1 2012-02-07 00:00:08 <midnightmagic> :-)  I'll put his name in the blockchain.
  2 2012-02-07 00:00:18 <BlueMatt> heh
  3 2012-02-07 00:00:23 <BlueMatt> for some reason I doubt he'll see that
  4 2012-02-07 00:00:32 <tcatm> BlueMatt: https://github.com/bitcoin/bitcoin.org/tree/contribute adds a subpage (contribute.html) which can be filled with content easily
  5 2012-02-07 00:00:34 <gmaxwell> "What do you mean you didn't feel welcome? We like totally put your name in the blockchain twice and everything"
  6 2012-02-07 00:00:37 <midnightmagic> "altamic come back okay" -> burn 0.01
  7 2012-02-07 00:00:39 <BlueMatt> email him https://github.com/altamic
  8 2012-02-07 00:00:58 <sipa> gmaxwell: how do you mean you've never been at alpha centauri?
  9 2012-02-07 00:01:46 <BlueMatt> gmaxwell: wait, you've never space traveled? damn you need to try that
 10 2012-02-07 00:01:49 <gmaxwell> "But the coinbase said beware the jaguar!" "Oh that, it's just a preamble"
 11 2012-02-07 00:02:03 <HostFat> another idea: can you add something like a blockchain clock / dynamic value ( something simple ) on the homepage?
 12 2012-02-07 00:02:27 <HostFat> it should be really simple, but it need to change
 13 2012-02-07 00:02:39 <HostFat> the block number is already a good idea
 14 2012-02-07 00:02:58 <BlueMatt> tcatm: cool, but remove the other three or so dev links already on the main page first (IMHO)
 15 2012-02-07 00:03:06 <tcatm> Can't be added easily.
 16 2012-02-07 00:03:14 <tcatm> @HostFat
 17 2012-02-07 00:04:28 <tcatm> BlueMatt: Yep. I don't have much time for it, though :/
 18 2012-02-07 00:05:04 <BlueMatt> mmm, alright
 19 2012-02-07 00:05:27 <Host> Away|good night :)
 20 2012-02-07 00:16:38 <BlueMatt> ooo, french court ordered lenovo to pay back the cost of windows licenses to a group who didnt want them, cheaper linux laptops here we come
 21 2012-02-07 00:18:42 <sipa> ... in france
 22 2012-02-07 00:19:48 <gribble> New news from bitcoinrss: TheBlueMatt opened pull request 16 on bitcoin/bitcoin.org <https://github.com/bitcoin/bitcoin.org/pull/16>
 23 2012-02-07 00:20:29 <BlueMatt> sipa: it can be used as precedent anywhere in the eu
 24 2012-02-07 00:22:38 <sipa> except the eu countries don't have precedent law :)
 25 2012-02-07 00:25:00 <gmaxwell> BlueMatt: I had this notion that if I bought a lenovo and then rejected the license they'd refund me the windows rather than take the cost of referbing a post consumer used good..
 26 2012-02-07 00:25:13 <gmaxwell> Nope. They happily let me refund the whole machine but wouldn't refund windows
 27 2012-02-07 00:25:32 <sipa> heh?
 28 2012-02-07 00:25:55 <BlueMatt> until you send a very legal-looking letter their way claiming unfair competition and monopolistic practices.  I guarantee you they would pay you the 50$ instead of getting sued
 29 2012-02-07 00:26:30 <BlueMatt> sipa: ok, not precedent, but I was still under the impression such a ruling could be applied in other eu countries
 30 2012-02-07 00:26:32 <gmaxwell> I wondered a bit what would happen if I then ordered 20. Opened them and returned them... but I didn't want to bet tens of thousands of dollars to troll them. :)
 31 2012-02-07 00:27:19 <BlueMatt> gmaxwell: how high did you get in the phone-ops chain?
 32 2012-02-07 00:27:30 <k9quaint> gmaxwell: btw, your nick evaluates now in my repl
 33 2012-02-07 00:27:51 <gmaxwell> Not very, I don't have have much patience for that sort of thing.
 34 2012-02-07 00:27:53 <k9quaint> and you should buy one, return, and then buy another, return, and just constantly have a lap top being shipped
 35 2012-02-07 00:27:55 <BlueMatt> did you while [ true ]; do echo "let me speak to your manager"; done
 36 2012-02-07 00:28:38 <gmaxwell> I did at least some of that. But they were quite eager to give me my money back.. and I knew taking the system back would be costly to them, so I still won
 37 2012-02-07 00:28:51 <BlueMatt> meh, fair enough
 38 2012-02-07 00:29:09 <gmaxwell> (I ended up buying a used one in any case, which I still ended up paying for windows on, but at least less)
 39 2012-02-07 00:29:15 <k9quaint> I just sold my windows key off my laptop
 40 2012-02-07 00:30:30 <gmaxwell> BlueMatt: they actually have a hidden link to buy laptops without windows, in fact.
 41 2012-02-07 00:30:33 <gmaxwell> (now)
 42 2012-02-07 00:30:45 <BlueMatt> they only offer it on a very few models (last I checked)
 43 2012-02-07 00:30:56 <BlueMatt> mostly old models
 44 2012-02-07 00:31:12 <gmaxwell> Correct. Thought not just old models. (well, depends on when you look I guess)
 45 2012-02-07 00:31:44 <k9quaint> so ironic, a chinese company being a stickler for IP :P
 46 2012-02-07 00:31:59 <BlueMatt> anyway, someone should do a mega-class-action against m$+vendors for monopolistic practices
 47 2012-02-07 00:32:27 <BlueMatt> because it really is ridiculous that you are forced to pay the license fee
 48 2012-02-07 00:32:31 <gmaxwell> k9quaint: not ironic, kickbacks for engaging in a business practice which has been found unlawful several times.
 49 2012-02-07 00:32:48 <k9quaint> gmaxwell: that too
 50 2012-02-07 00:33:07 <k9quaint> I just don't patronize those vendors
 51 2012-02-07 00:33:07 <sipa> BlueMatt: not sure you have a case, unless microsoft somehow forces them not to sell things without windows
 52 2012-02-07 00:33:26 <BlueMatt> sipa: how much you wanna bet that is in the contract with some of them?
 53 2012-02-07 00:33:37 <k9quaint> I bet it isnt in the contract
 54 2012-02-07 00:33:43 <sipa> BlueMatt: i'd guess it's off the contract :)
 55 2012-02-07 00:33:45 <sipa> yeah
 56 2012-02-07 00:33:58 <k9quaint> I bet its a nudge and a wink between SVPs
 57 2012-02-07 00:34:04 <BlueMatt> Id bet if you get the time to really get to discovery you could find some really nice emails
 58 2012-02-07 00:34:04 <gmaxwell> BlueMatt: what msft got in trouble for w/ US vendors was writing contracts that charged per system sold regardless of what they shipped on them.
 59 2012-02-07 00:34:28 <BlueMatt> you may have a case against eg dell
 60 2012-02-07 00:34:29 <k9quaint> lenovo is probaby just being lazy
 61 2012-02-07 00:35:26 <BlueMatt> someone with a business contract should start a side business of reselling lenovos, etc without windows licenses
 62 2012-02-07 00:35:31 <BlueMatt> (Im sure businesses get them)
 63 2012-02-07 00:35:39 <BlueMatt> so they can use their own site licenses
 64 2012-02-07 00:36:01 <gmaxwell> there is a company that does that (linuxlaptops or something) but their selection is very limited, they can't do CTO systems
 65 2012-02-07 00:36:16 <BlueMatt> mmm, thats annoying
 66 2012-02-07 00:37:35 <gmaxwell> might be more interesting to see someone buy them, 'unbundle' the windows licenses, and then invite msft to sue them (works best if your company consists of nothing but bored lawyers)
 67 2012-02-07 00:37:50 <BlueMatt> that one would be fun
 68 2012-02-07 00:38:26 <k9quaint> gmaxwell: sounds like a new business unit for Rambus
 69 2012-02-07 00:38:27 <BlueMatt> or start posting the licenses on pirate sites and see if m$ sues them (they are just a code and its not like the company had any eulas with m$ they signed)
 70 2012-02-07 00:38:34 <gmaxwell> esp post vernor v autodesk.
 71 2012-02-07 00:38:55 <k9quaint> is that the first sale doctrine = software case?
 72 2012-02-07 00:38:58 <gmaxwell> yes
 73 2012-02-07 00:39:13 <k9quaint> that one warmed my no-no place
 74 2012-02-07 00:40:39 <BlueMatt> that one was f'd up
 75 2012-02-07 00:59:50 <BlueMatt> gmaxwell: am I reading this correct that in https://github.com/bitcoin/bitcoin/compare/30999ec6f9...0b9a05a2bc you changed bitcoind to accept blank users?
 76 2012-02-07 01:00:04 <luke-jr> BlueMatt: it always has
 77 2012-02-07 01:00:16 <gmaxwell> BlueMatt: I changed it to not accept blank passwords.
 78 2012-02-07 01:00:27 <BlueMatt> ...
 79 2012-02-07 01:00:27 <gmaxwell> (it still accepts blank users"
 80 2012-02-07 01:00:28 <BlueMatt> wow
 81 2012-02-07 01:00:28 <gmaxwell> )
 82 2012-02-07 01:00:29 <BlueMatt> ...
 83 2012-02-07 01:00:47 <luke-jr> BlueMatt: for a long time, I used blank username on Eligius with a long password
 84 2012-02-07 01:00:51 <luke-jr> nobody will ever guess that ;)
 85 2012-02-07 01:02:32 <gmaxwell> Blame me if someone is mad that their complicated user with blank password causes them issues. .. but I will not feel very guilty about that.
 86 2012-02-07 01:09:45 <k9quaint> *poof* all bugs = features
 87 2012-02-07 01:15:27 <FROTUSCI> cool
 88 2012-02-07 01:19:32 <luke-jr> http://edgemarcvoiprouters.com/ <-- manuf site of some of the routers DDoSing Eligius -.-
 89 2012-02-07 02:01:35 <JFK911> how are the routers responsible
 90 2012-02-07 02:02:24 <luke-jr> JFK911: they're originating the DDoS
 91 2012-02-07 02:02:38 <JFK911> they are giving out machines that tickle your service?
 92 2012-02-07 02:03:07 <luke-jr> apparently the company has no security
 93 2012-02-07 02:03:13 <JFK911> oh
 94 2012-02-07 02:03:38 <cjd> so like someone owned a ton of those voip routers and are using them to send udp rape?
 95 2012-02-07 02:03:49 <cjd> or using their webserver?
