1 2012-02-11 00:00:07 <BlueMatt> the right arrow to the right of each result
  2 2012-02-11 00:00:09 <NxTitle> luke-jr: can you not just save as?
  3 2012-02-11 00:00:11 <BlueMatt> that you can hover over
  4 2012-02-11 00:00:15 <luke-jr> NxTitle: I don't *want* to save it
  5 2012-02-11 00:00:15 <NxTitle> then read the file from the disk
  6 2012-02-11 00:00:35 <NxTitle> ah
  7 2012-02-11 00:00:44 <NxTitle> well you could run it through urldecode or something, I dunno
  8 2012-02-11 00:02:13 <sipa> luke-jr, gmaxwell: i did a few minor changes to addrman, in particular when exiting; any of you want to run it for a while (possibly in gdb or valgrind, maybe with -DDEBUG_ADDRMAN) ?
  9 2012-02-11 00:02:37 <luke-jr> gmaxwell: LOL @ iPhone being "high-end"
 10 2012-02-11 00:02:52 <luke-jr> sipa: I'll readd it to next-test next rebuild
 11 2012-02-11 00:03:17 <sipa> ok, thanks
 12 2012-02-11 00:03:44 <luke-jr> sipa: did you change anythign that looked possibly a bug?
 13 2012-02-11 00:03:51 <gmaxwell> sipa: I'll test.
 14 2012-02-11 00:04:21 <sipa> luke-jr: yes, but only when exiting (it was possible the dump thread was killed while busy)
 15 2012-02-11 00:15:00 <gavinandresen> https://github.com/gavinandresen/bitcoin-git/tree/discourageblocks   could use code review before I turn it into a pull request
 16 2012-02-11 00:17:59 <luke-jr> +    // Discouraging a block means declining to announce it to peers
 17 2012-02-11 00:18:01 <luke-jr> 
 18 2012-02-11 00:18:02 <luke-jr> +    // (unless they ask about it) and refusing to build directly
 19 2012-02-11 00:18:04 <luke-jr> 
 20 2012-02-11 00:18:05 <luke-jr> +    // on top of it.
 21 2012-02-11 00:18:07 <luke-jr> 
 22 2012-02-11 00:18:08 <luke-jr> this definition is problematic
 23 2012-02-11 00:18:14 <luke-jr> if you refuse to build on top of them, then you make an orphan
 24 2012-02-11 00:18:33 <BlueMatt> gavinandresen: wait, we are supposed to get code review _before_ pull requesting? damn ive been doing it wrong all along
 25 2012-02-11 00:19:24 <Diablo-D3> luke-jr: no
 26 2012-02-11 00:19:28 <sipa> luke-jr: they are accepted as soon as someone else builds on top of them
 27 2012-02-11 00:19:29 <Diablo-D3> it depends on WHY it was rejected
 28 2012-02-11 00:19:40 <luke-jr> sipa: relevance?
 29 2012-02-11 00:19:56 <gavinandresen> luke-jr: "you make an orphan" -- you mean YOUR block will be an orphan, or the block you're discouraging?
 30 2012-02-11 00:20:04 <luke-jr> nobody else will take my block unless it's based on the "discouraged" block by default
 31 2012-02-11 00:20:08 <luke-jr> gavinandresen: your block
 32 2012-02-11 00:20:21 <gavinandresen> luke-jr: right, you need 50+% of the network with the same 'discourage' rules.
 33 2012-02-11 00:20:49 <sipa> gavinandresen: "unless they ask for it" ... where is that implemented?
 34 2012-02-11 00:20:52 <luke-jr> gavinandresen: safer to take it in steps IMO
 35 2012-02-11 00:21:39 <BlueMatt> gavinandresen: an idea: add an option to half-discourage ie discourage if its in a part of a chain which has the same height as another chain so that miners are more willing to accept it
 36 2012-02-11 00:21:39 <gavinandresen> sipa:  good catch, old comment from previous version of the code....
 37 2012-02-11 00:22:00 <BlueMatt> (obv has to be optional)
 38 2012-02-11 00:22:13 <sipa> BlueMatt: not sure i get that
 39 2012-02-11 00:22:29 <BlueMatt> ie dont discourage until someone else has discouraged
 40 2012-02-11 00:22:47 <sipa> how do you observe that someone else discouraged it?
 41 2012-02-11 00:22:52 <gavinandresen> ... until somebody else has discouraged and announced an alternative?
 42 2012-02-11 00:22:58 <BlueMatt> ie there is currently a fork both with the same height
 43 2012-02-11 00:23:08 <BlueMatt> yea, if there is an alternative with the same height
 44 2012-02-11 00:23:19 <BlueMatt> that way miners would be more willing to accept it
 45 2012-02-11 00:23:40 <gmaxwell> but an alternative won't exist if people don't create an orphan there.. alas.
 46 2012-02-11 00:23:57 <BlueMatt> as currently only eg p2pool users who dont know any better would add discourage imho
 47 2012-02-11 00:24:34 <BlueMatt> in fact, competitive mining pools may deliberately make discouraged blocks to give others orphans
 48 2012-02-11 00:25:27 <gavinandresen> Mmm.... I think convincing the big pools to agree on 'discourage' rules is the way to go.
 49 2012-02-11 00:25:40 <BlueMatt> agreed, Im just not sure how easy that would be to do
 50 2012-02-11 00:26:21 <gavinandresen> Well, assuming everybody agrees it is a good idea I don't think it will be hard.
 51 2012-02-11 00:26:32 <sipa> wild assumption
 52 2012-02-11 00:27:02 <gavinandresen> I'm open to other suggestions on handling the duplicate coinbase problem
 53 2012-02-11 00:27:27 <BlueMatt> well, no I think this is probably the way to go for now
 54 2012-02-11 00:27:32 <BlueMatt> but in general for discourage ideas
 55 2012-02-11 00:27:33 <sipa> make it illegal, deployed alongside bip16
 56 2012-02-11 00:27:42 <BlueMatt> an optional half-discourage might help adoption
 57 2012-02-11 00:27:45 <gavinandresen> luke-jr: re: take it in steps:   I think that's what discourage blocks does.  First, set a date by which we want 50+% discouraging.
 58 2012-02-11 00:28:00 <BlueMatt> sipa: not really possible (pruning becomes harder)
 59 2012-02-11 00:28:07 <gavinandresen> Then, set a much farther-out date by which 50+% will outright reject blocks with dup coinbases.
 60 2012-02-11 00:28:10 <BlueMatt> well ok possible, just not ideal
 61 2012-02-11 00:28:18 <sipa> BlueMatt: not true
 62 2012-02-11 00:28:27 <BlueMatt> you cant prune coinbases then
 63 2012-02-11 00:28:34 <sipa> BlueMatt: the rule should be: disallow duplicate coinbase if the previous wasn't completely spent
 64 2012-02-11 00:28:37 <BlueMatt> or coinbase hash
 65 2012-02-11 00:28:53 <BlueMatt> that doesnt solve the actual problem, does it?
 66 2012-02-11 00:29:05 <sipa> it does
 67 2012-02-11 00:29:16 <BlueMatt> by completely spent you mean x deep?
 68 2012-02-11 00:29:20 <sipa> no
 69 2012-02-11 00:29:26 <sipa> i mean all txouts marked spent
 70 2012-02-11 00:29:39 <sipa> one block deep suffices
 71 2012-02-11 00:30:15 <sipa> in that case, disconnectblock will bring the state back to exactly what it was before during a reorg
 72 2012-02-11 00:30:54 <gavinandresen> Alternative plan:  get 50+% of miners to agree to put previous-block-hash in their coinbase transactions, and discourage blocks that don't....
 73 2012-02-11 00:30:57 <BlueMatt> mmm, ok fair enough
 74 2012-02-11 00:31:14 <BlueMatt> gavinandresen: I kinda prefer that-ish
 75 2012-02-11 00:31:23 <BlueMatt> but coinbase is so useful for eg side chain
 76 2012-02-11 00:32:07 <gavinandresen> 8 bytes of the hash should be plenty enough to keep people from creating duplicates on purpose....
 77 2012-02-11 00:32:07 <gmaxwell> BlueMatt: thats not exclusive.
 78 2012-02-11 00:32:17 <gavinandresen> (if we're worried about space in the coinbase)
 79 2012-02-11 00:32:27 <BlueMatt> gmaxwell: true...
 80 2012-02-11 00:32:41 <BlueMatt> still not a "clean" solution
 81 2012-02-11 00:34:09 <BlueMatt> also... gavinandresen on after 5??? wow
 82 2012-02-11 00:34:48 <gavinandresen> yeah, I'm tired... about to turn into a pumpkin
 83 2012-02-11 00:35:19 <josephcp> if you can get 51%, can't you reject duplicate coinbases up to a depth of 6 blocks today? that would be a pretty strong discouragement because it's close to impossible
 84 2012-02-11 00:36:02 <josephcp> acutally 3 or 4 would probably make more sense
 85 2012-02-11 00:36:33 <sipa> adding 64 bits of the prevhash to the coinbase makes it 1.8*10^19 times harder to mine a block with a duplicate coinbase
 86 2012-02-11 00:36:41 <sipa> i think 16 bits suffices, actually
 87 2012-02-11 00:36:52 <sipa> wait
 88 2012-02-11 00:37:24 <sipa> never mind
 89 2012-02-11 00:37:32 <BlueMatt> 1.8*10^19x harder than nothing is still...nothing
 90 2012-02-11 00:37:37 <gmaxwell> thats not right.
 91 2012-02-11 00:37:51 <sipa> no, i'm wrong
 92 2012-02-11 00:38:31 <gmaxwell> It's just the odds that the value repeats. You only get to try it once per block.
