1 2011-06-26 00:00:50 <pakaran> am I the only one who sees the testnet block explorer as showing a very old block number?
  2 2011-06-26 00:03:43 <pakaran> BlueMatt, so those are alternate roots, like testnet?
  3 2011-06-26 00:34:11 <cam1235> Can someone please help me? :D
  4 2011-06-26 00:34:29 <cam1235> anyone with knowledge of bitcoind
  5 2011-06-26 00:35:15 <unclemantis> any iphone apps yet? either itunes official or hacked?
  6 2011-06-26 00:35:54 <unclemantis> cam1235 check out the bitcoin wiki
  7 2011-06-26 00:36:00 <unclemantis> seriously good eats in there
  8 2011-06-26 00:37:58 <cam1235> Tried...i installed it via ruby. Im using shared hosting, aka no ssh.
  9 2011-06-26 00:38:28 <BlueMatt> cam1235: ask a question dont ask to ask
 10 2011-06-26 00:40:24 <cam1235> Still, can someone please help me out? o.o
 11 2011-06-26 00:42:25 <unclemantis> is he gone?
 12 2011-06-26 00:42:55 <unclemantis> did cam1235 even try to pick up the fork and feed himself before he asked for help? Seriously?
 13 2011-06-26 01:00:16 <xtalmath> I have a strange non bitcoin related question, but really have no clue where else to ask: suppose I am a normal user on a normal ISP (kind of the case), suppose I am getting DDoSed, enough to chocke my network card, but much lower than chocking the ISP higher up in the routing chain. Will this data flowing to my computer be counted as download volume?
 14 2011-06-26 01:04:13 <xtalmath> or only if I "requested" data? how could an ISP determine this? if there is some metadata from which they assume I chose to download, then one could write software that pretends someone else is sending the data. I believe however that they do not analyse the stream to determine if it was wanted or received data, so DDoS would count as download costly
 15 2011-06-26 01:05:38 <upb> sure its counted
 16 2011-06-26 01:05:50 <gmaxwell> The former, and because they can't tell and even if they could it wouldn't be in their interest to do so
 17 2011-06-26 01:06:23 <xtalmath> suppose the fed tries the following attack: "DDoS" anyone who is visible on network, or send them enough data so that the user starts noticing on his ISP account page that his downloaded volumes are getting annoyingly high.
 18 2011-06-26 01:06:26 <upb> they would need to effectively do conntrack on all their customers otherwise :D
 19 2011-06-26 01:06:29 <upb> madness
 20 2011-06-26 01:07:21 <gmaxwell> Depending how how they count you could even be billed for more than the capacity of your link (though that would be easy to get out of)
 21 2011-06-26 01:07:21 <xtalmath> ISPs if customers complain ISPs would get annoyed and complain back
 22 2011-06-26 01:08:52 <xtalmath> also socially, a miner or bitcoin enthousiast may be sharing the internet connection with the rest of his family,... they wouldnt like it when they cant check their mails or surf comfortably
 23 2011-06-26 01:10:16 <xtalmath> then again, I assume ISPs have DDoS filters in place to shut down the connection (i.e. lots of messages for that IP) => trigger happy ddos filter shuts of your connection.
 24 2011-06-26 01:12:13 <xtalmath> you would want to be able to tell your ISP, to ignore incoming connections that you dont approve of, i.e. their could be a protocol with which your computer could communicate with the ISP that you only want traffic from IP you allow (those you see on bitcoin network, DNS servers, the IPs you got from your DNS server,...)
 25 2011-06-26 01:14:09 <upb> it really doesnt help the isp to 'shut down' a connection
 26 2011-06-26 01:14:14 <upb> they'd have to nullroute it
 27 2011-06-26 01:14:26 <xtalmath> no but it spares them angry customers
 28 2011-06-26 01:14:33 <upb> i seriously doublt they'd do that for a residential customers /32 or whatever
 29 2011-06-26 01:14:39 <xtalmath> hmm
 30 2011-06-26 01:15:13 <xtalmath> then again I dont know that much of internet protocol.
 31 2011-06-26 01:16:23 <xtalmath> perhaps normal data send transactions (not ping etc) need to be accepted by the receiving end?
 32 2011-06-26 01:16:44 <xtalmath> i mean, perhaps this is already in place?
 33 2011-06-26 01:18:45 <upb> if the dos is just to fill your pipe up with udp they dont have to be accepted
 34 2011-06-26 01:18:46 <xtalmath> like perhaps whenever we hear of DDoSed sites, its typically computers requesting a web page, and then its hard to discern if its a user or an attacker. while a computer that requests information could in theory know if it was requested.
 35 2011-06-26 01:19:38 <xtalmath> upb: so udp dont have to be accepted but TCP for example does?
 36 2011-06-26 01:19:43 <upb> yep
 37 2011-06-26 01:20:07 <xtalmath> so in theory you could ask your ISP to only allow TCP on your link?
 38 2011-06-26 01:20:18 <upb> if theyre trying to keep your webserver busy with half open or accepted connections
 39 2011-06-26 01:20:19 <jrmithdobbs> you'd still pay for that ddos bandwidth though
 40 2011-06-26 01:20:23 <jrmithdobbs> just fyi
 41 2011-06-26 01:20:37 <upb> heh they wont do anything like that xtalmath :)
 42 2011-06-26 01:20:51 <jrmithdobbs> and ya, if you do that there's other effective methods anyways
 43 2011-06-26 01:20:53 <upb> the best they would do is nullroute your ip at their upstreams
 44 2011-06-26 01:20:55 <jrmithdobbs> upb: actually, they will.
 45 2011-06-26 01:21:02 <jrmithdobbs> upb: not true any more
 46 2011-06-26 01:21:07 <upb> hmm
 47 2011-06-26 01:21:12 <xtalmath> most bitcoin enthousiasts are either liars or BTC poor, im not sure how many are willing to sacrifice their bandwith for bitcoin survival
 48 2011-06-26 01:21:25 <upb> for a home user ?:P
 49 2011-06-26 01:22:10 <jrmithdobbs> upb: home user no, but a colo/hosting facilities have some awesome capabilities these days
 50 2011-06-26 01:22:26 <upb> sure but were talking about home dsl/cable here
 51 2011-06-26 01:22:29 <jrmithdobbs> upb: like traffic pattern detection with direct hooks into their upstream to add selective null routes
 52 2011-06-26 01:22:32 <jrmithdobbs> oic
 53 2011-06-26 01:22:44 <jrmithdobbs> missed that
 54 2011-06-26 01:22:56 <upb> well thats what i assumed from xtalmath's question anyway :)
 55 2011-06-26 01:23:47 <xtalmath> well both home users and stronger connections are interesting questions, how big or small would the network be if they could seriously discourage home users?
 56 2011-06-26 01:24:17 <xtalmath> also note how running bitcoin is then no longer a personal stance but family wise or appartment wise etc, since your neighbours could start complaining
 57 2011-06-26 01:24:54 <xtalmath> the few dedicated home users and the dedicated strong connections, would easily be identified in cooperation with the ISPs
 58 2011-06-26 01:25:07 <jrmithdobbs> um
 59 2011-06-26 01:25:12 <xtalmath> hmm I guess TOR would solve the problem then
 60 2011-06-26 01:25:24 <jrmithdobbs> what was the original frame of your question (and the question itself)
 61 2011-06-26 01:25:31 <jrmithdobbs> because i don't think you're talking about what i thought you were at all
 62 2011-06-26 01:25:35 <dehuman> there are a couple of practical things to mitigate ddos depending upon the nature
 63 2011-06-26 01:25:41 <xtalmath> ah, just exploring the validity of an attack
 64 2011-06-26 01:25:43 <dehuman> there are services which proxy, you just point your dns at them
 65 2011-06-26 01:26:16 <dehuman> if attacker is focusing on a single ip its easy enough to change dns so users and either a backup server on another ip or migrate to new ip temporarily
 66 2011-06-26 01:26:51 <jrmithdobbs> dehuman: depends on scale and nature of attack
 67 2011-06-26 01:26:55 <dehuman> also you can throttle connections too
 68 2011-06-26 01:27:01 <dehuman> depending on which lever ddos occurs
 69 2011-06-26 01:27:01 <upb> 26 06:00 < xtalmath> I have a strange non bitcoin related question, but really have no clue where else to ask: suppose I am a normal user on a normal ISP (kind of the
 70 2011-06-26 01:27:04 <upb> case), suppose I am getting DDoSed, enough to chocke my network card, but much lower than chocking the ISP higher up in the routing chain. Will this
 71 2011-06-26 01:27:07 <jrmithdobbs> dehuman: that's actually almost always ineffective against competent attackers
 72 2011-06-26 01:27:07 <upb> data flowing to my computer be counted as download volume?
 73 2011-06-26 01:27:10 <upb> this was the original q :P
 74 2011-06-26 01:27:12 <dehuman> i think there are apache mods to mitigate ddos arent they
 75 2011-06-26 01:27:17 <dehuman> jrmithdobbs: which?
 76 2011-06-26 01:27:21 <jrmithdobbs> dehuman: both
 77 2011-06-26 01:27:55 <xtalmath> upb: thanks , was trying to rephrase, but I like the ascii art above :D
 78 2011-06-26 01:28:00 <jrmithdobbs> dehuman: they don't even really have to be competent, it's a matter of the size of their dronenet
 79 2011-06-26 01:28:04 <upb> haha
 80 2011-06-26 01:28:19 <jrmithdobbs> dehuman: 99% of ddoses are not very sophisticated
 81 2011-06-26 01:28:41 <dehuman> jrmithdobbs: but a proper dns based ddos mitigation service where it utilizes a large number of proxies seems effective enough
 82 2011-06-26 01:28:43 <phantomcircuit> jrmithdobbs, LOIC
 83 2011-06-26 01:28:48 <upb> well my friend is a cto at an isp and what they do in case of a ddos is they nullroute the ip and make the customer pay for the traffic
 84 2011-06-26 01:29:00 <upb> and if they dont want to pay when the ddos continues, they terminate him
 85 2011-06-26 01:29:05 <phantomcircuit> dehuman, proper ddos mitigation is via anycast but anyways...
