1 2013-04-11 13:53:49 <jsfsn> sfraise: You can also use blockchain.info or some other API-vendor
  2 2013-04-11 13:53:53 <sfraise> I guess my real question is if seller lists a product, then buy buys product, buyer enters payment info and hits "send", can site owner take a percentage of the transaction directly
  3 2013-04-11 13:54:03 <sydna> well, yes.
  4 2013-04-11 13:54:34 <sydna> before setting up a site, you might want to read the bitcoin wiki a bit, learn how transactions work and what they are
  5 2013-04-11 13:55:35 <sydna> not really. the user would send to a generated address, and you'd then do a TX from there to the servers and sellers wallets.
  6 2013-04-11 13:56:00 <sfraise> I've looked through it a little but I haven't really determined if you can actually do it directly that way
  7 2013-04-11 13:56:06 <sfraise> ic, so that answers my question then
  8 2013-04-11 13:56:43 <sfraise> my issue with that is transaction fees would be doubled
  9 2013-04-11 13:56:54 <sfraise> would then not? assuming a transaction fee was set
 10 2013-04-11 13:57:03 <sydna> it would be.
 11 2013-04-11 13:57:05 <sfraise> "they"
 12 2013-04-11 13:57:21 <sydna> the transaction fees are hardly an issue though
 13 2013-04-11 13:57:25 <new299> yay, someone cleared the ban on my ISP. Thank you ops.
 14 2013-04-11 13:57:29 <sydna> 0.0005BTC is next to nothing.
 15 2013-04-11 13:57:35 <sfraise> so... back to my original question, how do we do it on the fly and limit the trans fee to 1
 16 2013-04-11 13:57:41 <sydna> you don't.
 17 2013-04-11 13:57:48 <jsfsn> sydna: Espcially after today (;
 18 2013-04-11 13:57:58 <sfraise> hardly an issue means it's still an issue
 19 2013-04-11 13:58:18 <sfraise> right now maybe it's not a big issue, but pressing forward it seems like it's a bottle neck
 20 2013-04-11 13:58:22 <jsfsn> sfraise: Read the Bitcoin whitepaper and the wiki
 21 2013-04-11 13:58:24 <sydna> there's no way of doing it as you want, unless you asked them to send you a fee, and then a separate TX to the seller.
 22 2013-04-11 13:58:44 <sydna> go and understand how TX work, that would answer this question.
 23 2013-04-11 13:58:51 <sfraise> well actually maybe that's the way to go jsfsn
 24 2013-04-11 13:58:58 <jsfsn> You basiclly want to route payments with a fee, which is quite trivial
 25 2013-04-11 13:59:12 <sfraise> instead of taking the full amount and then paying the seller, maybe you just split the transaction into two
 26 2013-04-11 13:59:26 <serp> your site would have a wallet.  which would provide an address... the person originally paying would have to tx fee to send to your site.  then your site would pay 1 to send to the seller.
 27 2013-04-11 13:59:47 <sydna> serp: yeah, but he wants to do it in a single transaction, which isn't possible.
 28 2013-04-11 14:00:22 <sfraise> I'd also prefer for the actual payment not to hit my hands ever, just the fee I would charge for the service
 29 2013-04-11 14:00:34 <serp> not possible
 30 2013-04-11 14:00:54 <jsfsn> Well, it is
 31 2013-04-11 14:00:55 <serp> you could have it go directly to the sellers then ask them for fees
 32 2013-04-11 14:01:04 <jsfsn> He can monitor the sellers address as well as his own
 33 2013-04-11 14:01:08 <MC1984> you cant skim fees of a txn unless youre a miner
 34 2013-04-11 14:01:08 <sfraise> through a json string couldn't you just split a transaction into two?
 35 2013-04-11 14:01:12 <jsfsn> Ask the buyer to both pay the seller and his address
 36 2013-04-11 14:01:17 <sipa> certainly possible, but it meams 1) exposing the receiver address to the sender, and trust the sender to not just leave out your fee
 37 2013-04-11 14:01:30 <serp> the problem is the customer is the one doing the transaction.. not your site
 38 2013-04-11 14:01:34 <sydna> jsfsn: then they'd still be paying twoTX fees, which was the objection in the first place
 39 2013-04-11 14:01:55 <jsfsn> sydna: Indeed, two transactions has to happen
 40 2013-04-11 14:01:57 <sfraise> that's what I'm saying jsfsn, but the buyer wouldn't need to actually know he's paying two, you could do it in the code couldn't you?
 41 2013-04-11 14:02:06 <sydna> no.
 42 2013-04-11 14:02:07 <jsfsn> sfraise: no
 43 2013-04-11 14:02:08 <sipa> no
 44 2013-04-11 14:03:11 <jsfsn> Two tx-fees are inevitable
 45 2013-04-11 14:06:43 <sfraise> maybe I'm wrong, but say you have an ajax pay button where you gather the seller's wallet, buyer's wallet, and amount to pay, which sends to a php file where it splits the transaction down before actually doing the transaction where 99% is sent to seller's wallet and the other 1% is sent to site owner's wallet, you're saying it's impossible
 46 2013-04-11 14:07:20 <jsfsn> sfraise: I'm not even close to understanding the logic there
 47 2013-04-11 14:07:34 <jsfsn> sfraise: That is obviously not possible, how could it be?
 48 2013-04-11 14:08:05 <jsfsn> sfraise: Again, go read the wiki
 49 2013-04-11 14:09:35 <jsfsn> Actually, it could obviously be done if you had the sellers wallet, but I know you did not mean that
 50 2013-04-11 14:10:37 <pjorrit_> this sounds like a nightmare waiting to happen
 51 2013-04-11 14:10:44 <jsfsn> Indeed
 52 2013-04-11 14:11:10 <jsfsn> I hope your not planning to implement this your self, sfraise?
 53 2013-04-11 14:12:46 <sydna> bitcoin and PHP is a scary combination anyway
 54 2013-04-11 14:14:02 <helo> smoking and gas leak repair
 55 2013-04-11 14:14:46 <sydna> eyedrops and sulphuric acid in similar bottles
 56 2013-04-11 14:21:39 <abadr> sipa: do you have a reference for the discrepancies discovered while writing the haskell client? original website for it seems to be down
 57 2013-04-11 14:22:46 <MC1984> is there a help/suppot chan yet o is this still it
 58 2013-04-11 14:23:53 <midnightmagic> MC1984: There are topic-specific #bitcoin* channels depending on what you want to talk about.
 59 2013-04-11 14:24:53 <MC1984> im scared to /list on freenode :(
 60 2013-04-11 14:25:30 <wumpus> MC1984: #bitcoin-tech
 61 2013-04-11 14:26:03 <midnightmagic> MC1984: I think you can use wildcards? not sure though.
 62 2013-04-11 14:27:02 <midnightmagic> MC1984: My mistake. You can just list based on certain parameters like number of users..
 63 2013-04-11 14:27:27 <sipa> abadr: the guy who wrote it discovered the ability to use compressed public keys, key recovery, found the problem with duplicate transactions (see BIP30 and http://r6.ca/blog/20120206T005236Z.html), found the origins of the now-considered-useless OP_CODESEPARATOR, ... i wonder if i forget something)
 64 2013-04-11 14:27:45 <sipa> wait, not key recovery
 65 2013-04-11 14:27:49 <TD> yeah, but i think that's not really to do with haskell and more the fact that he was a genius
 66 2013-04-11 14:27:52 <TD> i mean, the two may be related
 67 2013-04-11 14:28:00 <TD> but haskell isn't a magical way to find any of those things
 68 2013-04-11 14:28:02 <pyrret> any news about bitstamp..?
 69 2013-04-11 14:28:13 <gonffen> midnightmagic: fyi that works on some networks like rizon
 70 2013-04-11 14:29:09 <gmaxwell> TD: I have 8 projects in coverity's open source stuff, it's a huge pain to get them to add a project, and so far it's never actually been useful for any of mine??? at least not over the available alternative. (Okay, it found a bug in one of the speex tests)  ... but if people think it would be useful I can request an account for Bitcoin.
 71 2013-04-11 14:29:19 <TD> oh
 72 2013-04-11 14:29:24 <TD> we used to use it on wine
 73 2013-04-11 14:29:26 <TD> it found quite a lot of things
 74 2013-04-11 14:29:30 <TD> but that was a huge codebase
 75 2013-04-11 14:29:40 <sipa> TD: i think it's partly related to the fact that Haskell is just *different* and required a different implementation approach than simply line-per-line transliterating the imperative satoshi code
 76 2013-04-11 14:29:41 <TD> of course, a lot of FPs too. i doubt bitcoin is particularly buggy in the ways it can detect
 77 2013-04-11 14:29:46 <midnightmagic> gonffen: Yeah and I think there are network services sometimes that will do that too.
 78 2013-04-11 14:29:54 <gmaxwell> (It's substantially similar to clang static analysis but with better false positive surpression and a good database for tracking long standing issues)
 79 2013-04-11 14:30:15 <TD> sipa: maybe. i guess it depends ... i don't see how functional vs imperative helps you discover compressed public keys. i mean, that is just how closely you examine your ecdsa implementation isn't it
 80 2013-04-11 14:30:33 <sipa> TD: it did mean he was forced to have a better understanding of the system
 81 2013-04-11 14:30:39 <TD> maybe so
 82 2013-04-11 14:31:40 <TD> who is spamming testnet with transactions that lack inputs?
 83 2013-04-11 14:32:14 <gmaxwell> There is either an attack or a bug. It's not clear what it is.
 84 2013-04-11 14:32:19 <sipa> gmaxwell: didn't you discover that it was BitcoinJ code relaying these 0-input transactions? or has that been cleared up?
 85 2013-04-11 14:32:41 <TD> bitcoinj doesn't accept inbound connections nor does it relay
 86 2013-04-11 14:32:44 <TD> so it can't be that
 87 2013-04-11 14:32:51 <TD> unless someone heavily modded it of course
 88 2013-04-11 14:32:52 <gmaxwell> There was some evidence of that.
