1 2013-09-20 00:04:20 <longcat> eligius or p2pool?
  2 2013-09-20 00:04:43 <wizkid057> eligius, ofc
  3 2013-09-20 00:04:51 <forrestv> p2pool
  4 2013-09-20 00:05:00 <gmaxwell> If you'd even consider using p2pool you should be using p2pool.
  5 2013-09-20 00:05:00 <wizkid057> no u!
  6 2013-09-20 00:05:46 <gmaxwell> wizkid057: go implement coinbase only mining with luke and you can have dibs on my first pass recommendations again. :P
  7 2013-09-20 00:06:08 <Luke-Jr> …
  8 2013-09-20 00:06:12 <Luke-Jr> this is #bitcoin-dev
  9 2013-09-20 00:06:26 <Luke-Jr> not #poolfight
 10 2013-09-20 00:06:29 <wizkid057> lol
 11 2013-09-20 00:18:41 <longcat> i am using p2pool but I didnt really do much consideration beforehand.  It seems the anti pool hopping mechanism hurts me when i want to use my 2x7990 to actually...  play some slideshows/games (ugh, amd drivers...)
 12 2013-09-20 00:19:15 <gmaxwell> uh. there is no anti-poolhopping that can hurt you in p2pool. It's simply immune to hopping.
 13 2013-09-20 00:19:40 <Luke-Jr> it is higher variance, that "hurts" hoppers (intentional or not) short-term
 14 2013-09-20 00:24:45 <gmaxwell> Luke-Jr: huh? that doesn't have anything to do if you're hopping or not.  Your return variance and expected return is the same regardless if you bunch up your work or spread it out.
 15 2013-09-20 00:26:05 <Luke-Jr> gmaxwell: 1 Th over 1 day is lower variance than 1 Th over 2 days
 16 2013-09-20 00:26:12 <Luke-Jr> measured in time
 17 2013-09-20 00:26:20 <gmaxwell> (there are pools that explictly penalize hoppers though)
 18 2013-09-20 00:27:43 <gmaxwell> Luke-Jr: I have no clue what you're saying there. If you are going to submit X shares there is a certian distribution of rewards you might recieve for those X shares, it doesn't matter when you submit them, the distribution is the same.
 19 2013-09-20 00:28:39 <Luke-Jr> gmaxwell: the same distribution across 1 day vs across 2 days is the same, yes, but the variance per day is higher
 20 2013-09-20 00:29:08 <gavinandresen> #@)*)(*(*&&*(#$)_:  Any advice on debugging this pull-tester-machine-upgrade problem:  https://gist.github.com/gavinandresen/6631723
 21 2013-09-20 00:30:02 <gmaxwell> Ok. This is still faires dancing on pinheads in any case. There are pools that seriously nail your expected income for "hopping" (though I think thats scummy)
 22 2013-09-20 00:30:20 <Neozonz> Disc|could someone help me out with the bitcoin php class?
 23 2013-09-20 00:30:34 <Neozonz> Disc|trying to use send many, but api doesn't seem to like my input
 24 2013-09-20 00:30:36 <Neozonz> Disc|$bitcoin->sendmany("", $QueuePayment);
 25 2013-09-20 00:30:49 <gmaxwell> gavinandresen: thats really useless looking.
 26 2013-09-20 00:30:52 <Neozonz> Disc|$QueuePayment is an array with address and amount
 27 2013-09-20 00:31:21 <gavinandresen> gmaxwell: I agree.  I ran dist-upgrade, there are no held packages....
 28 2013-09-20 00:32:08 <gmaxwell> http://askubuntu.com/questions/124845/eerror-pkgproblemresolverresolve-generated-breaks-this-may-be-caused-by-hel  not very helpful
 29 2013-09-20 00:32:32 <Luke-Jr> gavinandresen: personally, I always "upgrade" binary distros by a clean install and reconfig
 30 2013-09-20 00:33:43 <gavinandresen> gmaxwell: the dist-upgrade/apt.log file has … stuff…. in it....
