1 2011-03-01 00:00:22 <Lachesis> ok jgarzik, i got your pool set up
  2 2011-03-01 00:00:26 <Lachesis> now how do i interpret this db?
  3 2011-03-01 00:01:55 <theymos> Why is this transaction not in a block yet? Seems to me like a bug. http://www.bitcoin.org/smf/index.php?topic=3966.msg57098#msg57098
  4 2011-03-01 00:02:35 <ArtForz> hmmm.... too big?
  5 2011-03-01 00:03:34 <theymos> The dump is above that post. Doesn't look huge.
  6 2011-03-01 00:03:52 <ArtForz> hrrrm
  7 2011-03-01 00:04:24 <theymos> Maybe its priority was too low and it triggered the free limit at 4kB blocksize. But it should have gotten in a later block.
  8 2011-03-01 00:05:13 <ArtForz> I have that tx queued
  9 2011-03-01 00:05:33 <ArtForz> so it's there, but other stuff gets priority
 10 2011-03-01 00:05:44 <Blitzboom> lol, one transaction got through now
 11 2011-03-01 00:05:47 <jgarzik> Lachesis: that's an open ended question :)
 12 2011-03-01 00:05:51 <Blitzboom> after seven hours
 13 2011-03-01 00:05:56 <amiller> ArtForz, would you publicly display your queue?
 14 2011-03-01 00:05:57 <jgarzik> Lachesis: the mining pool thread includes some sample SQL
 15 2011-03-01 00:06:12 <Lachesis> jgarzik, alright - keeping some cards to yourself? :)
 16 2011-03-01 00:06:14 <theymos> ArtForz: OK. I'll post that in the thread. Thanks.
 17 2011-03-01 00:06:26 <jgarzik> Lachesis: yep
 18 2011-03-01 00:06:31 <ArtForz> question is, why the fuck is it getting pushed back
 19 2011-03-01 00:06:37 <jgarzik> Lachesis: you're welcome to post your insights to that thread...
 20 2011-03-01 00:06:42 <ArtForz> it doesnt look like it should have a that bad priority score
 21 2011-03-01 00:06:49 <Lachesis> jgarzik, thanks for the pool code
 22 2011-03-01 00:07:11 <Lachesis> it's a bit rough, but way better than i could have done from scratch
 23 2011-03-01 00:07:36 <jgarzik> Lachesis: it's a lot rough.  doesn't include some anti-cheating code, for example...
 24 2011-03-01 00:12:07 <tcatm> luke-jr: where's the subcent thread?
 25 2011-03-01 00:12:25 <luke-jr> tcatm: which one?
 26 2011-03-01 00:12:30 <luke-jr> the bugfix never had a thread
 27 2011-03-01 00:12:47 <luke-jr> http://www.bitcoin.org/smf/index.php?topic=3786.0 is the latest for precision
 28 2011-03-01 00:12:55 <tcatm> well it has a pull request now https://github.com/bitcoin/bitcoin/pull/85
 29 2011-03-01 00:14:36 <luke-jr> tcatm: thx
 30 2011-03-01 00:18:21 <phantomcircuit> http://codepad.org/xfNhNhkm
 31 2011-03-01 00:18:25 <phantomcircuit> that's from the official client
 32 2011-03-01 00:18:31 <phantomcircuit> and would appear to be a very wrong checksum
 33 2011-03-01 00:18:48 <ArtForz> that tx should have a score of 20567855
 34 2011-03-01 00:19:36 <ArtForz> so it falls short of the dPriority > COIN * 144 / 250 test
 35 2011-03-01 00:19:54 <phantomcircuit> am i just screwing up the checksum calculation very wrong or is there a regression in the mainline client here?
 36 2011-03-01 00:19:55 <ArtForz> and as tx size is > 4000, fAllowFree is never true for it
 37 2011-03-01 00:20:19 <theymos> Ah. So it will never confirm?
 38 2011-03-01 00:21:16 <ArtForz> probably not
 39 2011-03-01 00:21:46 <phantomcircuit> seems like the client shouldn't let you make transactions which will never go through :|
 40 2011-03-01 00:21:51 <ArtForz> yeah, why the fuck is the client even sending those
 41 2011-03-01 00:21:59 <ArtForz> s/sending/creating
 42 2011-03-01 00:22:16 <dirtyfilthy> phantomcircuit: what are you checksumming? the entire packet or just the payload?
 43 2011-03-01 00:23:50 <phantomcircuit> dirtyfilthy, just the payload
 44 2011-03-01 00:23:58 <phantomcircuit> you cant calculate a checksum for the entire packet
 45 2011-03-01 00:24:02 <phantomcircuit> it contains the checksum
 46 2011-03-01 00:24:18 <molecular> will the score of that transaction rise with it's age?
 47 2011-03-01 00:24:26 <ArtForz> hurrmmm... doesnt look like a recent client would create that one
 48 2011-03-01 00:24:55 <amiller> this is a really good discussion on escrow http://www.bitcoin.org/smf/index.php?topic=750.0
 49 2011-03-01 00:25:04 <ArtForz> I think I see the problem
 50 2011-03-01 00:25:15 <amiller> i'm interested right now in how piggybacking small transactions on large transactions into escrow would give them higher priority
 51 2011-03-01 00:25:20 <ArtForz> it will *eventually* go through
 52 2011-03-01 00:25:35 <ArtForz> but it'll take quite a while
 53 2011-03-01 00:26:08 <luke-jr> ?
 54 2011-03-01 00:27:27 <ArtForz> about... 190 more blocks
 55 2011-03-01 00:28:11 <theymos> I will post that in the tread. Might that be causing the other problems with confirmations?
 56 2011-03-01 00:28:22 <ArtForz> at that point its score will be > 57600000 = it passes the dPriority > COIN * 144 / 250 check
 57 2011-03-01 00:28:32 <dirtyfilthy> phantomcircuit: why is there a length of 27 and a payload length of 23 in your example?
 58 2011-03-01 00:28:55 <phantomcircuit> dirtyfilthy, the official client appears to include the length of the checksum in the length
 59 2011-03-01 00:29:11 <ArtForz> looks like the tx-sending code needs some fixing
 60 2011-03-01 00:30:30 <molecular> like not generate a tx with a score too low, or suggest to add a fee?
 61 2011-03-01 00:30:36 <ArtForz> yep
 62 2011-03-01 00:30:43 <molecular> sounds sane
 63 2011-03-01 00:30:53 <dirtyfilthy> phantomcircuit: i don't think that's the case dude
 64 2011-03-01 00:31:06 <ArtForz> check size and priority of transaction, if it's > 4kB and score < 57600000 require min fee
 65 2011-03-01 00:31:37 <phantomcircuit> dirtyfilthy, well that's a network dump of what im getting from the mainline client
 66 2011-03-01 00:32:32 <molecular> ArtForz, should I (or someone else) write up an issue on github or does this need more discussion?
 67 2011-03-01 00:32:36 <dirtyfilthy> phantomcircuit: addresses are 26 bytes long.... plus 1 byte for variable integer length... an address wouldn't fit in 23 bytes
 68 2011-03-01 00:33:37 <phantomcircuit> dirtyfilthy, lol yes im aware
 69 2011-03-01 00:35:28 <ArtForz> maybe reduce the size limit to 3.5k or so, otherwise it can only get into an otherwise empty block
 70 2011-03-01 00:36:05 <theymos> You might guess the size based on the last few blocks.
 71 2011-03-01 00:38:05 <phantomcircuit> dirtyfilthy, none the less it's what im getting
 72 2011-03-01 00:41:37 <molecular> well, even if the tx-generation code is fixed. doesn't this still open a possibility for attack?
 73 2011-03-01 00:42:07 <ArtForz> err... how?
 74 2011-03-01 00:42:30 <molecular> generate lots of low-prio transactions > 4k, no fee and have everyones ram used up?
 75 2011-03-01 00:42:37 <ArtForz> yes
 76 2011-03-01 00:42:58 <theymos> Are memory pool transactions stored in memory? I thought I heard somewhere that they were stored on disk.
 77 2011-03-01 00:43:07 <ArtForz> nope, in memory
 78 2011-03-01 00:43:19 <theymos> That is a pretty big attack, then.
 79 2011-03-01 00:43:22 <ArtForz> and I've been saying that a infinite tx cache is a bad idea for... quite a while now
 80 2011-03-01 00:45:52 <molecular> how could it be made finite?
 81 2011-03-01 00:47:29 <ArtForz> well, my node starts dropping tx lowest-prio first when tx cache gets > 5MB
 82 2011-03-01 00:47:52 <theymos> That seems smart. Honest nodes will rebroadcast, anyway.
 83 2011-03-01 00:47:54 <doublec> How customized is your node?
 84 2011-03-01 00:48:01 <doublec> it sounds like it has quite a few tweaks
 85 2011-03-01 00:48:12 <ArtForz> yeah, it's modified quite a bit
 86 2011-03-01 00:49:09 <molecular> well, if every node does this -> bad, no?
 87 2011-03-01 00:49:14 <ArtForz> not really
 88 2011-03-01 00:49:24 <ArtForz> nodes lose their tx cache on restart
 89 2011-03-01 00:50:15 <molecular> that's different, not all nodes are restarted at one point in time
 90 2011-03-01 00:50:47 <ArtForz> and not all nodes will have the same cached transactions
 91 2011-03-01 00:50:56 <molecular> hmm
 92 2011-03-01 00:51:37 <molecular> this must've been discussed before? is there a thread in forum?
 93 2011-03-01 00:52:03 <ArtForz> not sure, probably yes
 94 2011-03-01 00:52:10 <KuT-Sickness> I was wondering, I'm using poclbm so I can use my ati for calculation
 95 2011-03-01 00:52:21 <KuT-Sickness> but it uses up 100% of my gpu resources by default
 96 2011-03-01 00:52:28 <KuT-Sickness> is there any way to limit this to lets say 40%?
 97 2011-03-01 00:52:28 <molecular> but currently, this might be a doable attack on the network, right?
 98 2011-03-01 00:52:33 <ArtForz> yes
 99 2011-03-01 00:52:43 <ArtForz> I think I remember discussing it with gavin
100 2011-03-01 00:53:24 <ArtForz> back then consensus was something like "yeah, it's a possible DoS, we'll have to fix that somehow someday"
101 2011-03-01 00:53:28 <doublec> KuT-Sickness: I think tweaking the -f switch helps with how much resource it uses
102 2011-03-01 00:53:57 <molecular> if it's as simple as limiting the transaction queue...
103 2011-03-01 00:54:18 <ArtForz> well, iirc it's not quite as simple
104 2011-03-01 00:54:36 <sethsethseth> ya i put -f600 on one of my cores and it makes it like 85% usage
105 2011-03-01 00:54:43 <ArtForz> problem is, nodes pass on tx they dont have in queue already to their peers
106 2011-03-01 00:55:04 <KuT-Sickness> -f600?
107 2011-03-01 00:55:11 <ArtForz> so what happens when someone sends enough tx to drop his oldest off the queue...