 89 2013-04-11 14:33:10 <gmaxwell> what was that txn id again?
 90 2013-04-11 14:33:12 <TD> it might announce them, i guess, if somebody managed to get one into their wallet
 91 2013-04-11 14:33:16 <gonffen> midnightmagic: just wanted to make sure you knew you weren't losing it ;)
 92 2013-04-11 14:33:23 <sipa> TD: right
 93 2013-04-11 14:34:02 <gmaxwell> TD: the txn in question doesn't just have no inputs the whole thing is: (looks it up)
 94 2013-04-11 14:34:39 <gmaxwell> 01000000000000000000
 95 2013-04-11 14:34:54 <sipa> txid d21633ba23f70118185227be58a63527675641ad37967e2aa461559f577aec43
 96 2013-04-11 14:35:13 <TD> huh
 97 2013-04-11 14:35:18 <gmaxwell> http://code.google.com/p/bitcoinj/issues/attachmentText?id=350&aid=3500001000&name=coinbase-tx-2.log&token=wsD6ZTnKn_8yeY_TLh0FDNORzHk%3A1363719034057
 98 2013-04-11 14:35:18 <sipa> 03:32:09< gmaxwell> http://code.google.com/p/bitcoinj/issues/attachmentText?id=350&aid=3500001000&name=coinbase-tx-2.log&token=wsD6ZTnKn_8yeY_TLh0FDNORzHk%3A1363719034057
 99 2013-04-11 14:35:21 <sipa> 03:32:18< gmaxwell> I cite: W/System.err(26246): 6842310 [New I/O worker #13] INFO com.google.bitcoin.core.MemoryPool - [95.27.142.162]:8333: Announced new transaction [1] d21633ba23f70118185227be58a63527675641ad37967e2aa461559f577aec43
100 2013-04-11 14:35:53 <TD> that means the peer announced it
101 2013-04-11 14:35:57 <TD> not that the client announced it
102 2013-04-11 14:36:14 <TD> specifically that the IP 95.27.142.162 announced it and we downloaded it
103 2013-04-11 14:36:21 <gmaxwell> oh. darn.
104 2013-04-11 14:36:33 <gmaxwell> Then back to knowing nothing.
105 2013-04-11 14:36:49 <TD> it's not very clear, i agree. It should probably say Peer announced new transaction. i should whack some static tx.verify() calls into more places so peers that send us garbage are disconnected earlier.
106 2013-04-11 14:37:29 <sipa> at least we know whom not to blame anymore :)
107 2013-04-11 14:37:45 <abadr> sipa: why does he call it Midas Money?
108 2013-04-11 14:38:14 <sipa> abadr: good question :)
109 2013-04-11 14:38:24 <abadr> o_O
110 2013-04-11 14:39:33 <TD> i found out in the last few days why talking heads on rolling TV news never seem to know anything
111 2013-04-11 14:39:48 <TD> they try and find people who can make it to a studio at the time they go live, with about 6-8 hours notice
112 2013-04-11 14:39:56 <TD> (that's when they are busy trying to book someone)
113 2013-04-11 14:40:24 <TD> where "make it to a studio" eliminates anyone not in the USA or London, for the particular network in question. needless to say, this is a good way to find people who aren't actually involved with whatever is being reported on.
114 2013-04-11 14:41:19 <Plornt> Anyone got any ideas on some bitcoin services that have been requested to be made?
115 2013-04-11 14:41:29 <TD> Plornt: there are tons of things you could do, if you're looking for work
116 2013-04-11 14:42:15 <Plornt> Hmm any ideas on what there is to do or a list of things to do >.<?
117 2013-04-11 14:42:19 <Plornt> Im not looking to earn any money
118 2013-04-11 14:42:27 <Plornt> just want to help out in general with things
119 2013-04-11 14:42:48 <TD> depends what you mean by "services". web apps? p2p apps? core protocol infrastructure?
120 2013-04-11 14:43:24 <Plornt> Im more suited towards web development
121 2013-04-11 14:44:25 <TD> it'd be nice if there was an appstore-like catalogue of merchants and services, but one with a strong emphasis on editorial quality, high quality entries, maybe some user reviews etc
122 2013-04-11 14:44:43 <TD> there have been a bunch of attempts at this but none ever took off because they were kind of half-assed
123 2013-04-11 14:47:08 <topi`> sipa: the way I see it, I think that bitcoin reference client ought to be coded in Haskell, for the numerous merits of it regarding to bugs and misbehaviour
124 2013-04-11 14:47:37 <topi`> ...but that won't happen as long as most ppl are not familiar with Haskell
125 2013-04-11 14:49:03 <TD> i learned some haskell once, and came away with the strong impression that the costs weren't worth the benefits
126 2013-04-11 14:49:21 <sipa> topi`: if anything, you'd end up with a reference client that doesn't match what is actually deployed on the network
127 2013-04-11 14:50:20 <topi`> that's another problem. but I could easily see a situation where a proper Haskell client will outperform the C++ client
128 2013-04-11 14:50:49 <topi`> with projects like Data Parallel Haskell
129 2013-04-11 14:50:51 <sipa> topi`: so if there was a mismatch between them, it would be a bug in the actually used implementation for not following the standard set by the reference, but it would be the reference that is *wrong*, as consistency is more important for a consensus algorithm than 'correctness'
130 2013-04-11 14:51:02 <imsaguy> another user reporting a QT crash in #eligius
131 2013-04-11 14:51:37 <sipa> topi`: yes and no; getting really good performance in Haskell often needs non-ideomatic code, in my experience, and that is exactly what you'd want to avoid for an easy-to-understand reference implementation
132 2013-04-11 14:51:51 <topi`> sipa: that just means that we must emulate *bugs* as well :) in order to be compatible with the consensus.
133 2013-04-11 14:53:24 <topi`> sipa: I think the kinds of problems (matching tx's, calculating ECDSA etc) are in that range of problems that actually work really well with idiomatic code
134 2013-04-11 14:53:35 <pjorrit_> that's very difficult to do if you trip over things like a maximum allowed refered transactions in a block
135 2013-04-11 14:53:54 <topi`> that said, a qsort would be implemented faster using monads...
136 2013-04-11 15:00:28 <sfraise> is there any way to run bitcoind on a redhat linux box that will only update glibc to version 2.5 instead of 2.7?
137 2013-04-11 15:05:48 <wumpus> sfraise: doesn't it compile with glibc 2.5?
138 2013-04-11 15:06:20 <sfraise> no, i get an error that glibc_2.7 not found
139 2013-04-11 15:06:22 <sydna> sfraise: you shouldn't have issues using the binary from SF either
140 2013-04-11 15:07:22 <wumpus> not sure about the binary, but I don't think we use a any recent glibc features, so if you compile from source it should work
141 2013-04-11 15:09:23 <Diapolo> wumpus: I really lile Jonas work on the graphics :).
142 2013-04-11 15:10:10 <wumpus> Diapolo: yes, me too
143 2013-04-11 15:10:34 <wumpus> and he's mac developer too, that's great
144 2013-04-11 15:10:45 <Diapolo> you have a little more time for Bitcoin recently, that is nice
145 2013-04-11 15:10:56 <Diapolo> yeah indeed, Mac love was needed badly
146 2013-04-11 15:11:21 <wumpus> yes, somewhat more
147 2013-04-11 15:11:23 <sydna> still is needed.
148 2013-04-11 15:11:48 <wumpus> what is needed?
149 2013-04-11 15:11:55 <sydna> Mac love
150 2013-04-11 15:12:09 <wumpus> can you be specific?
151 2013-04-11 15:12:17 <Diapolo> When is 0.8.2 scheduled and what are the blockers now?
152 2013-04-11 15:12:37 <Diapolo> and will there be an RC phase?
153 2013-04-11 15:12:40 <sydna> in 0.8.1 if you close the main window, you can't get it back without restarting the client
154 2013-04-11 15:12:51 <sydna> if you change the proxy settings you can't change them back
155 2013-04-11 15:13:03 <sydna> and the dock menu is missing
156 2013-04-11 15:13:04 <wumpus> the window problem is known, yes
157 2013-04-11 15:13:16 <wumpus> dock menu issue should be fixed
158 2013-04-11 15:13:27 <wumpus> I don't know of an issue with the proxy settings
159 2013-04-11 15:13:39 <sydna> I think it's filed on github
160 2013-04-11 15:14:05 <Diapolo> that proxy stuff is on my list, IMHO we should not try to allow on-the-fly changes, as I think this will duplicate code from core in Qt though
161 2013-04-11 15:14:11 <sfraise> how do you compile bitcoind from source in redhat linux?
162 2013-04-11 15:14:23 <wumpus> Diapolo: I don't really know what is blocking 0.8.2
163 2013-04-11 15:14:26 <sydna> sfraise: make -f makefile.unix
164 2013-04-11 15:14:48 <sydna> Diapolo: excellent
165 2013-04-11 15:15:01 <Diapolo> wumpus: translations are nearly up to date, one last pull during a possible RC phase and I'm really satisfied
166 2013-04-11 15:15:44 <wumpus> sfraise: and follow the build instructions in doc/build-unix.txt
167 2013-04-11 15:16:51 <sfraise> that works for redhat fedora? I thought that was for ubuntu
168 2013-04-11 15:17:40 <Luke-Jr> 
ACTION ponders if it might make sense to make 0.9 simply be CodeShark's multiwallet + coin control and defer every other 0.9-planned feature for 0.10

169 2013-04-11 15:17:40 <wumpus> some parts may be for ubuntu, but most of the instructions hold for any unixy os
170 2013-04-11 15:18:17 <sfraise> no such file or directory
171 2013-04-11 15:18:17 <sydna> sfraise: if you are having difficulty reading build instructions, I seriously doubt you should be running a financial website.
172 2013-04-11 15:18:36 <Diapolo> Was CodeShark here recently? I'm a bit afraid such big refactorings break additional stuff badly... it need's to be better planned still.