 31 2013-09-20 00:36:58 <cfields> gavinandresen: i would guess jenkins is the problem there, that's the only 3rd-party repo i see, at a quick glance
 32 2013-09-20 00:37:24 <gavinandresen> mmm.  will commenting that out of sources.list work?
 33 2013-09-20 00:37:54 <gavinandresen> … and by "work", I mean can I comment it out, upgrade, and then not have horrible terrible things happen if I try to put it back?
 34 2013-09-20 00:38:21 <cfields> if that is indeed the problem, the jenkins repo has probably installed something that collides with the update
 35 2013-09-20 00:38:31 <gavinandresen> okey dokey
 36 2013-09-20 00:39:02 <cfields> so you'd want to check to see which packages have come from there
 37 2013-09-20 00:39:14 <gavinandresen> great!  How do I do that?
 38 2013-09-20 00:39:31 <gavinandresen> And once I do that…. then what?
 39 2013-09-20 00:41:01 <cfields> heh, i'm googling  :)
 40 2013-09-20 00:46:34 <Neozonz> Disc|is there a way to sendmany without a fromaccount?
 41 2013-09-20 00:47:47 <gavinandresen> Neozonz|Disc: no
 42 2013-09-20 00:48:21 <gmaxwell> Neozonz|Disc: ask what you're trying to accomplish not how you think you need to accomplish it.
 43 2013-09-20 00:49:09 <Neozonz> Disc|I am attempting to code a function that sends many miners payments as one transaction via the sendmany feature
 44 2013-09-20 00:49:36 <gmaxwell> okay, so use the empty string as your account, as thats the default account.
 45 2013-09-20 00:49:53 <gavinandresen> … and listaccounts will tell you if you have balances assigned to any other accounts
 46 2013-09-20 00:50:00 <Neozonz> Disc|I've been trying, but it seems to still reply with "sendmany requires an account"
 47 2013-09-20 00:50:02 <gavinandresen> if you do, you can use move to reassign them
 48 2013-09-20 00:50:22 <Neozonz> Disc| $bitcoin->sendmany('', $QueuePayment);
 49 2013-09-20 00:51:28 <Neozonz> Disc|I am using the following class -> https://github.com/mikegogulski/bitcoin-php/blob/master/src/bitcoin.inc
 50 2013-09-20 00:52:00 <gavinandresen> that's a bug in line 952 of that php code.  An empty account is fine.
 51 2013-09-20 00:52:48 <gavinandresen> should only error if account is null (I forget the weird php null versus empty-string versus false rules)
 52 2013-09-20 00:53:28 <Neozonz> Disc|ah ok, thank you so much
 53 2013-09-20 00:54:14 <cfields> gavinandresen: i really can't see the problem, sorry
 54 2013-09-20 00:55:24 <gavinandresen> okey dokey… I'll scroll back and find that java7 ppa and see if that works....
 55 2013-09-20 01:02:57 <Neozonz> Disc|am i required to do a json_encode , example $bitcoin->sendmany('', json_encode($QueuePayment));
 56 2013-09-20 01:09:30 <gavinandresen> Pull Tester is still Experiencing Technical Difficulties.  I'm giving up trying to fix it for now.
 57 2013-09-20 01:19:37 <BlueMatt> gavinandresen: yea, I think I can get a java6 version
 58 2013-09-20 01:19:39 <BlueMatt> give me a minute
 59 2013-09-20 01:20:24 <gavinandresen> BlueMatt: thanks!
 60 2013-09-20 01:24:28 <Neozonz> Disc|i give up on sendmany
 61 2013-09-20 01:24:34 <Neozonz> Disc|now im getting Didn't receive 200 OK from remote server.
 62 2013-09-20 01:26:34 <gavinandresen> somebody should implement a -debug=rpc that spits out exactly the JSON received….
