1 2012-06-12 00:19:43 <BlueMatt> ok, forget all that, got the cxfreeze (equivalent of py2exe) gitian-downloader exe to download and verify a gitian zip....now to sleep
  2 2012-06-12 00:21:20 <D34TH> woot
  3 2012-06-12 01:34:05 <devrandom> BlueMatt: great!
  4 2012-06-12 04:44:58 <gribble> New news from bitcoinrss: xanatos opened issue 1442 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/issues/1442>
  5 2012-06-12 06:51:05 <gribble> New news from bitcoinrss: laanwj opened pull request 1443 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1443>
  6 2012-06-12 08:08:08 <sagar> hi
  7 2012-06-12 08:14:22 <sagar> can any1 help me?
  8 2012-06-12 08:15:00 <sipa> not unless you ask your question
  9 2012-06-12 08:15:16 <sagar> right :)
 10 2012-06-12 08:15:36 <sagar> okay...how can i create a bitcoin web client?
 11 2012-06-12 08:15:59 <sagar> I went through this link...
 12 2012-06-12 08:16:00 <sagar> https://en.bitcoin.it/wiki/API_reference_%28JSON-RPC%29
 13 2012-06-12 08:16:33 <sagar> its good, but may be not telling the sequence that I want to know..
 14 2012-06-12 08:17:00 <sagar> I know PHP and my environment is Windows
 15 2012-06-12 08:17:30 <sagar> now if any1 can tell me the steps ti implement the web client.
 16 2012-06-12 08:17:40 <sipa> well people won't do it for you
 17 2012-06-12 08:18:14 <sagar> I know, I just eant to know the actual steps I must follow..
 18 2012-06-12 08:18:26 <sagar> *want
 19 2012-06-12 08:39:30 <sipa> sagar: feel free to ask around on the forums if you want assistance
 20 2012-06-12 08:40:30 <sagar> okay, can u plz give me the URL here..
 21 2012-06-12 08:40:41 <sipa> bitcointalk.org
 22 2012-06-12 08:41:08 <sagar> thanks Sipa, hope I will get my solution there :)
 23 2012-06-12 08:41:30 <sipa> but don't expect people to do your work for you
 24 2012-06-12 08:41:40 <sipa> ask specific questions
 25 2012-06-12 09:00:59 <sagar> yes as I just want to know the ways I must, follow and the reaons of showing the error
 26 2012-06-12 09:01:16 <sagar> and here u can see my new work on bitcoin
 27 2012-06-12 09:01:24 <sagar> zenwallet.org
 28 2012-06-12 10:29:04 <Eliel_> sagar: not a good idea to show qr-code for your payment address next to qr-code that gives access to the wallet...
 29 2012-06-12 10:29:23 <Eliel_> you're supposed to dare to show the payment address qr-code to someone who wants to pay you
 30 2012-06-12 10:29:41 <Eliel_> if they can steal your wallet instead, that's not nice.
 31 2012-06-12 10:31:16 <osxorgate> yeh so i was having problems earlier with bitcointools. from what i understand from http://docs.python.org/library/bsddb.html bsddb needs a db of version <=4.7 , now since 0.6.2 bitcoin uses 4.8 i think? I'm using python 2.7 if that makes a difference
 32 2012-06-12 10:47:04 <gmaxwell> osxorgate: what are you trying to use bitcoin tools for?
 33 2012-06-12 10:47:35 <osxorgate> gmaxwell i wish to analyze the blockchain for txs going to certain addrs. the rpc won't let me do that yet
 34 2012-06-12 10:48:21 <osxorgate> yesterday i asked what my options were, bitcointools was mentioned
 35 2012-06-12 10:48:53 <osxorgate> but i'm open to other suggestions :)
 36 2012-06-12 10:49:26 <gmaxwell> In any case, it should work fine you're reading outdated docs. the python bdb will work with whatever is on your system.
 37 2012-06-12 10:51:08 <osxorgate> the thing is that i use the examples given in https://github.com/gavinandresen/bitcointools readme.txt, but when i try to run dbdump.py --block=whatever, i get an error telling me to quit running bitcoin apps (none are running)
 38 2012-06-12 10:51:20 <osxorgate> i get this on my home machine win7 x64 and this laptop winxp32
 39 2012-06-12 10:51:25 <BlueMatt> gmaxwell: afaict, thats only true if you are using the python bdb that doesnt ship with python...
 40 2012-06-12 10:53:13 <gavinandresen> osxorgate: try running bitcoin with the -detachdb option, then shutting down, then running bitcointools.  And then please report back if that fixes it
 41 2012-06-12 10:53:28 <gavinandresen> (if it does, it means I need to fix bitcointools)
 42 2012-06-12 10:53:28 <osxorgate> gavinandresen ok i will try that
 43 2012-06-12 11:26:22 <BlueMatt> devrandom: sorry, I thought cx_Freeze complaining about not having libs caused more errors than just if you try to import them, fixed that
 44 2012-06-12 11:26:46 <BlueMatt> anyway, after those two pulls, gitian-downloader works pretty well on win32 with cx_Freeze
 45 2012-06-12 11:34:44 <osxorgate> gavinandresen that seems to have fixed that error, now another which i suspect is not your fault: https://gist.github.com/511a5ac5b4a93981378c
 46 2012-06-12 11:35:52 <osxorgate> (free space is 15gig)
 47 2012-06-12 11:36:15 <gavinandresen> it is trying, and failing, to map blk0001.dat into memory, and not finding enough free RAM
 48 2012-06-12 11:36:33 <gavinandresen> (or not enough contiguous free RAM, I don't remember how WIndows does memory management)
 49 2012-06-12 11:36:50 <osxorgate> ah RAM ok that makes sense
 50 2012-06-12 11:36:54 <osxorgate> let's see if i can fix that
 51 2012-06-12 11:43:03 <gmaxwell> gavinandresen: or rather not enough address space?   Anyone know if 32 bit windows is 2+2 split?
 52 2012-06-12 11:43:47 <osxorgate> my laptop has 2 gig ram, trying now on my home machine
 53 2012-06-12 11:44:41 <gavinandresen> gmaxwell: could be not enough address space on 32-bit windows, I know etotheipi had trouble with mmap on 32-bit windows
 54 2012-06-12 11:45:03 <sipa> gmaxwell: 3/1 split iirc
 55 2012-06-12 11:48:39 <osxorgate> same error on my desktop, although it has 9 gig of ram free
 56 2012-06-12 11:49:35 <BlueMatt> can we commit to keeping a changelog url up-to-date?
 57 2012-06-12 11:50:39 <gavinandresen> git shortlog is always up-to-date
 58 2012-06-12 11:51:03 <gavinandresen> Anything that requires manually updating... will never be up-to-date.
 59 2012-06-12 11:51:25 <t7> unless you update it pre-commit
 60 2012-06-12 11:51:29 <osxorgate> so hm.. maybe i'd be better off doing this in a linux vm?
 61 2012-06-12 11:51:33 <t7> lel
 62 2012-06-12 11:51:34 <BlueMatt> gavinandresen: I mean on release
 63 2012-06-12 11:51:44 <BlueMatt> we already do them for each release in the forum announcements
 64 2012-06-12 11:51:54 <BlueMatt> just an extra copy/paste
 65 2012-06-12 11:52:38 <gavinandresen> oh, so by "we" you mean "Mr. Release Announcement Person"
 66 2012-06-12 11:52:41 <gavinandresen> (aka me)
 67 2012-06-12 11:53:12 <BlueMatt> yea
 68 2012-06-12 11:53:39 <gavinandresen> archiving the release announcements on the wiki is a great idea, "we" should do that.
 69 2012-06-12 12:10:03 <BlueMatt> wumpus: why do we not support opening the debug logfile from the gui in non-win32?
 70 2012-06-12 12:19:00 <sipa> BlueMatt: how would you do it on other OS'es?
 71 2012-06-12 12:20:02 <wumpus> BlueMatt: too much hassle, what editor would you use etc?
 72 2012-06-12 12:20:14 <BlueMatt> is there no "open file with default editor" command in qt?
 73 2012-06-12 12:20:19 <wumpus> feel free to add support for it, but I didn't feel like it
 74 2012-06-12 12:20:23 <BlueMatt> it just has to call xdg-open
 75 2012-06-12 12:20:37 <wumpus> it's so simple?
 76 2012-06-12 12:20:46 <BlueMatt> on linux, yea
 77 2012-06-12 12:20:51 <BlueMatt> osx, I have nfc
 78 2012-06-12 12:20:55 <wumpus> I imagined having to implement a 'preferred editor' setting and nonsense like that
 79 2012-06-12 12:21:05 <BlueMatt> no, xdg-open handles it
 80 2012-06-12 12:21:11 <wumpus> which is just going too deep into the rabbithole
 81 2012-06-12 12:21:13 <wumpus> let's see
 82 2012-06-12 12:21:33 <wumpus> seems to work, opens gedit for me
 83 2012-06-12 12:22:04 <BlueMatt> its part of X, so everyone's got it, essentially
 84 2012-06-12 12:22:20 <gavinandresen> on osx it is /usr/bin/open
 85 2012-06-12 12:22:30 <gavinandresen> "open files and directories"
 86 2012-06-12 12:22:38 <sipa> open exists on linux as well, afaik
 87 2012-06-12 12:23:02 <gavinandresen> open -t opens in the preferred text editor on osx
 88 2012-06-12 12:23:24 <BlueMatt> sipa: on my system, /bin/open is a link to openvt, which "start a program on a new virtual terminal (VT)."
