1 2015-02-06 00:25:00 <Luke-Jr> hm, GAit is not here recently? did he change his nick?
  2 2015-02-06 00:34:23 <l33thax0r> I just did some analysis on gaivns proposal to increase block sizes.   I discovered that Larger blocks will lead to less full nodes since VPS costs go up, which means more centralization and people builing software to input/output data from the blockchain with redeemscripts and P2SH.
  3 2015-02-06 00:36:02 <l33thax0r> And I strongly suggest *not* having an algorithim that increases blocksizes automatically based on the size of past blocks.
  4 2015-02-06 00:43:02 <Luke-Jr> [00:41:49] <gdoteof> qt/test/../paymentrequestplus.h:8:31: fatal error: paymentrequest.pb.h: No such file or directory
  5 2015-02-06 00:43:08 <Luke-Jr> cfields: ^ would it be hard to get a configure error?
  6 2015-02-06 00:44:19 <cfields> Luke-Jr: configure should disable qt if protobuf isn't found
  7 2015-02-06 00:44:28 <cfields> how'd you manage to get it to try?
  8 2015-02-06 00:45:45 <Luke-Jr> gdoteof in #bitcoin said it was just a normal ./configure
  9 2015-02-06 00:46:38 <cfields> is it building out-of-tree ?
 10 2015-02-06 00:47:22 <gdoteof> Luke-Jr: hi
 11 2015-02-06 00:47:25 <Luke-Jr> gdoteof: [00:46:55] <cfields> is it building out-of-tree ?
 12 2015-02-06 00:47:33 <Luke-Jr> brb
 13 2015-02-06 00:47:37 <ak_> im running a fresh default install now.
 14 2015-02-06 00:47:39 <cfields> gdoteof: hi
 15 2015-02-06 00:47:40 <gdoteof> cfields: sorry; ... idk?
 16 2015-02-06 00:47:54 <cfields> gdoteof: did you run configure from inside the bitcoin dir?
 17 2015-02-06 00:48:05 <ak_> ^^
 18 2015-02-06 00:48:09 <gdoteof> i just     sudo apt-get install libqt4-dev libprotobuf-dev protobuf-compiler
 19 2015-02-06 00:48:15 <gdoteof> and am redoing ./configure
 20 2015-02-06 00:48:24 <gdoteof> cfields: i ran it originally from the root dir
 21 2015-02-06 00:48:29 <ak_> ahhhh
 22 2015-02-06 00:48:29 <gdoteof> ./autogen
 23 2015-02-06 00:48:31 <gdoteof> ./configure
 24 2015-02-06 00:48:50 <ak_> you need to run again in bitcoin dir
 25 2015-02-06 00:49:01 <cfields> gdoteof: configure should've disabled qt if protobuf wasn't found
 26 2015-02-06 00:49:09 <ak_> you have dropped some stuff in / maybe u need to cleanup?
 27 2015-02-06 00:49:27 <gdoteof> cfields: yeah.. i don't think it did.  or, idk; i ran ./configure and then i got the error
 28 2015-02-06 00:49:32 <cfields> ak_: pretty sure he meant repo root. if he's building from /, he can help himself :)
 29 2015-02-06 00:49:38 <ak_> lol
 30 2015-02-06 00:49:41 <ak_> yup
 31 2015-02-06 00:49:45 <gdoteof> cfields: yes the approot =P
 32 2015-02-06 00:50:40 <gdoteof> is there a way to switch the sync node, if there is such a thing
 33 2015-02-06 00:50:55 <gdoteof> i'm building rc4 only because this syncing is completely unreasonable
 34 2015-02-06 00:51:42 <gdoteof> and i hear i can get faster syncing in v.10
 35 2015-02-06 00:52:45 <gdoteof> now im getting like 1 block / 30 seconds
 36 2015-02-06 00:54:51 <gdoteof> even building the client; my cpu is barely hitting 30%
 37 2015-02-06 00:55:28 <sipa> use make -j4, if you want to build using 4 cores in parallel
 38 2015-02-06 00:55:36 <sipa> by default you'll only build using one thread
 39 2015-02-06 00:56:37 <gdoteof> sipa: thanks.  im really just bitcihng about the syncing in v.9.4 being unreasonably slow
 40 2015-02-06 00:57:19 <gdoteof> the only time my network goes about 100kbs is when i get a ton of orphan blocks
 41 2015-02-06 00:57:27 <gdoteof> i cand ownload those from the network at lightning speed
 42 2015-02-06 00:58:07 <gdoteof> all i really wanted to do was connect to a faster peer
 43 2015-02-06 00:59:42 <sipa> start with -connect=IP
 44 2015-02-06 00:59:54 <sipa> that'll connect you exclusively to the peer you want to
 45 2015-02-06 01:00:04 <gdoteof> now to find a fast peer that i want
 46 2015-02-06 01:00:16 <sipa> well feel free to test 0.10 :)
 47 2015-02-06 01:00:27 <sipa> it should deal much better with this
 48 2015-02-06 01:00:29 <gaantr-mobile> I am trying to create a spend on a wallet using bitcoin 0.12 and it is telling me my balance is 0 but I have unspent outputs worth something.