 63 2013-09-20 01:26:53 <Neozonz> Disc|+1
 64 2013-09-20 01:31:06 <Neozonz> Disc|enabled the debugger on bitcoin php class
 65 2013-09-20 01:31:07 <Neozonz> Disc|and got this
 66 2013-09-20 01:31:07 <Neozonz> Disc|{&quot;result&quot;:null,&quot;error&quot;:{&quot;code&quot;:-1,&quot;message&quot;:&quot;value is type str, expected real&quot;},&quot;id&quot;:null}
 67 2013-09-20 01:32:57 <gavinandresen> Ah.  You're passing numbers as strings instead of numbers.
 68 2013-09-20 01:33:27 <gavinandresen> Cast the amounts to float (or real or whatever php calls them)
 69 2013-09-20 01:34:07 <Neozonz> Disc|i did though :o
 70 2013-09-20 01:34:25 <Neozonz> Disc| doubleval($balance - $fees);
 71 2013-09-20 01:34:58 <gmaxwell> in python people often get tripped up because it needs a dict not an array.
 72 2013-09-20 01:35:01 <BlueMatt> gavinandresen: try the version in git
 73 2013-09-20 01:35:51 <gavinandresen> BlueMatt: will do....
 74 2013-09-20 01:38:24 <gavinandresen> BlueMatt: while I'm copying things… can I get rid of /mnt/test-scripts/FullBlockTestGenerator.java?  And any reason for the name change of BitcoinjBitcoindComparisonTool versus just BitcoindCompar....
 75 2013-09-20 01:38:42 <BlueMatt> that is the java file which is there purely so people can see the source
 76 2013-09-20 01:38:44 <BlueMatt> its out of date though
 77 2013-09-20 01:38:48 <BlueMatt> so...I dunno, remote it I guess
 78 2013-09-20 01:38:58 <BlueMatt> (/mnt/test-scripts is web-accessible)
 79 2013-09-20 01:39:20 <gavinandresen> I'll remove it.
 80 2013-09-20 01:39:45 <Neozonz> Disc|play nice with me rpc!
 81 2013-09-20 01:39:47 <Neozonz> Disc|>_<
 82 2013-09-20 01:39:51 <BlueMatt> gavinandresen: its still useful
 83 2013-09-20 01:39:59 <BlueMatt> the b** identifiers you can lookup in it should still be right
 84 2013-09-20 01:40:12 <gavinandresen> BlueMatt: where is the up-to-date version?
 85 2013-09-20 01:40:44 <BlueMatt> https://code.google.com/r/bluemattme-bitcoinj/source/browse/core/src/test/java/com/google/bitcoin/core/FullBlockTestGenerator.java?name=blocktester
 86 2013-09-20 01:40:52 <BlueMatt> https://bluemattme-bitcoinj.googlecode.com/git-history/blocktester/core/src/test/java/com/google/bitcoin/core/FullBlockTestGenerator.java <-- raw file
 87 2013-09-20 01:41:22 <gavinandresen> BlueMatt: thanks
 88 2013-09-20 01:42:46 <gavinandresen> BlueMatt: while I have you… is patch.txt needed now that we have -regtest mode?
 89 2013-09-20 01:43:07 <BlueMatt> shouldnt be
 90 2013-09-20 02:43:49 <BlueMatt> gavinandresen: sorry...thats what I get for not checking if there are errors...
 91 2013-09-20 02:43:54 <BlueMatt> gavinandresen: new version uploaded
 92 2013-09-20 03:07:09 <gavinandresen> cfields: still around?
 93 2013-09-20 03:07:20 <cfields> yep
 94 2013-09-20 03:07:33 <gavinandresen> New pull-tester error; ./configure not found in the linux-build directory
 95 2013-09-20 03:07:43 <gavinandresen> … but it is in linux-build/$DISTDIR
 96 2013-09-20 03:08:13 <cfields> hmm, got a log?
 97 2013-09-20 03:08:25 <gavinandresen> cfields: http://jenkins.bluematt.me/pull-tester/03572435afc6c1f9a78fb5acdf4815f32adcbbf0/
 98 2013-09-20 03:12:31 <cfields> gavinandresen: is it running the .py in git?
