1 2013-09-28 03:26:43 <jrmithdobbs> gmaxwell: you around?
2 2013-09-28 09:14:10 <warren> gmaxwell: https://github.com/bitcoin/bitcoin/issues/2770#issuecomment-25294214 damn... this is very similar to what we're hearing from this one litecoin mac user. clean shutdown, no sleep/suspend, and it comes back corrupted on next run.
3 2013-09-28 09:14:35 <warren> so there is a different mac issue than what we know...
4 2013-09-28 12:52:24 <sphered> I need to send a percentage of coins that are mined to a list of accounts. I know this can be done by modifying the source code but how? Any help would be appreciated.
5 2013-09-28 15:02:48 <_Sam--> hi, what is the best profitability calculator you like to use for mining hardware?
6 2013-09-28 15:03:54 <pankkake> http://mining.thegenesisblock.com/
7 2013-09-28 15:04:25 <_Sam--> thank you.
8 2013-09-28 15:13:02 <sphered> some development help needed. I need to send a percentage of coins that are mined to a list of accounts. I know this can be done by modifying the source code but how? using litecoin sourcecode
9 2013-09-28 15:13:46 <Graet> try #litecoin-dev
10 2013-09-28 15:14:02 <sphered> thanks
11 2013-09-28 15:14:49 <_Sam--> im trying to setup a stratum server for my own miners and have the source and have it running on ec2 cloud server, but where can i find more details about the options and what the accepted values are, or what some of them even mean, or do
12 2013-09-28 15:16:12 <_Sam--> and Graet : how did your 650 turn out?
13 2013-09-28 15:16:59 <Graet> good sam, needs a new chain and a tune, but thats the nature of neglecting bikes ;)
14 2013-09-28 15:17:10 <_Sam--> is it a 530 or 525 chain?
15 2013-09-28 15:17:43 <Graet> no idea, sorry, i'm a bit busy with my pol atm, and its very offtopic for -dev
16 2013-09-28 15:17:50 <Graet> pool*
17 2013-09-28 15:17:57 <_Sam--> sorry, which pool is yours?
18 2013-09-28 15:18:34 <Graet> Ozcoin
19 2013-09-28 15:18:59 <_Sam--> nice work.
20 2013-09-28 15:20:01 <_Sam--> send me a settings.py besides sample so i can understand more :P
21 2013-09-28 15:20:58 <Graet> cant atm, in middle of migrating servers - thats what i'm busy with]
22 2013-09-28 15:22:43 <_Sam--> i understand. if i can help you in any way, lmk. i have 18 years of senior linux experience.
23 2013-09-28 15:23:01 <_Sam--> including HA services.
24 2013-09-28 15:26:51 <Graet> thanks, we have a good team already :)
25 2013-09-28 15:35:08 <jgarzik> ACTION continues to be tempted to create a GUI make-your-own-altchain tool ;-)
26 2013-09-28 15:35:46 <jgarzik> maybe a checkbox for pre-mining
27 2013-09-28 15:35:59 <Graet> lol
28 2013-09-28 15:36:07 <chmod755> lol
29 2013-09-28 15:36:25 <chmod755> jgarzik, make it closed-source and sell it for BTC
30 2013-09-28 15:37:53 <jgarzik> Radio buttons for hashing algorithm: ( ) SHA256 ( ) scrypt ( ) obscure scheme based on prime numbers that no one feels is valuable or fast
31 2013-09-28 15:38:26 <chmod755> md4 (x)
32 2013-09-28 15:39:16 <sipa> jgarzik: ( ) all of the above
33 2013-09-28 15:39:23 <jgarzik> Money supply model: ( ) declining new supply, modelled after resource extraction, ( ) constant 50 coins per block ( ) Ben Bernanke
34 2013-09-28 15:39:37 <chmod755> hahaha
35 2013-09-28 15:39:43 <sipa> ( ) random
36 2013-09-28 15:40:01 <pankkake> I'm actually thinking of starting an altcoin named after Ben Bernanke
37 2013-09-28 15:40:10 <Diablo-D3> honestly
38 2013-09-28 15:40:10 <sipa> BBC!
39 2013-09-28 15:40:13 <Diablo-D3> if I made bitcoin
40 2013-09-28 15:40:13 <pankkake> the Bernankoin
41 2013-09-28 15:40:19 <Diablo-D3> I would have kept 50 coins per block
42 2013-09-28 15:40:33 <lianj> Diablo-D3: forever?
43 2013-09-28 15:41:02 <Diablo-D3> yes
44 2013-09-28 15:41:03 <pankkake> the number of coins per block would be directly related to difficulty
45 2013-09-28 15:41:26 <Diablo-D3> the rate of inflation would not remain linear
46 2013-09-28 15:41:33 <lianj> Diablo-D3: that would be a totally different coin then. and maybe had never taken off
47 2013-09-28 15:41:42 <sipa> and maybe it would
48 2013-09-28 15:41:46 <Diablo-D3> lianj: nope think about it
49 2013-09-28 15:41:53 <Diablo-D3> 50 coins would be big in the beginning
50 2013-09-28 15:41:57 <Diablo-D3> but very small right now
51 2013-09-28 15:41:57 <lianj> sipa: yea, maybe
52 2013-09-28 15:42:01 <sipa> we don't actually know what property or properties in bitcoin mde it take off
53 2013-09-28 15:42:11 <jgarzik> Diablo-D3, someone already created a 50-BTC-per-block-forever fork
54 2013-09-28 15:42:13 <Diablo-D3> the inflation in percent would reduce every block
55 2013-09-28 15:42:18 <Diablo-D3> jgarzik: yeah, so?
56 2013-09-28 15:42:19 <jgarzik> Diablo-D3, same chain genesis, same keys, etc.
57 2013-09-28 15:42:27 <Diablo-D3> jgarzik: wtf =/
58 2013-09-28 15:42:28 <Diablo-D3> why
59 2013-09-28 15:42:36 <lianj> sipa: i feel like having a fixed sum was one of the well at least selling points
60 2013-09-28 15:42:47 <jgarzik> Diablo-D3, they disagreed with the declining output
61 2013-09-28 15:42:49 <sipa> very early on, i think so
62 2013-09-28 15:42:53 <Diablo-D3> lianj: but its not a fixed sum if you use percentages
63 2013-09-28 15:42:56 <Diablo-D3> ;;blocks
64 2013-09-28 15:42:56 <sipa> right now, i'm less sure
65 2013-09-28 15:42:57 <gribble> 260592
66 2013-09-28 15:43:05 <Diablo-D3> we'd have 260592 * 50 coins atm
67 2013-09-28 15:43:28 <Diablo-D3> ;;calc 260592 * 50
68 2013-09-28 15:43:28 <gribble> 13029600
69 2013-09-28 15:43:29 <jgarzik> agree w/ sipa, RE early on
70 2013-09-28 15:43:38 <jgarzik> The main selling point today is "it works"
71 2013-09-28 15:43:45 <lianj> kinda :D
72 2013-09-28 15:43:55 <Diablo-D3> ;;calc 240000 * 50 + (260592 - 240000) * 25
73 2013-09-28 15:43:55 <gribble> 12514800
74 2013-09-28 15:44:06 <Diablo-D3> (I think I did that right)
75 2013-09-28 15:44:19 <jgarzik> it's the first global currency (no, gold doesn't count, you cannot transfer gold easily with full trust)
76 2013-09-28 15:44:19 <sipa> "total_amount" : 11764664.79571183
77 2013-09-28 15:44:42 <Diablo-D3> sipa: isnt yours involving lost coins or something
78 2013-09-28 15:45:07 <sipa> Diablo-D3: it includes overwritten coins and fees that were not claimed
79 2013-09-28 15:45:17 <sipa> but the effect of those is very small
80 2013-09-28 15:45:19 <Diablo-D3> yeah see, Im not calculating that
81 2013-09-28 15:45:26 <Diablo-D3> isnt it halved every 240k blocks?
82 2013-09-28 15:45:32 <sipa> a few 1000 BTC at most
83 2013-09-28 15:45:52 <lianj> 1176485000000000
84 2013-09-28 15:45:57 <jgarzik> sipa, BTW, been meaning to ask you: any last little bits of polish needed, now that prune-unspendables is merged?
