1 2012-10-19 02:10:40 <helo> ok, time for a quantal ubuntu ppa ;)
  2 2012-10-19 02:10:48 <helo> readysetgo
  3 2012-10-19 02:12:25 <helo> BlueMatt: if you need someone to test or anything, let me know
  4 2012-10-19 03:12:55 <bladezor> Hello, when trying to execute bitcoind from the command line in Ubuntu
  5 2012-10-19 03:13:01 <bladezor> when I execute bitcoind
  6 2012-10-19 03:13:05 <bladezor> It hangs
  7 2012-10-19 03:13:14 <bladezor> There's no I/O
  8 2012-10-19 03:15:08 <weex> do you mean no output from the command?
  9 2012-10-19 03:15:19 <weex> that would be normal
 10 2012-10-19 03:15:23 <weex> usually you run bitcoind &
 11 2012-10-19 03:15:25 <bladezor> Nm
 12 2012-10-19 03:15:34 <bladezor> It seems I wasn't running the daemon
 13 2012-10-19 03:15:39 <bladezor> And there's an exception
 14 2012-10-19 03:16:57 <lianj> -printconsole
 15 2012-10-19 03:17:14 <bladezor> EXCEPTION: 22DbRunRecoveryException
 16 2012-10-19 03:17:14 <lianj> eh, -printtoconsole
 17 2012-10-19 03:17:15 <bladezor> DbEnv::open: DB_RUNRECOVERY: Fatal error, run database recovery
 18 2012-10-19 03:19:47 <bladezor> ah
 19 2012-10-19 03:19:51 <bladezor> seems like the dbs were locked
 20 2012-10-19 03:31:32 <freewil> is it possible to link two multisig addresses together if they both share 1 (of 2) common keys
 21 2012-10-19 03:38:47 <BlueMatt> helo: wanna test in about an hour?
 22 2012-10-19 09:02:02 <UukGoblin> so there's 4 new bytes in the coinbase before "\\x{fa}\\x{be}mm" now? (compared to half a year ago)
 23 2012-10-19 09:02:34 <sipa> in version=2 blocks, yes
 24 2012-10-19 09:02:37 <sipa> see BIP 34
 25 2012-10-19 09:03:31 <UukGoblin> ah right
 26 2012-10-19 09:03:37 <UukGoblin> 4 bytes for 300 years
 27 2012-10-19 09:03:41 <UukGoblin> that should suffice
 28 2012-10-19 09:04:05 <UukGoblin> are all blocks version=2 now?
 29 2012-10-19 09:04:20 <sipa> i don't think so
 30 2012-10-19 09:04:23 <UukGoblin> meh, I guess ones mined by p2pool will be
 31 2012-10-19 09:14:28 <doublec> UukGoblin: this lists version 1/2 block split by some of the pools http://blockorigin.pfoe.be/top.php
 32 2012-10-19 09:16:12 <UukGoblin> doublec, that's virtually none! but p2pool is in good shape, as is eligius
 33 2012-10-19 09:17:37 <doublec> yes, elipse as well
 34 2012-10-19 09:22:09 <sipa> BIP34 will only be enforced when nVersion=2 occurs in 75% of the last 1000 blocks
 35 2012-10-19 09:22:14 <sipa> will probably still take a while
 36 2012-10-19 09:23:12 <UukGoblin> nodnod
 37 2012-10-19 09:23:33 <UukGoblin> just need to update chronobit ;-)
 38 2012-10-19 09:39:46 <bladezor> if I use testnet does it download the blockchain to seperate .dat files or does it overwrite the ones for main?
 39 2012-10-19 09:39:55 <bladezor> with bitcoind
 40 2012-10-19 09:44:27 <sipa> yes, the datafiles are in a subdirectory (~/.bitcoin/testnet3)
 41 2012-10-19 09:48:32 <bladezor> okay awesome, thanks
 42 2012-10-19 09:50:37 <UukGoblin> oh, p2pool also puts a comma between the block height and the MM coinbase
 43 2012-10-19 09:51:08 <UukGoblin> did namecoin not need any updates after BIP 0034?
 44 2012-10-19 10:18:39 <JyZyXEL> the generation inputs scriptsig is the so called coinbase?
 45 2012-10-19 10:18:48 <sipa> yes
 46 2012-10-19 10:19:08 <JyZyXEL> blockexplorer.com doesn't seem to have a decoder for it
 47 2012-10-19 10:19:28 <JyZyXEL> it just shows it as a long hex value
 48 2012-10-19 10:21:56 <sipa> there's also not really anything to decode
 49 2012-10-19 10:22:04 <sipa> except the height-in-coinbase as of BIP34
 50 2012-10-19 10:22:12 <sipa> the rest is freeform binary data
 51 2012-10-19 10:36:00 <_dr> have there been any reviews on adi shamirs paper yet?
 52 2012-10-19 10:37:10 <_dr> weren't it for his name, i'd think he doesn't have the slightest clue about the way bitcoin works
 53 2012-10-19 10:37:30 <sipa> yes, there is a thread about it on the forum
 54 2012-10-19 10:37:49 <sipa> some people mailed the authors, and even got feedback from them
 55 2012-10-19 10:39:15 <_dr> clearly iacr has a very poor reviewing process
 56 2012-10-19 10:57:14 <gmaxwell> _dr: My heart sings at how consistently and throughly negative everyone remotely clueful's review has been.
 57 2012-10-19 10:58:00 <gmaxwell> (there is so much dispute in the bitcoin community it's refreshing to see ~everyone agree on /something/)
 58 2012-10-19 10:58:40 <sipa> haha
 59 2012-10-19 11:06:11 <Joric> 78 percent of Bitcoin currency stashed under digital mattress!
 60 2012-10-19 11:06:16 <Joric> *facepalm*
 61 2012-10-19 11:07:05 <gmaxwell> Joric: god knows, it may be true; but the paper doesn't really show that. Oh well, we shouldn't complain; that claim is less bad for us that "1001% bitcoins used for drugs!"
 62 2012-10-19 11:09:19 <Joric> oddly theres no silk road analysis at all =)
 63 2012-10-19 11:11:20 <Joric> though they somehow figured out mtgox, instawallet and deepbit addresses
 64 2012-10-19 11:11:28 <Joric> http://eprint.iacr.org/2012/584.pdf
 65 2012-10-19 11:13:27 <gmaxwell> Joric: I know  mtgox, instawallet and deepbit addresses too.
 66 2012-10-19 11:14:03 <gmaxwell> Joric: did you see the instawallet response?
 67 2012-10-19 11:14:40 <Joric> nope what they say
 68 2012-10-19 11:15:46 <gmaxwell> "Hahaha"
 69 2012-10-19 11:16:39 <gmaxwell> Joric: https://bitcointalk.org/index.php?topic=118797.msg1280496#msg1280496
 70 2012-10-19 11:18:03 <Joric> myinstawallet
 71 2012-10-19 11:20:58 <gavin_NYC> Good morning y'all... I've got about half an hour free, I'm going to try to do all the announce-0.7.1 stuff...
 72 2012-10-19 11:23:11 <Joric> 'The figures about Instawallet do not seem correct, on May 13th there were 103,513 wallets having at least one incoming transaction, the paper mentions 23,649 different addresses' this paper is total crap
 73 2012-10-19 12:32:28 <sipa> i wonder how much all these people who want to use the raw tx api just for finding the "sender address" of transactions will like ultraprune
 74 2012-10-19 12:34:48 <gmaxwell> "Good"
 75 2012-10-19 12:39:39 <sipa> i hope will start create dummy outputs just to keep the data in the utxo set...