 99 2013-09-20 03:12:49 <gavinandresen> cfields: slightly different version.  Why?
100 2013-09-20 03:13:15 <gavinandresen> cfields: I don't see how there would be a ./configure in the linux-build directory….
101 2013-09-20 03:13:19 <cfields> just trying to understand what's changed. not sure how it could end up a layer too deep that way
102 2013-09-20 03:13:26 <gavinandresen> Looking at build-tests.sh
103 2013-09-20 03:13:52 <gavinandresen> (but I don't know what make distdir does)
104 2013-09-20 03:14:25 <cfields> gavinandresen: make distdir makes a bootstrapped dir named bitcoin-0.8.99. meaning no ./autogen.sh is needed in it
105 2013-09-20 03:14:46 <cfields> we then rename that to linux-build, and do a fresh build in there
106 2013-09-20 03:15:32 <gavinandresen> cfields: ah, move-as-rename, not move as move-into-the-existing-directory-called-linux-build
107 2013-09-20 03:16:19 <cfields> gavinandresen: correct. looks like it's very likely that it's doing the latter though
108 2013-09-20 03:16:34 <cfields> i suppose it could do that if the dir wasn't wiped
109 2013-09-20 03:17:27 <gmaxwell> gavinandresen: oops. 92f2c1fe Use boost::asio::deadline_timer for walletpassphrase timeout  breaks using walletpassphrase in the debug console on a node that doesn't have the rpc running.
110 2013-09-20 03:18:35 <gavinandresen> gmaxwell: ah… that makes sense.
111 2013-09-20 03:22:09 <cfields> gavinandresen: yea, i think that's what's happening. a bit hard to follow that py though
112 2013-09-20 03:22:19 <cfields> was it changed in a way that old builds could be hanging around?
113 2013-09-20 03:22:38 <gavinandresen> checkout_pull does:     run("rm -rf ${BUILD_DIR}")
114 2013-09-20 03:22:50 <gavinandresen> BUILD_DIR is the chroot-environment 'bitcoin' directory....
115 2013-09-20 03:23:07 <gavinandresen> I'm running through those manually, will see what I get.
116 2013-09-20 03:31:40 <gavinandresen> bah, I'm befuddled….
117 2013-09-20 03:33:10 <cfields> anything interesting?
118 2013-09-20 03:33:27 <gavinandresen> successfully building master in the chroot environment...
119 2013-09-20 03:34:53 <gavinandresen> … and make check is running the block tester nicely.
120 2013-09-20 03:35:44 <gavinandresen> I have no idea how pull-tester.py might start with a linux-build.  It removes the entire build directory before git clone / git merge / etc
121 2013-09-20 03:36:16 <gavinandresen> If there was more than one copy of it running at a time that would cause issues, but I wasn't doing that.
122 2013-09-20 03:36:19 <BlueMattBot> Project Bitcoin build #420: FAILURE in 22 sec: http://jenkins.bluematt.me/job/Bitcoin/420/
123 2013-09-20 03:36:30 <BlueMatt> :(
124 2013-09-20 03:36:50 <gavinandresen> BlueMatt: the jenkins build bot uses a completely different environment from the pull-tester, right?
125 2013-09-20 03:37:28 <BlueMatt> its not in the chroot, but it (used to) be the same set of scripts or so
126 2013-09-20 03:37:31 <cfields> gavinandresen: would it halt on failure of an rf -rf ?
127 2013-09-20 03:37:40 <BlueMatt> -f implies no
128 2013-09-20 03:37:49 <cfields> any chance it's running but not getting fully nuked, i mean
129 2013-09-20 03:38:04 <BlueMatt> permissions error?