108 2011-03-01 00:55:15 <KuT-Sickness> I thought the default value when not using the -f switch was 30
109 2011-03-01 00:55:24 <sethsethseth> right
110 2011-03-01 00:55:27 <Blitzboom> i use 64
111 2011-03-01 00:55:32 <Blitzboom> no lag
112 2011-03-01 00:55:35 <KuT-Sickness> so mine should be at 30
113 2011-03-01 00:55:42 <KuT-Sickness> going at 68.1Mhash/s
114 2011-03-01 00:55:44 <KuT-Sickness> thats normal?
115 2011-03-01 00:55:48 <Blitzboom> what gpu?
116 2011-03-01 00:55:51 <KuT-Sickness> hd4850
117 2011-03-01 00:55:52 <ArtForz> = you turned a memory consumption DoS into a massively multiplied network DoS
118 2011-03-01 00:56:07 <Blitzboom> https://en.bitcoin.it/wiki/Mining_hardware_comparison
119 2011-03-01 00:56:11 <KuT-Sickness> ty
120 2011-03-01 00:56:18 <ArtForz> so clients would still have to keep track of dropped TX hashes at least for a while
121 2011-03-01 00:56:20 <molecular> ArtForz, the transaction queue moves from node RAM to being passed around the network all the time
122 2011-03-01 00:56:28 <molecular> traffic increase
123 2011-03-01 00:57:03 <ArtForz> hrrrm... actually it's not worse than what we currently have
124 2011-03-01 00:57:11 <doublec> storing the tx's on disk rather than ram would help wouldn't it?
125 2011-03-01 00:57:47 <ArtForz> well, that wont really help if tx are accumulating faster than they can get into blocks...
126 2011-03-01 00:57:53 <molecular> what keeps the attacker from filling up the disks as well?
127 2011-03-01 00:58:20 <thecoreh> Hey, is this the place to ask about bitcoin exchange?
128 2011-03-01 00:58:40 <thecoreh> Or only about the development of the client?
129 2011-03-01 00:58:59 <doublec> thecoreh: exchanging bitcoins might be better in #bitcoin-otc
130 2011-03-01 00:59:10 <doublec> thecoreh: lots of people there ready to buy/sell
131 2011-03-01 00:59:10 <thecoreh> doublec: Thanks :)
132 2011-03-01 00:59:25 <ArtForz> we have the same possible network DoS currently - just send a shitload of different TX
133 2011-03-01 01:00:05 <molecular> That could even be done with a single node on crappy connection
134 2011-03-01 01:00:12 <ArtForz> a limited tx cache would stop it from also acting as a mem consumption DoS
135 2011-03-01 01:00:36 <molecular> that network DoS in discussion/awareness?
136 2011-03-01 01:00:47 <ArtForz> theres no good way to fix it
137 2011-03-01 01:00:58 <KuT-Sickness> doublec, any hints as to what to set the -f switch to?
138 2011-03-01 01:01:05 <KuT-Sickness> I've tried several values between 5 and 60
139 2011-03-01 01:01:09 <ArtForz> except ratelimit transactions, which iirc we already do
140 2011-03-01 01:01:11 <doublec> KuT-Sickness: try 1000
141 2011-03-01 01:01:15 <KuT-Sickness> nothing seems to stop the 99% gpu usage
142 2011-03-01 01:01:18 <doublec> KuT-Sickness: and work up/down from there
143 2011-03-01 01:01:36 <KuT-Sickness> k
144 2011-03-01 01:01:38 <molecular> the mem-attack would bring down the net quite quickly, though, right?
145 2011-03-01 01:01:50 <ArtForz> well... not too quickly
146 2011-03-01 01:02:14 <molecular> because of the rate-limit?
147 2011-03-01 01:02:18 <ArtForz> yep
148 2011-03-01 01:02:35 <ArtForz> can't find it right now
149 2011-03-01 01:02:37 <molecular> is that per-connection?
150 2011-03-01 01:02:55 <ArtForz> iirc we limit to 5 tx/peer/sec or so
151 2011-03-01 01:03:09 <molecular> that's already a lot
152 2011-03-01 01:03:15 <ArtForz> yep
153 2011-03-01 01:03:16 <molecular> I could make them big
154 2011-03-01 01:03:43 <molecular> how big could I make a transaction?
155 2011-03-01 01:04:02 <ArtForz> well, not too big, or you'll hit the GetMinFee limit
156 2011-03-01 01:04:22 <molecular> still, I can use many connections
157 2011-03-01 01:04:25 <ArtForz> looks like 9kB
158 2011-03-01 01:04:41 <ArtForz> beyond that nodes will simply ignore it
159 2011-03-01 01:04:45 <Lachesis> that's 158MB/hr/peer
160 2011-03-01 01:04:59 <Lachesis> would need several connections before it was a problem
161 2011-03-01 01:06:19 <molecular> why? *100 connections = 15GB
162 2011-03-01 01:06:24 <molecular> that'll take out most nodes
163 2011-03-01 01:08:16 <ArtForz> except most nodes dont have 100 connections, they have 8
164 2011-03-01 01:08:44 <ArtForz> so only useful for taking a single node out
165 2011-03-01 01:09:21 <molecular> well, with 8 connections, thats still 1.2GB/hour
166 2011-03-01 01:09:27 <ArtForz> yep
167 2011-03-01 01:09:58 <molecular> that's already trouble for some nodes. after 4 hours, shit starts hitting fan, no?
168 2011-03-01 01:10:04 <ArtForz> yep
169 2011-03-01 01:10:28 <molecular> as I see it now, we _must_ limit the tx cache
170 2011-03-01 01:10:34 <ArtForz> imo it's a practical attack to pretty much DoS the whole network
171 2011-03-01 01:11:00 <ArtForz> but it seems nowadays "nobody did it yet" is a good enough excuse to ignore stuff like that ...
172 2011-03-01 01:11:00 <molecular> well, nodes will restart and take another 2 hours
173 2011-03-01 01:11:06 <molecular> it might not completely down the net
174 2011-03-01 01:11:17 <molecular> uhm, we could try on testnet
175 2011-03-01 01:12:30 <KuT-Sickness> hmz guess I'm dropping bitcoin early, not worth it having my gpu @ 97% load all the time
176 2011-03-01 01:13:09 <TheKid> KuT-Sickness: good, difficulty will go down for the rest of us ;3
177 2011-03-01 01:13:27 <ArtForz> yay! more for me! err... I mean, well, sounds like a reasonable decision.
178 2011-03-01 01:13:35 <KuT-Sickness> lol
179 2011-03-01 01:14:09 <KuT-Sickness> pretty sure the hw will die early if you have it on such loads all the time, probably not worth it
180 2011-03-01 01:14:49 <lfm> KuT-Sickness: were you checking temps?
181 2011-03-01 01:15:02 <molecular> KuT-Sickness, yes, I fried a gpu once, by playing wolfenstein for 2 nights, it just blew, man. so dont do it
182 2011-03-01 01:15:14 <molecular> nobody here does it, either
183 2011-03-01 01:15:15 <KuT-Sickness> yes lfm
184 2011-03-01 01:15:41 <ArtForz> so far I didnt manage to kill a single GPU by mining alone
185 2011-03-01 01:15:48 <molecular> I was joking
186 2011-03-01 01:15:53 <KuT-Sickness> ofcourse
187 2011-03-01 01:16:03 <molecular> if you have stock Voltage and no overclocking, you're pretty much safe
188 2011-03-01 01:16:05 <ArtForz> thoguh it'S not too hard with overvolting + heavy OC ;)
189 2011-03-01 01:16:05 <KuT-Sickness> nobody spent more than 2 hours on that fail wolfenstein remake
190 2011-03-01 01:16:07 <omglolbbq> what do i do if my transactions arent being verified?
191 2011-03-01 01:16:27 <molecular> you _could_ replace the stupid fan-controller that lets things to up to 92 ???C on stock 5970
192 2011-03-01 01:16:43 <lfm> omg wait 5 days
193 2011-03-01 01:16:52 <ArtForz> omglolbbq: wait a few days or start paying fees
194 2011-03-01 01:16:52 <KuT-Sickness> tbh it didnt go above 85 here
195 2011-03-01 01:17:05 <molecular> omglolbbq, there's currently problems with transactions queueing up
196 2011-03-01 01:17:05 <omglolbbq> is that new?
197 2011-03-01 01:17:19 <lfm> omglolbbq: could be called a bug
198 2011-03-01 01:17:20 <molecular> yes, it's new as far as I know
199 2011-03-01 01:17:23 <ArtForz> we currently have a backlog of TX queuing up
200 2011-03-01 01:17:28 <ArtForz> 669 now
201 2011-03-01 01:17:41 <omglolbbq> is that meant to happen or? xD
202 2011-03-01 01:17:42 <molecular> wow, rising quite quickly
203 2011-03-01 01:17:52 <molecular> the problem is identified, but not fixed
204 2011-03-01 01:17:58 <molecular> and it wont be fixed quickly
205 2011-03-01 01:18:36 <molecular> to avoid the issue, use higher amounts or pay fee
206 2011-03-01 01:18:51 <omglolbbq> ok tnx for the update :)
207 2011-03-01 01:19:07 <molecular> this problem might be scaring people off
208 2011-03-01 01:19:38 <ArtForz> q is, why is it suddenly *that* bad
209 2011-03-01 01:20:18 <molecular> how large was the backlog usually before whenever this started today?
210 2011-03-01 01:20:39 <ArtForz> usually 0-20, a few 100 while tx spam was ongoing
211 2011-03-01 01:21:09 <molecular> maybe someone is actually testing a tx-spam-attack?
212 2011-03-01 01:21:41 <ArtForz> the tx that make it into blocks dont look exactly spammy
213 2011-03-01 01:22:29 <molecular> it doesnt look like there are more than usual transaction on bitcoinmonitor
214 2011-03-01 01:22:48 <bk128> wait, 669 tx's are queued?
215 2011-03-01 01:23:05 <ArtForz> 671 now
216 2011-03-01 01:23:31 <molecular> is it monotonically rising?
217 2011-03-01 01:23:34 <bk128> how do you see that?
218 2011-03-01 01:23:48 <molecular> the only one here that can see it is ArtForz with a special monitor-node he has
219 2011-03-01 01:23:58 <ArtForz> python half-a-node connected to main client
220 2011-03-01 01:24:05 <ArtForz> or just add a debug print to your client
221 2011-03-01 01:24:22 <bk128> ah.  tx's are encoded in blocks, right?  how many can be encoded in each?
222 2011-03-01 01:24:43 <ArtForz> free tx? about 10 on average
223 2011-03-01 01:24:44 <bk128> time to implement tx fees?
224 2011-03-01 01:25:09 <bk128> I've never fully understood how tx fees work
225 2011-03-01 01:26:01 <ArtForz> just print mapTransactions.size() at the end of CTransaction::AcceptToMemoryPool
226 2011-03-01 01:26:10 <hazek> is there a limit of how many txs fit into a single block?
