1 2011-10-27 00:10:47 <magma_> When nodes work on a block that has the same transactions this is a bit redundant, no?
  2 2011-10-27 00:11:43 <gmaxwell> magma_: No.
  3 2011-10-27 00:12:58 <gmaxwell> The actual validation of the transactions is quite cheap compared to the process of finding the solution to make the block good.
  4 2011-10-27 00:13:25 <gmaxwell> And the fact that everyone validate the transactions is why the system is decenteralized no one trusts that anyone else did the checking.
  5 2011-10-27 00:35:23 <Davincij15> tcatm: thanks
  6 2011-10-27 02:14:19 <orange-hand> Hi, does anyone know of any theoretical way to do a decentralised escrow?
  7 2011-10-27 02:15:09 <casascius> quick question, who runs the web content at bitcoin.org?
  8 2011-10-27 02:26:43 <jgarzik> casascius: tcatm, with approval from devs + a few other trusted people like theymos
  9 2011-10-27 02:28:15 <Davincij15> in the bitcoin code under the wallet.cpp file if remove the ablity to send coins would that prevent me from finding blocks?
 10 2011-10-27 02:28:41 <casascius> Wanted to propose that given to the idea that one can now go to Memory Dealers and buy bitcoins with credit cards and PayPal.  (maybe not advertising MD directly, but definitely including "credit card" and some sort of link that eventually gets people there)
 11 2011-10-27 02:29:19 <Davincij15> It does not look like it but some times some some deveolpers like to be cute and execut methods by their memory location.
 12 2011-10-27 02:32:32 <casascius> Assuming he doesn't get burned or shut down, that's a pretty big deal
 13 2011-10-27 02:32:56 <casascius> He has a decent chance of success, both because of his existing business volume, and the fact that he's only shipping physical coins.
 14 2011-10-27 02:33:07 <Davincij15> orange-hand: decentralized escro would only work between bitcoin based currency because the electronic curency you use today exits at the will of a bank and legislation not by any kind of rules of reality.
 15 2011-10-27 02:33:14 <nanotube> orange-hand: yes, just put in a transaction that requires 2 of 3 signatures to redeem, with 3 potentials being buyer, seller, and escrow agent.
 16 2011-10-27 02:33:50 <nanotube> so either buyer and seller agree to release, or escrow agent, after reviewing the case, sides with either buyer or seller to release
 17 2011-10-27 02:34:18 <orange-hand> ok very nice
 18 2011-10-27 02:34:22 <orange-hand> is this implemented?
 19 2011-10-27 02:34:26 <nanotube> no :)
 20 2011-10-27 02:34:28 <Davincij15> nanotube release what? bitcoins dollars?
 21 2011-10-27 02:34:35 <amiller> as for a theoretical solution orange-hand
 22 2011-10-27 02:34:42 <amiller> if the 'terms of release' can be defined in a very precise computable way
 23 2011-10-27 02:34:45 <nanotube> orange-hand: if you manually make such a transaction, and mine it into a block, it will work
 24 2011-10-27 02:34:50 <amiller> such as a message "BEARS WIN" signed by a particular key
 25 2011-10-27 02:34:54 <nanotube> but the default client won't let you create it
 26 2011-10-27 02:34:57 <amiller> you can use the blockchain itself as the escrow agent
 27 2011-10-27 02:34:58 <gmaxwell> orange-hand: well, its kinda implemented. There is a patch. The bitcoin _network_ already supports it.
 28 2011-10-27 02:35:02 <nanotube> Davincij15: no, release bitcoins.
 29 2011-10-27 02:35:19 <nanotube> the bitcoin network of course has no ability to affect dollar release
 30 2011-10-27 02:35:25 <orange-hand> oh very nice
 31 2011-10-27 02:35:38 <orange-hand> I am writing a paper on laundry methods
 32 2011-10-27 02:35:50 <nanotube> use tide, and color-safe bleach
 33 2011-10-27 02:35:58 <nanotube> heh
 34 2011-10-27 02:35:59 <amiller> orange-hand do you know anything about blindbitcoin.com
 35 2011-10-27 02:36:15 <orange-hand> and one interesting solution I thought of was chaining laundries in an onion
 36 2011-10-27 02:36:27 <gmaxwell> orange-hand: Before you write a paper on "laundry methods" please read up about what money laundering actually is, so I don't have to flame you into the earth.
 37 2011-10-27 02:36:42 <orange-hand> I'll have you guys review it
 38 2011-10-27 02:36:46 <orange-hand> so don't worry :D
 39 2011-10-27 02:37:02 <nanotube> well, it would be better to review something written by someone who knows what money laundering is :)
 40 2011-10-27 02:37:03 <gmaxwell> orange-hand: most bitcoiners think that bouncing funds between a few wallets is laundering.. but that actually has very little relation to what money laundering is in the real world.
 41 2011-10-27 02:37:06 <amiller> orange-hand, you should look up Chaumian blinding and the Lucre library
 42 2011-10-27 02:37:08 <orange-hand> amiller, no I don't, and I can't seem to reach the site
 43 2011-10-27 02:37:47 <amiller> orange-hand, well blindbitcoin was a failed/ultimately-compromised attempt at making a bitcoin 'anonymizing' service using chaumian blinding
 44 2011-10-27 02:37:55 <orange-hand> gmaxwell, yes, I am doing comprehensive (I think) research on the subject
 45 2011-10-27 02:37:56 <gmaxwell> We don't even really have a name for doing that (the bouncing thing) with cash, because its something that people constantly do with cash anyways. E.g. when you go buy a stick of gum with a $50 and get back two $20s, a $5 and a few ones.
