1 2010-10-25 00:14:59 <Diablo-D3> ArtForz: you here?
  2 2010-10-25 00:23:54 <forthright> does bitcoin use a btree or a recno Berkely DB?
  3 2010-10-25 00:25:38 <jgarzik> forthright: why is hash out of the question?  :)
  4 2010-10-25 00:25:50 <jgarzik> forthright: it uses DB_BTREE
  5 2010-10-25 00:29:15 <forthright> uh... hash is bad for file system afaik : P thanks
  6 2010-10-25 00:30:09 <jgarzik> forthright: it depends greatly on your dataset.  for a randomized dataset, an item lookup may require fewer disk reads in a hash database.
  7 2010-10-25 00:31:02 <jgarzik> forthright: on-disk, both hash and btree wind up looking quite similar to your kernel's filesystem driver.
  8 2010-10-25 00:32:28 <dgores> thanks (forthright rename)
  9 2010-10-25 00:43:55 <jgarzik> woot!  https://mtgox.com/users/sendMoney
 10 2010-10-25 00:44:27 <jgarzik> mtgox users can now send money to mtgox users (or even non-mtgox users)
 11 2010-10-25 00:45:50 <doublec> nice
 12 2010-10-25 00:46:45 <doublec> that makes mtgox much more useful
 13 2010-10-25 00:51:23 <Kiba> is it me or bitcoins got sold way below market rate at mtgox?
 14 2010-10-25 00:51:31 <Kiba> something sold for 0.0901
 15 2010-10-25 00:52:29 <doublec> sometimes there are weird trades like that
 16 2010-10-25 00:52:33 <doublec> I suspect bugs
 17 2010-10-25 00:53:38 <Kiba> mtgox got a lot of small bugs
 18 2010-10-25 00:54:15 <jgarzik> it should never be the case that fewer than ~25 bids or asks are shown, when the order book exists to support that size
 19 2010-10-25 00:54:50 <jgarzik> sometimes during large price movements, it will only display a few bids, when in reality there are a great many bids at lower prices
 20 2010-10-25 01:10:46 <AAA_awright> It would be nice if it errors, if you give an email not registered with the site
 21 2010-10-25 01:11:32 <AAA_awright> 0.13$ yikes
 22 2010-10-25 01:12:14 <AAA_awright> What's with the explosion in the last month?
 23 2010-10-25 01:34:05 <appamatto> AAA_awright I guess people are finding out
 24 2010-10-25 01:34:28 <appamatto> I think I've told about 10 people and none of them seemed ready to go out and buy BTC
 25 2010-10-25 01:38:15 <Kiba> don't know what is happening but I do see new people from time to time
 26 2010-10-25 01:40:10 <appamatto> noone in a libertarian chat I'm in had heard about BTC
 27 2010-10-25 01:41:07 <appamatto> I assume that liberty folks will form a large portion of the early adopters
 28 2010-10-25 01:41:45 <Kiba> no
 29 2010-10-25 01:41:51 <noagendamarket> I found an ecurrency provider that lets me buy liberty reserve from my post office lol
 30 2010-10-25 01:41:58 <Kiba> it's a special breed of liberty folks
 31 2010-10-25 01:42:10 <appamatto> what breed is that? :p
 32 2010-10-25 01:42:12 <Kiba> called.....
 33 2010-10-25 01:42:17 <Kiba> crypto-anarchists
 34 2010-10-25 01:42:32 <appamatto> Not sure crypto- is a necessary part of that
 35 2010-10-25 01:42:54 <noagendamarket> crypto-agorist :)
 36 2010-10-25 01:42:55 <Kiba> it's pretty good bet that most of us are hackers
 37 2010-10-25 01:43:17 <appamatto> heh
 38 2010-10-25 01:44:33 <Kiba> noagendamarket: what's taking you so long to build that web ads service?
 39 2010-10-25 01:45:06 <noagendamarket> well my developer has been moving across the state
 40 2010-10-25 01:45:18 <noagendamarket> now he is back It will be done pretty quickly
 41 2010-10-25 01:45:55 <Kiba> what does he think of bitcoin?
 42 2010-10-25 01:46:12 <Kiba> and how come you got a team of developers working for you?
 43 2010-10-25 01:47:39 <noagendamarket> lol they dont work for me
 44 2010-10-25 01:48:09 <Kiba> you say "my developer"
 45 2010-10-25 01:53:02 <Kiba> supicious
 46 2010-10-25 01:54:32 <noagendamarket> whats suspicious?
 47 2010-10-25 01:55:41 <Kiba> "my developer" and you say they don't work for me
 48 2010-10-25 02:10:58 <noagendamarket> well the developer working on the project
 49 2010-10-25 02:13:01 <Kiba> the developer working on the project waht?
 50 2010-10-25 02:14:10 <jgarzik> Ok you can send your mtgox BTC or USD to anyone else now. So people who want to withdraw funds just need to find someone that wants to add funds. This isn't the final solution of course but it should help people that need their money ASAP.
 51 2010-10-25 02:14:14 <jgarzik> </mtgox>
 52 2010-10-25 02:17:41 <nanotube> yea, saw that, jgarzik. seems like a decent stopgap, until something real like ach comes by.
 53 2010-10-25 02:18:16 <Kiba> can't ach or whatever it is can be disputed by scammers?
 54 2010-10-25 02:18:52 <OneFixt> Doesn't matter for withrawal.
 55 2010-10-25 02:19:02 <jgarzik> Kiba: MadHatter seems to think so
 56 2010-10-25 02:19:58 <jgarzik> IMO 90% of fraud prevention is avoiding hit-n-run scammers.  The other 10% are the nefarious "sleeper cells" that are smarter, and far more difficult to detect.
 57 2010-10-25 02:20:42 <Kiba> patient and impatient scammers
 58 2010-10-25 02:22:33 <nanotube> as usual... being patient makes you better at things. including scamming.
 59 2010-10-25 02:24:01 <Kiba> patience, determination, naughtyness
 60 2010-10-25 02:26:02 <noagendamarket> Ive found a way people can download a bartcode and pay for bitcoins at their local post office
 61 2010-10-25 02:26:12 <noagendamarket> *barcode
 62 2010-10-25 02:28:16 <Kiba> umm
 63 2010-10-25 02:28:27 <Kiba> it seem that mybitcoin required php?
 64 2010-10-25 02:28:31 <Kiba> I am using rails
 65 2010-10-25 02:33:09 <jgarzik> noagendamarket: that's pretty neat.  is this something kept to yourself, or gonna post it here / the forums?
 66 2010-10-25 02:37:30 <Kiba> so..did anybody try out http://soulplaying.com yet?
 67 2010-10-25 02:38:31 <noagendamarket> Ive only just discovered it.  I need to do more digging first.
 68 2010-10-25 02:39:14 <Kiba> hmm
 69 2010-10-25 02:39:24 <Kiba> it seem that mybitcoin is a bunch of php
 70 2010-10-25 02:39:40 <Kiba> and yo, I am also inexperienced with writing app that accept payment
 71 2010-10-25 02:47:06 <jgarzik> Kiba: replied on the forum.  summary: it's just HTML form values.  play around with MBC 'merchant tools', with encrypted-url feature disabled, to see what HTML form values are required.
 72 2010-10-25 02:48:39 <Kiba> well
 73 2010-10-25 02:48:45 <Kiba> this is my first time doing it
 74 2010-10-25 02:50:11 <noagendamarket> looks nice kiba
 75 2010-10-25 02:51:58 <Kiba> it's a WIP progress
 76 2010-10-25 02:52:02 <Kiba> work in progress...
 77 2010-10-25 02:53:06 <noagendamarket> bitcoin is a wip
 78 2010-10-25 02:53:10 <noagendamarket> lol
 79 2010-10-25 02:53:59 <Kiba> the money I make from that site is going to be converted to dollars so I can pay some bills and save for thing-o-matic
 80 2010-10-25 02:55:59 <appamatto> have any serious adjustments been made to bitcoin?
 81 2010-10-25 02:56:12 <appamatto> I keep hearing that it's imperfect, but I'm having trouble finding flaws :p
 82 2010-10-25 02:56:28 <jgarzik> appamatto: not really.  the basic structure presented in the paper exists in the software.
 83 2010-10-25 02:56:58 <jgarzik> appamatto: nothing's perfect.  the important thing is that flaws are known, eg. bitcoin's "over 50% of the cpu" weakness.
