1 2014-09-09 00:00:39 <gmaxwell> I'd be half tempted to leave it defaced, good way to finalize the move off of it. :P
  2 2014-09-09 00:00:55 <Luke-Jr> lol
  3 2014-09-09 00:01:12 <Anduck> then it may also be good to move the mailing lists elsewhere
  4 2014-09-09 00:01:21 <gmaxwell> sure sure.
  5 2014-09-09 00:01:30 <Anduck> what are the main channels for communicating about bitcoin any way?
  6 2014-09-09 00:01:38 <Luke-Jr> IRC mainly IMO
  7 2014-09-09 00:01:40 <sirius_> looks like it's fixed now?
  8 2014-09-09 00:01:51 <Luke-Jr> MLs worked for a while, but too many n00bs increasing SNR these days
  9 2014-09-09 00:01:57 <Luke-Jr> sirius_: "fixed"?
 10 2014-09-09 00:02:15 <Anduck> yeah sf bitcoin page looks fixed
 11 2014-09-09 00:02:35 <gmaxwell> I got more mail from him...
 12 2014-09-09 00:03:00 <Anduck> Luke-Jr: everyone was a noob
 13 2014-09-09 00:04:36 <Anduck> what was in the mail?
 14 2014-09-09 00:05:46 <gmaxwell> Anduck:  "Take me to dinner before you fuck me."
 15 2014-09-09 00:06:20 <sipa> seems reasonable
 16 2014-09-09 00:06:26 <petertodd> hot
 17 2014-09-09 00:06:46 <skinnkavaj> maybe the hacker got satoshis alarm key
 18 2014-09-09 00:07:00 <Luke-Jr> does that mean it's a female?
 19 2014-09-09 00:07:06 <Anduck> nah, i guess sn kept/keeps all keys offline anyway
 20 2014-09-09 00:07:25 <Luke-Jr> skinnkavaj: doubtful
 21 2014-09-09 00:07:32 <gmaxwell> skinnkavaj: it seems like it was a simple recovery style compromise, no biggie.
 22 2014-09-09 00:07:34 <Luke-Jr> unless Satoshi emailed it to Gavin…
 23 2014-09-09 00:07:46 <Anduck> if he did, he probs did it gpg-encrypted =P
 24 2014-09-09 00:07:50 <petertodd> Anduck: don't assume that "brilliant cryptographer" == "takes good care of day to day details"
 25 2014-09-09 00:07:53 <gmaxwell> Luke-Jr: unlikely that he got past emails.
 26 2014-09-09 00:08:05 <Anduck> petertodd: well, he did take good care of his anonymity
 27 2014-09-09 00:08:14 <petertodd> Anduck: or that "satoshi" == "brilliant cryptographer"
 28 2014-09-09 00:08:34 <theymos> gmaxwell: He got some. The email sent to me contained a quote of an email I sent to satoshin@gmx.com in March 2014.
 29 2014-09-09 00:08:37 <petertodd> Anduck: that's really not that hard you know...
 30 2014-09-09 00:08:42 <Anduck> well, satoshi must be some kind of brillian cryptographer to make bitcoin work
 31 2014-09-09 00:08:51 <Luke-Jr> theymos: satoshi was corresponding in March 2014? O.o
 32 2014-09-09 00:08:54 <Anduck> petertodd: sure
 33 2014-09-09 00:09:06 <Luke-Jr> Anduck: there is zero new cryptography in Bitcoin
 34 2014-09-09 00:09:12 <petertodd> theymos: evidence the account compromise happened some time ago
 35 2014-09-09 00:09:32 <Anduck> Luke-Jr: i know..
 36 2014-09-09 00:10:30 <theymos> Luke-Jr: I sent him an email then. I was telling him to contact me with a PGP signature if he wanted his forum account back (which I locked long ago). This was right after he posted to that p2pfoundation site.
 37 2014-09-09 00:10:44 <Anduck> theymos: the "i am not dorian" message at p2pfoundation was sent at march 7 -most likely the account was already compromised back then
 38 2014-09-09 00:10:45 <Luke-Jr> ah
 39 2014-09-09 00:10:58 <theymos> Anduck: Yes, seems likely.
 40 2014-09-09 00:11:02 <petertodd> theymos: plausible scenario is definitely someone had the idea to compromise the account at that time
 41 2014-09-09 00:11:29 <gmaxwell>  perhaps they finally decided they couldn't do anything that intersting with it.
 42 2014-09-09 00:11:39 <gmaxwell> they could have at least been funny. :(
 43 2014-09-09 00:11:55 <petertodd> gmaxwell: well school starts again for many right about now...
 44 2014-09-09 00:12:11 <Anduck> well, could've done some kind of "i destroyed all my bitcoins" or "i will dump 1M bitcoins"
 45 2014-09-09 00:12:35 <Luke-Jr> Anduck: oh my
 46 2014-09-09 00:12:43 <Luke-Jr> yes, that might have hurt the market a lot..
 47 2014-09-09 00:12:44 <btcdrak> I hate to say it, but gmx isnt exactly mega secure. I know a few people who've had their gmx accounts compromised.
 48 2014-09-09 00:12:52 <petertodd> Anduck: clever thing would be "I destroyed all my keys" - not specifically bitcoins
 49 2014-09-09 00:13:34 <sipa> btcdrak: and you couldn't have told that to satoshi 6 years ago? dude...
 50 2014-09-09 00:13:39 <gmaxwell> petertodd: better, ... post some crazy cryptogram.
 51 2014-09-09 00:14:08 <Luke-Jr> possible when the whole Dorin thing was going down, Satoshi leaked his gmx account just in case someone got closer to who he really was? :p
 52 2014-09-09 00:14:16 <petertodd> sipa: btcdrak == satoshi obviously, and found out the hard way
 53 2014-09-09 00:14:20 <gmaxwell> sipa: can you check if you have access on SF?
 54 2014-09-09 00:14:29 <btcdrak> sipa: I wish I was around 6 years ago -.-
 55 2014-09-09 00:14:49 <petertodd> Luke-Jr: If I were Satoshi I'd consider making some Satoshi document dropping hoax's myself to muddle the trail...