 96 2012-02-07 02:04:29 <JFK911> ive never heard of the company, so i was thinking maybe they are really small, and have the same 19 year old kid as a support rep and firmware developer
 97 2012-02-07 02:05:09 <k9quaint> cheap router firmware is notorious for being buggy and insecure
 98 2012-02-07 02:05:17 <k9quaint> same with printers
 99 2012-02-07 02:05:54 <BlueMatt> contact their host
100 2012-02-07 02:06:09 <cjd> http://www.edgewaternetworks.com/ <-- there's the "real" edgemark, looks like a reseller
101 2012-02-07 02:06:09 <JFK911> maybe the way to the bottom is to get one  of the devices, wait for it to be found, and see who tickles it
102 2012-02-07 02:06:11 <k9quaint> root them back, and install the null set as firmware
103 2012-02-07 02:06:17 <JFK911> haha
104 2012-02-07 02:06:42 <cjd> I gather it's their webserver seeing as it's showing a deface page
105 2012-02-07 02:21:16 <luke-jr> FWIW, just had a UDP flood from Deepbit's IP too
106 2012-02-07 02:21:24 <luke-jr> so someone is obviously trying to frame them
107 2012-02-07 02:23:44 <BlueMatt> you need to start sending out some spam abuse emails to hosts and hosts of hosts
108 2012-02-07 02:26:48 <luke-jr> random ISPs? :p
109 2012-02-07 02:27:01 <luke-jr> this attack seems to be using an exploit in DD-Wrt for amplification
110 2012-02-07 02:27:02 <NxTitle> luke-jr: are you still being attacked?
111 2012-02-07 02:27:05 <luke-jr> NxTitle: yes
112 2012-02-07 02:27:18 <ultra_> luke-jr: O.o what exploit?
113 2012-02-07 02:27:22 <ultra_> dd-wrt printer?
114 2012-02-07 02:27:31 <luke-jr> ultra_: apparently it has SNMP open to the public
115 2012-02-07 02:27:42 <luke-jr> so they send a request with Eligius's IP forged as the source
116 2012-02-07 02:27:45 <BlueMatt> if an isp is forwarding/accepting packets with spoofed ips, the isp's abuse email needs contacted
117 2012-02-07 02:27:46 <luke-jr> and they all flood Eligius with replies
118 2012-02-07 02:28:20 <NxTitle> is bitcoin.cz still attacking? o.O
119 2012-02-07 02:28:39 <luke-jr> NxTitle: never was I bet
120 2012-02-07 02:28:51 <NxTitle> well it was probably spoofed
121 2012-02-07 02:29:08 <luke-jr> Deepbit was spoofed last
122 2012-02-07 02:29:09 <ultra_> BlueMatt: how would the ISP tell if they were spoofed ips incoming? O.o
123 2012-02-07 02:30:55 <BlueMatt> http://en.wikipedia.org/wiki/Reverse_path_forwarding
124 2012-02-07 02:38:37 <gmaxwell> ultra_: you can tell fairly painfully by tracing the traffic backwards, fortunately monitoring tools like arbor automated that for providers
125 2012-02-07 02:40:03 <ultra_> gmaxwell: you're telling me they'll run arbor on every packet? :p
126 2012-02-07 02:40:07 <k9quaint> if you are getting a serious DoS, that means you did something to piss off assholes
127 2012-02-07 02:40:13 <k9quaint> so, keep it up ;)
128 2012-02-07 02:40:40 <gmaxwell> ultra_: nah, sampled netflow.. there are lots of packets, why bother looking at all of them. :)
129 2012-02-07 02:40:46 <BlueMatt> is it much traffic, or just from odd ips?
130 2012-02-07 02:41:04 <gmaxwell> ultra_: 1:1000 - 1:4000 sampling rates are pretty common.
131 2012-02-07 02:41:10 <ultra_> gmaxwell: sure, but that's not something that can be expected of an ISP automatically as BlueMatt suggested
132 2012-02-07 02:41:29 <BlueMatt> not really, but ips really should be
133 2012-02-07 02:41:47 <gmaxwell> ultra_: er I'm not following you.
134 2012-02-07 02:42:02 <gmaxwell> (1) any ISP with end user connectivity will be using URPF.
135 2012-02-07 02:42:13 <gmaxwell> It's frequently a requirement of peering/transit agreements.
136 2012-02-07 02:42:48 <gmaxwell> (2) anyone with customers who are isps (so they won't be strict-urpfed) will be running something like arbor so they can track back floods.
137 2012-02-07 02:43:21 <ultra_> gmaxwell: well, you pretty much need to be close to backbone to spoof afaik
138 2012-02-07 02:44:06 <gmaxwell> (and arbor is just an automated reporting system that runs on top sampled netflow which is just always running)
139 2012-02-07 02:45:26 <ultra_> sounds real effective
140 2012-02-07 02:50:27 <gmaxwell> ultra_: What do you say that?
141 2012-02-07 02:51:34 <TuxBlackEdo> hmm
142 2012-02-07 02:51:38 <TuxBlackEdo> i agree
143 2012-02-07 02:52:12 <ultra_> cause if it worked so well, the DDoSer would be gone and not DDoSing luke-jr's home website and shit like he was talking about
144 2012-02-07 02:56:15 <gmaxwell> ultra_: you're assuming anyone with the power to do anything about it cares.
145 2012-02-07 02:56:29 <gmaxwell> It's probably just a few hundred mbit/sec.. not enough to catch anyone's attention.
146 2012-02-07 03:02:32 <FROTUSCI> ?LLL:L
147 2012-02-07 03:03:27 <ultra_> lol, well if it's enough that it takes out the host without raising any eyebrows that shit is useless....
148 2012-02-07 03:07:04 <ultra_> you'd think they'd have something more effective >.<M
149 2012-02-07 03:11:27 <gmaxwell> ultra_: feh, block the target host. Done.
150 2012-02-07 03:11:35 <gmaxwell> You're not much of a BOFH are you?
151 2012-02-07 03:25:04 <ultra_> gmaxwell: what do you mean block the target host? nullroute them as happened in the eligius case.. and make the DDoS more effective?
152 2012-02-07 03:25:45 <gmaxwell> tada.
153 2012-02-07 03:25:48 <gmaxwell> Problem solved.
154 2012-02-07 03:26:40 <ultra_> lol, I spose :p
155 2012-02-07 04:23:29 <midnightmagic> i wonder if there are any legitimate uses of ip spoofing..
156 2012-02-07 04:23:51 <gmaxwell> midnightmagic: YES
157 2012-02-07 04:23:54 <gmaxwell> midnightmagic: ANYCAST.
158 2012-02-07 04:24:05 <gmaxwell> (esp anycast DNS)
159 2012-02-07 04:25:32 <midnightmagic> that's using normal bgp/etc to announce presence..  i mean classic ip spoof smurfers
160 2012-02-07 04:25:45 <gmaxwell> It's useful on the egress side even without anycast ingress.
161 2012-02-07 04:25:55 <cjd> hiding your location
162 2012-02-07 04:26:02 <midnightmagic> and like you say, properly announced anycast won't be filtered by ingress..
163 2012-02-07 04:26:20 <cjd> if you own the address, no reason why you can't spoof it from somewhere else
164 2012-02-07 04:26:41 <midnightmagic> cjd: the answer will come back to the "real" location though?
165 2012-02-07 04:26:46 <gmaxwell> I have a cute GSLB solution which works by one node getting your dns request.. then all DNS responders reply with the host near them (on the top of a 80ms clock)... and the first to reach you wins. :)
166 2012-02-07 04:27:17 <gmaxwell> but yea, there is no reason to let edge networks in general spoof... and the vast majority are not.
167 2012-02-07 04:27:46 <midnightmagic> surely a network capable of spoofing, and then doing it, would make the backbone admins a little testy?!
168 2012-02-07 04:28:12 <gmaxwell> midnightmagic: meh, why? just don't make my phone ring!
169 2012-02-07 04:28:23 <cjd> but ofc everyone (except cable/dsl) allows it because allowing it is easier than blocking it
170 2012-02-07 04:28:27 <midnightmagic> gmaxwell: I like that, that's a fun idea. Are they time-sync'd to answer simultaneously?
171 2012-02-07 04:28:40 <gmaxwell> midnightmagic: yes.
172 2012-02-07 04:28:48 <midnightmagic> gmaxwell: I used to have my phone ringing, and it sucked. I was angry.
173 2012-02-07 04:29:22 <gmaxwell> cjd: nah, not so. URPF is pretty widely deployed.
174 2012-02-07 04:30:10 <midnightmagic> gmaxwell: How do they time-sync? just everybody agrees to only answer at some future time as requested by the origination (queried) dns?
175 2012-02-07 04:30:30 <midnightmagic> "BTW, answer this within five seconds of my timestamp, or don't answer at all.."
176 2012-02-07 04:30:39 <gmaxwell> midnightmagic: yes, the master timestamps it and they round up enough to all answer at once.
177 2012-02-07 04:31:04 <midnightmagic> gmaxwell: Is the delay by some kind of consensus?
178 2012-02-07 04:31:32 <midnightmagic> gmaxwell: And why would you need such a thing? lol
179 2012-02-07 04:31:34 <gmaxwell> midnightmagic: just configuration. It's pretty easy to set safely enough.
180 2012-02-07 04:31:47 <midnightmagic> Like, you specifically..
181 2012-02-07 04:32:13 <gmaxwell> not me personally, people I've solved problems for.
182 2012-02-07 04:32:21 <midnightmagic> huh!
183 2012-02-07 04:32:38 <midnightmagic> that's a pretty neat mechanism..
184 2012-02-07 04:32:42 <gmaxwell> (When you absoltely positively need to serve traffic from the most close datacenter to the requestors recursive resolver!)
185 2012-02-07 04:33:47 <midnightmagic> i guess even secure dns wouldn't break that sort of thing, because the answer is controlled by the authority at all times anyway
186 2012-02-07 04:34:03 <gmaxwell> Yep.
187 2012-02-07 04:34:19 <midnightmagic> I wonder if any software gets irritated at the multiple responses..  usually that's a sign of a spoof..
188 2012-02-07 04:34:50 <midnightmagic> "Well fuck, these people are doing it on purpose..  back to the drawing board."
