1 2011-09-08 00:00:00 <trqlala> hmm same here
  2 2011-09-08 00:00:39 <trqlala> its too immediate code for me... perhaps satoshi did model checking, but then he didnt release enough info to gain the trust of cryptographers
  3 2011-09-08 00:00:43 <ymirhotfoot> and I myself have been doing just a tiny bit of studying so I can learn how it works, and yes,
  4 2011-09-08 00:01:04 <ymirhotfoot> a push should be made to aim for stuff near to formal model checking
  5 2011-09-08 00:01:09 <cjdelisle> nah, all of the algos are known and proven
  6 2011-09-08 00:01:25 <ymirhotfoot> sha256 good, I think that
  7 2011-09-08 00:01:27 <cjdelisle> I usually laugh at people when they invent their own crypto algo
  8 2011-09-08 00:01:41 <ymirhotfoot> trqlala meant the protocol/implementation.
  9 2011-09-08 00:01:59 <trqlala> i mean its all talk about secure bitcoin protocol, but I see zero formal verifications (not even attempts, and id be glad to see parts being checked, but nothing at all,... just waiting for the first flaw to show)
 10 2011-09-08 00:02:13 <ymirhotfoot> the basic bits, inner loops, signings, etc.. all bog standard nowadays
 11 2011-09-08 00:02:38 <ymirhotfoot> it is the Great Satoshi Blockchain that is new
 12 2011-09-08 00:02:50 <cjdelisle> It has been looked at by some security professionals and they found a number of "potential issues" but not one of them could actually be exploited.
 13 2011-09-08 00:03:08 <cjdelisle> It's not the best situation but it has gotten review.
 14 2011-09-08 00:03:11 <ymirhotfoot> trqlala, I am not worried in that direction, though more formality there would help,
 15 2011-09-08 00:03:27 <ymirhotfoot> I just want to see it laid out in a model I can grasp.
 16 2011-09-08 00:03:54 <gmaxwell> The usability of formal verification tools is overstated. Recursive function? Oops.. Too bad. can't prove anything about it.  Bignum? oops too bad, can't prove anything about it.  Pointer arithemetic. Oops too bad, can't prove anything about it.
 17 2011-09-08 00:04:04 <ymirhotfoot> Just a few hours ago I was thinking "Well crystallization, yes, the blockchain is perhaps an example.".
 18 2011-09-08 00:04:25 <ymirhotfoot> In general stat mech is important ;)
 19 2011-09-08 00:05:01 <ymirhotfoot> gmaxwell, as I said, inner parts of less interest to me with respect to Bitcoin,
 20 2011-09-08 00:05:02 <trqlala> gmaxwell: formal verification tool *usage* is not an exact science, you have to *model* the software before checking it, dont expect it to prove/disprove security of elliptic curve signing
 21 2011-09-08 00:05:16 <ymirhotfoot> Graet Outer Loop Crystallization
 22 2011-09-08 00:05:22 <gmaxwell> The actual bitcoin algorithim itself is pretty much intutively correct on its face. There isn't much to prove about it, provided the cryptographic functions do what they're supposted to do. :)
 23 2011-09-08 00:05:23 <ymirhotfoot> of great interest
 24 2011-09-08 00:05:59 <ymirhotfoot> What are the constitutive equations of the medium bitcoind creates/live-in?
 25 2011-09-08 00:06:00 <gmaxwell> (see Satoshi's byzantine wifi crackers example)
 26 2011-09-08 00:06:05 <ymirhotfoot> Yes.
 27 2011-09-08 00:06:17 <ymirhotfoot> yes, well known to me.
 28 2011-09-08 00:06:20 <trqlala> gmaxwell: im more worried about the small details like formal verification on choosing another branch cut in hindsight (even the wiki displays situation hell on that topic)
 29 2011-09-08 00:06:56 <ymirhotfoot> I am not clear on meaning of word "cut" here.
 30 2011-09-08 00:06:59 <trqlala> where can I find s's byzantine wifi crackers example?
 31 2011-09-08 00:07:00 <ymirhotfoot> Am newbie
 32 2011-09-08 00:07:28 <gmaxwell> http://www.mail-archive.com/cryptography@metzdowd.com/msg09997.html
 33 2011-09-08 00:07:32 <ymirhotfoot> Look at discussion following satoshi's publoication in late 2008, let me look it up
 34 2011-09-08 00:08:02 <trqlala> i meant, later more POW appears on a side history of blocks that split of earlier
 35 2011-09-08 00:08:43 <ymirhotfoot> http://www.mail-archive.com/cryptography@metzdowd.com/msg09997.html
 36 2011-09-08 00:08:56 <gmaxwell> trqlala: that happens from time to time (block reorgs).
 37 2011-09-08 00:09:26 <ymirhotfoot> ah, gmaxwell is faster than I
 38 2011-09-08 00:09:30 <bx_> lookign for skilled web devs exp with js (ajax), php, mysql, etc PM me ...  Also looking for Mobile devs (android/iphone) pay and equity in large bitcoin products & games, along with Contract-by-contract for small biz clients. PM me, Thanks ! :)
 39 2011-09-08 00:09:35 <trqlala> gmaxwell: yes but the code handling that seems much less "intutively correct" to me
 40 2011-09-08 00:10:06 <trqlala> here I would feel safer with formal checking of block reorgs
 41 2011-09-08 00:10:10 <gmaxwell> trqlala: sure, the _code_ might be buggy (though its been pretty heavily tested), but validating a model doesn't actually validate the code.
 42 2011-09-08 00:10:24 <gmaxwell> And validating the code is probably not possible with existing validation tools.
 43 2011-09-08 00:10:47 <ymirhotfoot> Quote: <gmaxwell> The actual bitcoin algorithim itself is pretty much intutively
 44 2011-09-08 00:10:48 <trqlala> one could validate parts
 45 2011-09-08 00:11:13 <gmaxwell> (because, e.g. the reorg does BDB lookups to find blocks nothing is going to formally validate something as complex as BDB)
 46 2011-09-08 00:11:41 <ymirhotfoot> Yes to quote, but, behind this main truth, there is actual stat mech, a family of stat mech models, whose behavior we might look at.
 47 2011-09-08 00:11:54 <ymirhotfoot> satoshi did some simulations, and
 48 2011-09-08 00:11:57 <trqlala> gmaxwell: thats exactly why a good model checker (the programmer who writes the model) would not use BDB code in the model...
 49 2011-09-08 00:12:08 <ymirhotfoot> I have a high opinion of satoshi's powers.
 50 2011-09-08 00:12:49 <ymirhotfoot> bdb is not to the point I think, gmaxwell, at least my point.
