1 2011-12-25 00:00:33 <CIA-100> bitcoin: Luke Dashjr blknotify * rd30b81b16cf9 bitcoind-personal/src/ (init.cpp main.cpp): Execute a command when best block changes (-blocknotify=<cmd>) http://tinyurl.com/cszuxd3
  2 2011-12-25 00:20:13 <CIA-100> bitcoin: various next-test * r37528c..d64b3d bitcoind-personal/ (10 files in 3 dirs): (8 commits) http://tinyurl.com/7vr93zh
  3 2011-12-25 02:10:17 <CIA-100> libbitcoin: Kamil Domanski cleanup * red232fa07e49 / (6 files in 3 dirs): connection and handshaking code cleanup http://tinyurl.com/c23ptyx
  4 2011-12-25 02:11:47 <Kiba`> hey guys
  5 2011-12-25 02:11:52 <Kiba`> I want to fund a game with bitcoin
  6 2011-12-25 02:26:40 <luke-jr> Kiba`: I hear Humble Bumble will accept Bitcoins next time around
  7 2011-12-25 02:34:00 <Kiba`> luke-jr: and you will donate a bitcoin to help get my project started
  8 2011-12-25 02:35:01 <Kiba`> luke-jr: http://kibabase.com/articles/the-city-a-modest-proposal
  9 2011-12-25 05:43:11 <Rabbit67890> MagicalTux: Thanks
 10 2011-12-25 07:01:08 <midnightmagic> HAPPY XMAS! :)
 11 2011-12-25 10:00:41 <kazey> Hi guys i am preety new to bitcoin concepts, i understand the P2P part but i am a bit confused about something. Lets say i receive BTC to a store, and its in the wallet, is there a way I can transfer it to the bitcoin desktop client to enable ease of transfer to whoever i want
 12 2011-12-25 12:46:42 <CIA-100> bitcoin: Gavin Andresen master * rfe35816 / src/main.cpp : Be more conservative: check all transactions in blocks after last checkpoint. - http://git.io/ZbaFUA https://github.com/bitcoin/bitcoin/commit/fe358165e3ed3656dcc501f1a585dd5eaecf9b45
 13 2011-12-25 13:18:18 <TuxBlackEdo> merry christmas gavinandresen
 14 2011-12-25 13:22:20 <luke-jr> gavinandresen: I thought you were gone for Christmas? :P
 15 2011-12-25 13:24:26 <sipa> "gone" is relative
 16 2011-12-25 13:26:52 <JFK911> gone: http://www.dailymail.co.uk/news/article-2078434/Mother-confesses-throwing-daughters-corpse-trash.html
 17 2011-12-25 13:48:13 <gavinandresen> I am gone for Christmas... Merry Christmas everybody!
 18 2011-12-25 13:48:46 <TuxBlackEdo> afk for christmas
 19 2011-12-25 13:50:22 <CIA-100> bitcoin: Luke Dashjr 0.5.x * r1a778b07496d bitcoind-stable/src/main.cpp: Merge branch '0.5.0.x' into 0.5.x http://luke.dashjr.org/programs/bitcoin/w/bitcoind/stable.git/commitdiff/1a778b07496d9ceb4527ba3b6a59e28154bb02c5
 20 2011-12-25 13:51:12 <gavinandresen> I would be completely gone if I didn't get a report of a remotely exploitable crashing bug in git HEAD last night...  sigh.
 21 2011-12-25 13:51:41 <luke-jr> heh, that one DID seem more serious than the commit message suggested <.<
 22 2011-12-25 13:52:16 <luke-jr> although 0.4 and 0.5 don't check inputs for "standardness", so I wonder just how bad it is for them
 23 2011-12-25 13:52:26 <gavinandresen> at least it was caught before we produced binaries-- but anybody running git HEAD on the production network should pull
 24 2011-12-25 13:52:30 <makomk> gavinandresen: sorry about that.
 25 2011-12-25 13:52:38 <gavinandresen> Doesn't affect 0.4/0.5 or my backported patches
 26 2011-12-25 13:53:00 <gavinandresen> makomk: no need to apologize, critical bugs need to get reported and fixed
 27 2011-12-25 13:53:02 <luke-jr> gavinandresen: oh, did 0.4/0.5 have a check for that condition somewhere else?