85 2013-09-28 15:46:06 <Diablo-D3> so
86 2013-09-28 15:46:09 <sipa> jgarzik: such as?
87 2013-09-28 15:46:10 <Diablo-D3> if I made a coin
88 2013-09-28 15:46:10 <Diablo-D3> that
89 2013-09-28 15:46:16 <Diablo-D3> used scrypt or that new thing
90 2013-09-28 15:46:19 <Diablo-D3> used the knobs correctly
91 2013-09-28 15:46:24 <Diablo-D3> had an easy to use client
92 2013-09-28 15:46:32 <Diablo-D3> had a working business gateway
93 2013-09-28 15:46:54 <sipa> included a unicorn for every user
94 2013-09-28 15:46:55 <Diablo-D3> and allowed clusterable bitcoinds
95 2013-09-28 15:47:02 <Diablo-D3> (shared replicated databases)
96 2013-09-28 15:47:14 <jgarzik> sipa, anything. We had at one time talked about adding some flag fields to gettxoutset indicating pruning, and other small details. Just making a general poke, searching around the dusty corners of human memory for the possibility of any last details like that.
97 2013-09-28 15:47:35 <sipa> jgarzik: yeah, right now, that hash field is not reliable
98 2013-09-28 15:47:38 <Diablo-D3> blocks happened every 5 minutes
99 2013-09-28 15:47:46 <Diablo-D3> 50 coins per block forever
100 2013-09-28 15:47:53 <Diablo-D3> would people mine on it?
101 2013-09-28 15:47:55 <sipa> jgarzik: unless reindexed since upgrading
102 2013-09-28 15:47:58 <jgarzik> sipa, rewritten, any pruning-related to-do items like that unfinished, and needing to be done for 0.9.x?
103 2013-09-28 15:48:09 <jgarzik> sipa, maybe 0.9 should force reindex?
104 2013-09-28 15:48:13 <jgarzik> for all
105 2013-09-28 15:48:18 <sipa> jgarzik: not worth it imho
106 2013-09-28 15:48:29 <devthedevpro> Hi
107 2013-09-28 15:48:43 <devthedevpro> How's everyone doing?
108 2013-09-28 15:48:54 <sipa> jgarzik: maybe we could just write a "version last reindexed at"
109 2013-09-28 15:48:56 <Diablo-D3> oh oh oh
110 2013-09-28 15:48:57 <Diablo-D3> and
111 2013-09-28 15:48:59 <Diablo-D3> I could write it
112 2013-09-28 15:49:00 <Diablo-D3> in erlang
113 2013-09-28 15:49:02 <Diablo-D3> <3
114 2013-09-28 15:49:04 <jgarzik> Diablo-D3, huhwhuh? bitcoind's purpose is creating a shared replicated database :)
115 2013-09-28 15:49:11 <devthedevpro> Can someone help me with something?
116 2013-09-28 15:49:17 <chmod755> ACTION makes a new fork of bitcoin and calls it litecoin so everyone is confused as f***
117 2013-09-28 15:49:20 <Diablo-D3> jgarzik: yes, but the wallet.dat isnt clustered.
118 2013-09-28 15:49:23 <devthedevpro> I'm trying to start up a LTC pool. Not sure where to start.
119 2013-09-28 15:49:25 <sipa> devthedevpro: do 't ask to ask
120 2013-09-28 15:49:30 <jgarzik> Diablo-D3, right now we have a database with a replication factor > 20000x ;p
121 2013-09-28 15:49:47 <Diablo-D3> jgarzik: I know =P
122 2013-09-28 15:49:48 <devthedevpro> Trying to set up a P2Pool node
123 2013-09-28 15:49:51 <sipa> devthedevpro: ask in #litecoin-dev
124 2013-09-28 15:49:51 <sipa> or in #bitcoin-mining perhaps
125 2013-09-28 15:49:57 <Diablo-D3> jgarzik: but that seems insane for, say, a cluster for a bank
126 2013-09-28 15:49:58 <jgarzik> Diablo-D3, stop trying to use bitcoind as a wallet if you want huge, sharded, scaled, failing over key storage
127 2013-09-28 15:50:01 <devthedevpro> kk
128 2013-09-28 15:50:03 <sipa> or in #p2pool
129 2013-09-28 15:50:16 <devthedevpro> Thanks guys
130 2013-09-28 15:50:58 <jgarzik> IMO Big Companies will have several bitcoind "border routers", and use non-bitcoind key management that involves multiple computers and quite different schemes from a single-computer design
131 2013-09-28 15:51:13 <sipa> jgarzik: how is no-wallet-mode doing?
132 2013-09-28 15:51:20 <sipa> haven't folledup on that
133 2013-09-28 15:51:30 <jgarzik> sipa, I think it's ready, there may be one last nit
134 2013-09-28 15:51:37 <jgarzik> testing is going well by multiple reports
135 2013-09-28 15:51:40 <chmod755> what is no-wallet-mode??
136 2013-09-28 15:51:46 <Diablo-D3> jgarzik: exactly
137 2013-09-28 15:51:50 <Diablo-D3> jgarzik: so why not have that built in
138 2013-09-28 15:51:51 <devthedevpro> sipa: Those channels are all dead :
139 2013-09-28 15:51:53 <sipa> i'm pondering about merging libsecp256k1, only enabled in an ezperimental build that disables mining and wallet
140 2013-09-28 15:51:58 <sipa> devthedevpro: patience
141 2013-09-28 15:52:05 <Diablo-D3> btw, question
142 2013-09-28 15:52:06 <sipa> not everyone is awake all the time
143 2013-09-28 15:52:41 <Diablo-D3> do we consider elliptic curves still secure?
144 2013-09-28 15:52:52 <sipa> i consider secp256k1 secure
145 2013-09-28 15:53:04 <Diablo-D3> yeah but what if the NSA broke it
146 2013-09-28 15:53:14 <jgarzik> sipa, ah yes. All that remains is shed-painting over command line option naming and handling. Gavin wants -wallet=0 / -nowallet, which, meh but OK, just a bit annoying to overload a filename param IMHO.
147 2013-09-28 15:53:18 <jgarzik> for no-wallet
148 2013-09-28 15:53:30 <jgarzik> everything else is tested and working
149 2013-09-28 15:53:33 <jgarzik> and ready for merge
150 2013-09-28 15:54:09 <jgarzik> in fact, I could probably pull the "add !pwalletMain checks" patch right now
151 2013-09-28 15:54:11 <sipa> jgarzik: i'll add some paint myself to the discussion then
152 2013-09-28 15:54:27 <Diablo-D3> sipa: like, okay, lets say I was making a new coin
153 2013-09-28 15:54:31 <Diablo-D3> why wouldnt I use ed25519 instead
154 2013-09-28 15:55:16 <sipa> Diablo-D3: i think i'd use ed25519 right now, indeed
155 2013-09-28 15:55:26 <jgarzik> sipa, ACK re libsecp256k1. I want to get it in the repo and build, even if disabled-by-default with safety checks (like those you mention)
156 2013-09-28 15:55:36 <jgarzik> *repo and build system
157 2013-09-28 15:55:39 <Diablo-D3> ACTION ponders.
158 2013-09-28 15:55:49 <Diablo-D3> sipa: I wonder if its too late to reinvent bitcoin
159 2013-09-28 15:56:01 <sipa> Diablo-D3: i don't know
160 2013-09-28 15:56:09 <jgarzik> Diablo-D3, of course not
161 2013-09-28 15:56:13 <Diablo-D3> I wish I knew how parts of it worked
162 2013-09-28 15:56:14 <sipa> Diablo-D3: there are many actually significant technical improvements possible
163 2013-09-28 15:56:21 <Diablo-D3> like, how do I actually send coins to someone else
164 2013-09-28 15:56:29 <Diablo-D3> I know the address is a hash of a public key
165 2013-09-28 15:56:36 <sipa> that are hard to inpossible to include into bitcoin
166 2013-09-28 15:56:46 <jgarzik> Diablo-D3, one possible endgame is that current bitcoin becomes a super-strong-mining-power chain with few transactions, timestamping and securing another chain with slightly different features.