 93 2012-02-11 00:38:32 <sipa> it means you'd just need to wait until the (prevhash % N) is equal to that of the coinbase you're trying to duplicaye
 94 2012-02-11 00:39:20 <gmaxwell> yes. We could alternatively use the height there.
 95 2012-02-11 00:39:29 <gmaxwell> And then that won't collide.
 96 2012-02-11 00:39:53 <sipa> gmaxwell: that's the solution!
 97 2012-02-11 00:40:17 <sipa> gavinandresen: ^
 98 2012-02-11 00:40:19 <gmaxwell> (and it's helpfully small for the forseeable future)
 99 2012-02-11 00:40:52 <gavinandresen> good idea
100 2012-02-11 00:41:28 <BlueMatt> ack
101 2012-02-11 00:42:14 <gmaxwell> luke-jr: is there a positioning requirement for mm that would keep us from sticking a counter in a particlar location?
102 2012-02-11 00:42:59 <BlueMatt> is it sad when someone asks you what today is on a 7-day scale and your first instinct is 13th of feb instead of friday?
103 2012-02-11 00:45:08 <BlueMatt> timeline for such a change?
104 2012-02-11 00:45:32 <Diablo-D3> blueMatt: yes, especially when its the 10th.
105 2012-02-11 00:45:44 <BlueMatt> Diablo-D3: 10 base 7 == 13
106 2012-02-11 00:45:50 <BlueMatt> 10 base 10 == 13 base 7
107 2012-02-11 00:45:59 <Diablo-D3> the.... what?
108 2012-02-11 00:46:04 <Diablo-D3> how...
109 2012-02-11 00:46:05 <Diablo-D3> jus
110 2012-02-11 00:46:07 <Diablo-D3> erjarwr
111 2012-02-11 00:46:08 <Diablo-D3> qawrjaprjoltqwt
112 2012-02-11 00:46:08 <luke-jr> gavinandresen: step 1 could be, when there's two competing blocks, prefer the one that isn't discouraged
113 2012-02-11 00:46:10 <Diablo-D3> qweotjpiwatjrawpjrw
114 2012-02-11 00:46:13 <Diablo-D3> wrfkaofkwoakakrfagtqwrasr
115 2012-02-11 00:46:16 <Diablo-D3> trpijrtoqjwrja
116 2012-02-11 00:46:23 <luke-jr> gmaxwell: I don't understand the question
117 2012-02-11 00:46:29 <gavinandresen> luke-jr: sure.  But there is rarely two competing blocks.
118 2012-02-11 00:46:40 <sipa> luke-jr: where is the mm hash placed in the coinbase?
119 2012-02-11 00:46:44 <sipa> luke-jr: does that matter?
120 2012-02-11 00:46:56 <BlueMatt> gotta go, see yall later
121 2012-02-11 00:46:58 <gmaxwell> luke-jr: in the coinbase, can the mm be at any location? e.g. can we require the counter to be first?
122 2012-02-11 00:47:06 <luke-jr> gavinandresen: there won't be, if step 2 (this) gets deployed anywhere
123 2012-02-11 00:47:21 <luke-jr> gmaxwell: it can be any place, provided the magic prefix is there
124 2012-02-11 00:47:26 <gmaxwell> great.
125 2012-02-11 00:47:40 <luke-jr> without the prefix, within the first 20 octets
126 2012-02-11 00:47:42 <gmaxwell> putting it first makes sense to me, using the bitcoin variable length integer coding, I guess.
127 2012-02-11 00:48:10 <gavinandresen> So it could be: discourage block if first 5 bytes of coinbase aren't push-4-bytes and then the 32-bit height
128 2012-02-11 00:48:31 <Diablo-D3> gavinandresen: er
129 2012-02-11 00:48:33 <Diablo-D3> wat?
130 2012-02-11 00:49:25 <gavinandresen> ... and I suppose it would make sense to start discouraging in block races right away... (only danger would be a little more code to write/test)
131 2012-02-11 00:49:41 <gavinandresen> Diablo-D3: what what?
132 2012-02-11 00:50:10 <luke-jr> gavinandresen: with step 1 in place, participating miners don't risk orphans before 50%, and there's no timely coordination required
133 2012-02-11 00:50:41 <gavinandresen> yes, but it gives almost zero protection from an attacker trying to split the chain
134 2012-02-11 00:51:14 <Diablo-D3> gavinandresen: I have no clue what you just said
135 2012-02-11 00:51:17 <Diablo-D3> what does that do?
136 2012-02-11 00:51:40 <gavinandresen> Putting the block height in the coinbase?  Prevents somebody from creating duplicates on purpose
137 2012-02-11 00:51:51 <Diablo-D3> oh, that thing gmaxwell found earlier?
138 2012-02-11 00:52:37 <Diablo-D3> I still dont get how thats useful, wont spend, dupe, spend dupe still get caught?
139 2012-02-11 00:52:38 <gmaxwell> roconner not me.
140 2012-02-11 00:52:40 <Diablo-D3> i mean normally
141 2012-02-11 00:52:46 <Diablo-D3> well, someone found it
142 2012-02-11 00:52:58 <gmaxwell> Diablo-D3: it's complicated.
143 2012-02-11 00:53:32 <k9quaint> gmaxwell: can you explain it with interpretive dance?
144 2012-02-11 00:55:07 <sipa> if you have a coinbase tx, and in a chain-split that is only seen by part of the network a dupe is created with the same coinbase tx afterwards
145 2012-02-11 00:55:27 <sipa> and that fork gets reorged away, the original coinbase tx is gone from the database
146 2012-02-11 00:55:37 <sipa> while it is still there for those that didn't see the fork
147 2012-02-11 00:56:01 <sipa> if someone now tries to spend it, you got a split
148 2012-02-11 00:58:01 <gavinandresen> I'm a pumpkin.  See y'all later.
149 2012-02-11 00:58:21 <sipa> cya
150 2012-02-11 00:58:49 <sipa> it's only a real problem if the majority of mining power didn't see the reorg and accepts the spend
151 2012-02-11 01:00:47 <Diablo-D3> sipa: but the spend is ...
152 2012-02-11 01:00:48 <Diablo-D3> oh right
153 2012-02-11 01:00:51 <Diablo-D3> the stupidity with tx ids
154 2012-02-11 01:42:28 <userggj> will this be possible?; https://bitcointalk.org/index.php?topic=750.0
155 2012-02-11 01:43:43 <luke-jr> already is
156 2012-02-11 01:45:31 <userggj> implemented?
157 2012-02-11 01:45:40 <userggj> on gui?
158 2012-02-11 01:45:44 <luke-jr> no
159 2012-02-11 01:46:06 <userggj> any reason why not?
160 2012-02-11 01:46:41 <luke-jr> nobody's taken the time to do it
161 2012-02-11 01:46:55 <userggj> ha ok
162 2012-02-11 02:01:21 <josephcp> userggj: a working example, https://blockchain.info/wallet/escrow
163 2012-02-11 02:01:45 <userggj> will see
164 2012-02-11 03:59:02 <etotheipi_> is there a way to use 20-byte addresses in P2SH scripts?
165 2012-02-11 03:59:48 <gmaxwell> '20 byte addresses'?
166 2012-02-11 03:59:59 <etotheipi_> hash160 strings
167 2012-02-11 04:00:48 <gmaxwell> etotheipi_: not usefully.
168 2012-02-11 04:00:58 <luke-jr> etotheipi_: P2SH is always hash160&
169 2012-02-11 04:01:14 <gmaxwell> no e.g. pay to pay to script to pay hash160 pubkey
170 2012-02-11 04:01:18 <etotheipi_> I meant, using hash160 instead of public keys
171 2012-02-11 04:01:29 <luke-jr> &
172 2012-02-11 04:01:30 <gmaxwell> which would be dumb unless you don't know the pubkey, but if you don't you can't write the redemption script.
173 2012-02-11 04:01:42 <gmaxwell> if we had key recovery you could usefully do that.
174 2012-02-11 04:02:10 <etotheipi_> oh, right
175 2012-02-11 04:02:30 <etotheipi_> wait, I haven't looked into it much
176 2012-02-11 04:03:24 <etotheipi_> doesn't the necessity to use full public keys detract from one of the reasons not to use full multi-sig to begin with?  transferring super long strings?
177 2012-02-11 04:03:57 <etotheipi_> wait, don't tell me how stupid I am, I need to read more before I start a real argument
178 2012-02-11 04:04:07 <gmaxwell> etotheipi_: I have you trained. Good.
179 2012-02-11 04:04:25 <etotheipi_> it's just a natural impulse to ask first...
180 2012-02-11 04:04:49 <luke-jr> :P
181 2012-02-11 04:08:12 <CRichard> gmaxwell: do you think it would be a useful addition to have something like a bayeux server implemented in full bitcoin nodes?
182 2012-02-11 04:08:32 <CRichard> gmaxwell: so that lightweight clients could talk to them via http requests and long polling
183 2012-02-11 04:10:44 <etotheipi_> so you do have to have everyone's public keys in order to construct the P2SH script
184 2012-02-11 04:11:37 <etotheipi_> does this not bother anyone else?
185 2012-02-11 04:11:45 <Diablo-D3> hrm
186 2012-02-11 04:11:54 <Diablo-D3> I think I might have gotten bitcoind to compile on alpine
187 2012-02-11 04:12:38 <etotheipi_> maybe I just haven't thought about it enough, but it bothers me that using P2SH requires a "different" addressing system
188 2012-02-11 04:13:10 <etotheipi_> as long as we were changing isStandard, why couldn't we also add a isStandard script for using hash160 strings in P2SH scripts?
189 2012-02-11 04:14:54 <etotheipi_> now I wish I hadn't waited to so long to start reading up on BIP 0016
190 2012-02-11 04:15:45 <Diablo-D3> nope, doesnt work
191 2012-02-11 04:17:50 <etotheipi_> gmaxwell, is there a reason for not allowing hash160-based multi-sig tx?  (besides the more-complicated internal scripts which will eventually be pruned)
192 2012-02-11 04:21:12 <gmaxwell> etotheipi_: Are you tired? You're asking a kind of stupid question.