130 2013-09-20 03:38:18 <gavinandresen> runs as root, shouldn't be any permissions errors
131 2013-09-20 03:38:27 <cfields> yea, that's what i was thinking
132 2013-09-20 03:38:33 <BlueMatt> gavinandresen: jenkins runs /mnt/test-scripts/build-script.sh still
133 2013-09-20 03:38:42 <BlueMatt> gavinandresen: maybe that needs fixed, I dunno how the new system works?
134 2013-09-20 03:38:47 <cfields> mainly that suspicious lock file: http://jenkins.bluematt.me/pull-tester/03572435afc6c1f9a78fb5acdf4815f32adcbbf0/bitcoin/linux-build/
135 2013-09-20 03:39:18 <gavinandresen> BlueMatt: fixing jenkins is on the list, after fixing pull-tester and getting it building win32 again
136 2013-09-20 03:39:25 <BlueMatt> ahhh
137 2013-09-20 03:39:29 <BlueMatt> so...things are borked
138 2013-09-20 03:39:45 <gavinandresen> BlueMatt: can you turn off jenkins for now?  I want to eliminate variables....
139 2013-09-20 03:40:20 <BlueMatt> gavinandresen: paused the polling so it wont build
140 2013-09-20 03:40:34 <gavinandresen> BlueMatt: thanks
141 2013-09-20 03:40:52 <BlueMatt> (go into configure and set it to poll scm if you want to reenable it this weekend into next week - Ill be gone for all of it)
142 2013-09-20 03:41:18 <gavinandresen> my brain is full… I just went to add a way to trace what pull-tester.py does and found I already done did that....
143 2013-09-20 03:44:17 <cfields> BlueMatt: does the checker use an h2 database?
144 2013-09-20 03:44:36 <gavinandresen> cfields: still failing, here are the commands that pull-tester.py is running:  https://gist.github.com/gavinandresen/6633037
145 2013-09-20 03:45:10 <BlueMatt> cfields: ummm...I think so, yes
146 2013-09-20 03:53:51 <cfields> gavinandresen: hmm, am i missing something? looks like it doesn't cd into /mnt/chroot-tmp/mnt/bitcoin after cloning it?
147 2013-09-20 03:54:21 <cfields> or if it's already in there, it'd be in a weird state if you rm -rf $(PWD) && mkdir -p $(PWD)
148 2013-09-20 03:54:35 <gavinandresen> cwd=build_dir
149 2013-09-20 03:54:42 <gavinandresen> … arguments to the run command
150 2013-09-20 03:55:37 <gavinandresen> what IS that 1.lock.db file in linux-build.....
151 2013-09-20 03:55:46 <cfields> i'm just looking at the output you pasted
152 2013-09-20 03:56:12 <Neozonz> Disc|hmmmmmmmmmmmmmmmmmmmm
153 2013-09-20 03:56:22 <BlueMatt> mmmmmmmmmmmmm
154 2013-09-20 03:56:25 <gavinandresen> cfields: mmm.  I should improve the TRACE code so it tells the directory if cwd is set....
155 2013-09-20 03:56:26 <cfields> it's a h2 database lock file, presumably created by the block checker
156 2013-09-20 03:56:50 <BlueMatt> cfields: h2 lock files come in the form dbname.h2.lock iirc
157 2013-09-20 03:57:24 <cfields> BlueMatt: just going by the layout of the lock-file itself
158 2013-09-20 03:57:35 <BlueMatt> ahh, well if it says hw in it, sure
159 2013-09-20 03:57:42 <BlueMatt> h2
160 2013-09-20 03:58:25 <cfields> contents: http://pastebin.com/raw.php?i=6Jw7NRHG
161 2013-09-20 04:00:24 <cfields> gavinandresen: ah, i see what you mean about cwd=build_dir. i was missing that part for sure
162 2013-09-20 04:00:37 <gavinandresen> ok, something weird is happening.
163 2013-09-20 04:01:00 <gavinandresen> I rm -rf linux-build and it immediately comes back
164 2013-09-20 04:01:22 <Neozonz> Disc|i hate you bitcoin rpc >_<
165 2013-09-20 04:01:28 <Neozonz> Disc|sendmany still doesn't work...