 84 2010-10-25 02:57:33 <appamatto> yeah, but if you check the wiki about what happens in the over 50% case, it's really not that bad
 85 2010-10-25 02:58:08 <appamatto> It's a ton of effort to go through just to be able to double-spend some bitcoins
 86 2010-10-25 02:58:22 <jgarzik> yup
 87 2010-10-25 02:59:03 <Kiba> the bigger the network, the bigger the effort
 88 2010-10-25 03:05:07 <Kiba> my life is a series of accidents that lead me to learn and accept bitcoin
 89 2010-10-25 11:50:44 <dgores> void resize(size_type n, value_type c=0)         { vch.resize(n + nReadPos, c); }
 90 2010-10-25 11:50:45 <dgores> void reserve(size_type n)                        { vch.reserve(n + nReadPos); }
 91 2010-10-25 11:50:59 <dgores> net.cpp:    vRecv.resize(nPos + nBytes);
 92 2010-10-25 11:51:07 <vezeena> nice
 93 2010-10-25 11:51:10 <vezeena> sexy code
 94 2010-10-25 11:51:39 <dgores> ... thanks.  so whats the deal there with the resize?
 95 2010-10-25 11:51:47 <dgores> (and reserve?)
 96 2010-10-25 11:53:35 <dgores> ... never mind I figured it out
 97 2010-10-25 12:18:44 <jgarzik> <gribble> BCM|   NEW TRD|PPUSD                  1000 @ $0.1000
 98 2010-10-25 12:18:48 <jgarzik> impressive
 99 2010-10-25 12:20:34 <MacRohard> yea.. bitcoin economy is nominally >$500k now
100 2010-10-25 12:21:07 <MacRohard> 524850 USD
101 2010-10-25 12:21:09 <MacRohard> heh
102 2010-10-25 12:36:44 <vezeena> this is sick amounts of inflation
103 2010-10-25 12:36:58 <vezeena> $1 = 8.3 btc
104 2010-10-25 12:38:09 <vezeena> hmm cant change my nick for some reason (genjix)
105 2010-10-25 12:38:20 <Malouin_> hehe
106 2010-10-25 12:38:33 <vezeena> a few days ago it was $1 = 15 btc
107 2010-10-25 12:40:21 <vezeena> i wonder when the bubble will burst
108 2010-10-25 12:40:30 <edcba> if ?
109 2010-10-25 12:40:55 <vezeena> it cant inflate forever at astronomical rates
110 2010-10-25 12:41:06 <edcba> why not ? :)
111 2010-10-25 12:41:22 <edcba> won't look astronomical after a while :)
112 2010-10-25 12:41:53 <MacRohard> deflation
113 2010-10-25 12:42:00 <vezeena> yep deflate sorry
114 2010-10-25 13:07:08 <jgarzik> just used intexchange.com.  easy, quick LREUR->LRUSD trade.
115 2010-10-25 13:12:04 <Diablo-D3> jgarzik: ooh
116 2010-10-25 13:12:08 <Diablo-D3> I never thought of that
117 2010-10-25 13:12:42 <jgarzik> one could also use intexchange.com to get liqpay for mtgox
118 2010-10-25 13:16:48 <Teppy> I think that Bitcoins will hyperdeflate (like they're doing now) until the number of users reaches saturation. So I'd imagine that PayPal growth (new accounts) has slowed, because that's probably reached saturation.
119 2010-10-25 13:18:00 <Teppy> Then it should go into a slight deflation mode, roughly approximating energy usage.
120 2010-10-25 13:18:12 <Teppy> (Energy usage across the globe.)
121 2010-10-25 13:28:11 <Diablo-D3> Teppy: "hyperdeflate"
122 2010-10-25 13:28:21 <Diablo-D3> you have no clue how much I like the sound of that word
123 2010-10-25 13:28:36 <Diablo-D3> it makes me warm and fuzzy inside... its almost disgusting, really
124 2010-10-25 13:28:59 <Malouin_> hahaha
125 2010-10-25 14:00:59 <dwdollar1> I f-ing hate bankers.
126 2010-10-25 14:02:33 <jgarzik> http://www.bitcoin.org/smf/index.php?topic=1561.0
127 2010-10-25 14:02:35 <bitbot> Buy and Sell Mt Gox USD
128 2010-10-25 14:02:47 <jgarzik> To help out mtgox, and hopefully move mtgox closer to permitting withdrawals, I am offering to buy some MTGOX-USD using PayPal funds.
129 2010-10-25 14:03:30 <jgarzik> dwdollar1: you're a banker now, too, ya know ;-)
130 2010-10-25 14:03:45 <dwdollar1> the assholes just denied my ACH
131 2010-10-25 14:04:04 <dwdollar1> access
132 2010-10-25 14:04:27 <jgarzik> dwdollar1: for BCM?
133 2010-10-25 14:04:58 <dwdollar1> yeah
134 2010-10-25 14:05:14 <dwdollar1> well, the new website I'm working on
135 2010-10-25 14:05:16 <jgarzik> dwdollar1: FirstACH or some other provider?
136 2010-10-25 14:05:24 <dwdollar1> yeah, FirstACH
137 2010-10-25 14:05:34 <dwdollar1> I might try the other one...
138 2010-10-25 14:06:14 <dwdollar1> ACHDirect
139 2010-10-25 14:06:15 <Teppy> What is the other website?
140 2010-10-25 14:06:50 <dwdollar1> I've been working on a new one that would have escrow and instant trades.
141 2010-10-25 14:07:00 <Teppy> Ah, that will be nice.
142 2010-10-25 14:07:19 <Teppy> Eric says that it's OK to talk about his website in here...
143 2010-10-25 14:07:56 <dwdollar1> They jerked me around for two weeks acting like it was going to be ok.
144 2010-10-25 14:08:01 <Teppy> So the big problem I've had with Dragon's Tale is that people want to try it but can't get Bitcoins if all they have is a credit card.
145 2010-10-25 14:08:17 <Teppy> And I can't do the credit card stuff because then that gets really close to gambling laws.
146 2010-10-25 14:08:51 <jgarzik> Teppy: I think most sane merchants will take "paypal or bitcoin" or "CC or bitcoin" etc.  Though I agree that gambling throws a big monkey wrench into such plans.
147 2010-10-25 14:08:51 <Teppy> So Eric is setting up a site that has one purpose: Buying Bitcoins (and then selling them back later) instantly, with credit cards.
148 2010-10-25 14:09:13 <jgarzik> Teppy: how will he avoid buybitcoins.com problem?
149 2010-10-25 14:09:25 <Teppy> Which specific problem?
150 2010-10-25 14:09:49 <jgarzik> Teppy: chargebacks, usually due to stolen credit card
151 2010-10-25 14:10:35 <Teppy> Ah - here's how: First off, eliminating chargebacks entirely won't happen. So we designed a system to reduce them by a lot:
152 2010-10-25 14:10:40 <jgarzik> Teppy: go to https://buybitcoins.com/
153 2010-10-25 14:11:55 <Teppy> The site will ask for (in addition to the normal stuff), the card issuing bank's 1-800 number on the back, and the phone number linked to that card.
154 2010-10-25 14:12:10 <Teppy> The Visa/Mastercard system doesn't track that phone number, but the issuing bank does.
155 2010-10-25 14:12:54 <jgarzik> Teppy: you verify that information manually, I guess?
156 2010-10-25 14:13:25 <Teppy> After doing a successful preauth, he'll call the issuing bank, identify himself as a merchant, verify that the card # is linked to that account, and that the associated phone# hasn't changed in the last 30 days.
157 2010-10-25 14:14:05 <Teppy> Then he'll call the phone# on a recorded line and ask for a 4-digit secret code that the website provided after the successful preauth.
158 2010-10-25 14:14:40 <jgarzik> Teppy: sounds decent.  That should help eliminate 80% of fraud, hit-n-run.  Then there's the other 20%...  http://www.bitcoin.org/smf/index.php?topic=1559.msg18495#msg18495
159 2010-10-25 14:14:43 <bitbot> Suggestion for dwdollar & other BTC traders: How to fix the SCAM problem : jgarzik: No system is foolproof, but ShadowOfHarbringer's suggestions are good.  That should eliminate 80% of fraud, which is "hit-and-run" scammers with stolen accounts, who must liquidate in a hurry before account owner notices.  The remaining frauds are <ul style="margin-top: 0; margin-bot...
160 2010-10-25 14:16:35 <Teppy> jgarzik: That doesn't solve the problem of normal people being able to pay instantly with a credit card.
161 2010-10-25 14:17:00 <Teppy> And I agree, it only hits 80% of the problem.
162 2010-10-25 14:17:24 <Teppy> The solution I just described is very close to foolproof.
163 2010-10-25 14:17:37 <Teppy> IOW, how would you compromise the system I just described?
164 2010-10-25 14:18:26 <jgarzik> Teppy: someone needs to figure out how to get ATM access to Visa/MC cards.  Average people can withdraw cash via their CCs at most worldwide ATMs...  why can't we do the same for bitcoins, I wonder?
