1 2010-11-26 00:10:29 <BostX> hi
  2 2010-11-26 00:11:17 <BostX> does anyone know where I can find the info who sent me a bc?
  3 2010-11-26 00:15:51 <OneFixt> BostX: you can't, that's the part what is anonymous
  4 2010-11-26 00:15:59 <OneFixt> that is*
  5 2010-11-26 00:16:24 <gribble> Error: "bc," is not a valid command.
  6 2010-11-26 00:16:24 <OneFixt> ;;bc, stats
  7 2010-11-26 00:16:28 <OneFixt> ;;bc,stats
  8 2010-11-26 00:16:30 <gribble> Current Blocks: 93851 | Current Difficulty: 6866.89864897 | Next Difficulty At Block: 94752 | Next Difficulty In: 901 blocks | Next Difficulty In About: 5 days, 20 hours, 55 minutes, and 42 seconds | Next Difficulty Estimate: 7317.04556505
  9 2010-11-26 00:16:31 <BostX> OneFixt, hmm :(
 10 2010-11-26 00:16:48 <OneFixt> you might want to ask around if you have an idea of who it might have been
 11 2010-11-26 00:17:56 <BostX> OneFixt, well, it was me... im getting to know what bc is and how it works. I sent 0.01 from one comp to another
 12 2010-11-26 00:18:13 <BostX> OneFixt, and now I want to send it back but...
 13 2010-11-26 00:18:38 <OneFixt> did it show up on the other computer?
 14 2010-11-26 00:18:41 <BostX> OneFixt, it's an insane to type the whole address on a command line ...
 15 2010-11-26 00:18:51 <OneFixt> oh, copy paste if you can
 16 2010-11-26 00:18:54 <BostX> OneFixt, yea that part works
 17 2010-11-26 00:19:24 <OneFixt> there's no option to "return" coins to whoever sent it
 18 2010-11-26 00:19:29 <BostX> OneFixt, well it took several hours to catch all the 90 000 blocks but finally it arrived
 19 2010-11-26 00:19:37 <OneFixt> =) welcome
 20 2010-11-26 00:19:49 <BostX> OneFixt, :) yea :)
 21 2010-11-26 00:20:11 <BostX> OneFixt, but the addresses is a pain...
 22 2010-11-26 00:20:15 <ByteCoin> OneFixt: In principle, all the information is there to allow you to return coins if you wish..
 23 2010-11-26 00:20:33 <BostX> ByteCoin, ? where?
 24 2010-11-26 00:20:50 <ByteCoin> BostX: Tell me the address that was credited with the coins and I'll tell you where they came from.
 25 2010-11-26 00:21:00 <ByteCoin> Then I'll tell you how to find out yourself
 26 2010-11-26 00:21:25 <BostX> and another thing - everytime I reboot or reconnect to inet I get yet another address...
 27 2010-11-26 00:23:03 <BostX> ByteCoin, ? do u wanna know the receiver's or sender's address
 28 2010-11-26 00:23:45 <ByteCoin> BostX: I thought you wanted to return them. Presumably the sender's address is what you're trying to find out?
 29 2010-11-26 00:23:57 <BostX> ByteCoin, right :)
 30 2010-11-26 00:24:03 <ByteCoin> I presume the receiver's address is you...
 31 2010-11-26 00:24:12 <ByteCoin> One of your addresses
 32 2010-11-26 00:24:45 <ByteCoin> Funnily enough, the information that you must tell me is the information you have rather than the stuff you don't know.
 33 2010-11-26 00:25:59 <BostX> well I sent the coin to ..... grrr, i cannot copy&paste it :(
 34 2010-11-26 00:26:06 <OneFixt> ByteCoin: true, technically
 35 2010-11-26 00:26:15 <ByteCoin> Best kind of true
 36 2010-11-26 00:26:19 <OneFixt> =)
 37 2010-11-26 00:26:35 <OneFixt> just not guaranteed that the person will still have the address etc.
 38 2010-11-26 00:26:39 <OneFixt> but in this case it should be fine
 39 2010-11-26 00:26:48 <ByteCoin> I only need six or so characters from the middle
 40 2010-11-26 00:26:54 <ByteCoin> No need to type the whole thing...
 41 2010-11-26 00:26:58 <ByteCoin> Of your address
 42 2010-11-26 00:27:46 <ByteCoin> You might as well tell me the amount transferred so I can check that the info is consistent
 43 2010-11-26 00:27:59 <ByteCoin> What's the delay?
 44 2010-11-26 00:28:38 <BostX> ok, the  receiver was 1PyDFXTSBxHE8Tu7BBRpzVoVmhvob8xzdZ
 45 2010-11-26 00:28:46 <BostX> 0.01
 46 2010-11-26 00:28:52 <ByteCoin> ok. wait a bit
 47 2010-11-26 00:29:04 <BostX> ByteCoin, how u gonna do it?
 48 2010-11-26 00:30:00 <BostX> OneFixt, no no I still keep the 0.01
 49 2010-11-26 00:30:02 <ByteCoin> Sent from 1EqZYt8zortMyMj2DLDA9r46xUVButWdWP
 50 2010-11-26 00:30:16 <BostX> ByteCoin, RIGHT!
 51 2010-11-26 00:30:26 <BostX> ByteCoin, how did you find it?
 52 2010-11-26 00:30:43 <ByteCoin> Go to https://blockexplorer.com/ and type part of the address in the search box
 53 2010-11-26 00:32:04 <BostX> ByteCoin, thx! but is it possible to get this info somehow on my local comp?
 54 2010-11-26 00:32:45 <BostX> ByteCoin, u know blockexplorer.com might get hacked
 55 2010-11-26 00:32:53 <ByteCoin> The raw info is on your local computer in blk0001.dat. The trick is to interpret it.
 56 2010-11-26 00:33:04 <ByteCoin> Oh, that is if you're running bitcoin
 57 2010-11-26 00:33:24 <BostX> ByteCoin, yea I do... leme c it
 58 2010-11-26 00:33:57 <ByteCoin> I could do it using my own software but it's not set up for it. blk0001.dat is a data file.
 59 2010-11-26 00:34:11 <ByteCoin> You can't open it in notepad etc..
 60 2010-11-26 00:34:34 <ByteCoin> blockexplorer will be fine.
 61 2010-11-26 00:34:56 <BostX> ByteCoin, well I tried 'head blk0001.dat'...
 62 2010-11-26 00:35:22 <ByteCoin> Unless you're a programmer, don't get involvedf
 63 2010-11-26 00:35:42 <BostX> that's the concent of my blk0001.dat
 64 2010-11-26 00:35:47 <BostX> yea Im
 65 2010-11-26 00:41:18 <BostX> btw what mean the '?' if I take look at the transactions on the blockexplorer?
 66 2010-11-26 00:41:43 <ByteCoin> It's a bit pants... It just facilitates mouseover information.
 67 2010-11-26 00:44:41 <BostX> oh that works really miserable... leme check it how it works in fireforx
 68 2010-11-26 00:44:48 <BostX> i use chrome
 69 2010-11-26 00:45:51 <BostX> firefox shows it better but it still has some potential for improvement
 70 2010-11-26 00:46:27 <ByteCoin> forum post theymos with improvements. Search bitcoin block explorer thread
 71 2010-11-26 00:47:52 <OneFixt> ByteCoin: could you help me solve something?
 72 2010-11-26 00:47:56 <OneFixt> http://www.blockexplorer.com/address/18LC8TVQ9EokM5i4aKH1DdJzSTB729AkAD
 73 2010-11-26 00:48:03 <BostX> uhm... I gotta go to bed now... c u later and thx!
 74 2010-11-26 00:48:28 <OneFixt> ByteCoin: I had 100 coins at this address - sent 50 and 50 one right after the other, and I ended up only receiving 50 instead of the 100.
 75 2010-11-26 00:48:31 <ByteCoin> I see it
 76 2010-11-26 00:48:41 <OneFixt> My bitcoind says balance is 0, so I can't re-send.
 77 2010-11-26 00:48:48 <OneFixt> Block explorer still says balance is 50.
 78 2010-11-26 00:49:43 <ByteCoin> Try closing bitcoind and restarting it.
 79 2010-11-26 00:49:53 <OneFixt> Ok, I'll try.
 80 2010-11-26 00:50:40 <OneFixt> Didn't help.
 81 2010-11-26 00:50:56 <ByteCoin> Hmm...
 82 2010-11-26 00:51:00 <OneFixt> My hunch is that I may have to re-download the blocks... but that's a pain.
 83 2010-11-26 00:51:18 <ByteCoin> It's worth a forum post perhaps...
 84 2010-11-26 00:51:20 <OneFixt> Or can bitcoind re-scan them to calculate the balance?
 85 2010-11-26 00:51:29 <ByteCoin> Put where and when you sent them
 86 2010-11-26 00:51:50 <ByteCoin> I'm not a great expert on the current software
 87 2010-11-26 00:52:29 <OneFixt> ok, thanks
 88 2010-11-26 00:52:55 <OneFixt> I am going to try one thing first - copying the wallet to a different location and re-downloading the blocks to see if it finds the coins.
 89 2010-11-26 00:53:57 <ByteCoin> That's a good first step.
 90 2010-11-26 00:54:29 <ByteCoin> Might be worth shutting down the bitcoind that sent the coins while the other new client is running
 91 2010-11-26 00:54:39 <ByteCoin> The old might contaminate the new
 92 2010-11-26 00:55:14 <OneFixt> I was thinking of doing it on another machine.
 93 2010-11-26 00:55:32 <ByteCoin> I know. Nevertheless, shut down the other bitcoind
 94 2010-11-26 00:56:00 <ByteCoin> It might get contaminated over the network.
 95 2010-11-26 00:56:17 <ByteCoin> It's somewhat unlikely but better to rule it out.
 96 2010-11-26 00:57:25 <OneFixt> hm... satoshi wrote something like this on the forum:
 97 2010-11-26 01:38:48 <tcatm> Kiba: ping?
 98 2010-11-26 01:40:26 <Kiba> bing
 99 2010-11-26 01:41:08 <tcatm> I can send you the xcf file
100 2010-11-26 01:41:47 <Kiba> tcatm: thanks.
101 2010-11-26 01:43:06 <Kiba> tcatm: hackerkiba@gmail.com
102 2010-11-26 01:44:13 <tcatm> on the way
103 2010-11-26 01:49:17 <Kiba> thanks
104 2010-11-26 02:02:47 <OneFixt> ByteCoin: Re-downloading the blocks didn't get the coins back =/
105 2010-11-26 02:06:40 <OneFixt> yikes, I think I had 2 bitcoin instances, on different computers, start out with the same empty wallets
106 2010-11-26 02:06:49 <OneFixt> but those wallets had keys...
107 2010-11-26 02:07:14 <OneFixt> and now when I send from one wallet, coins disappear in the other, or something like that
108 2010-11-26 02:11:20 <ByteCoin> OneFixt: I can't help you. Write everything in a forum post and they'll sort it out.
109 2010-11-26 02:12:03 <ByteCoin> The good news is that the block chain thinks you still have the coins.
110 2010-11-26 02:13:03 <OneFixt> I had another strange disappearance just happen
111 2010-11-26 02:13:23 <ByteCoin> Ooops. I tell a lie. The block chain says you've sent both to 1EWbkLgfzpqqYW1dd2dhbPrwQJYFTioUpK
112 2010-11-26 02:14:21 <ByteCoin> The last block of 50 was only incorporated into the chain about 10 minutes ago
113 2010-11-26 02:14:58 <ByteCoin> I wonder if this is because of the current spam flood.