 51 2011-09-08 00:12:54 <gmaxwell> trqlala: we need things like BDB (or its complexity) to have an implementation which scales to the real world. If you make a toy implementation that can be checked but can't be used, what does it matter if its correct but the software everyone uses is wrong?
 52 2011-09-08 00:13:00 <ymirhotfoot> let us ignore bdb
 53 2011-09-08 00:13:01 <trqlala> he would (over)simplify the database stuff, just to find logical flaws in handling reorgs
 54 2011-09-08 00:13:15 <noagendamarket> wonder if bitcoin was satoshis first block chain
 55 2011-09-08 00:13:34 <noagendamarket> If there is another test one he abandoned
 56 2011-09-08 00:13:38 <gmaxwell> noagendamarket: you mean if the current public chain was the first one he created? no obviously not.
 57 2011-09-08 00:13:38 <ymirhotfoot> they are differnt pieces of engineering, with different tools to investigate behavior
 58 2011-09-08 00:13:46 <noagendamarket> yea
 59 2011-09-08 00:13:55 <ymirhotfoot> satsohi said he did simulations.
 60 2011-09-08 00:13:57 <gmaxwell> noagendamarket: the public chain was provably created ~24 hours before the public release.
 61 2011-09-08 00:14:02 <ymirhotfoot> I do not doubt him.
 62 2011-09-08 00:14:04 <trqlala> gmaxwell: *no* verification human or automated will give complete proof of security,... that does not mean we cant model important parts of its behaviour individually
 63 2011-09-08 00:14:08 <gmaxwell> noagendamarket: since he didn't write and test the software in 24 hours& ;)
 64 2011-09-08 00:14:23 <noagendamarket> maybe hes still growing the other one lol
 65 2011-09-08 00:14:42 <noagendamarket> like an alternate universe
 66 2011-09-08 00:14:52 <gmaxwell> trqlala: I don't know what to say I've spent weeks of time with frama-c formally validating code far simpler than what we're talking about and I'm doubtful to the usefulness of the approach on bitcoin.
 67 2011-09-08 00:15:18 <ymirhotfoot> gmaxwell, I am nottalking about that level.
 68 2011-09-08 00:15:27 <trqlala> ymirhotfoot: sure I think he did simulations, he shows results in his paper, but thats very elementary probability simulation, not formal verification of protocols etc
 69 2011-09-08 00:15:30 <gmaxwell> (the reason for the ~24 hour delay is that it probably took him that long to solve the genesis block)
 70 2011-09-08 00:15:40 <ymirhotfoot> rather a family of models, with say 6 parameters, and some simulations
 71 2011-09-08 00:15:58 <ymirhotfoot> stat mech network models
 72 2011-09-08 00:16:03 <gmaxwell> ymirhotfoot: I'm talking to you and trqlala and you're concerned about different things.
 73 2011-09-08 00:16:06 <noagendamarket> http://twitter.com/#!/bitcoinmedia/status/111617120395214849  <-- retweet
 74 2011-09-08 00:16:14 <ymirhotfoot> I am not much interested in bugs in bitcoind.
 75 2011-09-08 00:16:18 <trqlala> gmaxwell: is it correct that currently only fees and possibly ineffectiveness prevent penny attack?
 76 2011-09-08 00:16:31 <gmaxwell> ymirhotfoot: for you I'm poingint out that the basic algorithim is intutively correct, and his paper already did the intresting analysis.
 77 2011-09-08 00:17:16 <gmaxwell> trqlala: 'only' ?  I don't see why you say only, it turns the act of holding bitcoin still into a reusable proof of work system. Other than the usablity warts its a pretty excellent measure.
 78 2011-09-08 00:17:30 <ymirhotfoot> Yes, as I mention, I know Satoshi has both foram estimates in his paper, and he did some simulations, but there is stuff to be done today, as always
 79 2011-09-08 00:17:42 <ymirhotfoot> funding main impediment to human advance
 80 2011-09-08 00:17:57 <ymirhotfoot> foram -> formal
 81 2011-09-08 00:18:20 <gmaxwell> trqlala: basically to make a free txn you must hold 1 BTC still for 24 hours (or a linear equal, e.g. 2 btc for 12 hours).
 82 2011-09-08 00:18:25 <trqlala> gmaxwell: lots of people were drawn to bitcoin because small transactions is a huge gap in transaction market, fees suddenly make it much less interesting
 83 2011-09-08 00:18:48 <gmaxwell> If you don't meet that criteria, or you have outputs smaller than 0.01 you have to pay 0.0005 BTC per transaction which makes attacks expensive fast.
 84 2011-09-08 00:18:50 <phantomcircuit> the fees are tiny
 85 2011-09-08 00:19:09 <trqlala> gmaxwell: so there ARE other methods i.e. the one you just mentioned
 86 2011-09-08 00:19:22 <gmaxwell> trqlala: Hm? I consider it one method.
 87 2011-09-08 00:19:31 <trqlala> are there more?
 88 2011-09-08 00:19:39 <ymirhotfoot> As I said, no disagreement, gmaxwell.  But stability against competitors might one worthwhile thing to look at, "evolutionarily stable strategy"s might be conside red.
 89 2011-09-08 00:21:14 <trqlala> gmaxwell: doesnt it make more sense to require proof of work per transaction (based on last block), this way people with tight budgets can spend soon after receiving
 90 2011-09-08 00:21:30 <ymirhotfoot> question of fees mainly comes in later
 91 2011-09-08 00:21:47 <ymirhotfoot> after we have working subsystems for various things,
 92 2011-09-08 00:21:53 <gmaxwell> trqlala: no because that would divert computational power from the whole system. bitcoins themselves are fixated, reusuabl proof of work.
 93 2011-09-08 00:21:53 <trqlala> itd be easy to verify (sha takes much less time than elliptic curve sigs)
 94 2011-09-08 00:22:14 <ymirhotfoot> working and are convincing to Old World, that is, pre-Bitcoin, merchants,
 95 2011-09-08 00:22:42 <ymirhotfoot> or our stuff is so good, new merchants come in and take away business from non-bitcoin using merchants
 96 2011-09-08 00:23:52 <trqlala> gmaxwell: thats not convincing, each transaction has elliptic curve sigs, the extra POW validation would cost nothing (a single combined hash), only the sender would have to deliver POW
 97 2011-09-08 00:24:11 <gmaxwell> trqlala: Not the validation, the computation.
 98 2011-09-08 00:24:38 <gmaxwell> (the sender)
 99 2011-09-08 00:24:48 <trqlala> which computation? the proof of work? thats on the burden of the sender, how does that divert computational power from the whole system?