193 2012-02-11 04:21:50 <etotheipi_> probably
194 2012-02-11 04:21:52 <gmaxwell> P2SH is a pay to H(spending script). So you have to know the spending script to create a P2SH address.
195 2012-02-11 04:21:56 <etotheipi_> what did I miss?
196 2012-02-11 04:22:20 <gmaxwell> The spending script must include the pubkey, because we couldn't validate the signature otherwise (lacking key recovery).
197 2012-02-11 04:22:39 <gmaxwell> So if you are authoring a p2sh address you must know the pubkey. Pay to hash160 makes no sense there.
198 2012-02-11 04:22:52 <etotheipi_> (I just concluded that wasn't the case from looking at the structure... but obviously I misinterpretted what I read)
199 2012-02-11 04:23:46 <CRichard> gmaxwell: did you catch my question above?
200 2012-02-11 04:24:32 <gmaxwell> CRichard: I didn't but I see it now.
201 2012-02-11 04:24:53 <gmaxwell> CRichard: the reference client is really not at all structured to be a thin client server today.
202 2012-02-11 04:25:20 <gmaxwell> CRichard: I don't have an opinions on protocols used to interconnect them once the work is done to restructure it to make that sort of usage reasonable.
203 2012-02-11 04:25:29 <CRichard> k
204 2012-02-11 04:26:16 <etotheipi_> gmaxwell, I'm not seeing it:  we are using a "regular" OP_CHECKMULTISIG TxOut script in the serialized script, and paying with an almost-regular OP_CHECKMULTISIG redemption script (a list of signatures)... I don't see why you can't replace those with the super-complicated pay-to-hash160 scripts
205 2012-02-11 04:27:51 <etotheipi_> gmaxwell, like the scripts shown in this thread:  https://bitcointalk.org/index.php?topic=56095.msg670577#msg670577
206 2012-02-11 04:28:15 <gmaxwell> etotheipi_: because you must spend with exactly that script.
207 2012-02-11 04:28:24 <gmaxwell> And you can't validate a signature without the pubkey.
208 2012-02-11 04:28:40 <etotheipi_> the public key and signature are supplied in the txin script
209 2012-02-11 04:28:55 <gmaxwell> you could write some script that says pushpubkey dup hash160 pushhash equals .. checksig..  but that would be moronic.
210 2012-02-11 04:29:19 <gmaxwell> etotheipi_: P2SH is pay to hash of the script.
211 2012-02-11 04:30:00 <etotheipi_> okay, there's clearly a misunderstanding here (most likely my own)... I will come up with a concrete example, and you can point out what I missed
212 2012-02-11 04:30:11 <gmaxwell> etotheipi_: sorry, I'm baffled too.
213 2012-02-11 04:32:03 <gmaxwell> etotheipi_: yea. Not sure where the disconnect is.  P2SH says "pay only to an input script with hash X"  ... so to come up with X you must know the pubkey in advance because any txin has to provide the pubkey and p2sh specifies the hash of the txnin.  Am I still making no sense to you?
214 2012-02-11 04:34:07 <etotheipi_> so your input script is [signatures] {2 pubkey0 pubkey1 pubkey2 3 OP_CHECKMULTISIG}  where everythign in {} is the serialized script
215 2012-02-11 04:34:37 <etotheipi_> the stuff in {} would be hashed an included in the TxOut script, to be spent by the tx in script above
216 2012-02-11 04:35:56 <etotheipi_> why couldn't you use:  [sig0] [pubkey0] [sig1] [pubkey1] {long script using only hash160 strings}
217 2012-02-11 04:36:35 <gmaxwell> Okay, none of that is p2sh.
218 2012-02-11 04:36:49 <gmaxwell> oh sorry, misread.
219 2012-02-11 04:37:34 <etotheipi_> "long script using only hash160 strings" would be what you see in the non-OP_CHECKMULTISIG unit tests here:   https://bitcointalk.org/index.php?topic=56095.msg670577#msg670577
220 2012-02-11 04:37:37 <gmaxwell> I follow you up to your questions.
221 2012-02-11 04:37:50 <gmaxwell> "why couldn't you use:  [sig0]" use as what? desert topping? :)
222 2012-02-11 04:38:27 <etotheipi_> oh, sorry... I messed up my syntax:   [sig0 pubkey0 sig1 pubkey1] {long script using only hash160 strings}
223 2012-02-11 04:39:09 <etotheipi_> the long script I am referring to is a valid script that verifies the supplied public keys match the hash160 strings in the original script, before checking signatures
224 2012-02-11 04:39:20 <etotheipi_> (most definitely not standard, and definitely not simple, but it works)
225 2012-02-11 04:39:37 <gmaxwell> 21:28 < gmaxwell> you could write some script that says pushpubkey dup hash160 pushhash equals .. checksig..   but that would be moronic.
226 2012-02-11 04:39:46 <gmaxwell> ^ am I misunderstanding that you're suggesting that?
227 2012-02-11 04:40:12 <gmaxwell> e.g. the script provides the pubkeys, and checks the very keys its providing with hash160s.. thus saving you nothing.
228 2012-02-11 04:40:21 <etotheipi_> well I wasn't clear exactly what you were saying there, since it doesn't look like the script I'm referring to, but probably
229 2012-02-11 04:40:57 <etotheipi_> in what I'm proposing, there are no public keys in the script that is serialized
230 2012-02-11 04:41:03 <gmaxwell> That can't work.
231 2012-02-11 04:41:15 <etotheipi_> and I'm not necessarily proposing it, I'm trying to understand why it's not possible
232 2012-02-11 04:41:32 <etotheipi_> okay, well clearly I missed something, I'll think about it and stop spending your time answering my stupid questions
233 2012-02-11 04:42:12 <etotheipi_> based on your reaction, I feel like it must be a stupid question, but I just need more time to absorb BIP 0016
234 2012-02-11 04:42:37 <gmaxwell> etotheipi_: ah! how about this: the seralized script is exected with a clean enviroment
235 2012-02-11 04:43:04 <gmaxwell> (are you expecting to feed extra data into it?)
236 2012-02-11 04:43:30 <Diablo-D3> >0016
237 2012-02-11 04:43:32 <Diablo-D3> >00
238 2012-02-11 04:43:40 <Diablo-D3> >implication we're ever going to have that many
239 2012-02-11 05:07:13 <etotheipi_> gmaxwell, http://pastebin.com/yNK2ehqF
240 2012-02-11 05:08:29 <etotheipi_> I feel like it's the same thing I said, just with more space and better formatting, to be abs clear... I still don't see how this wouldn't be valid if isStandard was not considered
241 2012-02-11 05:09:23 <etotheipi_> it seems like the TxOut script preparer only needs each party's regular hash160...
242 2012-02-11 05:10:31 <gmaxwell> it would fail to validate because there would be no public keys on the stack when the seralized script executes, so it would fail.
243 2012-02-11 05:11:19 <etotheipi_> why would there be no public keys on the stack?  the sigs are there, right?
244 2012-02-11 05:12:03 <gmaxwell> because they are not inside the seralized script.
245 2012-02-11 05:12:26 <etotheipi_> neither are the signatures, but they somehow made it into the script execution for OP_CHECKMULTISIG to succeed
246 2012-02-11 05:12:44 <trigliu> cool
247 2012-02-11 05:12:59 <trigliu> OP_RAND is next
248 2012-02-11 05:17:18 <luke-jr> lol
249 2012-02-11 05:18:56 <trigliu> op_checkmultisig is awesome
250 2012-02-11 05:19:02 <Diablo-D3> I fucking hate bitcoin's code
251 2012-02-11 05:19:03 <trigliu> ive been waiting for something like that'
252 2012-02-11 05:22:16 <etotheipi_> the serialized script is included in the TxIn script as one, large PUSHDATA op, right? and recognized to be expanded for evaluation because of the special sequence of opcodes in the TxOut script?
253 2012-02-11 05:23:16 <Diablo-D3> seriously, why in fucking hell is this code so goddamned ugly
254 2012-02-11 05:23:21 <Diablo-D3> I cant even tell what the fuck its doing
255 2012-02-11 05:23:28 <Diablo-D3> all I see is boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost boost v
256 2012-02-11 05:23:38 <Diablo-D3> some <>s that dont belong
257 2012-02-11 05:23:44 <Diablo-D3> and lots of other insane shit
258 2012-02-11 05:25:49 <etotheipi_> gmaxwell, as far as I can see, the only reason it is invalid is because of arbitrary restrictions put on BIP 0016, that could've been just as easily adapted to accommodate this use case if it wanted to (for instance, keeping everything on the stack before the serialized script)
259 2012-02-11 05:26:05 <Diablo-D3> no wonder bitcoin adoption is a fucking failure
260 2012-02-11 05:26:10 <Diablo-D3> who the fuck even wants to compile this shit
261 2012-02-11 05:26:35 <etotheipi_> (I don't mean to suggest the restrictions in BIP 16 are arbitrary, but it seems they could just as easily have supported this structure (or similar) without compromising functionality/security)
262 2012-02-11 05:27:04 <etotheipi_> if it still seems like a stupid question, I'll back into my Armory-development-meditative state
263 2012-02-11 05:40:40 <olp> I compiled it many times, but ya I dont like all the high level c++ as well
264 2012-02-11 06:29:06 <trigliu> OP_RAND is a highly tamper-resistant lottery
265 2012-02-11 06:29:32 <trigliu> unlike those phony powerball drawings where we're supposed to believe the numbers are fairly picked
266 2012-02-11 06:30:38 <trigliu> lotteries suck in general tho im sure there are other applications.