114 2010-11-26 02:15:48 <OneFixt> Now that's interesting.. because I sent it from a different bitcoind, on a different machine.
115 2010-11-26 02:16:09 <OneFixt> oh.. let me see if it could be spam
116 2010-11-26 02:16:40 <OneFixt> there doesn't seem to be spam currently (but there was a 50k transaction)
117 2010-11-26 02:17:10 <ByteCoin> I assure you, there's spam currenty
118 2010-11-26 02:17:24 <ByteCoin> currenty spam! yum!
119 2010-11-26 02:17:57 <OneFixt> What seems to have happened, is that I messed up my bitcoind installation by cloning it
120 2010-11-26 02:18:01 <OneFixt> and now things are strange
121 2010-11-26 02:30:13 <OneFixt> Looks like several computers had the same addresses, and all thought they had coins when in fact it was just one batch of coins that they had, and then when I sent from one, the others either tried to double-send or noticed that they had nothing left and updated the balance.
122 2010-11-26 02:30:21 <OneFixt> Looks like I won't lose coins, but will have to do some cleaning.
123 2010-11-26 02:30:44 <xulrunner42> Oh, you copied your wallet :x
124 2010-11-26 02:30:46 <xulrunner42> Mistake
125 2010-11-26 02:31:38 <OneFixt> yeah, it was empty, but it had pre-generated keys
126 2010-11-26 02:32:53 <OneFixt> I didn't copy any coins, but when one bitcoind received coins, the other thought that it also had those coins (as far as I can tell).
127 2010-11-26 02:33:41 <OneFixt> every time that you generate a block, does it get generated to one address, or to a new one?
128 2010-11-26 02:37:30 <xulrunner42> I think they come to Your Address
129 2010-11-26 02:37:34 <xulrunner42> But that's a totally uninformed guess
130 2010-11-26 02:37:54 <xulrunner42> Do you have a netbook?
131 2010-11-26 02:40:17 <xulrunner42> Are there any bitcoin droid apps yet
132 2010-11-26 02:40:27 <xulrunner42> http://martyfunkhouser.csh.rit.edu/~yebyen/
133 2010-11-26 02:46:35 <joe_1> i dont undesrstand that site
134 2010-11-26 02:48:47 <joe_1> files are .img (what is that), font is incredibly small, some things are crossed out, the setences make no sense
135 2010-11-26 02:50:54 <joe_1> ok i goggled img and now i knwo what it is. but i dont see any bitcoin apps
136 2010-11-26 02:51:00 <xulrunner42> Well, not yet
137 2010-11-26 02:51:09 <xulrunner42> I just read the forum post to get a bitcoin app going though
138 2010-11-26 02:51:17 <xulrunner42> They are disk images
139 2010-11-26 02:51:23 <xulrunner42> I had to build 20 just to get this far
140 2010-11-26 02:51:28 <xulrunner42> You can run android on your netbook :D
141 2010-11-26 02:51:38 <xulrunner42> That's why all of the cross-outs
142 2010-11-26 02:51:41 <xulrunner42> And small print
143 2010-11-26 02:51:53 <joe_1> wow
144 2010-11-26 02:52:23 <xulrunner42> I think the source tree is in a pretty good state now
145 2010-11-26 02:52:47 <xulrunner42> If you get the one with a star, it works pretty good... today's build actually brings the wifi up on boot, if you uncheck ethernet first
146 2010-11-26 02:53:07 <xulrunner42> I think that all of this is very hardware dependent
147 2010-11-26 02:53:37 <xulrunner42> There is a generic_x86 target as well, but I don't have any hardware that can run it
148 2010-11-26 02:54:16 <joe_1> yeah, i dont have a net book
149 2010-11-26 02:54:22 <xulrunner42> Bummar
150 2010-11-26 02:54:54 <xulrunner42> It might work on any i915 hardware
151 2010-11-26 02:55:00 <xulrunner42> I have yet to try it on my PowerEdge SC420
152 2010-11-26 02:55:52 <nanotube> OneFixt: every block generation comes to a new address.
153 2010-11-26 02:57:03 <nanotube> OneFixt: but if you have several clients trying to gen with a cloned wallet... they could be working on the same address. because the key for generation is chosen from the keypool.
154 2010-11-26 02:57:17 <nanotube> so you really shouldn't run the same wallet from different places. :)
155 2010-11-26 03:02:53 <jgarzik> indeed
156 2010-11-26 03:03:46 <nanotube> Kiba: i bet you can find some handy (hah!) tutorials on the internet for that
157 2010-11-26 03:04:02 <Kiba> the tutorial doesn't help that much
158 2010-11-26 03:18:35 <MT`AwAy> nameless|: I would too but I got tired of grumbling, now I just make lazy ugly websites
159 2010-11-26 03:18:50 <MT`AwAy> (like https://smsz.net/ which works nicely, but which has no design)
160 2010-11-26 03:19:13 <nanotube> anyone know what the new fee schedule is like in version .17? the announcement forum post said something about 'different free transaction limits'...
161 2010-11-26 03:20:14 <MT`AwAy> nanotube: http://www.bitcoin.org/wiki/doku.php?id=transaction_fee ?
162 2010-11-26 03:20:20 <ArtForz> doesnt look like it changed from r186
163 2010-11-26 03:20:35 <ArtForz> err r187
164 2010-11-26 03:20:45 <MT`AwAy> mh, nothing written there in fact :(
165 2010-11-26 03:22:33 <nanotube> ArtForz: the .17 announcement, http://www.bitcoin.org/smf/index.php?topic=1946.0 has one of the items as "free transaction limits" ... so what gives?
166 2010-11-26 03:22:49 <nanotube> also, the transaction_fee wiki mentions 0.3.16 - and that version never existed, it seems.
167 2010-11-26 03:23:31 <ArtForz> we reduced fee-free size to 27kB
168 2010-11-26 03:23:57 <nanotube> ah ic
169 2010-11-26 03:23:59 <MT`AwAy> "The UI transaction fee setting was easy since it was still there from 0.1.5 and all I had to do was re-enable it"
170 2010-11-26 03:24:05 <nanotube> it used to be 50kb or so?
171 2010-11-26 03:24:10 <ArtForz> also, only allow a few k of fee-less 0-confirm based TX
172 2010-11-26 03:24:29 <nanotube> any idea what happened with .16, ArtForz ? :)
173 2010-11-26 03:24:35 <ArtForz> never released I think
174 2010-11-26 03:24:57 <ArtForz> r188 was .16
175 2010-11-26 03:25:23 <nanotube> mmm
176 2010-11-26 03:25:54 <nanotube> so it seems that satoshi decided not to go for "all tx have fee at least 0.001 (or something to that effect)" ?
177 2010-11-26 03:26:03 <ArtForz> for now, no
178 2010-11-26 03:26:07 <ArtForz> looks like gavin is still working out the kinks in the account handling RPC stuff
179 2010-11-26 03:26:19 <nanotube> mm
180 2010-11-26 03:35:05 <Kiba> err
181 2010-11-26 03:35:07 <Kiba> new sketch
182 2010-11-26 03:37:17 <Kiba> I think I did a great job in revising the right hand
183 2010-11-26 03:41:43 <Kiba> Am I a pioneer or what?
184 2010-11-26 03:42:33 <Kiba> The first to short sell bitcoins, the first to sell art for bitcoins...
185 2010-11-26 03:52:15 <noagendamarket> I wish there was a way to tell if youve already downloaded a file
186 2010-11-26 03:52:42 <noagendamarket> I paid an extra .02 only to get the same one lol
187 2010-11-26 03:53:29 <Kiba> I am delighted that you single handely help me break even on my investment
188 2010-11-26 03:54:23 <MT`AwAy> 9 hosts updated to 0.3.17
189 2010-11-26 03:57:12 <theymos> Did we skip 0.3.16 or did I just forget to send out a mailing list message for that one?
190 2010-11-26 03:58:15 <MT`AwAy> theymos: skipped it seems
191 2010-11-26 03:58:31 <MT`AwAy> anyway the good thing is I could remove getwork patch from my ebuild :D
192 2010-11-26 04:06:09 <lazarus> ;;bc,stats
193 2010-11-26 04:06:11 <gribble> Current Blocks: 93879 | Current Difficulty: 6866.89864897 | Next Difficulty At Block: 94752 | Next Difficulty In: 873 blocks | Next Difficulty In About: 5 days, 16 hours, 18 minutes, and 6 seconds | Next Difficulty Estimate: 7330.30780793
194 2010-11-26 04:22:28 <nanotube> MT`AwAy: /me goes to update to .17... make that 10 hosts pretty soon. :)
195 2010-11-26 04:22:57 <ArtForz> btw, satoshis getwork seems incompatible with m0s
196 2010-11-26 04:23:58 <nanotube> ArtForz: in the .17 announcement he says "thanks m0mchil for the getwork"... i'd have thought it would be compatible :)
197 2010-11-26 04:24:47 <ArtForz> I think Diablo-D3 said the endianness is different
198 2010-11-26 04:25:37 <doublec> satoshi went through the differences here iirc http://www.bitcoin.org/smf/index.php?topic=1901.msg24149;
199 2010-11-26 04:26:09 <nanotube> mmm
200 2010-11-26 04:26:31 <theymos> ArtForz: You said something that made me believe that the SVN r184 "efficiently sort transaction dependencies in one pass" change only really makes a difference for people mining with lots of GPUs. Is that right?
201 2010-11-26 04:26:49 <ArtForz> not quite
202 2010-11-26 04:27:00 <ArtForz> with enough queued transactions it will affect CPU miners as well
203 2010-11-26 04:27:32 <theymos> Every miner multiplies it?
204 2010-11-26 04:27:46 <ArtForz> yep
205 2010-11-26 04:27:56 <theymos> Thanks.
206 2010-11-26 04:28:29 <ArtForz> CPU overhead basically scaled as num_miners * (num_transactions ^ 2)
207 2010-11-26 04:29:19 <ByteCoin> Why squared in the number of txns?
208 2010-11-26 04:30:02 <ArtForz> I think because it scanned all other cached transactions for each transaction
209 2010-11-26 04:30:16 <ByteCoin> *shudders*
210 2010-11-26 04:30:29 <ArtForz> = a twice nested loop over cached txes = O(n^2)
211 2010-11-26 04:31:46 <ArtForz> the changed algo should scale O(n)
212 2010-11-26 04:32:19 <ArtForz> and with satoshis getwork changes number of miners shouldnt affect it at all
213 2010-11-26 04:40:45 <LobsterMan> http://imgur.com/Yl6k9.jpg
214 2010-11-26 04:43:38 <nanotube> LobsterMan: lol where did you dig that up
215 2010-11-26 04:43:58 <LobsterMan> reddit :P
216 2010-11-26 04:44:19 <nanotube> ic.
217 2010-11-26 05:15:54 <davux> hi!
218 2010-11-26 05:16:37 <davux> is it true that the transaction history of any address can be deduced from the block chain?
219 2010-11-26 05:16:51 <ByteCoin> yes
220 2010-11-26 05:17:47 <davux> how hard would it be to program an interface that takes an address as input, and outputs a list of transactions?