 86 2011-06-26 01:29:07 <jrmithdobbs> dehuman: the only req is that they fill your upstream. so if they're able to fill *your* upstream's upstream, which is commonly not as hard as you'd think, it doesn't matter what you do
 87 2011-06-26 01:29:46 <dehuman> jrsmithdob: but they have to fill each point, if you have 100 ways in, and no one ever sees tha tway in, you have to fill up 100 different pipes
 88 2011-06-26 01:29:59 <jrmithdobbs> phantomcircuit: the only thing i've actually seen work *well* in the wild are traffic analysis + null route inserts far enough upstream (at level3/cogent/xo/etc) to prevent the attacker from being able to fill the pipe
 89 2011-06-26 01:30:09 <dehuman> and they can proxy your content
 90 2011-06-26 01:30:22 <dehuman> anyway there are ddos mitigation services and they can handle pretty sophisticated attacks
 91 2011-06-26 01:30:30 <xtalmath> so how workable would this attack be? generate enough traffic to bitcoin running households to piss them off by using up their financially payed for bandwith, both volume and speed as this happens? how resistant will bitcoin fan be against his family members complaining?
 92 2011-06-26 01:30:32 <jrmithdobbs> dehuman: you're missing the obvious flaw here
 93 2011-06-26 01:30:33 <dehuman> they obviously arent free and i'm no expert in ddos mitigation
 94 2011-06-26 01:30:41 <enquirer> bitcoin.exe is needlessly thrashing the disk - reads 32GB, writes 19 GB over 24 hours
 95 2011-06-26 01:30:48 <jrmithdobbs> dehuman: they may have multiple endpoints to come in through, but it's still hitting your backend somewhere
 96 2011-06-26 01:30:59 <jrmithdobbs> dehuman: they just have to fill that backend pipe, not the fronted of all the proxy endpoints
 97 2011-06-26 01:31:04 <dehuman> jrmithdobbs: not if its cached on the proxy
 98 2011-06-26 01:31:13 <jrmithdobbs> you can't cache the bitcoin protocol, for instance
 99 2011-06-26 01:31:19 <jrmithdobbs> which is what we're talking about
100 2011-06-26 01:31:22 <phantomcircuit> jrmithdobbs, yup, set the dns records to an anycast for the ddos service, which samples traffic and null routes upstream for bad peers
101 2011-06-26 01:31:30 <gmaxwell> enquirer: bdb uses a lot of fsync in order to not get corrupted by surprise shutdown.
102 2011-06-26 01:31:35 <lfm> enquirer: why do you say it is needless?
103 2011-06-26 01:31:47 <jrmithdobbs> dehuman: it's a very hard problem to solve well
104 2011-06-26 01:32:55 <gmaxwell> Lots of providers use arbor (or its varrious competition) to take sampled feeds of traffic from all over the network and then detect attacks. (and allows them to inject null routes that are tagged to kill the traffic _source_ via urpf abuse)
105 2011-06-26 01:34:23 <xtalmath> gmaxwell: about the fsync, would a minireimplementation of journalled filesystem work? like journalled wallet, with atomic operations?
106 2011-06-26 01:35:12 <lfm> xtalmath: "atomic" operation imply syncs don they?
107 2011-06-26 01:35:17 <enquirer> the db is only 300 MB ... so what if it is corrupted during power down?
108 2011-06-26 01:35:39 <xtalmath> gmaxwell, ive read your ECDSA deterministic wallet, i know how ecdsa works, but how safe do you feel this is in the case an attacker finds out multiple of your public keys?
109 2011-06-26 01:36:08 <xtalmath> right, of course :(
110 2011-06-26 01:36:23 <jrmithdobbs> gmaxwell: ya i worked somewhere that had one of the first netzentry deployments
111 2011-06-26 01:36:28 <jrmithdobbs> gmaxwell: the shit is fuckin cool
112 2011-06-26 01:36:42 <lfm> xtalmath: I dont think it matters at all if they see many public keys
113 2011-06-26 01:36:45 <xtalmath> what is arbor?
114 2011-06-26 01:37:08 <enquirer> its custom made db?? why not use sqlite
115 2011-06-26 01:37:09 <jrmithdobbs> xtalmath: system like i was mentioned, does traffic pattern analysis is capable of inserting nullroutes upstream
116 2011-06-26 01:37:15 <jrmithdobbs> s/was //
117 2011-06-26 01:37:28 <lfm> enquirer: sq is lots of overhead
118 2011-06-26 01:37:28 <xtalmath> enquirer its berkely db i think
119 2011-06-26 01:37:55 <pakaran> can someone send me testnet bitcoins?  my address is mqGaAX1dCC7VYzNG5h1uh9MQftjCU93zQU
120 2011-06-26 01:38:05 <pakaran> and the fountain is working, but it keeps showing no confirmations
121 2011-06-26 01:38:09 <pakaran> I think since nobody's mining
122 2011-06-26 01:38:11 <lfm> enquirer: bdb is well tested for many many years
123 2011-06-26 01:38:14 <jrmithdobbs> xtalmath: do you mean public keys or btc addresses, because the answer is very different depending, btw re: question to gmaxwell
124 2011-06-26 01:38:17 <enquirer> couldn't be less efficient ... with 19GB writes and 1 hour of CPU time over 24 hours, doing almost nothing
125 2011-06-26 01:38:35 <jrmithdobbs> xtalmath: the btc addresses do nothing to increase odds
126 2011-06-26 01:38:43 <lfm> enquirer: well of course sqlite could be less efficient
127 2011-06-26 01:38:51 <xtalmath> i mean public keys: the user doesnt only receive on the deterministic wallet but spends
128 2011-06-26 01:39:08 <xtalmath> so that reveals the past public keys
129 2011-06-26 01:39:26 <lfm> enquirer: ok if you dont run bitcoin at all then youd save a lot of processing
130 2011-06-26 01:39:36 <jrmithdobbs> xtalmath: and i think you'd need an extremely large number of public keys to increase odds, and I'm not sure it even would in that case, i'll let him speak to that. he's much better at math ;p
131 2011-06-26 01:39:45 <xtalmath> also my question relates to type2
132 2011-06-26 01:40:25 <lfm> other keys wouldnt help at all
133 2011-06-26 01:42:00 <jrmithdobbs> "at all" is a pretty strong statement
134 2011-06-26 01:42:05 <lfm> yup
135 2011-06-26 01:42:07 <phantomcircuit> hmm
136 2011-06-26 01:42:14 <phantomcircuit> i think the network is highly segmented
137 2011-06-26 01:42:30 <xtalmath> also the notation on the forum is pretty wordlike, it does mean that the private keys (the scalars, not the points) are linear right? (a+n*b) with a,b the 2 secret scalars which form the private keys as a function of n?
138 2011-06-26 01:42:35 <jrmithdobbs> phantomcircuit: why?
139 2011-06-26 01:42:47 <phantomcircuit> transactions are being broadcast by britcoin's bitcoind which never reach anybody
140 2011-06-26 01:42:57 <jrmithdobbs> ruh roh
141 2011-06-26 01:43:16 <phantomcircuit> i just tested it and my client with 8 connections isn't getting the transaction
142 2011-06-26 01:43:22 <phantomcircuit> but i can see bitcoind sent it
143 2011-06-26 01:43:28 <xtalmath> has this wallet been develloped/implemented in some patch?
144 2011-06-26 01:43:31 <phantomcircuit> and is connected to like 80 peers
145 2011-06-26 01:43:42 <jrmithdobbs> phantomcircuit: /msg
146 2011-06-26 01:43:47 <lfm> phatoso you're saying the net is fragmented?
147 2011-06-26 01:44:49 <lfm> phantomcircuit: britcoin isnt attached to the bitcoin net?
148 2011-06-26 01:45:08 <phantomcircuit> wait nvm
149 2011-06-26 01:45:12 <xtalmath> oof
150 2011-06-26 01:45:15 <phantomcircuit> my bitcoin was off when it was sent
151 2011-06-26 01:45:19 <phantomcircuit> and it isn't in a block yet
152 2011-06-26 01:45:23 <phantomcircuit> so of course i dont see it
153 2011-06-26 01:45:49 <phantomcircuit> lfm, it's very well connected actually
154 2011-06-26 01:47:01 <lfm> ok
155 2011-06-26 01:48:06 <gmaxwell> xtalmath: As far as I can tell the only risk is that they find out the rest of your keys if they also find some crazy relationship with the hash function that allows them to recover the secret.
156 2011-06-26 01:48:20 <xtalmath> the ECDSA based wallet is really nice and elegant in concept, so I trust the hashbased deterministic wallet much more, security through obscurity works if only you know that you are using the algorithm, damn you for publishing it
157 2011-06-26 01:48:35 <gmaxwell> xtalmath: here is how I'm convinced that it is no danger to your private key:
158 2011-06-26 01:48:47 <xtalmath> gmaxwell: was type2 implemented? I think I can find a flaw in it...
159 2011-06-26 01:48:59 <gmaxwell> xtalmath: I've implemented it outside of bitcoin.
160 2011-06-26 01:49:22 <jrmithdobbs> phantomcircuit: actually you should still get it on inv broadcast from your peers
161 2011-06-26 01:49:27 <lfm> they use /dev/random or /dev/urandom or some such right?
162 2011-06-26 01:49:33 <xtalmath> gmaxwell: ill think about it, maybe in a few days ill challenge you to give me sequence of public addresses generated by it
163 2011-06-26 01:49:35 <jrmithdobbs> phantomcircuit: since it should be in their inv and you haven't seen it before
164 2011-06-26 01:50:02 <gmaxwell> xtalmath: K.