166 2013-09-20 04:01:36 <gavinandresen> … so something is running on the pull-tester machine that is recreating that directory.
167 2013-09-20 04:01:41 <cfields> gavinandresen: sounds like you've got a process constantly remaking that lock
168 2013-09-20 04:02:09 <gavinandresen> cfields: ah, there is an old java BitcoindComparisonTool running
169 2013-09-20 04:02:55 <Neozonz> Disc|{
170 2013-09-20 04:02:56 <Neozonz> Disc|  1
171 2013-09-20 04:02:56 <Neozonz> Disc|&quot;method&quot;: &quot;sendmany&quot;,
172 2013-09-20 04:02:56 <Neozonz> Disc|&quot;params&quot;: [
173 2013-09-20 04:02:56 <Neozonz> Disc|  &quot;&quot;,
174 2013-09-20 04:02:56 <Neozonz> Disc|  &quot;{\&quot;1Pg6qB5gfibViLHpSxRD6UUMyVXhbTPgBe\&quot;:49.4}&quot;,
175 2013-09-20 04:02:58 <Neozonz> Disc|],
176 2013-09-20 04:03:00 <Neozonz> Disc|&quot;id&quot;: null
177 2013-09-20 04:03:02 <Neozonz> Disc|}
178 2013-09-20 04:03:04 <Neozonz> Disc|crap
179 2013-09-20 04:03:06 <Neozonz> Disc|sorry
180 2013-09-20 04:03:23 <cfields> gavinandresen: hmm, that's annoying.
181 2013-09-20 04:03:48 <cfields> BlueMatt: can you see if it's possible to have that lock created elsewhere?
182 2013-09-20 04:04:13 <gavinandresen> BlueMatt: s'all right, lock was from the java BitcoindComparisonTool running
183 2013-09-20 04:04:45 <Neozonz> Disc|anyone wanna help me with my sendmany problem :<
184 2013-09-20 04:04:49 <gavinandresen> I killed that old process and should be better.
185 2013-09-20 04:07:21 <gavinandresen> Neozonz|Disc: so that would be:  {"method":"sendmany", "params": ["", "{"1PG…":49.4",]     There should be no " before the { inside params.
186 2013-09-20 04:07:43 <gavinandresen> You're passing a string for the {address:amount} list, instead of an Object
187 2013-09-20 04:09:03 <Neozonz> Disc|orz
188 2013-09-20 04:09:08 <Neozonz> Disc|my savior
189 2013-09-20 04:09:13 <Neozonz> Disc|lemmie go fix that up
190 2013-09-20 04:09:46 <gavinandresen> keeping my fingers crossed that the eleventh time will be the charm….
191 2013-09-20 04:11:13 <gavinandresen> also wondering how make check could be taught to kill the java process if something goes haywire
192 2013-09-20 04:11:50 <gavinandresen> … and then thinking that all of this might have taken less time if I'd written a gitian script to do the pull-testing and installed lxc on the pull-tester machine so it could gitian build things.....
193 2013-09-20 04:18:04 <cfields> gavinandresen: it killed the process originally but you asked for it to be removed. To add it back just set TIMEOUT= in the check-local
194 2013-09-20 04:19:19 <gavinandresen> "set TIMEOUT= in the check-local" …..  not following.
195 2013-09-20 04:19:41 <cfields> jeez irc makes me sound dickish. the above should read: it honors a timeout env var if there's one set
196 2013-09-20 04:19:51 <gavinandresen> what is "it" ?
197 2013-09-20 04:20:05 <gavinandresen> Remember, my brain is full....
198 2013-09-20 04:20:44 <cfields> gavinandresen: in Makefile.am: @TIMEOUT=15 qa/pull-tester/run-bitcoind-for-test.sh $(JAVA) ...
199 2013-09-20 04:21:17 <gavinandresen> that is 15 seconds?  I think the tests take longer than that.
