1 2013-04-12 00:00:37 <MC1984> oh my node seems to be connected to one with a version of "eligius" thats mildly interesting
2 2013-04-12 00:00:41 <saracen> I'd say they were a form of compression. I wonder how well using something like snappy would perform. Or whether the encoding would work better with bits, rather than using byte boundaries.
3 2013-04-12 00:01:10 <lianj> MC1984: why, its lukes pool
4 2013-04-12 00:01:23 <MC1984> i know
5 2013-04-12 00:02:03 <MC1984> i wonder if i got one from b.i
6 2013-04-12 00:02:29 <MC1984> lots of "startingheight" well under 150,000, lots of new nodes
7 2013-04-12 00:03:38 <gmaxwell> saracen: general bitpacking is fairly costly. Conventional CPUs really prefer that you at least work a byte at a time.
8 2013-04-12 00:04:09 <gmaxwell> Though you can have bit-resolution encoding which efficiently works a word at a time.
9 2013-04-12 00:05:03 <RoboTeddy> would the mailing list proposals to prevent arbitrary data in the blockchain interfere with http://blog.cryptographyengineering.com/2013/04/zerocoin-making-bitcoin-anonymous.html ?
10 2013-04-12 00:05:28 <HM2> every time i see a major channel discussing bitcoin it still seems surreal
11 2013-04-12 00:06:23 <muadib> ya its mental
12 2013-04-12 00:06:26 <MC1984> they discuss it mostly from the sensationalist pov
13 2013-04-12 00:06:28 <muadib> and giving me a headache
14 2013-04-12 00:06:32 <barbarousrelic> I just got an "assertion failed" message when trying to start Bitcoin 0.8.1.0 Now it is reindexing the blocks.
15 2013-04-12 00:06:38 <MC1984> mystery currency created by shadowy figure
16 2013-04-12 00:06:44 <saracen> My indication that it would be "fine", is that I know game engines (id tech3, 4, source) all use "bit buffers" and send messages very very quickly because latency is such an issue with online play.
17 2013-04-12 00:09:44 <saracen> I guess the added benefit if the serialization was more module, is that it would allow for easier experimentation.
18 2013-04-12 00:10:10 <gmaxwell> RoboTeddy: doesn't matter, the scheme as designed is completely non-viable.
19 2013-04-12 00:10:18 <saracen> The UDP stuff somebody was working on, for example, maybe it would be better to split the packet sizes up to a maximum of 1500 for latency, to avoid fragmentation.
20 2013-04-12 00:10:34 <saracen> It might make sense to serialize things differently.
21 2013-04-12 00:11:00 <RoboTeddy> gmaxwell: ok, let's take as given that the scheme is non-viable; would a similarly-minded scheme that /was/ viable be prevented by the blockchain data rules?
22 2013-04-12 00:11:10 <gmaxwell> RoboTeddy: their one way accumulator things are very computationally expensive and result in very large transacitons.. e.g. 100k or so. they just wave their hands and say "uh.. keep the data in some DHT"
23 2013-04-12 00:11:39 <gmaxwell> RoboTeddy: but ignoring that it doesn't work, no??? it doesn't break that. They're already only storing a hash of data stored elsewhere.
24 2013-04-12 00:11:48 <graingert> gmaxwell: I shouldn't have gone back to university
25 2013-04-12 00:12:01 <RoboTeddy> gmaxwell: yeah, I noticed that -- definitely not something we can adopt, but it's interesting research. OK, good to know
26 2013-04-12 00:12:11 <graingert> Luke-Jr: you should have bought my timetable after all
27 2013-04-12 00:21:04 <Luke-Jr> graingert: I know, right?
28 2013-04-12 00:28:58 <graingert> Luke-Jr: I propose #graingerts-timetable-pricetalk
29 2013-04-12 00:29:50 <oo7pig> Hi. I have a big wallet around 250M. Each wallet Flushing takes around 4 seconds and blocks RPC calls. Do you know if there're any solutions? Thanks
30 2013-04-12 00:31:46 <gmaxwell> oo7pig: how did you get a 250mbyte wallet?!
31 2013-04-12 00:31:58 <gmaxwell> oo7pig: I don't have any short term solution for that.
32 2013-04-12 00:33:08 <oo7pig> I'm from btcchina. I wrote the code years ago and create an account in the wallet for each registered user
33 2013-04-12 00:33:20 <oo7pig> That's why the current wallet is so big
34 2013-04-12 00:34:16 <paulo_> oo7pig: manage your users without using the wallet file.
35 2013-04-12 00:34:38 <gmaxwell> paulo_: yes. thats nice, but he's already in this state.
36 2013-04-12 00:34:51 <oo7pig> yes I realized that but it will be a big change. Can I make the flushing not so frequently
37 2013-04-12 00:35:30 <graingert> oo7pig: you use an offline wallet for most of your funds right?
38 2013-04-12 00:35:35 <gmaxwell> oo7pig: Yes, you could remove most of the runtime flushing. Makes it more likely that a live backup would not be successful.
39 2013-04-12 00:36:46 <oo7pig> I see. How can I do that
40 2013-04-12 00:36:57 <oo7pig> @graingert sure
41 2013-04-12 00:37:02 <paulo_> no really a dev question, but I'm curious. How much does a trading platfor make?
42 2013-04-12 00:37:07 <graingert> Just checking
43 2013-04-12 00:37:30 <graingert> paulo_: multiply fee by throughout
44 2013-04-12 00:37:52 <graingert> Volume*
45 2013-04-12 00:38:32 <paulo_> hmmm not bad.
46 2013-04-12 00:41:02 <oo7pig> gmaxwell how can I do that and is there a negative impact besides live backup?
47 2013-04-12 00:42:36 <k9quaint> I gotta say, this has been an interesting week for Bitcoin
48 2013-04-12 00:43:48 <madb_> hello
49 2013-04-12 00:44:42 <gmaxwell> oo7pig: it would potentially increase the risk of data loss of account data in the event of a crash. (though bdb is supposted to be durable even without the flushing)
50 2013-04-12 00:45:25 <gmaxwell> oo7pig: you can set -flushwallet=0 I believe.
51 2013-04-12 00:48:10 <oo7pig> gmaxwell I think another way is to trim the wallet size. Is there a way to remove the internal 'move' transactions from the wallet?
52 2013-04-12 00:58:19 <bytecoin> ;;ticker
53 2013-04-12 00:58:21 <gribble> BTCUSD ticker | Best bid: 128.00000, Best ask: 129.94000, Bid-ask spread: 1.94000, Last trade: 130.00000, 24 hour volume: 139119.23267947, 24 hour low: 65.00000, 24 hour high: 188.70000, 24 hour vwap: 132.39586
54 2013-04-12 01:18:19 <danwalton> hello
55 2013-04-12 01:18:43 <danwalton> I just launched a little project and I would love some feedback. http://www.satoshibones.com thanks guys!
56 2013-04-12 01:19:04 <copumpkin> s/loose/lose/
57 2013-04-12 01:19:11 <copumpkin> also, I don't think this channel is the right place
58 2013-04-12 01:19:33 <danwalton> thanks
59 2013-04-12 01:19:54 <BlueMatt> copumpkin: do you send coins back the address that represents the source?
60 2013-04-12 01:20:13 <copumpkin> I don't do anything, but it looks like a satoshidice-alike
61 2013-04-12 01:20:23 <copumpkin> so I'd assume so
62 2013-04-12 01:20:25 <danwalton> right, I am trying to do something a little unique here
63 2013-04-12 01:20:27 <BlueMatt> ehh, danwalton
64 2013-04-12 01:20:31 <danwalton> http://www.satoshibones.com/invest
65 2013-04-12 01:20:41 <copumpkin> talking about a "source" is a bit of a mis-concept
66 2013-04-12 01:20:59 <BlueMatt> danwalton: yea, there is technically no source for a transaction
67 2013-04-12 01:21:05 <BlueMatt> and people using webwallets may break
68 2013-04-12 01:21:10 <BlueMatt> (ie not get the response coins)
69 2013-04-12 01:21:17 <copumpkin> I think that's what his warning means when he says some wallets may not be compatible
70 2013-04-12 01:21:34 <BlueMatt> oh, must've missed that
71 2013-04-12 01:21:40 <BlueMatt> sorry, apparently Im not paying attention
72 2013-04-12 01:21:42 <danwalton> I use the source transaction as part of the response transaction
73 2013-04-12 01:21:45 <copumpkin> you gotta admit, it angers the purists, but it's dead simple and has no overhead in terms of user-friendliness, if they don't lose their money
74 2013-04-12 01:22:38 <BlueMatt> copumpkin: "if they don't lose their money" seems like a pretty big caveat
75 2013-04-12 01:23:05 <gmaxwell> copumpkin: on that subject, you see that SD hasn't been returning bets? once there were 15k unpaid douglas modified the stats script to ignore the unpaid ones.
76 2013-04-12 01:23:25 <copumpkin> well, with the standard client, you will get the money back, right? and so people can keep gambling with zero effort and you can keep making money
77 2013-04-12 01:23:31 <copumpkin> it's the electronic equivalent of a slot machine
78 2013-04-12 01:23:34 <copumpkin> takes zero thought
79 2013-04-12 01:23:42 <copumpkin> gmaxwell: I didn't :/
80 2013-04-12 01:24:58 <Luke-Jr> danwalton: flood attacks like that aren't unique, just annoying
81 2013-04-12 01:26:04 <danwalton> Luke-Jr: appreciate the feedback ;)
82 2013-04-12 01:26:34 <Luke-Jr> danwalton: if you care about Bitcoin, don't abuse it. take deposits and send withdrawls out
83 2013-04-12 01:26:57 <gmaxwell> Luke-Jr: can you point him to your post proposing the ephemerial 'accounts'?