165 2011-06-26 01:50:04 <phantomcircuit> possibly all my peers are 0.3.23 and blocking it from being relayed
166 2011-06-26 01:50:10 <enquirer> hmm its also flushing the wallet on every db operation
167 2011-06-26 01:50:21 <lfm> enquirer: as it should
168 2011-06-26 01:50:37 <xtalmath> and just for the fun of it you could put some symbolic BTC on the accounts
169 2011-06-26 01:51:25 <phantomcircuit> probably going to have to wait for an eligius block or something
170 2011-06-26 01:51:31 <xtalmath> the best test to see if you really trust it yourself or designed a backdoor is to have you put a lot in it though ;)
171 2011-06-26 01:51:43 <phantomcircuit> ah there we go
172 2011-06-26 01:51:46 <phantomcircuit> 1/unconfirmed
173 2011-06-26 01:52:09 <enquirer> but wallet only changes when you generate new keys, isn't it?
174 2011-06-26 01:52:13 <gmaxwell> xtalmath: it's pretty easy to see that it doesn't endanger the private key. E.g. if it did, I could take one of _your_ public keys, and generate a bunch more public keys related to it, then perform whatever attack to get your private key.
175 2011-06-26 01:52:34 <gmaxwell> enquirer: no, it also changes when you send coins, since that generates new hidden change addresses.
176 2011-06-26 01:52:46 <lfm> enquirer: wallet balances are updated for every relevant txn
177 2011-06-26 01:52:58 <phantomcircuit> and what lfm said
178 2011-06-26 01:53:05 <jrmithdobbs> and addr.dat is updated every like 5ms it seems like
179 2011-06-26 01:53:13 <jrmithdobbs> ugh
180 2011-06-26 01:53:15 <gmaxwell> I assume everyone has seen this: https://forum.bitcoin.org/index.php?topic=22434
181 2011-06-26 01:53:18 <gmaxwell> Topic: [ANN] Bitcoin "No Forced TX Fee" mainline client fork
182 2011-06-26 01:53:30 <jrmithdobbs> seriously
183 2011-06-26 01:55:22 <jrmithdobbs> gmaxwell: why don't the idjits just stop storing like .01btc per address
184 2011-06-26 01:55:23 <jrmithdobbs> ugh
185 2011-06-26 01:55:39 <xtalmath> erm, does his fork start from scratch or just random money distribution of the current blockchain?
186 2011-06-26 01:55:59 <enquirer> but it should update wallet.dat only for those tx that reference one of keys in the wallet
187 2011-06-26 01:56:02 <jrmithdobbs> xtalmath: he just (badly) removes the fee req logic
188 2011-06-26 01:56:04 <jrmithdobbs> very badly
189 2011-06-26 01:56:07 <gmaxwell> it's just a modified client that doesn't prevent you from footgunning yourself.
190 2011-06-26 01:56:29 <jrmithdobbs> even though he'll send out txns that noone else will forward
191 2011-06-26 01:56:44 <jrmithdobbs> and you'll have to modify your wallet to fix it afterwards to send with official client
192 2011-06-26 01:56:47 <jrmithdobbs> lol
193 2011-06-26 01:56:58 <xtalmath> well if he takes the speculators and idjits with him thats all right
194 2011-06-26 01:57:24 <gmaxwell> The speculators all stay in the exchanges.
195 2011-06-26 01:57:24 <phantomcircuit> i think ill get a EAH6950
196 2011-06-26 01:57:31 <phantomcircuit> so i can run 8 monitors
197 2011-06-26 01:57:34 <xtalmath> I might download his sources just to see how to fork :P
198 2011-06-26 01:57:51 <jrmithdobbs> xtalmath: if the exchanges can't see your transactions it doesn't much matter
199 2011-06-26 01:57:56 <enquirer> gmaxwell: that's the same, generates new hidden change addresses = generate new key
200 2011-06-26 01:57:59 <lfm> ya speculators never actually use the block chain or the bitcoin net, they just use an exchange account
201 2011-06-26 01:58:12 <phantomcircuit> jrmithdobbs, even the most retarded transactions will eventually find their way to a miner who will include them
202 2011-06-26 01:58:15 <gmaxwell> enquirer: as lfm pointed out it also updates to keep your balance.
203 2011-06-26 01:58:20 <phantomcircuit> (especially if you know what you're doing ;)
204 2011-06-26 01:58:30 <gmaxwell> enquirer: but most people are surprised by the "new hidden change addresses = generate new key" because it's hidden.
205 2011-06-26 01:58:30 <jrmithdobbs> phantomcircuit: at this point the wait period is pretty huge, though
206 2011-06-26 01:58:55 <phantomcircuit> jrmithdobbs, heh yeah
207 2011-06-26 01:59:14 <xtalmath> lfm: if the majority of bitcoin holders are idiots, they start speculating with his coins (apparently bigger value) until things go wrong
208 2011-06-26 01:59:35 <gmaxwell> phantomcircuit: right, but it's not hard to create a txn that will not work well. Especially by violating the minimum output with no fee at all.
209 2011-06-26 02:00:00 <lfm> well Im not sure what percentage of bitocin users are ape idiots
210 2011-06-26 02:00:01 <phantomcircuit> gmaxwell, any transaction a miner includes will work
211 2011-06-26 02:00:10 <phantomcircuit> so long as the script is actually valid
212 2011-06-26 02:00:21 <phantomcircuit> it might just take a very long time
213 2011-06-26 02:01:16 <lfm> gmaxwell: I thot some miners like art forzz still accepted all txn even free ones
214 2011-06-26 02:02:00 <gmaxwell> Perhaps so, of course you actually have to get the transaction to him.
215 2011-06-26 02:02:16 <lfm> ya ok
216 2011-06-26 02:02:29 <gmaxwell> It's hard to get a feel for how long dust transaction take, since they often don't get forwarded at all you won't see them in the memory pool even if you take them.
217 2011-06-26 02:02:31 <phantomcircuit> eligius has a fixed linear fee schedule
218 2011-06-26 02:02:39 <enquirer> gmaxwell: almost all tx are not ours and don't affect wallet at all, so no need to flush the wallet
219 2011-06-26 02:03:28 <gmaxwell> enquirer: are you saying the wallet itself got $gigs of writes?
220 2011-06-26 02:03:31 <phantomcircuit> 0.00004096 BTC per 512 bytes
221 2011-06-26 02:03:54 <gmaxwell> phantomcircuit: yep, I pointed out eligius in my response there.
222 2011-06-26 02:03:54 <phantomcircuit> although i assume he has a floor
223 2011-06-26 02:04:05 <gmaxwell> 512 bytes is the floor.
224 2011-06-26 02:04:09 <phantomcircuit> since i've seen free transactions from eligius
225 2011-06-26 02:04:19 <phantomcircuit> or i think i have
226 2011-06-26 02:04:29 <lfm> phantomcircuit: not recent
227 2011-06-26 02:04:36 <gmaxwell> Only luke's crazy transactions.
228 2011-06-26 02:04:53 <gmaxwell> (you can make a txn which has a payment to eligius in it, and it'll mine that with no 'fee')
229 2011-06-26 02:05:21 <phantomcircuit> lol
230 2011-06-26 02:05:31 <phantomcircuit> so a fee that's only redeemable by eligius
231 2011-06-26 02:05:32 <phantomcircuit> nice
232 2011-06-26 02:05:47 <gmaxwell> And there is a patch to bitcoin which adds an OP_NOP and an eligius fee output, and always peers with a node that relays to eligius.
233 2011-06-26 02:06:06 <gmaxwell> The non-standard script discourages anyone else from mining it.
234 2011-06-26 02:06:14 <lfm> well if you have something to pay him for thats reasonable but itd be silly to do that instead of a fee
235 2011-06-26 02:07:39 <phantomcircuit> http://www.newegg.com/Product/Product.aspx?Item=N82E16824176177
236 2011-06-26 02:07:40 <phantomcircuit> http://www.newegg.com/Product/Product.aspx?Item=N82E16814121437
237 2011-06-26 02:07:42 <phantomcircuit> zomg want
238 2011-06-26 02:07:44 <phantomcircuit> x6
239 2011-06-26 02:07:54 <phantomcircuit> anybody have 10k i can uh borrow?
240 2011-06-26 02:07:56 <enquirer> gmaxwell: let's see wallet is 104kb, it updates every 30 seconds on average, with current (very low) volume of transactions, so its about 300mb of writes
241 2011-06-26 02:08:47 <gmaxwell> phantomcircuit: hm. That display looks nice!
242 2011-06-26 02:08:51 <enquirer> gmaxwell: so it's not the main source of writes, just most obvious
243 2011-06-26 02:08:59 <phantomcircuit> gmaxwell, yes now i just need 6 of them
244 2011-06-26 02:09:00 <phantomcircuit> hehe
245 2011-06-26 02:09:08 <jrmithdobbs> enquirer: you sure your numbers are coming from the wallet.dat and not addr.dat?
246 2011-06-26 02:09:14 <xtalmath> the same thread says BitCoin will soon be written in Qt, is this true? if so when? is this one guy trying to port it, or a shared goal by most developers?
247 2011-06-26 02:09:22 <jrmithdobbs> xtalmath: very soon
248 2011-06-26 02:09:35 <jrmithdobbs> xtalmath: code's already functional
249 2011-06-26 02:09:45 <phantomcircuit> i wonder if you could put 2 6950's in a single box and run 12 displays
250 2011-06-26 02:09:46 <xtalmath> how soon is that? will it be 0.3.24?
251 2011-06-26 02:09:59 <phantomcircuit> heh qt deserves a 0.4.0
252 2011-06-26 02:10:06 <jrmithdobbs> xtalmath: .4.0 or not long after
253 2011-06-26 02:10:10 <CIA-103> bitcoin: wyze * reaa8d67380e4 mining-proxy/ (8 files in 6 dirs): New pool status statistics, highlighting in various place http://tinyurl.com/62gxzda
254 2011-06-26 02:10:12 <CIA-103> bitcoin: Chris Howie * r7aec17fa9823 mining-proxy/ (8 files in 6 dirs): Merge pull request #22 from wyze/master http://tinyurl.com/697rf2v
255 2011-06-26 02:10:26 <jrmithdobbs> xtalmath: is my understanding
256 2011-06-26 02:10:35 <xtalmath> it will jump from 0.2.23 to 0.4.0?