189 2012-02-07 04:34:58 <gmaxwell> No. This has been used in production for some very very large sites. (thoug most people do gslb thats a bit less crazy)
190 2012-02-07 04:35:03 <gmaxwell> well right, maybe it caused that. :)
191 2012-02-07 06:38:30 <chmod755> hi #bitcoin-dev
192 2012-02-07 06:38:52 <chmod755> I have a problem with my wallet.dat
193 2012-02-07 06:40:35 <chmod755> I had a few orphaned coins and transfer my BTC to another wallet - now it seems that all of my coins are gone
194 2012-02-07 06:41:25 <chmod755> the network has rejected the transactions
195 2012-02-07 06:41:43 <chmod755> how can I access my BTC now?
196 2012-02-07 06:42:41 <chmod755> any ideas?
197 2012-02-07 06:43:10 <chmod755> (is there a tool to fix wallet.dat files?)
198 2012-02-07 08:15:10 <TuxBlackEdo> i know the bitcoin network is vulnurable to the 51% attack, but is that only of new mining power? for example lets say a 40% actively participating pool decides to stop mining to the main blockchain and build a secret (unannounced) forked blockchain, would that senario require less than 51-50% mining power?
199 2012-02-07 08:15:25 <chmod755> bitcoind listaccounts
200 2012-02-07 08:15:34 <TuxBlackEdo> chmod755, rescan
201 2012-02-07 08:15:37 <chmod755> my wallet is fucked
202 2012-02-07 08:15:38 <TuxBlackEdo> that's the fix
203 2012-02-07 08:15:43 <chmod755> TuxBlackEdo: didn't work
204 2012-02-07 08:15:44 <TuxBlackEdo> -rescan
205 2012-02-07 08:15:56 <TuxBlackEdo> export private keys maybe?
206 2012-02-07 08:16:06 <chmod755> I tried rescanning and reloading the entire blockchain
207 2012-02-07 08:16:07 <chmod755> k
208 2012-02-07 08:16:44 <chmod755> TuxBlackEdo: do you know how to export the keys?
209 2012-02-07 08:17:09 <edcba> TuxBlackEdo: you need >50% statically
210 2012-02-07 08:17:15 <chmod755> -13k lol, I wish I had 13k
211 2012-02-07 08:17:45 <justmoon> TuxBlackEdo: the 51% is of total mining power, basically randomness favors whoever has more mining power and over time, exponentially so - that's why you need more mining power than everyone else put together to mount an attack and that's where the 51% comes from
212 2012-02-07 08:32:03 <TuxBlackEdo> justmoon, imagine this
213 2012-02-07 08:32:09 <TuxBlackEdo> the bitcoin network is 100ghash
214 2012-02-07 08:32:21 <TuxBlackEdo> and inside the bitcoin network, deepbit is 50ghash
215 2012-02-07 08:32:47 <TuxBlackEdo> to mount a 51 percent attack if i am not deeptbit, I need 51ghash
216 2012-02-07 08:32:52 <TuxBlackEdo> but if i am deepbit
217 2012-02-07 08:32:57 <justmoon> no, you need 101 gh
218 2012-02-07 08:33:14 <justmoon> because the network is 100 ghash, so to outrun it, you need at least 101
219 2012-02-07 08:33:58 <TuxBlackEdo> yes
220 2012-02-07 08:33:59 <TuxBlackEdo> but
221 2012-02-07 08:34:13 <TuxBlackEdo> If I am already 50% of the network
222 2012-02-07 08:34:30 <TuxBlackEdo> i could just decide to pull my mining power and start my own secret blockchain
223 2012-02-07 08:34:42 <chmod755> #OccupyDeepbit WE ARE THE 51%
224 2012-02-07 08:34:53 <justmoon> sure, if you have more mining power than the rest of the network that's the 51% attack
225 2012-02-07 08:35:09 <justmoon> but it can't really be a pool, because if deepbit did that, its miners would start abandoning them
226 2012-02-07 08:35:20 <TuxBlackEdo> why?
227 2012-02-07 08:35:26 <TuxBlackEdo> they could do it secretive
228 2012-02-07 08:35:41 <TuxBlackEdo> it's not like deepbit couldn't keep paying the miners out of pocket
229 2012-02-07 08:36:05 <TuxBlackEdo> think of the 5-10% fee they charge
230 2012-02-07 08:36:07 <justmoon> a 51% attack would be very obvious
231 2012-02-07 08:36:23 <justmoon> there are dozens of people in this channel including myself who are running monitoring that would detect it
232 2012-02-07 08:36:38 <TuxBlackEdo> and what would be the plan?
233 2012-02-07 08:36:54 <justmoon> well, put a message in #bitcoin-mining
234 2012-02-07 08:37:12 <Eliel> and a message on the forums
235 2012-02-07 08:37:23 <Eliel> it'd spread like wildfire.
236 2012-02-07 08:38:11 <justmoon> miners would have to be afraid that even if deepbit "wins" the hashing war that somehow the network will go back to the legitimate chain and they would lose their entire income for the duration of the attack
237 2012-02-07 08:40:08 <TuxBlackEdo> i am sure they could pay out of pocket for days if not months worth of blocks
238 2012-02-07 08:40:26 <TuxBlackEdo> and if they succeed, they wouldn't even have to pay anything
239 2012-02-07 08:42:39 <justmoon> pay out of pocket with what? bitcoins?
240 2012-02-07 08:43:16 <justmoon> those would lose a ton of value if the attack succeeded
241 2012-02-07 08:45:07 <justmoon> I think a lot of miners hold bitcoins or have bitcoin investments
242 2012-02-07 08:45:17 <justmoon> we *know* that they have money invested in mining hardware
243 2012-02-07 08:45:34 <justmoon> if a 51% attack succeeds, the whole 1-cpu-1-vote idea is called into question
244 2012-02-07 08:45:57 <justmoon> and there may not be much point in mining anymore - making that expensive mining hardware worthless
245 2012-02-07 08:46:08 <justmoon> and of course, most people try to be ethical
246 2012-02-07 08:46:19 <justmoon> miners started switching away from deepbit when they got over 51%
247 2012-02-07 08:46:33 <justmoon> they didn't even abuse that fact, people switched away just as a precaution
248 2012-02-07 09:02:45 <TuxBlackEdo> justmoon, i think deepbit got pressured into using a higher fee to drive away miners
249 2012-02-07 09:03:00 <justmoon> they didn't raise the fee at that time
250 2012-02-07 09:03:14 <TuxBlackEdo> but it could turn out that deepbit owns BTCGuild as well
251 2012-02-07 09:03:56 <justmoon> point being?
252 2012-02-07 09:04:20 <TuxBlackEdo> all it would take is slush or btc guild to conspire with deepbit
253 2012-02-07 09:04:43 <justmoon> no, again, the pool operators can do what they want, it is the miners that control the actual mining power
254 2012-02-07 09:05:09 <TuxBlackEdo> a 51% attack would not have to last for days and days
255 2012-02-07 09:05:32 <TuxBlackEdo> just long enough to deposit some already gotten btc and then sell it on the market and take a wire
256 2012-02-07 09:05:40 <TuxBlackEdo> then announce the secret blockchain
257 2012-02-07 09:05:58 <TuxBlackEdo> with a earlier payment to yourself
258 2012-02-07 09:06:10 <TuxBlackEdo> instead of to mtgox
259 2012-02-07 09:06:12 <justmoon> so you're suggesting that slush (who I've met in person) would steal money in a public way
260 2012-02-07 09:06:40 <TuxBlackEdo> bitcoin is not at all about trust, is it?
261 2012-02-07 09:06:45 <TuxBlackEdo> there is no trust
262 2012-02-07 09:07:15 <justmoon> it's not that I trust him, it's that I know he's intelligent
263 2012-02-07 09:07:34 <justmoon> and I know who he is - if he stole money publicly, he'd go to jail
264 2012-02-07 09:08:00 <justmoon> and of course if you steal, the stolen money is confiscated and you can be sued for damages
265 2012-02-07 09:08:19 <chmod755> ;;gettrust TuxBlackEdo
266 2012-02-07 09:08:19 <gribble> Trust relationship from user chmod755 to user TuxBlackEdo: Level 1: 0, Level 2: 0 via 0 connections. Graph: http://serajewelks.bitcoin-otc.com/trustgraph.php?source=chmod755&dest=TuxBlackEdo
267 2012-02-07 09:08:34 <chmod755> (11:06:45 AM) TuxBlackEdo: there is no trust << yep
268 2012-02-07 09:09:50 <cjd> one reason a lot of the attacks which are theorized would never work is simply because people (in this channel) are watching
269 2012-02-07 09:13:00 <TuxBlackEdo> only would take 3600btc or 72blocks for the duration of the attack
270 2012-02-07 09:13:12 <TuxBlackEdo> if the attack isnt successful
271 2012-02-07 09:13:21 <TuxBlackEdo> if it took 1 day
272 2012-02-07 09:14:46 <TuxBlackEdo> plus you could short bitcoin while you are doing this because of all the fallout
273 2012-02-07 09:14:49 <justmoon> the cost of the actual blocks doesn't matter so much
274 2012-02-07 09:15:00 <TuxBlackEdo> just to pay off the miners
275 2012-02-07 09:15:10 <justmoon> what's discouraging is that your pool will be dead afterwards and you'll go to jail
276 2012-02-07 09:15:30 <TuxBlackEdo> just like the mybitcoin founder went to jail?
277 2012-02-07 09:15:45 <justmoon> he would have if we knew who he was
278 2012-02-07 09:15:49 <justmoon> we know who the pool ops are
279 2012-02-07 13:58:50 <luke-jr> I love how Gavin just ignores problems with his own pullreqs, yet refuses to merge others' which have no problems&
280 2012-02-07 14:10:45 <phantomcircuit> luke-jr, did you catch the the not doing the transaction checking introduced a vulnerability? called it
281 2012-02-07 14:10:57 <luke-jr> phantomcircuit: ?