 76 2012-10-19 12:39:48 <sipa> *not
 77 2012-10-19 12:42:14 <gmaxwell> We need to think more about minor fee calculation improvements to incentivize reducing the utxo set and coin selection improvements to benefit from them.
 78 2012-10-19 12:43:29 <sipa> a CalcUTXOSetSizeEffect(CTransaction) wouldn't be hard
 79 2012-10-19 12:48:16 <gmaxwell> only very weakly related???, I was contemplating how awful to implement (storage) and how useful higher order coin-days-destroyed would be; E.g. an input's priority wouldn't just be its age but age+x*(priority_of_parent*value/parent_output_value) for some small but non-zero x.
 80 2012-10-19 13:13:14 <sipa> the problem with rules based on the utxo state, is that in order for clients to predict it, they need the utxo set
 81 2012-10-19 13:13:23 <sipa> which you don't want to force
 82 2012-10-19 13:13:58 <sipa> you can approximate things of course, certainly where wallets have the prevout tx they arr spending from
 83 2012-10-19 13:14:06 <epscy> what is uxto?
 84 2012-10-19 13:14:31 <sipa> but the most interesting stuff, like the last txout of a tx being spent, can't necessarily be observed
 85 2012-10-19 13:14:47 <sipa> epscy: unspent transaction output (set)
 86 2012-10-19 13:15:26 <epscy> ahh, the spendable balance basically
 87 2012-10-19 13:15:52 <sipa> and i think we'd really like to enciurage spending the last output of a tx
 88 2012-10-19 13:16:15 <sipa> epscy: except it's not a balance at all, but yes
 89 2012-10-19 13:19:07 <gmaxwell> sipa: for the way ultraprune is structured does it really matter if it spends the last output or a non-last output?
 90 2012-10-19 13:20:15 <sipa> it's interbally a txid->list_of_outputs map
 91 2012-10-19 13:23:41 <jgarzik> ;;seen gavinandresen
 92 2012-10-19 13:23:42 <gribble> gavinandresen was last seen in #bitcoin-dev 1 day, 22 hours, 48 minutes, and 52 seconds ago: <gavinandresen> I wonder how much it costs a bank to process a cash-at-the-ATM deposit....
 93 2012-10-19 13:23:48 <jgarzik> hrm
 94 2012-10-19 13:23:54 <jgarzik> need to get that release out
 95 2012-10-19 13:27:36 <gmaxwell> 06:20 < gavin_NYC> Good morning y'all... I've got about half an hour free, I'm going to try to do all the announce-0.7.1 stuff...
 96 2012-10-19 13:27:51 <sipa> gmaxwell: so spending the last txout of a txid usually results in a far larger reduxtion in utxo size than any other
 97 2012-10-19 13:28:20 <sipa> especially as just the txid (the key) is far larger than the value data left at that point
 98 2012-10-19 13:28:32 <jgarzik> ah, good
 99 2012-10-19 13:30:26 <gmaxwell> sipa: oh right duh. for some blonde-moment reason I was thinking they were indexed by scriptpubkey instead of by txid:vout
100 2012-10-19 13:31:39 <gmaxwell> sipa: one way to make that score memoryless is to have a different cost for different vout indexes. E.g. spending index zero is much more likely to be spending the last index than index ten.
101 2012-10-19 13:34:07 <freewil> is it possible to link two multisig addresses together if they both share 1 (of 2) common keys
102 2012-10-19 13:34:36 <gmaxwell> freewil: Can you expand that question some? what does link mean?
103 2012-10-19 13:35:42 <freewil> i mean if you wanted to generate a bunch of multisig addresses using one common key, would someone who analyzes the blockchain be able to see that those multisig addresses with the common key have a common owner?
104 2012-10-19 13:36:44 <freewil> ... just based on the addresses
105 2012-10-19 13:36:45 <gmaxwell> Not until they're spent.
106 2012-10-19 13:36:45 <sipa> freewil: in case of p2sh, after being spent only
107 2012-10-19 13:36:53 <gmaxwell> Right what sipa said.
108 2012-10-19 13:37:43 <freewil> because the spending transaction would have inputs from two addresses?
109 2012-10-19 13:37:50 <freewil> or can you tell some other way?
110 2012-10-19 13:37:56 <KOLANICH> hi all
111 2012-10-19 13:38:21 <freewil> ...i mean inputs from two different multisig addresses
112 2012-10-19 13:39:39 <gmaxwell> freewil: because when its spent the script is disclosed and you can see the public keys.
113 2012-10-19 13:40:26 <freewil> ok - so if you have a multisig address c using two keys a and b, you'd be able to tell what a and b are once an output is spent that was sent to c?
114 2012-10-19 13:41:03 <gmaxwell> Yes.
115 2012-10-19 13:41:24 <gmaxwell> (of course, knowing a or b doesn't define 'ownership')
116 2012-10-19 13:41:37 <freewil> right
117 2012-10-19 13:41:41 <freewil> ok thanks!
118 2012-10-19 13:43:30 <jgarzik> 
ACTION wonders if the average age on the forums is <21, or even <18

119 2012-10-19 13:43:46 <jgarzik> a lot of posters just seem to have no clue at all about the legal system
120 2012-10-19 13:44:06 <gmaxwell> some people I would have pegged for 14 have turned out to be 41??? so I dunno.
121 2012-10-19 13:44:08 <edcba> 'the' ?
122 2012-10-19 13:45:28 <gmaxwell> edcba: it's a fair statement, law differs across the world, but people in our community have reasoning so detached from reality they dont even manage to catch the features which are universal.
123 2012-10-19 14:14:09 <jgarzik> bitcoin.org still wants updating for 0.7.1 release
124 2012-10-19 14:24:57 <sipa> is 0.7.1 released?
125 2012-10-19 15:00:54 <drizztbsd> lol last version of bitcoin for ubuntu => 0.3.24~dfsg-1
126 2012-10-19 15:03:12 <helo> for quantal?
127 2012-10-19 15:04:04 <drizztbsd> 12.04
128 2012-10-19 15:04:13 <drizztbsd> quantal is not lts (useless on servers)
129 2012-10-19 15:04:21 <drizztbsd> quantal has 0.6.2.2
130 2012-10-19 15:04:33 <drizztbsd> silly since 0.7.0 is out since month
131 2012-10-19 15:05:57 <FabianB> does wallet.dat just contain private keys or also adress book and done transactions?
132 2012-10-19 15:06:29 <Luke-Jr> drizztbsd: and 0.6.3 longer
133 2012-10-19 15:06:58 <gmaxwell> FabianB: yep.
134 2012-10-19 15:07:27 <andyrossy> i want to download all the bitcoins
135 2012-10-19 15:07:33 <andyrossy> how much hash do i need?
136 2012-10-19 15:07:36 <drizztbsd> does 0.7.1 contains some security patch?
137 2012-10-19 15:08:46 <FabianB> gmaxwell: both or just the private keys?
138 2012-10-19 15:10:07 <gmaxwell> FabianB: all of the above.
139 2012-10-19 15:10:31 <FabianB> gmaxwell: ok, thx
140 2012-10-19 15:19:27 <jgarzik> drizztbsd: it contains what the changelog says it contains ;p
141 2012-10-19 15:19:51 <jgarzik> anyway
142 2012-10-19 15:20:02 <jgarzik> next step: chase down Gavin for final ACK on ultraprune merge
143 2012-10-19 15:20:09 <jgarzik> then chase down sipa to merge it
144 2012-10-19 15:20:27 <jgarzik> I think there were some minor OSX build issues, that sounded easy to fix?