 89 2012-06-12 12:24:05 <sipa> right
 90 2012-06-12 12:24:12 <wumpus> heh, let's just use an ifdef then
 91 2012-06-12 12:25:15 <sipa> you could turn it into a file:// url as well
 92 2012-06-12 12:25:17 <BlueMatt> wumpus: looks like QDesktopServices::openUrl works
 93 2012-06-12 12:25:18 <BlueMatt> http://www.developer.nokia.com/Community/Wiki/How_to_launch_other_applications_in_Qt
 94 2012-06-12 12:25:26 <sipa> and have it opened in a browser
 95 2012-06-12 12:25:43 <wumpus> a browser.. hmm...
 96 2012-06-12 12:27:21 <BlueMatt> gavinandresen: oh, hey, look we have a changelog page. last update: 0.3.24...
 97 2012-06-12 12:27:21 <wumpus> yeah, why not, if browsers are good with big textfiles
 98 2012-06-12 12:27:23 <BlueMatt> https://en.bitcoin.it/w/index.php?title=Changelog&action=view
 99 2012-06-12 12:27:39 <BlueMatt> wumpus: afaict QDesktopServices::openUrl will open with the default file handler
100 2012-06-12 12:28:04 <gavinandresen> BlueMatt: nice! who's the lazy sod who stopped updating it?
101 2012-06-12 12:28:26 <BlueMatt> doesnt look like any devs were ever updating it, it was "Maged"
102 2012-06-12 12:28:54 <BlueMatt> anyway, Ill go update it later today
103 2012-06-12 12:28:59 <sipa> and initially nanotube
104 2012-06-12 12:29:05 <gavinandresen> BlueMatt: thanks
105 2012-06-12 12:40:15 <wumpus> voila https://github.com/bitcoin/bitcoin/pull/1444
106 2012-06-12 12:40:48 <wumpus> simply opens gedit on my ubuntu, so I think it calls xdg-open, I wonder what it does for mac
107 2012-06-12 12:41:17 <wumpus> (I mean QDesktopServices::openUrl does)
108 2012-06-12 12:41:22 <BlueMatt> nice
109 2012-06-12 12:42:15 <gavinandresen> nanotube: There was a 20% chance I'd end up being the lazy sod and completely forgot I'd ever written a changlog wiki page :)
110 2012-06-12 12:42:49 <BlueMatt> gavinandresen: I think there was a chance it was any one of us
111 2012-06-12 12:43:21 <nanotube> gavinandresen: haha
112 2012-06-12 12:44:08 <gribble> New news from bitcoinrss: laanwj opened pull request 1444 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1444>
113 2012-06-12 12:44:11 <bulletbill> hey guys, any eta on 0.7 release (or whatever number is next up). or should i be targeting 0.6.2 for the next few months?
114 2012-06-12 12:51:31 <BlueMatt> bulletbill: Id say 0.6.2 for a few months
115 2012-06-12 12:51:35 <gavinandresen> I'd like to get a 0.7 release candidate 1 out soon.  What are you 'targeting' ?
116 2012-06-12 12:52:11 <BlueMatt> there are a lot of outstanding pulls...
117 2012-06-12 12:52:15 <gavinandresen> BlueMatt: why do you say a few months?  Too many months between releases means too many features that won't be thoroughly tested
118 2012-06-12 12:52:17 <Diapolo> gavinandresen: GUI wise I think 0.7 RC is a little too early perhaps call it beta or alpha?
119 2012-06-12 12:52:31 <Diapolo> BlueMatt: ack
120 2012-06-12 12:52:33 <BlueMatt> everything is beta ;)
121 2012-06-12 12:52:40 <Diapolo> you know what I mean ^^
122 2012-06-12 12:52:44 <gavinandresen> What's not working in the GUI right now?
123 2012-06-12 12:52:59 <BlueMatt> there are a ton of minor pulls that should be pulled or close first imho
124 2012-06-12 12:53:16 <Diapolo> the new networking options should get in, to able to control them and minor stuff
125 2012-06-12 12:53:32 <bulletbill> well, currently, we build off 0.6.2 and then merge some other branches to get the feature set we want. however, i see some of those branches have been merged post-0.6.2, so am wondering if i should wait for 0.7 or just stick to my merging strategy for the near term
126 2012-06-12 12:53:34 <gavinandresen> minor stuff we don't have to wait for, in my opinion.
127 2012-06-12 12:53:57 <gavinandresen> bulletbill: who is "we" ?  what are you doing?  (or can't you say...)
128 2012-06-12 12:54:00 <BlueMatt> Im not saying wait, Im saying stuff that pull-able now, its not worth leaving those open imho
129 2012-06-12 12:54:43 <Diapolo> gavinandresen: are you fine with language updates in RC phase? if not I say wait ^^
130 2012-06-12 12:55:01 <BlueMatt> Diapolo: in the past, translation updates have happened right up to release
131 2012-06-12 12:55:01 <gavinandresen> BlueMatt: I'm too busy working on the multisignature RPC calls to test/pull lower priority stuff
132 2012-06-12 12:55:17 <gavinandresen> Diapolo: yes, language updates can happen right up until the end
133 2012-06-12 12:55:41 <Diapolo> alright
134 2012-06-12 12:56:34 <Diapolo> at least the GUI stuff is somehow problematic I think, because laanwj is the only one, who frequently tests, comments and merges them ... if he is busy or offline this let's GUI work progress slowly
135 2012-06-12 12:58:20 <gavinandresen> GUI improvements are not a high priority for me right now.  The high priorities for the project, for me, are wallet security and network stability.
136 2012-06-12 12:58:41 <gavinandresen> Oh, and cross-implementation compatibility.
137 2012-06-12 12:58:52 <luke-jr> gavinandresen: did you see the bestblock graph?
138 2012-06-12 12:59:13 <gavinandresen> luke-jr: I don't think so
139 2012-06-12 12:59:28 <luke-jr> gavinandresen: http://luke.dashjr.org/programs/bitcoin/files/charts/bestblocks.html
140 2012-06-12 12:59:55 <luke-jr> hmm, looks a bit odd during a thousand-blocks change
141 2012-06-12 13:01:01 <kinlo> is it a known bug that the splash page of the bitcoin client on mac does not go away once the program is loaded?
142 2012-06-12 13:01:11 <sipa> can you do something like "0-10 blocks off", "11-100 blocks off", "101-1000 blocks off", ...?
143 2012-06-12 13:01:17 <Diapolo> gavinandresen: I know what you think of the GUI in terms of priority ... well I think a good GUI makes it better usable for normal users (which in the end needs a secure and robust core, this IS indeed very true).
144 2012-06-12 13:01:26 <sipa> or powers of 16 if you prefer that :)
145 2012-06-12 13:01:51 <gavinandresen> http://bitcoinstatus.rowit.co.uk/  has "Servers block chain status" charts
146 2012-06-12 13:01:53 <Diapolo> kinlo: Take a look in the Git issue tracker, I think it was mentioned somewhere there.
147 2012-06-12 13:02:00 <luke-jr> gavinandresen: gmp_bip still needs pulling before 0.7; keep in mind that until bitcoind is BIP22 compliant, most p2pool users are vulnerable to attack
148 2012-06-12 13:02:15 <sipa> what attack?
149 2012-06-12 13:02:26 <luke-jr> sipa: well, the problem is, "off" is relative :P
150 2012-06-12 13:03:01 <luke-jr> sipa: since sigops are not provided, it's possible to make p2pool overflow the sigop limit unknowingly
151 2012-06-12 13:03:49 <osxorgate> so i think bitcoin-qt.exe does not properly close down the db (-detachdb fixes that) on winxp, should i file a bug somewhere?
152 2012-06-12 13:04:08 <gavinandresen> how does that attack work?  a p2pool user connects to a rogue bitcoind?  I thought p2pool users ran their own bitcoind's....
153 2012-06-12 13:04:28 <BlueMatt> gavinandresen: you send a p2pool user's bitcoind a ton of txes, it makes invalid blocks
154 2012-06-12 13:04:31 <sipa> osxorgate: it closes it perfectly, but it doesn't detach it from the database environment
155 2012-06-12 13:04:32 <gavinandresen> osxorgate: no, that's intended behavior
156 2012-06-12 13:04:52 <sipa> so you can't move/copy thr files around
157 2012-06-12 13:05:17 <osxorgate> ah ok.. not quite sure what that means but ok ;)
158 2012-06-12 13:05:19 <helo> luke-jr: do only ~25% of nodes have the latest 1000 blocks because so many people are syncing?
159 2012-06-12 13:05:24 <sipa> does p2pool keep its own mempool?