84 2013-04-12 01:27:11 <Luke-Jr> gmaxwell: I don't know where it is :x
85 2013-04-12 01:27:20 <danwalton> Luke-Jr: I care about bitcoin. It should be robust no matter what the use
86 2013-04-12 01:27:37 <Luke-Jr> danwalton: abuse != use
87 2013-04-12 01:28:01 <Luke-Jr> danwalton: Bitcoin can only be robust against abuse by ignoring/filtering it
88 2013-04-12 01:28:13 <danwalton> Possible
89 2013-04-12 01:28:31 <danwalton> i think bit coin stands up pretty well. its the trading markets that need support
90 2013-04-12 01:28:51 <Luke-Jr> danwalton: DP's flood is already costing the network dearly because miners have been too neglegent to filter it out
91 2013-04-12 01:28:52 <gmaxwell> danwalton: If only the world were so simple. Robustness is not a binary state. To be hyperbolic: "I care about america, it should be robust even if I bomb it with chemical weapons" .. and indeed, america would persist, but at what level.
92 2013-04-12 01:28:57 <danwalton> also, my main interest is in a bitcoin based stock exchange. something that is p2p and uses the existing network.
93 2013-04-12 01:29:01 <Luke-Jr> danwalton: do you really want to make that worse?
94 2013-04-12 01:29:21 <danwalton> i don't think I'm going to make it worse
95 2013-04-12 01:29:28 <copumpkin> "to ensure that america is reslient against bomb attacks, I shall bomb all its major cities"
96 2013-04-12 01:29:42 <copumpkin> "it's for the good of america"
97 2013-04-12 01:29:47 <danwalton> queue hitler analogy in 5,4,3,2
98 2013-04-12 01:29:47 <Luke-Jr> danwalton: probably not, but please don't try
99 2013-04-12 01:29:49 <copumpkin> <_<
100 2013-04-12 01:29:49 <gmaxwell> danwalton: The low cost of running a full node is essential to bootstrapping bitcoin, keeping it decenteralized while technology improves and while the justification for taking the cost of running it is built by bitcoin becoming accepted as a lasting valuable thing.
101 2013-04-12 01:30:04 <gmaxwell> danwalton: if the blockchain was 10GB on day 1 bitcoin would have gone nowhere.
102 2013-04-12 01:30:23 <danwalton> I'm flattered, but bitcoin will survive me.
103 2013-04-12 01:30:52 <Luke-Jr> danwalton: just because Bitcoin will survive you, does not justify you trying to harm it
104 2013-04-12 01:31:06 <gmaxwell> likewise today, if the growing cost of operating a bitcoin node outpaces bitcoin's established economic importance then everyone who continues to use it will do so via a couple of centeralized services??? and then it'll all be over when some attack on those services happens.
105 2013-04-12 01:32:03 <danwalton> I just asked satoshi, he likes my project. everything is cool.
106 2013-04-12 01:32:03 <gmaxwell> and sure, one piece of stupidity is not going to break it??? but all inefficiencies erode our 'startup capital' faster than need be, and decrease Bitcoin's chance of survival.
107 2013-04-12 01:32:26 <gmaxwell> danwalton: ...
108 2013-04-12 01:32:54 <GlitchNZ> danwalton: don't worry about these guys - they think innovation and free exploration of ideas are the devils work
109 2013-04-12 01:33:19 <Luke-Jr> GlitchNZ: ???
110 2013-04-12 01:33:29 <gmaxwell> GlitchNZ: are you british too?
111 2013-04-12 01:33:38 <gmaxwell> :P
112 2013-04-12 01:33:41 <GlitchNZ> lol
113 2013-04-12 01:34:02 <danwalton> not worried at all. criticism is acknowledgment of creation. and thats why I launched this today on irc
114 2013-04-12 01:34:06 <saracen> lol
115 2013-04-12 01:34:18 <Luke-Jr> the only thing innovative in the topic is social engineering ways around anti-spam - not sure that's useful
116 2013-04-12 01:34:33 <GlitchNZ> There are those that hold onto bitcoin as an ideology that is sacred, and all participants should not desicrate it, then there are those that think it is a gift to the world that should be explored every way possible
117 2013-04-12 01:35:00 <Luke-Jr> GlitchNZ: there is NOTHING useful about a stupid flood attack
118 2013-04-12 01:35:29 <Luke-Jr> all it does is destroy the system for everyone else
119 2013-04-12 01:35:30 <gmaxwell> GlitchNZ: funny, someone who creates an unoriginal clone of a system which is known to make bitcoin less usable for all the other users.. I wouldn't really call that valuable exploriation. Then again, it's all moot: there have been a bunch of clones none are successful.
120 2013-04-12 01:35:49 <gmaxwell> You can speculate as to why... considering that there should be no network effect in that business...
121 2013-04-12 01:35:55 <gmaxwell> but thats not a subject for this channel.
122 2013-04-12 01:36:00 <GlitchNZ> Luke: you see it as a flood attack that will take down bitcoin infrastructure - i see it as innovation that incourages development on stronger infrastructure that supports what the people want out of bitcoin
123 2013-04-12 01:36:19 <praetoriansentry> Does anyone else use quantmod + R + mtgox api?
124 2013-04-12 01:36:20 <Luke-Jr> GlitchNZ: well, I'll be blunt: you are an idiot for thinking so
125 2013-04-12 01:36:27 <lianj> sending out 0.00000001 is so awesome genius, great creation
126 2013-04-12 01:36:28 <GlitchNZ> if people didnt want it, then the sites would fail
127 2013-04-12 01:36:41 <GlitchNZ> if bitcoin cant handle it, someone will develop a system that can
128 2013-04-12 01:36:50 <gmaxwell> GlitchNZ: No one is saying it can't handle it.
129 2013-04-12 01:36:52 <Luke-Jr> GlitchNZ: the sites would fail, were it not for the money laundery and/or stock pump and dump going on
130 2013-04-12 01:37:18 <gmaxwell> GlitchNZ: it doesn't actually take many or even any to cause create a lot of traffic at such a site.
131 2013-04-12 01:37:19 <GlitchNZ> the question is simply, do we bend the people to work within the confines of the bitcoin system, or do we bend bticoin to work with what the people want
132 2013-04-12 01:37:33 <gmaxwell> GlitchNZ: Why not both?
133 2013-04-12 01:37:42 <Squeezle_> build an exchange into the wallet, add google maps plugin to find local participants
134 2013-04-12 01:37:45 <gmaxwell> Every system influences people and is influenced by people.
135 2013-04-12 01:37:50 <Luke-Jr> GlitchNZ: it is not even theoretically possible to simply "accept" flood attacks like this
136 2013-04-12 01:37:53 <GlitchNZ> sorry, i prefer to make software work for me, not e work for it
137 2013-04-12 01:38:30 <gmaxwell> GlitchNZ: You seem to be taking a rather narrow idelogically oriented view here.
138 2013-04-12 01:38:45 <lianj> danwalton: so you pay 0.0005 to send 0.00000001 saying you lost?
139 2013-04-12 01:38:47 <wumpus> yes, can we talk about development here
140 2013-04-12 01:38:58 <wumpus> too much noise
141 2013-04-12 01:39:07 <danwalton> lianj: yes
142 2013-04-12 01:39:08 <Luke-Jr> wumpus: you have something to say about development? :D
143 2013-04-12 01:39:16 <gmaxwell> Luke-Jr: he's right. pshaw.
144 2013-04-12 01:39:19 <wumpus> Luke-Jr: always
145 2013-04-12 01:39:23 <danwalton> lianj: gotta support the network
146 2013-04-12 01:39:24 <gmaxwell> The channel should be quiet when there is nothing to say in any case.
147 2013-04-12 01:39:32 <Luke-Jr> gmaxwell: IMO, off-topic is only off-topic when there's something on-topic to discuss ;)
148 2013-04-12 01:39:41 <danwalton> lianj: but you make a good point
149 2013-04-12 01:40:03 <GlitchNZ> I think people coming to a dev channel to talk about something they have develped, and getting opinions from their peers is on topic
150 2013-04-12 01:40:03 <gmaxwell> Luke-Jr: no, people follow this channel from the logs too. It's important to transparency, since we've mostly failed at having important discussions on the list except for final stuff.
151 2013-04-12 01:40:11 <Luke-Jr> danwalton: considering that your spam costs the network more than 0.0005 BTC to handle, you're not supporting anything
152 2013-04-12 01:40:29 <Luke-Jr> gmaxwell: true
153 2013-04-12 01:40:32 <lianj> danwalton: support the networks and annoy your users that end up with all that 0.00000001 their wallet tries to spend again and making their txs bigger to eventually pay more fee in bad cases?
154 2013-04-12 01:40:35 <Luke-Jr> GlitchNZ: so why are you here?
155 2013-04-12 01:40:35 <saracen> While bitcoin can handle it, aren't these transactions only annoying because it means that there's even less of a reason to run a node, or am I missing something?
156 2013-04-12 01:40:39 <wumpus> GlitchNZ: this channel is really about bitcoin client dev, not the infrastructure around it, there are channels such as #bitcoin-tech for that
157 2013-04-12 01:40:44 <gmaxwell> GlitchNZ: he got an opnion it is the common and generally uncontroversial opinion that people should avoid operating in that manner.
158 2013-04-12 01:40:52 <danwalton> ok
159 2013-04-12 01:41:32 <Luke-Jr> saracen: they're exactly the kind of flood that miners are supposed to protect against
160 2013-04-12 01:41:55 <saracen> Luke-Jr: So they're expected to filter these transactions?
161 2013-04-12 01:42:04 <Luke-Jr> ACTION ponders whether to work more on BFGMiner+Mesa, or Bitcoin-Qt "next-test"
162 2013-04-12 01:42:29 <lianj> danwalton: isn't it enough to show on the page that you lost? why send him 0.00000001, thats useless
163 2013-04-12 01:42:34 <Luke-Jr> saracen: yes; the original plan was to make it simply unaffordable to spam with transaction fees, but DP & co have figured out a way to socially engineer others into paying the cost for them
164 2013-04-12 01:42:36 <GlitchNZ> saracen, luke: that is exactly how the system has been designed - transaction fee's andwhat is 'acceptable' should be controlled by market economics
165 2013-04-12 01:42:47 <gmaxwell> saracen: right??? they result in perpetual increases in data storage required, more time to introduce a new node, more bandwidth used, slower confirmations for joe-blow and his regular bitcoin transactions, higher fees for random bitcoin users... etc.. Nothing fatal.. but not good either.
