1 2012-02-01 00:00:09 <luke-jr> BlueMatt: you can't guarantee what the scripts do
  2 2012-02-01 00:00:17 <luke-jr> without forcing IsStandard on protocol rules
  3 2012-02-01 00:00:34 <BlueMatt> yea you can, if (evalwithoutopcat() && evalwithopcat())
  4 2012-02-01 00:00:48 <gmaxwell> I just happened to make a list of features I want in bitcoin, and perhaps amusingly no script improvements made my list: https://en.bitcoin.it/wiki/User:Gmaxwell/features
  5 2012-02-01 00:00:54 <luke-jr> BlueMatt: good point
  6 2012-02-01 00:00:54 <[Tycho]> luke-jr: may be like OP_NEWSCRIPTS instead of OP_CAT_NEW :)
  7 2012-02-01 00:01:05 <luke-jr> [Tycho]: OP_HASKELL_EVAL
  8 2012-02-01 00:01:27 <[Tycho]> No one needs haskell.
  9 2012-02-01 00:01:29 <BlueMatt> OP_EVAL serializedscript
 10 2012-02-01 00:01:32 <BlueMatt> :)
 11 2012-02-01 00:01:34 <graingert> what about dart
 12 2012-02-01 00:01:39 <BlueMatt> then you can enable whatever the hell you want
 13 2012-02-01 00:01:47 <[Tycho]> Serialized scripts are ugly.
 14 2012-02-01 00:01:56 <BlueMatt> yea, but it works
 15 2012-02-01 00:02:18 <luke-jr> BlueMatt: so does BIP 17 ;)
 16 2012-02-01 00:02:25 <BlueMatt> ...
 17 2012-02-01 00:02:31 <BlueMatt> gmaxwell: scrypt wallet encryption - go do it, it shouldnt take more than like 20 minutes
 18 2012-02-01 00:03:13 <gmaxwell> BlueMatt: 1/4 of the stuff there isn't hard. Also not very important either.
 19 2012-02-01 00:03:18 <luke-jr> is there something wrong with AES? O.o
 20 2012-02-01 00:03:35 <gmaxwell> luke-jr: it would still be AES. Scrypt is the KDF not the actual encryption
 21 2012-02-01 00:03:36 <splatster> I have given up trying to understand the differences to between BIP12/16/17/
 22 2012-02-01 00:03:50 <gmaxwell> splatster: good. To most people there is no difference.
 23 2012-02-01 00:04:22 <BlueMatt> gmaxwell: you forgot a nice api to the core bitcoin block/tx store/validation/etc that all other parts use
 24 2012-02-01 00:04:23 <splatster> Well if there wasn't, no one would argue over it so much.
 25 2012-02-01 00:04:27 <BlueMatt> gmaxwell: oh wait, CBlockStore ;)
 26 2012-02-01 00:05:20 <gmaxwell> yea, I didn't mention any API things (except seperating the interface, I guess) because I've never had an opinion on API, and I try not to have one!
 27 2012-02-01 00:05:42 <BlueMatt> heh, ok
 28 2012-02-01 00:07:01 <BlueMatt> (before 0.6 is released)
 29 2012-02-01 00:08:48 <gmaxwell> etotheipi_: So on BIP10, one complication is that the offline client would be forced to show the change as an output when it asks for approval. Seems unfortunate to me.
 30 2012-02-01 00:09:23 <gmaxwell> etotheipi_: I suppose it could detect it by seeing if its a key it controls, but then it would potentially have to enumerate its wallet chains quite far to find it.
 31 2012-02-01 00:09:52 <etotheipi_> gmaxwell, that is true... I autodetect that in Armory (doing exactly what you suggested), but the logic wasn't trivial
 32 2012-02-01 00:10:54 <etotheipi_> but for an arbitrary transaction, you are right, we don't know the change output
 33 2012-02-01 00:12:11 <gmaxwell> I'm worried about the computation required to detect it. You could flag it and the device could show the flag but the flag might be a lie. :(
 34 2012-02-01 00:12:22 <etotheipi_> I was just about to say that
 35 2012-02-01 00:12:24 <gmaxwell> Your determinstic keys don't give random access.
 36 2012-02-01 00:12:32 <gmaxwell> otherwise you could just give the relevant index.
 37 2012-02-01 00:12:38 <graingert> have two sets of deterministic keys
 38 2012-02-01 00:12:39 <etotheipi_> what do yo umean?
 39 2012-02-01 00:12:44 <graingert> one for change and one for regualar
 40 2012-02-01 00:12:48 <etotheipi_> i can access keys by hash160 or index
 41 2012-02-01 00:13:14 <etotheipi_> it's just a matter of how you read them in from file into RAM data structures
 42 2012-02-01 00:13:23 <gmaxwell> etotheipi_: I mean if the change is to the 283498th index you're going to be doing a lot of computation on what may be some 10mw security hardened arm core.
 43 2012-02-01 00:16:53 <gmaxwell> You can buy tamperresistant single chip computers off the shelf thats what I'd really want my offline wallet to be. But they're not speed demons... fine for signing a transaction in subsecond times, not so much for enumerating long determinstic key chains.
 44 2012-02-01 00:18:33 <etotheipi_> gmaxwell, that was one of the downsides of moving from the "linear" to hash-based deterministic wallet
 45 2012-02-01 00:18:50 <etotheipi_> the old method, I could generate like 100,000 private keys per second
 46 2012-02-01 00:19:11 <etotheipi_> the new method is bottlenecked by the priv->pub compute time
 47 2012-02-01 00:19:12 <poiuh> what happens if allocatoin fails during unserialization
 48 2012-02-01 00:19:28 <poiuh> std::bad_alloc ?
 49 2012-02-01 00:19:47 <etotheipi_> poiuh, what unserialization?
 50 2012-02-01 00:20:23 <poiuh> serialize.h
 51 2012-02-01 00:20:24 <etotheipi_> gmaxwell, you have to make some assumption about the private keys... in my case I assume that there will never be a gap of more than 100 keys without one being used
 52 2012-02-01 00:20:40 <etotheipi_> (customizable for the application... could be 10,000 for a big distributor)
 53 2012-02-01 00:21:41 <etotheipi_> in my case, if I don't make *some* assumption like that, then I do run off into undefined-compute-time
 54 2012-02-01 00:22:14 <etotheipi_> I could be searching for a key I don't even have.... generate a couple-hundred-thousand addresses before the battery is drained
 55 2012-02-01 00:23:39 <poiuh> main.cpp:2657:    catch (std::ios_base::failure& e)
 56 2012-02-01 00:23:42 <poiuh> i guess it would be caught there
 57 2012-02-01 00:24:11 <poiuh> or rather at 2674:         catch (std::exception& e) {
 58 2012-02-01 00:37:48 <luke-jr> O.o BIP 22 isn't very good itself (blockchain fork), but does raise some interesting ideas
 59 2012-02-01 00:38:46 <luke-jr> am I correct in thinking that if a pubkey has extra crap on the end, it will be ignored by current clients?
 60 2012-02-01 00:39:25 <gmaxwell> umm.
 61 2012-02-01 00:39:26 <gmaxwell> wtf
 62 2012-02-01 00:39:45 <gmaxwell> it's embedding another scripting system in the scripting system, and its completely incompatible.
 63 2012-02-01 00:39:54 <gmaxwell> Why do we give BIPs for non-starter ideas? :(
 64 2012-02-01 00:40:18 <luke-jr> no idea, but think of this:
 65 2012-02-01 00:40:33 <luke-jr> what if all multisigs were A && (&) ?
 66 2012-02-01 00:40:38 <poiuh> is there a bip for DOUBLE_SPEND_NOTICE
 67 2012-02-01 00:40:40 <luke-jr> could they be hidden inside the pubkey?
 68 2012-02-01 00:40:58 <luke-jr> poiuh: no, if you want to write one, I think it'd be a good idea
 69 2012-02-01 00:40:58 <tcatm> We should have a BIP review process before they get assigned a number at all.
 70 2012-02-01 00:41:28 <etotheipi_> I have a reference implementation in Armory, but it's not very useful when the satoshi client drops double-broadcast tx before I see them :(
 71 2012-02-01 00:41:30 <gmaxwell> poiuh: no, it came up on the mailing list a while back.. a couple of non-starters were proposed (DOS vulnerable, attack magnets) and the provable type came up eventually.
 72 2012-02-01 00:41:56 <poiuh> interesting
 73 2012-02-01 00:42:03 <poiuh> gmaxwell: provable type?
 74 2012-02-01 00:42:29 <gmaxwell> https://en.bitcoin.it/wiki/User:Gmaxwell/features
 75 2012-02-01 00:42:30 <gmaxwell> Provable notice of doublespend (when a node sees a doublespend it prepares an alert indexed by the doublespent output, including at least two spends of it, and floods it, if a client receives a such a notice for a transaction sent to it, it shoots fireworks and greatly delays showing it as confirmed)
 76 2012-02-01 00:42:41 <poiuh> oh cool
 77 2012-02-01 00:43:10 <gmaxwell> The key there is that it actually shows its peers the double spending signatures, so it can't be faked. And if it's identified by input hash, it can't be too much of a flood storm creater
 78 2012-02-01 00:43:41 <poiuh> seems like a cool idea
 79 2012-02-01 00:45:02 <luke-jr> gmaxwell: anyway, do you think that's possible? BIP 22 restricted to A && (&) cases?
 80 2012-02-01 00:45:29 <luke-jr> actually
 81 2012-02-01 00:45:35 <luke-jr> that'd have the static analysis issue, wouldn't it?
 82 2012-02-01 00:45:46 <poiuh> beyond that.. far in the future, you might want a way to query a random node about a tx.  command="tx_info"   the node would respond with details if it saw the tx, if it considers it valid, if it was marked dblspend etc.    that way you can ping random nodes to get a feel for whether or not the tx you're interested in was propagated somewhere
 83 2012-02-01 00:46:49 <poiuh> the point being to increase confidence in quick payments without waiting for the next block
 84 2012-02-01 00:47:31 <gmaxwell> poiuh: oh that reminds me of another one
 85 2012-02-01 00:49:03 <gmaxwell> *P2Pool
 86 2012-02-01 00:49:04 <gmaxwell> **If P2Pools have a significant amount of hashpower, and their shares carry commitments to transactions, P2Pool can create sub-minute evidence of intent to mine  coupled with double-spend-alerts can create pretty good fast confidence.
 87 2012-02-01 00:49:20 <gmaxwell> (more a p2pool feature than a bitcoin one, but it's a point of integration)
 88 2012-02-01 00:49:29 <poiuh> hmm pretty cool
 89 2012-02-01 00:50:09 <gmaxwell> poiuh: your querying is a good idea too you don't even need a protocol feature for it. nodes won't remember double spends, if they remember the transaction then its the winner from their perspective.
 90 2012-02-01 00:50:15 <gmaxwell> So just ask nodes for the txn in question.