 46 2011-10-27 02:39:02 <orange-hand> from what I've read, what "laundering" is commonly done with bitcoins is really meant to be "layering"
 47 2011-10-27 02:39:08 <gmaxwell> I think it's actually fair to say that _true_ money laundering is currently impossible in bitcoin, because there are basically no high volume cashlike businesses to run the funds through. The only thing close is the exchanges, and you can be sure they're going to do nothing to protect your privacy against law enforcement. (nor should we as a community want them to)
 48 2011-10-27 02:39:28 <orange-hand> gmaxwell, I have a partial solution to that
 49 2011-10-27 02:39:29 <gmaxwell> orange-hand: yes, I suppose so. But without the final placement ::shrugs::
 50 2011-10-27 02:39:32 <orange-hand> incentive-based laundering
 51 2011-10-27 02:39:47 <orange-hand> that is, pay people to "invest"
 52 2011-10-27 02:40:01 <orange-hand> part of the commission could give a high 1 day investment return
 53 2011-10-27 02:40:26 <amiller> mix-servers can be paid for anonymizing service
 54 2011-10-27 02:40:26 <gmaxwell> orange-hand: I think people already know that one incentive to run bitcoin businesses is that you'll get more business than you deserve due to people trying to cycle bitcoin income that they'd rather keep private&
 55 2011-10-27 02:40:26 <orange-hand> say, .3%
 56 2011-10-27 02:40:27 <orange-hand> haha
 57 2011-10-27 02:40:29 <amiller> even blindbitcoin took a small commission
 58 2011-10-27 02:40:39 <orange-hand> amiller, how did they fail?
 59 2011-10-27 02:40:55 <gmaxwell> amiller: "mixing" is pretty much pointless. And duh you pay them... classical money laundering if pretty cash inefficient, you'll get back less than half you put in, unless you own the laundering infrastructure too.
 60 2011-10-27 02:41:43 <gmaxwell> There is still the mixer paradox: If your mixer is some uber anonymous party operating via hidden service .... then the rational thing to do will be to grow their bussiness to a comfortable size and then rip everyone off.
 61 2011-10-27 02:42:07 <gmaxwell> If the mixer is someone you can identify so you could send ninjas to beat them if they did this then your enemy will send ninjas to deanonymize you.
 62 2011-10-27 02:42:22 <orange-hand> heh
 63 2011-10-27 02:42:31 <nanotube> gmaxwell: the paradox is solved by elevating the mixer fee
 64 2011-10-27 02:42:40 <orange-hand> I am trying to compile all the desired properties of a mixing/laundering service
 65 2011-10-27 02:42:41 <amiller> well a properly implemented mixer provides a 'proof' of the correctness of its mixing
 66 2011-10-27 02:42:45 <amiller> and the blinding part, is done by the client
 67 2011-10-27 02:42:51 <nanotube> so that the ongoing revenue from operating the mixer is higher than the one-time lump sum from ripping people off
 68 2011-10-27 02:42:57 <orange-hand> and present multiple innovative solutions that partially solve some of those properties
 69 2011-10-27 02:43:12 <amiller> gmaxwell, so sure the mixer could leave with all the bitcions, but that wouldn't have to compromise your privacy, and they couldn't distinguish you to just rip off 'your' bitcoins
 70 2011-10-27 02:43:15 <gmaxwell> nanotube: yes, but it has to be quite high to do that, especailly since the mixer can so effectively hide the side of their standing pool.
 71 2011-10-27 02:43:53 <gmaxwell> amiller: the mixer can rob all of their customers at once, which is great.
 72 2011-10-27 02:44:20 <orange-hand> well, if you use the multiple signature transactions, the mixer can't steal, can it?
 73 2011-10-27 02:44:23 <nanotube> well, so say it is 10%. the present value of that would easily be greater than the one time ripoff.
 74 2011-10-27 02:44:27 <gmaxwell> Also, if your mixer figures out who you are  (after all you're not anonymous before the mixer or there wouldn't be a point for having a mixer) they can blackmail you.
 75 2011-10-27 02:44:33 <gmaxwell> nanotube: but its not a one time ripoff.
 76 2011-10-27 02:44:47 <nanotube> gmaxwell: start another mixer later?
 77 2011-10-27 02:45:03 <gmaxwell> nanotube: run them concurrently, and do the ripoff when the next in line is starting to pick up traffic.
 78 2011-10-27 02:45:08 <Davincij15> I did grep search of SendMoney and SendMoneyToBitcoinAddress none of these are used in generating blocks can anyone confirm this?
 79 2011-10-27 02:45:11 <nanotube> gmaxwell: right... i se. heh.
 80 2011-10-27 02:45:33 <gmaxwell> nanotube: their risk is only that they'll scare people away from mixing entirely.
 81 2011-10-27 02:45:37 <nanotube> orange-hand: no, but then you can steal from the mixer (get their coins, and fail to release yours?)
 82 2011-10-27 02:45:48 <amiller> maybe mixing is an activity you'd only want to participate in with your friends, web-of-trust style?
 83 2011-10-27 02:45:54 <nanotube> gmaxwell: indeed...
 84 2011-10-27 02:46:19 <nanotube> amiller: then it's not much of a mixer, since they can stick up your friends until they tell them who you are
 85 2011-10-27 02:46:21 <gmaxwell> amiller: nah, then you just identify a whole clique and you make them rat each other out.
 86 2011-10-27 02:46:54 <orange-hand> nanotube, in response to your question, the solution I am thinking of requires that I explain my theoretical mixing service scheme
 87 2011-10-27 02:46:56 <gmaxwell> Really the answer is that you simply _expect_ the mixer will be inefficient, that they'll rob you sometimes... and if your business model doesn't allow that, find a way to not need a mixer.
 88 2011-10-27 02:47:15 <nanotube> orange-hand: hehe ic. well then i'll look forward to reading about it in your paper, because i'm about to go to sleep :)
 89 2011-10-27 02:47:26 <orange-hand> heh, kk
 90 2011-10-27 02:47:34 <orange-hand> gn
 91 2011-10-27 02:47:40 <nanotube> o/ :)
 92 2011-10-27 02:47:47 <orange-hand> if anyone else is interested in hearing, I can continue
 93 2011-10-27 02:47:51 <amiller> cheers nanotube
 94 2011-10-27 02:47:55 <nanotube> gmaxwell: yep, costs of doing 'business' :)
 95 2011-10-27 02:48:05 <nanotube> amiller: ttyl
 96 2011-10-27 02:48:06 <gmaxwell> Mixing isn't actually all that important in any case. Making the funds appear to have a normal source is important. Because no matter how awesome your mixer is, if you're spending money the spending leaves evidence.