166 2013-04-12 01:42:55 <Squeezle_> I am unsure if my topic was on topic, should I be in tech?
167 2013-04-12 01:43:08 <danwalton> lianj: this is helpful feedback. I bet I can improve this.
168 2013-04-12 01:43:14 <Luke-Jr> Squeezle_: yours is on-topic
169 2013-04-12 01:43:38 <Luke-Jr> Squeezle_: I think, however, the difficulties in that are mainly legal
170 2013-04-12 01:43:43 <Squeezle_> Luke-Jr: thanks was a bit fuzzy :p
171 2013-04-12 01:43:44 <gmaxwell> danwalton: luke made a nice outline as to how such a service could work which should have equal or better usability but not involve so much transaction load.
172 2013-04-12 01:43:57 <gmaxwell> (and he'd run one if it didn't appear to be unlawful for him to do so)
173 2013-04-12 01:44:03 <danwalton> gmaxwell:where is it?
174 2013-04-12 01:44:21 <danwalton> sounds interesting
175 2013-04-12 01:44:35 <gmaxwell> danwalton: trying to find it.
176 2013-04-12 01:44:45 <danwalton> cool. thanks!
177 2013-04-12 01:44:57 <Squeezle_> that is probably true, I prefer to do the programming first and let other people worry about the paperwork, hence my move to bitcoin
178 2013-04-12 01:46:25 <gmaxwell> danwalton: but the general idea is that the player hits a webpage.. drops in a bitcoin address of their own, page respond back with an address to send to.. users sends.. then with no confirms.. the balance shows up and the user can just click click click to play as much as they like. When they're done they can either just close their browser, or can hit a checkout. The site then pays them their remaining balance using their original ...
179 2013-04-12 01:46:32 <gmaxwell> ... input... and perhaps requiring confirms on the input before the withdraw depending on the amounts and site policy.
180 2013-04-12 01:46:41 <gmaxwell> so the idea is that there is no accounts ... no sign up.. no waiting...
181 2013-04-12 01:47:03 <Luke-Jr> gmaxwell: strange, I can't find it with bctroll's search
182 2013-04-12 01:47:17 <gmaxwell> but there only needs to be 1-2 transactions per user per session.
183 2013-04-12 01:47:32 <gmaxwell> (maybe even less, if you make the timeout days)
184 2013-04-12 01:47:33 <Squeezle_> that is how I expected it to work before seeing what has been built, it seems the more natural approach, temp wallet
185 2013-04-12 01:48:15 <gmaxwell> and then you can do very high speed playing.. as fast as the user can click. No transaction fees per play.. and you can also offer games that are more complicated than just a couple of prefab options.
186 2013-04-12 01:48:32 <gmaxwell> The "ask for a refund address first" makes it work with _all_ wallets, including things like mtgox accounts.
187 2013-04-12 01:48:46 <Squeezle_> pay to play
188 2013-04-12 01:49:11 <Luke-Jr> (if you really want to support miners, you can always add up a transaction fee to put on the withdrawl)
189 2013-04-12 01:49:28 <gmaxwell> Luke-Jr: yea, but you don't need to have one on every play??? which is the big improvement.
190 2013-04-12 01:49:44 <Luke-Jr> ACTION nods
191 2013-04-12 01:49:51 <Squeezle_> and it improves your odds of transaction priority
192 2013-04-12 01:49:54 <feral> thats brilliant
193 2013-04-12 01:50:03 <gmaxwell> and if you're worried about security you can use exactly the same 'cryptographic' proof with a hmac tha sdice uses... nothing about that requires a blockchain.
194 2013-04-12 01:50:18 <lianj> feral: that how many btc games besides 1dice crap works
195 2013-04-12 01:50:32 <feral> circle is like that yeah?
196 2013-04-12 01:50:40 <gmaxwell> e.g. you just take the users refund address and add a counter that goes up with every play, hmac that to get your random numbers.
197 2013-04-12 01:51:41 <lianj> feral: yep. played several blackjack games (different) sites that all had that flow. no account, deposit in and withdraw when done.
198 2013-04-12 01:52:09 <gmaxwell> lianj: ha, I wasn't aware people were doing that ??? two years ago it seemed everything was account oriented!
199 2013-04-12 01:52:25 <saracen> are there any IRC games that revolve around bitcoins yet?
200 2013-04-12 01:52:30 <gmaxwell> well 'everything'
201 2013-04-12 01:52:38 <gmaxwell> Not like I looked at more than a couple things.
202 2013-04-12 01:52:56 <lianj> gmaxwell: one year ago i played it on a visually very shitty site but the btc thing worked very good there already
203 2013-04-12 01:53:03 <amiller> i really wish someone would make some irc games involving bitcoin
204 2013-04-12 01:53:18 <gmaxwell> amiller: we really need an IRC micropayment bot.
205 2013-04-12 01:53:22 <saracen> I might do so if it isn't overly done already.
206 2013-04-12 01:53:23 <amiller> i can't give someone $0.20 and tell them to go have fun
207 2013-04-12 01:53:23 <Squeezle_> I think the main reason it hasn't been done like that already is the advertising oppourtunitys that come with a subsription, but no one wants to actually see the ad's, no ads and the page would be fast, gamblers nightmare
208 2013-04-12 01:53:44 <lianj> amiller: pssh, was just about to restart mine
209 2013-04-12 01:54:04 <amiller> i wish there was something fun to do with tiny amounts of bitcoin besides gambling but that has instant gratification
210 2013-04-12 01:54:27 <amiller> donating to a cause doesn't count, buying silly things on bitmit or otc is kinda close
211 2013-04-12 01:54:32 <spine023u4230> btc needs a mud using satoshis instead of gold ^_^
212 2013-04-12 01:54:47 <gmaxwell> amiller: there has been some fun games in IRC in the past.. not many latey.
213 2013-04-12 01:55:00 <gmaxwell> e.g. people were doing "decrypt this mtgox code" for a while.
214 2013-04-12 01:55:05 <Squeezle_> it needs atm machines at the grocery store and near places that don't accept it yet
215 2013-04-12 01:55:30 <lianj> gmaxwell: #winBTC does that kind of challenges atm, fun.
216 2013-04-12 01:55:40 <amiller> atm's only let you connect bitcoin to things normal currencies already do, it's not the same as a pure bitcoin novelty
217 2013-04-12 01:55:43 <Luke-Jr> stupid freenode
218 2013-04-12 01:55:44 <betece> Is there any way to create a form of Bitcoin-OTC as a decentralized Exchange?
219 2013-04-12 01:55:50 <gmaxwell> I hadn't seen that channel. Cool.
220 2013-04-12 01:56:08 <Luke-Jr> gmaxwell: decided to rework rpc_priotxn - does it make sense to put the map Gavin suggested on the mempool object?
221 2013-04-12 01:56:10 <Wayward> does anyone know just what <patterns> vanitygen supports?
222 2013-04-12 01:56:14 <Wayward> I can't find it documented
223 2013-04-12 01:56:42 <Wayward> if it supports prefix-substrings only, or globbing patterns, or regular expressions
224 2013-04-12 01:57:03 <amiller> i'd like to give a website or an irc bot $0.20 in bitcoins and have htem mail me a post card
225 2013-04-12 01:57:05 <gmaxwell> amiller: I made a puzzle once where you had to de-rot13 then mmdecode .. then apply the burrows-wheeler transform. ... and not only one but _two_ people eventually solved it. Though it took over a day.
226 2013-04-12 01:57:16 <amiller> except even that would require me to divulege more personal information than i'd trust to a bot
227 2013-04-12 01:57:17 <Luke-Jr> Wayward: vanitygen -h
228 2013-04-12 01:57:25 <amiller> i'll tell you what the best idea is along these lines
229 2013-04-12 01:57:26 <gmaxwell> Luke-Jr: I think that makes sense.
230 2013-04-12 01:57:32 <amiller> is to hook up a bitcoin frontend to something done on mechanical turk
231 2013-04-12 01:57:46 <Wayward> Luke-Jr: still no mention
232 2013-04-12 01:57:47 <amiller> not an in-bitcoin mechanical turkl crowdsourcing platform - i mean to just use amazon mechanical turk in the background
233 2013-04-12 01:57:53 <Luke-Jr> Wayward: try reading it
234 2013-04-12 01:58:01 <Wayward> i just did again.
235 2013-04-12 01:58:11 <oiram> Can't they at least tweet that they're being ddos'd?
236 2013-04-12 01:58:12 <gmaxwell> amiller: you mean coinworker? or the other side?
237 2013-04-12 01:58:13 <Wayward> "By default, <pattern> is interpreted as an exact prefix." says nothing.
238 2013-04-12 01:58:14 <amiller> like you cuold upload a picture on imgur and give the bot the url, and the bot would hire someone on mechanical turk to like embellish your drawing with something
239 2013-04-12 01:58:26 <amiller> yeah not like coinworker
240 2013-04-12 01:58:39 <gmaxwell> okay the other side. oh yea, you'd mentioned that you used coinworker before.
241 2013-04-12 01:58:41 <Luke-Jr> Wayward: it answers your question
242 2013-04-12 01:58:53 <Wayward> Luke-Jr, where?
243 2013-04-12 01:58:55 <gmaxwell> amiller: seems really really trivial to make an actual btc to mt gateway.
244 2013-04-12 01:59:09 <amiller> the real pleasure is in making special pipelines
245 2013-04-12 01:59:18 <Wayward> "By default, windows sucks" ... solution: don't use defaults.