160 2012-06-12 13:05:27 <luke-jr> gavinandresen: send a txn with a lot of P2SH-added sigops to the miners
161 2012-06-12 13:05:52 <sipa> i thought it relied om gmp for getting a transaction list?
162 2012-06-12 13:06:12 <luke-jr> sipa: it does, but bitcoind doesn't tell it what the sigops for each txn is, so it can't remove any
163 2012-06-12 13:06:22 <sipa> ?
164 2012-06-12 13:06:30 <luke-jr> well, it needs the fee info to remove; sigop info to tell when it's necessary
165 2012-06-12 13:06:33 <sipa> so it does keep its owm mempool?
166 2012-06-12 13:06:37 <gavinandresen> bitcoind won't accept non-standard transactions, so there are a max of 3 sigops per txn....
167 2012-06-12 13:06:38 <luke-jr> not afaik
168 2012-06-12 13:06:59 <luke-jr> gavinandresen: yes, it requires a lot of spam
169 2012-06-12 13:06:59 <sipa> then what is the problem?
170 2012-06-12 13:07:08 <BlueMatt> helo: unlikely that so many nodes are in ibd, and, if so, ibd is slower than though...in any case, we don't really know, which is the problem
171 2012-06-12 13:07:20 <helo> oh, i guess that would be 25% with the last ~200 blocks
172 2012-06-12 13:07:20 <sipa> gmp does not create invalid block templates, right?
173 2012-06-12 13:07:34 <luke-jr> sipa: you forget p2pool adds its own sigops
174 2012-06-12 13:07:49 <luke-jr> for the generation
175 2012-06-12 13:07:49 <sipa> right, for the coinbase
176 2012-06-12 13:08:23 <luke-jr> and in any case, I don't see any further issues with gmp_bip
177 2012-06-12 13:08:45 <gavinandresen> so the attack is:  broadcast gazillions of transactions (paying transaction fees), then hope that a p2pool user generates a block over the sigop limits?  Is that right?
178 2012-06-12 13:08:56 <luke-jr> gavinandresen: pretty much, yes
179 2012-06-12 13:09:00 <gavinandresen> meh
180 2012-06-12 13:09:26 <BlueMatt> gavinandresen: it would cost a lot, but afaict it should work pretty well
181 2012-06-12 13:09:37 <luke-jr> gavinandresen: this came to my attention because it was basically happening on Eligius
182 2012-06-12 13:09:39 <BlueMatt> (have to refresh the txes after every new block)
183 2012-06-12 13:10:20 <gavinandresen> "doctor, it hurts when I accept non-standard transactions...."
184 2012-06-12 13:10:21 <luke-jr> not sure if SatoshiDICE accidentally can trigger it or what
185 2012-06-12 13:10:40 <BlueMatt> gavinandresen: dont have to use nonstd txes
186 2012-06-12 13:11:09 <gavinandresen> ok.  Well fixing a bug in getmemorypool I'm 100% behind.
187 2012-06-12 13:11:44 <luke-jr> anyhow, gmp_bip is ready and was acked for 0.7 a while ago; so what's the hold up?
188 2012-06-12 13:11:49 <BlueMatt> (Im saying the attack should work, albeit being a bit expensive, Im not farmiliar enough with BIP22 and the pull to comment on that directly)
189 2012-06-12 13:11:53 <sipa> no problems with including sigop information in gmp, as i said
190 2012-06-12 13:12:21 <gavinandresen> luke-jr: did you respond to the comments I made about the code this morning?
191 2012-06-12 13:12:28 <sipa> luke-jr: it got even more complex since then
192 2012-06-12 13:13:22 <sipa> and as it seems some optional features are quite fundamental anyway, i would really prefer to have them as the only way
193 2012-06-12 13:13:49 <sipa> yes if you have a miner that wants to work just as if it were runnung on getwork, all these things aren't necessary
194 2012-06-12 13:13:53 <luke-jr> gavinandresen: no, I hadn't seen them. will work on that.
195 2012-06-12 13:13:54 <gavinandresen> I have a minor issue with the commit saying "BIP 22 compliances" when BIP 22 is a work in progress, and likely to change.
196 2012-06-12 13:14:02 <luke-jr> sipa: gmp_bip has not gotten any more complex, no.
197 2012-06-12 13:14:44 <sipa> luke-jr: i can't remember seeing different possible transaction decompositions in there before
198 2012-06-12 13:15:22 <luke-jr> sipa: ok, but that was to actually address a real practical bug
199 2012-06-12 13:15:33 <sipa> i know that
200 2012-06-12 13:15:46 <gavinandresen> yeah, when did the transaction decomp stuff sneak into the RPC interface? That seems like over-complication to me
201 2012-06-12 13:16:08 <sipa> but what if 10 more such "bugfixes" are necessary to make all possible combinations of features useful?
202 2012-06-12 13:16:19 <sipa> i don't claim that this is the case
203 2012-06-12 13:16:41 <sipa> but most of bip22 has not been imolemented ever on either client or server side
204 2012-06-12 13:17:03 <gavinandresen> writing specs before implementing is a really bad idea.
205 2012-06-12 13:17:32 <sipa> so if bip22 grows overly complex to be able to meet its intended use cases, i'd rather not support it at all (and that would be a pity)
206 2012-06-12 13:17:57 <luke-jr> sipa: no, most of BIP22 has been implemented.
207 2012-06-12 13:18:12 <sipa> nonce ranges?
208 2012-06-12 13:18:19 <luke-jr> I said most, not all :p
209 2012-06-12 13:18:29 <sipa> transaction shortening?
210 2012-06-12 13:18:32 <luke-jr> (nonceranges have been implemented for getwork, however)
211 2012-06-12 13:18:36 <sipa> merkle tree stuff?
212 2012-06-12 13:18:46 <sipa> server priority lists?
213 2012-06-12 13:19:03 <luke-jr> server priority lists are supported by getwork minres
214 2012-06-12 13:19:19 <kinlo> sipa: I like the fact that bip22 includes a standard for all implementations while it is not required to be implemented by all
215 2012-06-12 13:19:31 <gribble> New news from bitcoinrss: Diapolo opened pull request 1445 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1445>
216 2012-06-12 13:19:43 <luke-jr> anyhow, the topic has changed back to BIP22 itself again, which isn't the point
217 2012-06-12 13:19:50 <kinlo> sipa: for example, server priority/failovers are not to be implemented in bitcoind, ti would not make sense
218 2012-06-12 13:19:51 <sipa> no that is the point
219 2012-06-12 13:20:15 <kinlo> sipa: but it would be very interesting to have it in the standard so miners would implement it and pools would use it
220 2012-06-12 13:20:42 <sipa> i never commented on the pull request, as i don't object at all to have bitcoind compliant to an agreed standard
221 2012-06-12 13:21:03 <sipa> but right now it feels that several parts will not be implemented at all
222 2012-06-12 13:21:07 <kinlo> I do agree it is perhaps a bit too complex, but I do think the complexity might be justified...
223 2012-06-12 13:21:24 <sipa> if no server supports them, no client will be implemented that uses them
224 2012-06-12 13:21:30 <kinlo> sipa: I disagree, it's not up to bitcoind to implement them, it's up to the pools...
225 2012-06-12 13:21:53 <kinlo> sipa: and rest assured that most things will get implemented by the pool software
226 2012-06-12 13:21:54 <sipa> kinlo: i am not talking about bitcoind
227 2012-06-12 13:21:56 <kinlo> well, not all
228 2012-06-12 13:21:58 <luke-jr> bitcoind's relevance in BIP22 is the small subset that makes sense for bitcoind
229 2012-06-12 13:22:02 <kinlo> but most
230 2012-06-12 13:22:14 <kinlo> noncerange is probably no longer relevant since gpu mining
231 2012-06-12 13:22:18 <sipa> kinlo: feel free to respond on the mailing list
232 2012-06-12 13:22:24 <luke-jr> kinlo: noncerange is important for GPUs too
233 2012-06-12 13:22:33 <sipa> oh come on
234 2012-06-12 13:22:40 <kinlo> sipa: I've been pondering, but I'll have to spend some time to investigate the bip properly
235 2012-06-12 13:22:40 <luke-jr> no GPU supports anywhere near 4 GH/s
236 2012-06-12 13:23:11 <sipa> kinlo: my point exactly: nobody bothers to discuss it properly because it's too complex
237 2012-06-12 13:23:15 <kinlo> luke-jr: you don't want a getwork every second :)
238 2012-06-12 13:23:24 <luke-jr> kinlo: that's why ntime is rolled.
239 2012-06-12 13:23:42 <sipa> i see absolutely no reason for noncerange
240 2012-06-12 13:23:59 <kinlo> luke-jr: so we have ntime rolling to make sure you have a *larger* noncerange, but noncerange is an option to get a *smaller* noncerange ?
241 2012-06-12 13:24:07 <sipa> certainly not in a protocol that is designed to move block generation to the client
242 2012-06-12 13:24:10 <kinlo> sipa: it made sense for cpumining, but it doesn't do so anymore
243 2012-06-12 13:24:43 <sipa> in gmp it makes even less sense than in getwork
244 2012-06-12 13:25:00 <luke-jr> sipa: you're neither a miner developer, nor poolserver developer. I don't see how your lack of need for noncerange is relevant.
