1 2011-01-16 00:00:00 <ArtForz> Athlon2 X2
  2 2011-01-16 00:00:05 <ArtForz> 2nd slot is only x4 electrical, but it's cheap
  3 2011-01-16 00:00:17 <hacim> ah
  4 2011-01-16 00:00:30 <Diablo-D3> hacim: phenom IIs without L3, but double the L2.
  5 2011-01-16 00:00:31 <ArtForz> and the slots are decently spaced
  6 2011-01-16 00:00:36 <Raulo> Artforz: Thanks
  7 2011-01-16 00:00:38 <devon_hillard> the GPUs don't really need more than 1x lanes for mining, do they?
  8 2011-01-16 00:00:47 <Diablo-D3> hacim: you typically buy dual core as athlon II and quad core as phenom II
  9 2011-01-16 00:00:54 <ArtForz> well, they lose a tiny bit of performance, but well < 1%
 10 2011-01-16 00:01:03 <ArtForz> theres also quadcore A2s
 11 2011-01-16 00:01:08 <devon_hillard> I saw people use a dremel tool on their 1x pci-ex slot to make a graphics card fit
 12 2011-01-16 00:01:19 <Diablo-D3> yes, theres AII x4s, and PII x2s
 13 2011-01-16 00:01:20 <ArtForz> (they're really PhenomIIs with disabled L3)
 14 2011-01-16 00:01:22 <Diablo-D3> both pointless
 15 2011-01-16 00:01:37 <Diablo-D3> ArtForz: no, they're AII mobile dies
 16 2011-01-16 00:01:42 <Diablo-D3> they still have 2x L2.
 17 2011-01-16 00:01:44 <hacim> why do you want that over actually having L3?
 18 2011-01-16 00:02:04 <Diablo-D3> hacim: the L3 does nothing on so few cores
 19 2011-01-16 00:02:11 <ArtForz> nope
 20 2011-01-16 00:02:22 <hacim> ah
 21 2011-01-16 00:02:23 <ArtForz> check out 645 for example
 22 2011-01-16 00:02:33 <Diablo-D3> hacim: you have slightly better performance per dollar on AII x2 than PII x2
 23 2011-01-16 00:02:48 <Diablo-D3> hacim: but AMD's gigantic dick size kicks in on PII x4
 24 2011-01-16 00:02:55 <ArtForz> theres really plenty of A2s with 512k L2
 25 2011-01-16 00:02:57 <luke-jr> tcatm: hurry up!
 26 2011-01-16 00:03:04 <Diablo-D3> man
 27 2011-01-16 00:03:08 <Diablo-D3> a 16 core Phenom II
 28 2011-01-16 00:03:16 <ArtForz> most of em are normal A2s with half L2 disabled
 29 2011-01-16 00:03:16 <Diablo-D3> would probably explode nutsacks
 30 2011-01-16 00:03:34 <ArtForz> well, there are 12-core opterons...
 31 2011-01-16 00:03:36 <Diablo-D3> you'd just be sitting there and POP
 32 2011-01-16 00:03:41 <Diablo-D3> there goes your nutsack
 33 2011-01-16 00:03:57 <Diablo-D3> ArtForz: but they cheat and have two memory controllers =P
 34 2011-01-16 00:04:26 <ArtForz> anyways, *some* A2-X4s are crippled PhenomIIs
 35 2011-01-16 00:05:05 <devon_hillard> when is the next AMD architecture planned? Q1 2011?
 36 2011-01-16 00:05:14 <Diablo-D3> it already came out, devon
 37 2011-01-16 00:05:20 <ArtForz> shortly after duke nukem forever is released *ducks*
 38 2011-01-16 00:05:24 <devon_hillard> there's one with the APUs and then there was AM3+?
 39 2011-01-16 00:05:26 <Diablo-D3> ArtForz: thats later this year.
 40 2011-01-16 00:05:30 <ArtForz> sorry, Fusion != new CPU arch
 41 2011-01-16 00:05:31 <hacim> i wonder what a cheap mining box would be, probably something that can do dual pci-e and run 5850s and above, with enough of a PSU to handle it
 42 2011-01-16 00:05:38 <ArtForz> it's still the same crappy K10 cores
 43 2011-01-16 00:05:46 <devon_hillard> APUs are for mobile or weak devices, right?
 44 2011-01-16 00:05:47 <niekie> Diablo-D3: I'll believe it when I see it.
 45 2011-01-16 00:05:53 <Diablo-D3> its a new arch in the sense that the GPU part is on die and a native HTX part.
 46 2011-01-16 00:06:05 <Diablo-D3> devon_hillard: no, they'll have opteron fusions too
 47 2011-01-16 00:06:12 <ArtForz> yeah, but not a new microarchitecture
 48 2011-01-16 00:06:15 <Diablo-D3> devon_hillard: like 16 cores and a 68xx worth of GPU
 49 2011-01-16 00:06:30 <ArtForz> = it's still the same CPU cores
 50 2011-01-16 00:06:36 <tcatm> luke-jr: code inserted (without testing)
 51 2011-01-16 00:06:37 <devon_hillard> but right now, only available for low-power stuff
 52 2011-01-16 00:06:38 <hacim> you probably dont need much CPU if you are just GPU mining
 53 2011-01-16 00:06:44 <ArtForz> and K10 is really showing it's age compared clock-for-clock with sandy bridge
 54 2011-01-16 00:06:54 <Diablo-D3> ArtForz: yeah, but sandy vagina is overpriced shit
 55 2011-01-16 00:06:59 <ArtForz> overpriced?
 56 2011-01-16 00:07:06 <Diablo-D3> its slow, its overheating, its expensive, and intel is going to go bankrupt
 57 2011-01-16 00:07:25 <Diablo-D3> maybe if they cut prices in half they can come close to beating AMD
 58 2011-01-16 00:07:28 <devon_hillard> they had enough cash to bribe dell for years :)
 59 2011-01-16 00:07:30 <devon_hillard> and still do
 60 2011-01-16 00:07:37 <ArtForz> it's as fast as a fucking i7-980X for most workloads FFS
 61 2011-01-16 00:07:46 <devon_hillard> and pay those high fines in the US and EU
 62 2011-01-16 00:07:47 <Diablo-D3> devon_hillard: you know a company is fucked if they have to bribe people
 63 2011-01-16 00:07:52 <luke-jr> [20:07:36] <bitcoincharts> TX 1LVndDxUHXAeH6bQ3j7jdvqmZW31kVLT4z 10 TBC, 1CQWT7zgb5gsLDgRkbkGSACZVBektX37Xm 1951424 uBTCents (b3fa1d5bae3df16b72317c5852694e8cc22f3c8f917ee04cc5b1c17dddd54a83)
 64 2011-01-16 00:07:55 <luke-jr> tcatm: looks good there
 65 2011-01-16 00:08:10 <devon_hillard> well, I loved the gall of Dell, who treated Intel like a personal ATM :)
 66 2011-01-16 00:08:11 <ArtForz> well, AMD is nice if you just want a cheap quadcore
 67 2011-01-16 00:08:11 <Diablo-D3> ubtc :D
 68 2011-01-16 00:08:23 <devon_hillard> making their targets, quarter after quarter...
 69 2011-01-16 00:08:25 <Diablo-D3> Intel is nice if you think your wallet is overweight
 70 2011-01-16 00:08:32 <ArtForz> but if you *need* a fast CPU, SB beats anything
 71 2011-01-16 00:08:36 <Diablo-D3> I can buy two AMD CPUs for the cost of an Intel
 72 2011-01-16 00:08:42 <luke-jr> Diablo-D3: #Bitcoin-monitor
 73 2011-01-16 00:08:48 <ArtForz> two *slow* AMD CPUs
 74 2011-01-16 00:08:48 <Diablo-D3> luke-jr: never.
 75 2011-01-16 00:08:55 <Diablo-D3> ArtForz: two FAST amd cpus
 76 2011-01-16 00:08:59 <ArtForz> hexcore PhenomIIs arent exactly cheap
 77 2011-01-16 00:09:06 <luke-jr> Diablo-D3: why not
 78 2011-01-16 00:09:07 <Diablo-D3> sandy vagina isnt fast.
 79 2011-01-16 00:09:14 <Diablo-D3> luke-jr: Im already on three bitcoin channels
 80 2011-01-16 00:09:22 <luke-jr> Diablo-D3: but this one is fun
 81 2011-01-16 00:09:27 <luke-jr> the game is simple:
 82 2011-01-16 00:09:29 <Diablo-D3> thats what these three said
 83 2011-01-16 00:09:38 <luke-jr> watch for tx
 84 2011-01-16 00:09:44 <luke-jr> guess which address is the receiving end
 85 2011-01-16 00:09:47 <ArtForz> it's not fast, it's the fastest per-core on the market
 86 2011-01-16 00:09:51 <luke-jr> then send some extra BTC to confuse them
 87 2011-01-16 00:10:04 <ArtForz> only thing coming close in total throughput are 12-core opterons and 6-core i7s
 88 2011-01-16 00:10:06 <Diablo-D3> ArtForz: straight out? sure... but it uses more watts and costs more dollars
 89 2011-01-16 00:10:09 <Diablo-D3> ArtForz: so why bother
 90 2011-01-16 00:10:20 <ArtForz> more dollars than a $1k i7 ?
 91 2011-01-16 00:10:20 <Diablo-D3> AMD still wins
 92 2011-01-16 00:10:24 <ArtForz> I rather think not
 93 2011-01-16 00:10:40 <Diablo-D3> more dollars than an AMD that scores better performance/dollar or performance/watt.
 94 2011-01-16 00:10:45 <hacim> what does this mean "Military Class/ X6 CPU supported/ CPU CoreUnlock"
 95 2011-01-16 00:10:48 <ArtForz> sorry, it doesn't
 96 2011-01-16 00:10:49 <hacim> http://www.newegg.com/Product/Product.aspx?Item=N82E16813130290&Tpk=MSI%20870-G45
 97 2011-01-16 00:10:55 <Diablo-D3> hacim: just MSI bullshit
 98 2011-01-16 00:11:01 <ArtForz> AMD rules, but they really dropped the ball
 99 2011-01-16 00:11:05 <tcatm> luke-jr: could be more confusing when you send to the change address ;)
100 2011-01-16 00:11:16 <hacim> ah
101 2011-01-16 00:11:24 <luke-jr> tcatm: you'd have to send more, then
102 2011-01-16 00:11:25 <Diablo-D3> hacim: its their solid state shit brand
103 2011-01-16 00:11:32 <Diablo-D3> hacim: gigabyte otoh calls theirs durable
104 2011-01-16 00:11:38 <ArtForz> a 2600K is nearly twice as fast as a PhenomII X6 1100T for heavily threaded apps
105 2011-01-16 00:11:38 <luke-jr> tcatm: since I'm sending 10 TBC quantities, I'm also screwing up the balance #s
106 2011-01-16 00:11:47 <ArtForz> more than wtwice as fast single threaded
107 2011-01-16 00:11:47 <luke-jr> ArtForz: not a 2600?
108 2011-01-16 00:11:48 <Diablo-D3> ArtForz: I doubt it.
109 2011-01-16 00:12:12 <ArtForz> I benchmarked it
110 2011-01-16 00:12:21 <Diablo-D3> and even if it was, its not enough to save intel from bankruptcy
111 2011-01-16 00:12:30 <Diablo-D3> too little too late
112 2011-01-16 00:12:35 <ArtForz> I like PhenomIIs, but SB is just fucking crazy fast
113 2011-01-16 00:12:36 <Diablo-D3> they should have came out with that 10 years ago.