167 2013-09-28 15:57:16 <Diablo-D3> but how do I, with the private key of that public key, make a transaction with those coins that were sent to me
168 2013-09-28 15:57:21 <jgarzik> in that case, "another chain" can have Lamport sigs or whatever people really like better
169 2013-09-28 15:57:31 <jgarzik> see /title links, hard fork wish list etc
170 2013-09-28 15:57:56 <sipa> gmaxwell has a wiki page with more exotic improvement ideas
171 2013-09-28 15:58:02 <Diablo-D3> jgarzik: itym /topic
172 2013-09-28 15:58:47 <jgarzik> sipa, quite seriously, I think it would be a nice exercise to run a testnet_x, with a test coin that includes such "we would do this, if we could clean slate" features.
173 2013-09-28 15:59:09 <jgarzik> just reset, when consensus says testnet_x should include a radical change
174 2013-09-28 15:59:19 <jgarzik> keep a living, working record of good ideas
175 2013-09-28 15:59:34 <Diablo-D3> how does bitcoin use signatures to prove ownership of an address
176 2013-09-28 15:59:46 <jgarzik> and provide a good base for anyone with a legitimate alt-coin use
177 2013-09-28 15:59:54 <jgarzik> (IMO, local coins are one such use)
178 2013-09-28 16:00:33 <sipa> Diablo-D3: by signing the spending transaction's hash with the private key corresponding to it
179 2013-09-28 16:00:53 <sipa> Diablo-D3: (after some non-trivial modifications)
180 2013-09-28 16:00:59 <Diablo-D3> sipa: yeah, but how does that prove it.
181 2013-09-28 16:01:13 <sipa> Diablo-D3: learn digital signature schemes
182 2013-09-28 16:01:23 <Diablo-D3> I tried that once, its confusing as hell
183 2013-09-28 16:01:34 <sipa> that's exactly what digital signatures do
184 2013-09-28 16:01:47 <Diablo-D3> I dont understand the math
185 2013-09-28 16:01:58 <sipa> well then either skip it, or learn it
186 2013-09-28 16:02:57 <Diablo-D3> and what sort of non-trivial modification
187 2013-09-28 16:03:02 <devthedevpro> Does anyone have a spare VPS?
188 2013-09-28 16:06:03 <Diablo-D3> sipa: so, if I know someone's address, and they make a tx, there is math where I can verify they made the tx?
189 2013-09-28 16:06:13 <sipa> Diablo-D3: actually, you don't
190 2013-09-28 16:06:25 <sipa> Diablo-D3: the full node implementation just verifies all signatures
191 2013-09-28 16:06:42 <sipa> which means the wallet just doesn't even get to see them if they're invalid
192 2013-09-28 16:06:50 <sipa> (or have some invalid transaction in their history)
193 2013-09-28 16:07:25 <Diablo-D3> sipa: see, this is what I dont get, how can bitcoin be secure if individual tx cannot be validated
194 2013-09-28 16:07:43 <sipa> they can be
195 2013-09-28 16:08:09 <sipa> but as we already have a system that verifies *everything*, there is no need to individually verify things
196 2013-09-28 16:08:25 <Diablo-D3> sipa: but I can store tx inside of the chain that are garbage
197 2013-09-28 16:08:41 <sipa> no, that would make the chain invalid
198 2013-09-28 16:08:47 <sipa> dude, how long have you been around?
199 2013-09-28 16:08:55 <Diablo-D3> so how are people storing plain text in the chain?
200 2013-09-28 16:08:57 <Diablo-D3> I know people have done it
201 2013-09-28 16:09:07 <sipa> they store it inside amounts
202 2013-09-28 16:09:18 <sipa> or create addresses with data in them, and spend to it
203 2013-09-28 16:09:27 <sipa> those are technically completely valid transactions
204 2013-09-28 16:09:31 <Diablo-D3> but if the amount cannot be valid, how can it be a valid tx
205 2013-09-28 16:09:38 <sipa> ...
206 2013-09-28 16:09:50 <sipa> why would the amount not be "valid" ?
207 2013-09-28 16:09:59 <sipa> they just store some data in the last decimals
208 2013-09-28 16:10:06 <sipa> (in binary)
209 2013-09-28 16:10:14 <Diablo-D3> yes, but wouldnt they be burning coins then?
210 2013-09-28 16:10:17 <sipa> they are
211 2013-09-28 16:10:24 <Diablo-D3> why =/
212 2013-09-28 16:10:30 <sipa> because it's cool, i guess
213 2013-09-28 16:10:47 <sipa> or construct an address that is not actually the hash of a public ket
214 2013-09-28 16:10:59 <sipa> nobody knows the private key doesn't exist, so those outputs cannot be spent
215 2013-09-28 16:11:06 <sipa> but the transactions crediting them are valid
216 2013-09-28 16:11:07 <Diablo-D3> yeah but that still burns coins
217 2013-09-28 16:11:09 <sipa> yes
218 2013-09-28 16:11:17 <chmod755> jgarzik, just heard that the bootstrap.dat can be downloaded at ~25MB/s - that's really nice
219 2013-09-28 16:11:34 <Diablo-D3> sipa: so why not require something where you have to announce a new address before being able to send to it
220 2013-09-28 16:11:45 <sipa> Diablo-D3: how would that help?
221 2013-09-28 16:11:55 <Diablo-D3> good point =/
222 2013-09-28 16:11:57 <sipa> Diablo-D3: it would require *everyone* to hear your broadcast
223 2013-09-28 16:12:05 <sipa> which is an impossibly hard thing to guarantee
224 2013-09-28 16:12:10 <sipa> unless you make it part of the chain
225 2013-09-28 16:12:13 <Diablo-D3> well, it'd be encoded in the chain
226 2013-09-28 16:12:13 <sipa> which means extra bloat
227 2013-09-28 16:12:17 <Diablo-D3> but that would be abusable
228 2013-09-28 16:12:22 <sipa> bingo
229 2013-09-28 16:12:30 <sipa> gmaxwell actually has a scheme that would prevent this
230 2013-09-28 16:12:42 <sipa> would it requires different address structure
231 2013-09-28 16:12:46 <Diablo-D3> sipa: is it possible to produce a proof that your signature is real?
232 2013-09-28 16:15:47 <Diablo-D3> sipa: sorry, I mean, is it possible to produce a proof that your address is real
233 2013-09-28 16:18:17 <Luke-Jr> define real
234 2013-09-28 16:18:29 <sipa> Diablo-D3: yes
235 2013-09-28 16:18:32 <Diablo-D3> Luke-Jr: not a fictional address used for burning coins
236 2013-09-28 16:18:40 <sipa> you can prove that you own the corresponding private key
237 2013-09-28 16:18:54 <Diablo-D3> sipa: so why dont we require that?
238 2013-09-28 16:19:04 <sipa> Diablo-D3: because it means bloat
239 2013-09-28 16:19:15 <Diablo-D3> how much bloat
240 2013-09-28 16:19:35 <Luke-Jr> it also require disclosing your pubkey..
241 2013-09-28 16:19:43 <Luke-Jr> sipa: only network bloat
242 2013-09-28 16:19:52 <Diablo-D3> Luke-Jr: I dont think it does
243 2013-09-28 16:20:23 <sipa> Luke-Jr: it would mean significantly reducing the script language's flexibility
244 2013-09-28 16:20:40 <Luke-Jr> sipa: well, it'd just be an IsStandard thing in any case
245 2013-09-28 16:20:47 <sipa> right
246 2013-09-28 16:20:58 <sipa> that's certainly easier
247 2013-09-28 16:21:12 <Luke-Jr> IMO just verifying it's a hash, and not arbitrary data, is good enough
248 2013-09-28 16:21:30 <sipa> right
249 2013-09-28 16:21:45 <Diablo-D3> so how would I do that?
250 2013-09-28 16:21:49 <Diablo-D3> is there a formula for that?