 97 2011-10-27 02:48:25 <gmaxwell> E.g. "Where did gmaxwell get that awesome boat from?"
 98 2011-10-27 02:48:35 <orange-hand> well, I am not interested in avoiding taxes
 99 2011-10-27 02:48:37 <gmaxwell> "er. I .. sold .. alpaca socks.. em"
100 2011-10-27 02:48:43 <orange-hand> more interested in cleaning very dirty funds
101 2011-10-27 02:48:53 <gmaxwell> orange-hand: laundering is usually not very tax efficient.
102 2011-10-27 02:49:19 <gmaxwell> Here is how you clean "dirty funds" in bitcoin: Sell them to a horder who isn't planning on spending them any time this decade.
103 2011-10-27 02:49:43 <orange-hand> lol
104 2011-10-27 02:49:47 <orange-hand> hmm
105 2011-10-27 02:49:51 <orange-hand> that is a good thought
106 2011-10-27 02:49:53 <amiller> hm, that's pretty good
107 2011-10-27 02:49:59 <orange-hand> lol
108 2011-10-27 02:50:15 <orange-hand> I'll add that to the incentive based laundry
109 2011-10-27 02:50:34 <orange-hand> higher incentive for coins that aren't used often?
110 2011-10-27 02:51:29 <gmaxwell> Well, you do the txn with them anonymously using a 2 of 2 escrow transaction. Then they stir the coins for the next decade but carefully keep them out of exchanges and real businesses.
111 2011-10-27 02:52:00 <orange-hand> that can easily be graphed though
112 2011-10-27 02:52:52 <gmaxwell> nah. I mean, yes you can always tell the coins came from someplace dirty in the past, but by the time they get a name.. ten years will have passed and it won't be completely clear exactly when that person got them.
113 2011-10-27 02:53:20 <orange-hand> yes but
114 2011-10-27 02:53:29 <orange-hand> hoarders can only really be trusted once
115 2011-10-27 02:53:41 <orange-hand> otherwise you will just be trading dirty money for dirty money
116 2011-10-27 02:53:51 <gmaxwell> They're stupid if they send them to the exchange a month later you tell them its dirty and thats why you're paying them a premium.
117 2011-10-27 02:54:07 <gmaxwell> oh.. well true, but I suppose you could investigate the coins they propose to pay you with.
118 2011-10-27 02:54:41 <gmaxwell> a key point here is that even if the horder eventually has to launder the dirty coin conventionally, they'd only have to do it in the _future_ when it would be presumably much easier to do that.
119 2011-10-27 02:54:53 <gmaxwell> (or really... _possible_ to do that)
120 2011-10-27 02:55:03 <gmaxwell> and besides, you shouldn't steal people's bitcoin. Shame on you.
121 2011-10-27 02:55:16 <orange-hand> yes ... this is beginning to sound like an extension to the incentive based trading broker
122 2011-10-27 02:55:31 <orange-hand> I haven't
123 2011-10-27 02:55:46 <orange-hand> I am exploring this for fun, and maybe future profit
124 2011-10-27 02:56:16 <orange-hand> and also at the same time, make more tracing tools
125 2011-10-27 02:56:20 <gmaxwell> The underworld is crappy business from a quality of life perspective.
126 2011-10-27 02:56:45 <gmaxwell> If you're actually smart enough to do something useful in that domain then you could do lots of completely above the board things in bitcoin that will be just as profitable if not moreso.
127 2011-10-27 02:57:03 <orange-hand> well, I can make a profit by running an incentive based laundry and skimming a commission
128 2011-10-27 02:57:27 <orange-hand> perhaps I will
129 2011-10-27 02:57:44 <orange-hand> this is just an idea I thought of, if you have profitable ideas, share them lol
130 2011-10-27 02:58:15 <gmaxwell> And any loss of profitablity will be offset by (0) having a _much_ larger customer base, (1) not having to deal with criminals, who tend to be shit customers, (2) not making yourself a target for entities far more powerful than you, (3) not depriving yourself of the shared protective forces lawful activities enjoy (courts, cops, public irc channels)
131 2011-10-27 02:59:49 <gmaxwell> orange-hand: you can basically take $RANDOM_POPULAR_THING and add "with bitcoin" and if bitcoin is successful then you have a good chance of being successful just through first mover advantage so long as you're basically competent. Obviously you can do better by being more intelligent than that.
132 2011-10-27 02:59:56 <orange-hand> as I said, if you have any profitable ideas, share them :D
133 2011-10-27 03:00:52 <orange-hand> true, but most things take startup money
134 2011-10-27 03:00:57 <orange-hand> laundering doesn't
135 2011-10-27 03:01:03 <orange-hand> or very little
136 2011-10-27 03:01:26 <gmaxwell> Well, I have before. For example, we could use a number of different merchant lubrication services e.g. things to secure fast transactions.
137 2011-10-27 03:01:47 <orange-hand> and truly, if I think of something legit, I will surely explore it too
138 2011-10-27 03:01:48 <gmaxwell> (and there are a number of ways to do that, and many of them need a semi-trusted service provider)
139 2011-10-27 03:02:57 <orange-hand> wouldn't such a provider need a large sum of money to startup?
140 2011-10-27 03:03:05 <orange-hand> regulations, fees etc.
141 2011-10-27 03:04:10 <gmaxwell> No. You don't have to have transactions go through you like mtgox.
142 2011-10-27 03:04:45 <gmaxwell> https://bitcointalk.org/index.php?topic=28565.5;wap2  < options (3) and (4) on my list here.
143 2011-10-27 03:04:46 <orange-hand> so what, just something like speedy confirmations?
144 2011-10-27 03:05:08 <gmaxwell> (3) needs some money to back your promises, but not a ton.
145 2011-10-27 03:06:30 <orange-hand> interesting
146 2011-10-27 03:06:45 <gmaxwell> Esp since you don't need to insure transactions for their full value. E.g. you could reply to a vendor "I'll insure  _this_ (8 BTC) tranaction against reversal for a price of 0.0001 BTC per BTC of insurance up to a maximum of 5 BTC, quarter price if you get a copy of their photo id"
147 2011-10-27 03:06:50 <orange-hand> how fast can the fastest transaction occur though; it has to be at minimum-worst-case 10 minutes
148 2011-10-27 03:06:53 <orange-hand> no?
