1 2012-05-13 00:01:02 <jgarzik> sipa: that's a good question...  how does IRC and IPv4 addresses interact?  do we use a new nick encoding, pick first local ipv4 address, ignore all ipv6 addresses, etc.?
  2 2012-05-13 00:01:17 <sipa> jgarzik: i say ignore; irc is deprecated
  3 2012-05-13 00:01:33 <jgarzik> sipa: agree
  4 2012-05-13 00:02:29 <jgarzik> issue #1273: "it will be nice to add commands like ???addnode??? and ???listnodes??? to PRC interface, so it will be unneded to restart bitcoind to manual node adding on, for example, testing IPv6 connectivity ;)"
  5 2012-05-13 00:02:32 <jgarzik> ^^ agreed
  6 2012-05-13 00:02:38 <jgarzik> in general some networking rpcs would be nice
  7 2012-05-13 00:03:10 <sipa> addnode is a bit hard, but a "open a connection NOW to <...>" is very doable
  8 2012-05-13 00:05:03 <jgarzik> sipa: the latter should be all that's needed
  9 2012-05-13 00:05:06 <jgarzik> imo
 10 2012-05-13 00:05:58 <BlueMatt> addnode shouldnt be that hard...just add it to the addnode vector?
 11 2012-05-13 00:06:23 <BlueMatt> or has addnode changed with ipv6?
 12 2012-05-13 00:06:27 <sipa> no, not really
 13 2012-05-13 00:06:45 <sipa> BlueMatt: gmaxwell had some complaint about its behaviour though (see above)
 14 2012-05-13 00:07:16 <sipa> jgarzik: anyway; no indeed, not hard; but "connect now to" seems more useful as an RPC
 15 2012-05-13 00:07:20 <BlueMatt> did I code that wrong, or did it change?
 16 2012-05-13 00:07:28 <BlueMatt> it used to try one per addnode arg
 17 2012-05-13 00:07:33 <BlueMatt> (ie once per dns name)
 18 2012-05-13 00:07:34 <sipa> i may have messed things up :)
 19 2012-05-13 00:07:44 <sipa> but i don't think i touched it too much
 20 2012-05-13 00:08:12 <BlueMatt> well maybe it was always broken...either way, needs fixed
 21 2012-05-13 00:08:27 <n0n00dz4u> whats up crackers
 22 2012-05-13 00:08:59 <Diablo-D3> n0n00dz4u: everyone is celebrating DMC's successful launch
 23 2012-05-13 00:09:18 <BlueMatt> dmc?
 24 2012-05-13 00:09:29 <BlueMatt> oh...
 25 2012-05-13 00:09:34 <Diablo-D3> Diablo Mining Company!
 26 2012-05-13 00:09:41 <Diablo-D3> 18 more shares and we pass 250
 27 2012-05-13 00:09:43 <n0n00dz4u> congrats!
 28 2012-05-13 00:09:50 <BlueMatt> tell me you havent made /that/ much money...
 29 2012-05-13 00:10:13 <Diablo-D3> blueMatt: yes, I am the proud custodian of a large sum of bitcoins.
 30 2012-05-13 00:10:36 <n0n00dz4u> hopefully I can secure (read beg) for a start up loan :)
 31 2012-05-13 00:10:55 <BlueMatt> I dont know what to think of that...on one hand, its better that you have more mining than deepbit, on the the other...
 32 2012-05-13 00:11:30 <Diablo-D3> bluematt: more? no, read the revised plan
 33 2012-05-13 00:11:42 <Diablo-D3> the $1m plan doesnt give me 3.5 thash on day one
 34 2012-05-13 00:11:43 <n0n00dz4u> congratulatory vlog to be released in honor of DMC
 35 2012-05-13 00:11:52 <n0n00dz4u> also live from flannigans
 36 2012-05-13 00:11:55 <BlueMatt> well, add some mining power which decreases deepbit's %
 37 2012-05-13 00:12:37 <Diablo-D3> blueMatt: if I do everything right, we get north of 1 thash
 38 2012-05-13 00:12:47 <BlueMatt> damn...
 39 2012-05-13 00:12:49 <Diablo-D3> but not on the first day of full power
 40 2012-05-13 00:13:08 <Diablo-D3> the $1m plancalls for 40 BFL MiniRigs
 41 2012-05-13 00:13:12 <Diablo-D3> which is 1008 ghash
 42 2012-05-13 00:13:33 <n0n00dz4u> tweeted
 43 2012-05-13 00:13:44 <n0n00dz4u> what am I drinking diablo?
 44 2012-05-13 00:13:48 <n0n00dz4u> glenlevit!
 45 2012-05-13 00:14:00 <n0n00dz4u> i'm thinking glenlevit
 46 2012-05-13 00:16:13 <Diablo-D3> you're talking to someone who doesnt drink.
 47 2012-05-13 00:16:34 <n0n00dz4u> Smoke pot?
 48 2012-05-13 00:16:41 <n0n00dz4u> *elbow jabs*
 49 2012-05-13 00:17:06 <Diablo-D3> nope.
 50 2012-05-13 00:17:22 <n0n00dz4u> well everyone has something
 51 2012-05-13 00:17:29 <n0n00dz4u> I enjoy being intoxicated
 52 2012-05-13 00:17:39 <n0n00dz4u> and in light of recent developments
 53 2012-05-13 00:17:41 <n0n00dz4u> and you
 54 2012-05-13 00:17:44 <n0n00dz4u> your success
 55 2012-05-13 00:17:53 <n0n00dz4u> I'm ordering a scotch and recording a vlog
 56 2012-05-13 00:18:07 <n0n00dz4u> thats how deathbylollipop treats their friends for the record
 57 2012-05-13 00:20:22 <jgarzik> Diablo-D3: "$1m USD"  wasn't it originally 1 million bitcoins?  or am I misremembering?
 58 2012-05-13 00:20:35 <Diablo-D3> it got downsized
 59 2012-05-13 00:20:43 <Diablo-D3> due to the poll of epic lamerity.
 60 2012-05-13 00:22:02 <jgarzik> Diablo-D3: do you live in Maine already?  (i.e. is DMC facility remote or local)
 61 2012-05-13 00:22:21 <BlueMatt> arent there cheaper states for power costs?
 62 2012-05-13 00:22:31 <Diablo-D3> jgarzik: yes, I am STUCK IN THIS FROZEN WASTELAND
 63 2012-05-13 00:22:36 <Diablo-D3> I MEAN, IM LOOKING OUT OF MY WINDOW
 64 2012-05-13 00:22:48 <Diablo-D3> AND I SEE A HUNTER RIDING A POLAR BEAR
 65 2012-05-13 00:23:00 <BlueMatt> heh, it was high 70s today in philli
 66 2012-05-13 00:23:06 <BlueMatt> also, somehow I doubt that...
 67 2012-05-13 00:23:26 <jgarzik> BlueMatt: DMC is apparently solar
 68 2012-05-13 00:23:32 <BlueMatt> ahhh, nice
 69 2012-05-13 00:23:39 <BlueMatt> wait, wtf?
 70 2012-05-13 00:23:41 <BlueMatt> solar in maine?
 71 2012-05-13 00:23:51 <BlueMatt> thats not efficient...
 72 2012-05-13 00:23:55 <jgarzik> BlueMatt: https://bitcointalk.org/index.php?topic=77469.0
 73 2012-05-13 00:23:58 <jgarzik> BlueMatt: no, it's not :)
 74 2012-05-13 00:24:31 <jgarzik> ideal is Iceland:  geothermal power and easy cooling ;)
 75 2012-05-13 00:24:41 <BlueMatt> yea...
 76 2012-05-13 00:25:31 <BlueMatt> ideal are a lot of places other than maine
 77 2012-05-13 00:26:53 <Diablo-D3> jgarzik: that statement is not entirely correct
 78 2012-05-13 00:27:20 <jgarzik> Diablo-D3: no statement ever is
 79 2012-05-13 00:27:32 <Diablo-D3> actually, it just fails all over the places
 80 2012-05-13 00:27:33 <BlueMatt> 89% of statistics are made-up
 81 2012-05-13 00:27:42 <Diablo-D3> I mean, one
 82 2012-05-13 00:27:53 <Diablo-D3> I REALLY want a 9000 foot tall fucking gigantic wind mill of DOOM
 83 2012-05-13 00:28:08 <Diablo-D3> and Im going to install speakers at the base of it, the outdoor weatherized kind
 84 2012-05-13 00:28:11 <BlueMatt> and since when is maine windy...?
 85 2012-05-13 00:28:15 <Diablo-D3> and have it play that music from n64
 86 2012-05-13 00:28:20 <Diablo-D3> er zelda n64
 87 2012-05-13 00:28:20 <jgarzik> regardless, wind power is quickly being proven stupid
 88 2012-05-13 00:28:36 <Diablo-D3> bluematt: coastal maine has some of the best wind in the USA
 89 2012-05-13 00:28:57 <BlueMatt> this whole lets place my new datacenter in maine because I happen to live here business doesnt seem right to me...
 90 2012-05-13 00:28:57 <Diablo-D3> jgarzik: "proven" stupid?
 91 2012-05-13 00:28:59 <Diablo-D3> how so?
 92 2012-05-13 00:29:22 <Diablo-D3> bluematt: no different than doing it in alaska, really
 93 2012-05-13 00:29:39 <BlueMatt> well wtf would you put a datacenter in alaska, either?
 94 2012-05-13 00:29:52 <BlueMatt> hmmm...didnt realize that...though then shouldnt you put your windmill off the coast in the ocean?
 95 2012-05-13 00:30:13 <Diablo-D3> coastal includes NOT in the water too
 96 2012-05-13 00:30:22 <luke-jr> sipa: addnode {'ip':&, 'port':&, 'immediate':bool, 'retain':bool} ?
 97 2012-05-13 00:30:30 <BlueMatt> yea, but in the water would be more efficient, more winds there...
 98 2012-05-13 00:30:32 <Diablo-D3> maine has surprisingly lots of wind
 99 2012-05-13 00:30:35 <Diablo-D3> people forget this
100 2012-05-13 00:30:51 <BlueMatt> or didnt know it, because...who travels to maine???
101 2012-05-13 00:31:16 <Diablo-D3> yeah
102 2012-05-13 00:31:19 <Diablo-D3> but this is the good part
103 2012-05-13 00:31:23 <jgarzik> Diablo-D3: in most places.  Maine may be an exception, I admit
104 2012-05-13 00:31:29 <Diablo-D3> land is cheap, its cold, and theres no people
105 2012-05-13 00:31:36 <Diablo-D3> and solar is NOT the worst in the nation
106 2012-05-13 00:31:42 <BlueMatt> theres no people for a reason...