145 2012-10-19 15:21:06 <Luke-Jr> jgarzik++
146 2012-10-19 15:21:14 <Luke-Jr> drizztbsd: not afaik
147 2012-10-19 15:26:44 <helo> so 0.8 will include ultraprune/leveldb... will it actually prune anything?
148 2012-10-19 15:27:23 <gmaxwell> helo: depends on what you mean by prune.
149 2012-10-19 15:28:08 <gmaxwell> Ultraprune is based on a maximally pruned txout index. The nodes still have all the data and are indistinguishable over the p2p from older nodes.
150 2012-10-19 15:30:55 <helo> spent txouts aren't needed, even for ibd?
151 2012-10-19 15:31:33 <gmaxwell> helo: We still have the spent txouts in ultraprune.
152 2012-10-19 15:31:38 <gmaxwell> They're just not indexed.
153 2012-10-19 15:31:47 <gmaxwell> But they're still part of the blocks.
154 2012-10-19 15:31:57 <helo> ahh, ok
155 2012-10-19 15:35:40 <Eliel_> will 0.8 have the option to delete the blockchain data on function with just the ultraprune database?
156 2012-10-19 15:40:29 <gmaxwell> Eliel: no. And it's not that simple.
157 2012-10-19 15:40:46 <gmaxwell> The ultraprune database alone couldn't possibly reorg.
158 2012-10-19 15:45:33 <Eliel> yes, so you'd need to keep a buffer of a number of the most recent blocks in addition. That won't be supported?
159 2012-10-19 15:47:49 <BlueMatt> ;;seen TD
160 2012-10-19 15:47:49 <gribble> TD was last seen in #bitcoin-dev 2 days, 20 hours, 43 minutes, and 23 seconds ago: <TD> (or at least close to real time)
161 2012-10-19 15:56:20 <UukGoblin> "Replace hard-coded maximum block size (1,000,000 bytes) and maximum number of signature operations per block (20,000) with ???." <- is there a discussion somewhere on "???"?
162 2012-10-19 15:57:23 <gmaxwell> UukGoblin: that page is dreaming, not a plan.
163 2012-10-19 15:58:03 <gmaxwell> UukGoblin: there are a bunch of silly threads on that subject in varrious places.
164 2012-10-19 15:59:19 <UukGoblin> well, I think it's an interesting problem and would basically like to know more about the available ideas... but not silly ones
165 2012-10-19 16:00:59 <UukGoblin> I can even see a solution where it could stay at 1MB/20k forever...
166 2012-10-19 16:01:13 <UukGoblin> but it'd harm decentralization and small end-users
167 2012-10-19 16:02:11 <gmaxwell> what? harm decentralization?? er. The biggest reason to not up the cap is to prevent decenteralization from being totally lost.
168 2012-10-19 16:02:42 <gmaxwell> If you allow really large blocks than the potential cost of validation can become quite large... so there would be more incentive to 'let someone else do the validation'.
169 2012-10-19 16:03:31 <edcba> ??? should not be a hard limit
170 2012-10-19 16:03:33 <UukGoblin> ah, well, I was thinking to basically leave transaction processing to large wallet companies
171 2012-10-19 16:03:37 <edcba> it should be computed limit
172 2012-10-19 16:04:03 <edcba> we shouldn't have to change a hard limit every x release
173 2012-10-19 16:04:09 <UukGoblin> all small and high-volume transactions are processed internally within wallet providers (and between wallet providers using contracts), and then wallet providers only make large adjustment transactions rarely
174 2012-10-19 16:04:27 <edcba> block size is more a bandwidth problem
175 2012-10-19 16:04:34 <UukGoblin> bandwidth + storage
176 2012-10-19 16:04:55 <edcba> ok maybe we should have some flags on blocks
177 2012-10-19 16:05:10 <UukGoblin> gmaxwell, if you leave the cap at where it is, and let transaction rates grow, you'll effectively boost up the transaction fees
178 2012-10-19 16:05:20 <edcba> so that each client could tell if he wants to raise some value or to decrease it
179 2012-10-19 16:05:22 <UukGoblin> making it hard for small users to transfer bitcoins
180 2012-10-19 16:05:23 <gmaxwell> edcba: that @#$#@
181 2012-10-19 16:05:39 <gmaxwell> edcba: so should we also not have a total limit on the number of coins.
182 2012-10-19 16:05:42 <UukGoblin> edcba, I saw a discussion on that earlier
183 2012-10-19 16:05:58 <edcba> gmaxwell: interesting question :)
184 2012-10-19 16:06:04 <gmaxwell> edcba: You fail. Allowing miners to adjust the size is an interesting suicide pact.
185 2012-10-19 16:06:11 <edcba> but i was only thinking about technicals limit :)
186 2012-10-19 16:06:46 <edcba> i forgot that blocks were tied to fees too
187 2012-10-19 16:06:46 <UukGoblin> so in order to let lots of small cheap transfers, you do them elsewhere and aggregate them
188 2012-10-19 16:06:50 <edcba> fucking miners :)
189 2012-10-19 16:07:11 <gmaxwell> edcba: no no, you can't hide behind that. The security of the system depends on scarcity of block space, and wide distribution of validation.  The economic consequences are technical consequences too. :P
190 2012-10-19 16:07:36 <edcba> yes ok i retract my suggestion !
191 2012-10-19 16:07:52 <edcba> fcking bitcoin client we can't modify at all :)
192 2012-10-19 16:07:53 <gmaxwell> Well it's complicated. I agree.
193 2012-10-19 16:08:38 <edcba> hmm
194 2012-10-19 16:08:49 <edcba> maybe we should implement some meta feedback loop then
195 2012-10-19 16:09:25 <edcba> the problem is now to define it :p
196 2012-10-19 16:10:11 <edcba> something preventing us to fall in commons tragedy or whatever it is called
197 2012-10-19 16:11:09 <edcba> i don't see easy solution :(
198 2012-10-19 16:11:34 <gmaxwell> Not that big of a deal.
199 2012-10-19 16:12:21 <gmaxwell> It's not like bitcoin could ever reasonably handle the worldwide cash transaction volume??? you'll always have to have secondary systems that have enough information hiding to be scalable.
200 2012-10-19 16:12:27 <UukGoblin> like, a subbitcoin network, of some sort. Maybe based on OpenTransactions.
201 2012-10-19 16:12:34 <sipa> jgarzik: why would you need me to merge it? :)
202 2012-10-19 16:13:22 <edcba> ok i know what is the problem of bitcoin ! it's sipa and likes :)
203 2012-10-19 16:13:28 <jgarzik> sipa: <shrug>  Seems like the usual procedure for "big ticket items" is to gather ACKs, then merge it yourself
204 2012-10-19 16:13:31 <UukGoblin> and then... have the actual bitcoin nodes validate the OT status...
205 2012-10-19 16:13:35 <jgarzik> but yes, it doesn't really matter who merges it
206 2012-10-19 16:13:37 <sipa> jgarzik: ah, sure :)
207 2012-10-19 16:13:42 <edcba> the mining clans should be built in bitcoin
208 2012-10-19 16:13:57 <edcba> ie more frequent rewards
209 2012-10-19 16:13:57 <sipa> edcba: i'm a problem for bitcoin? geez, thanks...