173 2013-04-11 15:19:03 <Luke-Jr> Diapolo: hence why to take it in small steps
174 2013-04-11 15:19:08 <sfraise> dude you can't compile using ubuntu commands on redhat fedora lol
175 2013-04-11 15:19:10 <Luke-Jr> CodeShark's here more than you are :P
176 2013-04-11 15:19:36 <Diapolo> Luke-Jr: maybe, but his help with fixing bugs was zero after wumpus merged his first big patch
177 2013-04-11 15:27:42 <wumpus> Diapolo: I think we had the worst part of the GUI changes now for multiwallet... ofc with his core changes, we should test it much more seriously before merging anyway
178 2013-04-11 15:31:34 <sfraise> my bad... was thinking the box I was on was fedora, it's enterprise 5
179 2013-04-11 15:34:58 <EvilPete> Luke-Jr: there appears to be a fair amount of community support for an officially "blessed" coin-control release.
180 2013-04-11 15:35:25 <sipa> EvilPete: i'm sure Luke-Jr will answer that nothing is official :)
181 2013-04-11 15:36:42 <EvilPete> sipa: of course
182 2013-04-11 15:37:24 <Luke-Jr> EvilPete: pretty sure the only real stopper has been code quality :p
183 2013-04-11 15:38:04 <sipa> EvilPete: anyway, if all goes well, i think we'll get coin control in 0.9.0
184 2013-04-11 15:38:13 <sipa> (no promise)
185 2013-04-11 15:39:22 <TD> i think gavin is heads down on payment protocol support atm
186 2013-04-11 15:40:45 <Luke-Jr> 
ACTION sure hopes the payment protocol offers a real solution to DP

187 2013-04-11 15:40:46 <wumpus> in my opinion multiwallet does have priority over coin control, but yeah it'd be nice to get both in
188 2013-04-11 15:41:01 <sipa> wumpus: agree
189 2013-04-11 15:41:19 <sipa> Luke-Jr: i wouldn't get my hopes up about them supporting it
190 2013-04-11 15:41:43 <Luke-Jr> sipa: I know, but as long as it *offers* a solution, people will be able to see more clearly how harmful it is
191 2013-04-11 15:41:56 <midnightmagic> EvilPete: You can already do coin control with listunspent and the rawtransaction interface.
192 2013-04-11 15:42:05 <Luke-Jr> 1st it = PP, 2nd it = DP
193 2013-04-11 15:42:43 <wumpus> yes, the non-ui part of coincontrol has been merged for a while
194 2013-04-11 15:44:06 <midnightmagic> The only thing missing is a way to graph-search for connectivity between addresses.
195 2013-04-11 15:44:34 <sipa> getaddressgroupings?
196 2013-04-11 15:45:42 <midnightmagic> sipa: I must misunderstand what that does.
197 2013-04-11 15:45:49 <midnightmagic> 
ACTION goes to read.

198 2013-04-11 15:46:16 <midnightmagic> sipa: You meant listaddressgroupings right?
199 2013-04-11 15:46:22 <sipa> right
200 2013-04-11 15:46:30 <sipa> i must admit i never used that command myself
201 2013-04-11 15:47:36 <midnightmagic> sipa: If that does an arbitrary-depth graph search then I guess voila :)
202 2013-04-11 15:51:43 <Diapolo> wumpus: you are probably right about Qt refactorings :) I'm out once more, looking forward to the final and mergable splash screen pull and 0.8.2 ;)
203 2013-04-11 16:00:52 <gmaxwell> midnightmagic: effectively, not very rapidly either.
204 2013-04-11 16:01:15 <midnightmagic> gmaxwell: Thanks!
205 2013-04-11 16:06:07 <MC1984> bitcoind memory 42mb
206 2013-04-11 16:06:10 <MC1984> wat
207 2013-04-11 16:06:55 <sipa> midnightmagic: how do you do that? :D
208 2013-04-11 16:06:59 <sipa> eh, MC1984
209 2013-04-11 16:07:27 <MC1984> i just saw it as low as 40mb
210 2013-04-11 16:07:29 <skinnkavaj> So are we ready to start rebuilding the economy once again?
211 2013-04-11 16:07:38 <MC1984> with 24 connections and 2000 mempool
212 2013-04-11 16:07:44 <skinnkavaj> And by that i mean further develop the bitcoin _protocol_
213 2013-04-11 16:08:32 <skinnkavaj> You should all take a look at this: http://www.reddit.com/r/Bitcoin/comments/1c4uxz/now_that_the_hype_is_gone_and_media_will_forget/
214 2013-04-11 16:09:11 <MC1984> !blocks
215 2013-04-11 16:09:13 <gribble> 230842
216 2013-04-11 16:09:19 <petertodd> skinnkavaj: that poster doesn't seem to get that #1 and #2 conflict with each other
217 2013-04-11 16:10:29 <petertodd> skinnkavaj: #3 is already happening, localbitcoins.com, #4 is a worthy goal, and is happening
218 2013-04-11 16:10:58 <sipa> MC1984: many blocks do you have...?
219 2013-04-11 16:11:17 <sipa> just having the block headers in memory for the whole chain takes around 40 MB on a 32-bit system
220 2013-04-11 16:11:26 <sipa> ;;bc,blocks
221 2013-04-11 16:11:27 <gribble> 230843
222 2013-04-11 16:11:27 <MC1984> all of them
223 2013-04-11 16:11:36 <MC1984> gone back up to 150mb now
224 2013-04-11 16:11:43 <MC1984> after a processblock
225 2013-04-11 16:12:51 <MC1984> peak usage is 400mb so thats about a magnitude of bouncing around on the memory usage lol
226 2013-04-11 16:14:34 <gmaxwell> sipa:  ... how does someone end up with a 200mbyte wallet?
227 2013-04-11 16:15:31 <petertodd> gmaxwell: it's really easy, send lots of transactions which are long-no-conf chains
228 2013-04-11 16:16:50 <gmaxwell> petertodd: nah, that doesn't work so well, go try it on testnet.
229 2013-04-11 16:17:07 <sipa> gmaxwell: doesn't really matter - it's ridiculous that it takes seconds to sync
230 2013-04-11 16:17:17 <gmaxwell> After about 30 it can't generate more than 1 per second... "Ratelimiting by inefficient algorithim"
231 2013-04-11 16:17:18 <petertodd> gmaxwell: That's exactly where I ran into the problem.
232 2013-04-11 16:17:59 <gmaxwell> sipa: oh I absolutely agree, but since you didn't ask I wondered if you might know already!
233 2013-04-11 16:18:41 <HM> gah stupid gcc
234 2013-04-11 16:18:44 <gmaxwell> I have no idea how the wallets get that big.
235 2013-04-11 16:18:59 <HM> gnu stdlibc++ doesn't support the locale sensitive version of isblank() yet -_-
236 2013-04-11 16:30:13 <gagecolton> psst
237 2013-04-11 16:30:36 <gagecolton> is there a way to get a bitcoin client working over ham radios?
238 2013-04-11 16:31:35 <petertodd> gagecolton: by far the easiest would be to get IP over ham working first
239 2013-04-11 16:33:05 <paulo_> hello
240 2013-04-11 16:33:07 <michagogo> Quick question... What does it mean when bitcoin returns "TX rejected (code -22)" from a sendrawtransaction?
241 2013-04-11 16:33:13 <michagogo> or, where is there documentation for this?
242 2013-04-11 16:33:31 <mungojelly> hey paulo_ yeah it's chaos over there :/
243 2013-04-11 16:33:44 <mungojelly> anyway!  let's make a blockchain that determines access to a forum, you have to buy an account on the forum through the blockchain.
244 2013-04-11 16:34:09 <sivu> forumcoin
245 2013-04-11 16:34:37 <michagogo> Alternately, is there a way to rebroadcast a transaction that's better than feeding the output of getrawtransaction into sendrawtransaction?
246 2013-04-11 16:35:08 <MC1984> i think b.i can broadcast raw txns
247 2013-04-11 16:35:42 <sipa> michagogo: it means bitcoind would not accept that transaction into its memory pool, if it were received from network
248 2013-04-11 16:35:44 <paulo_> I'm interested in alternative block chains, but I can't think of other applications.
249 2013-04-11 16:36:02 <sipa> michagogo: which is a very good indicator for the fact that other nodes wouldn't accept it either
250 2013-04-11 16:36:06 <michagogo> sipa: Hmm
251 2013-04-11 16:36:18 <MC1984> paulo_ theres one that tried to be a DNS root
252 2013-04-11 16:36:32 <michagogo> Does getrawtransaction not return what sendrawtransaction wants?
253 2013-04-11 16:36:32 <paulo_> what needs distributed concensus?
254 2013-04-11 16:36:35 <MC1984> or just a key/value database
255 2013-04-11 16:36:39 <paulo_> hmm
256 2013-04-11 16:36:40 <mungojelly> well we should be able to make a blockchain that supports some sort of general computing platform inside of it, but maybe we have to iterate through these special purpose chains first to comprehend it
257 2013-04-11 16:37:02 <sipa> michagogo: getrawtransaction works?
258 2013-04-11 16:37:22 <sipa> michagogo: that explains, it means bitcoin already has it in its memory pool (and it is likely already sent out)
259 2013-04-11 16:37:27 <michagogo> Ah
260 2013-04-11 16:37:31 <michagogo> Yeah, it was sent out
261 2013-04-11 16:37:35 <mungojelly> or i dunno let's just go for the brass ring.  it computes something on all of the nodes of the network.  you'd have to pay within the blockchain for that computing power.  hmm.
262 2013-04-11 16:37:49 <michagogo> sipa: Is there a way to make it forget it so I can rebroadcast it?
263 2013-04-11 16:37:58 <michagogo> e
264 2013-04-11 16:37:58 <michagogo> Specifically, after connecting to a new nod
265 2013-04-11 16:38:16 <sipa> michagogo: you're not going to like it: yes, by restarting the node
266 2013-04-11 16:38:29 <sipa> michagogo: but remember that your peers won't relay if they knew about it already
267 2013-04-11 16:38:46 <michagogo> sipa: So quit the client and restart it?