 91 2012-02-01 00:50:22 <poiuh> yeah makes sense
 92 2012-02-01 00:51:00 <gmaxwell> though personally I think someone should run a confirmation insurance service that uses stuff like that to offer vendors reversal insurance for a modest fee.
 93 2012-02-01 00:51:12 <poiuh> yeah
 94 2012-02-01 00:51:43 <poiuh> for small tx like $20 in gas, i think most vendors would be fine with some doublespend alerts / confidence in propagated tx
 95 2012-02-01 00:52:03 <BlueMatt> gmaxwell: amen
 96 2012-02-01 00:54:03 <gmaxwell> Well, unfortunately when I bring this up around bitcoiners a .. frightening number of them seem to assume such a service would involve contracts with miners to rewrite the chains. 0_o
 97 2012-02-01 00:54:14 <gmaxwell> must be some cultural gap about what "insurance" means.
 98 2012-02-01 00:54:38 <gmaxwell> In my world insurance involves pocket protectors and actuarials.
 99 2012-02-01 00:54:51 <k9quaint> insurance means geckos with australian accents
100 2012-02-01 00:54:59 <BlueMatt> or just a desire to think everything bitcoin is a conspiracy
101 2012-02-01 00:55:00 <k9quaint> I don't want geckos in my block chain
102 2012-02-01 00:55:29 <gmaxwell> BlueMatt: I don't even mean having negative opinions about it being that. 0_o
103 2012-02-01 00:56:08 <k9quaint> bluematt: I think Solidcoin serves a valid purpose by syphoning off the worst conspiracy nutjobs so BTC doesn't have to deal with em
104 2012-02-01 00:56:19 <BlueMatt> heh
105 2012-02-01 00:56:27 <gmaxwell> k9quaint: except we do, because it becomes their holy mission to attack bitcoin.
106 2012-02-01 00:56:32 <k9quaint> I mean, try to picture coinhunter trying to argue about BIPS
107 2012-02-01 00:56:37 <gmaxwell> k9quaint: http://solidcoinmafia.com/
108 2012-02-01 00:56:47 <gmaxwell> k9quaint: he's currently doing so, you haven't seen that thread? :)
109 2012-02-01 00:57:04 <k9quaint> you clicked on a thread with Coinhunter talking about code in it?
110 2012-02-01 00:57:08 <gmaxwell> becuase this BIP thing involves Luke and the altcoiners hate luke... well..
111 2012-02-01 00:57:08 <k9quaint> your problem, not mine
112 2012-02-01 00:57:25 <gmaxwell> k9quaint: you can't tell from the bitcointalk urls when people paste them in irc. :(
113 2012-02-01 00:57:37 <k9quaint> btw, I modded BTC so that gavin can't alert me
114 2012-02-01 00:57:49 <gmaxwell> you ninja you.
115 2012-02-01 00:58:19 <gmaxwell> k9quaint: you ever see what the alerts look like in bitcoind? they're cool
116 2012-02-01 00:58:33 <gmaxwell> They show up in that useless errors field in getinfo.
117 2012-02-01 00:59:25 <k9quaint> don't try to use your mind control lasers on me sir, they NO LONGER WORK
118 2012-02-01 01:15:28 <Graet> hmm interesting, imnsurance to me means someone pays me not to go to work for a bit longer
119 2012-02-01 01:38:24 <avivmas_> hi
120 2012-02-01 01:38:48 <avivmas_> whats up
121 2012-02-01 01:38:56 <avivmas_> nakamoto ded?
122 2012-02-01 01:41:02 <BlueMatt> you mean satoshi, no one has heard from him in months and months
123 2012-02-01 01:41:10 <BlueMatt> nearly a year in a few months
124 2012-02-01 01:43:36 <luke-jr> he left
125 2012-02-01 01:49:27 <Ferroh> ;;bc, stats
126 2012-02-01 01:49:28 <gribble> Error: "bc," is not a valid command.
127 2012-02-01 01:50:49 <luke-jr> grep fTimeReceivedIsTxTime src/*.cpp <-- makes me very confused :|
128 2012-02-01 01:55:56 <poiuh> etotheipi_ is a smart dude
129 2012-02-01 01:57:11 <etotheipi_> poiuh, where did that come from?
130 2012-02-01 01:57:24 <etotheipi_> are you listening to my conversation with my girlfriend?
131 2012-02-01 01:58:08 <etotheipi_> (thanks, btw)
132 2012-02-01 02:07:53 <etotheipi_> gmaxwell, do you have a recommendation for the change-output issue?  It seems to me that for a multi-sig tx, change isn't really an issue (the single input had a specific value), and for your own tx, you either would recognize the change output yourself, or Armory will be able to identify based on your wallet
133 2012-02-01 02:08:32 <etotheipi_> I guess, for your own tx, it could have a comment identifying which is which (in the TxDP)
134 2012-02-01 02:08:48 <etotheipi_> actually, that's one update to BIP 0010 that might be worth adding:  comment lines
135 2012-02-01 02:09:07 <etotheipi_> they could be used for communicating between parties, or clients could insert special strings for themselves to understand
136 2012-02-01 02:09:23 <gmaxwell> etotheipi_: my concern was mostly offline wallets where the private keys were limited devices and couldn't easily do the smart identification.
137 2012-02-01 02:09:38 <gmaxwell> A flag would be nice so they could at least be smarter about their search.
138 2012-02-01 02:09:47 <gmaxwell> and yes, I think a comment field is important.
139 2012-02-01 02:09:54 <gmaxwell> "which one was this? shit"
140 2012-02-01 02:14:34 <etotheipi_> Just hash-prefixed lines?
141 2012-02-01 02:15:03 <etotheipi_> since it is parsed line-by-line, I guess any line anywhere that starts with a '#' can be ignored
142 2012-02-01 02:15:24 <phantomcircuit> BlueMatt, well... nobody that is admitting it
143 2012-02-01 02:15:25 <etotheipi_> and then clients can put whatever they want there... human-readable, or just information for the offline system
144 2012-02-01 02:17:59 <BlueMatt> phantomcircuit: ???
145 2012-02-01 02:18:23 <phantomcircuit> <BlueMatt> you mean satoshi, no one has heard from him in months and months
146 2012-02-01 02:18:29 <BlueMatt> oh
147 2012-02-01 02:18:31 <BlueMatt> yea
148 2012-02-01 02:19:48 <poiuh> coiledcoin
149 2012-02-01 02:20:25 <etotheipi_> actually, while I'm at it, and before I do my first release, any other recommendations for BIP 0010?  The choice of formatting/strings might be suboptimal...
150 2012-02-01 02:21:25 <splatster> et
151 2012-02-01 02:21:28 <splatster> oops
152 2012-02-01 02:21:51 <splatster> etotheipi_: Is RAM now not so intensively used?
153 2012-02-01 02:25:15 <da2ce7> ooh https://en.bitcoin.it/wiki/BIP_0022
154 2012-02-01 02:25:46 <gmaxwell> da2ce7: that's dead on arrival. It requires a chain fork.
155 2012-02-01 02:26:10 <gmaxwell> da2ce7: it also embeds a seperate scripting language inside the scripting language.
156 2012-02-01 02:26:32 <da2ce7> ok.
157 2012-02-01 02:27:00 <da2ce7> maybe mike should update his BIP to indicate it requires a fork.
158 2012-02-01 02:27:02 <gmaxwell> etotheipi_: I don't think the BIP as written is rigorous, e.g. that without looking at your code I could write my own and and up successfully parsing the same inputs as you.
159 2012-02-01 02:27:13 <gmaxwell> da2ce7: it does, around page 10 of it or so.
160 2012-02-01 02:27:17 <splatster> Which one requires the least change for the protocol and all of the bitcoin nodes?
161 2012-02-01 02:28:02 <gmaxwell> splatster: BIP16 / BIP17 are basically equal. 17's patch is a bit smaller, but adding the checksig fix on top of it will probably make it equal.
162 2012-02-01 02:29:00 <splatster> Which one will be easiest for custom nodes to implement?
163 2012-02-01 02:29:26 <gmaxwell> Again, they're roughly equal.
164 2012-02-01 02:29:48 <splatster> So why do people prefer one over the other?
165 2012-02-01 02:30:29 <luke-jr> da2ce7: with some revision, it wouldn't require a fork; but it would break roconor's static analysis thing
166 2012-02-01 02:30:35 <gmaxwell> splatster: because the essence of being a geek is caring about things no one cares about.. for better or worse. Sometimes things no one should care about.
167 2012-02-01 02:30:51 <luke-jr> gmaxwell: checksig fix?
168 2012-02-01 02:30:59 <gmaxwell> luke-jr: the sigops counting.
169 2012-02-01 02:31:12 <luke-jr> oh, non-checkmultisig multisig?
170 2012-02-01 02:31:40 <gmaxwell> Is _that_ how it was proposed to fix it in 17? I didn't actually see the proposal for that.
171 2012-02-01 02:31:40 <luke-jr> it might.
172 2012-02-01 02:31:48 <gmaxwell> I just knew there was one.
173 2012-02-01 02:31:52 <luke-jr> gmaxwell: I was waiting for genjix to give me a BIP number& :p
174 2012-02-01 02:32:12 <splatster> I would consider myself a geek, I just don't obsess over stupid shit.
175 2012-02-01 02:32:13 <luke-jr> since it's a future problem, I don't really consider it a priority
176 2012-02-01 02:32:35 <gmaxwell> luke-jr: Can we have some pre-bip process? I don't think new ideas should be introduced via bips. It's really sad to see people do a lot of careful writing just to get the "ummm. no" :(
177 2012-02-01 02:32:52 <luke-jr> gmaxwell: I also posted for reviews in here a few days ago ;)
178 2012-02-01 02:32:56 <gmaxwell> splatster: whatever you are interested is not stupid to you, it might be stupid to other people.
179 2012-02-01 02:33:04 <gmaxwell> luke-jr: I saw you do that. sorry!
180 2012-02-01 02:33:06 <luke-jr> http://pastebin.com/cw59JfnU
181 2012-02-01 02:34:29 <gmaxwell> hm. it's a little more data, but not that much.
182 2012-02-01 02:35:27 <luke-jr> I think it evens out the data size when added to BIP 17
183 2012-02-01 02:35:35 <luke-jr> since BIP 17 used slightly less
184 2012-02-01 02:53:31 <etotheipi_> splatster, alpha release will still be full-RAM
185 2012-02-01 02:55:22 <etotheipi_> the RAM reduction is first priority after alpha, along with address books (right now it's a pain to move money between wallets, or find old addresses)
186 2012-02-01 03:15:42 <casascius> Random idea: any way to get the channel log moved to a better host that doesn't serve advertising that attempts exploits?  On multiple occasions, I have had the channel log page try to hijack my machine, such as with fake antivirus.