221 2010-11-26 05:17:56 <davux> like a web page for example
222 2010-11-26 05:18:05 <davux> it could be pretty useful
223 2010-11-26 05:18:11 <ArtForz> http://blockexplorer.com/
224 2010-11-26 05:18:12 <ByteCoin> like https://blockexplorer.com/
225 2010-11-26 05:18:55 <davux> oh :)
226 2010-11-26 05:19:22 <ByteCoin> Roll your own. His doesn't display transaction and block sizes...
227 2010-11-26 05:19:25 <davux> how comes the "from" appears, but doesn't in the official client?
228 2010-11-26 05:20:21 <ByteCoin> Probably because the from address can only be calculated if the coins you're using are from standard transactions
229 2010-11-26 05:20:47 <ByteCoin> The bitcoin client is probably interested in handling the general case
230 2010-11-26 05:21:09 <ArtForz> not to mention it's useless for general use
231 2010-11-26 05:21:22 <ByteCoin> yeah? explain...
232 2010-11-26 05:22:18 <ArtForz> well, whats the use case?
233 2010-11-26 05:22:53 <ByteCoin> Uhh.. all the transactions in the block chain
234 2010-11-26 05:23:10 <ArtForz> okay, whats the use case of DISPLAYING that?
235 2010-11-26 05:23:13 <ByteCoin> AFAIK they are all transactions from one address to another
236 2010-11-26 05:23:19 <ArtForz> you received 50 btc from xyz, which got those from a generation in block 123. hooray.
237 2010-11-26 05:24:08 <ByteCoin> I see your point.... but it does enable refunds!
238 2010-11-26 05:24:43 <ByteCoin> I don't know whether people would like to track transactions by having their payments have come from a certain address
239 2010-11-26 05:25:29 <ByteCoin> Suppose you have coins in 1Artforz12345... You might publish it as yours and have payments made to and from it
240 2010-11-26 05:25:59 <ByteCoin> Does any of this persuade you?
241 2010-11-26 05:26:10 <theymos> You can't send coins from a specific address currently. It would be nice if you could.
242 2010-11-26 05:26:35 <ByteCoin> Well you can by running multiple clients but I take your point
243 2010-11-26 05:26:56 <ByteCoin> The spammer seems to have some control over addresses in this fashion
244 2010-11-26 05:27:13 <ArtForz> as far as I can tell he's running 2 clients
245 2010-11-26 05:27:37 <ByteCoin> I'm writing a wallet handler and transaction generator. This will certainly be included functionality
246 2010-11-26 05:28:23 <ByteCoin> Is the spammer's IP discernable or is he using TOR or suchlike?
247 2010-11-26 05:29:32 <ArtForz> dunno, but if I were the spammer I'd use TOR or simialr to connect to a bunch of nodes
248 2010-11-26 05:30:14 <ByteCoin> Artforz: Have I made a good case for displaying where the coins in your payment come from? I am curious.
249 2010-11-26 05:30:21 <ArtForz> nope
250 2010-11-26 05:30:55 <theymos> It should be displayed when you double-click it. It's not really useful, but it's interesting.
251 2010-11-26 05:31:00 <ArtForz> yep
252 2010-11-26 05:31:00 <ByteCoin> Ok... perhaps the time isn't ripe for my wallet manager
253 2010-11-26 05:31:04 <davux> what spammer?
254 2010-11-26 05:31:05 <ArtForz> stuff it in a debug window
255 2010-11-26 05:32:22 <jgarzik> I suppose I could serialize, then get length of that
256 2010-11-26 05:32:23 <ArtForz> wast it something like getserializedsize or something?
257 2010-11-26 05:32:24 <ByteCoin> jgarzik: I would be tempted to issue the relevant getblock command over the TCP interface
258 2010-11-26 05:32:41 <ByteCoin> Then count the number of bytes coming back
259 2010-11-26 05:33:00 <ByteCoin> Seriously guys, I don't know why you like json-rpc so much
260 2010-11-26 05:33:12 <ArtForz> me neither
261 2010-11-26 05:33:14 <jgarzik> ByteCoin: "from within bitcoin"
262 2010-11-26 05:33:18 <ByteCoin> You get good info by pretending to be a client
263 2010-11-26 05:33:33 <ArtForz> ByteCoin: guess why I wrote that half-a-client?
264 2010-11-26 05:33:58 <ByteCoin> ArtForz:Hmm..... I wonder.....
265 2010-11-26 05:34:16 <ByteCoin> Could it be that you're not at the whim of someone else's software?
266 2010-11-26 05:34:31 <ByteCoin> Could it be that nobody can stop you doing what you want?
267 2010-11-26 05:34:51 <ByteCoin> Could it be that you can't really be lied to?
268 2010-11-26 05:35:13 <ArtForz> hint: try writing something to monitor txes and blocks as your node gets them
269 2010-11-26 05:35:14 <ByteCoin> Could it be that you're immune to a lot of versioning problems?
270 2010-11-26 05:36:00 <ArtForz> not to metnion it serves as a nice documention of the network protocol
271 2010-11-26 05:36:07 <tylergillies> will entering a fee into client make it any faster? or will that only affect speed later on?
272 2010-11-26 05:36:15 <nanotube> ArtForz: is your half-a-client code posted somewhere?
273 2010-11-26 05:36:24 <ArtForz> yep
274 2010-11-26 05:36:28 <ByteCoin> ArtForz: Yeah. I'm in the process. Didn't quite get the opoint of the hint
275 2010-11-26 05:36:43 <theymos> tylergillies: It makes your transactions faster when someone is spamming the network.
276 2010-11-26 05:36:46 <ByteCoin> I have monitored the txes and blocks
277 2010-11-26 05:36:57 <tylergillies> theymos: does that happen often?
278 2010-11-26 05:37:11 <ByteCoin> There wasn't anything THAT surprising.
279 2010-11-26 05:37:18 <ArtForz> http://pastebin.com/ZSM7iHZw
280 2010-11-26 05:37:53 <ByteCoin> I think I was a bit shocked that new blocks are relayed wholesale when all the clients have most of the transactions already
281 2010-11-26 05:38:15 <theymos> tylergillies: It's been happening more frequently lately. Personally, I would never pay a fee -- the worst that can happen is a 1-3 hour delay of your transactions.
282 2010-11-26 05:38:25 <ByteCoin> tylergillies: It's happening now but it's a low impact at the moment
283 2010-11-26 05:38:29 <ArtForz> and with the tx priritization even that's kidna unlikely
284 2010-11-26 05:38:50 <ArtForz> normal TXes now get priority over "spammy" looking ones
285 2010-11-26 05:39:17 <ByteCoin> Would you guys be surprised to find out that occasionally you get blocks >200k bytes?
286 2010-11-26 05:39:23 <ArtForz> nope
287 2010-11-26 05:39:25 <jgarzik> no
288 2010-11-26 05:39:31 <ArtForz> thats blocks from old miners
289 2010-11-26 05:39:47 <ByteCoin> Fair enough...
290 2010-11-26 05:40:15 <jgarzik> spam is up to 5 BTC in value, and now uses different to/from addresses
291 2010-11-26 05:40:23 <ByteCoin> Do you guys want to hear briefly about my "shunning" solution to spam?
292 2010-11-26 05:40:28 <jgarzik> spammer is playing with the priority system
293 2010-11-26 05:40:38 <ArtForz> except it wont help him much
294 2010-11-26 05:40:54 <ArtForz> priority = value * age of inputs
295 2010-11-26 05:41:52 <theymos> ByteCoin: Satoshi has already talked about "discouraged blocks" -- the network doesn't on blocks that don't follow your fee rules closely enough, but it will accept them if they get one block deep. Is that what you're thinking of?
296 2010-11-26 05:42:01 <theymos> doesn't build on*
297 2010-11-26 05:42:44 <ByteCoin> Hmm.. that sounds somewhat worrying.. I would have encouraged a free market for miners when it comes to fees
298 2010-11-26 05:42:51 <ByteCoin> No that's not it.
299 2010-11-26 05:43:22 <ByteCoin> theymos: Can you get a reference for that?
300 2010-11-26 05:43:29 <ByteCoin> Wait
301 2010-11-26 05:43:41 <ByteCoin> I'm douing a search for "discouraged blocks"
302 2010-11-26 05:44:12 <theymos> http://www.bitcoin.org/smf/index.php?topic=165.msg1595#msg1595
303 2010-11-26 05:44:19 <theymos> I was worried about it, too, at first. But it only works if a large portion of the network hates your rules.
304 2010-11-26 05:44:33 <theymos> And generators won't enforce rules that are bad for generators.
305 2010-11-26 05:45:09 <ByteCoin> Oh no. In the context he was talking about I approve
306 2010-11-26 05:45:45 <ByteCoin> I was going to post about the problem he mentions.
307 2010-11-26 05:46:32 <ByteCoin> It's possible to simplify miner software and minimise GPU downtime by just hashing blocks with your coinbase transaction in and NO other transactions.
308 2010-11-26 05:46:47 <ByteCoin> You get the 50BTC but don't take any trasnactions out of the pool.
309 2010-11-26 05:46:59 <ByteCoin> Not good for the network
310 2010-11-26 05:47:25 <ByteCoin> You wouldn't even need to run a full network node
311 2010-11-26 05:47:48 <ByteCoin> satoshi's method is the only real way to discourage this
312 2010-11-26 05:48:16 <theymos> The concept of discouraged blocks can be used for other issues. Spam transactions, odd fees, large blocks, etc. A good solution to the block size thing would be to set 1MB as "preferred" and up to 20MB as "discouraged", with sizes above that being refused as now.
313 2010-11-26 05:49:25 <ByteCoin> As I mention in my most recent post. If there's 20MB of transactions in your pool in memory, better that they be in a 20MB block and then hit the disk.
314 2010-11-26 05:49:36 <ByteCoin> Then your memory is clear
315 2010-11-26 05:49:47 <theymos> Just drop transactions and let the sender rebroadcast.
316 2010-11-26 05:49:50 <ByteCoin> I don't see any justification for pure block limits
317 2010-11-26 05:50:13 <ByteCoin> Theymos: What if they're not spam
318 2010-11-26 05:50:23 <ByteCoin> How does dropping them help things?
319 2010-11-26 05:50:24 <theymos> Then they'll be rebroadcast...
320 2010-11-26 05:50:34 <ByteCoin> That makes it better?
321 2010-11-26 05:50:59 <theymos> Yes. Spammers won't rebroadcast -- real people will.
322 2010-11-26 05:51:07 <ByteCoin> If they're legit transactions then forcing rebroadcasts would be a massive performance hit
323 2010-11-26 05:51:20 <ByteCoin> So why won't spammers rebroadcast?
324 2010-11-26 05:51:32 <theymos> They'd have to store every transaction, which is expensive.
325 2010-11-26 05:51:57 <ByteCoin> They're SPAMMERS. Do you think spamming is really free?
326 2010-11-26 05:52:27 <ArtForz> if your time and bandwith costs nothing... then yes
327 2010-11-26 05:52:35 <ByteCoin> Precisely
328 2010-11-26 05:52:37 <theymos> The rate of rebroadcast can be increased. 3 hours (current) is too long, I think.
329 2010-11-26 05:52:49 <ArtForz> 3 hours?
330 2010-11-26 05:52:56 <ArtForz> afair it's 1 block + 30-60 min
331 2010-11-26 05:53:13 <theymos> I thought it waited 3 hours for the first rebroadcast.