282 2012-02-07 14:13:35 <phantomcircuit> luke-jr, http://seclists.org/fulldisclosure/2012/Feb/0
283 2012-02-07 14:14:14 <phantomcircuit> that combined with being the first connection a new node makes could allow you to spend counterfeit coins
284 2012-02-07 14:14:16 <luke-jr> phantomcircuit: no, that's not a real problem
285 2012-02-07 14:14:37 <phantomcircuit> with the current initial download logic it is
286 2012-02-07 14:14:45 <phantomcircuit> it's extremely unlikely to be exploited
287 2012-02-07 14:14:48 <phantomcircuit> but it's possible
288 2012-02-07 14:15:00 <luke-jr> no
289 2012-02-07 14:15:18 <luke-jr> the download logic mentioned there was fixed in a release before it was posted
290 2012-02-07 14:15:33 <phantomcircuit> yes and people still join #bitcoin using old clients
291 2012-02-07 14:16:05 <phantomcircuit> that commit was reckless and people did call him on it despite the fd saying otherwise
292 2012-02-07 14:18:06 <luke-jr> phantomcircuit: so, what's the solution? now there's 4 examples of things merged that shouldn't have been :P
293 2012-02-07 14:18:18 <luke-jr> (and countless examples of things that should be merged, but haven't been)
294 2012-02-07 14:18:18 <phantomcircuit> i think you can guess :)
295 2012-02-07 14:19:10 <luke-jr> phantomcircuit: my guesses all seem non-viable. viability is a requirement to a solution
296 2012-02-07 14:23:36 <Eliel> phantomcircuit: it's only a problem with 0.5 and 0.5.1
297 2012-02-07 14:24:01 <phantomcircuit> im sure there are tons of people still running 0.5 and 0.5.1
298 2012-02-07 14:24:18 <phantomcircuit> the very fact that it was included after people voiced their opposition is a problem
299 2012-02-07 14:24:19 <Eliel> ok, let me look up the stats page for client versions
300 2012-02-07 14:24:39 <luke-jr> ^ examples of why lying about the client/version (currently merged it git master) is a bad idea
301 2012-02-07 14:25:06 <sipa> luke-jr: not being in the form you like it does not make it lying
302 2012-02-07 14:25:47 <luke-jr> sipa: I didn't complain about the form, I complained about the lying
303 2012-02-07 14:26:17 <luke-jr> I don't care about the form.
304 2012-02-07 14:27:29 <Eliel> what is this thing about lying? link?
305 2012-02-07 14:27:50 <luke-jr> sipa: what are your thoughts on the latest? (Gavin breaking non-spammy 'green addresses')
306 2012-02-07 14:28:42 <Eliel> phantomcircuit: there http://bitcoinstatus.rowit.co.uk/
307 2012-02-07 14:28:46 <sipa> luke-jr: which are you talking about now?
308 2012-02-07 14:29:17 <luke-jr> Eliel: https://github.com/bitcoin/bitcoin/pull/715
309 2012-02-07 14:29:33 <phantomcircuit> Eliel, that shows at least 199 vulnerable hosts
310 2012-02-07 14:29:35 <luke-jr> sipa: https://github.com/bitcoin/bitcoin/pull/769
311 2012-02-07 14:29:54 <phantomcircuit> no more like 700
312 2012-02-07 14:30:33 <luke-jr> sipa: also, I posted on the ML when he announced his intent-to-merge
313 2012-02-07 14:30:46 <luke-jr> in more details as to the practical solution it breaks
314 2012-02-07 14:30:48 <sipa> luke-jr: green addresses use spurious pushes?
315 2012-02-07 14:31:10 <Eliel> phantomcircuit: yes, that's a reasonably small fraction. Also, it's quite difficult to exploit.
316 2012-02-07 14:31:14 <luke-jr> sipa: green addresses right now require you send from a single address; so all merchants supporting it need to transfer funds to that address
317 2012-02-07 14:31:39 <phantomcircuit> Eliel, that's comical given gavin said ages ago that he would be very cautious
318 2012-02-07 14:31:51 <luke-jr> sipa: before pull 769, it was *possible* to replace that with just an extra signature, and then have miners strip that off when making a block
319 2012-02-07 14:32:21 <sipa> luke-jr: the reason for making them non-standard is because a miner could just strip those extra pushes, which would cause much more problems (people not recognizing their own transactions, e.g.)
320 2012-02-07 14:32:47 <luke-jr> sipa: people who send such transactions should expect that
321 2012-02-07 14:33:01 <luke-jr> it has no effect on people sending normal transactions
322 2012-02-07 14:33:36 <sipa> so, your point is that there was another solution, which nobody used, and now becomes impossible, because we fix another issue?
323 2012-02-07 14:34:00 <luke-jr> my point is that it doesn't fix any real issue, and breaks the only known solution to a real issue
324 2012-02-07 14:35:46 <sipa> but your solution - while better than the current green addresses - still boil down to putting extra data in the block chain
325 2012-02-07 14:35:58 <luke-jr> sipa: no, because miners can/should strip the extra data
326 2012-02-07 14:36:08 <luke-jr> so it's only present for relaying
327 2012-02-07 14:36:26 <sipa> that's even more broken to me
328 2012-02-07 14:36:39 <sipa> it isn't data that should be part of a transaction
329 2012-02-07 14:37:11 <luke-jr> tacking it onto the transaction is the only way to satisfy the requirements of the merchants using it
330 2012-02-07 14:37:35 <sipa> i know, but it isn't the kind of solution one should encourage, imho
331 2012-02-07 14:37:56 <Eliel> I think the best way to do green addresses is to have a side channel that provides the extra signatures.
332 2012-02-07 14:38:01 <sipa> indeed
333 2012-02-07 14:38:06 <luke-jr> they'll do it one way or another; the question is, do we want the bloat to be temporary/stripped, or permanent in the blockchain
334 2012-02-07 14:38:16 <Eliel> a simple DHT would work well. The data isn't needed for more than an hour in any case.
335 2012-02-07 14:38:51 <luke-jr> Eliel: will that work without Silk Road potentially revealing their IP? that seems to be the "use case" they want
336 2012-02-07 14:39:58 <Eliel> alternatively, add the extra signature to go with the transaction somewhere other than scriptSig. It has no business in the script.
337 2012-02-07 14:40:30 <luke-jr> Eliel: nor does it harm there
338 2012-02-07 14:40:37 <luke-jr> provided miners strip it off
339 2012-02-07 14:44:06 <diki> phantomcircuit:are you the owner of intersango?
340 2012-02-07 14:44:14 <phantomcircuit> diki, cofounder
341 2012-02-07 14:44:21 <diki> isnt that the same?
342 2012-02-07 14:44:49 <diki> anyway, SEPA is still working, right?
343 2012-02-07 14:44:59 <phantomcircuit> yes
344 2012-02-07 14:45:16 <phantomcircuit> diki, we are not an eMoney provider
345 2012-02-07 14:45:24 <phantomcircuit> unlike somepeople
346 2012-02-07 14:45:50 <diki> so 0.65% fee is 0.0065 from a single trade, right?
347 2012-02-07 14:46:42 <phantomcircuit> yes
348 2012-02-07 14:47:10 <phantomcircuit> i was going to build the loyalty program today but instead im having to deal with lloyds being broken
349 2012-02-07 14:50:17 <diki> Is there any reason why the price is lower there than in other exchanges?
350 2012-02-07 14:52:04 <gmaxwell> luke-jr: thats an interesting thought... but what do you suggest instead.
351 2012-02-07 14:52:09 <FROTUSCI> the bitcoins they sell are refurbished
352 2012-02-07 14:52:13 <gmaxwell> oops I was in scrollbck.
353 2012-02-07 14:52:29 <diki> FROTUSCI:can you rephrase?
354 2012-02-07 14:52:35 <FROTUSCI> j/k hehe
355 2012-02-07 14:52:40 <gmaxwell> luke-jr: if miners strip it then the txn ID will change.
356 2012-02-07 14:53:11 <luke-jr> gmaxwell: so?
357 2012-02-07 14:53:22 <luke-jr> gmaxwell: someone sending transactions with extra data should expect that
358 2012-02-07 14:53:43 <luke-jr> (heck, they can even precalculate the final txnid)
359 2012-02-07 14:54:03 <gmaxwell> also if people do start sending txn like that some miners will mine it.
360 2012-02-07 14:54:22 <luke-jr> that doesn't hurt more than how it is now at least
361 2012-02-07 14:54:25 <gmaxwell> luke-jr: how about this: step 1. block it everywhere. Step 2. make miners strip it. Step 3. permit it again.
362 2012-02-07 14:54:42 <luke-jr> gmaxwell: why step 1?
363 2012-02-07 14:54:55 <gmaxwell> this way laggard upgraders won't accidentally be mining a flood of garbage when the flood starts
364 2012-02-07 14:55:14 <gmaxwell> also gives us time to find better alternatives.
365 2012-02-07 14:55:26 <luke-jr> gmaxwell: the potential 'garbage' is still less than what they have live right now
366 2012-02-07 14:55:38 <luke-jr> right now, we have an extra transaction
367 2012-02-07 14:55:51 <sipa> that can immediately be pruned
368 2012-02-07 14:56:07 <luke-jr> sipa: never for full clients
369 2012-02-07 14:56:15 <gmaxwell> I'm surprised to see you advocating this approach, considering that people have rejected more sane (e.g. signmessage) based solutions simply because they want to pretend they don't know their payments are being made to underground drug markets.
370 2012-02-07 14:57:35 <luke-jr> gmaxwell: we can't force them to use an approach that doesn't meet their requirements
371 2012-02-07 14:57:46 <luke-jr> so the choice from my perspective is spammy or not spammy
372 2012-02-07 14:58:39 <gmaxwell> right now appeneding random data to transactions isn't just spammy (though it's spammy too) but it encourage abuse, because you can add crap to other people's transactions just to implement "omg saved in the blockchain forever" services
373 2012-02-07 14:59:07 <TuxBlackEdo> heh
374 2012-02-07 14:59:19 <sipa> gmaxwell: don't give TuxBlackEdo ideas
375 2012-02-07 14:59:30 <gmaxwell> E.g. evading our current anti-spam behavior by turning every user into a spammer like TuxBlackEdo.
376 2012-02-07 14:59:36 <TuxBlackEdo> ...
377 2012-02-07 14:59:48 <TuxBlackEdo> i did it like 2-3 times
378 2012-02-07 14:59:52 <gmaxwell> hehe
379 2012-02-07 14:59:57 <TuxBlackEdo> x_x
380 2012-02-07 15:02:11 <luke-jr> gmaxwell: better than adding an extra transaction
381 2012-02-07 15:03:15 <gmaxwell> luke-jr: ::cries::
382 2012-02-07 15:03:58 <luke-jr> gmaxwell: right now, if people started messing with others' txns, I bet we could get at least some of the bigger pools merge a patch to strip crap
383 2012-02-07 15:04:09 <gmaxwell> luke-jr: you'd have to get passed the anti-dos metric if you were using extra transactions.
384 2012-02-07 15:04:18 <FROTUSCI> yamatough
385 2012-02-07 15:05:03 <luke-jr> gmaxwell: they *are* using extra transactions
386 2012-02-07 15:05:12 <gmaxwell> luke-jr: a large txn with a few thousand btc that had not been recently moved could take about 40kbytes of extra data.
387 2012-02-07 15:05:28 <gmaxwell> for free to the spammer.