165 2010-10-25 14:19:03 <jgarzik> Teppy: no system is foolproof, but it does seem reasonably secure, if labor intensive.
166 2010-10-25 14:20:33 <Teppy> Western Union and (I forget the other one) do the equivalent of ATM access, but you can still do a chargeback (they eat those), and their fees are 15% or more, AND you pay a cash advance fee to your bank.
167 2010-10-25 14:21:27 <Teppy> Also, the recipient must pick up the money in person, which discourages fraud.
168 2010-10-25 14:22:42 <Teppy> There's one other cool feature of this system:
169 2010-10-25 14:23:25 <Teppy> Have you tried logging in to Dragon's Tale?
170 2010-10-25 14:24:13 <jgarzik> no
171 2010-10-25 14:24:22 <Teppy> You create a character, and then click "Login." It notices that you don't have BTC and gives 3 choices:
172 2010-10-25 14:24:43 <Teppy> 1. I Already Have some Bitcoins. Tell me where to send them."
173 2010-10-25 14:25:01 <Teppy> 2. "I just sent some Bitcoins. Check my balance again."
174 2010-10-25 14:25:06 <Teppy> 3. "Buy Bitcoins"
175 2010-10-25 14:25:47 <Teppy> If you select "Buy Bitcoins", then it launches a web browser over to bitcoingateway.com, and fills in a field with your Dragon's Tale Bitcoin address.
176 2010-10-25 14:26:09 <Teppy> So normal users don't even have to *see* a Bitcoin address.
177 2010-10-25 14:26:26 <Teppy> The popup itself says "Dragon's Tale uses Bitcoins as Chips"
178 2010-10-25 14:26:39 <Teppy> So that's a concept that anyone can wrap their head around.
179 2010-10-25 14:27:17 <jgarzik> definitely
180 2010-10-25 14:28:02 <Teppy> And then if other sites want to do a similar service, I can just randomly pick one when someone clicks "Buy Bitcoins"
181 2010-10-25 14:29:22 <Teppy> In fact, dwdollar1, if your new site has some way for normal users to instantly get Bitcoins with, say, PayPal, I could have the popup give options "Buy Bitcoins with PayPal", "Buy Bitcoins with a Credit Card", etc.
182 2010-10-25 14:30:25 <Teppy> But I want all the sites in the rotation to be dead-simple, users fill in some information and Bitcoins show up at Dragon's Tale.
183 2010-10-25 14:36:58 <UukGoblin> 172324 <+gribble> MTG|     TRADE|                     40.933 @ $0.9
184 2010-10-25 14:37:02 <UukGoblin> that happened again ;-]
185 2010-10-25 14:55:17 <jgarzik> weird
186 2010-10-25 14:55:32 <jgarzik> I simply do not understand why satoshi pushes so hard against multiple bitcoin instances on the same computer.
187 2010-10-25 14:55:46 <jgarzik> he is now creating a multiple-account system _inside_ bitcoin, to avoid this.
188 2010-10-25 15:01:27 <nanotube> jgarzik: what's a multiple-account system?
189 2010-10-25 15:01:44 <dwdollar> Teppy:  I'm going to stay away from PayPal, except for the person - to - person trading.
190 2010-10-25 15:01:47 <jgarzik> nanotube: multiple wallets
191 2010-10-25 15:02:21 <nanotube> jgarzik: ah hmm ic. yea interesting. i'd have thought just allowing multiple instances would be much easier
192 2010-10-25 15:02:25 <jgarzik> nanotube: "foo" has 1000.0 BTC and keys A, B, C.  "bar" has 1234.99 BTC and keys D, E, F.
193 2010-10-25 15:02:29 <jgarzik> nanotube: me too
194 2010-10-25 15:04:31 <Teppy> dwdollar: Any instant (or, say, within 10 minutes) ways to get Bitcoins for someone not yet known to your site? (Bank transfer? Don't know how those work.)
195 2010-10-25 15:05:22 <Teppy> I figure that nobody plans way ahead to gamble :)
196 2010-10-25 15:16:56 <dwdollar> Teppy:  Unfortunately, I don't think I can have anything that quick.  Adding a time delay for deposits helps combat fraud.
197 2010-10-25 15:19:11 <nanotube> jgarzik: would be nice though if he adds ability to export individual keys, and to view balances associated with individual keys.
198 2010-10-25 15:20:04 <ArtForz> actually that should be quite trivial to bolt on ...
199 2010-10-25 15:21:30 <nanotube> one would think so. :) it just hasn't been done.
200 2010-10-25 15:22:28 <ArtForz> modifying the send routine to allow specifying which keys to use as inputs would be more involved, but still not really 'hard'
201 2010-10-25 15:46:00 <dust__> some btc sold for .9 USD on mtgox!?
202 2010-10-25 15:51:40 <nanotube> dust__: yea, not the first time, either. seems there are some bugs in bid/ask matching code on mtg
203 2010-10-25 15:52:10 <dust__> so did that trade actually go through or is it just a bug in reporting?
204 2010-10-25 15:53:29 <nanotube> it must have gone through - it's up to the counterparties to report the bug or not...
205 2010-10-25 15:53:50 <nanotube> jgarzik: bitcoinwatch says '1 US dollar is worth 2.00 BTC.' which seems to be quite incorrect...
206 2010-10-25 16:18:13 <kencausey> less than 50BTC at $0.90 is interesting but not exactly earthshaking.  It looks like the market is simply clearing, although I'm not sure things are working right as I have an offer that's not showing up in the depth chart.
207 2010-10-25 16:22:17 <kencausey> In fact it's higher than the lowest bid showing, I guess I'll email mtgox
208 2010-10-25 16:25:44 <kencausey> actually nevermind, there are no bids shown at all, it's all asks near 0.90 and above.  The culling logic is throwing things off I guess.
209 2010-10-25 16:25:50 <kencausey> I am emailing mtgox though
210 2010-10-25 16:28:35 <nanotube> yea, please do kencausey it def looks like there are some bugs to work out
211 2010-10-25 16:35:05 <kencausey> Already sent
212 2010-10-25 16:39:26 <jgarzik> nanotube: it's an average between BCM and mtgox prices
213 2010-10-25 16:41:09 <jgarzik> nanotube: ah, it's due to mtgox's "last trade == 0.9"
214 2010-10-25 16:51:00 <nanotube> jgarzik: ah ic
215 2010-10-25 17:12:45 <kencausey> mtgox depth of market is more sane now
216 2010-10-25 17:12:51 <kencausey> somewhat
217 2010-10-25 17:35:27 <ArtForzZz> played a bit more with my bitcoin-double-sha256 vhdl core
218 2010-10-25 17:35:38 <Diablo-D3> heh
219 2010-10-25 17:35:42 <Diablo-D3> ArtForz: so, tell me something
220 2010-10-25 17:35:47 <Diablo-D3> ArtForz: how exactly do the miners work?
221 2010-10-25 17:36:11 <ArtForz> erm... what do you mean?
222 2010-10-25 17:37:09 <Diablo-D3> well
223 2010-10-25 17:37:36 <Diablo-D3> that python script that gives me the hash and the block header
224 2010-10-25 17:37:43 <Diablo-D3> how did the miner produce that to begin with?
225 2010-10-25 17:38:18 <ArtForz> well, block header is 80 bytes
226 2010-10-25 17:39:24 <ArtForz> 4 byte version, 32 byte hash of prev block, 32 byte merkle root hash of transactions in block, 4 byte unix timestamp, 4 byte nBits (= target), 4 byte nonce
227 2010-10-25 17:40:46 <ArtForz> each miner thread has it's own pubkey in the coinbase transaction, so each miner works on a slightly different set of transactions -> each miner has a different hashmerkleroot in the block header it's working on
228 2010-10-25 17:41:11 <ArtForz> version is constant, hash of prev block is... the hash of the previous block
229 2010-10-25 17:41:32 <ArtForz> nTime is simply the current UTC unix timestamp
230 2010-10-25 17:41:43 <Diablo-D3> yes, but what exactly is...