267 2012-02-11 06:47:15 <cjd> pay to whoever can guess the lot bits of the hash of the next block
268 2012-02-11 06:47:25 <cjd> *most low bits
269 2012-02-11 06:56:08 <trigliu> cool cjd
270 2012-02-11 07:20:38 <gribble> New news from bitcoinrss: rm44377 opened issue 820 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/820>
271 2012-02-11 07:42:19 <sipa> not syure anyone thought of this already, but having the block height in the coinbase allows for much stricter bounds on the difficulty and time in orphan blocks
272 2012-02-11 07:42:32 <sipa> gmaxwell: ^
273 2012-02-11 08:23:47 <cande> can i se which bitcoin address a tx is sent from?
274 2012-02-11 08:31:01 <cande> this could be a good feater for the bitcoin gui, to show which address the tx was sent from
275 2012-02-11 08:31:33 <bd__> cande: not really. bitcoin txns often have multiple inputs
276 2012-02-11 08:31:43 <bd__> and/or those inputs can be internal wallet keys that are used only for splits
277 2012-02-11 08:32:02 <bd_> so it wouldn't really be meaningful
278 2012-02-11 08:34:14 <cande> so if i want to send the bitcoins back, look here: http://blockexplorer.com/tx/55275a829ad993acbabb2d865e82ac46bb28d9fc4029beb8372cfdf46f3aaf56#o1
279 2012-02-11 08:34:37 <cande> would 1LHYE63... be the correct address?
280 2012-02-11 08:35:41 <bd_> probably. although the recipient might find it coming back at an unexpected address
281 2012-02-11 08:36:49 <cande> how do you mean?
282 2012-02-11 08:38:03 <sipa> cande: do not assume that that address is owned by the sender
283 2012-02-11 08:38:28 <sipa> if you want to do a refund, ask for an address
284 2012-02-11 08:41:44 <cande> hm
285 2012-02-11 08:42:10 <sipa> for example if they use an e-wallet service, sending something there may not credit your sender's account
286 2012-02-11 08:43:09 <cande> ah
287 2012-02-11 08:43:52 <cande> but if he sent it from bitcoin-qt?
288 2012-02-11 08:43:55 <bd_> sipa: but the e-wallet service would get the money at least!
289 2012-02-11 08:44:25 <cande> bd_, good for the e-wallet service :)
290 2012-02-11 08:44:36 <sipa> cande: he may still have switched wallets, lost his computer, ...
291 2012-02-11 08:45:35 <sipa> in general it is bad practice to send something to an address without the receiver's request to send it there
292 2012-02-11 08:45:37 <cande> bd_, how about those internal wallet keys?
293 2012-02-11 08:45:53 <bd_> dunno, I don't actually use bitcoin, I just hang out here ;)
294 2012-02-11 08:45:59 <cande> hah :)
295 2012-02-11 08:46:23 <sipa> those will work, cande, in bitcoin-qt, but it is bad practice
296 2012-02-11 08:46:24 <bd_> If I were writing the client, it'd handle that. But I'm not.
297 2012-02-11 08:48:24 <cande> sipa,  I understand, i just got one customer who wanted a refund of an order, and when i asked him about a bitcoin address, he promptly said, send it back to the same address.
298 2012-02-11 08:48:47 <sipa> ok, in that case, no problem
299 2012-02-11 08:49:21 <cande> so 1LHYE63... is a safe bet
300 2012-02-11 08:49:41 <sipa> yes, both are fine
301 2012-02-11 08:50:08 <cande> 19ZPqU.. also?
302 2012-02-11 08:50:50 <sipa> yes
303 2012-02-11 08:51:22 <cande> cool :)
304 2012-02-11 08:51:24 <cande> thx alot
305 2012-02-11 09:06:40 <marf_away> maybe
306 2012-02-11 09:07:03 <marf_away> the customer is using a e-qallet and doesnt know he cant get monney back that way?
307 2012-02-11 09:19:56 <gribble> New news from bitcoinrss: laanwj opened issue 821 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/821>
308 2012-02-11 10:31:10 <newfrozzen> Hi =)
309 2012-02-11 10:31:36 <gribble> New news from bitcoinrss: xanatos opened issue 822 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/822>
310 2012-02-11 10:41:47 <gribble> New news from bitcoinrss: sipa opened pull request 823 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/823>
311 2012-02-11 12:52:45 <newfrozzen> #bitcoin-mining
312 2012-02-11 12:57:46 <gribble> New news from bitcoinrss: xanatos opened issue 824 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/824>
313 2012-02-11 13:28:38 <gribble> New news from bitcoinrss: sipa opened pull request 825 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/825>
314 2012-02-11 14:17:48 <gribble> The operation succeeded.
315 2012-02-11 14:17:48 <sipa> ;;later tell BlueMatt it seems the old 'key' records are not removed when encrypting a wallet in git head; where was this supposed to happen?
316 2012-02-11 14:18:19 <jamescarr> sipa, git head or git master?
317 2012-02-11 14:18:40 <jamescarr> of course, I guess you could say head of master
318 2012-02-11 14:18:57 <UukGoblin> where can I read up about how merged mining works in detail? I mean, if I wanted to start a new chain that I'd like to have linked to bitcoin - what should getauxblock return, what are the details...?
319 2012-02-11 14:19:47 <UukGoblin> oh found a wiki page
320 2012-02-11 14:19:58 <sipa> jamescarr: https://github.com/bitcoin/bitcoin.git branch master's HEAD
321 2012-02-11 14:34:47 <UukGoblin> is there a list of chain IDs somewhere?
322 2012-02-11 14:39:46 <gribble> New news from bitcoinrss: laanwj opened pull request 826 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/826>
323 2012-02-11 14:44:49 <gribble> New news from bitcoinrss: sipa opened pull request 827 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/827>
324 2012-02-11 14:51:13 <lianj> .
325 2012-02-11 15:05:04 <gribble> New news from bitcoinrss: sipa opened pull request 828 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/828>
326 2012-02-11 15:12:46 <gribble> The operation succeeded.
327 2012-02-11 15:12:46 <sipa> ;;later tell BlueMatt nevermind, fixed in pullreq 827; want to have a look at pullreq 828?
328 2012-02-11 15:14:21 <lianj> headmerge!
329 2012-02-11 16:25:10 <sipa> i'm surprised that ThreadCleanWalletPassphrase even worked at all... it calculated its number of seconds left to sleep as (GetTime() - nWalletUnlockTime) ...
330 2012-02-11 16:26:36 <sipa> while it should have been 1000*(nWalletUnlockTime - GetTime())
331 2012-02-11 16:27:24 <sipa> *milliseconds
332 2012-02-11 16:33:15 <userggf> hi, i was looking escrow transactions on blockchain.info
333 2012-02-11 16:33:42 <userggf> i ask; is it possible instead of 2 of 3
334 2012-02-11 16:34:01 <userggf> is possible something with percentage
335 2012-02-11 16:35:06 <userggf> like 60% of all on a pre definied keys
336 2012-02-11 16:37:39 <sipa> you could probably do something with up to 20 keys, and have some weight assigned to each
337 2012-02-11 16:37:43 <Idiot___> userggf: i beleive their working on all even numbered keys or all pretty looking ones :P
338 2012-02-11 16:37:48 <sipa> and a minimum required weight
339 2012-02-11 16:38:08 <sipa> but those would be non-standard scripts
340 2012-02-11 16:41:18 <userggf> no one is working on escrow transactions on satoshi client?
341 2012-02-11 16:41:55 <sipa> sure - 0.6 will be the first client with preliminary support for them
342 2012-02-11 16:42:22 <userggf> on gui?
343 2012-02-11 16:42:38 <userggf> any screenshot?
344 2012-02-11 16:42:51 <sipa> no, no gui
345 2012-02-11 16:43:31 <sipa> only making a few types standard (so they get relayed by the network), and an RPC call to create addresses corresponding to multisigs
346 2012-02-11 16:44:16 <userggf> ok
347 2012-02-11 16:55:13 <luke-jr> userggf: be sure to get a BIP 17 enabled miner if you solo/p2pool mine
348 2012-02-11 16:55:17 <luke-jr> to help enable it
349 2012-02-11 16:56:26 <PK> when I request work from the bitcoin client and solve it. Which address will get the block? Is there any way to control that?
350 2012-02-11 16:56:48 <luke-jr> PK: I have a pullreq that lets you control it, but mainline hasn't merged it yet
351 2012-02-11 16:57:03 <luke-jr> PK: they claim not enough people need to control it
352 2012-02-11 16:57:25 <luke-jr> if you want to voice your support, you can comment on https://github.com/bitcoin/bitcoin/pull/719
353 2012-02-11 16:57:27 <PK> luke-jr: I'm fine with any address getting the block, really. I just like to know which one it is :)
354 2012-02-11 16:57:47 <luke-jr> I don't know any easy way to find that out
355 2012-02-11 16:58:39 <sipa> the next unused pool address
356 2012-02-11 16:58:48 <sipa> but there is no way to request that in advance
357 2012-02-11 17:00:32 <PK> ok, so it makes a new address for a new block (at least new to me, the user since I don't see it in advance).
358 2012-02-11 17:00:48 <sipa> indeed
359 2012-02-11 17:01:09 <Graet> crazy that eh, i can set an address in one bitcoind to generate to another, but not know the adress on the same bitcoind :)
360 2012-02-11 17:02:27 <sipa> ?
361 2012-02-11 17:02:39 <PK> I'd really like an easy way to get th private key out too. Maybe a QR picture so I can use my wallet on the bitcoin client and on the mobilephone. Any plans in that direction already?
362 2012-02-11 17:03:20 <Graet> sipa, the pool is setup with mining nodes and a wallet to generate blocks to, the mining nodes dont ever see any coins
363 2012-02-11 17:04:01 <PK> luke-jr: I could ask for a getwork request. That should show me the address, right?