107 2012-05-13 00:31:55 <Diablo-D3> like, people went MOVE TO OREGON OR WASHINGTON
108 2012-05-13 00:32:01 <Diablo-D3> those states have _worse_ solar than maine
109 2012-05-13 00:32:05 <jgarzik> Diablo-D3: the turbines involve huge capital costs, the maintenance is huge, and the current tech constantly breaks down (go go zero power output)
110 2012-05-13 00:32:12 <Diablo-D3> because you keep fucking yourselves over with the insane weather patterns
111 2012-05-13 00:32:20 <Diablo-D3> it rains EVERY SINGLE DAY there
112 2012-05-13 00:32:21 <Diablo-D3> SINGLE
113 2012-05-13 00:32:22 <Diablo-D3> DAY
114 2012-05-13 00:32:28 <Diablo-D3> when the fuck does the SUN come out
115 2012-05-13 00:32:34 <Diablo-D3> how the fuck do you people even LIVE
116 2012-05-13 00:32:40 <jgarzik> subsidies are the only thing keeping wind power alive in ~80% of the installed US locations
117 2012-05-13 00:32:43 <BlueMatt> same way as in england ;)
118 2012-05-13 00:32:55 <Diablo-D3> jgarzik: fuck the subsidies
119 2012-05-13 00:33:12 <Diablo-D3> and bullshit, current tech DOESNT constantly break down
120 2012-05-13 00:33:23 <BlueMatt> jgarzik: hey, look at germany, half the country has solar on its roof....and you are north of maine with the cloud cover of the uk...
121 2012-05-13 00:33:24 <Diablo-D3> every break down Im aware of in the past 5 years has been 20 year old turbines.
122 2012-05-13 00:33:30 <jgarzik> Diablo-D3: your DMC prospectus should include the quote "it rains EVERY SINGLE DAY there" in the section on solar power
123 2012-05-13 00:33:35 <jgarzik> under risks
124 2012-05-13 00:33:46 <Diablo-D3> jgarzik: in maine it isnt really bad weather
125 2012-05-13 00:33:55 <Diablo-D3> and 1200kwh/yr is Maine specific
126 2012-05-13 00:34:02 <Diablo-D3> its what has been measured on real panels
127 2012-05-13 00:34:11 <jgarzik> Diablo-D3: really?  well done, then.
128 2012-05-13 00:34:27 <Diablo-D3> I mean, I really cant find a better place than maine without moving to canada
129 2012-05-13 00:34:32 <luke-jr> Diablo-D3: solar panels produce more power water-cooled, fwiw
130 2012-05-13 00:34:43 <Diablo-D3> Im actually in luck that I already live here
131 2012-05-13 00:35:33 <BlueMatt> luck is a rather broad term, in this case...
132 2012-05-13 00:36:39 <Diablo-D3> bluematt: panels are getting better
133 2012-05-13 00:36:45 <BlueMatt> still...
134 2012-05-13 00:36:53 <Diablo-D3> btw, where I said 1200khw/yr per panel, thats a 1kw panel
135 2012-05-13 00:36:57 <Diablo-D3> just to be perfectly clear
136 2012-05-13 00:37:09 <Diablo-D3> they manufacture panels of lots of sizes
137 2012-05-13 00:37:13 <Diablo-D3> some much smaller than that
138 2012-05-13 00:37:21 <Diablo-D3> so adjust your measurements accordingly
139 2012-05-13 00:37:25 <luke-jr> I would think Florida has much better solar than Maine
140 2012-05-13 00:37:29 <luke-jr> but it does rain a lot here, sometimes
141 2012-05-13 00:37:31 <Diablo-D3> luke-jr: its not the best
142 2012-05-13 00:37:41 <Diablo-D3> theres a part in arizona that gets insane solar
143 2012-05-13 00:37:41 <JFK911> what panel gives 1kw
144 2012-05-13 00:37:49 <Diablo-D3> JFK911: they make 1kw panels
145 2012-05-13 00:37:52 <Diablo-D3> they're rather huge
146 2012-05-13 00:37:57 <JFK911> what is its dimensions and mass
147 2012-05-13 00:38:22 <JFK911> i have a pallet of kyocera 85 watt things
148 2012-05-13 00:38:32 <Diablo-D3> arent those just cells?
149 2012-05-13 00:38:41 <JFK911> no they are assembled panels
150 2012-05-13 00:38:46 <Diablo-D3> wtf how big are they
151 2012-05-13 00:38:48 <Diablo-D3> they must be tiny
152 2012-05-13 00:38:48 <JFK911> you can buy indiv cells on ebay
153 2012-05-13 00:38:55 <JFK911> they are roof sized
154 2012-05-13 00:38:59 <JFK911> ive got more than 40 on the pallet
155 2012-05-13 00:39:32 <Diablo-D3> yeah, the 1kw ones Ive seen are like 3 by 6 foot or something
156 2012-05-13 00:39:37 <JFK911> 40 x 26 inches
157 2012-05-13 00:39:47 <JFK911> 20 lbs each
158 2012-05-13 00:39:59 <n0n00dz4u> phones charging
159 2012-05-13 00:40:00 <JFK911> mine are old but still sold today
160 2012-05-13 00:40:05 <n0n00dz4u> hey diablo
161 2012-05-13 00:40:06 <JFK911> and not cheaply either
162 2012-05-13 00:40:13 <Diablo-D3> n0n00dz4u: hey
163 2012-05-13 00:40:17 <n0n00dz4u> how man bitcoins do you think I can mine on my blackberry torch?
164 2012-05-13 00:40:38 <n0n00dz4u> 1 million possibly?
165 2012-05-13 00:41:04 <Diablo-D3> JFK911: they have like 500 cells depending on the make I think
166 2012-05-13 00:41:28 <JFK911> google for type kc85
167 2012-05-13 00:42:30 <Diablo-D3> JFK911: that says 72 cells per panel
168 2012-05-13 00:42:34 <Diablo-D3> and those are pretty funny shaped cells
169 2012-05-13 00:42:51 <JFK911> you were expecting hex shape?
170 2012-05-13 00:43:03 <Diablo-D3> no, the more square ones Ive seen
171 2012-05-13 00:43:51 <JFK911> well i dont have connectors
172 2012-05-13 00:43:58 <JFK911> thats why i havent used any yet
173 2012-05-13 00:49:25 <jgarzik> Diablo-D3: what number of sold shares is DMC's "start moving" point?
174 2012-05-13 00:50:12 <jgarzik> Diablo-D3: if you are set on wind, sounds like that's the first power source?  because obviously solar won't get you running 24/7 without overcapacity + storage
175 2012-05-13 00:50:31 <Diablo-D3> jgarzik: Im not send on wind, btw
176 2012-05-13 00:50:41 <Diablo-D3> and btw, theres no storage
177 2012-05-13 00:50:46 <Diablo-D3> we just sell it back to the grid
178 2012-05-13 00:51:06 <jgarzik> Diablo-D3: yeah that works as long as you are net positive
179 2012-05-13 00:51:22 <n0n00dz4u> vlog recorded
180 2012-05-13 00:52:47 <Diablo-D3> jgarzik: yes
181 2012-05-13 00:53:01 <jgarzik> other options for storage -- excluding batteries, that is -- include compressed air, elevated water, fuel cell production, ...
182 2012-05-13 00:53:12 <Diablo-D3> yeah yeah yeah, it costs money.
183 2012-05-13 00:53:17 <Diablo-D3> its cheaper just to overbuy cells.
184 2012-05-13 00:53:22 <jgarzik> probably
185 2012-05-13 00:53:25 <gmaxwell> jgarzik: flywheels are fun!
186 2012-05-13 00:53:33 <jgarzik> ohhh I forgot flywheels
187 2012-05-13 00:53:43 <Diablo-D3> money, it costs.
188 2012-05-13 00:54:10 <n0n00dz4u> I'm promoting for Diablo just so everyone knows
189 2012-05-13 00:54:20 <jgarzik> yeah grid tie + sufficient photovoltaic to keep your bill >0 definitely makes a lot of sense
190 2012-05-13 00:54:34 <gmaxwell> <=0
191 2012-05-13 00:54:42 <gmaxwell> darn boundary conditions.
192 2012-05-13 00:55:05 <gmaxwell> Greedy people like negative bills too.
193 2012-05-13 00:55:07 <jgarzik> if a miner explodes... power profit!
194 2012-05-13 00:55:20 <jgarzik> but really
195 2012-05-13 00:55:23 <Diablo-D3> yeah what gmaxwell said
196 2012-05-13 00:55:26 <jgarzik> with $1m just do an ASIC
197 2012-05-13 00:55:28 <Diablo-D3> negative bills means....
198 2012-05-13 00:55:35 <Diablo-D3> profits.
199 2012-05-13 00:55:36 <Diablo-D3> fuck yeah.
200 2012-05-13 00:55:40 <Diablo-D3> jgarzik: and no
201 2012-05-13 00:55:44 <Diablo-D3> thats $10m
202 2012-05-13 00:55:47 <Diablo-D3> and
203 2012-05-13 00:55:48 <jgarzik> hardly
204 2012-05-13 00:55:52 <Diablo-D3> thats just to GET them
205 2012-05-13 00:55:57 <jgarzik> nope
206 2012-05-13 00:56:05 <Diablo-D3> I have to sell all of them not to go out of business
207 2012-05-13 00:56:08 <luke-jr> jgarzik: ASICs are for sale, not use ;)
208 2012-05-13 00:56:16 <Diablo-D3> jgarzik: dude, sasic and shit asic is shit
209 2012-05-13 00:56:21 <Diablo-D3> Im not going to fuck with that
210 2012-05-13 00:56:30 <Diablo-D3> either I lay waste to all that I can see, or Im not going to do it
211 2012-05-13 00:56:34 <luke-jr> jgarzik: gmaxwell proposes if someone makes ASICs and won't sell them, we change the block hashing algo on them
212 2012-05-13 00:56:57 <Diablo-D3> gmaxwell can, like
213 2012-05-13 00:57:16 <Diablo-D3> I dont disagree with him actually
214 2012-05-13 00:57:19 <luke-jr> (selling them prevents a 51% attack, and ensures there are enough people upset over such a change that it would fail)
215 2012-05-13 01:05:03 <luke-jr> jgarzik: ping
216 2012-05-13 01:05:41 <luke-jr> jgarzik: where is the general locking mechanism for the multithreaded JSON-RPC? or are the calls all threadsafe now?
217 2012-05-13 01:08:45 <JFK911> Diablo-D3: dude you are by the ocean, use wave energy
218 2012-05-13 01:09:11 <Diablo-D3> I love how people think ocean front property is cheap, and navigating bureaucracy is easy
219 2012-05-13 01:11:15 <Diablo-D3> sometime in the future? who knows
220 2012-05-13 01:11:20 <Diablo-D3> but not today, thats for sure
221 2012-05-13 01:12:11 <gmaxwell> just build your stuff in standard shipping containers (they're cheap in the US because we don't export enough anyways)... then your farm can be easily moved
222 2012-05-13 01:12:30 <gmaxwell> Cheap power in antartica? no problem.
223 2012-05-13 01:12:48 <Diablo-D3> gmaxwell: I'd like to keep it in the US
224 2012-05-13 01:13:12 <gmaxwell> Cheap power in alaska? no problem.