231 2010-10-25 17:41:56 <Diablo-D3> ArtForz: okay
232 2010-10-25 17:42:14 <Diablo-D3> Im more confused
233 2010-10-25 17:42:18 <ArtForz> nBits is a slightly weird compact representastion of the target value
234 2010-10-25 17:42:18 <Diablo-D3> than I was several minutes ago
235 2010-10-25 17:42:28 <Diablo-D3> WHAT is the miner doing
236 2010-10-25 17:42:35 <ArtForz> nNonce is whast each miner increases by 1 for each hash
237 2010-10-25 17:42:42 <ArtForz> sha256(sha256(block headeR))
238 2010-10-25 17:42:49 <ArtForz> and compare the result to the target value
239 2010-10-25 17:43:10 <Diablo-D3> yes, but where does the sequential checking come in
240 2010-10-25 17:43:27 <ArtForz> you start with nonce = 0 or whatever
241 2010-10-25 17:43:39 <ArtForz> do the hsahs compare against target, <= target, YAY, block found
242 2010-10-25 17:43:49 <ArtForz> > target, incearse nonce by 1, hash again
243 2010-10-25 17:44:31 <ArtForz> every few 1000 hashes check if a second elapsed, if yes update nTime and reset nonce to 0
244 2010-10-25 17:44:39 <ArtForz> or 1, or 0xdeadbeef or whatever
245 2010-10-25 17:44:51 <ArtForz> doesn't matter one bit
246 2010-10-25 17:45:00 <Diablo-D3> okay so
247 2010-10-25 17:45:08 <Diablo-D3> you know that getwork patch m0's miner uses?
248 2010-10-25 17:45:12 <Diablo-D3> what does it give me
249 2010-10-25 17:46:15 <ArtForz> I think the block header and the state after SHA256ing the first 64 byte block of the header
250 2010-10-25 17:46:18 <Diablo-D3> because if it gives me everything I need but the 4 bytes at the end, Im golden
251 2010-10-25 17:46:41 <ArtForz> sec
252 2010-10-25 17:46:49 <ArtForz> 'block' = block header
253 2010-10-25 17:47:24 <ArtForz> 'state' = sha256 state after SHA256 over first 64 bytes of header
254 2010-10-25 17:47:35 <ArtForz> 'target' = hash target
255 2010-10-25 17:48:42 <Diablo-D3> okay so
256 2010-10-25 17:49:01 <Diablo-D3> 64 bytes, then append time, target, and nonce, and hash the rest of the way?
257 2010-10-25 17:49:07 <ArtForz> the 'state' ius just a optimzation as the first 64 bytes of the block only change when transactions are added or a work on a new block starts
258 2010-10-25 17:49:33 <Diablo-D3> wait, damnit
259 2010-10-25 17:49:43 <Diablo-D3> m0's shit doesnt use the state
260 2010-10-25 17:49:46 <Diablo-D3> I dont think
261 2010-10-25 17:49:48 <Diablo-D3> or maybe it does
262 2010-10-25 17:49:59 <ArtForz> it does
263 2010-10-25 17:50:09 <Diablo-D3> okay so
264 2010-10-25 17:50:12 <Diablo-D3> I take the state
265 2010-10-25 17:50:23 <Diablo-D3> then continue on round 1 where it left off?
266 2010-10-25 17:50:26 <ArtForz> it also has the optimization of precalcing the first 3 rounds of sha256 of the last 16 bytesd of the header
267 2010-10-25 17:50:51 <ArtForz> as we only iterate nNonce = last 4 bytes of header in the inner miner loop
268 2010-10-25 17:52:03 <Diablo-D3> yeah but
269 2010-10-25 17:52:35 <Diablo-D3> okay, all those sharound things? are they just done on the next byte, or does each round consume another byte?
270 2010-10-25 17:52:51 <ArtForz> sharound "consumes" one dword
271 2010-10-25 17:53:16 <Diablo-D3> you mean an int?
272 2010-10-25 17:53:26 <ArtForz> whatever,  bytes
273 2010-10-25 17:53:28 <ArtForz> *4
274 2010-10-25 17:53:34 <ArtForz> those 3 sharounds calc the sha256 state after first 3 rounds of 2nd part of header
275 2010-10-25 17:53:46 <ArtForz> sha256 operates on 64 byte blocks
276 2010-10-25 17:54:10 <ArtForz> so the 80 byte header is padded to 128 bytes = 2 sha256 rounds
277 2010-10-25 17:55:17 <ArtForz> sha256 is pretty damn simple
278 2010-10-25 17:57:28 <ArtForz> anyways, my current VHDL miner code should now get ~10Mh/s on a Altera EP3C25
279 2010-10-25 17:58:43 <ArtForz> even with only chip cost thats > 4x $/Mh compared to a 5970
280 2010-10-25 17:58:57 <Malouin_> ouch
281 2010-10-25 17:59:04 <Diablo-D3> heh
282 2010-10-25 17:59:07 <ArtForz> and we still need PCB, power supply, interface, ...
283 2010-10-25 18:00:31 <ArtForz> I dont have a physical Altera chip of that gen, so no clue on power efficiency
284 2010-10-25 18:00:41 <Diablo-D3> ArtForz: so what is the shader supposed to output when its done, 8 ints?
285 2010-10-25 18:01:21 <ArtForz> no lcue
286 2010-10-25 18:01:46 <ArtForz> but considering we'd need about 64 of em to match a single OCed 5970 @ 370W ...
287 2010-10-25 19:07:57 <Teppy> Just a quick poll - which OS does everyone in here use on their desktop?
288 2010-10-25 19:09:07 <Diablo-D3> debian for the past decade.
289 2010-10-25 19:09:31 <ArtForz> debian sid
290 2010-10-25 19:12:16 <soultcer> Debian, thinking of switching to ubuntu
291 2010-10-25 19:12:20 <Diablo-D3> heh, Ive been using debian for so long, sid didnt exist yet
292 2010-10-25 19:12:23 <Diablo-D3> soultcer: do it, and I will kill you
293 2010-10-25 19:12:43 <ArtForz> well, I think my first debian desktop was woody
294 2010-10-25 19:12:50 <ArtForz> or was it potato
295 2010-10-25 19:12:52 <soultcer> I installed Ubuntu on one of my VPSes and I love how up-to-date the packages are
296 2010-10-25 19:13:14 <ArtForz> potato
297 2010-10-25 19:13:28 <ArtForz> later dist-upgraded to woody
298 2010-10-25 19:14:05 <ArtForz> debian unstable also has surprisingly up-to-date packages
299 2010-10-25 19:14:13 <Diablo-D3> heh, I had slink
300 2010-10-25 19:14:17 <Diablo-D3> then went to potato unstable
301 2010-10-25 19:14:45 <Diablo-D3> when potato went stable, unstable became woody
302 2010-10-25 19:15:02 <Diablo-D3> and then when woody went stable, unstable became perma-sid
303 2010-10-25 19:15:02 <soultcer> So when did they introduce sid?
304 2010-10-25 19:15:07 <Diablo-D3> bam
305 2010-10-25 19:15:13 <ArtForz> sounds right
306 2010-10-25 19:15:29 <soultcer> I started after woody. Etch if I recall right
307 2010-10-25 19:17:45 <ArtForz> I relal don't like ubuntu
308 2010-10-25 19:18:06 <Diablo-D3> ubuntu needs to fucking die
309 2010-10-25 19:18:25 <ArtForz> debian sid with no non-x86 archs and random UI changes FNAR
310 2010-10-25 19:18:38 <Zarutian> Diablo-D3: why? Is it a 'hot-pocket' distro?
311 2010-10-25 19:19:17 <Diablo-D3> a what?
312 2010-10-25 19:19:22 <ArtForz> I *like* being able to run debian on a SGI Indy
313 2010-10-25 19:19:36 <Diablo-D3> I like being able to run a distro not requiring me to kill devs
314 2010-10-25 19:21:35 <ArtForz> yes, Indy, as in, 100MHz MIPS CPU, 128MB ram, 2 * 4.3GB SCSI HDs
315 2010-10-25 19:22:26 <Diablo-D3> lol
316 2010-10-25 19:23:09 <ArtForz> it actually is pretty usable considering its specs
317 2010-10-25 19:26:58 <soultcer> Debian on my NSLU2 rocks too
318 2010-10-25 19:27:16 <Keefe> Diablo-D3: devs as in developers?
319 2010-10-25 19:27:37 <Zarutian> what I hate about todays software is how bloated it has becomen
320 2010-10-25 19:27:41 <Malouin_> Win 7 Ultimate 64bit
321 2010-10-25 19:27:45 <Malouin_> *hiding himself*
322 2010-10-25 19:27:56 <Keefe> i'm still a windows desktop user, but i like debian for non-desktop stuff
323 2010-10-25 19:28:32 <Keefe> such as the home gateway (old machine running debian), vps's, now a dedicated bitcoin miner
324 2010-10-25 19:29:44 <Keefe> it's not that i dislike linux for desktop so much as that i'mjust happy with XP
325 2010-10-25 19:30:03 <Keefe> might try Win7 dsomeday
326 2010-10-25 19:30:06 <ArtForz> I stuck with 2k before switching to linux
327 2010-10-25 19:32:05 <ArtForz> hrmmm... actually what were significant changes from 2k to XP, except for wlan support and the bubblegum theme?