268 2013-04-11 16:38:52 <sipa> michagogo: yes
269 2013-04-11 16:39:00 <michagogo> Why do you assume I won't like it?
270 2013-04-11 16:39:45 <mungojelly> is the plan to integrate coin coloring into the main Bitcoin client once it's more stable?  i'm wondering what's the time-table for that transition.  i know it's resisted by some people, are we still pretending it's not inevitable? ;)
271 2013-04-11 16:40:04 <BlueMatt> mungojelly: uhh...wat?
272 2013-04-11 16:40:04 <petertodd> mungojelly: no
273 2013-04-11 16:40:54 <Mylon> mungo: what would in your opinion be the added value of color coding computer bits?
274 2013-04-11 16:41:14 <Mylon> it would require an additional 8 bits (if not more) so it needs to have a purpose ;)
275 2013-04-11 16:41:14 <paulo_> "coin coloring"
276 2013-04-11 16:41:43 <michagogo> Ah, I see -- and if a transaction's already in the blockchain, it gets a different error
277 2013-04-11 16:41:50 <michagogo> "transaction already in block chain (code -5)"
278 2013-04-11 16:42:05 <michagogo> Which makes me wonder, why is there not an error for "I already know this"
279 2013-04-11 16:42:07 <Mylon> that would imo just be a GUI enhancement
280 2013-04-11 16:42:07 <sipa> when it's already in the mempool, there should be a clearer error
281 2013-04-11 16:42:14 <paulo_> If i was to make a blockchain for key/value pairs, what would be the incentive for extending the chain?
282 2013-04-11 16:42:15 <mungojelly> Mylon: hm?  no it doesn't require any protocol changes as the bitcoinx folks are designing it.
283 2013-04-11 16:42:17 <sipa> and perhaps even allow to force-rebroadcast
284 2013-04-11 16:42:47 <mungojelly> paulo_: namecoin is a blockchain for key/value pairs.  the incentive is that some people are buying namecoin, for some reason. :/
285 2013-04-11 16:43:10 <michagogo> sipa: Good to know that what I figured out by myself works as long as I restart my client
286 2013-04-11 16:43:19 <mungojelly> coin coloring as i understand it is just a matter of looking over the blockchain and seeing which transactions descend in a proper way from a chosen genesis transaction.
287 2013-04-11 16:43:31 <sipa> michagogo: feel free to file a bug about this, so it's not forgotten
288 2013-04-11 16:43:38 <paulo_> I don't undesrtand how namecoin is a currency yet a key/value storage system.
289 2013-04-11 16:44:01 <michagogo> I'll do that as soon as you point me to the bugtracker :-)
290 2013-04-11 16:44:07 <mungojelly> paulo_: it's cool, even though i don't much like how namecoin is implemented in its details.  the system itself sells you names.
291 2013-04-11 16:44:13 <petertodd> paulo_: easy, it's Bitcoin + a special transaction type that is defined as a key/value definition
292 2013-04-11 16:44:28 <michagogo> paulo_: It has a mechanism to destroy coins in exchange for putting a value into the blockchain
293 2013-04-11 16:44:44 <michagogo> Well, not destroy, but take them out of use as coins
294 2013-04-11 16:44:55 <petertodd> paulo_: You could implement exactly what namecoin is on Bitcoin itself, although doing so won't make anyone very pleased with you.
295 2013-04-11 16:45:01 <mungojelly> paulo_: Bitcoin only knows how to sell people Bitcoin transactions (so far!?), but it does the same thing if you think about it.  it autonomously collects fees for transactions and then gives them to miners in exchange for blocks, it acts as a vendor of transactions.
296 2013-04-11 16:45:16 <Luke-Jr> paulo_: namecoin isn't really a currency.
297 2013-04-11 16:45:35 <michagogo> Hmm, does bitcoin-qt automatically drop peers?
298 2013-04-11 16:45:41 <mungojelly> i've seen people buy things for namecoin.  not sure why but it happens.
299 2013-04-11 16:45:52 <michagogo> I launched, a few seconds later I had about 17 peers
300 2013-04-11 16:45:53 <sipa> michagogo: only if they misbehave
301 2013-04-11 16:45:58 <michagogo> Then it dropped to 12
302 2013-04-11 16:46:04 <sipa> michagogo: but your peers can disconnect for some reasons too
303 2013-04-11 16:46:08 <michagogo> Ah
304 2013-04-11 16:46:16 <BenderCoin> Luke-Jr, please explain why you think  namecoin is not really a currency
305 2013-04-11 16:46:20 <Luke-Jr> michagogo: people can buy things with anything, doesn't make it a currency ;P
306 2013-04-11 16:46:22 <michagogo> Like what?
307 2013-04-11 16:46:25 <Luke-Jr> like carrots
308 2013-04-11 16:46:52 <Luke-Jr> BenderCoin: it's a key/value system
309 2013-04-11 16:46:54 <michagogo> [21:42:32] <@sipa> michagogo: feel free to file a bug about this, so it's not forgotten
310 2013-04-11 16:46:54 <michagogo> Where would I do that?
311 2013-04-11 16:47:01 <mungojelly> well, i'd say it's not black or white.  it does make carrots a little more currencyful.  if people bought stuff for carrots all day long, it would add up to their being a currency.
312 2013-04-11 16:47:04 <sipa> well, namecoin is a weird currency in any sense, because it has a fixed exchange rate to being used as tokens for registering names
313 2013-04-11 16:47:10 <BenderCoin> Luke-Jr, no, namecoin is bitcoin + name value system extension.
314 2013-04-11 16:47:23 <sipa> if it becomes too useful for registering names, it will become terrible expensive as a currency
315 2013-04-11 16:47:34 <BenderCoin> namecoins have exactl properties as bitcoins, same code everything.
316 2013-04-11 16:47:49 <sipa> if it becomes too exchange as a currency, it won't be used to register names
317 2013-04-11 16:47:52 <michagogo> !google bitcoin bug tracker
318 2013-04-11 16:47:53 <gribble> Bitcoin: <http://bitcoin.org/>; Bitcoin version 0.5.1 released: <http://bitcoin.org/releases/2011/12/15/v0.5.1.html>; Common Vulnerabilities and Exposures - Bitcoin: <https://en.bitcoin.it/wiki/Common_Vulnerabilities_and_Exposures>
319 2013-04-11 16:47:54 <Luke-Jr> BenderCoin: you're ignoring colours
320 2013-04-11 16:47:58 <michagogo> hmm
321 2013-04-11 16:48:02 <mungojelly> anyway obviously you can write a blockchain to do anything (that can be computed by all the nodes of the network) in exchange for anything within the system.
322 2013-04-11 16:48:06 <michagogo> !google file bitcoin bug
323 2013-04-11 16:48:07 <gribble> Bitcoin - Browse /Bitcoin/bitcoin-0.7.1 at SourceForge.net: <http://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.7.1/>; Bitcoin - Browse /Bitcoin/bitcoin-0.8.0 at SourceForge.net: <http://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.8.0/>; Bitcoin - Browse /Bitcoin/bitcoin-0.7.2 at SourceForge.net: (1 more message)
324 2013-04-11 16:48:19 <sipa> michagogo: https://github.com/bitcoin/bitcoin/issues
325 2013-04-11 16:48:23 <BenderCoin> namecoin db records are held and exchanged as a separate mechanism from namecoins themselves. so you could say namecoin is not a currency, but namecoins are
326 2013-04-11 16:48:23 <michagogo> thanks
327 2013-04-11 16:48:44 <mungojelly> namecoin is a proof of concept.  it doesn't work especially well but it does work.
328 2013-04-11 16:49:12 <paulo_> what about a storage of magnet links for torrents.
329 2013-04-11 16:49:44 <BenderCoin> mungojelly, yeah, the spec says 1023 bytes of data per name, I lost a load of good names due to the bug where any data value over 500 bytes locks your name and you lose it
330 2013-04-11 16:50:26 <BenderCoin> Luke-Jr, no idea what 'ignoring colours' means
331 2013-04-11 16:50:31 <mungojelly> BenderCoin: ugh i didn't know that bug, i've only glanced at it, but it just doesn't feel well implemented. :(   weird that people still buy it; there aren't enough choices apparently.
332 2013-04-11 16:51:41 <mungojelly> paulo_: well just storing data, generally, as long as the model is that the keyholder can change the data, we might as well just generalize that
333 2013-04-11 16:52:03 <mungojelly> paulo_: which i guess is kinda what the namecoin system does but apparently badly :/
334 2013-04-11 16:53:04 <BenderCoin> mungojelly, there are no active developers so its getting rusty. it needs a killer app, like maybe a cross platform wallet, and some good user apps and maybe it could gain traction.
335 2013-04-11 16:53:23 <BenderCoin> traction=users+devs+$
336 2013-04-11 16:53:40 <mungojelly> BenderCoin: it's written in c++?  i don't think i want to learn c++ again until i have a larger brain :/
337 2013-04-11 16:54:42 <mungojelly> that's another thing, could we get a cryptocurrency network written in Python or Scheme or something that humans can comprehend? ;)  we'd have a lot more kinds of coin if people could make them by tweaking in Python.
338 2013-04-11 16:55:07 <sipa> armory and electrum are bitcoin clients written in python
339 2013-04-11 16:55:14 <sipa> neither implements the network rules though
340 2013-04-11 16:55:23 <sipa> (they're not full clients on themself)
341 2013-04-11 16:55:26 <BenderCoin> mungojelly, the bitcoin-ruby guys are doing some good stuff with namecoin also.