246 2013-04-12 01:59:32 <amiller> i want like a bitcoin midway
247 2013-04-12 01:59:32 <Wayward> where's the "use regular expressions" anti default
248 2013-04-12 01:59:36 <amiller> carnival games and souvenirs
249 2013-04-12 01:59:38 <Luke-Jr> Wayward: there's options listed below that change the format
250 2013-04-12 01:59:49 <amiller> the campier the better really especially if it keeps stakes low and thereby is accessible to more people
251 2013-04-12 01:59:57 <Luke-Jr> -r Use regular expression match instead of prefix
252 2013-04-12 01:59:58 <Luke-Jr> (Feasibility of expression is not checked)
253 2013-04-12 02:00:02 <Luke-Jr> -i Case-insensitive prefix search
254 2013-04-12 02:00:05 <Wayward> hmmm
255 2013-04-12 02:00:10 <betece> is there any project on creating a decentralized exchange for bitcoin ?
256 2013-04-12 02:00:21 <Luke-Jr> betece: #bitcoin-otc
257 2013-04-12 02:00:33 <Wayward> I guess oclvanitygen.exe doesn't have -r
258 2013-04-12 02:00:36 <Wayward> or not documented
259 2013-04-12 02:01:00 <betece> Luke-Jr: bitcoin otc is to nerdy compared to mtgox
260 2013-04-12 02:02:17 <Wayward> oclvanitygen.exe is the one that uses graphics hardware, right? normal vanitygen does not?
261 2013-04-12 02:02:50 <Wayward> seems oclvanitygen.exe does not infact support -r :/
262 2013-04-12 02:06:56 <Luke-Jr> gmaxwell: are we allowed to use 'auto' type in bitcoind yet? <.<
263 2013-04-12 02:22:24 <Wayward> what's the function to convert Difficulty and Hash/sec into Expected duration?
264 2013-04-12 02:23:32 <Wayward> is difficulty estimated mean number of hashes to match?
265 2013-04-12 02:24:58 <cyphase> http://blog.cryptographyengineering.com/2013/04/zerocoin-making-bitcoin-anonymous.html
266 2013-04-12 02:25:02 <cyphase> zerocoin ^
267 2013-04-12 02:27:47 <cyphase> oh, i guess i'm late (looking at chat logs)
268 2013-04-12 02:30:32 <lianj> Wayward: ((0xffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff / Bitcoin.decode_compact_bits(0x1a022fbe)) / 270_000_000) / 60 / 60 / 24 => 1412 days @ 270Mh
269 2013-04-12 02:31:28 <Wayward> where does Bitcoin.decode_compact_bits(0x1a022fbe) come from?
270 2013-04-12 02:31:45 <lianj> oh, thats from my code
271 2013-04-12 02:35:39 <lianj> Wayward: its to turn the compact bits (from the block header) into a bignum again
272 2013-04-12 02:37:18 <Wayward> is there a calculator that you know of to convert vanitygen difficulty with hashrate into duration?
273 2013-04-12 02:38:58 <lianj> i think thats something different
274 2013-04-12 02:39:13 <lianj> not sure, never looked into vanitygen
275 2013-04-12 03:14:31 <_dr> venitygen difficulty with hashrate into duration?
276 2013-04-12 03:15:18 <_dr> the process adheres to a pdf, so there's no way to get a real eta, but there's a pretty good eta estimation based on probablility built into vanitygen, isn't there?
277 2013-04-12 03:15:52 <_dr> eta estimation, nice
278 2013-04-12 03:24:34 <forrestv> Wayward, difficulty*2^32 is expected number of hashes
279 2013-04-12 03:25:13 <Wayward> ok
280 2013-04-12 04:15:25 <Luke-Jr> BlueMatt: can you look into http://jenkins.bluematt.me/pull-tester/1d149393d8dcbd1593b7cb979d46879ad10b95aa/test.log ? it builds fine for me..
281 2013-04-12 04:26:31 <BlueMatt> Luke-Jr: bad timing, Im just headed off to bed...can you try running exactly as in http://jenkins.bluematt.me/pull-tester/files/build-script.sh (it build leveldb first, probably needs reverted) and tell me what happens (either script needs changed or more research needs done)
282 2013-04-12 04:26:50 <Luke-Jr> BlueMatt: no rush
283 2013-04-12 04:27:34 <BlueMatt> also, why has github still not gotten their shit together and returned all the pull requests in its api instead of like 20
284 2013-04-12 04:28:59 <Luke-Jr> CodeShark: are you up?
285 2013-04-12 04:29:13 <CodeShark> Luke-Jr: yer!
286 2013-04-12 04:29:19 <CodeShark> how can I be of service?
287 2013-04-12 04:29:33 <Luke-Jr> CodeShark: can you explain the state of your multiwallet pullreqs? :P
288 2013-04-12 04:29:44 <Luke-Jr> it "feels" like there are some redundant still open?
289 2013-04-12 04:29:55 <CodeShark> yes, indeed...so let me explain...
290 2013-04-12 04:30:52 <CodeShark> #2124 was an attempt to add full support - but it backfired on me because the master branch continued to move forward, which made it very difficult to merge
291 2013-04-12 04:31:18 <CodeShark> the build works for me and I'm actually running it on a couple machines right now
292 2013-04-12 04:31:30 <CodeShark> but obviously, the goal isn't to maintain a separate branch :p
293 2013-04-12 04:31:55 <CodeShark> therefore, I started pull request #2407
294 2013-04-12 04:32:16 <CodeShark> which is an attempt to restructure the commits of #2124 in a way that is much easier to check and to test
295 2013-04-12 04:32:18 <CodeShark> incrementally
296 2013-04-12 04:33:26 <CodeShark> so for instance, #2407 touches none of the RPC
297 2013-04-12 04:33:36 <CodeShark> nor anything in main.cpp
298 2013-04-12 04:34:05 <CodeShark> furthermore, the CWalletManager class has been added in a new pair of source files
299 2013-04-12 04:34:13 <CodeShark> so as not to be mucking with wallet.h/wallet.cpp
300 2013-04-12 04:34:39 <CodeShark> make sense?
301 2013-04-12 04:35:36 <Luke-Jr> so #2407 is part of #2124
302 2013-04-12 04:35:59 <CodeShark> yes, although in a very differently structured commit
303 2013-04-12 04:36:25 <CodeShark> 2407 is a requirement for 2124 to work - you can think of it that way
304 2013-04-12 04:36:55 <CodeShark> but added in a way that touches the minimal number of source files and changes as little behavior as possible
305 2013-04-12 04:37:39 <Luke-Jr> 2407 doesn't conflict with 2124?
306 2013-04-12 04:37:54 <CodeShark> oh, I wouldn't try to merge them :p
307 2013-04-12 04:38:06 <CodeShark> the requirement is the CWalletManager class
308 2013-04-12 04:38:28 <CodeShark> but it's been added in different source files
309 2013-04-12 04:39:09 <CodeShark> basically, I want to completely restructure the commits from 2124 and gradually merge them into master
310 2013-04-12 04:39:15 <Luke-Jr> is it possible to get the features of 2124 via 2407+others?
311 2013-04-12 04:39:27 <CodeShark> yes, the idea is to do the others
312 2013-04-12 04:39:31 <CodeShark> which haven't been done yet
313 2013-04-12 04:39:53 <CodeShark> I'll be taking chunks from 2124 and making incremental pull requests
314 2013-04-12 04:40:00 <CodeShark> 2124 is too big a change to be safely merged
315 2013-04-12 04:41:18 <CodeShark> so to answer your question, Luke-Jr, yes
316 2013-04-12 04:41:59 <CodeShark> 2407 is the way forward, not 2124 - 2124 will be cherrypicked
317 2013-04-12 04:42:21 <Luke-Jr> but not today..
318 2013-04-12 04:42:50 <CodeShark> lol - were you hoping on getting it all done today?
319 2013-04-12 04:42:58 <CodeShark> no, not today
320 2013-04-12 04:43:15 <CodeShark> I would like to fix any issues with 2407 and get it merged first
321 2013-04-12 04:43:39 <CodeShark> then worry about further pull requests
322 2013-04-12 04:44:47 <Luke-Jr> CodeShark: I was hoping to get the full featureset into next-test today, at least :P
323 2013-04-12 04:45:09 <CodeShark> I'll probably add the RPC stuff last
324 2013-04-12 04:45:10 <CodeShark> lol
325 2013-04-12 04:45:16 <Luke-Jr> where does 2184 fit in?
326 2013-04-12 04:46:03 <CodeShark> 2184 focused on exposing the multiwallet functionality not merely in the RPC but in the GUI as well
327 2013-04-12 04:46:26 <CodeShark> however, after doing this I realized that the two parts should be tackled separately
328 2013-04-12 04:46:57 <CodeShark> core and GUI
329 2013-04-12 04:47:40 <CodeShark> I believe 2184 will build and run...haven't done extensive tests on it - but it's a dead end
330 2013-04-12 04:47:55 <CodeShark> that branch is dead
331 2013-04-12 04:48:40 <CodeShark> #2220 was the way forward there
332 2013-04-12 04:49:01 <Luke-Jr> i c
333 2013-04-12 04:51:56 <CodeShark> anything else I can help clarify?
334 2013-04-12 04:54:50 <Luke-Jr> CodeShark: dunno yet
335 2013-04-12 05:08:50 <CodeShark> on a separate note, I don't know what the most appropriate channel for this discussion might be - but this is by far the bitcoin-related channel with the highest signal-to-noise ratio so I'll give it a shot:
336 2013-04-12 05:09:18 <CodeShark> We need to make sure that the serious blunders with the exchanges and the markets do not tarnish the image of bitcoin in the long run
337 2013-04-12 05:09:31 <n1c> I don't think that's possible.
338 2013-04-12 05:09:36 <n1c> The press like headlines.
339 2013-04-12 05:09:45 <MaxValor> and people like an i told you so story
340 2013-04-12 05:09:51 <n1c> +1
341 2013-04-12 05:09:54 <CodeShark> oh, in the short term they'll have a field day with this bs
342 2013-04-12 05:09:57 <CodeShark> no doubt
343 2013-04-12 05:09:59 <MaxValor> play through the pain and think long term
344 2013-04-12 05:10:13 <CodeShark> I'm thinking once all this media frenzy passes
345 2013-04-12 05:10:49 <CodeShark> I believe that the core ideas behind bitcoin are sufficiently sound that they will prove themselves out in the end - but there are people in this channel who could have a big impact on the PR side of things
346 2013-04-12 05:11:33 <CodeShark> surely the media will want interviews and quotes and all that wonderful stuff
347 2013-04-12 05:11:44 <MaxValor> energy best spent else where for the time being: until the radar goes quiet any positive pr publicity will be disregarded atm
348 2013-04-12 05:14:02 <gmaxwell> CodeShark: example question from the press: "Will bitcoin's crash today make it harder for you to buy drugs?"