225 2012-05-13 01:13:14 <freewil> shipping container is actually a really good idea
226 2012-05-13 01:13:24 <Diablo-D3> Im not moving to alaska
227 2012-05-13 01:13:27 <freewil> just move it whenever electricity costs change
228 2012-05-13 01:13:28 <Diablo-D3> they arent even part of the US anymore
229 2012-05-13 01:13:37 <Diablo-D3> freewil: okay, so, how do I move my solar and wind farm?
230 2012-05-13 01:13:45 <Diablo-D3> you cant get cheaper than free, anyways.
231 2012-05-13 01:14:29 <freewil> you build the solar panels ontop of a collapsing platform that folds into the shipping container :)
232 2012-05-13 01:14:35 <gmaxwell> Diablo-D3: you sell your farm and move where the cheap power is. Don't be so location biased. There are people waiting to be exploited all over the world.
233 2012-05-13 01:15:52 <Diablo-D3> gmaxwell: Im not location biased
234 2012-05-13 01:15:55 <Diablo-D3> but why would I _sell_ it
235 2012-05-13 01:15:58 <Diablo-D3> that seems short sighted
236 2012-05-13 01:18:32 <freewil> you could ship it to china and set it up in some dirty ass coal city
237 2012-05-13 01:18:39 <freewil> probably wont get much solar there though
238 2012-05-13 01:19:30 <Diablo-D3> my hardware doesnt leave the US.
239 2012-05-13 01:20:45 <jgarzik> luke-jr: CRPCTable::execute() ... LOCK2(cs_main, pwalletMain->cs_wallet)
240 2012-05-13 01:21:20 <luke-jr> jgarzik: is there any existing way to unlock those inside the scope?
241 2012-05-13 01:21:37 <jgarzik> luke-jr: you would change that to something like "if strMethod == getwork" { call actor } else { LOCK2()... call actor }
242 2012-05-13 01:22:12 <jgarzik> conditional locking is ugly, but sometimes I'm an ugly guy ;)
243 2012-05-13 01:22:35 <jgarzik> maybe put a big "this code is not your role model" comment
244 2012-05-13 01:23:48 <freewil> i hate comments like that, it's like "hey, im actually a better programmer than this, but i dont give a fuck right now"
245 2012-05-13 01:24:04 <jgarzik> Diablo-D3: re chips, you can do a highly specialized chip, full ASIC, for around $1m if you partner with an existing house that makes a lot of ASICs anyway.  deals like that are common.
246 2012-05-13 01:24:26 <gmaxwell> freewil: thats not always what that means.
247 2012-05-13 01:24:44 <Diablo-D3> jgarzik: that takes time and money
248 2012-05-13 01:24:53 <Diablo-D3> Im simply not interested in it.
249 2012-05-13 01:24:57 <Diablo-D3> let somebody else with $1m do it
250 2012-05-13 01:25:02 <gmaxwell> freewil: sometimes it's just "this isn't a great thing to do, but there were complicated tradeoffs here and in this case it was the right decision. That doesn't mean you should copy this elsewhere, kid"
251 2012-05-13 01:25:14 <jgarzik> precisely :)
252 2012-05-13 01:25:41 <freewil> yeah ok justifications are always good i suppose
253 2012-05-13 01:25:42 <jgarzik> code gets cut-n-pasted everywhere in open source
254 2012-05-13 01:25:49 <n0n00dz4u> he bought four I front him five more
255 2012-05-13 01:25:52 <Diablo-D3> NOW
256 2012-05-13 01:25:56 <n0n00dz4u> 9 piece straight 8balls
257 2012-05-13 01:26:00 <Diablo-D3> if someone did asic and it was cheaper mh/$ than the best option?
258 2012-05-13 01:26:04 <Diablo-D3> I'd buy them
259 2012-05-13 01:26:08 <jgarzik> in the kernel, we see bugs in one driver get copied into 5 other drivers
260 2012-05-13 01:26:11 <jgarzik> it's awful
261 2012-05-13 01:26:11 <n0n00dz4u> :D
262 2012-05-13 01:26:21 <jgarzik> you need a license for cut-n-paste coding
263 2012-05-13 01:26:24 <luke-jr> anyone know what lock is supposed to protect pindexBest?
264 2012-05-13 01:26:25 <Diablo-D3> jgarzik: I fucking hate people like that
265 2012-05-13 01:27:03 <jgarzik> luke-jr: honestly, I would benchmark the current solution before assuming that further patches were needed
266 2012-05-13 01:27:16 <luke-jr> jgarzik: I'm not trying to optimize anything O.o
267 2012-05-13 01:27:37 <luke-jr> oh, I see your line of thought
268 2012-05-13 01:27:43 <luke-jr> jgarzik: my aim right now is implementing longpolling
269 2012-05-13 01:28:15 <jgarzik> luke-jr: LP + current m-t design will leave a lot of idle threads sitting there
270 2012-05-13 01:28:18 <jgarzik> a -lot-
271 2012-05-13 01:28:23 <freewil> for rpc?
272 2012-05-13 01:28:27 <luke-jr> jgarzik: even if there's only one LP connection?
273 2012-05-13 01:28:49 <jgarzik> luke-jr: one idle thread per client
274 2012-05-13 01:29:00 <jgarzik> luke-jr: you don't want thousands of idle threads
275 2012-05-13 01:29:00 <luke-jr> I think that's OK in this case. for now.
276 2012-05-13 01:29:19 <luke-jr> I can't imagine why there would be more than one getmemorypool longpoll client on bitcoind
277 2012-05-13 01:30:41 <luke-jr> I think in every case for bitcoind, there's only one connection
278 2012-05-13 01:43:19 <luke-jr> hmm
279 2012-05-13 01:43:27 <luke-jr> I don't see any mutex protecting pindexBest :o
280 2012-05-13 01:44:22 <BlueMatt> bitcoin doesnt protect setting pointers in a few cases
281 2012-05-13 01:45:31 <BlueMatt> uhhh...anyone else getting 500 when accessing their fork of bitcoin/bitcoin, or is it my fork?
282 2012-05-13 01:48:05 <BlueMatt> luke-jr: see-also: https://github.com/TheBlueMatt/bitcoin/commit/46154c5589f69f7a8f024c70dd6e75c5a7044b75
283 2012-05-13 01:48:10 <BlueMatt> (mostly the comments)
284 2012-05-13 01:48:46 <luke-jr> i c
285 2012-05-13 01:53:02 <luke-jr> BlueMatt: did you consider using a different (non-IPC non-recursive) mutex just for that?
286 2012-05-13 01:53:17 <BlueMatt> I didnt spend any real time looking into it
287 2012-05-13 01:53:29 <BlueMatt> at the time there was a ton more to do, and I just saw that old bitcoin did it, so I left it
288 2012-05-13 01:55:15 <jgarzik> sipa: any pref for CAddrDB location?  (a) back in db.*, (b) a tiny addrdb.*, (c) inside addrman.*, (d) other?
289 2012-05-13 01:59:30 <luke-jr> why are we using IPC mutexes anyhow?
290 2012-05-13 02:04:48 <n0n00dz4u> listening to see you in hell 2002 by suicide commando on grooveshark
291 2012-05-13 02:04:53 <n0n00dz4u> go diablo!
292 2012-05-13 02:04:54 <n0n00dz4u> go diablo!
293 2012-05-13 02:04:59 <n0n00dz4u> its your birthday :)
294 2012-05-13 02:39:48 <jgarzik> ewwww
295 2012-05-13 02:39:52 <jgarzik> GetDataDir is just ugly
296 2012-05-13 02:40:35 <jgarzik> instead, we should call GetDataDirAndDoThatStuffOnce, and then elsewhere in the code refer directly to pathDataDir variable.
297 2012-05-13 02:48:05 <luke-jr> #936 (getmemorypool updates for BIP 22) needs re-reviewing
298 2012-05-13 02:58:12 <jgarzik> datadir is certainly read-only for the duration of the program
299 2012-05-13 03:09:17 <luke-jr> jgarzik: but there's two of them
300 2012-05-13 03:10:38 <jgarzik> luke-jr: two of what?
301 2012-05-13 03:10:47 <luke-jr> datadirs
302 2012-05-13 03:10:55 <jgarzik> luke-jr: can you be more specific?
303 2012-05-13 03:11:11 <luke-jr> jgarzik: there's the dir with the config file, and the dir with the databases
304 2012-05-13 03:11:22 <luke-jr> jgarzik: for mainnet, they're identical, but for testnet different
305 2012-05-13 03:11:39 <luke-jr> at least, I think that's what the whole GetDataDir is about
306 2012-05-13 03:12:20 <jgarzik> luke-jr: the data directory value is static for the entire runtime of the program, whether it is testnet or not
307 2012-05-13 03:12:40 <luke-jr> sure, just noting there'd need to be two such variables
308 2012-05-13 03:12:44 <jgarzik> you must restart to switch data directories to/from testnet
309 2012-05-13 03:12:55 <jgarzik> luke-jr: no, there is no need for two variables
310 2012-05-13 03:13:06 <jgarzik> there is only one datadir for the entire runtime of the program
311 2012-05-13 03:13:35 <luke-jr> testnet uses the same directory as mainnet for some things (config file, at least), and another for others (databases, debug.log)
312 2012-05-13 03:22:20 <jgarzik> luke-jr: ah, good point
313 2012-05-13 03:23:54 <jgarzik> looks like just config file is called with GetDataDir(false)
314 2012-05-13 05:14:07 <gribble> New news from bitcoinrss: jgarzik opened pull request 1281 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1281>
315 2012-05-13 05:19:40 <gribble> New news from bitcoinrss: jgarzik opened pull request 1282 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1282>
316 2012-05-13 05:29:49 <gribble> New news from bitcoinrss: laanwj opened pull request 1283 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1283>
317 2012-05-13 06:53:23 <sipa> jgarzik: it can go it db.*; that's already doing several things at once
318 2012-05-13 06:55:09 <sipa> jgarzik: also, splitting off the block database to a separate file sounds dangerous, as you need atomicity on transactions spanning tx and blk indexes
319 2012-05-13 06:58:55 <sipa> luke-jr: a non-IPC mutex?
320 2012-05-13 07:14:51 <sirk390> hi, does anyone what use-case COINBASE_MATURITY prevents?
321 2012-05-13 07:15:55 <ThomasV> you the verb
322 2012-05-13 07:15:56 <sirk390> wouldn't it work exactly as well without this test?