257 2011-06-26 02:11:02 <jrmithdobbs> we're on .3.23
258 2011-06-26 02:11:04 <xtalmath> how soon is that in weeks?
259 2011-06-26 02:11:05 <phantomcircuit> select deposit_accounts.account_id,currencies.currency_abbreviation from deposit_accounts,currencies where deposit_accounts.currency_id=currencies.currency_id;
260 2011-06-26 02:11:07 <phantomcircuit> oops
261 2011-06-26 02:11:11 <jrmithdobbs> xtalmath: no idea yet
262 2011-06-26 02:11:11 <phantomcircuit> plz2ignore
263 2011-06-26 02:11:13 <xtalmath> i meant 3.23
264 2011-06-26 02:11:23 <jrmithdobbs> xtalmath: "soon"
265 2011-06-26 02:11:29 <jrmithdobbs> xtalmath: i'd guess a month or so
266 2011-06-26 02:11:37 <xtalmath> ok
267 2011-06-26 02:12:06 <lfm> so itl prolly be 12 weeks
268 2011-06-26 02:13:21 <nhodges> when i'm working with the mbc sci api, do i need to use gpg? i seem to be generating a blank paypage url each time
269 2011-06-26 02:13:32 <nhodges> like the function to encrypt the string is breaking
270 2011-06-26 02:13:35 <jrmithdobbs> the what
271 2011-06-26 02:13:40 <nhodges> mybitcoin sci api
272 2011-06-26 02:13:44 <phantomcircuit> jrmithdobbs, god i realized i haven't been checking user_id at all
273 2011-06-26 02:13:46 <upb> phantomcircuit: you dont use ansi joins ?:P
274 2011-06-26 02:13:48 <phantomcircuit> facepalm
275 2011-06-26 02:14:03 <jrmithdobbs> phantomcircuit: fix that shit
276 2011-06-26 02:14:08 <jrmithdobbs> ;p
277 2011-06-26 02:14:11 <phantomcircuit> upb, why bother im already using postgresql RETURNING anyways
278 2011-06-26 02:14:17 <upb> haha
279 2011-06-26 02:14:33 <phantomcircuit> seriously
280 2011-06-26 02:14:40 <phantomcircuit> INSERT ... RETURNING is fucking brilliant
281 2011-06-26 02:14:52 <jrmithdobbs> SQL92 4 LYFE
282 2011-06-26 02:14:56 <gmaxwell> Yea, some of the postgresisms are really nice.
283 2011-06-26 02:14:56 <phantomcircuit> with mysql you have to do separate transactions
284 2011-06-26 02:15:16 <phantomcircuit> which is not only much slower, but could result in an inconsistent db state
285 2011-06-26 02:15:30 <jrmithdobbs> just use oracle and be done with it
286 2011-06-26 02:15:41 <phantomcircuit> i refuse
287 2011-06-26 02:15:43 <phantomcircuit> XD
288 2011-06-26 02:15:56 <jrmithdobbs> oh ya
289 2011-06-26 02:15:57 <phantomcircuit> jrmithdobbs, also im obviously not talking about the live exchange
290 2011-06-26 02:16:03 <jrmithdobbs> wait until you have to scale to real numbers ;p
291 2011-06-26 02:16:19 <upb> i didnt actually mean for portability but becaue its more readable :)
292 2011-06-26 02:16:20 <jrmithdobbs> phantomcircuit: ya i know what you've been working on
293 2011-06-26 02:16:35 <phantomcircuit> if this has to scale past 1k tps im just going to write a realtime exchange in c++
294 2011-06-26 02:17:00 <jrmithdobbs> you'll still need a datastore with functioning HA
295 2011-06-26 02:17:16 <jrmithdobbs> which pretty much means rac whether you like it or not
296 2011-06-26 02:17:19 <phantomcircuit> yeah but that would be nothing but a journal
297 2011-06-26 02:17:32 <jrmithdobbs> if you want something tested/vetted, anyways
298 2011-06-26 02:17:49 <jrmithdobbs> and by that point your shareholders will demand such ;p
299 2011-06-26 02:18:13 <phantomcircuit> lol
300 2011-06-26 02:18:35 <phantomcircuit> <-- going to make sure company is structured that idgaf what they think
301 2011-06-26 02:18:36 <phantomcircuit> xD
302 2011-06-26 02:19:09 <phantomcircuit> jrmithdobbs, anyways it turns out forex type stuff is actually significantly harder
303 2011-06-26 02:19:32 <phantomcircuit> the contention is such that you basically cant scale horizontally
304 2011-06-26 02:19:49 <phantomcircuit> GBP/BTC pretty much has to run on a single system
305 2011-06-26 02:20:19 <phantomcircuit> you throw other systems into the mix and you're just wasting time with network latency
306 2011-06-26 02:20:34 <phantomcircuit> even if it's micro seconds
307 2011-06-26 02:22:48 <jrmithdobbs> i'm not sure i buy that assertion
308 2011-06-26 02:23:18 <jrmithdobbs> especially with 10gig
309 2011-06-26 02:23:39 <jrmithdobbs> assuming sufficiently powerful systems on both ends
310 2011-06-26 02:24:11 <upb> interesting, how would you scale order matching horizontally?
311 2011-06-26 02:24:46 <phantomcircuit> jrmithdobbs, the actual order matching shouldn't take more than a few microseconds
312 2011-06-26 02:25:16 <phantomcircuit> order in, buy up/sell down, journal result
313 2011-06-26 02:26:58 <phantomcircuit> the slow part is the journal
314 2011-06-26 02:27:59 <pakaran> why does it seem that no new testnet blocks are showing up?
315 2011-06-26 02:28:22 <pakaran> everything I've done on testnet shows up unconfirmed/0
316 2011-06-26 02:29:27 <phantomcircuit> pakaran, someone jacked up the difficulty and left
317 2011-06-26 02:29:36 <phantomcircuit> so it takes forever to generate a new block
318 2011-06-26 02:29:52 <pakaran> and the difficulty isnt going to adust for awhile?
319 2011-06-26 02:29:54 <xtalmath> are testnet coins being sold?
320 2011-06-26 02:30:02 <xtalmath> pakaran: correct
321 2011-06-26 02:30:10 <CIA-103> bitcoin: Neil Kistner * rf1a8503f9d16 mining-proxy/htdocs/views/admin/dashboard.view.php: Removed the word "shares" after the number of shares http://tinyurl.com/5v62pl7
322 2011-06-26 02:30:13 <CIA-103> bitcoin: Chris Howie * r670de6138b0e mining-proxy/htdocs/config.inc.php.sample: Whitespace http://tinyurl.com/5tq3gtl
323 2011-06-26 02:30:17 <CIA-103> bitcoin: Chris Howie * r05ace9fddafe mining-proxy/htdocs/views/admin/dashboard.view.php: Merge remote branch 'wyze/remove-shares' http://tinyurl.com/68j4tvw
324 2011-06-26 02:30:36 <xtalmath> who is CIA-103 and why does he post cryptic messages?
325 2011-06-26 02:30:42 <pakaran> is it possible to run a cpu miner on testnet/
326 2011-06-26 02:30:46 <pakaran> becase i'll do it
327 2011-06-26 02:30:57 <gmaxwell> xtalmath: it's a SCM to IRC relay.
328 2011-06-26 02:31:08 <xtalmath> SCM?
329 2011-06-26 02:31:20 <gmaxwell> source code management. e.g. svn.
330 2011-06-26 02:31:39 <xtalmath> not supply chain management or software configuration management
331 2011-06-26 02:31:53 <xtalmath> M seems to definitely stand for management according to you and google
332 2011-06-26 02:34:21 <pakaran> xtalmath, I'll give you testnet coins... you just won't be able to do anything with them until a block gets generated... sigh
333 2011-06-26 02:35:18 <xtalmath> I dont want any, just thought that if an exchange was setup once difficulty has lowered, the difficulty would be maintained more smoothly
334 2011-06-26 02:35:42 <jrmithdobbs> why is cia even watching mining-proxy?
335 2011-06-26 02:35:48 <pakaran> is 26601 the current testnet block number btw?
336 2011-06-26 02:35:54 <pakaran> xtalmath, that's a thought.
337 2011-06-26 02:35:59 <xtalmath> testnet coins are bound to get lost, since well, theyre for testing purpouses, so it might pique interest of speculators
338 2011-06-26 02:36:13 <enquirer> in 8 minutes there were 400 operations on addr.dat, out of this 10 flushes
339 2011-06-26 02:36:17 <pakaran> i've seen people offering to sell 1k testnet coins for .05 real ones, etc
340 2011-06-26 02:36:20 <pakaran> not sure how close that is to the minig difficulty ratio though
341 2011-06-26 02:36:29 <jrmithdobbs> enquirer: ya, addr.dat and related code are crap
342 2011-06-26 02:36:47 <jrmithdobbs> enquirer: watch it on osx, p sure the ops double or more
343 2011-06-26 02:36:52 <pakaran> xtalmath, and some of those speculators will be mining... and thus the amout of new blocks on testnet should be the same as the "exchange rate"?
344 2011-06-26 02:37:19 <xtalmath> no amount of new blocks should remain 10minutes per block
345 2011-06-26 02:37:20 <enquirer> blk0001.dat was opened and closed 1158 times
346 2011-06-26 02:37:36 <pakaran> right
347 2011-06-26 02:37:41 <pakaran> that's what i was trying to say, right
348 2011-06-26 02:37:54 <pakaran> because mining on testnet would have the same expected bitcoins as mining on the real net
349 2011-06-26 02:37:59 <samlander> what's the point in buying testnet coins
350 2011-06-26 02:38:06 <pakaran> on the other hand, some of the speculators would also abuse the testnet faucet like crazy
351 2011-06-26 02:38:22 <pakaran> samlander, setting up a market would encourage people to mine on testnet
352 2011-06-26 02:38:28 <samlander> i understand that the source is open and anyone can start a new currency, but at what cost?
353 2011-06-26 02:38:44 <pakaran> so we wouldn't have the situation like now where the new block is taking forever because it was one person mining heavily, who left
354 2011-06-26 02:38:46 <samlander> why dillute the value of the currently established currency by fracturing the source?