251 2013-09-28 16:21:59 <Luke-Jr> Diablo-D3: new address format is SHA256(pubkey) instead of RIPEMD(SHA256(pubkey))
252 2013-09-28 16:22:02 <sipa> but it does mean you need to be able to identify every data element inside the script that may be at some point used as a pubkey hash
253 2013-09-28 16:22:22 <Luke-Jr> Diablo-D3: you still use RIPEMD in blocks/transactions, but you must prove it's a hash by providing SHA256(pubkey) on the netwrok
254 2013-09-28 16:22:25 <sipa> which is quite hard in combination with P2SH, iirc, as it means also looking inside the embedded script in the spending transaction
255 2013-09-28 16:22:34 <sipa> but inside IsStandard(), that's doable
256 2013-09-28 16:22:38 <Diablo-D3> Luke-Jr: yeah, but how does that help
257 2013-09-28 16:22:50 <Luke-Jr> Diablo-D3: miners refuse to include transactions unless they know the preimage
258 2013-09-28 16:23:08 <Diablo-D3> that only proves the ripemd160 is real
259 2013-09-28 16:23:12 <Diablo-D3> not the sha256
260 2013-09-28 16:23:13 <sipa> indeed
261 2013-09-28 16:23:25 <sipa> but that does mean you can't use it for arbitrary data storage already
262 2013-09-28 16:23:26 <Luke-Jr> Diablo-D3: yes, that's good enough for all cases I'm aware of now
263 2013-09-28 16:23:32 <sipa> it doesn't mean you're not burning coins
264 2013-09-28 16:23:36 <sipa> but who cares about that
265 2013-09-28 16:23:40 <Diablo-D3> Luke-Jr: so I can now spam fake sha256s
266 2013-09-28 16:23:42 <Diablo-D3> for data storage
267 2013-09-28 16:23:42 <Luke-Jr> you can't prove I'm not burning coins..
268 2013-09-28 16:23:44 <Diablo-D3> without burning coins
269 2013-09-28 16:23:50 <Luke-Jr> Diablo-D3: the SHA256 isn't stored.
270 2013-09-28 16:23:52 <sipa> Diablo-D3: plus those fake sha256s aren't stored
271 2013-09-28 16:24:00 <sipa> so yes, it's spam, but not stored spam
272 2013-09-28 16:24:05 <Diablo-D3> sipa: okay so
273 2013-09-28 16:24:06 <Diablo-D3> uh
274 2013-09-28 16:24:12 <Diablo-D3> if I send a tx
275 2013-09-28 16:24:22 <Diablo-D3> that uses a ripemd160 address
276 2013-09-28 16:24:35 <sipa> you broadcast (TX(ripemd(sha(pubkey)),sha(pubkey))
277 2013-09-28 16:24:37 <Diablo-D3> I must also have sent a sha256 of my private key recently?
278 2013-09-28 16:24:42 <Diablo-D3> er
279 2013-09-28 16:24:43 <sipa> of your pubkey
280 2013-09-28 16:24:44 <Diablo-D3> public key
281 2013-09-28 16:24:49 <sipa> which you always have
282 2013-09-28 16:25:11 <Luke-Jr> Diablo-D3: RIPEMD160 addresses would be unsupported
283 2013-09-28 16:25:24 <sipa> oh
284 2013-09-28 16:25:31 <Diablo-D3> Luke-Jr: why?
285 2013-09-28 16:25:32 <sipa> you need to have the pubkey of the recipient, indeed
286 2013-09-28 16:25:39 <Luke-Jr> Diablo-D3: because you don't have the preimage for it
287 2013-09-28 16:25:51 <Diablo-D3> sipa: yeah exactly, its why I dont get it
288 2013-09-28 16:25:57 <Luke-Jr> Diablo-D3: the address would encode the SHA256, and the sender would RIPEMD160 that, while it has the preimage
289 2013-09-28 16:26:00 <Diablo-D3> Luke-Jr: ahh
290 2013-09-28 16:26:12 <Diablo-D3> so it'd make addresses longer
291 2013-09-28 16:26:14 <Luke-Jr> yep
292 2013-09-28 16:26:25 <Luke-Jr> .. to do it with bitcoin
293 2013-09-28 16:26:28 <Diablo-D3> so why dont I ripemd160 the ripemd160?
294 2013-09-28 16:26:39 <Luke-Jr> an altcoin could just do RIPEMD160(RIPEMD160(SHA256(pubkey)
295 2013-09-28 16:26:40 <Luke-Jr> right
296 2013-09-28 16:27:20 <Diablo-D3> but I'd need to know the preimage of the reciever
297 2013-09-28 16:27:46 <Diablo-D3> so to be able to send to someone, I'd need a pubkey preimage
298 2013-09-28 16:30:29 <Diablo-D3> sipa, Luke-Jr: am I wrong/
299 2013-09-28 16:30:44 <Luke-Jr> Diablo-D3: you'd need the preimage of the final hash
300 2013-09-28 16:30:54 <Luke-Jr> not necessarily the full pubkey
301 2013-09-28 16:31:02 <Diablo-D3> this is what Im talking about
302 2013-09-28 16:31:08 <Diablo-D3> how would I get such a thing
303 2013-09-28 16:31:20 <Luke-Jr> Diablo-D3: that's what you'd make your address format encode
304 2013-09-28 16:31:47 <Diablo-D3> but wouldn't I just be allowing people to put plaintext in the chain some way or another?
305 2013-09-28 16:32:03 <Luke-Jr> no
306 2013-09-28 16:32:13 <Luke-Jr> the address-encoded data would never be in the chain
307 2013-09-28 16:32:51 <Diablo-D3> so the address would be the encoded preimage?
308 2013-09-28 16:33:04 <sipa> indeed
309 2013-09-28 16:33:07 <Diablo-D3> but how does that prevent people from burning coins
310 2013-09-28 16:33:11 <sipa> it doesn't
311 2013-09-28 16:33:18 <sipa> but why do you care about others burning coins?
312 2013-09-28 16:33:25 <Diablo-D3> sipa: because its dumb =/
313 2013-09-28 16:33:31 <sipa> it's the same as if they'd decide to just throw away a privkey
314 2013-09-28 16:33:38 <Diablo-D3> meh
315 2013-09-28 16:33:42 <sipa> or even to just never spend any coins
316 2013-09-28 16:34:40 <Diablo-D3> how the hell did satoshi do all of this alone
317 2013-09-28 16:34:42 <Diablo-D3> its impossible
318 2013-09-28 16:34:52 <Diablo-D3> it had to be a team of people
319 2013-09-28 16:35:20 <Luke-Jr> â¦
320 2013-09-28 16:35:34 <Luke-Jr> Diablo-D3: what it prevents is people abusing the blockchain as data storage
321 2013-09-28 16:35:42 <Diablo-D3> Luke-Jr: true
322 2013-09-28 16:36:24 <Diablo-D3> so whats stored in the chain is ripmed160 of ripmed160(sha256(pubkey)) and the address is ripmed160(sha256(pubkey))?
323 2013-09-28 16:36:31 <jgarzik> I could see a single person doing a crappy Windows version of a fantastically thought through idea ;p
324 2013-09-28 16:36:33 <sipa> for example
325 2013-09-28 16:37:00 <Luke-Jr> who wants to lart "Hive" for displaying "from" addresses?
326 2013-09-28 16:37:14 <Luke-Jr> (or are those addresses at all? looks hexish!)
327 2013-09-28 16:37:48 <jgarzik> for the record I do think Satoshi is a single person. I also am probably the least interested in Satoshi-as-a-person. Reporters love to ask about it, but that's non-technical stuff and therefore boring and irrelevant.
328 2013-09-28 16:37:59 <Diablo-D3> sipa: so the tx is signed by the keypair, then bitcoind scans every tx for a hash it knows is a private key of it's?
329 2013-09-28 16:38:20 <Diablo-D3> er, sender keypair, reciever hash, respectively
330 2013-09-28 16:38:22 <sipa> Diablo-D3: every wallet knows which addresses it controls
331 2013-09-28 16:38:32 <Diablo-D3> sipa: exactly
332 2013-09-28 16:38:42 <sipa> Diablo-D3: whether it actually has the privkey is irrelevant (in case the wallet is encrypted, it doesn't actuallt have them)
333 2013-09-28 16:39:13 <Diablo-D3> I mean reciever hash of public key
334 2013-09-28 16:40:00 <Diablo-D3> sipa: see, what I dont get is, how can a new tx be created from outputs
335 2013-09-28 16:40:23 <sipa> ?