100 2011-09-08 00:24:59 <gmaxwell> And people who are not attacking should already have aged bitcoins. Attacking itself diminishes your supply of aged bitcoins, but doesn't diminish your cpu power.
101 2011-09-08 00:25:14 <trqlala> and it would be much lower than the POW for block generation of course
102 2011-09-08 00:25:17 <noagendamarket> http://www.theatlanticwire.com/business/2011/09/paul-krugman-incites-bitcoin-cyber-geek-infighting/42188/
103 2011-09-08 00:25:24 <Diablo-D3> who?
104 2011-09-08 00:25:31 <Diablo-D3> I wish the media would quit making people up
105 2011-09-08 00:25:41 <Diablo-D3> theres no such person
106 2011-09-08 00:25:45 <gmaxwell> trqlala: if its sufficiently lower then it's completely ineffective as a security measure.
107 2011-09-08 00:25:57 <ymirhotfoot> When will the Scientific American article appear?
108 2011-09-08 00:26:46 <gmaxwell> trqlala: look at it this way: It's _already_ a ontime pow system: you can do the computation to mine the bitcoin that you use for fees and/or standing balance.
109 2011-09-08 00:27:25 <gmaxwell> trqlala: if you use a totally different pow its either much cheaper, in which case not as secure, or its diverting power from being used to create the security of the overall blockchain.
110 2011-09-08 00:28:00 <trqlala> sufficiently lower would still be a very effective security measure, youd need MANY transaction POWs to cripple the network compared to the steady x times as fast as block generation
111 2011-09-08 00:28:53 <gmaxwell> trqlala: we can make the anti-dos costs as low as we like by lowering the priority required and/or reducing the minfee.
112 2011-09-08 00:29:08 <ymirhotfoot> I see some parameters that should be defined in the implicit models under discussion bygmaxwell and trqlala:
113 2011-09-08 00:29:09 <trqlala> and the same for POW sending
114 2011-09-08 00:29:17 <gmaxwell> (though recent expirence has show that it shouldn't be any lower than it currently is, I think)
115 2011-09-08 00:29:29 <ymirhotfoot> what is the statistical shape of the two networks under discussion?
116 2011-09-08 00:29:46 <gmaxwell> trqlala: sure, but using bitcoin as the mechenism for pow here has two important properties:
117 2011-09-08 00:29:46 <ymirhotfoot> how strong are the attackers in the two cases?
118 2011-09-08 00:29:48 <ymirhotfoot> etc...
119 2011-09-08 00:30:16 <gmaxwell> (1) you can let people use their standing balances to avoid paying fees (attackers don't have standing balances, by virtue of the attack, so this differentiates many normal users from attackers in a lower cost way)
120 2011-09-08 00:30:30 <ymirhotfoot> Heree Leibniz's old dream has come partly true:
121 2011-09-08 00:30:32 <gmaxwell> (2) every bit of cpu time spend on the blockchain increases its security.
122 2011-09-08 00:30:45 <ymirhotfoot> make up your two competeing modesl, and
123 2011-09-08 00:30:50 <ymirhotfoot> run the simulations
124 2011-09-08 00:31:01 <trqlala> it seems much harder to me too cripple the network through creating MANY POW transactions compared to hijacking block generation, even for modest txn POW...
125 2011-09-08 00:31:25 <ymirhotfoot> Numbers, trqlala, then
126 2011-09-08 00:31:32 <ymirhotfoot> Numbers, gmaxwell
127 2011-09-08 00:31:49 <gmaxwell> trqlala: err.. I don't see why you think that.
128 2011-09-08 00:32:10 <ymirhotfoot> we'll teach gribble to run some various styles of network evolution models.
129 2011-09-08 00:32:58 <gmaxwell> (nor do I see how that has has anything to do with the form the pow takes)
130 2011-09-08 00:32:58 <trqlala> gmaxwell: the fundament of our discussion revolves around how many attacker penny txns / block would cripple the network
131 2011-09-08 00:33:28 <gmaxwell> trqlala: oh, well for one "cripple the network" is probably the wrong threat at least right now.
132 2011-09-08 00:33:56 <trqlala> so the fees/sleeptime may be lowered? :)
133 2011-09-08 00:34:01 <gmaxwell> The bigger risk now is premature blockchain bloat making it too costly to run full validating nodes before bitcoin has matured enough to fund its operation at full scale.
134 2011-09-08 00:34:20 <ymirhotfoot> One could teach a good one year course in stat mech using Peter Whittle's book networks in stochastic equilibrium, and
135 2011-09-08 00:34:35 <ymirhotfoot> with Bitcoin as an example
136 2011-09-08 00:34:42 <ymirhotfoot> Bitcoin and Bitcoin like things
137 2011-09-08 00:35:45 <trqlala> gmaxwell: I still dont see how you can claim that having POW sending would reduce security of the network, most senders of money are not miners anyway
138 2011-09-08 00:35:59 <gmaxwell> The second biggest risk now is DOS against the maximum block size. E.g. filling the memory pool with junk txn so that no ones free transactions go through in reasonable time, making bitcoin painful to use for many people.  This would only take a sustained rate of a couple transactions per second.
139 2011-09-08 00:36:32 <trqlala> and algorithmic complexity dictates that the crypto signing with elliptic curves scales much faster than a constant time hash function
140 2011-09-08 00:36:33 <gmaxwell> trqlala: it doesn't matter _who_ the miner is.
141 2011-09-08 00:36:38 <ymirhotfoot> am newbie, but I thought gmaxwell's attack, or something near, seemede maybe possible.
142 2011-09-08 00:37:07 <ymirhotfoot> What does signing have to do with Bitcoin Classic here?
143 2011-09-08 00:37:08 <gmaxwell> trqlala: The reason it doesn't matter is that when btc is the POW it increases demand for BTC which increases mining.
144 2011-09-08 00:37:11 <Diablo-D3> no, gmaxwell still hasnt discoverd a useful attack
145 2011-09-08 00:37:26 <Diablo-D3> everything still requires using THE ATTACK, which is basically nigh impossible now.
146 2011-09-08 00:37:27 <trqlala> gmaxwell: if the sender is not a miner the cpu cycles wasted on POW does not influence the blockchain security
147 2011-09-08 00:37:31 <ymirhotfoot> Diablo-D3, I am glad to hear this.
148 2011-09-08 00:37:55 <ymirhotfoot> Is there a three line, suitable for ignorant newbie, demostartion of why the attack fails?
149 2011-09-08 00:37:58 <Diablo-D3> THE ATTACK, n., using more than 50% hashing power than the rest of the network, and the network is at 14 thash/sec
150 2011-09-08 00:38:06 <Diablo-D3> or several thousand 5850s.
