1 2010-10-03 00:04:20 <yebyen> i wonder if i published my bitcrapsTEST binary
  2 2010-10-03 00:04:24 <yebyen> the modified bitcoin
  3 2010-10-03 00:19:01 <Keefe> does bitcoin wait until the next block before recording a generated block in the wallet?
  4 2010-10-03 00:19:53 <yebyen> i think it waits for a confirmation
  5 2010-10-03 00:20:12 <echelon> could the bitcoin network support a system of accumulating interest on your balance for coin generation instead of a flat 50 btc's?
  6 2010-10-03 00:21:15 <yebyen> no man, that's crazy :D
  7 2010-10-03 00:21:39 <yebyen> someone's gotta be able to come in with a bigger computer and get as much money as you even if you're rich
  8 2010-10-03 00:21:41 <Keefe> in other words, the reward for generating be directly related to your balance instead of always 50btc (currently)?
  9 2010-10-03 00:22:19 <echelon> yeah
 10 2010-10-03 00:22:21 <Keefe> that would mess up the concept of planned-in-advance inflation
 11 2010-10-03 00:22:39 <Keefe> well, partly at least
 12 2010-10-03 00:22:41 <theymos> Yeah; that'd be terrible.
 13 2010-10-03 00:23:11 <echelon> well.. the network can determine how much the interest rate can be
 14 2010-10-03 00:23:16 <theymos> Keefe: I think it only waits 1 block to display it.
 15 2010-10-03 00:24:14 <theymos> echelon: Interest is earned when you have actually gained value on your investment. This would not be "interest", but a giant inflation subsidy.
 16 2010-10-03 00:24:16 <yebyen> ferite is italian or some other language for protected
 17 2010-10-03 00:24:21 <yebyen> limbiajul jocurilor de noroc partea a doua
 18 2010-10-03 00:24:25 <yebyen> i did not know that
 19 2010-10-03 00:25:03 <yebyen> you could build a bitcoin bank
 20 2010-10-03 00:25:05 <yebyen> that pays interest
 21 2010-10-03 00:25:17 <yebyen> but there's really no point in offering interest if you don't have a deposit
 22 2010-10-03 00:25:23 <yebyen> and mtgox is not doing it currently :)
 23 2010-10-03 00:26:52 <theymos> The bank could only pay interest if it's doing something risky with your money.
 24 2010-10-03 00:27:20 <yebyen> or if there was a government offering it bailouts all the time
 25 2010-10-03 00:29:34 <yebyen> america, fuck yeah
 26 2010-10-03 00:40:08 <echelon> lol
 27 2010-10-03 00:40:19 <nanotube> theymos: kermit: hmm interesting, all my previously-unconfirmed transactions are now confirmed, and with 244 confirmations, to boot. guess those made it over to artforz sometime earlier, then....
 28 2010-10-03 00:41:30 <ArtForz> neat
 29 2010-10-03 00:42:55 <yebyen> so if nobody on the bitcoin network is generating, no transactions get confirmed, right?
 30 2010-10-03 00:43:02 <ArtForz> yup
 31 2010-10-03 00:43:11 <yebyen> even if they're all connected
 32 2010-10-03 00:43:15 <ArtForz> sounds right, I set my fees to 0 about 3 days ago
 33 2010-10-03 00:43:22 <nanotube> ArtForz: cool. :)
 34 2010-10-03 00:43:22 <yebyen> fees?
 35 2010-10-03 00:43:24 <theymos> nanotube: I guess propagation isn't as bad as we thought.
 36 2010-10-03 00:43:31 <yebyen> how do you set your fees
 37 2010-10-03 00:43:35 <nanotube> there's still one that's unconfirmed, even though it's from 9/29
 38 2010-10-03 00:43:52 <nanotube> yebyen: you have to compile a custom client
 39 2010-10-03 00:43:58 <yebyen> well i do that
 40 2010-10-03 00:44:02 <nanotube> though ideally... that would be a cli switch. :)
 41 2010-10-03 00:44:10 <nanotube> what's everyone's thought on that ---^
 42 2010-10-03 00:44:20 <yebyen> nah
 43 2010-10-03 00:44:28 <ArtForz> main.h CTransaction::GetMinFee
 44 2010-10-03 00:44:44 <yebyen> do you only get to charge a fee if i need you to confirm my transaction for some reason?
 45 2010-10-03 00:44:53 <yebyen> or only if it's very big
 46 2010-10-03 00:44:54 <yebyen> or what
 47 2010-10-03 00:45:10 <nanotube> yebyen: there is a set of rules in the 'official' client about fees.