323 2012-05-13 07:17:07 <sirk390> e.g I'm talking about the test in ConnectInputs "If prev is coinbase, check that it's matured"
324 2012-05-13 07:17:32 <freewil> a coinbase block is considered confirmed after 120 blocks
325 2012-05-13 07:17:39 <freewil> everything after is only 6
326 2012-05-13 07:17:44 <freewil> im guessing that is what thats about
327 2012-05-13 07:19:04 <sirk390> yes but why? the 6 is not enforced at all
328 2012-05-13 07:20:17 <sirk390> you can even spend a transaction that is  in the same block
329 2012-05-13 07:20:53 <sipa> sirk390: if a coinbase transaction is reorged, it is lost
330 2012-05-13 07:21:09 <sipa> contrary to another transaction which is at least attempted to move to the new chain
331 2012-05-13 07:21:54 <sipa> that means that if you'd spend a coinbase output, after 6 confirmations, and it would still get reorganised... the transactions depending on them are lost as well
332 2012-05-13 07:22:21 <sipa> for that reason there is a very conservative requirement of 120 confirmations
333 2012-05-13 07:25:21 <sirk390> sipa: ok I see thanks
334 2012-05-13 07:47:51 <gribble> New news from bitcoinrss: laanwj opened pull request 1284 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1284>
335 2012-05-13 08:11:25 <Eliel> freewil: actually, the actual hard limit for a coinbase maturity is 100 blocks. However, there's an additional soft limit of 20 blocks on top of that.
336 2012-05-13 08:11:57 <freewil> what do you mean by soft limit
337 2012-05-13 08:12:33 <sipa> the protocol requires 100 confirmations
338 2012-05-13 08:12:39 <sipa> but the client enforces 120
339 2012-05-13 08:13:34 <Eliel> as in a block that spends a coinbase after 100 confirmations is accepted but the transaction doing the spend would not get relayed or included in blocks before 120.
340 2012-05-13 08:17:19 <jine> Crap, thanks for the reminder. I've forgotten to patch main.cpp when i upgrade to 0.6.2 a few days ago. (We're using blocks with 100 confirmations, not 120)
341 2012-05-13 08:18:21 <jine> Had to check when you mentioned it... :)
342 2012-05-13 08:40:00 <gribble> New news from bitcoinrss: sipa opened pull request 1285 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1285>
343 2012-05-13 08:40:31 <sipa> but the client enforces 120
344 2012-05-13 10:33:32 <gribble> New news from bitcoinrss: sipa opened pull request 1286 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1286>
345 2012-05-13 11:10:06 <gribble> New news from bitcoinrss: sipa opened pull request 1287 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1287>
346 2012-05-13 11:45:40 <gribble> New news from bitcoinrss: Diapolo opened pull request 1288 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1288>
347 2012-05-13 12:11:20 <gribble> New news from bitcoinrss: Diapolo opened pull request 1289 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1289>
348 2012-05-13 12:27:27 <kinlo> is there a way to go to a public address from a private address using the official client?  Or to find the public address of an address in your wallet?
349 2012-05-13 12:34:21 <freewil> kinlo,
350 2012-05-13 12:34:35 <freewil> there is the rpc call dumpprivkey <bitcoinaddress>
351 2012-05-13 12:34:43 <gmaxwell> kinlo: Not that I can think of, and "public key" not "public address".
352 2012-05-13 12:35:15 <freewil> i guess what i said is opposite of what you want to do
353 2012-05-13 12:35:31 <kinlo> gmaxwell: right, sorry for the misuse of words, but indeed, I can't find it either so I assumed I was overlooking something
354 2012-05-13 12:36:25 <gmaxwell> kinlo: I can't think of any reason we'd have that.
355 2012-05-13 12:36:28 <sipa> kinlo: validateaddress will tell you the public key of an address, if known
356 2012-05-13 12:36:48 <gmaxwell> Ha.
357 2012-05-13 12:37:23 <kinlo> gmaxwell: I'm trying to use addmultisigaddress but how can I do that when I don't want both keys that are to be seperated are required to be in the same wallet
358 2012-05-13 12:37:42 <kinlo> gmaxwell: but I read in the source addmultisigaddress also seems to accept public keys
359 2012-05-13 12:38:42 <sipa> yes, use validateaddress to find all public keys, gather them, and turn them into a multisigaddress
360 2012-05-13 12:39:07 <kinlo> sipa: just a bit stupid that that functionality is hidden
361 2012-05-13 12:39:21 <kinlo> a quick how-to for the multisig addresses might be in place
362 2012-05-13 12:40:28 <sipa> meh; they're not very useful right now anyway, as spending them requires having all keys in the same wallet
363 2012-05-13 12:42:29 <kinlo> oh
364 2012-05-13 12:42:40 <kinlo> that was definatly going to be one of the followup questions :p
365 2012-05-13 12:43:35 <luke-jr> sipa: boost::mutex instead of boost::interprocess::*
366 2012-05-13 12:43:46 <gmaxwell> Well not 'all' you have to satisify the rules you can spend a 1 of 2 for example.
367 2012-05-13 12:44:16 <kinlo> well
368 2012-05-13 12:44:17 <luke-jr> sirk390: it'd be possible to treat coinbase spends as coinbases, but enforcing confirmations is simpler
369 2012-05-13 12:44:37 <kinlo> the whole idea is that you can have several keys working together *not* being in the same wallet
370 2012-05-13 12:44:49 <kinlo> so for now securing a high-value wallet is not yet an option? :)
371 2012-05-13 12:45:42 <luke-jr> sipa: boost::interprocess::* are I presume fork-safe
372 2012-05-13 12:45:49 <luke-jr> sipa: but we shouldn't need that
373 2012-05-13 12:47:57 <sipa> luke-jr: i think boost::mutex is just oldr
374 2012-05-13 12:48:35 <luke-jr> sipa: well, boost::interprocess::* apparently is incompatible with boost::condition_variable :p
375 2012-05-13 12:49:39 <sipa> you may be right
376 2012-05-13 12:50:46 <gmaxwell> kinlo: No, it's not. And it wouldn't be if we had the rest of the functionality simply because BIP16 deployment isn't ~100% yet.
377 2012-05-13 12:51:50 <luke-jr> " Boost.Threads implements these synchronization utilities between threads inside the same process. Boost.Interprocess implements similar mechanisms to synchronize threads from different processes."
378 2012-05-13 12:51:54 <luke-jr> http://www.boost.org/doc/libs/1_47_0/doc/html/interprocess/synchronization_mechanisms.html
379 2012-05-13 12:52:35 <sipa> luke-jr: i suppose in practice, many of them are simply implemented by pthreads
380 2012-05-13 12:52:49 <sipa> but it's a good point: why are we using interprocess syncing in that case?
381 2012-05-13 12:52:59 <luke-jr> "In boost 1.39.0, there is only specific support for pthreads. On all other platforms, it becomes a busy-loop with a yield call in the middle (essentially the same system that you describe)."
382 2012-05-13 12:53:18 <luke-jr> "What this means is that a contended boost::interprocess::mutex on windows is VERY expensive - although the uncontended case is almost free."
383 2012-05-13 12:54:00 <sipa> interesting
384 2012-05-13 12:54:11 <luke-jr> boost::threads has actual win32 implementation
385 2012-05-13 12:54:23 <luke-jr> this might address that other bug you guys were fighting
386 2012-05-13 12:54:34 <sipa> although that bug was only on BSD/OSX
387 2012-05-13 12:55:48 <sipa> but it does have a condition variable...
388 2012-05-13 12:56:26 <luke-jr> I use the conditional variable to implement BIP 22 longpolling
389 2012-05-13 12:56:58 <sipa> although it's trivial to implement a semaphore using condition variables
390 2012-05-13 12:57:34 <sipa> (but the win32 boost::threads implementation uses low-level semaphores to implement condition variables)
391 2012-05-13 12:57:40 <luke-jr> XD
392 2012-05-13 12:58:22 <luke-jr> http://boost.cppll.jp/BDTJ_1_29/libs/thread/doc/faq.html#question10
393 2012-05-13 13:00:52 <sipa> so: i want better than only mutexes; i use interprocess's condition variables; this spinlocks on OSX; it rewrite everything using semaphores; this spinlocks on OSX; i write a polling implementation of semaphores for OSX
394 2012-05-13 13:01:15 <sipa> and now it seems like there's a chance we could have just used boost::threads::condition_variable all along
395 2012-05-13 13:01:28 <sipa> (though the code in general has improved by the rewrites)
396 2012-05-13 13:02:32 <luke-jr> XD
397 2012-05-13 13:02:46 <luke-jr> sorry, I haven't been paying attention to the whole OSX debacle
398 2012-05-13 13:03:25 <sipa> seems like i should write a CSemaphore that uses boost::threads::condition_variable, and try that on OSX
399 2012-05-13 13:46:22 <jgarzik> sipa: as a code comment indicates, BDB can do cross-file transactions
400 2012-05-13 13:46:41 <jgarzik> sipa: precisely so that people don't do stupid stuff like Satoshi did, and glom everything into a single key/value db
401 2012-05-13 13:46:49 <sipa> jgarzik: i know it can; but you'll need to implement it :)
402 2012-05-13 13:46:59 <jgarzik> sipa: yes
403 2012-05-13 13:47:06 <jgarzik> sipa: hence the code comment :)
404 2012-05-13 13:48:18 <sipa> still, i think it's ugly to split that data into two files: together they define a consistent view on the current blockchain state
405 2012-05-13 13:50:38 <gmaxwell> bleh, the Mac OS X link on the site doesn't go to bitcoin-0.6.2.2 as the macros are all setup to make all the versions the same. Should I do a one off change to the macros with the risk that we'll forget to change it back with the next version?
406 2012-05-13 13:50:59 <gmaxwell> sipa: then the blockchain split from the index is ugly too.
407 2012-05-13 13:51:30 <gmaxwell> sipa: what if we had a $DATADIR/blockchain/  which had the db enviroment for all blockchain stuff as well as the chain files?
408 2012-05-13 13:52:27 <sipa> gmaxwell: good point
409 2012-05-13 13:54:18 <sipa> combine that with moving to ipaddr.dat, and moving away from bdb wallets :)
410 2012-05-13 13:55:48 <luke-jr> gmaxwell: make it so we can't forget to change it back :p
411 2012-05-13 13:55:49 <sipa> anyone with the ability to compile on OSX or BSD?
412 2012-05-13 13:56:53 <kinlo> compile bitcoin?
413 2012-05-13 13:57:16 <sipa> kinlo: you know, running a compiler and stuffz
414 2012-05-13 13:57:36 <kinlo> sipa: I have a compiler, I have osx, I just haven't yet successfully compiled it yet
415 2012-05-13 13:57:45 <sipa> never mind then
416 2012-05-13 13:59:39 <gmaxwell> Do we have a release procedure page someplace? (e.g. something that reminds us to nag bluematt to update the ppa and such?)
417 2012-05-13 14:00:03 <sipa> https://github.com/bitcoin/bitcoin/blob/master/doc/release-process.txt ?