364 2012-02-11 17:04:30 <luke-jr> PK: no
365 2012-02-11 17:04:53 <luke-jr> Graet: you can't set where you generate in bitcoind without coinbaser ;)
366 2012-02-11 17:05:42 <Graet> yes, still surprising with all these very clever that we cant tell if we generate to same bitcoind...
367 2012-02-11 17:05:53 <Graet> what the addy will be :)
368 2012-02-11 17:09:16 <sipa> PK: 0.6.0 will have dumpprivkey RPC call
369 2012-02-11 17:09:59 <PK> nice, but feels dangerous to allow that from remote :O
370 2012-02-11 17:10:15 <PK> imho better have a privatekey import call. Much safer
371 2012-02-11 17:10:26 <sipa> eh
372 2012-02-11 17:10:29 <sipa> you can't compare those
373 2012-02-11 17:10:36 <sipa> you need both an import and an export call
374 2012-02-11 17:10:44 <sipa> one is useless without the other
375 2012-02-11 17:11:07 <sipa> also, if you consider dumpprivkey unsafe, sendto is also unsafe...
376 2012-02-11 17:11:08 <PK> not really. Only if you generate the other key with a bitcoind.
377 2012-02-11 17:11:17 <PK> true
378 2012-02-11 17:18:20 <userggf> the last post here https://bitcointalk.org/index.php?topic=40264.400
379 2012-02-11 17:18:32 <userggf> is already possible?
380 2012-02-11 17:19:24 <sipa> 100 keys is a lot... that means a transaction of at least 2-3 KiB
381 2012-02-11 17:19:46 <sipa> the network supports it, but you'd need to find a miner that will accept such a non-standard transaction
382 2012-02-11 17:20:07 <userggf> is this a problem?
383 2012-02-11 17:21:47 <sipa> probably yes -- the network will not relay such a transaction for you, and you will certainly need a fee
384 2012-02-11 17:22:02 <BlueMatt> likely a pretty big fee too
385 2012-02-11 17:22:23 <userggf> hum, this is bad
386 2012-02-11 17:22:39 <BlueMatt> why would you want to do such a huge tx anyway?
387 2012-02-11 17:23:13 <userggf> i imagine the situation of:
388 2012-02-11 17:23:51 <userggf> my english is bad, but i'll try to explaing
389 2012-02-11 17:24:31 <userggf> you know closed places where you livesnd pay a month fee
390 2012-02-11 17:24:41 <userggf> live and
391 2012-02-11 17:25:06 <userggf> or that you don't live but have house on it
392 2012-02-11 17:26:29 <userggf> many people pay and don't know where the money goes
393 2012-02-11 17:26:57 <BlueMatt> sipa: so "while (GetTime() < nWalletUnlockTime) Sleep(GetTime() - nWalletUnlockTime);"  has a race condition?
394 2012-02-11 17:27:01 <userggf> so who admin the close place condominium
395 2012-02-11 17:27:18 <sipa> BlueMatt: you were checking nWalletUnlockTime outside of the CS
396 2012-02-11 17:27:41 <userggf> should before by things or service, will need an aval first
397 2012-02-11 17:27:57 <userggf> from who own houses on it
398 2012-02-11 17:28:38 <userggf> before buy
399 2012-02-11 17:30:03 <userggf> understand?
400 2012-02-11 17:30:23 <sipa> yes
401 2012-02-11 17:30:48 <sipa> that's a potential application for multisigs, but one that becomes hard for many keys
402 2012-02-11 17:30:48 <userggf> make sense?
403 2012-02-11 17:31:18 <userggf> ok
404 2012-02-11 17:31:40 <sipa> over time, such transactions will probably become possible, if there is demand for it
405 2012-02-11 17:32:26 <userggf> good to know
406 2012-02-11 17:33:05 <userggf> my freak was a answer like techinical impossible
407 2012-02-11 17:33:56 <sipa> BlueMatt: the trick to doing such things is always do all calculation that is based on the shared variable inside a CS, but release the mutex right before sleeping
408 2012-02-11 17:34:34 <sipa> otherwise the check may use one value, but the delay for sleeping use another
409 2012-02-11 17:34:54 <BlueMatt> that wouldnt cause a problem though
410 2012-02-11 17:35:25 <BlueMatt> even if you are setting nWalletUnlockTime at the worst possible clock, I still see little way for that to cause problems
411 2012-02-11 17:35:58 <BlueMatt> s/little/no/
412 2012-02-11 17:36:53 <BlueMatt> maybe on a 32-bit computer, if you set the upper 32-bits of nWalletUnlockTime before the sleep, and the lower after the sleep has been calculated, the sleep could go on for too long
413 2012-02-11 17:37:39 <sipa> your initial check is outside a CS; the thread could think it is non-zero, but immediately afterwards it is set to zero by another thread
414 2012-02-11 17:37:59 <sipa> that would mean it only increments nWalletUnlockTime, and exists
415 2012-02-11 17:38:09 <sipa> *exits
416 2012-02-11 17:38:17 <sipa> and the wallet would never get locked again
417 2012-02-11 17:38:26 <sipa> i'm sure there are tons of such combinations possible
418 2012-02-11 17:40:21 <BlueMatt> if its non-0 on the check and 0 on the sleep, the sleep will get called with a negative number, so it should just not do anything, and this thread will exit, no problem
419 2012-02-11 17:40:48 <sipa> i'm not talking about the sleep
420 2012-02-11 17:41:29 <sipa> i'm talking about the initial if (nWalletUnlockTime == 0) test on line 1542
421 2012-02-11 17:41:36 <sipa> which happens outside a CS
422 2012-02-11 17:42:01 <sipa> which means another thread could simultaneously be setting it to zero
423 2012-02-11 17:43:06 <BlueMatt> I dont think that can happen, as the wallet would have to have been unlocked, then set to zero while still being unlocked, then locked.  and ThreadCleanWalletPassphrase wont be started if the wallet is already unlocked
424 2012-02-11 17:43:08 <sipa> so you end up in the else branch, thinking that an another thread will do the actual wait-until-time-passed, but that other thread is exiting
425 2012-02-11 17:43:23 <sipa> see walletlock()
426 2012-02-11 17:43:31 <sipa> that sets nWalletUnlockTime to 0
427 2012-02-11 17:43:45 <BlueMatt> oh, that sets it after lock...
428 2012-02-11 17:44:01 <BlueMatt> meh, ok
429 2012-02-11 17:44:23 <sipa> there may or may not be other potential issues, but the code now is absolutely thread-safe :)
430 2012-02-11 17:45:02 <BlueMatt> fair enough, though maybe replace the cs....Enter(...__FILE__, __LINE__) with a define somewhere?
431 2012-02-11 17:45:15 <BlueMatt> make it look a bit cleaner?
432 2012-02-11 17:45:20 <BlueMatt> ;)
433 2012-02-11 17:45:21 <sipa> maybe, it's also used in net.h in BeginMessage() afaik
434 2012-02-11 17:52:02 <luke-jr> http://paste.pocoo.org/show/549398/
435 2012-02-11 18:06:33 <userggf> sipa, see the piuk answer: https://bitcointalk.org/index.php?topic=40264.msg743209;topicseen#msg743209
436 2012-02-11 18:07:44 <sipa> userggf: he is right - it is not practically feasible right now, but the internal protocol allows it
437 2012-02-11 18:08:03 <userggf> right
438 2012-02-11 18:08:45 <sipa> BlueMatt: better now?
439 2012-02-11 18:11:59 <BlueMatt> sipa: looks nice, whats the reason for the do {} while(0);?
440 2012-02-11 18:13:05 <sipa> BlueMatt: that's the only way to write a macro that always behaves like a function call, even in combination with nested if-then-elses
441 2012-02-11 18:13:27 <BlueMatt> hmm, odd anyway, looks good
442 2012-02-11 18:13:52 <sipa> it's something you often see inside system headers
443 2012-02-11 18:14:21 <BlueMatt> hmm, fun
444 2012-02-11 18:15:40 <cjd> should be a -Wunbracketed-ifs
445 2012-02-11 18:18:14 <denisx> can someone pls add netinet.h and sys/socket.h to protocol.cpp
446 2012-02-11 18:18:20 <denisx> this is needed on freebsd to build
447 2012-02-11 18:19:01 <sipa> denisx: can you submit a pull request?
448 2012-02-11 18:19:07 <luke-jr> sipa: is there a reason to use a macro instead of an inline?
449 2012-02-11 18:19:34 <sipa> luke-jr: for ENTER_CRITICAL_SECTION? yes, to make __FILE__ and __LINE__ work
450 2012-02-11 18:19:40 <luke-jr> i c
451 2012-02-11 18:40:48 <gmaxwell> sipa: I couldn't figure out if using height was more cool because of orphan filtering, or if using the hash of the inputs to the inputs of the transactions in the prev block (to prevent people from mining who didn't really validate the prior block) was.
452 2012-02-11 20:44:00 <Stupid____> phantomcircuit you banned me from #bitcoin ?
453 2012-02-11 20:44:18 <phantomcircuit> yes your appeal can be filed in /dev/null
454 2012-02-11 20:44:22 <phantomcircuit> very good compression
455 2012-02-11 20:44:28 <Stupid____> thats not fair
456 2012-02-11 20:44:29 <Stupid____> i wasnt the only one
457 2012-02-11 20:44:33 <Stupid____> i didnt even start it
458 2012-02-11 20:45:53 <Stupid____> please unban me
459 2012-02-11 20:47:14 <Stupid____> phantomcircuit>
460 2012-02-11 20:47:15 <Stupid____> ?