245 2012-06-12 13:25:02 <sipa> as switching to a new coinbase scriptsig is trivial
246 2012-06-12 13:25:21 <luke-jr> kinlo: ntime rolling isn't a larger nonce range, it is rolling the ntime header like you're supposed to :p
247 2012-06-12 13:25:24 <sipa> luke-jr: then can you pleasw give me a concrete use case?
248 2012-06-12 13:25:44 <luke-jr> sipa: I did in my reply to you, and BIP 22's Rationale section&
249 2012-06-12 13:25:52 <sipa> most of your arguments seem to be "some pool may like to"
250 2012-06-12 13:26:01 <luke-jr> switching to a new coinbase scriptsig is NOT ALWAYS trivial
251 2012-06-12 13:26:02 <kinlo> luke-jr: it is a way to allow fewer getworks, to re-use the same nonces
252 2012-06-12 13:27:02 <luke-jr> switching to a new coinbase scriptsig is why 10% of Eligius blocks are 1-txn
253 2012-06-12 13:28:54 <kinlo> I don't really get why you get 10% of those 1-txn's while I haven't had a single one
254 2012-06-12 13:29:49 <luke-jr> kinlo: you have less miners and more servers I think?
255 2012-06-12 13:29:50 <kinlo> there probably is a reason, but kinda surprises me, I'd expected to have several myself too
256 2012-06-12 13:30:36 <kinlo> true, my servers are better scaled, but does that change anything?  afaik all miners are sent first a longpoll with a 1-txn, then a longpoll with a full-txn
257 2012-06-12 13:31:06 <kinlo> there are only a few seconds between those 2 longpolls, or does that takes like > 1 minute for you?
258 2012-06-12 13:31:35 <kinlo> to get on average 10% of 1-txn's, shouldn't there be an average delay of exactly 1 minute?
259 2012-06-12 13:33:09 <Quaix> I was looking at my disk usage while downloading the blockchain.. I was seeing about 200-400 disk writes/sec, each write 10-20k on average. I'm concerned about this needlessly wearing out my SSD
260 2012-06-12 13:35:21 <luke-jr> kinlo: about a minute, yeah
261 2012-06-12 13:37:36 <kinlo> luke-jr: seriously?  you're running 1 minute at 100% cpu to create merkle root's?
262 2012-06-12 13:37:40 <sipa> one minute to calculate a merkle root...?
263 2012-06-12 13:38:00 <kinlo> luke-jr: how many getworks/second do you do?
264 2012-06-12 13:38:03 <luke-jr> sipa: to create a few ten-thousand
265 2012-06-12 13:38:16 <sipa> oh of course
266 2012-06-12 13:39:12 <kinlo> luke-jr: merkle root generation is single-threaded ofcourse, so you can only use one core... I've solved that with multiple backends, I can use all my cores
267 2012-06-12 13:39:37 <sipa> why is it single threaded?
268 2012-06-12 13:39:40 <kinlo> my merkle-root recalculations are all below 5 seconds
269 2012-06-12 13:39:50 <sipa> sounds perfectly parallellizable
270 2012-06-12 13:40:36 <kinlo> sipa: it should be, but afaik it is implemented as a single thread
271 2012-06-12 13:40:41 <sipa> where?
272 2012-06-12 13:40:54 <kinlo> we're discussing eloipool here
273 2012-06-12 13:40:56 <sipa> ok
274 2012-06-12 13:53:52 <Quaix> no one else finds these excessive hard disk writes when downloading the blockchain a concern?
275 2012-06-12 13:54:06 <Diablo-D3> Quaix: a lot of people do
276 2012-06-12 13:54:10 <Diablo-D3> theres work being done to fix that
277 2012-06-12 13:54:17 <Quaix> awesome
278 2012-06-12 14:02:50 <gavinandresen> Diablo-D3: who is working on further optimizing disk writes (beyond what we've already done with the bdb cache changes) ?
279 2012-06-12 14:04:00 <BlueMatt> gavinandresen: https://github.com/bitcoin/bitcoin/pull/1437#issuecomment-6274621 roughly what you wanted on the test plan?
280 2012-06-12 14:06:21 <gavinandresen> BlueMatt: yes, perfect, something a non-developer tester can run themselves
281 2012-06-12 14:15:25 <luke-jr> gavinandresen: made those changes to gmp_bip
282 2012-06-12 14:17:51 <BlueMatt> HTTP451 Unavailable for Legal Reasons...FUCK
283 2012-06-12 14:18:26 <BlueMatt> why we need such a thing is just sad
284 2012-06-12 14:18:27 <luke-jr> BlueMatt: ?
285 2012-06-12 14:18:51 <BlueMatt> http://www.tbray.org/tmp/draft-tbray-http-legally-restricted-status.html#anchor3
286 2012-06-12 14:19:53 <luke-jr> sounds reasonable
287 2012-06-12 14:20:08 <BlueMatt> reasonable to work around broken governments who dont understand the internet
288 2012-06-12 14:20:27 <luke-jr> nothing broken
289 2012-06-12 14:20:31 <luke-jr> internet is not special
290 2012-06-12 14:20:48 <BlueMatt> no, but trying to block something on it not only doesnt work, but just contributes to its use
291 2012-06-12 14:21:48 <BlueMatt> I agree illegal things on the internet shouldnt be tolerated, but govts consistently fail to understand the Streisand effect and the ability of people to easily circumvent blocks
292 2012-06-12 14:21:52 <luke-jr> blocking CP seems to work pretty well
293 2012-06-12 14:22:23 <luke-jr> ability to circumvent the rules, doesn't change the usefulness of the rules
294 2012-06-12 14:22:25 <BlueMatt> last I heard people use tor hidden services and get it
295 2012-06-12 14:22:45 <luke-jr> yes, but you don't have to worry about coming across it accidentally for example
296 2012-06-12 14:22:47 <BlueMatt> and its not like tor hidden services are easily blocked
297 2012-06-12 14:23:16 <BlueMatt> true, but I dont think thats really the point of blocking it
298 2012-06-12 14:23:23 <luke-jr> that's part of it
299 2012-06-12 14:23:32 <luke-jr> an important part IMO
300 2012-06-12 14:24:29 <BlueMatt> an important part when it comes to CP, sure, but when it comes to piracy which, lets face it, is the main target for something like this, it really doesnt help in the slightest
301 2012-06-12 14:25:27 <luke-jr> it helps establish intent
302 2012-06-12 14:25:44 <luke-jr> you can't claim you didn't know it was illegal, if you have to bypass a block
303 2012-06-12 14:25:45 <BlueMatt> for cp, sure, for piracy no
304 2012-06-12 14:26:03 <BlueMatt> I was linked to http://pirateproxy.com/goto?thepiratebay
305 2012-06-12 14:28:59 <luke-jr> This domain is for sale, call us at 1-888-694-6735 or Buy this Domain Now!
306 2012-06-12 14:29:12 <BlueMatt> an example...
307 2012-06-12 14:29:23 <drizztbsd> http://labaia.ws
308 2012-06-12 14:29:40 <BlueMatt> https://tpb.pirateparty.org.uk/
309 2012-06-12 14:29:48 <BlueMatt> run by the uk pirate party
310 2012-06-12 14:37:12 <JFK911> cool will the NBP help too
311 2012-06-12 14:37:16 <JFK911> er BNP
312 2012-06-12 14:41:20 <Diablo-D3> 02:50] <gavinandresen> Diablo-D3: who is working on further optimizing disk writes (beyond what we've already done with the bdb cache changes) ?
313 2012-06-12 14:41:35 <Diablo-D3> gavinandresen: dunno, sipa and gmaxwell and bluematt keep talking about stuff
314 2012-06-12 14:42:05 <gavinandresen> that's why I asked, talking isn't the same as doing, and I thought all of that work was done.
315 2012-06-12 14:42:23 <BlueMatt> disk writes? Im not sure what you're talking about?
316 2012-06-12 14:42:29 <BlueMatt> there is txindex pruning: https://github.com/bitcoin/bitcoin/pull/1405
317 2012-06-12 14:42:47 <gavinandresen> that won't cut down on disk writes, though, that would increase disk writes, yes?
318 2012-06-12 14:42:51 <BlueMatt> oh, and tx pre-loading in the first stage of parallel block checking https://github.com/TheBlueMatt/bitcoin/commit/b2c6602
319 2012-06-12 14:43:05 <BlueMatt> gavinandresen: correct, but it will decrease lookup times in txindex when finding transactions
320 2012-06-12 14:43:39 <BlueMatt> and jgarzik's stuff on splitting txindex up
321 2012-06-12 14:43:56 <BlueMatt> aside from those, Im not aware of anything
322 2012-06-12 14:44:06 <Diablo-D3> gavinandresen: I will admit, I dont quite pay attention sometimes
323 2012-06-12 14:44:16 <BlueMatt> and sipa's stuff to split off addr/wallet from bdb
324 2012-06-12 14:44:20 <BlueMatt> esp addr could help a ton
325 2012-06-12 14:44:34 <BlueMatt> was sipa working on addr split, or was that jgarzik?