328 2010-10-25 19:33:55 <ArtForz> right, themes, fast user switching, cleartype, WLAN and PPPoE support
329 2010-10-25 19:34:25 <bonsaikitten> haha debian
330 2010-10-25 19:34:31 <bonsaikitten> that's funny
331 2010-10-25 19:35:05 <Diablo-D3> [05:34:54] --- [bonsaikitten] (~quassel@gentoo/developer/bonsaikitten)
332 2010-10-25 19:35:06 <Diablo-D3> fail
333 2010-10-25 19:35:11 <bonsaikitten> what, I'm lazy
334 2010-10-25 19:35:22 <bonsaikitten> can't be bothered to unf*ck the init scripts so I can use a machine
335 2010-10-25 19:35:52 <Diablo-D3> I was using debian before gentoo existed, and I will be using debian long after that gentoo fad goes away
336 2010-10-25 19:37:57 <ArtForz> no big miracle
337 2010-10-25 19:38:17 <bonsaikitten> Diablo-D3: I'll take you seriously once debian grows a package manager
338 2010-10-25 19:38:20 <ArtForz> debian is nearly as old as slack
339 2010-10-25 19:38:20 <bonsaikitten> :)
340 2010-10-25 19:39:15 <Diablo-D3> debian's package manager predates gentoo.
341 2010-10-25 19:39:32 <ArtForz> by quite a bit I think
342 2010-10-25 19:39:34 <soultcer> Wait, the gentoo guy says debian needs a package manger?
343 2010-10-25 19:41:46 <soultcer> Whoops, now he's silent.
344 2010-10-25 19:41:47 <soultcer> Guess he's still compiling his answer.
345 2010-10-25 19:42:05 <ArtForz> hehehe
346 2010-10-25 19:43:01 <ArtForz> not to mention debian had sane package dependency and conflict handling while most other distros were still stuck in RPM dependency hell
347 2010-10-25 19:46:37 <Diablo-D3> ArtForz: so wait
348 2010-10-25 19:46:42 <Diablo-D3> how the fuck do I test if my miner works
349 2010-10-25 19:46:49 <ArtForz> magic!
350 2010-10-25 19:46:57 <ArtForz> or you write a simple test framework
351 2010-10-25 19:58:33 <bonsaikitten> Diablo-D3: debian still can't handle downgrades, and the package manager things it's an rc system ...
352 2010-10-25 19:58:39 <bonsaikitten> but I guess people like playing games
353 2010-10-25 19:59:04 <bonsaikitten> soultcer: I'm in too many places on IRC :) causes some lag every now and then
354 2010-10-25 19:59:24 <soultcer> ;-)
355 2010-10-25 20:23:50 <Keefe> ArtForz: have you ever watched the execution time stats for opencl kernels, using clGetEventProfilingInfo? i wonder why it's not constant.
356 2010-10-25 20:26:09 <ArtForz> not really
357 2010-10-25 20:26:50 <noagendamarket_> #bitcoin-otc
358 2010-10-25 20:27:05 <ArtForz> I just use gettimeofday and # of runs to get avergae execution time
359 2010-10-25 20:28:26 <Keefe> i was hoping it would be constant so i could figure out the call overhead
360 2010-10-25 20:29:06 <ArtForz> as time spent in the opencl runtime and transfer time is also a large factor I rather trust wall clock time taken for X runs over internal counters
361 2010-10-25 20:29:09 <Keefe> but it varies by at leas a couple percent
362 2010-10-25 20:29:27 <ArtForz> yup
363 2010-10-25 20:30:00 <Keefe> i've been counting actual hashes done per wall clock time, to determine mhps
364 2010-10-25 20:30:30 <ArtForz> yeah, that works best I think
365 2010-10-25 20:31:15 <Keefe> when you said you were doing 340 mhps/core @ 910, was that rounded to nearest 10?
366 2010-10-25 20:31:24 <ArtForz> nope
367 2010-10-25 20:32:56 <ArtForz> actual average over >24h is 339973 kh/s
368 2010-10-25 20:34:48 <Keefe> i've never seen aticonfig show more than "99%" gpu usage. is it even possible to see 100% from aticonfig?
369 2010-10-25 20:34:57 <ArtForz> I don't think so
370 2010-10-25 20:39:38 <Keefe> does anyone know if some (or all) exchangers are able to deposit and withdraw USD from Liberty Reserve directly, at a 1:1 rate less possible fees?
371 2010-10-25 20:40:06 <Keefe> noagendamarket_? your recent forum post got me thinking about that
372 2010-10-25 20:40:24 <jgarzik> Keefe: poke nanotube, I think he did research on becoming an LR exchanger
373 2010-10-25 20:40:41 <jgarzik> Keefe: it sounded like 1:1, once identity verification checks panned out, IIRC
374 2010-10-25 20:41:12 <jgarzik> probably have some sort of minimum transaction size
375 2010-10-25 20:41:26 <Keefe> ok, makes me feel better about LR. that means it's not a totally separate currency from USD only psychologically related
376 2010-10-25 20:42:45 <Keefe> without exchangers being able to do that, it would be like trading old us silver certificates with no ability to actually redeem them for silver from the us treasury
377 2010-10-25 20:43:54 <noagendamarket_> I can deposit to LR at the post office
378 2010-10-25 20:44:20 <noagendamarket_> https://e-forexgold.com/efx2/exchange_rates
379 2010-10-25 20:44:26 <noagendamarket_> I use that site
380 2010-10-25 20:44:57 <jgarzik> Keefe: on exchangezone, I've done 1:1 LR:PP transactions (less PP fees)
381 2010-10-25 20:45:26 <Keefe> that's not my point...
382 2010-10-25 20:45:31 <jgarzik> sometimes it seems like LRUSD is worth more than PPUSD on exchangezone, though
383 2010-10-25 20:45:38 <jgarzik> I wonder if that's due to chargebacks
384 2010-10-25 20:45:53 <Keefe> it's one thing to trade LRUSD for other USD with another third party
385 2010-10-25 20:46:18 <Keefe> it's another to actually be able to "redeem" LRUSD for USD from Liberty Reserve themselves
386 2010-10-25 20:47:16 <Keefe> or put another way, is the amount of LRUSD in circulation constant? or is there a means to increase or decrease circulation quantity
387 2010-10-25 20:50:56 <Diablo-D3> [05:58:34] <bonsaikitten> Diablo-D3: debian still can't handle downgrades, and the package manager things it's an rc system ...
388 2010-10-25 20:50:59 <Diablo-D3> since when?
389 2010-10-25 20:51:13 <Keefe> Liberty Reserve claims that LRUSD is backed by USD. that should imply that at least some people are able to increase circulation of LRUSD by sending USD (such as from a traditional bank account) to Liberty Reserve in exchange for a credit to their LR account
390 2010-10-25 20:51:22 <Keefe> and vice versa
391 2010-10-25 20:51:39 <bonsaikitten> Diablo-D3: since as long as I can remember back
392 2010-10-25 20:51:41 <Keefe> and we know that the general public cannot do so
393 2010-10-25 20:51:58 <bonsaikitten> no, auto-starting or restarting daemons is not the correct thing to do for a package manager.
394 2010-10-25 20:52:14 <bonsaikitten> see the puppet init script on debian as an example why that's crazy :)
395 2010-10-25 20:53:04 <Diablo-D3> I do downgrades frequently
396 2010-10-25 20:53:06 <Diablo-D3> so uh
397 2010-10-25 20:53:07 <Diablo-D3> wtf?
398 2010-10-25 20:53:15 <Diablo-D3> I suspect you just dont know the syntax
399 2010-10-25 20:53:27 <Keefe> noagendamarket_: it appears that e-forexgold is like other exchangers, where you're just trading LRUSD/USD with the exchange and not with Liberty Reserve. am i correct?
400 2010-10-25 20:53:34 <bonsaikitten> I guess that's why there are packages with names like "qt-4.7-really-4.4.2" :)
401 2010-10-25 20:54:15 <bonsaikitten> which is awesomely hilarious, but ... then it gets better with the random strings of letters after that :)
402 2010-10-25 20:54:21 <noagendamarket_> No on e can exchange directly with liberty reserve
403 2010-10-25 20:54:22 <Diablo-D3> bonsaikitten: thats due to uncaught API changes and its too late to fix
404 2010-10-25 20:54:30 <Diablo-D3> apt-get wont rollback versions automatically
405 2010-10-25 20:54:32 <bonsaikitten> Diablo-D3: so downgrade ... oh wait. derp :D
406 2010-10-25 20:54:33 <Diablo-D3> it WILL do so when ordered
407 2010-10-25 20:54:57 <bonsaikitten> I do wonder what happens when the real 4.7 is released ... oh, uhm, extra-derp!