355 2011-06-26 02:39:11 <xtalmath> actually comparing the current block number of testnet  with current block of real-block of real at introduction of testnet might be interesting to graph over time
356 2011-06-26 02:39:12 <jrmithdobbs> pakaran: who was that
357 2011-06-26 02:39:16 <enquirer> blkindex 2919 writes
358 2011-06-26 02:39:27 <pakaran> we dont know
359 2011-06-26 02:39:44 <pakaran> i'd just mine on testnet rather than complain
360 2011-06-26 02:39:56 <pakaran> but all i have is a dual core lapptop cpu that gets about 600 khash
361 2011-06-26 02:40:03 <enquirer> databaselog   1038 flushes to disk
362 2011-06-26 02:40:06 <pakaran> and i don't know how to run my cpu miner on testnet without a pool
363 2011-06-26 02:41:16 <samlander> my god i wonder how many blocks i would solve pointing my 1 GH/s at it
364 2011-06-26 02:41:43 <upb> theres this new exchange doing testing on testnet
365 2011-06-26 02:41:47 <upb> that probably caused it :)
366 2011-06-26 02:41:53 <pakaran> about ten an hour, samlander
367 2011-06-26 02:42:05 <pakaran> and if you are willing to generate 2 blocks, a lot of us would be grateful
368 2011-06-26 02:42:12 <pakaran> like i could send my testnet coins back to the faucet
369 2011-06-26 02:42:16 <pakaran> and not feel guilty anymore
370 2011-06-26 02:42:41 <jrmithdobbs> pakaran: that's funny
371 2011-06-26 02:42:49 <enquirer> so it's mostly databaselog flushes that are responsible for most IO
372 2011-06-26 02:42:49 <jrmithdobbs> time to restart testnet? ;p
373 2011-06-26 02:43:24 <xtalmath> pakaran: if he points his GH/s at it it would be 24 blocks an hour
374 2011-06-26 02:44:09 <nhodges> mybitcoin SCI api is not posting back anything to my receiver
375 2011-06-26 02:44:10 <nhodges> :/
376 2011-06-26 02:44:13 <jrmithdobbs> and fuck up difficulty more!
377 2011-06-26 02:44:14 <jrmithdobbs> lol
378 2011-06-26 02:46:05 <enquirer> ah no, thats 173 log flushes
379 2011-06-26 02:46:37 <lfm> enquirer: in how long?
380 2011-06-26 02:47:08 <enquirer> 8 minutes
381 2011-06-26 02:47:17 <lfm> minor problem
382 2011-06-26 02:47:39 <enquirer> suspisious why  blk0001.dat was opened and closed 1158 times
383 2011-06-26 02:48:07 <lfm> in 8 minutes/
384 2011-06-26 02:48:15 <lfm> ?
385 2011-06-26 02:48:19 <enquirer> yes
386 2011-06-26 02:48:24 <lfm> hmm , good question
387 2011-06-26 02:48:28 <enquirer> let's see
388 2011-06-26 02:49:15 <gmaxwell> enquirer: you're not still syncing up with the blockchain are you?
389 2011-06-26 02:51:06 <enquirer> 133310 blocks
390 2011-06-26 02:51:09 <pakaran> the problem is testnet difficulty won't go *down* until enough blocks are generated for a difficulty update.
391 2011-06-26 02:51:12 <enquirer> running it over 24 hours
392 2011-06-26 02:51:44 <lfm> pakaran: yup, It is the problem the old testnet faced too
393 2011-06-26 02:52:06 <lfm> ;;bc,blocks
394 2011-06-26 02:52:07 <gribble> 133310
395 2011-06-26 02:52:10 <pakaran> hmm
396 2011-06-26 02:52:16 <pakaran> how do i point a cpu miner to testnet?
397 2011-06-26 02:52:22 <lfm> enquirer: so you are caught up now
398 2011-06-26 02:52:22 <pakaran> i can run one without a pool, right?
399 2011-06-26 02:52:52 <lfm> pakaran: point miner at localhost and run bitcoind -testnet
400 2011-06-26 02:53:29 <pakaran> and how do i enter the key that the generated coins should be given to?
401 2011-06-26 02:53:56 <lfm> pakaran: you dont, bitcoind will make up a key and put it in a local wallet.dat
402 2011-06-26 02:54:43 <pakaran> i have a testnet miner running
403 2011-06-26 02:54:55 <pakaran> but you're saying i just enter localhost instead of the mining pool name, with no username or pass?
404 2011-06-26 02:54:57 <pakaran> i'm confused
405 2011-06-26 02:55:05 <pakaran> i mean, i have a testnet *cllient* running
406 2011-06-26 02:55:29 <lfm> pakaran: use same rpcuser and rpcpassword from your ~/.bitcoin/bitcoin.cong
407 2011-06-26 02:55:35 <lfm> .conf
408 2011-06-26 02:56:37 <lfm> pakaran: make one then
409 2011-06-26 02:56:48 <pakaran> ...
410 2011-06-26 02:56:49 <lfm> eeeeiiiii flood
411 2011-06-26 02:56:50 <pakaran> i'm very confused
412 2011-06-26 02:56:55 <pakaran> is this going to be too much of a fuss?
413 2011-06-26 02:57:34 <lfm> make file ~/.bitcoin/bitcoin.conf and put in two lines :rpcuser=username and rpcpass=password1234
414 2011-06-26 02:58:16 <lfm> rpcpassword=pass1234
415 2011-06-26 02:58:57 <pakaran> in my testnet directory, you mean?
416 2011-06-26 02:59:01 <lfm> well if makeing two lines in a text editor is too much ...
417 2011-06-26 03:00:12 <lfm> not in testnet dir afaik
418 2011-06-26 03:01:35 <lfm> whats challenge.txt for?
419 2011-06-26 03:03:08 <pakaran> hmm
420 2011-06-26 03:03:17 <pakaran> minerd doesn't take the -testnet option
421 2011-06-26 03:03:25 <lfm> right
422 2011-06-26 03:03:48 <pakaran> so i just point it at the testnet port, but what's that?
423 2011-06-26 03:04:08 <lfm> rpc port does not change for testnet
424 2011-06-26 03:04:20 <lfm> only the p2p port changes
425 2011-06-26 03:05:01 <pakaran> so i need to not be running a regular client at all?
426 2011-06-26 03:05:13 <lfm> huh?
427 2011-06-26 03:05:27 <lfm> oh ya shut down your regular bitcoin
428 2011-06-26 03:05:27 <pakaran> so that it connects to the testnet client, and mines for testnet?
429 2011-06-26 03:05:57 <pakaran> ok
430 2011-06-26 03:06:07 <pakaran> that was the problem, i think
431 2011-06-26 03:06:38 <pakaran> it's still not working.
432 2011-06-26 03:06:40 <pakaran> nathan@pakaran ~/cpuminer-1.0 $ minerd  --url http://localhost --userpass foobar:32768
433 2011-06-26 03:07:27 <lfm> pakaran: let the bitcoin run for a while befor starting the miner to make sure you have whole testnet block chain
434 2011-06-26 03:08:08 <lfm> nm looks like its running
435 2011-06-26 03:10:05 <lfm> pakaran: blocks should be 25729,
436 2011-06-26 03:10:07 <lfm> or more
437 2011-06-26 03:10:31 <pakaran> i have 26603?
438 2011-06-26 03:10:55 <pakaran> anyhow, i edited the config file, and i'm still getting could not connect to rpc server
439 2011-06-26 03:11:07 <lfm> ok I have 26603, too now
440 2011-06-26 03:11:31 <lfm> pakaran: di you put that user and password on your miner then?
441 2011-06-26 03:12:10 <pakaran> oh crap
442 2011-06-26 03:12:20 <pakaran> it's rpcpassword=, not rpcpass=
443 2011-06-26 03:12:30 <lfm> ya sorry
444 2011-06-26 03:13:30 <pakaran> i'm STILL getting http request failed
445 2011-06-26 03:13:41 <pakaran> i do want to just put in http://localhost for the url, right?
446 2011-06-26 03:13:42 <pakaran> no port?
447 2011-06-26 03:13:48 <lfm> pakaran: did you put that user and password on your miner then?
448 2011-06-26 03:14:00 <pakaran> yes, and i checked them several times
449 2011-06-26 03:14:02 <lfm> maybe need to use 127.0.0.1 not localhost
450 2011-06-26 03:14:31 <pakaran> but i didn'tk now it was possible to use a regular client as a mining pool though?
451 2011-06-26 03:14:43 <pakaran> and it doesn't connect to 127.0.0.1 either
452 2011-06-26 03:15:05 <lfm> ok are you running bitcoind or bitcoin?
453 2011-06-26 03:15:14 <pakaran> wxbitcoin
454 2011-06-26 03:15:21 <lfm> ok add -server option
455 2011-06-26 03:15:55 <lfm> I never heard of wxbitcoin actually
456 2011-06-26 03:16:03 <pakaran> still not working
457 2011-06-26 03:16:14 <lfm> restart minerd?
458 2011-06-26 03:16:24 <pakaran> i just did
459 2011-06-26 03:16:28 <pakaran> still http request failed
460 2011-06-26 03:17:37 <lfm> so minerd userpass "username:pass123" maybe
461 2011-06-26 03:17:43 <pakaran> ok
462 2011-06-26 03:17:45 <lfm> so minerd userpass="username:pass123" maybe
463 2011-06-26 03:17:47 <pakaran> putting in the port did work
464 2011-06-26 03:17:52 <pakaran> in the miner
465 2011-06-26 03:17:55 <pakaran> it's saying it's mining
466 2011-06-26 03:18:17 <pakaran> presumably mining for one of my addresses in my testnet client?
467 2011-06-26 03:18:26 <lfm> port 8332?
468 2011-06-26 03:18:33 <pakaran> yeah
469 2011-06-26 03:18:56 <lfm> whats your khas/sec then?