326 2012-06-12 14:44:59 <gavinandresen> addr is already only written once per... (minute?  I forget)
327 2012-06-12 14:45:23 <BlueMatt> last I heard that was still a huge amount of traffic compared to writing it ourselves once per minute
328 2012-06-12 14:46:49 <gavinandresen> I like to work from hard numbers, like "doing it this way is X% faster than the old way" -- in my experience, developers have widely different opinions on what constitutes "a huge amount"
329 2012-06-12 14:47:07 <gavinandresen> "it saved 2%, that's HUGE!"
330 2012-06-12 14:47:59 <jgarzik> txindex split, when leaving it as DB_BTREE, saved 10% of space
331 2012-06-12 14:48:14 <jgarzik> otherwise untuned/unchanged from current, vis a vis db config settings
332 2012-06-12 14:49:04 <BlueMatt> https://github.com/bitcoin/bitcoin/pull/1233 used to be almost 30% time savings, but I havent checked it since redoing it, and it does offer some pretty good chances for improvements aside from just itself
333 2012-06-12 14:49:33 <BlueMatt> txindex.dat pruning saves significant space, but not really any time
334 2012-06-12 14:50:10 <BlueMatt> (1/2 of CTxIndex's could be removed, and with -autoprune you could lose almost 1/2 of the space, too)
335 2012-06-12 14:50:48 <BlueMatt> no, 1/3 space savings
336 2012-06-12 14:50:59 <BlueMatt> with more recent checkpoints you could probably get a ton more
337 2012-06-12 14:52:01 <BlueMatt> gavinandresen: what were you looking for, specifically?
338 2012-06-12 14:53:12 <gavinandresen> what do you mean, what am I looking for?
339 2012-06-12 14:53:34 <BlueMatt> "who is working on further optimizing disk writes (beyond what we've already done with the bdb cache changes) ?"
340 2012-06-12 14:53:46 <BlueMatt> were you looking for something, or just wondering what was being done?
341 2012-06-12 14:54:06 <gavinandresen> wondering if something was being done specifically to optimize for disk writes on SSD devices
342 2012-06-12 14:54:26 <gavinandresen> (because I didn't think anybody was working on that)
343 2012-06-12 14:54:40 <BlueMatt> hmm...what are you looking for us to do for ssds?
344 2012-06-12 14:54:51 <BlueMatt> aside from cutting down write volume, not sure theres much there to do?
345 2012-06-12 14:55:20 <gavinandresen> I don't really care about optimizing disk writes on SSD devices, but Diablo-D3 seemed to say that somebody else was working on that
346 2012-06-12 14:55:41 <BlueMatt> (the discussion of how much was being written to addr.dat was someone using an ssd commenting that it was writing something ugly like 10x the size of the file on every rewrite, IIRC)
347 2012-06-12 14:57:29 <Diablo-D3> sooner I get lugh done, sooner I can get other shit done
348 2012-06-12 14:59:05 <xorgate> about my issue earlier, regarding bitcointools. it seems the problem was using 32bit python, i downloaded 64bit and it works now.
349 2012-06-12 15:02:31 <Quaix> BlueMatt, I was wondering if it's possible to combine the writes into bigger ones instead of having a ton of 10-20k writes
350 2012-06-12 15:03:17 <BlueMatt> Quaix: am I correct in remembering that it was writing some large volume to addr.dat per minute, was that you?
351 2012-06-12 15:04:09 <Quaix> must've been someone else
352 2012-06-12 15:04:34 <BlueMatt> but, yea we could combine some block writing to blk0001.dat
353 2012-06-12 15:04:40 <BlueMatt> in theory, at least
354 2012-06-12 15:05:24 <Quaix> from what I understand, SSDs have to rewrite the whole 128k or 512k block or something like that each time
355 2012-06-12 15:05:59 <Quaix> not sure if SSDs optimize that somehow or not
356 2012-06-12 15:06:48 <BlueMatt> every drive does that
357 2012-06-12 15:07:24 <BlueMatt> though not every drive has a kb block size
358 2012-06-12 15:07:48 <jgarzik> most modern SSDs write anew (not rewrite same), when "updating" a block on "disk"
359 2012-06-12 16:52:35 <luke-jr> IMO, #936, #1245, #1431 should be blockers for 0.7.0 - and nice-to-have #1355, #1416, #1246, #1393, #1240, #1409
360 2012-06-12 17:02:19 <BlueMatt> theres something odd about that list...oh wait, they're all yours
361 2012-06-12 17:03:50 <BlueMatt> anyway, before we freeze, I think we need to merge a ton, there's no point stopping and letting pulls sit around and get old for no reason
362 2012-06-12 17:04:45 <BlueMatt> sipa: what scenario are you talking about at https://github.com/bitcoin/bitcoin/pull/973#issuecomment-6279391 ?
363 2012-06-12 17:05:09 <sipa> BlueMatt: you stop processing incoming data when the output buffer to a node is full, right?
364 2012-06-12 17:06:20 <sipa> the scenario is probably unrealistic enough to be a problem
365 2012-06-12 17:07:18 <BlueMatt> oh, I see what you're saying...still if we do get such an inv, we would have to send a getdata in response to do anything with it, so...
366 2012-06-12 17:07:56 <BlueMatt> and Im not sure that scenario is entirely realistic given tcp...
367 2012-06-12 17:08:00 <sipa> agree
368 2012-06-12 17:08:08 <sipa> seems to work fine, too
369 2012-06-12 17:10:38 <sipa> can someone explain me the zoo #1002, #1023, #1437 ?
370 2012-06-12 17:10:59 <sipa> outdated/superceded/dependencies/conflicts?
371 2012-06-12 17:11:32 <xorgate> is there a doc somewhere describing the internals of the bitcoind/bitcoin-qt apps? for reading up should one desire to fix some bugs maybe
372 2012-06-12 17:14:44 <BlueMatt> 1437 is the minimum required to do win32 uris, 1002 adds some checking to addresses before throwing them in gui and pops up to the user in that case (should be merged), 1023 does a few unnecessary/redundant things and a few things that are kinda just changes for the sake of change (/that were required to allow for some old stuff that was in that pull, but is no longer) but also adds a few sanity checks that may be useful
373 2012-06-12 17:14:46 <BlueMatt> sipa: ^
374 2012-06-12 17:15:21 <BlueMatt> ie IMPLEMENT_RANDOMIZE_STACK around ipc thread
375 2012-06-12 17:15:35 <sipa> ok, thanks
376 2012-06-12 17:16:32 <sipa> xorgate: wumpus maintains a doxygen of the sourcecode
377 2012-06-12 17:19:29 <xorgate> sipa thanks i'll try to find it :0
378 2012-06-12 17:19:33 <wumpus> https://dev.visucore.com/bitcoin/
379 2012-06-12 17:19:40 <xorgate> that was fast
380 2012-06-12 17:19:57 <xorgate> You don't have permission to access /bitcoin/ on this server.
381 2012-06-12 17:20:06 <wumpus> eh
382 2012-06-12 17:20:13 <wumpus> https://dev.visucore.com/bitcoin/doxygen/
383 2012-06-12 17:20:43 <xorgate> great thanks
384 2012-06-12 17:29:54 <sipa> BlueMatt: what do you think about #1271 right now?
385 2012-06-12 17:33:12 <luke-jr> BlueMatt: yeah, strange how everyone else's get merged and mine just sit there ready :/
386 2012-06-12 17:33:39 <BlueMatt> luke-jr: I dont think any pull of large size aside from -qt ones have been pulled for quite some time...
387 2012-06-12 17:33:58 <luke-jr> BlueMatt: not all those are large
388 2012-06-12 17:34:08 <luke-jr> but anyhow, the point isn't to complain, but to try to get things moving
389 2012-06-12 17:34:15 <sipa> i think rebroad also has a track record of non-merged pulls :)
390 2012-06-12 17:34:18 <BlueMatt> by large, I mean like 5 lines...
391 2012-06-12 17:34:20 <luke-jr> git master has been pretty quiet for a week or two now
392 2012-06-12 17:34:26 <luke-jr> sipa: yeah, but those have reasons :P
393 2012-06-12 17:34:53 <luke-jr> a lot of mine, I sit there talking to myself :p
394 2012-06-12 17:35:45 <sipa> i've acked 1245/1246 already!
395 2012-06-12 17:36:01 <sipa> i'm going through the list of pull requests now
396 2012-06-12 17:37:37 <luke-jr> hmm, I should update #1240 with my fix
397 2012-06-12 17:40:26 <BlueMatt> sipa: #1271 could be done cleaner, but it should work
398 2012-06-12 17:40:37 <BlueMatt> (not significantly, but slightly)
399 2012-06-12 17:40:58 <BlueMatt> oh, and it does add one debug line that gavin may not like
400 2012-06-12 17:41:10 <BlueMatt> src/net.cpp:530
401 2012-06-12 17:42:43 <BlueMatt> but it largely looks good
402 2012-06-12 18:07:12 <luke-jr> there, rebased #1240 with the fix embedded
403 2012-06-12 18:08:29 <luke-jr> sipa: btw, dooglus (coin sel refactor + coincontrol) basically told me he got tired of rebasing and didn't have the time to keep doing it
404 2012-06-12 18:27:35 <skittixch> Hey, I'm wondering if anyone can help me figure out how to get my bitcoind working again.  When I run bitcoind -daemon, it tells me it's probably already running.  I check with ps aux and I don't see any bitcoin related things.  I've inlcuded the last few lines of my debug.log file here. http://codepad.org/SAO0bd1w  The problem persists after I restart my server...any ideas?