227 2011-03-01 01:26:19 <ArtForz> yes
228 2011-03-01 01:26:22 <hazek> oh
229 2011-03-01 01:26:31 <ArtForz> theres several
230 2011-03-01 01:26:43 <bk128> https://en.bitcoin.it/wiki/File:Lfm_fee.png I've never seen that message before
231 2011-03-01 01:27:13 <molecular> ArtForz, can you see the sizes of the queued transactions? I bet there are many big ones
232 2011-03-01 01:27:26 <ArtForz> hmmm... no
233 2011-03-01 01:27:45 <hazek> but wait gusy, isn't this a big problem?
234 2011-03-01 01:27:47 <molecular> "no, can't see", or "no, there arent"?
235 2011-03-01 01:27:52 <ArtForz> can't see
236 2011-03-01 01:27:58 <hazek> or is there a solution available?
237 2011-03-01 01:28:11 <ArtForz> and if I add code to print it I have to restart my monitor node... which also only keeps tx in memory
238 2011-03-01 01:30:02 <molecular> would add some printf to my node and compile bitcoin for my first time, but it's getting late and I have to get up tomorrow...
239 2011-03-01 01:30:23 <bk128> can you request to pay a tx fee so your tx is processed faster?
240 2011-03-01 01:31:16 <ArtForz> yes
241 2011-03-01 01:31:40 <bk128> in the normal client?
242 2011-03-01 01:31:59 <molecular> I dont even know how to build bitcoin
243 2011-03-01 01:32:04 <bk128> me either :p
244 2011-03-01 01:32:25 <molecular> oh, "make"
245 2011-03-01 01:32:28 <ArtForz> paytxfee option
246 2011-03-01 01:32:46 <molecular> urgh, this will take longer than 10 minutes...
247 2011-03-01 01:32:58 <hazek> I don't mean to sound a panic alarm but if we already have a queue at this stage how bad is it going to get if this currecie ever gets picked up by 10 or 50 times more people
248 2011-03-01 01:33:02 <bk128> doesn't look like it's accessible using the bitcoin.org gui client
249 2011-03-01 01:33:06 <molecular> am off to bed and hoping bitcoin is still there tomorrow ;)
250 2011-03-01 01:33:11 <hazek> currency*
251 2011-03-01 01:33:30 <ArtForz> bk128: it's a commandline switch/config file option
252 2011-03-01 01:33:38 <molecular> it's in options
253 2011-03-01 01:33:39 <ArtForz> iirc paytxfee=0.01 = pay 0.01 per kB
254 2011-03-01 01:33:48 <molecular> settings/options/pay transaction fee
255 2011-03-01 01:33:58 <ArtForz> oh, right
256 2011-03-01 01:34:05 <bk128> ArtForz: so how long do you think tx's are going to be queued?
257 2011-03-01 01:34:09 <amiller> ArtForz, would you tell me if this transaction is in your queue b675a2b3429a40d47a75501aac442c3f27eda34e9aff51ec8b40cef00f3d6f16
258 2011-03-01 01:34:13 <ArtForz> sec
259 2011-03-01 01:34:21 <amiller> it has a .03 btc fee and is 12.00 btc in value
260 2011-03-01 01:34:35 <ArtForz> yep
261 2011-03-01 01:35:36 <molecular> amiller, my node received your tx before block 111136 was generated
262 2011-03-01 01:36:10 <amiller> since it has a .03 fee and it's several hours old now wouldn't it be a higher priority
263 2011-03-01 01:36:36 <ArtForz> it should be... how big is it?
264 2011-03-01 01:36:49 <amiller> i don't know how to check
265 2011-03-01 01:36:59 <amiller> i'm using 'bitcoin listtransacftions '' 100'
266 2011-03-01 01:37:13 <dirtyfilthy> there's an option in settings->options for paying a default tx fee in the gui
267 2011-03-01 01:38:02 <bk128> not on the mac client
268 2011-03-01 01:38:05 <ArtForz> I have a theory why it's not getting into a block
269 2011-03-01 01:38:22 <bk128> oh, it's under bitcoin>preferences
270 2011-03-01 01:38:46 <hazek> ArtForz what's the limit of txs per block?
271 2011-03-01 01:39:06 <amiller> ArtForz, do go on
272 2011-03-01 01:39:07 <ArtForz> hazek: theres no fixed limit
273 2011-03-01 01:39:44 <hazek> hmmmm
274 2011-03-01 01:39:46 <hazek> [03:26] <hazek> is there a limit of how many txs fit into a single block? [03:26] <ArtForz> yes
275 2011-03-01 01:39:54 <hazek> which is it?
276 2011-03-01 01:39:55 <hazek> :D
277 2011-03-01 01:40:19 <ArtForz> 500kB
278 2011-03-01 01:40:35 <bk128> DiabloMiner keeps locking up on my debian computer
279 2011-03-01 01:40:46 <ArtForz> amiller: client doesnt take into account block size when determining fee at send of transaction
280 2011-03-01 01:41:00 <hazek> please explain the logic behind it, numbers don't tell me anything because I don't understand the code..
281 2011-03-01 01:41:13 <ArtForz> so automatically calculated min fee is the bare minimum so that the tx would go into a otherwise empty block
282 2011-03-01 01:41:37 <ArtForz> thanks to the backlog of free tx... theres never a otherwise empty block
283 2011-03-01 01:41:48 <amiller> how big is this particular transaction block?
284 2011-03-01 01:41:56 <amiller> er not transaction block
285 2011-03-01 01:41:58 <amiller> just transaction
286 2011-03-01 01:42:21 <ArtForz> which means we have to fix CreateTransaction. soon.
287 2011-03-01 01:42:40 <molecular> is there an issue for that on github?
288 2011-03-01 01:42:51 <ArtForz> I don't think so
289 2011-03-01 01:43:00 <molecular> should I make it?
290 2011-03-01 01:43:16 <ArtForz> I'd say yes
291 2011-03-01 01:43:32 <molecular> ok, will
292 2011-03-01 01:43:52 <ArtForz> otherwise the current fee handling (or lack of it...) in CreateTransaction will cause major problems
293 2011-03-01 01:44:52 <molecular> damnit, github doesn't like me... can't create an issue or even see issues ("loading...")
294 2011-03-01 01:45:10 <molecular> ah, works now
295 2011-03-01 01:46:19 <molecular> ok, title: "CreateTransaction: should suggest fee for low-priority transactions" ?
296 2011-03-01 01:48:58 <lfm> hazek: you know the meaning of the term "bug"?
297 2011-03-01 01:49:08 <hazek> sure do
298 2011-03-01 01:49:24 <hazek> didn't realize this was the reason for this problem
299 2011-03-01 01:49:29 <bk128> ArtForz: so what's the fix for createtransaction?
300 2011-03-01 01:49:41 <hazek> all my questions are answered, ty ;)
301 2011-03-01 01:49:41 <lfm> ok
302 2011-03-01 01:51:10 <ArtForz> 1. use a nBlockSize a lottle more realistic than 1 for GetMinFee
303 2011-03-01 01:51:15 <molecular> ArtForz, allow me to quote some of what you said here in the github issue?
304 2011-03-01 01:51:25 <ArtForz> sure
305 2011-03-01 01:51:28 <bk128> so there will be mandatory fees now?
306 2011-03-01 01:51:50 <lfm> bk128: nope
307 2011-03-01 01:52:04 <ArtForz> 2. calculate tx priority score and size, do the fAllowFree calculation like CreateNewBlock
308 2011-03-01 01:53:29 <lfm> bk128: there may be a new fee for certain odd txn
309 2011-03-01 01:53:41 <bk128> ok
310 2011-03-01 01:53:53 <hazek> odd tx as in big or as in super small
311 2011-03-01 01:54:13 <hazek> or does the ammount not play a role
312 2011-03-01 01:54:21 <lfm> big as in near the limit
313 2011-03-01 01:54:46 <hazek> is that big in terms of bitcoins or interms of bite size
314 2011-03-01 01:54:52 <hazek> byte*
315 2011-03-01 01:54:56 <molecular> https://github.com/bitcoin/bitcoin/issues/86 <- go vote
316 2011-03-01 01:55:22 <molecular> and/or comment
317 2011-03-01 01:55:26 <lfm> big in terms of byte size I think mainly influenced by the number of inputs
318 2011-03-01 01:55:28 <mmarker> *grumble*
319 2011-03-01 01:55:31 <mmarker> Sometimes, I hate C
320 2011-03-01 01:55:40 <hazek> cool ty
321 2011-03-01 01:57:53 <molecular> how can I label an issue on github?
322 2011-03-01 01:59:31 <dirtyfilthy> mmarker: was it you that needed a copy of bouncycastle built for android?
323 2011-03-01 01:59:43 <mmarker> Nope, got it
324 2011-03-01 01:59:47 <mmarker> I took the code from APG
325 2011-03-01 01:59:52 <dirtyfilthy> sweet as
326 2011-03-01 01:59:58 <mmarker> they have a set of scripts that trims up a BC
327 2011-03-01 02:00:12 <mmarker> But right now, I'm trying to collect a bounty
328 2011-03-01 02:00:14 <mmarker> :D
329 2011-03-01 02:00:22 <mmarker> failing HARD, but trying!
330 2011-03-01 02:00:26 <dirtyfilthy> which one?
331 2011-03-01 02:00:35 <mmarker> porting a win32 miner to Linux
332 2011-03-01 02:00:46 <dirtyfilthy> good luck
333 2011-03-01 02:00:48 <mmarker> Getting a segfault in my ASM, which always spells "FUN TO DEBUG"
334 2011-03-01 02:02:33 <lfm> mmarker: no always, sometimes gdb will poit you right to it
335 2011-03-01 02:02:41 <mmarker> lfm: It is
336 2011-03-01 02:02:43 <mmarker> actually
337 2011-03-01 02:02:59 <mmarker> It's trying to figure out why my leaf function is messing up
338 2011-03-01 02:03:09 <mmarker> most likely me f'in up the x86_64 calling convention
339 2011-03-01 02:03:40 <mmarker> This is bad: mov eax,[rax]...and rax has the value 0x20
340 2011-03-01 02:03:59 <mmarker> I dont think memory address 0x20 is...urm...a good place to be poking about
341 2011-03-01 02:04:17 <lfm> ya sometimes it is near impossible to find out how you trashed some addres
342 2011-03-01 02:04:28 <mmarker> I see how I did it
343 2011-03-01 02:04:35 <mmarker> Remember kids:
344 2011-03-01 02:04:48 <mmarker> int nonce; f(nonce) passes the value of the nonce
345 2011-03-01 02:05:01 <mmarker> int nonce; f(&nonce), that passes the pointer
346 2011-03-01 02:05:03 <mmarker> DOH
347 2011-03-01 02:06:55 <mmarker> Ok, now comes funtime #2, fix one sigsev, get another
348 2011-03-01 02:08:12 <mmarker> Oh. Snap
349 2011-03-01 02:08:25 <mmarker> lea
350 2011-03-01 02:08:44 <mmarker> OOPS
351 2011-03-01 02:08:46 <lfm> ya I suppose segv is the most common asm bug, less common in higher langs
352 2011-03-01 02:09:08 <mmarker> Note kids, if you're programming in 64 bit
353 2011-03-01 02:09:16 <mmarker> it helps to, like, use 64 bits for memory access!