 56 2014-09-09 00:16:25 <sipa> gmaxwell: doesn't seem to be the case
 57 2014-09-09 00:18:04 <gmaxwell> yea, so I think the trouble maker undefaced it without fixing back up the access.
 58 2014-09-09 00:19:07 <sipa> https://sourceforge.net/projects/bitcoin/ -> "Brought to you by: s_nakamoto"
 59 2014-09-09 00:19:10 <dgenr8> Anduck: well sf would know when the email had become re-registerable.  iirc that msg was sent preceisely when dorian was out to lunch w/ reporter.  the intrigue.
 60 2014-09-09 00:20:48 <gmaxwell> please take the satoshi speculation elsewere. :)
 61 2014-09-09 00:21:14 <owowo> #satosh-speculation <--
 62 2014-09-09 00:21:19 <sipa> speculation belongs in #bitcoin-pricetalk :D
 63 2014-09-09 00:25:13 <dgenr8> gmx rather.  and of course they would know when it was actually re-registered
 64 2014-09-09 00:31:48 <gmaxwell> seems there are multiple people using that email account. :P
 65 2014-09-09 00:34:45 <gmaxwell> y'all may want to download all the content off sf.net in case it's deleted. (e.g. old downloads and svn repo)
 66 2014-09-09 00:37:32 <midnightmagic> gmaxwell: do you know of anyone who got copies of the mailing lists on it?
 67 2014-09-09 00:38:01 <petertodd> midnightmagic: mail-archive.org's one is pretty complete
 68 2014-09-09 00:38:12 <petertodd> midnightmagic: I've now got a copy of the -security archives :P
 69 2014-09-09 00:38:43 <midnightmagic> petertodd: ah, thanks.
 70 2014-09-09 00:39:35 <saivann> Well as for release notes, I should have imported all of them from sourceforge on bitcoin.org, only formatting has been slightly modified.
 71 2014-09-09 00:49:59 <midnightmagic> for my copy the last checkin of the sf bitcoin svn repo dated Date:   Mon Feb 7 14:46:53 2011 +0000, by gavin, "Added 320 fresh seednodes to replace the old ones."  ? look right?
 72 2014-09-09 00:54:28 <midnightmagic> obviously not.. lol
 73 2014-09-09 01:00:27 <petertodd> midnightmagic: I thought the SVN history had already been imported into git?
 74 2014-09-09 01:01:07 <midnightmagic> petertodd: yeah but having built .. well probably a half dozen scm-specific conversion utils, they are never accurate, esp. when the source is svn.
 75 2014-09-09 01:01:39 <petertodd> midnightmagic: though on git that commit is by Daniel Folkinshteyn and predates the gavin one, which makes me think we were running a conversion util git -> svn by that point
 76 2014-09-09 01:04:51 <midnightmagic> sounds right. Turns out sourceforge lets one rdump it, so I just got that.
 77 2014-09-09 01:05:29 <petertodd> interesting: just got another forwarded email from 2011 - indicates the account was hijacked rather than re-registered (or re-registered a long time ago)
 78 2014-09-09 01:15:33 <midnightmagic> oh sweet, a direct rsync to the svn root also works.
 79 2014-09-09 01:19:52 <midnightmagic> rsync://bitcoin.svn.sourceforge.net:/svn/bitcoin/*
 80 2014-09-09 02:24:43 <Gnomethrower> https://twitter.com/petertoddbtc/status/509141236100263936
 81 2014-09-09 02:24:52 <Gnomethrower> christ...
 82 2014-09-09 02:26:26 <gwern> Gnomethrower: it gets worse, the hacker didn't even redact his screenshots right
 83 2014-09-09 02:26:39 <Gnomethrower> gwern: oh, neat; example?
 84 2014-09-09 02:26:58 <gwern> you can manipulate the second screenshot to get the redacted name and address
 85 2014-09-09 02:27:09 <Gnomethrower> I haven't seen any screenshots yet
 86 2014-09-09 02:27:38 <gwern> oh, http://pastebin.com/7gbPi8Qr
 87 2014-09-09 02:29:43 <Gnomethrower> gwern: huh
 88 2014-09-09 02:31:18 <lewellyn> gwern: this is why i redact in a bitmap editor and draw a solid color box. :P
 89 2014-09-09 02:31:29 <Gnomethrower> how the hell do you unredact that
 90 2014-09-09 02:31:30 <gwern> lewellyn: as you should!
 91 2014-09-09 02:31:52 <Gnomethrower> also lewellyn: fuck yes, why does no-one do that?!
 92 2014-09-09 02:32:04 <lewellyn> gwern: i gave up years ago explaining to people what "photoshop is not a bitmap editor" means. :P
 93 2014-09-09 02:34:24 <Gnomethrower> gwern: any hints as to how to manipulate the second screenshot?
 94 2014-09-09 02:35:01 <gwern> Gnomethrower: that's no fun
 95 2014-09-09 02:35:33 <Gnomethrower> gwern: well, as far as I can tell Satoshi doesn't live in St Louis, Missouri ;)
 96 2014-09-09 02:35:44 <cystic> Anyone have a mirror to the images, they won't load for me.
 97 2014-09-09 02:35:47 <Gnomethrower> so I'm very curious as to what's underneath
 98 2014-09-09 02:35:49 <gwern> Gnomethrower: you know where satoshi lives? INCROYABLE
 99 2014-09-09 02:36:16 <Gnomethrower> seeing that is reassuring, because I'm fairly sure he's safe now
100 2014-09-09 02:36:21 <Gnomethrower> whereas before I was a little uncertain
101 2014-09-09 02:36:53 <Gnomethrower> gwern: I didn't say that ;)
102 2014-09-09 02:36:58 <Gnomethrower> but I know where he doesn't.
103 2014-09-09 02:37:19 <gwern> how's that?
104 2014-09-09 02:37:41 <Gnomethrower> gwern: that's no fun
105 2014-09-09 02:37:42 <Gnomethrower> ;)
106 2014-09-09 02:37:47 <Apocalyptic> gwern, we all know it's McCaleb anyway
107 2014-09-09 02:38:00 <gwern> mccaleb is too incompetent to be satoshi. vide the mtgox password hashing fiasco
108 2014-09-09 02:38:27 <Gnomethrower> http://imgur.com/a/PNwWj
109 2014-09-09 02:38:33 <Gnomethrower> ( @cystic)
110 2014-09-09 02:38:40 <gmaxwell> please take the satisho speculation stuff out of this channel.