149 2011-10-27 03:07:32 <gmaxwell> orange-hand: no, it can happen _instantly_ because you don't wait for it to go into the blockchain. Instead you make sure it got propagated to the big miners. And you listen to any conflicting txn.
150 2011-10-27 03:07:43 <gmaxwell> And you're taking a little risk, but you're charging a fee.
151 2011-10-27 03:07:53 <orange-hand> ok
152 2011-10-27 03:08:25 <orange-hand> hehe
153 2011-10-27 03:08:35 <orange-hand> my evil mind is thinking about how I would try to attack that
154 2011-10-27 03:08:39 <gmaxwell> The thing is, as the insurance vendor you have a much better ability to assess the risk at every instant, so it's worth it for vendors to pay you to take it instead of taking the risk themselves.
155 2011-10-27 03:08:49 <gmaxwell> Thats the kind of mind needed to offer the service.
156 2011-10-27 03:09:08 <gmaxwell> Joe blow store vendor can't really handle the risk because he can't understand the attacks.
157 2011-10-27 03:10:14 <orange-hand> ok cool
158 2011-10-27 03:10:17 <orange-hand> I'll definitely look into it
159 2011-10-27 03:10:18 <orange-hand> still going to finish the paper
160 2011-10-27 03:10:22 <gmaxwell> Have fun.
161 2011-10-27 03:10:29 <orange-hand> it identifies a lot of vulnerabilities (I think) in current "laundries"
162 2011-10-27 03:12:12 <orange-hand> ok, I am going afk
163 2011-10-27 03:12:19 <orange-hand> good talking to you
164 2011-10-27 04:58:04 <diki> the bitcoin wiki is down
165 2011-10-27 05:05:48 <phantomcircuit> diki, hardware failure
166 2011-10-27 05:20:13 <diki> thats very unusual
167 2011-10-27 05:20:36 <JFK911> hacked
168 2011-10-27 05:22:45 <nathan7> greetings, earthlings.
169 2011-10-27 05:22:56 <nathan7> ohshit.
170 2011-10-27 05:23:01 <nathan7> that's bad.
171 2011-10-27 05:24:14 <diki> i actually need some stuff from the wiki
172 2011-10-27 05:25:27 <neofutur> (09:13) < phantomci> MagicalTux, is there a backup of the wiki?
173 2011-10-27 05:25:30 <neofutur> fyi
174 2011-10-27 05:25:41 <diki> weird
175 2011-10-27 05:25:42 <diki> i cant
176 2011-10-27 05:25:52 <neofutur> not yet
177 2011-10-27 05:28:42 <neofutur> (09:14) <@MagicalTu> the ddos protected ips are down right now
178 2011-10-27 05:28:49 <neofutur> (09:14) <@MagicalTu> because they don't route to the right rack
179 2011-10-27 05:29:33 <neofutur> just wait, dont worry and /ignore JFK911 troll :p
180 2011-10-27 05:34:00 <JFK911> tux is also hosting the wiki?
181 2011-10-27 05:34:52 <FellowTraveler1> hi all.
182 2011-10-27 05:35:34 <neofutur> JFK911: actually he created it afaik
183 2011-10-27 05:37:12 <neofutur> will you complain because people try to help bitcoin and actually _do_ things ?
184 2011-10-27 05:43:51 <jav__> Hi there! Does someone here have some testnet coins to spare? mrBvHhWXjt8L5v43ejvWAcnzddfh3i7R2b .. I tried to get some from the faucet, but even so it claimed to have sent them, I don't think it did (I didn't receive anything and testnet blockexplorer showed nothing either) (?)
185 2011-10-27 05:50:15 <phantomcircuit> jav__, it can take a *very* long time to generate a testnet block
186 2011-10-27 05:51:38 <jav__> phantomcircuit: I tried to get coins from the faucet yesterday.. or is *very* long several days?
187 2011-10-27 05:51:52 <phantomcircuit> that seems too long
188 2011-10-27 05:52:02 <phantomcircuit> i have some... somewhere
189 2011-10-27 06:08:56 <jav__> nice... just a second ago I received all the faucet coins I had requested
190 2011-10-27 07:13:14 <tcatm> casascius: the first "call to action" link on bitcoin.org links to http://www.weusecoins.com/ so that site might be a good place to add a link to MD
191 2011-10-27 07:16:04 <jav__> Question: If I'm looking at a script of the form "<sig> <pubKey> OP_DUP ..." then <sig> is a signature of a hash of a simplified transaction, right? does that simplified transcation include <pubKey>?
192 2011-10-27 08:15:39 <benne> regarding translation of the bitcoin channel - who to contact?
193 2011-10-27 08:17:43 <SomeoneWeird> channel?
194 2011-10-27 08:17:45 <SomeoneWeird> client?
195 2011-10-27 08:18:47 <benne> wow - sorry
196 2011-10-27 08:18:48 <benne> client :)
197 2011-10-27 08:19:10 <tcatm> <-
198 2011-10-27 10:25:37 <CIA-101> libbitcoin: genjix * r4e8e7dc2ba6a / (3 files in 3 dirs): Completed EC key class + unit test.
199 2011-10-27 11:05:29 <Diablo-D3> ffs
200 2011-10-27 11:05:33 <Diablo-D3> will someone temp kline him?
201 2011-10-27 11:08:07 <nathan7> wel damnit
202 2011-10-27 11:08:09 <nathan7> *well
203 2011-10-27 11:11:55 <edcba> ...
204 2011-10-27 11:12:05 <nathan7> Someone remind me how to ban someone with a note.
205 2011-10-27 11:12:30 <edcba> just ban him and kick him with a note :)
206 2011-10-27 11:12:37 <gmaxwell> nathan7: /msg sucker You suck
207 2011-10-27 11:16:35 <Graet> /ban midnightmagic fix your internet!
208 2011-10-27 11:22:39 <nathan7> hrmpf.