418 2012-05-13 14:00:14 <kinlo> sipa: I'll try that release process file btw
419 2012-05-13 14:00:19 <kinlo> perhaps it will just work ;)
420 2012-05-13 14:00:40 <kinlo> I never tried before
421 2012-05-13 14:02:51 <n0n00dz4u> congrats to diablod3 for the release of DMC
422 2012-05-13 14:03:04 <n0n00dz4u> promotional vlog to be released today on youtube
423 2012-05-13 14:03:08 <n0n00dz4u> links in 5
424 2012-05-13 14:03:10 <n0n00dz4u> and uh
425 2012-05-13 14:03:14 <n0n00dz4u> fuck bitcointalk.org
426 2012-05-13 14:03:42 <jgarzik> sipa: I don't care about the files.  We can put whatever in whatever file you like
427 2012-05-13 14:04:15 <jgarzik> sipa: the important goal is having a single key/value database for block index entries (key=hash, value=data) and a single key/value database for transactions (key=hash, value=data)
428 2012-05-13 14:04:27 <jgarzik> sipa: andmaking those DB_HASH
429 2012-05-13 14:04:44 <jgarzik> sipa: (speaking of, I wonder if BDB doesn't like mixing DB_BTREE and DB_HASH in same file)
430 2012-05-13 14:05:36 <jgarzik> for bitcoin's two HUGE key/value lookups, we don't need the current crap (heavyweight keys, mixed namespace)
431 2012-05-13 14:05:55 <jgarzik> everything else (picking which files stores the data) is incidental
432 2012-05-13 14:06:50 <gmaxwell> jgarzik: watch out for https://bugzilla.redhat.com/show_bug.cgi?id=523698
433 2012-05-13 14:07:41 <sipa> jgarzik: i have no problem with separating the databases, putting them in separate files under blockchain/, or using DB_HASH
434 2012-05-13 14:07:49 <gmaxwell> oh nevermind they're just getting caught by the regular backward but not forward incompatiblity.
435 2012-05-13 14:08:08 <gmaxwell> (I thought they were saying that DB_HASH didn't have backward compatiblity either)
436 2012-05-13 14:08:09 <sipa> jgarzik: splitting the file may have cache-related performance improvements
437 2012-05-13 14:13:38 <kinlo> sipa: I'm usink fink instead of macports, seems to be a daunting task to get bitcoin compiled now
438 2012-05-13 14:14:02 <sipa> kinlo: unless you're interested in being able to do so yourself, i wouldn't botger
439 2012-05-13 14:14:28 <kinlo> sipa: macosx is just an os I use to ssh to real os'es in, so not really :p
440 2012-05-13 14:14:52 <kinlo> besides, I'm doing my exercises
441 2012-05-13 14:14:58 <kinlo> should concentrate
442 2012-05-13 14:15:02 <kinlo> and you should do the same :p
443 2012-05-13 14:21:29 <Graet> n0n00dz4u, whats your problem, spamming me in pm is really ...um pointless
444 2012-05-13 14:22:54 <Graet> [00:22:26] <n0n00dz4u> FUCK BITCOINTALK.ORG YOU DUMB CRACKER
445 2012-05-13 14:23:03 <Graet> feel free to ban the jkerk anyone with ops
446 2012-05-13 14:23:44 <Graet> he's got it on 1sec repeat
447 2012-05-13 14:25:31 <Graet> oh well ignored it
448 2012-05-13 14:28:35 <n0n00dz4u> anyway
449 2012-05-13 14:28:45 <n0n00dz4u> suddenly got quiet in here
450 2012-05-13 14:28:49 <n0n00dz4u> let me go check bitcointalk to see what the latest update to the bitcoinica incident is
451 2012-05-13 14:28:53 <n0n00dz4u> all of you suck at computers
452 2012-05-13 14:28:55 <n0n00dz4u> just saying
453 2012-05-13 14:31:08 <Someguy123> going to go mad, anyone able to help me with this? http://img1.uploadscreenshot.com/images/orig/5/13303193247-orig.png
454 2012-05-13 14:31:13 <Someguy123> I did a -rescan already, didn't fix it
455 2012-05-13 14:31:24 <Someguy123> I've had this TX for over an hour, and it's refusing to broadcast to any nodes
456 2012-05-13 14:31:55 <sipa> what is that?
457 2012-05-13 14:32:47 <Someguy123> sipa, it's my webUI for bitcoin,
458 2012-05-13 14:32:52 <Someguy123> but the point is, I have a TX that's stuck
459 2012-05-13 14:32:56 <Someguy123> refusing to broadcast to any nodes
460 2012-05-13 14:33:07 <Someguy123> and a -rescan didn't fix it
461 2012-05-13 14:33:29 <sipa> rescanning won't help
462 2012-05-13 14:34:17 <sipa> nornmally, it should always rebroadcast every 15-30 minutes
463 2012-05-13 14:34:27 <sipa> until it is confirmed
464 2012-05-13 14:35:35 <n0n00dz4u> I don't what you're talking about
465 2012-05-13 14:35:37 <n0n00dz4u> I'm here to develop my business
466 2012-05-13 14:35:39 <n0n00dz4u> I don't even know who you are
467 2012-05-13 14:35:41 <n0n00dz4u> um
468 2012-05-13 14:35:43 <n0n00dz4u> not me.
469 2012-05-13 14:35:44 <luke-jr> Someguy123: why do you assume it isn't broadcast?
470 2012-05-13 14:35:45 <n0n00dz4u> yeah about that
471 2012-05-13 14:35:48 <gmaxwell> Someguy123: I'm assuming you modified your client to remove the fee rules? Sucks to be you.
472 2012-05-13 14:35:50 <n0n00dz4u> I haven't done anything wrong
473 2012-05-13 14:35:53 <n0n00dz4u> lmfao
474 2012-05-13 14:35:55 <Someguy123> gmaxwell, no, it was a 0.005 fee
475 2012-05-13 14:35:56 <Someguy123> =_=
476 2012-05-13 14:36:38 <Someguy123> gmaxwell, 1024
477 2012-05-13 14:36:39 <Someguy123> oops
478 2012-05-13 14:36:43 <Someguy123> http://img1.uploadscreenshot.com/images/orig/5/13302452747-orig.png
479 2012-05-13 14:36:46 <gmaxwell> Someguy123: then it will automatically rebroadcast.
480 2012-05-13 14:37:07 <sipa> and it already did broadcast
481 2012-05-13 14:37:11 <sipa> (to 12 nodes, even)
482 2012-05-13 14:37:23 <Someguy123> sipa, that was invalid it seems
483 2012-05-13 14:37:29 <Someguy123> after a -rescan it stuck at 0 unconfirmewd
484 2012-05-13 14:37:33 <Someguy123> and refused to broadcast anymore
485 2012-05-13 14:37:37 <sipa> rescan has nothing to do with this
486 2012-05-13 14:37:40 <sipa> stop using it
487 2012-05-13 14:37:46 <Someguy123> I used it once =_=
488 2012-05-13 14:37:50 <sipa> it won't matter
489 2012-05-13 14:38:02 <sipa> all it does is find missing wallet transactions
490 2012-05-13 14:38:20 <sipa> transactions that *are* in the block chain, but not in your wallet
491 2012-05-13 14:38:57 <sipa> why do you think it's invalid? if it was invalid back then, it still is
492 2012-05-13 14:40:14 <Someguy123> but still, sipa, what am I supposed to do now?
493 2012-05-13 14:40:22 <gmaxwell> Someguy123: just wait.
494 2012-05-13 14:40:28 <Someguy123> gmaxwell, it's been 3 hours
495 2012-05-13 14:40:33 <Someguy123> it still hasn't been sent.
496 2012-05-13 14:40:44 <gmaxwell> Someguy123: why do you keep saying it hasn't been sent?
497 2012-05-13 14:40:44 <Someguy123> doesn't show up in any blockexplorer or any other bitcoind
498 2012-05-13 14:40:51 <gmaxwell> it hasn't been mined yet.
499 2012-05-13 14:40:57 <luke-jr> gmaxwell: blockchain.info doesn't see it either: 942f208d93caca3af9643e2d6a854792ddd38bf0e55bff87b257756e2b92cf36
500 2012-05-13 14:41:08 <Someguy123> luke-jr, yeah I noticed that
501 2012-05-13 14:41:19 <sipa> any chance the tx is a double spend?
502 2012-05-13 14:41:33 <Someguy123> sipa, no idea
503 2012-05-13 14:41:35 <Someguy123> shouldn't be...
504 2012-05-13 14:41:36 <gmaxwell> No mention of it in my logs either. Not even as a reject.
505 2012-05-13 14:41:44 <gmaxwell> Someguy123: are you running multiple bitcoinds with the same wallet?
506 2012-05-13 14:41:51 <Someguy123> gmaxwell, sort-of
507 2012-05-13 14:41:57 <Someguy123> I have some of my keys on blockchain.info
508 2012-05-13 14:42:01 <gmaxwell> ...
509 2012-05-13 14:42:02 <Someguy123> but that shouldn't matter
510 2012-05-13 14:42:10 <sipa> wallets are not meant to be shared
511 2012-05-13 14:42:17 <Someguy123> whenever I recieve or send from blockchain.info, it always ends in my main bitcoin
512 2012-05-13 14:42:41 <sipa> if one node tries to spend, and the other didn't hear about it yet, it may try to spend the came coin
513 2012-05-13 14:42:52 <gmaxwell> run the gettransaction 942f208d93caca3af9643e2d6a854792ddd38bf0e55bff87b257756e2b92cf36 rpc on that node and tell us the output.
514 2012-05-13 14:44:01 <Someguy123> gmaxwell, http://pasteitt.aa.am/showpaste.php?id=21&key=j8OFHpG
515 2012-05-13 14:45:24 <gmaxwell> oh darn, you're not running git.
516 2012-05-13 14:45:48 <Someguy123> yeah, current windows build...
517 2012-05-13 14:46:07 <luke-jr> Someguy123: getmemorypool
518 2012-05-13 14:46:16 <gmaxwell> (the git version of bitcoin decodes the transaction so we can see what input its spending)
519 2012-05-13 14:46:53 <luke-jr> gmaxwell: Eloipool's modules are useful for decoding binary txn data :D
520 2012-05-13 14:47:07 <luke-jr> Someguy123: pastebin getmemorypool :P
521 2012-05-13 14:47:14 <luke-jr> I gotta run, but that will probably include it
522 2012-05-13 14:47:16 <Someguy123> luke-jr, http://pasteitt.aa.am/showpaste.php?id=22&key=HXMZeei
523 2012-05-13 14:47:17 <luke-jr> &maybe
524 2012-05-13 14:47:35 <luke-jr> hmm, no easy way to tell which one it is
525 2012-05-13 14:47:55 <luke-jr> will analyze when I get back if still a problem
526 2012-05-13 14:47:59 <gmaxwell> luke-jr: actually, it won't include it if its a double spend.
527 2012-05-13 14:48:11 <luke-jr> gmaxwell: well, can at least confirm whether it is or not
528 2012-05-13 14:48:14 <n0n00dz4u> had to restart for updates
529 2012-05-13 14:48:25 <Raignemy> Helllo! Did I understand right, that addr.dat file content everithing that i need to backup about addresses? (my adresses and their keys, my "phone book"?)
530 2012-05-13 14:48:35 <gmaxwell> NO.
531 2012-05-13 14:48:37 <Someguy123> no.