336 2013-09-28 16:40:35 <Diablo-D3> sipa: why cant I sign a tx stealing other peoples outputs
337 2013-09-28 16:40:35 <Luke-Jr> it's funny how the public seems to have this misconception of Satoshi as a programming wizard who wrote Bitcoin's original code exploit-free etc
338 2013-09-28 16:40:44 <sipa> Diablo-D3: because you don't have their private key?)
339 2013-09-28 16:40:56 <sipa> creating a signature requires a private key
340 2013-09-28 16:41:02 <Diablo-D3> sipa: but I dont need their private key if their isnt private data
341 2013-09-28 16:41:02 <sipa> validating it only requires a public key
342 2013-09-28 16:41:09 <sipa> ?
343 2013-09-28 16:41:18 <sipa> private keys are always private
344 2013-09-28 16:41:37 <Diablo-D3> well, isnt a tx out only previous tx + amount?
345 2013-09-28 16:42:11 <sipa> it's a list of previous outputs being consumed, and a list of new outputs being produced
346 2013-09-28 16:42:18 <Diablo-D3> sipa: yes
347 2013-09-28 16:42:27 <sipa> the ones consumed contain a signature with their private key
348 2013-09-28 16:42:38 <Diablo-D3> sipa: okay so
349 2013-09-28 16:42:40 <sipa> outputs contain a list of public keys being sent to (or hashes thereof)
350 2013-09-28 16:42:44 <Diablo-D3> lets say I have three addresses
351 2013-09-28 16:42:49 <Diablo-D3> X sends 3 to A
352 2013-09-28 16:42:53 <Diablo-D3> Y sends 2 to B
353 2013-09-28 16:42:57 <Diablo-D3> Z sends 1 to C
354 2013-09-28 16:43:02 <Diablo-D3> I now have 6 coins
355 2013-09-28 16:43:11 <Diablo-D3> because A, B, and C are my addresses
356 2013-09-28 16:43:13 <sipa> you have 3 coins
357 2013-09-28 16:43:23 <Diablo-D3> you know what I mean
358 2013-09-28 16:43:23 <sipa> with values 3 BTC, 2 BTC and 1 BTC respectively
359 2013-09-28 16:43:42 <Diablo-D3> my wallet's entirety is 3 tx worth 6 BTC
360 2013-09-28 16:43:47 <sipa> 3 txouts, yes
361 2013-09-28 16:43:59 <Diablo-D3> I then write a tx sending 6 BTC to someone else
362 2013-09-28 16:44:09 <Diablo-D3> how am I allowed to do that
363 2013-09-28 16:44:18 <sipa> your tx will have 3 inputs
364 2013-09-28 16:44:31 <sipa> referring to the 3 previous txouts that credited your addresses
365 2013-09-28 16:44:48 <sipa> the first one, spending X's 3 BTC sent to A, will contain a signature made with A's private key
366 2013-09-28 16:45:00 <sipa> the second one, spending Y's 2 BTC sent to B, will contain a signature made with B's private key
367 2013-09-28 16:45:13 <sipa> the third one, spending Z's 1 BTC sent to C, will contain a signature made with C's private key
368 2013-09-28 16:46:05 <sipa> anyone validating it, will look up the 3 previous txouts being spent
369 2013-09-28 16:46:12 <sipa> check that the public keys match
370 2013-09-28 16:46:19 <sipa> and that the signatures match those public keys
371 2013-09-28 16:46:29 <sipa> and that the sum of the values is not less than 6 BTC
372 2013-09-28 16:47:24 <Diablo-D3> sipa: okay, so each txout has a signature?
373 2013-09-28 16:47:31 <sipa> each txin has a signature
374 2013-09-28 16:47:34 <Diablo-D3> errr right
375 2013-09-28 16:47:38 <sipa> and each txin refers to exactly one previous txout
376 2013-09-28 16:48:05 <sipa> a txout is (amount, pubkey), a txin is (txouthash, txoutindex, signature)
377 2013-09-28 16:50:11 <sipa> *txhash
378 2013-09-28 16:51:19 <Diablo-D3> sipa: and I dont have to understand how signatures work, right? I can just treat them like a black box?
379 2013-09-28 16:51:30 <sipa> that's just ecdsa
380 2013-09-28 16:51:45 <Diablo-D3> yeah but I still dont understand the math on those
381 2013-09-28 16:51:52 <sipa> depending on whether you're implementing ecdsa yourself (don't!), you treat them as a black box or not :)
382 2013-09-28 16:52:09 <Diablo-D3> so
383 2013-09-28 16:52:11 <sipa> you need to understand what is being signed
384 2013-09-28 16:52:16 <Diablo-D3> to make bitcoin work
385 2013-09-28 16:52:18 <sipa> but you just hand it to the crypto library
386 2013-09-28 16:52:22 <Diablo-D3> 1) make a keypair
387 2013-09-28 16:52:43 <Diablo-D3> 2) ripmed160(sha256(pubkey)) is my address
388 2013-09-28 16:53:06 <Diablo-D3> 3) ripemd160(my address) is what people use for txout
389 2013-09-28 16:53:13 <Luke-Jr> no, that's only for an altcoin
390 2013-09-28 16:53:24 <Luke-Jr> for bitcoin, you need to use sha256(pubkey) for address
391 2013-09-28 16:53:28 <Diablo-D3> you know what I mean
392 2013-09-28 16:53:37 <sipa> or ripemd160(sha256(pubkey))
393 2013-09-28 16:54:10 <Diablo-D3> 4) then my bitcoind scans for tx's that use a txout target hash that is mine
394 2013-09-28 16:54:39 <Diablo-D3> 6) then I construct a tx with txins that reference txouts in previous tx's, and I sign per txin
395 2013-09-28 16:55:12 <sipa> you still need things like coin selection, dealing with change, ...
396 2013-09-28 16:55:24 <Diablo-D3> yes, but that part I already understand
397 2013-09-28 16:55:35 <sipa> integrating in something that will actually gives you transactions
398 2013-09-28 16:55:43 <sipa> (an SPV or full node, for example)
399 2013-09-28 16:55:54 <Diablo-D3> aaaaaaaaaaand
400 2013-09-28 16:56:00 <Diablo-D3> hrm
401 2013-09-28 16:56:18 <Diablo-D3> sipa: how do I verify a signature?
402 2013-09-28 16:56:29 <sipa> for every transaction
403 2013-09-28 16:56:33 <sipa> for each of its inputs
404 2013-09-28 16:56:37 <sipa> find the txout being spent
405 2013-09-28 16:56:45 <Diablo-D3> sipa: yes but how do I verify it without knowing the public key
406 2013-09-28 16:56:48 <sipa> you can't
407 2013-09-28 16:57:01 <sipa> your spending transaction contains the public key
408 2013-09-28 16:57:07 <sipa> if that isn't in the address
409 2013-09-28 16:57:17 <sipa> or in other words
410 2013-09-28 16:57:22 <Diablo-D3> the address is a hash of the public key
411 2013-09-28 16:57:34 <sipa> a bitcoin signature is (ecdsa signature, ecdsa pubkey)
412 2013-09-28 16:57:43 <sipa> a bitcoin pubkey is (hash(ecdsa pubkey))
413 2013-09-28 16:57:54 <Diablo-D3> so...
414 2013-09-28 16:58:05 <Diablo-D3> why are we hashing pubkeys if we need to disclose the real pubkey anyhow?
415 2013-09-28 16:58:19 <sipa> originally, because it made addresses shorter
416 2013-09-28 16:58:33 <sipa> but it has other advantages, like making the UTXO set smaller
417 2013-09-28 16:58:42 <sipa> and having some marginally better cryptographic properties
418 2013-09-28 16:58:57 <Diablo-D3> but if theres a flaw in ecdsa
419 2013-09-28 16:59:05 <sipa> then we're doomed
420 2013-09-28 16:59:07 <Diablo-D3> where I can produce privkeys from pubkeys
421 2013-09-28 16:59:09 <sipa> doomed!