114 2011-01-16 00:12:39 <ArtForz> and they OC like hell
115 2011-01-16 00:13:02 <ArtForz> my 2600K gets 4.4 on air using the stock cooler
116 2011-01-16 00:13:38 <ArtForz> far from the 5GHz+ reviewers are claiming (handpicked press samples?), but from 3.4 stock it's pretty damn decent
117 2011-01-16 00:14:08 <luke-jr> ArtForz: K cant' do VT-d
118 2011-01-16 00:14:10 <Raulo> Is MSI 870A-G54A just an updated version of 870-G45?
119 2011-01-16 00:14:13 <ArtForz> yep
120 2011-01-16 00:14:22 <ArtForz> ....checks
121 2011-01-16 00:14:45 <Raulo> 870-G45s are hard to find in my country
122 2011-01-16 00:15:21 <ArtForz> sems to be the successor model
123 2011-01-16 00:15:45 <ArtForz> newer southbridge, USB3, 6Gb sata
124 2011-01-16 00:16:22 <hacim> Raulo: buying a mining machine?
125 2011-01-16 00:16:38 <ArtForz> btw, why the FUCK do mainboard mfgs put the 1st x16 slot as slot #2?
126 2011-01-16 00:16:40 <Raulo> It's more expensive but I can't find any reasonable 870-G45s
127 2011-01-16 00:16:41 <LobsterMan> ;;bc,stats
128 2011-01-16 00:16:43 <gribble> Current Blocks: 102889 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 1942 blocks | Next Difficulty In About: 1 week, 4 days, 22 hours, 26 minutes, and 42 seconds | Next Difficulty Estimate: 21028.90627601
129 2011-01-16 00:16:51 <LobsterMan> ;;bc,calc 108000
130 2011-01-16 00:16:52 <gribble> The average time to generate a block at 108000 Khps, given current difficulty of 18437.64439217 , is 1 week, 1 day, 11 hours, 40 minutes, and 32 seconds
131 2011-01-16 00:16:55 <Raulo> Thanks for the help.
132 2011-01-16 00:17:16 <Raulo> I hope I would find an old box that works with 5xxx, though
133 2011-01-16 00:17:21 <ArtForz> it makes no sucking fence
134 2011-01-16 00:17:24 <ArtForz> *fucking sense
135 2011-01-16 00:17:30 <hacim> sucking fence, lol
136 2011-01-16 00:19:28 <ArtForz> seriously, I sometimes want to hit those fuckwits over the head with the fucking ATX spec
137 2011-01-16 00:19:35 <tcatm> luke-jr: UnicodeEncodeError: 'ascii' codec can't encode character u'ue9dc' in position 65: ordinal not in range(128)
138 2011-01-16 00:19:47 <hacim> Raulo: if you find a cheap configuration that works with the 5xxx cards, i'm interested to know what you find
139 2011-01-16 00:19:55 <ArtForz> preferably engraved in a nice solid tungsten plate
140 2011-01-16 00:21:15 <ArtForz> as I suspect plain steel won't have the neccesary oomph to get the message through
141 2011-01-16 00:25:15 <luke-jr> tcatm: do you understand that?
142 2011-01-16 00:25:30 <luke-jr> tcatm: sounds like a problem with the IRC end
143 2011-01-16 00:26:59 <luke-jr> tcatm: what are you using for that?
144 2011-01-16 00:27:33 <tcatm> luke-jr: supybot
145 2011-01-16 00:28:33 <luke-jr> tcatm: plz paste your irc.reply or whatevr lin
146 2011-01-16 00:28:34 <luke-jr> line
147 2011-01-16 00:28:48 <tcatm> self.irc.queueMsg(ircmsgs.privmsg(self.plugin.channel, msg))
148 2011-01-16 00:29:28 <luke-jr> self.irc.queueMsg(ircmsgs.privmsg(self.plugin.channel, msg.encode('utf-8')))
149 2011-01-16 00:29:29 <luke-jr> try that
150 2011-01-16 00:30:03 <tcatm> changed
151 2011-01-16 00:30:10 <luke-jr> tcatm: what's your donation address?
152 2011-01-16 00:30:35 <tcatm> 17asVKkzRGTFvvGH9dMGQaHe78xzfvgSSA
153 2011-01-16 00:31:04 <luke-jr> sent you 1?? TBC
154 2011-01-16 00:31:29 <tcatm> seems to work
155 2011-01-16 00:31:34 <luke-jr> yep
156 2011-01-16 00:31:49 <luke-jr> [20:31:21] <bitcoincharts> TX 17asVKkzRGTFvvGH9dMGQaHe78xzfvgSSA 1?? TBC, 1JntVnVu42ip6xPmLke8GnK3CSaQhsquFD 1164992 uBTCents afb99fffbed3f594cd49f7bbed107fb15c06b43f92ffa4a9486dbf5db49a2cc1
157 2011-01-16 00:33:21 <tcatm> which client are you using?
158 2011-01-16 00:33:38 <luke-jr> the original one, with bugs fixed
159 2011-01-16 00:33:50 <tcatm> patches?
160 2011-01-16 00:33:54 <luke-jr> yeah
161 2011-01-16 00:34:06 <tcatm> are they public?
162 2011-01-16 00:34:11 <luke-jr> yep
163 2011-01-16 00:34:24 <luke-jr> lp:~luke-jr/+junk/bitcoin-tonal
164 2011-01-16 00:34:55 <luke-jr> bzr branch lp:~luke-jr/+junk/bitcoin-tonal && cd bitcoin-tonal && qmake && make
165 2011-01-16 00:35:02 <luke-jr> and restart bitcoind with the new one
166 2011-01-16 00:35:23 <luke-jr> tested only on Gentoo so far
167 2011-01-16 00:35:39 <luke-jr> the old Makefiles should still work tho
168 2011-01-16 00:36:03 <luke-jr> it's not really tonal, you still need to specify BTC
169 2011-01-16 00:36:12 <luke-jr> but it allows fractions that have tonal values
170 2011-01-16 00:36:22 <luke-jr> and avoids fees
171 2011-01-16 00:37:04 <luke-jr> fyi, the amount I sent you is properly spoken "ton-vy bitcoins"
172 2011-01-16 00:38:19 <tcatm> mkay ;)
173 2011-01-16 00:39:15 <Slix`> How are bitcoin addresses encoded? Base64?
174 2011-01-16 00:39:22 <tcatm> custom Base58
175 2011-01-16 00:39:48 <Slix`> Why Base58?
176 2011-01-16 00:40:07 <luke-jr> Slix`: I learned by now not to ask "why" questions about bitcoin
177 2011-01-16 00:40:16 <luke-jr> a lot of it is quite frankly retarded XD
178 2011-01-16 00:40:32 <tcatm> Slix`: to avoid confusion with similiar letters like l and I
179 2011-01-16 00:41:00 <Slix`> Ahh. I guess that's a decent reason.
180 2011-01-16 00:41:19 <luke-jr> why is bitcoin protocol mixed with big and little endian?
181 2011-01-16 00:41:28 <ArtForz> it is?
182 2011-01-16 00:41:36 <luke-jr> yes
183 2011-01-16 00:41:47 <ArtForz> last time I checked we used LE everywhere except for ip+port in addr()
184 2011-01-16 00:41:59 <luke-jr> LE & except & is mixing :P
185 2011-01-16 00:42:30 <ArtForz> you'd rather have big-endian *IPs* ?!?
186 2011-01-16 00:42:34 <ArtForz> *liitle.-.endian
187 2011-01-16 00:42:38 <luke-jr> rather than mixed, yes
188 2011-01-16 00:42:45 <ArtForz> thats fucking backwards
189 2011-01-16 00:42:50 <ArtForz> big-endian IP makes sense
190 2011-01-16 00:42:51 <luke-jr> LE is, yes
191 2011-01-16 00:43:06 <ArtForz> because in octets it's the right way around
192 2011-01-16 00:43:18 <luke-jr> can the protocol negotiate the rest?
193 2011-01-16 00:43:21 <ArtForz> having BE port is... weird
194 2011-01-16 00:43:29 <luke-jr> eg, have new versions be pure BE?
195 2011-01-16 00:43:37 <ArtForz> errr... no
196 2011-01-16 00:43:53 <ArtForz> Block and TX data pretty much has to stay LE unless yo uwant to break the chain
197 2011-01-16 00:44:09 <luke-jr> can't convert on-the-fly?
198 2011-01-16 00:44:18 <ArtForz> sure, or byteswap everytime you [de]serialize
199 2011-01-16 00:44:59 <ArtForz> but imo "LE everywhere, except for ip/port pairs" is simple enough to support
200 2011-01-16 00:46:31 <slush> http://img17.allegroimg.pl/photos/400x300/14/06/48/75/1406487515
201 2011-01-16 00:47:09 <ArtForz> never seen something like that
202 2011-01-16 00:47:20 <ArtForz> this is either some weird prototype or a 'shop
203 2011-01-16 00:47:25 <slush> 12x display port and 3 slots to high
204 2011-01-16 00:48:24 <ArtForz> looks like a prototype eyefinity 5970
205 2011-01-16 00:49:07 <ArtForz> yep
206 2011-01-16 00:49:12 <slush> http://www.engadget.com/2010/04/30/powercolor-hd5970-eyefinity-12-makes-six-screens-yesterdays-new/
207 2011-01-16 00:49:21 <slush> yes, looks like this is even same picture
208 2011-01-16 00:49:35 <ArtForz> yep, crop from 1st pic
209 2011-01-16 00:49:51 <slush> oh, cool, the third row of display ports can be dismounted
210 2011-01-16 00:50:28 <ArtForz> also looks like it has 2*2GB
211 2011-01-16 00:50:49 <ArtForz> normal 5970 only has 8 ram chips per GPU
212 2011-01-16 00:51:15 <slush> is 360 euro fair price for this?
213 2011-01-16 00:51:47 <Diablo-D3> slush: its a powercolor
214 2011-01-16 00:51:49 <Diablo-D3> fuck. that.
215 2011-01-16 00:52:09 <ArtForz> not to mention it looks like it never went into production
216 2011-01-16 00:52:12 <slush> well, chips are the same, right?
217 2011-01-16 00:52:26 <Diablo-D3> and why the fuck are they selling that shit that way
218 2011-01-16 00:52:41 <slush> I see one in Czech bazaar
219 2011-01-16 00:52:44 <luke-jr> http://blockexplorer.com/tx/a288fec5559c3f73fd3d93db8e8460562ebfe2fcf04a5114e8d0f2920a6270dc
220 2011-01-16 00:52:46 <luke-jr> WTF Is this
221 2011-01-16 00:52:50 <luke-jr> ArtForz?
222 2011-01-16 00:52:53 <slush> so it was probably in production...
223 2011-01-16 00:52:57 <molec> just got kicked
224 2011-01-16 00:53:03 <Diablo-D3> slush: its dumb.
225 2011-01-16 00:53:13 <ArtForz> "a transaction"
226 2011-01-16 00:53:18 <Diablo-D3> you'd be better off buying two cards with 6
227 2011-01-16 00:53:29 <ArtForz> we have 2 or 3 of these in the block chain
228 2011-01-16 00:53:30 <molec> ArtForz, what's your timezone?
229 2011-01-16 00:53:33 <luke-jr> ArtForz: a blackhole transaction?
230 2011-01-16 00:53:40 <ArtForz> some joker tried to consume a lot of CPU with duplicate OP_CHECKSIGS
231 2011-01-16 00:53:47 <ArtForz> GMT+1
232 2011-01-16 00:53:58 <molec> ArtForz, ah, hello!
233 2011-01-16 00:54:10 <slush> everybody in GMT+1 should be in bed now
234 2011-01-16 00:54:13 <slush> including me
235 2011-01-16 00:54:20 <molec> slush, no!