187 2012-02-01 03:15:46 <casascius> This happened again just now
188 2012-02-01 03:16:51 <BlueMatt> casascius: tell cdecker about that
189 2012-02-01 03:16:56 <BlueMatt> (he runs it)
190 2012-02-01 03:17:17 <BlueMatt> Im sure he'll be more than happy to deal with it
191 2012-02-01 03:22:16 <casascius> thanks
192 2012-02-01 03:27:30 <poiuh> satoshi officially endorses coiledcoin
193 2012-02-01 03:29:57 <luke-jr> does that mean I should stop demanding 100 CLC fees on txns?
194 2012-02-01 03:30:12 <poiuh> haha
195 2012-02-01 03:32:10 <nanotube> casascius: i can put up gribble's -dev chanlog as a backup if you prefer. :)
196 2012-02-01 03:34:16 <casascius> no prob by me, it is not somehting i access frequently, just something i thought maybe nobody was aware of
197 2012-02-01 04:15:36 <etotheipi_> so wtf is this tonal bitcoin stuff?
198 2012-02-01 04:16:00 <etotheipi_> I have no idea what it offers anyone, besides a lot of confusion...
199 2012-02-01 04:17:03 <etotheipi_> is there a reason I should consider figuring it out?
200 2012-02-01 04:17:25 <BlueMatt> no
201 2012-02-01 04:17:32 <BlueMatt> (unless you are luke)
202 2012-02-01 04:21:07 <casascius> all you need to know about tonal is 2 things, 1 luke promotes it as easier for humans, and 2, in tonal, 5+5=9
203 2012-02-01 04:21:37 <gmaxwell> This is all you need to know: https://www.youtube.com/watch?v=SXx2VVSWDMo
204 2012-02-01 04:21:57 <gmaxwell> etotheipi_: it's just base 16 with funny symbols.
205 2012-02-01 04:23:12 <casascius> and one of those funny symbols is the glyph "9" and has a value of ten
206 2012-02-01 04:23:30 <etotheipi_> what is the point of it?
207 2012-02-01 04:23:31 <gmaxwell> oh wow, that video doesn't have the part of the song where he does it in base 8.
208 2012-02-01 04:23:35 <casascius> i caught luke getting confused by this in one of his posts
209 2012-02-01 04:24:13 <gmaxwell> https://www.youtube.com/watch?v=DfCJgC2zezw  < oh this one has it.
210 2012-02-01 04:24:45 <gmaxwell> etotheipi_: http://en.wikipedia.org/wiki/Tonal_System#Tonal_System_.28Hexadecimal.29
211 2012-02-01 04:24:53 <etotheipi_> there must be some justifiable reason for coming up with it
212 2012-02-01 04:25:37 <casascius> the justifiable reason is based on the notion that fractions are more natural when based on powers of two than ten
213 2012-02-01 04:26:01 <gmaxwell> etotheipi_: twos rich bases make subdividing easier.
214 2012-02-01 04:26:10 <gmaxwell> This sort of thinking is why there are 360 degrees in a circle.
215 2012-02-01 04:26:33 <gmaxwell> (2*2*2*3*3*5, nice and factor rich)
216 2012-02-01 04:27:07 <etotheipi_> gotcha
217 2012-02-01 04:27:10 <luke-jr> etotheipi_: http://books.google.com/books?id=aNYGAAAAYAAJ&printsec=frontcover
218 2012-02-01 04:27:49 <gmaxwell> It's not a bad idea, except for the whole being-compatible-is-more-important-than-everything detal that makes lots of things choose lesser options.
219 2012-02-01 04:28:51 <casascius> i would find it less peculiar if 1 TBC = 1 BTC and that TBC was just BTC written in hex (e.g. 1000 TBC is just 4096 BTC)... but he has his definition pegged to the satoshi, so rather, it's 0.00001000 TBC = 0.00004096 BTC
220 2012-02-01 04:28:53 <etotheipi_> I can see, if you grew up learning that and nothing else (in a vacuum), it might be slightly beneficial... but in a world full of established number system...
221 2012-02-01 04:29:05 <casascius> and 1 TBC is like 42.94967296 BTC (2^32 satoshi)
222 2012-02-01 04:29:21 <gmaxwell> etotheipi_: it's probably beneficial to be fluent in multiple systems however gives you more mental tools for working with things.
223 2012-02-01 04:29:22 <casascius> i fully agree, if we were in the stone age inventing numbers, it would be a great thing
224 2012-02-01 04:29:39 <luke-jr> 1 TBC = 0.00065536 BTC fwiw
225 2012-02-01 04:29:49 <gmaxwell> (and, while I know nothing about the tonal symbols, I'm comfortable with doing arithemetic in base-2 and 16 and I find it useful)
226 2012-02-01 04:29:55 <casascius> yeah, I knew it was something like that, but it's so damn confusing =)
227 2012-02-01 04:30:14 <TuxBlackEdo> mmm bitcoin bongs
228 2012-02-01 04:31:07 <casascius> what will they grow up to be?
229 2012-02-01 04:31:10 <TuxBlackEdo> i can only imagine how luke-jr's children will turn out
230 2012-02-01 04:31:34 <luke-jr> casascius: well-rounded polymaths hopefully? :p
231 2012-02-01 04:32:06 <casascius> why not just use hexadecimal instead of tonal
232 2012-02-01 04:32:38 <casascius> using tonal symbols is like writing english with Mormon Deseret Alphabet characters (which look very much like tonal)
233 2012-02-01 04:32:46 <luke-jr> casascius: because tonal was already there and ready to use. I didn't want to invent something new.
234 2012-02-01 04:32:47 <gmaxwell> luke-jr: my girlfriend recommends this: http://en.wikipedia.org/wiki/Sideways_Arithmetic_From_Wayside_School
235 2012-02-01 04:33:07 <casascius> isn't hexadecimal already there and easy to use, and "9" always means nine as a bonus?
236 2012-02-01 04:33:12 <luke-jr> inventing something new would mean I'd *really* be the only one using it
237 2012-02-01 04:33:30 <luke-jr> casascius: no, hexadecimal doesn't have pronunciations, units of measure, etc
238 2012-02-01 04:33:39 <luke-jr> calendar, etc etc
239 2012-02-01 04:33:47 <gmaxwell> casascius: common hex usage doesn't have the negative exponent usage either.
240 2012-02-01 04:34:13 <gmaxwell> The closest you get is binary, where binary floating point is common, but we always end up poorly converting it to dec and thus don't have names for it.
241 2012-02-01 04:34:19 <luke-jr> casascius: plus, letters make for bad digits
242 2012-02-01 04:34:54 <casascius> Why do letters make bad digits?  Will someone confuse 0xdeadbeef with meat?
243 2012-02-01 04:35:32 <luke-jr> someone might confuse a fish with 1 fish
244 2012-02-01 04:35:46 <casascius> sure is better than someone confusing "9" (0x09) with "9" (0x0a)
245 2012-02-01 04:35:59 <gmaxwell> yes. I mean .. 2ax=3 .. is that 2*a*x=3 or 2a*x = 3?
246 2012-02-01 04:36:09 <gmaxwell> luke-jr: whats the reason for the 9swap btw?
247 2012-02-01 04:36:24 <luke-jr> casascius: that's minor considering the decimal digits are recycled at all; I'd personally prefer 0x10 new digits entirely
248 2012-02-01 04:36:44 <casascius> if we considered "a" to mean ten, we'd never use it as a variable in algebra
249 2012-02-01 04:36:44 <luke-jr> gmaxwell: I presume to mirror the low-bit digit 6
250 2012-02-01 04:37:03 <luke-jr> gmaxwell: the high-bit digits are inverted low-bit digits to some extent
251 2012-02-01 04:37:05 <gmaxwell> luke-jr: clever
252 2012-02-01 04:37:06 <casascius> anymore than one would ask, 2+9=3, solve for 9
253 2012-02-01 04:37:44 <gmaxwell> casascius: A number was found where a variable was expected -`solve'
254 2012-02-01 04:37:59 <luke-jr> I suspect the original reason he kept the decimal digits was the intention of replacing decimal overnight globally
255 2012-02-01 04:38:01 <casascius> but 6 inverted is nine, not ten.  so why shouldn't "9" mean nine if that's the case
256 2012-02-01 04:38:16 <luke-jr> casascius: &
257 2012-02-01 04:38:23 <gmaxwell> luke-jr: yea, actually at the time it was proposed it was probably possible to think of that.
258 2012-02-01 04:38:46 <luke-jr> gmaxwell: the whole reason decimal won out in SI was because the treaties establishing SI specified it -.-
259 2012-02-01 04:38:55 <luke-jr> casascius: http://upload.wikimedia.org/wikipedia/commons/e/e6/Hexadecimal_Clock_by_Nystrom.jpg
260 2012-02-01 04:39:09 <gmaxwell> luke-jr: to be fair, decimal was being widely used before then.
261 2012-02-01 04:39:11 <luke-jr> casascius: observe how the right side mirrors the left
262 2012-02-01 04:39:29 <luke-jr> gmaxwell: sure, but so was imperial units
263 2012-02-01 04:39:38 <luke-jr> and other dozenal and tonal native units
264 2012-02-01 04:39:56 <casascius> i have a mind to manufacture you that clock on my laser engraver
265 2012-02-01 04:39:57 <gmaxwell> yea, we still do use dozenal units too.
266 2012-02-01 04:40:37 <luke-jr> casascius: I'm seriously looking for gears to build a tonal grandfather clock if you know where I could get those :p
267 2012-02-01 04:40:45 <luke-jr> it's surprisingly difficult to find
268 2012-02-01 04:40:49 <casascius> do you have the face of that clock
269 2012-02-01 04:40:57 <gmaxwell> why would you need gears? you just change the face?
270 2012-02-01 04:41:18 <gmaxwell> or is the tonal minute a different length? I guess it would be.
271 2012-02-01 04:41:20 <etotheipi_> he'll only take gears if the number of teeth is a power of 2
272 2012-02-01 04:41:22 <gmaxwell> Does tonal time use the SI second?
273 2012-02-01 04:41:26 <luke-jr> gmaxwell: seconds hand is the problem
274 2012-02-01 04:41:27 <luke-jr> no
275 2012-02-01 04:41:38 <gmaxwell> Ah. meh. You should fix that.
276 2012-02-01 04:41:46 <luke-jr> tonal timmill is 1.318359375 seconds
277 2012-02-01 04:41:54 <gmaxwell> SI second is the hardest one to change.
278 2012-02-01 04:42:02 <gmaxwell> bleh I guess you can't.
279 2012-02-01 04:42:03 <luke-jr> also, the dozenal clock has 20; hours per day
280 2012-02-01 04:42:26 <casascius> what size do said gears need to be
281 2012-02-01 04:42:26 <luke-jr> tonal clock is 10 tims per day
282 2012-02-01 04:42:39 <luke-jr> casascius: it's been a while& I'd have to redo the math I suspect
283 2012-02-01 04:42:56 <casascius> i would have no problem creating a metallic face of a tonal clock
284 2012-02-01 04:43:01 <gmaxwell> etotheipi_: We're all familar with the theory behind gears with teeth which are natural numbers in cardinality...