422 2013-09-28 16:59:12 <Diablo-D3> well
423 2013-09-28 16:59:14 <Diablo-D3> fuck =/
424 2013-09-28 16:59:36 <Diablo-D3> sipa: okay so
425 2013-09-28 16:59:39 <sipa> however, such an evolution is very unlikely to happen instantaeously
426 2013-09-28 16:59:44 <Diablo-D3> this makes addresses smaller BUT
427 2013-09-28 16:59:48 <Diablo-D3> it doesnt make signatures smaller
428 2013-09-28 17:00:12 <Diablo-D3> since addresses take up 160, but signatures take up whatever the signature format uses
429 2013-09-28 17:00:29 <Diablo-D3> so if I used lamport signatures
430 2013-09-28 17:00:35 <sipa> meaning we'll see cryptographic breaks of ECDSA that reduce its security, without posing a practical danger, before there is an actual exploit
431 2013-09-28 17:00:37 <Diablo-D3> txins would be fucking massive
432 2013-09-28 17:00:41 <sipa> yes
433 2013-09-28 17:00:56 <Diablo-D3> also, ECDSA may already be 100% compromised
434 2013-09-28 17:01:00 <Diablo-D3> because of the NSA
435 2013-09-28 17:01:45 <Diablo-D3> sipa: so
436 2013-09-28 17:01:48 <Diablo-D3> whats more damaging
437 2013-09-28 17:01:52 <Diablo-D3> spamming tx to people
438 2013-09-28 17:01:56 <Diablo-D3> er nm
439 2013-09-28 17:01:59 <Diablo-D3> its the same thing
440 2013-09-28 17:02:28 <Diablo-D3> so if I use lamport sigs
441 2013-09-28 17:02:31 <Diablo-D3> say goodbye to my harddrive?
442 2013-09-28 17:02:33 <sipa> jgarzik: jgarzik@bitpay.com seems to bounce
443 2013-09-28 17:02:49 <jgarzik> GRUMBLE
444 2013-09-28 17:03:03 <jgarzik> Google Google Google
445 2013-09-28 17:03:10 <Diablo-D3> sipa: is there a secure signature that is computationally intensive?
446 2013-09-28 17:03:18 <sipa> jgarzik: stephen@bitpay.com too
447 2013-09-28 17:04:09 <jgarzik> sipa, are you directly emailing, or looking at listman bounces?
448 2013-09-28 17:04:39 <jgarzik> sipa, trying to figure out if you have a trace that does not go through a mailing list remailer
449 2013-09-28 17:04:43 <Diablo-D3> sipa: Im assuming no?
450 2013-09-28 17:04:44 <jgarzik> s/trace/bounce/
451 2013-09-28 17:05:01 <sipa> jgarzik: bounce from bitcoin-dev mailinglist
452 2013-09-28 17:05:12 <sipa> jgarzik: forwarded you the bounce notification
453 2013-09-28 17:05:15 <sipa> in case you don
454 2013-09-28 17:05:19 <sipa> 't get it yourself
455 2013-09-28 17:05:51 <jgarzik> got your forward, and related, just had to re-enable mailing list membership due to listman complaining
456 2013-09-28 17:06:24 <sipa> Diablo-D3: use a computationally intensive message hash?
457 2013-09-28 17:06:32 <sipa> but why?
458 2013-09-28 17:06:48 <Diablo-D3> sipa: so people cant spam txes
459 2013-09-28 17:06:52 <jgarzik> 451 4.3.0 try again. t10si3441343vcr.35 - gsmtp:
460 2013-09-28 17:06:52 <jgarzik> . Please
461 2013-09-28 17:06:52 <jgarzik> retry timeout exceeded
462 2013-09-28 17:06:54 <Diablo-D3> and fill up people's hds
463 2013-09-28 17:07:01 <jgarzik> looks like Google ain't werkin
464 2013-09-28 17:07:11 <sipa> heh
465 2013-09-28 17:08:49 <jgarzik> exmulti.com never had this problem, but bitpay.com does. both used Google Apps hosted solution.
466 2013-09-28 17:08:58 <jgarzik> humdinger.
467 2013-09-28 17:09:04 <sipa> has it happened before?
468 2013-09-28 17:24:24 <Diablo-D3> sipa: okay so, the txin of a tx contains a tx hash and the output index of that tx?
469 2013-09-28 17:27:48 <Diablo-D3> sipa: on top of the sig stuff I mean
470 2013-09-28 17:27:58 <elevatioN> i have bitcoins for sale, cash in td bank
471 2013-09-28 17:28:04 <elevatioN> omg sorry wrong room!
472 2013-09-28 17:28:07 <elevatioN> dont kick me =)
473 2013-09-28 17:43:39 <Diablo-D3> https://www.evernote.com/shard/s248/sh/ee8aff03-2ecf-41b8-b6ee-63af30dd08c4/8031998db2bce43c3a02722b2c8aa217
474 2013-09-28 17:43:48 <Diablo-D3> sipa, Luke-Jr: did I get anything wrong/
475 2013-09-28 17:44:50 <michagogo> cloud|Diablo-D3: I suggest you read the Script article and the Transaction article on the wiki
476 2013-09-28 17:45:01 <Diablo-D3> michagogo|cloud: nope
477 2013-09-28 17:45:17 <michagogo> cloud|Hmm?
478 2013-09-28 17:45:18 <Diablo-D3> Im not interested in reviving scripts for the altcoin Im never going to write
479 2013-09-28 17:45:30 <Diablo-D3> and Ive read the original satoshi paper ;)
480 2013-09-28 17:45:45 <sipa> Diablo-D3: in that case, i'm not interested in helping you write an altcoin :)
481 2013-09-28 17:45:57 <Diablo-D3> sipa: good, because Im never going to write it =P
482 2013-09-28 17:46:00 <michagogo> cloud|Diablo-D3: The Script article also gives examples
483 2013-09-28 17:46:04 <sipa> i assumed as muc
484 2013-09-28 17:46:14 <michagogo> cloud|Explains what the verification does
485 2013-09-28 17:46:17 <Diablo-D3> sipa: if I write one, its just to test out new ideas that may someday make it back into bitcoin
486 2013-09-28 17:46:33 <michagogo> cloud|Also, take a transaction from the network and getrawtransaction txid 1
487 2013-09-28 17:47:20 <Diablo-D3> sipa: unlike, say, the 9000 altcoins that just replaced the genesis block and the network port and did NOTHING ELSE
488 2013-09-28 17:47:37 <Diablo-D3> sipa: or the ones that did nothing else, but swapped out sha256 in mining for scrypt
489 2013-09-28 17:47:45 <Diablo-D3> sipa: or the ones that did nothing else and just made blocks shorter
490 2013-09-28 17:50:47 <jgarzik> ACTION tries to recall gmaxwell's nickname for amiller
491 2013-09-28 17:51:24 <jgarzik> "one of the world's experts in authenticated data structures"?
492 2013-09-28 17:51:29 <jgarzik> or was it provable data structures
493 2013-09-28 17:52:07 <jgarzik> ACTION is trying to find a generic English phrase for bitcoin's data structure use
494 2013-09-28 17:53:08 <jgarzik> provable sounds more correct than authenticated
495 2013-09-28 18:10:47 <sipa> jgarzik: authenticated, afaik
496 2013-09-28 18:11:05 <sipa> i'm not sure what a provable data structure would be; data doesn't claim anything
497 2013-09-28 18:11:27 <sipa> it's authenticated in that you can use a short authenticated code that commits to the data
498 2013-09-28 18:11:56 <jgarzik> sipa, the logic: you can prove it valid, or not. authenticated implies to me a security procedure invoked by a human.
499 2013-09-28 18:12:21 <jgarzik> in the sense of a math "proof"
500 2013-09-28 18:13:40 <sipa> jgarzik: hmm, i think i ascribe different meanings to those terms, but who am i :)
501 2013-09-28 18:18:57 <swulf--> sipa: is there a trivial way to disable blockchain downloading in bitcoind?