405 2012-06-12 18:28:18 <BlueMatt> rm /home/skittixch/.bitcoin/.lock
406 2012-06-12 18:28:24 <BlueMatt> oh, nvm bind
407 2012-06-12 18:28:31 <BlueMatt> you have something else listening on 8333
408 2012-06-12 18:28:49 <BlueMatt> netstat -lnp | grep 8333
409 2012-06-12 18:29:49 <skittixch> it's thinking pretty hard about that one :\n3721521
410 2012-06-12 18:30:30 <BlueMatt> can happen if you have stale nfs mounts or something...
411 2012-06-12 18:30:54 <skittixch> it's been hanging a lot lately.  I'm not savvy enough to know how to freshen that up...
412 2012-06-12 18:31:13 <BlueMatt> try lsof -n | grep LISTEN | grep 8333
413 2012-06-12 18:31:18 <BlueMatt> might need sudo
414 2012-06-12 18:31:28 <skittixch> yeah, my ssh is hung it seems
415 2012-06-12 18:31:53 <Diapolo> those Linux stuff is magic stuff whenever I read it ^^
416 2012-06-12 18:33:16 <skittixch> did you mean that I should type "try lsof -n | grep LISTEN | grep 8333" without quotes?
417 2012-06-12 18:33:18 <skittixch> or without the try?
418 2012-06-12 18:33:23 <BlueMatt> without try
419 2012-06-12 18:33:29 <skittixch> it doesn't like it
420 2012-06-12 18:33:35 <skittixch> lsof command not found
421 2012-06-12 18:33:38 <skittixch> I'm on debian squeeze
422 2012-06-12 18:33:43 <BlueMatt> well stick with netstat then
423 2012-06-12 18:33:58 <skittixch> trying again as root
424 2012-06-12 18:34:53 <skittixch> still just hanging there
425 2012-06-12 18:35:09 <BlueMatt> odd...not sure Ive ever seen it hang with lnp
426 2012-06-12 18:35:31 <BlueMatt> do you get anything without grep?
427 2012-06-12 18:36:52 <skittixch> hangs on netstat -lnp :\n3721571
428 2012-06-12 18:37:00 <skittixch> I know I've done netstat before
429 2012-06-12 18:37:19 <BlueMatt> thats really odd...its just listening/numeric (no dns)/processes
430 2012-06-12 18:37:30 <skittixch> :\n3721581
431 2012-06-12 18:38:11 <skittixch> so just netstat -ln?
432 2012-06-12 18:38:32 <BlueMatt> yea
433 2012-06-12 18:39:22 <skittixch> heeey
434 2012-06-12 18:39:24 <skittixch> so that worked
435 2012-06-12 18:39:52 <BlueMatt> does it show anything on 8333?
436 2012-06-12 18:40:56 <skittixch> it does, but I don't see any identifying info... http://codepad.org/1ktCGBdJ
437 2012-06-12 18:41:30 <BlueMatt> yea without p its kinda useless...
438 2012-06-12 18:41:36 <BlueMatt> p doesnt work with root though?
439 2012-06-12 18:42:14 <skittixch> it hung last time I did it with both root and user
440 2012-06-12 18:42:31 <skittixch> I think...should I try again as user? or root?
441 2012-06-12 18:42:36 <BlueMatt> /proc and /sys are mounted properly?
442 2012-06-12 18:42:59 <skittixch> I'd hope so.  I'm a newbie admin.  This is my learning machine, so anything's possible
443 2012-06-12 18:43:03 <Diapolo> 7103 is that the correct number for testnet blocks?
444 2012-06-12 18:43:41 <BlueMatt> maybe "fuser 8333/tcp"
445 2012-06-12 18:44:22 <skittixch> fuser 8333/tcp seems to have hung it :(
446 2012-06-12 18:44:27 <skittixch> as user, not root
447 2012-06-12 18:44:32 <skittixch> haven't tried root yet
448 2012-06-12 18:44:36 <BlueMatt> something is kinda f'd up here...
449 2012-06-12 18:44:41 <skittixch> seems like it
450 2012-06-12 18:44:48 <BlueMatt> whats your mount line on /proc ?
451 2012-06-12 18:45:15 <skittixch> I don't follow, I'm sorry
452 2012-06-12 18:45:26 <BlueMatt> mount | grep proc
453 2012-06-12 18:46:27 <skittixch> proc on /proc type proc (rw,noexec,nosuid,nodev)
454 2012-06-12 18:46:54 <BlueMatt> what about "mount | grep sys"
455 2012-06-12 18:47:11 <skittixch> sysfs on /sys type sysfs (rw,noexec,nosuid,nodev)
456 2012-06-12 18:49:29 <skittixch> do you see anything funky about those results?
457 2012-06-12 18:49:35 <BlueMatt> nope
458 2012-06-12 18:49:41 <skittixch> well that's good at least...
459 2012-06-12 18:50:29 <BlueMatt> well Im outta ideas, dont know if you can get to listening sockets directly from /proc/pid, probably can but I dunno how
460 2012-06-12 18:50:36 <skittixch> trying netstat -lnp again, and just holding on...either it'll work, or putty will time out.  I'll let you know how it goes
461 2012-06-12 18:50:50 <skittixch> thanks for the help so far...at least I know my whole system probably isn't borked...just a crucial part of it ;)
462 2012-06-12 18:50:51 <BlueMatt> wait, putty is timing out not the command itself?
463 2012-06-12 18:50:59 <skittixch> the command itself
464 2012-06-12 18:51:03 <luke-jr> ARGH SOMEONE FIND A BLOCK
465 2012-06-12 18:51:05 <skittixch> just hangs there
466 2012-06-12 18:51:10 <BlueMatt> luke-jr: make it p2pool....
467 2012-06-12 18:51:20 <BlueMatt> skittixch: and Cntrl-C doesnt kill it?
468 2012-06-12 18:51:30 <luke-jr> BlueMatt: nah, p2pool is apparently poor at finding blocks
469 2012-06-12 18:51:39 <BlueMatt> over the past few days...bigtime
470 2012-06-12 18:51:51 <skittixch> no, nor does pause|break or ctrl d
471 2012-06-12 18:52:04 <BlueMatt> try running the command in screen?
472 2012-06-12 18:52:15 <luke-jr> BlueMatt: someone analyzed their lifetime data; if they were SMPPS like Eligius, they'd have like 3000 BTC extra credit to make up
473 2012-06-12 18:52:33 <BlueMatt> yea, the orphans are killing it
474 2012-06-12 18:52:37 <skittixch> like, not through ssh?
475 2012-06-12 18:52:46 <BlueMatt> skittixch: ssh -> screen -> shell(s)
476 2012-06-12 18:53:14 <luke-jr> BlueMatt: speaking of which, Eligius got 3 orphans in a row a few days ago :/
477 2012-06-12 18:53:28 <skittixch> I'm in ssh right now.  I've been entering commands in ssh via putty...it's the only command line access I've got at work
478 2012-06-12 18:53:56 <BlueMatt> putty -> ssh -> screen (on the remote server) -> shell(s) (on the remote server)
479 2012-06-12 18:55:07 <Diapolo> I use p2pool sometimes, and I know it was showing accepted shares, for which I never got paid ... dunno what was wrong. It worked before.
480 2012-06-12 18:55:29 <skittixch> I'm not sure if you're telling me to do something differently, or just making sure I know my terminology...
481 2012-06-12 18:56:22 <BlueMatt> Im suggesting you checkout screen and use it: http://www.kuro5hin.org/story/2004/3/9/16838/14935
482 2012-06-12 18:56:27 <BlueMatt> (its a command)
483 2012-06-12 18:58:38 <skittixch> gotcha, reading through now.  Thank BlueMatt
484 2012-06-12 19:04:41 <skittixch> fwiw I was dicking around in samba, and it seems like I left it in a "very bad inconsistent state", so that could...maybe...have something to do with something...
485 2012-06-12 19:17:35 <luke-jr> hmm
486 2012-06-12 19:17:40 <luke-jr> how do I force bitcoind to try reorging?
487 2012-06-12 19:17:52 <BlueMatt> give it a block that will trigger a reorg
488 2012-06-12 19:18:09 <luke-jr> it didn't
489 2012-06-12 19:18:12 <BlueMatt> s/will trigger a reorg/has higher height than current tip/
490 2012-06-12 19:18:16 <BlueMatt> then thats a bug ;)
491 2012-06-12 19:18:19 <luke-jr> 0000000000000379ab05f7f5495500679240545bbca74c3ae9afaf21e3403306 has just found
492 2012-06-12 19:18:31 <luke-jr> ProcessBlock: ORPHAN BLOCK, prev=0000000000000a368115
493 2012-06-12 19:19:12 <BlueMatt> and then it will request blocks up to it, did your node not respond?