342 2013-04-11 16:55:37 <sipa> jgarzik's pynode is a full node written in python
343 2013-04-11 16:56:11 <mungojelly> BenderCoin: i hadn't heard of it, my Ruby is rusty but it's a beautiful language so maybe i'd like to help with that :D
344 2013-04-11 16:56:35 <michagogo> sipa: Done, https://github.com/bitcoin/bitcoin/issues/2512
345 2013-04-11 16:56:38 <Luke-Jr> BenderCoin: http://ansuz.sooke.bc.ca/entry/23
346 2013-04-11 16:56:51 <michagogo> And I apologize in advance for the fact that I'm not great at explaining stuff
347 2013-04-11 16:57:27 <mungojelly> i'm surprised there are so few cryptocurrencies.  are there really only like dozens of cryptocurrencies ever, including dead ones?  why?  can't any of you start one in an afternoon by tweaking this code you know well???
348 2013-04-11 16:58:00 <sipa> mungojelly: yes, we could, but any cryptocurrency that is just a few tweaks isn't worth experimenting with in my opinion
349 2013-04-11 16:58:04 <pjorrit_> most who know the code well prefer to focus on bitcoin
350 2013-04-11 16:58:07 <Luke-Jr> mungojelly: you end up creating a pump-and-dump scam because there's no reason to adopt your new "cryptocurrency"
351 2013-04-11 16:58:36 <mungojelly> i can understand i guess why there's not many serious attempts, though i'm surprised by that too really, but what really surprises me is that there's hardly any experimentation, play, fooling around
352 2013-04-11 16:58:42 <Luke-Jr> adoption is necessary to the legitimacy of a cryptocurrency
353 2013-04-11 16:58:52 <sipa> mungojelly: there are very interesting things that could be investigated in new cryptocurrencies, but all of them at least require understanding the existing system and its issues well, and significant or even from-scratch implementations because they're so different
354 2013-04-11 16:58:59 <Luke-Jr> mungojelly: there's plenty of experimentation, mostly on bitcoin and its testnets
355 2013-04-11 16:59:20 <sipa> mungojelly: any other simple change is better integrated in bitcoin itself if possible :)
356 2013-04-11 16:59:27 <mungojelly> Luke-Jr: ok well maybe there's some testnet experimentation i'm unfamiliar with :)
357 2013-04-11 16:59:54 <mungojelly> sipa: well for instance a coin i thought of a couple weeks ago is GambleCoin-- sometimes balances just randomly go up and down
358 2013-04-11 17:00:02 <Luke-Jr> O.o;;
359 2013-04-11 17:00:10 <mungojelly> now obviously that would not be an especially valuable coin.  i hope.  it's just a silly fun idea.
360 2013-04-11 17:00:49 <michagogo> mungojelly: Except, where would the entropy for the randomness come from?
361 2013-04-11 17:00:52 <sipa> i'd just call it silly, but that's perhaps just me :)
362 2013-04-11 17:01:01 <michagogo> Also, is there anything I forgot to add in https://github.com/bitcoin/bitcoin/issues/2512 ?
363 2013-04-11 17:01:49 <sipa> michagogo: seems fine to me
364 2013-04-11 17:02:05 <mungojelly> michagogo: i dunno, i thought we had plenty of entropy to do random things, do we not?  run something in the system that makes entropy?  i hear there's some cellular automata that are useful.  that's what A New Kind Of Science said, anyway. :D
365 2013-04-11 17:02:24 <michagogo> mungojelly: But how would it be decentralized?
366 2013-04-11 17:02:55 <mungojelly> michagogo: take data for the entropy from a bunch of blocks in a row?
367 2013-04-11 17:03:05 <sipa> mungojelly: that means miners can control it
368 2013-04-11 17:03:10 <sipa> blocks aren't random
369 2013-04-11 17:03:29 <mungojelly> well but if it's chaotic and it's determined by a bunch of blocks it would make it too expensive.
370 2013-04-11 17:03:57 <pjorrit_> well if the last of that bunch would be deciding
371 2013-04-11 17:04:10 <mungojelly> if you're taking the data from the hash of the block it's crazy expensive even for one block, isn't it??
372 2013-04-11 17:04:31 <sipa> mungojelly: depends for what
373 2013-04-11 17:04:40 <mungojelly> or like impossible.  it has to be 0s at the beginning, if you're taking data from the end too.. you'd have to do the whole thing.
374 2013-04-11 17:05:00 <sipa> please be more concrete
375 2013-04-11 17:05:27 <sipa> if you're really planning on doing this, think about it well, work out the math and formula's you'd use and come back
376 2013-04-11 17:05:38 <BlackPrapor> Luke-Jr: Hello Luke. It appears that greedy ddosers are more dangerous to bitcoin, than atlcoins -_-. Do you think its possible to do something like p2p based exchange? I've seen people discussing this today
377 2013-04-11 17:05:47 <pjorrit_> or just code it up and launch it, ppl will eat it up ;p
378 2013-04-11 17:06:11 <Luke-Jr> BlackPrapor: it definitely is possible - see #bitcoin-otc
379 2013-04-11 17:06:17 <eklass> BlackPrapor: if there was a "p2p exchanbge", who would hold the fiat deposits?
380 2013-04-11 17:06:19 <mungojelly> well it's a general class of systems, cryptocurrency systems that develop some entropy, you could use randomness for lots of things.  i don't think it seems too difficult but yes you'd have to make really sure you'd really mixed shit up enough.
381 2013-04-11 17:06:42 <BlackPrapor> eklass: interesting question =)
382 2013-04-11 17:07:02 <Luke-Jr> BlackPrapor: another concept p2p exchange is ripple ;)
383 2013-04-11 17:07:08 <eklass> BlackPrapor: i think we've identified a major hurdle ;)
384 2013-04-11 17:07:13 <Luke-Jr> at least the original ripple
385 2013-04-11 17:07:14 <pjorrit_> you could go the #opentransactions route for that
386 2013-04-11 17:07:30 <pjorrit_> and have someone issuing tokens to represent the btc or dollars
387 2013-04-11 17:07:43 <pjorrit_> but then you need to trust that issuer to have the real goods
388 2013-04-11 17:07:44 <BlackPrapor> Luke-Jr: otc doesn't have a way to have a single exchange rate, does it?
389 2013-04-11 17:08:01 <Luke-Jr> BlackPrapor: so? :p
390 2013-04-11 17:08:55 <BlackPrapor> Luke-Jr: so, its hard to exchange when there is no agreeable price to exchange at
391 2013-04-11 17:09:33 <BlackPrapor> Luke-Jr: + all the problems with trust, all those scammers
392 2013-04-11 17:09:34 <eklass> BlackPrapor: people set the prices, exchanges don't. it's no different if you're exchanging between 1 seller and 1 buyyer or 1000 sellers and a 1000 buyers
393 2013-04-11 17:09:55 <Luke-Jr> ^
394 2013-04-11 17:10:22 <topace> hmm, im having problems with my data center bandwidth, whats the best way to limit bandwidth used by bitcoind ?
395 2013-04-11 17:10:25 <BlackPrapor> eklass: yes, so it would be awesome to have a way to see the average price all otc users exchange at..
396 2013-04-11 17:10:32 <topace> will keeping the connection count low (but reasonable) help ?
397 2013-04-11 17:10:50 <Luke-Jr> topace: the same way you limit bandwidth used by any application
398 2013-04-11 17:11:01 <mungojelly> BlackPrapor: in my experience people have been mostly trading based on the mtgox price everywhere :/
399 2013-04-11 17:11:03 <eklass> BlackPrapor: bitcoincharts has an API for rolling weighted prices
400 2013-04-11 17:11:20 <BenderCoin> MTGox == Comex. One price rules the world. NYSPOT is GOXSPOT
401 2013-04-11 17:11:28 <helo> if each block is allowed to be as large as typical node's bandwidth/cpu/mem permit, wouldn't bootstrapping a new node in log(n) time require an order of magnitude more capacity than the average node?
402 2013-04-11 17:11:39 <BlackPrapor> eklass: another problem is that you have to find a ton of traders if you want to make a large buy/sell
403 2013-04-11 17:11:41 <topace> Luke-Jr: and that is ?
404 2013-04-11 17:12:01 <HM2> meh
405 2013-04-11 17:12:04 <eklass> there's no way you could calculate on all OTC transactions when there's no central information store to record the transaction
406 2013-04-11 17:12:06 <helo> sorry, abusing the term "bootstrapping" to mean "syncing"
407 2013-04-11 17:12:08 <Luke-Jr> topace: shrug, I use my OpenWrt router's QoS
408 2013-04-11 17:12:11 <HM2> in big markets exchanges synchronise due to arbitrage
409 2013-04-11 17:12:23 <HM2> the exchanges are too crappy and the liquidity too low atm
410 2013-04-11 17:12:46 <topace> doesnt the linux kernel have some throttleing built into it these days?
411 2013-04-11 17:12:57 <BlackPrapor> HM2: because in most exchanges its hard to get the money in
412 2013-04-11 17:12:59 <_dr> when did this become #bitcoin-price? ;)
413 2013-04-11 17:13:04 <Luke-Jr> topace: probably, but you'll want a tool to configure it
414 2013-04-11 17:13:16 <helo> liquidity problems are characteristic of a manic market
415 2013-04-11 17:13:24 <HM2> if an asset is completely fungible then you'd be a fool to not to trade at the market price.
416 2013-04-11 17:13:58 <mungojelly> it seems to me the currency markets will decentralize themselves by turning into bots.  they're starting off doing a bit of arbitrage & ridiculous predictive trading, but naturally they have to expand.
417 2013-04-11 17:15:05 <helo> the only solution is for people to stop being irrational and buying at ephemeral sky-high prices. pavlov says everyone needs to experience a big crash to keep their expectations sane.
418 2013-04-11 17:16:34 <mungojelly> people are barely able now to comprehend dead wallets, but they are going to have to deal with having a zillion agents actively dealing with various assets for them, because it's not going to stay anywhere near a human level of complexity.
419 2013-04-11 17:17:32 <paulo_> wow, speculation discussion here too?
420 2013-04-11 17:17:47 <mungojelly> the problem isn't that it's hard to make decentralized exchanges.  the problem is making a decentralized exchange *simple enough for a human*.
421 2013-04-11 17:18:05 <helo> barring spamminess or actual dev discussion, offtopicness is somewhat tolerated
422 2013-04-11 17:18:42 <_dr> so? emacs or vim?