151 2011-09-08 00:38:07 <jgarzik> gmaxwell: "filling the memory pool"   <<--  THAT is the attack, actually
152 2011-09-08 00:38:11 <gmaxwell> trqlala: yes it does, because they're not buying those bitcoins that would be used in fees from miners, they're not giving those bitcoins back to miners in the form of fees, and they're not increasing demand for bitcoin by keeping standing balances.
153 2011-09-08 00:38:20 <jgarzik> TX cache is unbounded at present
154 2011-09-08 00:38:23 <trqlala> gmaxwell: so now btc as POW is not about security at all but a bitcoin economic decision, trying to increase BTC demand?
155 2011-09-08 00:38:25 <ymirhotfoot> Well, that attack, yes, but .... gmaxwell, step in if you want
156 2011-09-08 00:38:39 <gmaxwell> trqlala: ::cries::
157 2011-09-08 00:39:00 <ymirhotfoot> jgarzik, so something is less dfended than it should be, or other way round?
158 2011-09-08 00:39:03 <luke-jr_> oh
159 2011-09-08 00:39:07 <luke-jr_> I see what jgarzik is saying
160 2011-09-08 00:39:32 <luke-jr_> POW-sending effectively eliminates the only guaranteed* Bitcoin market: transaction fees
161 2011-09-08 00:39:37 <gmaxwell> trqlala: I'm only pointing out that with BTC as the (reusuable)POW unit it ultimately means that whatever N units of computation are spent on POW are ultimately put into blockchain security.
162 2011-09-08 00:39:37 <trqlala> also, if it was a computational POW, then a rich attacker has no advantage over a poor attacker
163 2011-09-08 00:39:47 <gmaxwell> Even if the user is not a miner.
164 2011-09-08 00:40:15 <gmaxwell> trqlala: haha. Sure they do, they buy computing time to attack (perhaps from establsihed miners). Or steal it.
165 2011-09-08 00:40:22 <ymirhotfoot> I am sure I do not understand what trqlala is proposing.
166 2011-09-08 00:40:36 <trqlala> gmaxwell: you mean that fees are there to support mining?
167 2011-09-08 00:40:58 <gmaxwell> trqlala: No, they currently prevent DOS attacks.
168 2011-09-08 00:41:07 <trqlala> thats what I completely agree with, but preventing penny attack should be done cryptographically not financially
169 2011-09-08 00:41:15 <jgarzik> TX cache and CPU DoS are the current near-term worries
170 2011-09-08 00:41:16 <gmaxwell> trqlala: it's fungable.
171 2011-09-08 00:41:30 <ymirhotfoot> trqlala, ah the subtlety of most working financial systems,
172 2011-09-08 00:41:41 <ymirhotfoot> some peieces are logically not like other peieces.
173 2011-09-08 00:41:49 <gmaxwell> trqlala: To prevnt the attack N units of CPU must be spent, N depends on the wealth of the attackers (and the value they gain on attacking)
174 2011-09-08 00:41:51 <ymirhotfoot> Satoshi is a subtle being.
175 2011-09-08 00:41:56 <trqlala> wait a minute: i am NOT proposing do away with fees, I am proposing not to prevent penny attack with fees
176 2011-09-08 00:42:26 <gmaxwell> trqlala: if N takes the form of bitcoin, then those cycles make bitcoin more secure overall by encouraging more mining.  If N is something else, it works the same for POW purposes, but doesn't have the side benefit.
177 2011-09-08 00:43:45 <noagendamarket> fees will support miners after the reward runs out. How can you do away with them ?
178 2011-09-08 00:43:56 <gmaxwell> This is true no matter what N is.  If you make N executations of SHA1() then I can _still_ pay miners to stop mining and to instead compute attack POWs for me.  Either way the attack costs X bitcoin at the end of the day, X is determined by how big the users and developers have agreed to set N to.
179 2011-09-08 00:44:04 <trqlala> gmaxwell: so you claim that impulse response of an attack on the system through financial motivation is so quick that they will buy and install new hardware fast enough to prevent the attack?
180 2011-09-08 00:44:32 <ymirhotfoot> wait I see some words I understand!
181 2011-09-08 00:44:37 <gmaxwell> trqlala: we're using our installed base of hardware already... the bitcoin infrastructure.
182 2011-09-08 00:44:46 <trqlala> noagendamarket: none of us is proposing to do away with fees to reward transaction processing
183 2011-09-08 00:45:21 <gmaxwell> noagendamarket: trqlala thinks that instead of using coin aging and fees to prevent penny DOS we should instead use bitcoin-orthorgonal hashcash.
184 2011-09-08 00:45:27 <osmosis> Do transactions have a timestamp associated with them in the blockchain? Or is it just new blocks that are timestamped?
185 2011-09-08 00:45:27 <ymirhotfoot> trqlala, the error you refer is foundational o the "free market" "deregulate" project, which has led to many bad things.
186 2011-09-08 00:45:37 <trqlala> ymirhotfoot: your obvious physicist/engineer/mathematician, I guess the former since you referenced statistical mechanics
187 2011-09-08 00:45:40 <ymirhotfoot> Often economists do not include reaction times.
188 2011-09-08 00:46:07 <ymirhotfoot> Some day counterbalancing forces will come and stop damage.
189 2011-09-08 00:46:11 <gmaxwell> I'm pointing out that the two are fundimentally equal except using bitcoin as the proof of work has a collateral benefit of increasing the security of bitcoin overall.
190 2011-09-08 00:46:11 <trqlala> gmaxwell how is it bitcoin orthogonal?
191 2011-09-08 00:46:29 <gmaxwell> trqlala: not part of the bitcoin distributed algorithim (not mining).
192 2011-09-08 00:47:35 <ymirhotfoot> trqlala, these days I mainly do a little Perl
193 2011-09-08 00:47:40 <ymirhotfoot> in Scheme.
194 2011-09-08 00:47:46 <gmaxwell> osmosis: blocks themselves.
195 2011-09-08 00:47:57 <trqlala> gmaxwell: taxing small highturnover transactions in the valid economy in general will also promote mining and hence security, my problem is that the method is indiscriminate and just tax
196 2011-09-08 00:47:58 <noagendamarket> well no one uses a cpu for hashing anymore...you could do that somehow
197 2011-09-08 00:48:07 <gmaxwell> osmosis: in essense the blockchain is what gives consensus times to the transactions.
198 2011-09-08 00:48:16 <osmosis> gmaxwell, got it
199 2011-09-08 00:48:35 <noagendamarket> like hashcash before you can send a tx
200 2011-09-08 00:48:38 <noagendamarket> ?