111 2014-09-09 02:38:55 <Gnomethrower> gmaxwell, two things
112 2014-09-09 02:38:57 <Gnomethrower> 1) it's relevant
113 2014-09-09 02:39:00 <Gnomethrower> 2) where to?
114 2014-09-09 02:39:05 <jgarzik> #bitcoin
115 2014-09-09 02:39:23 <gmaxwell> you'll get kicked out of there. there was a satoshi speculation channel created previously.
116 2014-09-09 02:39:23 <Gnomethrower> fair
117 2014-09-09 02:39:41 <jgarzik> IMO it is fair #bitcoin material at the present time
118 2014-09-09 02:40:06 <Gnomethrower> jgarzik: that would make sense..
119 2014-09-09 02:40:45 <gmaxwell> it will just irritate people if it takes over the channel again-- for every person who thinks this is the most interesting thing ever there is another who thinks its profoundly boring and irrelevant, but fine.
120 2014-09-09 02:41:20 <Gnomethrower> uh, what the hell just happened
121 2014-09-09 02:41:23 <Gnomethrower> why did I get kickbanned
122 2014-09-09 02:41:38 <Apocalyptic> you posted an imgur link
123 2014-09-09 02:41:45 <gmaxwell> see the topic
124 2014-09-09 02:41:45 <Gnomethrower> that's not allowed?
125 2014-09-09 02:41:53 <Gnomethrower> ...oh
126 2014-09-09 02:41:55 <Gnomethrower> *facepalm*
127 2014-09-09 02:43:46 <gwern> gmaxwell: if nothing else, these little affairs illustrate interesting principles: in this case, that expiring email addresses are *serious* vulnerabilities, that redaction must be done with flat images and solid colors... hm what else
128 2014-09-09 02:48:30 <gmaxwell> gwern: I'd argue the vulnerabilty is email based account recovery.
129 2014-09-09 02:48:38 <gmaxwell> I no longer believe expiration was involved.
130 2014-09-09 02:49:03 <gwern> gmaxwell: either way there's a useful lesson here, I figure
131 2014-09-09 02:49:11 <gmaxwell> email based account recovery makes compromising a single thing, which might not be well protected, into a much broader compromise.
132 2014-09-09 02:49:47 <gmaxwell> sure, though I think it's one we already knew: e.g. we'd been good about removing old access from github. (even though it created some stupid drama)
133 2014-09-09 02:50:07 <gwern> you guys didn't have the ability to remove him from the SF project?
134 2014-09-09 02:50:45 <gmaxwell> I'm not sure we really knew he was still there, or at least I didn't... we don't really use the sf project for anything anymore... except a couple mailing lists.
135 2014-09-09 02:50:58 <gmaxwell> one problem with removing privs is remembering to do it everywhere. :(
136 2014-09-09 02:51:30 <phantomcircuit> gmaxwell, and this is why i now have a secret domain name
137 2014-09-09 02:51:43 <phantomcircuit> :/
138 2014-09-09 02:52:36 <gmaxwell> domain names a pretty easy to capture... :( and a lot of sites will manage to leak the email -- or at least the domain.
139 2014-09-09 02:52:42 <gmaxwell> gah, offtopic.
140 2014-09-09 03:05:35 <warren> I made a backport branch of 0.9.3 + watch-only
141 2014-09-09 03:05:41 <warren> anyone want to help test?
142 2014-09-09 03:10:43 <Luke-Jr> lame :P
143 2014-09-09 03:12:33 <gmaxwell> warren: really more testing of the release candates + recent fixes would be more productive.
144 2014-09-09 03:59:38 <dabura667> Hmm
145 2014-09-09 04:42:01 <warren> sigh.  something about gitian makes my system go haywire.  The CTRL button behaves like it is always pressed.
146 2014-09-09 05:04:26 <warren> spooky.  "unset DISPLAY" avoids the problem.
147 2014-09-09 05:16:28 <michagogo> warren: o_O
148 2014-09-09 05:17:36 <Luke-Jr> it's annoying how GNU Screen behaves differently inside consoles if you started X inside a screen
149 2014-09-09 05:17:55 <Luke-Jr> (it silently opens a new window in the session X is running in)
150 2014-09-09 05:18:02 <Luke-Jr> (workaround: unset STY)
151 2014-09-09 06:40:00 <aschildbach_> What's the issue with Sourceforge?
152 2014-09-09 06:40:20 <sipa> none anymore
153 2014-09-09 06:40:42 <aschildbach_> So what *was* the issue? (.:
154 2014-09-09 06:40:54 <sipa> satoshi's email account and sf.net account were hacked, and used to temporarily deface the project page
155 2014-09-09 06:41:20 <aschildbach_> Neat! Bitcoin is growing up (-:
156 2014-09-09 06:41:56 <sipa> all links from bitcoin.org to sf were removed
157 2014-09-09 06:42:34 <sipa> and through sf staff intervention, satoshi's account was deactivated
158 2014-09-09 06:43:43 <lewellyn> i guess this is proof that satoshi is gone
159 2014-09-09 06:53:13 <gmaxwell> we weren't using it for much of anything, links back to old announcements, and some mailing lists... etc. most everything else has been on bitcoin.org and github for some time.
160 2014-09-09 06:53:41 <gmaxwell> though you should all check your security against password recovery attacks; presumably this has given a bunch of people the idea that they should try this stuff more.
161 2014-09-09 06:56:12 <Luke-Jr> someone recently tried to social engineer their way into Eligius servers by impersonating me :p
162 2014-09-09 06:56:27 <warren> Luke-Jr: were you fooled?
163 2014-09-09 06:57:20 <gmaxwell> luke did not believe this other person was luke.
164 2014-09-09 06:57:24 <gmaxwell> warren: missed you around these parts.