532 2012-05-13 14:48:39 <Someguy123> completely wrong
533 2012-05-13 14:48:52 <Someguy123> you want to backup the wallet.dat (while bitcoin is NOT running if possible)
534 2012-05-13 14:49:02 <Someguy123> addr.dat is not needed and will be regenerated automatically
535 2012-05-13 14:49:07 <gmaxwell> Raignemy: addr.dat records the p2p peers.  What you want is in wallet.dat.
536 2012-05-13 14:49:26 <gmaxwell> Someguy123: we have a backup function in the client, file->backup ... you don't need to shut down to use that.
537 2012-05-13 14:49:35 <Someguy123> yeah, but I'm just checking
538 2012-05-13 14:49:40 <Raignemy> My private keys also stored in wallet.dat? so, i will able to receive funds on same addresses?
539 2012-05-13 14:50:17 <luke-jr> Raignemy: private keys are ONLY stored in wallet.dat
540 2012-05-13 14:50:21 <gmaxwell> Raignemy: your wallet.dat contains all the irreplacable data that distinguishes your bitcoin install from everyone elses...  address book, private keys, your transactions and lables.. etc.
541 2012-05-13 14:50:30 <luke-jr> Raignemy: and wallet.dat CANNOT be safely shared between different computers
542 2012-05-13 14:50:46 <Raignemy> luke-jr: gmaxwell: oh, thanks, I get it
543 2012-05-13 14:50:53 <Raignemy> so simple
544 2012-05-13 14:50:57 <luke-jr> guard wallet.dat with your life :P
545 2012-05-13 14:51:19 <luke-jr> Raignemy: anyhow, who told you or where did you read that addr.dat was of any importance?
546 2012-05-13 14:51:29 <Raignemy> and if I will loose all other files, I will able to get everithing back with wallet.dat?
547 2012-05-13 14:51:41 <luke-jr> right
548 2012-05-13 14:51:46 <luke-jr> assuming bitcoin wasn't running
549 2012-05-13 14:51:53 <luke-jr> if bitcoin is running, database/ dir is important
550 2012-05-13 14:52:06 <luke-jr> bbl
551 2012-05-13 14:52:14 <kinlo> yeah, to backup wallet.dat you need to use the backupwallet command, or first shut it down, then copy the file
552 2012-05-13 14:52:24 <Raignemy> okay, thanks =)
553 2012-05-13 14:57:43 <Someguy123> so gmaxwell any ideas on what I should do yet?
554 2012-05-13 14:57:45 <Someguy123> I've just left it for now
555 2012-05-13 15:01:20 <Raignemy> btw, I want to complain about mtgox. In their private API they banned all request methods exept POST. And I spent few days to figure that out.
556 2012-05-13 15:01:49 <kinlo> complain on their own channel, #mtgox
557 2012-05-13 15:02:07 <Raignemy> already =)
558 2012-05-13 15:18:41 <sipa> twmz: care to try building another solution for the 100% cpu issue?
559 2012-05-13 15:19:48 <twmz> sure
560 2012-05-13 15:21:17 <sipa> twmz: https://github.com/sipa/bitcoin/tree/threadsync
561 2012-05-13 15:25:08 <twmz> building...
562 2012-05-13 15:36:50 <twmz> bitcoin-qt crashes shortly after startup.  trying bitcoind to see if it gives a better error.  I don't know how to debug crashes on OS X.
563 2012-05-13 15:39:18 <twmz> hmm, bitcoind is not crashing.  waiting for 8 connections to see if the 100% cpu problem is present or fixed
564 2012-05-13 15:42:26 <twmz> sipa: cpu is fine (~2%) with bitcoind after 8 connections and a few minutes.  bitcoin-qt crashes.  Going to rebuild and try again.
565 2012-05-13 15:48:41 <twmz> getting the stacktrace from gdb now.  In the mean time, this is the apple diagnostics report.  not sure if there is anything useful in here.  Not sure if this was your change or some other 0.7.0 change that causes it either.  http://hastebin.com/xubekolufe.txt
566 2012-05-13 15:50:59 <sipa> twmz: i could backport the change to 0.6.2
567 2012-05-13 15:58:51 <twmz> I don't understand the crash.  It appears to be in the thread handling RPC requests.  I don't see any recent changes nor do I understand why it would be different between bitcoind and bitcoin-qt.  It's possible that my QT build environment is just messed up and my bitcoind environment is not.
568 2012-05-13 15:59:48 <sipa> it's very reassuring that it works on bitcoind
569 2012-05-13 16:01:04 <twmz> it's been running for 10 minutes as bitcoind and no problems
570 2012-05-13 16:02:43 <twmz> I don't really ever use the GUI.  I only was using it for the convenience of having a precompiled binary.  I'll run with this bitcoind for a while and see if anything crops up over night, etc.
571 2012-05-13 16:21:42 <BladeMcCool> i'd really love to beg for some help to install 0.6.2 on ubuntu 10.04LTS over top of an old 0.3 version. i have no idea why "apt-get install bitcoin" tells me i have the latest version. i don't know how to punch it in the face and make it upgrade to 0.6.2
572 2012-05-13 16:22:49 <gmaxwell> BladeMcCool: do you have the ppa setup?
573 2012-05-13 16:23:28 <sipa> BladeMcCool: try installing bitcoind or bitcoin-qt instead
574 2012-05-13 16:23:39 <sipa> i think they're separate packages now
575 2012-05-13 16:23:46 <BladeMcCool> i think i do. but maybe not correctly. i added some lines to sources.list that i found for lucid on https://launchpad.net/~bitcoin/+archive/bitcoin
576 2012-05-13 16:24:19 <BladeMcCool> i will try that sipa, ty for suggestion
577 2012-05-13 16:24:49 <gmaxwell> Maybe we should try contacting canonical and point out that they're distributing remotely exploitable finance software and that perhaps they should stop.
578 2012-05-13 16:25:45 <dusty_> Hi all, I need some help in a testnet issue. Again :-/
579 2012-05-13 16:26:54 <dusty_> I'm trying to understand why block #46728 (http://blockexplorer.com/testnet/block/0000000019eb783f972474131e29c652a365c2a9a0ce619c1fcbe4843e8b618c) has difficulty bits 1c2336a4
580 2012-05-13 16:27:22 <splatster> gmaxwell: They can't/won't/don't care.
581 2012-05-13 16:27:28 <dusty_> I've carefully tried to interpret code in main.cpp : GetNextWorkRequired(const CBlockIndex* pindexLast, const CBlock *pblock)
582 2012-05-13 16:28:04 <dusty_> and it seems to me that the answer should be 0x1d0fffff and not 0x1c2336a4
583 2012-05-13 16:28:14 <BladeMcCool> i have progress! (thanks!) .. now to get past error of "trying to overwrite '/usr/bin/bitcoind', which is also in package bitcoin 0:0.3.24-ppa2~lucid" .. i probably have to remove the old one first yes? whats the proper way to do that (assuming thats what i need to do) and not lose the blockchain?
584 2012-05-13 16:28:26 <dusty_> please, can someone give me a hint?
585 2012-05-13 16:28:31 <sipa> BladeMcCool: apt-get purge bitcoin
586 2012-05-13 16:29:47 <sipa> dusty_: it's only 14 minutes after the previous block
587 2012-05-13 16:29:57 <sipa> dusty_: so it uses the normal difficulty rule
588 2012-05-13 16:33:29 <BladeMcCool> sipa, gmaxwell: woot! thank you! ("version" : 60200, "blocks" : 180003)
589 2012-05-13 16:34:35 <gribble> 180003
590 2012-05-13 16:34:35 <sipa> ;;bc,blocks
591 2012-05-13 16:37:56 <dusty_> sipa: ouch... how have I missed that? thanks
592 2012-05-13 16:40:52 <n0n00dz4u> happy mothers day crackers
593 2012-05-13 16:41:28 <BladeMcCool> hey its deathbylollipop
594 2012-05-13 16:41:49 <BladeMcCool> happy mothers day to you too. i should call ma now. thanks for the reminder
595 2012-05-13 16:42:05 <gmaxwell> n0n00dz4u: Are you the same person who distributed a link to a malwared miner on the forum?
596 2012-05-13 16:42:38 <n0n00dz4u> gmaxwell if I cared about anything you had to say
597 2012-05-13 16:42:53 <n0n00dz4u> I would tell you I am a swastika carying member of a neo nazi organization
598 2012-05-13 16:43:12 <gmaxwell> BladeMcCool: thanks.
599 2012-05-13 16:43:13 <BladeMcCool> rofl
600 2012-05-13 16:43:17 <BladeMcCool> :D np
601 2012-05-13 16:43:44 <jgarzik> back
602 2012-05-13 16:44:10 <jgarzik> sipa: yes, splitting the file also permits careful tuning of page size (cluster size), which is constant within a single file (and thus, any databases within that file)
603 2012-05-13 16:44:40 <jgarzik> gmaxwell: noted.  I had heard about that DB_HASH issue, but that URL is useful
604 2012-05-13 16:49:06 <osmosis> it would be nice if the send and receive addresses created locally showed a date, and a date column to sort by.
605 2012-05-13 16:53:00 <Someguy123> so gmaxwell any ideas on what I should do yet?
606 2012-05-13 16:53:36 <Someguy123> still hasn't budged from 0 confirms
607 2012-05-13 16:53:39 <Someguy123> and it's been 6 hours
608 2012-05-13 16:54:27 <gmaxwell> Someguy123: seems likely that you've managed to doublespend an input through the use of sharing private keys between multiple clients.
609 2012-05-13 16:54:45 <gmaxwell> Someguy123: the easiest way to recover is to restore a wallet backup from shortly before that transaction.
610 2012-05-13 16:54:57 <Someguy123> hmm
611 2012-05-13 16:55:28 <Someguy123> gah, last backup was 5 days ago
612 2012-05-13 16:55:57 <gmaxwell> well, if you haven't done more than 100 transactions / getnewaddresses in that time.. then that would be okay.
613 2012-05-13 16:56:06 <sipa> gmaxwell: or generated 100 blocks ;)
614 2012-05-13 16:56:29 <Someguy123> gmaxwell, yeah I've created a few new addresses I think in that time
615 2012-05-13 16:56:52 <gmaxwell> created a few is okay, so long is a few << 100.
616 2012-05-13 16:57:35 <dusty_> sipa: I've a new one for you :) Block http://blockexplorer.com/testnet/block/00000000A29BEA9EE5B1716D7EBD6E33691D92C4C8BB518DB380D952840BFE54 have time before previous block, but difficulty resets to minimum
617 2012-05-13 16:57:47 <dusty_> any hint on why that behaviour?
618 2012-05-13 16:57:49 <sipa> dusty_: that's a bug :)
619 2012-05-13 16:57:53 <dusty_> ok :)
620 2012-05-13 16:57:54 <Someguy123> I backed up the screwedup wallet just to be sure anyway
621 2012-05-13 16:58:03 <sipa> dusty_: a known one, but fixing it would mean another testnet split
622 2012-05-13 16:58:05 <dusty_> an overflow I suppose
623 2012-05-13 16:58:08 <Someguy123> anyway... loading that old backup now
624 2012-05-13 16:59:00 <dusty_> sipa: in your opinion should I put an exception on my code for that block or simply re-implement the same bug?