332 2010-11-26 05:53:29 <ByteCoin> Spammers will patch their software so that, if they suspect their transactions are being dropped they will connect to different clients and rebroadcast.
333 2010-11-26 05:54:16 <jgarzik> nNextTime = GetTime() + GetRand(30 * 60);
334 2010-11-26 05:54:39 <theymos> I guess that's not too bad, then.
335 2010-11-26 05:54:49 <ArtForz> okay... so it checks randomly every 0-30 min
336 2010-11-26 05:55:06 <jgarzik> it waits one iteration, first time through
337 2010-11-26 05:55:13 <ArtForz> if there's a new block, and the tx is older than 5 min, rebroadcast
338 2010-11-26 05:55:54 <theymos> The lowest-fee transactions would be the first ones dropped from memory. If everyone is rebroadcasting, the effect is essentially the same as including in blocks (but a bit slower).
339 2010-11-26 05:56:24 <theymos> Just the responsibility for storage is moved from the network to the senders.
340 2010-11-26 05:56:46 <ByteCoin> And the bandwidth for new transaction propagation is massively reduced.
341 2010-11-26 05:56:54 <ArtForz> yep, one could use a similar scoring system for dropping tx to what we have now for selecting tx for a block
342 2010-11-26 05:56:56 <ByteCoin> Sounds like a really bad "solution"
343 2010-11-26 05:57:10 <jgarzik> tx cache pruning?
344 2010-11-26 05:57:16 <ArtForz> yep
345 2010-11-26 05:57:39 <jgarzik> would be nice to have spam be the most likely cache entries to be pruned
346 2010-11-26 05:57:44 <ByteCoin> I bet that this "transaction dropping" will be the source of major bitcoin brownouts if it gets popular
347 2010-11-26 05:57:57 <ArtForz> why?
348 2010-11-26 05:58:03 <theymos> Generators will be powerful in the future. Network and disk space is unimportant.
349 2010-11-26 05:58:13 <ArtForz> upping the cache sizes is a simple change
350 2010-11-26 05:58:21 <ByteCoin> Because the bandwidth for propagating new transactions is massively reduced as I said
351 2010-11-26 05:58:37 <ArtForz> huh?
352 2010-11-26 05:59:00 <ByteCoin> New transactions have to compete with retransmitting old transactions to get across the net
353 2010-11-26 05:59:28 <ByteCoin> This means that new transactions are dropped... and so on ad infinitum
354 2010-11-26 05:59:37 <ArtForz> if that new transaction has a decent value it weill get propagated just fine
355 2010-11-26 05:59:42 <ArtForz> if it looks psammy... not so much
356 2010-11-26 06:00:05 <ByteCoin> Ok.. there is that
357 2010-11-26 06:00:22 <ArtForz> if actual legit transaction rate gets close to the cache size limits, up the cache size
358 2010-11-26 06:00:53 <ByteCoin> Artforz: In that case, not much will help you
359 2010-11-26 06:01:05 <ArtForz> huh?
360 2010-11-26 06:01:40 <ByteCoin> Well if the transaction rate increases over the ability to store them in blocks then BOOM
361 2010-11-26 06:02:05 <ArtForz> BOOM "tx will take a bit longer until enough clients increased cache sizes"
362 2010-11-26 06:02:22 <davux> how will transactions be confirmed when block generation stops?
363 2010-11-26 06:02:28 <ByteCoin> There's a problem of bandwidth where there are lots of nodes generating trasnsactions and not so many miners.
364 2010-11-26 06:02:28 <davux> or gets really slow?
365 2010-11-26 06:02:49 <theymos> jgarzik: My wild guess is 10 "backbone" companies, plus a few hundred hobbyist generators. Everyone else will be connected to one or two of the backbone nodes.
366 2010-11-26 06:02:55 <ByteCoin> Block generation is designed to happen every 10 mins on average
367 2010-11-26 06:03:06 <theymos> davux: It never stops. When it gets slow, you have to wait.
368 2010-11-26 06:03:20 <davux> but it will stop eventually, won't it?
369 2010-11-26 06:03:24 <ArtForz> nope
370 2010-11-26 06:03:32 <davux> since the total number of bitcoins is limited to 21 million
371 2010-11-26 06:03:40 <ArtForz> the amount of minted coins per block will get lower and lower
372 2010-11-26 06:03:45 <jgarzik> tx fees
373 2010-11-26 06:03:52 <ArtForz> blocks will still get generated at 10 min/block
374 2010-11-26 06:04:09 <ArtForz> at what difficulty... who knows
375 2010-11-26 06:04:33 <ArtForz> incentive for miners then will be a) tx fees and b) maintaining the network
376 2010-11-26 06:04:56 <davux> so even when there's no bitcoin left (or nearly) to be generated, blocks will continue to be generated?
377 2010-11-26 06:05:01 <ArtForz> yep
378 2010-11-26 06:05:01 <davux> i thought blocks and bitcoins were nearly the same thing
379 2010-11-26 06:05:08 <ArtForz> not really
380 2010-11-26 06:05:13 <nanotube> my wild guess is, there will be a 'pooled generator' running, so all the plebes can contribute. :)
381 2010-11-26 06:05:30 <ArtForz> yep
382 2010-11-26 06:05:51 <ArtForz> I expect a hub-leaf config is the only option if bitcoin gets really big
383 2010-11-26 06:06:12 <ByteCoin> My guess is that GPU generation will be seen to be a bad thing and will be made to cease
384 2010-11-26 06:06:20 <ArtForz> lol
385 2010-11-26 06:06:34 <nanotube> ByteCoin: that's not possible... how would you detect what source cpu power comes from? :P
386 2010-11-26 06:06:52 <ByteCoin> You'd have to change the proof of work...
387 2010-11-26 06:06:53 <jgarzik> my guess is that we'll have way too way zeroes in a hash
388 2010-11-26 06:07:01 <ArtForz> nah
389 2010-11-26 06:07:02 <davux> :)
390 2010-11-26 06:07:02 <theymos> Electricity cost competition will at some point make even pooled mining unprofitable.
391 2010-11-26 06:07:43 <ByteCoin> The problem with the current proof of work is that specialized hardware has better performance/cost characteristics
392 2010-11-26 06:07:47 <davux> i stopped generating, because my client was running at 400 khashes/s, and it seemed like it would take 2+ years
393 2010-11-26 06:07:51 <davux> so there was no point
394 2010-11-26 06:07:59 <nanotube> theymos: it's not about profit. people compute for the various @home profit at a clear cost to them.
395 2010-11-26 06:08:08 <nanotube> s/not about/not all about/ :)
396 2010-11-26 06:08:09 <davux> i think many people are in the same situation
397 2010-11-26 06:08:13 <ArtForz> ByteCoin: and thats true in the general case
398 2010-11-26 06:08:30 <ArtForz> theres really not many algorithms dedicated hardware can't speed up
399 2010-11-26 06:08:30 <ByteCoin> Artforz: Not so much...
400 2010-11-26 06:08:58 <ByteCoin> Hard to make specialized hardware if your proof of work requires a slab of memory too
401 2010-11-26 06:09:01 <jgarzik> surely there is a business model for pooled mining, even at higher TX rates
402 2010-11-26 06:09:15 <ByteCoin> A pretty big slab
403 2010-11-26 06:09:31 <AAA_awright> What if there's a maths breakthrough that renders blocks no longer necessary?
404 2010-11-26 06:09:39 <ByteCoin> Don't forget that an average computer has at least 1 GIGABYTE of memory
405 2010-11-26 06:09:43 <ArtForz> so?
406 2010-11-26 06:10:00 <theymos> nanotube: You'll need an enterprise-level Internet connection to generate at some point. Maybe some organizations will run donation nodes, but @home will be impossible.
407 2010-11-26 06:10:06 <AAA_awright> Point is, I don't see the protocol staying the same
408 2010-11-26 06:10:07 <ArtForz> nope
409 2010-11-26 06:10:17 <ArtForz> theymos: for pooled egenration the miners only need the block header
410 2010-11-26 06:10:20 <ByteCoin> So how would your gpu cope if each proof of work required processing several MB of information
411 2010-11-26 06:10:23 <jgarzik> indeed
412 2010-11-26 06:10:28 <nanotube> theymos: why, the server just distributes the work to the nodes
413 2010-11-26 06:10:58 <ByteCoin> Your GPU would be memory througput bound really soon
414 2010-11-26 06:11:01 <theymos> Forgot about that...
415 2010-11-26 06:11:03 <ArtForz> lol
416 2010-11-26 06:11:52 <jgarzik> it's such a pain manually manage external memory and CPU (read: GPU)
417 2010-11-26 06:12:22 <ArtForz> compare the memory bandwidth of a CPU to a mid-range GPU
418 2010-11-26 06:12:45 <bd_> Actually, it'd probably be sufficient to make the proof of work require a lot of branch operations
419 2010-11-26 06:12:54 <ByteCoin> I presume that the GPU has better bandwidth but by how many times?
420 2010-11-26 06:12:56 <bd_> Memory bandwidth alone isn't that much of a problem for a GPU
421 2010-11-26 06:13:06 <jgarzik> oh sure.  doesn't decrease the programmer's annoyance level any :)
422 2010-11-26 06:14:25 <ArtForz> about a factor of 4
423 2010-11-26 06:14:41 <ByteCoin> Put it this way. A memory-irrelevant POW like sha256 is much faster on a GPU than something memory bandwidth bound
424 2010-11-26 06:14:49 <ArtForz> yep
425 2010-11-26 06:15:19 <ByteCoin> So if your GPU only got you a factor of 4 improvement then it wouldn't be so compelling
426 2010-11-26 06:16:07 <ArtForz> so.. you want to make generation easiest on general-purpose CPUs
427 2010-11-26 06:16:19 <ByteCoin> You'd have to choose your POW so that it's something that CPUs do BEST of all hardware
428 2010-11-26 06:16:31 <ArtForz> congratulations, you just made it easier to break bitcoin with a botnet. you win one internets.
429 2010-11-26 06:17:12 <jgarzik> as tech moves forward everybody will have a useful GPU
430 2010-11-26 06:17:13 <ByteCoin> GPU miners are invulnurable from botnetting? Artforz - lose one internet
431 2010-11-26 06:17:24 <ArtForz> a lot less vulnerable
432 2010-11-26 06:17:37 <ByteCoin> Just because you have the knowhow?
433 2010-11-26 06:17:50 <ArtForz> your average rooted windows box doesn't have a very powerful GPU
434 2010-11-26 06:18:03 <ByteCoin> OK. I take your point
435 2010-11-26 06:18:29 <ArtForz> though it probably has a C2D at least
436 2010-11-26 06:18:45 <ByteCoin> Still, the current situation makes it more likely that bitcoin will be broken by a motivated attacker with specialized hardware
437 2010-11-26 06:19:01 <ArtForz> yep
438 2010-11-26 06:19:05 <ByteCoin> so ASIC on one side, botnet on the other
439 2010-11-26 06:19:19 <ArtForz> specialized hardware = major initial investment
440 2010-11-26 06:19:41 <theymos> The person with specialized hardware is competent and actually cares.