165 2014-09-09 06:57:31 <warren> gmaxwell: yeah, been busy
166 2014-09-09 06:57:34 <Luke-Jr> warren: no, the datacenter knows better
167 2014-09-09 06:57:48 <gmaxwell> they also claimed to be "luke sash junior"
168 2014-09-09 06:58:02 <warren> gmaxwell: after I rebase litecoin + watchonly I'm going to fix your patch to get rid of the website IP lookups
169 2014-09-09 06:58:03 <Luke-Jr> :p
170 2014-09-09 06:58:20 <sipa> "hi. i am you, but came here in a flying delorean from the year 2044. i need access to your server."
171 2014-09-09 06:58:40 <Luke-Jr> XD
172 2014-09-09 06:58:44 <gmaxwell> I'm working full time on bitcoin now, I'd suggest leaving that to me. I rewrote the patch to make it easier to test. Likely would have started the tests on it today, except for the dramaz. :)
173 2014-09-09 06:58:45 <warren> the Delorean still looks futuristic today ...
174 2014-09-09 06:59:05 <Luke-Jr> gmaxwell: ah, you started today? nice
175 2014-09-09 06:59:20 <warren> gmaxwell: ah, you  understood what the failure was when I tested it?
176 2014-09-09 06:59:50 <gmaxwell> I think so. (well I think I did when I rewrote it, I did the rewrite a month or so ago)
177 2014-09-09 06:59:53 <Luke-Jr> I guess "stopped" may be a better verb there :p
178 2014-09-09 07:00:18 <phantomcircuit> <Luke-Jr> someone recently tried to social engineer their way into Eligius servers by impersonating me :p
179 2014-09-09 07:00:24 <phantomcircuit> someone tries that nearly once a week
180 2014-09-09 07:01:04 <Luke-Jr> phantomcircuit: well, they first need to identify what datacenter to contact in this case
181 2014-09-09 07:01:10 <gmaxwell> Luke-Jr: last day at Mozilla was the 5th. My productivity isn't super high yet because I basically have no place setup at home to work yet, and have been interleaving doing that while juggling other things. (uh, I have a lot of non-work 'debt' built up from these last couple years of being oversubscribed)
182 2014-09-09 07:01:36 <gmaxwell> (well I don't need to tell you that, you've been in my maze of cardboard boxes^w^w^w^whouse)
183 2014-09-09 07:01:38 <Luke-Jr> >_<
184 2014-09-09 07:01:48 <Luke-Jr> gmaxwell: surely it's improved since then :P
185 2014-09-09 07:02:05 <sipa> they've been replaced by plastic bixes
186 2014-09-09 07:02:09 <Luke-Jr> haha
187 2014-09-09 07:03:39 <gmaxwell> I have no more miners in the kitchen now!
188 2014-09-09 07:04:08 <Luke-Jr> gmaxwell: aww, how will you cook your food?
189 2014-09-09 07:04:18 <phantomcircuit> ha
190 2014-09-09 07:04:30 <warren> 
ACTION wonders if any non-data centers are mining anymore

191 2014-09-09 07:05:03 <Luke-Jr> warren: does my house and my parents' house count as a datacenter? <.<
192 2014-09-09 07:05:10 <gmaxwell> oh I still have miners, but they are only in one room, instead of three.
193 2014-09-09 07:05:32 <phantomcircuit> warren, the data centers are like
194 2014-09-09 07:05:35 <phantomcircuit> tier -1 now
195 2014-09-09 07:05:49 <phantomcircuit> you too can be the proud owner of a warehouse in a cold climate!
196 2014-09-09 07:05:54 <Luke-Jr> wish I could find someone who wanted to figure out pool heating with miners
197 2014-09-09 07:06:02 <Luke-Jr> got a bunch I have to retire soon otherwise :/
198 2014-09-09 07:06:06 <gmaxwell> all the real mining is on rafts out in lakes in the arctic circle or something. :)
199 2014-09-09 07:06:09 <warren> There any fiber links to Antarctica?
200 2014-09-09 07:06:20 <Luke-Jr> gmaxwell: stop melting ice! you're going to sink my house :<
201 2014-09-09 07:06:24 <warren> ... and a nuclear power plant ...
202 2014-09-09 07:06:29 <phantomcircuit> warren, sure but no electricity :P
203 2014-09-09 07:07:43 <phantomcircuit> cold climate, cheap power, sane governments, pick two
204 2014-09-09 07:08:15 <Luke-Jr> phantomcircuit: sane governments don't exist, so the choice is obvious
205 2014-09-09 07:08:29 <warren> doesn't Iceland qualify for all three?
206 2014-09-09 07:09:00 <justanotheruser> warren: their govt doesn't like bitcoin
207 2014-09-09 07:09:00 <phantomcircuit> warren, the last point is a maybe
208 2014-09-09 07:09:01 <justanotheruser> or any foreign currency for that matter
209 2014-09-09 07:09:22 <warren> they seemed to like lots of foreign currency until recently
210 2014-09-09 07:09:23 <phantomcircuit> they have capital controls in place and IBM specifically chose not to put a dc there because of their "liberal" democracy
211 2014-09-09 07:09:28 <phantomcircuit> (ie sticky fingered)
212 2014-09-09 07:10:15 <phantomcircuit> also afaict there is no more power available from the power companies
213 2014-09-09 07:10:27 <phantomcircuit> somebody *cough* bought it all up
214 2014-09-09 07:10:39 <sipa> there are only 300k icelandic people... i guess such a small number easily leads to a different style of government
215 2014-09-09 07:14:05 <michagogo> jgarzik: what's that hash(?) you tweeted?
216 2014-09-09 07:17:25 <michagogo> gmaxwell: what's 'debt'?
217 2014-09-09 07:18:15 <michagogo> To clarify: What's "'debt'"?
218 2014-09-09 07:19:04 <Luke-Jr> michagogo: he went to bed; I interpreted it as "code I said I'd help with, but never got time for"
219 2014-09-09 07:19:57 <michagogo> Ahhh
220 2014-09-09 07:20:28 <michagogo> Oh, I guess that makes sense
221 2014-09-09 07:21:14 <michagogo> Meh.com -- it's what woot.com used to be. Same person, I think.
222 2014-09-09 07:21:22 <michagogo> Er, wrong channel... Sorry.
223 2014-09-09 07:35:52 <gmaxwell> hm. hit a leveldb assert on shutdown in master - a few days + some patches bitcoind: db/version_set.cc:803: leveldb::VersionSet::~VersionSet(): Assertion `dummy_versions_.next_ == &dummy_versions_' failed.
