meaningless.
 563 2011-01-25 03:37:19 <molecular> is it?
 564 2011-01-25 03:37:33 <Diablo-D3> if you're on a real bitcoind, it just means the chain changed while you wernt looking
 565 2011-01-25 03:38:27 <molecular> hm, I see. actual erroneous calculations will be caught by "( H == 0 )" farther out
 566 2011-01-25 03:38:42 <Diablo-D3> yeah, it'll emit an error if the calc is actually wrong
 567 2011-01-25 03:39:08 <molecular> ok, I retreat my suggestion then ;)
 568 2011-01-25 03:39:32 <molecular> It's late here, I'll be going to bed...
 569 2011-01-25 03:40:07 <Diablo-D3> also, hrrrrm
 570 2011-01-25 03:40:13 <Diablo-D3> Im getting 2996 khash/sec
 571 2011-01-25 03:40:22 <molecular> how big the loop?
 572 2011-01-25 03:40:26 <Diablo-D3> 1024
 573 2011-01-25 03:40:35 <Diablo-D3> and I changed my mhash meter to * 1024.
 574 2011-01-25 03:40:38 <molecular> if you multiply that, that would be a lot
 575 2011-01-25 03:40:41 <molecular> ah, ok
 576 2011-01-25 03:41:02 <molecular> doesn't look so good. are we overlooking something?
 577 2011-01-25 03:41:42 <molecular> I got 300*256 = 76800 before, but the 300 are extremely rough, the hash-meter jittered like hell
 578 2011-01-25 03:41:56 <Diablo-D3> because kernel execution is too long
 579 2011-01-25 03:42:09 <Diablo-D3> it wasnt meant to handle extremely long kernel executions
 580 2011-01-25 03:42:11 <molecular> yeah
 581 2011-01-25 03:42:15 <Diablo-D3> so the minimum size is too big
 582 2011-01-25 03:52:16 <tcatm> Diablo-D3: just ported art's changes to m0's miner: 561 Mhash/s on a stock 5970
 583 2011-01-25 03:52:28 <Diablo-D3> heh nice
 584 2011-01-25 03:52:35 <Diablo-D3> I wonder if this just shits on loops
 585 2011-01-25 03:53:23 <tcatm> i.e. 350 Mhash/s on a 5870 @ 900 MHz
 586 2011-01-25 03:53:54 * Diablo-D3 takes the loop out
 587 2011-01-25 03:54:06 <tcatm> let's try without vectors
 588 2011-01-25 03:54:22 <tcatm> 336 Mhash/s
 589 2011-01-25 03:54:26 <tcatm> (5870 @ 900)
 590 2011-01-25 03:54:29 <Diablo-D3> okay whats
 591 2011-01-25 03:54:36 <Diablo-D3> 189052229 / 1024
 592 2011-01-25 03:54:53 <tcatm> 185 Mhash/s?
 593 2011-01-25 03:54:58 <Diablo-D3> I DONT THINK SO TIM
 594 2011-01-25 03:55:10 <Diablo-D3> thats over double the fucking hash rate
 595 2011-01-25 03:55:30 <tcatm> divide by 2?
 596 2011-01-25 03:56:00 <hacim> that seems slightly better than m0n's miner
 597 2011-01-25 03:56:23 <hacim> oh, wait, tcatm you ported art's changes to m0's?
 598 2011-01-25 03:56:29 <tcatm> hacim: yep
 599 2011-01-25 03:56:39 <hacim> tcatm: nice! got a link to the changes?
 600 2011-01-25 03:56:44 <Diablo-D3> something isnt fucking right here
 601 2011-01-25 03:56:57 <tcatm> hacim: some pastebin somewhere, don't have the link anymore
 602 2011-01-25 03:57:06 <hacim> tcatm: no, I mean to your modified m0
 603 2011-01-25 03:57:09 <hacim> :)
 604 2011-01-25 03:57:21 <hacim> tcatm: what were you getting before with your 5870 @ 900 Mhz?
 605 2011-01-25 03:57:21 <tcatm> I'm not done.
 606 2011-01-25 03:57:29 <hacim> ah cool
 607 2011-01-25 03:57:31 <tcatm> ~345
 608 2011-01-25 03:57:50 <hacim> yeah, thats what I'm getting... more around 325
 609 2011-01-25 03:58:02 <tcatm> using vectors (-v flag)?
 610 2011-01-25 03:58:10 <hacim> yeah, -f 5 -d 0 -w 256 -v
 611 2011-01-25 03:58:22 <tcatm> try -f 1 -v -w 128
 612 2011-01-25 03:58:24 <Diablo-D3> I dont use vectors
 613 2011-01-25 03:58:32 <hacim> 1300Mhz memory
 614 2011-01-25 03:58:33 <Diablo-D3> and btw, 64 seems most optimum on such hardware
 615 2011-01-25 03:58:56 <Diablo-D3> something is seriously fucked
 616 2011-01-25 03:58:57 <tcatm> Diablo-D3: depends on the kernel
 617 2011-01-25 03:59:05 <Diablo-D3> its saying
 618 2011-01-25 03:59:07 * tcatm tries with -w 64
 619 2011-01-25 03:59:19 <tcatm> 335
 620 2011-01-25 03:59:37 <tcatm> so -w 128 is optimum
 621 2011-01-25 03:59:44 <Diablo-D3> 513852
 622 2011-01-25 03:59:52 <hacim> with -f 1 -v -w 128 i'm getting 333
 623 2011-01-25 04:00:30 <hacim> 321 with -w 64
 624 2011-01-25 04:01:19 <hacim> does seem like -w 128 I get more
 625 2011-01-25 04:01:39 <hacim> but I can't quite make 345
 626 2011-01-25 04:01:55 <tcatm> I'm using a hacked SDK 2.1
 627 2011-01-25 04:02:04 <tcatm> my compiler is more efficient
 628 2011-01-25 04:02:04 <Diablo-D3> okay this is wrong
 629 2011-01-25 04:02:07 <Diablo-D3> and Ive broken something
 630 2011-01-25 04:02:10 <hacim> tcatm: did you tweak your memory mhz?
 631 2011-01-25 04:02:16 <tcatm> hacim: no
 632 2011-01-25 04:02:22 <hacim> tcatm: I have mine at 1300
 633 2011-01-25 04:02:24 <tcatm> the code doesn't even touch the memory
 634 2011-01-25 04:02:30 <Diablo-D3> 509183/1024 is 496
 635 2011-01-25 04:02:38 <Diablo-D3> 496 khash.
 636 2011-01-25 04:02:43 <hacim> tcatm: interesting, what does your --od-getclocks show for your memory mhz?
 637 2011-01-25 04:02:57 <tcatm> 1200
 638 2011-01-25 04:03:06 <hacim> i wonder if I scaled mine back to 1200 I might get more
 639 2011-01-25 04:03:11 <hacim> seems unlikely
 640 2011-01-25 04:03:21 <tcatm> as I said. no memory access in kernel
 641 2011-01-25 04:04:01 <hacim> i wonder why you are able to get 10k more than I am
 642 2011-01-25 04:04:15 <tcatm> hacked SDK ;)
 643 2011-01-25 04:04:18 <hacim> what
 644 2011-01-25 04:04:21 <hacim> are you using 2.1?
 645 2011-01-25 04:04:30 <tcatm> 2.1 with compiler from 2.2
 646 2011-01-25 04:04:36 <hacim> whoa nelly
 647 2011-01-25 04:04:47 <Diablo-D3> okay now its saying 259 mhash
 648 2011-01-25 04:04:47 <hacim> i'm using 2.3
 649 2011-01-25 04:04:57 <tcatm> so performance from 2.2 but works with multigpu
 650 2011-01-25 04:05:27 <hacim> you mean performance from 2.1, but works with multigpu?
 651 2011-01-25 04:06:17 <tcatm> compiler from 2.2, libopencl from 2.1
 652 2011-01-25 04:06:29 <hacim> (my understanding is that stream sdk 2.1 is much more performant)
 653 2011-01-25 04:06:49 <tcatm> no, 2.2 is faster
 654 2011-01-25 04:08:04 * Diablo-D3 takes the * 1024 out
 655 2011-01-25 04:08:13 <Diablo-D3> something is fucked here
 656 2011-01-25 04:08:22 <tcatm> you can't code :P
 657 2011-01-25 04:08:24 <Diablo-D3> its saying Im doing absurd levels of mhash
 658 2011-01-25 04:10:04 <Diablo-D3> I think I broke my local work size management code
 659 2011-01-25 04:10:18 * Diablo-D3 switches to old kernel to make sure
 660 2011-01-25 04:10:43 <tcatm> I wonder if this works... :)
 661 2011-01-25 04:10:50 ducki2p has quit (Ping timeout: 240 seconds)
 662 2011-01-25 04:11:01 <Diablo-D3> yeah it "works" its listing 55 mhash/sec
 663 2011-01-25 04:11:09 <Diablo-D3> wait there we go, back up to 74
 664 2011-01-25 04:11:34 <tcatm> what should you get on what card?
 665 2011-01-25 04:11:42 <Diablo-D3> about that.
 666 2011-01-25 04:12:00 <Diablo-D3> the highest consistent score Ive seen is 74.6
 667 2011-01-25 04:13:36 <Diablo-D3> http://pastebin.com/WFh5VHGD
 668 2011-01-25 04:13:40 <Diablo-D3> thats what mine is now
 669 2011-01-25 04:16:38 <hacim> tcatm: are you just setting your LD_LIBRARY_PATH to the 2.1 lib dir for the openCL?
 670 2011-01-25 04:17:03 <tcatm> hacim: mixed files from both SDKs
 671 2011-01-25 04:17:20 <tcatm> LogicError: clSetKernelArg failed: invalid arg size
 672 2011-01-25 04:17:22 <tcatm> ??? :)
 673 2011-01-25 04:17:45 <Diablo-D3> tcatm: heh
 674 2011-01-25 04:18:13 ducki2p has joined
 675 2011-01-25 04:18:46 <Diablo-D3> tcatm: wheres that happening?
 676 2011-01-25 04:19:03 <hacim> hm, i dont notice any appreciable difference using 2.2
 677 2011-01-25 04:19:47 <tcatm> Diablo-D3: when settings args :P
 678 2011-01-25 04:19:54 <Diablo-D3> tcatm: yeah but check the headers
 679 2011-01-25 04:19:55 <Diablo-D3> identical
 680 2011-01-25 04:20:00 <tcatm> yep
 681 2011-01-25 04:20:10 <Diablo-D3> so obviously thats wrong
 682 2011-01-25 04:20:30 <hacim> whoa, 924 khash/s is way worse
 683 2011-01-25 04:20:43 <tcatm> three cards?
 684 2011-01-25 04:21:02 <Diablo-D3> tcatm: not only that, its not saying that on mine
 685 2011-01-25 04:21:15 <Diablo-D3> oh durrrrr
 686 2011-01-25 04:21:26 <hacim> wow, using the 2.2 libs really hosed the hash rate
 687 2011-01-25 04:21:33 <Diablo-D3>       output[0] = it ^ tnonce;
 688 2011-01-25 04:21:35 <Diablo-D3> hehhh
 689 2011-01-25 04:21:43 <Diablo-D3> hacim: yes it has if you dont use the hack
 690 2011-01-25 04:21:46 <Diablo-D3> its consuming all your cpu time
 691 2011-01-25 04:22:36 <hacim> whats this 'hack'?
 692 2011-01-25 04:22:46 <Diablo-D3> what miner are you on?
 693 2011-01-25 04:23:20 <hacim> Diablo-D3: i've been trying m0
 694 2011-01-25 04:23:31 <Diablo-D3> his doesnt use it
 695 2011-01-25 04:23:45 <Diablo-D3> tcatm: now that I fixed the obvious oversight
 696 2011-01-25 04:24:01 <Diablo-D3> I get 74.8 on his without loop, 74.6 on mine without loop
 697 2011-01-25 04:24:14 <Diablo-D3> actually closer to 74.9
 698 2011-01-25 04:24:33 <Diablo-D3> so about half a percent
 699 2011-01-25 04:24:34 <Diablo-D3> lol
 700 2011-01-25 04:27:34 <hacim> tcatm: alright, I give, how did you mix those two?
 701 2011-01-25 04:28:09 <tcatm> hacim: don't remember what files I copied
 702 2011-01-25 04:28:11 <Diablo-D3> lol okay
 703 2011-01-25 04:28:14 <Diablo-D3> now that shits working
 704 2011-01-25 04:28:17 <Diablo-D3> with the loop on
 705 2011-01-25 04:28:21 <Diablo-D3> Im getting about
 706 2011-01-25 04:28:32 <Diablo-D3> 48 mhash
 707 2011-01-25 04:28:32 <Diablo-D3> lol
 708 2011-01-25 04:28:36 <hacim> hehe
 709 2011-01-25 04:28:46 <tcatm> that's not bad
 710 2011-01-25 04:28:53 <tcatm> meanwhile I'm at 352 @ 900
 711 2011-01-25 04:29:03 <Diablo-D3> 48? down from 75?
 712 2011-01-25 04:29:20 <hacim> damn
 713 2011-01-25 04:29:36 <hacim> you've gone up 10k more, so now are 20k ahead of me
 714 2011-01-25 04:29:53 <mrb_> picture of the future dual-GPU HD 6990: http://www.fudzilla.com/graphics/item/21635-amds-antilles-card-pixellized
 715 2011-01-25 04:30:04 <Diablo-D3> mrb_: OH GOD I CAME
 716 2011-01-25 04:30:06 <Diablo-D3> someone get me a towel
 717 2011-01-25 04:30:48 <mrb_> two 8-pin connectors. wonder if officially it means it will be a 375W card... and what are the repercussions for AMD breaking the PCIe specs wrt. max wattage
 718 2011-01-25 04:30:53 <mrb_> Diablo-D3: eww
 719 2011-01-25 04:31:00 <tcatm> woah, finally a good cooler design
 720 2011-01-25 04:31:10 <hacim> is that stephen colbert?
 721 2011-01-25 04:31:22 <Diablo-D3> no
 722 2011-01-25 04:31:26 <Diablo-D3> hes not that ugly
 723 2011-01-25 04:31:32 <mrb_> the card seems shorter than a 5970
 724 2011-01-25 04:32:03 <hacim> tcatm: did you find this hydra lib/compiler mix somewhere? i'm wondering how you figured this out
 725 2011-01-25 04:32:31 <tcatm> hacim: no, did it myself
 726 2011-01-25 04:32:33 <Diablo-D3> I wonder if I drop the loop size it'll get better
 727 2011-01-25 04:33:47 <hacim> damn
 728 2011-01-25 04:34:11 <Diablo-D3> lets try 16
 729 2011-01-25 04:35:25 noagendamarket has quit (Ping timeout: 240 seconds)
 730 2011-01-25 04:35:36 <Diablo-D3> 72.7
 731 2011-01-25 04:35:58 <Diablo-D3> lets try 8
 732 2011-01-25 04:38:17 <Diablo-D3> 70.5
 733 2011-01-25 04:38:21 <Diablo-D3> lets try 4
 734 2011-01-25 04:39:24 <echelon> when specifying your own nodes using addnode, does bitcoin still attempt to connect to the hardcoded nodes?
 735 2011-01-25 04:39:53 <Diablo-D3> 67
 736 2011-01-25 04:40:17 <Diablo-D3> lets try 32
 737 2011-01-25 04:41:27 <Diablo-D3> 70.6
 738 2011-01-25 04:41:44 <tcatm> echelon: at least it'll connect to other nodes
 739 2011-01-25 04:41:48 <Diablo-D3> lets try 2
 740 2011-01-25 04:41:49 <Diablo-D3> lets try 24
 741 2011-01-25 04:41:57 <echelon> huh
 742 2011-01-25 04:43:31 <Diablo-D3> 73.1
 743 2011-01-25 04:43:56 <Diablo-D3> lets try 20
 744 2011-01-25 04:44:58 <jgarzik> echelon: hardcoded nodes are just for seeding the address list.  that is separate from the connection logic.
 745 2011-01-25 04:45:31 <jgarzik> echelon: -connect specifies that you should -only- connect to specified nodes, and none other.  otherwise, -addnode, IRC seeds and hardcoded seeds all go into the address poll, from which connections are made.
 746 2011-01-25 04:45:46 <echelon> i used -addnode and -noirc
 747 2011-01-25 04:46:00 <tcatm> 566 Mhash/s on a stock 5970
 748 2011-01-25 04:46:04 <Diablo-D3> about 73.3
 749 2011-01-25 04:46:15 <jgarzik> echelon: in that case, it will attempt connections to hardcoded nodes If And Only If there are no other addresses in your database.
 750 2011-01-25 04:46:30 <echelon> ohh kk
 751 2011-01-25 04:51:12 <echelon> wait, you would use -connect instead of -addnode?
 752 2011-01-25 04:58:10 Cusipzzz has quit (Quit: KVIrc 4.0.2 Insomnia http://www.kvirc.net/)
 753 2011-01-25 05:04:06 <tcatm> Error:E010:Irreducible ControlFlow Detected
 754 2011-01-25 05:04:08 <tcatm> fsck :)
 755 2011-01-25 05:05:54 <hacim> ok, so its been 120 blocks since the block I found, and I still don't have 50btc
 756 2011-01-25 05:07:46 <hacim> am I doing something wrong?
 757 2011-01-25 05:08:51 <tcatm> Diablo-D3: btw, the loop decreases performance for me, too
 758 2011-01-25 05:11:11 <Diablo-D3> what are you on for hardware?