423 2013-04-11 17:18:43 <mungojelly> if the target audience is bots, it's a different story.  it can be a very complex interface where you have to determine the trustworthiness of thousands and thousands of other agents by various complex methods repeatedly in real time.
424 2013-04-11 17:19:14 <helo> although i'd be happy if it hadn't been, and my somewhat viable question had been addressed by someone :/
425 2013-04-11 17:19:28 <mungojelly> i've been using both emacs and vim but i hate both of them. :(  i hate them most when they work best for me, because they make me feel specially powerful and separate and not like i have power that i can share with all of humanity. :(
426 2013-04-11 17:19:35 <HM2> we saw a 10-12 fold increase in trading price inside a month
427 2013-04-11 17:19:39 <sipa> and one fish actually bites
428 2013-04-11 17:19:43 <HM2> yet people still expect that to continue
429 2013-04-11 17:20:34 <HM2> I'm glad i find the technical aspects of bitcoin more interesting than the social and economic implications
430 2013-04-11 17:20:47 <BlackPrapor> mungojelly: so, how do you think it would be possible to decentralise exchanges, and still be able to let everyone know the real supply/demand values, and spot price?
431 2013-04-11 17:21:13 <michagogo> HM2: I'm 100% with you on that
432 2013-04-11 17:21:28 <michagogo> Personally, I think the whole system is ingenious
433 2013-04-11 17:22:48 <MC1984> how the price of weed get s
434 2013-04-11 17:22:51 <MC1984> set?
435 2013-04-11 17:23:25 <MC1984> kinda commodity with no central price arbiter obv
436 2013-04-11 17:23:46 <mungojelly> BlackPrapor: um well you can't have 100% transparency of those things in any system can you?  i'm assuming it's going to be a chaotic mess.  but what i'm imagining is a network of untrusting peer agents.
437 2013-04-11 17:23:49 <MC1984> varies from place to place but its roughly within a price envelope
438 2013-04-11 17:23:57 <HM2> weed isn't fungible
439 2013-04-11 17:24:10 <HM2> it varies in quality
440 2013-04-11 17:24:15 <BenderCoin> if the drug for fiat market was as centralized as bitcoin with mtgox, the price of weed would be the Amsterdam Spot.
441 2013-04-11 17:24:28 <MC1984> yeah and each quality has its price
442 2013-04-11 17:24:48 <HM2> quality is hard to quanitify
443 2013-04-11 17:25:01 <HM2> personal opinion and taste
444 2013-04-11 17:25:10 <BenderCoin> Gox Strains: USD EUR etc
445 2013-04-11 17:25:17 <MC1984> well its price comes from somewhere
446 2013-04-11 17:25:21 <poggy> There is a saying as well, in many places the price doesn't vary but 'the cup suffers'
447 2013-04-11 17:25:23 <mungojelly> weed usually gets classified into a few different qualities, precisely to force that fake fungibility on it because the market needs it
448 2013-04-11 17:25:34 <poggy> meaning they just change how much you get
449 2013-04-11 17:26:03 <poggy> very common in poorer African countries
450 2013-04-11 17:26:12 <mungojelly> so there's no forethought from the main dev community about how coloring is going to be integrated into the whole system?  it seems important, to me. :/
451 2013-04-11 17:26:55 <MC1984> how does otc find its price
452 2013-04-11 17:27:02 <MC1984> just goes from gox or what
453 2013-04-11 17:28:46 <poggy> They don't really have a price
454 2013-04-11 17:28:47 <BlackPrapor> mungojelly: I was thinking, maybe it could be possible to keep the trust with fiat money in current exchanges. Then they'd issue a suplemental cryptoUSD (premined and distributed among major exchanges, and run offline, but blockchain would be available with some delay on Internet). Then BTC would be traded with cryptoUSD, but everyone would have to use a registered btc addresses with major xchange
455 2013-04-11 17:28:49 <sipa> otc uses whatever price people agree to
456 2013-04-11 17:28:55 <sipa> just like mtgox, actually
457 2013-04-11 17:29:14 <MC1984> i spose thats what it comes down to
458 2013-04-11 17:29:23 <poggy> It would be a lot easier to say they did if their bot was stricter about the way people listed offers
459 2013-04-11 17:29:24 <MC1984> the price is whatever the buy and sller agree
460 2013-04-11 17:29:46 <BlackPrapor> that way every peer would be able to trade fast, see real price, supply demand, and trust each other
461 2013-04-11 17:31:20 <_dr> well, you could always come up with a decentralized system like p2pool for trading. no need to keep old blocks, decentralized, use some crypto to prevent cheating. but you still need a central place that takes your $$$
462 2013-04-11 17:32:31 <BlackPrapor> _dr: correct. fiat currencies could be still handled by those companies, which already know how to deal with it, and comply with AML and etc
463 2013-04-11 17:33:07 <_dr> but look at all the people screaming for a central authority to shut down trading once the price crashes
464 2013-04-11 17:33:31 <pjorrit_> what? pussies ;p
465 2013-04-11 17:33:32 <_dr> even bender bending rodriguez would see the irony in that
466 2013-04-11 17:34:06 <BlackPrapor> the need for cryptoUSD (cryptoEUR and etc) is just to be able to trade p2p and see trade stats.
467 2013-04-11 17:43:39 <jok> hi@all
468 2013-04-11 17:43:53 <jok> Question. scriptPubKey=OP_DUP OP_HASH160 26cda7bbceabd8fe6208c31518ba74ec6dbb0dd9 OP_EQUALVERIFY OP_CHECKSIG
469 2013-04-11 17:44:33 <jok> is it right that "26cda7bbceabd8fe6208c31518ba74ec6dbb0dd9" is an bitcoin address ?
470 2013-04-11 17:44:45 <gribble> The bot responds when you start a line with the ! character. A good starting point for exploring the bot is the !facts command. You can also visit the bot's website for a list of help topics and documentation: http://gribble.sourceforge.net/
471 2013-04-11 17:44:45 <jok> !help
472 2013-04-11 17:45:09 <sipa> jok: in hex, yes
473 2013-04-11 17:45:17 <sipa> jok: it's typically encoded in base58check
474 2013-04-11 17:47:13 <jok> sipa, good, but if make pack() (like "26" => \\x26) and then EncodeBase58() i get "YMZihgNGMrQgoFLeXvyqdAvR3Sx"
475 2013-04-11 17:47:39 <jok> it is shorted then default bitcoin address
476 2013-04-11 17:47:50 <sipa> jok: base58check = base58 + 32-bit checksum
477 2013-04-11 17:48:23 <Luke-Jr> base58(format identifier + data + checksum)
478 2013-04-11 17:48:25 <sipa> jok: it means, before encoding as base58, you append the first 4 bytes of the double-sha256 hash of the data preceeding it
479 2013-04-11 17:48:28 <jok> thks
480 2013-04-11 17:48:43 <sipa> and indeed, there a version byte in front as (which is taken into account for the checksum)
481 2013-04-11 17:48:50 <Luke-Jr> format identifier 0 means OP_DUP OP_HASH160 <data> OP_EQUALVERIFY OP_CHECKSIG
482 2013-04-11 17:53:57 <jok> ok. is there is a function in bitcoind source which do convertation from such hex to right BTC address ?
483 2013-04-11 17:55:38 <Luke-Jr> jok: not likely
484 2013-04-11 17:56:35 <gonffen> !facts
485 2013-04-11 17:56:36 <gribble> To see a nice sortable web view of all factoids, click here: http://gribble.dreamhosters.com/viewfactoids.php?db=%23bitcoin-dev || To see a list of the most popular factoids, run !rank || To search factoids, run !factoids search <yoursearchterm>
486 2013-04-11 17:56:42 <dsal> Hey, what would cause a transaction to hang for a very long time?
487 2013-04-11 17:56:46 <dsal> http://blockchain.info/tx/ef7251a123301310fff1b502b5f6d459676f17892cc2c6c4d56417ba417a3565
488 2013-04-11 17:56:59 <dsal> I'm curious as to why this would be unconfirmed after nearly two hours.
489 2013-04-11 17:57:40 <gmaxwell> dsal: wrong channel, ask in #bitcoin
490 2013-04-11 17:57:57 <Scrat> #bitcoin-tech
491 2013-04-11 17:58:21 <gmaxwell> Scrat: no, that channel is for apps and tools stuff.
492 2013-04-11 17:58:37 <Scrat> ah, sorry
493 2013-04-11 18:01:33 <jok> Luke-Jr, it's strange, course we can send to btc-address in rigth form
494 2013-04-11 18:02:18 <michagogo> gmaxwell: Not what the topic says
495 2013-04-11 18:02:20 <michagogo> 'Technical aspects of bitcoin that don't belong in #bitcoin-dev friendlier than #bitcoin :)'
496 2013-04-11 18:08:00 <Diablo-D3> https://developer.amazon.com/sdk/coins/landing.html?ref_=pe_132830_29076940
497 2013-04-11 18:08:08 <Diablo-D3> gmaxwell, sipa: what the fuck am I looking at
498 2013-04-11 18:11:44 <diki> what kind of changes are needed in bitcoin to have special private keys that can be owned by a single entitiy only? This should basically allow for people to sell private keys and subsequently lose access to them even if they know it?
499 2013-04-11 18:12:07 <diki> s/know it?/know it.
500 2013-04-11 18:12:10 <jouke> diki: it is called bitcoin.
501 2013-04-11 18:12:18 <oiram> I'm having trouble finding an intelligent btc IRC conversation since all the scammers poluted the channels.  Is there a new channel for people who's IQs are larger than 5?
502 2013-04-11 18:13:02 <oiram> They are just trying to generate fear on the other channels. ...and with the bad english, it is really suspicious.
503 2013-04-11 18:13:17 <diki> jouke:huh?
504 2013-04-11 18:15:20 <helo> diki: private keys, by virtue of being data, can be copied/owned by multiple people.