210 2012-10-19 16:14:25 <edcba> ok more diplomatically you are the solution we need to internalize in bitcoin :p
211 2012-10-19 16:14:55 <edcba> aka we need to make you redundant so we can fire you :)
212 2012-10-19 16:15:52 <Eliel> I think Proof of Stake type of voting might work for deciding the maximum block size.
213 2012-10-19 16:16:24 <edcba> hmm
214 2012-10-19 16:16:27 <Eliel> although, perhaps not a direct vote
215 2012-10-19 16:16:29 <edcba> 
ACTION seeks that...

216 2012-10-19 16:16:33 <UukGoblin> Eliel, how often do you hold the elections then?
217 2012-10-19 16:17:17 <Eliel> UukGoblin: people with lots of bitcoins have an incentive to keep people using Bitcoin and to keep the security up.
218 2012-10-19 16:17:41 <UukGoblin> Eliel, and really, again... blocksize should stay small for /small/ users, and by holding proof-of-stake elections, you're immediately favouring BIG users
219 2012-10-19 16:18:18 <Eliel> UukGoblin: define your BIG and small use.
220 2012-10-19 16:18:37 <UukGoblin> small = can't prove stake
221 2012-10-19 16:18:46 <edcba> BIG : |------------------| small : |-----|
222 2012-10-19 16:18:46 <UukGoblin> for the purpose of this subject. ;-]
223 2012-10-19 16:18:50 <edcba> not to scale of course
224 2012-10-19 16:19:15 <UukGoblin> edcba, thank you, greatly, for your insightful input.
225 2012-10-19 16:19:21 <Eliel> UukGoblin: that's the whole point. To be able to make a big effect, you need to have big risks on the results too.
226 2012-10-19 16:20:15 <Wollo> hello
227 2012-10-19 16:20:48 <UukGoblin> oh, and also: current fee structure is kinda weird.
228 2012-10-19 16:20:57 <edcba> yes fees shouldn't event exist imo
229 2012-10-19 16:21:10 <edcba> even
230 2012-10-19 16:21:18 <UukGoblin> it encourages people to make lots of small transactions rather than few big ones
231 2012-10-19 16:21:30 <edcba> hmm
232 2012-10-19 16:21:33 <UukGoblin> effectively resulting in more "spam", i.e. useless data being transferred and stored
233 2012-10-19 16:21:46 <edcba> not sure about that
234 2012-10-19 16:21:55 <edcba> tragedy of commons too
235 2012-10-19 16:22:02 <Eliel> UukGoblin: it might do that in some situations, sure.
236 2012-10-19 16:22:40 <Eliel> UukGoblin: but even more than that, it encourages people to receive coins at reasonable sizes compared to their needs.
237 2012-10-19 16:22:59 <Wollo> I'm curious about a thing. Is it possible to "track" a bitcoin to retrive a list of the addresses of all of his previous owners?
238 2012-10-19 16:23:07 <edcba> yeah fees should be externalized to bitcoin
239 2012-10-19 16:23:18 <edcba> ie negotiating them between peers
240 2012-10-19 16:23:20 <UukGoblin> Wollo, yes, definitely possible
241 2012-10-19 16:23:45 <edcba> in same way than block exchange protocol
242 2012-10-19 16:23:46 <Wollo> thank you UukGoblin
243 2012-10-19 16:23:48 <Eliel> Wollo: yes, but it's somewhat of a fuzzy tracking.
244 2012-10-19 16:24:18 <Wollo> thank you too Eliel
245 2012-10-19 16:24:46 <edcba> "process my tx and i'll give you block notifications"
246 2012-10-19 16:25:00 <edcba> or peer with me and i'll give you some btc
247 2012-10-19 16:25:02 <UukGoblin> while "owners" imo can get fuzzy, you can definitely track all the addresses a bitcoin has been to since being mined
248 2012-10-19 16:25:16 <edcba> ie maybe we should just rewards peers
249 2012-10-19 16:25:28 <edcba> good acting peers of course
250 2012-10-19 16:25:34 <BlueMatt> gmaxwell: what ever came of the lets-overhaul-fees discussion late summer?
251 2012-10-19 16:25:35 <Eliel> when there's a transaction with more than one input, it doesn't even make sense to try to pin the bitcoin to one of those inputs.
252 2012-10-19 16:26:16 <UukGoblin> Eliel, no, you pin it to all of them... ok, sort of fuzzy, then ;-]
253 2012-10-19 16:26:19 <edcba> so i imagine a bitcoin client having a lot more peers than now...
254 2012-10-19 16:26:40 <jgarzik> BlueMatt: for one, we sort blocks by fee-per-kb
255 2012-10-19 16:27:44 <BlueMatt> jgarzik: did we (finally) change the actual priority algo?
256 2012-10-19 16:28:11 <BlueMatt> jgarzik: also, you gonna be at pycarolinas?
257 2012-10-19 16:30:17 <jgarzik> BlueMatt: see c555400ca134991e39d5e3a565fcd2215abe56f6 -blockminsize -blockmaxsize -blockprioritysize
258 2012-10-19 16:30:43 <gmaxwell> We didn't change the priority algo, just the fee handling and block sizing.
259 2012-10-19 16:30:44 <jgarzik> BlueMatt: gavin had a gist somewhere describing these changes, but I do not seem to have that link (sipa? gmaxwell?)
260 2012-10-19 16:31:02 <BlueMatt> jgarzik: yea, but the gist went way further than just that commit
261 2012-10-19 16:31:20 <BlueMatt> jgarzik: I was wondering how much more was done than just sort by fee-per-kb
262 2012-10-19 16:31:29 <jgarzik> gmaxwell: well, the size of the priority area is now easily malleable
263 2012-10-19 16:31:57 <jgarzik> BlueMatt: much more has been proposed... but the above is all that's been pushed upstream
264 2012-10-19 16:32:08 <jgarzik> e.g. looking at block history for fee calculation help
265 2012-10-19 16:32:10 <BlueMatt> mmm, shame
266 2012-10-19 16:33:35 <UukGoblin> ah, cool stuff
267 2012-10-19 16:35:25 <jgarzik> BlueMatt: don't recall anything about pycarolinas
268 2012-10-19 16:35:43 <BlueMatt> jgarzik: guess not then...its tomorrow
269 2012-10-19 17:50:27 <edcba> i guess everyone saw that link : http://eprint.iacr.org/2012/584 ?
270 2012-10-19 17:52:22 <jgarzik> edcba: Yes.  https://bitcointalk.org/index.php?topic=118797.0
271 2012-10-19 17:52:38 <jgarzik> edcba: that thread outlines some of the paper's flaws
272 2012-10-19 18:06:48 <helo> i find myself kind of enjoying all of these new "revelations" about bitcoin, and the inevitable proclamations of doom
273 2012-10-19 18:09:43 <edcba> i still wonder about the inflation/deflation stuff
274 2012-10-19 18:09:45 <helo> bitcoin is so radical that most criticism is just accepted at face value... so many people are just waiting to hear why bitcoin won't work, despite the fact that it is already working
275 2012-10-19 18:10:55 <helo> edcba: yeah me too... an economy based around a deflationary currency is probably very unlike our present economy
276 2012-10-19 18:11:45 <helo> the fun/scary part is that without draconian government crackdown, bitcoin is likely to force itself to prominence and we'll find out of the deflationary caution is valid
277 2012-10-19 18:12:49 <helo> out *if the
278 2012-10-19 18:14:35 <helo> the transition from "~stable inflationary currency economy" to "~stable deflationary currency economy" is going to be so bumpy that it will seem that the warnings about deflation were true
279 2012-10-19 18:17:38 <root2> I thought the problem with deflation was mostly divisibility, and loans(borrowing money and having to pay it back when it is worth more)
280 2012-10-19 18:20:36 <helo> most economics refer to deflationary spiral as the big problem, afaik
281 2012-10-19 18:21:00 <helo> i kind of like the idea of people not being under pressure to blow all of their cash on something, anything before it loses its value
282 2012-10-19 18:26:54 <gmaxwell> helo: hey, if you're in debt denominated in dollars than inflation is your friend!