 759 2011-01-25 05:11:24 <tcatm> 5870/5970
 760 2011-01-25 05:11:58 <Diablo-D3> art says it sfaster for him
 761 2011-01-25 05:12:13 <tcatm> Using my custom miner it is faster, too
 762 2011-01-25 05:12:18 <tcatm> strange
 763 2011-01-25 05:12:30 <tcatm> still 352 Mhash/s on a 5870 isn't bad :)
 764 2011-01-25 05:13:24 <Diablo-D3> lets switch to 2.3
 765 2011-01-25 05:13:30 <Diablo-D3> 67 without loop
 766 2011-01-25 05:15:08 <Diablo-D3> actually. 68.5
 767 2011-01-25 05:16:09 <Diablo-D3> lets tru 20 loops
 768 2011-01-25 05:16:50 <hacim> ah damn, i miscalculated 120
 769 2011-01-25 05:17:17 <Diablo-D3> 68.4 with 20
 770 2011-01-25 05:17:23 <Diablo-D3> sorry, 68.5
 771 2011-01-25 05:17:25 <tcatm> total gain: 70 Mhash/s on my cluster
 772 2011-01-25 05:17:41 <Diablo-D3> lets try 1024
 773 2011-01-25 05:19:24 <Diablo-D3> tt
 774 2011-01-25 05:19:27 <Diablo-D3> 52
 775 2011-01-25 05:21:34 <Diablo-D3> ooh its reached 51
 776 2011-01-25 05:21:35 <Diablo-D3> er 61
 777 2011-01-25 05:22:07 <Diablo-D3> lets try with sync objects
 778 2011-01-25 05:22:50 Kiba has quit (Ping timeout: 240 seconds)
 779 2011-01-25 05:23:54 <Diablo-D3> about the same
 780 2011-01-25 05:23:54 <Diablo-D3> heh
 781 2011-01-25 05:24:01 <Diablo-D3> so loop just isnt worth it for me
 782 2011-01-25 05:24:44 kron has joined
 783 2011-01-25 05:24:47 <kron> NOOOOOOOOO
 784 2011-01-25 05:24:55 <kron> I LOST ALL MY BITCOINS GAMBLING
 785 2011-01-25 05:24:58 <kron> FML
 786 2011-01-25 05:29:14 <echelon> lol
 787 2011-01-25 05:29:25 <echelon> how much did you have
 788 2011-01-25 05:30:30 <tcatm> hacim: http://pastebin.com/g8QJ1R3r
 789 2011-01-25 05:37:42 <hacim> tcatm: hm, i get various hunk failures
 790 2011-01-25 05:37:49 <hacim> i wonder if I am using a different version than you are
 791 2011-01-25 05:39:00 <tcatm> probably
 792 2011-01-25 05:39:22 <tcatm> patch it manually :)
 793 2011-01-25 05:39:36 noagendamarket has joined
 794 2011-01-25 05:39:50 noagendamarket has quit (Changing host)
 795 2011-01-25 05:39:50 noagendamarket has joined
 796 2011-01-25 05:40:14 <hacim> i'd prefer to make sure I have the latest
 797 2011-01-25 05:42:41 <hacim> seems I have the latest git HEAD
 798 2011-01-25 05:42:58 <tcatm> maybe i didn't have the latest
 799 2011-01-25 05:46:29 * hacim monkey patches
 800 2011-01-25 05:50:39 <kron> i had 300
 801 2011-01-25 05:52:39 <kron> if anyone wants to send me some cause im sad =[
 802 2011-01-25 05:52:39 <kron>  1NYTPaUfig9wuXd3ToTDrUsn7Yh9wREYqR
 803 2011-01-25 05:52:42 <mrb_> kron: want to recover them? pick a number between 1 and 10.
 804 2011-01-25 05:52:57 <kron> 7!
 805 2011-01-25 05:53:05 <mrb_> incorrect. I was thinking of 5.
 806 2011-01-25 05:53:09 <mrb_> you owe me 300 BTC.
 807 2011-01-25 05:53:16 <kron> -.-
 808 2011-01-25 05:53:17 <mrb_> :)
 809 2011-01-25 05:53:33 <mrb_> where were you gambling?
 810 2011-01-25 05:53:39 <nanotube> kron: well... let that be a lesson to you for the future. :P
 811 2011-01-25 05:53:42 <kron> the double or trouble
 812 2011-01-25 05:53:43 <kron> lol
 813 2011-01-25 05:54:04 <kron> they must be raking in the bitcoins
 814 2011-01-25 05:54:19 <nanotube> as is any gambling establishment. the house edge.
 815 2011-01-25 05:55:06 <kron> right
 816 2011-01-25 05:55:10 <kron> shoulda quite while i was ahead
 817 2011-01-25 05:55:13 <kron> of course
 818 2011-01-25 05:55:32 * nanotube is starting to think that maybe he should run a gambling site. hehe
 819 2011-01-25 05:55:50 <kron> i was thinking the same thing lol
 820 2011-01-25 05:55:51 <OneFixt> Is there any problem with getting rid of the rounding in AmountFromValue? Will all nodes accept transactions of .01xxxxxx?
 821 2011-01-25 05:56:00 <nanotube> OneFixt: yes
 822 2011-01-25 05:56:03 <nanotube> they will
 823 2011-01-25 05:56:42 <OneFixt> nanotube: thanks
 824 2011-01-25 05:56:50 noagendamarket has quit (Ping timeout: 240 seconds)
 825 2011-01-25 05:57:51 <kron> ;_;
 826 2011-01-25 05:58:07 <mrb_> "Facebook confirms plans to make Credits the mandatory "universal currency - http://venturebeat.com/2011/01/24/facebook-credits-mandatory/
 827 2011-01-25 05:58:26 <kron> if facebook adopted bitcoins
 828 2011-01-25 05:58:28 <kron> it would be over
 829 2011-01-25 05:58:29 <kron> lol
 830 2011-01-25 05:58:55 * mrb_ wonders if there is some business opportunity related to BTC <-> FB credits
 831 2011-01-25 05:59:50 <tcatm> ölr
 832 2011-01-25 05:59:54 <tcatm> dd
 833 2011-01-25 06:00:04 <lfm> can you convert fb credits back to local money?
 834 2011-01-25 06:00:37 <nanotube> mrb_: wow, 30% fee on buying fb credits. that's /bold/
 835 2011-01-25 06:02:14 doublec has quit (Quit: Leaving)
 836 2011-01-25 06:03:35 molecular has quit (Ping timeout: 276 seconds)
 837 2011-01-25 06:03:43 <hacim> tcatm: hm, i'm getting some crazy compilation errors
 838 2011-01-25 06:04:01 <tcatm> hacim: you probably need to fix it for plain sdk 2.1
 839 2011-01-25 06:04:05 molecular has joined
 840 2011-01-25 06:04:49 <hacim> tcatm: http://pastebin.com/xSHHirKQ
 841 2011-01-25 06:05:25 <tcatm> that's strange
 842 2011-01-25 06:05:58 <tcatm> try this kernel http://pastebin.com/RF5R9d10
 843 2011-01-25 06:06:12 <kron> quick question, how long will it take me to generate coins with a quad core AMD and a nvidia 8800gtx
 844 2011-01-25 06:06:22 <kron> should i join a pool?
 845 2011-01-25 06:06:27 <hacim> oddly it works with 2.2, but not 2.1 or 2.3
 846 2011-01-25 06:06:38 <tcatm> oh wait
 847 2011-01-25 06:06:47 <tcatm> change line 20 to #define rotr(x, y) amd_bitalign((u)x, (u(x, (u)y)
 848 2011-01-25 06:06:49 <tcatm> er...
 849 2011-01-25 06:07:01 <tcatm> change line 20 to #define rotr(x, y) amd_bitalign((u)x, (u)x, (u)y)
 850 2011-01-25 06:08:18 dduane has quit (Ping timeout: 255 seconds)
 851 2011-01-25 06:09:06 <hacim> hm, that doesn't seem to help
 852 2011-01-25 06:09:32 <hacim> oh wait, you said try that kernel
 853 2011-01-25 06:09:40 <tcatm> yes and change that line
 854 2011-01-25 06:11:34 <Keefe> ;;bc,calc 30000
 855 2011-01-25 06:11:35 <gribble> The average time to generate a block at 30000 Khps, given current difficulty of 18437.64439217 , is 4 weeks, 2 days, 13 hours, 13 minutes, and 55 seconds
 856 2011-01-25 06:11:40 <Keefe> kron: ^
 857 2011-01-25 06:12:06 <nanotube> kron: in other words... yes you should. :)
 858 2011-01-25 06:12:50 <Keefe> a pool will give you about 50 btc over that time, but you'll get a little each day instead of all at once
 859 2011-01-25 06:13:08 <nanotube> ;;bc,stats
 860 2011-01-25 06:13:09 <gribble> Current Blocks: 104460 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 371 blocks | Next Difficulty In About: 2 days, 3 hours, 31 minutes, and 40 seconds | Next Difficulty Estimate: 21827.79160716
 861 2011-01-25 06:13:10 <hacim> tcatm: interesting that seems to work, although it blurts this out before it starts hashing: http://pastebin.com/jxR5dbgh
 862 2011-01-25 06:13:28 <tcatm> that's okay...
 863 2011-01-25 06:13:43 <tcatm> how much did you get before that patch?
 864 2011-01-25 06:13:49 <kron> nice i have to get that set up
 865 2011-01-25 06:14:50 jwalck has quit (Ping timeout: 246 seconds)
 866 2011-01-25 06:14:53 <hacim> tcatm: 332663, now I'm getting 336162
 867 2011-01-25 06:15:03 <hacim> not the increase you were seeing
 868 2011-01-25 06:15:20 <tcatm> at least it's an improvement
 869 2011-01-25 06:15:25 rlifchitz has quit (Read error: Operation timed out)
 870 2011-01-25 06:16:01 rlifchitz has joined
 871 2011-01-25 06:16:01 rlifchitz has quit (Changing host)
 872 2011-01-25 06:16:01 rlifchitz has joined
 873 2011-01-25 06:16:07 <hacim> why is it blurting that out before it hashes?
 874 2011-01-25 06:16:42 <nanotube> it's just a warning, shouldn't affect anything.
 875 2011-01-25 06:16:48 <tcatm> I didn't care to do some calculations "the correct way"
 876 2011-01-25 06:17:26 <hacim> you dont worry that doing them the less than correct way could impact potential results?
 877 2011-01-25 06:17:52 larsivi has quit (Remote host closed the connection)
 878 2011-01-25 06:18:05 <hacim> probably not an impact to speed, but potentially to accuracy or validity?
 879 2011-01-25 06:18:23 <tcatm> I verified it and it's working correct.
 880 2011-01-25 06:25:23 <hacim> tcatm: in case you want to update to the latest git, this is the diff: http://pastebin.com/v98gWWLc
 881 2011-01-25 06:26:31 <tcatm> thanks
 882 2011-01-25 06:27:24 <hacim> now if I could only ramp up to 350
 883 2011-01-25 06:31:43 skeledrew1 has quit (Quit: Instantbird 0.3a1pre)
 884 2011-01-25 06:32:48 Slix` has quit (Read error: Connection reset by peer)
 885 2011-01-25 06:33:13 testtest has quit (Quit: Page closed)
 886 2011-01-25 06:34:30 skeledrew has joined
 887 2011-01-25 06:37:57 molecular has quit (Read error: Connection reset by peer)
 888 2011-01-25 06:40:19 molecular has joined
 889 2011-01-25 06:48:08 dduane has joined
 890 2011-01-25 06:51:10 bk128 has joined
 891 2011-01-25 07:03:12 nevezen has quit (Quit: leaving)
 892 2011-01-25 07:04:17 <Diablo-D3> hmm
 893 2011-01-25 07:04:24 <Diablo-D3> wonder if I should push new version of my miner
 894 2011-01-25 07:12:04 <joe_1> YESSSSSSSSSS diablo released a new miner if i can get it first i can take over the world
 895 2011-01-25 07:12:32 mrb_ has quit (Ping timeout: 265 seconds)
 896 2011-01-25 07:13:32 <Diablo-D3> ?
 897 2011-01-25 07:13:33 <tcatm> too late, already released a patch for m0's
 898 2011-01-25 07:13:36 larsivi has joined
 899 2011-01-25 07:13:45 <Diablo-D3> tcatm: yeah but its not as good as mine
 900 2011-01-25 07:14:11 <tcatm> What does yours do better?
 901 2011-01-25 07:14:26 <Diablo-D3> I think its a little late to be asking that.
 902 2011-01-25 07:14:27 <tcatm> (hint: with testing only H==0 in kernel I don't see any improvement)
 903 2011-01-25 07:15:47 <tcatm> Can you make a binary that doesn't define rotate using bitalign?
 904 2011-01-25 07:16:02 <tcatm> Then I could run your miner and see how fast it is.
 905 2011-01-25 07:16:12 <Diablo-D3> erm, mine doesnt if you dont have it
 906 2011-01-25 07:16:32 <tcatm> Problem is: I have bitalign, but I don't have amd_bitalign()
 907 2011-01-25 07:16:49 <tcatm> i.e. my compiler will translate rotate() into bitalign() on its own
 908 2011-01-25 07:17:01 <Diablo-D3> see above.
 909 2011-01-25 07:17:13 <tcatm> ?
 910 2011-01-25 07:17:34 <Diablo-D3> my miner already works on targets that dont have bitalign
 911 2011-01-25 07:17:41 <Diablo-D3> otherwise I could never run it myself
 912 2011-01-25 07:18:13 <tcatm> it says #if BITALIGN... right on top of your kernel!
 913 2011-01-25 07:18:32 <Diablo-D3> yes, and look very closely at that code!
 914 2011-01-25 07:18:52 <tcatm> it defines rot() using bitalign when that flag is set?
 915 2011-01-25 07:19:01 <Diablo-D3> yes
 916 2011-01-25 07:19:09 <tcatm> whthat's wrong
 917 2011-01-25 07:19:37 <tcatm> or... your java code shouldn't define that flag when running on my miner
 918 2011-01-25 07:19:51 <Diablo-D3> how so? if you dont have amd_media_ops, then my miner never defines BITALIGN
 919 2011-01-25 07:20:11 <Diablo-D3> so it does rotate()
 920 2011-01-25 07:21:02 <tcatm> is it possible to change your kernel when using your binaries?
 921 2011-01-25 07:21:13 <Diablo-D3> yes, just edit it
 922 2011-01-25 07:21:31 <Diablo-D3> but I dont get why you're having a problem
 923 2011-01-25 07:21:56 <Diablo-D3> if you dont have cl_amd_media_ops, then it never uses amd_bitalign
 924 2011-01-25 07:22:04 * jgarzik needs to update cpuminer with that 4way optimization
 925 2011-01-25 07:22:16 TheAncientGoat has joined
 926 2011-01-25 07:22:28 * tcatm needs to update 4way with art's opencl optimization
 927 2011-01-25 07:22:35 <jgarzik> http://www.bitcoin.org/smf/index.php?topic=820.msg40022#msg40022
 928 2011-01-25 07:23:15 <Diablo-D3> heh
 929 2011-01-25 07:23:31 <tcatm> I should probably rewrite 4way. I can think of many improvements after writing a few opencl miner
 930 2011-01-25 07:23:45 ArtForzZz is now known as ArtForz
 931 2011-01-25 07:24:08 <Pegasus-Rider> BTW, 4-way is for 64-bit machines only, right?
 932 2011-01-25 07:24:17 <tcatm> yep
 933 2011-01-25 07:24:48 <jgarzik> well
 934 2011-01-25 07:24:58 <Diablo-D3> ArtForz: I now have your kernel in my miner
 935 2011-01-25 07:25:38 <ArtForz> neat
 936 2011-01-25 07:26:23 <tcatm> I ported them to m0's. 567 MHash/s on stock 5970
 937 2011-01-25 07:26:28 <Diablo-D3> I get about a half of a percent increase on my 4850
 938 2011-01-25 07:26:35 <jgarzik> 4way works just fine on 32-bit
 939 2011-01-25 07:26:56 <ArtForz> well, I never bothered testing how it performs on 4xxx
 940 2011-01-25 07:27:07 <ArtForz> 4xxx is hopeless anyways
 941 2011-01-25 07:27:28 <lfm> but its ionly faster on machines that also support 64 bit even if you are only running in 32 bit mode
 942 2011-01-25 07:28:19 <tcatm> Diablo-D3: how do I tell your miner which GPU to use?
 943 2011-01-25 07:28:31 <jgarzik> true
 944 2011-01-25 07:28:36 <ArtForz> 567, not too shabby
 945 2011-01-25 07:28:37 <lfm> -d
 946 2011-01-25 07:29:10 <Diablo-D3> tcatm: cant yet.
 947 2011-01-25 07:29:22 <Diablo-D3> it uses all
 948 2011-01-25 07:30:09 <tcatm> that's stupid
 949 2011-01-25 07:30:17 <Diablo-D3> not really
 950 2011-01-25 07:30:26 <Diablo-D3> most people would never want to run a miner just on one
 951 2011-01-25 07:31:33 <tcatm> your miner does 308 Mhash/s on a 5870 @ 900
 952 2011-01-25 07:32:10 <ArtForz> thats... bad
 953 2011-01-25 07:32:14 <tcatm> Mkay, it just jumped to 332
 954 2011-01-25 07:32:28 <CyanDynamo> http://arstechnica.com/web/news/2011/01/philanthroper-hopes-to-boost-charity-of-the-day-1-at-a-time.ars
 955 2011-01-25 07:32:28 <ArtForz> should be closer to 350
 956 2011-01-25 07:32:36 <tcatm> never seen such a jumpy hashmeter on my GPUs before
 957 2011-01-25 07:33:05 <Diablo-D3> its only a 30 second average
 958 2011-01-25 07:33:38 <tcatm> ArtForz: m0's with your kernel does ~352
 959 2011-01-25 07:33:52 <Diablo-D3> tcatm: also, give it time
 960 2011-01-25 07:33:53 <tcatm> Diablo-D3: make a binary of your new kernel
 961 2011-01-25 07:33:58 <Diablo-D3> you can wait.