236 2011-01-16 00:54:39 <ArtForz> who cares, it's saturday... err... sunday
237 2011-01-16 00:54:46 <slush> :-D
238 2011-01-16 00:54:53 <molec> I tried my girlfriends yoga-lesson last week. since then, I can't sleep before 5:30
239 2011-01-16 00:55:04 <Diablo-D3> http://www.engadget.com/2011/01/15/man-discovers-glasses-free-3d-tech-in-the-blink-of-an-eye-video/
240 2011-01-16 00:55:05 <slush> Diablo-D3: well, is there any specific reason why reject this card? Except you don't like manufacturer? :)
241 2011-01-16 00:55:07 <Diablo-D3> this is disturbing
242 2011-01-16 00:55:14 <Diablo-D3> slush: cant trust the driver suport
243 2011-01-16 00:55:39 <ArtForz> shrug, I'd get one for the novelty value, but BIOs and driver support are probably flaky if existing at all
244 2011-01-16 00:55:41 <Diablo-D3> they're using the 6 display port controllers off the 2nd gpu
245 2011-01-16 00:55:45 <ArtForz> yep
246 2011-01-16 00:55:53 <Diablo-D3> it was never meant to be used that way
247 2011-01-16 00:55:54 <luke-jr> ArtForz: like you?
248 2011-01-16 00:56:05 <ArtForz> does 12-screen eyefinity with 2 5870s work?
249 2011-01-16 00:56:11 <Diablo-D3> also they're fucks
250 2011-01-16 00:56:16 <luke-jr> ArtForz: cmon, we know you're the only one capable of generating such tx
251 2011-01-16 00:56:21 <Diablo-D3> they should have put it on the vent slot
252 2011-01-16 00:56:23 <ArtForz> nope, wasnt me
253 2011-01-16 00:56:29 <Diablo-D3> and then put vents between each display port plug
254 2011-01-16 00:56:40 <ArtForz> I created a bunch of weird tx on testnet though
255 2011-01-16 00:56:58 <slush> Diablo-D3: ok, I accept it, quite good reasons
256 2011-01-16 00:57:09 <slush> too bad there are only few 5970 on market
257 2011-01-16 00:57:29 <Diablo-D3> slush: well, they can do the same trick with a 6990 anyhow
258 2011-01-16 00:57:30 <ArtForz> ?
259 2011-01-16 00:57:39 <Diablo-D3> they just connected the displayport pins
260 2011-01-16 00:57:42 <ArtForz> german etailers still have fucktons of em
261 2011-01-16 00:58:25 <slush> ArtForz: but for what price? In czech new cards are fucking expensive and old card are... who knows where?
262 2011-01-16 00:58:35 <ArtForz> 415EUR incl. tax
263 2011-01-16 00:58:42 <slush> ArtForz: new?
264 2011-01-16 00:59:09 <ArtForz> http://www.hoh.de/PC-Komponenten/Grafikkarten/ATI/PCIe/HD5970/Sapphire-HD-5970-2048MB-GDDR5-PCIe-Full-Retail_i9199_124724.htm
265 2011-01-16 00:59:18 <slush> in czech 550 euro and more
266 2011-01-16 01:00:10 <ArtForz> wow, that sucks
267 2011-01-16 01:00:32 <ArtForz> but yeah, seems .cz gets raped for tech prices even more than .de
268 2011-01-16 01:00:40 <ArtForz> still better than australia though ;)
269 2011-01-16 01:00:42 <slush> well, just find somebody who knows german and its done
270 2011-01-16 01:00:51 <slush> mostly because we don't have euro, I think
271 2011-01-16 01:01:13 <ArtForz> not really, iirc I saw a bunch in france for 380EUR
272 2011-01-16 01:01:32 <slush> ohh, nice price
273 2011-01-16 01:02:10 <ArtForz> hmm... looks like HOH is actually the last shop with the 400EUR 5970s in stock
274 2011-01-16 01:02:25 <ArtForz> there were about half a dozen shops that still had em in stock like a week ago
275 2011-01-16 01:02:39 <slush> lol, just found eshop with 850euro for 5970
276 2011-01-16 01:03:09 <slush> (in czech, of course)
277 2011-01-16 01:03:15 <ArtForz> wow.
278 2011-01-16 01:03:32 <ArtForz> "normal" 5970s are ~500EUR in .de
279 2011-01-16 01:03:38 <ArtForz> well, 500-550
280 2011-01-16 01:03:51 <ArtForz> only the sapphires were that cheap for a while
281 2011-01-16 01:04:22 <ArtForz> as sapphire is AMDs fav. OEM, I guess they got the last remaining stock of cypress chips
282 2011-01-16 01:04:38 <Diablo-D3> s/AMD/ATI/ s/ATI/AMD/ *cough*
283 2011-01-16 01:04:58 <ArtForz> s/AMD/DAAMIT/
284 2011-01-16 01:05:08 <Diablo-D3> hee
285 2011-01-16 01:05:38 <ArtForz> we're red! we're green! we're in between!
286 2011-01-16 01:05:48 <slush> http://www.anacomp.cz/gigabyte-hd5870-eyefinityx6-edition-gv-r597d5-2gd-b-2gb-gddr5-256bit-pci-e-6xmini-display-port_d44148.html
287 2011-01-16 01:05:58 <slush> I ask because it is cheap and new and in czech :)
288 2011-01-16 01:06:28 <ArtForz> anyways, sapphire builds all of AMD/ATI/he-who-shall-not-be-nameds reference cards
289 2011-01-16 01:06:49 <ArtForz> thats a normal x6 eyefinity
290 2011-01-16 01:06:52 <ArtForz> 5870
291 2011-01-16 01:07:05 <slush> oh, its 5870, soooryyy
292 2011-01-16 01:07:23 <slush> That explains a lot (at least price)
293 2011-01-16 01:09:23 <ArtForz> hmmm... looks like 5850s might actually be a better deal
294 2011-01-16 01:09:33 <ArtForz> 160EUR in .de
295 2011-01-16 01:10:03 <ArtForz> not bad considering a 5970 is only 2.23x as fast
296 2011-01-16 01:11:32 <ArtForz> so you need about twice the mainboards+CPU+ram, same amount of PSUs (4*5850 on one 800W shouldnt be a problem)
297 2011-01-16 01:12:43 <slush> maybe cheaper, but what about power efficiency?
298 2011-01-16 01:12:57 <ArtForz> 5850 is nearly as good as 5970
299 2011-01-16 01:13:01 <ArtForz> better than 5870
300 2011-01-16 01:13:50 <ArtForz> the official 151W figure for 5850 is too high, they're closer to 130-135W
301 2011-01-16 01:14:22 <ArtForz> guess AMD wanted some headroom if they had bad chips that needed higher voltage to be stable or something
302 2011-01-16 01:15:20 <ArtForz> cheap dual x16 board + cpu + ram is ~120 EUR
303 2011-01-16 01:15:32 <ArtForz> so 60 EUR per card
304 2011-01-16 01:16:03 <molec> ArtForz, wohnst du in DE?
305 2011-01-16 01:16:12 <ArtForz> yep, nordbayern
306 2011-01-16 01:16:25 <molec> w??rz burg?
307 2011-01-16 01:16:26 <ArtForz> yeah, 5850 should actually end up cheaper
308 2011-01-16 01:16:41 <ArtForz> n st??ck s??dlich von regensburg
309 2011-01-16 01:16:54 <cdecker> Nice I see we have a lot of German guys here :D
310 2011-01-16 01:17:26 <ArtForz> yeah, kinda crazy how many germans like bitcoin
311 2011-01-16 01:17:33 <cdecker> hehe
312 2011-01-16 01:17:52 <cdecker> Euro's not doing so fine
313 2011-01-16 01:18:06 <cdecker> I shouldn't complain
314 2011-01-16 01:18:14 <cdecker> I live in CHF land :D
315 2011-01-16 01:18:33 <molec> cdecker, where in Hamburg?
316 2011-01-16 01:18:49 <cdecker> Long time since I lived in Hamburg
317 2011-01-16 01:18:55 <molec> am in eimsb??ttel
318 2011-01-16 01:19:04 <cdecker> But I still like to call myself a hamburger :D
319 2011-01-16 01:19:05 <molec> allright
320 2011-01-16 01:19:36 <molec> tcatm: not much choice... 100km north of hamburg.
321 2011-01-16 01:19:39 <ArtForz> anyways, from US PoV we're supposed to be liberal hippie commies or something ;)
322 2011-01-16 01:19:44 <cdecker> Nice should meet some day :D
323 2011-01-16 01:19:52 <cdecker> Anyway got to go
324 2011-01-16 01:19:58 <cdecker> See you soon ^^
325 2011-01-16 01:20:23 <molec> when I was in the stats in 1992, some people seriously asked me, whether we had (electrical) power in germany.
326 2011-01-16 01:20:32 <molec> cu, cdecker
327 2011-01-16 01:20:44 <ArtForz> don't forget about the medical death boards ;)
328 2011-01-16 01:20:45 <molec> s/stats/states
329 2011-01-16 01:21:28 <ArtForz> maybe something with the hacker (counter)culture in .de ?
330 2011-01-16 01:21:52 <molec> not sure, something in their past about avoiding authority??
331 2011-01-16 01:22:00 <ArtForz> yep
332 2011-01-16 01:22:18 <molec> ArtForz, would love to meet you some time in person
333 2011-01-16 01:22:20 <ArtForz> overall german hackers seem to like decentralized crypto-anything ;)
334 2011-01-16 01:22:27 <molec> yeah
335 2011-01-16 01:22:27 <slush> hmm, 5850 is nowhere in bazaars and new one is more than 50% of 5970 here. So when I count in also more needed slots and CPUs, it perform worse than 5970 for me
336 2011-01-16 01:22:40 <molec> also: you should demo your ASICs next year at 28c3
337 2011-01-16 01:22:42 <ArtForz> freenet, tor, I2P, now bitcoin
338 2011-01-16 01:22:45 <slush> yes, also tor is very popular in german
339 2011-01-16 01:22:54 <molec> also popular in iran
340 2011-01-16 01:22:58 <ArtForz> I maybe will
341 2011-01-16 01:23:15 <ArtForz> also, the guys presentation on DES on FPGAs @ 27C3 was ... bad
342 2011-01-16 01:23:20 <molec> if you want any help, I'm here!
343 2011-01-16 01:23:25 <ArtForz> several HUGE factual errors
344 2011-01-16 01:23:40 <molec> didn't see that (yet)
345 2011-01-16 01:23:58 <ArtForz> his claimed numbers for DES on GPU were 1/20 of what I could get on a 5970 without even trying
346 2011-01-16 01:24:58 <slush> but yes, 5850 looks like better deal than 5870, definitely
347 2011-01-16 01:25:35 <molec> can you point me to the vid of the talk?
348 2011-01-16 01:25:51 <ArtForz> I only watched the live stream
349 2011-01-16 01:25:59 <molec> allright... they guy's name?
350 2011-01-16 01:26:07 <luke-jr> ;;bc,calc 4205
351 2011-01-16 01:26:08 <gribble> The average time to generate a block at 4205 Khps, given current difficulty of 18437.64439217 , is 31 weeks, 0 days, 23 hours, 8 minutes, and 43 seconds
352 2011-01-16 01:26:10 <ArtForz> http://events.ccc.de/congress/2010/Fahrplan/events/4203.en.html
353 2011-01-16 01:26:47 <molec> thx
354 2011-01-16 01:27:24 <ArtForz> I kinda liked his idea of using JTAG for communication
355 2011-01-16 01:27:53 <molec> nice idea to harvest old electronics for fpgas
356 2011-01-16 01:28:01 <ArtForz> yep
357 2011-01-16 01:28:21 <ArtForz> it's pretty slow, but you pretty much can't get lower pincount, and finding JTAG on old boards is usually pretty easy
358 2011-01-16 01:28:34 <molec> but less effective than buying asics if you have the $
359 2011-01-16 01:28:40 <ArtForz> depends
360 2011-01-16 01:28:50 <ArtForz> if you get a good price, it could be on par
361 2011-01-16 01:28:57 <molec> u think?