 49 2015-02-06 01:00:40 <sipa> i know of no bitcoin 0.12
 50 2015-02-06 01:00:42 <gdoteof> yeah yeah :).  i will test it in a minute .  it look slike it will build correctly this time
 51 2015-02-06 01:01:01 <gaantr-mobile> sipa sorry bitcoinj
 52 2015-02-06 01:01:24 <sipa> gaantr-mobile: perhaps those outputs are not sufficiently confirmed yet?
 53 2015-02-06 01:04:59 <gdoteof> okay now this is going at a pretty good rate
 54 2015-02-06 01:05:26 <gdoteof> cpu is maxed out
 55 2015-02-06 01:05:30 <gdoteof> feels good
 56 2015-02-06 01:07:27 <fanquake> ;;blocks
 57 2015-02-06 01:07:27 <gribble> 342161
 58 2015-02-06 01:08:49 <gdoteof> well.. to whoever cares about release candidates.. when i didn't have the protodb stuff installed, ./configure did not catch it and `make` failed.  after installing the protodb dependencies it went without a hitch
 59 2015-02-06 01:10:30 <gdoteof> now i am getting 2-8 blocks a second
 60 2015-02-06 01:10:45 <fanquake> gdoteof Open an issue with all the info, logs os etc, an someone will take a look at it https://github.com/bitcoin/bitcoin/issues/new
 61 2015-02-06 01:13:13 <phantomcircuit> sipa, what autotools magic do i need to do to get libsecpk validation
 62 2015-02-06 01:14:48 <sipa> phantomcircuit: no such thing
 63 2015-02-06 01:15:05 <sipa> the USE_SECP256K1 in the bitcoin core is outdated, and should probably just be removed
 64 2015-02-06 01:16:09 <phantomcircuit> sipa, separate branch?
 65 2015-02-06 01:19:24 <sipa> phantomcircuit: i have no branch for bitcoin core that does libsecp256k1-based validation currentpy
 66 2015-02-06 01:19:37 <phantomcircuit> ah
 67 2015-02-06 01:19:37 <phantomcircuit> k
 68 2015-02-06 01:19:57 <sipa> *currently
 69 2015-02-06 02:03:12 <rgenito> nkuttler, fanquake thank you guys for the testnet coins, i really appreciate it :)
 70 2015-02-06 02:39:44 <morcos> sipa: any reason USE_SECP256K1 should be removed?  i find it very useful.  makes some of my intensive testing run twice as fast
 71 2015-02-06 02:40:35 <sipa> morcos: does it even work? :o
 72 2015-02-06 02:41:12 <morcos> it required a few small changes
 73 2015-02-06 02:41:54 <morcos> functions changed names and i had to pass different flags to some configure or start function
 74 2015-02-06 02:42:09 <sipa> right
 75 2015-02-06 02:42:33 <sipa> i just don't feel comfortable with people using it for validation now
 76 2015-02-06 02:42:51 <SubCreative> <3
 77 2015-02-06 02:43:59 <morcos> sipa: yeah, just leave it unsupported, but don't make it harder..   i figured it would only be helpful bug testing if i got some kind of error using it for validation
 78 2015-02-06 03:05:17 <Luke-Jr> morcos: if you don't want it removed, I'd suggest you submit the patches to make it at least compile?