 962 2011-01-25 07:34:08 * nathan7 hops around
 963 2011-01-25 07:43:14 * Diablo-D3 is further optimizing it
 964 2011-01-25 07:44:05 <ArtForz> how the F do you optimize for 5xxx on a 4xxx ?
 965 2011-01-25 07:44:27 <Diablo-D3> ArtForz: not specifically for 5xxx, just in general
 966 2011-01-25 07:44:48 <ArtForz> good luck with that
 967 2011-01-25 07:53:15 <Diablo-D3> its now running about 76.2
 968 2011-01-25 07:58:12 mrb_ has joined
 969 2011-01-25 08:04:04 <Diablo-D3> it now does about 76.6
 970 2011-01-25 08:07:26 <tcatm> You should really get a 5870 or 5970
 971 2011-01-25 08:15:04 <Diablo-D3> I dont think its getting any faster than this for the time being
 972 2011-01-25 08:20:01 <Diablo-D3> so lets say
 973 2011-01-25 08:20:32 <Diablo-D3> a 3% increase in speed on my 4850
 974 2011-01-25 08:20:42 <Diablo-D3> that seems about right
 975 2011-01-25 08:22:22 <Diablo-D3> ArtForz: isnt that what you got on your shit vs m0's?
 976 2011-01-25 08:23:06 <ArtForz> somewhere around there
 977 2011-01-25 08:23:09 bk128 has quit (Quit: bk128)
 978 2011-01-25 08:23:23 <Diablo-D3> heh, interesting, your kernel did very well on 4xxx then
 979 2011-01-25 08:24:22 <ArtForz> well, converting W from array to variables seems to help the compiler quite a bit
 980 2011-01-25 08:26:14 <CIA-98> DiabloMiner: Patrick McFarland master * rdc5b12e / (3 files in 3 dirs): Switch to ArtForz's kernel, 3% speed increase - http://bit.ly/exOmWo
 981 2011-01-25 08:28:19 <Diablo-D3> tcatm: there
 982 2011-01-25 08:28:22 <Diablo-D3> new binary uploaded
 983 2011-01-25 08:32:10 Daviey has quit (Excess Flood)
 984 2011-01-25 08:42:16 altamic has joined
 985 2011-01-25 08:43:42 <tcatm> Diablo-D3: did only the kernel change?
 986 2011-01-25 08:44:33 <sipa> seems a bit faster on my 4870 too
 987 2011-01-25 08:45:34 <Diablo-D3> tcatm: and the stuff to support it, yes
 988 2011-01-25 08:45:42 <sipa> aha, you enable the looping trick only on 5xxx
 989 2011-01-25 08:45:53 <Diablo-D3> sipa: it wont work on 4xxx
 990 2011-01-25 08:46:01 <sipa> ok
 991 2011-01-25 08:46:11 <Diablo-D3> for small loops, it doesnt win, for large loops, it shits itself
 992 2011-01-25 08:46:14 Daviey has joined
 993 2011-01-25 08:46:25 <Diablo-D3> and I suspect nvidia cant win at all here
 994 2011-01-25 08:50:25 <tcatm> Diablo-D3: 342 Mhash/s on 5870 @ 900MHz
 995 2011-01-25 08:50:39 altamic has quit (Ping timeout: 272 seconds)
 996 2011-01-25 08:51:58 <Diablo-D3> what was the target?
 997 2011-01-25 08:52:09 <tcatm> diff 1
 998 2011-01-25 08:52:13 <Diablo-D3> no I meant
 999 2011-01-25 08:52:14 <tcatm> using the pool
1000 2011-01-25 08:52:17 <Diablo-D3> what was I supposed to hit
1001 2011-01-25 08:52:23 <tcatm> don't know?
1002 2011-01-25 08:52:38 <Diablo-D3> well, the wiki ways 314 mhash
1003 2011-01-25 08:52:40 <sipa> over 9000!
1004 2011-01-25 08:53:03 <Diablo-D3> so thats a 9% increase
1005 2011-01-25 08:53:05 <tcatm> for 850 MHz IIRC
1006 2011-01-25 08:53:05 <Diablo-D3> and hard to believe
1007 2011-01-25 08:53:15 <Diablo-D3> ahh yes 850
1008 2011-01-25 08:53:24 <tcatm> so 323 Mhash/s @ 850 MHz
1009 2011-01-25 08:53:34 <Diablo-D3> so a 3% increase
1010 2011-01-25 08:53:43 <tcatm> still 9 Mhash/s worse than m0's
1011 2011-01-25 08:54:13 <sipa> than m0's own kernel, or than m0 with Art's kernel?
1012 2011-01-25 08:54:24 <tcatm> m0's with art's
1013 2011-01-25 08:54:38 <tcatm> though it's pretty much the same kernel as Diablo-D3 uses
1014 2011-01-25 08:54:54 <Diablo-D3> did you move the stuff out of the kernel like I did?
1015 2011-01-25 08:54:58 <tcatm> sure
1016 2011-01-25 08:55:02 <Diablo-D3> heh
1017 2011-01-25 08:55:31 <Diablo-D3> you're running mine with -w 64 -f 1?
1018 2011-01-25 08:55:34 <tcatm> What's that "Attempt x found" stuff?
1019 2011-01-25 08:55:39 <Diablo-D3> H == 0.
1020 2011-01-25 08:55:41 <tcatm> -w 128 -f 1
1021 2011-01-25 08:55:53 <Diablo-D3> -w 64 should be faster
1022 2011-01-25 08:56:00 * tcatm tries
1023 2011-01-25 08:56:06 <tcatm> for m0's 128 is faster
1024 2011-01-25 08:57:16 <tcatm> no change
1025 2011-01-25 08:57:27 <Diablo-D3> heh, interesting
1026 2011-01-25 08:57:40 <tcatm> you should smooth the hashmeter
1027 2011-01-25 08:57:56 <Diablo-D3> it used to be 100% smooth
1028 2011-01-25 08:57:59 <Diablo-D3> but people bitched about it
1029 2011-01-25 08:58:09 <tcatm> it's jumping +/- 70 mhash/s
1030 2011-01-25 08:58:22 <Diablo-D3> with -f 1? Im not surprised
1031 2011-01-25 08:58:33 <tcatm> what's that -f 1 anyway?
1032 2011-01-25 08:58:36 <Diablo-D3> my miner tends to destroy the system clock due to interrupt spam
1033 2011-01-25 08:58:42 <Diablo-D3> it runs one kernel per second.
1034 2011-01-25 08:59:01 <tcatm> can I tell it to run one kernel after the other?
1035 2011-01-25 08:59:08 <Diablo-D3> huh?
1036 2011-01-25 08:59:27 <tcatm> just enqueue the next kernel after the previous finished and you read the buffer
1037 2011-01-25 08:59:41 <Diablo-D3> erm, it already does that?
1038 2011-01-25 08:59:49 <tcatm> so it should output a constant hashrate
1039 2011-01-25 08:59:51 <Diablo-D3> and its already running 3 concurrently?
1040 2011-01-25 08:59:59 <tcatm> on 3 GPUs, yep
1041 2011-01-25 09:00:05 <Diablo-D3> on _1_ GPU.
1042 2011-01-25 09:00:12 <Diablo-D3> it forces the driver queue to be stuffed.
1043 2011-01-25 09:01:24 <tcatm> strange
1044 2011-01-25 09:01:52 <tcatm> when using the hardware clock as timing reference the hashrate is smooth
1045 2011-01-25 09:02:05 <tcatm> 345
1046 2011-01-25 09:02:05 <Diablo-D3> yeah, because it fucks over time
1047 2011-01-25 09:02:28 <Diablo-D3> happens on art's shit too
1048 2011-01-25 09:03:45 <cosurgi> hiya in the morning.
1049 2011-01-25 09:03:58 <Diablo-D3> so thats more like a 4% increase
1050 2011-01-25 09:04:02 <cosurgi> Diablo-D3: it just 'git pull'ed your code. Does it work?
1051 2011-01-25 09:04:09 <cosurgi> s/it/I.
1052 2011-01-25 09:04:16 <Diablo-D3> cosurgi: one could assume
1053 2011-01-25 09:04:16 <tcatm> cosurgi: yep
1054 2011-01-25 09:04:34 <cosurgi> ok, let me give it a try :)
1055 2011-01-25 09:05:08 <Diablo-D3> tcatm: dont you have a 5970?
1056 2011-01-25 09:05:16 <tcatm> I have a few, yes.
1057 2011-01-25 09:05:23 <tcatm> I usually do testing on the 5870, though.
1058 2011-01-25 09:06:14 <Diablo-D3> well, this would get 552 on a 5970 according to your numbers
1059 2011-01-25 09:06:20 <Diablo-D3> which is almost as much as art's 556
1060 2011-01-25 09:06:21 <tcatm> let's try
1061 2011-01-25 09:09:12 <cosurgi> Diablo-D3: radeon 5830, before: 213 Mh, now: 178 Mh (but it seems to go up)
1062 2011-01-25 09:09:45 <Diablo-D3> wait for it to settle after awhile
1063 2011-01-25 09:09:53 <cosurgi> ok. 184 now.
1064 2011-01-25 09:09:55 <Diablo-D3> also, who knows how this'll act on a 5830
1065 2011-01-25 09:10:02 <Diablo-D3> thats a rather weird chip
1066 2011-01-25 09:10:06 <tcatm> ~560 on 5970
1067 2011-01-25 09:10:15 <Diablo-D3> tcatm: niiiiiiice
1068 2011-01-25 09:10:30 <sipa> maybe the looping should be disabled separately>
1069 2011-01-25 09:10:36 <cosurgi> 202 Mh ....
1070 2011-01-25 09:10:58 <tcatm> sometimes it even shows a total of 2000000 khash/s for 2 cards ;)
1071 2011-01-25 09:11:15 <Diablo-D3> yeah its probably sticking somewhere
1072 2011-01-25 09:11:31 <cosurgi> and back down to 186 Mh.
1073 2011-01-25 09:11:35 <Diablo-D3> I should throw in a check kernel work size check
1074 2011-01-25 09:11:47 <tcatm> does that cause problems?
1075 2011-01-25 09:11:54 <Diablo-D3> it'll run with 0
1076 2011-01-25 09:11:55 <cosurgi> ok. I'm rolling back to previous version, and check on dual 5850
1077 2011-01-25 09:11:57 <Diablo-D3> which its nuts
1078 2011-01-25 09:12:02 <ArtForz> I gave up on dynamically adjusting globalworksize
1079 2011-01-25 09:15:28 slush1 has joined
1080 2011-01-25 09:16:29 <cosurgi> dual 5850, before: 518 Mh, now: 452 Mh. Seems that it stopped going up.
1081 2011-01-25 09:16:44 <Diablo-D3> cosurgi: try -w 64 or -w 128
1082 2011-01-25 09:16:51 <cosurgi> I am doing -w 64
1083 2011-01-25 09:17:01 <cosurgi> whoa, now it's 397 Mh
1084 2011-01-25 09:17:04 <tcatm> Diablo-D3: your miner is nice for multi gpu use, though. much easier to track than one process/gpu
1085 2011-01-25 09:17:07 <cosurgi> I'll try -w 128
1086 2011-01-25 09:17:29 <Diablo-D3> cosurgi: what sdk is this? 2.1?
1087 2011-01-25 09:17:38 <cosurgi> yes. I didn't update
1088 2011-01-25 09:17:47 <cosurgi> should I?
1089 2011-01-25 09:17:48 <sipa> tcatm: do you enable the vectorization in m0/art on 5970?
1090 2011-01-25 09:18:04 <Diablo-D3> sipa: it doesnt make it faster.
1091 2011-01-25 09:18:33 <tcatm> sipa: yep
1092 2011-01-25 09:18:45 <tcatm> depends on kernel
1093 2011-01-25 09:19:09 <cosurgi> 07:01 < hacim> oddly it works with 2.2, but not 2.1 or 2.3
1094 2011-01-25 09:19:23 <Diablo-D3> cosurgi: the real problem is, your systems could be fucked
1095 2011-01-25 09:19:25 <tcatm> It's all about using as little registers as possible while doing as much as work in a thread as possible
1096 2011-01-25 09:19:29 <Diablo-D3> cosurgi: you're lower than you should be
1097 2011-01-25 09:19:30 <cosurgi> hacim: the new Diablo + ArtForz kernel works best with 2.2 SDK ?
1098 2011-01-25 09:19:52 <sipa> tcatm: could you give a number?
1099 2011-01-25 09:20:04 <Diablo-D3> cosurgi: dual 5850 should be about 482 or more
1100 2011-01-25 09:20:21 <cosurgi> 441 Mh now, -w 128. With your previous miner it was 518 Mh.\
1101 2011-01-25 09:20:33 <Diablo-D3> with what -f?
1102 2011-01-25 09:20:39 <tcatm> sipa: 352 with vectors, 335 without (5870 @ 900)
1103 2011-01-25 09:20:46 <cosurgi> I didn't set any -f, what's the default?
1104 2011-01-25 09:20:51 <Diablo-D3> 60
1105 2011-01-25 09:20:53 <sipa> tcatm: both with art's kernel?
1106 2011-01-25 09:21:03 <tcatm> Diablo-D3: your miner now shows close to 601 Mhash/s for stock 5970
1107 2011-01-25 09:21:15 <tcatm> sipa: yep
1108 2011-01-25 09:21:29 <Diablo-D3> ArtForz: whats your high score now?
1109 2011-01-25 09:21:30 <sipa> looks like Diablo-D3 should implement the vector trick too :)
1110 2011-01-25 09:21:39 <Diablo-D3> sipa: erm, I implemented it first
1111 2011-01-25 09:21:43 <sipa> oh
1112 2011-01-25 09:21:45 <Diablo-D3> I got rid of it awhile ago
1113 2011-01-25 09:21:47 <sipa> i'm sorry
1114 2011-01-25 09:22:07 <sipa> i'll reformulate: looks like Diablo-D3 should re-implement the vector trick :)
1115 2011-01-25 09:22:09 <Diablo-D3> the only platform it was legitimately faster on was 4xxx, and AMD fixed that
1116 2011-01-25 09:22:23 <Diablo-D3> the compiler currently autovectorizes correctly.
1117 2011-01-25 09:23:04 <ArtForz> 601Mh/s on stock 5970 using OCL? not likely.
1118 2011-01-25 09:23:08 <sipa> apparently not, if i can believe tcatm's numbers
1119 2011-01-25 09:23:26 <tcatm> ArtForz: yep, his hashmeter is broken
1120 2011-01-25 09:23:30 <Diablo-D3> it is not.
1121 2011-01-25 09:23:33 <cosurgi> heh, it's easier to believe that it's slower :)
1122 2011-01-25 09:23:44 <cosurgi> which is my case ;-)
1123 2011-01-25 09:24:01 <sipa> i think on cosurgi's hardware maybe the looping slows things down
1124 2011-01-25 09:24:14 <cosurgi> I tried -w 64 -f 60, and -w 64 0f 5.
1125 2011-01-25 09:24:17 <cosurgi> omg, wait.
1126 2011-01-25 09:24:23 <Diablo-D3> sipa: unlikely
1127 2011-01-25 09:24:32 <cosurgi> -w 64 -f 5 gives 534 Mh
1128 2011-01-25 09:24:56 <cosurgi> buuu, it went down to 539 Mh.
1129 2011-01-25 09:25:08 <tcatm> it shows 1988477 khash/s for 2x 5970 @ 810 MHz
1130 2011-01-25 09:25:15 <cosurgi> but still higher than original 518 Mh
1131 2011-01-25 09:25:23 <sipa> 534 -> 539, nice slowdown!
1132 2011-01-25 09:25:30 <Diablo-D3> I love slowdowns like that
1133 2011-01-25 09:25:52 <cosurgi> oh, fuck, it went up, and I can't read
1134 2011-01-25 09:25:54 <slush1> Diablo-D3: is the binary also updated to artforz kernel?
1135 2011-01-25 09:25:58 <Diablo-D3> 1988 mhash? wtf
1136 2011-01-25 09:26:03 <Diablo-D3> now thats just impossible
1137 2011-01-25 09:26:09 <Diablo-D3> slush1: yes
1138 2011-01-25 09:26:17 <slush1> great, i'll try it too
1139 2011-01-25 09:26:39 <ArtForz> I'll have to count operations, but iirc with 100% VLIW usage and optimizing out all constant calcs, ~670Mh/s is max on stock 5970
1140 2011-01-25 09:27:01 <ArtForz> *570
1141 2011-01-25 09:27:21 <cosurgi> ok, it seems to be steady at 530 Mh now.
1142 2011-01-25 09:27:33 <ArtForz> theoretical max with BFI_INT is ~620
1143 2011-01-25 09:27:46 <cosurgi> um, now it's really jumping around, 540 now, 539, 591 ???? 550  528, wtf?
1144 2011-01-25 09:28:08 <Diablo-D3> cosurgi: at low -f values, it bends time
1145 2011-01-25 09:28:13 <cosurgi> ok.