502 2013-09-28 18:19:24 <michagogo> cloud|swulf--: why?
503 2013-09-28 18:19:43 <michagogo> cloud|Not without source editing and recompiling.
504 2013-09-28 18:19:48 <sipa> swulf--: -connect=0.0.0.0 :P
505 2013-09-28 18:19:48 <swulf--> that's fine
506 2013-09-28 18:19:53 <michagogo> cloud|(Or having no connections)
507 2013-09-28 18:20:15 <swulf--> Basically, I'd just like a node to be running so I can push transactions
508 2013-09-28 18:20:22 <michagogo> cloud|Ah
509 2013-09-28 18:20:22 <sipa> ah
510 2013-09-28 18:20:33 <michagogo> cloud|Stick a return somewhere
511 2013-09-28 18:20:42 <michagogo> cloud|(And ignore michagogo)
512 2013-09-28 18:20:42 <michagogo> swulf--: As the topic used to say, what are you trying to do?
513 2013-09-28 18:20:46 <swulf--> michagogo: I can track down that 'somewhere', but perhaps you guys can save me some time :)
514 2013-09-28 18:20:57 <michagogo> cloud|swulf--: well, actually...
515 2013-09-28 18:21:08 <michagogo> cloud|You'll also need to patch validation
516 2013-09-28 18:21:09 <sipa> just never send a getdata
517 2013-09-28 18:21:26 <sipa> and indeed, usually sending a transactions means just pushing it to your local mempool
518 2013-09-28 18:21:29 <michagogo> cloud|You need the utxo set to validate the transaction
519 2013-09-28 18:21:36 <sipa> which will fail if you don't have an up-to-date chain
520 2013-09-28 18:21:38 <swulf--> micha: sendrawtransaction will do some tx validation before sending?
521 2013-09-28 18:21:41 <michagogo> cloud|Yes
522 2013-09-28 18:21:42 <sipa> yes
523 2013-09-28 18:21:46 <swulf--> ok, easy to fix
524 2013-09-28 18:21:52 <michagogo> It behaves as idf the tx came over the network
525 2013-09-28 18:21:55 <michagogo> if*
526 2013-09-28 18:22:08 <swulf--> fwiw, this is to push transactions to litecoin, not bitcoin
527 2013-09-28 18:22:11 <michagogo> Maybe not *that* easy
528 2013-09-28 18:22:15 <michagogo> Ahhh
529 2013-09-28 18:22:18 <swulf--> litecoin doesn't seem to have any push pages for transactions
530 2013-09-28 18:22:21 <michagogo> SO that's why you can't run a lite node
531 2013-09-28 18:22:32 <michagogo> (unless those exist for litecoin)
532 2013-09-28 18:22:43 <sipa> push pages?
533 2013-09-28 18:22:48 <michagogo> like pushtx
534 2013-09-28 18:22:50 <swulf--> sipa: like blockchain.info/pushtx
535 2013-09-28 18:22:51 <sipa> ?
536 2013-09-28 18:22:54 <sipa> oh
537 2013-09-28 18:22:55 <sipa> meh :)
538 2013-09-28 18:23:05 <swulf--> yeah..
539 2013-09-28 18:24:05 <swulf--> Easy: AlreadyHave() can return true for MSG_BLOCK always
540 2013-09-28 18:24:11 <michagogo> swulf--: You could always just take the network part of the code and use that to independantly fetch a list of peers from a dnsseed, connect to several, send a tx, and disconnect
541 2013-09-28 18:24:36 <michagogo> IMO running all of bitcoind's code is overkill
542 2013-09-28 18:24:41 <sipa> yeah
543 2013-09-28 18:24:50 <sipa> some of the bitcoin network libraries may be easier
544 2013-09-28 18:25:14 <sipa> connect to 8, broadcast to 4, wait until you here it from any of the other 4
545 2013-09-28 18:25:14 <swulf--> yeah, probably
546 2013-09-28 18:25:19 <sipa> *hear
547 2013-09-28 20:11:32 <Diablo-D3> goddamnit real life
548 2013-09-28 20:12:13 <Diablo-D3> sipa: so how many people would want me dead if my newcoin did not have scripts
549 2013-09-28 20:13:06 <michagogo> Diablo-D3: What would you do instead?
550 2013-09-28 20:19:22 <Diablo-D3> michagogo: exactly one tx type
551 2013-09-28 20:19:25 <Diablo-D3> you send money to someone else
552 2013-09-28 20:19:38 <michagogo> Well
553 2013-09-28 20:19:52 <michagogo> That takes away a huge amount of flexibility
554 2013-09-28 20:19:55 <michagogo> no multisig, etc
555 2013-09-28 20:20:01 <Diablo-D3> that no one is using anyhow
556 2013-09-28 20:20:13 <michagogo> false
557 2013-09-28 20:20:20 <michagogo> people are starting to use that more now
558 2013-09-28 20:20:51 <michagogo> I know shesek made something that mostly works that makes multisig-based dispute mediation work pretty easily
559 2013-09-28 20:23:33 <sipa> if i'd create an altcoin now, it'd be P2SH only
560 2013-09-28 20:23:44 <sipa> so you'd only have one type of address
561 2013-09-28 20:24:58 <sipa> with addresses that encode SHA256(script), and txouts that contain SHA256^2(script)
562 2013-09-28 20:27:25 <sipa> (which also means block explorer-like sites can't reverse-engineer an address from the blockcain, which is perfect)
563 2013-09-28 20:27:29 <warren> It sounds like we have different mac corruption issue
564 2013-09-28 20:27:44 <gmaxwell> sipa: I'd leave a hashtype flag in it for paranoia against future hash weaknessses. :P But yea.
565 2013-09-28 20:27:58 <michagogo> so, p2shh?
566 2013-09-28 20:44:52 <BlueMatt> sipa: ping
567 2013-09-28 20:45:40 <sipa> BlueMatt: pong
568 2013-09-28 20:47:10 <BlueMatt> sipa: working on getting headersfirst+compariontool working...
569 2013-09-28 20:47:24 <BlueMatt> cant get it to pass b12/13/14
570 2013-09-28 20:47:39 <sipa> which does what , exactly?
571 2013-09-28 20:47:58 <sipa> there may of course be a bug in headersfirst
572 2013-09-28 20:48:18 <BlueMatt> for some reason when I send it an inv of b13 before b12, it requests headers for b12+b13 (as it should) but doesnt request the actual block b12 and then somehow accepts b13 without b12...
573 2013-09-28 20:48:27 <BlueMatt> then never requests b14 when it gets an inv for it
574 2013-09-28 20:49:02 <BlueMatt> I havent actually read headersfirst yet, but thats what appears to be happening from debug.log
575 2013-09-28 20:49:27 <sipa> you have that log somewhere?
576 2013-09-28 20:50:18 <sipa> it should request b13 before b12, i expect
577 2013-09-28 20:50:24 <sipa> but request b12 too
578 2013-09-28 20:50:40 <sipa> unless b13 isn't actually the new longest chain
579 2013-09-28 20:51:21 <Diablo-D3> ACTION has no clue whats being discussed
580 2013-09-28 21:05:47 <nanotube> so it seems that http://bitcoinstatus.rowit.co.uk/ hasn't been updated in a while... is there something similar which tracks bitcoin network health?
581 2013-09-28 21:07:57 <gmaxwell> nanotube: that thing was never very useful or accurate.
582 2013-09-28 21:08:49 <BlueMatt> has there ever been a particularly accurate + useful network node monitor?
583 2013-09-28 21:11:21 <gmaxwell> the seeds are the closest I think but they don't make nifty over-time graphs.
584 2013-09-28 21:13:08 <BlueMatt> nor are they particularly correct last I checked...last time I looked the number varied randomly through time with no clear evidence node counts were actually changing
585 2013-09-28 21:14:13 <nanotube> hmm... so that's a no then eh. :)
586 2013-09-28 21:15:35 <mappum> wow, so has the network size really declined that much?
587 2013-09-28 21:15:52 <mappum> is that because of changes in the mining distribution?
588 2013-09-28 21:17:18 <BlueMatt> mappum: p(the data you're looking being bad) = +/- .9
589 2013-09-28 21:19:39 <mappum> BlueMatt: negative probability?