388 2012-02-07 15:08:31 <helo> block_hash/target_hash histogram has slope ~0 because we're way out in the tail of the normal hash distribution?
389 2012-02-07 15:12:10 <edcba> eliel ?
390 2012-02-07 15:13:15 <FROTUSCI> transacticorn
391 2012-02-07 15:15:50 <luke-jr> could we add some kind of signmessage-based message distribution to the protocol, separate from txns?
392 2012-02-07 15:16:30 <sipa> maybe, but why does the whole world need to know it?
393 2012-02-07 15:17:14 <Eliel> not everyone needs it. So it ought to be disabled by default.
394 2012-02-07 15:17:24 <gmaxwell> sipa: because the whole world knows about transactions.
395 2012-02-07 15:17:35 <luke-jr> sipa: because the people using green addresses insist on not knowing each others servers for stupid reasons
396 2012-02-07 15:17:51 <sipa> what about giving a signed "I, ScamWallet Pro, guarantee that transaction <txid> comes from me and I will not try to double spend it" to the customer when doing the transaction
397 2012-02-07 15:18:06 <sipa> the customer gives that signature to the payee, and voila
398 2012-02-07 15:18:16 <sipa> ok, probably too inconvenient to be actually used... :(
399 2012-02-07 15:18:32 <gmaxwell> Yea, it's not automagic.
400 2012-02-07 15:19:01 <gmaxwell> luke-jr: how can we make such a system spam resistant and more private than greenaddresses?
401 2012-02-07 15:19:44 <gmaxwell> (one problem with greenaddresses is that they totally blow up privacy on on side the alternative can be more attractive by not having that policy)
402 2012-02-07 15:19:58 <gmaxwell> s/policy/problem/
403 2012-02-07 15:20:18 <sipa> gmaxwell: public determinstic key chain for a provider?
404 2012-02-07 15:20:55 <gmaxwell> still blows up privacy. Otherwise ...
405 2012-02-07 15:21:06 <sipa> ok, not much more anonimous, as that too will result in the whole world knowing what comes from which provider
406 2012-02-07 15:21:12 <edcba> Eliel: what dht thing ?
407 2012-02-07 15:21:33 <FROTUSCI> dHT
408 2012-02-07 15:23:17 <gmaxwell> sipa: I guess it's not possible to have both at once. alas. but at least keeping it seperate would make it a little more ephemeral.
409 2012-02-07 15:24:10 <sipa> it is information that goes from the provider to the payee
410 2012-02-07 15:24:19 <gmaxwell> sipa: your ScamWalletPro solution solves it all.
411 2012-02-07 15:24:42 <sipa> either that happens via a point-to-point communication between them, or via the customer, or via the whole world
412 2012-02-07 15:24:53 <gmaxwell> You could do both... have the ScamWalletPro method, plus a distribution method.. and you can choose to manually move the sigs (private) or announce them (easy)
413 2012-02-07 15:25:15 <gmaxwell> but the sigs would be the same way either case.
414 2012-02-07 15:26:08 <gmaxwell> e.g. "Send signature to ???"  with options "everyone" "manual" "http://destination.site/payments.cgi"
415 2012-02-07 15:27:11 <gmaxwell> the last would use a payment protocol to send it, the first would use some flooding network. The middle would just give you an extended "TXNID"
416 2012-02-07 15:27:25 <Eliel> edcba: distribute hash table
417 2012-02-07 15:30:06 <edcba> i know what it is
418 2012-02-07 15:30:12 <edcba> but for what ?
419 2012-02-07 15:31:21 <luke-jr> hi [Tycho]
420 2012-02-07 15:31:34 <[Tycho]> Hello.
421 2012-02-07 15:31:51 <FROTUSCI> hey
422 2012-02-07 15:31:52 <luke-jr> [Tycho]: someone is trying to frame you FYI
423 2012-02-07 15:32:23 <luke-jr> by flooding Eligius from your IP
424 2012-02-07 15:32:51 <luke-jr> but they did it from slush's IP too, and UDP is spoofable, so I don't think it's you guys
425 2012-02-07 15:40:34 <Eliel> edcba: side channel for signatures that basically do the same thing as green transactions right now.
426 2012-02-07 15:50:04 <[Tycho]> luke-jr: is Eligius unavailable because of this ?
427 2012-02-07 15:51:45 <luke-jr> [Tycho]: yes
428 2012-02-07 15:53:17 <[Tycho]> So there is some evil mastermind :)
429 2012-02-07 15:53:49 <gavinandresen> Can I play the part of the Brave Hero who rides in and saves the day?
430 2012-02-07 15:56:24 <luke-jr> gavinandresen: you have a solution to DDoS? ;)
431 2012-02-07 15:56:26 <gribble> New news from bitcoinrss: luke-jr opened pull request 805 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/805>
432 2012-02-07 15:57:10 <gavinandresen> luke-jr: no, I don't.  Darn, I guess i'll have to play the part of an innocent bystander.
433 2012-02-07 15:57:18 <gavinandresen> (and hope I don't get shot)
434 2012-02-07 16:07:28 <gavinandresen> luke-jr: Can you modify the "Fetch and build inputs" section in doc/release-process.txt as part of your deps-win32.yml change?
435 2012-02-07 16:08:01 <luke-jr> gavinandresen: aha, sure. wish I'd noticed that section before I started :P
436 2012-02-07 16:12:16 <luke-jr> gavinandresen: pushed
437 2012-02-07 16:12:30 <gavinandresen> thanks
438 2012-02-07 16:13:31 <gavinandresen> FYI all:  I'll be gitian-building 0.6 release candidate 1 this afternoon, if all goes well I hope to push the tag and upload binaries when it's done.
439 2012-02-07 16:14:29 <luke-jr> gavinandresen: is there a reason behind the hold-up on rpc_keepalive, or just too much for 0.6?
440 2012-02-07 16:15:56 <gavinandresen> luke-jr: "This is quick and dirty code, it may not work for you"  does not inspire confidence....
441 2012-02-07 16:16:04 <gavinandresen> (comment at the top of bitcoinrpc.cpp)
442 2012-02-07 16:16:20 <gavinandresen> And donation addresses in code are inappropriate
443 2012-02-07 16:16:34 <luke-jr> gavinandresen: well, that was before it was merged with my code, and the comment only allows removing it when "accepted into the main distribution"
444 2012-02-07 16:16:37 <gavinandresen> And has it actually been tested on any non-mining sites that do lots of RPC?
445 2012-02-07 16:17:30 <luke-jr> not sure; I know at least gmaxwell and I tried to break it with flooding other methods
446 2012-02-07 16:18:15 <gavinandresen> The high-level feature it gives is "able to handle more RPC requests per second" ?
447 2012-02-07 16:18:36 <luke-jr> gavinandresen: as well as multiple JSON-RPC connections (with HTTP/1.1 keepalive)
448 2012-02-07 16:19:09 <luke-jr> this is, AIUI, the one "must have" to fix solo mining
449 2012-02-07 16:19:48 <luke-jr> but I might be underestimating the other optimizations
450 2012-02-07 16:20:06 <gavinandresen> solo mining with more than how many GPUs?
451 2012-02-07 16:20:44 <luke-jr> I don't know the current limitations, just that it's lower than the requirements needed to really solo mine
452 2012-02-07 16:20:56 <luke-jr> (ie, find blocks once a week on average)
453 2012-02-07 16:21:06 <gavinandresen> -- trying to judge risk/benefit here, if the risk is "introduces a new, rare deadlock that makes merchants' sites break" but the benefit is "makes it easier for 0.01% of miners to solo mine" ...
454 2012-02-07 16:21:23 <gavinandresen> then it ain't worth it.
455 2012-02-07 16:22:08 <luke-jr> perhaps it'd be a good idea to find a merchant using bitcoind to test
456 2012-02-07 16:22:53 <luke-jr> it *shouldn't* have much risk there, since it locks all methods except getwork
457 2012-02-07 16:22:57 <gavinandresen> So, my short answer would be:  too risky for
458 2012-02-07 16:23:00 <luke-jr> but would be nice to know
459 2012-02-07 16:23:00 <Mqrius> Does ThomasV have an IRC presence?
460 2012-02-07 16:23:01 <luke-jr> for sure
461 2012-02-07 16:23:05 <ThomasV> yes
462 2012-02-07 16:23:18 <Mqrius> ThomasV: You missed some files in the latest update, for example electrum.py
463 2012-02-07 16:23:24 <ThomasV> omg
464 2012-02-07 16:23:36 <ThomasV> will fix asap
465 2012-02-07 16:23:41 <Mqrius> kay
466 2012-02-07 16:24:33 <ThomasV> Mqrius: no, there's no electrum.py anymore
467 2012-02-07 16:24:45 <Mqrius> Oh, it's gui.py now?
468 2012-02-07 16:24:55 <ThomasV> the executable is called electrum
469 2012-02-07 16:25:59 <Mqrius> ThomasV: I don't follow. There's no "electrum" in there
470 2012-02-07 16:26:12 <ThomasV> oh indeed
471 2012-02-07 16:26:17 <ThomasV> moment
472 2012-02-07 16:26:45 <luke-jr> gavinandresen: since you don't seem to be merging Coinbaser as a whole for 0.6, and have deferred it so long that probably nobody who needs it will be using bitcoind by 0.7, would you merge setworkaux by itself for 0.6?
473 2012-02-07 16:27:00 <ThomasV> Mqrius: try now
474 2012-02-07 16:27:38 <Mqrius> ThomasV: That's better :)
475 2012-02-07 16:28:19 <gavinandresen> luke-jr: not high enough on my development priorities list to review/test/pull it.
476 2012-02-07 16:30:01 <Mqrius> ThomasV: Did you fix this floatingpoint bug user "marked" mentioned?
477 2012-02-07 16:30:03 <luke-jr> gavinandresen: sipa's review and Eligius/EclipseMC/etc test change that at all?
478 2012-02-07 16:30:10 <ThomasV> Mqrius: yes
479 2012-02-07 16:30:18 <Mqrius> okay
480 2012-02-07 16:30:30 <ThomasV> Mqrius: fyi, there's a channel for #electrum
481 2012-02-07 16:30:35 <Mqrius> Oh right
482 2012-02-07 16:31:34 <gavinandresen> luke-jr: nope.  By the way, what's up with BIP 19?
483 2012-02-07 16:32:35 <luke-jr> gavinandresen: what about it?
484 2012-02-07 16:32:48 <gavinandresen> you going to do a sample implementation to shake out the bugs?
485 2012-02-07 16:33:06 <gavinandresen> ... or start a discussion about it on the -dev mailing list?