1146 2011-01-25 09:28:25 <tcatm> Diablo-D3: what about using the hardware clock as reference?
1147 2011-01-25 09:28:39 <Diablo-D3> tcatm: it uses whatever your kernel gives it
1148 2011-01-25 09:28:50 <tcatm> which is probably cpu time
1149 2011-01-25 09:29:19 <cosurgi> so the problem might be with hasmeter. It could give wrong value.... ? How wrong?
1150 2011-01-25 09:29:21 <tcatm> or you could use those new HPET
1151 2011-01-25 09:29:36 <Diablo-D3> new?
1152 2011-01-25 09:29:40 <Diablo-D3> they havent been new for like 10 years
1153 2011-01-25 09:29:52 <Diablo-D3> cosurgi: the problem is with time itself.
1154 2011-01-25 09:30:07 <cosurgi> you mean -reading the hardware clock?
1155 2011-01-25 09:30:11 <cosurgi> could it be that my current 539 is in fact slower than previous 518 ?
1156 2011-01-25 09:31:09 <Diablo-D3> cosurgi: doubt it
1157 2011-01-25 09:31:28 <cosurgi> ok.
1158 2011-01-25 09:31:46 <tcatm> whoa 3784683 khash/s :D
1159 2011-01-25 09:32:16 <cosurgi> wow, I got a tx fee 0.03 BTC.
1160 2011-01-25 09:33:23 davout has joined
1161 2011-01-25 09:33:38 <cosurgi> Diablo-D3: would it speed things up if I replaced K[..] with constants in the code below?
1162 2011-01-25 09:33:45 <cosurgi> __constant uint K[64] = {
1163 2011-01-25 09:33:49 <tcatm> cosurgi: no
1164 2011-01-25 09:33:53 <Diablo-D3> no
1165 2011-01-25 09:33:55 <tcatm> they're already constants
1166 2011-01-25 09:34:04 <cosurgi> ok, so it's optimized away.
1167 2011-01-25 09:34:28 davout has quit (Changing host)
1168 2011-01-25 09:34:28 davout has joined
1169 2011-01-25 09:35:27 <cosurgi> do you know what is the kernel size limit? hypothethically thinking about unrolling inner loop, or parts of it, from original ArtForz release.
1170 2011-01-25 09:35:46 <cosurgi> perhaps doesn't make sens, but what is the kernel size?
1171 2011-01-25 09:36:33 <Diablo-D3> there isnt a limit
1172 2011-01-25 09:36:43 <sipa> cosurgi: that would amount to just creating multiple instances in parallel within the same worker
1173 2011-01-25 09:36:44 <Diablo-D3> but the code cache size is small
1174 2011-01-25 09:36:55 <Diablo-D3> sipa: no it wouldnt
1175 2011-01-25 09:36:59 <sipa> sorry, in series
1176 2011-01-25 09:37:02 <sipa> not in parallel
1177 2011-01-25 09:37:03 <Diablo-D3> it'd just slow the kerneld down badly
1178 2011-01-25 09:37:15 <cosurgi> I see. what is the cache size?
1179 2011-01-25 09:37:27 <ArtForz> 64kB iirc
1180 2011-01-25 09:37:35 <cosurgi> ok. Not much.
1181 2011-01-25 09:38:42 <cosurgi> what -f do you use now?
1182 2011-01-25 09:38:55 <cosurgi> I am trying from 5 to 8...
1183 2011-01-25 09:40:11 <Diablo-D3> little changes like that wont matter
1184 2011-01-25 09:40:16 <Diablo-D3> try multiples of 60 and/or 1
1185 2011-01-25 09:40:33 <cosurgi> ok.
1186 2011-01-25 09:42:11 <cosurgi> lol, 5830 -f 120 gives 6 Mh
1187 2011-01-25 09:42:22 <cosurgi> dual 5850 -f 120 gives 17 Mh
1188 2011-01-25 09:47:35 devon_hillard has joined
1189 2011-01-25 09:51:30 echelon has left ("Leaving")
1190 2011-01-25 09:54:27 asdf30 has quit (Ping timeout: 240 seconds)
1191 2011-01-25 09:54:48 <slush1> -w 64 -f 1 gives me slightly lower performance
1192 2011-01-25 09:55:16 <cosurgi> slush1: than which one?
1193 2011-01-25 09:55:38 <slush1> gpus are not on 99%, which was with previous core
1194 2011-01-25 09:55:43 <slush1> oh, now....
1195 2011-01-25 09:56:05 <slush1> +15mhash per 5970
1196 2011-01-25 09:56:39 <slush1> and gpus back at 99%
1197 2011-01-25 09:59:02 satamusic has quit (Ping timeout: 276 seconds)
1198 2011-01-25 10:05:45 altamic has joined
1199 2011-01-25 10:18:13 davout has quit (Quit: i <3 pork (http://dev.ojnk.net))
1200 2011-01-25 10:20:20 davout has joined
1201 2011-01-25 10:20:50 davout has quit (Changing host)
1202 2011-01-25 10:20:51 davout has joined
1203 2011-01-25 10:21:19 AAA_awright_ has joined
1204 2011-01-25 10:21:27 m0mchil has joined
1205 2011-01-25 10:23:13 AAA_awright has quit (Ping timeout: 240 seconds)
1206 2011-01-25 10:26:08 <tcatm> hey m0mchil. I made a patch for your miner.
1207 2011-01-25 10:27:09 <m0mchil> yup, thanks... trying it right now
1208 2011-01-25 10:27:31 <m0mchil> having strange issue with 'longobject.c:336: bad argument to internal function'
1209 2011-01-25 10:27:55 <tcatm> that's indeed strange
1210 2011-01-25 10:27:58 <m0mchil> could you please make a ptach against my latest github?
1211 2011-01-25 10:28:36 <tcatm> the patch on forum should be against your latest github
1212 2011-01-25 10:29:20 <m0mchil> hm ok
1213 2011-01-25 10:33:48 satamusic has joined
1214 2011-01-25 10:37:40 zylche has quit (Quit: zylche)
1215 2011-01-25 10:40:52 <sipa> fails to apply here
1216 2011-01-25 10:41:02 <sipa> 2 hunks failed in .cl and 2 failed in .py
1217 2011-01-25 10:41:23 <tcatm> patch by hand
1218 2011-01-25 10:41:44 <sipa> sure, looks doable
1219 2011-01-25 10:41:53 <tcatm> the .cl is replaced completely I think and it's just a few lines in .py
1220 2011-01-25 10:47:04 zylche has joined
1221 2011-01-25 10:59:17 dukeleto has quit (Excess Flood)
1222 2011-01-25 10:59:51 dukeleto has joined
1223 2011-01-25 11:03:52 darsk1ez has joined
1224 2011-01-25 11:19:05 <necrodearia> http://www.tomshardware.com/news/OpenGL-4.1-Catalyst-FirePro-FireGL-unified-driver,12062.html
1225 2011-01-25 11:42:57 tg has quit (Read error: Connection reset by peer)
1226 2011-01-25 11:43:43 tg has joined
1227 2011-01-25 11:44:04 RazielZ has joined
1228 2011-01-25 11:52:31 <necrodearia> From discussion in another channel that I posted to: https://github.com/bitcoin/bitcoin - It may be helpful to pose as a contributor to that staging tree and to prepare a patch that has a kind of flaw that you recognize but may not be recognizable by others so easily or at least not immediately and as soon as your patch is accepted and included in the staging tree and also eventually the main release, then you can take advantage of the e
1229 2011-01-25 11:52:31 <necrodearia> xploit and getbalance() and sendtoaddress() compromising all currently active bitcoin instances ^_^
1230 2011-01-25 11:52:47 <necrodearia> In regards to possible exploitation or compromise
1231 2011-01-25 11:53:11 TheAncientGoat has quit (Ping timeout: 276 seconds)
1232 2011-01-25 11:53:11 TheAncientGoat_ has joined
1233 2011-01-25 11:54:18 <necrodearia> all currently active bitcoin instances running the malicious version of the client
1234 2011-01-25 11:56:00 <necrodearia> "good reason to keep the wallet/bitcoind featurelight"
1235 2011-01-25 11:56:01 * sipa assumes getting a patch accepted in the main line won't be that easy
1236 2011-01-25 11:56:55 <necrodearia> Even the staging tree should have some concern for not getting accepted so easily.
1237 2011-01-25 11:56:55 <tcatm> If it adds useful functionality or even simplifies existing code it could be easy
1238 2011-01-25 12:01:40 <necrodearia> I sent reference ( http://veritas.maximilianeum.ch/bitcoin/irc/logs/2011/01/25 ) to gavin
1239 2011-01-25 12:02:52 * tcatm digs out an arm based computer and builds a secure bitcoin box with it
1240 2011-01-25 12:02:58 darsk1ez has quit (Quit: Ex-Chat)
1241 2011-01-25 12:06:47 TheAncientGoat_ has quit (Quit: No Ping reply in 180 seconds.)
1242 2011-01-25 12:07:03 TheAncientGoat has joined
1243 2011-01-25 12:20:00 <lfm> necrodearia: that ati driver has been out for ages since dec 13 by the time stamp
1244 2011-01-25 12:20:05 noagendamarket has joined
1245 2011-01-25 12:54:04 <lfm> ;;bc,stats
1246 2011-01-25 12:54:06 <gribble> Current Blocks: 104503 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 328 blocks | Next Difficulty In About: 1 day, 22 hours, and 40 seconds | Next Difficulty Estimate: 21769.48532616
1247 2011-01-25 12:54:14 kermit has quit (Quit: Leaving.)
1248 2011-01-25 13:00:35 <CIA-98> DiabloMiner: Patrick McFarland master * r6ae8542 / src/main/java/com/diablominer/DiabloMiner/DiabloMiner.java : Add specialized hashmeter mode at -f 1, don't allow 0 run kernels - http://bit.ly/gbbd3s
1249 2011-01-25 13:05:11 molecular has quit (Ping timeout: 240 seconds)
1250 2011-01-25 13:05:37 molecular has joined
1251 2011-01-25 13:07:49 <tcatm> Diablo-D3: binary updated?
1252 2011-01-25 13:11:14 Diablo-D3 has quit (Ping timeout: 255 seconds)
1253 2011-01-25 13:12:59 <MacRohard> ;;bc,stats
1254 2011-01-25 13:13:02 <gribble> Current Blocks: 104504 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 327 blocks | Next Difficulty In About: 1 day, 21 hours, 52 minutes, and 15 seconds | Next Difficulty Estimate: 21758.78735681
1255 2011-01-25 13:21:24 <slush1> I see specific GPU's Bus IDs in xorg.conf in my miner. Can it be the reason why computer freeze when I move cards to another slots?
1256 2011-01-25 13:21:49 <ArtForz> yep
1257 2011-01-25 13:21:54 <slush1> oh, great
1258 2011-01-25 13:22:08 <slush1> I spent whole Sunday on this
1259 2011-01-25 13:23:35 <slush1> graphic drivers for ati are crazy. It does not init displays when xorg is missing and totally freeze when there are another BusIDs
1260 2011-01-25 13:23:56 <slush1> so I have to remove xorg.conf, reboot, connect with ssh, call aticonfig --initial and then try to restart
1261 2011-01-25 13:23:57 <slush1> insane
1262 2011-01-25 13:24:16 <tcatm> add something like aticonfig --initial --force to /etc/rc.local
1263 2011-01-25 13:24:50 <slush1> does not break reinitializing on startup another settings made by aticonfig?
1264 2011-01-25 13:25:27 <ArtForz> not really
1265 2011-01-25 13:25:28 <tcatm> you can set them after calling --initial --force
1266 2011-01-25 13:25:30 <slush1> but interesting idea, fan speeds and clocks are not persisted in xorg.conf at all
1267 2011-01-25 13:25:50 <slush1> thanks for tip, tcatm! I'll give it a chance
1268 2011-01-25 13:26:02 <ArtForz> I generally just delete xorg.xonf before powering off for card swapping
1269 2011-01-25 13:26:02 <tcatm> man rc.local is something like: aticonfig --initial --force; X :0 -ac&; sleep 5; /start_miners.sh
1270 2011-01-25 13:26:10 <tcatm> s/man/my/
1271 2011-01-25 13:26:33 noagendamarket has quit (Remote host closed the connection)
1272 2011-01-25 13:27:03 <slush1> what is 'X :0 -ac" part?
1273 2011-01-25 13:27:15 <sipa> X is the X server
1274 2011-01-25 13:27:18 <sipa> :0 is the display
1275 2011-01-25 13:27:35 <slush1> oh, nice
1276 2011-01-25 13:27:55 <sipa> X is normally started by your display manager (the graphical login prompt) under typical desktop distros
1277 2011-01-25 13:27:59 <sipa> but you can start it yourself
1278 2011-01-25 13:29:06 <slush1> I tried to install bare system with X too, but because many troubles with startup freezing, I was unsure where the problem is. So I reinstall it to normal distro.
1279 2011-01-25 13:29:13 Pegasus-Rider has quit (Changing host)
1280 2011-01-25 13:29:13 Pegasus-Rider has joined
1281 2011-01-25 13:29:22 <slush1> thanks for your tips
1282 2011-01-25 13:29:43 <slush1> Pretty bad that my miner is 50km far, so every tiny change is 1 hour by travelling
1283 2011-01-25 13:30:06 <slush1> (on hw part)
1284 2011-01-25 13:30:36 <tcatm> once set up there should be no reason to change anything
1285 2011-01-25 13:30:59 <slush1> of course
1286 2011-01-25 13:32:27 <slush1> but it is still new miner. Yesterday I found that fan is not in good position and I have to move it a little. So for now I had to underclock one card :) Every tiny change is THE problem
1287 2011-01-25 13:33:30 <sipa> anyone an idea why one of the two cores of a 5970 sometimes lowers its speed?
1288 2011-01-25 13:33:53 <tcatm> sipa: overheating?
1289 2011-01-25 13:33:59 <ArtForz> VRMs overheating
1290 2011-01-25 13:34:24 <sipa> i believe the temperature sensor said something around 75 degrees or so
1291 2011-01-25 13:35:20 <sipa> still possible the VRMs are overheating?
1292 2011-01-25 13:35:34 <niekie> <slush1> what is 'X :0 -ac" part? <-- -ac disables the airconditioning in your computer, making it blow up. </trollmode>
1293 2011-01-25 13:35:38 <niekie> ^^
1294 2011-01-25 13:35:44 <ArtForz> very unlikely
1295 2011-01-25 13:35:47 kron has quit (Ping timeout: 265 seconds)
1296 2011-01-25 13:36:06 <ArtForz> VRMs overheating is usually a problem > 90°C core or so
1297 2011-01-25 13:36:39 <sipa> any other explanations?
1298 2011-01-25 13:36:48 <molecular> morning
1299 2011-01-25 13:37:06 <molecular> tcatm, did you and Diablo get art's kernel going and some measurements?
1300 2011-01-25 13:37:29 <sipa> molecular: Diablo's miner is updated ans user ArtForz's kernel now
1301 2011-01-25 13:37:37 <sipa> 3% speed gain
1302 2011-01-25 13:37:39 <molecular> speed?
1303 2011-01-25 13:37:40 <molecular> oh
1304 2011-01-25 13:37:41 <tcatm> molecular: 567 Mhash/s on stock 5970 with hacked SDK
1305 2011-01-25 13:37:53 <sipa> define hached sdk?
1306 2011-01-25 13:38:13 <molecular> tcatm: I think you might have to help me some time with that backporting of compiler
1307 2011-01-25 13:38:14 <tcatm> 2.2 compiler hacked into 2.1 opencl
1308 2011-01-25 13:38:28 <molecular> you said 2.3 -> 2.1 before
1309 2011-01-25 13:38:29 <tcatm> i.e. multigpu support with 2.2
1310 2011-01-25 13:38:54 <tcatm> molecular: might have mistyped it
1311 2011-01-25 13:38:59 <molecular> ok
1312 2011-01-25 13:40:31 <tcatm> problem with hacking SDKs is that it doesn't support amd_bitalign() anymore but it will magically use bitalign for rotate()
1313 2011-01-25 13:40:55 <davout> hmm interesting
1314 2011-01-25 13:45:25 <davout> so how do you tweak the sdk ? put 2.1 libs in 2.2 sdk or opposite ?
1315 2011-01-25 13:46:01 <tcatm> I don't remember what exactly I did. I think I copied some libs an binaries from 2.2 to 2.1
1316 2011-01-25 13:47:43 <molecular> hm, and headers from 2.1?
1317 2011-01-25 13:48:28 <tcatm> looks like I changed builtins_amdil_opt.bc somehow
1318 2011-01-25 13:48:42 <tcatm> but I can't find where I took it from (comparing md5sums)
1319 2011-01-25 13:49:49 <tcatm> I copied at least clc from 2.2 to 2.1
1320 2011-01-25 13:50:48 <tcatm> libglut.so is also from 2.2
1321 2011-01-25 13:52:18 jwalck has joined
1322 2011-01-25 13:52:35 <tcatm> also copied lib/gpu/*.bc to lib/x86_64/
1323 2011-01-25 13:53:01 <tcatm> and copied builtins-evergreen.bc to builtins_amdil_opt.bc
1324 2011-01-25 13:53:06 <molecular> Diablo-D3 new miner runs extremely slow here. 210mhash/s (560 before)
1325 2011-01-25 13:53:31 <molecular> hmm, does he #define LOOPS from java-code somehow?