354 2011-03-01 02:12:20 <mmarker> hmm, well...yasm is failing me
355 2011-03-01 02:12:34 <bk128> anyone know how to find out why diablominer keeps locking up after a few hours?
356 2011-03-01 02:18:42 <bk128> maybe I'll have to switch to the python miner
357 2011-03-01 02:19:17 <amiller> i got worse performance with diablominer than the python miner - but with an nvidia card so it might not be relevant
358 2011-03-01 02:19:45 <Necr0s> I only ever ran the python one.
359 2011-03-01 02:19:50 <amiller> it wasn't very much better, 14Mhps vs ~13.8Mhps
360 2011-03-01 02:22:20 <bk128> Diablo runs fine for a while, then just locks up.  it's doing about 680mhash/sec on 2 5870's OC'd a bit on Debian
361 2011-03-01 02:22:28 <bk128> maybe I'll have to install the sun java
362 2011-03-01 02:22:38 <bk128> oracle java*
363 2011-03-01 02:23:09 <lfm> bk128: what java do you have
364 2011-03-01 02:23:39 <bk128> java version "1.6.0_18" OpenJDK Runtime Environment (IcedTea6 1.8.3) (6b18-1.8.3-2+squeeze1)
365 2011-03-01 02:23:40 <bk128> OpenJDK 64-Bit Server VM (build 16.0-b13, mixed mode)
366 2011-03-01 02:25:16 <Necr0s> Couldn't hurt to try.
367 2011-03-01 02:25:34 <mmarker> grr
368 2011-03-01 02:26:21 <mmarker> so, something bizzare is happening with this grand asm, isn't lea reg, [reg1 + reg2 * 4] reg = value in reg1 + 4 * value in reg2?
369 2011-03-01 02:27:10 <ArtForz> no
370 2011-03-01 02:27:29 <ArtForz> wait, it is
371 2011-03-01 02:27:34 <mmarker> ArtForz, what should it be then...the number I see in the register isn't that
372 2011-03-01 02:27:36 <ArtForz> reg = reg1 + reg2 * 4
373 2011-03-01 02:27:44 <mmarker> which is what's causing me some mindboggling
374 2011-03-01 02:31:45 <mmarker> hmm
375 2011-03-01 02:31:51 <mmarker> the disassembly is right
376 2011-03-01 02:31:55 <mmarker> something is amiss
377 2011-03-01 02:36:50 <mmarker> Oh BINGO
378 2011-03-01 02:37:01 <mmarker> I'm clobbering the register somewhere
379 2011-03-01 02:37:04 <mmarker> I am a dumbass
380 2011-03-01 02:41:29 <mmarker> Yea, think I got it
381 2011-03-01 02:41:38 <mmarker> add esi, 16 blanks out the high order bits?
382 2011-03-01 02:42:04 <ArtForz> iirc, yes
383 2011-03-01 02:42:15 <mmarker> Ok, that's my bug for THIS pass
384 2011-03-01 02:42:21 <mmarker> now let's see if we generate proper SHA-2
385 2011-03-01 02:42:54 <mmarker> Well, it runs, but incorrectly!
386 2011-03-01 02:42:58 <mmarker> Woot!
387 2011-03-01 02:43:08 <mmarker> HashMeter(0): 17377216 hashes, 7639.96 khash/sec
388 2011-03-01 02:43:12 <andrewh> if there's anyone that wants any of these, i'll give them to you at 50% their price. http://www.bitcoin.org/smf/index.php?topic=3934.0
389 2011-03-01 02:43:36 <mmarker> hmm, wait
390 2011-03-01 02:43:41 <validus> whats a gtx 460 hit mining
391 2011-03-01 02:43:54 <mmarker> validus, my 1gig hits ~68k
392 2011-03-01 02:44:07 <ArtForz> yup, ~68k
393 2011-03-01 02:44:07 <validus> anda 6970 hits 300k?
394 2011-03-01 02:44:18 <ArtForz> yup
395 2011-03-01 02:44:20 <validus> iirc
396 2011-03-01 02:44:30 <validus> hmm...
397 2011-03-01 02:45:56 <validus> i plan on mining but not getting a video card just for the sake of mining so trying to figure out the best economical choice
398 2011-03-01 02:46:02 <mmarker> It's the difference between design, right Art. nVidia focus on floating point, ATI, integer?
399 2011-03-01 02:46:11 <andrewh> ;;bc,calc 300000
400 2011-03-01 02:46:12 <gribble> The average time to generate a block at 300000 Khps, given current difficulty of 55590.23763914 , is 1 week, 2 days, 5 hours, 4 minutes, and 20 seconds
401 2011-03-01 02:46:14 <andrewh> wai
402 2011-03-01 02:46:15 <andrewh> t
403 2011-03-01 02:46:17 <andrewh> 300k?
404 2011-03-01 02:46:32 <validus> 300 000
405 2011-03-01 02:46:37 <andrewh> or did you mean 300M/
406 2011-03-01 02:46:41 <ArtForz> 300k khash aka 300M
407 2011-03-01 02:46:43 <andrewh> damn keyboard
408 2011-03-01 02:46:45 <Lachesis> mmarker, what's getting you 7.6 MH/s?
409 2011-03-01 02:46:49 <andrewh> heh
410 2011-03-01 02:47:10 <mmarker> Lachesis: Magic
411 2011-03-01 02:47:11 <validus> its really gonna depend on framerate i guess
412 2011-03-01 02:47:18 <validus> 480 still trumps alot of gpu's in fps
413 2011-03-01 02:47:26 <Lachesis> mmarker, i see - what cpu?
414 2011-03-01 02:47:29 <andrewh> every time I snap my fingers, it's 1Ghash
415 2011-03-01 02:47:37 <mmarker> Lachesis: Don't trust my numbers yet
416 2011-03-01 02:47:39 <andrewh> ;p
417 2011-03-01 02:47:43 <Lachesis> mmarker, alright
418 2011-03-01 02:47:48 <Lachesis> mmagic, you there?
419 2011-03-01 02:47:49 <mmarker> since I dont think I'm calculating correctly
420 2011-03-01 02:47:58 <Lachesis> mmagic, gribble has never seen you
421 2011-03-01 02:47:59 <mmarker> WAIT until I get proof of work :D
422 2011-03-01 02:48:11 <andrewh> http://www.youtube.com/watch?v=5_sfnQDr1-o
423 2011-03-01 02:48:23 <Lachesis> mmagic, and your name starts with mma, which screws over my autocomplete when i try to talk to mmarker
424 2011-03-01 02:48:23 <validus> so im building a new system with 2 grand. what video card decent for mining and gaming both
425 2011-03-01 02:49:18 <validus> ;;bc,calc 68000
426 2011-03-01 02:49:19 <gribble> The average time to generate a block at 68000 Khps, given current difficulty of 55590.23763914 , is 5 weeks, 5 days, 15 hours, 19 minutes, and 10 seconds
427 2011-03-01 02:49:33 <TheKid> validus: 5970
428 2011-03-01 02:49:37 <TheKid> barring that, dual 5870
429 2011-03-01 02:49:43 <TheKid> or wait for the 6990
430 2011-03-01 02:49:50 <validus> im also lookin at fps in games
431 2011-03-01 02:49:57 <validus> and the 3dmark scores
432 2011-03-01 02:50:08 <Necr0s> same, 5970s.
433 2011-03-01 02:50:09 <Lachesis> validus, yeah, 5970 is a fast card
434 2011-03-01 02:50:18 <Lachesis> 6990 will be killer too
435 2011-03-01 02:50:27 <validus> im not lookin to drop 700 bucks on a gpu
436 2011-03-01 02:50:30 <ArtForz> 6990 should pwn for 3D
437 2011-03-01 02:50:30 <TheKid> validus: I know, I told you what was the best
438 2011-03-01 02:50:38 <TheKid> validus: then don't ask without giving a price range
439 2011-03-01 02:50:44 <ArtForz> dual 6950 also isnt too bad
440 2011-03-01 02:50:46 <validus> i did. i said i have 2 grand to build a new pc
441 2011-03-01 02:50:51 <TheKid> yeah
442 2011-03-01 02:50:51 <validus> that includes monitor and all though
443 2011-03-01 02:50:55 <Lachesis> $2,000 and you won't spend 700 on a GC?
444 2011-03-01 02:50:57 <TheKid> that is plenty of cash
445 2011-03-01 02:51:02 <TheKid> for a kickass video card
446 2011-03-01 02:51:06 <Necr0s> yes
447 2011-03-01 02:51:06 <validus> im thinking 27 inch monitor and 5.1 surround sound then making up the rest
448 2011-03-01 02:51:07 <TheKid> unless you are building intel
449 2011-03-01 02:51:08 <TheKid> lol
450 2011-03-01 02:51:11 <validus> i do want the 100 gaming keyboard :P
451 2011-03-01 02:51:13 <Lachesis> validus, oh fair enough
452 2011-03-01 02:51:14 <validus> $100*
453 2011-03-01 02:51:20 <TheKid> validus: get the Das Keyboard Ultimate
454 2011-03-01 02:51:26 <TheKid> best keyboard in existence
455 2011-03-01 02:51:35 <validus> i just like the one i saw that was backlit
456 2011-03-01 02:51:39 <validus> not even the g19
457 2011-03-01 02:52:02 <validus> the razor lycosa
458 2011-03-01 02:52:05 <ArtForz> meh, I prefer linear spring + foil contact
459 2011-03-01 02:52:24 <Necr0s> $700 vid card, $500 display, $300 motherboard, that's still $500 left.
460 2011-03-01 02:52:43 <validus> true, prob go with amd quad core black edition
461 2011-03-01 02:52:44 <mmarker> Hmm
462 2011-03-01 02:52:47 <validus> 4 - 8 gig of ram
463 2011-03-01 02:52:53 <validus> then at least 2 tb. blu ray burner
464 2011-03-01 02:53:00 <TheKid> get an SSD, don't be a scrub
465 2011-03-01 02:53:03 <validus> no
466 2011-03-01 02:53:08 <mmarker> is there a test miner which, if I connect to it, will be able to test to see if I'm returning valid work?
467 2011-03-01 02:53:09 <validus> i will only buy ssd when i can afford the best
468 2011-03-01 02:53:12 <TheKid> why a blu ray burner?
469 2011-03-01 02:53:13 <Necr0s> Yes, SSD is a must.
470 2011-03-01 02:53:22 <TheKid> and you can, 700 is more than enough for a top of the line 256 gig drive
471 2011-03-01 02:53:25 <validus> because i have crap tons of data and their not that expensive
472 2011-03-01 02:53:27 <TheKid> which is more than enough for games
473 2011-03-01 02:53:35 <Necr0s> Yeah, leave out the BD-RW and get SSD instead.