224 2014-09-09 07:36:06 <gmaxwell> this seems to not be super informative.
225 2014-09-09 07:37:34 <warren> gmaxwell: yeah, happens to me too
226 2014-09-09 07:37:36 <warren> gmaxwell: with 0.9.3
227 2014-09-09 07:37:51 <wumpus> that error is reported with leveldb a lot
228 2014-09-09 07:37:55 <wumpus> not just in the context of bitcoin
229 2014-09-09 07:38:04 <sipa> strange, i have never seen it
230 2014-09-09 07:38:16 <wumpus> me neither, but it's been reported before
231 2014-09-09 07:39:27 <wumpus> hmm see https://code.google.com/p/leveldb/issues/detail?id=136
232 2014-09-09 07:39:51 <wumpus> even a link to a patch to leveldb
233 2014-09-09 07:40:23 <sipa> oh, we sometimes had iterators open at close, but that was fixed in master very recently
234 2014-09-09 07:40:46 <wumpus> it can *also* happen then, but it seems there is a reference counting issue too
235 2014-09-09 07:40:55 <sipa> ah
236 2014-09-09 07:43:08 <gmaxwell> sipa: any idea where the aformentioned fix in master was? I can see if I have it.
237 2014-09-09 07:43:21 <gmaxwell> likely unless it was merged like... this weekend.
238 2014-09-09 07:50:32 <sipa> gmaxwell: #4656
239 2014-09-09 07:53:44 <gmaxwell> sipa: yep, my node was running with that.
240 2014-09-09 07:53:52 <gmaxwell> ohoh. so one interesting point.
241 2014-09-09 07:54:01 <gmaxwell> I'd been running a loop of all the rpcs over and over again
242 2014-09-09 07:54:14 <gmaxwell> and we have a "verifychain" rpc... perhaps a leak there.
243 2014-09-09 07:55:00 <gmaxwell> (incidentally, under valgrind verifychain is slow enough that I lose all my peers :P )
244 2014-09-09 07:55:11 <sipa> gettxoutsetinfo iterates over the database directly
245 2014-09-09 07:55:18 <sipa> verifychain does not
246 2014-09-09 07:55:32 <gmaxwell> or there, indeed I called that too, though thats not a new rpc.
247 2014-09-09 07:56:02 <gmaxwell> in any case, back to bed for me.
248 2014-09-09 07:56:11 <sipa> seems early for you?
249 2014-09-09 07:56:19 <sipa> :p
250 2014-09-09 08:03:07 <wumpus> warren: do you have the versionset problem consistently? if so, can you try with https://github.com/google/leveldb/pull/3 applied to see if that fixes it?
251 2014-09-09 08:04:27 <warren> wumpus: not consistently, and I'm linking against fedora's system leveldb
252 2014-09-09 08:04:34 <warren> oh wait, I'm not
253 2014-09-09 08:04:48 <warren> i'll try it soon
254 2014-09-09 08:05:12 <wumpus> ehh.. don't link against system's leveldb!
255 2014-09-09 08:05:21 <wumpus> we include it for good reason
256 2014-09-09 08:20:15 <phantomcircuit> gmaxwell, any idea why Avalon is pre-calculating the first 3 stages of the expansion function? https://ehash.com/product/a3222-20pcs/
257 2014-09-09 08:31:17 <warren> Anyone familiar with the code that bans an IP address, how fine grained is it when banning ipv6?
258 2014-09-09 08:31:50 <sipa> just a single ip, afaik
259 2014-09-09 08:31:55 <sipa> even for ipv6
260 2014-09-09 08:32:37 <phantomcircuit> warren, that is an issue with the ipv6 stuff
261 2014-09-09 08:32:51 <phantomcircuit> it's basically impossible to tell the difference between someone with a /48 at home
262 2014-09-09 08:32:59 <phantomcircuit> and a university with a /48
263 2014-09-09 08:33:33 <timothy> well you can ban an IPv6 range :P
264 2014-09-09 08:33:42 <lewellyn> /48 is huge though. you may as well say /16 ipv4.
265 2014-09-09 08:33:52 <michagogo> If only there were some kind of standard for that
266 2014-09-09 08:33:56 <warren> how big a range is given to home users?
267 2014-09-09 08:34:04 <timothy> usually /64
268 2014-09-09 08:34:05 <lewellyn> /64 is not uncommon
269 2014-09-09 08:34:06 <michagogo> warren: it really varies by ISP
270 2014-09-09 08:34:11 <lewellyn> maybe like a /56.
271 2014-09-09 08:34:13 <timothy> it's the only range usable with radvd
272 2014-09-09 08:34:37 <lewellyn> but realistically no smaller than /64 as llap and friends won't work on smaller.
273 2014-09-09 08:34:38 <michagogo> ;;google what size ipv6 range do home users get
274 2014-09-09 08:34:39 <gribble> IPv6 - Wikipedia, the free encyclopedia: <http://en.wikipedia.org/wiki/IPv6>; IP address - Wikipedia, the free encyclopedia: <http://en.wikipedia.org/wiki/IP_address>; RFC 6177 - IPv6 Address Assignment to End Sites: <http://tools.ietf.org/html/rfc6177>
275 2014-09-09 08:34:56 <lewellyn> and a /64 is huuuuge anyhow.
276 2014-09-09 08:35:05 <warren> https://togami.com/~warren/temp/example-bitcoind-dos-mitigation-via-iptables.txt    <---- a ipv6 variation of this with a subnet and rate limiting could mitigate ipv6 connection abuse
277 2014-09-09 08:37:42 <timothy> lewellyn: /64 is needed to use router advertisement, and some OS does not support dhcpv6 yet :)
278 2014-09-09 08:38:10 <lewellyn> timothy: that's what i said. ;)
279 2014-09-09 08:39:04 <lewellyn> 
ACTION compares ipv6 to appletalk and ipx a lot, btw.

280 2014-09-09 08:39:23 <lewellyn> it's closer to those than to ipv4, in many respects.