 28 2011-12-25 13:53:19 <luke-jr> in any case, I backported the part that fit with ConnectInputs
 29 2011-12-25 13:54:01 <gavinandresen> luke-jr: the bug is I messed up with I split the old ConnectInputs into FetchInputs/ConnectInputs
 30 2011-12-25 13:55:44 <gavinandresen> makomk: You don't happen to have an intelligent input-fuzzing tool, do you?
 31 2011-12-25 13:55:57 <luke-jr> doh, you're right, I just doubled up the check&
 32 2011-12-25 13:56:38 <makomk> gavinandresen: sadly not; discovered this completely by accident.
 33 2011-12-25 13:58:24 <nanotube> i bet there are such things living on sourceforge
 34 2011-12-25 13:58:28 <makomk> (I managed to accidentally introduce a bug that mangled transactions in a different way which triggered the same crash.)
 35 2011-12-25 13:58:48 <luke-jr> LLVM has a nice thing for testing
 36 2011-12-25 13:59:29 <luke-jr> I don't know how to use it tho
 37 2011-12-25 13:59:48 <makomk> gavinandresen: also, see your e-mail...
 38 2011-12-25 14:00:16 <CIA-100> bitcoin: Luke Dashjr 0.5.x * ra91a40febd06 bitcoind-stable/src/main.cpp: Merge branch '0.5.0.x' into 0.5.x http://tinyurl.com/coewtvq
 39 2011-12-25 14:00:55 <luke-jr> I think I'm going to EOL 0.5.0.x as soon as 0.6.0 is out
 40 2011-12-25 14:01:02 <luke-jr> too redundant
 41 2011-12-25 14:01:09 <gavinandresen> nanotube: I want a bitcoin-specific fuzzer, that does things like fuzzes a transaction then computes a valid hash....
 42 2011-12-25 14:01:44 <nanotube> gavinandresen: well, then some work will probably be required :)
 43 2011-12-25 14:01:50 <luke-jr> LLVM's tool simulates every possible input to the program ;)
 44 2011-12-25 14:02:21 <nanotube> luke-jr: haha well, 99.9-> percent of inputs will not have a valid hash then. ;)
 45 2011-12-25 14:02:31 <nanotube> but if time is not a constraint.... :)
 46 2011-12-25 14:02:38 <luke-jr> nanotube: sure, but you'll catch crashes
 47 2011-12-25 14:02:52 <luke-jr> nanotube: you're assuming it's slow, brute-force ;)
 48 2011-12-25 14:03:05 <gavinandresen> makomk: that'll teach me to try to fix bugs between opening presents and breakfast....
 49 2011-12-25 14:03:20 <luke-jr> nanotube: rather, AIUI, it allows memory to contain quantum states basically
 50 2011-12-25 14:03:33 <gavinandresen> makomk: (I got the assert condition backwards)
 51 2011-12-25 14:03:42 <luke-jr> nanotube: so when there's a branch, it follows both paths (constrained by the quantum state of the condition)
 52 2011-12-25 14:03:51 <nanotube> heh
 53 2011-12-25 14:03:53 <makomk> gavinandresen: ah, that explains it.
 54 2011-12-25 14:05:05 <luke-jr> I suppose the hashing might slow it down
 55 2011-12-25 14:05:30 <gavinandresen> You'll never accidently get a valid SHA256 hash
 56 2011-12-25 14:06:41 <gavinandresen> ... so a dumb input fuzzer will never exercise any code after a "does the hash match the data I was expecting to get" check
 57 2011-12-25 14:07:12 <gavinandresen> Ditto for fuzzing blocks-- a smart block fuzzer should fuzz the block, then do the proof-of-work so the block has a valid hash....
 58 2011-12-25 14:07:20 <gavinandresen> (dumb fuzzing should also be done, of course)
 59 2011-12-25 14:08:07 <CIA-100> bitcoin: Gavin Andresen master * r60835d9 / src/main.cpp : assert condition in previous commit was backwards - http://git.io/-IEC6Q https://github.com/bitcoin/bitcoin/commit/60835d96276549ce17fe163a18c2f78dc5b267bc
 60 2011-12-25 14:08:53 <cjdelisle> perhaps it would make a nice unit test, just generate a random tx/block and call the verification function on it
 61 2011-12-25 14:09:10 <cjdelisle> then fuzzing is just a matter of running that unit test in a loop
 62 2011-12-25 14:09:16 <makomk> gavinandresen: yep, that seems to work nicely now the assertion's fixed :-)
 63 2011-12-25 14:10:35 <gavinandresen> cjdelisle: I think a speaks-at-the-network-protocol-level fuzzer would be worth the investment of time, because it will test any implementation
 64 2011-12-25 14:10:46 <gavinandresen> (unit tests are always a good idea, too....)