201 2011-09-08 00:48:44 <IO-> is BlueMatt ever around anymore?
202 2011-09-08 00:48:53 <gmaxwell> trqlala: it's not indiscriminate because it gives aged inputs a free pass which strikes _directly_ to a fundimental difference between most users and all attackers.
203 2011-09-08 00:49:20 <trqlala> i obviously mean indiscriminate with valid vs attack transactions
204 2011-09-08 00:49:30 <gmaxwell> trqlala: and there is no real escape otherwise. Requiring N units of computation is exactly equal to requring X bitcoins where X is the amount you could have mined instead if you were mining.
205 2011-09-08 00:50:00 <ymirhotfoot> BlueMatt was here yeasterday, or maybe in #bitcoin.  I spoke with him briefly.
206 2011-09-08 00:50:13 <IO-> thanks
207 2011-09-08 00:50:26 <trqlala> gmaxwell that too is indiscriminate: a valid user does not have a mining rig and just wants to sends money around, hes not going to wait ages until he mines a block just so he can pay his bitcoin taxes
208 2011-09-08 00:50:40 <gmaxwell> trqlala: it's not though. Most regular usage doesn't have super high turnaround and a zero balance. All attackers do.  A _few_ non attack applications do, unfortunately, but then they get the fees which is completely isomorphic to a local pow except for the collateral benefit.
209 2011-09-08 00:50:46 <gmaxwell> trqlala: almost no one mines solo.
210 2011-09-08 00:51:02 <gmaxwell> trqlala: you can mine and recieve bitcoins on any computer.
211 2011-09-08 00:51:03 <ymirhotfoot> THE SOLO MINER
212 2011-09-08 00:51:11 <ymirhotfoot> COMING TO SYFY
213 2011-09-08 00:51:15 <ymirhotfoot> THIS WINTER
214 2011-09-08 00:51:20 <trqlala> if you look at the real world, people live with tight budgets, the large fraction of users just comes by with his income
215 2011-09-08 00:52:08 <gmaxwell> trqlala: I think you might be overestimating how much aging is required but regardless ignore that.  Because a straight POW can not improve that in the slighest.
216 2011-09-08 00:52:33 <ymirhotfoot> trqlala, I know I do not grasp your proposal, but, what does poverty of a person using/mining bitcoin have to do with it.
217 2011-09-08 00:52:36 <gmaxwell> People always have the option of just paying a fee to have a txn. Paying a pow would be no imrpovement.
218 2011-09-08 00:52:41 <trqlala> why is it so hard for you to imagine that valid internet transactions will typically be small value and high turnover
219 2011-09-08 00:52:43 <gmaxwell> er improvement.
220 2011-09-08 00:53:07 <trqlala> ok whatever
221 2011-09-08 00:53:41 <ymirhotfoot> ad micropayments: most electric companies and individuals arrange their business by means of a systm of micropayments.
222 2011-09-08 00:54:02 <ymirhotfoot> I have never understood why folks often used to say "micropayments will never work".
223 2011-09-08 00:54:05 <ymirhotfoot> they do.
224 2011-09-08 00:54:43 <gmaxwell> trqlala: If the average txn is is .1 BTC and you spend 1 BTC/day, then you need to have a balance of 10x your daily spending in order to avoid paying fees on average. And if you do pay, it's about a half cent in cost.
225 2011-09-08 00:54:52 <gmaxwell> I don't think thats especially burdensome.
226 2011-09-08 00:55:03 <gmaxwell> But regardless, pow still can't improve it at all.
227 2011-09-08 00:56:08 <gmaxwell> luke-jr_: I've got a bussines idea for you.
228 2011-09-08 00:56:14 <trqlala> how about this: what if ever the sleep time of coins or the fees prove to be too small, ... who will change the value in time? after the attack?
229 2011-09-08 00:56:34 <luke-jr_> gmaxwell: ?
230 2011-09-08 00:56:47 <gmaxwell> luke-jr_: setup your pool so that modified bitcoin clients can connect and mine for free. But in doing to they accrew a balance with you, which you'll use to pay transaction fees for txn they send you directly.
231 2011-09-08 00:57:49 <gmaxwell> (well there is no 'pay' you mine those txn for free, because they came from a use that mined for you)
232 2011-09-08 00:57:57 <ymirhotfoot> gmaxwell, have you been converted to trqlala's view?
233 2011-09-08 00:58:06 <gmaxwell> ymirhotfoot: No.
234 2011-09-08 00:58:16 <ymirhotfoot> not even a little bit?
235 2011-09-08 00:58:29 <ymirhotfoot> OK
236 2011-09-08 00:58:41 <gmaxwell> Oh well, I wasn't in total disagreement with him to begin with.
237 2011-09-08 00:58:55 <ymirhotfoot> ;)
238 2011-09-08 00:59:07 <gmaxwell> The point he made that right now mining isn't equally distributed enough is a real issue, and a much bigger one than this subject.
239 2011-09-08 00:59:17 <ymirhotfoot> I do not understand this issue/issues very well,
240 2011-09-08 00:59:24 <ymirhotfoot> ah your last just came in.
241 2011-09-08 00:59:41 <gmaxwell> the quantization of solo mining means that "stop whining about fees, just mine a little and its quickly paid for"  isn't currently a good answer.
242 2011-09-08 00:59:43 <ymirhotfoot> I nowthink I have a dim hazy view of this nexus of issiues.
243 2011-09-08 00:59:55 <ymirhotfoot> thanks, gmaxwell and trqlala!
244 2011-09-08 01:00:50 <gmaxwell> Mining on random desktop cpus would _easily_ pay a random users complete transaction fee burden and then some, even assuming they were living hand to mouth.
245 2011-09-08 01:01:17 <gmaxwell> But there is an interesting question about how to get them mining, and how to do so without creating more tiny transactions.
246 2011-09-08 01:01:29 <ymirhotfoot> gmaxwell, last of interest to me.  OK
247 2011-09-08 01:02:06 <gmaxwell> I pointed out that its economically equal if they mine or someone else does,  but a lot of people would probably feel better about doing it themselves rather than paying (probably because their CPU is already there and idle)
248 2011-09-08 01:02:23 <ymirhotfoot> I must be away from the sctreen for a bit, but I will let my irc agaent run.
249 2011-09-08 01:02:29 <gmaxwell> so thus my suggestion to luke (who is a rather creative mining pool operator)
250 2011-09-08 01:06:32 <ymirhotfoot> just one thing, there are markets where if you have many seats on the exchange, the chance of winning soon is greater,
251 2011-09-08 01:07:19 <ymirhotfoot> but no matter how many your expectation is the same, see recent Massachusetts state games
252 2011-09-08 01:07:42 <ymirhotfoot> perhaps gmaxwell's suggested protocol is a like game?