349 2013-04-12 05:14:08 <CodeShark> lol
350 2013-04-12 05:14:37 <MaxValor> probably not
351 2013-04-12 05:14:44 <MaxValor> folks want something, they buy it
352 2013-04-12 05:14:45 <MaxValor> simple enough
353 2013-04-12 05:17:18 <n1c> Someone the other day said the bitcoin foundation should have a bigger press-presence which is probably a good idea.
354 2013-04-12 05:17:22 <n1c> Have "official" statements etc.
355 2013-04-12 05:17:26 <n1c> "We are not mt.gox, it's a third party"
356 2013-04-12 05:17:27 <n1c> etc.
357 2013-04-12 05:20:01 <MaxValor> yea bitcoin itself is just the users
358 2013-04-12 05:20:11 <MaxValor> the exchanges would be wise to keep their heads down
359 2013-04-12 05:20:25 <MaxValor> anything they say at this point is going to be used against them
360 2013-04-12 05:20:51 <MaxValor> they should however be communicating via fb/web prescence youtube explaining what is happening and what they are doing to try and counter it
361 2013-04-12 05:21:04 <MaxValor> at the very least letting us know they are going down for maint more than 5 mins before they do
362 2013-04-12 05:21:09 <n1c> Greed is a strong power.
363 2013-04-12 05:21:09 <santoscork> Does anyone know what the rough cost of setting up a multi GPU system is? I know about ASICs, like butterfly but I am curious about multi GPU. Any ideas on cost?
364 2013-04-12 05:21:24 <n1c> That's too vague santoscork.
365 2013-04-12 05:21:29 <n1c> It costs what you pay for the hardware?
366 2013-04-12 05:21:30 <MaxValor> n1c greed and fear
367 2013-04-12 05:21:34 <n1c> There is no guessing.
368 2013-04-12 05:21:50 <santoscork> n1c yes, in the most basic terms of cost, just the investment on hardware
369 2013-04-12 05:22:05 <Graet> MaxValor, i have been following them on fb...
370 2013-04-12 05:22:05 <n1c> I think try #bitcoin-mining
371 2013-04-12 05:23:44 <MaxValor> same here
372 2013-04-12 05:23:53 <MaxValor> lots of learning going on across the board I'd say
373 2013-04-12 05:24:14 <MaxValor> to be expected. but folks need to take the lessons to heart, the forex and stock exchanges already learned these things
374 2013-04-12 05:24:59 <CodeShark> seems like bitcoin could use an influx of real financial genius
375 2013-04-12 05:25:13 <CodeShark> in the sense of people with extensive knowledge and understanding of markets
376 2013-04-12 05:25:30 <MaxValor> they are already here
377 2013-04-12 05:25:41 <CodeShark> we need them doing more of the running of the show
378 2013-04-12 05:25:42 <CodeShark> lol
379 2013-04-12 05:25:42 <MaxValor> the problem is that to many laymen have instant nearly feeless access
380 2013-04-12 05:25:48 <CodeShark> and less sitting back
381 2013-04-12 05:26:02 <MaxValor> and the internet generation is all about immediate gratification so lots of arm chair speculation
382 2013-04-12 05:26:15 <MaxValor> the pros take advantage of that to drive price and volume in their favor as they choose
383 2013-04-12 05:26:37 <CodeShark> so the solution would be to require minimum sales/purchases
384 2013-04-12 05:26:39 <CodeShark> minimum size orders
385 2013-04-12 05:26:53 <MaxValor> well now you are regulating
386 2013-04-12 05:26:59 <MaxValor> which is anti everything about bitcoin
387 2013-04-12 05:27:09 <CodeShark> it doesn't have to be dictated top-down
388 2013-04-12 05:27:27 <MaxValor> I do think their are solutions but you have to be careful otherwise you fall in line with all the other currencies in the world
389 2013-04-12 05:27:29 <CodeShark> for instance, say there existed another exchange which only handles larger transactions - but has near 100% availability
390 2013-04-12 05:27:40 <CodeShark> and the orders are executed near-instant
391 2013-04-12 05:27:58 <CodeShark> it would do a LOT to help smoothen out these fluctuations from gox
392 2013-04-12 05:28:05 <MaxValor> I think what you are saying in a nutshell, is that the consumer drives the market and since the exchanges all failed in the last 72 hours, more will pop up
393 2013-04-12 05:28:14 <MaxValor> and because of that these problems will become less drastic
394 2013-04-12 05:28:21 <CodeShark> oh, no doubt more will pop up
395 2013-04-12 05:28:25 <CodeShark> this is the end of gox dominanc
396 2013-04-12 05:28:26 <MaxValor> having one major exchange does not work in a digital world
397 2013-04-12 05:28:31 <MaxValor> I agree
398 2013-04-12 05:28:58 <MaxValor> that and you have to think how much gox made
399 2013-04-12 05:29:07 <MaxValor> their was 10x the volume possible trading that never got to
400 2013-04-12 05:29:21 <MaxValor> so new exchanges will pay for them selves and make some folks some money if done correctly
401 2013-04-12 05:29:44 <CodeShark> I'm sure there are quite a number of capable people out there thinking along these lines, MaxValor
402 2013-04-12 05:29:50 <Graet> still interesting that our community thinks 40% s too big for a pool but happily will allow an exchange to 80% :P
403 2013-04-12 05:30:15 <MaxValor> yep the hurdle for exchange creation will be with legitimate banks and not scaring the shit out of them that this is nothing but a money laundering scheme
404 2013-04-12 05:30:27 <MaxValor> graet that is the discussion that has been brought to light
405 2013-04-12 05:30:33 <MaxValor> and I agree its laughable
406 2013-04-12 05:30:45 <Graet> there iusnt a big enough marker cap or volume for real money launderrs to be interested yet
407 2013-04-12 05:30:49 <MaxValor> all it takes is one person at mt gox to decide and manipulate the market for gain and bam
408 2013-04-12 05:31:02 <Graet> yep
409 2013-04-12 05:31:05 <MaxValor> no but the laws in most countries are very strict about this
410 2013-04-12 05:31:09 <MaxValor> and severe penalties
411 2013-04-12 05:31:15 <Graet> indeed
412 2013-04-12 05:31:21 <MaxValor> so setting up an exchange in some places is a nightmare
413 2013-04-12 05:31:31 <Graet> i'm well aware :)
414 2013-04-12 05:31:36 <MaxValor> esp when you factor that some banks cant/wont send usd to some countries or other currencies
415 2013-04-12 05:31:49 <Graet> luckily i dont deal in usd :)
416 2013-04-12 05:32:13 <MaxValor> well I do, but its not bad since I'm in the USA
417 2013-04-12 05:32:23 <MaxValor> but if your prefered cash out was usd their are some countries you would be sol
418 2013-04-12 05:32:31 <Graet> sure
419 2013-04-12 05:32:43 <MaxValor> fiat is fiat many wash to cash out and exchange and make even more money if you wanted to play multiple markets
420 2013-04-12 05:33:09 <MaxValor> I think however their are better ways to make yourself money
421 2013-04-12 05:35:02 <Graet> yep
422 2013-04-12 05:35:33 <jsfsn> The laws in europe does not seem very strict regarding bitcoin exchange
423 2013-04-12 05:35:43 <MaxValor> probably wont ever be
424 2013-04-12 05:35:53 <MaxValor> however you could have banks in a country stop allowing transactions
425 2013-04-12 05:35:57 <MaxValor> like Israel did earlier
426 2013-04-12 05:36:05 <MaxValor> but their are ways around that to for now
427 2013-04-12 05:36:21 <jsfsn> Transactions to other accounts?
428 2013-04-12 05:36:31 <MaxValor> yea
429 2013-04-12 05:36:35 <MaxValor> dwolla, etc
430 2013-04-12 05:36:47 <MaxValor> several other options to swap it around to bring it into a country
431 2013-04-12 05:36:54 <MaxValor> also could just setup a bank account in another country
432 2013-04-12 05:37:03 <MaxValor> from what I read poland is all about this bitcoin craze
433 2013-04-12 05:37:06 <jsfsn> I think there need to be more local exchanges
434 2013-04-12 05:37:09 <jsfsn> Simplicity of use
435 2013-04-12 05:37:09 <MaxValor> bank to bank xfers are still good to go
436 2013-04-12 05:37:18 <jsfsn> need/s
437 2013-04-12 05:37:22 <MaxValor> I agree, but I think that is something that comes with the maturity of bitcoin
438 2013-04-12 05:37:25 <CodeShark> exchanges need decentralization
439 2013-04-12 05:37:40 <jsfsn> CodeShark, yes
440 2013-04-12 05:37:58 <jsfsn> CodeShark, at least not one exchange with monopol
441 2013-04-12 05:38:07 <CodeShark> decentralization of exchanges is the next logical step in the evolution of digital currency
442 2013-04-12 05:38:23 <MaxValor> yep was an expensive mistake
443 2013-04-12 05:38:51 <MaxValor> but a lot of new venture capitalist esp from sillicon valley see this and are going to buy into the system by creating infrastructure to correct it and thus make themselves a ton of money
444 2013-04-12 05:39:00 <jsfsn> I'm not sure how it would be possible with a real decentrialzed exchange though
445 2013-04-12 05:39:47 <MaxValor> just create more and more
446 2013-04-12 05:39:56 <MaxValor> until the volume is able to be handled
447 2013-04-12 05:40:10 <jsfsn> Low valume exchanges are not good either
448 2013-04-12 05:40:17 <jsfsn> We somehow need price discovery
449 2013-04-12 05:41:11 <jsfsn> Or just give everyone asics and let them mine bitcoin themself (:
450 2013-04-12 05:41:35 <gmaxwell> jsfsn: it's not. Fundimentally the dollar is not decentrialzed. At best you could do is have centeralized dollar tokens that you could exchange, and then you're going to find regulators coming at you with guns ablazing.