470 2011-06-26 03:19:00 <lfm> khash
471 2011-06-26 03:19:01 <pakaran> the only "issue" is there's no longpoll
472 2011-06-26 03:19:06 <pakaran> it hasn't displayed yet
473 2011-06-26 03:19:15 <pakaran> which is also odd since it does every minute when mining for a pool
474 2011-06-26 03:19:37 <lfm> true there is no longpoll but since it is localhost it should be ok, not much slack
475 2011-06-26 03:20:02 <lfm> pakaran: the first one can be a while
476 2011-06-26 03:20:17 <upb> haha i hadnt read this thread http://forum.bitcoin.org/index.php?topic=20535.0;all
477 2011-06-26 03:20:21 <pakaran> so when a block is found it should update pretty much immediately?
478 2011-06-26 03:20:27 <upb> 'We had no intention of getting this to happen, and we have followed every industry standard to make this secure.'
479 2011-06-26 03:20:59 <lfm> pakaran: naw it can still take 120 blocks before you can spend it
480 2011-06-26 03:21:53 <pakaran> i know that.
481 2011-06-26 03:21:58 <pakaran> it needs to wait for confirmations.
482 2011-06-26 03:22:15 <pakaran> and in any case, at current testnet difficulty...  i think i'll take about an hour to cpu mine a block
483 2011-06-26 03:24:35 <lfm> Difficulty: 45.5018    --- est next difficulty: 69.58, in 7.33 days
484 2011-06-26 03:25:03 <gribble> Error: "bc" is not a valid command.
485 2011-06-26 03:25:03 <pakaran> ;;bc gen 600 45
486 2011-06-26 03:25:15 <pakaran> so how long will it take my 600 khash to generate a block?
487 2011-06-26 03:25:21 <pakaran> i don't really know gribble syntax
488 2011-06-26 03:25:30 <lfm> ;;bc,gend 600 45
489 2011-06-26 03:25:31 <gribble> The expected generation output, at 600 Khps, given the supplied difficulty of 45, is 13.4110450745 BTC per day and 0.558793544769 BTC per hour.
490 2011-06-26 03:25:44 <gribble> The average time to generate a block at 600 Khps, given the supplied difficulty of 45, is 3 days, 17 hours, 28 minutes, and 42 seconds
491 2011-06-26 03:25:44 <lfm> ;;bc,calcd 600 45
492 2011-06-26 03:25:55 <pakaran> so likely not going to happen, thanks.
493 2011-06-26 03:26:14 <lfm> youd prolly need to let it run a few days ya
494 2011-06-26 03:57:14 <FellowTraveler> hi all.
495 2011-06-26 03:57:29 <lfm> FellowTraveler: hi
496 2011-06-26 05:05:06 <CIA-103> bitcoin: Daniel Folkinshteyn * r37041fe28826 supybot-bitcoin-marketmonitor/GPG/ (plugin.py test.py): GPG: add information on current login status to gpg info command. http://tinyurl.com/43n58d9
497 2011-06-26 05:52:53 <stuhood> ahh& testnet must be using the new irc channels now, eh?
498 2011-06-26 05:55:39 <stuhood> hm, no& #bitcoinTEST isn't partitioned.
499 2011-06-26 05:56:53 <stuhood> http://blockexplorer.com/testnet is just a little behind i guess.
500 2011-06-26 07:44:13 <minus> hm, bitcoin crashes on the testnet
501 2011-06-26 07:44:29 <minus> bitcoin: serialize.h:1057: CDataStream& CDataStream::write(const char*, int): Assertion `nSize >= 0' failed.
502 2011-06-26 07:47:40 <sipa> which version?
503 2011-06-26 07:49:16 <minus> latest
504 2011-06-26 07:49:34 <minus> linux x86 in a VM, mining
505 2011-06-26 07:51:24 <minus> 06/26/2011 09:34:40 Flushing wallet.dat \n Flushed wallet.dat 5ms
506 2011-06-26 07:51:34 <minus> that's the last thing that happened
507 2011-06-26 07:51:44 <minus> maybe because keypool=0?
508 2011-06-26 07:52:10 <minus> i set keypool=10 now and continue mining
509 2011-06-26 08:04:43 <CIA-103> bitcoin: Pieter Wuille master * rd3800d5 / (src/wallet.cpp src/wallet.h):
510 2011-06-26 08:04:44 <CIA-103> bitcoin: Merge pull request #347 from sipa/delkeyuser
511 2011-06-26 08:06:40 <diki> is it possible to run two bitcoinD daemons on one machine(linux)?
512 2011-06-26 08:06:56 <diki> both of them in -server mode
513 2011-06-26 08:07:09 <minus> should be
514 2011-06-26 08:07:34 <diki> i dont know though...how will both of them work with the same port
515 2011-06-26 08:07:41 <minus> yea
516 2011-06-26 08:07:48 <minus> that's what iwas just thinking about
517 2011-06-26 08:08:16 <minus> you can change the RPC port but the p2p port, no idea
518 2011-06-26 08:09:24 <sipa> -port
519 2011-06-26 08:09:30 <upb> if you sent one as non listening you could
520 2011-06-26 08:10:08 <minus> -port isn't in there https://en.bitcoin.it/wiki/Running_Bitcoin
521 2011-06-26 08:10:15 <minus> lousy work D:
522 2011-06-26 08:11:40 <minus> $ ./bitcoin -port 8334
523 2011-06-26 08:11:44 <minus> uhm
524 2011-06-26 08:11:59 <wumpus> minus: then add it it's a wiki
525 2011-06-26 08:12:03 <minus> i'm confused now
526 2011-06-26 08:12:32 <sipa> there are a few other undocumented options as well
527 2011-06-26 08:12:56 <minus> oh sweet, port in the config works
528 2011-06-26 08:13:09 <sipa> yes, all options can be both in config file or on the command line
529 2011-06-26 08:13:41 <minus> if you put option in the cmdline does it not parse the config file anymore?
530 2011-06-26 08:13:48 <minus> wumpus: i'm on it now
531 2011-06-26 08:14:14 <sipa> command line overrides whatever is in the config file
532 2011-06-26 08:14:17 <wumpus> minus: prolly should have been -port=8334
533 2011-06-26 08:14:36 <minus> hmm
534 2011-06-26 08:14:50 <minus> so bitcoin is not using getopt
535 2011-06-26 08:15:24 <wumpus> no it's using some boost option library
536 2011-06-26 08:18:30 <minus> is there a place you can get the whole blockchain from?
537 2011-06-26 08:19:06 <minus> like the full blocks, not the compacted ones
538 2011-06-26 08:19:46 <sipa> BlueMatt has a download on his site
539 2011-06-26 08:20:09 <minus> and that would be?
540 2011-06-26 08:25:16 <zamgo> http://lmgtfy.com/?q=bluematt+blockchain
541 2011-06-26 08:26:39 <prof7bit> is my assumption correct that a non-validating client like bitcoinj can be allowed to relay unconfirmed transactions to other clients but should *NEVER* relay any blocks to other clients in order to not weaken the integrity of the block chain?
542 2011-06-26 08:27:39 <minus> lol zamgo i did search for it, i must have used the wrong keywords
543 2011-06-26 08:28:58 <zamgo> ;)
544 2011-06-26 08:30:42 <prof7bit> if i write a lightweight client that does the simplified verification and it would become so immense popular that the majority of users would prefer it over the original client, would I be able to accidentally destroy the entire bitcoin network by releasing such a client to the unwashed masses?
545 2011-06-26 08:32:08 <sipa> as long as lightweight clients connect to a majority of honest full nodes, or connect to one trusted full node, there shouldn't be a problem
546 2011-06-26 08:32:09 <minus> prof7bit: ofc not
547 2011-06-26 08:32:30 <minus> hm
548 2011-06-26 08:33:05 <minus> i have no problem with the fullblown client on my desktop but i'd need something lightweight for my server
549 2011-06-26 08:33:08 <prof7bit> if i blindly relay blocks without verifying them?
550 2011-06-26 08:33:13 <minus> cause i cant really spare 70MB on it
551 2011-06-26 08:34:05 <sipa> prof7bit: lightweight nodes do not advertize as full nodes
552 2011-06-26 08:34:16 <sipa> so other lightweight nodes should ignore them
553 2011-06-26 08:34:24 <wumpus> sipa: https://github.com/bitcoin/bitcoin/issues/348
554 2011-06-26 08:34:32 <prof7bit> if they play by the ruls
555 2011-06-26 08:34:35 <prof7bit> rules
556 2011-06-26 08:35:28 <sipa> wumpus: meh, it's a mess
557 2011-06-26 08:36:12 <wumpus> yep usually I can fix these things b ut I have to admit this one is above my head :-)
558 2011-06-26 08:36:18 <sipa> wumpus: serialize.h depends on wallet.h ?
559 2011-06-26 08:36:55 <wumpus> huh it doesn't
560 2011-06-26 08:37:13 <prof7bit> http://code.google.com/p/bitcoinj/source/browse/trunk/src/com/google/bitcoin/core/VersionMessage.java
561 2011-06-26 08:37:15 <wumpus> let me see, maybe I did something wrong xD already thought it didn't make sense
562 2011-06-26 08:37:20 <prof7bit> line 30
563 2011-06-26 08:37:41 <sipa> wumpus: you're rebasing bitcoin-qt to master?
564 2011-06-26 08:37:42 <prof7bit> bitcoinj advertises itself as NODE_NETWORK
565 2011-06-26 08:38:31 <wumpus> sipa: yes
566 2011-06-26 08:38:35 <sipa> nice
567 2011-06-26 08:38:56 <prof7bit> and this bit should not be set when a client does not have verified blocks
568 2011-06-26 08:39:15 <sipa> prof7bit: hmm, talk to TD
569 2011-06-26 08:39:28 <wumpus> sipa: ok can close the issue, this was me making a mistake (though it's still a mess xD)
570 2011-06-26 08:39:55 <sipa> it's quite possible there is some dependency loop
571 2011-06-26 08:40:23 <wumpus> yeah very possible
572 2011-06-26 08:40:27 <sipa> that just happens to not be a problem because of the particular order of includes in headers.h
573 2011-06-26 08:41:02 <wumpus> indeed, and external code might not want to include the whole headers.h shebang
574 2011-06-26 08:41:53 <sipa> we should get rid of headers.h anyway
575 2011-06-26 08:42:02 <sipa> ... someday
576 2011-06-26 08:42:05 <wumpus> right
577 2011-06-26 08:42:24 <yorick> how many calls to getreceivedbyaddress can I make in a minute?