474 2011-03-01 02:53:36 <TheKid> optical media= fail
475 2011-03-01 02:53:38 <validus> im having trouble with teh 2.5 tb i have now
476 2011-03-01 02:53:39 <Necr0s> Optical media is dead.
477 2011-03-01 02:53:46 <TheKid> seriously are you new at building computers
478 2011-03-01 02:53:54 <TheKid> dual WD20EADS
479 2011-03-01 02:53:54 <validus> nope. but i do lots of HD and blu ray
480 2011-03-01 02:53:59 <Necr0s> You need a file server.
481 2011-03-01 02:53:59 <TheKid> if you need space
482 2011-03-01 02:54:04 <validus> i need a crapt ton of space
483 2011-03-01 02:54:05 <TheKid> or fileserver, true
484 2011-03-01 02:54:05 <validus> lol
485 2011-03-01 02:54:11 <validus> like 30tb+
486 2011-03-01 02:54:12 <validus> :/
487 2011-03-01 02:54:17 <TheKid> then build a file server
488 2011-03-01 02:54:20 <TheKid> and don't half ass it
489 2011-03-01 02:54:21 <ArtForz> yep
490 2011-03-01 02:54:22 <Necr0s> Big server.
491 2011-03-01 02:54:24 <validus> well that comes later
492 2011-03-01 02:54:25 <mmarker> anyone?
493 2011-03-01 02:54:33 <Necr0s> Fetching opticaL discs is fail.
494 2011-03-01 02:54:35 <mmarker> :D
495 2011-03-01 02:54:37 <TheKid> mmarker: just connect to testnet and gen
496 2011-03-01 02:54:38 <validus> im just going to get me a decent gaming pc for school and games
497 2011-03-01 02:54:40 <Necr0s> It needs to be all online all the time.
498 2011-03-01 02:54:42 <TheKid> or connect to slush's pool
499 2011-03-01 02:54:49 <mmarker> I'm on slush's pool
500 2011-03-01 02:54:59 <mmarker> but I don't know if I've hit a winner
501 2011-03-01 02:55:05 <TheKid> validus: look, let me link you how to not suck at building computers
502 2011-03-01 02:55:07 <TheKid> gimme 10 minutes
503 2011-03-01 02:55:14 <mmarker> I'm hashing...
504 2011-03-01 02:55:21 <Necr0s> A 20tb server using raid6 shouldn't be too difficult to build.
505 2011-03-01 02:55:22 <validus> i know how not to suck i was just asking opinions so im not spending 24 hrs doing research
506 2011-03-01 02:55:25 <ArtForz> my local fileserver currently has about 17T
507 2011-03-01 02:55:42 <validus> i spent 12 hrs researching ati then 12 hrs on nvidia for a $50 budget card
508 2011-03-01 02:55:46 <validus> you dont understand :P
509 2011-03-01 02:55:54 <ArtForz> and thats with a bunch of 1T drives in the mix
510 2011-03-01 02:56:03 <mmarker> Ok, if this is working, which I doubt. I'm hitting 6500 khash/sec on one core
511 2011-03-01 02:56:16 <mmarker> BUT, can't tell if it's working...guess I need to testnet
512 2011-03-01 02:57:19 <ArtForz> 2 6*2T raid6s, 6*1T raid6, 2*1T raid1
513 2011-03-01 02:57:25 <lfm> validus: I think you should do 24 hours research
514 2011-03-01 02:57:45 <validus> i usually do longer. i will make an effort to find everything about something cuz im frugal like that
515 2011-03-01 02:57:49 <validus> but have expensive tastes
516 2011-03-01 02:59:08 <validus> but i know everybody here loves ati, and thast best to find info about ati and real world tests than just trolling forums and half flame threads
517 2011-03-01 02:59:51 <TheKid> validus: http://www.newegg.com/Product/Product.aspx?Item=N82E16820231308
518 2011-03-01 02:59:58 <TheKid> http://www.newegg.com/Product/Product.aspx?Item=N82E16817153106
519 2011-03-01 03:00:04 <TheKid> http://www.newegg.com/Product/Product.aspx?Item=N82E16813130295
520 2011-03-01 03:00:08 <TheKid> http://www.newegg.com/Product/Product.aspx?Item=N82E16822136514
521 2011-03-01 03:00:18 <TheKid> http://www.newegg.com/Product/Product.aspx?Item=N82E16811146065&Tpk=nzxt%20vulcan
522 2011-03-01 03:00:23 <validus> a pm would have been better so your not spamming everywhere lol
523 2011-03-01 03:00:26 <TheKid> http://www.newegg.com/Product/Product.aspx?Item=N82E16814121388R
524 2011-03-01 03:00:28 <TheKid> whatever
525 2011-03-01 03:00:32 <TheKid> http://www.newegg.com/Product/Product.aspx?Item=N82E16819103808
526 2011-03-01 03:00:39 <mmarker> Ok, how the f does -testnet work. getwork is giving me an error :(
527 2011-03-01 03:00:39 <TheKid> http://www.newegg.com/Product/Product.aspx?Item=N82E16820148349
528 2011-03-01 03:00:49 <TheKid> though it occurs to me this might be over budget, second
529 2011-03-01 03:00:51 <amiller> TheKid, that's expensive for a 5870
530 2011-03-01 03:00:55 <mmarker> course, I have 0 blocks. Frack
531 2011-03-01 03:00:59 <TheKid> amiller: 5870x2
532 2011-03-01 03:01:18 <lfm> validus: do you want performance on "real world tests" or good biycoin mining speed?
533 2011-03-01 03:01:32 <[Tycho]> His bitcoin fell on the floor and shattered badly, but he has some glue ! :) http://blockexplorer.com/tx/bdcb569c164ac271ec0eade20a101a2988f7793034340506db8bdec12393bdff
534 2011-03-01 03:01:49 <validus> im doing both i guess you could say. the real world tests are the mining speed
535 2011-03-01 03:01:59 <TheKid> validus: okay, that's slightly over budget
536 2011-03-01 03:01:59 <validus> but im also lookin at how well the card handles tessla, fps, and cost
537 2011-03-01 03:02:20 <lfm> tessla?
538 2011-03-01 03:02:22 <TheKid> but why do you want a 5.1 surround system when you could just get a g930
539 2011-03-01 03:02:24 <validus> directx 11 thing
540 2011-03-01 03:02:26 <TheKid> lfm: tesselation
541 2011-03-01 03:02:26 <[Tycho]> Tesselation.
542 2011-03-01 03:02:36 <lfm> k
543 2011-03-01 03:02:40 <TheKid> validus: http://www.newegg.com/Product/Product.aspx?Item=N82E16826104399&Tpk=g%20930
544 2011-03-01 03:02:53 <mmarker> I need to add some debugging dumps
545 2011-03-01 03:03:07 <TheKid> and that's assuming you get everything from newegg, buying on ebay for some of that and the price will drop considerably
546 2011-03-01 03:03:09 <mmarker> getting a lot of hashes, no hits on slush's pool
547 2011-03-01 03:03:10 <validus> i already have a ncie set of headphones
548 2011-03-01 03:03:15 <validus> just not 5.1 surround
549 2011-03-01 03:03:29 <validus> but thse generic speakers i have suck ass. and its only like 60 bucks for some speakers
550 2011-03-01 03:03:35 <mmarker> Yea, something must be wrong
551 2011-03-01 03:03:42 <validus> if i was buying hardcore headphones id get the new bose
552 2011-03-01 03:04:01 <lfm> sure, go for those 7.1 headphones
553 2011-03-01 03:04:07 <validus> i got these for like 20 i think http://www.newegg.com/Product/Product.aspx?Item=N82E16826106455&cm_re=hifi_headphones-_-26-106-455-_-Product
554 2011-03-01 03:04:10 <TheKid> 7.1 wireless headphones dude
555 2011-03-01 03:04:15 <TheKid> they are awesome
556 2011-03-01 03:04:20 <TheKid> and better than most speakers you can buy
557 2011-03-01 03:04:21 <validus> sound quality is awesome on those
558 2011-03-01 03:04:38 <TheKid> meh, I'm no audiophile so I won't argue that
559 2011-03-01 03:04:54 <validus> im not either . but for 20 bucks those headphones are golden
560 2011-03-01 03:04:55 <validus> lol
561 2011-03-01 03:05:00 <lfm> monster headphone cables
562 2011-03-01 03:05:06 <mmarker> Hang on, I have a sample getwork...
563 2011-03-01 03:05:14 <validus> most monster cables are a makerting ploy
564 2011-03-01 03:05:15 <TheKid> anyway, that whole setup comes up to a little over 2k
565 2011-03-01 03:05:17 <validus> marketing*
566 2011-03-01 03:05:29 <validus> TheKid: just slightly over dont work. i only have 2 grand and i mean exact
567 2011-03-01 03:05:30 <TheKid> and that's assuming you can't reuse any of your old drives
568 2011-03-01 03:05:32 <quellhorst> anyone have to convert moex to 6 pin for video cards?
569 2011-03-01 03:05:38 <TheKid> validus: save one more paycheck then noob
570 2011-03-01 03:05:45 <validus> i dont have a paycheck
571 2011-03-01 03:05:48 <TheKid> quellhorst: I did
572 2011-03-01 03:05:48 <validus> thats half the problem
573 2011-03-01 03:06:05 <lfm> quellhorst: I generally prefer to get the right psu
574 2011-03-01 03:06:05 <TheKid> well wait
575 2011-03-01 03:06:06 <validus> im using college loan to build me a nice system hehe
576 2011-03-01 03:06:07 <TheKid> for like
577 2011-03-01 03:06:11 <TheKid> that's stupid
578 2011-03-01 03:06:16 <validus> not really since i need it
579 2011-03-01 03:06:27 <TheKid> if you're taking college loans you're gonna be in crazy debt and poor
580 2011-03-01 03:06:33 <TheKid> you don't want to waste it on a ridiculous computer
581 2011-03-01 03:06:38 <validus> well thats the only way i can get to college atm
582 2011-03-01 03:06:40 <TheKid> http://www.ncix.com/products/index.php?sku=58434&vpn=LX%2ERAC02%2E041&manufacture=Acer
583 2011-03-01 03:06:43 <validus> and im not wasting it
584 2011-03-01 03:06:44 <TheKid> right
585 2011-03-01 03:06:47 <TheKid> buy that laptop
586 2011-03-01 03:06:50 <TheKid> play everything on high
587 2011-03-01 03:06:53 <validus> nope. no lappys
588 2011-03-01 03:06:55 <TheKid> it's great
589 2011-03-01 03:07:01 <TheKid> well, you're making a poor life choice
590 2011-03-01 03:07:05 <TheKid> enjoy
591 2011-03-01 03:07:08 <lfm> mmarker so?
592 2011-03-01 03:07:10 <validus> nah, not really
593 2011-03-01 03:07:35 <validus> poor life choice will be staying with this school after a year. i plan on transferring locally and finishing up my degree
594 2011-03-01 03:07:39 <validus> but this is neccessary evil
595 2011-03-01 03:07:50 <mmarker> lfm: I need to fix my code.