209 2011-10-27 11:24:37 <nathan7> I am being driven insane
210 2011-10-27 11:26:00 <imsaguy2> its across all his channels
211 2011-10-27 11:26:02 <imsaguy2> it sucks
212 2011-10-27 11:26:03 <nathan7> Yeah.
213 2011-10-27 11:26:29 <imsaguy2> services should catch on to an issue like this and resolve it
214 2011-10-27 11:26:36 <nathan7> Yes, they should
215 2011-10-27 11:26:44 <imsaguy2> its every ~30 seconds
216 2011-10-27 11:26:47 <manveru> what's he doing?
217 2011-10-27 11:26:57 <imsaguy2> his client is probably trying to join too many channels at once
218 2011-10-27 11:27:04 <imsaguy2> and then freenode kills him off
219 2011-10-27 11:27:15 <imsaguy2> so it just keeps repeating until his client runs out of retries
220 2011-10-27 11:27:20 <imsaguy2> (if it even does)
221 2011-10-27 11:28:11 <nathan7> mhm
222 2011-10-27 11:28:16 <manveru> ah, i filter join/part/quit out, so i don't even see it
223 2011-10-27 11:28:47 <imsaguy2> I like knowing who is coming in the door
224 2011-10-27 11:29:05 <imsaguy2> what makes it worse is he has long enough to auth
225 2011-10-27 11:29:15 <imsaguy2> so the cloaking doubles the join/part
226 2011-10-27 11:31:18 <Eliel> !later tell gavinandresen if we're adding a new transaction type, I'd suggest we use an error correcting code for the checksum. If the code is 4 bytes, then a 4 byte code could detect errors up to 4 bytes and fix up to 2 bytes.
227 2011-10-27 11:31:18 <gribble> The operation succeeded.
228 2011-10-27 11:31:44 <Eliel> !later tell gavinandersen uh, this is for the new address type, not the transaction.
229 2011-10-27 11:31:44 <gribble> The operation succeeded.
230 2011-10-27 11:32:01 <imsaguy2> this is what makes freenode suck
231 2011-10-27 11:32:06 <imsaguy2> I mention it in #freenode
232 2011-10-27 11:32:14 <imsaguy2> the reply: <+jbroome> imsaguy2: he's not in here, talk to the ops of those channels
233 2011-10-27 11:32:31 <imsaguy2> so basically, even though its a network issue, each channel should have to ban him
234 2011-10-27 11:37:14 <lianj> aw, whenhe is back at his session he may wonder why all of freenode turned on him :D
235 2011-10-27 11:37:44 <imsaguy2> well its frakking annoying
236 2011-10-27 11:43:08 <nathan7> hrmpf.
237 2011-10-27 11:44:58 <imsaguy2> he's done
238 2011-10-27 11:45:59 <nathan7> hmm, seems to be over
239 2011-10-27 11:46:15 <ThomasV> how does bitcoin choose the input addresses it uses in a transaction ? does it simply try to minimize the length of the tx, or does it take other things into consideration?
240 2011-10-27 11:46:33 <imsaguy2> ThomasV, the current input selection code sucks
241 2011-10-27 11:46:36 <imsaguy2> to put it bluntly
242 2011-10-27 11:47:00 <ThomasV> that's a vague answer
243 2011-10-27 11:48:05 <imsaguy2> It really doesn't take into account transaction size
244 2011-10-27 11:48:11 <imsaguy2> or anything else that matters
245 2011-10-27 11:48:21 <ThomasV> so what does it do ?
246 2011-10-27 11:48:47 <imsaguy2> gmaxwell ^
247 2011-10-27 11:49:10 <gmaxwell> It's actually quite smart but it's quite smartly doing the wrong thing.
248 2011-10-27 11:49:56 <ThomasV> will you guys tell me more ?
249 2011-10-27 11:50:02 <gmaxwell> In particular the selection code knows basically nothing about the anti-ddos priority. So it will happily make a selection that causes you to have to pay the nussiance fee when otherwise was possible.
250 2011-10-27 11:50:40 <ThomasV> what is the nussiance fee ?
251 2011-10-27 11:50:53 <gmaxwell> Go read the code?  It's not that complicated. It tries to minimize the data size, taking several passes first starting off with txn which have 6+ confirms (or 1 confirm if they are yours often the source of the fee from what I've seen).
252 2011-10-27 11:51:42 <gmaxwell> ThomasV: The 0.0005 BTC you need to pay when your transaction has very low priority.
253 2011-10-27 11:52:47 <gmaxwell> (https://en.bitcoin.it/wiki/Transaction_fees)
254 2011-10-27 11:54:09 <ThomasV> gmaxwell: minimizing the data size is not simple. if you always minimize it for the current tx, I guess you end up with lots of small amounts on various addresses, and if you later need to send a large amount, you will need to use multiple inputs, which you could have avoided. so there seems to be a tradeoff between present and future tx sizes
255 2011-10-27 11:56:28 <gmaxwell> It's not the right metric anymore in any case. Or rather, it's the net most important thing after priority.
256 2011-10-27 11:56:28 <ThomasV> I guess my question is, does it try to minimize the current tx size, or does it also try to keep a small number of addresses in order to save on future fees
257 2011-10-27 11:57:09 <ThomasV> gmaxwell: btw, the wiki is down
258 2011-10-27 11:57:13 <gmaxwell> Yes, no.  And I'm not so sure that keeping a small number of addresses saves future fees. (within bounds)
259 2011-10-27 11:57:38 <ThomasV> I see.
260 2011-10-27 11:58:04 <gmaxwell> What I _think_ would be better is finding the lowest priority input greater than the fee minimum. But thats a more complex objective.
261 2011-10-27 11:58:18 <gmaxwell> That would tend to sweep out inputs as it got a chance.
262 2011-10-27 11:59:21 <ThomasV> what is the fee minimum ?
263 2011-10-27 11:59:30 <imsaguy2> 0
264 2011-10-27 11:59:35 <imsaguy2> then .0005
265 2011-10-27 11:59:53 <ThomasV> I mean, how is it related to the data size
266 2011-10-27 12:01:12 <gmaxwell> ThomasV: The priority of a txn is calculated as SUM_OVER_INPUTS( value * confirms ) / txn_bytes  where value is in base units.