362 2011-01-16 01:29:03 <ArtForz> yep
363 2011-01-16 01:29:12 <ArtForz> problem is you really need a quite big FPGA to fit a unrolled sha256 engine, let alone 2
364 2011-01-16 01:30:00 <molec> if u use a small fpga @(say)200mhz, say, too small to fit sha-256 pipeline, what will you get?
365 2011-01-16 01:30:20 <ArtForz> one bitcoinhash is 2 sha256 blocks
366 2011-01-16 01:30:40 <ArtForz> so if you can fit 1/2 of a block pipelined, you get 1 hash / 4 clocks
367 2011-01-16 01:30:56 <molec> 1 has / 4 clocks sounds good
368 2011-01-16 01:31:18 <ArtForz> getting >120MHz out of a spartan6 is a small miracle :/
369 2011-01-16 01:31:28 <ArtForz> and spartan3 is a lot slower still
370 2011-01-16 01:32:16 <molec> does the max clock somehow depend on the logic implemented on the fpga?
371 2011-01-16 01:32:24 <ArtForz> yes
372 2011-01-16 01:32:56 <ArtForz> a single layer of logic+flipflop can get 350MHz+ on a S6
373 2011-01-16 01:33:08 <molec> do you have some old fpga dev-board laying around that I could use??
374 2011-01-16 01:33:11 <molec> ... just to get the hang=
375 2011-01-16 01:33:14 <ArtForz> nope
376 2011-01-16 01:33:18 <molec> s/=/.
377 2011-01-16 01:33:21 <molec> ok
378 2011-01-16 01:33:24 <ArtForz> I only have 3
379 2011-01-16 01:33:35 <molec> can you recommend how I could search on ebay?
380 2011-01-16 01:33:53 <ArtForz> thats kinda tricky, the talk explains it pretty well
381 2011-01-16 01:34:16 <molec> I really enjoiyed making that 8bit cpu in vhdl and seeing that run at UNI-stgt
382 2011-01-16 01:34:36 <ArtForz> yep, seeing a VHDL design work in hardware is pretty amazing
383 2011-01-16 01:34:54 <ArtForz> at least for the first few times
384 2011-01-16 01:34:59 <molec> only ran @1Mhz, but still, it was like a baby born and living
385 2011-01-16 01:35:16 <molec> 1khz?
386 2011-01-16 01:35:30 <molec> dunno, it was really slow
387 2011-01-16 01:35:40 <molec> some bug in the c-code that setup the shit, not my fault ;I)
388 2011-01-16 01:35:48 <ArtForz> kinda like building your first electric vehicle, or completely overhauling your first engine
389 2011-01-16 01:36:03 <molec> never did one of these
390 2011-01-16 01:36:15 <ArtForz> did all of them
391 2011-01-16 01:36:19 <molec> might be easier to just fuck some girl ;)
392 2011-01-16 01:36:54 <molec> ArtForz, if you ever around Hamburg, and looking for place to sleep, call 0176 6126597
393 2011-01-16 01:36:57 <ArtForz> completely rebuilt the diesel engine in my '87 VW T3
394 2011-01-16 01:37:22 <molec> sorry, got the nubmer wrong
395 2011-01-16 01:37:24 <ArtForz> ex-BW sani-bus
396 2011-01-16 01:37:26 <molec> 0176 61265697
397 2011-01-16 01:39:24 <ArtForz> my daily driver is a '89 ford transit ;)
398 2011-01-16 01:40:03 <ArtForz> old car = 1.5kEUR/yr+ tax
399 2011-01-16 01:40:24 <ArtForz> old "small truck" ~ 120EUR/yr tax
400 2011-01-16 01:42:20 <necrodearia> http://www.youtube.com/user/stefbot#p/u/14/7R4ppEADvm4
401 2011-01-16 01:43:12 <molec> what's your work to require a daily drive?
402 2011-01-16 01:43:23 <molec> you seem to be making enough BTC to live from it
403 2011-01-16 01:44:04 <necrodearia> and http://www.youtube.com/user/stefbot#p/u/12/Cu9HUBs5mp0
404 2011-01-16 01:44:23 <ArtForz> I'm doing electronic/control design for printing/binding machines
405 2011-01-16 01:44:43 <ArtForz> = on-demand one-off book runs
406 2011-01-16 01:45:21 <ArtForz> pretty cool tech, and I have pretty much the coolest boss ever
407 2011-01-16 01:46:32 <ArtForz> no fixed work hours *at all*
408 2011-01-16 01:47:16 <ArtForz> work at home, come in 2 days a week at 2-7AM, who cares as long as the job gets done
409 2011-01-16 01:48:30 <molec> you come in at 02:00-0700 ?
410 2011-01-16 01:48:46 <ArtForz> if I want to, yes
411 2011-01-16 01:49:05 <molec> you have this insanely nice job
412 2011-01-16 01:49:10 <ArtForz> or work 9-5 too or just work from home if I want to
413 2011-01-16 01:49:12 <molec> envy!
414 2011-01-16 01:49:22 <molec> arsch!
415 2011-01-16 01:49:24 <molec> :)
416 2011-01-16 01:49:31 <ArtForz> like I said. coolest. boss. ever.
417 2011-01-16 01:49:51 <molec> nice
418 2011-01-16 01:50:03 <molec> what more can you ask for?
419 2011-01-16 01:50:15 <ArtForz> nothing really, I'm pretty happy
420 2011-01-16 01:50:45 <molec> we were not put in this world to be happy
421 2011-01-16 01:51:03 <ArtForz> well, if you're not happy, you're doing it wrong ;)
422 2011-01-16 01:51:11 <molec> there must be something to solve/do/fuck
423 2011-01-16 01:51:34 <ArtForz> aka strive for greatness, be content with mediocrity
424 2011-01-16 01:51:35 <molec> I'm happy.... just not happy enough
425 2011-01-16 02:01:07 <luke-jr> [21:44:28] <bitcoincharts> TX 1NHeFo9B5VUsAWw42P6f2Q7nhdh6CJbQy 1239744 uBTCents, 1KpLXJqQLSx87sjns33AKgZDP9cSDuKBXx ???? TBC bdbea3ba47130140214620ae1775fd334ac07478e2abe58d91e8350351d97877
426 2011-01-16 02:01:36 <tcatm> Is there anyone else using TBC?
427 2011-01-16 02:01:43 <ArtForz> I like this guy
428 2011-01-16 02:02:05 <luke-jr> tcatm: with time :
429 2011-01-16 02:02:18 <luke-jr> tcatm: right now, lack of clients makes it difficult
430 2011-01-16 02:02:26 <tcatm> luke-jr: So you're the only one? :P
431 2011-01-16 02:02:38 <luke-jr> tcatm: I've been giving prizes in TBC, and such :
432 2011-01-16 02:03:05 <tcatm> => lots of subcent change that'll end up in fees. Good for miners :)
433 2011-01-16 02:03:19 <luke-jr> yeah, until people get fixed clients :P
434 2011-01-16 02:06:58 <luke-jr> I must say, marioxcc's idea would have been a really bad idea :P
435 2011-01-16 02:07:11 <tcatm> What is his idea?
436 2011-01-16 02:07:13 <luke-jr> doing it the way he wanted, would have made an algorithm to detect BTC vs TBC impossible
437 2011-01-16 02:07:17 <luke-jr> he wanted 1 BTC = 1 TBC
438 2011-01-16 02:09:55 <niekie> What on earth is TBC.
439 2011-01-16 02:10:04 <luke-jr> niekie: Tonal BitCoin
440 2011-01-16 02:10:08 <luke-jr> as opposed to Decimal BitCoin
441 2011-01-16 02:10:16 <nameless> |I want 1NBC
442 2011-01-16 02:10:19 <niekie> So using another numbering base?
443 2011-01-16 02:10:24 <luke-jr> niekie: yes
444 2011-01-16 02:10:26 <nameless> |1NBC = 9999999999999999999999999999999999999999999BTC
445 2011-01-16 02:10:33 <niekie> What exactly are the advantages in that?
446 2011-01-16 02:10:34 <luke-jr> nameless|: impossible then
447 2011-01-16 02:10:38 <luke-jr> niekie: decimal sucks
448 2011-01-16 02:10:40 <nameless> |luke-jr: nonsense
449 2011-01-16 02:10:49 <nameless> |I just need more processing power than all of you combined
450 2011-01-16 02:10:59 <luke-jr> nameless|: BTC is capped at 21mil
451 2011-01-16 02:11:06 <nameless> |luke-jr: Like I said
452 2011-01-16 02:11:10 <luke-jr> tcatm: I thought I did
453 2011-01-16 02:11:23 <tcatm> (Tera Bitcoin)
454 2011-01-16 02:11:36 <luke-jr> lame SI/Decimal
455 2011-01-16 03:14:27 <newsham> *raises hands in air and yells*KHANNNNNN!!!!
456 2011-01-16 04:32:30 <malfy> why is there a connection to an irc network/channel from the bitcoin daemon?
457 2011-01-16 04:33:09 <nanotube> it's used for bootstrapping.
458 2011-01-16 04:34:08 <malfy> i dont understand
459 2011-01-16 04:34:24 <malfy> isn't it supposed to be a p2p network?
460 2011-01-16 04:35:13 <malfy> there should be no need for a huge channel
461 2011-01-16 04:36:18 <nanotube> yes, but any p2p network needs a method to bootstrap, to find other peers.
462 2011-01-16 04:36:26 <nanotube> you don't need irc if you don't want it, you can run with -noirc. there's a hardcoded list of seednodes.
463 2011-01-16 04:36:39 <malfy> ok
464 2011-01-16 05:02:34 <presence> any mining.bitcoin.cz guys on?
465 2011-01-16 05:06:13 <presence> it seems that your account pw and your worker pws have to align
466 2011-01-16 05:06:30 <presence> yet it displays the worker pw on the account url...which I would prefer not to reveal.
467 2011-01-16 05:19:26 <andrew12> aww
468 2011-01-16 05:19:36 <andrew12> I wanted to say something witty like 'people can mine for you ;)'
469 2011-01-16 06:03:00 <hacim> what kind of wattage is needed in a PSU to fire 5xxx cards?
470 2011-01-16 06:05:55 <RazielZ> One 5xxx card?
471 2011-01-16 06:05:58 <RazielZ> And which one?
472 2011-01-16 06:05:59 <RazielZ> ._.
473 2011-01-16 06:06:00 <mrb_> hacim: depends. a 5450 only needs 19W for itself. a 5970 needs 300W for itself. a power-efficient machine may draw no more than 80W under full load so in theory, even with a high-end 5970, a 380W PSU is sufficient.
474 2011-01-16 06:06:15 <RazielZ> 380W?
475 2011-01-16 06:06:20 <RazielZ> Yeah, go for atleast 800
476 2011-01-16 06:06:23 <RazielZ> :D
477 2011-01-16 06:06:39 <mrb_> I have done that myself (400W PSU for a 5970)
478 2011-01-16 06:06:53 <RazielZ> What the nts
479 2011-01-16 06:06:54 <RazielZ> nuts even
480 2011-01-16 06:07:24 <mrb_> only possible to do safely... if you have a clamp-meter and knows exactly how many Amps you are drawing :)
481 2011-01-16 06:07:24 <RazielZ> I'd be too scareed to power my PC with anything less than 750W
482 2011-01-16 06:07:25 <RazielZ> ._.
483 2011-01-16 06:07:54 <mrb_> I have also run 4 x 5970 at full load with only two 560W PSU
484 2011-01-16 06:08:12 <RazielZ> 4 5970
485 2011-01-16 06:08:17 <RazielZ> want.