285 2012-02-01 04:43:46 <casascius> i have no problem with the fascination with tonal numbers to begin with, of course i dont think it belongs in bitcoin, but i'd make you a tonal clock if it was as easy as pushing a button
286 2012-02-01 04:44:06 <casascius> do you have your tonal clock face in a vectorized form
287 2012-02-01 04:44:47 <luke-jr> well, this is primarily a school project, so I was planning to have my children draw one on cardboard <.<
288 2012-02-01 04:45:16 <luke-jr> I'm not aware of any vectorized clock face
289 2012-02-01 04:45:40 <casascius> you should vectorize it and i'll burn it on a piece of wood.  good luck finding a movement in the right units, but it will at least look neat
290 2012-02-01 04:45:44 <luke-jr> I suppose it shouldn't be too difficult to make though& I do have the tonal characters in SVG
291 2012-02-01 04:46:37 <etotheipi_> luke-jr, aren't there more useful things you could be doing with your time?
292 2012-02-01 04:46:46 <etotheipi_> I mean, it's a cool concept...
293 2012-02-01 04:46:46 <luke-jr> etotheipi_: more useful than teaching my children?
294 2012-02-01 04:47:06 <etotheipi_> okay, more useful things you could be teaching your children
295 2012-02-01 04:47:09 <luke-jr> that's my primary duty in life, everything else revolves around it
296 2012-02-01 04:47:20 <luke-jr> well, numbers isn't their only subject ofc
297 2012-02-01 04:47:38 <luke-jr> I'm behind on non-English languages, partly because I suck at them
298 2012-02-01 04:48:39 <poiuh> L-i
299 2012-02-01 04:50:18 <NxTitle> SuprTiggr: I don't know if you're solely a bot, but if you are can you _please_ stop telling me being logged in as root is a bad thing?
300 2012-02-01 04:50:36 <NxTitle> I'm not running as root, that's merely my fake ident
301 2012-02-01 04:50:55 <luke-jr> LOL
302 2012-02-01 04:51:02 <gmaxwell> SuprTiggr: can you please give me that script so I can run it too?
303 2012-02-01 04:51:28 <luke-jr> gmaxwell: I didn't know you used mIRC!
304 2012-02-01 04:51:31 <gmaxwell> NxTitle: just /ignore it in your settings. Come on.
305 2012-02-01 04:51:45 <luke-jr> :p
306 2012-02-01 04:51:46 <NxTitle> gmaxwell: oh right >.<
307 2012-02-01 04:52:10 <gmaxwell> luke-jr: "BitchX" forever, yo.
308 2012-02-01 04:52:17 <NxTitle> s/X//
309 2012-02-01 04:52:54 <NxTitle> so what's up with all this bitcoin drama?
310 2012-02-01 04:53:21 <gmaxwell> Nah, there is no drama.
311 2012-02-01 04:53:35 <NxTitle> gavin wants to kick out luke-jr or something?
312 2012-02-01 04:57:16 <BlueMatt> he did, but luke is calming down(ish) since gavin said that
313 2012-02-01 04:57:18 <BlueMatt> so maybe not
314 2012-02-01 04:57:44 <NxTitle> ah, I see
315 2012-02-01 04:59:43 <NxTitle> that's good
316 2012-02-01 05:00:36 <NxTitle> while I disagree with lots of viewpoints of his (might as well be honest :P), he's still contributed lots to bitcoin and it would suck to see him go
317 2012-02-01 05:09:39 <MrTiggr> gmaxwell - RootWarner plugin for SupyBot is the script
318 2012-02-01 05:09:49 <MrTiggr> is python so u can prolly reuse it in other bot ;)
319 2012-02-01 05:09:56 <NxTitle> lol
320 2012-02-01 05:14:16 <MrTiggr> i hope ur using an IRC client that is 100% secure with no vulnerabilities in it NxTitle ;) ... otherwise, SuprTiggr's warnings may be warranted ;)  [BTW - i am SuprTiggr's botherder]
321 2012-02-01 05:14:48 <NxTitle> MrTiggr: fake ident, I'm not running as root
322 2012-02-01 05:14:54 <NxTitle> it's the user setting in irssi
323 2012-02-01 05:15:19 <MrTiggr> lol nice
324 2012-02-01 05:15:45 <MrTiggr> i'll see if i cant tell supr to let you be ..... i'm not turning RootWarner off tho
325 2012-02-01 05:16:42 <NxTitle> but don't worry - I run ssh without a password, use windows XP SP0 without a firewall, enter my CC into sketchy websites, and donate to nigerians
326 2012-02-01 05:16:46 <NxTitle> so I'm secure
327 2012-02-01 05:16:57 <NxTitle> sshd, I might add
328 2012-02-01 05:18:27 <etotheipi_> Joric, you should look for splatster and figure out Armory compiling
329 2012-02-01 05:18:36 <etotheipi_> we made a breakthrough
330 2012-02-01 05:18:41 <etotheipi_> but I couldn't help anymore
331 2012-02-01 05:18:44 <splatster> yes we did
332 2012-02-01 05:19:16 <splatster> /me breaks out pastebin for some disgusting logs
333 2012-02-01 05:19:43 <NxTitle> I compiled armory on windows this morning
334 2012-02-01 05:19:52 <splatster> this is for OS X
335 2012-02-01 05:19:53 <NxTitle> took me a hell of a lot of time
336 2012-02-01 05:19:57 <NxTitle> oh
337 2012-02-01 05:20:06 <etotheipi_> NxTitle, yeah there is nothing pleasant about the build instructions for it
338 2012-02-01 05:20:22 <etotheipi_> unfortunately I can't compile full binaries until I figure out a nasty doesn't-really-shutdown bug
339 2012-02-01 05:20:40 <NxTitle> the build instructions are actually very well written
340 2012-02-01 05:20:42 <splatster> Full log of latest make swig
341 2012-02-01 05:20:44 <splatster> http://pastebin.com/fJHAJFYP
342 2012-02-01 05:20:48 <NxTitle> it's just that I'm often an idiot :P
343 2012-02-01 05:21:12 <etotheipi_> however, I can go halfway, and provide all the python code with a precompiled cppblockutils module -- the user will only need to download and install python, twisted, and pyqt4 (which should be fairly straightforward)
344 2012-02-01 05:21:28 <Joric> i made a breakthrought too
345 2012-02-01 05:21:32 <etotheipi_> in fact, I was just getting ready to do that...
346 2012-02-01 05:21:34 <splatster> Joric: So as I said last night, etotheipi_ sent me his premade .cxx and .py files that were causing trouble
347 2012-02-01 05:21:42 <NxTitle> indeed - I got it packaged up into py2exe and everything too
348 2012-02-01 05:21:45 <Joric> totally fucked up all my multiboot
349 2012-02-01 05:22:08 <Joric> but now i'm on 10.6.7 + xcode 4.2
350 2012-02-01 05:22:13 <splatster> So some tweaking of the make has left us with the log I put above
351 2012-02-01 05:22:25 <etotheipi_> NxTitle, be careful with the .exe... sometimes it leaves the python environment (with all the RAM) open in the background
352 2012-02-01 05:22:38 <splatster> Joric: Don't destroy your system, I'll keep doing this on mine
353 2012-02-01 05:22:49 <splatster> Any ideas what to do though?
354 2012-02-01 05:23:00 <etotheipi_> I cannot, for my life, figure out how to get qt4reactor to shutdown completely, reliably, on windows, yet
355 2012-02-01 05:23:12 <NxTitle> etotheipi_: will look out for that, however I have 16GB of ram on my main box :P
356 2012-02-01 05:23:34 <Joric> splatster, pastebin is empty
357 2012-02-01 05:23:46 <NxTitle> also, have you looked at the bug that one guy mentioned on the forum where it cannot connect to the local bitcoind when trying to send a transaction?
358 2012-02-01 05:23:51 <etotheipi_> this is my current system: http://dl.dropbox.com/u/1139081/memusage.png
359 2012-02-01 05:23:53 <Joric> oh sorry it's okay
360 2012-02-01 05:24:04 <etotheipi_> NxTitle, I don't have a way to look at it
361 2012-02-01 05:24:09 <Joric> splatster, i've seen it two days agy
362 2012-02-01 05:24:12 <Joric> ago
363 2012-02-01 05:24:21 <NxTitle> I used the workaround by grabbing the transaction from the debug output and broadcasting it otherwise
364 2012-02-01 05:24:24 <NxTitle> however yeah
365 2012-02-01 05:24:32 <NxTitle> etotheipi_: what version of python and architecture are you using?
366 2012-02-01 05:24:41 <Joric> i compiled swig and made cxx/i myself on the mac but got the same linking errors
367 2012-02-01 05:24:43 <etotheipi_> oh yeah, I meant to add a popup about that, as an alternative
368 2012-02-01 05:24:45 <NxTitle> I was doing python 2.6 and 32-bit
369 2012-02-01 05:24:48 <MrTiggr> NxTitle: i have turned off rootwarner for this channel only - i figure us dev's know what we're doing :)
370 2012-02-01 05:24:50 <etotheipi_> nibor created a tool
371 2012-02-01 05:25:07 <etotheipi_> err.. webpage for broadcasting it
372 2012-02-01 05:25:14 <splatster> Joric: I'm lost
373 2012-02-01 05:25:23 <etotheipi_> but so far, I'm either connected, or I'm in offline mode, I can't get in between like you guys :(
374 2012-02-01 05:25:28 <NxTitle> MrTiggr: haha, alright - though everyone knows the safest way to run a box is as root :P
375 2012-02-01 05:25:46 <MrTiggr> LOL
376 2012-02-01 05:25:57 <Joric> splatster, so, did you manage to run it on on mac or not?
377 2012-02-01 05:25:59 <MrTiggr> windows users sure know that ;)
378 2012-02-01 05:26:02 <splatster> no
379 2012-02-01 05:26:10 <NxTitle> Disable UAC
380 2012-02-01 05:26:15 <NxTitle> profit
381 2012-02-01 05:26:22 <splatster> I can't get passed those errors
382 2012-02-01 05:26:24 <MrTiggr> hehehehe UAC==Graphical sudo ...done wrong
383 2012-02-01 05:26:32 <Joric> wow 15.7 GiB RAM
384 2012-02-01 05:26:37 <NxTitle> lol :P
385 2012-02-01 05:26:42 <Joric> 4 cpus
386 2012-02-01 05:27:04 <Joric> finally the system that suits etotheipi_ client just fine
387 2012-02-01 05:27:08 <Joric> no offence
388 2012-02-01 05:27:12 <NxTitle> MrTiggr: also the fact that you can just wrap an existing installer with a trojan - it asks for auth to install a legit program and bam, you are owned
389 2012-02-01 05:27:48 <MrTiggr> worser still ... we all know the stickykeys hack to pwn a doze box no !?