 65 2011-12-25 14:10:49 <cjdelisle> good point..
 66 2011-12-25 14:11:56 <cjdelisle> I'm too lazy/pressed to write lots of unit tests for my project so I end up writing one unit test which tests a whole lot of code at once, kind of a unit/functional test
 67 2011-12-25 14:15:57 <luke-jr> gavinandresen: LLVM's quantum-like thing will get past it I think
 68 2011-12-25 14:17:09 <gavinandresen> luke-jr: cool. I don't know nuthin about LLVM, you willing to tackle that project?
 69 2011-12-25 14:18:49 <makomk> gavinandresen: actually, I have a feeling the patch may not be enough on second thoughts.
 70 2011-12-25 14:19:15 <gavinandresen> makomk: what did I miss?
 71 2011-12-25 14:19:15 <luke-jr> gavinandresen: unfortunately, I haven't had time to even look at it for my paid work yet :/
 72 2011-12-25 14:19:26 <luke-jr> (I also know nothing about LLVM)
 73 2011-12-25 14:19:45 <makomk> if (inputsRet.count(prevout.hash))
 74 2011-12-25 14:19:46 <makomk> continue; // Got it already
 75 2011-12-25 14:20:28 <makomk> ^ it's only checked for the first input from that input transaction.
 76 2011-12-25 14:21:50 <gavinandresen> Hmm?  In FetchInputs() ?  that's in a for () loop over all inputs
 77 2011-12-25 14:22:08 <gavinandresen> Oh, I see what you're saying....
 78 2011-12-25 14:22:31 <gavinandresen> ... need to duplicate the check before continuing...
 79 2011-12-25 14:23:10 <makomk> Yep, or something. Just fired up and crashed a node with latest git head.
 80 2011-12-25 14:23:17 <gavinandresen> makomk: good catch.
 81 2011-12-25 14:24:03 <nanotube> gavinandresen: maybe you should save this for after breakfast :)
 82 2011-12-25 14:25:19 <makomk> Yeah, that might be a good idea. The world probably won't catch fire in the next hour or so. I hope.
 83 2011-12-25 14:27:38 <Diablo-D3> help
 84 2011-12-25 14:27:42 <Diablo-D3> er
 85 2011-12-25 14:27:43 <Diablo-D3> welp
 86 2011-12-25 14:27:54 <Diablo-D3> I guess my bid to save bitcoin has failed.
 87 2011-12-25 14:34:06 <gavinandresen> makomk: Can you git am the patch I just sent via email and test?
 88 2011-12-25 14:34:40 <makomk> gavinandresen: sure!
 89 2011-12-25 14:37:05 <makomk> gavinandresen: yeah, that seems to fix it.
 90 2011-12-25 14:37:43 <kiba> hey guys, I wanna make a game! Fund me with some bitcoin!
 91 2011-12-25 14:37:51 <kiba> http://kibabase.com/articles/the-city-a-modest-proposal
 92 2011-12-25 14:37:53 <gavinandresen> makomk: great, thanks.
 93 2011-12-25 14:38:10 <CIA-100> bitcoin: Gavin Andresen master * r61977f9 / src/main.cpp : Check all prevout.n if one transaction provides multiple inputs - http://git.io/ksFPOQ https://github.com/bitcoin/bitcoin/commit/61977f956c6bacd6e05e3211bf2a90e5b5672a94
 94 2011-12-25 14:38:15 <makomk> gavinandresen: no problem, hope you enjoy the rest of your Christmas!
 95 2011-12-25 14:38:24 <gavinandresen> ... one of these years I'll manage to get everything right the first time... (hah!)
 96 2011-12-25 15:40:32 <finway> Merry Chrismas!
 97 2011-12-25 15:40:56 <cjdelisle> And Merry Christmas to you, and Merry Yesterday and Merry Tomorrow :)
 98 2011-12-25 15:41:55 <finway> Thank you for your hard work, all  the developers, it's my hornor to know you guys.