486 2011-01-16 06:09:22 <RazielZ> Really, if you have any spares you can send em here, I won't mind
487 2011-01-16 06:14:37 <hacim> ok, so two 5870s would be fine on a 500w or so, just shopping around for a PSU
488 2011-01-16 06:15:39 <mrb_> hacim: watch the max Amp on the 12V rail. two 5870s = 2*188W = 31 Amp at 12V
489 2011-01-16 06:15:59 <mrb_> I bet some crappy 500W PSU may not even be capable of supplying 31A@12V
490 2011-01-16 06:16:15 <mrb_> on the other hand some high-end 400W ones may be able to do it
491 2011-01-16 06:16:21 <mrb_> read the specs.
492 2011-01-16 06:16:47 <hacim> mrb_: what is the max amp on the 12v?
493 2011-01-16 06:18:03 <hacim> so, eg. this corsair 500w is +12V@34A, so it could handle the two 5870s
494 2011-01-16 06:18:21 <mrb_> yep this one seems good
495 2011-01-16 06:18:45 <mrb_> although that leaves only 3 Amps for the rest of the system
496 2011-01-16 06:18:56 <mrb_> 3*12 = 36 Watt
497 2011-01-16 06:19:18 <mrb_> if you have an energy efficient CPU, that may be ok
498 2011-01-16 06:19:56 <RazielZ> also if you don't plan on overclocking
499 2011-01-16 06:20:09 <mrb_> on the other hand you are unlikely to hit the maximum amp on two 5870, even when mining 24/7
500 2011-01-16 06:20:38 <mrb_> maximum *theoretical* amp
501 2011-01-16 06:20:56 <hacim> hmm, so maybe that is cutting it too close?
502 2011-01-16 06:21:13 <mrb_> I dunno, what's your hardware?
503 2011-01-16 06:22:25 <mrb_> if you want to be comfortable, leave 10 Amps to the rest of the system. 120W will account for a 95W CPU at full load plus a ~30W on the mobo/ram/hdd
504 2011-01-16 06:22:49 <hacim> mrb: haven't bought anything yet, but this is what I was looking at: https://micah.riseup.net/shots/2011.01.16-02.21.1295162515.xTjPuKdczM.png
505 2011-01-16 06:23:18 <hacim> i welcome any suggestions for improvements
506 2011-01-16 06:23:25 <mrb_> 95W CPU.
507 2011-01-16 06:24:05 <mrb_> yeah look for a PSU capable of at least 40A@12V
508 2011-01-16 06:24:15 <hacim> perhaps I can get a cheaper/lower power cpu
509 2011-01-16 06:24:16 <RazielZ> that RAM is cheap D:
510 2011-01-16 06:24:29 <hacim> RazielZ: yeah, I'm worried it might be shit
511 2011-01-16 06:24:40 <RazielZ> I don't know...
512 2011-01-16 06:24:44 <mrb_> your config at CPU and GPU full load would draw (95+188*2)/12 = 39.25 Amp
513 2011-01-16 06:24:55 <RazielZ> crucial is usually decent I think
514 2011-01-16 06:24:59 <RazielZ> we don't really have it around here
515 2011-01-16 06:25:18 <RazielZ> but the whole PC is cheap so sure, why not
516 2011-01-16 06:25:31 <RazielZ> I'm not really familiar with the lower end of components >.>
517 2011-01-16 06:25:52 <hacim> i could drop that CPU down to 65W
518 2011-01-16 06:26:15 <mrb_> good idea
519 2011-01-16 06:26:58 <RazielZ> I hate your prices so much. Things are way more expensive here
520 2011-01-16 06:34:22 <kiba> price hating, RazielZ?
521 2011-01-16 06:52:02 <RazielZ> Yeah.
522 2011-01-16 08:09:44 <lfm> ;;bc,calcd 150000 373.70862152
523 2011-01-16 08:09:45 <gribble> The average time to generate a block at 150000 Khps, given the supplied difficulty of 373.70862152, is 2 hours, 58 minutes, and 20 seconds
524 2011-01-16 08:11:50 <nanotube> lfm: why 373?
525 2011-01-16 08:12:18 <lfm> testnet
526 2011-01-16 08:12:43 <nanotube> ah
527 2011-01-16 08:12:51 <nanotube> testnet diff is based on 1==0.4
528 2011-01-16 08:12:54 <nanotube> 0.5 that is
529 2011-01-16 08:13:00 <nanotube> so mainnet equivalent difficulty is 186
530 2011-01-16 08:13:08 <nanotube> point something
531 2011-01-16 08:14:47 <lfm> well i dont understand that
532 2011-01-16 08:15:57 <lfm> does that mean it averages to 5 min per block then?
533 2011-01-16 08:15:58 <nanotube> so... in mainnet, when difficulty is 1.0, the hextarget is 0x00000000FFFF0000000000000000000000000000000000000000000000000000
534 2011-01-16 08:16:20 <nanotube> on testnet, difficulty of 1.0, hextarget is twice that. (lshift by one more digit)
535 2011-01-16 08:17:18 <nanotube> it still 'aims' at 10min per block, just that 'minimum diff' is essentially 0.5, rather than 1
536 2011-01-16 08:17:55 <nanotube> so as far as actual 'probability to find hash', tn diff of 373 (meaning, 373 times more difficult than at 1.0), is actually only 186 times more difficult than mainnet 1.0
537 2011-01-16 08:18:26 <nanotube> because mainnet 1.0 is twice as hard as testnet 1.0
538 2011-01-16 08:18:39 <nanotube> ... hope that makes sense. :)
539 2011-01-16 08:19:42 <nanotube> but anyway, the final takeaway, is that if you want to know the actual avg time between blocks... use "bc,calcd 150000 186.x" :)
540 2011-01-16 08:21:16 <lfm> seems like it would make it less useful as a test [platform to have different rules
541 2011-01-16 08:23:41 <nanotube> not sure why min diff was chosen to be different.. probably just to make it easier to generate blocks for testing. (which, now that artforz has driven the difficulty up, is now a moot point :)
542 2011-01-16 08:23:49 <nanotube> )
543 2011-01-16 08:24:14 <lfm> well it is still easiuer
544 2011-01-16 08:24:36 <nanotube> yes
545 2011-01-16 08:24:45 <nanotube> twice as easy, in fact. ;)
546 2011-01-16 08:25:26 <lfm> no I mean even 373 is easier than 18000
547 2011-01-16 08:26:09 <nanotube> ah yes indeed. and in terms of 'real difficulty', testnet is about 100 times easier than mainnet currently.
548 2011-01-16 08:26:22 <nanotube> which is still plenty difficult, for a cpu. :)
549 2011-01-16 08:28:38 <joe_1> theymos is your bitcoin address checker at http://www.bitcoin.org/smf/index.php?topic=1844.0 valid and working?
550 2011-01-16 08:29:48 <lfm> I think testnet could just be thrown out and reset and restarted
551 2011-01-16 08:39:00 <mrb_> who accidentally left his GPUs running on testnet again?
552 2011-01-16 08:44:43 <joe_1> does a bitcoin address contain 160 bits?
553 2011-01-16 08:46:30 <lfm> mrb_ why you think it was accident?
554 2011-01-16 08:52:44 <joe_1> the i2p community wants a couple minor changes to the network protocol code to get it to talk to i2p
555 2011-01-16 08:53:34 <ducki2p> thats a bit of a harsh statement :)
556 2011-01-16 08:54:23 <ducki2p> at this point I am exploring what would need to be done to get BC to work on I2P
557 2011-01-16 08:55:01 <lfm> joe_1 so would it still be able to talk to old bitcoin versions?
558 2011-01-16 08:55:41 <joe_1> no --- good point. the i2p version would only talk to other i2p users
559 2011-01-16 08:55:55 <joe_1> so there would need to be an option to talk to both i2p and non-i2p users, so that the network will have these "bridges" to the real network
560 2011-01-16 08:56:07 <xelister> what kind of latences can bitcoin sustain to work properly?
561 2011-01-16 08:56:25 <xelister> is 30 minute lag in posting the transaction order ok?
562 2011-01-16 08:56:28 <lfm> joe_1 so you would also need Id think some sort of gaeway then to bridge the two types
563 2011-01-16 08:56:57 <xelister> are transaction orders tied to anything, like time or last block, or are they just untied messages (untill incorportated in some block)?
564 2011-01-16 08:57:29 <lfm> xelister the latter I think. no timestamp in txns
565 2011-01-16 08:58:11 <lfm> xelister so long as there are no delays propogating "solved" blocks
566 2011-01-16 08:58:43 <joe_1> 30 minute lag is probably unacceptable for miners, but users willing to wait 2 hours for 1 confirmation on sent and received transactions will be OK
567 2011-01-16 08:59:30 <lfm> well users would still not like extra delays but the system would still work I think
568 2011-01-16 08:59:56 <joe_1> of course i2p does not have 30 minute lag, it has something like 45 second lag
569 2011-01-16 09:00:29 <lfm> even 45 second lag would be bad for blcok propogation
570 2011-01-16 09:00:43 <lfm> block
571 2011-01-16 09:00:48 <xelister> actually i2p can be the fastest afair
572 2011-01-16 09:00:56 <xelister> you can configure chain length
573 2011-01-16 09:02:18 <xelister> tunnel
574 2011-01-16 09:02:23 <lfm> ya I think youd want to keep chains quite short for bitcoin
575 2011-01-16 09:02:28 <ducki2p> joe_1: most of my connections have a RTT of 1-2 seconds
576 2011-01-16 09:02:31 <ducki2p> (that is with 2 hops)
577 2011-01-16 09:03:07 <lfm> 2 seconds should be ok for bitcoin
578 2011-01-16 09:03:19 <joe_1> it would be bad for block propagation but another question is, would the addition of high latency peers decrease the efficiency of existing mining on the low latency network that already exists?
579 2011-01-16 09:04:14 <xelister> higher latency for miners just mean that more of the generated blocks would turn out to be chain branches that are rejected later
580 2011-01-16 09:04:19 <lfm> if a lot of the node were on high l;atency you would have more often races and blocks rejected
581 2011-01-16 09:05:23 <xelister> ArtForz: would you please put say 4, then 2, then 1 5970 on testnet to "fix it" (get diff back to 20 usable for testing ;) ?
582 2011-01-16 09:05:25 <joe_1> also, the high-latency net could build chains fast and reject chains that the low latency net had been working on, if the bridge went out for a few minutes then came back online to deliver a longer chain.
583 2011-01-16 09:05:37 <lfm> miners would probably prefer to be on direct connections to minimize chances of their blocks getting rejkected
584 2011-01-16 09:07:47 <lfm> new mined blocks are already quite annonymous on the existing bitcoin net so I doubt there would be much call for extra measues
585 2011-01-16 09:08:30 <lfm> measures
586 2011-01-16 09:10:12 <xelister> lfm: bitcoin is quite tracable
587 2011-01-16 09:10:46 <lfm> ok tell me who won the last three blocks?
588 2011-01-16 09:11:02 <nanotube> xelister: it only takes 22 min to gen a block on testnet with a 5970, at current difficulty. not a big deal for anyone with a decent gpu.
589 2011-01-16 09:11:14 <xelister> lfm: give me 10.000 BTC and I will
590 2011-01-16 09:11:28 <xelister> .. and I will e ready to respond to such question for next blocks
591 2011-01-16 09:11:32 <xelister> *will be
592 2011-01-16 09:11:54 <xelister> remember that 10.000 usd is not even 1% of LANUNCH MONEY for irs etc
593 2011-01-16 09:12:10 <xelister> or [other] criminals that would like to trace your activity
594 2011-01-16 09:12:16 <lfm> the info is not worth 10 btc to me
595 2011-01-16 09:12:22 <xelister> yeah for me neither
596 2011-01-16 09:12:27 <xelister> but it may be for someone in the future
597 2011-01-16 09:12:44 <xelister> eg what if IRS says - every bitcoiner, pay income tax for btcs earned/mined
598 2011-01-16 09:12:49 <xelister> or like "bitcoin is now illegal"
599 2011-01-16 09:13:23 <lfm> so how you gonna tell if a btc is mined in usa or not?