408 2010-10-25 20:55:15 <noagendamarket_> The only way to trade LR is through exchangers
409 2010-10-25 20:55:47 <Diablo-D3> bonsaikitten: nothing
410 2010-10-25 20:55:58 <Diablo-D3> it works as normal
411 2010-10-25 20:55:59 <bonsaikitten> but there's a package named 4.7
412 2010-10-25 20:56:00 <Keefe> not even the big exchangers can exchange directly with Liberty Reserve? so LRUSD is not truly backed by USD and the quantity of LRUSD in circulation is constant?
413 2010-10-25 20:56:06 <T_X-> wow, what the fuck is wrong with mtgox!
414 2010-10-25 20:56:09 <Diablo-D3> there is a PACKAGE named 4.7
415 2010-10-25 20:56:11 <bonsaikitten> so ... now I must not depend on packages claiming to be 4.7
416 2010-10-25 20:56:11 <Diablo-D3> not a version
417 2010-10-25 20:56:28 <Diablo-D3> bonsaikitten: you CANT depend on packages... you're not a debian maint.
418 2010-10-25 20:56:29 <bonsaikitten> because 4.7 is 4.4 and 4.7 at the same time
419 2010-10-25 20:56:45 <T_X-> no more bids left, wow
420 2010-10-25 20:56:52 <Diablo-D3> only debian maints can create debs and expect anything they do to matter.
421 2010-10-25 20:56:53 <bonsaikitten> Diablo-D3: I hope my example of random insanity shows you why I prefer having a package manager instead
422 2010-10-25 20:56:53 <Keefe> T_X-: it's just a display problem
423 2010-10-25 20:56:56 <Diablo-D3> this isnt redhat
424 2010-10-25 20:57:01 <T_X-> ah, okay, hmm
425 2010-10-25 20:57:06 <Diablo-D3> bonsaikitten: uh, but apt-get IS a package manager you fucking idiot
426 2010-10-25 20:57:13 <bonsaikitten> I have been able to do proper downgrades since ca. 2000
427 2010-10-25 20:57:21 <T_X-> I already was in bed, but that got me straight back to my laptop for checking :D
428 2010-10-25 20:57:22 <Diablo-D3> no wonder Gentoo sucks if all Gentoo devs are fucking morons like you
429 2010-10-25 20:57:23 <bonsaikitten> Diablo-D3: as much as tar is a package manager
430 2010-10-25 20:57:30 <T_X-> Keefe: thanks for the info
431 2010-10-25 20:57:31 <ArtForz> and it was a package manager before portage even existed...
432 2010-10-25 20:57:48 <bonsaikitten> you refuse to see structural flaws as issues, because ... ur not a dev lulz
433 2010-10-25 20:58:12 <bonsaikitten> oh, one good thing. debian bash doesn't come pre-broken anymore
434 2010-10-25 20:58:19 <Diablo-D3> "structural flaws"
435 2010-10-25 20:58:21 <bonsaikitten> so there may be hope :)
436 2010-10-25 20:58:23 <Diablo-D3> yes, it'd be nice if upstream didnt suck dick
437 2010-10-25 20:58:25 <Diablo-D3> it happens
438 2010-10-25 20:58:30 <Diablo-D3> there is a policy on how to get it fixed
439 2010-10-25 20:58:31 <bonsaikitten> hey, I just downgrade
440 2010-10-25 20:58:39 <Diablo-D3> yes, YOU downgrade
441 2010-10-25 20:58:43 <Diablo-D3> portage doesnt for you
442 2010-10-25 20:58:46 <Diablo-D3> and if it does, its broken
443 2010-10-25 20:58:56 <bonsaikitten> nah, I tell it emerge -uND world - and it'll downgrade
444 2010-10-25 20:59:04 <Diablo-D3> then portage is broken
445 2010-10-25 20:59:09 <bonsaikitten> because that's the proper way to handle it, not renaming things randomly
446 2010-10-25 20:59:12 <bonsaikitten> how is that broken?
447 2010-10-25 20:59:22 <Diablo-D3> why the fuck would I want a package manager that downgrades packages without my express permission
448 2010-10-25 20:59:36 <bonsaikitten> well, you TELL it to do that
449 2010-10-25 20:59:42 <bonsaikitten> so, uhm ... what was your point again?
450 2010-10-25 20:59:49 <bonsaikitten> apart from "OMG NOT APT FUUUUU" ?
451 2010-10-25 21:00:00 <Diablo-D3> you're trying to argue it cant downgrade
452 2010-10-25 21:00:01 <Diablo-D3> it can.
453 2010-10-25 21:00:08 <Diablo-D3> it just doesnt fuck over my system and do it automatically
454 2010-10-25 21:00:10 <Diablo-D3> NOT ONLY THAT
455 2010-10-25 21:00:15 <bonsaikitten> you claim portage satisfying the dep graph is wrong
456 2010-10-25 21:00:20 <bonsaikitten> which is kinda ... confusing
457 2010-10-25 21:00:20 <Diablo-D3> libqt4.7 is a _package_ not a _version_
458 2010-10-25 21:01:05 <noagendamarket_> Keefe I am getting an account at Technocash http://www.technocash.com/
459 2010-10-25 21:01:20 <bonsaikitten> Diablo-D3: hmm, what distinguishes a package from a version?
460 2010-10-25 21:01:25 <Diablo-D3> the name.
461 2010-10-25 21:01:25 <noagendamarket_> People will be able to deposit cash in a bank for btc
462 2010-10-25 21:01:28 <ArtForz> around here we seperate package name and version
463 2010-10-25 21:01:33 <bonsaikitten> Diablo-D3: is that another restriction of the package manager getting in the way?
464 2010-10-25 21:01:49 <bonsaikitten> hmm, but then qt is the package, and 4.7 the version
465 2010-10-25 21:01:49 <Diablo-D3> the version would be something like 4.7.0-really-4.4, from your example
466 2010-10-25 21:02:00 <Diablo-D3> or actually
467 2010-10-25 21:02:10 <Diablo-D3> 4.7.0-0.0-really-4.4
468 2010-10-25 21:02:20 <bonsaikitten> *twitch*
469 2010-10-25 21:02:22 <Diablo-D3> which then gets replaced with 4.7.0-1
470 2010-10-25 21:02:27 <bonsaikitten> *stab*
471 2010-10-25 21:02:33 <bonsaikitten> sorry, reflex
472 2010-10-25 21:02:36 <Diablo-D3> yes, god forbid we have a correctly designed packaging policy
473 2010-10-25 21:02:38 <bonsaikitten> been around Java people too long
474 2010-10-25 21:02:40 <Diablo-D3> and the great part is WE HAVE ONE
475 2010-10-25 21:02:45 <Diablo-D3> unlike Gentoo
476 2010-10-25 21:02:55 <bonsaikitten> doesn't mean it's coherent or sane
477 2010-10-25 21:02:57 <Diablo-D3> Gentoo is just randomly shove shit into portage, and hope it works
478 2010-10-25 21:03:03 <bonsaikitten> wanker.
479 2010-10-25 21:03:05 <Diablo-D3> thats worse than RPM hell
480 2010-10-25 21:03:13 <Diablo-D3> your shit was stupid 5 years ago, its stupid now
481 2010-10-25 21:03:16 <bonsaikitten> hah
482 2010-10-25 21:03:35 <Diablo-D3> why do you think I turned down the offer to become a gentoo dev? because I think your distro is shit
483 2010-10-25 21:03:44 <bonsaikitten> it's always fun to see a lack of arguments turns into an ad hominem
484 2010-10-25 21:03:53 <bonsaikitten> I think you're just incompetent
485 2010-10-25 21:04:03 <bonsaikitten> but that's a random opinion of someone you don't care about
486 2010-10-25 21:04:09 <bonsaikitten> so don't feel obliged to respond :)
487 2010-10-25 21:04:31 <theymos> I just compile all of my software on Linux. Then I don't have to worry about all of that package management nonsense -- I put stuff where I want it.
488 2010-10-25 21:04:40 <bonsaikitten> theymos: I'm too lazy for that
489 2010-10-25 21:04:50 <Diablo-D3> theymos: yeah, or I can just have the deb guys do it, since they already do it right
490 2010-10-25 21:04:53 <ArtForz> LFS FTW!
491 2010-10-25 21:04:56 <bonsaikitten> was a silly idea the first time I tried to get LFS working
492 2010-10-25 21:05:02 <Diablo-D3> bonsaikitten: oh, and a small question
493 2010-10-25 21:05:05 <bonsaikitten> Diablo-D3: see puppet :D
494 2010-10-25 21:05:17 <bonsaikitten> Diablo-D3: that is packaging insanity, according to policy and spec
495 2010-10-25 21:05:17 <Diablo-D3> why talk about that specific case, which almost never gets outside of unstable?
496 2010-10-25 21:05:27 <Diablo-D3> why are you running unstable if you dont want to specifically deal with upstream breakage?