451 2013-04-12 05:41:50 <MaxValor> soon as those hit the mining floor mining becomes nearly non profitable
452 2013-04-12 05:42:05 <MaxValor> as the difficulty will adjust
453 2013-04-12 05:42:32 <jsfsn> gmaxwell: Response to my "I do not know how it coud be possible"?
454 2013-04-12 05:42:37 <jsfsn> coud / could
455 2013-04-12 05:43:03 <jsfsn> I really do think that local exchanges are the easiest way of solving this issue
456 2013-04-12 05:43:22 <MaxValor> in an utopia yes
457 2013-04-12 05:43:24 <jsfsn> If it is an issue; I'm not sure
458 2013-04-12 05:43:26 <gmaxwell> jsfsn: yea, true decenteralized exchanges isn't possible.. at least not usefully. A dollar in my hand is not the same thing as a dollar in yours. When we put our dollars into mtgox they become the same thing.
459 2013-04-12 05:43:29 <MaxValor> in reality no. Regulators will cash in
460 2013-04-12 05:43:37 <jsfsn> gmaxwell: I agree
461 2013-04-12 05:43:49 <TD> i think the ripple exchange concept is as decentralized as yo can get
462 2013-04-12 05:43:58 <warren> and yet it isn't
463 2013-04-12 05:44:06 <jsfsn> I'm not sure why it is needed
464 2013-04-12 05:44:13 <TD> how is it not?
465 2013-04-12 05:44:15 <jsfsn> The concept of Bitcoin has nothing to do with other currencies
466 2013-04-12 05:44:18 <MaxValor> http://en.wikipedia.org/wiki/Gresham%27s_law
467 2013-04-12 05:44:28 <MaxValor> when you force local exchanges
468 2013-04-12 05:44:33 <jsfsn> It is not a problem for Bitcoin to solve that is
469 2013-04-12 05:44:33 <MaxValor> the bad money kills the good money
470 2013-04-12 05:44:40 <MaxValor> exactly jsfsn
471 2013-04-12 05:44:45 <TD> bank accounts form a peer to peer system, in a sense. just an appallingly bad one.
472 2013-04-12 05:45:01 <TD> so if you can find a way to get people moving money directly between their accounts with no middleman beyond the banks that already exist, that's a big step forward
473 2013-04-12 05:45:08 <TD> of course the banks won't like it ....
474 2013-04-12 05:45:22 <jsfsn> TD: Sounds like Bitcoin to me
475 2013-04-12 05:45:27 <MaxValor> basically bitcoin undermines all fiat
476 2013-04-12 05:45:29 <warren> banks extend lines of credit to each other
477 2013-04-12 05:45:43 <TD> my preferred decentralized exchange design is this one: https://en.bitcoin.it/wiki/Ripple_currency_exchange
478 2013-04-12 05:45:45 <warren> Ripple relies upon complete trust of issuers
479 2013-04-12 05:45:47 <TD> i don't think you can do better
480 2013-04-12 05:45:53 <TD> at least i haven't seen one
481 2013-04-12 05:46:08 <jrmithdobbs> i'm positive much better could be done
482 2013-04-12 05:46:09 <jsfsn> TD: Why do we even need to care about other currencies?
483 2013-04-12 05:46:14 <TD> not unless you do as gmaxwell says and issue USDcoins, but then you still need an institution that backs 1 USDcoin:1 dollar in a bank
484 2013-04-12 05:46:18 <jrmithdobbs> I've not seen one described adequately though
485 2013-04-12 05:46:24 <TD> jsfsn: seriously? because that's what people already have, maybe?
486 2013-04-12 05:46:58 <jsfsn> TD: Yes, for adoption of Bitcoin, but not for the system it self
487 2013-04-12 05:47:08 <warren> Bitstamp and Weex already issue what is effectively USDcoins
488 2013-04-12 05:47:12 <TD> i'm not sure you can separate the two like that.
489 2013-04-12 05:47:23 <jsfsn> TD: maybe not
490 2013-04-12 05:47:48 <warren> most ripple trade currently is between XRP and USD(bitstamp) or USD(weex)
491 2013-04-12 05:47:51 <jsfsn> TD: It is ironic, however, that we try to come up with a solution to a solution that solves the problem we started out to solve in the first place
492 2013-04-12 05:47:59 <TD> warren: well, what i was thinking of was a genuine merge mined alt-chain that lets you do atomic coin swaps using the hashcoin/luxgladius protocol
493 2013-04-12 05:48:02 <warren> or XRP and BTC(bitstamp) or BTC(weex)
494 2013-04-12 05:48:16 <TD> warren: once you get some USDcoins you can then trade them against bitcoins in a pure p2p network
495 2013-04-12 05:48:23 <TD> of course, it just pushes the underlying problems around, not really solves them
496 2013-04-12 05:48:31 <warren> yes, that's the real problem
497 2013-04-12 05:48:36 <jsfsn> TD: That just moves the problem
498 2013-04-12 05:48:36 <warren> you still have to trust an issuer
499 2013-04-12 05:49:01 <warren> If you have to trust the issuer, you might as well just centralize trade there.
500 2013-04-12 05:49:17 <jsfsn> A decentrialzed exchange for BTC / LTC could me bade though
501 2013-04-12 05:49:17 <MaxValor> which is why I don't like ripple
502 2013-04-12 05:49:30 <jsfsn> bade / made
503 2013-04-12 05:49:53 <TD> this is why i say that direct paypal/cash/wire transfers between participants in the market is as decentralized as you can get (assisted by a new p2p network of course)
504 2013-04-12 05:50:01 <wumpus> libertyreserve is basically 'USDcoin' (minus the block chain) :p
505 2013-04-12 05:50:02 <robbak> Q - what is the schedule for the 8.2 update? Any changes that we should be aware of?
506 2013-04-12 05:50:05 <TD> the problem is if your traffic gets too high, you risk bank account closure.
507 2013-04-12 05:52:52 <MaxValor> that could be an issue since banks can and will do it with no notice
508 2013-04-12 05:52:58 <MaxValor> would be bad to just lose your junk
509 2013-04-12 05:55:54 <jsfsn> Anyway, I'm trying to get my head around raw transactions
510 2013-04-12 05:56:28 <jsfsn> Would it be possible to create a cold wallet, fill the key pool up to a million, encrypt it and use it as a hot one?
511 2013-04-12 05:56:40 <CodeShark> how do you spend?
512 2013-04-12 05:56:42 <jsfsn> Can it create raw transactions unencrypted?
513 2013-04-12 05:56:57 <CodeShark> how do you sign the transactions without the keys?
514 2013-04-12 05:56:59 <jsfsn> CodeShark, you take the raw transaction created by the hot one, move it to the cold and sign it and back
515 2013-04-12 05:57:29 <jsfsn> ie. the decryption key never gets online
516 2013-04-12 05:57:36 <CodeShark> yes, you can do that - although I'm not sure I'd call that a hot wallet
517 2013-04-12 05:57:56 <jsfsn> No, I probably call it a cold one, but for this explination I found it simplier (:
518 2013-04-12 05:58:22 <jsfsn> Is it a stable solution?
519 2013-04-12 05:58:39 <CodeShark> you still have the bottleneck of having to send the transactions to the offline machine for signing
520 2013-04-12 05:58:39 <jsfsn> as long as the keypool does not need to be extended
521 2013-04-12 05:58:50 <CodeShark> computing the inputs and outputs to the transaction is the cheap part
522 2013-04-12 05:58:51 <jsfsn> Yes
523 2013-04-12 05:59:21 <CodeShark> so I'm not sure that really solves much of the problem
524 2013-04-12 06:00:06 <jsfsn> Well, it solves the problem the server with the hot wallet being hacked
525 2013-04-12 06:00:10 <CodeShark> the purpose of what I would call a hot wallet is to allow for quick processing of payments and withdrawals - without the bottleneck of offline signing
526 2013-04-12 06:00:10 <jsfsn> "hot" that is
527 2013-04-12 06:00:32 <jsfsn> A hot wallet will exist as well
528 2013-04-12 06:00:50 <jsfsn> This procedure is to fill up the real hot wallet with funds
529 2013-04-12 06:01:31 <jsfsn> I realize that there might be a better solution though
530 2013-04-12 06:01:53 <warren> Put the keys on a computer that is firewall limited to do only ONE thing on the network, sync the block chain. Is that good enough?
531 2013-04-12 06:02:37 <jrmithdobbs> no
532 2013-04-12 06:02:41 <jrmithdobbs> man wipe
533 2013-04-12 06:03:06 <jrmithdobbs> if it's not installed and you don't have access to ubuntu manpages it's worth installing just to read
534 2013-04-12 06:03:09 <jrmithdobbs> fyi
535 2013-04-12 06:03:25 <jrmithdobbs> (best.man.page.ever.)
536 2013-04-12 06:03:52 <jrmithdobbs> warren: semi-serious re: relevence.
537 2013-04-12 06:03:56 <warren> (how is this relevant?)
538 2013-04-12 06:04:19 <jrmithdobbs> it addresses your question directly, believe or not, read it
539 2013-04-12 06:05:22 <jrmithdobbs> warren: http://linux.die.net/man/1/wipe
540 2013-04-12 06:05:36 <warren> I'm reading it. I don't see how this is relevant.
541 2013-04-12 06:06:32 <jrmithdobbs> did you not get to the DHS conspiracy theory part yet?
542 2013-04-12 06:06:45 <jrmithdobbs> re: tampered with hardware?
543 2013-04-12 06:07:14 <jrmithdobbs> the guy's a bit nuts (that's why it's entertaining) but his points are valid
544 2013-04-12 06:07:32 <warren> If you lose physical control all bets are off.
545 2013-04-12 06:07:44 <jrmithdobbs> you never had it to lose, is the point.