281 2014-09-09 08:39:23 <timothy> welcome to 1970s
282 2014-09-09 08:44:10 <phantomcircuit> <michagogo> ;;google what size ipv6 range do home users get
283 2014-09-09 08:44:12 <phantomcircuit> depends
284 2014-09-09 08:44:25 <phantomcircuit> i can get a /48 from he
285 2014-09-09 08:44:35 <phantomcircuit> the problem is as timothy said
286 2014-09-09 08:44:50 <phantomcircuit> a /64 is effectively the smallest routable unit because of how radvd works
287 2014-09-09 08:44:54 <phantomcircuit> which is lol ridiculous
288 2014-09-09 08:45:28 <timothy> they use mac addresses translated in 64 bit (EUI-64), so you need 64 bit to map the mac address :P
289 2014-09-09 08:45:39 <timothy> it's quite silly instead
290 2014-09-09 08:45:44 <phantomcircuit> timothy, except a mac address is only 48 bits >.>
291 2014-09-09 08:46:00 <timothy> infact I sayd "translated in 64 bit" :P
292 2014-09-09 08:46:04 <timothy> said*
293 2014-09-09 08:46:07 <phantomcircuit> i know
294 2014-09-09 08:46:17 <timothy> i don't know the reason
295 2014-09-09 08:46:40 <phantomcircuit> timothy, because someone doesn't understand maths
296 2014-09-09 08:47:06 <phantomcircuit> no really it's because they use a hash function
297 2014-09-09 08:47:18 <phantomcircuit> which is difficult to make map 1:1
298 2014-09-09 08:47:58 <phantomcircuit> timothy, but once again that's silly since any system with any sense uses random end results
299 2014-09-09 08:48:03 <timothy> they also use the "privacy" extension (by default in windows), so the don't use mac address either
300 2014-09-09 08:48:43 <warren> You have native ipv6 at home?
301 2014-09-09 08:49:03 <phantomcircuit> warren, i actually turned it off because it's broken
302 2014-09-09 08:49:05 <phantomcircuit> thanks comcast
303 2014-09-09 08:49:18 <timothy> no, but I work for a company that builds access gateways (CPE)
304 2014-09-09 08:50:18 <phantomcircuit> i do hae a /64 on a server though
305 2014-09-09 08:50:37 <btcdrak> JetBrains finally released a C++ IDE <3 http://www.jetbrains.com/clion/
306 2014-09-09 08:50:39 <warren> do folks still use the 6 over 4 tunnels?
307 2014-09-09 08:50:47 <btcdrak> I totally love their python IDE editor
308 2014-09-09 08:51:30 <warren> I've mostly ignored ipv6 all this time, but now I want to test mitigation measures against it...
309 2014-09-09 08:51:52 <michagogo> Hm, just did some reading, sounds like a subnet is defined as /64
310 2014-09-09 08:52:26 <michagogo> With the rfc suggesting that a customer get multiple subnets, e.g. /56
311 2014-09-09 08:53:26 <wumpus> anyone ever tried building/running bitcoind with fsplit-stack?
312 2014-09-09 08:53:42 <warren> what is that?
313 2014-09-09 08:53:45 <gribble> Split Stacks - GCC - GNU: <https://gcc.gnu.org/wiki/SplitStacks>; mathieu lacage - using -fsplit-stack - GCC - GNU: <http://gcc.gnu.org/ml/gcc-help/2012-11/msg00223.html>; Ian Lance Taylor - Re: using -fsplit-stack - GCC - GNU: <http://gcc.gnu.org/ml/gcc-help/2012-11/msg00224.html>
314 2014-09-09 08:53:45 <michagogo> ;;google fsplit-stack
315 2014-09-09 08:53:55 <lewellyn> michagogo: that's what i said. 01:34  lewellyn: /64 is not uncommon 01:34  lewellyn: maybe like a /56.
316 2014-09-09 08:54:53 <wumpus> it's a way to avoid allocating a block of memory per thread
317 2014-09-09 11:03:56 <skinnkavaj> https://www.youtube.com/watch?v=bma1p-DfocM
318 2014-09-09 11:04:12 <skinnkavaj> #Bitcoin2014 - Feat. Presentation: Taking Bitcoin Development to the Next Level by Mike Hearn
319 2014-09-09 11:15:26 <hearn> skinnkavaj: where'd you find the link?
320 2014-09-09 11:42:04 <skinnkavaj> hearn: Reddit frontpage
321 2014-09-09 11:42:19 <hearn> oh yeah
322 2014-09-09 11:43:14 <hearn> skinnkavaj: what did you think of it?
323 2014-09-09 11:44:50 <skinnkavaj> hearn: Was kinda boring to see the old screenshots of lighthouse, would rather want to see how lighthouse is now when released
324 2014-09-09 11:44:56 <hearn> right
325 2014-09-09 11:45:16 <hearn> will be talking to olivier this afternoon, is the plan. i'm impatient to get the source code out there.
326 2014-09-09 11:45:21 <skinnkavaj> hearn: Whats left to do with it now?
327 2014-09-09 11:46:00 <hearn> needs a bit more polish, a few edge case bugs being resolved. perhaps more features. olivier wants it to be as solid a product as possible to do end user release, so that's one of the things we're discussing at the moment
328 2014-09-09 11:46:04 <hearn> what exactly is needed for beta, for 1.0 etc
329 2014-09-09 11:48:57 <skinnkavaj> hearn: You should make an awesome trailer for it, like the Open Bazaar one. Seems to have attracted a lot of interested developers helping out with Open Bazaar development
330 2014-09-09 11:49:02 <hearn> yeah
331 2014-09-09 11:49:07 <hearn> i know :) there's lots of things to do
332 2014-09-09 11:49:27 <hearn> the source code will come first and a nice website + video + end user docs + native packages etc will come later
333 2014-09-09 11:49:37 <hearn> actually native packages are mostly done already. some people already have alpha builds that auto update
334 2014-09-09 11:50:06 <hearn> i'm not totally sure when the right time to do a video is. i don't want to make a "trailer" for something you can't download, but then again as more and more people get access to the app, i guess someone will make a video at some point anyway.
335 2014-09-09 11:50:08 <hearn> need to think about it
336 2014-09-09 11:51:04 <brisque> hearn: thinking about all the Bitcoin-related products that have had trailers, are you sure that's a good idea?
337 2014-09-09 11:51:14 <hearn> sure what's a good idea?