625 2012-05-13 16:59:18 <dusty_> I suppose there will be some other blocks of this kind later in the chain
626 2012-05-13 16:59:51 <dusty_> so it's better to "implement" the bug, until upstream fixes it
627 2012-05-13 16:59:52 <Someguy123> okay, gmaxwell I ahve 0.48BTC in this wallet, is it safe to send it to my blockchain.info wallet this time?
628 2012-05-13 17:00:15 <dusty_> sipa: do you know if there is a plan to fix it before next testnet reset?
629 2012-05-13 17:00:19 <Someguy123> any way to prevent a doublespend before I do this?
630 2012-05-13 17:00:38 <gmaxwell> Someguy123: don't send until you're fully synced and have no unconfirmed transaction in any overlapping wallet.
631 2012-05-13 17:00:45 <Someguy123> fully synced and no unconfirmed here
632 2012-05-13 17:00:54 <Someguy123> with full connections
633 2012-05-13 17:01:04 <Someguy123> 8 connections and slowly growing
634 2012-05-13 17:01:08 <gmaxwell> (no unconfirmed on the other wallet either)
635 2012-05-13 17:01:17 <Someguy123> gmaxwell, other wallet has no unconfirmed
636 2012-05-13 17:01:31 <gmaxwell> okay you should be safe to send it then.
637 2012-05-13 17:01:34 <sipa> dusty_: is it worth it? testnet is quite broken anyway right now; a full reset will fix things, but that will be done simultaneously with fixing this bug
638 2012-05-13 17:01:36 <Joric> it's usually capped to 8 connections
639 2012-05-13 17:02:17 <sipa> Joric: 8 outgoing, yes
640 2012-05-13 17:02:23 <dusty_> sipa: ok, thanks
641 2012-05-13 17:02:53 <Someguy123> gmaxwell, https://blockchain.info/tx-index/5601121/1f0b294fbfa6217f8e88e135614ba1f66ac8fb5c362d7e2bb8ed6b8b143f698f
642 2012-05-13 17:02:56 <Someguy123> it worked atleast :)
643 2012-05-13 17:03:25 <Someguy123> okay great, 1.15 BTC on my blockchain wallet
644 2012-05-13 17:03:30 <Someguy123> sounds right now :), thanks gmaxwell
645 2012-05-13 17:03:38 <gmaxwell> Someguy123: no problem.
646 2012-05-13 17:04:34 <dusty_> sipa: do you know if transactions and block with version != 1 are accepted in prodnet, right now?
647 2012-05-13 17:04:48 <sipa> version numbers are ignored
648 2012-05-13 17:05:01 <dusty_> sipa: in both testnet and prodnet?
649 2012-05-13 17:07:01 <sipa> afaik, yes
650 2012-05-13 17:10:24 <kinlo> you're going to tell me I can just inject a version 5 block in the chain and no client is going to complain?
651 2012-05-13 17:10:35 <sipa> yes
652 2012-05-13 17:10:51 <kinlo> doesn't sound something that is good to do
653 2012-05-13 17:11:30 <sipa> no
654 2012-05-13 17:14:44 <dusty_> sipa: as always, thanks for being so helpful
655 2012-05-13 17:15:08 <dusty_> my next stop will be the hell to understand BIP16 :D
656 2012-05-13 17:15:35 <Eliel> dusty_: it's not so difficult :)
657 2012-05-13 17:16:01 <Eliel> at least, if you already understand the old-style transactions.
658 2012-05-13 17:16:21 <dusty_> Eliel: I'm not really expert there
659 2012-05-13 17:16:33 <dusty_> you mean the transactions to IP address?
660 2012-05-13 17:16:36 <sipa> i haven't implemented BIP16 myself, and i'm sure the devil is in the details :)
661 2012-05-13 17:16:47 <sipa> dusty_: no, regular pay-to-pubkey-hash transactions
662 2012-05-13 17:17:06 <dusty_> sipa: the devil is always in the details :)
663 2012-05-13 17:17:18 <t7> what happened with bitonica then?
664 2012-05-13 17:17:33 <dusty_> sipa: which language are you implementing in ?
665 2012-05-13 17:19:44 <dusty_> t7: https://bitcointalk.org/index.php?topic=81045.0;topicseen
666 2012-05-13 17:19:59 <osmosis> It would be nice if I could right click to lock an address, so that my sends wouldnt come from that address.
667 2012-05-13 17:21:11 <t7> wow ???500,000
668 2012-05-13 17:25:37 <sipa> dusty_: i work on the bitcoin client, so C++, but gavin implemented BIP16
669 2012-05-13 17:28:12 <gmaxwell> osmosis: what for?
670 2012-05-13 17:28:58 <gmaxwell> osmosis: (I admit I've had times I would have used that but I'm curious if you'd use it for the same stuff)
671 2012-05-13 17:29:08 <osmosis> gmaxwell, to make sure money isnt spend from a particular account. In this case, I want all my funds from mining to stay in that address.
672 2012-05-13 17:31:05 <gmaxwell> osmosis: Yes, but why?
673 2012-05-13 17:31:52 <osmosis> because im neurotic?
674 2012-05-13 17:35:50 <osmosis> For privacy reasons, Id like to hold onto my coins which came from more anonymous sources.
675 2012-05-13 17:36:58 <osmosis> i suppose this could also be accomplished with some interface that allowed you to select which accounts to pull from when doing a send.
676 2012-05-13 17:37:35 <osmosis> but that would be more complicated
677 2012-05-13 17:38:05 <sipa> i think you're confusing address and account
678 2012-05-13 17:48:18 <gmaxwell> osmosis: accounts are only bookkeeping there is no actual coins assigned to them.
679 2012-05-13 17:48:54 <osmosis> yes, addresses
680 2012-05-13 17:49:16 <osmosis> is that whole account thing going to be pulled out. Ive never heard of anyone actually using it.
681 2012-05-13 18:11:36 <gribble> New news from bitcoinrss: Diapolo opened issue 1290 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1290>
682 2012-05-13 18:31:16 <Diapolo> hello all
683 2012-05-13 18:34:59 <Diapolo> has anyone an IPv6 node running currently?
684 2012-05-13 18:37:06 <sipa> Diapolo: not right now
685 2012-05-13 18:37:33 <Diapolo> sipa: well it seems things are looking better now
686 2012-05-13 18:41:59 <Diapolo> sipa: but I get 2x Warning:dereferencing pointer 'sockaddr' does break strict-aliasing rules [-Wstrict-aliasing] 1 in netbase.cpp line 340 and the other in net.cpp line 1699
687 2012-05-13 18:42:44 <jgarzik> sipa gmaxwell: tempting to make block index a flat file.  all we do is (a) load 100% of data into mapBlockIndex at startup, and (b) periodically write an index entry.  never use erase.
688 2012-05-13 18:43:02 <jgarzik> that model better fits an append log than a key/value db
689 2012-05-13 18:43:14 <jgarzik> however...  it does use a lot of RAM to keep all those block headers
690 2012-05-13 18:43:55 <jgarzik> I lean towards keeping key/value index for hash->blockindex, but storing it only on disk, eliminating mapBlockIndex
691 2012-05-13 18:44:14 <jgarzik> that way, older blocks do not waste RAM
692 2012-05-13 18:44:55 <jgarzik> already have kernel and bdb cache backing us
693 2012-05-13 18:46:22 <jgarzik> eliminating mapBlockIndex would also eliminate a long startup procedure
694 2012-05-13 18:49:08 <phantomcircuit> jgarzik, optimally the block index would be a mmap'd region
695 2012-05-13 18:49:16 <phantomcircuit> but that turns into a horrible hack very quickly :)
696 2012-05-13 18:51:41 <sipa> jgarzik: the block index is already loaded into memory entirely
697 2012-05-13 18:52:22 <jgarzik> sipa: yes I know... that's what I was just talking about
698 2012-05-13 18:52:46 <sipa> i've already considered dumping a serialized vector of CDiskBlockIndexes into bdb instead of 180000 key/values
699 2012-05-13 18:53:03 <jgarzik> <jgarzik> "all we do is" ... "load 100% of data into mapBlockIndex at startup"
700 2012-05-13 18:53:19 <jgarzik> <jgarzik> "tempting to make block index a flat file"
701 2012-05-13 18:53:30 <jgarzik> however
702 2012-05-13 18:53:50 <jgarzik> mapBlockIndex pointlessly wastes RAM, because most older blocks are never accessed
703 2012-05-13 18:54:07 <jgarzik> we keep almost 200,000 block headers in RAM, and access 10% of that
704 2012-05-13 18:54:36 <jgarzik> thus
705 2012-05-13 18:54:59 <jgarzik> _deleting_ mapBlockIndex would reclaim that RAM
706 2012-05-13 18:55:30 <sipa> you need to make sure its transactionality follows that of the CTxIndex bdb transactions
707 2012-05-13 18:55:43 <sipa> which is trickier if you have a flat file
708 2012-05-13 18:56:35 <sipa> (the block connection system relies on being able to abort)
709 2012-05-13 18:56:39 <jgarzik> sipa: true, however, fundamentally it is (a) append only and (b) only appended when we add a block to another non-bdb file, blk%d.dat
710 2012-05-13 18:57:02 <sipa> it's not append only
711 2012-05-13 18:57:15 <jgarzik> sipa: yes, it is.  we never erase, only add.
712 2012-05-13 18:57:29 <sipa> reorganisations change prevBlockHash
713 2012-05-13 18:57:42 <sipa> eh, hashNextBlock
714 2012-05-13 18:59:03 <jgarzik> sipa: ah, good point
715 2012-05-13 18:59:12 <jgarzik> it is updated
716 2012-05-13 18:59:29 <sipa> it may be possible to do without, actually
717 2012-05-13 18:59:31 <jgarzik> easy enough to make that append-only
718 2012-05-13 18:59:47 <sipa> by only.storing back pointers in the index file
719 2012-05-13 19:00:01 <sipa> and storing a bestblock pointer im bdb still
720 2012-05-13 19:00:04 <Diapolo> sipa: accepted connection 50.97.128.202:36647 Added time data, samples 2, offset +5742120752267649946 (+95702012537794165 minutes) should this occur on IPv6?
721 2012-05-13 19:00:08 <jgarzik> yeah
722 2012-05-13 19:00:13 <sipa> Diapolo: no
723 2012-05-13 19:00:29 <jgarzik> bbiab, gotta ride bicycle before rain comes
724 2012-05-13 19:05:17 <Diapolo> sipa: see http://pastebin.com/Fk3R3e3c
725 2012-05-13 19:06:54 <sipa> Diapolo: looks like expected
726 2012-05-13 19:07:06 <sipa> anything in particular i should notice?
727 2012-05-13 19:09:46 <Diapolo> sipa: it should be IPv6 only, why are the connections listing IPv4 addresses?
728 2012-05-13 19:10:20 <sipa> which version of the code is this?