390 2012-02-01 05:28:58 <NxTitle> xD
391 2012-02-01 05:29:16 <MrTiggr> 1.) grab a copy of cmd  2.) boot to linux live cd of your chosing 3) mount the windoze drive 4.) replace stickeykeys exe with renamed cmd exe 5.) reboot to windows, hit shift 5 times and enjoy your new root/admin commandline :D
392 2012-02-01 05:29:19 <etotheipi_> no worries, Joric .  I know that it's a PITA for resources... but I wanted to get it out there so people with the RAM *can* use it while I bring down the ram
393 2012-02-01 05:29:52 <etotheipi_> the blockchain doubled in size since I started the project :(
394 2012-02-01 05:29:59 <NxTitle> etotheipi_: another good way to use lots of memory is to pipe yes into yes - why I'm mentioning this, I don't know
395 2012-02-01 05:30:08 <etotheipi_> haha
396 2012-02-01 05:31:02 <etotheipi_> btw, Joric... Armory is not CPU intensive... just RAM
397 2012-02-01 05:31:09 <Joric> i made another chart using my extrapolation skills ) http://goo.gl/80CSr
398 2012-02-01 05:31:23 <NxTitle> well, it maxes out one core for a while, however on multicore that doesn't matter much
399 2012-02-01 05:32:10 <etotheipi_> so NxTitle, do you have the same tx-broadcast issue?  perpetually?
400 2012-02-01 05:33:01 <NxTitle> so far it's been constant
401 2012-02-01 05:33:11 <NxTitle> however I've only tested it on one single machine
402 2012-02-01 05:33:12 <etotheipi_> argh... because I haven't had this problem once
403 2012-02-01 05:33:24 <NxTitle> using python 2.6 x86, with and without py2exe
404 2012-02-01 05:33:26 <etotheipi_> between all my VMs and host machine
405 2012-02-01 05:33:42 <NxTitle> it's possible you're using a different python/arch combo
406 2012-02-01 05:33:46 <NxTitle> and somehow that has to do with it
407 2012-02-01 05:33:50 <etotheipi_> well I'm using 2.6 x64
408 2012-02-01 05:34:02 <NxTitle> there are many possible factors, even down to twisted or pyqt4 versions
409 2012-02-01 05:34:41 <NxTitle> it's also possible this is a windows problem
410 2012-02-01 05:34:50 <NxTitle> though everything should be fairly cross platform
411 2012-02-01 05:34:59 <etotheipi_> I've been testing it in Windows (and haven't heard any other reports)
412 2012-02-01 05:34:59 <splatster> etotheipi_: here's something interesting:
413 2012-02-01 05:35:01 <splatster> $ sudo python ArmoryQt.py Traceback (most recent call last):
414 2012-02-01 05:35:32 <etotheipi_> splatster, that's not "interesting," that's "whoops, I don't have PyQt4 installed"
415 2012-02-01 05:35:36 <NxTitle> sokyou need to install pyqt4 then :P
416 2012-02-01 05:35:42 <etotheipi_> (installed properly, at least)
417 2012-02-01 05:35:44 <splatster> so I am reinstalling py27-pyqt4
418 2012-02-01 05:35:45 <NxTitle> oops, fail tab
419 2012-02-01 05:35:53 <NxTitle> but yeah, what ethsaid
420 2012-02-01 05:36:02 <splatster> But it got farther than it normally would have
421 2012-02-01 05:36:07 <NxTitle> also, > sudo
422 2012-02-01 05:36:08 <NxTitle> why?
423 2012-02-01 05:36:48 <splatster> sudo because permissions problems are the #1 problem I have run into
424 2012-02-01 05:36:55 <etotheipi_> alright, just merged into Master for the first time since early december
425 2012-02-01 05:37:09 <splatster> not really but just in general it doesn't hurt
426 2012-02-01 05:37:25 <etotheipi_> I think it's ready for RC1, just to make sure I didn't botch something stupid in my last-minute polishing
427 2012-02-01 05:37:28 <NxTitle> odd - nothing in armory should require root permissions
428 2012-02-01 05:37:35 <splatster> Oh so now I am trying to compile old shit, yes?
429 2012-02-01 05:37:40 <NxTitle> at least in the code I've read
430 2012-02-01 05:37:46 <etotheipi_> splatster, don't worry about it
431 2012-02-01 05:37:59 <etotheipi_> I've polished the GUI, but the version you had still works
432 2012-02-01 05:38:13 <etotheipi_> and everything I've committed since your last checkout is all python-code updates:  no recompiling necessary
433 2012-02-01 05:38:20 <NxTitle> anyways it's nearly midnight, so night all
434 2012-02-01 05:38:44 <NxTitle> etotheipi_: I'll come back on tomorrow or something and help diagnose the issue deeper
435 2012-02-01 05:39:00 <etotheipi_> that would be appreciated NxTitle
436 2012-02-01 05:39:01 <NxTitle> so far all I know is self.proto is not defined and that is triggering the DNE message
437 2012-02-01 05:39:08 <NxTitle> well, not set
438 2012-02-01 05:39:23 <etotheipi_> self.proto gets set when the connection is established
439 2012-02-01 05:39:33 <NxTitle> ah
440 2012-02-01 05:39:39 <etotheipi_> I'll check to see if it resets to "None" on disconnect
441 2012-02-01 05:39:54 <NxTitle> and oddly enough, it says in the debug that it successfully connects to satoshi client
442 2012-02-01 05:39:55 <etotheipi_> do you know if it says "Offline" in the bottom right corner?
443 2012-02-01 05:40:20 <NxTitle> I don't remember if it did, however I don't recall seeing it
444 2012-02-01 05:40:27 <poiuh> mangus
445 2012-02-01 05:40:30 <NxTitle> seems like something that would have stood out
446 2012-02-01 05:41:12 <etotheipi_> okay, talk more tomorrow
447 2012-02-01 05:41:13 <etotheipi_> thanks
448 2012-02-01 05:41:56 <NxTitle> etotheipi_: do you think it's a good idea to try re-establishing the connection ion send if it appears to be disconnected?
449 2012-02-01 05:42:01 <NxTitle> ion = on
450 2012-02-01 05:42:31 <etotheipi_> NxTitle, that *is* what I'm doing... are you suggesting I shouldn't?
451 2012-02-01 05:42:57 <etotheipi_> usually when I leave Armory running for a long time, I see the "Handshake Complete.  Connection Open!" message a few times
452 2012-02-01 05:43:01 <NxTitle> oh o.O I'm definitely not suggesting you shouldn't
453 2012-02-01 05:43:03 <etotheipi_> maybe once an hour
454 2012-02-01 05:43:23 <etotheipi_> this is why I'm confused
455 2012-02-01 05:43:40 <etotheipi_> if it couldn't connect the first time, it goes into offlien mode... if it does connect, then it will keep trying to reconnect
456 2012-02-01 05:43:43 <NxTitle> however I mean when you hit send, the slot receives the action, and if self.proto is None then attempt to re-establish
457 2012-02-01 05:43:55 <etotheipi_> and if it connected the first time, then it should be able to reconnect
458 2012-02-01 05:44:05 <splatster> Joric: Do you think a combo update would fix anything I might have broken in this whole proccess?
459 2012-02-01 05:44:24 <NxTitle> yeah, I just have to figure out where it's being set to None if anyywhere
460 2012-02-01 05:44:35 <etotheipi_> it doesn't get set in many places...
461 2012-02-01 05:45:23 <Joric> splatster, combo update where? i comboupdated to 10.6.7 it doesn't fix anything it only adds more mess )
462 2012-02-01 05:45:44 <etotheipi_> I don't see how it could be none, unless it never connected to begin with
463 2012-02-01 05:45:44 <NxTitle> yeah, I'll have to take a deeper look tomorrow
464 2012-02-01 05:45:48 <etotheipi_> okay, sounds good
465 2012-02-01 05:45:49 <NxTitle> for now though, goodnight
466 2012-02-01 05:48:52 <splatster> I think a 10.7.2 combo will solve any problems that become of all this shit
467 2012-02-01 06:54:47 <splatster> etotheipi_: You still up?
468 2012-02-01 06:55:27 <splatster> Loading Armory Engine:
469 2012-02-01 06:55:28 <splatster> Armory Version:       0.50
470 2012-02-01 06:55:42 <splatster> File "ArmoryQt.py", line 39, in <module>
471 2012-02-01 06:55:43 <splatster> from armoryengine import *
472 2012-02-01 06:55:53 <splatster> File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/posixpath.py", line 68, in join
473 2012-02-01 06:55:54 <splatster> AttributeError: 'NoneType' object has no attribute 'endswith'
474 2012-02-01 06:55:55 <splatster> Any ideas?
475 2012-02-01 09:52:32 <pentarh> namecoin fundamental design error https://bitcointalk.org/index.php?topic=62017.msg725651#msg725651
476 2012-02-01 10:01:51 <diki> the forum suddenly isn't opening for me...
477 2012-02-01 10:07:46 <poiuh> ok
478 2012-02-01 10:45:27 <UukGoblin> so considering coinbase transactions are not prunable (or at least their hashes aren't), is it actually any different whether you store your aux data in coinbase or as a tx output?
479 2012-02-01 10:46:15 <sipa> how do you mean, not prunable?
480 2012-02-01 10:46:58 <UukGoblin> sipa, a verifying node has to remember all coinbase transactions (or at least their hashes) to prohibit double spending in the "identical coinbase" scenario
481 2012-02-01 10:47:20 <sipa> it doesn't have to prohibit that
482 2012-02-01 10:47:40 <UukGoblin> I believe it's a hard rule
483 2012-02-01 10:47:45 <sipa> it isn't
484 2012-02-01 10:48:19 <sipa> if you create a duplicate coinbase tx, you are stupid
485 2012-02-01 10:48:54 <UukGoblin> stupid or not, other nodes will have to prohibit the second spending of a duplicate txid
486 2012-02-01 10:49:15 <sipa> i don't see how that is different from any other tx?
487 2012-02-01 10:49:15 <UukGoblin> at least in my current understanding
488 2012-02-01 10:49:46 <sipa> for each tx, you remember which txouts are unredeemed
489 2012-02-01 10:49:54 <poiuh> zouba
490 2012-02-01 10:49:59 <sipa> if you encounter a spending of a redeemed one, call it invalid
491 2012-02-01 10:50:08 <sipa> otherwise, mark its inputs redeemed
492 2012-02-01 10:50:40 <UukGoblin> sipa, ok, it's different in this scenario: mine a block X with a coinbase txid 01abcd, spend txid 01abcd to someone, then mine another block Y with the same coinbase 01abcd
493 2012-02-01 10:50:54 <sipa> then you can spend it again?
494 2012-02-01 10:50:59 <UukGoblin> no, you can't
495 2012-02-01 10:51:04 <sipa> why not?