494 2012-06-12 19:24:03 <luke-jr> BlueMatt: does it log?
495 2012-06-12 19:24:55 <BlueMatt> luke-jr: try -debug
496 2012-06-12 19:28:33 <luke-jr> BlueMatt: nothing new there; this isn't with master btw
497 2012-06-12 19:28:42 <BlueMatt> oh, try master
498 2012-06-12 19:28:42 <luke-jr> trying to debug the 0.4.x P2SH issue
499 2012-06-12 19:29:17 <luke-jr> would defeat the point
500 2012-06-12 19:29:33 <luke-jr> I do have pfrom->PushGetBlocks(pindexBest, GetOrphanRoot(pblock2));
501 2012-06-12 19:32:05 <luke-jr> it's being ignored by // Filter out duplicate requests
502 2012-06-12 19:36:27 <rainfly_x> If I were to ask for some technical advice, would I be interrupting?
503 2012-06-12 19:37:04 <luke-jr> no
504 2012-06-12 19:37:26 <rainfly_x> Cool.
505 2012-06-12 19:37:54 <rainfly_x> I'm working on the theory side of a project, which is gonna involve Bitcoin contracts.
506 2012-06-12 19:38:45 <rainfly_x> The overall intent of the site is to fund open source development through micropayments for specific features and bugfixes.
507 2012-06-12 19:39:08 <BlueMatt> luke-jr: did it not get a response the first time?
508 2012-06-12 19:39:27 <luke-jr> dunno
509 2012-06-12 19:39:42 <BlueMatt> that debug.log should show
510 2012-06-12 19:40:10 <rainfly_x> And since part of the idea is to make it so anyone can donate to a sort of "get-it-done" fund, without requiring technical knowledge, a big element of it would be using third party mediators.
511 2012-06-12 19:40:46 <helo> rainfly_x: bitcoin is not ideally suited for micropayments
512 2012-06-12 19:41:30 <rainfly_x> helo: how far from ideally suited? I was thinking $.75 average, would overhead eat too much of that?
513 2012-06-12 19:44:40 <helo> rainfly_x: $0.75 average is not too micro... is feasible, at least
514 2012-06-12 19:46:00 <BlueMatt> you can throw a ton of tiny transactions at bitcoin, but each block has limited free-tx space, and larger transactions (somewhat) take priority
515 2012-06-12 19:46:44 <BlueMatt> if you dont mind your donations taking a day to confirm, you probably wont have too much problem at current transaction volumes and growth for the next while (assuming we dont get another satoshidice)
516 2012-06-12 19:47:21 <rainfly_x> That sounds fine to me.
517 2012-06-12 19:49:50 <rainfly_x> Basically, what I need is some sort of contract system, where a person can pledge X amount of money to Y developer, where that developer is usually chosen by a third-party mediator Z (chosen by the pledger), but the pledger can override.
518 2012-06-12 19:50:28 <rainfly_x> And there's a deadline to become one of the developers in the eligible pool.
519 2012-06-12 19:51:22 <rainfly_x> That way we can avoid some of the lazier forms of mediator fraud and such.
520 2012-06-12 19:51:26 <luke-jr> rainfly_x: someone did something like that once
521 2012-06-12 19:51:52 <rainfly_x> luke-jr: Yeah? Do you remember what happened to/with it?
522 2012-06-12 19:52:00 <luke-jr> rainfly_x: nobody used it :P
523 2012-06-12 19:52:41 <rainfly_x> That's a shame. I would have, but I never even heard of it.
524 2012-06-12 19:52:48 <BlueMatt> it wasnt contract-based, and was bitcoin-only irrc?
525 2012-06-12 19:52:54 <BlueMatt> or are you thinking of a different one?
526 2012-06-12 19:52:58 <luke-jr> dunno, it's been a while
527 2012-06-12 19:53:01 <BlueMatt> ie for development on bitcoin
528 2012-06-12 19:54:56 <BlueMatt> in terms of doing it with bitcoin contracts, you may end up with an unacceptable amount of caveats
529 2012-06-12 19:55:12 <rainfly_x> Like what?
530 2012-06-12 19:55:13 <BlueMatt> but it could work with multisig...
531 2012-06-12 19:59:04 <BlueMatt> lemme think, try user -> multisig donate address 2/2 of User+Mediator -> second tx that the user+mediator have to agree on which pays to dev
532 2012-06-12 19:59:18 <BlueMatt> but then you need both user and mediator to agree to payout
533 2012-06-12 19:59:24 <BlueMatt> but neither can get the money back if it fails
534 2012-06-12 20:01:42 <BlueMatt> also, read https://en.bitcoin.it/wiki/Contracts
535 2012-06-12 20:02:00 <BlueMatt> written largely by TD
536 2012-06-12 20:13:52 <BlueMatt> hmm...should I put stable maintenance on en.bitcoin.it/wiki/Changelog ?
537 2012-06-12 20:14:01 <rainfly_x> BlueMatt: Already had that open in another tab :)
538 2012-06-12 20:14:42 <rainfly_x> I'll reread it a few times though... it's one of those things that take several shots before they sink into one's head.
539 2012-06-12 20:15:24 <luke-jr> BlueMatt: seems to me, it should be split into sections per branch
540 2012-06-12 20:15:32 <luke-jr> and updated :p
541 2012-06-12 20:15:45 <BlueMatt> up to you to keep it up-to-date, but Ill throw it up today...
542 2012-06-12 20:21:27 <BlueMatt> luke-jr: what are the latest versions?
543 2012-06-12 20:21:58 <luke-jr> BlueMatt: 0.4.6, 0.5.5, 0.6.0.7, and 0.6.2
544 2012-06-12 20:22:34 <luke-jr> 0.5.0.6 was the last 0.5.0.x release (it's no longer supported)
545 2012-06-12 20:22:48 <BlueMatt> wait, we had 0.5.0.X too? god this gets confusing...
546 2012-06-12 20:22:53 <luke-jr> 0.5.3.1 was a one-off
547 2012-06-12 20:23:22 <BlueMatt> what other 4-digit releases were there?
548 2012-06-12 20:24:01 <luke-jr> v0.5.0.1 v0.5.0.2 v0.5.0.3 v0.5.0.4 v0.5.0.5 v0.5.0.6 v0.5.3.1 v0.6.0.7 v0.6.2.1 v0.6.2.2
549 2012-06-12 20:24:19 <luke-jr> I think the 0.6.2.x were just Mac
550 2012-06-12 20:24:30 <BlueMatt> well this is gonna take a while
551 2012-06-12 20:24:46 <luke-jr> maybe it makes sense to just link the forum posts instead of copying them all
552 2012-06-12 20:24:50 <BlueMatt> was there no 0.4.2 release announcement?
553 2012-06-12 20:25:42 <luke-jr> BlueMatt: google can't find one
554 2012-06-12 20:25:52 <BlueMatt> hence my asking ;)
555 2012-06-12 20:25:58 <luke-jr> IIRC there was a 0.4.x I didn't bother announcing
556 2012-06-12 20:25:59 <luke-jr> so probably
557 2012-06-12 20:26:10 <BlueMatt> meh, whatever
558 2012-06-12 20:27:11 <BlueMatt> what was the last 0.4.X?
559 2012-06-12 20:30:18 <luke-jr> BlueMatt: 0.4.6
560 2012-06-12 20:30:59 <BlueMatt> I have no known forum post for 0.4.5 and 0.4.2
561 2012-06-12 20:32:18 <luke-jr> I know there must have been one for 0.4.5 since that was the BIP16 release
562 2012-06-12 20:32:58 <luke-jr> hmm
563 2012-06-12 20:33:06 <luke-jr> or not? I don't seem to even have 0.4.5 binaries
564 2012-06-12 20:33:24 <BlueMatt> Ill put never released
565 2012-06-12 20:33:35 <luke-jr> there were rcs at least
566 2012-06-12 20:33:44 <BlueMatt> well, whatever
567 2012-06-12 20:35:56 <luke-jr> amazing how long it takes to get orphan chains fixed
568 2012-06-12 20:41:23 <BlueMatt> luke-jr: I cant find forum posts or evidence of the existence of 0.5.0.2+
569 2012-06-12 20:41:30 <BlueMatt> and no forum post for 0.5.0.1 aside from a bug report
570 2012-06-12 20:41:55 <luke-jr> BlueMatt: 0.5.0.x and 0.6.0.x are short-term stable branches, and I don't usually announce them
571 2012-06-12 20:42:03 <luke-jr> ie, to use them you need to pull from git
572 2012-06-12 20:42:06 <BlueMatt> were binaries ever released?