253 2011-09-08 01:08:27 <ymirhotfoot> ah, no, OK, you assign bitcoins mined by hashes done, so no.
254 2011-09-08 01:10:44 <CIA-92> bitcoin: Con Kolivas * r9b5cd618868a cgminer/ (README adl.c main.c miner.h): Implement accepting a range of engine speeds as well to allow a lower limit to be specified on the command line.
255 2011-09-08 01:13:59 <luke-jr_> gmaxwell: miners can submit a transaction id with their difficulty-256 shares to get whitelisted without a txn fee
256 2011-09-08 01:21:24 <gmaxwell> luke-jr_: ohh. :) down side of that is that it doesn't let you store value.  e.g. I can't accrue value with you and then make a txn without waiting.
257 2011-09-08 01:21:40 <luke-jr_> true
258 2011-09-08 01:21:55 <luke-jr_> the bigger problem IMO is getting a bitcoin client to send them w/o a fee
259 2011-09-08 01:21:56 <luke-jr_> ;)
260 2011-09-08 01:26:48 <gmaxwell> if bitcoin ever gets some kind of plugin system, istm this would be a pretty obvious plugin. Add some plugin and you'll do some CPU mining to pay your TXN fees.
261 2011-09-08 01:30:42 <CIA-92> bitcoin: Con Kolivas * rb50041c88089 cgminer/ (README adl.c main.c): Allow per-device fan ranges to be set and use them in auto-fan mode.
262 2011-09-08 01:55:41 <CIA-92> bitcoin: Con Kolivas * rd6f3bd056441 cgminer/adl.c: Display which GPU has overheated in warning message.
263 2011-09-08 01:55:43 <CIA-92> bitcoin: Con Kolivas * re0a9f1aae387 cgminer/ (adl.c adl.h main.c): Allow temperature targets to be set on a per-card basis on the command line.
264 2011-09-08 02:25:39 <CIA-92> bitcoin: Con Kolivas * r355b24cef35b cgminer/README: Document the temperature command line changes.
265 2011-09-08 02:25:41 <CIA-92> bitcoin: Con Kolivas * r0e1c2916242d cgminer/adl.c: Display fan range in autofan status.
266 2011-09-08 02:25:43 <CIA-92> bitcoin: Con Kolivas * rf1cf79ae9d31 cgminer/main.c: With many cards, the GPU summaries can be quite long so use a terse output line when showing them all.
267 2011-09-08 02:29:52 <bx_> lookign for skilled web devs exp with js (ajax), php, mysql, etc PM me ...  Also looking for Mobile devs (android/iphone) pay and equity in large bitcoin products & games, along with Contract-by-contract for small biz clients. PM me, Thanks ! :)
268 2011-09-08 02:35:40 <CIA-92> bitcoin: Con Kolivas * r4b43efceca03 cgminer/main.c: Use a terser device status line to show fan RPM as well when available.
269 2011-09-08 02:39:09 <osmosis> How does this apply to bitcoin?  http://en.wikipedia.org/wiki/Export_of_cryptography_in_the_United_States#Current_status
270 2011-09-08 02:40:30 <gmaxwell> osmosis: bitcoin (at least current released versions) do not use any encryption at all.
271 2011-09-08 02:41:10 <gmaxwell> It uses authentication, but export restrictions generally explicitly exclude cryptographic authentication.
272 2011-09-08 02:48:56 <cjdelisle> and ofc nobody cares about exporting crypto nor has for a long time
273 2011-09-08 02:52:09 <gmaxwell> cjdelisle: ITAR is still enforced against commercial entities.
274 2011-09-08 02:53:40 <cjdelisle> yea but not against open source hippies
275 2011-09-08 03:02:34 <bonsaikitten> that's why you do like OpenBSD and develop the crypto bits in civilized countries
276 2011-09-08 03:02:40 <bonsaikitten> import is not an issue ...
277 2011-09-08 03:05:11 <Eliel> cjdelisle: well, it's a bit obviously useless to try to enforce it against open source. Easy enough to route around it.
278 2011-09-08 03:07:29 <cjdelisle> yea and it just makes no sense anymore, when does a hash function become a cryptographic function? when you add the XOR to it?
279 2011-09-08 03:07:47 <cjdelisle> just old sillyness
280 2011-09-08 04:09:24 <bx_> lookign for skilled web devs exp with js (ajax), php, mysql, etc PM me ...  Also looking for Mobile devs (android/iphone) pay and equity in large bitcoin products & games, along with Contract-by-contract for small biz clients. PM me, Thanks ! :)
281 2011-09-08 04:13:58 <gmaxwell> cjdelisle: meh, it's not silly.
282 2011-09-08 04:14:35 <gmaxwell> cjdelisle: Policy goal: Keep random people defending their country from a US invasion from getting cheap encrypted radios that the US can't monitor.
283 2011-09-08 04:14:52 <gmaxwell> Policy solution: regulate large scale commercial transactions of cryptographic devices.
284 2011-09-08 04:15:17 <gmaxwell> Does it matter that _in theory_ these people in a warzone could somehow managed to macguyver their own radios with open source and duct tape? No.
285 2011-09-08 04:15:45 <gmaxwell> Does it matter that they could sneak a couple radios in from other countries? No, not really.
286 2011-09-08 04:16:01 <gmaxwell> Policy doesn't have to be completely effective to be effective enough.
287 2011-09-08 04:16:52 <gmaxwell> Does DRM stop you from copying a game? Probably not. Does it stop my 67 year old mother (an avid WoW player), absolutely even the smallest friction is enough to drive her to buy rather than download.
288 2011-09-08 04:18:10 <ymirhotfoot> gmaxwell, it is your tone of sweet reasonableness that, if the other side is good, will cause them to come to your house early in the roundup.
289 2011-09-08 04:18:20 <bonsaikitten> gmaxwell: but if any one person manages to circumvent it it's liberated for everyone
290 2011-09-08 04:18:43 <bonsaikitten> DRM doesn't work, that's the funny bit
291 2011-09-08 04:19:23 <gmaxwell> bonsaikitten: it's not because the crack shows up on random sites, some of which are of ill-repute and ship malware too. My mom can't tell the sites apart... figuring out what are good trustworthy sources and what aren't is not a costless operation, and because its unlawful it must remain underground.
292 2011-09-08 04:19:57 <ymirhotfoot> But DRM, indeed Palladium, does work on Apples hand helds, for most of these devices out there.
293 2011-09-08 04:20:13 <ymirhotfoot> Nothing gets on those devices except what Apple allows.