1326 2011-01-25 13:53:43 <tcatm> molecular: yep
1327 2011-01-25 13:53:48 <molecular> aha
1328 2011-01-25 13:58:44 <molecular> he really made some changes in his java code... or maybe I had an older version before.
1329 2011-01-25 13:59:56 <sipa> yes, ArtForz' kernel uses more precomputed values
1330 2011-01-25 14:00:50 <molecular> I see he turns sets LOOP dynamically based on wether bitalign is available
1331 2011-01-25 14:01:46 <npouillard> ;;bc,calc 235000
1332 2011-01-25 14:01:47 <gribble> The average time to generate a block at 235000 Khps, given current difficulty of 18437.64439217 , is 3 days, 21 hours, 36 minutes, and 14 seconds
1333 2011-01-25 14:01:47 <molecular> Hmm, don't understand why it runs so slowly, though (5970). It ran "normally" when I tried it last night
1334 2011-01-25 14:03:00 <npouillard> ;;bc,calc 243000
1335 2011-01-25 14:03:00 <gribble> The average time to generate a block at 243000 Khps, given current difficulty of 18437.64439217 , is 3 days, 18 hours, 31 minutes, and 20 seconds
1336 2011-01-25 14:08:39 <molecular> is there a pplib-cmd that sets the fan back to auto-control?
1337 2011-01-25 14:09:09 <cosurgi> molecular: try -f 1 or -f 5
1338 2011-01-25 14:09:17 testtest has joined
1339 2011-01-25 14:09:37 <hacim> cosurgi: i haven't tried the new diablo, but what you quoted of me was a bit out of context
1340 2011-01-25 14:09:50 <hacim> cosurgi: 2.2 seems to make things *really* slow
1341 2011-01-25 14:09:51 <cosurgi> hacim: ok, I'm sorry.
1342 2011-01-25 14:10:13 <cosurgi> I misunderstood then.
1343 2011-01-25 14:10:35 <molecular> cosurgi: "aticonfig --pplib-cmd 'set fanspeed 0 80'" <- put "-f" where?
1344 2011-01-25 14:10:45 <cosurgi> molecular: not here.
1345 2011-01-25 14:10:57 <cosurgi> ./DiabloMiner-Linux.sh -w 64 -f 5 ....
1346 2011-01-25 14:11:20 <molecular> ah you mean to get speed from the new version with artforz' kernel?
1347 2011-01-25 14:11:26 <cosurgi> yes.
1348 2011-01-25 14:11:33 <molecular> ok, will try
1349 2011-01-25 14:11:35 <cosurgi> ArtForz's kernel is crazy slow with default -f 60
1350 2011-01-25 14:12:10 <molecular> cosurgi, thanks for the hint, measuring...
1351 2011-01-25 14:12:12 <npouillard> ;;bc,calc 251000
1352 2011-01-25 14:12:14 <gribble> The average time to generate a block at 251000 Khps, given current difficulty of 18437.64439217 , is 3 days, 15 hours, 38 minutes, and 14 seconds
1353 2011-01-25 14:13:16 Maccer has quit (Ping timeout: 240 seconds)
1354 2011-01-25 14:13:46 kermit has joined
1355 2011-01-25 14:14:37 <molecular> it's better now: around 490mhash/s, thanks cosurgi. can't really tell the speed. hashmeter pretty unstable
1356 2011-01-25 14:14:38 <hacim> I'm only able to get 334 on at 5870 @ 900
1357 2011-01-25 14:15:22 <molecular> maybe one could let in run for exactly some time then count the pooled shares found ;)
1358 2011-01-25 14:22:27 <cosurgi> ;;bc,calc 754082
1359 2011-01-25 14:22:28 <gribble> The average time to generate a block at 754082 Khps, given current difficulty of 18437.64439217 , is 1 day, 5 hours, 10 minutes, and 13 seconds
1360 2011-01-25 14:22:49 <cosurgi> with ArtForz's kernel I am one hour better.
1361 2011-01-25 14:22:50 Kiba has joined
1362 2011-01-25 14:24:12 <altamic> http://bit.ly/e7VBgn
1363 2011-01-25 14:24:31 <Kiba> hello
1364 2011-01-25 14:25:46 <molecular> so noone knows how to set fan back to autocontrol using aticonfig?
1365 2011-01-25 14:27:42 <cosurgi> molecular: impossible, you must reboot.
1366 2011-01-25 14:27:42 <molecular> cosurgi, thanks.
1367 2011-01-25 14:27:42 <cosurgi> You are welcome :)
1368 2011-01-25 14:28:43 <hacim> tcatm: tried copying clc, libglut.so from 2.2 to 2.1, lib/gpu/*.bc to lib/x86_64/; copied builtins-evergreen.bc to builtins_amdil_opt.bc
1369 2011-01-25 14:29:15 <hacim> tcatm: then I set my ATISTREAMSDKROOT, ATISTREAMSDKSAMPLESROOT, LD_LIBRARY_PATH env vars to the new location, but I get 1339 khash/sec with that setup :P
1370 2011-01-25 14:31:25 <molecular> what's the best way to get EUR from my btc to germany?
1371 2011-01-25 14:32:46 <molecular> mtgox says: "If you live in the US or EU we can send it to you by direct deposit." they will charge 2%, no limit
1372 2011-01-25 14:33:00 <molecular> i mean: no minimum
1373 2011-01-25 14:33:38 <sipa> ;;bc,calcd 82000 1
1374 2011-01-25 14:33:39 <gribble> The average time to generate a block at 82000 Khps, given the supplied difficulty of 1, is 52 seconds
1375 2011-01-25 14:34:48 <CIA-98> bitcoin: Gavin Andresen master * rb1a657a / build-unix.txt : Explicitly state that make -f makfile.unix is the way to build - http://bit.ly/dFWGyU
1376 2011-01-25 14:35:29 <hacim> tcatm: could you maybe put your hacked sdk up somewhere so I can compare them?
1377 2011-01-25 14:35:54 <comboy> ;;bc,calcd 600000 40000
1378 2011-01-25 14:35:54 <gribble> The average time to generate a block at 600000 Khps, given the supplied difficulty of 40000, is 3 days, 7 hours, 32 minutes, and 11 seconds
1379 2011-01-25 14:36:03 <comboy> ;;bc,calcd 600000 80000
1380 2011-01-25 14:36:03 <gribble> The average time to generate a block at 600000 Khps, given the supplied difficulty of 80000, is 6 days, 15 hours, 4 minutes, and 22 seconds
1381 2011-01-25 14:43:30 noagendamarket has joined
1382 2011-01-25 14:57:16 ducki2p has quit (Ping timeout: 240 seconds)
1383 2011-01-25 14:57:46 dukeleto has quit (Ping timeout: 246 seconds)
1384 2011-01-25 14:58:16 dukeleto has joined
1385 2011-01-25 14:58:41 <luke-jr> :O
1386 2011-01-25 14:58:55 <luke-jr> nobody told me AMD's blob crap requires turning off kernel preemption!
1387 2011-01-25 14:59:28 dukeleto has quit (Excess Flood)
1388 2011-01-25 14:59:35 <bonsaikitten> luke-jr: it's proprietary, it needs to be speciuhl
1389 2011-01-25 15:00:00 dukeleto has joined
1390 2011-01-25 15:00:35 <Kiba> Teppy1: yo
1391 2011-01-25 15:09:44 ducki2p has joined
1392 2011-01-25 15:21:59 xelister has quit (Remote host closed the connection)
1393 2011-01-25 15:28:40 larsivi has quit (Ping timeout: 240 seconds)
1394 2011-01-25 15:43:37 jrabbit has quit (Ping timeout: 246 seconds)
1395 2011-01-25 15:49:40 noagendamarket has quit (Ping timeout: 240 seconds)
1396 2011-01-25 15:49:44 jago25_98 has joined
1397 2011-01-25 15:56:30 jrabbit has joined
1398 2011-01-25 16:12:51 <npouillard> ;bc,stats
1399 2011-01-25 16:13:10 <npouillard> ;;bc,stats
1400 2011-01-25 16:13:12 <gribble> Current Blocks: 104525 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 306 blocks | Next Difficulty In About: 1 day, 19 hours, 10 minutes, and 48 seconds | Next Difficulty Estimate: 21765.62402128
1401 2011-01-25 16:21:07 TheAncientGoat has quit (Ping timeout: 276 seconds)
1402 2011-01-25 16:24:37 <rapacity> things are getting really difficult :< it's almost impossible to mine buttcoins now
1403 2011-01-25 16:25:17 brunner has quit (Ping timeout: 255 seconds)
1404 2011-01-25 16:26:57 molecular has quit (Ping timeout: 240 seconds)
1405 2011-01-25 16:28:38 davout has quit (Quit: i <3 pork (http://dev.ojnk.net))
1406 2011-01-25 16:36:04 akem has quit (Read error: Connection reset by peer)
1407 2011-01-25 16:41:31 molecular has joined
1408 2011-01-25 16:47:54 jago25_98 has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
1409 2011-01-25 16:51:53 m0mchil has quit ()
1410 2011-01-25 16:52:21 devon_hillard_ has joined
1411 2011-01-25 16:54:53 andrew12^school has joined
1412 2011-01-25 16:55:16 devon_hillard has quit (Ping timeout: 240 seconds)
1413 2011-01-25 16:55:26 <andrew12^school> hi yes
1414 2011-01-25 16:55:29 <andrew12^school> :p
1415 2011-01-25 16:55:48 <andrew12^school> LOL
1416 2011-01-25 16:55:58 <andrew12^school> bitcoin.org is blocked at school for "P2P/File sharing"
1417 2011-01-25 16:56:12 <andrew12^school> half of it is correct
1418 2011-01-25 16:59:03 <nanotube> hahaha
1419 2011-01-25 16:59:15 <nanotube> submit a complaint.
1420 2011-01-25 16:59:17 <andrew12^school> but they didn't block the proxy i usually use
1421 2011-01-25 16:59:42 <andrew12^school> nanotube: i've tried that before. you have to have a school email (i.e. you have to be a teacher) to even get beyond the form
1422 2011-01-25 17:00:06 <andrew12^school> but I can't get any connections anyways because of the way the network is set up
1423 2011-01-25 17:00:14 <andrew12^school> gotta get ptunnel working :P
1424 2011-01-25 17:00:49 <nanotube> heh or just set up an ssh tunnel to a server of your choice.
1425 2011-01-25 17:01:05 <andrew12^school> can't do that
1426 2011-01-25 17:01:25 <andrew12^school> the only ports that arent blocked are http, http ssl, and ping
1427 2011-01-25 17:01:29 <ArtForz> or implement ip-over-http-header
1428 2011-01-25 17:01:48 <andrew12^school> ArtForz: or use ptunnel.
1429 2011-01-25 17:02:04 <andrew12^school> ptunnel uses ICMP packets (ping packets)
1430 2011-01-25 17:02:23 <ArtForz> yes, and any halfway locked down network has icmp filtered
1431 2011-01-25 17:02:44 <andrew12^school> nope, ping works fine
1432 2011-01-25 17:03:08 <ArtForz> yay for incompetent admins
1433 2011-01-25 17:03:10 <andrew12^school> also on some of the laptops here, if you log in to the laptop with the username "test" with no password, you can install stuff.
1434 2011-01-25 17:03:39 <andrew12^school> for example this one
1435 2011-01-25 17:03:42 <andrew12^school> :P
1436 2011-01-25 17:10:24 molecular has quit (Ping timeout: 250 seconds)
1437 2011-01-25 17:10:46 molecular has joined
1438 2011-01-25 17:21:14 brunner has joined
1439 2011-01-25 17:24:25 andrew12^school has quit (Ping timeout: 265 seconds)
1440 2011-01-25 17:25:26 <luke-jr> andrew12: SSH on port 143
1441 2011-01-25 17:25:47 <luke-jr> ArtForz: ICMP filtering is illegal
1442 2011-01-25 17:26:24 nevezen has joined
1443 2011-01-25 17:28:44 <sipa> illegal? by what law?
1444 2011-01-25 17:29:01 <luke-jr> IP :p
1445 2011-01-25 17:48:48 gusti has joined
1446 2011-01-25 17:49:57 gusti has quit (Client Quit)
1447 2011-01-25 17:50:17 bertodsera has quit (Ping timeout: 272 seconds)
1448 2011-01-25 18:05:20 chaord has joined
1449 2011-01-25 18:10:57 slush_ has joined
1450 2011-01-25 18:13:22 <slush_> hi, when I see "CrossFire chain for adapter 0, status: disabled", it is enough?
1451 2011-01-25 18:13:39 <slush_> Or need I remove CF chains completely?
1452 2011-01-25 18:16:10 <slush_> I see all adapters in aticonfig, but only one device is detected by DiabloMiner
1453 2011-01-25 18:16:28 <slush_> so I'm curious what else i broken
1454 2011-01-25 18:16:41 <slush_> tcatm: and thanks for advice with --initial in rc.local, it works!
1455 2011-01-25 18:20:40 bertodsera has joined
1456 2011-01-25 18:21:38 <slush_> ArtForz: ---6^^
1457 2011-01-25 18:22:00 <slush_> ArtForz: please, it will be something really easy...
1458 2011-01-25 18:23:54 <ArtForz> dont forget DISPLAY=:0
1459 2011-01-25 18:26:50 <slush_> I have DISPLAY=:0 almost everywhere :)
1460 2011-01-25 18:27:06 <slush_> Diablo's detect one core and it run on 200mhash, which indicate there is something messed up
1461 2011-01-25 18:27:34 <ArtForz> does xorg.conf list all devices?
1462 2011-01-25 18:28:26 <slush_> wow, what's the magic? THere is only one device listed
1463 2011-01-25 18:28:46 <slush_> But aticonfig --initial is started in rc.local
1464 2011-01-25 18:28:59 <slush_> And in previous restart, it detected everything succesfully
1465 2011-01-25 18:29:15 <ArtForz> aticonfig --initial --adapter=all -f
1466 2011-01-25 18:29:21 <slush_> fuck
1467 2011-01-25 18:29:25 <slush_> forgot to adapter=all
1468 2011-01-25 18:29:33 <slush_> I'm retarted :/
1469 2011-01-25 18:30:32 <slush_> You saved my ass, again.
1470 2011-01-25 18:34:52 xelister has joined
1471 2011-01-25 18:34:52 xelister has quit (Changing host)
1472 2011-01-25 18:34:52 xelister has joined
1473 2011-01-25 18:35:02 <slush_> ArtForz: well, xorg is now correct, crossfires are disabled, but issue is the same; still only one device in miner
1474 2011-01-25 18:35:56 <ArtForz> hrrrm
1475 2011-01-25 18:36:45 <xelister> slush_: and you have DISPLAY=0:0 ?
1476 2011-01-25 18:36:53 <ArtForz> display=:0 you mean
1477 2011-01-25 18:36:56 <ArtForz> err DISPLAY
1478 2011-01-25 18:36:58 <xelister> yea
1479 2011-01-25 18:37:21 <slush_> yes I have
1480 2011-01-25 18:38:19 <ArtForz> what does aticonfig --lsch say?
1481 2011-01-25 18:38:47 <slush_> when I try --odgt --adapter X, it works only for 0 and 1. Not for >1
1482 2011-01-25 18:39:23 <slush_> --lsch: CF chain for adapter 0, status: disabled
1483 2011-01-25 18:39:38 <slush_> and then two lines with busIDs of two devices
1484 2011-01-25 18:39:42 <ArtForz> aticonfig --adapter=all --cfd
1485 2011-01-25 18:39:45 <slush_> but there are 4 5970
1486 2011-01-25 18:40:01 <slush_> hmm... the same
1487 2011-01-25 18:40:04 <ArtForz> may have to do that with X stopped
1488 2011-01-25 18:40:08 <slush_> sorry
1489 2011-01-25 18:40:30 <slush_> CrossFIre chains deleted
1490 2011-01-25 18:40:37 <slush_> Warning: X needs to be restarted ....
1491 2011-01-25 18:40:58 <ArtForz> restart X, see if aticonfig --lsch now says no chains defined
1492 2011-01-25 18:41:05 <ArtForz> it should
1493 2011-01-25 18:41:13 <slush_> oh, now --lsch shows No crossfire chains defined
1494 2011-01-25 18:41:28 <slush_> shoud I add --cfd to rc.local too?
1495 2011-01-25 18:41:32 <ArtForz> any change for --adapter=all --odgt ?
1496 2011-01-25 18:41:43 <slush_> great
1497 2011-01-25 18:41:58 <slush_> oh well, two temperatures and 6x error message
1498 2011-01-25 18:42:08 <ArtForz> what error message?