600 2011-01-16 09:13:45 <xelister> for sums like 10k USD bitcoin can be traced
601 2011-01-16 09:13:56 <xelister> its not something impossible
602 2011-01-16 09:14:09 <xelister> even average small time hacker/badguy could do it
603 2011-01-16 09:15:16 <lfm> easy to say
604 2011-01-16 09:15:41 <xelister> you say this is hard to do for 10 kbtc?
605 2011-01-16 09:20:45 <lfm> i yhink youd be lucky to figure out where the next block came from and if you were not prepared you could not figure out where the last block came from no matter how skilled you were
606 2011-01-16 09:37:17 <niekie> Hm. Amazon offers GPU instances. This is new to me.
607 2011-01-16 09:38:00 <niekie> I guess paying $2.10/hour for them is not currently worth it though.
608 2011-01-16 09:39:21 <niekie> Specs are 'Cluster GPU Quadruple Extra Large 22 GB memory, 33.5 EC2 Compute Units, 2 x NVIDIA Tesla
609 2011-01-16 09:43:33 <mrb_> lfm: accidental, because I fail to see why anybody would want to put so much resources mining on testnet
610 2011-01-16 09:44:11 <mrb_> niekie: this has been discussed in the past. if you do the math, ec2 instances are not worth it.
611 2011-01-16 09:44:36 <mrb_> both gpu and cpu instances.
612 2011-01-16 09:45:09 <lfm> mrb_ maybe it was a test to see how badly they could break testnet
613 2011-01-16 09:46:20 <mrb_> I must admit I once mined 20 blocks in < 5 minutes on testnet
614 2011-01-16 09:46:36 <nevezen> is testnet a simulation?
615 2011-01-16 09:47:01 <niekie> Yes and no.
616 2011-01-16 09:47:22 <niekie> It's a real network, and real clients are looking for blocks on it. But it's used for testing purposes.
617 2011-01-16 09:47:25 <lfm> ya, I think it only took 20 min for art to mine 2016 blocks
618 2011-01-16 09:47:36 <mrb_> it's another instance of the peer-to-peer bitcoin network that operate on another port
619 2011-01-16 09:48:45 <mrb_> by convention people declare it as the "test network" but it's really no different than the real network other than people does not recognize coins on testnet as having any value
620 2011-01-16 09:51:33 <lfm> and since testnet is not valued it could theoreticlly be reset back to zero
621 2011-01-16 09:52:35 <mrb_> yeah, a software update would have to be published so that all bitcoin nodes cooperate and agree to reset it at a determined time.
622 2011-01-16 09:53:31 <mrb_> but really, just not doing anything would cause it to reset itself anyway. since it's gonna take a while to mine the next 2016 blocks
623 2011-01-16 09:54:44 <lfm> yes once the 2016 blocks are found at the hi difficulty then the difficulty will drop back to a more reasonable level but it could take a long time
624 2011-01-16 09:55:24 <mrb_> if you try to mine the blocks somewhate quickly, then the difficulty will not decrease much
625 2011-01-16 09:55:49 <mrb_> the longer it takes to mine the block, the bigger the difficulty drop will be
626 2011-01-16 09:56:00 <lfm> yup
627 2011-01-16 09:58:48 <slush> current testnet difficulty is pretty high. Is there any plan to reset it?
628 2011-01-16 09:59:20 <lfm> no real plan but there are a few suggestions to reset it
629 2011-01-16 09:59:25 <slush> I mean, testnet is mainly for testing, but currently it takes too long to test finding a block, even on decent gpu
630 2011-01-16 09:59:33 <slush> I definitely vote for it
631 2011-01-16 10:00:07 <lfm> the question is, is it a design flaw revealed that it needs to be reset?
632 2011-01-16 10:00:11 <xelister> we could ask ArtForz to put ther 4, then 2, then 1 gpu, and it would reset it reasonably quick
633 2011-01-16 10:00:12 <slush> For example, I need tens or even hundreds of solved blocks for full pool testing. With current difficulty, it will take ages on 5970
634 2011-01-16 10:00:15 <xelister> lfm: yes
635 2011-01-16 10:00:33 <xelister> lfm: the difficulty should be reevaluated not only each 2016 blocks, but also each 48 hours
636 2011-01-16 10:01:23 <slush> I can kindly ask ArtForz to not test cartel mode too often ;), but I doubt it will have an effect
637 2011-01-16 10:01:38 <slush> + some people began to speculate on testnet value
638 2011-01-16 10:01:40 <xelister> well we wrecked testnet :|
639 2011-01-16 10:01:45 <slush> which put more miners here
640 2011-01-16 10:05:50 <lfm> so if its a design flaw will we have the same problem on the main net if a big miner drops out and the block rate slows to a crawl?
641 2011-01-16 10:08:27 <johnyh> btw, anyone have testnet bitcoins lying around? over 10,000 best
642 2011-01-16 10:08:51 <lfm> I think art has some
643 2011-01-16 10:09:58 <joe_1> im getting scared
644 2011-01-16 10:10:22 <lfm> why joe?
645 2011-01-16 10:10:29 <joe_1> what if the test net takes over
646 2011-01-16 10:10:39 <joe_1> all my wealth is in main net coins
647 2011-01-16 10:11:28 <lfm> I think people are silly to value testnet. it coulkd be reset tommorow and all the old balances would be dead
648 2011-01-16 10:11:53 <joe_1> only if people update...
649 2011-01-16 10:13:26 <slush> maybe automatic chain reset after...20 000 blocks... can be good idea for testnet?
650 2011-01-16 10:14:21 <joe_1> actually i dont think it could ever take over,
651 2011-01-16 10:14:35 <joe_1> bitcoin has value by virtue that it is the first
652 2011-01-16 10:15:04 <lfm> i think its best to be manual reset. if testnet gets too many special cases it will lose value as a testing ground
653 2011-01-16 10:15:37 <slush> who can decide testnet reset?
654 2011-01-16 10:17:42 <joe_1> theoretically, nobody. practically, the bitcoin client developers
655 2011-01-16 10:17:42 <lfm> I think satoshi or someone just has to make new block zero and then get people to stop uing the old testnet chain. might need new version, not sure really
656 2011-01-16 10:19:02 <slush> I hope they don't release version with block reset with forgotten condition to testnet ;)
657 2011-01-16 10:20:19 <joe_1> well it's resilient enough that that wouldn't be too big of a problem. it would only take one person who didn't upgrade, to re-populate the network once they realized the mistake and released the fixed version
658 2011-01-16 10:20:19 <lfm> "forgotten condition"?
659 2011-01-16 10:21:07 <lfm> joe if its a different block zero then the old chain wouldnt work any more
660 2011-01-16 10:21:30 <joe_1> i'm saying after they fixed it back.
661 2011-01-16 10:22:15 <lfm> huh?
662 2011-01-16 10:22:30 <joe_1> slush was saying i hope they don't accidentally reset the main chain.
663 2011-01-16 10:22:33 <slush> lfm: "forget to make condition which perform block reset only on testnet"
664 2011-01-16 10:23:17 <lfm> oh, ya I guess we dont really wanna reset the main net now
665 2011-01-16 10:23:32 <lfm> even if it is still beta
666 2011-01-16 10:24:13 <joe_1> is it a coincidence that satoshi used the biblical term "genesis" to refer to the first block?
667 2011-01-16 10:25:38 <lfm> um, no, it is the beginning, thats what genisis means
668 2011-01-16 10:25:43 <joe_1> meaning that God himself handed down the hash to the first block, so the main bitcoin chain is the "in god we trust" bitcoin currency.
669 2011-01-16 10:26:15 <lfm> not neccesarily anything to do with god
670 2011-01-16 10:26:37 <joe_1> yeah, but i think that's what we should tell other people, so nobody gets any wise ideas about starting new chains.
671 2011-01-16 10:27:10 <lfm> genisis is from the same root as generate
672 2011-01-16 10:42:00 <noagendamarket> joe satoshi himself says people should start new block chains to setrup competing currencies
673 2011-01-16 10:42:53 <joe_1> i dont think that's right
674 2011-01-16 10:43:24 <joe_1> first, it splits up valuable CPU power needed to fend off attacks from governments
675 2011-01-16 10:43:58 <joe_1> second, if people see that currencies are coming and going then none of them will ever have value, since value is based on expectation of future value.
676 2011-01-16 10:47:25 <joe_1> when an ordinary person looks at a candidate currency and asks themselves if it's valueable, they're asking what makes it forever and what makes it self evident, to the rest of the world, that it's a currency?
677 2011-01-16 10:48:17 <joe_1> for bitcoin, it's the first application of the proof-of-work chain cryptocurrency.
678 2011-01-16 10:50:22 <joe_1> if i wake up tomorrow and start my own chain, someone asks me why is it a currency? "it's a proof-of-work chain based cryptocurrency that I created yesterday"
679 2011-01-16 10:52:58 <joe_1> i could theoretically imagine 2 bitcoin chains, but nothing more than 2. The reason i say that, is that we already accept gold and silver-- 2 metals-- as currency, when we should really only have 1 metal currency.
680 2011-01-16 10:54:57 <joe_1> maybe i should start a thread asking satoshi about this. or look for any threads already on there about this subject
681 2011-01-16 11:07:05 <lfm> if you stand behind your currency it will have value. if you make standing offer to trade it for another currency of anything of value, gold, cookies, your labor, if you back it up then it will have value
682 2011-01-16 11:17:42 <T_X> slush: thanks for the awesome extra statstics :). that somehow gives a good feeling for having selected a %-donation value :)
683 2011-01-16 11:18:07 <slush> thanks :)
684 2011-01-16 11:19:42 <slush> that's my first graphs in javascript, it was real pain
685 2011-01-16 11:21:59 <T_X> :D. but looks nice :)
686 2011-01-16 11:36:42 <T_X> on mtgox, only trades on the market cost 0.65%, right? I mean, sending btc to the mtgox account and withdrawing them again does not have a transaction fee, right?
687 2011-01-16 11:36:54 <slush> correct
688 2011-01-16 11:41:23 <T_X> kk. actually what I would find useful is the possiblity to share an mtgox wallet over multiple accounts for one thing and having the possibility for reviewing transactions from and to that wallet publically.
689 2011-01-16 11:43:09 <slush> is it related to trading somehow? This feature would be more useful on mybitcoin, which is more 'bank' thank exchange
690 2011-01-16 11:43:24 <T_X> hmm, okay
691 2011-01-16 11:43:41 <sipa> T_X: sending them euros does incur a 2% fee as well
692 2011-01-16 11:44:04 <T_X> no, there's no plan to send EUR
693 2011-01-16 11:44:20 <T_X> just to have a deposit which is not controlled by a single person
694 2011-01-16 11:44:38 <sipa> ?
695 2011-01-16 11:44:51 <slush> both are talking about different things :)
696 2011-01-16 11:45:09 <T_X> looks like :D
697 2011-01-16 11:45:13 <slush> sipa talks about mtgox fees
698 2011-01-16 11:45:21 <slush> not about shared wallets
699 2011-01-16 11:45:32 <sipa> oh, i hadn't read everything afterwards yet :)
700 2011-01-16 11:46:48 <T_X> and maybe a certain amount of elected people within this community were having the right to make transactions out of this account. however, it shall be visible for everyone else in the community who did any of those transactions
701 2011-01-16 11:59:37 <T_X> mybitcoin does not have such nice donation buttons as mtgox has, has it?
702 2011-01-16 12:09:36 <Diablo-D3> http://www.youtube.com/watch?v=cTQNwMxqM3E
703 2011-01-16 12:09:38 <Diablo-D3> what
704 2011-01-16 12:45:55 <npouillard> I'm looking for info about the use of testnet, any pointers?