441 2010-11-26 06:19:43 <ByteCoin> I actually had an idea that you should have two proofs of work and take it in turns. The difficulties of both would vary independently//
442 2010-11-26 06:19:56 <ArtForz> and I mean MAJOR, you'd pretty much have to go ASIC to get a decent edge over GPUs
443 2010-11-26 06:20:19 <ByteCoin> One should be GPU based one CPU. Yes we're talking ASIC to get better than GPU
444 2010-11-26 06:20:27 <ArtForz> = FPGA wont cut it
445 2010-11-26 06:20:35 <ByteCoin> FPGA slow
446 2010-11-26 06:20:40 <ArtForz> slow and expensive
447 2010-11-26 06:20:50 <ByteCoin> for what it does
448 2010-11-26 06:21:05 <ArtForz> and beats GPUs 'only' by about 1.3-5 on Mhash/W
449 2010-11-26 06:21:12 <ArtForz> *1.3-1.5
450 2010-11-26 06:21:44 <ByteCoin> What do you think about two proofs of work?
451 2010-11-26 06:22:03 <ByteCoin> One would be suitable for GPUs and the other for CPUs
452 2010-11-26 06:22:25 <ArtForz> that could work
453 2010-11-26 06:23:01 <jgarzik> what is "suitable" for CPU yet not GPU?
454 2010-11-26 06:23:19 <ArtForz> anything with lots of branching and random memory access patterns
455 2010-11-26 06:23:22 <ByteCoin> jgarzik: Something memory bandwidth bound - or at least more so
456 2010-11-26 06:23:38 <ByteCoin> What artforz said
457 2010-11-26 06:23:48 <jgarzik> branching will get better as GPGPU catches on
458 2010-11-26 06:23:59 <jgarzik> it's a clear need already
459 2010-11-26 06:24:01 <ByteCoin> There are some fundamental limitations
460 2010-11-26 06:24:15 <jgarzik> such as?
461 2010-11-26 06:24:18 <ArtForz> I doubt it, as you make the SIMD vectors smaller your size/speed advantage over CPUs diminishes
462 2010-11-26 06:24:38 <ByteCoin> What artforz said
463 2010-11-26 06:24:55 <ArtForz> in the end you end up with something like larrabee
464 2010-11-26 06:25:03 <ArtForz> dozens of cpu-ish cores on a chip
465 2010-11-26 06:25:12 <jgarzik> nod, to Larrabee
466 2010-11-26 06:25:31 <ByteCoin> Memory bandwidth is a bottleneck. Why are supercomputers so expensive? Is it the processor?
467 2010-11-26 06:25:41 <ArtForz> nope, the interconnect
468 2010-11-26 06:25:48 <ByteCoin> A nice whizzy processor makes a supercomputer?
469 2010-11-26 06:25:59 <ByteCoin> Artforz wins back his internets
470 2010-11-26 06:26:51 <ArtForz> of course a decent ASIC implementation can hit TB/s
471 2010-11-26 06:27:12 <ByteCoin> ASIC implementation of what?
472 2010-11-26 06:27:23 <ByteCoin> SHA256?
473 2010-11-26 06:27:29 <ArtForz> of a memory controller
474 2010-11-26 06:27:45 <ArtForz> = getting a shitload of bandwidth isn't much of a problem, just add channels
475 2010-11-26 06:27:56 <ByteCoin> If that's the case why isn't it more common?
476 2010-11-26 06:28:17 <ArtForz> because nothing needs that much bandwidth
477 2010-11-26 06:28:46 <ArtForz> well, except for stuff like large crossbar switches
478 2010-11-26 06:28:52 <ByteCoin> Hmmm..... But some systems are running on 8 cores... I think memory is still a problem there mainly due to the lack of bandwidth
479 2010-11-26 06:29:27 <ByteCoin> Ok I won't argue
480 2010-11-26 06:29:29 <ArtForz> the problem with CPUs is that making a high-speed bus work with CPU and memory socketed is not easy
481 2010-11-26 06:29:33 <theymos> With the new tx fee changes, generators will always pay no fee on transactions under 10kB, even though they know the actual blocksize. That sucks.
482 2010-11-26 06:30:12 <ArtForz> you sure?
483 2010-11-26 06:30:13 <ByteCoin> theymos: I don't understand. What has the actual blocksize got to do with anything?
484 2010-11-26 06:30:23 <theymos> ByteCoin: Changes the fee.
485 2010-11-26 06:30:32 <ByteCoin> And whgy would generators pay a fee?
486 2010-11-26 06:30:47 <jgarzik> ByteCoin: generators receive the fee :)
487 2010-11-26 06:30:54 <ArtForz> hmmm., you'Re right
488 2010-11-26 06:31:12 <theymos> ArtForz: You think it's a bug, or intentional?
489 2010-11-26 06:31:21 <ByteCoin> Oh no.. Do I have to read that bit of source AGAIN....
490 2010-11-26 06:31:25 <ArtForz> I think thats a bug
491 2010-11-26 06:32:41 <theymos> I'll email Satoshi about it.
492 2010-11-26 06:33:16 <ArtForz> or maybe because miners will set fallowFree if a tx scores high enough?
493 2010-11-26 06:33:52 <ArtForz> hmmm... nope... they'll still hit the 27k limit
494 2010-11-26 06:34:20 <ArtForz> so yeah, I'd say bug
495 2010-11-26 06:34:41 <jgarzik> ?  nBlockSize>1 appears to be passed to GetMinFee in CreateNewBlock()?
496 2010-11-26 06:34:59 <ArtForz> yep
497 2010-11-26 06:35:22 <theymos> jgarzik: It's called without parameter when sending bitcoins, though, which causes you to ignore the blocksize and always pay no fees on transactions below 10 kB.
498 2010-11-26 06:35:49 <ByteCoin> Ugh....
499 2010-11-26 06:36:12 <LobsterMan> does anyone know what svn revision bitcoin 0.3.17 is based on?
500 2010-11-26 06:36:19 <jgarzik> theymos: apologies...  I'm a bit confused.  How does one know the block size when creating a TX?
501 2010-11-26 06:36:40 <theymos> jgarzik: Generators can look at the size of their own temporary block, which is likely to be accurate.
502 2010-11-26 06:37:03 <ArtForz> r191
503 2010-11-26 06:37:26 <ByteCoin> I think the whole fees thing needs to be overhauled. You only know how much fee you SHOULD have sent too late.
504 2010-11-26 06:37:31 <ByteCoin> There has to be a better way
505 2010-11-26 06:37:39 <jgarzik> theymos: and?  at that point, the TX (and any associated fee) has already been sent.  I thought fee was bundled with each TX at creation time, by having input>output?
506 2010-11-26 06:37:42 <LobsterMan> thanks ArtForz
507 2010-11-26 06:38:16 <ByteCoin> The way that code leverages default arguments to essentially be two different functions depending on where it's called from is ugly
508 2010-11-26 06:39:12 <theymos> jgarzik: Before creating the transaction, generators can know the approximate blocksize and therefore guess a more-likely-to-be-correct fee amount to send.
509 2010-11-26 06:39:39 <ByteCoin> I'm sure non-generators know this also from their transaction cache
510 2010-11-26 06:39:50 <jgarzik> theymos: it looks at recent history of block sizes?
511 2010-11-26 06:40:45 <theymos> jgarzik: No. Generators look at their own temporary block that they're hashing.
512 2010-11-26 06:42:50 <jgarzik> theymos: just to be clear, "generator" == creator of TX, or creator of block?
513 2010-11-26 06:43:06 <theymos> jgarzik: Anyone who has "generate coins" turned on.
514 2010-11-26 06:43:30 <theymos> People who don't attempt to create blocks don't collect transactions and can't guess at the block size.
515 2010-11-26 06:43:54 <ByteCoin> Can't they guess the size just by looking at their trasnaction pool?
516 2010-11-26 06:44:04 <theymos> Non-generators don't have a transaction pool.
517 2010-11-26 06:44:29 <ByteCoin> They receive them. They don't store them?
518 2010-11-26 06:44:41 <jgarzik> OK, then I don't understand "Before creating the transaction, generators can know[...]" -- because generators don't create transactions, nor do they "send fees" anywhere except their own wallet....
519 2010-11-26 06:44:46 <theymos> I don't think they even request them after getting the inv (or they shouldn't).
520 2010-11-26 06:45:09 <ByteCoin> I thought they didn't relay what they've already seen so they have to store (at least the hash) of what they've seen
521 2010-11-26 06:45:10 <jgarzik> creator of TX pays TX fees, and is the one who needs to know how much to send
522 2010-11-26 06:45:31 <theymos> jgarzik: I have "generate coins" turned on. I also send transactions from time to time. Those transactions that I send can use correct fees because I know which transactions will be in the next block.
523 2010-11-26 06:46:01 <ByteCoin> How do clients relay transactions correctly if they don't have a transaction pool?
524 2010-11-26 06:46:07 <LobsterMan> is changing the getwork interval from 5s to anything else likely to change anything significant?
525 2010-11-26 06:46:08 <theymos> ByteCoin: Yeah; maybe they store the hashes for relay purposes.
526 2010-11-26 06:46:11 <ByteCoin> Non-generating clientss
527 2010-11-26 06:46:17 <jgarzik> theymos: ahhh.  ok, understand.  an increasingly-uncommon case, IMO.
528 2010-11-26 06:46:19 <jgarzik> tnx
529 2010-11-26 06:46:40 <theymos> This benefit is actually the only reason that I do generate.
530 2010-11-26 06:47:29 <jgarzik> How many follow the same practice, and are likely to do so into the future?
531 2010-11-26 06:47:39 <theymos> Probably no one.
532 2010-11-26 06:47:55 <ByteCoin> theymos: How do you use this knowledge to calculate fees. I presume your client is non-standard
533 2010-11-26 06:48:13 <ByteCoin> It's something you've set up.
534 2010-11-26 06:48:22 <theymos> ByteCoin: No. The client does it automatically if you're a generator.
535 2010-11-26 06:48:54 <ByteCoin> How is are you made aware of it? Where is the "you should include this fee" information displayed?
536 2010-11-26 06:49:32 <theymos> GetMinFee is used for both generation and transaction creation. It looks at the blocksize.
537 2010-11-26 06:50:30 <ByteCoin> But as a user how do you tell. Is it displayed somewhere?
538 2010-11-26 06:51:01 <tylergillies> wooho just got docs.google.com to throw a 500 error
539 2010-11-26 06:51:04 <ByteCoin> Or are your outgoing transactions fee'd at the discretion of your generating client?
540 2010-11-26 06:51:05 <tylergillies> first time ive seen that
541 2010-11-26 06:51:26 <theymos> ByteCoin: Right. It's the fee that will show up in the "you must pay this fee" dialog.
542 2010-11-26 06:51:40 <ByteCoin> Ok. I understand.
543 2010-11-26 07:00:16 <jgarzik> shoot, not many bitcoins at all living at bitcoinmarket anymore
544 2010-11-26 07:01:07 <jgarzik> no one wants MoneyBookers USD at all [for bitcoins] apparently, and I'm the only one selling BTC for PGAU
545 2010-11-26 07:04:12 <LobsterMan> it's interesting to look at the heat trends of my cards between different bitcoin and miner versions
546 2010-11-26 07:08:17 <LobsterMan> LOL
547 2010-11-26 07:08:20 <LobsterMan> oh dear
548 2010-11-26 07:09:55 <LobsterMan> http://imgur.com/FwSsr.png
549 2010-11-26 07:10:05 <LobsterMan> and then it crashes
550 2010-11-26 07:10:43 <LobsterMan> if i use -v
551 2010-11-26 07:11:06 <LobsterMan> http://pastebin.com/R4TBuvBK
552 2010-11-26 07:23:53 <OneFixt> LobsterMan: is that a bug or are you actually getting 3.8Gh/s?