200 2013-09-20 04:22:33 <cfields> gavinandresen: https://github.com/bitcoin/bitcoin/blob/master/Makefile.am#L121 . See line 121 and 149
201 2013-09-20 04:22:37 <gavinandresen> In any case, seems like setting a timeout will just make problems harder to debug in the future.  "It's broken! debug debug… wait, no, it works…"
202 2013-09-20 04:22:58 <cfields> yes, that's seconds. You can add a TIMEOUT=1200 or so in the 2nd invocation if you want a watchdog for it
203 2013-09-20 04:24:52 <gavinandresen> afk for a while.
204 2013-09-20 04:28:57 <Neozonz> Disc|}
205 2013-09-20 04:28:57 <Neozonz> Disc|  ["1Pg6qB5gfibViLHpSxRD6UUMyVXhbTPgBe"]=>
206 2013-09-20 04:28:57 <Neozonz> Disc|array(1) {
207 2013-09-20 04:28:57 <Neozonz> Disc|  float(49.4)
208 2013-09-20 04:29:31 <Neozonz> Disc|seems php puts the " around string arrays...
209 2013-09-20 04:37:20 <Luke-Jr> Neozonz|Disc: why are you making a string array?
210 2013-09-20 04:42:18 <jgarzik> SD is certainly odd
211 2013-09-20 04:42:28 <jgarzik> gets sold, and traffic drops to nil
212 2013-09-20 04:42:41 <jrmithdobbs> oh?
213 2013-09-20 04:42:45 <jgarzik> certainly seems to support the notion that a lot of the traffic was generated
214 2013-09-20 04:42:54 <Neozonz> Disc|Luke-Jr, trying to sendmany with bitcoin.inc.php
215 2013-09-20 04:43:11 <Neozonz> Disc|https://github.com/mikegogulski/bitcoin-php/blob/master/src/bitcoin.inc
216 2013-09-20 05:02:34 <Neozonz> Disc|Luke-Jr, any ideas?
217 2013-09-20 05:10:54 <gmaxwell> Luke-Jr: Feel like writing a BIP? https://bitcointalk.org/index.php?topic=297904.0  seems everyone is implementing signmessage in really horribly wrong ways.
218 2013-09-20 05:13:06 <Luke-Jr> maybe
219 2013-09-20 05:19:40 <jgarzik> gmaxwell, BIP for a message format?  multibit has a PGP-like format.
220 2013-09-20 05:20:28 <gmaxwell> jgarzik: For the well established widely deployed signmessage functionality we already have, because apparently coding a var int is too hard to get right for some people. :)
221 2013-09-20 05:22:38 <gmaxwell> hm? multibit uses the normal signmessage stuff.
222 2013-09-20 05:22:50 <gmaxwell> The incompatible with everything else pgp like thing is armory.
223 2013-09-20 05:24:11 <Neozonz> Disc|does this library even work for sendmany?
224 2013-09-20 05:24:22 <Neozonz> Disc|looks like my formatting is all correct
225 2013-09-20 05:24:38 <Neozonz> Disc|but keep getting {"result":null,"error":{"code":-1,"message":"value is type str, expected obj"},"id":null}
226 2013-09-20 05:24:44 <Neozonz> Disc|driving me bonkers here
227 2013-09-20 05:34:03 <gavinandresen> BlueMatt: Block tester is getting stuck in a loop: 05:32:58 10 BitcoindComparisonTool.main: bitcoind still hasn't requested block b9 with hash 03c6e8ab3cd9c17dc694b3fb99b8b7ab5e9c8fa648a5bbd89d14d49a3f91dc9b
228 2013-09-20 05:37:07 <gavinandresen> BlueMatt: end of the debug.log:   https://gist.github.com/gavinandresen/6633672
229 2013-09-20 05:38:30 <gavinandresen> Neozonz|Disc: Your array should look like :  { "1Pg…" => 49.4 }
230 2013-09-20 05:38:48 <Neozonz> Disc|it does
231 2013-09-20 05:39:29 <gavinandresen> ok.  Then maybe your PHP-JSON library is doing the wrong thing with Objects nested in Objects, but that seems unlikely.