338 2014-09-09 11:51:31 <brisque> well, a trailer.
339 2014-09-09 11:51:58 <hearn> oh, you mean, doing one at all?
340 2014-09-09 11:52:10 <hearn> normally i always want to see screenshots of an app i'm about to download.
341 2014-09-09 11:52:32 <hearn> so i guess a video is a logical thing people will want too. it's just a quick way to judge "is this app worth my time"?
342 2014-09-09 11:53:04 <skinnkavaj> Hard to not want to use Open Bazaaar after watching this video: https://www.youtube.com/watch?v=zAGij8FSrNc
343 2014-09-09 11:55:17 <brisque> it's entirely possible I'm alone in it, but to me all of the products with big flashy videos never amount to anything.
344 2014-09-09 11:56:38 <brisque> screenshots of a product, sure, that's a different beast to a promotional video.
345 2014-09-09 11:57:24 <dbrock> I think the person who made the psychedelic Mycelium trailer should be hired to make every single future crypto-related video ever
346 2014-09-09 11:59:53 <brisque> dbrock: that's almost certainly sending the wrong message.
347 2014-09-09 12:02:03 <dbrock> brisque: what's wrong with being sexy?
348 2014-09-09 12:04:25 <hearn> skinnkavaj: well that's a bit over the top ....
349 2014-09-09 12:05:04 <hearn> i'd probably do a (quieter) training video at least.
350 2014-09-09 12:05:07 <hearn> anyway. we'll see.
351 2014-09-09 12:05:11 <brisque> dbrock: you don't see the problem with Bitcoin being described by a hippie having a 60s fueled acid trip?
352 2014-09-09 12:07:42 <skinnkavaj> hearn: Tell me more about your ideas how to speed up development and reach concensus easier, you have had a lot of these ideas.. Have you successfully implemented them in lighthouse?
353 2014-09-09 12:07:56 <hearn> lighthouse only implements sighash_anyonecanpay type contracts
354 2014-09-09 12:08:04 <hearn> it doesn't require any particular kind of project
355 2014-09-09 12:08:08 <hearn> it's just a way to move money around
356 2014-09-09 12:08:40 <jsfsn> The initial users are most likely more used to text explaining what they are about to use rather than a video
357 2014-09-09 12:08:51 <hearn> the idea is, there can be websites or aggregator sites that gather project (files) together.
358 2014-09-09 12:08:53 <jsfsn> regarding lighthouse
359 2014-09-09 12:09:07 <hearn> so it's kind of a blend of a p2p desktop app and web stuff. the web is easier for things like commenting, sharing, exploring, etc.
360 2014-09-09 12:09:26 <hearn> jsfsn: sure, i didn't mean as a replacement for real documentation :)    which is one thing I still need to write ....
361 2014-09-09 12:09:30 <hearn> though ideally the app wouldn't need any.
362 2014-09-09 12:10:03 <hearn> i don't think it's hard to use at the moment, but a few hints scattered around would help. anyway post open sourcing i'm going to try and recruit some people who enjoy UI programming to help with things like that
363 2014-09-09 12:10:04 <jsfsn> ideally :)
364 2014-09-09 13:56:01 <wumpus> what do projects use to host their mailinglists in the post-sourceforge era?
365 2014-09-09 14:03:14 <pigeons> i've seen some google groups being used, not endorsing it
366 2014-09-09 14:03:25 <pigeons> seems to work fine though
367 2014-09-09 14:31:32 <andytoshi> for pcb/geda we used a listserv hosted on dj's server http://www.delorie.com/listserv/
368 2014-09-09 14:31:57 <andytoshi> idk how much hassle that is for him to maintain
369 2014-09-09 14:32:14 <bsm117532> So I've been tasked with creating an alt-coin, and I am loth to duplicate some of the jokers I've seen in the alt-coin world.  Therefore I wonder what people might think of this proposal:
370 2014-09-09 14:32:34 <bsm117532> bitcoin while also contributing to this new coin.  The coin, initially will not have any coinbase reward and the only nodes/miners will be run by the company issuing it.
371 2014-09-09 14:32:34 <bsm117532> What if we make a direct clone of bitcoin and maintain a private blockchain?  This will allow me to follow and contribute to
372 2014-09-09 14:33:07 <bsm117532> By establishing a market for the coin first (while having a peg to BTC) we hope to avoid the speculation & pump/dump that plagues altcoins, since they have no inherent value.
373 2014-09-09 14:33:07 <bsm117532> Its value will be pegged to BTC or some other asset like USD.  Atsome point in the future we could choose to "release" the coin from its peg, add a coinbase reward, and recruit miners.
374 2014-09-09 14:33:28 <bsm117532> Thoughts?
375 2014-09-09 14:34:13 <andytoshi> bsm117532: you can't ask altcoin questions here ... -wizards might be more appropriate ... but thoughts are, you would not make a private blockchain by cloning bitcoin
376 2014-09-09 14:34:52 <bsm117532> Sorry, thought this would be the most appropriate place.  Intention is to work on/follow bitcoin.
377 2014-09-09 14:35:24 <andytoshi> -wizards is fine if you wanna chat about the peg mechanism
378 2014-09-09 14:35:35 <bsm117532> I'll ask there...but it seems out of scope relative to ECDSA conversations.  ;-)
379 2014-09-09 16:22:37 <jgarzik> gavinandresen, gmaxwell: Maybe some review for the stupid and forgetful (me)...   what are the goals served by remembering orphan transactions?
380 2014-09-09 16:22:51 <jgarzik> IMO you should only need to remember wallet orphans
381 2014-09-09 16:22:59 <jgarzik> others will get retransmitted (or not)
382 2014-09-09 16:23:07 <sipa> being able to deal with out of order announced transactions
383 2014-09-09 16:23:15 <sipa> only wallet transactions are reannounced
384 2014-09-09 16:23:36 <brisque> jgarzik: I don't think you can assume that in the current network. the vast majority of wallets in use never rebroadcast transactions.
385 2014-09-09 16:23:36 <sipa> if they depend on other unconfirmed transaction that your node misses, you can't process them
386 2014-09-09 16:23:56 <jgarzik> correction: wallet tx's and their deps
387 2014-09-09 16:24:09 <jgarzik> any reason to remember unrelated?  ie. some might crop up during a reorg.