553 2010-11-26 07:29:47 <OneFixt> oh, that's Terahash, lol; bug it is
554 2010-11-26 07:32:27 <OneFixt> ;;bc,estimate
555 2010-11-26 07:32:28 <gribble> 7341.58529730
556 2010-11-26 07:32:32 <OneFixt> ;;bc,stats
557 2010-11-26 07:32:35 <gribble> Current Blocks: 93902 | Current Difficulty: 6866.89864897 | Next Difficulty At Block: 94752 | Next Difficulty In: 850 blocks | Next Difficulty In About: 5 days, 12 hours, 30 minutes, and 24 seconds | Next Difficulty Estimate: 7341.58529730
558 2010-11-26 07:47:20 <LobsterMan> i think it's a bug
559 2010-11-26 07:47:21 <LobsterMan> lol
560 2010-11-26 07:51:58 <LobsterMan> it looks like it was counting total hashes
561 2010-11-26 07:51:59 <LobsterMan> not the rate
562 2010-11-26 07:54:20 <jrabbit> http://tech.slashdot.org/story/10/11/26/0238248/UK-Police-To-Get-Major-New-Powers-To-Seize-Domains
563 2010-11-26 07:57:27 <LobsterMan> is my gfc card fan more likely to burn out running at 100% for long periods of time as opposed to like 80%?
564 2010-11-26 07:57:29 <LobsterMan> gfx*
565 2010-11-26 08:31:18 <tylergillies> I work at readwriteweb.com and im good friends with the co-editor, im gonna try to get him to write a story about bitcoin for our wikipedia efforts, gonna make that part of the story
566 2010-11-26 08:44:39 <theymos> Cool. I should buy some more BTC before your story is released. ;)
567 2010-11-26 08:44:45 <tylergillies> heh
568 2010-11-26 08:57:11 <jgarzik> http://www.bitcoin.org/smf/index.php?topic=1958.0
569 2010-11-26 08:57:12 <jgarzik> Bitcoin press hits, notable sources
570 2010-11-26 08:57:27 <jgarzik> contributions (URLs) requested.
571 2010-11-26 09:27:37 <LobsterMan> i added a new link to that thread
572 2010-11-26 09:28:13 <LobsterMan> i don't know if it's "legit" or whatnot but it has people commenting on it
573 2010-11-26 09:42:28 <LobsterMan> everyone should send the EFF like 10 btc :]
574 2010-11-26 09:42:29 <LobsterMan> https://www.eff.org/helpout
575 2010-11-26 09:42:38 <LobsterMan> if you live in the US
576 2010-11-26 09:45:05 <joe_1> why does it matter if u live in the us
577 2010-11-26 09:46:52 <LobsterMan> or anyone who supports them :]
578 2010-11-26 09:47:08 <joe_1> o
579 2010-11-26 09:48:04 <LobsterMan> they're a us based organization
580 2010-11-26 09:48:05 <LobsterMan> http://en.wikipedia.org/wiki/Electronic_Frontier_Foundation
581 2010-11-26 09:48:35 <altamic> LobsterMan: done
582 2010-11-26 09:48:51 <LobsterMan> nice
583 2010-11-26 09:48:54 <LobsterMan> i sent them 10 heh
584 2010-11-26 09:49:02 <LobsterMan> maybe will send more in the future
585 2010-11-26 09:49:12 <altamic> idem
586 2010-11-26 09:49:40 <LobsterMan> i sent some to m0mchil for his miner too lol
587 2010-11-26 09:54:48 <joe_1> 1NFVpieFXFj8iiYMbckUsuVaizFhSjXjQQ
588 2010-11-26 09:57:22 <LobsterMan> :P
589 2010-11-26 09:57:49 <joe_1> thanks
590 2010-11-26 09:58:21 <joe_1> have u tried my bitcoin site yet
591 2010-11-26 09:58:24 <LobsterMan> no
592 2010-11-26 09:58:38 <joe_1> it's at cashcow.no-ip.org
593 2010-11-26 10:02:26 <joe_1> i'll start you off with 20 coins if you let me know the username
594 2010-11-26 10:06:47 <joe_1> im starting a financial website as well in the near future
595 2010-11-26 10:07:51 <joe_1> a forex type site to allow people to buy dollars / sell dollars / short sell / etc
596 2010-11-26 10:13:58 <LobsterMan> ahh hmm
597 2010-11-26 10:20:36 <theymos> I just enabled compression on BBE. For those who've used it before: does it seem faster now? (It seems slower to me, but it's on my LAN.) http://blockexplorer.com/
598 2010-11-26 10:20:38 <altamic> joe_1: I will code for BTC :)
599 2010-11-26 10:21:51 <LobsterMan> seems about the same to me
600 2010-11-26 10:21:54 <LobsterMan> maybe a bit faster?
601 2010-11-26 10:22:05 <LobsterMan> i never really noticed it was especially slow before
602 2010-11-26 10:22:24 <altamic> fast as usual here
603 2010-11-26 10:24:10 <theymos> Thanks. I'll leave it compressed for now.
604 2010-11-26 11:06:39 <tylergillies> its kinda lonely in #bitcoin on i2p
605 2010-11-26 11:06:43 <tylergillies> its just me and myself
606 2010-11-26 11:52:02 <MacRohard> you might as well leave since you just blew your cover ;)
607 2010-11-26 12:06:21 <LobsterMan> http://english.aljazeera.net/video/americas/2010/11/201011269019909253.html
608 2010-11-26 12:06:52 <doublec> what's with all the 0.004 mtgox trades?
609 2010-11-26 12:07:26 <doublec> #bitcoin-market is showing a ton of them
610 2010-11-26 12:07:43 <LobsterMan> not sure, likely a spammer
611 2010-11-26 12:07:53 <LobsterMan> it stopped for a few hours last night
612 2010-11-26 12:07:57 <LobsterMan> but it's started up again
613 2010-11-26 12:15:00 <doublec> weird trade amounts going on
614 2010-11-26 12:15:15 <doublec> maybe they're trying to find or exploit some rounding bug
615 2010-11-26 12:16:00 <doublec> or someone's trading bot went crazy and they're going to be disappointed when they check on it later...
616 2010-11-26 12:16:03 <LobsterMan> lots of small trades recently
617 2010-11-26 12:17:02 <doublec> lots of 9.744 @ $0.282
618 2010-11-26 12:17:13 <doublec> followed by 10 @ $0.2764
619 2010-11-26 12:17:30 <doublec> interesting how the price is always the same for the given number of btc
620 2010-11-26 12:18:25 <LobsterMan> yeah mtgox has some shit to sort out
621 2010-11-26 12:19:21 <remmy> The transaction values of the two are almost the same too
622 2010-11-26 12:19:29 <doublec> might just be two trading bots competing
623 2010-11-26 12:20:16 <remmy> It does make MTG look like a busy place :)
624 2010-11-26 12:32:16 <tylergillies> is gpu built into main client now? whats this getwork people are talking about?
625 2010-11-26 12:52:10 <ByteCoin> test
626 2010-11-26 13:21:45 <UukGoblin> Diablo-D3, I mvn package'd your miner, and now when I try to start it on a 64-bit ubuntu lucid I get: Exception in thread "main" java.lang.UnsatisfiedLinkError: /home/bitcoin/DiabloMiner/target/libs/natives/linux/liblwjgl.so: /home/bitcoin/DiabloMiner/target/libs/natives/linux/liblwjgl.so: wrong ELF class: ELFCLASS32 (Possible cause: architecture word width mismatch)
627 2010-11-26 13:22:07 <UukGoblin> (dunno why it uses a wrong lib)
628 2010-11-26 13:22:16 <UukGoblin> (it should use liblwjgl64.so
629 2010-11-26 13:22:24 <UukGoblin> (which is there)
630 2010-11-26 13:23:03 <UukGoblin> ahrm, the *64.so tries to link to jawt.so which /isn't/ there I think
631 2010-11-26 13:23:09 <UukGoblin> libjawt.so I mean
632 2010-11-26 13:27:25 <UukGoblin> question 2: is there a way to further overclock a HD 5770 - ati's drivers only allow up to 960MHz, and my card is still only 61 deg C warm
633 2010-11-26 13:40:36 <Diablo-D3> back
634 2010-11-26 13:40:42 <Diablo-D3> UukGoblin: because you're on ubuntu 64
635 2010-11-26 13:40:43 <Diablo-D3> right?
636 2010-11-26 13:40:46 <Diablo-D3> riiiiight?
637 2010-11-26 13:40:56 <Diablo-D3> riiiiiiiiiiiight?
638 2010-11-26 13:44:36 <UukGoblin> Diablo-D3, right
639 2010-11-26 13:44:45 <UukGoblin> I am on ubuntu 64
640 2010-11-26 13:44:53 <UukGoblin> is ubuntu wrong or is 64 wrong? ;-P
641 2010-11-26 13:45:08 <Diablo-D3> UukGoblin: the combination of both
642 2010-11-26 13:45:12 <Diablo-D3> ubuntu 32 works, debian 64 works
643 2010-11-26 13:45:16 <Diablo-D3> but not ubuntu 64
644 2010-11-26 13:45:19 <Diablo-D3> and I cant figure out why
645 2010-11-26 13:45:29 <UukGoblin> hm.
646 2010-11-26 13:45:32 <UukGoblin> sucks.
647 2010-11-26 13:45:33 <Diablo-D3> and it effects, theoretically, any lwjgl using opencl app (but not opengl apps)
648 2010-11-26 13:45:36 <Diablo-D3> yes it does
649 2010-11-26 13:45:44 <Diablo-D3> because now three of you ubuntu faggots have reported that bug
650 2010-11-26 13:45:50 <UukGoblin> I'm happy to fiddle with it
651 2010-11-26 13:45:52 <Diablo-D3> the only cure is to quit using ubuntu altogether
652 2010-11-26 13:46:02 <Diablo-D3> because its clearly an ubuntu bug
653 2010-11-26 13:46:12 <UukGoblin> firstly, when I ldd that liblwjgl64.so, I get a not found on libjawt.so
654 2010-11-26 13:46:30 <Diablo-D3> UukGoblin: ubuntu does package sun java
655 2010-11-26 13:46:32 <Diablo-D3> try using it
656 2010-11-26 13:46:36 <UukGoblin> I am using it
657 2010-11-26 13:46:51 <UukGoblin> one from "deb http://archive.canonical.com/ lucid partner"
658 2010-11-26 13:46:54 <Diablo-D3> orly?
659 2010-11-26 13:46:57 <Diablo-D3> not openjdk?
660 2010-11-26 13:47:11 <Diablo-D3> java -version should agree with you
661 2010-11-26 13:47:13 <UukGoblin> I had openjdk first, but maven was bitching about javac, so I got the sun one
662 2010-11-26 13:47:24 <UukGoblin> java version "1.6.0_22"
663 2010-11-26 13:47:33 <UukGoblin> Java(TM) SE Runtime Environment (build 1.6.0_22-b04)
664 2010-11-26 13:47:37 <Diablo-D3> yeah thats the right one
665 2010-11-26 13:47:41 <Diablo-D3> I wonder what the hell ubuntu did
666 2010-11-26 13:48:15 <UukGoblin> so - any clue as to what libjawt is?