486 2012-02-07 16:33:07 <NxTitle> http://pastebin.com/raw.php?i=BUB3dygQ
487 2012-02-07 16:33:23 <luke-jr> I could, but it didn't seem urgent since there's no sigop shortage.
488 2012-02-07 16:33:53 <luke-jr> IMO makes more sense to wait and see whether BIP 16 or 17 gets adopted on the protocol first, since BIP 19 only needs to be implemented later
489 2012-02-07 16:34:41 <luke-jr> unless I'm missing something?
490 2012-02-07 16:35:04 <gavinandresen> sipa and coblee had other ideas on generic transaction forms, kind of wondered why you bothered to write a BIP without talking about it first.
491 2012-02-07 16:35:47 <luke-jr> we did discuss it here first
492 2012-02-07 16:35:54 <gavinandresen> ... and a little bothered that nobody seems to be following the BIP 0001 process
493 2012-02-07 16:37:21 <luke-jr> sorry, I'll try to remember to involve the ML first next time
494 2012-02-07 16:38:34 <BlueMatt> gavinandresen: would you mind marking https://github.com/bitcoin/bitcoin/issues/802 high priority since it is a very, very low-risk security vuln but is a block for 0.6 (IMO), that way it wont get forgotten
495 2012-02-07 16:38:47 <luke-jr> BlueMatt: it's fixed
496 2012-02-07 16:38:54 <BlueMatt> oh?
497 2012-02-07 16:39:01 <luke-jr> BlueMatt: see pull 805
498 2012-02-07 16:39:04 <BlueMatt> oh, nice
499 2012-02-07 16:40:04 <BlueMatt> was it tested for determinism?
500 2012-02-07 16:40:28 <gavinandresen> I'm building dependencies right now....
501 2012-02-07 16:40:43 <luke-jr> BlueMatt: the results from my test produced the same output as before I changed it
502 2012-02-07 16:40:54 <BlueMatt> what does that mean?
503 2012-02-07 16:41:04 <luke-jr> pre-805 binaries match post-805 binaries
504 2012-02-07 16:41:19 <BlueMatt> no, was the tar from the deps build deterministic?
505 2012-02-07 16:41:28 <BlueMatt> answer: no it doesnt even use faketime
506 2012-02-07 16:41:38 <BlueMatt> follow up: why was it merged/pull-requested?
507 2012-02-07 16:41:57 <luke-jr> are the middle states supposed to be deterministic?
508 2012-02-07 16:42:01 <BlueMatt> yes
509 2012-02-07 16:42:04 <BlueMatt> have to be
510 2012-02-07 16:42:14 <BlueMatt> (or gitian will complain about different inputs)
511 2012-02-07 16:42:18 <luke-jr> I would think only the finished binaries that get distributed matter
512 2012-02-07 16:42:31 <BlueMatt> gitian will still complain
513 2012-02-07 16:42:46 <luke-jr> how will gitian know? O.o
514 2012-02-07 16:43:00 <BlueMatt> the input hashes are in the signed file
515 2012-02-07 16:43:08 <luke-jr> oh
516 2012-02-07 16:43:53 <luke-jr> so touch every file with a reference date, and use gzip -m I think
517 2012-02-07 16:44:10 <luke-jr> hmm
518 2012-02-07 16:44:26 <BlueMatt> usually Im just lazy and do a zip with faketime
519 2012-02-07 16:44:52 <gavinandresen> BlueMatt: why was it merged: because when I went to gitian compile the Windows release a little while ago I was missing dependencies...  and we can work out the gitian issues between rc1 and final release
520 2012-02-07 16:44:54 <luke-jr> ideal would be to strip out all the unnecessary files (.o etc) too
521 2012-02-07 16:45:34 <luke-jr> BlueMatt: is there a way to tell gitian to not kill the VM after it finishes?
522 2012-02-07 16:45:55 <luke-jr> I'd like to check access times to see which files/dirs bitcoin actually needs
523 2012-02-07 16:45:59 <BlueMatt> no, but if the build fails, it wont kill it and you can export PATH=$PATH:libexec; on-target
524 2012-02-07 16:46:05 <BlueMatt> to be ssh'd into the vm
525 2012-02-07 16:46:31 <BlueMatt> ie just add exit 1 in the script somewhere
526 2012-02-07 16:46:40 <luke-jr> good idea
527 2012-02-07 16:46:57 <HostFat> hi
528 2012-02-07 16:47:32 <HostFat> question: will v0.6 support the same URI features that Electrum has added here: https://bitcointalk.org/index.php?topic=50936.msg735942#msg735942
529 2012-02-07 16:47:42 <HostFat> is it already supporting them?
530 2012-02-07 16:48:37 <FROTUSCI> cool
531 2012-02-07 16:48:41 <BlueMatt> we were discussing those features a couple days ago, I argued both of them are not worth implementing
532 2012-02-07 16:48:49 <HostFat> ah
533 2012-02-07 16:48:56 <BlueMatt> the signing thing is cool, and could be, but aliases not so much
534 2012-02-07 16:49:57 <BlueMatt> if you go find the logs, the discussing is somewhat interesting
535 2012-02-07 16:50:24 <HostFat> I'll give a look, thank you :)
536 2012-02-07 16:50:52 <HostFat> anyway, do you think that you will make a decision before the next release?
537 2012-02-07 16:51:25 <BlueMatt> I certainly wont be implementing either of those before 0.6
538 2012-02-07 16:51:30 <BlueMatt> and I doubt anyone else will
539 2012-02-07 16:52:36 <HostFat> ok
540 2012-02-07 16:54:04 <gavinandresen> BlueMatt:   66f8a64421c1499908687dee4573c20582ed06afaeda2a8e5e0643abe6d5bcbd  bitcoin-deps-0.0.1.tbz2
541 2012-02-07 16:56:07 <luke-jr> e945221b55feece70f3f94b0d80f051f301072fa4bc22b5b51e91c3fbd17b7f4  gitian-inputs/bitcoin-deps-0.0.1.tbz2
542 2012-02-07 16:56:41 <BlueMatt> gavinandresen: sorry, dont have time to build atm, Ill build later today
543 2012-02-07 16:57:19 <BlueMatt> it is timestamped as it currently sits afaict
544 2012-02-07 16:58:28 <BlueMatt> gmaxwell did ack the non-default irc change, correct?
545 2012-02-07 16:59:11 <gmaxwell> BlueMatt: Yes. It appears to be okay.
546 2012-02-07 16:59:19 <gmaxwell> I did two more tests and the world did not end.
547 2012-02-07 16:59:29 <BlueMatt> ok, just wanted to make sure someone did the research
548 2012-02-07 16:59:41 <gmaxwell> (well, I mean, I ack the behavior change I didn't actually check gavin's patch)
549 2012-02-07 16:59:57 <gmaxwell> (hadn't gotten that far yet)
550 2012-02-07 17:00:08 <BlueMatt> meh, good enough
551 2012-02-07 17:00:18 <luke-jr> BlueMatt: gmaxwell tested without IRC on a VPS I setup that had never had a bitcoin node on it
552 2012-02-07 17:00:35 <BlueMatt> yea, I just never heard the full conclusion of the tests
553 2012-02-07 17:00:59 <gmaxwell> No, not really good enough. Every patch needs actual review, not just of what its supposted to do but of what it does. :)
554 2012-02-07 17:01:07 <gmaxwell> but it's good enough for now.
555 2012-02-07 17:02:00 <BlueMatt> well its good enough for me, since I really dont have time to read every commit, I have to settle for reading commitmsgs
556 2012-02-07 17:02:15 <BlueMatt> (not that me reading commits would help much, but meh)
557 2012-02-07 17:02:23 <gmaxwell> If you only have enough time to read one, read the patch. For short patches the patch doesn't take much longer to read. :)
558 2012-02-07 17:21:50 <luke-jr> gavinandresen: btw, is 0.6 going to wait for a bugfix for that chain forking exploit? (i'm not sure anyone has a good solution yet)
559 2012-02-07 17:22:37 <gavinandresen> the duplicate coinbase might-possibly-be-a-chainforking-exploit-if-you-can-mine-a-bunch-of-blocks thing?
560 2012-02-07 17:23:04 <gmaxwell> gavinandresen: it looks like mining a bunch is likely not required. If it works.
561 2012-02-07 17:23:34 <gavinandresen> gmaxwell: what does it cost the attacker?
562 2012-02-07 17:23:38 <gmaxwell> (you use your duplicate coinbases to make duplicate transactions. And use the transactions in the reorg to make a mess.
563 2012-02-07 17:24:21 <gmaxwell> gavinandresen: mine two, perhaps three blocks with delays in between. And either create a small reorg or take advantage of one.
564 2012-02-07 17:24:37 <gmaxwell> Without actually _trying_ I'm not sure if there aren't constraints that make it harder though.
565 2012-02-07 17:24:59 <gavinandresen> gmaxwell: invalidate how many blocks in the reorg?
566 2012-02-07 17:25:02 <gmaxwell> And consequence is that offline nodes that didn't see the reorg  may be willing to follow a chain that online nodes that did will not.
567 2012-02-07 17:25:06 <gavinandresen> (just one, I assume, would be enough)
568 2012-02-07 17:25:22 <gmaxwell> Yes. Just one.
569 2012-02-07 17:25:54 <luke-jr> gavinandresen: roconnor found a way to make the dupe coinbase attack cheap
570 2012-02-07 17:25:55 <gmaxwell> But since the more restrictive of the nodes are almost certantly going to have more hashpower I'm not currently losing sleep even if its our current worst case understanding.
571 2012-02-07 17:25:56 <gavinandresen> Seems like a weak attack:  you have to reorg yourself, and the most damage you do is to create a split between miners where are/aren't seeing the whole history.  BUT... should be fixed
572 2012-02-07 17:26:12 <BlueMatt> in theory, its not a big deal unless you can reverse it - online nodes following a chain that offline nodes will not
573 2012-02-07 17:26:25 <gmaxwell> Right.
574 2012-02-07 17:26:38 <josephcp> i can foresee non-satoshi implementations of bitcoin screwing up with duplicate transaction ids too, it's just so unintuitive
575 2012-02-07 17:26:49 <josephcp> hashes i mean
576 2012-02-07 17:26:55 <gmaxwell> I can't currently see how to do that.
577 2012-02-07 17:27:01 <luke-jr> gavinandresen: if you connect to every node, you can wait to release your attack block until it will cause a blockchain split
578 2012-02-07 17:27:26 <gmaxwell> luke-jr: the split, however, will resolve.