496 2012-02-01 10:51:05 <UukGoblin> it's the same txid
497 2012-02-01 10:51:15 <sipa> so, it's redeemable again
498 2012-02-01 10:51:15 <UukGoblin> nodes will prohibit spending the same txid twice
499 2012-02-01 10:51:27 <UukGoblin> nope it's not
500 2012-02-01 10:51:37 <UukGoblin> you can't redeem the same txid twice
501 2012-02-01 10:52:05 <UukGoblin> we've talked about it earlier with roconnor and gmaxwell
502 2012-02-01 10:55:10 <sipa> looks to me that a second tx with the same hash will overwrite the previous one
503 2012-02-01 11:07:15 <UukGoblin> vuln in sudo: http://i.imgur.com/AjceU.png
504 2012-02-01 11:07:39 <upb> ahahahaha, gives that pic a whole new meaning :D
505 2012-02-01 11:08:02 <UukGoblin> sipa, well, I'm not sure, but from what I heard, people who genereated duplicate coinbases weren't able to spend them
506 2012-02-01 11:08:23 <sipa> UukGoblin: hmm, could be
507 2012-02-01 11:12:15 <UukGoblin> so wouldn't that make the coinbases special and actually a worse place to put extra data in?
508 2012-02-01 11:12:50 <sipa> well, if it is true, coinbases aren't special
509 2012-02-01 11:13:02 <Joric> UukGoblin: mother of god!
510 2012-02-01 11:13:07 <sipa> but it is extremely unlikely to generate a duplicate non-coinbase tx
511 2012-02-01 11:17:18 <UukGoblin> sipa, I think short of a hash collision, it should be impossible to generate a duplicate non-coinbase tx
512 2012-02-01 11:17:26 <sipa> indeed
513 2012-02-01 11:58:41 <Moron__> whats a coinbase?
514 2012-02-01 12:00:10 <sipa> Moron__: the first transaction in each block is the one that introduces new coins in the system (the miner's reward)
515 2012-02-01 12:00:16 <sipa> it's called the coinbase
516 2012-02-01 12:00:39 <sipa> it is special as it does not spend anything
517 2012-02-01 12:01:10 <Moron__> oh
518 2012-02-01 12:03:53 <lianj> all coinbase are belongs to the miner
519 2012-02-01 12:04:11 <Moron__> does the genesis block have a coinbase?
520 2012-02-01 12:04:16 <midnightmagic> often, people will also refer to the actual "coinbase" field itself in the generation transaction as a "coinbase" too. You can see an example of a generation txn which shows a coinbase, here: http://blockexplorer.com/rawtx/fca294ed69a8cb2be96228e1dcb1d132bc2c941198178d21a8f95f7afb05ad15
521 2012-02-01 12:04:17 <lianj> sure
522 2012-02-01 12:04:30 <sipa> Moron__: it only has a coinbase
523 2012-02-01 12:05:01 <sipa> here is it: http://blockchain.info/tx-index/1/4a5e1e4baab89f3a32518a88c31bc87f618f76673e2cc77ab2127b7afdeda33b
524 2012-02-01 12:05:20 <midnightmagic> http://blockexplorer.com/rawblock/000000000019d6689c085ae165831e934ff763ae46a2a6c172b3f1b60a8ce26f also
525 2012-02-01 12:05:21 <sipa> notice the text in it
526 2012-02-01 12:06:02 <Moron__> thats the genesis block ?
527 2012-02-01 12:06:19 <Moron__> what does n mean?
528 2012-02-01 12:07:09 <sipa> my link shows the coinbase transaction of the genesis block
529 2012-02-01 12:07:20 <Moron__> oh ok
530 2012-02-01 12:07:28 <sipa> not sure what n you talk about
531 2012-02-01 12:07:41 <Moron__> "n":4294967295
532 2012-02-01 12:07:49 <Moron__> sory im not too familiar with the blockchain format
533 2012-02-01 12:08:32 <sipa> well, the actual "coinbase" field (which contains some arbitrary miner-specified data) is actually in the same place as the input script
534 2012-02-01 12:08:54 <sipa> so coinbase transactions have actually one "fake" input, which counts as coinbase (it's not interpreted as input)
535 2012-02-01 12:09:23 <sipa> by definition, it must consume "from" txout 000000...:4294967295
536 2012-02-01 12:10:22 <Moron__> oh, so the n is usually for input?
537 2012-02-01 12:11:00 <Moron__> strange
538 2012-02-01 12:11:51 <Moron__> first time it opened in some wierd bracket format, now its in a html format with graphics
539 2012-02-01 12:12:17 <Moron__> oh well i think i get theidea
540 2012-02-01 12:18:41 <poiuh> coinface mc hargalarg
541 2012-02-01 12:25:20 <Moron__> how hard are bitbills/casascius coins to counterfeit or steal the private key from?
542 2012-02-01 12:26:14 <marf_away> try it
543 2012-02-01 12:26:27 <marf_away> biggest problem are the holographs
544 2012-02-01 12:27:33 <UukGoblin> if you can print holograms, I guess it's easy
545 2012-02-01 12:27:43 <UukGoblin> I'm not sure how hard it is though
546 2012-02-01 12:31:06 <Moron__> so in theory you could print fakes with just a public key and no valid private key?
547 2012-02-01 12:31:16 <marf_away> yes
548 2012-02-01 12:31:31 <marf_away> you can print dollars too
549 2012-02-01 12:31:32 <marf_away> ;)
550 2012-02-01 12:31:42 <Moron__> yeh but im scared of the secret service
551 2012-02-01 12:32:00 <marf_away> maybe mr bitbill sends a pitbull after you
552 2012-02-01 12:32:02 <marf_away> ;D
553 2012-02-01 12:32:07 <Moron__> :P
554 2012-02-01 12:33:44 <Moron__> i figure I could make a killing selling them to bitcoin hopefuls at carboot sales aand the like?
555 2012-02-01 12:34:27 <marf_away> sure
556 2012-02-01 12:35:07 <marf_away> they are just toys
557 2012-02-01 12:35:19 <Moron__> :)
558 2012-02-01 12:35:23 <marf_away> not intended for real use
559 2012-02-01 12:35:38 <edcba> you can trade mtg cards too
560 2012-02-01 12:36:02 <edcba> they are toy, doesn't mean they are worthless
561 2012-02-01 12:36:24 <sipa> a bitbull?
562 2012-02-01 12:37:14 <marf_away> a bit dog
563 2012-02-01 12:37:17 <marf_away> big
564 2012-02-01 12:37:48 <poiuh> "obama caught operating dog fighting ring in white house basement?"
565 2012-02-01 12:39:49 <edcba> it is not a ring it's a chain !!!
566 2012-02-01 12:57:09 <poiuh> heyro
567 2012-02-01 13:17:53 <Moron__> heyro
568 2012-02-01 13:17:55 <Moron__> to you too
569 2012-02-01 13:48:38 <gmaxwell> "The libcoin/bitcoind client downloads the entire block chain 3.5 times faster than the bitcoin/bitcoind client. This is less than 90 minutes on a modern laptop!"
570 2012-02-01 13:48:46 <gmaxwell> It would be nice to know where it got that speedup from.
571 2012-02-01 13:49:22 <gmaxwell> I see that it has code to disable fsync from libdb... but it's commented out.
572 2012-02-01 13:49:27 <sipa> i looked at the source briefly-- it is still very recognizable
573 2012-02-01 13:49:53 <gmaxwell> Likewise (thus the comment on the code to disable fsync).
574 2012-02-01 13:53:26 <UukGoblin> what's libcoin/bitcoind? google's being unhelpful again
575 2012-02-01 13:53:39 <luke-jr> UukGoblin: new library/daemon that was just released
576 2012-02-01 13:53:44 <luke-jr> UukGoblin: see last post to dev ML
577 2012-02-01 13:55:46 <UukGoblin> announcement looks cool...
578 2012-02-01 13:55:47 <gmaxwell> It's a refactor of the reference bitcoin code.
579 2012-02-01 13:56:12 <UukGoblin> "See the ponzicoin example for how you define your own chain" heh, fun
580 2012-02-01 13:57:59 <gavinandresen> Nice!  Glad somebody did it...
581 2012-02-01 13:58:26 <UukGoblin> I was secretly thinking about it and someone beat me to it... happened AGAIN ;-]
582 2012-02-01 13:58:39 <luke-jr> they should probably merge up to 0.5 at least
583 2012-02-01 13:58:44 <luke-jr> looks like based on 0.4
584 2012-02-01 13:58:54 <luke-jr> but I guess that explains their first version number
585 2012-02-01 14:03:04 <phantomcircuit> bleh how to diff :|
586 2012-02-01 14:06:18 <Joric> finally cmake
587 2012-02-01 14:06:34 <Joric> https://github.com/ceptacle/libcoin
588 2012-02-01 14:09:18 <Joric> went building armory on mac
589 2012-02-01 14:35:50 <Turingi> certificate authorities, web of trust, all that is obsolete!
590 2012-02-01 14:36:12 <sipa> how so?
591 2012-02-01 14:36:13 <Turingi> with the bitcoin block chain (of arbitrary age), you can get that
592 2012-02-01 14:36:28 <sipa> bitcoin does not give you trust between arbitrary parties
593 2012-02-01 14:36:43 <sipa> and certainly no proof of identity
594 2012-02-01 14:36:45 <Turingi> it's an append-only database, ultra-redundant
595 2012-02-01 14:37:03 <sipa> it is not append-only
596 2012-02-01 14:37:14 <Turingi> it is for a chain that's old enough
597 2012-02-01 14:38:21 <gmaxwell> Turingi: bitcoin is indeed very exciting, but it's easy to get carried away.
598 2012-02-01 14:41:41 <nathan7> exciting things
599 2012-02-01 14:41:44 <nathan7> what's up, bitcoinverse
600 2012-02-01 14:47:29 <luke-jr> Turingi: perhaps you mean namecoin?
601 2012-02-01 14:47:56 <Turingi> luke-jr: I don't understand why namecoin has to be a separate system
602 2012-02-01 14:48:11 <Turingi> it can all be strapped to the canonical chain
603 2012-02-01 14:48:18 <sipa> Turingi: because it does not belong in the bitcoin block chain
604 2012-02-01 14:48:22 <luke-jr> Turingi: because people don't want crap bloating Bitcoin? :p
605 2012-02-01 14:48:35 <sipa> which is, as you said, ultra-redundant, and costs a ton
606 2012-02-01 14:49:19 <Turingi> sipa, luke-jr: that may be, but the presence of this killer feature will render it being used like that
607 2012-02-01 14:49:35 <sipa> i don't see why it is a killer feature
608 2012-02-01 14:49:39 <helo> (bitcoin's ability to scale as blockchain size grows is arguably its biggest weakness)
609 2012-02-01 14:49:52 <sipa> there are much more convenient storage systems available
610 2012-02-01 14:50:01 <Turingi> besides, bitcoin has critical mass, namecoin doesn't
611 2012-02-01 14:50:28 <Turingi> the worst that happens is some people just destroy their coins in exchange for creating preformatted transactions
612 2012-02-01 14:50:35 <Turingi> or preformatted keys
613 2012-02-01 14:50:43 <Turingi> which increases the value of the remaining BTC
614 2012-02-01 14:51:15 <sipa> at the cost of *everyone* running a full node to keep a copy of it, for eternity
615 2012-02-01 14:51:25 <Turingi> that will happen anyway
616 2012-02-01 14:51:54 <JRWR> What is the expected file size in 20 years?