461 2012-02-11 20:47:33 <phantomcircuit> no you've been warned by multiple ops over multiple days to stop
462 2012-02-11 20:47:44 <Stupid____> ill stop i swear
463 2012-02-11 20:47:54 <Stupid____> your the only one who warned me
464 2012-02-11 20:47:58 <luke-jr> lol
465 2012-02-11 20:48:05 <luke-jr> Stupid____: get a life
466 2012-02-11 20:48:22 <Stupid____> :(
467 2012-02-11 20:48:35 <Stupid____> thats it, im selling all my bitcoins
468 2012-02-11 20:48:44 <luke-jr> all 2?
469 2012-02-11 20:48:49 <phantomcircuit> i saw gmaxwell kick you several times yesterday
470 2012-02-11 20:49:03 <Stupid____> he was just playing phantomcircuit
471 2012-02-11 20:49:08 <Stupid____> its a little thing we have going ;)
472 2012-02-11 20:49:36 <BlueMatt> if you feel like appealing a ban, please do it in pm
473 2012-02-11 20:50:10 <luke-jr> BlueMatt: phantomcircuit already said /dev/null
474 2012-02-11 20:50:20 <BlueMatt> oh, sorry appeal there then
475 2012-02-11 20:50:26 <Stupid____> its not responding
476 2012-02-11 20:51:28 <phantomcircuit> /ignore Stupid_*!*@* ALL
477 2012-02-11 20:51:38 <Stupid____> phantom your being unreasonable
478 2012-02-11 20:51:57 <Stupid____> whos in charge of the irc channels?
479 2012-02-11 20:52:28 <BlueMatt> phantomcircuit
480 2012-02-11 20:52:34 <Stupid____> thats BS
481 2012-02-11 20:52:42 <BlueMatt> shame
482 2012-02-11 20:52:53 <Stupid____> oh well, i guess i can keep you guys company
483 2012-02-11 20:53:26 <savage_> if you are careful...
484 2012-02-11 20:53:35 <phantomcircuit> i have no interest in arguing about something im not going to change
485 2012-02-11 20:53:46 <phantomcircuit> he's been warned ALOT to stop and refuses
486 2012-02-11 20:53:47 <Stupid____> phantomcircuit, you banned me without any warning whatsoever
487 2012-02-11 20:53:50 <Stupid____> im just asking for a second chance
488 2012-02-11 20:54:08 <Stupid____> you banned me and not any one else too
489 2012-02-11 20:54:10 <phantomcircuit> if he can find at least one other person to make a case i'll listen
490 2012-02-11 20:54:22 <Stupid____> gmaxwell, can you help me out please?!
491 2012-02-11 20:57:43 <Stupid____> u kno what... im gonna mess up the blockchain
492 2012-02-11 20:57:46 <Stupid____> first chance i get :)
493 2012-02-11 20:58:48 <sipa> what has he done, except being stupid?
494 2012-02-11 20:58:54 <pingdrive> master troll my ass
495 2012-02-11 20:58:56 <Stupid____> sipa, i did nothing
496 2012-02-11 20:59:07 <Stupid____> other people were trolling
497 2012-02-11 20:59:09 <Stupid____> i made a few jokes
498 2012-02-11 20:59:13 <Stupid____> and i was banned
499 2012-02-11 20:59:31 <gmaxwell> #bitcoin has been idiotic all day, I haven't been looking at it.
500 2012-02-11 20:59:42 <Stupid____> hey gmaxwell your here : )
501 2012-02-11 20:59:46 <gmaxwell> "Stupid____> u kno what... im gonna mess up the blockchain" < not making a good case for you.
502 2012-02-11 20:59:48 <sipa> is it sometimes non-idiotic there?
503 2012-02-11 20:59:57 <gmaxwell> sipa: there are degrees.
504 2012-02-11 21:00:01 <Stupid____> ok i take that back, i was angry
505 2012-02-11 21:00:08 <Stupid____> did you get my pm gmaxwell?
506 2012-02-11 21:00:20 <Joric> phantomcircuit what have you done
507 2012-02-11 21:00:36 <BlueMatt> can you guys move to #bitcoin-meta or smth? I hate having any disk io (even irc log writing) while benchmarking cblockstore on disk
508 2012-02-11 21:00:49 <sipa> haha
509 2012-02-11 21:01:31 <pingdrive> the channel that noone ever visits
510 2012-02-11 21:01:49 <pingdrive> i guess you are fucked Stupid____
511 2012-02-11 21:02:33 <pingdrive> what dude i did what i could, i have no clout myself
512 2012-02-11 21:02:38 <pingdrive> i said you were legit
513 2012-02-11 21:03:23 <Stupid____> i am
514 2012-02-11 21:03:27 <pingdrive> phantomcircuit run the show so its up to him
515 2012-02-11 21:03:36 <Stupid____> thats not fair
516 2012-02-11 21:03:58 <Stupid____> gmaxwell u get my pm?
517 2012-02-11 21:04:03 <phantomcircuit> Stupid____, you'll be unbanned like tomorrow ish
518 2012-02-11 21:04:07 <pingdrive> he told you find more supporters, do it, he is giving you a chance
519 2012-02-11 21:04:31 <userggf> from all pools, what is the % being mined by p2p pool?
520 2012-02-11 21:04:33 <Stupid____> phantomcircuit: could you unban me a little sooner-ish?
521 2012-02-11 21:04:47 <gribble> 220629019.584
522 2012-02-11 21:04:47 <sipa> ;;bc,p2pool
523 2012-02-11 21:05:01 <Stupid____> phantomcircuit: if i promise to behave
524 2012-02-11 21:05:16 <PK> Stupid____: it's all about timezones. in the UK it's only 2 hours until tomorrow.
525 2012-02-11 21:05:23 <pingdrive> will you join his troll army?
526 2012-02-11 21:05:25 <gmaxwell> userggf: not much, p2pool only recently started growing a lot.
527 2012-02-11 21:05:36 <sipa> ;;calc 100*[bc,p2pool]/([bc,nethash]*1000000)
528 2012-02-11 21:05:38 <gribble> 2.26636331025
529 2012-02-11 21:05:45 <sipa> 2.266% :)
530 2012-02-11 21:05:45 <Stupid____> i guess PK,
531 2012-02-11 21:05:48 <BlueMatt> Stupid____: seriously, shut up no one cares, deal with this in pm or something less intrusive
532 2012-02-11 21:05:56 <Stupid____> ok
533 2012-02-11 21:06:04 <Stupid____> :(
534 2012-02-11 21:06:08 <userggf> 2,26%?
535 2012-02-11 21:06:19 <BlueMatt> shame, p2pool needs to grow
536 2012-02-11 21:06:28 <sipa> it's growing nicely
537 2012-02-11 21:06:33 <gmaxwell> It's been growing very quicky.
538 2012-02-11 21:06:57 <gmaxwell> BlueMatt: http://u.forre.st/p2pool/320.png
539 2012-02-11 21:07:06 <Stupid____> sipa: whats that complicated formula for?
540 2012-02-11 21:07:15 <userggf> why p2pool doesn't appear here http://blockchain.info/pools
541 2012-02-11 21:07:25 <pingdrive> so what is the whole deal in the animosity between p2pool and deepbit?
542 2012-02-11 21:07:33 <forrestv> userggf, i see it on the left
543 2012-02-11 21:07:44 <gmaxwell> userggf: the figures on blockchain.info are largely junk, fwiw.
544 2012-02-11 21:07:47 <BlueMatt> gmaxwell: nice, but again my expectation is that it gets to 50% tomorrow, so thats still a bit slow...
545 2012-02-11 21:07:57 <gmaxwell> pingdrive: huh?
546 2012-02-11 21:08:05 <Stupid____> what gets to 50%?
547 2012-02-11 21:08:22 <gmaxwell> Stupid____: go read the conversation that you've been interupting.
548 2012-02-11 21:08:33 <pingdrive> gmaxwell, i got an impression from the forums that p2pool is against deepbit
549 2012-02-11 21:08:59 <Stupid____> oic
550 2012-02-11 21:09:04 <BlueMatt> the whole point of p2pool is to move bitcoin mining off such high pool-dependency
551 2012-02-11 21:09:14 <BlueMatt> which kinda kills deepbit's business model...
552 2012-02-11 21:09:46 <Stupid____> not really BlueMatt
553 2012-02-11 21:10:12 <userggf> now i see p2pool on the graphic
554 2012-02-11 21:10:18 <marf_away> hello! how can i manipulate the data in my coinbase?
555 2012-02-11 21:10:28 <gmaxwell> pingdrive: I have no clue what you're talking about. But, of course, p2pool users advocate p2pool over centeralized pools.
556 2012-02-11 21:10:34 <marf_away> so i can write stuff inside like satoshi did
557 2012-02-11 21:10:44 <pingdrive> i gotta find it
558 2012-02-11 21:10:55 <pingdrive> there were som anti deepbit ads running
559 2012-02-11 21:11:20 <Stupid____> hey... i was just thinking
560 2012-02-11 21:11:23 <gmaxwell> marf_away: make a one line patch to current bitcoind. Look for COINBASE_FLAGS in the source.
561 2012-02-11 21:11:25 <pingdrive> i am prolly misinterpreting something
562 2012-02-11 21:11:27 <Stupid____> has anyone thought of writing a decentralised email server?
563 2012-02-11 21:11:38 <BlueMatt> ...
564 2012-02-11 21:11:46 <pingdrive> yeh i2p
565 2012-02-11 21:11:57 <gmaxwell> pingdrive: there is one or two people on the forum with some deepbit = evil signature, but I don't see how that has anything to do with p2pool.
566 2012-02-11 21:12:13 <pingdrive> oh yeh that
567 2012-02-11 21:12:15 <Stupid____> i2p has email?