 79 2015-02-06 03:05:28 <Luke-Jr> having code that doesn't even compile in there simply doesn't make sense
 80 2015-02-06 03:10:59 <morcos> well thats potentially what i was going to suggest, but sounds unclear if sipa would want that
 81 2015-02-06 03:13:38 <morcos> also realistically, sipa or someone more familiar with it is going to get more work as a result of that, even if its just review work
 82 2015-02-06 03:14:53 <morcos> personally i think it would be better to stay part of the code and just be warned against, since theoretically we're eventually switching to it, but i can see the argument against as well
 83 2015-02-06 04:29:39 <sipa> heh, github no longer classifies bitcoin code as typescript
 84 2015-02-06 04:29:46 <sipa> instead,  it now believes it is 95% C
 85 2015-02-06 04:29:55 <sipa> and 3.6% C++
 86 2015-02-06 04:30:04 <sipa> oh, crap
 87 2015-02-06 04:30:12 <sipa> nevermind; i was looking at libsecp256k1, not bitcoin
 88 2015-02-06 04:30:30 <sipa> it looks pretty good now
 89 2015-02-06 05:06:58 <nerdcore> I've successfully compiled bitcoind and bitcoin-cli on OpenBSD 5.6 amd64
 90 2015-02-06 05:07:39 <nerdcore> but test_bitcoin fails to link against /usr/lib/crt0.o (undefined reference to 'main')
 91 2015-02-06 05:09:07 <nerdcore> 0.9.4
 92 2015-02-06 05:10:13 <gmaxwell> nerdcore: it would probably be more useful to try 0.10rc ... just because whatever issue that is may already be fixed. (sorry, I have no idea if it is or isn't so that might be a waste of your time; but perhaps not)
 93 2015-02-06 10:59:37 <michagogo> sipa: yeah, they finally fixed that bug in Linguist that stops interpreting qt translations as typescript
 94 2015-02-06 11:03:40 <michagogo> https://github.com/github/linguist/pull/1842
 95 2015-02-06 11:24:49 <juliane_> Log Samples from Apache
 96 2015-02-06 11:24:51 <juliane_> Apache generally has two log files: apache error log and apache access log. <br /> The format of these files can be changed in the configuration, but these are<br /> the default formats:
 97 2015-02-06 11:24:53 <juliane_> Apache access log (success - code 200):
 98 2015-02-06 11:24:56 <juliane_> 192.168.2.20 - - [28/Jul/2006:10:27:10 -0300] "GET /cgi-bin/try/ HTTP/1.0" 200 3395
 99 2015-02-06 11:24:58 <juliane_> 127.0.0.1 - - [28/Jul/2006:10:22:04 -0300] "GET / HTTP/1.0" 200 2216
100 2015-02-06 11:25:00 <juliane_> Apache access log (failure - code 4xx):
101 2015-02-06 11:25:02 <juliane_> 127.0.0.1 - - [28/Jul/2006:10:27:32 -0300] "GET /hidden/ HTTP/1.0" 404 7218
102 2015-02-06 11:25:05 <juliane_> Apache unnacepted request methods (caused by TortoiseSVN):
103 2015-02-06 11:25:07 <juliane_> x.x.x.90 - - [13/Sep/2006:07:01:53 -0700] "PROPFIND /svn/[xxxx]/Extranet/branches/SOW-101 HTTP/1.1" 401 587
104 2015-02-06 11:25:09 <juliane_> x.x.x.90 - - [13/Sep/2006:07:01:51 -0700] "PROPFIND /svn/[xxxx]/[xxxx]/trunk HTTP/1.1" 401 587