388 2014-09-09 16:24:39 <sipa> just for ehatever reason if you only receive a second one in a chain
389 2014-09-09 16:24:43 <sipa> maybe you just connected
390 2014-09-09 16:25:05 <jgarzik> the consequences of failing that are low, and have to be handled anyway
391 2014-09-09 16:25:10 <sipa> maybe the first one was one your peer already knew, so it didn't relay it
392 2014-09-09 16:25:14 <jgarzik> it doesn't add much value
393 2014-09-09 16:25:35 <sipa> i would like to test that assumption first
394 2014-09-09 16:25:52 <sipa> see mempool growth in a double test for examplr
395 2014-09-09 16:26:42 <sipa> note that in general, nodes only announce each tx only once
396 2014-09-09 16:27:37 <sipa> if you're not connecting to a rebroadcasting wallet node dirextly, and for whatever reason you miss the first announcement, or it happens in different order from different peers, you will miss that transaction without orphan pool
397 2014-09-09 16:34:31 <blast_> hmm
398 2014-09-09 16:47:43 <jgarzik> sipa, sure, but it is a partial effort in any case
399 2014-09-09 16:48:02 <jgarzik> sipa, there are many cases where you come online at a different time, and simply miss those tx altogether
400 2014-09-09 16:49:56 <jgarzik> people typically have an economic incentive to get transactions confirmed, and will manually rebroadcast via bitcoind restart (or blockchain.info push)
401 2014-09-09 16:50:32 <jgarzik> it is difficult to see a case where a random third party remembering random third party orphans adds much value
402 2014-09-09 16:57:21 <sipa> jgarzik: i'm just saying that it has a use case; i don't know how often it used used or needed, but let's not just throw it away without having numbers
403 2014-09-09 17:24:40 <reticular> a lot of brains in here for it to be so quiet
404 2014-09-09 17:27:51 <reticular> anyone looking for new development?  probably Java or Node, restful interface
405 2014-09-09 17:28:09 <sipa> not here
406 2014-09-09 17:32:18 <reticular> Coinsafe solution for decreased transaction times is a fantastic step forward, but 10 seconds is still too long.  We should strive for sub-second confirmations with 99% accuracy
407 2014-09-09 17:33:54 <sipa> bitcoin-the-technology simply can't do that due to the laws of physics (information does not travel across the globe fast enough)
408 2014-09-09 17:34:00 <reticular> a decentralized model that provides probabilistic guarantees.  That probability increases as you get closer to 10 minutes of course.
409 2014-09-09 17:34:29 <reticular> let the customer decide what probability they deem worth the risk.  95% , 99%  99.9999%
410 2014-09-09 17:34:52 <sipa> even confirmations in the blockchain are probabilistic
411 2014-09-09 17:35:07 <sipa> anything below that is not even going to give you a number
412 2014-09-09 17:35:32 <reticular> Exposing that probability as the transactions are recorded can have value
413 2014-09-09 17:35:56 <sipa> do you have any actual ideas on how to accomplish that?
414 2014-09-09 17:36:57 <gavinandresen> sipa: I do…  (have miners gossip about the blocks they’re working on)
415 2014-09-09 17:37:03 <reticular> Yes, I have actual ideas on how to accomplish that.  Will my ideas work... no idea.
416 2014-09-09 17:37:13 <sipa> gavinandresen: if that includes PoW, sure
417 2014-09-09 17:37:32 <sipa> you're doing something like p2pool style shares that offer sub-confirmations
418 2014-09-09 17:37:37 <gavinandresen> exactly
419 2014-09-09 17:37:59 <sipa> reticular: then talk about those :)
420 2014-09-09 17:38:24 <Luke-Jr> the main (only?) downside I see to that method is that lucky blocks won't get the same benefit as not-so-lucky ones. but maybe not a real issue.
421 2014-09-09 17:39:19 <sipa> Luke-Jr: hmm?
422 2014-09-09 17:40:07 <Luke-Jr> sipa: if your block is the first one satisfying the gossip PoW
423 2014-09-09 17:40:15 <gavinandresen> Luke’s saying that if a block is found really quickly, before a miner even has a chance to broadcast “here’s what I’m working on”, it won’t propagate as fast
424 2014-09-09 17:40:21 <reticular> interesting with the "egging" attitude.  I've been a developer for way to many years to be impacted by developer arrogance.
425 2014-09-09 17:41:03 <gavinandresen> … but we can hyper-optimize block propagation in any case, whether they’re “this is what I’m working on and I ALMOST solved it” or satisfies-full-POW blocks.
426 2014-09-09 17:41:38 <sipa> reticular: i just mean you're just talking about sole vaulgue and hard goal that obviously everyone would want if it had no cost. the interesting part is how to do it :)
427 2014-09-09 17:41:46 <sipa> *some vague
428 2014-09-09 17:42:40 <sipa> gavinandresen, Luke-Jr: there are two separate goals here; faster propagation and more granular confirmation
429 2014-09-09 17:43:40 <gavinandresen> sipa: yes, two separate problems. But I think solving them together incentivizes miners to run the new code that can support much finer-grain confirmation confidence
430 2014-09-09 17:43:49 <sipa> right
431 2014-09-09 17:45:17 <Luke-Jr> I'd worry about people using it as evidence of confirmation for unconfirmed transactions, but otherwise sounds good to me
432 2014-09-09 17:45:26 <sipa> well, it is
433 2014-09-09 17:45:58 <Luke-Jr> sipa: not much more than watching for conflicting relays
434 2014-09-09 17:46:02 <gavinandresen> Glancing at http://www.verizonenterprise.com/about/network/latency/ …. it might be possible to get reasonable confidence about low-value transactions in under a second.
435 2014-09-09 17:46:08 <Luke-Jr> it's still completely vulnerable to a finney attack
436 2014-09-09 17:46:29 <Luke-Jr> which is sadly just a matter of cost today
437 2014-09-09 17:47:15 <sipa> Luke-Jr: right, it is proof.of work, but not work others can build on, and there is no value in it, so also nothing to lose by not building on it
438 2014-09-09 17:49:30 <petertodd> sipa: it's proof that you did some mining, without much evidence that you won't change your mind later