705 2011-01-16 12:47:58 <sipa> what do you need to know about it?
706 2011-01-16 12:50:56 <sipa> how often does https://smsz.net/btcStats/bitcoin.kml get updated?
707 2011-01-16 12:53:36 <npouillard> http://www.bitcoin.org/smf/index.php?PHPSESSID=fe6b3d0c1739bb2914976ffe1209fbb9&topic=363.0;all <== this thread looks quite outdated no?
708 2011-01-16 14:13:55 <sipa> ;;bc,calc 600000
709 2011-01-16 14:13:56 <gribble> The average time to generate a block at 600000 Khps, given current difficulty of 18437.64439217 , is 1 day, 12 hours, 39 minutes, and 41 seconds
710 2011-01-16 14:13:59 <sipa> ;;bc,calc 1200000
711 2011-01-16 14:14:00 <gribble> The average time to generate a block at 1200000 Khps, given current difficulty of 18437.64439217 , is 18 hours, 19 minutes, and 50 seconds
712 2011-01-16 14:25:21 <theymos> joe_1: My address checker is working.
713 2011-01-16 14:29:45 <niekie> ;;bc
714 2011-01-16 14:29:46 <gribble> Error: "bc" is not a valid command.
715 2011-01-16 14:29:48 <niekie> ;;help
716 2011-01-16 14:29:49 <gribble> The bot responds when you start a line with the ! character. A good starting point for exploring the bot is the !facts command. You can also visit the bot's website for a list of help topics and documentation: http://gribble.sourceforge.net/
717 2011-01-16 14:31:03 <niekie> ;;bc,blocks
718 2011-01-16 14:31:04 <gribble> 102998
719 2011-01-16 14:31:13 <niekie> ;;bc,stats
720 2011-01-16 14:31:15 <gribble> Current Blocks: 102998 | Current Difficulty: 18437.64439217 | Next Difficulty At Block: 104831 | Next Difficulty In: 1833 blocks | Next Difficulty In About: 1 week, 3 days, 10 hours, 30 minutes, and 36 seconds | Next Difficulty Estimate: 22596.89665695
721 2011-01-16 14:31:18 <tcatm> hey theymos. did the coin arrive?
722 2011-01-16 14:31:35 <theymos> Yes. It looks pretty cool! Thanks.
723 2011-01-16 14:31:51 <tcatm> cool :)
724 2011-01-16 14:35:59 <tcatm> theymos: can blockexplorer be modified to include TX that aren't in blocks yet? I have a python script that'll connect to a bitcoin node and output them as soon as they arrive
725 2011-01-16 14:39:54 <theymos> Yes. Can you post the script? I might not implement it for a while, though -- there are some other features I want to do first.
726 2011-01-16 14:41:51 <tcatm> Might as well use it to have realtime update for blocks ;)
727 2011-01-16 14:43:24 <tcatm> save this as bitcoin.py http://pastebin.com/JZM6Hemt
728 2011-01-16 14:44:37 <tcatm> this will dump blocks and TX as soon as it sees them http://pastebin.com/LsSebetx
729 2011-01-16 14:47:10 <theymos> Thanks. Probably I'll just use block messages from this to trigger block updates from getblock, since I already have getblock parsing coded.
730 2011-01-16 15:22:24 <nanotube> <lfm> [02:51:34] and since testnet is not valued it could theoreticlly be reset back to zero  <-- i beg to differ about the 'not valued' bit. there are several outstanding bid offers for testnet btc on OTC, so clearly testnet does have value.
731 2011-01-16 15:22:38 <nanotube> resetting it willy-nilly is not a good idea, i think.
732 2011-01-16 15:23:15 <sipa> it has value if someone wants to pay for it
733 2011-01-16 15:23:27 <sipa> the question is whether it is useful that it has value
734 2011-01-16 15:24:00 <ArtForz> oh, resetting testnet is easy
735 2011-01-16 15:24:33 <ArtForz> <12h with something like my cluster
736 2011-01-16 15:24:49 <tcatm> how would you reset it? fake nTime?
737 2011-01-16 15:24:54 <ArtForz> yep
738 2011-01-16 15:25:09 <sipa> any sentence which contains "my cluster", said by ArtForz, is by definition not "easy"
739 2011-01-16 15:25:38 <ArtForz> basically restart from genesis, play with block times to first drive difficulty way up, then let is settle back down
740 2011-01-16 15:25:55 <ArtForz> new chain has higher total work than current chain = takes over
741 2011-01-16 15:26:21 <tcatm> sounds like something worth to do
742 2011-01-16 15:26:40 <ArtForz> and total testnet work only represents about 12h @ 16Gh/s
743 2011-01-16 15:26:59 <ArtForz> unless my total work calculation is off for testnet
744 2011-01-16 15:27:25 <sipa> essentially allowing everyone to create their own testnet
745 2011-01-16 15:27:54 <sipa> -test
746 2011-01-16 15:28:30 <ArtForz> lets see if the math works out
747 2011-01-16 15:29:30 <tcatm> would even be interesting to see how clients react to all coinbases becoming invalid
748 2011-01-16 15:29:51 <sipa> ?
749 2011-01-16 15:30:20 <tcatm> sipa: to create your own testnet only connect to your nodes
750 2011-01-16 15:30:28 <ArtForz> we have about 16M seconds to work with
751 2011-01-16 15:32:02 <ArtForz> we can stuff about 5 blocks in a second thanks to the median-of-prev-11-plus-1 rule
752 2011-01-16 15:33:07 <ArtForz> = you can easily put 2016 blocks in 600 seconds to drive difficulty up without wasting too much time
753 2011-01-16 15:34:33 <tcatm> => difficulty of 1008?
754 2011-01-16 15:34:40 <ArtForz> ?
755 2011-01-16 15:34:58 <ArtForz> testnet is also limited by /4 *4 per period
756 2011-01-16 15:35:12 <tcatm> right, forgot about that
757 2011-01-16 15:44:11 <INEEDMONEY> so how's the number-crunching going?
758 2011-01-16 15:44:40 <ArtForz> yep, should work
759 2011-01-16 15:47:05 <ArtForz> 4 * 2016 blocks in 2400s -> get difficulty up to 32, 2016 block in 1209600s, 3* 2016 block in 4838400s each -> final difficulty is 0.5
760 2011-01-16 15:47:50 <ArtForz> total work in that chain is 732738600566784, testnet chain is 674195433361632 or so
761 2011-01-16 15:48:16 <ArtForz> that chain needs 15727200s after genesis
762 2011-01-16 15:49:31 <ArtForz> and yeah, it's < 2/3 the # of blocks of the current testnet chain, yet has higher sum-of-work
763 2011-01-16 15:50:26 <ArtForz> 16128 blocks
764 2011-01-16 15:50:57 <nanotube> wouldn't it be cheaper/faster to just let the diff on the current testnet come down?
765 2011-01-16 15:51:18 <ArtForz> well... that will take a while
766 2011-01-16 15:51:38 <molec> there's something I don't understand about cooperative mining: what keeps a miner from keeping the block for himself in case he finds one instead of submitting it to the cooperative?
767 2011-01-16 15:51:57 <nanotube> ;;bc,wiki pooled mining
768 2011-01-16 15:51:58 <gribble> https://en.bitcoin.it/wiki/Pooled_mining | Dec 30, 2010 ... Pooled mining is an approach where multiple generating clients contribute to the generation of a block, and then split the block reward ...
769 2011-01-16 15:52:02 <nanotube> molec: --^
770 2011-01-16 15:52:12 <molec> nanotube, thx
771 2011-01-16 15:52:13 <ArtForz> well... easier... yes
772 2011-01-16 15:52:30 <ArtForz> we just need a bit more hashpower on testnet
773 2011-01-16 15:52:39 <nanotube> molec: in short, blocks that clients are working on contain the tx giving btc to the pool. 'keeping it' is useless.
774 2011-01-16 15:52:43 <ArtForz> so we get ~2400s/block for maximum slowing
775 2011-01-16 15:53:09 <tcatm> how long would it take?
776 2011-01-16 15:53:29 <tcatm> ;;bc,calc 300000 42
777 2011-01-16 15:53:30 <gribble> Error: invalid syntax (<string>, line 1)
778 2011-01-16 15:53:35 <tcatm> ;;bc,calcd 300000 42
779 2011-01-16 15:53:36 <gribble> The average time to generate a block at 300000 Khps, given the supplied difficulty of 42, is 10 minutes and 1 second
780 2011-01-16 15:53:37 <ArtForz> 8 weeks to get to the next quartering
781 2011-01-16 15:54:22 <tcatm> instead of 1-2 days when restarting the chain
782 2011-01-16 15:54:29 <ArtForz> yep
783 2011-01-16 15:54:39 <ArtForz> and then we're still at 42 diff ...
784 2011-01-16 15:54:53 <ArtForz> = another 8 weeks to get to 10.something
785 2011-01-16 15:55:22 <ArtForz> restarting the chain somehow seems *easier*
786 2011-01-16 15:55:26 <tcatm> so only nTime needs to be carefully adjusted (i.e. ++ every 5s)?
787 2011-01-16 15:55:36 <ArtForz> well... kinda
788 2011-01-16 15:55:39 <ArtForz> but yeah
789 2011-01-16 15:55:59 <ArtForz> basically ntime +=1 every 5 blocks for the first 4*2016 blocks
790 2011-01-16 15:56:11 <ArtForz> then ntime += 600 for 2016 blocks
791 2011-01-16 15:56:20 <ArtForz> and ntime += 2400 for 3*2016 blocks
792 2011-01-16 15:57:26 <tcatm> ,,bc,calcd 300000 0.5
793 2011-01-16 15:57:27 <gribble> (bc,calcd <an alias, 2 arguments>) -- Alias for "echo The average time to generate a block at $1 Khps, given the supplied difficulty of $2, is [time elapsed [math calc 1/((2**224-1)/$2*$1*1000/2**256)]]".
794 2011-01-16 15:57:30 <tcatm> ;;bc,calcd 300000 0.5
795 2011-01-16 15:57:31 <gribble> The average time to generate a block at 300000 Khps, given the supplied difficulty of 0.5, is 7 seconds
796 2011-01-16 15:57:46 <ArtForz> would certainly be fun to see what happens when everyones balance goes *poof*
797 2011-01-16 15:59:47 <nanotube> mm yea.. fun.... O_.
798 2011-01-16 16:00:07 <ArtForz> well, it's called TESTnet for a reason
799 2011-01-16 16:00:41 <nanotube> heh true
800 2011-01-16 16:01:18 <nanotube> so how long would it take for someone with a few Thps to do the same to the mainnet?
801 2011-01-16 16:02:05 <ArtForz> mainnet work is ... 657980749827397368
802 2011-01-16 16:02:28 <ArtForz> 182h @ 1Thps
803 2011-01-16 16:03:04 <ArtForz> + a few more h for the work mainnet is doing in those 182h
804 2011-01-16 16:03:17 <ArtForz> so... about 200h @ 1Thps
805 2011-01-16 16:03:30 <tcatm> what's the easiert way to calculate network work?
806 2011-01-16 16:04:09 <ArtForz> errr
807 2011-01-16 16:04:17 <nanotube> ArtForz: mmm not so encouraging. let's hope that it'll be a while before adversaries with Thps's to throw around takes note. :)
808 2011-01-16 16:04:48 <ArtForz> mainline uses sum(2**256 / (block[i].target + 1))
809 2011-01-16 16:04:52 <kiba> gene guy seem to have a strong hatred against multinational corporations
810 2011-01-16 16:05:18 <ArtForz> and my code does, too
811 2011-01-16 16:05:44 <ArtForz> it's about the same as sum(block[i].difficulty * 2**32)
812 2011-01-16 16:06:05 <ArtForz> or simpler: sum(block[i].difficulty) * 2**32
813 2011-01-16 16:06:51 <ArtForz> that should come out to roughly the same
814 2011-01-16 16:09:05 <tcatm> so sum... target+1 is the accurate work?