1499 2011-01-25 18:42:17 <slush_> Get temperature failed for Adapter X
1500 2011-01-25 18:42:35 <ArtForz> hrrrm... I also had that quite a few times
1501 2011-01-25 18:42:50 <ArtForz> seems related to mode switching or something
1502 2011-01-25 18:43:17 <slush_> ok, I'll restart X with CF deleted
1503 2011-01-25 18:43:26 <ArtForz> yep
1504 2011-01-25 18:44:04 <ArtForz> and check if it stays deleted, those crossfire chains are harder to kill than vampires
1505 2011-01-25 18:44:30 <slush_> I added it to rc.local, hope it helps
1506 2011-01-25 18:53:38 jackmcbarn has quit (Ping timeout: 272 seconds)
1507 2011-01-25 18:54:16 sneak has quit (Ping timeout: 272 seconds)
1508 2011-01-25 18:56:32 <xelister> slush_: if all else fails, nuke /etc/ati from low orbit and reinstall drivers
1509 2011-01-25 18:57:03 <xelister> slush_: or leave voicemail to Ati software department, reminding them to get their shit togeather
1510 2011-01-25 18:57:36 <slush_> hm,after restart it completely freezed
1511 2011-01-25 18:57:47 <xelister> some operations must be from console with X stopped, others (most aticonfig uses, perhaps not all) must be done with X running, within the X (with proper DISPLAY)
1512 2011-01-25 18:58:13 <slush_> so now I have troubles how to get at least console
1513 2011-01-25 18:59:27 <xelister> slush_: try rescue mode
1514 2011-01-25 19:01:49 <slush_> i tried to install ext2 drivers to this notebook, but there's no driver compatible with win7 :/
1515 2011-01-25 19:02:46 <luke-jr> slush_: trust me, that's a good thing
1516 2011-01-25 19:02:53 <luke-jr> all the Windows ext2 drivers like to corrupt crap
1517 2011-01-25 19:03:27 <xelister> slush_: if nothing else works then ping me / pm me
1518 2011-01-25 19:03:51 <slush_> I'm going crazy, nothing work properly here
1519 2011-01-25 19:04:03 <midnightmagic> you have to ensure that xorg attaches a screen to the device.
1520 2011-01-25 19:04:21 <midnightmagic> even if it's invisible and you can't move a mouse to it (because coords are detached from main screen)
1521 2011-01-25 19:05:04 <midnightmagic> ah, i'm answering stuff from a couple days ago, just ignore me..
1522 2011-01-25 19:05:25 jackmcbarn has joined
1523 2011-01-25 19:10:12 dishwara has joined
1524 2011-01-25 19:10:50 sneak has joined
1525 2011-01-25 19:11:17 sneak is now known as Guest50437
1526 2011-01-25 19:13:32 devon_hillard_ has quit (Quit: Leaving)
1527 2011-01-25 19:13:41 devon_hillard has joined
1528 2011-01-25 19:14:45 <slush_> ArtForz: I absolutely don't understand why, but after restart with CF diabled, computer freeze when I start X
1529 2011-01-25 19:15:12 <slush_> ArtForz: I can go to recovery mode, check xorg.conf and so, but once I start X :0 -cf, it hangs
1530 2011-01-25 19:15:37 <slush_> aticonfig show all devices
1531 2011-01-25 19:16:10 <ArtForz> xorg.conf ok?
1532 2011-01-25 19:16:17 <ArtForz> = li9siting all 8 devices
1533 2011-01-25 19:16:21 <slush_> yes, all 8 devices here
1534 2011-01-25 19:17:10 <ArtForz> hmm weird
1535 2011-01-25 19:17:33 <ArtForz> does xorg create a log?
1536 2011-01-25 19:19:04 <slush_> ArtForz: no
1537 2011-01-25 19:19:06 <slush_> it's empty
1538 2011-01-25 19:19:17 <ArtForz> hmmm... thats really weird
1539 2011-01-25 19:19:34 <ArtForz> is it really hanging on X start or earlier?
1540 2011-01-25 19:19:49 <slush_> Now I'm in recovery console
1541 2011-01-25 19:20:06 <slush_> Immediately after # X :0 it hangs
1542 2011-01-25 19:21:54 <slush_> lspci shows me cards as "ATI Technologies Inc Device 689c". Don't know if it's OK
1543 2011-01-25 19:22:17 <cosurgi> slush_: you can try confirming that by running startx, last time I tried, startx worked remotely, so in shell session you'd see messages, right before it hangs.
1544 2011-01-25 19:22:58 <cosurgi> but it was some time ago, and I might remember wrong, that startx runs remotely...
1545 2011-01-25 19:23:23 <slush_> hmm, only notebook with win7 here :(
1546 2011-01-25 19:23:31 <cosurgi> putty ?
1547 2011-01-25 19:24:12 <slush_> oh, I thought I need X forwarding on remote machine
1548 2011-01-25 19:24:15 <slush_> I'll try
1549 2011-01-25 19:24:31 <cosurgi> no, you don't need.
1550 2011-01-25 19:25:39 dishwara has quit (Read error: Connection reset by peer)
1551 2011-01-25 19:25:57 dishwara has joined
1552 2011-01-25 19:26:08 <slush_> (WW) fglrx: No matching Device section for instance (BusID PCI:0@19:0:1) found
1553 2011-01-25 19:26:14 <slush_> many lines like this
1554 2011-01-25 19:26:23 <slush_> then freeze and putty disconnect
1555 2011-01-25 19:26:41 <cosurgi> at least something.
1556 2011-01-25 19:26:44 <ArtForz> thats really weird
1557 2011-01-25 19:27:05 <cosurgi> does lspci list PCI:0@19:0:1 and is there a match for it in xorg.conf ?
1558 2011-01-25 19:27:13 <slush_> looks like xorg has wrong busids
1559 2011-01-25 19:27:29 <ArtForz> the "no matching device section" is normal, should get for all pci devices NOT listed in xorg.conf
1560 2011-01-25 19:27:31 <slush_> I'll take a look, once it start to recovery shell again
1561 2011-01-25 19:29:35 <slush_> hm
1562 2011-01-25 19:29:44 <slush_> no busid 19 in lspci
1563 2011-01-25 19:29:59 <ArtForz> lspci is hex, xorg.conf is dec
1564 2011-01-25 19:31:29 <slush_> hm, then it looks OK
1565 2011-01-25 19:32:09 <slush_> anybody interested in my four 5970 cards?
1566 2011-01-25 19:32:38 <slush_> :)
1567 2011-01-25 19:33:05 <slush_> I'm out of my fantasy, what happen here
1568 2011-01-25 19:34:06 <slush_> Before an hour, it does not work in miner, but it started. Now after playing with CF, it just hangs
1569 2011-01-25 19:34:22 <ArtForz> to start over, just delete xorg.conf, delete /etc/ati/amdpcsdb, run aticonfig --initial --adapter=all -f, ...
1570 2011-01-25 19:37:14 <slush_> no change
1571 2011-01-25 19:39:02 malfy has quit (Remote host closed the connection)
1572 2011-01-25 19:39:43 dishwara has quit (Read error: Connection reset by peer)
1573 2011-01-25 19:40:26 RazielZ has quit ()
1574 2011-01-25 19:45:37 <slush_> great. I deleted xorg and amdpcsdb and started computer with one card. It works
1575 2011-01-25 19:45:43 <slush_> let's do it with second one
1576 2011-01-25 19:46:13 Myckel has joined
1577 2011-01-25 19:46:38 malfy has joined
1578 2011-01-25 19:47:02 larsivi has joined
1579 2011-01-25 19:50:38 malfy has quit (Read error: Connection reset by peer)
1580 2011-01-25 19:51:23 asdf30 has joined
1581 2011-01-25 19:53:45 <slush_> well, two cards working. But miner again see only one device
1582 2011-01-25 19:54:16 <slush_> (mining worked well on two cards today, before I plugged next cards into)
1583 2011-01-25 19:55:50 <slush_> but now even aticonfig throws errors, when asking for temperature of 3rd and 4th device
1584 2011-01-25 20:02:11 chaord has left ()
1585 2011-01-25 20:05:13 <slush_> ArtForz: I got two cards running back
1586 2011-01-25 20:05:30 <slush_> ArtForz: But when I connected third, white stripes appeared on boot
1587 2011-01-25 20:07:52 <ArtForz> hurrrm, weird
1588 2011-01-25 20:12:19 <andrew12> hi
1589 2011-01-25 20:13:38 andrew12 has quit (Changing host)
1590 2011-01-25 20:13:38 andrew12 has joined
1591 2011-01-25 20:14:51 Zarutian has joined
1592 2011-01-25 20:17:59 <slush_> ArtForz: well, I move to next level. I have three cards connected, aticonfig works well, no CF defined. When I start miner, computer freeze
1593 2011-01-25 20:18:05 <slush_> This looks like poor PSU
1594 2011-01-25 20:18:21 <slush_> but I have two Corsair 850W
1595 2011-01-25 20:20:57 tg has quit (Read error: Connection reset by peer)
1596 2011-01-25 20:21:16 <ArtForz> hummm
1597 2011-01-25 20:21:48 <slush_> computer not freeze, just X hangs down
1598 2011-01-25 20:22:00 <ArtForz> weird, it does sound somewhat like a power issue
1599 2011-01-25 20:22:00 <slush_> but it freeze when I try to start them again
1600 2011-01-25 20:22:01 tg has joined
1601 2011-01-25 20:22:31 <ArtForz> but then it might just be fglrx fucking up as usual
1602 2011-01-25 20:25:16 <slush_> I tried to connect two cards to one PSU and one card + PC to second one. Does not help
1603 2011-01-25 20:26:34 <ArtForz> hrrrm, should be plenty power
1604 2011-01-25 20:26:54 <ArtForz> can you pastebin your xorg.conf?
1605 2011-01-25 20:27:37 <slush_> I can, once it start up again
1606 2011-01-25 20:31:51 <slush_> I started PC back with two cards. Now it is broken again! It makes me mad!
1607 2011-01-25 20:32:18 <slush_> I didn't changed anything, just two restarts
1608 2011-01-25 20:33:12 <slush_> http://pastebin.com/GpqSJrE3
1609 2011-01-25 20:33:24 <slush_> for two cards, which does not work now
1610 2011-01-25 20:33:49 <ArtForz> yup, thats fine
1611 2011-01-25 20:39:25 <devon_hillard> http://www.techpowerup.com/139121/AMD-Slashes-Prices-of-Radeon-HD-6800-and-HD-6900-Series-Products.h
1612 2011-01-25 20:40:22 <devon_hillard> The 6xxx series uses dual-GPU with internal crossfire, right?
1613 2011-01-25 20:42:10 <ArtForz> nope
1614 2011-01-25 20:42:31 <ArtForz> only dual gpu 6xxx will be 6990, should launch february or so
1615 2011-01-25 20:44:25 <ArtForz> for mining 6850 is between 5770 and 5830, 6870 a bit below 5850, 6950 a bit above 5850, 6970 about the same as 5870
1616 2011-01-25 20:44:42 <ArtForz> so 6990 will probably end up about equal to 5970
1617 2011-01-25 20:44:53 lyspooner has joined
1618 2011-01-25 20:45:06 <andrew12> damn it
1619 2011-01-25 20:45:15 <andrew12> broke my friends vps :(
1620 2011-01-25 20:45:19 <andrew12> by updating the packages
1621 2011-01-25 20:45:25 <andrew12> "FATAL: kernel too old" for like every command now
1622 2011-01-25 20:46:57 <devon_hillard> ArtForz: so the 6xxx series is just better at graphics loads?
1623 2011-01-25 20:47:13 <ArtForz> yep
1624 2011-01-25 20:47:42 <ArtForz> they pretty much improved everything except raw shader throughput :/
1625 2011-01-25 20:48:37 <ArtForz> and mining perf/W actually dropped
1626 2011-01-25 20:48:57 RichardG_ has joined
1627 2011-01-25 20:49:06 AAA_awright_ is now known as AAA_awright
1628 2011-01-25 20:49:29 bertodsera has quit (Remote host closed the connection)
1629 2011-01-25 20:49:30 <ArtForz> perf/W they're on 5770 level, not too bad but no 5970
1630 2011-01-25 20:49:42 RichardG_ is now known as RichardG
1631 2011-01-25 20:50:09 malfy has joined
1632 2011-01-25 20:54:51 Myckel has quit (Quit: Ik ga weg)
1633 2011-01-25 20:54:57 <xelister> andrew12: no, you did not
1634 2011-01-25 20:55:07 <xelister> andrew12: you broke your ex-friend vps <_<
1635 2011-01-25 20:55:36 <andrew12> by updating the packages.
1636 2011-01-25 20:55:55 <andrew12> [03:50:34p] <@Liam_> it's unmanaged
1637 2011-01-25 20:55:55 <andrew12> [03:50:39p] <@Liam_> which means it's <censored> for evar
1638 2011-01-25 20:56:09 <xelister> thats what I just said
1639 2011-01-25 20:58:16 <andrew12> pretty sure we'll still be friends as this isn't my fault
1640 2011-01-25 20:59:55 <ducki2p> andrew12: I once caused an ISP to run chmod -R 644 in the root of one of his main servers.. he didnt understand that he had to run it in a specific subdir
1641 2011-01-25 21:00:18 <andrew12> HAHAHAHAHA
1642 2011-01-25 21:00:42 <xelister> ducki2p: good isp
1643 2011-01-25 21:00:42 <ducki2p> that pretty much messed up the entire system... but we got restore it with a couple hours of work
1644 2011-01-25 21:00:48 <ducki2p> he had to smoke a lot of ciggies though
1645 2011-01-25 21:01:04 * Kiba works his tail off
1646 2011-01-25 21:02:41 bertodsera has joined
1647 2011-01-25 21:04:25 <andrew12> well... i need a new vps so I don't end up fucking my friends up.
1648 2011-01-25 21:05:18 <andrew12> plus it'd be nice to have my own :P
1649 2011-01-25 21:07:53 <xelister> you can buy some vps by utc
1650 2011-01-25 21:07:58 <xelister> *btc
1651 2011-01-25 21:10:23 bk128 has joined
1652 2011-01-25 21:10:32 <andrew12> yeah
1653 2011-01-25 21:12:10 <Kiba> blah
1654 2011-01-25 21:12:35 gavinandresen has joined
1655 2011-01-25 21:12:49 <andrew12> hi gavinandresen
1656 2011-01-25 21:12:58 <gavinandresen> Howdy
1657 2011-01-25 21:13:00 <Kiba> fabianhjr: why are you...so absent?
1658 2011-01-25 21:13:18 <Kiba> hi gavinandresen
1659 2011-01-25 21:13:23 <Kiba> gavinandresen: got any work for me?
1660 2011-01-25 21:13:40 <lfm> ;;seen fabianhjr
1661 2011-01-25 21:13:41 <gribble> fabianhjr was last seen in #bitcoin-dev 1 day, 4 hours, 11 minutes, and 48 seconds ago: <fabianhjr> I hate those. Already lost a chance because of it. xD
1662 2011-01-25 21:13:53 <gavinandresen> Kiba: do you know anything about ubuntu vmbuilder?
1663 2011-01-25 21:13:58 <Kiba> no
1664 2011-01-25 21:14:33 <Kiba> apperantly, the lack of skills prevent my employment
1665 2011-01-25 21:14:41 <gavinandresen> How about creating USB memory stick images?  I've got an idea for a bitcoin business that doesn't involve writing code....
1666 2011-01-25 21:15:02 <donpdonp> Kiba: what kind of work do you want?
1667 2011-01-25 21:15:12 <UukGoblin> gribble, later tell mtgox any idea why I'm getting 'not logged in' when I go to https://mtgox.com/code/getFunds.php?name=goblin&pass=password'?
1668 2011-01-25 21:15:41 <andrew12> UukGoblin: iirc you can't send gribble commands with "gribble,"
1669 2011-01-25 21:15:43 <andrew12> you have to use ;;
1670 2011-01-25 21:16:10 <lfm> uukgoblin cookies?
1671 2011-01-25 21:16:25 <UukGoblin> lfm, the API specification doesn't mention any cookies
1672 2011-01-25 21:16:32 <UukGoblin> andrew12, thanks
1673 2011-01-25 21:16:41 <UukGoblin> ;;later tell mtgox any idea why I'm getting 'not logged in' when I go to https://mtgox.com/code/getFunds.php?name=goblin&pass=password ?
1674 2011-01-25 21:16:41 <gribble> The operation succeeded.
1675 2011-01-25 21:16:51 <andrew12> UukGoblin: ftr its not working for me either
1676 2011-01-25 21:16:54 <UukGoblin> affirmative
1677 2011-01-25 21:17:40 <Kiba> donpdonp: I can do some drawing, writing, and programming
1678 2011-01-25 21:17:57 * Kiba needs more practice in drawing apperantly
1679 2011-01-25 21:24:09 * molecular sent 37$ in cash to bitcoin4cash.com in canada and received 62.99. that's a rate of 0.58 $/BTC !! can anyone explain why I'm getting such a good rate while the page currently advertises 0.41?
1680 2011-01-25 21:25:00 * sipa hopes molecular is being ironic
1681 2011-01-25 21:25:12 <xelister> molecular: huh?
1682 2011-01-25 21:25:24 <molecular> how much is one USD worth (roughly) currently in BTC?
1683 2011-01-25 21:25:39 <Kiba> I thik 2.50
1684 2011-01-25 21:26:10 <sipa> it's worth as much as people are willing to give for it, but on mtgox that's currently around $0.41/BTC
1685 2011-01-25 21:26:20 <xelister> 2.1 .. 2.5 depending
1686 2011-01-25 21:26:23 <molecular> oh, so I got a really bad deal
1687 2011-01-25 21:26:30 <Kiba> 37 / 0.41
1688 2011-01-25 21:26:52 <lfm> ;;calc 37/0.41
1689 2011-01-25 21:26:53 <gribble> 37 / 0.41 = 90.2439024
1690 2011-01-25 21:27:16 <lfm> lotta service charges there
1691 2011-01-25 21:27:55 <UukGoblin> hrm
1692 2011-01-25 21:27:55 <molecular> currently it says rate 0.41, fee 2%
1693 2011-01-25 21:28:14 <UukGoblin> I'm getting weird read background on and below the "Amount to Sell" on mtgox
1694 2011-01-25 21:28:22 <UukGoblin> s/read/red/
1695 2011-01-25 21:29:15 <Kiba> molecular: maybe you sent the dollars so earlier
1696 2011-01-25 21:29:19 <Kiba> so much*
1697 2011-01-25 21:29:29 <Kiba> so they exchange at the rate of five days ago or something
1698 2011-01-25 21:29:30 <Kiba> I don't know
1699 2011-01-25 21:29:40 <sipa> it's never been 0.58
1700 2011-01-25 21:31:07 <molecular> 37 / 62.99 = 0.587 <- that's the rate he used ;(
1701 2011-01-25 21:31:12 <gavinandresen> molecular: ask them, might be just an honest mistake.