294 2011-09-08 04:20:25 <ymirhotfoot> Apple, after has root on the devices.
295 2011-09-08 04:20:36 <bonsaikitten> gmaxwell: and it shows up on chinese and russian and indian markets where you can buy the DVD for ~2$
296 2011-09-08 04:20:36 <ymirhotfoot> And root hath its privileges.
297 2011-09-08 04:21:43 <gmaxwell> bonsaikitten: 99% of the buyers in those markets would never have bought at the whole price. In the case of software the illicit copying is often an asset: you avoid adding the people who wouldn't buy to the userbase of the (OSS) competition.
298 2011-09-08 04:22:26 <bonsaikitten> gmaxwell: since prices have gone up over the last years I wouldn't buy games anymore
299 2011-09-08 04:22:47 <bonsaikitten> luckily I also lack the time, so it's a non-issue anyway
300 2011-09-08 04:31:36 <ymirhotfoot> Conference in New York City 10-12 September 2011:
301 2011-09-08 04:31:40 <ymirhotfoot> http://openvideoconference.org/alternative-currencies-and-transaction-models/
302 2011-09-08 04:31:59 <ymirhotfoot> Bitcoin will be at the center of the discussion.
303 2011-09-08 04:32:59 <gmaxwell> Oh, I'll be at OVC.
304 2011-09-08 04:33:59 <ymirhotfoot> It would be good to have authors of Bitcoin code at this meeting, also businessfolk, miners, bankers, and Bitcoin Tribes Folk without further specification.
305 2011-09-08 04:34:04 <ymirhotfoot> Come on down!
306 2011-09-08 04:34:25 <ymirhotfoot> Rick Moen^W^WBrewster Kahle will be there.
307 2011-09-08 04:34:53 <gmaxwell> Does this mean archive.org will take bitcoin donations soon? ;)
308 2011-09-08 04:35:08 <ymirhotfoot> ah, I know not
309 2011-09-08 04:35:17 <ymirhotfoot> I am just passing the notice along
310 2011-09-08 04:35:32 <ymirhotfoot> I have nothing to do with the organization of this conference
311 2011-09-08 04:36:38 <ymirhotfoot> gmaxwell, I am glad you will come, be'ezrat ha'Shem
312 2011-09-08 04:55:02 <bx_> lookign for skilled web devs exp with js (ajax), php, mysql, etc PM me ...  Also looking for Mobile devs (android/iphone) pay and equity in large bitcoin products & games, along with Contract-by-contract for small biz clients. PM me, Thanks ! :)
313 2011-09-08 05:53:20 <EskimoBob> Hi, has anyone used bitcoin-central code to build a actual working *coin exchange?
314 2011-09-08 06:29:05 <genjix> how can i contact theymos? his theymos@hotmail.com email no longer exists...
315 2011-09-08 06:32:31 <Graet> pm on the forums?
316 2011-09-08 06:32:42 <Graet> he only ircs occasionally
317 2011-09-08 06:36:08 <osmosis> "Error: This transaction requires a transaction fee of at least 0.001 because of its amount, complexity, or use of recently received funds "
318 2011-09-08 06:36:22 <osmosis> the error description doesnt really say what is happening
319 2011-09-08 06:36:24 <genjix> thanks Graet ... :/
320 2011-09-08 06:36:26 <genjix> who wants contributor access to be able to make new posts on bitcoinmedia.com?
321 2011-09-08 06:36:53 <genjix> a platform for your writings :)
322 2011-09-08 06:37:25 <osmosis> Whats actually happening is there is not enough money left in my wallet for the fee.
323 2011-09-08 06:41:07 <Graet> no worries genjix - wish i was a writer ;)
324 2011-09-08 06:53:13 <noagendamarket> Graet you can do a press release for the aussie bitcoin conference :)
325 2011-09-08 06:53:43 <noagendamarket> bitcoinpr <-----
326 2011-09-08 06:53:51 <Graet> noagendamarket sure - once i get a press secratary
327 2011-09-08 06:53:56 <noagendamarket> haha
328 2011-09-08 06:54:01 <Graet> or however u spell iy
329 2011-09-08 06:54:05 <Graet> it
330 2011-09-08 06:54:13 <Graet> see why :/ ?
331 2011-09-08 06:54:16 <Graet> lol
332 2011-09-08 06:54:20 <noagendamarket> lol
333 2011-09-08 06:54:32 <noagendamarket> Lucky Im a good spellchecker :)-
334 2011-09-08 06:54:54 <noagendamarket> ha
335 2011-09-08 06:55:11 <noagendamarket> I ususally end up ranting about the government
336 2011-09-08 06:55:36 <Graet> thats ok, someone has to... i cbf ;)
337 2011-09-08 06:55:41 <noagendamarket> lol
338 2011-09-08 06:55:44 <Graet> :P
339 2011-09-08 07:01:34 <genjix> noagendamarket: why your smileys have a beard?
340 2011-09-08 07:02:36 <noagendamarket> lawl
341 2011-09-08 07:02:49 <noagendamarket> its a tongue :p
342 2011-09-08 07:03:14 <noagendamarket> :)-
343 2011-09-08 07:03:21 <noagendamarket> acgtually its abeard
344 2011-09-08 07:03:23 <noagendamarket> lol
345 2011-09-08 07:03:29 <genjix> or a skinny neck
346 2011-09-08 07:03:37 <noagendamarket> a hitler moustache
347 2011-09-08 07:03:44 <noagendamarket> haha
348 2011-09-08 07:04:04 <noagendamarket> :-)
349 2011-09-08 07:04:51 <genjix> that's:
350 2011-09-08 07:04:55 <genjix> :^-)
351 2011-09-08 08:23:00 <EskimoBob> Hi, has anyone used bitcoin-central code to build a actual working *coin exchange?
352 2011-09-08 08:24:32 <osmosis> That was interesting. The transaction went through twice...debited my wallet twice. I only hit send one time though.
353 2011-09-08 08:25:15 <osmosis> It happened just after saying no to a previous transaction that wanted a fee.
354 2011-09-08 08:26:19 <lfm> osmosis: you sure the previos one with the fee didnt go thru?
355 2011-09-08 08:27:04 <osmosis> lfm, the previous one with the fee would have been a totally different amount.  So no, it wasnt the previous one. For some reason, it sent twice even though I only hit send once.
356 2011-09-08 08:30:22 <gmaxwell> Seems unlikely to me there is nothing in the backend code that could do that. (a truly duplicate txn would have the same inputs and end up being the same txid, and wouldn't really be a duplicate0
357 2011-09-08 08:31:09 <UukGoblin> sometimes a mouse sends two clicks? :-P
358 2011-09-08 08:32:11 <Rozz> everything is possible in winapi
359 2011-09-08 08:32:14 <lfm> UukGoblin: switches like buttons can bounce and if the mouse firmware is crappy it oculd be debounceing poorly.