505 2013-04-11 18:17:05 <helo> diki: a leger system like bitcoin with strong one-way transfer of value seems to do exactly what you describe
506 2013-04-11 18:19:49 <diki> What if a user generates a very hard vanity address, how would he sell it's private key, but also providing some ensurance he doesn't have a copy(and they always do :S)
507 2013-04-11 18:20:07 <sipa> diki: use EC math magic
508 2013-04-11 18:20:30 <wumpus> oiram: #bitcoin-tech
509 2013-04-11 18:20:42 <diki> sipa:If you are talking about adding/multiplying private keys, that works with something like a pool
510 2013-04-11 18:20:43 <sipa> diki: if i have a public key of yours, i can generate a number that has to be added to your *private* key in order to obtain an address with desired properties
511 2013-04-11 18:20:59 <diki> i.e you must provide a public keys
512 2013-04-11 18:21:03 <diki> *key
513 2013-04-11 18:21:33 <diki> But in my case let's say I've generated a very hard address by myself with vanitygen/oclvanitygen and I'd like to sell it, as it will take a while to produce.
514 2013-04-11 18:21:54 <diki> this is just an example, I personally don't plan on doing something like this.
515 2013-04-11 18:22:02 <jaakkos> it would be impossible to prove you don't have the private key.
516 2013-04-11 18:22:14 <jaakkos> because you do.
517 2013-04-11 18:22:19 <diki> exactly.
518 2013-04-11 18:22:39 <diki> One day people will want to roll with custom addresses.
519 2013-04-11 18:22:50 <diki> Like how you can buy a custom address for a car etc
520 2013-04-11 18:22:57 <jaakkos> but if they know what address they want, they can make other people compute it for them.
521 2013-04-11 18:23:12 <sipa> i hope one year from now, no human will need to see a bitcoin address at all anymore
522 2013-04-11 18:23:13 <ali1234> well, how about using namecoin to map fixed names to addresses?
523 2013-04-11 18:23:15 <pjorrit_> yea there's a distributed service for that isnt there?
524 2013-04-11 18:23:19 <jaakkos> yes
525 2013-04-11 18:23:19 <sipa> making vanity addresses obsolete
526 2013-04-11 18:23:24 <pjorrit_> in juts one year?
527 2013-04-11 18:23:28 <diki> sipa:Oh, having plans?
528 2013-04-11 18:23:33 <sipa> if the payment protocol gets adopted, sure
529 2013-04-11 18:23:39 <diki> bip?
530 2013-04-11 18:23:43 <sipa> you'll just pay to a URL
531 2013-04-11 18:23:58 <sipa> https://gist.github.com/gavinandresen/4120476
532 2013-04-11 18:26:17 <diki> and this would allow chargebacks?
533 2013-04-11 18:30:27 <sipa> no
534 2013-04-11 18:31:23 <jok> Luke-Jr, base58.h class CBase58Data, method ToString(). Yep!
535 2013-04-11 18:36:21 <BlueMatt> wtf? Winklevoss...
536 2013-04-11 18:36:35 <phantomcircuit> wat
537 2013-04-11 18:36:39 <BlueMatt> http://www.theverge.com/2013/4/11/4213956/whats-cooler-than-a-million-dollars-winklevoss-twins-own-1-percent-of
538 2013-04-11 18:36:49 <phantomcircuit> oh
539 2013-04-11 18:37:00 <diki> BlueMatt:Attack of the clones
540 2013-04-11 18:37:43 <phantomcircuit> lol they're about to get flooded with nonsense
541 2013-04-11 18:38:27 <jouke> lol
542 2013-04-11 18:39:30 <MC1984> oh so thats the bitcoin elite
543 2013-04-11 18:39:31 <phantomcircuit> 1064658 trades on mtgox upto Tue, 22 Nov 2011 02:09:35 GMT
544 2013-04-11 18:39:41 <phantomcircuit> this is going to take a while
545 2013-04-11 18:42:41 <jgarzik> Large, naive portions of this community will be shocked when the rich and powerful... buy into bitcoin and become the Bitcoin Rich And Powerful.
546 2013-04-11 18:43:14 <denisx> if they own 1% of all bitcoins that would be more than 10 million $ and not 1.3 million $
547 2013-04-11 18:43:19 <denisx> something fishy with the repotz
548 2013-04-11 18:43:22 <denisx> report
549 2013-04-11 18:43:48 <ThomasV> "The Winklevii ? as they are popularly known ? say they own nearly 1 percent of that, or some $11 million."
550 2013-04-11 18:44:02 <ThomasV> something fishy with your eyes :)
551 2013-04-11 18:44:44 <denisx> The brothers claim to have amassed one of the largest portfolios of Bitcoin in the world, worth around $1.3 million, or about 1 percent of the entire currency's dollar value equivalent
552 2013-04-11 18:44:47 <ThomasV> oh, sry, I wasn't reading the same link
553 2013-04-11 18:44:51 <ThomasV> http://dealbook.nytimes.com/2013/04/11/as-big-investors-emerge-bitcoin-gets-ready-for-its-close-up/
554 2013-04-11 18:46:31 <denisx> ThomasV: that report looks alot better
555 2013-04-11 18:50:55 <k9quaint> man, #litecoin is almost as crazy sauce as #solidcoin was
556 2013-04-11 18:51:27 <k9quaint> rabid amateurs flailing about in a swimming pool filled with maple syrup and pinecones
557 2013-04-11 18:52:06 <Scrat> it's like trolling in a sea of trolls
558 2013-04-11 18:52:25 <MaxValor> i like the swimming pool metaphor better
559 2013-04-11 18:52:26 <gonffen> lol
560 2013-04-11 18:52:33 <MaxValor> throw in some random barbed wire too
561 2013-04-11 18:53:08 <k9quaint> I was trying to have a conversation about the pros and cons of blockchain speed
562 2013-04-11 18:59:09 <_maniac_> Hello. I'm experiencing high swap usage for bitcoind. Ubuntu 12.04. both version 0.8.0 and 0.8.1.
563 2013-04-11 19:01:01 <sipa> _maniac_: feel free to try git head, if you can compile from source; it has much better memory usage
564 2013-04-11 19:01:03 <denisx> _maniac_: compile HEAD or just buy more ram
565 2013-04-11 19:01:13 <_maniac_> thx, will try.
566 2013-04-11 19:03:23 <jok> Luke-Jr, base58(format identifier + data + checksum) . so we don't need any additional information besides 19034a421bf9f6a14657f39c806944d3e197d1cf ( ....OP_HASH160 19034a421bf9f6a14657f39c806944d3e197d1cf OP_EQUALVERIFY...)
567 2013-04-11 19:03:36 <jok> ?
568 2013-04-11 19:04:23 <sipa> jok: indeed
569 2013-04-11 19:07:18 <jok> sipa, ok
570 2013-04-11 19:19:07 <jok> sipa, i am confused with these lines: "class CKeyID : public uint160",     CKeyID keyID = newKey.GetID();, CBitcoinAddress(keyID).ToString();
571 2013-04-11 19:19:45 <cyphase> any thoughts on the issue of changing the crypto in bitcoin? if there's some breakthrough that renders ecdsa easily crackable, even if support was added for something else, only newly transfered bitcoins would be protected by the new crypto
572 2013-04-11 19:20:00 <sipa> jok: the keyid is that 160-bit hex string
573 2013-04-11 19:20:21 <sipa> jok: CBitcoinAddress is the class for converting to string representation of an address
574 2013-04-11 19:20:33 <sipa> jok: addresses can also be send-to-script instead of send-to-pubkeyhash
575 2013-04-11 19:21:52 <jok> sipa, what do you mean in last sentence? is it right that we can took address from Script by callinf some f ?
576 2013-04-11 19:22:29 <sipa> jok: where do you see a script?
577 2013-04-11 19:23:06 <bitcoinxc> hello - I am making a website to replace the bitcoin faucet - It will give awqay 4-12 satoshi to all visitors for free - can I get some help testing the website?
578 2013-04-11 19:23:42 <sipa> bitcoinxc: seems very hard to do
579 2013-04-11 19:23:46 <graingert_> bitcoinxc: you bet
580 2013-04-11 19:23:51 <graingert_> bitcoinxc: link?
581 2013-04-11 19:23:58 <jok> class CScript:     std::string ToString()
582 2013-04-11 19:24:01 <sipa> bitcoinxc: spending such a coin would require more fees than what they receive...
583 2013-04-11 19:24:09 <sipa> jok: that isn't called
584 2013-04-11 19:25:03 <bitcoinxc> the website is www.bitcoinxc.com - it is early alpha
585 2013-04-11 19:25:18 <sipa> coins less than 0.0001 aren't useful, really
586 2013-04-11 19:26:04 <bitcoinxc> I think giving away a few satoshi at a time will encourage people to get wallets
587 2013-04-11 19:26:21 <bitcoinxc> and I may increase teh give aways in the future.
588 2013-04-11 19:26:56 <sipa> bitcoinxc: you don't understand
589 2013-04-11 19:27:06 <sipa> you can't spend these
590 2013-04-11 19:27:18 <bitcoinxc> I understand perfectly
591 2013-04-11 19:27:40 <sipa> then what is the point?
592 2013-04-11 19:27:59 <bitcoinxc> the value of a few satoshi is too low to buy anything - I will increase the giveaway later. I am in alpha - give me a break
593 2013-04-11 19:28:08 <sipa> oh, sorry
594 2013-04-11 19:28:21 <sipa> i misunderstood - i thought this was the intention :)
595 2013-04-11 19:28:58 <sipa> bitcoinxc: i'd suggest testing on testing with realistic amounts, though
596 2013-04-11 19:29:40 <bitcoinxc> the goal is to get people to think and act - think about bitcoins and get a wallet - then if they have done the first steps it is that much easier to fund their accounts later - I am just getting ppl started
597 2013-04-11 19:30:49 <sipa> bitcoinxc: sure, i'm not arguing against the usefulness of a faucet in general
598 2013-04-11 19:31:48 <bitcoinxc> so if you hit submit - did your page blank - did it reload the page or did it give you a confirmation that the bitcoin was sent?