1702 2011-01-25 21:31:13 <molecular> never been there on mtgox
1703 2011-01-25 21:31:19 <molecular> yeah, I will ask
1704 2011-01-25 21:31:34 <molecular> maybe I miscounted the money, too, who knows...
1705 2011-01-25 21:32:02 satamusic_ has joined
1706 2011-01-25 21:32:02 <UukGoblin> aa
1707 2011-01-25 21:32:05 <molecular> hmm, seems like $10 might be missing... 27 / 62.99 = 0.42
1708 2011-01-25 21:32:19 <UukGoblin> there's a "Please enter a value greater to or equal to 10" written with a red font on a red background
1709 2011-01-25 21:32:39 <andrew12> hehe yeah
1710 2011-01-25 21:33:02 <xelister> $money_to_give = $amount * $rate;
1711 2011-01-25 21:33:11 <xelister> $money_to_give *= 0.91; // honest mistake
1712 2011-01-25 21:33:15 <xelister> <_<
1713 2011-01-25 21:33:42 <molecular> I'm pretty sure I made the mistake now ;) (shame on me)
1714 2011-01-25 21:33:44 <Kiba> where is fabianjhr when I need him?
1715 2011-01-25 21:33:45 <sipa> if that were open source, it would actually be "honest" :)
1716 2011-01-25 21:34:01 <molecular> I went to the form on the page and chrome remembers I entered "27".
1717 2011-01-25 21:34:06 <molecular> just remembered the amount wrongly
1718 2011-01-25 21:34:41 <sipa> so they used 0.429
1719 2011-01-25 21:34:46 <sipa> that's not bad
1720 2011-01-25 21:35:08 <molecular> yeah
1721 2011-01-25 21:35:48 <molecular> they must've used something lower considering the 2% fee
1722 2011-01-25 21:36:05 <molecular> (27. *0.98) / 62.99 = 0.4200
1723 2011-01-25 21:36:10 <sipa> 0.42
1724 2011-01-25 21:36:12 <sipa> indeed
1725 2011-01-25 21:39:03 <molecular> the automatic fan control of my 5970 lets the temp get a little too high. (91°C). Does someone have a script handy to regulate the fan to match a target temp?
1726 2011-01-25 21:40:24 <sipa> no, but you can set it higher manually
1727 2011-01-25 21:41:34 <lfm> you can force fan to full speed
1728 2011-01-25 21:41:35 <molecular> I know. but I'd like it automated to adjust for room temp / freq / ...
1729 2011-01-25 21:41:41 <molecular> that's too loud
1730 2011-01-25 21:41:41 zorg987 has joined
1731 2011-01-25 21:42:09 <lfm> 91 degrees is like fan stopped
1732 2011-01-25 21:42:14 zorg987 has quit (Client Quit)
1733 2011-01-25 21:42:18 <molecular> uh
1734 2011-01-25 21:42:22 <molecular> 47%
1735 2011-01-25 21:42:54 <sipa> you can tweak the speed until you have a nice compromise between loudness and temperature
1736 2011-01-25 21:42:57 zorg987 has joined
1737 2011-01-25 21:43:05 <sipa> and then hope that load/room temp don't change :)
1738 2011-01-25 21:43:20 <lfm> or write your own script to read the temp and set the fan speed
1739 2011-01-25 21:43:31 <molecular> I can also make a script to check the temp and change speed accordingly. just thinking should I use bash or python or java...
1740 2011-01-25 21:43:39 * sipa would use bash
1741 2011-01-25 21:43:43 <zorg987> Hi guys
1742 2011-01-25 21:43:46 <ArtForz> I implemented a simple PI controller in C
1743 2011-01-25 21:44:03 <molecular> PI?
1744 2011-01-25 21:44:22 <ArtForz> Proportional + Integral
1745 2011-01-25 21:44:31 <andrew12> 3.14159265358979..
1746 2011-01-25 21:44:31 <lfm> ya bash would work fine, its not like you need microsecond reponse
1747 2011-01-25 21:44:33 <ArtForz> using ADL lib directly bypasses a lot of aticonfig braindamage
1748 2011-01-25 21:45:09 <molecular> bash is likely quickest to impl. doing it in java/py would allow for easier integration into miners ;)
1749 2011-01-25 21:45:37 <sipa> π
1750 2011-01-25 21:45:42 <sipa> more accurate!
1751 2011-01-25 21:45:49 <zorg987> I seem to have found a disappearing transaction ... sent myself 50BTC, but they never show up. In depth investigation of the block chain seem to show everything's the way it should (transaction is in a blokc, block is confirmed, etc), but the UI stubbornly refuses to adjust my balance ?
1752 2011-01-25 21:46:11 <sipa> zorg987: well, your balance shouldn't change :)
1753 2011-01-25 21:46:12 <molecular> Artforz ADL lib ?
1754 2011-01-25 21:46:16 <Kiba> how many blocks do you have downloaded?
1755 2011-01-25 21:46:17 <ArtForz> yep
1756 2011-01-25 21:46:29 * molecular googling adl
1757 2011-01-25 21:46:39 <ArtForz> http://developer.amd.com/gpu/ADLSDK/Pages/default.aspx
1758 2011-01-25 21:46:58 <molecular> thanks
1759 2011-01-25 21:47:03 <zorg987> sipa: Ahah :) I meant: sent from another wallet on another machine.
1760 2011-01-25 21:47:05 <ArtForz> the zip contains html documentation
1761 2011-01-25 21:47:24 <sipa> zorg987: it really is another wallet, not a copied one?
1762 2011-01-25 21:47:24 <ArtForz> I originally wanted to implement a python wrapper for it, but got lazy
1763 2011-01-25 21:47:35 <molecular> /usr/lib/libatiadlxx.so ??
1764 2011-01-25 21:47:41 <ArtForz> yup
1765 2011-01-25 21:48:07 <ArtForz> using ADL you can get/set clocks, read temps, get/set fanspeed, ...
1766 2011-01-25 21:48:22 <zorg987> Kiba: I have all the blocks up to 104561, and the tx is old (appears in block 104518)
1767 2011-01-25 21:48:26 <ArtForz> prety much everything aticonfig does while X is running is done via ADL
1768 2011-01-25 21:48:38 <molecular> cool
1769 2011-01-25 21:48:47 <molecular> that C controller you implemented uses adl?
1770 2011-01-25 21:48:51 <ArtForz> yep
1771 2011-01-25 21:48:59 <molecular> can you paste that?
1772 2011-01-25 21:49:05 <Kiba> hmm
1773 2011-01-25 21:49:07 <Kiba> I don't know
1774 2011-01-25 21:50:01 <zorg987> sipa: yes
1775 2011-01-25 21:50:16 <zorg987> sipa: different machine, different wallet.
1776 2011-01-25 21:50:29 prax has joined
1777 2011-01-25 21:50:36 <ArtForz> sec
1778 2011-01-25 21:51:32 <zorg987> My question is: given the list of private keys in a wallet, shouldn't it be possible to scan the whole block chain and "replay" all the transactions in the wallet ?
1779 2011-01-25 21:51:34 <ArtForz> http://pastebin.com/STAF2VZ7
1780 2011-01-25 21:51:45 <sipa> zorg987: yes, 0.3.20 adds an option for that
1781 2011-01-25 21:52:01 <ArtForz> play with Kp/Ki/Kd to get a nice fan response curve
1782 2011-01-25 21:52:31 <zorg987> sipa: you mean the ReacceptWalletTransactions() method ?
1783 2011-01-25 21:52:37 <zorg987> sipa: or something else ?
1784 2011-01-25 21:52:42 <molecular> artforz: thanks, I'll check it out
1785 2011-01-25 21:53:00 <sipa> zorg987: --rescan
1786 2011-01-25 21:53:30 <gavinandresen> -rescan  (don't get me started on single-dash options versus double-dash.....)
1787 2011-01-25 21:53:39 <zorg987> sipa: oOooOo. sweet. let me go play with that and see what happens. thx for the tip.
1788 2011-01-25 21:53:48 <ArtForz> iirc 1/0.5/1 works decently for 5970s
1789 2011-01-25 21:54:08 gavinandresen has quit (Quit: gavinandresen)
1790 2011-01-25 21:54:09 <zorg987> gavin: is the "disappearing" tx problem a common one ?
1791 2011-01-25 21:54:11 <ArtForz> 5770s want a faster response as they have a lot less thermal mass
1792 2011-01-25 21:54:44 <ArtForz> I should probably make the loop params ... params
1793 2011-01-25 21:55:12 bertodsera has quit (Read error: Connection reset by peer)
1794 2011-01-25 21:55:47 <zorg987> sipa + gavin: and what do you know ... it worked !!!! I feel the urge to send bitcoins to someone (assuming I have an address :) )
1795 2011-01-25 21:55:53 <ArtForz> also implements a safety feature, if core temp is > targettemp + 5°C, it sets fanspeed to max
1796 2011-01-25 21:56:04 <andrew12> zorg987: send some to 1NkqZE4DA8h6Xbtz9GtJcfE1FcA8Z5yRJN :P
1797 2011-01-25 21:56:44 <prax> I accept bitcoins 12cP16TGxKcmBTRTChJYEXoR3VKZq52eyc
1798 2011-01-25 21:56:44 <ArtForz> response is a bit slow, so going idle -> load can cause it to overshoot target temp by quite a bit
1799 2011-01-25 21:56:45 <zorg987> andrew12: <grin>
1800 2011-01-25 21:56:59 <ArtForz> never bothered to really tune the params
1801 2011-01-25 21:57:06 <molecular> too bad, I have the adl lib, but no header for it. can I just use the .h from the adl_sdk with the lib from the stream sdk??
1802 2011-01-25 21:57:24 <ArtForz> erm, libatiadlxx is part of fglrx
1803 2011-01-25 21:57:31 <ArtForz> just use the headers from adl sdk
1804 2011-01-25 21:57:40 bertodsera has joined
1805 2011-01-25 21:58:23 noagendamarket has joined
1806 2011-01-25 21:58:27 <ArtForz> just compile and link -latiadlxx
1807 2011-01-25 21:58:47 noagendamarket has quit (Changing host)
1808 2011-01-25 21:58:48 noagendamarket has joined
1809 2011-01-25 21:59:03 <molecular> ok, fglrx is called "ati-drivers" on gentoo. but no adl header file comes with it, only glx.h, glxext.h, ...
1810 2011-01-25 21:59:08 <ArtForz> yes
1811 2011-01-25 21:59:13 <ArtForz> the .so comes with fglrx
1812 2011-01-25 21:59:18 <molecular> ok, I'll try with the headers from your link
1813 2011-01-25 21:59:19 <ArtForz> headers come with adl dsk
1814 2011-01-25 21:59:22 <molecular> allright
1815 2011-01-25 21:59:28 <ArtForz> *sdk
1816 2011-01-25 22:00:08 <ArtForz> the master/slave GPU detection for 5970 is hackish magic
1817 2011-01-25 22:00:20 <lyspooner> can i use blockexplorer to sort by highest transaction fee paid?
1818 2011-01-25 22:05:31 bertodsera has quit (Ping timeout: 255 seconds)
1819 2011-01-25 22:07:29 <zorg987> So here's another question for you guys: how do I consolidate all my public keys in a single wallet ?
1820 2011-01-25 22:11:35 <sipa> i believe there is no general wallet export/import function yet
1821 2011-01-25 22:11:51 gavinandresen has joined
1822 2011-01-25 22:12:02 <prax> what is it called, wallet.dat?
1823 2011-01-25 22:12:07 <sipa> yes
1824 2011-01-25 22:12:10 <prax> I need to know too now that I have BTC
1825 2011-01-25 22:12:22 <prax> should I toss that on a memory stick?
1826 2011-01-25 22:12:25 <sipa> make backups of that file
1827 2011-01-25 22:12:26 <sipa> yes
1828 2011-01-25 22:12:29 <prax> and possibly into a fireproof safe
1829 2011-01-25 22:12:33 Diablo-D3 has joined
1830 2011-01-25 22:12:38 <sipa> although know that it's not encrypted
1831 2011-01-25 22:12:55 <sipa> so anyone with access to that file, has access to the funds corresponding to the keys it contains
1832 2011-01-25 22:13:15 lyspooner has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
1833 2011-01-25 22:13:35 <prax> how do I tackle that then?
1834 2011-01-25 22:13:47 <sipa> encrypt the backup yourself, eg.
1835 2011-01-25 22:21:08 zorg987 has quit (Quit: Leaving)
1836 2011-01-25 22:24:30 brunner has quit (Ping timeout: 276 seconds)
1837 2011-01-25 22:28:23 tg has quit (Read error: Connection reset by peer)
1838 2011-01-25 22:28:50 tg has joined
1839 2011-01-25 22:35:02 fabianhjr has joined
1840 2011-01-25 22:35:28 <fabianhjr> Did anyone heard about the protests in Egypt turning into a revolution?
1841 2011-01-25 22:36:00 <Kiba> no
1842 2011-01-25 22:36:08 <Kiba> but did you read my email?
1843 2011-01-25 22:36:39 <Kiba> how much time do you have on Sunday, fabianhjr
1844 2011-01-25 22:36:57 <fabianhjr> Depends which sunday. Generally at least one hour. :/
1845 2011-01-25 22:38:34 <Kiba> we need more time to finish up and package the magazine...
1846 2011-01-25 22:39:01 <molecular> ArtForz: "undefined reference to `ADL_Main_Control_Create(void* (*)(int), int)", I think I got the wrong headers?
1847 2011-01-25 22:39:08 <Kiba> fabianhjr: PM me with your email address
1848 2011-01-25 22:39:10 <Kiba> err
1849 2011-01-25 22:39:13 <Kiba> bitcoin address
1850 2011-01-25 22:39:33 Zarutian has quit (Quit: Zarutian)
1851 2011-01-25 22:39:44 <andrew12> 1DonTwAnTYoURBiTCoInSsAvEtHEM
1852 2011-01-25 22:41:29 <ArtForz> weird
1853 2011-01-25 22:41:35 <Kiba> Teppy1: when can I expect a payment?
1854 2011-01-25 22:41:36 <ArtForz> is that a compiler or linker error
1855 2011-01-25 22:41:51 <molecular> linker. it's defined "extern"
1856 2011-01-25 22:41:55 <molecular> does that make sense?
1857 2011-01-25 22:41:59 <ArtForz> -latiadlxx
1858 2011-01-25 22:42:17 <molecular> yeah, have that
1859 2011-01-25 22:42:42 <ArtForz> thats... weird
1860 2011-01-25 22:42:53 <molecular> how can I check what symbols are in a lib?
1861 2011-01-25 22:43:29 <ArtForz> hrrrm
1862 2011-01-25 22:43:34 * Kiba goes whirling around
1863 2011-01-25 22:44:12 <lfm> molecular try nm command
1864 2011-01-25 22:44:53 doublec has joined
1865 2011-01-25 22:45:09 <ArtForz> works just fine here
1866 2011-01-25 22:45:23 davout has joined
1867 2011-01-25 22:45:29 <molecular> "nm -D /usr/lib/libatiadlxx.so | grep ADL_Main_Control" -> "0001d200 T ADL_Main_Control_Create"
1868 2011-01-25 22:45:29 <ArtForz> gcc -O2 -Wall -latiadlxx -o fancontrol fancontrol.c
1869 2011-01-25 22:45:29 slush_ has left ()
1870 2011-01-25 22:45:30 <molecular> hmmm
1871 2011-01-25 22:46:07 <Kiba> a new issue is out!
1872 2011-01-25 22:46:20 <Kiba> http://fabianhjr.co.cc/theBitcoinTimes/issue.pdf
1873 2011-01-25 22:46:22 <ArtForz> ldconfig should pick up stuff in /usr/lib
1874 2011-01-25 22:46:23 <Kiba> read it for free!
1875 2011-01-25 22:46:38 <fabianhjr> Kiba: :P wrong formatting.
1876 2011-01-25 22:46:39 <nanotube> 404 not found. great content! :)
1877 2011-01-25 22:46:47 <fabianhjr> issue00x.[pdf|odt]
1878 2011-01-25 22:46:56 <fabianhjr> or you can list the contents if you please.
1879 2011-01-25 22:47:00 <ArtForz> but it's a totally new 404!
1880 2011-01-25 22:47:09 <nanotube> heh
1881 2011-01-25 22:47:17 <molecular> ArtForz, hm, it said: "warning: libGL.so.1, needed by /usr/.../libatiadlxx.so, not found (try using -rpath or -rpath-link)", so I added -lGL
1882 2011-01-25 22:47:22 <fabianhjr> nanotube: http://fabianhjr.co.cc/theBitcoinTimes/issue001.odt
1883 2011-01-25 22:47:31 <fabianhjr> nanotube: http://fabianhjr.co.cc/theBitcoinTimes/issue001.pdf
1884 2011-01-25 22:47:34 <Kiba> http://fabianhjr.co.cc/theBitcoinTimes/issue001.pdf
1885 2011-01-25 22:47:48 <Kiba> fabianhjr: you made a big boo boo in your post
1886 2011-01-25 22:47:48 <ArtForz> weird, mine doesn't complain about that
1887 2011-01-25 22:48:01 <fabianhjr> ArtForz: you don't get a !404?