 48 2010-10-03 00:45:21 <Keefe> generators can require a fee to publish your txn in their blocks
 49 2010-10-03 00:45:38 <nanotube> yebyen: http://www.bitcoin.org/wiki/doku.php?id=transaction_fee
 50 2010-10-03 00:45:39 <theymos> nanotube: The problem with that is that the required fees would be unpredictable. Best to do that kind of thing only once we have just a few big generating companies that can publish their rules.
 51 2010-10-03 00:46:14 <yebyen> hmm
 52 2010-10-03 00:46:19 <yebyen> my bitcoin client is not connecting
 53 2010-10-03 00:46:22 <Keefe> if you haven't offered a big enough fee (whatever they choose to require), the generator ignores and doesn't publish your txn
 54 2010-10-03 00:46:37 <yebyen> neat
 55 2010-10-03 00:46:38 <nanotube> theymos: well... most people would still run with "defaults", so it should be fine, i think.
 56 2010-10-03 00:47:20 <yebyen> it sounds like a feature where if everyone subscribes to islam, then everyone has to pay 2.5% zakat
 57 2010-10-03 00:47:29 <theymos> nanotube: If 10% run above defaults, though, then transactions will take 10% longer, which sucks.
 58 2010-10-03 00:47:41 <yebyen> and if most people subscribe to islam, then some peoples transactions get ignored and not recognized
 59 2010-10-03 00:47:54 <Keefe> not ignored forever
 60 2010-10-03 00:47:58 <yebyen> really?
 61 2010-10-03 00:47:59 <yebyen> oh
 62 2010-10-03 00:48:03 <nanotube> theymos: but if 10% of others run below defaults... it evens out :)
 63 2010-10-03 00:48:06 <Keefe> all it takes is one person to not require a fee
 64 2010-10-03 00:48:15 <yebyen> because a random client is the one who gets to publish his solved block
 65 2010-10-03 00:48:28 <Keefe> and when that no-fee-generator eventually makes a block and includes your txn, all's good for you
 66 2010-10-03 00:48:39 <theymos> nanotube: No, that doesn't even it out. It just allows more transactions that would otherwise be impossible.
 67 2010-10-03 00:48:53 <Keefe> as long as your txn propagated thru the network to that generator
 68 2010-10-03 00:49:15 <yebyen> but if there's so much cpu power that my no-fee atom n270 can never generate a block
 69 2010-10-03 00:49:16 <nanotube> theymos: ah hehe true
 70 2010-10-03 00:49:21 <yebyen> i'm one who can't help you
 71 2010-10-03 00:49:23 <Keefe> it's not "never"
 72 2010-10-03 00:49:39 <Keefe> just extremely unlikely in a given small time range
 73 2010-10-03 00:50:06 <yebyen> it's cool that you can set a flag that says "i don't want to pay any taxes"
 74 2010-10-03 00:50:16 <Keefe> heh
 75 2010-10-03 00:50:16 <yebyen> and it can mean that your transactions don't get recognized :)
 76 2010-10-03 00:50:35 <yebyen> does that mean you would lose your bitcoins?
 77 2010-10-03 00:50:42 <yebyen> they don't come back, right
 78 2010-10-03 00:50:45 <Keefe> doesn't get published (included in a block) not that it's not recognized once published by someone
 79 2010-10-03 00:50:53 <theymos> yebyen: You could theoretically double-spend them.
 80 2010-10-03 00:50:56 <yebyen> you would have to know how to configure your client to recover them
 81 2010-10-03 00:51:14 <yebyen> now i have 7 connections and no blocks
 82 2010-10-03 00:51:20 <yebyen> hum
 83 2010-10-03 00:51:54 <Keefe> if a txn doesn't include a high enough fee, do nodes not keep that txn in memory to prevent double spending?
 84 2010-10-03 00:52:12 <ArtForz> they keep it in memory
 85 2010-10-03 00:52:12 <yebyen> not if they don't publish it
 86 2010-10-03 00:52:24 <theymos> Keefe: The network will forget about it over time.
 87 2010-10-03 00:52:44 <ArtForz> problem is, they never relay it
 88 2010-10-03 00:53:12 <Keefe> hmm, seems pointless to remember it but not relay it
 89 2010-10-03 00:53:14 <yebyen> i think that it would be cruel to write the client rule to say "i saw you try to spend that money, and you didn't pay taxes, so I didn't recognize it, but now, you don't have that money anymore."
 90 2010-10-03 00:53:16 <ArtForz> so if the originating client stops resending it, it drops off the network as clients are restarted
 91 2010-10-03 00:53:21 <theymos> Maybe that's why they don't rebroadcast it -- so it can be replaced.