283 2012-10-19 18:27:34 <D34TH> bitcoin is using so much memory
284 2012-10-19 18:27:56 <D34TH> 500MB
285 2012-10-19 18:27:59 <D34TH> D:
286 2012-10-19 18:28:19 <gmaxwell> D34TH: are you looking at virtual or actual in use?
287 2012-10-19 18:28:35 <D34TH> MEM% in htop, so actual
288 2012-10-19 18:28:40 <D34TH> VIRT is 1090M
289 2012-10-19 18:28:57 <gmaxwell> here I'm seeing 167m ... :P But this is not on a public listener.
290 2012-10-19 18:29:22 <D34TH> "connections" : 74,
291 2012-10-19 18:29:43 <gmaxwell> yea, that would be it.
292 2012-10-19 18:29:59 <D34TH> D:
293 2012-10-19 18:34:24 <sipa> which version?
294 2012-10-19 18:34:32 <D34TH> "version" : 70100,
295 2012-10-19 18:34:37 <sipa> ok
296 2012-10-19 18:34:41 <D34TH> i.e. HEAD
297 2012-10-19 18:35:04 <D34TH> im generally always running the latest commit
298 2012-10-19 18:35:33 <kjj_> odd.  my node is hanging when I try use the move RPC command on 0.7.1rc1
299 2012-10-19 18:36:05 <jgarzik> The economics bits of bitcoin are rather far less interesting than the technical bits.  But anyway...  RE deflation:  the huge caveat critics often miss is that most models and studies of deflation (or gold-backed currencies) involve nation-state currencies where everybody is locked in.
300 2012-10-19 18:36:28 <jgarzik> The motivations and uses and economics change radically, when one may slip in and out of a private currency with ease.
301 2012-10-19 18:37:52 <kjj_> hmm.  just before that, listaccounts logged hundreds of "Unknown transaction type found"
302 2012-10-19 18:38:31 <sipa> yes, i think anyone who considers bitcoin's succes outcome to be replacing a nation's money...
303 2012-10-19 18:38:58 <sipa> lkjjthat sounds bad...
304 2012-10-19 18:39:11 <sipa> kjj_: that sounds bad
305 2012-10-19 18:39:43 <helo> gmaxwell: a new deflationary currency is doubly so :)
306 2012-10-19 18:39:51 <kjj_> sipa: it seems bad
307 2012-10-19 18:40:18 <jgarzik> it is bad
308 2012-10-19 18:41:31 <sipa> kjj_: is your wallet sane?
309 2012-10-19 18:41:53 <kjj_> heh, it was
310 2012-10-19 18:42:06 <sipa> bleh
311 2012-10-19 18:42:24 <kjj_> at this point, I have no idea
312 2012-10-19 18:48:24 <kjj_> pywallet dump is huge, no errors so far, at least not that I can tell
313 2012-10-19 18:57:13 <D34TH> hmm, reimporting my blockchain from bootstrap.dat lets see what happens
314 2012-10-19 18:59:59 <kjj_> hmm.  I had generated a raw transaction, collecting a bunch of small outputs
315 2012-10-19 19:00:00 <gmaxwell> D34TH: watcha doing that for?
316 2012-10-19 19:00:12 <D34TH> https://bitcointalk.org/index.php?topic=22785.msg1283806#msg1283806
317 2012-10-19 19:00:21 <kjj_> I've done a bunch of those, but this one was twice the usual size, around 7k bytes
318 2012-10-19 19:00:45 <kjj_> I see it in listtransactions
319 2012-10-19 19:01:37 <kjj_> I see it in getrawmempool
320 2012-10-19 19:01:41 <D34TH> whats the average time for importing bootstrap.dat
321 2012-10-19 19:03:32 <sipa> D34TH: depends a lot on your system, in particulat I/O
322 2012-10-19 19:03:49 <D34TH> heh, load breaking 3.0
323 2012-10-19 19:04:45 <sipa> tmpfs > SSD > HDD > truecrypt HDD > USB
324 2012-10-19 19:05:08 <D34TH> its going to my raid
325 2012-10-19 19:05:09 <D34TH> D:
326 2012-10-19 19:05:30 <sipa> raid doesn't really help with write latency, which is the worst problem on BDB
327 2012-10-19 19:08:19 <midnightmagic> repends on the raid type and whether you have hardware raid assistance.
328 2012-10-19 19:08:44 <midnightmagic> also, tests here for larger datasets strongly suggest SSD are outperformed by HDD in the long run.
329 2012-10-19 19:09:17 <sipa> depends what kind of load
330 2012-10-19 19:10:13 <midnightmagic> to end-run around ssd performance constraints, the only solution is to buy an ssd which has wayyyyyy more in its ring buffers than it advertises it does to the host, and then do lots of garbage collection in the background. however, if you keep writing to it, eventually you will catch up with the garbage collector and write performance will suffer. there is basically no way to avoid it.
331 2012-10-19 19:10:31 <midnightmagic> sipa: This is with bdb-derived database backend.
332 2012-10-19 19:10:56 <sipa> ok
333 2012-10-19 19:11:07 <Gladamas> anyone here know a lot about advanced linux openvpn connection issues?
334 2012-10-19 19:11:41 <jurov> Gladamas, i know about some related to shitty dsl routers or mikrotik
335 2012-10-19 19:11:57 <jurov> dunnof if that's sufficiently advanced
336 2012-10-19 19:12:38 <Gladamas> ah. idk, I'm just relaying someone else's request
337 2012-10-19 19:13:01 <midnightmagic> Gladamas: Why aren't they asking for themselves?
338 2012-10-19 19:13:21 <Gladamas> i referred them here, i just was curious if anyone here knew anything first
339 2012-10-19 19:13:39 <midnightmagic> I use openvpn extensively, but that doesn't mean I know what you're talking about. :(
340 2012-10-19 19:14:12 <helo> Gladamas: i recommend #openvpn
341 2012-10-19 19:14:31 <helo> seems relatively active
342 2012-10-19 19:14:36 <Gladamas> thanks
343 2012-10-19 19:20:21 <D34TH> all imported
344 2012-10-19 19:21:30 <D34TH> uh oh
345 2012-10-19 19:21:34 <D34TH> something bad happened
346 2012-10-19 19:21:41 <D34TH> boost::filesystem::remove: Read-only file system:
347 2012-10-19 19:21:59 <D34TH> crap
348 2012-10-19 19:22:02 <sipa> ha, it tries to remove bootstrap.dat after loading
349 2012-10-19 19:22:12 <sipa> it shouldn't break though
350 2012-10-19 19:22:15 <D34TH> well now i cant even reboot my box
351 2012-10-19 19:22:20 <sipa> but it should be imported...
352 2012-10-19 19:22:22 <sipa> ?