 99 2011-12-25 16:04:12 <Red_> hi
100 2011-12-25 16:04:21 <Red_> In the Bitcoin addresses generation - what means step 4 Add network byte ?
101 2011-12-25 16:07:10 <[Tycho]> May be it's the version number ?
102 2011-12-25 16:08:11 <sipa> it is
103 2011-12-25 16:17:39 <Red_> hm
104 2011-12-25 16:18:40 <Red_> Does Network ID Byte mean version number?
105 2011-12-25 16:25:04 <[Tycho]> Yes.
106 2011-12-25 16:25:34 <[Tycho]> Currently there are two version numbers allowed in bitcoin addresses.
107 2011-12-25 16:25:52 <[Tycho]> And some are used by alt chains or testnet.
108 2011-12-25 16:28:52 <Red_> thanks for the info .cya
109 2011-12-25 18:39:04 <Diablo-D3> https://bitcointalk.org/index.php?topic=51371.0
110 2011-12-25 18:39:12 <Diablo-D3> CAUTION: DO NOT INSERT INTO A MEMORY SLOT!!!
111 2011-12-25 18:42:00 <lianj> blockchain overflow
112 2011-12-25 18:43:48 <cjdelisle> pretty neat, I hope someone makes something like that w/ gb ethernet socket because it would have other uses apart from mining
113 2011-12-25 18:47:21 <Diablo-D3> gmaxwell: did someone update the top 10 lately?
114 2011-12-25 18:47:29 <Diablo-D3> I dont remember it looking like that
115 2011-12-25 18:49:28 <Diablo-D3> oh well time to update
116 2011-12-25 19:00:32 <Joric> damnit those boards still need air cooling
117 2011-12-25 19:07:10 <Diablo-D3> emc is down goddamnit
118 2011-12-25 19:07:14 <Diablo-D3> oh wait there it goes
119 2011-12-25 19:17:16 <Diablo-D3> there top ten done
120 2011-12-25 19:18:02 <Diablo-D3> deepbit, slush, btc guild, abcpool, eligius, bitclockers, emc, bitcoins.lc, ars, btcmine
121 2011-12-25 19:20:42 <CIA-100> libbitcoin: genjix * r33c0cc31a703 /doc/organize/ (begin.svg index.html pool.png tree.png): Added organize document. http://tinyurl.com/7npghvh
122 2011-12-25 19:22:23 <Joric> Diablo-D3, do you have an url to the hashrate distribution chart?
123 2011-12-25 19:23:38 <Diablo-D3> Joric: no
124 2011-12-25 19:23:56 <Diablo-D3> its probably 60% deepbit, 40% everybody else.
125 2011-12-25 19:40:11 <Diablo-D3> https://imgur.com/gallery/0czi8
126 2011-12-25 19:40:12 <Diablo-D3> :3
127 2011-12-25 19:41:53 <cjdelisle> must be a lot of people in here who clicked because they claim to be "over capacity"
128 2011-12-25 19:42:22 <Rabbit67890> haha :3
129 2011-12-25 19:42:35 <Diablo-D3> fuck.
130 2011-12-25 19:42:46 <Diablo-D3> bitcoin: ruining your internets since 2010.
131 2011-12-25 19:42:53 <Rabbit67890> Don't really see the image, but theres a image saying its over capacity&. :3
132 2011-12-25 19:43:07 <Diablo-D3> keep trying
133 2011-12-25 19:43:11 <Rabbit67890> damnit. this is #bitcoin-dev
134 2011-12-25 19:43:19 <Diablo-D3> THIS IS SPARTAAAAAAAAA
135 2011-12-25 19:43:50 <cjdelisle> didjya here? irc.anonops is back, the wild ddos days are here again
136 2011-12-25 19:43:56 <Rabbit67890> D:
137 2011-12-25 19:44:10 <cjdelisle> yesterday sweden's link to efnet kept splitting
138 2011-12-25 19:44:11 <cjdelisle> heh
139 2011-12-25 19:44:17 <Rabbit67890> heh
140 2011-12-25 19:44:19 <Diablo-D3> dude
141 2011-12-25 19:44:20 <Diablo-D3> its efnet
142 2011-12-25 19:44:29 <Rabbit67890> they DDOS imgur now...
143 2011-12-25 19:44:31 <Diablo-D3> efnet splits more than ... than... fuck.