599 2013-04-11 19:32:45 <bitcoinxc> what do you think of page design? I could use help on the CSS
600 2013-04-11 19:33:42 <bitcoinxc> can anyone give me feedback on my new bitcoin faucet?
601 2013-04-11 19:34:09 <lianj> yes, where is it?
602 2013-04-11 19:34:29 <bitcoinxc> http://www.bitcoinxc.com/
603 2013-04-11 19:34:56 <bitcoinxc> tell me if the page blanks, reloads, or gives an error?
604 2013-04-11 19:36:17 <_maniac_> I personally find that animated favicon annoying.
605 2013-04-11 19:36:31 <lianj> bitcoinxc: no errors but also no status message that it succeeded
606 2013-04-11 19:36:40 <MC1984> browser mining?
607 2013-04-11 19:36:45 <bitcoinxc> what browser are you using?
608 2013-04-11 19:36:53 <jok> bitcoinbulletin,  no blank pages or errors
609 2013-04-11 19:36:53 <paulbohm> Impromptu Opensource Bitcoin Exchange Software Design Hack-Day at my place in SF/Mission today. Feel free to join if you feel like it!
610 2013-04-11 19:37:19 <bitcoinxc> uh - browser mining will be a way to suport the site through java mining
611 2013-04-11 19:37:23 <bitcoinxc> its not done yet
612 2013-04-11 19:37:28 <gonffen> bitcoinxc: is it supposed to confirm you submitted an address?
613 2013-04-11 19:37:45 <bitcoinxc> depends on the browser
614 2013-04-11 19:38:15 <bitcoinxc> what browser are you using? I use chrome and internet explorer
615 2013-04-11 19:38:27 <MC1984> do you realise how non viable java mining is
616 2013-04-11 19:38:35 <_maniac_> firefox. no message whatsoever upon submitting address.
617 2013-04-11 19:38:50 <bitcoinxc> yes but ..... If a milliion people java mine for 5 minutes each...
618 2013-04-11 19:39:03 <MC1984> youll get 10 cents or something
619 2013-04-11 19:39:27 <bitcoinxc> and 10 cents will enable me to give out millioins of satoshi
620 2013-04-11 19:39:35 <bitcoinxc> thus = success
621 2013-04-11 19:40:14 <MC1984> youll never find a block on your own
622 2013-04-11 19:40:19 <bitcoinxc> whats the best forum for bitcoin development? I am using php.
623 2013-04-11 19:40:23 <bitcoinxc> I pool
624 2013-04-11 19:40:27 <MC1984> and im not sure youll eve reach the payout threshold of any pool either
625 2013-04-11 19:40:45 <gonffen> or ever a damn share
626 2013-04-11 19:40:46 <bitcoinxc> I think your missing the point of the website
627 2013-04-11 19:40:59 <lianj> bitcoinxc: 12 satoshi meaning 0.00000012 btc?
628 2013-04-11 19:41:00 <bitcoinxc> the goal is not to get rich on java mining.
629 2013-04-11 19:41:19 <bitcoinxc> whatever teh smallest unit is
630 2013-04-11 19:41:35 <bitcoinxc> I will increase payout after I have the site fully operational
631 2013-04-11 19:41:40 <bitcoinxc> I am looking for some help
632 2013-04-11 19:41:47 <lianj> bitcoinxc: please dont send me that shit, costs more to send than its worth
633 2013-04-11 19:41:49 <Luke-Jr> bitcoinxc: use testnet
634 2013-04-11 19:41:54 <lianj> that too
635 2013-04-11 19:42:29 <lianj> 15 minutes mining gives you more than 0.00000012
636 2013-04-11 19:42:38 <_maniac_> can't you spend one satoshi coin with some other, bigger transaction?
637 2013-04-11 19:42:40 <bitcoinxc> I am having a hard enough time using bitcoin...
638 2013-04-11 19:42:50 <MC1984> _maniac_ doesnt work like that
639 2013-04-11 19:43:05 <graingert_> _maniac_: no, but you can spend one satoshi with some other transaction
640 2013-04-11 19:43:17 <bitcoinxc> you can tip ppl on reddit with satoshi
641 2013-04-11 19:43:48 <_maniac_> graingert_: that's what I meant, yes.
642 2013-04-11 19:43:49 <graingert_> bitcoinxc: it's not worth the time to collect
643 2013-04-11 19:43:54 <MC1984> graingert_ dont think it works like that either?
644 2013-04-11 19:43:57 <bitcoinxc> where is teh best bitcoin dev forums?
645 2013-04-11 19:44:27 <graingert_> MC1984: yes it does, it still lowers your priority
646 2013-04-11 19:44:44 <_maniac_> I mean, transaction consists of inputs and outputs. if I'm already paying for something from bigger coin, can't I put one satoshi in it (just to get rid of it)
647 2013-04-11 19:45:06 <graingert_> _maniac_: yes, but it will increase your fee more than it's worth
648 2013-04-11 19:45:09 <graingert_> maybe
649 2013-04-11 19:45:11 <graingert_> I think?
650 2013-04-11 19:45:15 <graingert_> gmaxwell: :(
651 2013-04-11 19:45:16 <lianj> _maniac_: sure, you pay only for the final tx size
652 2013-04-11 19:46:34 <cononicalmonster> whats going on in here
653 2013-04-11 19:46:36 <MC1984> an input is an input though wheter 1 satishi or 1 coin
654 2013-04-11 19:46:51 <MC1984> still uneconomical, youd olny do it if you rally wanted to get rid
655 2013-04-11 19:46:55 <jspilman> I think it increases size, which will count against you, but size is cheap.  Ive read people who claimed they could add additional high priority coins into a transaction, and use the dust without paying a fee that way
656 2013-04-11 19:47:06 <jspilman> but you're paying either way
657 2013-04-11 19:47:21 <gmaxwell> jspilman: sure, if you're already qualifying as free and have extra priority you can sweep dust that way.
658 2013-04-11 19:47:29 <jspilman> paying with priority is exactly the same as paying a fee with BTC, economically
659 2013-04-11 19:47:31 <cononicalmonster> what are thoughts on what happens when gox goes back up
660 2013-04-11 19:47:44 <jspilman> thoughts are - off topic? :-)
661 2013-04-11 19:47:45 <gmaxwell> jspilman: the difference is that priority is not conserved and btc is.
662 2013-04-11 19:48:00 <MC1984> gmaxwell couldnt the client have logic to sweep dust like that automatically where possible, to keep the utxo down
663 2013-04-11 19:48:05 <gmaxwell> cononicalmonster: Thats totally offtopic here, go to the cesspool at #bitcoin-pricetalk for that.
664 2013-04-11 19:48:07 <bitcoinxc> well my goal is to get people using bitcoin - not give away millions of $$$ - if you have a better suggestion let me know - I can always increase teh payout
665 2013-04-11 19:48:08 <jspilman> right, the miner can't spend your priority... hmm... :-)
666 2013-04-11 19:48:11 <bitcoinxc> http://www.bitcoinxc.com/
667 2013-04-11 19:48:32 <cononicalmonster> so what are we discussing:? mining?
668 2013-04-11 19:48:45 <gmaxwell> MC1984: sure. But it's not clear how much that will help??? esp the people with lots of dust tend to not have much priority at all.. and increasingly blocks are full of fee driven txn.
669 2013-04-11 19:49:04 <bitcoinxc> I created a new bitcoin faucet and peopel are complaining I dont give enough away...
670 2013-04-11 19:49:16 <cononicalmonster> ic
671 2013-04-11 19:49:17 <jspilman> does 'priority' ultiamtely go away at some point?
672 2013-04-11 19:49:27 <gmaxwell> jspilman: yes, basically.
673 2013-04-11 19:49:33 <jspilman> bitcoinxc: sounds about right
674 2013-04-11 19:49:36 <bitcoinxc> what is a good amount to give away then?
675 2013-04-11 19:49:46 <jspilman> maybe call it bitcoin trickle
676 2013-04-11 19:49:53 <gmaxwell> jspilman: and then you have fee.. but fee is conserved, and so a utxo can truly be uneconomical.
677 2013-04-11 19:49:56 <Konnichiwa> bitcoin drip
678 2013-04-11 19:50:15 <bitcoinxc> bitcoin nano payouts
679 2013-04-11 19:50:21 <jspilman> are any miners doing the merkle of uxto?
680 2013-04-11 19:50:37 <gmaxwell> jspilman: I came up with a way of solving the uneconomical utxo issue, but its probably too complicated a change for bitcoin. :(  (it's also a hardfork)
681 2013-04-11 19:50:57 <gmaxwell> jspilman: no, no software has been written for it. it's just an idea now, and not an especially concrete one.
682 2013-04-11 19:51:50 <_maniac_> sorry, what is "utxo"?
683 2013-04-11 19:51:56 <jspilman> bitcoinxc: well, if the faucet produces literally unspendable output, then it's just broken - it's not like you can 'accumulate' dust and eventually get something that's spendable - you just have to wait it out until BTC is $10,000 / coin and the dust has meaningful value... which could take a while!
684 2013-04-11 19:52:04 <Konnichiwa> is MtGox using bitcoind RPC for its trading or what?
685 2013-04-11 19:52:27 <jspilman> best writeup I found was here: https://en.bitcoin.it/wiki/User:DiThi/MTUT
686 2013-04-11 19:52:47 <jgarzik> Konnichiwa: MtGox has its own bitcoin platform software and trading engine.  Try #mtgox for specific questions
687 2013-04-11 19:52:52 <gmaxwell> (The way you solve the uneconomical unspent transaction output problem is that you make every output created reduce the permitted blocksize of the block creating it by the size required to redeem it.. and when it gets redeemed you increase the permitted blocksize by that amount.)
688 2013-04-11 19:52:53 <jspilman> are people still convinced it would be useful?
689 2013-04-11 19:53:02 <bitcoinxc> do I need to run bitcoind in shell on my server or can we do php api calls without it? I am on a shared server...