617 2012-02-01 14:52:28 <helo> 1ZB
618 2012-02-01 14:52:37 <Turingi> all I'm describing would happen with or without special use
619 2012-02-01 14:52:46 <Turingi> the block chain will grow
620 2012-02-01 14:52:49 <UukGoblin> JRWR, max possible is 1TB
621 2012-02-01 14:52:55 <Turingi> at some point it becomes impractical for everyone to keep a copy of it
622 2012-02-01 14:52:59 <UukGoblin> JRWR, (if you mean the blockchain size)
623 2012-02-01 14:53:03 <JRWR> yes
624 2012-02-01 14:53:19 <UukGoblin> that's assuming every single block from now on was full
625 2012-02-01 14:53:22 <JRWR> UukGoblin: does any type of compression help? or is the blockchain just too random for that
626 2012-02-01 14:53:22 <sipa> Turingi: then why would you want to encourage that?
627 2012-02-01 14:53:33 <sipa> the block chain is quite compressible
628 2012-02-01 14:53:38 <Turingi> sipa: if bitcoin is not resilient against that, it will fail
629 2012-02-01 14:53:56 <sipa> bitcoin is what its users want it to
630 2012-02-01 14:53:58 <Turingi> sipa: and it's better if fails faster, else more capital will be poured into it
631 2012-02-01 14:54:07 <Turingi> but I don't think it would fail
632 2012-02-01 14:54:19 <sipa> if some users want it to be a storage system, and others want it to be a currency, there is a problem
633 2012-02-01 14:54:49 <Turingi> sipa: it can be both and perhaps other things too
634 2012-02-01 14:55:27 <sipa> it may be other things, but i strongly disagree it can be both
635 2012-02-01 14:56:50 <Turingi> sipa: do you at least agree that if bitcoin can't survive prepared transactions that it's too vulnerable as a currency?
636 2012-02-01 14:57:03 <sipa> what are prepared transactions?
637 2012-02-01 14:57:46 <Turingi> transactions and keys shaped to have a special meaning
638 2012-02-01 14:57:51 <JRWR> ah
639 2012-02-01 14:58:05 <JRWR> like a storage in the blockchain
640 2012-02-01 14:58:05 <sipa> have you read BIP16? :)
641 2012-02-01 14:58:09 <Turingi> BTC is spent toward a bogus key or transaction
642 2012-02-01 14:58:27 <Turingi> you can't recover the BTC, but you get to use the block chain as storage
643 2012-02-01 14:59:30 <Turingi> hm... although you don't strictly need prepared transactions
644 2012-02-01 14:59:47 <sipa> how about using dropbox? :p
645 2012-02-01 14:59:51 <JRWR> couldn't you just transfer the amount between two IDs that exist and use that as datastorage
646 2012-02-01 14:59:53 <Turingi> you can pass back and forth a certain ammount of bitcents
647 2012-02-01 14:59:56 <helo> there's a maximum blocksize, so as that is approached miners will start only mining the transactions with the best fee/size ratio, increasing the cost for people to store data in the blockchain
648 2012-02-01 14:59:58 <Turingi> and that can convey information
649 2012-02-01 15:00:18 <Turingi> so you still shape the block chain
650 2012-02-01 15:00:29 <Turingi> but not spend any BTC (until transaction costs start to matter)
651 2012-02-01 15:00:54 <Turingi> does that make sense?
652 2012-02-01 15:02:11 <sipa> i still don't see why you would want to do so, except for proving you can
653 2012-02-01 15:02:44 <Turingi> sipa: the point is not storage size, but non-repudiation
654 2012-02-01 15:02:56 <sipa> then put a hash in a coinbase
655 2012-02-01 15:02:58 <JRWR> I'm surpised that someone hasnt flooded the blockchain yet with a ton of transactions
656 2012-02-01 15:02:59 <gmaxwell> Turingi: the system already resists bulk data storage fairly successfully through anti-dos rule that require you to apply transaction fees when your activity starts to look spammy.
657 2012-02-01 15:03:02 <sipa> via a merkle tree
658 2012-02-01 15:03:02 <Turingi> sipa: and having a reference that can't be corrupted
659 2012-02-01 15:03:16 <gmaxwell> JRWR: People did that, and the bitcoin community solved it.
660 2012-02-01 15:03:19 <sipa> Turingi: that adds no data to the chain, and has the same security properties
661 2012-02-01 15:04:35 <helo> it's pretty expensive to solve a block... i wonder if miners will ever start selling their coinbase space for applications like that
662 2012-02-01 15:04:56 <gmaxwell> sipa: even better, in fact because its more private. No one can tell the data exists until you disclose the connecting hash.
663 2012-02-01 15:05:03 <sipa> gmaxwell: indeed!
664 2012-02-01 15:05:45 <gmaxwell> helo: they do, in a fashion. It's how merged mining works. (but the ability of tree commitments are more general than that by far)
665 2012-02-01 15:06:17 <gmaxwell> helo: though because the amount it can store is ~infinite the proper price for that is pretty low.
666 2012-02-01 15:07:00 <gmaxwell> sipa: this is why a notary service is on my features list to discourage wrongheaded abuses.
667 2012-02-01 15:07:01 <helo> ~infinite because a hash can represent an arbitrarily large amount of data?
668 2012-02-01 15:07:12 <Moron__> whats a merkle tree?
669 2012-02-01 15:07:18 <gmaxwell> helo: Yes, a hash tree.
670 2012-02-01 15:08:34 <vsrinivas> helo: selling coinbase space would be kinda neat, yea...
671 2012-02-01 15:09:49 <gmaxwell> Moron__: See wikipedia
672 2012-02-01 15:09:58 <k9quaint> Moron__: its a google result
673 2012-02-01 15:10:21 <Joric> wow macosx client is the prettiest
674 2012-02-01 15:11:26 <gmaxwell> I wrote a halfass spec about how you'd do notary trees connected to coinbase the other day: http://pastebin.com/1fVRZEBk
675 2012-02-01 15:14:56 <Turingi> helo: just imagine, one hash can represent the entire universe!
676 2012-02-01 15:14:59 <Turingi> :)
677 2012-02-01 15:15:08 <Moron__> how Turingi?
678 2012-02-01 15:15:41 <Turingi> helo>~infinite because a hash can represent an arbitrarily large amount of data?
679 2012-02-01 15:15:54 <Moron__> but is the universe data?
680 2012-02-01 15:15:58 <sipa> countably infinite amount of data
681 2012-02-01 15:16:07 <Turingi> it depends
682 2012-02-01 15:16:12 <sipa> the universe may well not be representable in a countably infinite dataset
683 2012-02-01 15:16:14 <Turingi> the visible universe is finite
684 2012-02-01 15:16:25 <Turingi> it may be infinite, but unreachable
685 2012-02-01 15:16:34 <Turingi> beyong some limit
686 2012-02-01 15:16:37 <Moron__> also the universe is expanding
687 2012-02-01 15:16:41 <sipa> what precision does the location of an electron have?
688 2012-02-01 15:16:43 <Turingi> s/beyong/beyond/
689 2012-02-01 15:16:46 <Moron__> so it may not be possible to collect the information required
690 2012-02-01 15:16:54 <sipa> finite? probabilistic? infinite?
691 2012-02-01 15:16:59 <Moron__> what about black holes et
692 2012-02-01 15:17:00 <Moron__> etc
693 2012-02-01 15:17:06 <Turingi> although, things like quantum uncertainty may make hashing problematic
694 2012-02-01 15:17:40 <JRWR> I always wondered if a collsion could happen on the blockchain, and how would you even fix it, or detect it
695 2012-02-01 15:17:46 <Turingi> so you could at most take a hash of all possible quantum states in their totality, but not the exact configuration the universe is in :)
696 2012-02-01 15:17:48 <Moron__> a collision?
697 2012-02-01 15:17:59 <luke-jr> Turingi: why not use namecoin?
698 2012-02-01 15:18:10 <Moron__> im hoping someone will hash the internet :P
699 2012-02-01 15:18:16 <JRWR> collision
700 2012-02-01 15:18:23 <sipa> Moron__: done; the hash is 42
701 2012-02-01 15:18:27 <Moron__> :P
702 2012-02-01 15:18:36 <Moron__> wow is that sha256?
703 2012-02-01 15:19:08 <Turingi> sha001
704 2012-02-01 15:19:09 <sipa> yes, modulo some not further explored small integer
705 2012-02-01 15:19:31 <Moron__> :)
706 2012-02-01 15:21:54 <JRWR> A good read on the issue: http://stackoverflow.com/questions/4014090/is-it-safe-to-ignore-the-possibility-of-sha-collisions-in-practice
707 2012-02-01 15:22:20 <JRWR> It "could" happen, just very very x 10^100000 hard to do
708 2012-02-01 15:27:43 <Moron__> the laws of statistics break down where 2^x and x is a very large number
709 2012-02-01 15:27:51 <Moron__> or a very small number
710 2012-02-01 15:28:14 <sipa> ehh no
711 2012-02-01 15:28:31 <sipa> your ability to visualize may, though
712 2012-02-01 15:28:58 <Moron__> yeh... etc whats the probability of putting a bomb in a bunch of aircraft debris and getting a functioning plane after an explosion?
713 2012-02-01 15:29:42 <sipa> small enough
714 2012-02-01 15:30:06 <Moron__> lol
715 2012-02-01 15:30:36 <theorbtwo> Small enough that if you've been doing it once per second since the beginning of the universe, it's less then 50% probable to have happened by now.
716 2012-02-01 15:31:18 <Moron__> is it theorbtwo? or is the probability in some way attenuated by newtonian physics?
717 2012-02-01 15:32:22 <theorbtwo> Well, I suppose it depends on how long you are willing to wait after the explosion.
718 2012-02-01 15:32:24 <edcba> but when you mean second is that the second relative to the current time or the second like now ?
719 2012-02-01 15:33:58 <JRWR> but it CAN happen, and what if it does?
720 2012-02-01 15:34:05 <Moron__> whats the probability of putting an orange in the microwave... turning it on, and getting an apple out?
721 2012-02-01 15:38:51 <Moron__> why do older versions than 0.5.2 harm the  network?
722 2012-02-01 15:38:56 <Moron__> is this an attempt to force p2sh on us?
723 2012-02-01 15:39:08 <gmaxwell> ...
724 2012-02-01 15:39:41 <gmaxwell> Moron__: Please don't be a bad person. Making that kind of accusation is harmful.
725 2012-02-01 15:39:50 <gmaxwell> Moron__: it's also stupid 0.5.2 doesn't have any p2sh.
726 2012-02-01 15:39:58 <Moron__> oh sorry
727 2012-02-01 15:40:04 <Moron__> mind if i ask the reason why an upgrade is needed?
728 2012-02-01 15:40:17 <gmaxwell> Moron__: No _released_ version of bitcoin has p2sh of any kind.