144 2011-12-25 19:44:35 <cjdelisle> &
145 2011-12-25 19:44:41 <Diablo-D3> I have nothing.
146 2011-12-25 19:45:16 <cjdelisle> noone can beat the mighty lukedots
147 2011-12-25 19:46:14 <Rabbit67890> haha
148 2011-12-25 19:46:47 <Rabbit67890> You have been disconnected. (net.split.bitcoin.freenode.net:6667)
149 2011-12-25 19:47:45 <cjdelisle> *.net and *.split are not server names , I dowt caw wut you say
150 2011-12-25 21:40:46 <CIA-100> bitcoin: Con Kolivas * r59293a37d6ab cgminer/util.c: Only pthread_join when pthread_cancel does not return an error. http://luke.dashjr.org/programs/bitcoin/w/cpuminer/cgminer.git/commitdiff/59293a37d6abaa879e6f24e26b206a3b318b2f79
151 2011-12-25 21:47:55 <onelineproof> Can a length 33 bitcoin address be valid?
152 2011-12-25 21:49:48 <wladston> anyone out there using django-bitcoin ?
153 2011-12-25 21:51:49 <onelineproof> I making a c program to print out address and private key pairs. So far everything looks good except some addresses are of length 33 instead of 34. I compared with pywallet and they also have those addresses as length 33, so I'm guessing it's fine.
154 2011-12-25 21:52:21 <wladston> it's fine, there are even smaller addresses
155 2011-12-25 21:52:42 <onelineproof> Can private key be less than 51 length?
156 2011-12-25 21:52:58 <onelineproof> in standard format
157 2011-12-25 21:53:12 <wladston> I don't think so
158 2011-12-25 21:54:14 <wladston> but I'm not that much into the cripto part
159 2011-12-25 21:54:20 <onelineproof> ya so far I only get 51, but I may take care of the other cases in case it is
160 2011-12-25 21:55:07 <cjdelisle> that's a private key in base58?
161 2011-12-25 21:55:10 <wladston> looks like you get smaller adresses if you have zeroes at the beggining of your key
162 2011-12-25 21:55:28 <onelineproof> ya in bitcoin base 58 format
163 2011-12-25 21:55:36 <onelineproof> starts with a 5
164 2011-12-25 21:55:58 <cjdelisle> from what I know of base58, don't count on anything
165 2011-12-25 21:57:19 <onelineproof> ya its not such a big deal, I can take care of the smaller cases.
166 2011-12-25 21:57:34 <onelineproof> Ill put the code on github in case someone wants to use it
167 2011-12-25 21:57:42 <onelineproof> tomorrow maybe
168 2011-12-25 21:59:25 <cjdelisle> also, and this is just me, I'd be inclined to represent private keys in hex because then it's blatently clear that you are not to be pasting them to people asking for money
169 2011-12-25 21:59:58 <cjdelisle> it's a "Whow, that is not a bitcoin address" message
170 2011-12-25 21:59:58 <onelineproof> no but I want to make a paper wallet
171 2011-12-25 22:00:07 <onelineproof> in qr format also
172 2011-12-25 22:00:16 <cjdelisle> ahh
173 2011-12-25 22:00:47 <cjdelisle> I wish you luck and I hope you use C++ because "print wallet" is something bitcoin needs
174 2011-12-25 22:00:57 <onelineproof> its in c
175 2011-12-25 22:01:09 <onelineproof> only depends on libdb and libcrypto
176 2011-12-25 22:01:38 <cjdelisle> well if I ever fork bitcoin, I'm certainly going to integrate it
177 2011-12-25 22:01:49 <onelineproof> theres already pywallet that does it, but i just wanted to make it simpler
178 2011-12-25 22:02:11 <onelineproof> and kind of gain a better understanding :)
179 2011-12-25 22:02:16 <gmaxwell> cjdelisle: BAM https://github.com/sipa/bitcoin/tree/showwallet
180 2011-12-25 22:02:22 <gmaxwell> cjdelisle: do I get a prize?
181 2011-12-25 22:03:35 <cjdelisle> very nice
182 2011-12-25 22:03:56 <wladston> well, anyone knows Jeremias Kangas ? he is the guy who coded django-bitcoin
183 2011-12-25 22:04:28 <cjdelisle> Here's your golden star    you're going to have to pretend it's golden because this channel is +c
184 2011-12-25 22:05:01 <Diablo-D3> bwhah
185 2011-12-25 22:12:38 <onelineproof> what's the difference between "key" and "wkey" in the wallet.dat?