579 2012-02-07 17:27:37 <luke-jr> gmaxwell: yes, but it will resolve with clients in inconsistent states
580 2012-02-07 17:27:48 <gmaxwell> (because the more restrictive nodes have more hash power under reasonable assumptions)
581 2012-02-07 17:27:49 <luke-jr> and then you just release a transaction that causes a blockchain fork based on that
582 2012-02-07 17:28:00 <gmaxwell> hm. it could lead to future splits, true, they'd also resolve but yea.
583 2012-02-07 17:28:15 <luke-jr> gmaxwell: by splitting up the dupe coinbase, you can make each side exclusive
584 2012-02-07 17:28:47 <gmaxwell> luke-jr: Ah. I'm not aware of that txn pattern.
585 2012-02-07 17:29:12 <gavinandresen> I nominate luke to figure out how to fix it. Because he didn't like my "discourage blocks" patch.
586 2012-02-07 17:29:38 <gmaxwell> In any case, all this depends on mining a duplicate coinbase. Which the discourage code, would make into a much lower risk until we work out fixing it better.
587 2012-02-07 17:33:26 <gmaxwell> Discouraging dupe coinbases makes sense even if we fix reorgs, simply because dupecoinbases allow dupe transactions which may be used abusively to confuse third party code even if our handling is fine (and ours probably isn't, even once once we fix reorgs)
588 2012-02-07 17:34:43 <gavinandresen> https://github.com/gavinandresen/bitcoin-git/commit/69945318d1130045f5e0424a6e0e5511c0c76d5e if anybody wants to get a head start on discouraging duplicate coinbases....
589 2012-02-07 17:36:03 <josephcp> that patch ignores the block only if it's at the highest height in the chain?
590 2012-02-07 17:36:47 <gavinandresen> yes, once it is built on it is accepted
591 2012-02-07 17:38:15 <BlueMatt> "// Never discourage our own blocks" why?
592 2012-02-07 17:38:30 <BlueMatt> if a user is modifying bitcoin to do stupid things, lets make them not relay their own blocks
593 2012-02-07 17:38:47 <gavinandresen> Because if you discourage your own blocks it makes testing with -discourageall really hard
594 2012-02-07 17:40:08 <BlueMatt> heh, oh ok
595 2012-02-07 17:44:38 <sipa> BlueMatt: https://github.com/bitcoin/bitcoin/pull/806/files
596 2012-02-07 17:44:50 <sipa> a bit clearer, imho :)
597 2012-02-07 17:45:08 <BlueMatt> nice, but why me?
598 2012-02-07 17:45:24 <sipa> you seem to have added quite some of them :)
599 2012-02-07 17:45:33 <BlueMatt> heh, ok that is true
600 2012-02-07 17:45:33 <sipa> and the other relevant people get an email :p
601 2012-02-07 17:45:45 <gribble> New news from bitcoinrss: sipa opened pull request 806 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/806>
602 2012-02-07 17:47:39 <luke-jr> hmm, interesting
603 2012-02-07 17:48:36 <luke-jr> at no time during bitcoin-qt's gitian build are libz, libpng, or libqrencode accessed :/
604 2012-02-07 17:48:40 <luke-jr> that seems wrong
605 2012-02-07 17:49:10 <BlueMatt> gavinandresen: sorry, I know you hated the idea of variable prefixes but I wanted to get bip21 out the door before 0.6 starts going into rc, and I saw no better ideas so thats what happened... https://github.com/bitcoin/bitcoin/pull/807
606 2012-02-07 17:49:27 <BlueMatt> does it come out with qrcodes enabled?
607 2012-02-07 17:50:32 <gavinandresen> good thing to test, I should have binaries ready for sanity testing very soon
608 2012-02-07 17:50:57 <luke-jr> BlueMatt: dunno, don't got Windoze
609 2012-02-07 17:50:59 <gribble> New news from bitcoinrss: TheBlueMatt opened pull request 807 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/807>
610 2012-02-07 17:51:30 <gavinandresen> wumpus:  does bitcoin-qt need a qmake USE_QRCODES or something to be compiled with qr-code support?  If it does, can you update the doc/release-process.txt doc?
611 2012-02-07 17:51:52 <luke-jr> gavinandresen: it does, but I see it in the gitian script already :/
612 2012-02-07 17:52:40 <sipa> gavinandresen: USE_UPNP, USE_QRCODE, USE_SSL, USE_DBUS are defined
613 2012-02-07 17:53:58 <gavinandresen> sipa: ok, thanks
614 2012-02-07 17:58:43 <sipa> wow, 1228 people on github are watching bitcoin/bitcoin
615 2012-02-07 18:01:37 <luke-jr> gitian win32 does not support QRcodes
616 2012-02-07 18:03:32 <Moron__> burp
617 2012-02-07 18:04:58 <luke-jr> wait
618 2012-02-07 18:05:05 <luke-jr> I'm testing with 0.5.2 <.<
619 2012-02-07 18:05:09 <luke-jr> lemme retry with master
620 2012-02-07 18:06:49 <gavinandresen> 0.6 rc1 binaries ready for sanity testing: https://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.6.0/test/   (Linux/windows, no mac yet)
621 2012-02-07 18:07:23 <FROTUSCI> woohoo
622 2012-02-07 18:14:05 <helo> has nBits ever been the same for two non-adjacent blocks?
623 2012-02-07 18:15:11 <sipa> huh?
624 2012-02-07 18:15:37 <sipa> two non-adjecent blocks of 2016 blocks, you mean maybe?
625 2012-02-07 18:15:54 <luke-jr> helo: not afaik
626 2012-02-07 18:16:03 <luke-jr> helo: *maybe* early on
627 2012-02-07 18:16:41 <sipa> well yes, before the difficulty increases above 1
628 2012-02-07 18:16:53 <helo> sipa: i didn't want to say "has any two blocks had the same nBits", as every block of 2016 has the same nBits... so i settled on the above :)
629 2012-02-07 18:17:05 <luke-jr> sipa: did it go back down to 1?
630 2012-02-07 18:17:28 <sipa> no, but there were several blocks of 2016 blocks with difficulty 1
631 2012-02-07 18:17:44 <luke-jr> those blocks were all adjacent to each other :P
632 2012-02-07 18:18:38 <helo> hah
633 2012-02-07 18:19:21 <sipa> 8, actually
634 2012-02-07 18:19:57 <gavinandresen> 0.6rc1 mac build just finished uploading to sourceforge
635 2012-02-07 18:20:14 <sipa> need gitian builds?
636 2012-02-07 18:21:13 <gavinandresen> "need" might be too strong a word, but sure, it'd be good to know if the builds can be reproduced
637 2012-02-07 18:21:30 <gavinandresen> I'm away for an hour or so, be back later.
638 2012-02-07 18:35:44 <wumpus> gavinandresen: seems it is described in build-unix.txt, but indeed not in release process
639 2012-02-07 18:37:33 <wumpus> hm, gitian.yml and gitian-win32.yml already have USE_QRENCODE=1 embedded by default, why would we explicitly want to mention it?
640 2012-02-07 18:38:22 <luke-jr> wumpus: nm, it was a problem with me building 0.5.2 tag
641 2012-02-07 18:39:39 <wumpus> it should be added to readme-qt.rst though...
642 2012-02-07 18:40:14 <sipa> wumpus: USE_QRENCODE or USE_QRCODE ?
643 2012-02-07 18:40:24 <sipa> the qt buildfile and the source code use the latter
644 2012-02-07 18:41:34 <wumpus> yes it's USE_QRCODE
645 2012-02-07 19:12:04 <XMPPwocky> hey, can someone give me the IP of their bitcoin node?
646 2012-02-07 19:12:37 <edcba> try 127.0.0.1
647 2012-02-07 19:12:50 <sipa> XMPPwocky: bitcoin.sipa.be
648 2012-02-07 19:14:10 <XMPPwocky> sipa: thanks
649 2012-02-07 19:21:05 <XMPPwocky> socat -x TCP4-LISTEN:8333 TCP4:217.133.49.52:8333 2>testpackets/session
650 2012-02-07 19:21:07 <XMPPwocky> o/
651 2012-02-07 19:22:52 <TripleSpeeder> Whats going on? Last block mined 47 minutes ago?
652 2012-02-07 19:26:04 <phantomcircuit> TripleSpeeder, well within statistical variance
653 2012-02-07 19:26:10 <sipa> ;;bc,probd [calc 1000000*[bc,nethash]] 47m
654 2012-02-07 19:26:12 <gribble> Error: There's really no reason why you should have underscores or brackets in your mathematical expression.  Please remove them.
655 2012-02-07 19:26:29 <sipa> ;;calc 1000000*[bc,nethash]
656 2012-02-07 19:26:30 <gribble> 9954538293
657 2012-02-07 19:26:37 <gribble> Error: "bc,probdf" is not a valid command.
658 2012-02-07 19:26:37 <sipa> ;;bc,probdf
659 2012-02-07 19:26:38 <gribble> (bc,probd <an alias, at least 2 arguments>) -- Alias for "math calc 1-exp(-$1*1000 * [seconds $*] / (2**32* $2))".
660 2012-02-07 19:26:38 <sipa> ;;bc,probd
661 2012-02-07 19:27:20 <sipa> nanotube: what am i doing wrong? :)
662 2012-02-07 19:27:36 <TripleSpeeder> Hope you're right. Still i'm scared somethings going on here :-(
663 2012-02-07 19:28:19 <sipa> TripleSpeeder: an hour is not uncommon
664 2012-02-07 19:28:24 <Eliel> ;;bc,probd 9954538293 47m
665 2012-02-07 19:28:24 <gribble> Error: There's really no reason why you should have underscores or brackets in your mathematical expression.  Please remove them.
666 2012-02-07 19:28:42 <Eliel> ;;bc,probd 9954538293 47
667 2012-02-07 19:28:42 <gribble> Error: There's really no reason why you should have underscores or brackets in your mathematical expression.  Please remove them.
668 2012-02-07 19:28:49 <gribble> Error: There's really no reason why you should have underscores or brackets in your mathematical expression.  Please remove them.
669 2012-02-07 19:28:49 <sipa> ;;bc,probd [calc 1000000*[bc,nethash]] 47 min
670 2012-02-07 19:28:54 <TripleSpeeder> Okay :-) Never noticed such long times before...