729 2012-05-13 19:10:24 <Diapolo> -blocknet="IPv4" -discover=0 and -externalip="6to4_addr"
730 2012-05-13 19:10:29 <Diapolo> current master network-wise
731 2012-05-13 19:12:32 <sipa> oh, just noticed; blocknet doesn't prevent listening by default on such networks
732 2012-05-13 19:12:39 <sipa> i'll fix that
733 2012-05-13 19:13:25 <sipa> use -blocknet=ipv4 -bind=youripv6
734 2012-05-13 19:13:50 <Diapolo> with -externalip specified, too?
735 2012-05-13 19:14:20 <sipa> if you have discover on, it will add bound addresses automatically as local
736 2012-05-13 19:15:11 <Diapolo> I disabled discover to ensure no IPv4 add is used, but that's not needed with blocknet then?
737 2012-05-13 19:15:37 <sipa> no, discover will not addlocal ipv4 addresses if ipv4 is blocked
738 2012-05-13 19:17:27 <Diapolo> okay
739 2012-05-13 19:17:30 <sipa> hmm, would a -onlynet= be more useful than a -blocknet= ?
740 2012-05-13 19:18:43 <Diapolo> I use it to really be able to test IPv6 connectivity only, but perhaps that makes sense yes.
741 2012-05-13 19:19:18 <sipa> (for now, the only usable networks in master are ipv4 and ipv6, but soon tor may be added as well)
742 2012-05-13 19:20:11 <sipa> hmm, who was (free)bsd user here?
743 2012-05-13 19:20:15 <denisx> me
744 2012-05-13 19:20:40 <sipa> denisx: would you mind trying to compile my threadsync branch, and see if it exhibits 100% cpu after getting 8 connections?
745 2012-05-13 19:20:55 <denisx> sipa: sure
746 2012-05-13 19:21:29 <denisx> sipa: link?
747 2012-05-13 19:21:38 <sipa> https://github.com/sipa/bitcoin/commits/threadsync
748 2012-05-13 19:23:06 <denisx> sipa: is that another fix to the 100% than the MAC_OSX fix?
749 2012-05-13 19:23:51 <sipa> denisx: yes, you don't need to patch anything
750 2012-05-13 19:25:33 <denisx> my testmachine has the port open, should I close it?
751 2012-05-13 19:27:52 <sipa> which port?
752 2012-05-13 19:29:03 <denisx> sipa: the port where the connections come in for bittorrent
753 2012-05-13 19:29:46 <denisx> whatever, right now it run with the port open
754 2012-05-13 19:29:52 <denisx> we will see whap happens
755 2012-05-13 19:30:01 <sipa> doesn't matter
756 2012-05-13 19:30:04 <denisx> ok
757 2012-05-13 19:30:25 <denisx> but I get more than 8 connections! ;)
758 2012-05-13 19:31:30 <sipa> oh, run with -nolisten if you want to check that
759 2012-05-13 19:33:13 <denisx> doesnt look good, right now it runs with 115%
760 2012-05-13 19:33:20 <denisx> but it is still downloading some blocks
761 2012-05-13 19:33:43 <denisx> ok, now it runs at 100%, all blocks loaded
762 2012-05-13 19:33:54 <sipa> :(
763 2012-05-13 19:34:14 <denisx> sipa: but your macosx did run fine for me on freebsd
764 2012-05-13 19:34:26 <sipa> denisx: yes, that just avoided the issue using polling
765 2012-05-13 19:34:30 <sipa> not a real/nice solution
766 2012-05-13 19:35:01 <Diapolo> sipa: with blocknet IPv4 and discover=1 this is happening: GetMyExternalIP() received [88.XX.XX.XX] 88.XX.XX.XX:0 GetMyExternalIP() returned 88.XX.XX.XX AddLocal(88.XX.XX.XX:8333,5)
767 2012-05-13 19:36:05 <sipa> Diapolo: how old is the code you're running (i fixed that today)
768 2012-05-13 19:38:42 <Diapolo> it should be up to date I did a clean full compilation round about 1-2 hours ago and have all commits in
769 2012-05-13 19:38:57 <sipa> hmmm
770 2012-05-13 19:39:11 <Diapolo> if I set discover=0 it does not do this
771 2012-05-13 19:41:27 <Diapolo> sipa: to which commit are you referring to, I'll double check its in my build
772 2012-05-13 19:43:20 <sipa> Diapolo: 1653f97c8f
773 2012-05-13 19:43:29 <sipa> Do not consider blocked networks local
774 2012-05-13 19:44:15 <sipa> Diapolo: doesn't work...
775 2012-05-13 19:44:16 <sipa> indeed
776 2012-05-13 19:45:03 <Diapolo> sipa: you made me feel like a newb :) I'm happy I was right with my observation :D
777 2012-05-13 19:46:00 <Diapolo> I can confirm that commit is in my build.
778 2012-05-13 19:46:47 <Diablo-D3> Diapolo!
779 2012-05-13 19:46:51 <Diablo-D3> why have you not bought DMC shares!
780 2012-05-13 19:47:00 <Diapolo> sipa: And I get 2x Warning:dereferencing pointer 'sockaddr' does break strict-aliasing rules [-Wstrict-aliasing]. 1 in netbase.cpp line 340 and the other in net.cpp line 1699
781 2012-05-13 19:47:20 <Diapolo> DMC what shares? sorry I kinda seldom read through the forums these days
782 2012-05-13 19:47:28 <Diapolo> reading
783 2012-05-13 19:48:01 <Diablo-D3> https://bitcointalk.org/index.php?topic=77469.0
784 2012-05-13 19:48:37 <Diapolo> are none left?
785 2012-05-13 19:48:56 <sipa> Diapolo: my mistake
786 2012-05-13 19:48:57 <Diablo-D3> no, theres a few thousand left
787 2012-05-13 19:50:10 <Diapolo> Diablo-D3: sounds pretty intersting but I'm currently using https://mybitcointrade.com/en/?content=/sparbuch/sparbuch with most of my coins ...
788 2012-05-13 19:50:42 <Diablo-D3> what an ugly website
789 2012-05-13 19:50:46 <Diablo-D3> and I have no clue what it says
790 2012-05-13 19:50:47 <Diapolo> indeed
791 2012-05-13 19:51:24 <Diapolo> it gives for 30 Days 6.56 % (+ 0.15 %)
792 2012-05-13 19:52:10 <sipa> why is that site half-german?
793 2012-05-13 19:52:17 <Diapolo> I tried it and it worked 2 times ... well currently there is not much I can do with the few coins I own, so I chose this.
794 2012-05-13 19:52:48 <Diapolo> sipa: I guess becuase it's from a german? I even didn't want to advertise!
795 2012-05-13 19:53:39 <Diapolo> sipa: Ideas for the compilation-warnings?
796 2012-05-13 19:53:53 <Diablo-D3> Diapolo: well, I wish everyone would invest in DMC so I can finally save Bitcoin from the clutches of evil
797 2012-05-13 19:54:37 <Diapolo> Diablo-D3: that sounds a bit god-like, no? I will have to read your posting and I'll do this I promise.
798 2012-05-13 19:54:59 <sipa> Diapolo: ignore them for now, they're harmless
799 2012-05-13 19:55:03 <Diablo-D3> no, I just should have budgeted a superhero costume
800 2012-05-13 19:55:12 <Diapolo> :-P
801 2012-05-13 19:56:27 <Diapolo> sipa: at least with bind="local IPv6 addr" no IPv4 connections are coming :)
802 2012-05-13 19:58:13 <sipa> denisx: can you retry building threadsync?
803 2012-05-13 19:58:18 <denisx> sipa: sure
804 2012-05-13 20:00:33 <Diapolo> I'm off, bye.
805 2012-05-13 20:04:16 <denisx> sipa: what did you change?
806 2012-05-13 20:05:03 <denisx> ok, got it
807 2012-05-13 20:06:28 <sipa> denisx: if it still goes to 100%, can you try checking git master as well?
808 2012-05-13 20:06:44 <denisx> yes yes, one after another!
809 2012-05-13 20:06:46 <denisx> ;)
810 2012-05-13 20:07:32 <sipa> thanks!
811 2012-05-13 20:09:37 <denisx> 9 connections and it still looks good
812 2012-05-13 20:11:09 <denisx> this one looks very promising, I will let it runs for some minutes
813 2012-05-13 20:11:59 <sipa> can you try with -nolisten?
814 2012-05-13 20:12:12 <sipa> (that way we're sure you really have 8 outbounds)
815 2012-05-13 20:17:27 <denisx> 8 connections and it is running fine
816 2012-05-13 20:17:54 <sipa> 8 outgoing ones?
817 2012-05-13 20:18:00 <denisx> yep
818 2012-05-13 20:18:05 <sipa> ok, great
819 2012-05-13 20:18:07 <sipa> thanks
820 2012-05-13 20:18:40 <denisx> what is the fix? is it nice or a hack?
821 2012-05-13 20:18:55 <sipa> it's nice :)
822 2012-05-13 20:19:31 <sipa> using boost::condition_variable instead of boost::interprocess::sync::interprocess_semaphore
823 2012-05-13 20:19:56 <sipa> the interprocess locking mechanisms are sometimes implemented as spinlocks, it seems
824 2012-05-13 20:20:05 <denisx> this would work under osx too I assume?
825 2012-05-13 20:20:21 <sipa> yes, someone already tested it
826 2012-05-13 20:20:45 <sipa> but i'm surprised you still got 100% a few minutes ago
827 2012-05-13 20:20:57 <sipa> i don't think much changed in between that could have affected that
828 2012-05-13 20:21:22 <denisx> hmm, maybe I messed something up
829 2012-05-13 20:21:29 <denisx> git is still a beast for me
830 2012-05-13 20:22:41 <sipa> do this: git remote add sipa https://sipa@github.com/sipa/bitcoin.git; git fetch sipa; git checkout threadsync; git reset --hard sipa/threadsync
831 2012-05-13 20:23:04 <sipa> eh
832 2012-05-13 20:23:18 <sipa> drop the sipa@ there
833 2012-05-13 20:24:22 <Dagger2> any chance we could change CNetAddr::ToStringIP() to use getnameinfo(), so it prints canonical addresses (i.e. :: and 2a02:348:5e:5a29::1 rather than 0:0:0:0:0:0:0:0 and 2a02:348:5e:5a29:0:0:0:1)?
834 2012-05-13 20:25:07 <sipa> Dagger2: sure; i originally used the naive way to avoid compatibility issues
835 2012-05-13 20:25:08 <denisx> sipa: when you said I should retry building threadsync, did you change anything?
836 2012-05-13 20:25:17 <sipa> denisx: something changed yes
837 2012-05-13 20:25:37 <sipa> (i rebased against new master, which had some extra bugfixes, but those should have been unrelated)
838 2012-05-13 20:29:27 <Dagger2> sipa: you mean failing if getnameinfo() is unavailable? I assume getaddrinfo and getnameinfo come as a set everywhere, and netbase.cpp already uses gai(), so I suspect it would already fail to compile in any situation where the build enviroment doesn't have gni()