353 2012-10-19 19:22:23 <D34TH> bash: /sbin/reboot: Input/output error
354 2012-10-19 19:22:45 <sipa> waah
355 2012-10-19 19:22:55 <sipa> i think your disk or fs died
356 2012-10-19 19:23:02 <D34TH> disk is alive
357 2012-10-19 19:23:03 <sipa> (well, its driver)
358 2012-10-19 19:23:08 <D34TH> ahh
359 2012-10-19 19:23:17 <sipa> check dmesh
360 2012-10-19 19:24:00 <D34TH> nothing close to current time
361 2012-10-19 19:24:32 <sipa> that's very strange
362 2012-10-19 19:25:00 <D34TH> ahh
363 2012-10-19 19:25:05 <D34TH> kern.log
364 2012-10-19 19:25:20 <D34TH> failed command: WRITE FPDMA QUEUED
365 2012-10-19 19:25:29 <D34TH> (ATA bus error)
366 2012-10-19 19:29:05 <kjj_> sipa: after a restart, node seems to be running fine.  nothing complaining about the wallet so far, but I haven't tried doing the move again
367 2012-10-19 20:16:36 <midnightmagic> Gladamas: Was the guy you were asking on behalf of named JimBags ?
368 2012-10-19 20:17:01 <Gladamas> no
369 2012-10-19 20:17:11 <Gladamas> sjuxax
370 2012-10-19 20:17:24 <midnightmagic> Gladamas: I was going to mock you for your bad choice in friends.
371 2012-10-19 20:18:31 <Gladamas> :P lol.
372 2012-10-19 21:04:27 <gmaxwell> Luke-Jr: I feel kinda bad for making CP's product sound bad esp since the BFL's figures may all be lies, but you may enjoy the bomb I tossed into that silly argument: https://bitcointalk.org/index.php?topic=79637.msg1284513#msg1284513
373 2012-10-19 21:30:19 <midnightmagic> gmaxwell: You missed one more ephemeral property in your breakeven calculations.
374 2012-10-19 21:30:25 <midnightmagic> gmaxwell: The currency is deflationary. :)
375 2012-10-19 21:30:48 <midnightmagic> "The nature of whatever bitcoins are" is deflationary.
376 2012-10-19 21:31:48 <gmaxwell> Well, I was trying to take the bitcoins out of it as much as possible.
377 2012-10-19 21:33:03 <gmaxwell> Both produce bitcoins, so at least in the one vs another analysis you can factor them out (ignoring ship times). Realistically both will become worthless when someday someone produces a new one on 14nm 3d-transistor graphite gate or whatever is the next great hotness.
378 2012-10-19 21:34:03 <midnightmagic> also, where did this number come from: $13.62 for your bfl operating costs figure?
379 2012-10-19 21:34:50 <gmaxwell> it's the forgone income from not investing $1300 at 8% and paying for 60w of power at $0.12/kwh
380 2012-10-19 21:36:09 <gmaxwell> (for one month). E.g. at the beginning of the month you have $1300 you can either have a BFL or stock market average returns.  So choosing the BFL costs you about $8.36 plus the power.
381 2012-10-19 21:36:37 <midnightmagic> ah, "plus the power". ok.
382 2012-10-19 21:36:53 <midnightmagic> duh
383 2012-10-19 21:37:04 <gmaxwell> I'll replace the $13.62 with the sum
384 2012-10-19 21:40:33 <midnightmagic> the foregone investment "loss" is less for the btcfpga for similar hashrate, so it makes sense to add in the cost of the potential "other" profits in there, just as an explanation of what people could be doing with their money, rather than just using the ratio of the costs..
385 2012-10-19 21:40:54 <midnightmagic> yeah, makes sense now..
386 2012-10-19 21:44:27 <gmaxwell> Perhaps I should have also included the full analysis with mining profit assumptions??? under my assumptions it makes mining fairly unattractive in the long term vs conventional investment... but making any assumptions about future difficulty is the stuff that argument are made of...
387 2012-10-19 21:51:21 <midnightmagic> LOL you're accounting for february aren't you? with the average hours-in-a-month?
388 2012-10-19 21:51:55 <gmaxwell> yea, I'm mostly accounting for leap-years there. Not fully though.
389 2012-10-19 21:52:07 <midnightmagic> LOL that's hilariously accurate
390 2012-10-19 21:52:31 <sipa> just use 365.2425 days per year
391 2012-10-19 21:52:33 <Gladamas> BFL ASIC pics, ASICs to ship Nov/Dec: https://bitcointalk.org/index.php?topic=119351.0;topicseen
392 2012-10-19 21:54:14 <gmaxwell> I wonder if those surprisingly non-descript packages are just dummies for oven testing and thermal measurements.
393 2012-10-19 21:55:09 <midnightmagic> sipa: I wish my power company sent me bills that regularly! :-)
394 2012-10-19 21:55:25 <sipa> MC1984: found a bug - bootstrap.dat was indeed not working in ultraprune (but it was caused by a bug by me)
395 2012-10-19 21:59:44 <MC1984> oh i was right
396 2012-10-19 22:00:10 <sipa> yup :)
397 2012-10-19 22:00:23 <gmaxwell> MC1984 hurray for testing
398 2012-10-19 22:00:25 <MC1984> omg i helped lol
399 2012-10-19 22:00:31 <sipa> thx!
400 2012-10-19 22:00:47 <midnightmagic> "price would have to be _negative_ to make it more attractive" <-- LOL
401 2012-10-19 22:01:51 <midnightmagic> This is why he's so anxious to get his devices done sooner..
402 2012-10-19 22:02:04 <midnightmagic> He was saying something about "sparing no expense" a few weeks ago.
403 2012-10-19 22:02:24 <sipa> he?
404 2012-10-19 22:03:58 <gmaxwell> cablepair.
405 2012-10-19 22:04:36 <gmaxwell> If he beats BFL to shipping??? regardless of my analysis??? a lot of people will probably cancel bfl orders and switch to him. It's _very_ much in his interest to ship early.
406 2012-10-19 22:04:49 <Diablo-D3> cancel bfl? lol
407 2012-10-19 22:04:55 <Diablo-D3> I dont think you understand gmaxwell
408 2012-10-19 22:04:57 <Diablo-D3> they're not preorders
409 2012-10-19 22:04:57 <sipa> ah, must be a nickname :)
410 2012-10-19 22:04:59 <Diablo-D3> they're investors
411 2012-10-19 22:05:18 <Diablo-D3> all those people who keep calling themselves customers and bfl calls customers arent
412 2012-10-19 22:05:19 <Diablo-D3> they're investors
413 2012-10-19 22:05:25 <Diablo-D3> and they got burned the moment they handed the cash over
414 2012-10-19 22:05:37 <gmaxwell> Diablo-D3: Inaba says otherwise. Which also seemed unbelevable to me. But they've accepted cancellations so far.
415 2012-10-19 22:06:06 <gmaxwell> If BFL ships first then I expect that cablepair will have a hard time selling all his stock and will make a lot less money.
416 2012-10-19 22:06:09 <jgarzik> which one is cablepair?  basic?  btcfpga?
417 2012-10-19 22:06:18 <Luke-Jr> bASIC
418 2012-10-19 22:06:41 <Diablo-D3> gmaxwell: thats how a scam works btw
419 2012-10-19 22:06:47 <Diablo-D3> you refund a few, but then jew the rest
420 2012-10-19 22:07:15 <sipa> is 'to jew' a verb?