267 2011-10-27 12:01:51 <ThomasV> gmaxwell: why ?
268 2011-10-27 12:02:02 <imsaguy2> basically 1 btc takes 1 day
269 2011-10-27 12:02:31 <gmaxwell> If this is under 51m (which for typical txn sizes is 1 BTC /day) then nodes won't forward unless there is at least a 0.0005 fee.
270 2011-10-27 12:02:53 <gmaxwell> This is a fairly effective and hard to avoid anti-DOS measure.
271 2011-10-27 12:03:43 <gmaxwell> (since it basically uses _stationary bitcoins_ as a proof-of-work to permit the txn. ... and lots of stationary bitcoins is not what you have if you're trying to do an attach that involves moving them a lot)
272 2011-10-27 12:04:14 <ThomasV> I see
273 2011-10-27 12:05:03 <gmaxwell> you'll also get hit with the 0.0005 fee if your txn has any outputs 0.01  (making open txn set bloating attacks costly)... and there are per-kb fees for very large txn, but I think that gets hit less often or at least people complain less often.
274 2011-10-27 12:05:24 <ThomasV> but this priority is used for forwarding the tx ; do miners also take priority into account, or do they just try to maximize their wealth ?
275 2011-10-27 12:05:40 <gmaxwell> though after the attacks on some of the other blockchain cryptocurrencies I've noticed very few complaints people do seem to believe that it prevents DOS attacks now. :)
276 2011-10-27 12:06:00 <imsaguy2> miners will include high priority before normal before low priority
277 2011-10-27 12:06:07 <gmaxwell> ThomasV: miners try to maximize their wealth, but they do take the free txn in priority order. Thats an argument against minimizing it.
278 2011-10-27 12:06:08 <imsaguy2> unless its been modified
279 2011-10-27 12:06:46 <gmaxwell> though you could make the decision dynamically if you're a full client: e.g. what priority should have gotten me into the last block assuming that miner used the same rules I'd use.
280 2011-10-27 12:11:14 <ThomasV> I am not a full client :-)
281 2011-10-27 12:15:18 <gmaxwell> ah, well then.
282 2011-10-27 12:15:30 <gmaxwell> (hey, at least I was thoughtful to suggest the possibility that you might not be one!)
283 2011-10-27 12:39:41 <gavinandresen> Eliel:  see https://bitcointalk.org/index.php?topic=46538.msg584239#msg584239  for my thoughts on address checksum improvements
284 2011-10-27 13:57:39 <gmaxwell> UI fail:
285 2011-10-27 13:57:44 <gmaxwell> What does a user do when they see error: {"code":-13,"message":"Error: Please enter the wallet passphrase with walletpassphrase first."}
286 2011-10-27 13:57:48 <gmaxwell> ?
287 2011-10-27 13:58:02 <gmaxwell> they do this:
288 2011-10-27 13:58:05 <gmaxwell> to: Mnamecoinaddress" error: {"code":-32601,"message":"Method not found"}
289 2011-10-27 14:00:29 <jrmithdobbs> ThomasV: crazy you ever figure out what it was?
290 2011-10-27 14:00:49 <ThomasV> jrmithdobbs: not really
291 2011-10-27 14:01:39 <ThomasV> jrmithdobbs: but I found that my addresses actually work, except for one of them, which is the one I was testing
292 2011-10-27 14:01:57 <jrmithdobbs> haha, hate when that happens
293 2011-10-27 14:02:31 <ThomasV> I need to investigate more
294 2011-10-27 14:02:58 <ThomasV> or I can give you the private key, and the bitcents will be for you if you can get them :-)
295 2011-10-27 14:04:12 <jrmithdobbs> ThomasV: i don't have time to fiddle with it today tbqh, was just wondering if you found the issue
296 2011-10-27 14:04:38 <ThomasV> no, but the issue is not what I thought in the first place
297 2011-10-27 14:04:56 <jrmithdobbs> ThomasV: you should test with testnet addresses
298 2011-10-27 14:05:16 <jrmithdobbs> ThomasV: i had an obo in my addr generation code that was causing 1/32 or something like that to be invalid
299 2011-10-27 14:05:16 <ThomasV> yeah, but it's a few cents, not a big deal
300 2011-10-27 14:05:28 <ThomasV> obo ?
301 2011-10-27 14:05:35 <jrmithdobbs> doing it on testnet and testing with 100ish addresses at once makes those easier to spot
302 2011-10-27 14:05:38 <jrmithdobbs> off by one
303 2011-10-27 14:05:58 <BlueMatt> graingert: see gribble msg
304 2011-10-27 14:07:05 <graingert> BlueMatt: I don't see one
305 2011-10-27 14:07:17 <graingert> BlueMatt: I see it now :p
306 2011-10-27 14:07:23 <BlueMatt> hehe
307 2011-10-27 14:07:29 <graingert> odd I was online 3 minutes ago...
308 2011-10-27 14:07:42 <BlueMatt> gribble doesnt send it to you until after you speak after the later tell
309 2011-10-27 14:07:47 <BlueMatt> (as it is assumed you might be afk)
310 2011-10-27 14:18:38 <BlueMatt> actually, Ill repost that here: if any trusted developers (ie gavin trusts you to do release builds) want to help with the bitcoin ubuntu ppa, please tag me
311 2011-10-27 14:44:58 <graingert> BlueMatt: I thought gittian stopped that
312 2011-10-27 14:45:14 <BlueMatt> not if you do ppas, launchpad has to do the building
313 2011-10-27 14:45:15 <graingert> and also the ubuntu build process needs you to build on their servers with a public build log
314 2011-10-27 14:45:27 <graingert> BlueMatt: can you not integrate them
315 2011-10-27 14:45:39 <graingert> launchpad will load gitian
316 2011-10-27 14:45:43 <BlueMatt> no, most of the launchpad building is on vms
317 2011-10-27 14:45:49 <graingert> for example
318 2011-10-27 14:45:52 <BlueMatt> (which dont have native vt-x support, which is effectively required)
319 2011-10-27 14:45:56 <graingert> ah
320 2011-10-27 14:46:04 <BlueMatt> yea, but if you are on the team you can still push arbitrary code to launchpad which it will build and distribute
321 2011-10-27 14:46:35 <graingert> meh I'd like to be on the team - but I'm not a trusted developer
322 2011-10-27 14:47:06 <BlueMatt> Id talk to gavin, he would probably be willing to add quite a few people
323 2011-10-27 14:47:22 <BlueMatt> but Im gonna defer to his judgment here
324 2011-10-27 14:49:25 <nanotube> BlueMatt: it may be helpful to note what kind of help is needed with the ppa
325 2011-10-27 14:49:59 <BlueMatt> graingert wants nightly/beta/stable packages, and I dont really have the time to commit to keeping those up-to-date
326 2011-10-27 14:50:02 <gavinandresen> graingert: email me some information about you-- experience, where you are, etc.  And a phone number I can call to talk with you.  I'm just extra paranoid when it comes to the bitcoin release process....