568 2012-02-11 21:12:32 <pingdrive> Stupid____, yes i never used it tho
569 2012-02-11 21:13:31 <pingdrive> i think the same person who is advocating p2pool is saying deepbit is evil
570 2012-02-11 21:13:50 <pingdrive> i thought the was some conflict, i guess there isnt
571 2012-02-11 21:13:51 <pingdrive> nvm
572 2012-02-11 21:13:55 <Stupid____> i donno whats wrong with deepbit, i advocate it to everyone
573 2012-02-11 21:14:07 <luke-jr> Stupid____: advocate Eligius instead
574 2012-02-11 21:14:17 <Stupid____> i might if i get unbanned luke-jr
575 2012-02-11 21:14:21 <pingdrive> deepbit = evil is a pretty strong slogan
576 2012-02-11 21:14:23 <PK> Stupid____: I was thinking about writing a decentralised forum. Each post signed and unalterable, undeleable. True freedom of speech.
577 2012-02-11 21:14:42 <luke-jr> pingdrive: a number of months ago, deepbit had a 50% hold on Bitcoin, which they could have abused
578 2012-02-11 21:14:51 <luke-jr> pingdrive: they stopped advertising on Google, and are only like 1/3 now
579 2012-02-11 21:14:51 <Stupid____> hasnt that been done before PK?
580 2012-02-11 21:14:52 <PK> and advocate btcserv, they can need all the support they can get after getting hacked.
581 2012-02-11 21:14:54 <marf_away> thx gmaxwell
582 2012-02-11 21:15:01 <BlueMatt> pingdrive: deepbit may not be, but to many the idea of bitcoin being centralized is to many people
583 2012-02-11 21:15:12 <BlueMatt> and the heavy centralization of mining is, to many people, evil
584 2012-02-11 21:15:47 <PK> Stupid____: not afaik. But please provide me a URL to such a project. It would be typical for me if my great ideas got implemented already. :) I'm jinxed like that.
585 2012-02-11 21:15:48 <pingdrive> okay, assumingly that all the miners have unified agenda
586 2012-02-11 21:15:55 <pingdrive> miners of Deepbit
587 2012-02-11 21:16:05 <Stupid____> i thought there were a few on tor
588 2012-02-11 21:16:11 <gmaxwell> deepbit producing 27% of the recent transactions and the smartass "trying to promote lite clients" justification for contiuing to do that and not use sendmany is a little evil.
589 2012-02-11 21:16:13 <Stupid____> but my knowledge on that is a bit flaky
590 2012-02-11 21:16:23 <PK> Stupid____: decentralized forums on tor?
591 2012-02-11 21:16:35 <Stupid____> one of the darknets
592 2012-02-11 21:16:42 <Stupid____> might not be tor
593 2012-02-11 21:17:39 <pingdrive> kay got it, somebody is promoting p2pool by saying deepbit is evil
594 2012-02-11 21:18:30 <gmaxwell> pingdrive: can you point me to who is doing that, and I'll tell them to generalize their complaint a bit.
595 2012-02-11 21:18:46 <Stupid____> i think they are more concerned at deepbit getting 51% of the hashing power
596 2012-02-11 21:18:56 <Stupid____> since p2pool cant exploit that, its thought to be better
597 2012-02-11 21:19:08 <Stupid____> personaly, i hope deepbit gets as much hashing power as they can
598 2012-02-11 21:19:28 <cjd> in p2pools defense, they are unique in that no one entity has ultimate power over what the blocks will look like
599 2012-02-11 21:19:35 <gmaxwell> deepbit has _has_ a majority of hash power, and [tycho] didn't give a shit until people DDOSed him off the air. :-/
600 2012-02-11 21:20:12 <gmaxwell> cjd: whats to defend p2pool can't control if someone who promotes it uses a weaksauce argument, though I did just offer to tell them to tone it down.
601 2012-02-11 21:20:25 <pingdrive> gmaxwell, umm i need to look, i am just going of that signature, its owner is the one who is driving the agenda
602 2012-02-11 21:20:26 <gmaxwell> damnit. _has had_
603 2012-02-11 21:20:38 <gmaxwell> pingdrive: there is no p2pool "owner"
604 2012-02-11 21:20:45 <cjd> /nod
605 2012-02-11 21:20:49 <pingdrive> gmaxwell, jsut wanted to see what is the current state of affair and not really do the work
606 2012-02-11 21:21:03 <sipa> Stupid____: what kind of reasoning is that? because p2pool cannot exploit a supermajority, you prefer that one who does gets as much hashrate as possible?
607 2012-02-11 21:21:11 <pingdrive> gmaxwell, okay i guess i dont undestand how it works
608 2012-02-11 21:21:25 <forrestv> pingdrive, i'm the closest thing to the owner, and i've never said that :P
609 2012-02-11 21:21:38 <gmaxwell> pingdrive: forest's (the orignal and primary developer of p2pool) signature just has a donation address.
610 2012-02-11 21:21:45 <PK> there may not be an owner of p2pool, but certainly some main/core developers that could be seen as owners.
611 2012-02-11 21:22:17 <pingdrive> ^ PK
612 2012-02-11 21:22:21 <gmaxwell> PK: forrestv is pretty much the only substantial developer. And he's said no such thing.
613 2012-02-11 21:22:33 <pingdrive> if p2pool is so great how come all the miners arent mining on it
614 2012-02-11 21:22:36 <PK> he is? didn't know that
615 2012-02-11 21:22:52 <pingdrive> it is easy i just got to look up who has the signature
616 2012-02-11 21:23:02 <forrestv> pingdrive, because people make stupid arguments against it like "why isn't everyone using it?"
617 2012-02-11 21:23:05 <gmaxwell> pingdrive: because its relatively new. It only reached 'full' maturity (mining transactions and such) a couple months ago.
618 2012-02-11 21:23:07 <PK> pingdrive: afaik, I was told that p2pool has a fee that goes directly to the main developer. Sort of a donation. Which would be you then?
619 2012-02-11 21:23:11 <gmaxwell> and that.
620 2012-02-11 21:23:23 <gmaxwell> PK: It has a fee like cgminer has a fee.
621 2012-02-11 21:23:34 <PK> an optional one you can turn off?
622 2012-02-11 21:23:37 <forrestv> yes
623 2012-02-11 21:23:37 <gmaxwell> (you can optionally donate)
624 2012-02-11 21:23:37 <pingdrive> PK, i dont own anything
625 2012-02-11 21:23:40 <gmaxwell> pingdrive: yes.
626 2012-02-11 21:23:51 <gmaxwell> er PK
627 2012-02-11 21:24:25 <PK> well, that little detail was left out last time someone in this channel told me about p2pool ^^
628 2012-02-11 21:24:26 <BlueMatt> forrestv: does it handle fee distribution fairly, or are they donated?
629 2012-02-11 21:24:34 <gmaxwell> 14:20 < pingdrive> gmaxwell, jsut wanted to see what is the current state of affair and not really do the work
630 2012-02-11 21:24:42 <gmaxwell> ^ you just wanted to spread FUD and not back it up? :-/
631 2012-02-11 21:24:46 <BlueMatt> PK: I believe it used to be non-optional
632 2012-02-11 21:24:52 <pingdrive> forrestv, how what you said makes sense?
633 2012-02-11 21:25:02 <gmaxwell> BlueMatt: distributed just like the subsidy.
634 2012-02-11 21:25:08 <BlueMatt> nice
635 2012-02-11 21:25:26 <pingdrive> well, considering my current bandwidth, running a browser is not one of my options ATM
636 2012-02-11 21:25:52 <Stupid____> pingdrive: you cant even run a browser? what kinda connection do you have?
637 2012-02-11 21:25:55 <pingdrive> so it does seem like i am spreading FUD (whatsever that means)
638 2012-02-11 21:25:58 <forrestv> pingdrive, half joking, but people ask that question a lot ("why isn't everyone using it?"), so it becomes a bit of a self-fulfilling question
639 2012-02-11 21:26:19 <gmaxwell> BlueMatt: I saw a 22 btc fee in my memory pool the other day, I was sad p2pool didn't get the next block. :)
640 2012-02-11 21:26:45 <BlueMatt> 22 btc fee being freely distributed? wow...
641 2012-02-11 21:26:57 <BlueMatt> gmaxwell: do you p2pool mine?
642 2012-02-11 21:27:00 <gmaxwell> There is even a user on the form who's posts have almost 100% been spreading fud about p2pool. Pretty interesting.
643 2012-02-11 21:27:06 <gmaxwell> BlueMatt: yes.
644 2012-02-11 21:27:11 <BlueMatt> nice
645 2012-02-11 21:27:24 <Stupid____> really gmaxwell?
646 2012-02-11 21:27:28 <Stupid____> im gonna have to study up on his posts
647 2012-02-11 21:27:48 <pingdrive> Stupid____, lol
648 2012-02-11 21:27:50 <gmaxwell> BlueMatt: I like to brag that I was the start of the big p2pool hashrate upswing. :)
649 2012-02-11 21:27:54 <pingdrive> Stupid____, you're a fucking troll
650 2012-02-11 21:28:06 <Stupid____> im not a troll
651 2012-02-11 21:28:07 <BlueMatt> gmaxwell: heh, well congrats
652 2012-02-11 21:28:12 <BlueMatt> Stupid____: yes you are
653 2012-02-11 21:28:22 <Stupid____> why am i a troll?
654 2012-02-11 21:28:30 <cjd> 22btc fee sounds like a mistake, or someone's using it for some "proof of loss" scheme
655 2012-02-11 21:28:39 <gmaxwell> cjd: https://bitcointalk.org/index.php?topic=61984.0
656 2012-02-11 21:28:45 <pingdrive> Stupid____, how is you studying some stupid ass posts relevant?
657 2012-02-11 21:29:01 <Stupid____> well, to try and stop people using p2p
658 2012-02-11 21:29:05 <Stupid____> p2pool
659 2012-02-11 21:29:15 <pingdrive> and why would you wnat to do that?