578 2011-06-26 08:42:30 <yorick> without risking bitcoind freezing up?
579 2011-06-26 08:42:39 <upb> what does this dependency loop matter tho, it does have include guards right ?
580 2011-06-26 08:42:54 <sipa> yorick: it shouldn't freeze up no matter what, if it does, you found a bug
581 2011-06-26 08:43:04 <sipa> but there are known reports of these, indeed
582 2011-06-26 08:43:23 <prof7bit> the "official" protocol definition is quite vague and it does not yet explicitly mention the problems or requirements or and detailed behaviour what to do when a client advertises itself not as a full node, so i thought i throw some collective brain power from this channel at this question.
583 2011-06-26 08:43:43 <wumpus> include guards only guard against recursive disasters
584 2011-06-26 08:43:44 <wumpus> you can still have issues with dependency loops, missing types etc
585 2011-06-26 08:45:13 <wumpus> the problem basically was that in serialize.h I got an error about the definition of CWalletTx being an incomplete type, so I naively added  an include for the wallet type... but this is template code so it works differently
586 2011-06-26 08:45:36 <upb> ahhh
587 2011-06-26 08:46:34 <sipa> also, apparently nobody on the dev mailing list wants to test autotools...
588 2011-06-26 08:47:42 <wump> db.h needs wallet.h it seems
589 2011-06-26 08:47:51 <sipa> yes
590 2011-06-26 08:48:49 <sipa> one solution would be to create an abstract class 'LoadableWallet' in a separate file, which db.h depends on, and cwallet inherits from
591 2011-06-26 08:51:11 <gm> any interest in a nicer OS X bitcoin client? i'm thinking of writing one
592 2011-06-26 08:51:33 <sipa> gm: there was already someone working on that
593 2011-06-26 08:51:39 <wump> maybe I should just stop worrying and love headers.h
594 2011-06-26 08:51:46 <gm> sipa: who?
595 2011-06-26 08:51:57 <sipa> wump: for now, i think that's the best option
596 2011-06-26 08:52:02 <gm> i was thinking of adding support for stuff like storing private keys in the OS X keychain
597 2011-06-26 08:52:19 <wump> :D
598 2011-06-26 08:52:33 <sipa> gm: not sure, but in general, i'd suggest you to wait until wump's bitcoin-qt is merged in mainline, solving the hardest part of creating an alternate gui
599 2011-06-26 08:52:56 <gm> actually i was going to start by controlling bitcoin via RPC
600 2011-06-26 08:52:57 <prof7bit> ok, forget what I said about bitcoinj, I did not read the code correctly. It will not set this bit. I need more coffee. Line 57:        localServices = 0;
601 2011-06-26 08:53:23 <sipa> gm: then have a look at spesmilo
602 2011-06-26 08:53:24 <wump> gm: I don't think you can achieve private key magic by just using JSON API
603 2011-06-26 08:53:48 <sipa> indeed, most inner workings are not exposed via rpc
604 2011-06-26 08:55:31 <gm> wump: that comes later, after bitcoin-qt is merged
605 2011-06-26 08:56:19 <wump> gm: yes it might be best to keep a level of abstraction so you can switch how you interface later
606 2011-06-26 08:56:21 <sipa> gm: you have a working osx dev environnement for bitcoin now?
607 2011-06-26 08:56:38 <prof7bit> what qt version is required for bitcoin-qt? I hope it is not the very latest bleeding edge?
608 2011-06-26 08:56:47 <gm> sipa: i'm not actually planning on modifying bitcoin at this stage
609 2011-06-26 08:57:01 <wump> prof7bit: it shouldn't require the bleeding edge
610 2011-06-26 08:57:15 <sipa> gm: ok, so you don't?
611 2011-06-26 08:57:29 <gm> sipa: haven't really tried
612 2011-06-26 08:57:31 <wump> prof7bit: (it still might, though, but feel free to let me know if there are incompatiblities with your qt version, as long as it's not some 2005 version :P)
613 2011-06-26 08:58:33 <prof7bit> please make it compile with relatively old qt4, maybe set up your development machine with an older qt4.
614 2011-06-26 08:58:42 <deSzadou> Guys if i am mining than i have to run bitcoin app too or just mining tool ?
615 2011-06-26 08:58:48 <wump> I have to rely on other people for that
616 2011-06-26 08:58:58 <gm> deSzadou: if you're mining in a pool, just the miner
617 2011-06-26 08:59:18 <prof7bit> a developer should not have bleeding edge libs on his machine.
618 2011-06-26 08:59:26 <deSzadou> so i dont have to run in cmd "bitcoin -server"
619 2011-06-26 08:59:28 <wump> don't tell me what I should and should not do
620 2011-06-26 08:59:31 <wump> you're not paying me for this
621 2011-06-26 08:59:54 <deSzadou> guys please answer me : D
622 2011-06-26 08:59:54 <prof7bit> i just wanted to help you
623 2011-06-26 09:00:05 <wump> okay, thanks then
624 2011-06-26 09:00:06 <deSzadou> i dont have to run "bitcoin -server" in cmd if i am mining in pool?
625 2011-06-26 09:00:10 <doublec> deSzadou: gm did asnwer you
626 2011-06-26 09:00:11 <sipa> deSzadou: correct
627 2011-06-26 09:00:14 <deSzadou> great
628 2011-06-26 09:00:14 <wump> you can help by testing :)
629 2011-06-26 09:00:16 <deSzadou> thanks guys !
630 2011-06-26 09:01:07 <prof7bit> because i know what happens when your bug tracker is suddenly full of people who complain that it wont run on current debian stable, etc.
631 2011-06-26 09:01:33 <wump> in that case I simply port back things
632 2011-06-26 09:02:25 <wump> I already had a mail that some frobnotz was not compatible with qt 4.6, so I fixed it and made another user happy xD
633 2011-06-26 09:02:56 <prof7bit> i've had the same problem with TorChat and the available python and wxpython versions on the various distributions
634 2011-06-26 09:03:41 <prof7bit> i decided to use really old versions only to make stop people complaining
635 2011-06-26 09:03:44 <wump> yes I've been into a lot of fights with python version incompatiblities as well :/
636 2011-06-26 09:04:21 <wump> yea ideally you set a barrier 'Qt X.X required'.. but I'm not that far yet
637 2011-06-26 09:04:35 <prof7bit> until 2 years ago I even still used windows 98 to build the windows version, now i am doing this on windows xp
638 2011-06-26 09:05:31 <wump> hehe
639 2011-06-26 09:05:59 <prof7bit> and now have sporadic complaints again about missing win98 compatibility! imagine! 2011!
640 2011-06-26 09:06:15 <wump> windows xp is the last windows version I touched, so I  tend to have the inverse problem
641 2011-06-26 09:06:46 <wump> lol, yes especially in China windows 98 is still being used
642 2011-06-26 09:07:39 <prof7bit> until now it seems everything I build on XP runs just fine on vista and 7, so i will keep using winxp for a long time
643 2011-06-26 09:09:11 <wump> yes that's a safe bet
644 2011-06-26 09:09:24 <sipa> wump: which OS do you use?
645 2011-06-26 09:09:28 <sipa> to develop
646 2011-06-26 09:09:43 <b4epoche> OSX
647 2011-06-26 09:09:46 <wump> ubuntu on my desktop and kubuntu on my laptop
648 2011-06-26 09:09:52 <prof7bit> unless i really badly need some bleeding edge feature of some bleeding edge api (what actually never hapopens) i can happily live with linking against old stuff
649 2011-06-26 09:10:25 <sipa> b4epoche: have you built bitcoin from source there yet?
650 2011-06-26 09:10:46 <b4epoche> sure...  b4epoche = ericmock.
651 2011-06-26 09:11:05 <sipa> ah!
652 2011-06-26 09:11:19 <b4epoche> trying out a new IRC client and a new nick ;-)
653 2011-06-26 09:11:34 <sipa> b4epoche: care to try to build autotools-based bitcoin?
654 2011-06-26 09:11:42 <wump> prof7bit: with Python I usually tend to write to the newest version docs, then fallback where needed for older versions. But for a static language like c++ that's a bit harder :/
655 2011-06-26 09:12:03 <b4epoche> sure, I tried b4 but it died during configure
656 2011-06-26 09:12:16 <sipa> oh, it didn't change
657 2011-06-26 09:12:17 <sipa> so nvm
658 2011-06-26 09:15:14 <Choko> according to the block format the nonce is only 4 bytes, this will mean at some time when the difficulty is sufficient large there will be a great chance that no nonce exists for a new block? https://en.bitcoin.it/wiki/Protocol_specification#block
659 2011-06-26 09:15:47 <sipa> there is a second nonce inside the coinbase transaction
660 2011-06-26 09:15:52 <sipa> of arbitrary size
661 2011-06-26 09:16:09 <sipa> otherwise an 5970 would run out of nonces every 5s
662 2011-06-26 09:16:31 <Choko> exactly my thought
663 2011-06-26 09:17:09 <sipa> so, that's not a problem
664 2011-06-26 09:21:08 <sipa> b4epoche: what was the problem again, exactly?
665 2011-06-26 09:21:35 <b4epoche> I was wondering that myself ;-)
666 2011-06-26 09:21:39 <b4epoche> I can't remember
667 2011-06-26 09:23:30 <b4epoche> I can try again.  but it doesn't look like I have the files anymore.  URL?
668 2011-06-26 09:24:19 <sipa> https://github.com/jaromil/bitcoin/commits/autotools3
669 2011-06-26 09:26:48 <b4epoche> command sequence?  autoconf, automake?