546 2013-04-12 06:07:52 <jrmithdobbs> so what are you betting on? :)
547 2013-04-12 06:10:33 <warren> If you lose physical control all bets are off.
548 2013-04-12 06:14:13 <warren> btw, does the reference client let you symlink or bind mount your wallet.dat to a different directory?
549 2013-04-12 06:14:39 <warren> (This sort of thing breaks in other apps because they create a new file and rename.)
550 2013-04-12 06:22:35 <pjorrit_> wow love that wipe man page :)
551 2013-04-12 06:25:37 <jsfsn> warren: I'm not sure I like a cold wallet to have internet access at all
552 2013-04-12 06:26:07 <warren> jsfsn: doesn't need internet access, only access to a node
553 2013-04-12 06:26:17 <warren> jsfsn: and that's the only permitted network traffic
554 2013-04-12 06:26:46 <jsfsn> It still needs to be "online"
555 2013-04-12 06:27:09 <sipa> real cold storage isn't connected to any network
556 2013-04-12 06:27:18 <jsfsn> exactly
557 2013-04-12 06:27:26 <sipa> you create transaction-to-be-signed on another machine
558 2013-04-12 06:27:39 <sipa> transport them without network to the cold storage
559 2013-04-12 06:27:41 <sipa> sign them
560 2013-04-12 06:27:45 <sipa> transport them back
561 2013-04-12 06:27:46 <jsfsn> Thats my thought as well
562 2013-04-12 06:27:48 <sipa> and broadcast them
563 2013-04-12 06:28:04 <jsfsn> And just keep the key pool pregenerated
564 2013-04-12 06:28:14 <jsfsn> Are there other issues than that?
565 2013-04-12 06:28:19 <sipa> use a client that has deterministic wallets
566 2013-04-12 06:29:12 <warren> can one instance of the deterministic wallet be public key only?
567 2013-04-12 06:29:17 <jsfsn> Armory then, is it wise to not use the reference client in production?
568 2013-04-12 06:29:29 <jsfsn> warren: Yes
569 2013-04-12 06:29:42 <gmaxwell> you can't use armory without the reference software.
570 2013-04-12 06:29:42 <sipa> warren: type-2 deterministic wallets can, which is the only type that is actually used
571 2013-04-12 06:29:57 <warren> that sounds great
572 2013-04-12 06:30:01 <jsfsn> gmaxwell: *only* the reference client that is
573 2013-04-12 06:30:51 <gmaxwell> for a manual totally offline wallet? It doesn't really facilitate that use case today.
574 2013-04-12 06:31:07 <gmaxwell> you can do it, but it's a bit kludgy.
575 2013-04-12 06:31:46 <warren> There's a thread on trolltalk, it sounds like people's server-side stored blockchain.info wallets are being brute forced cracked. People with weaker passwords are admitting it. Given you have no rate limit in brute forcing a copy of the encrypted wallet, if their encrypted wallet storage has been copied then many are at risk.
576 2013-04-12 06:31:55 <warren> apparently many wallets were broken and sent out in a single 25KB tx
577 2013-04-12 06:32:05 <warren> 108 BTC
578 2013-04-12 06:33:33 <warren> I have a bunch of previously used keys in blockchain wallet. I won't reuse them. Now I need a solution to keep them in retirement but to be notified if something comes in.
579 2013-04-12 06:34:47 <CodeShark> that application has actually been at the forefront of most of my involvement in bitcoin development, warren :p
580 2013-04-12 06:35:25 <CodeShark> or not that exact use case - but the ability to monitor specific keys and be alerted
581 2013-04-12 06:35:40 <warren> oh? in which app?
582 2013-04-12 06:36:13 <CodeShark> I'm currently using this in my projects: https://github.com/CodeShark/CoinClasses/tree/master/examples/listener2
583 2013-04-12 06:36:27 <CodeShark> I attach a filter to it and some sort of streaming/queue/messaging server
584 2013-04-12 06:37:17 <CodeShark> then I connect it to a trusted bitcoind instance via p2p
585 2013-04-12 06:38:37 <TD> warren: i was wondering about the anti-brute forcing logic there for a long time. can you obtain a wallet using just a username? or username+password too ...
586 2013-04-12 06:38:56 <TD> anti-bruteforcing for passwords is a remarkably hard problem, especially if you don't want to let people DoS each others accounts
587 2013-04-12 06:40:30 <warren> TD: yes, with just a username
588 2013-04-12 06:40:57 <warren> there's a way to use blockchain.info without syncing your wallet to the server, I think
589 2013-04-12 06:41:07 <warren> I previously assumed "eh, nobody could know this username"
590 2013-04-12 06:41:17 <warren> but *one* compromise of the server can copy them all
591 2013-04-12 06:41:38 <TD> just a username?
592 2013-04-12 06:41:42 <TD> uhhh ... that's not good
593 2013-04-12 06:41:53 <TD> i wonder how the wallets are encrypted. what kind of kdf they're using.
594 2013-04-12 06:42:00 <warren> you type in a username and password. It downloads the encrypted wallet and opens it locally.
595 2013-04-12 06:42:09 <TD> the server doesn't check the password
596 2013-04-12 06:42:12 <CodeShark> yeah, I thought it was all browser-based
597 2013-04-12 06:42:13 <warren> nope
598 2013-04-12 06:42:32 <warren> It has some ind of 2nd factor login, but I'm not sure what that does
599 2013-04-12 06:42:37 <TD> i suppose it can't. otherwise they'd have the users passwords and could decrypt the wallets indeed.
600 2013-04-12 06:42:39 <TD> ok
601 2013-04-12 06:42:51 <jonass> can anyone help creating a Qt4.8 Kit in QTCreator 2.7 (mac)?
602 2013-04-12 06:43:31 <warren> I like blockchain's implementation and functionality. If I started a new account that was *never* uploaded to the server it might be just fine.
603 2013-04-12 06:43:32 <CodeShark> so the exploit is only for weak passwords? i.e. dictionary attacks?
604 2013-04-12 06:43:37 <warren> It lets you download their app to run locally.
605 2013-04-12 06:43:54 <warren> exploit is brute forcing encrypted wallets in general
606 2013-04-12 06:44:01 <warren> if you stolen many wallets
607 2013-04-12 06:44:17 <warren> the cost of cracking them is likely much cheaper than the coins inside
608 2013-04-12 06:44:26 <CodeShark> sure
609 2013-04-12 06:44:42 <CodeShark> well, you'll probably crack at least a couple
610 2013-04-12 06:44:45 <TD> it shouldn't be possible to brute force encrypted wallets though, unless you used a password like "passwodr"
611 2013-04-12 06:44:48 <CodeShark> even if you haven't stolen that many :p
612 2013-04-12 06:44:57 <TD> but i guess if the KDF isn't strong enough, it could be an issue
613 2013-04-12 06:45:02 <TD> and blockchain has to calculate it in javascript
614 2013-04-12 06:45:04 <warren> the complaints in forums were dictoinary words plus number type of passwords
615 2013-04-12 06:45:07 <TD> so i suppose it must be somewhat weaker than native code
616 2013-04-12 06:45:12 <TD> hmm
617 2013-04-12 06:45:13 <TD> yeah. ok.
618 2013-04-12 06:45:44 <warren> still, that scared me enough that I won't put money anymore. I rather not give anyone the opportunity to brute force my key.
619 2013-04-12 06:45:52 <TD> things are so crazy lately. massive ddos attacks against the exchanges, the pools, brute forcing of wallets. sigh. i wish bitcoin would scale up slowly
620 2013-04-12 06:46:05 <warren> it gets worse!
621 2013-04-12 06:46:08 <TD> warren: how to manage backups of wallets is something that the community still hasn't really nailed, for sure
622 2013-04-12 06:46:14 <CodeShark> even with a good KDF, if you've got a thousand wallets, you can get 1000 times the cracking throughput since testing the key on all of them is cheap
623 2013-04-12 06:46:25 <TD> CodeShark: they're supposed to be salted to prevent that. no?
624 2013-04-12 06:46:26 <warren> XSS expoits on random links in BTC-e trollbox stole 34 bitcoins
625 2013-04-12 06:46:37 <CodeShark> I would think so, TD - haven't looked at the source
626 2013-04-12 06:46:39 <TD> warren: that looked more like a generic exploit kit rather than XSS
627 2013-04-12 06:47:00 <CodeShark> I mean, I would think so if it were properly implemented
628 2013-04-12 06:47:32 <warren> I'm thinking to put most of my coins in cold storage safe deposit box. so I don't think about trading.
629 2013-04-12 06:48:34 <TD> it looks like b.i might be using 10 rounds of sha256 as their kdf
630 2013-04-12 06:48:49 <TD> https://github.com/blockchain/My-Wallet/blob/master/wallet.js#L30
631 2013-04-12 06:49:38 <jonass> anyone time for help gettint Qt4 Kit ready in qtcreator 2.7?
632 2013-04-12 06:51:38 <CodeShark> salted, TD?
633 2013-04-12 06:51:39 <CodeShark> hm
634 2013-04-12 06:51:41 <kadoban> TD: huh...if that's true, isn't that quite low?
635 2013-04-12 06:51:42 <CodeShark> let me look
636 2013-04-12 06:51:58 <TD> i might be wrong, it's a bit hard to follow
637 2013-04-12 06:52:00 <TD> i'm still looking
638 2013-04-12 06:52:24 <CodeShark> lines 253-257?
639 2013-04-12 06:52:40 <TD> kadoban: 10 rounds of sha256 is not something that's going to stop password brute forcing, no. the new wallet encryption in bitcoinj uses scrypt which is supposed to be much better and a very large number of rounds, such that it takes JIT compiled java (i.e. native code) a few seconds
640 2013-04-12 06:52:48 <CodeShark> var round_data = Crypto.SHA256(sharedKey + dpassword, {asBytes: true});
641 2013-04-12 06:52:58 <TD> well i'm not sure what "double encryption" is meant to be
642 2013-04-12 06:53:19 <TD> line 2204
643 2013-04-12 06:53:32 <TD> it's sending the password to some crypto library and talking about pbkdf2
644 2013-04-12 06:53:45 <CodeShark> where does sharedKey come from?