729 2012-02-01 15:40:23 <gavinandresen> gmaxwell: the announcement in this IRC channel says old versions harm the network....
730 2012-02-01 15:40:56 <gmaxwell> Moron__: the harms message is related to version prior to 0.3.24
731 2012-02-01 15:41:02 <luke-jr> Moron__: not necessarily *older than 0.5.2*, but *OLD*
732 2012-02-01 15:41:04 <gavinandresen> ... which is true, if you're running a really old bitcoin that had the drop-clients-downloading-the-chain bug
733 2012-02-01 15:41:16 <Moron__> oic
734 2012-02-01 15:41:25 <gavinandresen> Probably the message should be changed, latest versions don't try to download the chain from those really old versions
735 2012-02-01 15:41:31 <gmaxwell> Moron__: older versions (still widely used on the network :( ) would randomly hang up on clients when they tried to pull the chain from them.
736 2012-02-01 15:41:36 <luke-jr> I suggested that a while back
737 2012-02-01 15:42:12 <k9quaint> good vid btw gavin
738 2012-02-01 15:42:50 <gmaxwell> Perhaps instead of harms.. *UPGRADING IS GOOD FOR YOUR SECURITY AND THE HEALTH OF THE NETWORK*
739 2012-02-01 15:43:09 <gmaxwell> (I agree that we can remove or relax it)
740 2012-02-01 15:43:09 <Moron__> yeh i agree, your version is less negative gmaxwell
741 2012-02-01 15:43:17 <k9quaint> don't say harms, say poisons :)
742 2012-02-01 15:43:32 <gmaxwell> Moron__: well the negative absolutely was justified a couple months ago. It matters less now.
743 2012-02-01 15:43:50 <Moron__> what happened a couple months ago?
744 2012-02-01 15:44:05 <k9quaint> Moron__: thanksgiving happened
745 2012-02-01 15:44:29 <luke-jr> it was added around the time of the network fix and then wallet encryption bug
746 2012-02-01 15:44:32 <luke-jr> IIRC
747 2012-02-01 15:45:01 <gmaxwell> Moron__: the fix to the chain pulling bug at the time is was actually becoming hard to bootstrap a node because you would just get disconnected over and over again.
748 2012-02-01 15:45:27 <gmaxwell> since then enough people have deployed fixed nodes that it's not a big issue.
749 2012-02-01 15:45:46 <sipa> etotheipi_: there?
750 2012-02-01 15:45:56 <Moron__> would it be a good idea to detect versions automatically and refuse to send the blockchain to old versions?
751 2012-02-01 15:48:49 <gmaxwell> Moron__: no.
752 2012-02-01 15:49:15 <gmaxwell> Moron__: if we really want to get people's attention the developers can send an alert.
753 2012-02-01 15:49:15 <sipa> that would make the users of those clients vulnerable
754 2012-02-01 15:49:30 <k9quaint> can't send me an alert!
755 2012-02-01 15:49:36 <gmaxwell> except to k9quaint
756 2012-02-01 15:49:40 <Moron__> bitcoin has an alert feature?
757 2012-02-01 15:49:45 <edcba> new unicode 6.1 with bitcoin symbol !
758 2012-02-01 15:50:18 <gavinandresen> k9quaint: ALERT: This is a test of the k9quaint Emergency Alert System
759 2012-02-01 15:50:26 <gmaxwell> Moron__: yes.
760 2012-02-01 15:50:34 <edcba> i'm kidding (i think)
761 2012-02-01 15:50:50 <luke-jr> edcba: Unicode has had the Bitcoin symbol for a while
762 2012-02-01 15:50:56 <Joric> etotheipi_: i managed to build and run bitcoin armory on osx
763 2012-02-01 15:51:04 <luke-jr> B??
764 2012-02-01 15:51:05 <edcba> but there are a lot of animals
765 2012-02-01 15:51:05 <k9quaint> gavinandresen: can I get a different font for my alerts?
766 2012-02-01 15:51:10 <edcba> no there is a bath symbol
767 2012-02-01 15:51:19 <edcba> stop spreading lies ! :)
768 2012-02-01 15:51:37 <luke-jr> edcba: B?? is the common Bitcoin symbol
769 2012-02-01 15:52:00 <Moron__> it doesnt work here
770 2012-02-01 15:52:11 <edcba> ok new money symbols : money bag
771 2012-02-01 15:52:11 <luke-jr> that's another matter :p
772 2012-02-01 15:52:11 <Moron__> its like a light b nd a box
773 2012-02-01 15:52:32 <edcba> glyph may show any currency symbol instead of a dollar sign
774 2012-02-01 15:52:48 <edcba> else there is currency exchange symbol and heavy dollar sign
775 2012-02-01 15:59:59 <edcba> http://www.fileformat.info/info/unicode/char/1f4b0/index.htm
776 2012-02-01 16:01:13 <edcba> but it doesn't seem new
777 2012-02-01 16:10:12 <BlueMatt> ;;seen genjix
778 2012-02-01 16:10:13 <gribble> genjix was last seen in #bitcoin-dev 19 hours, 20 minutes, and 5 seconds ago: <genjix> ok off. cya all.
779 2012-02-01 16:11:54 <gavinandresen> BlueMatt: I'm testing your UPnP patch now.
780 2012-02-01 16:12:34 <gavinandresen> BlueMatt: I'm rewriting the -upnp/-noupnp argument handling a bit, the way it is now the wallet setting always overrides the command-line arg
781 2012-02-01 16:12:47 <BlueMatt> ;;later tell genjix can you merge https://github.com/genjix/bips/pull/2 ?
782 2012-02-01 16:12:47 <gribble> The operation succeeded.
783 2012-02-01 16:13:08 <sipa> gavinandresen: i encountered something similar to the error you got in valgrind
784 2012-02-01 16:13:10 <BlueMatt> gavinandresen: ihmo it should be the other way around
785 2012-02-01 16:13:24 <sipa> the CAddrMan error in osx, i mean
786 2012-02-01 16:13:35 <BlueMatt> gavinandresen: well wallet should be set by cli if its there, then it can reset if you change at runtime
787 2012-02-01 16:13:35 <gavinandresen> BlueMatt: wallet settings should override command-line arguments?
788 2012-02-01 16:13:55 <BlueMatt> oh, sorry missed "the way it is now"
789 2012-02-01 16:14:08 <gavinandresen> no
790 2012-02-01 16:14:11 <gavinandresen> err np
791 2012-02-01 16:15:32 <sipa> what is he telling about?
792 2012-02-01 16:16:10 <BlueMatt> dequeus
793 2012-02-01 16:16:15 <BlueMatt> so, pretty damn boring
794 2012-02-01 16:16:29 <Moron__> so im wondering, how many bitcoins does an average bitcoin dev own?
795 2012-02-01 16:16:58 <lianj> ^^
796 2012-02-01 16:17:03 <luke-jr> I have about 400 BTC right now
797 2012-02-01 16:17:12 <JRWR> I have 0.56
798 2012-02-01 16:17:24 <BlueMatt> heh
799 2012-02-01 16:17:42 <JRWR> mostly from testing CPU/GPU miners
800 2012-02-01 16:19:01 <sipa> Moron__: i have around 200
801 2012-02-01 16:19:30 <BlueMatt> gavinandresen: re: list of "high prio" bugs, #711.  I got the guy's wallet and wasnt able to get anything useful back.  afaict there was some corruption somewhere in some of the privkeys and probably the mkey as well.  Im gonna assume it is a fluke from fs corruption or something, so unless you want to make ie a second file and store a backup copy of the mkey there, I dont know what could be done about it
802 2012-02-01 16:20:21 <gavinandresen> BlueMatt: ok, sounds like a "close because we cannot reproduce"
803 2012-02-01 16:20:28 <BlueMatt> gavinandresen: yep
804 2012-02-01 16:23:36 <gmaxwell> perhaps bitcoin should run backupwallet automatically and keep a couple old wallet files in rotation?
805 2012-02-01 16:23:56 <BlueMatt> (be careful about encryption/passphrase changes though)
806 2012-02-01 16:24:03 <BlueMatt> but, yea
807 2012-02-01 16:24:14 <gmaxwell> (we know that lossy disks are out there we can't make the user have a real backup strategy but we can do that.)
808 2012-02-01 16:24:28 <sipa> not a bad idea, imho
809 2012-02-01 16:24:45 <gmaxwell> Also saves our own butts should someday we ship a bug that gobbles wallets.
810 2012-02-01 16:24:50 <gmaxwell> BlueMatt: delete them all on passphrase change, I guess.
811 2012-02-01 16:25:26 <Moron__> shouldnt the backups be encrypted just incase?
812 2012-02-01 16:25:31 <sipa> s/password change/CDB::Rewrite/
813 2012-02-01 16:25:35 <gmaxwell> Moron__: they're encrypted.
814 2012-02-01 16:25:44 <gmaxwell> well, if the user has encryption the backups will also be encrypted.
815 2012-02-01 16:25:54 <gmaxwell> Moron__: backup is probably the wrong word to use here.
816 2012-02-01 16:26:02 <gmaxwell> These aren't backups. They oopswallets.
817 2012-02-01 16:26:08 <gmaxwell> er they are.
818 2012-02-01 16:26:25 <gmaxwell> I don't even think we'd have any code to read from them. just code to write them.
819 2012-02-01 16:27:24 <gavinandresen> I'd really like the oopswallet to be a piece of paper in a safe deposit box.
820 2012-02-01 16:27:53 <gmaxwell> gavinandresen: Have a scripting language that runs on the user? :)
821 2012-02-01 16:28:39 <gmaxwell> In any case, yea "have good backups" is the right answer, but its a little like "don't have a trojan compromised machine" ... we know not all users will follow it. :)
822 2012-02-01 16:29:29 <gmaxwell> I wouldn't even advertise the oopswallets to the user, and certantly wouldn't call them backups.
823 2012-02-01 16:29:44 <gavinandresen> We talking GUI users or bitcoind users?
824 2012-02-01 16:30:08 <gmaxwell> Both but mostly GUI users, of course.
825 2012-02-01 16:30:31 <gmaxwell> just keep wallet.old0 .. wallet.oldN .. and every time the user goes through 90% of the keypool size, overwrite the oldest one. Delete all of them on resilver, after the resilver is successful.
826 2012-02-01 16:30:50 <gavinandresen> GUI users I agree; we need wallet backup in the GUI.  With auto-prompts "at the right time"
827 2012-02-01 16:31:15 <gmaxwell> This is orthorgonal, though now that you bring it up it's a sin we don't have a file->backup in the GUI.
828 2012-02-01 16:31:39 <sipa> 18:27:24 < gavinandresen> I'd really like the oopswallet to be a piece of paper in a safe deposit box.   <-- determinstic wallets? :)
829 2012-02-01 16:31:57 <gmaxwell> ^ the use there is we only have to nag them to backup onceish.
830 2012-02-01 16:32:05 <gavinandresen> yes, deterministic wallet is the way to go I think