670 2011-06-26 09:27:22 <sipa> autoreconf -i, autoconf, configure, make
671 2011-06-26 09:28:35 <b4epoche> ah, right:
672 2011-06-26 09:28:37 <b4epoche> ./configure: line 7126: syntax error near unexpected token `SSL,'
673 2011-06-26 09:29:26 <gribble> The operation succeeded.
674 2011-06-26 09:29:26 <sipa> ;;later tell jaromil b4epoche had problem with autotools on OSX: './configure: line 7126: syntax error near unexpected token `SSL,'', './configure: line 7126: `PKG_CHECK_MODULES(SSL, libssl >= 0.9, :,''
675 2011-06-26 09:29:36 <prof7bit> does anybody know what is the current state of the private key encryption in the official client? is this: https://github.com/bitcoin/bitcoin/pull/232 the current state of the discussion or are there alternatives? how is the progress on this?
676 2011-06-26 09:29:48 <sipa> prof7bit: BlueMatt is working on a rewrite
677 2011-06-26 09:42:28 <minus> mining on the testnet is so much more fun
678 2011-06-26 09:42:35 <minus> i actually generate blocks!
679 2011-06-26 09:42:56 <minus> wtf is up with that guy
680 2011-06-26 09:54:34 <zamgo> minus: try namecoin testnet.  Difficulty 1
681 2011-06-26 09:55:02 <b4epoche> wtf what guy?
682 2011-06-26 09:55:04 <minus> that'd be like one block every 10 seconds
683 2011-06-26 09:55:05 <sipa> try testnet-in-a-box
684 2011-06-26 09:58:06 <pasky> can i switch my bitcoind to testnet by just enabling that option, or will that wipe out / rebuild my database and i'll have to re-download the whole blockchain again next time i switch back to normal net?
685 2011-06-26 09:58:23 <sipa> it keeps the testnet db in a separate directory
686 2011-06-26 09:59:26 <gm> minus: someone needs to start a testnet mining pool :P
687 2011-06-26 10:01:29 <prof7bit> someone needs to pay dollars for test coins
688 2011-06-26 10:01:46 <sipa> they have been traded for real bitcoins in the past
689 2011-06-26 10:02:13 <minus> gm: all the power of deepbit or another pool should work on the testnet for a short period of time
690 2011-06-26 10:02:23 <minus> probably results in 5 blocks per second
691 2011-06-26 10:02:32 <sipa> what is tn difficulty now?
692 2011-06-26 10:02:37 <minus> 45
693 2011-06-26 10:02:45 <minus> wait no
694 2011-06-26 10:02:47 <sipa> ;;bc,deepbit
695 2011-06-26 10:02:50 <gribble> 3846942000
696 2011-06-26 10:02:50 <minus> it was 45 yesterday
697 2011-06-26 10:03:02 <minus> blockexplorer testnet doesnt update
698 2011-06-26 10:03:06 <gribble> Error: '0.025120411007' is not a valid integer.
699 2011-06-26 10:03:06 <sipa> ;;bc,calcd [bc,deepbit] 22.5
700 2011-06-26 10:03:09 <minus> *didn't, since yesterady
701 2011-06-26 10:03:22 <sipa> ;;bc,calcd 3846942000 22.5
702 2011-06-26 10:03:23 <gribble> Error: '0.025120411007' is not a valid integer.
703 2011-06-26 10:03:28 <gribble> (bc,calcd <an alias, 2 arguments>) -- Alias for "echo The average time to generate a block at $1 Khps, given the supplied difficulty of $2, is [time elapsed [math calc 1/((2**224-1)/$2*$1*1000/2**256)]]".
704 2011-06-26 10:03:28 <sipa> ;;bc,calcd 3846942000
705 2011-06-26 10:04:35 <gm> minus: i sat here hammering 400 megahashes at it, made 250 btc in 10 mins or so
706 2011-06-26 10:05:00 <minus> yeah, i'm at like 1200BTC now
707 2011-06-26 10:10:35 <kinlo> how can I see when a block is generated?
708 2011-06-26 10:11:03 <lfm> kinlo: watch ~/.bitcoin/debug.log
709 2011-06-26 10:13:08 <kinlo> is there a way to do it trough rpc?
710 2011-06-26 10:13:23 <sipa> listtransactions shows generations
711 2011-06-26 10:13:55 <kinlo> yeah but there is no usable timestamp in listtransactions
712 2011-06-26 10:13:55 <lfm> kinlo try the getgetblockcountcommand
713 2011-06-26 10:13:59 <lfm> kinlo try the getgetblockcount command
714 2011-06-26 10:14:06 <minus> getblockcount
715 2011-06-26 10:14:17 <lfm> ya what he sed
716 2011-06-26 10:14:22 <kinlo> :)
717 2011-06-26 10:14:29 <kinlo> lemme try that
718 2011-06-26 10:14:40 <gribble> 133362
719 2011-06-26 10:14:40 <lfm> ;;bc,blocks
720 2011-06-26 10:14:53 <kinlo> mmmz
721 2011-06-26 10:15:29 <kinlo> no, see, what I would like to do is to get the exact time that I've generated a block, but not at this moment, but afterwards
722 2011-06-26 10:15:35 <kinlo> and exact time = up to the second
723 2011-06-26 10:15:41 <kinlo> so now I'm using listtransactions
724 2011-06-26 10:15:51 <lfm> kino grep generated ~/.bitcoin/debug.log
725 2011-06-26 10:16:02 <sipa> debug.log doesn't contain timestamps
726 2011-06-26 10:16:11 <lfm> yes it does!
727 2011-06-26 10:16:15 <kinlo> but listtransactions shows the time in the header from the block, which is incorrect
728 2011-06-26 10:16:24 <lfm> the generated messages have a date and time
729 2011-06-26 10:16:38 <sipa> kinlo: how so?
730 2011-06-26 10:16:45 <kinlo> lfm: 06/20/11 12:00 generated 50.00
731 2011-06-26 10:16:53 <lfm> exactly
732 2011-06-26 10:16:54 <kinlo> lfm: that does not contain seconds
733 2011-06-26 10:17:03 <kinlo> and is a hell to parse :)
734 2011-06-26 10:17:09 <sipa> why is time in the header of the block incorrect?
735 2011-06-26 10:17:11 <unclemantis> what is the neatest paper bitcoin solution you have all seen?
736 2011-06-26 10:17:23 <lfm> well you only have 1 block in 10 min, the seconds dont matter
737 2011-06-26 10:17:24 <sipa> unclemantis: bitbills ?
738 2011-06-26 10:17:51 <kinlo> sipa: your worker does getwork, it contains a timestamp, then works for minutes, submits the block with that old timestamp and then the block is found
739 2011-06-26 10:17:52 <unclemantis> i did some research and there are some homebrew solutions based on publishing the private key
740 2011-06-26 10:18:03 <lfm> the time is always correct for the person that made it! (not)
741 2011-06-26 10:18:18 <kinlo> so you have a delay, and I don't know how long a worker can do on a block, that might be a long time
742 2011-06-26 10:18:40 <lfm> the year went from 1999 to 19100
743 2011-06-26 10:18:42 <sipa> kinlo: minutes :o
744 2011-06-26 10:18:51 <unclemantis> egad!
745 2011-06-26 10:18:54 <sipa> you should get new work every few seconds
746 2011-06-26 10:18:56 <unclemantis> I'm late!
747 2011-06-26 10:19:14 <kinlo> sipa: can I trust this to be correct? :)
748 2011-06-26 10:19:28 <sipa> how do you mean?
749 2011-06-26 10:19:49 <lfm> does anyone really know what time it is?
750 2011-06-26 10:20:03 <lfm> does anybody care?
751 2011-06-26 10:20:08 <sipa> 12:20 GMT
752 2011-06-26 10:20:09 <kinlo> if you do not trust your workers, can you trust the time from those workers?
753 2011-06-26 10:20:20 <sipa> kinlo: workers do not set the time themselves, the pool does
754 2011-06-26 10:20:33 <sipa> unless using X-Roll-Ntime
755 2011-06-26 10:20:44 <lfm> well the workers should be allowed to change the time
756 2011-06-26 10:20:56 <sipa> yes, that's what X-Roll-Ntime is for
757 2011-06-26 10:21:18 <lfm> sipa huh? why cant they just change it?
758 2011-06-26 10:21:32 <sipa> some pools don't accept client-modified ntime
759 2011-06-26 10:21:41 <lfm> so they are wrong
760 2011-06-26 10:21:43 <sipa> i agree clients should be allowed to
761 2011-06-26 10:21:58 <sipa> but it's up to pools to decide what they accept, no?
762 2011-06-26 10:22:10 <lfm> ya, but they are wrong
763 2011-06-26 10:22:16 <sipa> they're not wrong
764 2011-06-26 10:22:25 <sipa> they may be stupid
765 2011-06-26 10:22:47 <lfm> theyre changing the getwork protocol that the standard client uses
766 2011-06-26 10:23:19 <sipa> i never read a specification of getwork that states what or what not clients are allowed to change
767 2011-06-26 10:23:24 <lfm> so they're incompatible with the standard, I call that wrong
768 2011-06-26 10:23:40 <lfm> it is implied standard
769 2011-06-26 10:23:46 <sipa> i disagree :)
770 2011-06-26 10:23:46 <unclemantis> either an rpc client like my bank uses, or a bitcoin app on the phone iteself
771 2011-06-26 10:24:18 <lfm> sipa ok so you are wrong too then
772 2011-06-26 10:25:06 <sipa> afaik no miner except poclbm in very early versions did change ntime on itself, and that was changed because pools didn't support it
773 2011-06-26 10:25:19 <sipa> without any written specification of the standard, it is defined by its common usage
774 2011-06-26 10:25:34 <lfm> diablo miner changed the time in some early version too
775 2011-06-26 10:25:41 <sipa> that's possible
776 2011-06-26 10:25:48 <lfm> it is an unwriten standard
777 2011-06-26 10:25:53 <lfm> de-facto
778 2011-06-26 10:26:37 <sipa> de-facto = from the facts :)
779 2011-06-26 10:27:01 <lfm> when there is differences between some yob of a pool operator and the "standard" bitcoin the standard defines the standard by definition