327 2011-10-27 14:50:29 <nanotube> paranoia++.
328 2011-10-27 14:50:39 <BlueMatt> its a good idea, just not something I can do
329 2011-10-27 15:11:57 <BlueMatt> alright, another round of building and some minor updates and ppa:bitcoin/bitcoin should be ready (at least for stable releases)
330 2011-10-27 15:17:16 <graingert> BlueMatt: w00t
331 2011-10-27 15:22:18 <jgarzik> BlueMatt: see my note from yesterday?
332 2011-10-27 15:33:48 <BlueMatt> jgarzik: no
333 2011-10-27 15:35:29 <BlueMatt> jgarzik: leave it with gribble in a later tell, thanks
334 2011-10-27 15:51:40 <CIA-101> bitcoinj: miron@google.com * r255 /trunk/src/com/google/bitcoin/examples/PrintPeers.java: Fix another Java-6ism
335 2011-10-27 15:55:12 <graingert> !later tell BlueMatt how long does a "round of building" take?
336 2011-10-27 15:55:12 <gribble> The operation succeeded.
337 2011-10-27 15:56:53 <gribble> The operation succeeded.
338 2011-10-27 15:56:53 <nanotube> ;;later tell BlueMatt jgarzik said 'that sounds like a good ppa setup' - that's the only line of his from yesterday that contained your nick :)
339 2011-10-27 16:02:06 <graingert> gavinandresen: you sound very different on the phone than from other times I've heard you
340 2011-10-27 16:02:44 <gavinandresen> graingert: maybe the Skype connection?  through the microphone in my mac... I didn't bother trying to get a good quality connection...
341 2011-10-27 16:03:52 <gavinandresen> graingert: I told BlueMatt I trust you are who you say you are-- thanks for offering to help!
342 2011-10-27 16:04:23 <graingert> awesome :)
343 2011-10-27 16:04:31 <graingert> yeah it sounded like digital artifacts
344 2011-10-27 16:04:33 <cuqa> when do orphans turn out to be orphans?
345 2011-10-27 16:04:45 <cuqa> pretty fast or can it still happen at like 80 confirmations
346 2011-10-27 16:04:52 <graingert> cuqa: orphan blocks?
347 2011-10-27 16:05:05 <cuqa> yes, orphan blocks, not orphan kids :<
348 2011-10-27 16:05:13 <graingert> cuqa: it can happen whenever
349 2011-10-27 16:05:18 <gmaxwell> cuqa: almost always by the next block..
350 2011-10-27 16:05:21 <graingert> it just becomes more and more likely
351 2011-10-27 16:05:23 <cuqa> ok
352 2011-10-27 16:05:29 <graingert> not to happen
353 2011-10-27 16:05:35 <cuqa> alright
354 2011-10-27 16:05:37 <graingert> vanishingly small at 80
355 2011-10-27 16:05:50 <cuqa> and at 120 its locked?
356 2011-10-27 16:05:50 <graingert> and limited to some blocks hard coded in the client
357 2011-10-27 16:05:57 <graingert> cuqa: it's never 100%
358 2011-10-27 16:06:00 <cuqa> oh
359 2011-10-27 16:06:08 <graingert> unless it's included in the client
360 2011-10-27 16:06:18 <cuqa> and that happened ever?
361 2011-10-27 16:06:23 <cuqa> probably not
362 2011-10-27 16:06:30 <graingert> I think the max is 2
363 2011-10-27 16:06:49 <gmaxwell> There was intentionally a deep fork created due to a fix for a software bug. But otherwise I don't think I've seen anyone mention observing greater than 2.
364 2011-10-27 16:07:25 <graingert> whs
365 2011-10-27 16:07:27 <cuqa> what I dont really get
366 2011-10-27 16:07:33 <gmaxwell> (a txn that created a zillion bitcoin got mined once, and the software was fixed and automatically forked the chain before the junk transaction and eventually overtook it)
367 2011-10-27 16:07:47 <cuqa> or dont understand how I would do it is determining the chance that a block becomes invalid
368 2011-10-27 16:08:00 <graingert> cuqa: you want to shift to #bitcoin-mining?
369 2011-10-27 16:08:15 <gmaxwell> cuqa: go read the bitcoin paper. there is software for calculating that.
370 2011-10-27 16:08:32 <cuqa> thx
371 2011-10-27 16:08:49 <gmaxwell> https://bitcoin.org/bitcoin.pdf
372 2011-10-27 16:33:43 <diki> tcatm:you are the owner of bitcoincharts, right?
373 2011-10-27 16:33:47 <diki> i saw a small typo
374 2011-10-27 16:35:46 <tcatm> diki: yes
375 2011-10-27 16:36:16 <diki> its in the "Out of coins" message
376 2011-10-27 16:36:32 <diki> 
377 2011-10-27 16:36:44 <diki> i think it should be just "relic"
378 2011-10-27 16:40:42 <CIA-101> libbitcoin: genjix * r967ec4deb468 /src/types.cpp: optionally allow non space separated hex input.