232 2013-09-20 05:40:48 <Neozonz> Disc|(
233 2013-09-20 05:40:48 <Neozonz> Disc|)
234 2013-09-20 05:40:48 <Neozonz> Disc|    [1PJjNS7KNV4i1FK1e9G6yk21gH9LBXViLg] => 49.4
235 2013-09-20 05:40:48 <Neozonz> Disc|Array
236 2013-09-20 05:42:23 <Neozonz> Disc|i guess the library isn't working as expected...
237 2013-09-20 05:48:18 <gavinandresen> Why is 1PJjN in an array?
238 2013-09-20 05:48:48 <Neozonz> Disc|?
239 2013-09-20 05:49:00 <gavinandresen> Neozonz|Disc:  you have an array of addresses as the key
240 2013-09-20 05:49:20 <gavinandresen> Should be Array ( "1PJjNS…" => 49.4 )
241 2013-09-20 05:49:26 <Neozonz> Disc|:O
242 2013-09-20 05:49:37 <Neozonz> Disc|oops
243 2013-09-20 05:50:25 <gavinandresen> … and if you have more than on address, Array ( "firstaddress" => amount, "secondaddress" => amount )
244 2013-09-20 05:50:37 <Neozonz> Disc|yeah i understand
245 2013-09-20 05:50:53 <Neozonz> Disc|looks like I was taking information the improper way from the first array
246 2013-09-20 05:51:12 <Neozonz> Disc|I just wanted the key coin_address, and ended up putting the array
247 2013-09-20 05:51:15 <gavinandresen> BlueMatt: … and of course NOW it is working perfectly
248 2013-09-20 05:52:29 <gavinandresen> cfields: running make -j 1 seems to be working better, I think there is some parallelism issue with the make check with blocktester rules
249 2013-09-20 06:02:42 <cfields> gavinandresen: hmm, i'll have a look
250 2013-09-20 06:03:22 <cfields> gavinandresen: any specific clashes?
251 2013-09-20 06:53:07 <Neozonz> gavinandresen, apparently [] signifies keys
252 2013-09-20 06:53:21 <Neozonz> asked some php gurus
253 2013-09-20 06:53:29 <Neozonz> this sucks... im giving up
254 2013-09-20 07:03:31 <Neozonz> i give up
255 2013-09-20 07:04:02 <Neozonz> if someone can test sendmany with bitcoin.inc.php @ https://github.com/mikegogulski/bitcoin-php/commits/master         please and thank you
256 2013-09-20 07:24:49 <warren> I'm having trouble finding the exact example that jgarzik pointed at of inline asm
257 2013-09-20 07:31:36 <warren> gmaxwell: after I have a clean generic cpuid detector, do you want me to try multiple sha256 runtimes with autodetect?
258 2013-09-20 12:16:40 <nkuttler> if i want to create a private testnet can i just change hashGenesisBlock, or what are those pchMessageStart things? or what's actually hashed?
259 2013-09-20 12:17:15 <jgarzik> warren, this?  http://gtf.org/garzik/misc/mmx-test.c
260 2013-09-20 12:18:10 <jgarzik> gmaxwell, no, multibit signs every message with PGP-like ASCII armor thingy
261 2013-09-20 12:20:48 <gmaxwell> jgarzik: hm? I went and tested it before talking to you. Plus if it weren't the standard format you couldn't use it for any of the services that need it.
262 2013-09-20 12:22:25 <jgarzik> gmaxwell, to be clear I should have said "multibit author"
263 2013-09-20 12:23:15 <gmaxwell> ah okay, so he's just using some convention for wrapping up the regular signatures? Fair enough.
264 2013-09-20 12:24:10 <jgarzik> gmaxwell, example, https://bitcointalk.org/index.php?topic=143274.msg3074097#msg3074097