1888 2011-01-25 22:48:04 <Kiba> the link doesn't work
1889 2011-01-25 22:48:30 <lfm> there it is
1890 2011-01-25 22:48:31 <fabianhjr> It does, I am just reading it and redownload it to make sure. :/
1891 2011-01-25 22:51:31 <fabianhjr> Feedback is appreciated even if it is just curses. So, Please give it. :)
1892 2011-01-25 22:51:51 <Kiba> you'll find that 3/4 of the content is written by me :D
1893 2011-01-25 22:52:46 * davout is reading
1894 2011-01-25 22:52:59 <Kiba> if you like it, send me a bitcoin! :D
1895 2011-01-25 22:53:33 asdf30 has quit (Ping timeout: 260 seconds)
1896 2011-01-25 22:54:05 <fabianhjr> No, send it to me! Support Mexico and Siesta! :D
1897 2011-01-25 22:55:08 riX2000 has quit (Quit: Leaving)
1898 2011-01-25 22:55:08 fabianhjr has quit (Read error: Connection reset by peer)
1899 2011-01-25 22:55:14 fabianhjr_ has joined
1900 2011-01-25 22:55:17 fabianhjr_ is now known as fabianhjr
1901 2011-01-25 22:55:23 <davout> i'll just send it to myself, support red wine and beautiful (but annoying) women
1902 2011-01-25 22:55:23 <fabianhjr> My apologies. xD
1903 2011-01-25 22:55:28 <davout> :D
1904 2011-01-25 22:55:35 <Kiba> I have like 499.22 BTC
1905 2011-01-25 22:55:43 <davout> Kiba: really?
1906 2011-01-25 22:55:45 <fabianhjr> davout: Siesta is more important than anything else. <3
1907 2011-01-25 22:55:51 <Kiba> davout: yeah
1908 2011-01-25 22:55:51 CyanDynamo has quit (Quit: Leaving.)
1909 2011-01-25 22:55:52 <davout> Kiba: address?
1910 2011-01-25 22:55:56 <Kiba> no
1911 2011-01-25 22:56:11 <Kiba> didn't you earn a boatload of cash from jgarizik?
1912 2011-01-25 22:56:14 <Kiba> you're richer than me!
1913 2011-01-25 22:56:14 <davout> i would've sent you the .78 missing :)
1914 2011-01-25 22:56:32 <davout> yes i am, but i spent more time coding than you did :3
1915 2011-01-25 22:56:43 <Kiba> 14Ghffp7ubgdYkPWEoRSEfjcPm4V5Fc5Cp
1916 2011-01-25 22:56:55 <fabianhjr> :P
1917 2011-01-25 22:57:05 <Kiba> I will get 83 more BTC this evening if I am lucky
1918 2011-01-25 22:57:14 <Kiba> plus an additional 25 BTC once teppy pay me
1919 2011-01-25 22:57:43 <fabianhjr> 1AMdyQwhswHknQfNGhSS7f1tqDv89Vpib2 <- Send all your bitcoins there for an _opportunity_ to get twice back!
1920 2011-01-25 22:57:58 <davout> with the flip a coin thing ?
1921 2011-01-25 22:58:18 <davout> i like this issue of the btc times, haven't finished yet tho
1922 2011-01-25 22:58:24 <fabianhjr> Yeah, except I bet all and return you the half. :P
1923 2011-01-25 22:58:45 <fabianhjr> davout: really appreciated. <3
1924 2011-01-25 22:58:53 <davout> like what if i send 10.50000001 BTC
1925 2011-01-25 22:58:58 <davout> and win ? :)
1926 2011-01-25 22:58:58 <donpdonp> its not even a fair toss, 49% win
1927 2011-01-25 22:59:13 bertodsera has joined
1928 2011-01-25 22:59:44 <fabianhjr> donpdonp: you get a ~95% return. Half the time you get the double.
1929 2011-01-25 22:59:55 <davout> KIba: i like round numbers much better
1930 2011-01-25 23:00:10 <donpdonp> fabianhjr: it should be half the time, but its 49% of the time
1931 2011-01-25 23:00:32 <donpdonp> also they dont say how they generate the random numbers
1932 2011-01-25 23:00:33 <edcba> we only get an opportunity or do you actually send twice back sometimes ?
1933 2011-01-25 23:00:33 * davout is getting very annoyed at his own fucking captchas
1934 2011-01-25 23:00:37 <nanotube> yep, casinos don't build their fancy buildings out of fair returns. :)
1935 2011-01-25 23:00:49 <nanotube> davout: nix the captchas
1936 2011-01-25 23:00:55 <davout> nix?
1937 2011-01-25 23:01:01 <nanotube> take out, kill, etc.
1938 2011-01-25 23:01:03 <nanotube> ;;ud nix
1939 2011-01-25 23:01:03 <gribble> http://www.urbandictionary.com/define.php?term=nix | To keep nix is an old english boarding school term for keeping a lookout. It is thought to derive from the word that was shouted by said ...
1940 2011-01-25 23:01:04 <davout> oh
1941 2011-01-25 23:01:04 <fabianhjr> 49 * 2 98% return. What are you complaining about? Oh yeah, he only gives 198% when you win. :P
1942 2011-01-25 23:01:12 <nanotube> mmm not that definitiol heh
1943 2011-01-25 23:01:15 <nanotube> n
1944 2011-01-25 23:01:18 <davout> yea their removal has already been commited
1945 2011-01-25 23:01:32 <edcba> how much can i send ?
1946 2011-01-25 23:01:34 <davout> not deployed since i'm shoving lots of refactoring in
1947 2011-01-25 23:01:36 <donpdonp> nanotube: yup. though im still trying to figure out of its a profitable venture to have real 50/50 payout and 198% return instead of 200%
1948 2011-01-25 23:01:50 <edcba> what are min/max ?
1949 2011-01-25 23:02:00 <davout> whats that technique to beat those games again?
1950 2011-01-25 23:02:01 <davout> like
1951 2011-01-25 23:02:06 <davout> bet X
1952 2011-01-25 23:02:09 <davout> if win stop
1953 2011-01-25 23:02:14 <edcba> bet x if lose bet 2*x
1954 2011-01-25 23:02:15 <davout> if fail bet 2*X
1955 2011-01-25 23:02:21 satamusic has quit (Ping timeout: 265 seconds)
1956 2011-01-25 23:02:24 <edcba> start with x small :)
1957 2011-01-25 23:02:25 <nanotube> right. a good way to run out of money :)
1958 2011-01-25 23:02:34 <davout> isn't that game vulnerable to this flaw?
1959 2011-01-25 23:02:40 <edcba> of course
1960 2011-01-25 23:02:47 <edcba> depends on min/max accepted
1961 2011-01-25 23:02:51 <molecular> ArtForz: idea: the symbols I'm missing are the ones declared extern in your code (line 8 thru 18). except one (ADL_Main_Control_Destroy). All others have some int arguments. sume sizeof(int) problem?
1962 2011-01-25 23:02:55 <molecular> s/sume/some
1963 2011-01-25 23:03:03 satamusic has joined
1964 2011-01-25 23:03:03 <Kiba> fabianhjr: the link on the forum is STILL broken
1965 2011-01-25 23:03:07 <nanotube> donpdonp: should be the same, whether you adjust probability down, or adjust payout down.
1966 2011-01-25 23:03:16 <davout> well if you start out with 1 and the max is 100 you're pretty unlikely to lose?
1967 2011-01-25 23:03:28 <Kiba> how are our audience going to be able to read the issue?
1968 2011-01-25 23:03:31 <ArtForz> shouldnt be
1969 2011-01-25 23:03:34 <lfm> davout if you have more money than the bank you might be able to do that
1970 2011-01-25 23:03:40 <ArtForz> compiles fine in 32 and 64 bit mode here
1971 2011-01-25 23:04:12 <lfm> davout ya you dont lose often but you lose big when you do lose
1972 2011-01-25 23:04:13 <donpdonp> nanotube: k. the site used to say 198% payback but switched their wording to 49% chance. it got me interested in random.org's TrueRNG
1973 2011-01-25 23:04:14 <molecular> is it possible I have a 64 bit lib?
1974 2011-01-25 23:04:22 <fabianhjr> Kiba: woops. My mistake. Fixed now. :P
1975 2011-01-25 23:04:36 <molecular> how to tell gcc to compile for 64 bit
1976 2011-01-25 23:04:39 <davout> jfm: that doesn't make sense, that would mean that at some point you'd be able to win, but the bank wouldn't be able to pay
1977 2011-01-25 23:04:43 <Kiba> so the editor will start work tomorrow
1978 2011-01-25 23:04:43 <davout> *lfm
1979 2011-01-25 23:04:45 <Kiba> hmm
1980 2011-01-25 23:05:03 satamusic___ has joined
1981 2011-01-25 23:05:08 * davout is looking up this particular exploit :)
1982 2011-01-25 23:05:21 <lfm> davout if you have bad loing streak and hit 100 limit you have lost a lot
1983 2011-01-25 23:06:04 satamusic___ has quit (Read error: Connection reset by peer)
1984 2011-01-25 23:06:55 satamusic___ has joined
1985 2011-01-25 23:06:57 <nanotube> donpdonp: davout http://en.wikipedia.org/wiki/Martingale_%28betting_system%29
1986 2011-01-25 23:07:15 <davout> start with one
1987 2011-01-25 23:07:58 <davout> nanotube: thank you!
1988 2011-01-25 23:08:08 * Kiba ponders where he can find more work
1989 2011-01-25 23:08:12 <nanotube> davout: so you have 7 doublings until you get to 128... about a 1% chance to lose it all.
1990 2011-01-25 23:08:12 <Kiba> jobs!
1991 2011-01-25 23:08:22 <nanotube> davout: np :) enjoy the read. :)
1992 2011-01-25 23:08:24 <lfm> davout ya start with 1, if you win start with 1 again, if you lose double your bet, after each win you will be 1 ahead. unless you hit the 100 limit then you can lose big and you cant double any more
1993 2011-01-25 23:08:29 satamusic_ has quit (Ping timeout: 246 seconds)
1994 2011-01-25 23:08:51 <dirtyfilthy> hahaha martingale
1995 2011-01-25 23:08:55 satamusic has quit (Read error: Connection reset by peer)
1996 2011-01-25 23:09:37 <dirtyfilthy> what could possible go wrong?
1997 2011-01-25 23:09:41 satamusic___ is now known as satamusic
1998 2011-01-25 23:10:29 <fabianhjr> Hey, the martingale article is brilliant. I hadn't tough of that. :P The question is, can anyone prove it by trying to beat ProbiWon?
1999 2011-01-25 23:10:51 <cosurgi> ;;seen cosurgi
2000 2011-01-25 23:10:51 <gribble> cosurgi was last seen in #bitcoin-dev 3 hours, 43 minutes, and 45 seconds ago: <cosurgi> does lspci list PCI:0@19:0:1 and is there a match for it in xorg.conf ?
2001 2011-01-25 23:10:58 <cosurgi> ;;seen cosurgi
2002 2011-01-25 23:10:58 <gribble> cosurgi was last seen in #bitcoin-dev 6 seconds ago: <cosurgi> ;;seen cosurgi
2003 2011-01-25 23:11:01 <cosurgi> lol
2004 2011-01-25 23:11:12 <cosurgi> goodnight
2005 2011-01-25 23:11:26 <davout> fabianhjr: if he's done things right he won't let you place bets that get him broke
2006 2011-01-25 23:11:37 <dirtyfilthy> "The martingale strategy fails even with unbounded stopping time, as long as there is a limit on earnings or on the bets (which are also true in practice).[2] It is only with unbounded wealth, bets and time that the martingale strategy can succeed."
2007 2011-01-25 23:11:51 <lfm> so gribble reponse time is 6 sec, thats not too good
2008 2011-01-25 23:12:00 <fabianhjr> davout: he has a 15 BTC bet limit, however, Imagine if someone bet 15 and got a 3500% back!? xD
2009 2011-01-25 23:12:33 <dirtyfilthy> i remember thinking i was going to be rich when i figured out martingale by myself when i was 16 or so... sadly, no
2010 2011-01-25 23:12:34 <fabianhjr> Well, that is not a really restrictive number now that I think of it. :/
2011 2011-01-25 23:12:44 <donpdonp> nanotube: excellent, thanks for the martingale link
2012 2011-01-25 23:12:56 <Kiba> soo...how's the magazine?
2013 2011-01-25 23:12:59 <fabianhjr> dirtyfilthy: you are obvusly not signing books about it and lying about the riches and losses you took. :p
2014 2011-01-25 23:13:31 <lfm> ya, you CAN make money on it, write a book
2015 2011-01-25 23:13:45 <fabianhjr> See! I told you so! :P
2016 2011-01-25 23:14:08 RichardG has quit (Ping timeout: 264 seconds)
2017 2011-01-25 23:14:15 <fabianhjr> Just make it with one of those "For Newbs" covers.
2018 2011-01-25 23:14:32 <lfm> ya gambling for dummies
2019 2011-01-25 23:14:37 <dirtyfilthy> SURE FIRE TIPS FOR UNLIMITED WEALTH
2020 2011-01-25 23:14:45 <fabianhjr> OMG, My GFX blocked right after I upgraded MS Security Essentials. Damn MicroShaft. D:<
2021 2011-01-25 23:14:51 <fabianhjr> Going to reboot.
2022 2011-01-25 23:14:59 fabianhjr has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
2023 2011-01-25 23:15:04 <Diablo-D3> ArtForz: I have found a problem
2024 2011-01-25 23:15:08 <Diablo-D3> this ew kernel is now so efficient
2025 2011-01-25 23:15:15 <Diablo-D3> its raping my 4xxx
2026 2011-01-25 23:15:50 <nanotube> ;;latency
2027 2011-01-25 23:15:51 <gribble> 0.06 seconds.
2028 2011-01-25 23:15:54 <nanotube> ;;ping
2029 2011-01-25 23:15:54 <gribble> pong
2030 2011-01-25 23:16:00 <nanotube> lfm: --^ seems plenty fast to me. :)
2031 2011-01-25 23:16:12 <lfm> ;;seen lfm
2032 2011-01-25 23:16:12 <gribble> lfm was last seen in #bitcoin-dev 1 minute and 39 seconds ago: <lfm> ya gambling for dummies
2033 2011-01-25 23:16:25 <nanotube> lfm: it doesn't include your requesting msg.
2034 2011-01-25 23:16:26 <lfm> never mind
2035 2011-01-25 23:16:34 <nanotube> :)
2036 2011-01-25 23:19:54 <davout> http://en.wikipedia.org/wiki/St._Petersburg_paradox
2037 2011-01-25 23:20:08 * davout is trapped in wikipedia
2038 2011-01-25 23:21:05 <dirtyfilthy> as long as you don't go to tvtropes you'll be fine
2039 2011-01-25 23:21:55 asdf30 has joined
2040 2011-01-25 23:22:18 kermit has quit (Quit: Leaving.)
2041 2011-01-25 23:22:54 devon_hillard has quit (Ping timeout: 240 seconds)
2042 2011-01-25 23:25:07 devon_hillard has joined
2043 2011-01-25 23:25:07 devon_hillard has quit (Changing host)
2044 2011-01-25 23:25:07 devon_hillard has joined
2045 2011-01-25 23:26:50 sabalaba has joined
2046 2011-01-25 23:27:29 <luke-jr> :|
2047 2011-01-25 23:27:35 <luke-jr> why is nobody else bidding on this?
2048 2011-01-25 23:28:13 <luke-jr> Sapphire Radeon Toxic HD 5850 1 GB DDR5 Dual
2049 2011-01-25 23:29:04 <lfm> theyre looking to snipe it?
2050 2011-01-25 23:29:38 <luke-jr> but still, it's been 24 hours and no competing bids
2051 2011-01-25 23:29:48 <luke-jr> LOL
2052 2011-01-25 23:29:51 <luke-jr> they sniped it
2053 2011-01-25 23:29:53 <luke-jr> XD
2054 2011-01-25 23:29:53 asdf58 has joined
2055 2011-01-25 23:30:23 dukeleto has quit (Excess Flood)
2056 2011-01-25 23:30:54 asdf30 has quit (Remote host closed the connection)
2057 2011-01-25 23:30:57 dukeleto has joined
2058 2011-01-25 23:32:18 <luke-jr> Gigabyte ATI Radeon HD5850 1 GB DDR5 PCI-E Video Card
2059 2011-01-25 23:32:23 <luke-jr> let's see if they snipe this one too :x
2060 2011-01-25 23:33:20 molecular has quit (Ping timeout: 264 seconds)
2061 2011-01-25 23:33:31 molecular has joined
2062 2011-01-25 23:40:44 larsivi has quit (Remote host closed the connection)
2063 2011-01-25 23:42:10 <luke-jr> meh, there's always more auctions
2064 2011-01-25 23:46:11 fabianhjr has joined
2065 2011-01-25 23:47:16 Kiba` has joined
2066 2011-01-25 23:48:13 Kiba has quit (Ping timeout: 250 seconds)
2067 2011-01-25 23:48:56 <fabianhjr> Ok, it is fixed now. :)
2068 2011-01-25 23:51:25 riush has quit (Remote host closed the connection)
2069 2011-01-25 23:57:59 <Diablo-D3> hey ArtForz, whats the theoretical maximum on my 4850 again? 78 something?
2070 2011-01-25 23:59:08 <Diablo-D3> because that'd mean I'm about 2% or so from max efficiency