421 2012-10-19 22:07:22 <jgarzik> Looks like BTCFPGA to me: https://bitcointalk.org/index.php?topic=79637.0
422 2012-10-19 22:07:24 <Diablo-D3> sipa: yes
423 2012-10-19 22:07:29 <Diablo-D3> btw, yes, its btcfpga
424 2012-10-19 22:07:32 <jgarzik> sipa: to anti-semites, I suppose
425 2012-10-19 22:07:41 <Diablo-D3> basic I think is deepbit?
426 2012-10-19 22:07:59 <Diablo-D3> jgarzik: heh, I went to school with a jewish kid, its 100% accurate from the stories hes said about his family
427 2012-10-19 22:08:29 <jgarzik> 
ACTION already paid for a BTCFPGA one

428 2012-10-19 22:08:36 <firethief> Diablo here knows all about them jews.
429 2012-10-19 22:08:47 <sturles> midnightmagic: gmaxwell forgot an important fact in his calculations.  The miners lose value over time.
430 2012-10-19 22:08:52 <sipa> "In my country we have problem..."
431 2012-10-19 22:08:54 <Diablo-D3> sturles: depends
432 2012-10-19 22:09:07 <sturles> Moore's Law, etc.
433 2012-10-19 22:09:12 <Diablo-D3> you cant predict future difficulty or future btc value
434 2012-10-19 22:09:24 <Diablo-D3> operating costs and (some) purchase price is in dollars
435 2012-10-19 22:09:31 <sturles> Indepndent of future difficulty or BTC value.
436 2012-10-19 22:09:43 <Diablo-D3> sturles: I write off hardware purchases 100%.
437 2012-10-19 22:09:47 <Diablo-D3> so maybe you do math differently
438 2012-10-19 22:09:54 <Diablo-D3> its worth $0 the day I open the box
439 2012-10-19 22:10:14 <jgarzik> disappointing, that bitmit is closing.  Apparently you can buy bitmit, on bitmit: https://www.bitmit.net/en/trade/i/8175-bitmit-net/description   Reason for closing: "Financial supervision requires special licences for businesses which holds customer funds"
440 2012-10-19 22:10:34 <sturles> Diablo-D3: Read this and the post I quoted from for context: https://bitcointalk.org/index.php?topic=79637.msg1284615#msg1284615
441 2012-10-19 22:10:44 <Luke-Jr> ;;bc,diff
442 2012-10-19 22:10:45 <gribble> 3072321.7320208
443 2012-10-19 22:11:12 <Diablo-D3> sturles: I dont see a problem with your math
444 2012-10-19 22:11:13 <Diablo-D3> so whats the issue
445 2012-10-19 22:11:56 <sturles> If small differences in power consumption matters over price per Ghash/s for ASIC miners.
446 2012-10-19 22:12:02 <Diablo-D3> it does
447 2012-10-19 22:12:05 <Diablo-D3> did someone say it didnt?
448 2012-10-19 22:12:12 <Diablo-D3> 
ACTION has done this math like thousands of times

449 2012-10-19 22:12:15 <sturles> Yes, my math. :-)
450 2012-10-19 22:13:07 <Diablo-D3> hell, Ive even calculated that underclocking ram on 58xx from 1000/1200 to 3xx is usually enough to get you to break even a month or so earlier (depending on btc prices, electricity prices, etc)
451 2012-10-19 22:13:24 <sturles> My math says that price per Ghash is much more important than power consumption as long as power consuption and price is at this level.
452 2012-10-19 22:13:41 <Diablo-D3> sturles: atm? yes
453 2012-10-19 22:13:46 <sturles> True for GPUs, yes.
454 2012-10-19 22:13:49 <Diablo-D3> but divide profit by 100
455 2012-10-19 22:13:55 <Diablo-D3> thats the asic difficulty wall
456 2012-10-19 22:14:30 <Diablo-D3> if you've already purchased asics (key word here is already) and will have them mining dec/jan/feb
457 2012-10-19 22:14:42 <Diablo-D3> price per ghash flat out controls profit
458 2012-10-19 22:15:06 <Diablo-D3> once that wall kicks in, its power usage per ghash because you're looking at a 2 year climb to profit
459 2012-10-19 22:15:35 <Diablo-D3> sturles: so I think we agree
460 2012-10-19 22:16:08 <sturles> For ASIC miners the power is only a very small part of the costs.  You have to mine for many years until the cost of power has surpassed your initial investment.'
461 2012-10-19 22:16:42 <Diablo-D3> at current difficulty? yes
462 2012-10-19 22:16:58 <Diablo-D3> at 100x difficulty? fffffffffffff--
463 2012-10-19 22:16:59 <sturles> Independent of difficulty.
464 2012-10-19 22:17:15 <Diablo-D3> its going to be 1000x in 2 years btw
465 2012-10-19 22:17:29 <Diablo-D3> but the climb will be much slower once all the first asics get into action
466 2012-10-19 22:17:46 <sturles> Difficulty and power consumption per Ghash has nothing to do with each other.
467 2012-10-19 22:17:55 <Diablo-D3> sturles: it does.
468 2012-10-19 22:18:07 <Diablo-D3> unless your electric company accepts bitcoins
469 2012-10-19 22:18:08 <gmaxwell> sturles: ... you should go read the analysis I linked to.
470 2012-10-19 22:18:28 <Diablo-D3> _now_, if btc prices skyrocket and climb with difficulty? no, they're not related
471 2012-10-19 22:18:41 <Diablo-D3> but I dont think thats going to happen
472 2012-10-19 22:18:52 <gmaxwell> We're also not talking about small difference in power consumption, we're talking about factors like 3-5x.
473 2012-10-19 22:18:53 <sturles> gmaxwell: What link?
474 2012-10-19 22:19:14 <Diablo-D3> if difficulty increases 100x, your profit per watt decreases 100x.
475 2012-10-19 22:19:20 <Diablo-D3> you're spending 100x power to get the same dollars
476 2012-10-19 22:19:24 <gmaxwell> 16:04 < gmaxwell> Luke-Jr: I feel kinda bad for making CP's product sound bad esp since the BFL's figures may all be lies, but you may enjoy the bomb I  tossed into that silly argument: https://bitcointalk.org/index.php?topic=79637.msg1284513#msg1284513
477 2012-10-19 22:19:55 <gmaxwell> sturles: I proposed a way of reasoning about the prices between mining options that fully factors out the bitcoin market side of the deal.
478 2012-10-19 22:20:14 <Diablo-D3> gmaxwell: a functioning price per mhash metric?
479 2012-10-19 22:20:30 <sturles> Diablo-D3: Yeah, but I would have to mine with a BFL Single for 25 years until I have paid the same amount powering it as I did for the device itself.  This makes power consumption pretty much irrelevant compared to the cost of the device.
480 2012-10-19 22:20:56 <Diablo-D3> sturles: yes, but what if difficulty increases 100x? thats .25 years.
481 2012-10-19 22:21:02 <gmaxwell> Diablo-D3: I turn the price into a monthly cost by comparing it with what the same amount would return on average invested in stock market indexes.
482 2012-10-19 22:21:19 <sturles> Diablo-D3: It is like comparing two expensive tablets for what they will cost on your power bill.
483 2012-10-19 22:21:26 <gmaxwell> So then power and initial price are both just monthly costs, and you can compare the monthly costs per hashrate.
484 2012-10-19 22:21:37 <Diablo-D3> gmaxwell: heh.
485 2012-10-19 22:21:51 <sturles> DerCoin: Difficulty has no effect on the price of electricity in my country.