667 2010-11-26 13:48:20 <Diablo-D3> yes, its part of java.
668 2010-11-26 13:48:21 <UukGoblin> I don't see it in apt
669 2010-11-26 13:48:48 <Diablo-D3> both sun and openjdk ship it
670 2010-11-26 13:48:56 <Diablo-D3> as does gcj
671 2010-11-26 13:48:59 <Diablo-D3> without it, no awt
672 2010-11-26 13:49:38 <UukGoblin> ah, OK, I have it in /usr/lib/jvm/java-6-sun-1.6.0.22/jre/lib/amd64/libjawt.so
673 2010-11-26 13:50:15 <UukGoblin> which is not in ld.so.conf
674 2010-11-26 13:50:27 <Diablo-D3> it shouldnt be
675 2010-11-26 13:50:28 <UukGoblin> which is fine - java probably knows where to look for it
676 2010-11-26 13:50:35 <Diablo-D3> yes
677 2010-11-26 13:50:42 <Diablo-D3> I suspect this isnt a java problem at all
678 2010-11-26 13:50:44 <UukGoblin> ok
679 2010-11-26 13:50:51 <Diablo-D3> but its a problem with their broken security shot no one else uses
680 2010-11-26 13:51:02 <Diablo-D3> since only ubuntu 64 suffers from it
681 2010-11-26 13:51:15 <UukGoblin> I'll make liblwjgl.so point at liblwjgl64.so
682 2010-11-26 13:51:27 <Diablo-D3> doesnt fix it
683 2010-11-26 13:51:34 <Diablo-D3> thats already been tried
684 2010-11-26 13:51:54 <Diablo-D3> lwjgl works by trying to link liblwjgl.so first, then the 64 bit one
685 2010-11-26 13:52:19 <Diablo-D3> problem is, there something very broken in ubuntu thats making attempting to link the 32 bit one fail
686 2010-11-26 13:52:50 <UukGoblin> indeed, but now it gives a different error: Exception in thread "main" java.lang.UnsatisfiedLinkError: /home/bitcoin/DiabloMiner/target/libs/natives/linux/liblwjgl.so: /usr/lib/jvm/java-6-sun-1.6.0.22/jre/lib/amd64/libjawt.so: symbol awt_FreeDrawingSurface, version SUNWprivate_1.1 not defined in file libmawt.so with link time reference
687 2010-11-26 13:53:31 <UukGoblin> libmawt.so?
688 2010-11-26 13:54:05 <Diablo-D3> yes even more awesome javary
689 2010-11-26 13:55:13 <UukGoblin> so libmawt.so is in headless, xawt and motif21
690 2010-11-26 13:55:43 <Diablo-D3> UukGoblin: now heres the really fun part
691 2010-11-26 13:55:49 <Diablo-D3> someone tried downloading sun's from sun
692 2010-11-26 13:55:56 <Diablo-D3> it STILL happens
693 2010-11-26 13:56:02 <Diablo-D3> so whatever it is, its not in java
694 2010-11-26 13:56:16 <Diablo-D3> and its not in ubuntu's packaging of java
695 2010-11-26 13:56:21 <Diablo-D3> its in ubuntu somewhere else
696 2010-11-26 13:56:26 <Diablo-D3> this bug makes no sense
697 2010-11-26 13:56:55 <Diablo-D3> UukGoblin: I bet if you install the 32 bit java package (I think ubuntu has one), it'll work
698 2010-11-26 13:57:00 <Diablo-D3> which is absolutely nuts
699 2010-11-26 13:57:19 <UukGoblin> so what's awt?
700 2010-11-26 13:58:01 <Diablo-D3> windowing toolkit shit
701 2010-11-26 13:58:34 <Diablo-D3> lwjgl uses it for opengl apps to open the window, but for nothing else (unless you use lwjgl as a widget in awt/swing apps)
702 2010-11-26 13:58:36 <UukGoblin> could you do: nm /path/to/jre/lib/amd64/libjawt.so | grep awt_FreeDrawing
703 2010-11-26 13:58:37 <UukGoblin> ?
704 2010-11-26 13:59:36 <Diablo-D3> U awt_FreeDrawingSurface@@SUNWprivate_1.1
705 2010-11-26 13:59:44 <UukGoblin> same as here
706 2010-11-26 13:59:56 <UukGoblin> so it must be defined somewhere with that weird version
707 2010-11-26 14:00:03 <Diablo-D3> yes because ubuntu screwed with shit
708 2010-11-26 14:00:35 <UukGoblin> yes
709 2010-11-26 14:00:37 <Diablo-D3> and like I said
710 2010-11-26 14:00:41 <UukGoblin> question is, which shit that is
711 2010-11-26 14:00:43 <Diablo-D3> download the .bin from sun
712 2010-11-26 14:00:45 <Diablo-D3> and it still fails
713 2010-11-26 14:00:57 <Diablo-D3> ubuntu uses really retarded security shit in the kernel
714 2010-11-26 14:01:05 <Diablo-D3> it'll probably break cleverly written code
715 2010-11-26 14:01:15 <Diablo-D3> this is why linus has rejected it upstream everytime its come up
716 2010-11-26 14:01:42 <Diablo-D3> UukGoblin: btw
717 2010-11-26 14:01:46 <Diablo-D3> everyone else whos had this bug
718 2010-11-26 14:01:48 <Diablo-D3> had to use a 32 bit jvm
719 2010-11-26 14:01:57 <Diablo-D3> like, people not even using lwjgl have had this bug
720 2010-11-26 14:02:45 <UukGoblin> shiit.
721 2010-11-26 14:02:48 <UukGoblin> that sounds deep.
722 2010-11-26 14:02:56 <UukGoblin> as in, that the bug is deep
723 2010-11-26 14:02:58 <Diablo-D3> I wonder if awt/swt is subtly broken somewhere on 64bit linux
724 2010-11-26 14:03:22 <Diablo-D3> because almost all the natural occurrences of the ELFCLASS32 bug are in awt/swt shit
725 2010-11-26 14:05:39 <UukGoblin> so which lib or deb defines awt_FreeDrawingSurface@@SUNWprivate_1.1 ?
726 2010-11-26 14:05:49 <Diablo-D3> its internal to java
727 2010-11-26 14:09:55 <UukGoblin> so awt_FreeDrawingSurface (however, without the @@SUNWprivate_1.1 shit) is deifned in xawt/libmawt.so and motif21/libmawt.so but not headless/libmawt.so
728 2010-11-26 14:10:13 <Diablo-D3> UukGoblin: yeah but
729 2010-11-26 14:10:16 <Diablo-D3> java does shit
730 2010-11-26 14:13:08 <UukGoblin> heh
731 2010-11-26 14:13:23 <UukGoblin> added DISPLAY=:0 and it almost works :-]
732 2010-11-26 14:14:23 <UukGoblin> Diablo-D3, http://pastebin.com/dnsBcnnW
733 2010-11-26 14:14:57 <UukGoblin> was /that/ reported too?
734 2010-11-26 14:15:47 <UukGoblin> (does running m0's miner in background hurt? where do I set which device I want to use with your miner?)
735 2010-11-26 14:17:56 <UukGoblin> now, that is crazy
736 2010-11-26 14:18:03 <UukGoblin> I added -w 1000
737 2010-11-26 14:18:09 <UukGoblin> it added two Junipers
738 2010-11-26 14:18:17 <UukGoblin> and I'm now getting 165428000 khash/sec
739 2010-11-26 14:18:25 <UukGoblin> even 183778571 khash/sec
740 2010-11-26 14:18:58 <UukGoblin> with m0's running in background and getting a tiny penalty (175Mhash -> 170Mhash on OC'ed card and 155Mhash -> 150Mhash on non-OC'ed)
741 2010-11-26 14:19:24 <UukGoblin> (also, I do prefer seeing separate hashrates for different cards as I want to know the effects of overclocking)
742 2010-11-26 14:20:00 <UukGoblin> (I'm guessing your khash is wrongly calculated when -w is specified)
743 2010-11-26 14:21:09 <UukGoblin> (and I definitely like the option of choosing which devices I want to run shit on)
744 2010-11-26 14:35:29 <Diablo-D3> b9acj
745 2010-11-26 14:35:41 <Diablo-D3> UukGoblin: -w cannot exceed the max hardware size
746 2010-11-26 14:35:46 <Diablo-D3> and it has to be a power of 2
747 2010-11-26 14:36:24 <Diablo-D3> UukGoblin: it should be catching that exception and dying properly
748 2010-11-26 14:37:45 <UukGoblin> Diablo-D3, it's not then
749 2010-11-26 14:38:03 <Diablo-D3> UukGoblin: wait
750 2010-11-26 14:38:04 <UukGoblin> and it also dies when it tries to determine it on its own
751 2010-11-26 14:38:07 <Diablo-D3> how did you fix the bug?
752 2010-11-26 14:39:26 <Diablo-D3> UukGoblin: well the default should work fine.
753 2010-11-26 14:39:44 <UukGoblin> the ubuntu thing? 1. I copied all 3 64-bit libs in target/libs/natives/linux as the non-64 ones, 2. I added DISPLAY=:0 (with running X)
754 2010-11-26 14:39:57 <UukGoblin> Diablo-D3, the default gives http://pastebin.com/dnsBcnnW
755 2010-11-26 15:16:28 <Diablo-D3> UukGoblin: hrm
756 2010-11-26 15:16:38 <Diablo-D3> UukGoblin: something sint right on your machine them
757 2010-11-26 15:16:52 <Diablo-D3> that method cannot segfault unless your shit is screwed
758 2010-11-26 15:17:56 <Diablo-D3> UukGoblin: again, give me your java -version
759 2010-11-26 15:18:09 <Diablo-D3> your hack cant work
760 2010-11-26 15:42:31 <nanotube> ;;bc,stats
761 2010-11-26 15:42:33 <gribble> Error: invalid syntax (<string>, line 1)
762 2010-11-26 15:42:38 <nanotube> ooh
763 2010-11-26 15:42:47 <nanotube> whatsup with that...
764 2010-11-26 15:44:45 <nanotube> looks like bbe is down...
765 2010-11-26 15:45:12 <nanotube> http://blockexplorer.com/q/estimate is, anyway
766 2010-11-26 15:45:27 <gribble> The operation succeeded.
767 2010-11-26 15:45:27 <nanotube> ;;later tell theymos http://blockexplorer.com/q/estimate is down.
768 2010-11-26 15:46:05 <UukGoblin> Diablo-D3, java version "1.6.0_22"
769 2010-11-26 15:46:06 <UukGoblin> Java(TM) SE Runtime Environment (build 1.6.0_22-b04)
770 2010-11-26 15:46:21 <UukGoblin> it's not a hack it's a workaround. ;-P
771 2010-11-26 15:52:46 <altamic> any macos user around?
772 2010-11-26 16:10:04 <Kiba> hello
773 2010-11-26 16:10:28 <tcatm> hi Kiba
774 2010-11-26 16:11:12 <Kiba> hi
775 2010-11-26 16:11:19 <zc00gii> wonder if any miners are around that work better on my machine then the 4way miner
776 2010-11-26 16:11:34 <tcatm> zc00gii: what hardware do you have?