671 2012-02-07 19:29:02 <gribble> 2820
672 2012-02-07 19:29:02 <sipa> ;;seconds 47m
673 2012-02-07 19:29:06 <Eliel> no underscores or brackets in my expression :P
674 2012-02-07 19:29:43 <gribble> Error: invalid syntax (<string>, line 1)
675 2012-02-07 19:29:43 <sipa> ;;calc 1-exp(-[bc,nethash]*1000000000*[seconds 47m] / (2**32*$2))
676 2012-02-07 19:30:02 <Eliel> you forgot to replace $2
677 2012-02-07 19:30:59 <sipa> ah!
678 2012-02-07 19:31:06 <BTC_Bear> math calc uses precision, i believe
679 2012-02-07 19:31:09 <sipa> ;;bc,prob [calc 1000000*[bc,nethash]] 47m
680 2012-02-07 19:31:10 <gribble> 0.991238813431
681 2012-02-07 19:31:39 <sipa> ;;calc 1/(1-[bc,prob [calc 1000000*[bc,nethash]] 47m])
682 2012-02-07 19:31:40 <gribble> 114.142221208
683 2012-02-07 19:31:44 <diki> That's a nice probability. I like it!
684 2012-02-07 19:31:53 <Eliel> now, what do these numbers mean? :)
685 2012-02-07 19:31:56 <sipa> TripleSpeeder: once every 114 blocks, it takes over 47 minutes
686 2012-02-07 19:32:12 <TripleSpeeder> Hehe, nice math :)
687 2012-02-07 19:32:14 <diki> Eliel:it's a secret ;)
688 2012-02-07 19:35:36 <Graet> and whose reality ;)
689 2012-02-07 19:35:40 <Graet> an hour now :)
690 2012-02-07 19:36:11 <sipa> ;;calc 1/(1-[bc,prob [calc 1000000*[bc,nethash]] 1h])
691 2012-02-07 19:36:12 <gribble> 423.190859077
692 2012-02-07 19:36:59 <TripleSpeeder> What was the longest break we had so far in the blockchain?
693 2012-02-07 19:37:37 <gmaxwell> TripleSpeeder: it's a bit hard to tell accurately because the timestamps are not all accurate.
694 2012-02-07 19:38:02 <Eliel> Graet: the actual block times as measured from the blocks that make up the blockchain. I don't know whos reality to label that :P
695 2012-02-07 19:38:11 <Graet> :D
696 2012-02-07 19:38:11 <TripleSpeeder> ;-)
697 2012-02-07 19:38:32 <gmaxwell> TripleSpeeder: e.g. a block with a timestamp 2 hours in the future did not always take two hours. :)
698 2012-02-07 19:38:51 <TripleSpeeder> hmm
699 2012-02-07 19:39:11 <sipa> longest delay between two blocks is over a day
700 2012-02-07 19:39:38 <Eliel> Graet: of course, depending on who you are, the exact times you first see any given block can vary quite a bit :)
701 2012-02-07 19:39:40 <gmaxwell> sipa: and since the difficulty has been greater than 1?
702 2012-02-07 19:39:54 <gmaxwell> sipa: it's no fair to count back when it was 1 because the system couldn't adapt below 1.
703 2012-02-07 19:39:59 <sipa> agree
704 2012-02-07 19:40:19 <Graet> lol Eliel true :)
705 2012-02-07 19:41:04 <Graet> be interesting to see the level of panic if it were to take 24 hours between blocks :P
706 2012-02-07 19:41:09 <luke-jr> maybe Eligius isn't the only pool to be DDoS'd
707 2012-02-07 19:41:11 <luke-jr> ?
708 2012-02-07 19:41:38 <luke-jr> BTCServ, ozcoin, and Eligius all down pretty much
709 2012-02-07 19:41:54 <Graet> ozcoin isnt down from ddos tho :)
710 2012-02-07 19:42:01 <BlueMatt> is eligius down from the ddos?
711 2012-02-07 19:42:04 <sipa> gmaxwell: 6 hours; Sun Aug 15 23:53:59 UTC 2010
712 2012-02-07 19:42:21 <gmaxwell> thats not fair either.
713 2012-02-07 19:42:29 <gmaxwell> Thats the return bug fix.
714 2012-02-07 19:42:41 <Eliel> ok, happened pretty fast.
715 2012-02-07 19:42:58 <gmaxwell> (er, value overflow, sorry)
716 2012-02-07 19:43:07 <sipa> ok, next one is 2.3 hours: Thu Oct 13 08:40:50 UTC 201
717 2012-02-07 19:43:13 <sipa> 2011
718 2012-02-07 19:43:22 <gmaxwell> Okay, I can't temember an incident from then.
719 2012-02-07 19:43:23 <sipa> not too long ago, actually
720 2012-02-07 19:43:32 <Eliel> 2.3 hours... that's quite some waiting
721 2012-02-07 19:43:32 <TripleSpeeder> Yep, not soo long ago.
722 2012-02-07 19:43:38 <gribble> Error: '2.3h' is not a valid argument.
723 2012-02-07 19:43:38 <sipa> ;;calc 1/(1-[bc,prob [calc 1000000*[bc,nethash]] 2.3h])
724 2012-02-07 19:44:01 <sipa> ;;calc 1/(1-[bc,prob [calc 1000000*[bc,nethash]] 138m])
725 2012-02-07 19:44:04 <gribble> 999999.999971
726 2012-02-07 19:44:35 <Eliel> ok... I don't think I'm quite buying that one as pure chance.
727 2012-02-07 19:44:37 <sipa> someone playing with timestamps
728 2012-02-07 19:44:52 <gmaxwell> sipa: actually I know that block that was some eligius weirdness I think.
729 2012-02-07 19:45:30 <Eliel> sipa: what are the blocks right before and after that one?
730 2012-02-07 19:45:43 <gmaxwell> at least if its the one I'm thinking of.
731 2012-02-07 19:46:22 <Eliel> perhaps you can get some bounds on the real time for the block from those.
732 2012-02-07 19:46:32 <Graet> lol i remember that gmaxwell
733 2012-02-07 19:47:45 <BlueMatt> ;;later tell genjix ok, probably last one and a very minor one at that (just a clarification) https://github.com/genjix/bips/pull/4
734 2012-02-07 19:47:45 <gribble> The operation succeeded.
735 2012-02-07 19:49:14 <gmaxwell> It should probably worry me that I know things about specific interesting blocks. At least I don't remember their heights.
736 2012-02-07 19:50:28 <[Tycho]> :)
737 2012-02-07 19:50:45 <BlueMatt> hey, at least someone is paying attention to individual blocks
738 2012-02-07 19:50:47 <nanotube> sipa: bc,probd takes three args (hashrate, difficulty, and time). bc,prob takes two args (hashrate, and time)
739 2012-02-07 19:50:55 <[Tycho]> I like some blocks.
740 2012-02-07 19:51:01 <sipa> nanotube: thanks, realized it already :)
741 2012-02-07 19:51:03 <[Tycho]> But mostly I'm interested in strange TXes.
742 2012-02-07 19:51:41 <sipa> actually
743 2012-02-07 19:52:15 <sipa> ;;bc,probd [calc 2**32/600/1000] 1h 1
744 2012-02-07 19:52:16 <gribble> Error: There's really no reason why you should have underscores or brackets in your mathematical expression.  Please remove them.
745 2012-02-07 19:52:28 <BlueMatt> heh
746 2012-02-07 19:52:29 <nanotube> sipa: ah yea, caught up on the scrollback :)
747 2012-02-07 19:52:31 <nanotube> also, time last.
748 2012-02-07 19:52:37 <sipa> ;;bc,probd [calc 2**32/600/1000] 1 1h
749 2012-02-07 19:52:38 <gribble> 0.997521247823
750 2012-02-07 19:52:49 <luke-jr> BlueMatt: yes
751 2012-02-07 19:52:55 <gribble> 403.38846309
752 2012-02-07 19:52:55 <sipa> ;;calc 1/(1-[bc,probd [calc 2**32/600/1000] 1 1h])
753 2012-02-07 19:53:06 <BlueMatt> luke-jr: in response to?
754 2012-02-07 19:53:13 <luke-jr> [15:42:01] <BlueMatt> is eligius down from the ddos?
755 2012-02-07 19:53:14 <BlueMatt> luke-jr: oh, nvm
756 2012-02-07 19:53:19 <BlueMatt> damn that sucks
757 2012-02-07 19:53:58 <luke-jr> I think it's just the datacenter's null routing of our IPs tho
758 2012-02-07 19:54:00 <luke-jr> at this point
759 2012-02-07 19:54:28 <BlueMatt> damn
760 2012-02-07 19:55:40 <[Tycho]> Yeah, I hate nullrouting.
761 2012-02-07 19:55:52 <[Tycho]> Otherwise there would be no problems with DDoS.
762 2012-02-07 19:56:47 <BlueMatt> what?
763 2012-02-07 19:58:25 <luke-jr> today's headlines: Tycho is OK with DDoSing his pool
764 2012-02-07 19:58:26 <luke-jr> :p
765 2012-02-07 19:58:43 <BlueMatt> yep, everyone go ddos deepbit
766 2012-02-07 19:59:01 <luke-jr> "Tycho is OK with DDoSing his pool; blames datacenter"
767 2012-02-07 19:59:17 <BlueMatt> heh
768 2012-02-07 19:59:40 <BlueMatt> who will fund the cash?
769 2012-02-07 19:59:42 <gmaxwell> "In soviet russia, pool dosses you"
770 2012-02-07 19:59:46 <luke-jr> LOL
771 2012-02-07 20:00:20 <luke-jr> "no problems with DDoS." -Tycho ;)
772 2012-02-07 20:00:58 <BlueMatt> lets get this ddos going, who knows botnet ops?
773 2012-02-07 20:01:08 <luke-jr> when I finally see Gavin in person, I owe him a brownie for the nice test suite
774 2012-02-07 20:07:03 <TD> ah nice
775 2012-02-07 20:07:11 <TD> somebody contributed wallet tx completion to bitcoinj
776 2012-02-07 20:07:16 <TD> the api is shaping up quite nicely
777 2012-02-07 20:08:45 <sipa> "completion" ?
778 2012-02-07 20:08:56 <TD> you create a Transaction object and add some outputs of whatever form you want
779 2012-02-07 20:09:04 <TD> then you give it to the wallet which adds inputs and a change output if necessary
780 2012-02-07 20:09:12 <BlueMatt> nice
781 2012-02-07 20:09:17 <sipa> sounds clean
782 2012-02-07 20:09:23 <luke-jr> +
783 2012-02-07 20:09:43 <TD> makes it a lot easier to create multi-sends and should make it easy to experiment with contracts
784 2012-02-07 20:09:50 <TD> though right now the script editor api is pretty woeful