486 2012-10-19 22:21:56 <sturles> Diablo-D3:
487 2012-10-19 22:22:03 <Diablo-D3> sturles: exactly.
488 2012-10-19 22:22:13 <Diablo-D3> sturles: so you're screwed.
489 2012-10-19 22:22:35 <sturles> Diablo-D3: Can you please explain why difficulty changes anything?
490 2012-10-19 22:22:39 <Diablo-D3> sturles: profit.
491 2012-10-19 22:22:49 <Diablo-D3> how much did that single cost you?
492 2012-10-19 22:22:51 <sturles> I haven't even mentiond profit.
493 2012-10-19 22:23:02 <Diablo-D3> oh wait, you're mining for fun?
494 2012-10-19 22:23:03 <Diablo-D3> screw that.
495 2012-10-19 22:23:05 <Diablo-D3> you're on your own
496 2012-10-19 22:23:08 <Detritus> Difficulty goes up, therefor the ammount of btc you generate for a given hash rate goes down
497 2012-10-19 22:23:15 <sturles> I mak absolutely no attempt of prdicting profit.
498 2012-10-19 22:23:23 <Diablo-D3> sturles: then why buy the hardware
499 2012-10-19 22:23:31 <Diablo-D3> its very important to see if you can even profit at all
500 2012-10-19 22:23:37 <Diablo-D3> asics are not profitable imo
501 2012-10-19 22:23:47 <sturles> Diablo-D3: I will try to explain this once more with a teaspoon.
502 2012-10-19 22:24:16 <sturles> Diablo-D3: If you pay 1000 for a miner and 50 to power it for a year.
503 2012-10-19 22:24:25 <sturles> Diablo-D3: Does power consumption matter?
504 2012-10-19 22:24:52 <Diablo-D3> sturles: yes because of factors that exist that you did not mention in your problem.
505 2012-10-19 22:24:55 <sipa> sturles: if your income makes it pay it off after 20 years only, yes
506 2012-10-19 22:24:57 <Diablo-D3> miner output is not linear.
507 2012-10-19 22:24:59 <sturles> It it is 25 or 100, and the expcted lifetime is 2 or 3 years, does it matter?
508 2012-10-19 22:25:10 <Diablo-D3> miner output is not linear.
509 2012-10-19 22:25:23 <sturles> ?
510 2012-10-19 22:25:29 <Diablo-D3> I do not pay for ghashes, I pay for future btc.
511 2012-10-19 22:25:45 <Diablo-D3> whatever gets me the most future btc for the least dollars it the winner.
512 2012-10-19 22:25:46 <sturles> Right.
513 2012-10-19 22:25:59 <Diablo-D3> the way you are describing the problem is wrong.
514 2012-10-19 22:26:04 <sturles> I can't predict future BTC, and make absolutely no attempt of doing it.
515 2012-10-19 22:26:09 <Diablo-D3> I can.
516 2012-10-19 22:26:19 <Diablo-D3> difficulty is going to skyrocket because of asic adoption
517 2012-10-19 22:26:19 <sturles> ASIC miners may be profitable, and may be not.
518 2012-10-19 22:26:45 <Diablo-D3> profitability, unless you're in on the full three months of dec/jan/feb, is going to be hard.
519 2012-10-19 22:27:04 <sturles> Profitablility changes nothing in this equation.
520 2012-10-19 22:27:12 <Diablo-D3> in yours? no
521 2012-10-19 22:27:14 <Diablo-D3> in mine? very yes.
522 2012-10-19 22:27:20 <sturles> Right.
523 2012-10-19 22:30:02 <vazakl> sup coin bros
524 2012-10-19 22:30:13 <sturles> I don't even try to predict future prices or difficulty or new ASICs coming on the market.  I am just trying to explain that when the power consumtion over a year for a current device costs 1/25th of what I pay to buy the device, it is totaly not important.  The price per Ghash/s is what's important.
525 2012-10-19 22:32:04 <gmaxwell> sturles: You can keep saying this, but go walk through the figures on the thread. It matters.
526 2012-10-19 22:32:13 <sturles> gmaxwell: If you invest in a stock market index, you don't lose the money.  You can sell your investment and get your money back.
527 2012-10-19 22:32:34 <sturles> You will not get your money back for an old miner.
528 2012-10-19 22:35:01 <sturles> gmaxwell: Your coice is between investing in a fund which normally will become worth more over time VS something which (hopefully) will make money but itself will become less worth over time.
529 2012-10-19 22:36:06 <sturles> gmaxwell: In your calculation you forgot to write off the value of the miner over time.
530 2012-10-19 22:36:23 <gmaxwell> I don't forget I intentionally do not do so??? and I even explain why.
531 2012-10-19 22:36:32 <gmaxwell> Because the parameters for doing it are highly debatable.
532 2012-10-19 22:37:02 <sturles> Yes, but this will not make the parameters go away.
533 2012-10-19 22:37:18 <gmaxwell> You responded doing so??? and I think you got it wrong, because you didn't give a longer life to the device with lower power.  ... but never the less you basically agreed with me: You show the devices _tie_ at 149 watts. No one expects the basic to be that low.
534 2012-10-19 22:37:38 <gmaxwell> my own expectation was more like 300 watts.
535 2012-10-19 22:38:04 <gmaxwell> The power _matters_. It's not the only factor, but it absolutely does matter.
536 2012-10-19 22:38:19 <sturles> Well, Tom says it will be "competetive".
537 2012-10-19 22:38:29 <gmaxwell> My own thinking is that it's better to support other vendors because we'll all win from competition.. but power still matters.
538 2012-10-19 22:39:13 <sturles> It doesn't matter much to m anyway.  My house needs heating all year, and electric power is the cheapest way to heat it.  A heat pump would be more efficient, but a heat pump doesn't make money.
539 2012-10-19 22:39:14 <gmaxwell> sturles: well, I've run numbers on 130nm designs, and I'd probably take a bet that he can't meet 54GH at 150w.  There are plenty of people in that thread arguing that power is totally irrelevant.
540 2012-10-19 22:39:31 <gmaxwell> sure, I note mining for heating in my post.
541 2012-10-19 22:39:33 <sturles> Arhg.  My e-key is broken.  Very unreliable.
542 2012-10-19 22:40:10 <gmaxwell> (I do that too??? totally miner heated although I have a heatpump the miners are better. :P )
543 2012-10-19 22:40:29 <sturles> :-)
544 2012-10-19 22:41:15 <sturles> What do you think about BFL's claims of 1 W / Ghash/s?
545 2012-10-19 22:41:57 <sturles> If you think 150 W is impossble?
546 2012-10-19 22:42:39 <gmaxwell> I think if BFL isn't lying completely they're on 45nm.
547 2012-10-19 22:43:20 <gmaxwell> And if they are then their claims are reasonable. But it also means that if it turns out they have a bad mask they'll probably vanish with everyone's money.
548 2012-10-19 22:43:45 <gmaxwell> (Inaba says they are not funded at all by preorders but I .. well. I wouldn't place a large bet against it, but nor do I believe it)
549 2012-10-19 23:13:09 <MC-Eeepc> le reddit integrated bitcoin into their comment system?
550 2012-10-19 23:14:10 <gmaxwell> MC-Eeepc: no, some reddit user is running a bot.
551 2012-10-19 23:14:13 <jgarzik> it's a plugin bot
552 2012-10-19 23:14:44 <MC-Eeepc> reddit has plugins?
553 2012-10-19 23:18:29 <MC-Eeepc> oh everyone has to trust one guy