596 2011-03-01 03:08:01 <validus> besides that how is it poor. i have no credit. no job. nodda. so therefore im still doing better as i have no other way to pay for school
597 2011-03-01 03:08:03 <TheKid> validus: explain how a 2k computer on a college loan is necessary
598 2011-03-01 03:08:04 <validus> or anything that it brings
599 2011-03-01 03:08:12 <mmarker> lfm: Need to figure out how to insert this test.
600 2011-03-01 03:08:23 <validus> its not neccessary but a new pc is. so thats what i gave myself as leave way
601 2011-03-01 03:08:24 <TheKid> what will it do that the laptop I linked can't do
602 2011-03-01 03:08:29 <TheKid> okay
603 2011-03-01 03:08:33 <TheKid> have fun, like I said
604 2011-03-01 03:09:05 <validus> new pc is definate. for that is a must. if im getting something new im not going bottom of the barrel shit sticks that i gotta spend money on
605 2011-03-01 03:09:09 <validus> ill hate myself later for it
606 2011-03-01 03:09:34 <quellhorst> lfm: yeah, kinda too late for me now
607 2011-03-01 03:09:45 <validus> kinda like paying for alienware thats just an overpriced dell
608 2011-03-01 03:09:52 <quellhorst> TheKid: where did you get the converter cable?
609 2011-03-01 03:10:18 <quellhorst> TheKid: i'm thinking of getting 2x of these. http://www.newegg.com/Product/Product.aspx?Item=N82E16812887001&cm_re=molex_to_6_pin-_-12-887-001-_-Product
610 2011-03-01 03:10:53 <validus> and grants are also paying for over 1/2 of my school. but new pc is no question so i dont see it as a bad decision
611 2011-03-01 03:11:00 <validus> i see staying in that online program a bad decision though
612 2011-03-01 03:11:01 <validus> heh
613 2011-03-01 03:11:46 <mmarker> lfm: however, if I'm spinning 24000khash/sec
614 2011-03-01 03:11:55 <mmarker> I should getting H==0 at SOME point
615 2011-03-01 03:12:00 <TheKid> quellhorst: mine came with my card
616 2011-03-01 03:12:16 <lfm> mmarker: cool
617 2011-03-01 03:12:24 <mmarker> lfm: No, NOT COOL
618 2011-03-01 03:12:53 <mmarker> Yea, it's going fast...but if the sha256 routine is wrong, who cares I can so 6500 kcalcs/sec on a core
619 2011-03-01 03:12:56 <mmarker> if it's garbage
620 2011-03-01 03:13:13 <luke-jr> mmarker: whatcha writing?
621 2011-03-01 03:13:22 <lfm> ok
622 2011-03-01 03:13:23 <mmarker> luke-jr: porting the SSE2 miner to Linux
623 2011-03-01 03:13:34 <luke-jr> doesn't cpuminer do that?
624 2011-03-01 03:13:42 <quellhorst> TheKid: lol! my card has those in it too
625 2011-03-01 03:13:45 <mmarker> luke-jr: using GCC intrinsics
626 2011-03-01 03:13:46 <quellhorst> where i thought it only had a dvd
627 2011-03-01 03:13:52 <mmarker> which are SO SO BAD
628 2011-03-01 03:14:05 <luke-jr> mmarker: how about an AVX miner instead?
629 2011-03-01 03:14:06 <quellhorst> what is a crossfire?
630 2011-03-01 03:14:18 <mmarker> Send me a Sandy Bridge...
631 2011-03-01 03:15:15 <luke-jr> mmarker: can't get those anymore ;P
632 2011-03-01 03:15:21 <mmarker> Heh
633 2011-03-01 03:15:24 <luke-jr> mmarker: want a shell account on mine?
634 2011-03-01 03:15:37 <mmarker> Not right now
635 2011-03-01 03:15:40 <luke-jr> aww
636 2011-03-01 03:15:42 <mmarker> I have too many projects! :(
637 2011-03-01 03:15:57 <mmarker> But an AVX miner would be simple after I debug this crap
638 2011-03-01 03:16:05 <luke-jr> you can run it with your own bitcoind/pool account on one of my cores <.<
639 2011-03-01 03:17:02 <amiller> luke-jr, can i have a shell on your machine
640 2011-03-01 03:17:14 <TheKid> quellhorst: when you link two AMD graphics cards together, it is called CrossfireX
641 2011-03-01 03:17:17 <luke-jr> amiller: why?
642 2011-03-01 03:17:44 <amiller> tinker with mining programs
643 2011-03-01 03:17:52 <amiller> store a separate wallet somewhere
644 2011-03-01 03:18:07 <luke-jr> amiller: are you qualified to write an AVX miner? :P
645 2011-03-01 03:19:18 <amiller> no, but i'm qualified to write opencl code and make meaningless comparisons using stream sdk
646 2011-03-01 03:19:57 <luke-jr> useful OpenCL code? :p
647 2011-03-01 03:20:01 <mmarker> ok, time to haxxxor some more
648 2011-03-01 03:20:05 <mmarker> last shot before bed
649 2011-03-01 03:22:30 <amiller> potentially useful opencl code
650 2011-03-01 03:23:01 <amiller> i've been writing opencl code for graphics projects and realtime computer vision for several years
651 2011-03-01 03:23:58 <luke-jr> amiller: could you improve ArtForz's kernel in poclbm? :P
652 2011-03-01 03:24:29 <quellhorst> TheKid: do i need to do that for a mining rig?
653 2011-03-01 03:24:47 <ArtForz> amiller: several years? opencl has only been out for like 2 ...
654 2011-03-01 03:24:53 <luke-jr> lol
655 2011-03-01 03:25:02 <amiller> well, glsl and cuda before that
656 2011-03-01 03:25:05 <amiller> but opencl made it way easier
657 2011-03-01 03:25:12 <TheKid> quellhorst: many people think that it ruins mining results
658 2011-03-01 03:25:19 <TheKid> I haven't ever had a problem with it
659 2011-03-01 03:25:22 <TheKid> but no, you don't
660 2011-03-01 03:26:01 <luke-jr> amiller: basically, my thinking is this: I know that this Radeon can take down my system easily, so there's more risk than an ordinary user account; but if I might gain significant MH/s extra, it might be an ok risk to take
661 2011-03-01 03:26:54 <amiller> by take down your system, you mean by overuse and heating
662 2011-03-01 03:27:09 <mmarker> le sigh. The pools should be asking for proof of work at the easiest level, no?
663 2011-03-01 03:27:18 <TheKid> mmarker: i BELIEVE SO
664 2011-03-01 03:27:20 <TheKid> whoops caps
665 2011-03-01 03:27:22 <amiller> do you have something that watches your gpu temp and shuts down your machine before something like that would happen
666 2011-03-01 03:27:34 <amiller> or shuts down the user's processes anyway
667 2011-03-01 03:27:58 <luke-jr> amiller: or crash
668 2011-03-01 03:28:05 <luke-jr> no
669 2011-03-01 03:28:19 <mmarker> hmm, what's to bot command to see average time for a difficulty? pushing 24000 khash/sec
670 2011-03-01 03:28:24 <mmarker> want to see how long it'll take
671 2011-03-01 03:28:27 <amiller> luke-jr, how expensive is a crash to you, is this a server that would just boot back up right away
672 2011-03-01 03:28:47 <amiller> i crash my system somewhat frequently when i'm tinkering with opencl
673 2011-03-01 03:28:51 <ArtForz> ;;bc,calcd 24000 1
674 2011-03-01 03:28:52 <gribble> The average time to generate a block at 24000 Khps, given the supplied difficulty of 1, is 2 minutes and 58 seconds
675 2011-03-01 03:28:52 <luke-jr> amiller: it's my desktop system.
676 2011-03-01 03:29:06 <mmarker> ArtForz: Yea, something is wrong
677 2011-03-01 03:29:14 <mmarker> No proof of work yet
678 2011-03-01 03:29:22 <luke-jr> amiller: cost of a crash varies.
679 2011-03-01 03:29:29 <mmarker> Give it until 11:30
680 2011-03-01 03:29:42 <luke-jr> accidentally mining on my Intel GPU is almost certain to crash it
681 2011-03-01 03:29:51 <amiller> luke-jr, well what would you want out of it, do you already run pobclm and it crashes your computer
682 2011-03-01 03:30:14 <luke-jr> poclbm only crashes me if it picks my Intel GPU
683 2011-03-01 03:30:33 <luke-jr> but I have a fairly consistent mining style right now
684 2011-03-01 03:30:38 <mmarker> Ok, I seriously can't be this unlucky
685 2011-03-01 03:30:41 <luke-jr> when I was setting things up, it crashed a bit
686 2011-03-01 03:33:25 <amiller> i could be pretty careful not to pick the intel gpu, i would only use pyopencl - i'd still probably crash it in another way
687 2011-03-01 03:33:25 <luke-jr> XD
688 2011-03-01 03:34:04 <mmarker> Ok, I really can't be this unlucky!
689 2011-03-01 03:35:51 <amiller> luke-jr, hard to tell, it's simple code
690 2011-03-01 03:36:07 <bk128> can anyone give me a quick idea of how to use screen so I don't have to keep my ssh session open to keep diablominer running?
691 2011-03-01 03:36:12 <amiller> luke-jr, not any apparent computation redundancy, although the code itself could be simplified by metaprogramming
692 2011-03-01 03:37:03 <amiller> bk128, http://www.kuro5hin.org/story/2004/3/9/16838/14935
693 2011-03-01 03:37:13 <bk128> amiller: thanks
694 2011-03-01 03:37:23 <amiller> i vouch for that guide, i've followed it and it is easy enough
695 2011-03-01 03:39:38 <lfm> bk128: you dont need screen, just use "&"
696 2011-03-01 03:40:14 <bk128> lfm: am I ever able to view the output then? khash/sec and blocks found?
697 2011-03-01 03:41:00 <lfm> bk128: depends, if you redirect it to a log file then you can always look in the log file
698 2011-03-01 03:41:02 <amiller> logging out stops backgrounded jobs anyway, lfm
699 2011-03-01 03:41:22 <mmarker> Ok, I give up for now
700 2011-03-01 03:41:27 <lfm> amiller: not always
701 2011-03-01 03:41:28 <mmarker> Tired, head isn't clear
702 2011-03-01 03:43:13 <quellhorst> with an atx 24 pin power connector, is it also required to plug in an atx 4-pin? it says the 4 pin is used to power the cpu.
703 2011-03-01 03:43:21 <quellhorst> my power suppoly doesn't seem to have a 4 pin
704 2011-03-01 03:43:59 <ArtForz> yes
705 2011-03-01 03:44:09 <quellhorst> yes its required?
706 2011-03-01 03:44:10 <TheKid> quellhorst: it has a 4 pin
707 2011-03-01 03:44:18 <TheKid> and yes it's required, of course
708 2011-03-01 03:44:21 <ArtForz> a psu without a 4pin would be... really weird
709 2011-03-01 03:44:24 <quellhorst> man, wonder why i don't see it
710 2011-03-01 03:44:31 <TheKid> it may be an 8 pin that seperates into dual 4 pin
711 2011-03-01 03:44:35 <TheKid> that is increasingly common
712 2011-03-01 03:44:39 <ArtForz> yup
713 2011-03-01 03:44:55 <quellhorst> yes there is an 8 pin
714 2011-03-01 03:45:39 <quellhorst> so i guess i need an 8 to 4 pin adapter?