 92 2010-10-03 00:53:39 <ArtForz> except replacement was disabled a few revisions back ...
 93 2010-10-03 00:53:53 <yebyen> replacement?
 94 2010-10-03 00:53:56 <ArtForz> yep
 95 2010-10-03 00:53:57 <theymos> ArtForz: Replaced because it was forgotten, I mean.
 96 2010-10-03 00:54:17 <yebyen> you mean the client was configured before to automatically recover its coins
 97 2010-10-03 00:54:27 <yebyen> if a transaction failed for long enough
 98 2010-10-03 00:54:39 <ArtForz> you used to be able to replace a in-memory TX with a one with the same inputs but different outputs
 99 2010-10-03 00:54:57 <yebyen> oh drat
100 2010-10-03 00:55:13 <theymos> Why was that ever thought to be secure? It makes it super easy to double-spend.
101 2010-10-03 00:55:23 <yebyen> i don't understand
102 2010-10-03 00:55:25 <ArtForz> no it doesnt
103 2010-10-03 00:55:28 <theymos> Why?
104 2010-10-03 00:55:43 <ArtForz> in-memory TXes are by definition in no block yet
105 2010-10-03 00:56:13 <ArtForz> the client actually using 0/unconfirmed TXes to build outgoing transactions was imo a bug
106 2010-10-03 00:56:26 <yebyen> do you know the conditions to trigger the tax in the lowest possible configuration?
107 2010-10-03 00:56:27 <theymos> It makes the "snack machine" situation impossible, though.
108 2010-10-03 00:56:29 <Keefe> so let's say i broadcast a no-fee txn to 10 peers, all of which don't like it so they don't publish and don't relay... i realize the problem and connect to a different set of peers and spend the same coins again but with a suitable fee, and that txn gets propagated and published... but won't the first 10 peers consider the block that included the second version txn to be a invalid block for double-spending?
109 2010-10-03 00:56:43 <yebyen> if there are 3 clients on the network, and one has all 170k coins for example
110 2010-10-03 00:56:58 <yebyen> and tries to send them all
111 2010-10-03 00:57:11 <yebyen> will he try to pay a tax?
112 2010-10-03 00:57:12 <ArtForz> Keefe: no, I don't think so
113 2010-10-03 00:57:19 <ArtForz> the Tx that gets into a block first wins
114 2010-10-03 00:57:32 <Keefe> ah, right
115 2010-10-03 00:58:05 <theymos> ArtForz: Peers do relay transactions that don't meet their fee requirements?
116 2010-10-03 00:58:42 <Keefe> so even with the ability to modify an unpublished txn removed from the stock client, all that one needs to do is connect to other peers and try again
117 2010-10-03 00:59:09 <Keefe> using a modified client i suppose, to allow the coins to be re-spent
118 2010-10-03 00:59:18 <ArtForz> theymos: yes
119 2010-10-03 00:59:44 <kermit> nanotube: darn, too bad i deleted my old wallets
120 2010-10-03 00:59:57 <yebyen> hg pull!
121 2010-10-03 00:59:59 <Keefe> oh? i thought we were just saying that txns aren't relayed if the fee isn't enough
122 2010-10-03 01:00:08 <yebyen> who wants to work on a bitcoin client in go
123 2010-10-03 01:00:43 <Keefe> are we talking about two different rules here?
124 2010-10-03 01:01:23 <theymos> Keefe: Peers ignore fees for the purposes of relaying transactions. (This is what I thought before, but I wasn't sure.)
125 2010-10-03 01:01:51 <Keefe> <Keefe> if a txn doesn't include a high enough fee, do nodes not keep that txn in memory to prevent double spending?
126 2010-10-03 01:01:54 <Keefe> <ArtForz> they keep it in memory
127 2010-10-03 01:01:57 <Keefe> <ArtForz> problem is, they never relay it
128 2010-10-03 01:02:10 <ArtForz> yes
129 2010-10-03 01:02:17 <ArtForz> only if the original client resends it
130 2010-10-03 01:02:25 <theymos> They relay it once, right?
131 2010-10-03 01:02:28 <ArtForz> = the one that has the tx in his wallet
132 2010-10-03 01:02:50 <ArtForz> basically a normal node never resends a TX it didn't create on it's own
133 2010-10-03 01:03:28 <Keefe> so a no-fee txn will propagate thru the whole net even if everyone requires a fee to publish?
134 2010-10-03 01:03:36 <ArtForz> yes
135 2010-10-03 01:04:12 <Keefe> ok, guess i got off on wrong tangent
136 2010-10-03 01:07:55 <theymos> Bitcoin does rebroadcast alerts, so it must have been a conscious design choice not to rebroadcast transactions.