497 2010-10-25 21:05:37 <Diablo-D3> this isnt gentoo.
498 2010-10-25 21:05:40 <bonsaikitten> Diablo-D3: because those are the ones where I usually end up compiling things by hand to work around idiocy
499 2010-10-25 21:06:04 <bonsaikitten> I really learned to love debian braindamage when I admin'ed something around 300 machines
500 2010-10-25 21:06:05 <Diablo-D3> <bonsaikitten> I love to bitch about things breaking and blame the wrong people, I'm a total asshole! wee!
501 2010-10-25 21:06:05 <theymos> I've been using LFS for about a year now, and it's been the only Linux "distro" that I have ever enjoyed using. Other distros kept surpirising me with strange behavior.
502 2010-10-25 21:06:15 <bonsaikitten> Diablo-D3: you must have a small penis
503 2010-10-25 21:06:18 <Diablo-D3> bonsaikitten: I also, now, understand why nenolod hates you.
504 2010-10-25 21:06:41 <bonsaikitten> hmm, nenolod ... sounds familiar ... do I have to know that entity?
505 2010-10-25 21:07:08 <bonsaikitten> I guess another "gentoo sucks so much that I use it" person
506 2010-10-25 21:07:29 <bonsaikitten> oh! I remember now! what a funny person.
507 2010-10-25 21:07:45 <Diablo-D3> ArtForz: I think I shall expand my ignore list by one.
508 2010-10-25 21:07:52 <Diablo-D3> wee!
509 2010-10-25 21:07:54 <bonsaikitten> teh horror :D
510 2010-10-25 21:08:24 <ArtForz> seriously, it's a fucking distro, not a religion
511 2010-10-25 21:08:36 <Diablo-D3> yeah, but cant tell the gentoo fags that
512 2010-10-25 21:08:41 <bonsaikitten> well, getting with the times is at times quite ... nice?
513 2010-10-25 21:08:43 <Diablo-D3> they get all ass burgers and shit when you do that
514 2010-10-25 21:08:45 <Diablo-D3> and flap their arms
515 2010-10-25 21:09:22 <bonsaikitten> "ooh ooh I are Diablo-D3 and everything sucks because I don't read manpages" :)
516 2010-10-25 21:12:11 <ArtForz> yeah, I get it, gentoo package management is infallible, gentoo users shit doesnt stink, and other distros having decent package management systems before gentoo existed are just imagination...
517 2010-10-25 21:13:23 <bonsaikitten> who claimed that?
518 2010-10-25 21:13:26 <Diablo-D3> ArtForz: hey, how many registers does a radeon 4850 have per whatever?
519 2010-10-25 21:13:29 <Diablo-D3> 16384?
520 2010-10-25 21:13:38 <Diablo-D3> or is that the whole device?
521 2010-10-25 21:13:44 <ArtForz> sexc
522 2010-10-25 21:13:45 <ArtForz> sec
523 2010-10-25 21:16:04 <ArtForz> 16384 per CU
524 2010-10-25 21:16:12 <Diablo-D3> and how many CU do I have?
525 2010-10-25 21:16:56 <ArtForz> 80 stream cores per CU so... 10
526 2010-10-25 21:17:55 <Diablo-D3> and I need about, say, 72 registers per shader execution?
527 2010-10-25 21:18:05 <ArtForz> not sure if those 16384 are 32-bit or 128-bit regs
528 2010-10-25 21:18:13 <ArtForz> soudns about right
529 2010-10-25 21:20:27 <Diablo-D3> hrrrm
530 2010-10-25 21:21:26 <Diablo-D3> ArtForz: and once I find a valid hash, I can immediately bail, right?
531 2010-10-25 21:22:11 <ArtForz> well, "immediately"
532 2010-10-25 21:22:30 <Diablo-D3> well, Im not likely to find multiple valid hashes, right?
533 2010-10-25 21:22:47 <ArtForz> yes
534 2010-10-25 21:23:28 <Diablo-D3> huh.
535 2010-10-25 21:24:20 <ArtForz> well, if you only do one iteration per shader it's no problem
536 2010-10-25 21:31:55 <Keefe> my code loses about 0.4% of winning hashes because i chose to minimize code branching/conditionals
537 2010-10-25 21:33:19 <Keefe> the odds of finding two winners in a single kernel execution is small enough i don't care
538 2010-10-25 21:33:30 <ArtForz> I take some slowdown by just setting a "at least one with H==0" bit for each batch of 1024 nonces and re-checking on the host
539 2010-10-25 21:34:09 <jgarzik> interesting.  my little VPS, which runs blk.bitcoinwatch.com and bitcoind itself, got pummeled on port 25 (SMTP).  exim handled it without a hitch, but it makes me wonder if it was an attack on bitcoin-related servers.
540 2010-10-25 21:34:27 <Diablo-D3> no, probably just normal spam
541 2010-10-25 21:34:29 <ArtForz> more likely random spammers
542 2010-10-25 21:34:39 <Keefe> i set an output mem location to the nonce when the hash is below the diff=1 target
543 2010-10-25 21:35:13 <Keefe> cause i also log diff=1 hits as a way of verifying correct math
544 2010-10-25 21:35:18 <ArtForz> my private SMTP gets hit by spammers a lot
545 2010-10-25 21:35:50 <ArtForz> it also gets complaints from "helpful" braindead people scanning for open relays
546 2010-10-25 21:36:18 <Keefe> since i use a global workgroup of 32M, it's a matter of the chance of finding more than one diff=1 winner in 32M hashes
547 2010-10-25 21:36:32 <jgarzik> denyhosts sure is handy.  I deny new hosts every day :)
548 2010-10-25 21:36:58 <ArtForz> = it appears to happily accept mail destined for external domains
549 2010-10-25 21:37:25 <soultcer> ArtForz: So what do you do what that mail for external domains? Read it?
550 2010-10-25 21:37:52 <Keefe> "diff=1 winner" = "H==0"
551 2010-10-25 21:38:12 <ArtForz> used to, but nowadays just goes straight to the bitbucket
552 2010-10-25 21:38:21 <ArtForz> yep
553 2010-10-25 21:38:40 <soultcer> Careful, my /dev/null once gut full and I had to reboot to fix it
554 2010-10-25 21:40:02 <ArtForz> apropos bitbucket: I *really* expected one above bit 5 here: http://www.youtube.com/watch?v=GcDshWmhF4A
555 2010-10-25 21:41:09 <Keefe> soultcer: you're joking, right?
556 2010-10-25 21:43:33 <soultcer> No I'm not
557 2010-10-25 21:43:49 <soultcer> I accidentally mounted my USB stick to /dev/ instead of /mnt :D
558 2010-10-25 22:06:32 <mtgox> .
559 2010-10-25 22:06:40 <Keefe> hi mtgox
560 2010-10-25 22:07:04 <Keefe> should ach withdrawals be available any day now?
561 2010-10-25 22:08:00 <Diablo-D3> omg its aliiive
562 2010-10-25 22:11:32 <mtgox> Hey Keefe: the first ACH processor I applied at turned me down. I started another application today. I'm going to implement LR tomorrow though
563 2010-10-25 22:14:02 <Kiba> hey people
564 2010-10-25 22:15:02 <theymos> mtgox: Will it be possible to withdraw current balances to LR?
565 2010-10-25 22:15:05 <Keefe> awesome! when USD can be withdrawn from mtgox to LR, i'll probably use mtgox much more often. i've been using BCM almost exclusively since the paypal freeze
566 2010-10-25 22:16:08 <mtgox> theymos: yeah when I get it integrated it will be. I'm trying to decide if it needs to be a 3rd currency though
567 2010-10-25 22:17:17 <Keefe> as in not LRUSD? or do you mean another tradable currency within mtgox?
568 2010-10-25 22:18:09 <jgarzik> polygamous trading between BTC<->LRUSD<->LREUR<->PGAU would be quite fun
569 2010-10-25 22:19:04 <mtgox> Keefe: like if we need to have BTC <-> USD <-> LRUSD
570 2010-10-25 22:20:02 <Keefe> would that be 2 or 3 currency pairs?
571 2010-10-25 22:20:12 <mtgox> 3
572 2010-10-25 22:20:20 <Keefe> nice
573 2010-10-25 22:21:06 <Keefe> but there still has to be a way for at least some people to deposit/withdraw "USD", not LRUSD
574 2010-10-25 22:22:31 <Keefe> or "USD" (i like to refer to it currently as MTGUSD) is worthless at mtgox. i believe the only reason the exchange rate at mtgox hasn't gone sky high is people believe direct usd withdrawal will be restored someday
575 2010-10-25 22:22:52 <mtgox> sure
576 2010-10-25 22:23:07 <Keefe> i guess that's what the attempt at setting up ACH is abpuout
577 2010-10-25 22:23:09 <Keefe> about*
578 2010-10-25 22:23:59 <Keefe> thanks for the update mtgox
579 2010-10-25 22:25:08 <jgarzik> it takes some doing, but I can buy bitcoins w/ PP @ mtgox.  it's a little bit of an escape valve, though not much.