715 2011-03-01 03:45:41 <lfm> quellhorst: sounds like you need a psu update
716 2011-03-01 03:47:43 <quellhorst> lfm: just got this on in new
717 2011-03-01 03:47:55 <TheKid> quellhorst: plug in half of the 8 pin
718 2011-03-01 03:49:02 <ArtForz> btw, good ref page: http://www.playtool.com/pages/psuconnectors/connectors.html
719 2011-03-01 03:49:04 <lfm> quellhorst: seems you go tthe wrong one
720 2011-03-01 03:49:12 <validus> lots of cards ship with proper plugins and attachments for a wide variety of things so ppl dont have to keep ordering diff converters
721 2011-03-01 03:49:17 <validus> your manual should tell you
722 2011-03-01 03:49:22 <bk128> quellhorst: sure the 8 pin doesnt separate?
723 2011-03-01 03:49:27 <bk128> to 2 4-pin connectors?
724 2011-03-01 03:50:49 <quellhorst> weird, this motherboard has a detachable 4 pin on the 20 pin connector
725 2011-03-01 03:51:13 <quellhorst> oh right. it does look like this 8 pin will come apart
726 2011-03-01 03:51:20 <lfm> ya its called a 24 pin
727 2011-03-01 03:52:53 <ArtForz> and I'm guessing next iteration of ATX standard will either introduce a 28-pin ATX connector or another 4 or 6-pin 12V connector
728 2011-03-01 03:53:00 <quellhorst> been too long since i did a custom build... never actually had high power vid cards
729 2011-03-01 03:53:13 <quellhorst> mostly did server builds
730 2011-03-01 03:54:01 <ArtForz> up to 7 PCIe slots, each with a max allowed 12V current draw of 6.25A ... from 2 12V pins on the ATX connector
731 2011-03-01 03:54:21 <ArtForz> errr... 3.125A
732 2011-03-01 03:54:45 <ArtForz> 6.25 is only for double-wide cards
733 2011-03-01 03:55:34 <ArtForz> thats > 21A over 2 pins rated at 6A each ...
734 2011-03-01 03:59:15 <ArtForz> mainboard makers are already adding 5.25" molexes, PCIe power receptables or a 2nd EPS12V on highend gamer boards to get around that
735 2011-03-01 03:59:53 <lfm> some mb already have the 4pin cpu power upped to a 8 pin
736 2011-03-01 04:00:02 <ArtForz> well, they need that
737 2011-03-01 04:00:27 <ArtForz> 4pin is only rated for 8A = 96W
738 2011-03-01 04:00:42 <ArtForz> though the connector can handle twice that
739 2011-03-01 04:01:57 <ArtForz> same shit as PCIe 6-pin, connector rated for > 250W, spec says 75W
740 2011-03-01 04:03:05 <AAA_awright> There needs to be a Bitcoin Convention
741 2011-03-01 04:03:24 <AAA_awright> With talks and speeches and products and innovation
742 2011-03-01 04:03:31 <AAA_awright> Things I can hold in my hands
743 2011-03-01 04:03:43 <AAA_awright> Kiba!
744 2011-03-01 04:03:51 <AAA_awright> >:(
745 2011-03-01 04:06:24 <jgarzik> <gribble> BC |     TRADE|LRUSD                1000.0 @ $0.95
746 2011-03-01 04:06:30 <jgarzik> nice.  glad to see volume @ bitcoin-central
747 2011-03-01 04:07:16 <lfm> those pci specs are more to do with traces than the plugs I spoze
748 2011-03-01 04:07:35 <ArtForz> no clue really
749 2011-03-01 04:07:56 <ArtForz> because the PCIe 8-pin isn't underrated quite as heavily
750 2011-03-01 04:08:36 <ArtForz> ahhh... I see why
751 2011-03-01 04:08:57 <ArtForz> 6-pin uses conenctors rated for 8A/pin, and only has 2 required +12 pins
752 2011-03-01 04:09:01 <bk128> is there a way to reattach to screen automatically when I ssh in?
753 2011-03-01 04:09:18 <ArtForz> kinda crazy, that'd *still* be 192W
754 2011-03-01 04:09:56 <ArtForz> PCIe 8pin requires 3 +12 lines and terminals rated for 13A/pin ...
755 2011-03-01 04:10:42 <ArtForz> so overrated with a similar crazy safety factor
756 2011-03-01 04:10:59 <ArtForz> 468W connector rating for a 150W rated standard...
757 2011-03-01 04:11:12 <bk128> overrated because they know the psu manufacturers are as cheap with everything as possible
758 2011-03-01 04:11:26 <amiller> has anyone looked at this, is there a reason it hasn't been followed up on since last year  http://www.bitcoin.org/smf/index.php?topic=847.20
759 2011-03-01 04:11:38 <lfm> they assume it will be assembled by overworked suicidal kids
760 2011-03-01 04:12:47 <ArtForz> guess they didnt want to repeat the problems seen with the ATX 20/24pin ...
761 2011-03-01 04:13:16 <bk128> I don't think i've ever seen a nicely soldered psu.  mainly where they bundle all the 12v wires and gnd wires together and solder them to the board
762 2011-03-01 04:14:05 <bk128> I've had a cheap apc surge protector where the hot wire was stripped and just floating around inside the case.  must have forgotten to solder that one :)
763 2011-03-01 04:16:41 <jgarzik> wow, major buying on mtgox
764 2011-03-01 04:17:24 <bk128> price is going up
765 2011-03-01 04:25:25 <quellhorst> whew, just got my first rig built
766 2011-03-01 04:25:35 <quellhorst> is there a linux iso available?
767 2011-03-01 04:26:02 <quellhorst> amd processor + hd 5870
768 2011-03-01 04:26:20 <lfm> lots
769 2011-03-01 04:27:21 <amiller> bitcoin is really rewarding to try as a 'toy currency'
770 2011-03-01 04:27:28 <amiller> there are all sorts of neat things to play with
771 2011-03-01 04:27:52 <amiller> i would feel very proud giving friends a bitcoin gift
772 2011-03-01 04:28:15 <amiller> i guess a lot of people alraedy agree that that's something it's already very good for
773 2011-03-01 04:28:36 <quellhorst> lfm: where do i find them?
774 2011-03-01 04:28:58 <bk128> quellhorst: no specific bitcoin distro.  a lot of people use ubuntu or debian though
775 2011-03-01 04:29:19 <quellhorst> ok, so i should just get ubuntu desktop 10.10?
776 2011-03-01 04:29:31 <amiller> that's what i'm using!
777 2011-03-01 04:29:36 <quellhorst> and, its ok to boot off usb? :)
778 2011-03-01 04:29:41 <quellhorst> thinking of using a usb stick
779 2011-03-01 04:29:49 <bk128> eh, better to install if you're going to be using it for mining a lot
780 2011-03-01 04:30:00 <quellhorst> ok
781 2011-03-01 04:30:01 <bk128> don't want to have to reconfigure everything after a reboot
782 2011-03-01 04:30:03 <amiller> i bet it wouldn't make a big difference
783 2011-03-01 04:30:19 <bk128> performance, none
784 2011-03-01 04:30:20 <lfm> ya, you should be able to boot off a usb cd
785 2011-03-01 04:30:31 <bk128> think he means a flash drive
786 2011-03-01 04:30:34 <amiller> you can have a readwrite usb drive can't you
787 2011-03-01 04:30:52 <quellhorst> yeah
788 2011-03-01 04:31:17 <bk128> it'd be slow though
789 2011-03-01 04:31:27 <lfm> ya flash drives work too I hear
790 2011-03-01 04:31:31 <amiller> it's not like much disk space is used for mining
791 2011-03-01 04:31:35 <bk128> use an old hard drive if you have one
792 2011-03-01 04:31:36 <amiller> or disk i/o
793 2011-03-01 04:31:56 <amiller> actually it seems like there's no disk i/o at all
794 2011-03-01 04:32:06 <lfm> there is some
795 2011-03-01 04:32:22 <quellhorst> i guess it would suck if your disk died... you loose your coins?
796 2011-03-01 04:32:32 <bk128> quellhorst: not if you mine in a pool
797 2011-03-01 04:32:40 <bk128> if not, always backup your wallet
798 2011-03-01 04:32:46 <lfm> quellhorst: ya, wallet backups are good idea
799 2011-03-01 04:32:48 <quellhorst> also, for the hd 5870 what SW do i need to install?
800 2011-03-01 04:32:55 <bk128> every time you generate or receive
801 2011-03-01 04:33:03 <quellhorst> can i setup my own pool?
802 2011-03-01 04:33:07 <bk128> on linux, use fglrx and stream sdk 2.1
803 2011-03-01 04:34:57 <lfm> quellhorst: you can just mine for yourself, no actual pool needed
804 2011-03-01 04:35:50 <quellhorst> lfm: but if i had multiple systems?
805 2011-03-01 04:36:00 <quellhorst> bk128: what about the bitcoin client to run?
806 2011-03-01 04:36:24 <bk128> quellhorst: pools are for multiple users pooling their bitcoins together and then splitting the coins generated
807 2011-03-01 04:36:28 <lfm> run the standard client as a server with any miner
808 2011-03-01 04:36:47 <quellhorst> ahh
809 2011-03-01 04:37:28 <quellhorst> and as for the gpu mining. i thought you needed a special version of the client for that to work.
810 2011-03-01 04:37:37 <bk128> quellhorst: make sure you make a bitcoin.conf file with your username and password
811 2011-03-01 04:38:15 <lfm> quellhorst: you just need a gpu miner, they all work with the server mode
812 2011-03-01 04:40:38 <quellhorst> lfm: is there a wiki page listing them?
813 2011-03-01 04:41:48 <lfm> not sure
814 2011-03-01 04:43:12 <bk128> quellhorst: look at http://www.bitcoin.org/smf/index.php?topic=2636
815 2011-03-01 04:43:27 <bk128> or you can use DiabloMiner https://www.bitcoin.org/smf/index.php?topic=1721.100
816 2011-03-01 04:46:19 <bk128> quellhorst: sorry, second link is https://www.bitcoin.org/smf/index.php?topic=1721.0
817 2011-03-01 04:47:26 <quellhorst> hmm. does it matter if i do ubuntu server or desktop?
818 2011-03-01 04:47:45 <quellhorst> found a usb stick with ubuntu server iso alread on it
819 2011-03-01 04:51:02 <bk128> you need x, desktop might be easier but you can always just apt-get install ubuntu-desktop
820 2011-03-01 04:51:08 <bk128> I think
821 2011-03-01 05:18:02 <bk128> is the queue still full of tx's?