815 2011-01-16 16:09:13 <ArtForz> yes
816 2011-01-16 16:09:36 <tcatm> what was the formula to calculate target from nbits again?
817 2011-01-16 16:10:27 <ArtForz> off the top of my head, (nbits & 0xFFFFFF) << (8 * (((nbits >> 24) & 0xFF) - 3))
818 2011-01-16 16:11:02 <tcatm> oh wait, getblock outputs diff instead of nbits :/
819 2011-01-16 16:12:15 <ArtForz> submit a feature request for BBE? ;)
820 2011-01-16 16:12:36 <theymos> What feature?
821 2011-01-16 16:12:40 <tcatm> bitcoincharts keeps the blockchain in a sqltable
822 2011-01-16 16:12:46 <tcatm> mysql> select sum(difficulty)*pow(2,32) from bitcoin_block;
823 2011-01-16 16:12:51 <tcatm> 6.58050765980394e+17
824 2011-01-16 16:12:53 <ArtForz> total chain work for blocks
825 2011-01-16 16:13:24 <ArtForz> @ block 103010, it should be 658059940115424900
826 2011-01-16 16:14:05 <ArtForz> error comes from 1.0 diff != 2**224
827 2011-01-16 16:14:27 <tcatm> can that be compensated?
828 2011-01-16 16:14:32 <ArtForz> not easily
829 2011-01-16 16:14:58 <ArtForz> sec
830 2011-01-16 16:16:12 <tcatm> theymos: can you do pow(2,256)/sum((nbits & 0xFFFFFF) << (8 * (((nbits >> 24) & 0xFF) - 3)) + 1) in your block database?
831 2011-01-16 16:16:13 <ArtForz> try using 4295032833 instead of 2**32
832 2011-01-16 16:16:33 <tcatm> 6.58060807191439e+17
833 2011-01-16 16:16:46 <ArtForz> thats a lot closer
834 2011-01-16 16:16:48 <tcatm> yep
835 2011-01-16 16:16:49 <tcatm> thanks
836 2011-01-16 16:17:13 <ArtForz> it's 4295032833.0000153... really (0x100010001.000100010001...)
837 2011-01-16 16:17:53 <ArtForz> = 2**256 / ((0xFFFF << 208) + 1)
838 2011-01-16 16:18:38 <ArtForz> you're within 0.001% or something anyways
839 2011-01-16 16:18:42 <theymos> tcatm: Not easily.
840 2011-01-16 16:18:45 <ArtForz> good enough for a rough estimate ;)
841 2011-01-16 16:19:20 <hacim> is memory or CPU resources ever a concern WRT to running a GPU miner?
842 2011-01-16 16:19:32 <ArtForz> nope
843 2011-01-16 16:21:14 <ArtForz> tcatm: your calculation is 0.0001318% high ... using floating point... so who cares ;)
844 2011-01-16 16:21:35 <nanotube> so... i guess this basically means that if we want to "prevent blockchain competition" from any other competing chains (and thereby 'preserve the mainnet chain value'), we could just 'hunt down' any startup new chains by throwing a bunch of our gen power at them, and 'resetting' their net. heh...
845 2011-01-16 16:21:47 <ArtForz> you can't do it with mainnet
846 2011-01-16 16:21:53 <nanotube> yes, hardcoded checkpoints
847 2011-01-16 16:21:58 <ArtForz> yep
848 2011-01-16 16:22:01 <nanotube> but you can do it starting from latest checkpoint
849 2011-01-16 16:22:06 <ArtForz> yep
850 2011-01-16 16:22:09 <tcatm> ArtForz: and the unit is? hashes?
851 2011-01-16 16:22:14 <ArtForz> yes
852 2011-01-16 16:22:17 <nanotube> e.g., if you did this since block 70k-something... that'd still be pretty havoc-y
853 2011-01-16 16:22:30 <ArtForz> unit for work is bitcoin block hashes
854 2011-01-16 16:22:39 <ArtForz> = about 2*sha256
855 2011-01-16 16:22:59 <ArtForz> or roughly 1500 core cycles using SSE2
856 2011-01-16 16:23:01 <hacim> cool, so I can get just the bare minimum for my mining machine
857 2011-01-16 16:24:04 <ArtForz> = mainnet chain is equivalent to about 987090 peta sse2 ops
858 2011-01-16 16:24:14 <theymos> If a "reset" happened on the real network, a new version would be released that locked in the real main chain. You'll still have the blocks -- they're not deleted when they become orphans.
859 2011-01-16 16:24:25 <ArtForz> or about 987 exaops
860 2011-01-16 16:24:35 <ArtForz> yep
861 2011-01-16 16:25:37 <ArtForz> so in a while we'll have the equivalent of 1 zetta-sse-ops in the main chain
862 2011-01-16 16:26:22 <ArtForz> another 6 orders of magnitude and we run out of SI prefixes :P
863 2011-01-16 16:26:45 <UukGoblin> hmm, what prevents LRUSD<->USD exchange rate from drifting far off 1.0?
864 2011-01-16 16:27:45 <ArtForz> wild guess: LR offering premium exchangers close to 1.0 exchange rate for LRUSD<->USD
865 2011-01-16 16:27:48 <nanotube> UukGoblin: 'official lr exchangers'
866 2011-01-16 16:27:56 <nanotube> right
867 2011-01-16 16:28:29 <nanotube> they can buy lr at close to 1.0 (if not exactly 1.0) from lr.
868 2011-01-16 16:29:04 <nanotube> due to competition, their premiums to others are generally single-digit percents.
869 2011-01-16 16:29:14 <UukGoblin> ah
870 2011-01-16 16:29:34 <UukGoblin> so LR /does/ sell LRUSD but only to '(un)official exchangers'
871 2011-01-16 16:29:53 <nanotube> s/(un)//
872 2011-01-16 16:30:05 <tcatm> ()official?
873 2011-01-16 16:30:13 <nanotube> you have to go through an official registration process, to be able to buy lr from them.
874 2011-01-16 16:30:21 <UukGoblin> "In order to add or remove funds from your LR account you should use one of the independent exchange providers from the list below. Please note that businesses listed on this page are independent and in no way related to Liberty Reserve, and Liberty Reserve does not guarantee any outcome that you may experience with these companies."
875 2011-01-16 16:30:26 <nanotube> tcatm: haha, ok, s/(un)//
876 2011-01-16 16:30:50 <nanotube> UukGoblin: yes, they're not legally related, but they're 'official' in that they have to register, and lr has to approve
877 2011-01-16 16:31:10 <UukGoblin> heh
878 2011-01-16 16:31:12 <UukGoblin> thanks niekie
879 2011-01-16 16:31:19 <niekie> No problem :)
880 2011-01-16 16:31:25 <nanotube> woo more ops. :)
881 2011-01-16 16:31:34 <niekie> Exactly.
882 2011-01-16 16:32:00 <nanotube> or ... wait i thought i did. heh
883 2011-01-16 16:32:04 <niekie> Nope :)
884 2011-01-16 16:32:11 <nanotube> in that case, tx. :)
885 2011-01-16 16:38:00 <kiba> the bitcoin job market sucks
886 2011-01-16 16:38:52 <UukGoblin> kiba, no wonder. paying income tax off bitcoin must be a pain...
887 2011-01-16 16:39:54 <nanotube> UukGoblin: haha i doubt that that is the reason. :)
888 2011-01-16 16:40:07 <kiba> so we got the base economy
889 2011-01-16 16:40:10 <kiba> Hosting
890 2011-01-16 16:40:18 <kiba> that's like the first order capital goods
891 2011-01-16 16:40:32 <kiba> and then we got a few sites
892 2011-01-16 16:40:47 <UukGoblin> yup, I guess it's mostly due to lack of popularity :-)
893 2011-01-16 16:40:59 <kiba> well
894 2011-01-16 16:41:11 <kiba> we have a thriving "bounty" system
895 2011-01-16 16:41:25 <kiba> but nobody cares about getting bitcoin from doing work
896 2011-01-16 16:41:42 <kiba> I think everybody just buy bitcoin rather than work for it
897 2011-01-16 16:42:00 <tcatm> I got about 3000 BTC from programming/webdesign jobs
898 2011-01-16 16:42:20 <UukGoblin> "bounty" isn't good enough for me... I need a job that'll provide me a constant income for a fixed number of hours of work per day
899 2011-01-16 16:43:00 <theymos> Bitcoin Block Explorer is finally back to normal! HTTPS is enabled again, and updates happen at the normal 1-minute frequency. Page loading should also be faster. I changed some things, so let me know if there are any bugs.
900 2011-01-16 16:43:03 <kiba> 1,200 dollars for tcatm
901 2011-01-16 16:43:18 <kiba> everyone is a programmer so they don't want to hire other programmers
902 2011-01-16 16:43:20 <Cusipzzz> theymos: nice, I sent a contrib yesterday.
903 2011-01-16 16:43:31 <theymos> Cusipzzz: Thanks!
904 2011-01-16 16:44:11 <Cusipzzz> been doing a lot of testing and BBE has been invaluable
905 2011-01-16 16:45:32 <tcatm> ArtForz: What rpm would you recommened for milling solid copper (instead of FR4)?
906 2011-01-16 16:46:29 <slush> whats mode +ooo ?
907 2011-01-16 16:46:40 <hacim> slush: three individual +o
908 2011-01-16 16:46:41 <niekie> Mass-opping.
909 2011-01-16 16:46:49 <hacim> slush: ie. ops for those three people
910 2011-01-16 16:46:50 <slush> wow, thanks :)
911 2011-01-16 16:47:46 <slush> is any help for irc admin commands lying around?
912 2011-01-16 16:47:48 <UukGoblin> hmm, rpm... as long as it spins... ;-)
913 2011-01-16 16:48:17 <niekie> slush: basic commands are /kick which takes 2 arguments, username and reason.
914 2011-01-16 16:48:50 <omglolbbq> i just started bitcooin and i get "WARNING: Displayed transactions may not be correct! You may need to upgrade, or other nodes may need to upgrade"... whats going on?
915 2011-01-16 16:49:00 <UukGoblin> slush, it's unlikely you'll have to use any commands... freenode is such a friendly place
916 2011-01-16 16:49:01 <slush> wow
917 2011-01-16 16:49:03 <niekie> slush: yeah, it works like that.
918 2011-01-16 16:49:12 <niekie> Now you lost op though ;)
919 2011-01-16 16:49:23 <slush> :-D
920 2011-01-16 16:49:28 <theymos> omglolbbq: How many blocks do you have?
921 2011-01-16 16:50:05 <omglolbbq> 102882
922 2011-01-16 16:50:48 <ArtForz> ;;bc,blocks
923 2011-01-16 16:50:56 <gribble> 103013
924 2011-01-16 16:51:11 <omglolbbq> i have 8 connections but it's not downloading new blocks
925 2011-01-16 16:51:37 <ArtForz> weird
926 2011-01-16 16:51:47 <ArtForz> what client version?
927 2011-01-16 16:51:51 <omglolbbq> .19
928 2011-01-16 16:52:09 <theymos> How strange. That's the error for "longer invalid chain".
929 2011-01-16 16:52:19 <omglolbbq> i had 102881 blocks when i started the client, it downloaded 1 block, then gave that message and stopped downloading
930 2011-01-16 16:52:35 <theymos> Try deleting your block database files. Maybe they're corrupted.
931 2011-01-16 16:52:55 <omglolbbq> ok that will take some time :)
932 2011-01-16 16:53:32 <ArtForz> yep, prob corrupted block chain
933 2011-01-16 16:53:48 <niekie> Make a backup though ;)
934 2011-01-16 16:54:12 <nevezen> where does it download newer blocks from?
935 2011-01-16 16:54:24 <sipa> the network