580 2010-10-25 22:26:13 <Kiba> mtgox: are you going to enforce higher standard for trading?
581 2010-10-25 22:26:14 <Keefe> that's the wrong direction
582 2010-10-25 22:27:00 <Keefe> if you can only deposit to mtgox via paypal, and not withdraw, that doesn't support the value of mtgusd
583 2010-10-25 22:27:19 <jgarzik> Keefe: no, I offer PP to somebody who wants to withdraw USD from mtgox
584 2010-10-25 22:27:36 <jgarzik> I get bitcoins, and help somebody out
585 2010-10-25 22:28:30 <Keefe> Kiba: why are higher trading standards needed? trading within mtgox has always been safe, to all i understand
586 2010-10-25 22:28:38 <mtgox> I have emails from other people wanting to buy mtgox USD also
587 2010-10-25 22:28:44 <Kiba> Keefe: to prevent drive-by scamming
588 2010-10-25 22:29:21 <Keefe> i think you're saying higher standards are needed for usd being depoited into mtgox. i thought you were referring to trading
589 2010-10-25 22:30:37 <Keefe> jgarzik: so you're offering ppusd in exchange for others' mtgusd? because you'll eventually be able to withdraw that mtgusd from mtgox as usd
590 2010-10-25 22:31:03 <mtgox> he wants to withdraw it as BTC I think
591 2010-10-25 22:31:12 <jgarzik> Keefe: I'm offering PPUSD in exchange for others' MTGUSD, correct.  I use it to buy & withdraw BTC.
592 2010-10-25 22:31:52 <jgarzik> s/offering/offered/
593 2010-10-25 22:31:52 <Keefe> which you then sell for usd elsewhere? or are you using this opportunity to buy btc as a net result?
594 2010-10-25 22:32:00 <MacRohard> do you get a discount? or are you just losing money? ;)
595 2010-10-25 22:32:57 <jgarzik> I do a 1:1 exchange, with the other party dealing with any PP fees
596 2010-10-25 22:33:33 <MacRohard> and then you buy bitcoins at a 2 cent premium at your loss
597 2010-10-25 22:33:41 <theymos> MacRohard: It's probably cheaper to get money into MtGox by doing a 1:1 trade like that.
598 2010-10-25 22:34:34 <Keefe> BTC can be bought for PPUSD at BCM for a lower price
599 2010-10-25 22:34:36 <twobitcoins> Are you doing this because you actually want to get money into Mt. Gox, or just to be nice?
600 2010-10-25 22:35:21 <jgarzik> a bit of both.  the BTC purchased is more expensive than at BCM, but BCM bitcoin purchase via PP is not as simple, and I would be responsible for fees.
601 2010-10-25 22:36:00 <jgarzik> I imagine there's a charity premium in there, but I think BTC demand will continue, so IMO it's a decent investment (albeit in my "stupidly risky" portfolio category)
602 2010-10-25 22:36:30 <theymos> The value of MTG-USD will also likely go up.
603 2010-10-25 22:36:44 <jgarzik> being able to charge bitcoin purchase to my CC via PP is quite handy
604 2010-10-25 22:37:07 <Keefe> ah
605 2010-10-25 22:37:21 <appamatto> jgarzik enough to offset fees?
606 2010-10-25 22:37:27 <Keefe> by putting the fees on the recipient of your PP transfer, you can use CC without cost
607 2010-10-25 22:37:34 <jgarzik> ding
608 2010-10-25 22:37:37 <Keefe> whereas at BCM you're responsible to cover the fees
609 2010-10-25 22:37:41 <jgarzik> appamatto: what Keefe noted
610 2010-10-25 22:37:50 <appamatto> but they'll just give you fewer BTC
611 2010-10-25 22:37:51 <Kiba> I finally have 12 dollars now! Yay!
612 2010-10-25 22:38:16 <Keefe> basically, jgarzik is not offering 1:1 PP for MTG USD
613 2010-10-25 22:38:40 <Keefe> because he is going to fund the PP transfer by CC and thereby charge the other party about 3%
614 2010-10-25 22:39:18 <jgarzik> most recent trade (won't divulge more details), fee was $5 on $1000 traded
615 2010-10-25 22:39:42 <mtgox> have you guys used this exchangeZone to buy and sell LR?
616 2010-10-25 22:39:46 <twobitcoins> Ah, now I see.  I have plenty of Mt. Gox USD that I'd like to withdraw, so I'd do a deal like that as a personal transfer, but I'm not really interested in taking a 3% loss since I believe Mt. Gox will support withdrawals eventually.
617 2010-10-25 22:40:00 <Keefe> i have seen but not yet used EZ
618 2010-10-25 22:40:19 <jgarzik> mtgox: I've done several LRUSD->PPUSD trades, getting rid of LRUSD.  Works great.  For the other direction, PPUSD->LRUSD, you gotta send PPUSD then wait 7 days for LRUSD.
619 2010-10-25 22:40:35 <twobitcoins> Hmm, I'm not sure PayPal CC fees can be $5 on $1000.
620 2010-10-25 22:40:49 <theymos> I'm not giving my personal details to some unknown company that's using cloaked whois info (exchangeZone).
621 2010-10-25 22:40:50 <Keefe> ya, i wonder about that too
622 2010-10-25 22:40:56 <jgarzik> theymos: ditto
623 2010-10-25 22:40:59 <Keefe> the $5 fee thing
624 2010-10-25 22:41:18 <jgarzik> they do escrow
625 2010-10-25 22:41:49 <mtgox> jgarzik: what is to stop the person that paypals you from reversing the trade?
626 2010-10-25 22:42:20 <mtgox> and are the trades normally 1:1 ?
627 2010-10-25 22:42:26 <Kiba> only 4 ask?
628 2010-10-25 22:42:31 <jgarzik> mtgox: only the barriers EZ erects to keep out scammers:  requiring verification (passport + utility bill), and 7 day waiting period
629 2010-10-25 22:42:55 <jgarzik> mtgox: LRUSD seems to tend to be worth a bit more than PPUSD.
630 2010-10-25 22:43:09 <mtgox> more? weird
631 2010-10-25 22:43:45 <Keefe> PPUSD is chargeback'able, so it makes sense to me that LRUSD would be worth more
632 2010-10-25 22:44:01 <Kiba> ain't LRUSD gold and silver?
633 2010-10-25 22:44:01 <mtgox> but it is worth less on BCM
634 2010-10-25 22:44:07 <jgarzik> mtgox: my trades were 1:1, excluding PP fees, which I had to eat.  but I was impatient and took the first or second bid, rather than waiting for entire bidding process to conclude.  EZ lets you accept a bid at any time, to begin an exchange even before X days is complete.
635 2010-10-25 22:44:16 <Keefe> but i guess it depends on the expected loss rate
636 2010-10-25 22:44:19 <jgarzik> Kiba: no
637 2010-10-25 22:45:57 <Keefe> Kiba: LRUSD is backed by USD. there's also LR gold grams and LR EUR
638 2010-10-25 22:46:18 <jgarzik> I haven't found any way to get LR gold grams; it seems to be deprecated in LR's API, too.
639 2010-10-25 22:46:39 <Keefe> nanotube confirmed to me that the exchangers can actually trade directly with Liberty Reserve, so it really is backed
640 2010-10-25 22:47:18 <mtgox> why do you care about PGAU?
641 2010-10-25 22:48:23 <jgarzik> mtgox: It's a commodity, not directly attached to any nation-state fiat currencies.  Useful attributes.
642 2010-10-25 22:59:34 <dwdollar> mtgox:  Did FirstACH turn you down?
643 2010-10-25 23:09:02 <mtgox> dwdollar: yeah
644 2010-10-25 23:09:36 <dwdollar> they turned me down too
645 2010-10-25 23:14:47 <MacRohard> presumably you could use.. paper cheques.
646 2010-10-25 23:15:18 <Netsniper> howdy MacRohard
647 2010-10-25 23:16:56 <MacRohard> hey
648 2010-10-25 23:42:38 <Teppy> mtgox: You still around?
649 2010-10-25 23:43:20 <mtgox> hey
650 2010-10-25 23:44:03 <Teppy> Not sure if it helps, but I think I came up with a way to do credit cards that's pretty much bulletproof...
651 2010-10-25 23:44:25 <Teppy> Chaord is implementing this on bitcoingateway, but it may be useful for MtGox.
652 2010-10-25 23:46:50 <Kiba> the search for nonchargebackable form of exchange continued