822 2011-03-01 05:18:50 <lfm> ya I got 105 in queue
823 2011-03-01 05:19:07 <lfm> 106
824 2011-03-01 05:19:18 <bk128> not as bad
825 2011-03-01 05:19:19 <Blitzboom> one of them is mine, lol
826 2011-03-01 05:19:38 <lfm> not sure, I restarted so I might not have them all
827 2011-03-01 05:19:46 <Blitzboom> when did you restart?
828 2011-03-01 05:20:18 <lfm> when I put in the patch to display the queue size
829 2011-03-01 05:21:11 <lfm> abou an hour ago it seems
830 2011-03-01 05:21:28 <Blitzboom> my transaction is & over ten hours old
831 2011-03-01 05:21:38 <lfm> yup
832 2011-03-01 05:26:56 <amiller> i can't tell if my transactions are making it through yet either
833 2011-03-01 05:27:00 <amiller> glacial!
834 2011-03-01 05:40:32 <quellhorst> cooler master cases sure are neat
835 2011-03-01 05:40:50 <lfm> quellhorst: I like em too
836 2011-03-01 05:40:56 <bk128> quellhorst: i like my lian-li case
837 2011-03-01 05:41:19 <lfm> I got 3 of the elite 330 models
838 2011-03-01 05:43:52 <bk128> http://www.tomshardware.com/reviews/midi-towers,1011-24.html
839 2011-03-01 05:43:56 <bk128> that's what I have now
840 2011-03-01 05:44:13 <quellhorst> i have a 912 haf
841 2011-03-01 05:44:50 <bk128> havent used a cooler master case in a while
842 2011-03-01 05:45:28 <lfm> http://www.memoryexpress.com/Products/PID-MX14200%28ME%29.aspx
843 2011-03-01 05:45:32 <quellhorst> bk128: haha, that has the equiv of a cold air intake for the cpu
844 2011-03-01 05:45:57 <bk128> http://www.newegg.com/Product/Product.aspx?Item=N82E16811112264
845 2011-03-01 05:45:59 <quellhorst> bk128: what/how many cards are you running on that?
846 2011-03-01 05:46:34 <bk128> quellhorst: no, my bitcoin miner is built in a mdf box http://imgur.com/a/xxALX
847 2011-03-01 05:46:57 <quellhorst> when that new gigabyte board comes out... would like to see how many cards i could get in. you need some type of pcie extender cables?
848 2011-03-01 05:47:20 <bk128> yeah, you dont want to stack the cards too close
849 2011-03-01 05:47:47 <quellhorst> bk128: mdf? = particle board? :)
850 2011-03-01 05:47:53 <bk128> quellhorst: this is ArtForz's http://bayimg.com/KAAeaaAdp
851 2011-03-01 05:48:09 <quellhorst> lol
852 2011-03-01 05:48:13 <bk128> mdf=medium density fiberboard
853 2011-03-01 05:48:22 <quellhorst> what power supplies are those?
854 2011-03-01 05:48:35 <bk128> I think he uses 2 1500watt psu's
855 2011-03-01 05:48:49 <quellhorst> i'm pretty sure i see 2x 1000 watt
856 2011-03-01 05:49:01 <quellhorst> zoom in http://image.bayimg.com/kaaeaaadp.jpg
857 2011-03-01 05:49:15 <quellhorst> super howler 1000w i think
858 2011-03-01 05:49:31 <bk128> oh, one is a "super flower" brand
859 2011-03-01 05:49:36 <bk128> europe only I think
860 2011-03-01 05:50:06 <bk128> well, maybe 1 is a 1500watt that powers 2 cards, the mobo, and hdd, and the other 1000w just powers the other 2 cards
861 2011-03-01 05:50:18 <bk128> cant remember
862 2011-03-01 05:50:22 <quellhorst> might try custom next rig
863 2011-03-01 05:50:58 <quellhorst> but if im gonna be doing 5870s... might need to do more than 4 cards/ rig?
864 2011-03-01 05:51:03 <lfm> ==noname
865 2011-03-01 05:51:08 <bk128> you might be able to.  haven't seen it
866 2011-03-01 05:51:19 <bk128> it's kinda tricky to use 2 psu's.  you have to pull the 12v lines off the pci-e extender and attach them directly to the psu
867 2011-03-01 05:53:25 <ArtForz> 2*1000
868 2011-03-01 05:53:41 <ArtForz> and thats only because I used to run the cards overvolted to near-5870 levels
869 2011-03-01 05:54:24 <bk128> ArtForz: think you could stack 6 5870's on one motherboard w/ extenders?
870 2011-03-01 05:54:45 <ArtForz> in theory yes, in practice fglrx croaks
871 2011-03-01 05:54:50 <bk128> oh
872 2011-03-01 05:54:54 <ArtForz> see mrbs blog
873 2011-03-01 05:55:37 <ArtForz> AMD says it might be fixed in fglrx 11.2, but I dont really want to try
874 2011-03-01 05:56:25 <ArtForz> for 4*5970 @ stock V, 2*800W should be plenty
875 2011-03-01 05:56:39 <bk128> ok.  do you think the 6990's will ever be faster than the 5970's?
876 2011-03-01 05:56:48 <ArtForz> for mining? no.
877 2011-03-01 05:56:48 <hwolf> yep
878 2011-03-01 05:56:57 <ArtForz> for 3D? hell yes.
879 2011-03-01 05:57:05 <hwolf> why not for mining?
880 2011-03-01 05:57:08 <bk128> probably going to wait for those since they'll have a decent resale value
881 2011-03-01 05:57:21 <bk128> hwolf: less shaders than the 5900 series
882 2011-03-01 05:57:40 <ArtForz> mainly because 6950/70 isnt faster than 5850/70, but uses more power
883 2011-03-01 05:58:04 <hwolf> 6990 has more shaders then 5970 from specs I read
884 2011-03-01 05:58:06 <ArtForz> = 6990 will be even more limited by the 300W wall than 5970
885 2011-03-01 05:58:08 <ArtForz> errr... no
886 2011-03-01 05:58:31 <ArtForz> 5970 has 2*1600, 6990 will probably have 2*1536
887 2011-03-01 05:58:42 <[Tycho]> ArtForz, how much RAM is in your boxes ?
888 2011-03-01 05:58:49 <ArtForz> 1GB each
889 2011-03-01 05:59:09 <hwolf> 6990 specs are 3840 right?
890 2011-03-01 05:59:16 <hwolf> 2*1900 something
891 2011-03-01 05:59:23 <ArtForz> errr... not as far as I know
892 2011-03-01 05:59:40 <ArtForz> 6990 probably wont be more than 2*6970 shaderwise
893 2011-03-01 05:59:53 <ArtForz> which would be 2*1536
894 2011-03-01 05:59:59 <hwolf> guess I was going by leaked spec (2*1920)
895 2011-03-01 06:00:20 <hwolf> might be hype to mess with nvidia or something
896 2011-03-01 06:00:26 <ArtForz> iirc the 1920 was what originally was planned for the 32nm chip
897 2011-03-01 06:01:38 <hwolf> i talked with somebody from xfx yesterday and they said there will be very limited supply of 6990 that they can get their hands on.
898 2011-03-01 06:01:44 <ArtForz> yes
899 2011-03-01 06:01:53 <hwolf> probably way over priced also
900 2011-03-01 06:01:56 <ArtForz> basically same shit as when 58xx/59xx was launched
901 2011-03-01 06:02:17 <ArtForz> TSMC/AMD can't supply enough chips
902 2011-03-01 06:03:27 <ArtForz> so basically once the stock built up for launch is sold off, supply will be very thin
903 2011-03-01 06:04:13 <ArtForz> 6990 should be pretty awesome for 3D though
904 2011-03-01 06:04:25 <ArtForz> 69xx scales way better in crossfire than 58xx
905 2011-03-01 06:05:06 <ArtForz> and thanks to the dynamic power limiting they can run the cards way closer to the 300W limit
906 2011-03-01 06:06:02 <quellhorst> hmm, didn't realize without the gfx cards even doing much, they will put out a good amount of heat
907 2011-03-01 06:06:23 <ArtForz> 5970 is ~ 300W in furmark, more like 200-230 for 3D stuff
908 2011-03-01 06:06:34 <hwolf> so lets say you wanted to have tsmc make you a bunch of specialized chips.  do they have a pretty big min order requirement?
909 2011-03-01 06:06:54 <ArtForz> errr... kinda
910 2011-03-01 06:07:25 <[Tycho]> dynamic power limiting ?
911 2011-03-01 06:07:26 <ArtForz> they'll be happy to run a single wafer for you, you still need a full mask set
912 2011-03-01 06:07:54 <quellhorst> are there any new cards in the pipeline that will be better for mining than the current cards?
913 2011-03-01 06:07:56 <ArtForz> which is well >$1M at 45nm
914 2011-03-01 06:08:10 <bk128> is tsmc their foundry?
915 2011-03-01 06:08:21 <pogden> $300,000 for crude processes
916 2011-03-01 06:08:42 <ArtForz> eh? $300k? at 45nm?
917 2011-03-01 06:08:54 <pogden> no
918 2011-03-01 06:09:07 <pogden> like >1micron
919 2011-03-01 06:09:20 <pogden> crude processes
920 2011-03-01 06:09:27 <ArtForz> that actually sounds a bit high for >1u stuff
921 2011-03-01 06:09:53 <lfm> how old is your info?
922 2011-03-01 06:12:01 <ArtForz> guess it depends a lot on what kind of process you want
923 2011-03-01 06:12:39 <quellhorst> lucked out. system powered on and ubuntu is installing first time around.
924 2011-03-01 06:12:40 <bk128> what process are the structured asics on?
925 2011-03-01 06:12:47 <ArtForz> 65nm
926 2011-03-01 06:12:53 <bk128> not too bad
927 2011-03-01 06:13:31 <ArtForz> yup
928 2011-03-01 06:13:43 <ArtForz> only drawback is you still have massive losses in routing
929 2011-03-01 06:14:03 <quellhorst> why did they stop making the 5970?
930 2011-03-01 06:14:17 <ArtForz> erm, because they stopped making cypress chips
931 2011-03-01 06:14:49 <ArtForz> because they need the capacity to make 6xxxs
932 2011-03-01 06:14:52 <quellhorst> does that mean the 5870 will have the same fate?
933 2011-03-01 06:14:57 <ArtForz> yep
934 2011-03-01 06:15:00 <quellhorst> damn
935 2011-03-01 06:15:04 <quellhorst> then whats the next card to get?
936 2011-03-01 06:15:32 <ArtForz> my guess is 5750/70 will stay available for a while
937 2011-03-01 06:15:58 <ArtForz> probably a 69xx
938 2011-03-01 06:16:10 <ArtForz> depends mainly on pricing
939 2011-03-01 06:17:33 <ArtForz> or just grab used 5970s on ebay after 6990 launches ;)
940 2011-03-01 06:18:35 <lfm> you guys will prolly be bidding against each other. I wonder how high you'll go! grin
941 2011-03-01 06:18:51 <[Tycho]> Lots of 5970 will be available cheaply after bitcoin crash :)
942 2011-03-01 06:20:14 <quellhorst> haha