660 2012-02-11 21:29:22 <PK> Stupid____: why would you want to stop people from using p2pool?
661 2012-02-11 21:29:26 <PK> or what pingdrive said
662 2012-02-11 21:29:51 <Stupid____> because eventually one of the larger pools will win out and their greed will allow them use 51% attacks to mess up the blockchain
663 2012-02-11 21:30:04 <PK> and that's why makes you a troll I guess.
664 2012-02-11 21:30:09 <pingdrive> gmaxwell is making an observatiopn he is not instructing you do shit
665 2012-02-11 21:30:24 <pingdrive> if you are ready for action join the marines
666 2012-02-11 21:31:12 <pingdrive> from what i am hearing p2pool does not threaten anyone with 51% stake in bitcoin
667 2012-02-11 21:31:33 <cjd> that is weird
668 2012-02-11 21:31:34 <Stupid____> pk, im no troll, i have my reasons
669 2012-02-11 21:31:43 <BlueMatt> the ability of miners using p2pool to 51%-attack is equivalent to each individual miner's ability to do so
670 2012-02-11 21:32:23 <cjd> what's probably the most weird about that transaction is just how "normal" it is, asside from the massive fee
671 2012-02-11 21:32:25 <gmaxwell> Stupid____: As far as I can tell, you're taking a contrived position purely for the purpose of getting people to argue with you.
672 2012-02-11 21:32:38 <BlueMatt> aka a troll
673 2012-02-11 21:32:55 <etotheipi_> so then... if 100% of computing power was using P2Pool, we'd be back at regular Bitcoin network, but with all rewards distributed proportionally instead of like a lottery
674 2012-02-11 21:33:09 <BlueMatt> etotheipi_: pretty much
675 2012-02-11 21:33:18 <etotheipi_> what's the downside?
676 2012-02-11 21:33:22 <gmaxwell> etotheipi_: well, p2pool shares are a lottery too, but a much more frequent one with more stable pay.
677 2012-02-11 21:33:24 <Stupid____> etotheipi_: what about the exchanges? their another point of centralisation?
678 2012-02-11 21:33:34 <pingdrive> forrestv gets the fee
679 2012-02-11 21:33:36 <pingdrive> thats the dowside
680 2012-02-11 21:33:43 <BlueMatt> pingdrive: not true
681 2012-02-11 21:33:57 <forrestv> pingdrive, it's an optional donation as of months ago
682 2012-02-11 21:34:18 <BlueMatt> etotheipi_: p2pool may not be be able to scale to quite the size of bitcoin's total mining power, but if that were fixed up a bit with some rule changes, there isnt much of one from a whoelistic pov
683 2012-02-11 21:34:23 <pingdrive> so no fees?
684 2012-02-11 21:34:30 <BlueMatt> s/whoelistic/wholeisit/
685 2012-02-11 21:34:32 <BlueMatt> s/whoelistic/wholeisitc/
686 2012-02-11 21:34:35 <BlueMatt> whatever
687 2012-02-11 21:34:38 <etotheipi_> lol
688 2012-02-11 21:34:38 <gmaxwell> pingdrive: this was answered to you several minutes ago.
689 2012-02-11 21:34:48 <BlueMatt> pingdrive: no, fees are distributed like the regular 50BTC
690 2012-02-11 21:35:06 <Stupid____> gmaxwell: btw just because i like to debate (or argue) with people doesnt mean i should get banned
691 2012-02-11 21:35:12 <pingdrive> can i mine behind tor?
692 2012-02-11 21:35:16 <Stupid____> not does it mean im a troll
693 2012-02-11 21:35:22 <Stupid____> not=nor
694 2012-02-11 21:35:36 <gmaxwell> pingdrive: yes, though you may expirence higher stale rates due to tor latency.
695 2012-02-11 21:35:51 <marf_away> i imagine verry high stale rates
696 2012-02-11 21:35:57 <pingdrive> pingdrive, i am doing fin enow, 1% stale
697 2012-02-11 21:36:03 <gmaxwell> marf_away: tor latency can be pretty good, it differs.
698 2012-02-11 21:36:15 <marf_away> nice to know
699 2012-02-11 21:36:16 <marf_away> :)
700 2012-02-11 21:36:38 <gmaxwell> marf_away: if you wanted you could constrain your circuit path to lower latency nodes.. compromising your anonymity some but improving latency.
701 2012-02-11 21:37:17 <pingdrive> 1.58311346 to be exact, 90 rejected shares out of 5685
702 2012-02-11 21:37:27 <marf_away> kk
703 2012-02-11 21:37:29 <pingdrive> i think its a pretty damn good compromise
704 2012-02-11 21:38:06 <Stupid____> how does p2pool work exactly? i heard it has its own blockchain?
705 2012-02-11 21:38:16 <Stupid____> are there any weaknesses in it?
706 2012-02-11 21:38:20 <BlueMatt> gmaxwell: ok, so Im about as confused as can be with this whole cblockstore performance thing.  Im doing some tests of downloading ~600 blocks at a time and it appears to have a "small but measurable increase in performance"
707 2012-02-11 21:38:33 <BlueMatt> whereas all the very long tests appear to have the opposite
708 2012-02-11 21:38:35 <gmaxwell> Stupid____: it works like bitcoin works. tada.
709 2012-02-11 21:38:51 <Stupid____> is it like a version of bitcoin with the block generation speeded up?
710 2012-02-11 21:38:59 <gmaxwell> Stupid____: no, it's not a currency.
711 2012-02-11 21:39:14 <BlueMatt> gmaxwell: (Ive only ever done long tests on tmpfs->tmpfs, which may be a factor...)
712 2012-02-11 21:39:54 <gmaxwell> Stupid____: it's pretty straight forward. You attempt to mine bitcoin, if your attempts have hash values low enough they are shares. There is a chain of shares. P2pool nodes perform their payout according to the last N shares in the p2pool share chain.
713 2012-02-11 21:40:08 <cjd> if the vast majority of hashing happened in p2pools, it would be nice since you could get weak/early confirmation on transactions just by listening for them in the pool.
714 2012-02-11 21:40:17 <gmaxwell> (N is 24 hours worth of 3*PPLNS, whichever is shorter)
715 2012-02-11 21:40:18 <Stupid____> what are shares?
716 2012-02-11 21:40:30 <gmaxwell> Stupid____: attempted bitcoin solutions.
717 2012-02-11 21:40:34 <Stupid____> oh
718 2012-02-11 21:40:42 <sipa> Stupid____: http://bitcoin.stackexchange.com/questions/1505/what-is-a-share-can-i-find-it-while-mining-solo-or-only-when-pool-mining
719 2012-02-11 21:41:13 <Stupid____> oh
720 2012-02-11 21:41:51 <BlueMatt> gmaxwell: you said you did a bit of cblockstore benchmarking, did you actually look at the time differences or just try to look for potential poor-performance-causing-code
721 2012-02-11 21:42:09 <gmaxwell> The latter.
722 2012-02-11 21:42:26 <gmaxwell> FWIW, litecoin p2pool is about 6% of the litecoin hashrate
723 2012-02-11 21:42:49 <gmaxwell> oh actually 8% now.
724 2012-02-11 21:43:10 <BlueMatt> arg...
725 2012-02-11 21:43:58 <gmaxwell> BlueMatt: have you tried measuring while on a spinning disk rather than tmpfs?
726 2012-02-11 21:44:05 <BlueMatt> thats what Im doing now
727 2012-02-11 21:44:24 <BlueMatt> but I really dont feel like doing enough tests on 5-hour-long benchmarks to get good readings...
728 2012-02-11 21:44:50 <Stupid____> in p2pool who pays you for your hashing?
729 2012-02-11 21:44:52 <BlueMatt> and short block downloads either are hiding the performance impact in disk variations or it just doesnt exist
730 2012-02-11 21:44:57 <sipa> Stupid____: bitcoin
731 2012-02-11 21:45:13 <sipa> same as with other miners
732 2012-02-11 21:45:13 <Stupid____> but i thought payments only come in 50btc per block goes
733 2012-02-11 21:45:23 <etotheipi_> so, about P2Pool... wouldn't there be a "problem" with advancements like BIP 0016, where it seems we're actually dependent on high-level of coordination with 50%+ of the hashpower
734 2012-02-11 21:45:44 <etotheipi_> for instance, how would we roll out something like BIP 0016 if everyone was on P2Pool?
735 2012-02-11 21:45:53 <sipa> Stupid____: yes, and with p2pool you agree to create blocks which are distributed over other people, proporational to their hashing rate
736 2012-02-11 21:45:58 <sipa> Stupid____: and they do the same for you
737 2012-02-11 21:46:02 <BlueMatt> you would have to wait for 50% of p2pool users to upgrade
738 2012-02-11 21:46:07 <BlueMatt> (if p2pool had 100% hash power)
739 2012-02-11 21:46:37 <Diablo-D3> what broke now
740 2012-02-11 21:46:38 <gmaxwell> etotheipi_: by putting it in new software and making the actual activation triggered by the coinbases automatically. P2pool could also intentionally stop rewarding shares to nodes that aren't upgraded.
741 2012-02-11 21:46:57 <Stupid____> im not gonna upgrade :)
742 2012-02-11 21:47:17 <etotheipi_> gmaxwell, your last sentence suggests that one person, or a small group can "intentionally" decide something
743 2012-02-11 21:47:20 <etotheipi_> who is that person/group?
744 2012-02-11 21:47:22 <sipa> Stupid____: any reason why you wouldn't update?
745 2012-02-11 21:47:28 <pingdrive> yeh fuck that
746 2012-02-11 21:47:39 <pingdrive> sipa, just to prove a point
747 2012-02-11 21:47:41 <Stupid____> because i got banned from the bitcoin channel