186 2011-12-25 22:16:43 <onelineproof> im guessing it only applies to some other crypto currencies like litecoin or namecoin, because I dont see it in my wallet, but it shows up as a case in pywallet
187 2011-12-25 22:22:11 <Joric> i wrote pywallet
188 2011-12-25 22:23:09 <Joric> well, the very first version of it )
189 2011-12-25 22:24:27 <Joric> wkey is 'wallet key', [public key, private key, created, expires, comment]
190 2011-12-25 22:25:27 <onelineproof> in my wallet i don't see that as a "type" in wallet.dat
191 2011-12-25 22:25:42 <onelineproof> but it's not such a big deal, i can still add that case
192 2011-12-25 22:26:56 <Joric> yeah there's no wkeys in my wallet as well just tried to trace those
193 2011-12-25 22:28:15 <Joric> only 'key' entries, no 'wkey's
194 2011-12-25 22:29:08 <sipa> not sure what wkey entries were used for
195 2011-12-25 22:29:13 <sipa> i've never seen them used
196 2011-12-25 22:29:33 <Joric> even gavin doesn't know what they are for afaik
197 2011-12-25 22:29:39 <onelineproof> ya its not just in pywallet, it was originally written by gavin in his bitcointools, and maybe even in the original client
198 2011-12-25 22:30:03 <Joric> maybe they were planned for 'lightweight' client or something
199 2011-12-25 22:30:06 <sipa> well, what they are for: wallet keys that expire
200 2011-12-25 22:30:10 <sipa> just for security
201 2011-12-25 22:30:13 <sipa> is my guess
202 2011-12-25 22:30:33 <onelineproof> db.cpp:854:            else if (strType == "key" || strType == "wkey")
203 2011-12-25 22:30:34 <Joric> maybe a future feature they hold creation date and shit
204 2011-12-25 22:31:00 <sipa> onelineproof: the base58 private key format can be shorten than 51 characters i believe
205 2011-12-25 22:31:03 <sipa> *shorter
206 2011-12-25 22:31:09 <onelineproof> ok cool
207 2011-12-25 22:31:46 <sipa> just base58 decode them - bitcoin uses a weird scheme for doing so
208 2011-12-25 22:31:58 <sipa> that doesn't gauarantee constant length
209 2011-12-25 22:32:21 <Joric> do you guys ever celebrate christmas? :D
210 2011-12-25 22:32:33 <sipa> it's december 26 over here ;)
211 2011-12-25 22:42:22 <[Tycho]> Depending on the christmas.
212 2011-12-25 22:43:48 <cjdelisle> tycho are your miner boxes fpga based?
213 2011-12-25 22:44:02 <[Tycho]> No.
214 2011-12-25 22:44:06 <cjdelisle> asic?
215 2011-12-25 22:44:18 <[Tycho]> Which boxes are you talking about ?
216 2011-12-25 22:44:31 <cjdelisle> the ones you were talking about building and selling a while back
217 2011-12-25 22:44:35 <[Tycho]> Yes, ASIC.
218 2011-12-25 22:45:14 <cjdelisle> k, I'm just curious about anything that's got fpgas connected to ethernet sockets and has been driven down in price by miners buying them
219 2011-12-25 22:45:19 <cjdelisle> for little crypto boxen
220 2011-12-25 22:45:28 <[Tycho]> VPN ?
221 2011-12-25 22:45:39 <cjdelisle> like that but all my own code
222 2011-12-25 22:46:03 <[Tycho]> Do you know that you don't need FPGAs for that ? Ordinary CPUs are fast enough.
223 2011-12-25 22:46:43 <Joric> ASIC? where do you get those, are they non-reprogrammable?
224 2011-12-25 22:46:52 <[Tycho]> Also you can use either any development board with ethernet or connect some ethernet transceiver to ngzhang's board.
225 2011-12-25 22:46:59 <[Tycho]> Joric: yes.
226 2011-12-25 22:47:14 <cjdelisle> I'm getting ~25Mb/s on my desktop running to a kvm instance on my desktop but that's with 3 layers of crypto and most nodes will only be doing one.
227 2011-12-25 22:47:47 <[Tycho]> ASICs are never reprogrammable.