573 2012-06-12 20:42:47 <luke-jr> 0.6.0.7rc1 and 0.6.0.8rc1 were, but Gavin asked not to have binaries for 0.6.0.7 final so 0.6.2 would get more attention
574 2012-06-12 20:45:06 <BlueMatt> meh, Im gonna ignore 0.X.Y.N entirely in changelog
575 2012-06-12 20:47:21 <luke-jr> k
576 2012-06-12 20:54:40 <bayleef> BlueMatt: BTW any ideas? Just tried verifying all blocks at level 6, still didn't fix it. It's still trying to connect blocks 000000000000052c233f..00000000000002c5eaa8 and can't verify some sig
577 2012-06-12 20:56:19 <sipa> checkblocks only verifies your current chain
578 2012-06-12 20:56:48 <sipa> it does not check for corruption in side chains that can potentially be extended to a new best chain
579 2012-06-12 20:58:36 <luke-jr> hmm
580 2012-06-12 20:59:04 <luke-jr> sipa: I have a (0.4.6) node that seems stuck, but it's actually just very slow to complete the orphan chain
581 2012-06-12 20:59:16 <luke-jr> it has all the latest unstuck branches merged
582 2012-06-12 21:00:04 <sipa> there's still a serious inefficiency in the latest unstuck thing
583 2012-06-12 21:00:24 <BlueMatt> bayleef: sorry, didnt realize what sipa just said, looks like theres no fix atm
584 2012-06-12 21:00:44 <BlueMatt> ok, wtf is up with 0.6.1? the changelog is the same as 0.6.2 and there is no 0.6.1 release announce thread
585 2012-06-12 21:00:49 <sipa> but very slow seems.strange, unless you accidentally always pick a bad peer
586 2012-06-12 21:00:58 <sipa> BlueMatt: 0.6.1 was never released
587 2012-06-12 21:01:15 <sipa> so 0.6.2 was released as successor to 0.6.0
588 2012-06-12 21:01:19 <BlueMatt> ahh...well probably wanna delete it from sf...http://sourceforge.net/projects/bitcoin/files/Bitcoin/bitcoin-0.6.1/
589 2012-06-12 21:03:06 <bayleef> BlueMatt: Ah... So I just wait for it to fix itself? Or something
590 2012-06-12 21:03:37 <BlueMatt> bayleef: yea, wait for a fix...for now, just delete blockchain...
591 2012-06-12 21:03:46 <BlueMatt> ok, I think the changelog is  up-to-date
592 2012-06-12 21:04:55 <bayleef> BlueMatt: Again? Just did that about a week ago, and got this. Got it the first time I downloaded the blockchain, too. Except that time I just grabbed an archive lol
593 2012-06-12 21:05:38 <BlueMatt> bayleef: I know, and Im really curious as to why this has happened to you, sure you memory/disk/processor/etc are ok?
594 2012-06-12 21:06:57 <bayleef> BlueMatt: Nothing else seems wrong, no IO errors or sudden lockups. Beyond that though idk
595 2012-06-12 21:07:22 <gribble> New news from bitcoinrss: TheBlueMatt opened pull request 1446 on bitcoin/bitcoin <https://github.com/bitcoin/bitcoin/pull/1446>
596 2012-06-12 21:08:23 <sipa> bayleef: i see that problem from time to time
597 2012-06-12 21:08:45 <sipa> i myself probably resync from network a few times per week as benchmark
598 2012-06-12 21:08:50 <luke-jr> sipa: well, I'm trying to debug why it's getting stuck on the first P2SH block; so it thinks this chain is bad
599 2012-06-12 21:09:10 <sipa> i haven't ever encountered such a problem
600 2012-06-12 21:09:53 <luke-jr> afaict, it's a problem with the backport
601 2012-06-12 21:10:22 <bayleef> It also happened to someone on the bitcoin stackexchange, but it mysteriously fixed itself for him/her. :(
602 2012-06-12 21:10:56 <luke-jr> people who encountered it before told me they switched to 0.6 temporarily to get past it
603 2012-06-12 21:11:06 <luke-jr> the weird thing is, it doesn't seem to be deterministic
604 2012-06-12 21:11:14 <luke-jr> (I couldn't reproduce it the first try)
605 2012-06-12 21:12:10 <sipa> bayleef: we've seen several reports of people being stuck on recent versions, but everytime we ask to send us their blockchain, everything seems perfectly fine to us
606 2012-06-12 21:12:38 <bayleef> sipa: Exactly what happened with me, too
607 2012-06-12 21:12:52 <luke-jr> sipa: stuck blocks don't seem to be a regression, FWIW
608 2012-06-12 21:13:08 <luke-jr> sipa: the ratio of 0.6.x stuck seems to be the same as the ratio of 0.6.x nodes overall
609 2012-06-12 21:13:12 <luke-jr> or close anyhow
610 2012-06-12 21:13:14 <sipa> no, we've seen such reports for a long time
611 2012-06-12 21:13:36 <sipa> not saying it is a regression, but it's certainly an issue
612 2012-06-12 21:13:51 <luke-jr> sipa: is there some way I can cheat and modify my bitcoind to try reconnecting the "bad block" again without waiting for the orphan chain to finish downloading? :/
613 2012-06-12 21:14:01 <luke-jr> safely, I mean
614 2012-06-12 21:14:16 <bayleef> fwiw this has happened with 0.6.0, 0.6.2 and current master from github.com/bitcoin/bitcoin
615 2012-06-12 21:14:18 <sipa> what do you mean by orphan chain?
616 2012-06-12 21:14:22 <BlueMatt> bayleef: no, not what happened to you
617 2012-06-12 21:14:36 <BlueMatt> on your first chain, I saw the same issue you had, but it wasnt stuck, it was just reorging forever
618 2012-06-12 21:14:37 <luke-jr> http://pastebin.com/MS5yVJAE <-- last reorg attempt
619 2012-06-12 21:14:59 <BlueMatt> (ie had to reconnect several thousand blocks, so it appeared stuck since it took quite some time
620 2012-06-12 21:15:08 <sipa> luke-jr: ok, side chain
621 2012-06-12 21:15:13 <luke-jr> sipa: apparently if you get the latest block before you're done syncing, it makes the blockchain download work backward, very slow
622 2012-06-12 21:15:39 <luke-jr> or maybe it's related to the chain failing to reorg, or both
623 2012-06-12 21:15:44 <sipa> luke-jr: removing all cdiskblockindex entries from blkindex.dat that are not in the main chain, should do the trick
624 2012-06-12 21:16:09 <luke-jr> sipa: the main chain *is* the "orphan chain" here
625 2012-06-12 21:16:27 <sipa> well, the actual main chain is
626 2012-06-12 21:16:31 <luke-jr> besides, editing dbs is beyond my competence
627 2012-06-12 21:16:46 <sipa> but to that node, it is a side chain
628 2012-06-12 21:16:56 <luke-jr> hmm
629 2012-06-12 21:17:24 <luke-jr> so there's no way to have init.c do "lookup block 177618 (by hash) and try to accept it"? :P
630 2012-06-12 21:18:04 <sipa> well, obviously there is something wrong with the database, as it tries to connect that block all the time
631 2012-06-12 21:18:10 <sipa> and it fails
632 2012-06-12 21:19:31 <luke-jr> no, there is something wrong with the code probably
633 2012-06-12 21:19:55 <sipa> if you feed it to 0.6, it corrects?
634 2012-06-12 21:20:56 <luke-jr> that's what I hear
635 2012-06-12 21:21:13 <sipa> it's not particularly hard to test, is it?
636 2012-06-12 21:25:02 <luke-jr> it'd take 15 minutes, and I have no reason to doubt the two independent reports
637 2012-06-12 21:29:10 <sipa> ok
638 2012-06-12 21:33:13 <luke-jr> sipa: do I trust bug reporters too easily? XD
639 2012-06-12 21:36:46 <luke-jr> yay for decompositions
640 2012-06-12 21:37:20 <luke-jr> found the bug
641 2012-06-12 21:43:17 <luke-jr> sipa: were you going to address those bugs I found in d68dcf7 btw?
642 2012-06-12 21:44:09 <sipa> hmm, link?
643 2012-06-12 21:44:19 <sipa> (not on a pc right now)
644 2012-06-12 21:44:20 <luke-jr> I just mentioned them on IRC :P
645 2012-06-12 21:44:22 <luke-jr> 
646 2012-06-12 21:44:23 <luke-jr> 
647 2012-06-12 21:45:02 <sipa> right
648 2012-06-12 21:45:24 <sipa> yes, i should at least check what consequences that has
649 2012-06-12 21:46:05 <luke-jr> fixing this BIP16 issue, I should roll a new stable release soon; would be nice if I could include that too
650 2012-06-12 21:48:33 <luke-jr> fix for stable P2SH bug was hiding in& e679ec9 OP_EVAL implementation
651 2012-06-12 21:48:38 <luke-jr> no wonder I missed it
652 2012-06-12 21:56:16 <luke-jr> https://gitorious.org/+bitcoin-stable-developers/bitcoin/bitcoind-stable/commit/9a48f56fb02338b7f68ab9fd469abc1abe0011c3
653 2012-06-12 22:00:42 <luke-jr> Db::get: Cannot allocate memory
654 2012-06-12 22:00:47 <luke-jr> I have plenty of memory :/
655 2012-06-12 23:24:56 <luke-jr> sipa: is there any harm in setting bdb max locks/objects higher to reorg leaps without d68dcf7?