777 2010-11-26 16:12:46 <zc00gii> tcatm: crappy hardware
778 2010-11-26 16:12:51 <zc00gii> uh, no GPU mining
779 2010-11-26 16:12:54 <zc00gii> a
780 2010-11-26 16:13:00 <zc00gii> {System info for kamino} {CPU} [Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz] {Cores} [2] {Usage} [22%] {Memory} [783MB/2005MB] {Swap} [13MB/7632MB] {Running} [Gentoo GNU/Linux with kernel version 2.6.35.1-gotta-go-get-that-GTA+ on a x86_64 processor] {Storage} [/dev/sda4 mounted at / 5.4G out of 265G used] [/dev/sdb1 mounted at /home 56G out of 932G used] {Uptime} [12:12:59 up 6 days, 1:08, 1 user] {Load average} [2.25, 2.21, 2.18]
781 2010-11-26 16:13:06 <zc00gii> there's everything else
782 2010-11-26 16:13:37 <tcatm> normal miner might work better than 4way
783 2010-11-26 16:14:38 <zc00gii> tcatm: hmm
784 2010-11-26 16:14:58 <zc00gii> tcatm: thing is, I'm waiting for a confirmation
785 2010-11-26 16:15:23 <zc00gii> tcatm: if I stop it and start the normal miner, will I lose the transaction confirmation?
786 2010-11-26 16:15:36 <tcatm> does it already have one confirmation?
787 2010-11-26 16:17:02 <zc00gii> no
788 2010-11-26 16:17:30 <Kiba> or not
789 2010-11-26 16:17:49 <tcatm> Kiba: that's like 200 downloads?
790 2010-11-26 16:18:12 <Kiba> .02 excuse me
791 2010-11-26 16:18:18 <Diablo-D3> back
792 2010-11-26 16:18:22 <Kiba> hmm, I am sooo obscure
793 2010-11-26 16:18:38 <Kiba> my work ain't worth pirating anyway
794 2010-11-26 16:18:41 <Diablo-D3> UukGoblin: huh.
795 2010-11-26 16:18:43 <Diablo-D3> UukGoblin: strange.
796 2010-11-26 16:18:46 <tcatm> zc00gii: well, you can always restart bitcoin but you should try to keep it running until it has at least one confirmation. it doesn't matter wheter you're computer is mining or not.
797 2010-11-26 16:19:01 <Diablo-D3> UukGoblin: and you're not getting an error it cant find the opencl libs?
798 2010-11-26 16:19:51 <UukGoblin> Diablo-D3, oh, for that, you need to set the environment right I think... ATISTREAMSDKROOT... er.... actually, it's set wrong...
799 2010-11-26 16:20:20 <Diablo-D3> UukGoblin: export LD_LIBRARY_PATH=/path/to/ati-stream-sdk-v2.2-lnx64/lib/x86_64/:$LD_LIBRARY_PATH
800 2010-11-26 16:20:21 <Diablo-D3> or whatnot
801 2010-11-26 16:20:21 <UukGoblin> that means it worked on ati stream 2.1
802 2010-11-26 16:20:27 <UukGoblin> yeah, I have it
803 2010-11-26 16:20:30 <Diablo-D3> it works on 2.1 too
804 2010-11-26 16:22:27 <Diablo-D3> UukGoblin: how are you running my miner? just with -u -p and -w?
805 2010-11-26 16:22:38 <UukGoblin> via DiabloMiner.sh
806 2010-11-26 16:22:47 <UukGoblin> and yes, exactly the parameters you mentioned
807 2010-11-26 16:22:50 <Diablo-D3> hrm
808 2010-11-26 16:22:55 <Diablo-D3> something isnt right on your box.
809 2010-11-26 16:22:57 <UukGoblin> and DISPLAY=:0
810 2010-11-26 16:23:04 <Diablo-D3> you shouldnt need DISPLAY=:0
811 2010-11-26 16:23:08 <UukGoblin> and the lib*64 swapped
812 2010-11-26 16:23:09 <Diablo-D3> the default already should be :0
813 2010-11-26 16:23:37 <UukGoblin> with DISPLAY="" I get this linking thing (Exception in thread "main" java.lang.UnsatisfiedLinkError: /home/bitcoin/DiabloMiner/target/libs/natives/linux/liblwjgl.so: /usr/lib/jvm/java-6-sun-1.6.0.22/jre/lib/amd64/libjawt.so: symbol awt_FreeDrawingSurface, version SUNWprivate_1.1 not defined in file libmawt.so with link time reference)
814 2010-11-26 16:23:45 <Diablo-D3> no
815 2010-11-26 16:23:48 <Diablo-D3> you cant blank it out
816 2010-11-26 16:23:59 <Diablo-D3> the default already should be :0
817 2010-11-26 16:23:59 <UukGoblin> it's blank by default man ;-]
818 2010-11-26 16:24:06 <UukGoblin> m0's miner works with DISPLAY=""
819 2010-11-26 16:24:53 <Diablo-D3> no it doesnt.
820 2010-11-26 16:24:56 <Diablo-D3> at least it shouldnt
821 2010-11-26 16:24:57 <Diablo-D3> and no
822 2010-11-26 16:24:59 <UukGoblin> but I set it to :0 for your miner because the headless/libmawt.so doesn't link to that weird awt_FreeDrawingSurface symbol (and it tries to use it)
823 2010-11-26 16:25:00 <Diablo-D3> its not blank by default
824 2010-11-26 16:25:09 <UukGoblin> it's blank when you run it via ssh.
825 2010-11-26 16:25:30 <Diablo-D3> any term opened in X is set to DISPLAY=:n where n is the nth X server running
826 2010-11-26 16:25:30 <UukGoblin> it's a headless and diskless system that I have.
827 2010-11-26 16:25:32 <zc00gii> it's been like, 20 minutes and still no confirmation
828 2010-11-26 16:25:37 <Diablo-D3> UukGoblin: ahh
829 2010-11-26 16:25:38 <UukGoblin> I don't have any terms open in X
830 2010-11-26 16:25:43 <Diablo-D3> UukGoblin: then yes, you need to set :0
831 2010-11-26 16:25:49 <UukGoblin> with m0's, I don't.
832 2010-11-26 16:25:55 <Diablo-D3> UukGoblin: its an lwjgl thing
833 2010-11-26 16:26:00 <UukGoblin> (but I do need X running)
834 2010-11-26 16:26:04 <Diablo-D3> lwjgl starts GL no matter if I use it or not
835 2010-11-26 16:26:17 <UukGoblin> I think it needs it for the awt shit
836 2010-11-26 16:26:39 <UukGoblin> ask the people who reported this bug (and tried swapping the 64-bit libs) on ubuntu 64, if they had DISPLAY=:0
837 2010-11-26 16:26:45 <Diablo-D3> UukGoblin: no, it uses awt for the opengl shit
838 2010-11-26 16:26:56 <Diablo-D3> UukGoblin: most people dont run headless miners
839 2010-11-26 16:27:02 <Diablo-D3> and anyone who does, thats their own problem
840 2010-11-26 16:27:25 <Diablo-D3> and Im actually surprised m0's works without DISPLAY set
841 2010-11-26 16:27:41 <Diablo-D3> because it respects :0.x notation
842 2010-11-26 16:27:51 <Diablo-D3> (it being the ati sdk)
843 2010-11-26 16:28:17 <Diablo-D3> UukGoblin: so, did you get it working right?
844 2010-11-26 16:29:17 <bitanarchy> Is there a possiblity to check the ballance my bitcoin address without exposing my wallet to a computer that is connected to the internet?
845 2010-11-26 16:29:36 <Diablo-D3> bitanarchy: no, not without knowing every address you've ever used
846 2010-11-26 16:30:22 <zc00gii> yay, it's confirmed
847 2010-11-26 16:30:29 <Diablo-D3> bitanarchy: the information in the chain can tell you the current balance of a specific address... which is meaningless
848 2010-11-26 16:30:43 <UukGoblin> Diablo-D3, well it seems to work with -w, but reports crazily large khash (hundreds of millions of khash)
849 2010-11-26 16:30:52 <Diablo-D3> UukGoblin: yeah, thats called "fail mode"
850 2010-11-26 16:30:57 <Diablo-D3> the kernel isnt being ran
851 2010-11-26 16:30:57 <UukGoblin> and I can't set which device I want to use / I don't see the per-device khash rate
852 2010-11-26 16:31:06 <UukGoblin> aaah.
853 2010-11-26 16:31:08 <Diablo-D3> there is no per device usage
854 2010-11-26 16:31:16 <Diablo-D3> it selects all available devices and uses them
855 2010-11-26 16:31:30 <Diablo-D3> something that has been improved over m0's
856 2010-11-26 16:31:32 <UukGoblin> yeah, which is what is fine for most purposes, but for OC and debug it isn't
857 2010-11-26 16:31:45 <Diablo-D3> well, for debug it still is useful BUT
858 2010-11-26 16:31:51 <UukGoblin> at least the ability to limit which cards it uses would be cool
859 2010-11-26 16:31:53 <Diablo-D3> ati names all the cards the same fucking thing
860 2010-11-26 16:32:05 <Diablo-D3> thx ati
861 2010-11-26 16:32:13 <UukGoblin> doesn't matter, I can map ATI's name directly to physical devices using PCI IDs (and aticonfig --lsa)
862 2010-11-26 16:32:18 <bitanarchy> Diablo-D3: why do you say the the information of the amount of bitcoins a certain address owns is meaningless? People may want to know how much bitcoins they own without having to expose their wallet to potential thiefs.
863 2010-11-26 16:32:28 <Diablo-D3> UukGoblin: how? all it says is, say, Juniper 4 times.
864 2010-11-26 16:32:35 <UukGoblin> (and by checking the temperature on them using a pistol thermometer)
865 2010-11-26 16:32:41 <UukGoblin> aticonfig --lsa
866 2010-11-26 16:32:46 <Diablo-D3> I mean the name opencl gives to me
867 2010-11-26 16:32:56 <Diablo-D3> Added ATI RV770 (10 CU, 1x vector, local work size of 64)
868 2010-11-26 16:33:02 <Diablo-D3> as such
869 2010-11-26 16:33:03 <UukGoblin> run aticonfig --lsa
870 2010-11-26 16:33:10 <UukGoblin> you'll see bus IDs
871 2010-11-26 16:33:16 <UukGoblin> or PCI whatever it's called
872 2010-11-26 16:33:17 <Diablo-D3> you're missing the point, UukGoblin
873 2010-11-26 16:33:25 <Diablo-D3> OpenCL doesnt give me that.
874 2010-11-26 16:33:35 <UukGoblin> ah, but with opencl you should have order of devices
875 2010-11-26 16:33:43 <UukGoblin> m0's has -d <number>
876 2010-11-26 16:33:49 <Diablo-D3> his is kind of random
877 2010-11-26 16:33:58 <UukGoblin> and the number is the same as the aticonfig --lsa + 1
878 2010-11-26 16:33:59 <Diablo-D3> its _still_ whatever the sdk returns to him
879 2010-11-26 16:34:10 <Diablo-D3> if the sdk suddenly decided to reverse the order
880 2010-11-26 16:34:17 <UukGoblin> I'm pretty sure the sdk always returns them in the same order
881 2010-11-26 16:34:27 <Diablo-D3> it does as long as the sdk doesnt change
882 2010-11-26 16:34:31 <UukGoblin> (and I can physically check by temperature)
883 2010-11-26 16:34:40 <Diablo-D3> the only thing I can do is check if multiple devices are named the same
884 2010-11-26 16:34:43 <Diablo-D3> which is absurd