228 2011-12-25 22:47:56 <cjdelisle> still interested in pushing that into the thousandish range
229 2011-12-25 22:48:26 <Joric> i suppose they only know how to calculate sha256? how many at once?
230 2011-12-25 22:48:42 <[Tycho]> Joric: how many of what ?
231 2011-12-25 22:48:56 <cjdelisle> sha circuits?
232 2011-12-25 22:49:04 <[Tycho]> cjdelisle: thousands of Mb/s ?
233 2011-12-25 22:49:33 <cjdelisle> That's the hope, these are going to be running between colo'd nodes if everything works out right
234 2011-12-25 22:49:58 <cjdelisle> so nodes are going to be routing for other nodes which drives up the mb/s peretty fast
235 2011-12-25 22:50:02 <[Tycho]> Why would you need that ?
236 2011-12-25 22:50:15 <cjdelisle> replacing the internet ;)
237 2011-12-25 22:51:51 <cjdelisle> an interesting property of this algorithm is that most routing can be done with no routing table lookups, the packet tells the router (which I call a "switch") which interface to send it down.
238 2011-12-25 22:52:06 <cjdelisle> Which means it can run on nonstandard hardware
239 2011-12-25 22:53:23 <[Tycho]> Replacing internet and running over it ?
240 2011-12-25 22:54:31 <cjdelisle> everything has to start somewhere
241 2011-12-25 22:55:11 <CIA-100> bitcoin: Con Kolivas * rb69aa23470ce cgminer/ (miner.h util.c): Use control_lock to protect thr->pth for thread creation/destruction. http://tinyurl.com/cxdmo5s
242 2011-12-25 22:55:28 <cjdelisle> the end vision is no more vpn links over "legacy" networks
243 2011-12-25 22:56:41 <[Tycho]> What are the benefits of your system ?
244 2011-12-25 22:57:17 <[Tycho]> And what is the level ?
245 2011-12-25 22:57:41 <cjdelisle> preformance, the routing table is split over the nodes so deaggregation is not an issue, to the point where nodes can safely use the hash of their public key as their ipv6 address
246 2011-12-25 22:58:52 <cjdelisle> privacy and integrity, all traffic is encrypted from end to end, ipv6 headers are encrypted from router to router and (optionally) switch headers are encrypted from switch to switch.
247 2011-12-25 22:59:14 <luke-jr> how is it better than MIPv6?
248 2011-12-25 22:59:24 <[Tycho]> Is it replacement for IP or it's on upper level ?
249 2011-12-25 22:59:39 <cjdelisle> it is a replacement for level3 and level2
250 2011-12-25 23:00:10 <[Tycho]> Will it require specialized hardware to work ? Like FPGA
251 2011-12-25 23:00:24 <cjdelisle> no but it will allow it to run faster
252 2011-12-25 23:00:51 <cjdelisle> it's a little bit like Cisco's LISP
253 2011-12-25 23:01:16 <cjdelisle> except instead of nodes looking up the "locator" they lookup the path and label the switch header with the path
254 2011-12-25 23:02:57 <cjdelisle> re MIPv6 (just looked it up) it doesn't really have the same goals
255 2011-12-25 23:04:08 <cjdelisle> another feature is it's intended ease of use, you shouldn't have to tell each router what it's unique position is in a network, you should be able to put them in those positions and let them discover their roll on their own.
256 2011-12-25 23:04:33 <cjdelisle> re no more configuring of OSPF areas
257 2011-12-25 23:04:33 <[Tycho]> Isn't replacing level2 a bit too far ?
258 2011-12-25 23:05:14 <cjdelisle> it needs it's own "level2", that said it can run over anything which connects 2 nodes, be it ethernet, wlan, udp/ip, etc.
259 2011-12-25 23:06:20 <[Tycho]> Strange...
260 2011-12-25 23:07:00 <cjdelisle> if you find it interesting, there's a partial half whitepaper/napkin here https://raw.github.com/cjdelisle/cjdns/master/rfcs/Whitepaper.txt
261 2011-12-25 23:08:36 <[Tycho]> May be it's using existing level2 ?
262 2011-12-25 23:09:38 <cjdelisle> yes, it can use ethernet or whatever but only as a point to point link
263 2011-12-25 23:10:36 <cjdelisle> the routers need to switch cloud to bring them together just as the switch cloud needs the routers to label packets in a way that they can switch them