590 2013-09-28 21:20:05 <BlueMatt> ~.09
591 2013-09-28 21:20:06 <BlueMatt> .9
592 2013-09-28 21:20:13 <mappum> i know :P
593 2013-09-28 21:20:43 <devthedevpro> So, I'm making a Mining pool. lol
594 2013-09-28 21:20:51 <devthedevpro> And, I've set it up on Koding.com. lol
595 2013-09-28 21:21:03 <devthedevpro> I wonder if it'll be stable enough
596 2013-09-28 21:27:01 <BlueMatt> has anyone figured out why autotools made building so much slower yet?
597 2013-09-28 21:31:10 <nanotube> well it's not called fasttools, it's called autotools. no speed promises. >_> :)
598 2013-09-28 21:31:33 <gmaxwell> mappum: the high numbers in that chart were BS.. though yes, it does appear that the network size has declined.
599 2013-09-28 21:31:53 <gmaxwell> I think it has less to do with mining (though some) than it does with the rise of webwallets and spv clients.
600 2013-09-28 21:32:04 <gmaxwell> and the increasing resources required to run a full node.
601 2013-09-28 21:32:44 <gmaxwell> not to mention bitcoin.org promoting thin clients over full nodes. ::shrugs::
602 2013-09-28 21:32:45 <mappum> ah. i guess it would be better to look at a graph of volume to look at real usage
603 2013-09-28 21:32:59 <BlueMatt> nanotube: no, I mean the actual compile step is slower
604 2013-09-28 21:33:09 <gmaxwell> mappum: volume?
605 2013-09-28 21:33:27 <mappum> gmaxwell: total of outputs
606 2013-09-28 21:33:36 <BlueMatt> also a terrible metric
607 2013-09-28 21:33:44 <gmaxwell> outputs?
608 2013-09-28 21:34:40 <mappum> of txs...
609 2013-09-28 21:34:55 <nanotube> mappum: transaction counts/values graphs exist on blockchain.info
610 2013-09-28 21:35:08 <mappum> that's what i'm looking at
611 2013-09-28 21:35:56 <nanotube> but those are kind of orthogonal to network health as far as full node counts and inbound connection slot availability
612 2013-09-28 21:36:38 <mappum> oh, i'm just interested in adoption, i'm not talking about network health
613 2013-09-28 21:36:48 <gmaxwell> mappum: you can't measure adoption that way.
614 2013-09-28 21:37:13 <mappum> i guess i mean, usage, not adoption
615 2013-09-28 21:37:33 <gmaxwell> One bozo with a shell script might be bouncing around coins constantly ... vs ... a million actual users trading between mtgox accounts or inputs.io creating on transactions at all.
616 2013-09-28 21:37:52 <gmaxwell> usage either.
617 2013-09-28 21:38:14 <mappum> true, but the system is big enough that you should still see trends
618 2013-09-28 21:38:51 <BlueMatt> not really, in this case
619 2013-09-28 21:39:02 <BlueMatt> at least not reliably
620 2013-09-28 21:39:23 <nanotube> gmaxwell: as far as .org page, it basically says how it is - mainline is resource intensive and takes a while to sync, other stuff is 'fast and easy'. that said, if you have suggestions for improving the choose your wallet page, i bet people would consider it.
621 2013-09-28 21:40:19 <gmaxwell> nanotube: I think the reference client needs to be made smoother to install first. people howling about it being a week or whatever is not acceptable.
622 2013-09-28 21:40:44 <gmaxwell> Once sipa's headers first stuff is in things will be greatly improved.
623 2013-09-28 21:40:46 <nanotube> mappum: no single metric is going to be sufficient, but putting together tx counts, values, bitcoin days destroyed (which includes coin age), bitpay merchant counts and tx volume... altogether can give you some idea.
624 2013-09-28 21:40:46 <warren> a week ?
625 2013-09-28 21:41:08 <nanotube> warren: maybe if you get stuck pulling from a really slow peer.
626 2013-09-28 21:41:28 <nanotube> or have a really slow hd? or insufficient ram so that things start getting paged to disk?
627 2013-09-28 21:41:44 <nanotube> yea week seems high to me... but it's not out of the realm of possibility
628 2013-09-28 21:41:49 <gmaxwell> warren: it can happen, if you have crap peer selection luck plus a slow system.
629 2013-09-28 21:42:00 <warren> BlueMatt: I thought builds were slower with autotools until I actually measured it, it's not much different if building clean vs. clean. Perhaps it is slower for cached objects though.
630 2013-09-28 21:42:06 <gmaxwell> some of these folks are turning their computers off and counting that in the time too.
631 2013-09-28 21:42:15 <nanotube> haha
632 2013-09-28 21:42:34 <nanotube> "i turned my computer off and went to sleep. when i woke up and turned it on, it still wasn't synced!"
633 2013-09-28 21:42:36 <BlueMatt> warren: it also misses cached objects for me sometimes
634 2013-09-28 21:42:39 <gmaxwell> User expirence is not always a fair thing: at the end of the day it doesn't matter how good our stuff is if the user is still unhappy.
635 2013-09-28 21:43:25 <BlueMatt> warren: especially across checkouts, which confuses it too much
636 2013-09-28 21:50:17 <sipa> did someone already write a script to do github merges with GPG-signed commits?
637 2013-09-28 21:50:29 <sipa> should be easy to do, and removes any ability for github to mess with our repo
638 2013-09-28 21:51:00 <michagogo> question: can I `gpg --import secring.gpg`?
639 2013-09-28 21:51:07 <sipa> i think so
640 2013-09-28 21:53:51 <michagogo> What is/are Luke's key(s)? I appear to have D53E9583 and 21F4889F in gpg --list-keys
641 2013-09-28 21:54:00 <michagogo> and both say [expired: 2013-06-09]
642 2013-09-28 22:08:44 <Diablo-D3> sipa: Ive figured out what to call digital signatures, btw
643 2013-09-28 22:08:49 <Diablo-D3> sipa: mathemagical
644 2013-09-28 22:36:50 <diki> Which is the most lightweight thin client?
645 2013-09-28 22:37:05 <michagogo> electrum?
646 2013-09-28 22:37:15 <sipa> do you consider a webbrowser that can display blockchain.info a "client" ?
647 2013-09-28 22:37:20 <michagogo> Ah, or that :-P
648 2013-09-28 22:37:34 <sipa> (not that i would suggest that, but "most lightweight" isn't very well defined)
649 2013-09-28 22:37:47 <diki> blockchain.info is insecure
650 2013-09-28 22:37:47 <sipa> if you require it to validate the chain, electrum
651 2013-09-28 22:37:59 <sipa> if you require it to be a network node, multibit
652 2013-09-28 22:38:01 <diki> I can tell from first glance from all the threads with stolen coins and bruteforced accounts
653 2013-09-28 22:38:08 <sipa> if you require it to validate transactions, bitcoin-qt
654 2013-09-28 22:38:28 <diki> I just need to spend some coins from my wallet w/o downloading the blockchain(mainly due to space limitations)
655 2013-09-28 22:39:06 <lianj> how do you do coin selection without downloading the blockchain?
656 2013-09-28 22:39:33 <sipa> coin selection only requires you to know your own spendable outputs
657 2013-09-28 22:39:36 <sipa> not everyone's
658 2013-09-28 22:39:53 <diki> Erm, I don't want to select anything.
659 2013-09-28 22:40:04 <sipa> the client will do coin selection for you
660 2013-09-28 22:40:08 <sipa> every wallet implementation does
661 2013-09-28 22:41:43 <diki> I've always been a full node, but since my HDD space is not unlimited, I have to purge the blockchain or use a thin client.
662 2013-09-28 22:45:25 <michagogo> [01:38:18] <diki> I just need to spend some coins from my wallet w/o downloading the blockchain(mainly due to space limitations)
663 2013-09-28 22:45:30 <michagogo> Multibit works, then.
664 2013-09-28 22:46:05 <michagogo> (so does upgrading your HD)
665 2013-09-28 22:46:49 <michagogo> (or if on a desktop, adding another HD)