137 2010-10-03 01:11:22 <ArtForz> of course, to stop stale TXes from clogging up the network
138 2010-10-03 01:11:59 <ArtForz> the question is, shouldnt a client also resend the not-yet-in-block TXes a TX it owns depends on when it does resendwallettransactiosn() ?
139 2010-10-03 01:13:22 <Keefe> i'd say certainly, if it's resending a txn that depends on another unpublished txn, it needs to resend both
140 2010-10-03 01:13:44 <ArtForz> well, I think currently it doesn't
141 2010-10-03 01:13:44 <Keefe> the stock client doesn't do that now?
142 2010-10-03 01:13:57 <Keefe> hey mizerydearia!
143 2010-10-03 01:14:07 <yebyen> norm!
144 2010-10-03 01:14:10 <ArtForz> wait, it does
145 2010-10-03 01:14:23 <Keefe> been missing nullvoid for a while. any way i can help?
146 2010-10-03 01:14:23 <mizerydearia> hi Keefe
147 2010-10-03 01:14:27 <ArtForz> RelayWalletTransaction
148 2010-10-03 01:14:39 <mizerydearia> Keefe, Sure!  I need to find an apartment to move to.
149 2010-10-03 01:14:59 <mizerydearia> May I move in with you?
150 2010-10-03 01:15:16 <yebyen> mizerydearia: are you in NY?
151 2010-10-03 01:15:18 <Keefe> i already live with others, so no, sorry
152 2010-10-03 01:15:26 <yebyen> i've got an empty room, it's only 325 :)
153 2010-10-03 01:15:27 <ArtForz> so... why then are there still 0/unconfirmed transactions around?
154 2010-10-03 01:15:27 <mizerydearia> Nope =/  I'm freezing in Wisconsin.
155 2010-10-03 01:15:38 <Keefe> but maybe i can host your site temporarily?
156 2010-10-03 01:15:41 <mizerydearia> yebyen, 325 bitcoins?
157 2010-10-03 01:15:54 <yebyen> mizerydearia: can you give me at least 3250?
158 2010-10-03 01:16:05 <mizerydearia> yebyen, I don't have that much =/
159 2010-10-03 01:16:15 <yebyen> well, 325 would probably take care of it
160 2010-10-03 01:16:23 <yebyen> i just got a job
161 2010-10-03 01:16:28 <yebyen> and they might keep me
162 2010-10-03 01:16:45 <yebyen> it's enough to pay for that empty room, and i'm already stuck with the electric bill :)
163 2010-10-03 01:16:50 <yebyen> i live in the attic
164 2010-10-03 01:17:04 <tsilly> Does anybody here do this with an 8800GT, that can tell me about how many hashes per second they get on the CUDA generator?
165 2010-10-03 01:17:26 <yebyen> mizerydearia: can you code in go or ferite?
166 2010-10-03 01:17:27 <tsilly> I'm trying to decide whether it is worth my time to reinstall Windows.
167 2010-10-03 01:17:31 <mizerydearia> Keefe, Well, I attempted to put the site on a vps, however, I found that there are some memory issues and due to limited memory on vps my site caused all other processes including bitbot to be killed periodically.  So, until I fix it on the vps or find a place to move to...
168 2010-10-03 01:17:48 <mizerydearia> yebyen: I haven't heard of either, but I can most likely learn.
169 2010-10-03 01:17:53 <yebyen> mizerydearia: i have a 1TB athlon 64 on RIT campus
170 2010-10-03 01:18:00 <yebyen> mizerydearia: ferite is very close to c
171 2010-10-03 01:18:03 <yebyen> but garbage collected
172 2010-10-03 01:18:08 <yebyen> and no pointers
173 2010-10-03 01:18:16 <mizerydearia> Ah, then in  that case it may be difficult for me since I do not understand c/c++.
174 2010-10-03 01:18:19 <yebyen> you get void types though :)
175 2010-10-03 01:18:47 <Keefe> mizerydearia, what do you need, besides httpd and php?
176 2010-10-03 01:18:48 <yebyen> you got object orientation down though?
177 2010-10-03 01:19:12 <yebyen> mizerydearia: it runs elive/freebsd though
178 2010-10-03 01:19:14 <yebyen> i haven't decided which
179 2010-10-03 01:19:36 <yebyen> so as long as your services are portable
180 2010-10-03 01:19:37 <yebyen> i can host em
181 2010-10-03 01:19:42 <mizerydearia> Keefe: Python and a Bitcoin server to extract data from
182 2010-10-03 01:19:54 <yebyen> mizerydearia: how about a chroot?