645 2013-04-12 06:54:30 <TD> https://github.com/blockchain/My-Wallet/blob/master/bitcoinjs-lib/src/crypto-js/pbkdf2.js
646 2013-04-12 06:54:31 <TD> yeah
647 2013-04-12 06:54:34 <TD> looks like 10 rounds of sha1 even
648 2013-04-12 06:54:37 <TD> not sure
649 2013-04-12 06:55:16 <CodeShark> line 3565 sharedKey = body.data('sharedkey');
650 2013-04-12 06:55:27 <TD> kadoban: that said apparently litecoin managed to stimulate the development of scrypt GPU acceleration, so perhaps it's not quite as strong as we'd hoped.
651 2013-04-12 06:55:37 <TD> kadoban: which is annoying.
652 2013-04-12 06:56:08 <TD> kadoban: well, client-side wallet encryption is only a short term thing anyway. longer term the only solution is either that people use secure operating systems or that they use secure hardware like trezor
653 2013-04-12 06:56:11 <CodeShark> it seems the problem with litecoin's use of scrypt is that difficulty is set the same way as with bitcoin rather than by upping the memory requirements on the algorithm
654 2013-04-12 06:56:12 <kadoban> TD: heh, ya that was maybe short sighted. ugh on the 10 rounds though, maybe javascript is too slow on some browsers for more, but man would i not trust that
655 2013-04-12 06:56:25 <TD> but in the meantime it could cause a lot of pain
656 2013-04-12 06:56:26 <CodeShark> there's a target hash, a valid hash must be smaller than that
657 2013-04-12 06:56:42 <CodeShark> the memory requirements or iterations or other parameters to the scrypt algorithm are hard-coded
658 2013-04-12 06:56:59 <kadoban> of course i guess if you use a /strong/ passphrase it's probably fine
659 2013-04-12 06:57:13 <CodeShark> of course, if they weren't it would make verification harder as well as mining
660 2013-04-12 06:57:39 <CodeShark> but the benefits of scrypt over sha256 in expense of hardware is mostly in the memory requirements
661 2013-04-12 06:57:57 <CodeShark> not in the target hash
662 2013-04-12 06:58:10 <CodeShark> I mean, in increasing memory requirements
663 2013-04-12 06:58:15 <CodeShark> to increase difficulty
664 2013-04-12 06:58:39 <TD> yes, i haven't looked at how the gpu scrypt acceleration works
665 2013-04-12 06:58:56 <TD> gfx cards have so much texture mem these days i'm not sure you can really build a hashing algorithm that is robust against it
666 2013-04-12 07:04:18 <wumpus> you'd want some algorithm that uses data structures with lots of levels of indirection, and wildly data dependent control flow
667 2013-04-12 07:04:29 <CodeShark> yeah, lots and lots and lots of branching
668 2013-04-12 07:04:33 <gmaxwell> "
669 2013-04-12 07:04:33 <gmaxwell> ???We have elected to put our money and faith in a mathematical framework that is free of politics and human error,??? Tyler Winklevoss said.
670 2013-04-12 07:04:36 <gmaxwell> "
671 2013-04-12 07:04:37 <gmaxwell> 0_o
672 2013-04-12 07:04:49 <CodeShark> haha
673 2013-04-12 07:04:51 <gmaxwell> Major points for getting it.
674 2013-04-12 07:04:52 <TD> lol
675 2013-04-12 07:05:40 <gmaxwell> (context: http://dealbook.nytimes.com/2013/04/11/as-big-investors-emerge-bitcoin-gets-ready-for-its-close-up/?hp which I'd missed until now)
676 2013-04-12 07:05:56 <TD> it's funny how they're both simultaneously right and wrong
677 2013-04-12 07:07:02 <gmaxwell> Yes, well they get the idea, of course our execution of that idea is far from perfect. :) But it's arguably superior to at least have that goal.
678 2013-04-12 07:08:02 <TD> my respect for them definitely increased after reading that article
679 2013-04-12 07:09:19 <CodeShark> gox's "we're victims of our own success" memo will continue to come back to bite them in the ass
680 2013-04-12 07:10:06 <CodeShark> but yeah, glad to see the winklevosses are behind bitcoin
681 2013-04-12 07:11:07 <jsfsn> Something with the matching algo @ gox seem weird
682 2013-04-12 07:11:24 <jsfsn> It is not trivial, but they does not seem to handle more than 1 match / s
683 2013-04-12 07:11:36 <_dr> since we're already OT, what's the reason for gox to be down right now? the usual maintenance/ddos/incompetence?
684 2013-04-12 07:11:37 <CodeShark> there are many serious technical problems with gox's system which cannot just be patched over
685 2013-04-12 07:12:00 <gmaxwell> _dr: dunno ask in #mtgox ??? there were some really major (apparently >80gbit/sec) dos attacks earlier
686 2013-04-12 07:12:35 <_dr> nice. i'm getting really curious about who's behind this
687 2013-04-12 07:12:51 <TD> presumably whoever it is has an AML verified account at the exchange
688 2013-04-12 07:12:59 <TD> but there's probably no way to link them
689 2013-04-12 07:13:11 <jsfsn> I'm not sure
690 2013-04-12 07:13:12 <gmaxwell> _dr: I assume it's the standard model that happens to banks, pools, and, esp., casinos: "Pay us protection money or we dos you"
691 2013-04-12 07:13:25 <CodeShark> lol
692 2013-04-12 07:13:25 <TD> or as they already said, attempts to push down the price and buy low
693 2013-04-12 07:13:26 <jsfsn> Is the price drop really correlated with the ddos's?
694 2013-04-12 07:13:37 <TD> i doubt it's protection money
695 2013-04-12 07:13:38 <_dr> fud
696 2013-04-12 07:13:46 <gmaxwell> TD: or both.
697 2013-04-12 07:13:51 <TD> "please wire us $1M to this account .... oh crap the police are here"
698 2013-04-12 07:13:56 <CodeShark> the price drop is most certainly correlated with Gox being taken offline whether it was deliberate or not
699 2013-04-12 07:14:00 <TD> "please send us bitcoins to address 1abc .... oh crap, how do i sell them"
700 2013-04-12 07:14:05 <jsfsn> I'd be more worried about ddos on the biggest pools
701 2013-04-12 07:14:35 <Scrat> gmaxwell: in which case you have utterly failed as a software engineer because ddos shouldn't be allowed to reach the trading engine
702 2013-04-12 07:14:43 <gmaxwell> TD: I know they get demands for protection money, hell even smallish pools do. But I don't know if any of those people are really much of a threat to gox since they have commercial anti-ddos service.
703 2013-04-12 07:15:21 <gmaxwell> Scrat: multiple layers of attack??? the ddos is mostly blocked by prolexic until they're totally overrun. The trading engine issue is obviously another kind of attack to whatever extent there is an attack at all.
704 2013-04-12 07:15:21 <TD> Scrat: lay off it. it's easy to be an armchair CEO. but there are only so many hours in each day and many things that must be done to run a place like mtgox.
705 2013-04-12 07:15:38 <gmaxwell> Scrat: also whats with the "you"?
706 2013-04-12 07:15:45 <jsfsn> The 0.002 trades are probably worse than the ddoses
707 2013-04-12 07:15:51 <Scrat> gmaxwell: well not you, in general
708 2013-04-12 07:15:51 <wumpus> could be just a user stampede attack
709 2013-04-12 07:15:56 <n1c> I guess, it is a form of attack.
710 2013-04-12 07:16:03 <gmaxwell> Scrat: I look forward to trading on your exchange. :)
711 2013-04-12 07:16:50 <CodeShark> place an order for 0.0001, cancel repeat
712 2013-04-12 07:17:10 <jsfsn> They could switch to an in-memory algo, but not trivial to keep the state sane in case of a drop
713 2013-04-12 07:17:20 <Scrat> TD: alright I won't mention it again
714 2013-04-12 07:17:48 <CodeShark> jsfsn, there are many far larger exchanges out there who seem to cope with the volume
715 2013-04-12 07:17:52 <wumpus> I hope the problems with mtgox will cause people to look for alternatives and make the exchange landscape a bit more varied
716 2013-04-12 07:18:04 <jsfsn> CodeShark, yes, I'm just not sure how safe they really are
717 2013-04-12 07:18:21 <CodeShark> the gox system simply was not designed to deal with this level of traffic/orders/trades/transactions/site visits
718 2013-04-12 07:18:33 <CodeShark> and no amount of patching it up will really address the fundamental problems
719 2013-04-12 07:18:46 <jsfsn> The fundamental problem is quite small I believe
720 2013-04-12 07:18:52 <CodeShark> oh yeah?
721 2013-04-12 07:18:53 <jsfsn> It is the single point of order matching that locks
722 2013-04-12 07:19:05 <jsfsn> Small != easy though
723 2013-04-12 07:19:22 <CodeShark> but that's just one of their problems
724 2013-04-12 07:19:37 <jsfsn> I would say that is the critical one
725 2013-04-12 07:19:40 <CodeShark> I mean, that seems to be a fundamental problem with their trading engine
726 2013-04-12 07:19:43 <TD> i got quoted by the economist as a "Bitcoin expert"
727 2013-04-12 07:19:47 <TD> ACTION can die happy now
728 2013-04-12 07:19:48 <gmaxwell> wumpus: lets hope lots of people don't land on an exchange that turns out to be a scam.
729 2013-04-12 07:19:52 <CodeShark> TD: nice :)
730 2013-04-12 07:20:08 <jsfsn> Well, the trading engines problem is, most likely, due to the wide lock while matching orders
731 2013-04-12 07:20:29 <CodeShark> yes, probably, jsfsn
732 2013-04-12 07:20:37 <CodeShark> which is why canceling an order takes an eternity
733 2013-04-12 07:20:53 <CodeShark> the other problem is the prioritization
734 2013-04-12 07:20:54 <jsfsn> I'm obviously not sure, but my guess is that they do inserts for each trade