360 2011-09-08 08:32:24 <UukGoblin> yeah
361 2011-09-08 08:33:16 <lfm> But the actual send button on the gui will not take two mouse clicks in a row and send two txn
362 2011-09-08 08:33:18 <osmosis> Rozz, ubuntu
363 2011-09-08 08:33:25 <osmosis> UukGoblin, sure, could have been two clicks.
364 2011-09-08 08:33:47 <UukGoblin> lfm, ok, don't know that, I don't use the GUI
365 2011-09-08 08:33:49 <osmosis> Rozz, actually....a ubuntu virtualbox...running in windows. so it could have been windows still.
366 2011-09-08 08:34:15 <osmosis> or even virtualbox for that matter
367 2011-09-08 08:34:17 <lfm> if you try to double click the send button it doesnt send two txn
368 2011-09-08 08:34:26 <osmosis> that makes sense then
369 2011-09-08 08:34:44 <osmosis> so nothing to do with the previous canceled transaction that wanted a fee.
370 2011-09-08 08:35:51 <osmosis> should there be some protection from double clicking the send button maybe?
371 2011-09-08 08:36:09 <lfm> I think the most likely answer is operator error
372 2011-09-08 08:36:32 <lfm> osmosis: you can't double click the send button now
373 2011-09-08 08:36:51 <osmosis> lfm, I cant?
374 2011-09-08 08:37:03 <lfm> osmosis: just try it! send to yourself
375 2011-09-08 08:40:22 <osmosis> lfm: I clicked send, then looked in the other window to watch the transaction come in..takes about 3 seconds. I watched, the transaction came in twice. I dont see how I could have possibly operator errored that.
376 2011-09-08 08:41:04 <osmosis> confused, i looked back from my sending client, and there the transaction was listed twice.
377 2011-09-08 08:41:27 <osmosis> anyways, im i didnt lose anything.
378 2011-09-08 08:41:55 <osmosis> it is 3:40am here, but thats my story.
379 2011-09-08 08:42:31 <lfm> osmosis well unless you can show how it could happen, it doesnt seem like any sort of network error or anything could cause it.
380 2011-09-08 08:43:07 <osmosis> lfm, thats why i thought maybe it was related to having said no to a previous transaction with a fee.
381 2011-09-08 08:43:59 <UukGoblin> osmosis, saying 'no' will still send the transaction, just without a fee, I guess
382 2011-09-08 08:44:18 <UukGoblin> that's probably not the problem you're seeing though
383 2011-09-08 08:44:49 <osmosis> UukGoblin, no it doesnt send. Thats what its asking, if I want to send and pay the fee.
384 2011-09-08 08:45:02 <UukGoblin> hmm
385 2011-09-08 08:45:10 <lfm> osmosis: are you using a standard version? regular version downloaded from bitcoin.org?
386 2011-09-08 08:45:13 <UukGoblin> ok I don't use the gui so not really sure
387 2011-09-08 08:45:18 <osmosis> lfm, correct
388 2011-09-08 08:45:58 <osmosis> it doesnt take you back to the main window when you say no though. It just goes back to the sending dialog. So I changed the amount of the transaction to see if it would go without a fee.
389 2011-09-08 08:46:22 <osmosis> im trying a few more times, see if i can duplicate at all.
390 2011-09-08 08:47:06 <osmosis> yep, it happened again. I can duplicate it.
391 2011-09-08 08:47:26 <gmaxwell> dork. You're sending to your local wallet?
392 2011-09-08 08:47:32 <gmaxwell> You're seeing it both out and in.
393 2011-09-08 08:47:35 <osmosis> gmaxwell, nope...two machines.
394 2011-09-08 08:48:05 <osmosis> ugh...false. not duplicated
395 2011-09-08 08:48:07 <gmaxwell> So what exact version of bitcoin, on what operating system, and what sequence of operations?
396 2011-09-08 08:48:39 <gmaxwell> From what you're saying you're not doing something that a zillion other bitcoin users don't do every day.
397 2011-09-08 08:51:14 <osmosis> 0.3.24-beta  ubuntu virtualbox, running on win7.
398 2011-09-08 08:51:53 <osmosis> Im going to send coins, but when it asks if I want to pay a fee I say no and then modify the amount and try to resend.
399 2011-09-08 08:52:03 <osmosis> Ill let you know if i can duplicate it or if it happens again.
400 2011-09-08 09:22:19 <lfm> @cad
401 2011-09-08 09:39:57 <lfm> http://www.isi.edu/gost/info/NetCheque/
402 2011-09-08 09:45:44 <CIA-92> bitcoin: Con Kolivas * r6c8341f1331f cgminer/ (main.c miner.h): Define max gpudevices in one macro.
403 2011-09-08 09:45:45 <CIA-92> bitcoin: Con Kolivas * r758afbf00e42 cgminer/configure.ac: Add -lpthread, not -pthread and remove -lm which is not required.
404 2011-09-08 11:19:51 <piuk> Could anyone explain what the "subscribe" message does in CNode::Subscribe()? I could find anything in the protocol spec is it redundant?
405 2011-09-08 11:20:34 <edcba> redudant with ?
406 2011-09-08 11:20:57 <piuk> as in obsolete? not used anymore
407 2011-09-08 11:23:11 <edcba> do you know there are dev tools able to tell you if such method is called somewhere in code ? :)
408 2011-09-08 11:23:52 <gavinandresen> I'm probably misremembering, but I believe it is a half-implemented feature Satoshi was working on... something about merchants and clients and a more anonymous meet-in-the-middle way of paying for things.
409 2011-09-08 11:24:02 <gavinandresen> It aught to be removed, in my opinion.
410 2011-09-08 11:24:57 <UukGoblin> gavinandresen, 3rd or so day with -noirc and no lockups :-)
411 2011-09-08 11:25:11 <UukGoblin> so that seems to have helped
412 2011-09-08 11:25:20 <piuk> Ok thanks, as far as i can tell it's not used anywhere so it's probably best to remove dead code
413 2011-09-08 11:31:34 <shadders> q: is magic number escaped if it occurs somewhere in the middle of a message?
414 2011-09-08 12:12:01 <makomk> Is it just me or is the gitian build process for Bitcoin kind of broken right now?
415 2011-09-08 12:12:24 <ThomasV> hmm, what is the correct way to test a git pull request ?
416 2011-09-08 12:13:04 <lfm> if there was a "correct way" to test then youd never find any of the interesting bugs