379 2011-10-27 18:00:52 <CIA-101> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * rd28d370 / daemon/init.js : Fixed duplicate helpful error message. (Once is plenty, thanks.) - http://git.io/WqnC3Q
380 2011-10-27 18:00:53 <CIA-101> bitcoinjs/node-bitcoin-p2p: Stefan Thomas master * rd5f26b8 / package.json : Updated to latest node-irc. Fixes #38. - http://git.io/PwTKqA
381 2011-10-27 21:05:48 <CIA-101> libbitcoin: genjix * rb0aa5ee4cacb / (3 files in 3 dirs): suvertx: priv app for new keypair, signing and verifying
382 2011-10-27 21:19:36 <freewil> what happened to v0.5.0
383 2011-10-27 21:31:32 <Eliel> freewil: it's not released yet
384 2011-10-27 21:32:16 <freewil> well i know that
385 2011-10-27 21:32:26 <freewil> is rc1 still being tested?
386 2011-10-27 21:35:17 <Eliel> yes
387 2011-10-27 21:39:26 <graingert> I was under the impression that rc2 was being tested
388 2011-10-27 21:40:08 <freewil> there's only a rc1 tagged on github
389 2011-10-27 21:41:25 <graingert> https://github.com/bitcoin/bitcoin/tags
390 2011-10-27 21:41:27 <graingert> so it is
391 2011-10-27 21:41:32 <graingert> perhaps I was seeing things
392 2011-10-27 22:15:46 <CIA-101> bitcoin: Con Kolivas * rd8770654f5db cgminer/configure.ac: Add x86_64 w64 mingw32 target courtesy of dukrat.
393 2011-10-27 22:21:08 <BlueMatt> graingert: ping
394 2011-10-27 22:22:37 <graingert> pong
395 2011-10-27 22:22:54 <BlueMatt> do you have a launchpad account?
396 2011-10-27 22:22:59 <graingert> I do
397 2011-10-27 22:23:07 <BlueMatt> name?
398 2011-10-27 22:23:25 <graingert> https://launchpad.net/~tagrain
399 2011-10-27 22:23:31 <BlueMatt> <gribble> Sent 6 hours and 17 minutes ago: <gavinandresen> (graingert)... so I trust him.
400 2011-10-27 22:23:48 <BlueMatt> added
401 2011-10-27 22:23:58 <graingert> fair enough
402 2011-10-27 22:24:00 <BlueMatt> you should now be able to push to ppa:bitcoin/bitcoin
403 2011-10-27 22:24:02 <graingert> thanks :)
404 2011-10-27 22:24:12 <BlueMatt> (the scripts I use are on my debian-build branch)
405 2011-10-27 22:24:37 <BlueMatt> ask if you need the miniupnpc stuff (Im too lazy to package that up nicely...)
406 2011-10-27 22:25:29 <graingert> it might be an idea to grab ~stretch
407 2011-10-27 22:25:56 <graingert> for the build team
408 2011-10-27 22:25:59 <graingert> as he's been doing fine so far
409 2011-10-27 22:26:09 <graingert> well he's gone on holiday recently
410 2011-10-27 22:29:31 <BlueMatt> thanks launchpad: "Hello Matt Corallo, graingert (tagrain) has been added as a member of Bitcoin (bitcoin) by Matt Corallo (bluematt)"
411 2011-10-27 22:29:38 <BlueMatt> oh really, I added someone?
412 2011-10-27 22:29:41 <BlueMatt> I didnt realize that
413 2011-10-27 22:29:57 <graingert> :p
414 2011-10-27 22:30:14 <BlueMatt> in other news, the oneiric build is broken because they dont have libdb4.8++-dev, only libdb4.8-dev...
415 2011-10-27 22:30:17 <graingert> BlueMatt: I guess it's in case someone gets authed as you
416 2011-10-27 22:30:43 <BlueMatt> and I cant decide whether to upload a copy of it, or link bitcoin against libdb5.1++-dev
417 2011-10-27 22:31:14 <graingert> http://packages.ubuntu.com/libdb4.8++ is down for me
418 2011-10-27 22:31:19 <BlueMatt> comments, questions?
419 2011-10-27 22:31:36 <BlueMatt> is down for you?
420 2011-10-27 22:31:45 <graingert> I'd rather have less packages installed per repo
421 2011-10-27 22:32:02 <graingert> less packages in each repo
422 2011-10-27 22:32:04 <BlueMatt> I agree, but Im also very against linking against libdb5.1
423 2011-10-27 22:32:26 <graingert> can you discover the reason for pulling 4.8++
424 2011-10-27 22:32:37 <BlueMatt> I googled a bit, but didnt find anything
425 2011-10-27 22:32:43 <BlueMatt> would probably have to ask the maintainer...
426 2011-10-27 22:33:05 <graingert> can I vote for renaming the bitcoin ppa, dev
427 2011-10-27 22:33:23 <BlueMatt> you mean rename bitcoin/bitcoin to like bitcoin/bitcoin-dev?
428 2011-10-27 22:33:33 <graingert> I was thinking bitcoin/dev
429 2011-10-27 22:33:51 <BlueMatt> but the idea is to link bitcoin/bitcoin for people to use
430 2011-10-27 22:33:57 <graingert> yep
431 2011-10-27 22:34:01 <BlueMatt> (on the download page)
432 2011-10-27 22:34:01 <graingert> that would be for release
433 2011-10-27 22:34:06 <BlueMatt> so its not a dev repo...
434 2011-10-27 22:34:19 <graingert> well currently it's RC
435 2011-10-27 22:34:26 <BlueMatt> (the only reason there are rcs now is because I dont want to do 0.4)
436 2011-10-27 22:34:39 <BlueMatt> so Im just gonna upgrade it when 0.5 hits beta
437 2011-10-27 22:35:03 <graingert> okay, just uncomfortable with having dev builds in what will be the release PPA
438 2011-10-27 22:35:14 <BlueMatt> I would say its fine
439 2011-10-27 22:35:25 <BlueMatt> (isnt it the way the chrome ppa is?)
440 2011-10-27 22:35:45 <graingert> really what need to be discussed is a release schedule
441 2011-10-27 22:35:46 <CIA-101> libbitcoin: genjix * r9b26b10b5833 / (5 files in 5 dirs): Bitcoin address from private key.