105 2015-02-06 11:25:11 <juliane_> x.x.x.90 - - [13/Sep/2006:07:00:53 -0700] "PROPFIND /svn/[xxxx]/[xxxx]/2.5 HTTP/1.1" 401 587
106 2015-02-06 11:25:13 <juliane_> x.x.x.90 - - [13/Sep/2006:07:00:53 -0700] "PROPFIND /svn/[xxxx]/Extranet/branches/SOW-101 HTTP/1.1" 401 587
107 2015-02-06 11:25:15 <juliane_> x.x.x.90 - - [13/Sep/2006:07:00:21 -0700] "PROPFIND /svn/[xxxx]/[xxxx]/trunk HTTP/1.1" 401 587
108 2015-02-06 11:25:18 <juliane_> x.x.x.90 - - [13/Sep/2006:06:59:53 -0700] "PROPFIND /svn/[xxxx]/[xxxx]/2.5 HTTP/1.1" 401 587
109 2015-02-06 11:25:20 <juliane_> x.x.x.90 - - [13/Sep/2006:06:59:50 -0700] "PROPFIND /svn/[xxxx]/[xxxx]/trunk HTTP/1.1" 401 587
110 2015-02-06 11:25:22 <juliane_> x.x.x.90 - - [13/Sep/2006:06:58:52 -0700] "PROPFIND /svn/[xxxx]/[xxxx]/trunk HTTP/1.1" 401 587
111 2015-02-06 11:25:24 <juliane_> x.x.x.90 - - [13/Sep/2006:06:58:52 -0700] "PROPFIND /svn/[xxxx]/Extranet/branches/SOW-101 HTTP/1.1" 401 587
112 2015-02-06 11:25:27 <juliane_> Apache error log:
113 2015-02-06 11:25:29 <juliane_> [Fri Dec 16 01:46:23 2005] [error] [client 1.2.3.4] Directory index forbidden by rule: /home/test/
114 2015-02-06 11:25:31 <juliane_> [Fri Dec 16 01:54:34 2005] [error] [client 1.2.3.4] Directory index forbidden by rule: /apache/web-data/test2
115 2015-02-06 11:25:33 <juliane_> [Fri Dec 16 02:25:55 2005] [error] [client 1.2.3.4] Client sent malformed Host header
116 2015-02-06 11:25:35 <juliane_> [Mon Dec 19 23:02:01 2005] [error] [client 1.2.3.4] user test: authentication failure for "/~dcid/test1": Password Mismatch
117 2015-02-06 11:25:38 <juliane_> Apache error log (startup) 3 examples:
118 2015-02-06 11:25:40 <juliane_> ** Normal (v2.x)
119 2015-02-06 11:25:42 <juliane_> [Sat Aug 12 04:05:51 2006] [notice] Apache/1.3.11 (Unix) mod_perl/1.21 configured -- resuming normal operations
120 2015-02-06 11:25:44 <juliane_> [Thu Jun 22 14:20:55 2006] [notice] Digest: generating secret for digest authentication ...
121 2015-02-06 11:25:46 <juliane_> [Thu Jun 22 14:20:55 2006] [notice] Digest: done
122 2015-02-06 11:25:49 <juliane_> [Thu Jun 22 14:20:55 2006] [notice] Apache/2.0.46 (Red Hat) DAV/2 configured -- resuming normal operations
123 2015-02-06 11:25:51 <juliane_> ** Restart by HUP signal (optional suEXEC)
124 2015-02-06 11:25:53 <juliane_> [Sat Aug 12 04:05:49 2006] [notice] SIGHUP received.  Attempting to restart
125 2015-02-06 11:25:55 <juliane_> [Sat Aug 12 04:05:51 2006] [notice] suEXEC mechanism enabled (wrapper: /usr/local/apache/sbin/suexec)
126 2015-02-06 11:25:57 <juliane_> ** after 'unclean' shutdown (left over PID file)
127 2015-02-06 11:26:00 <juliane_> [Sat Jun 24 09:06:22 2006] [warn] pid file /opt/CA/BrightStorARCserve/httpd/logs/httpd.pid overwritten -- Unclean shutdown of previous Apache run?
128 2015-02-06 11:26:02 <juliane_> [Sat Jun 24 09:06:23 2006] [notice] Apache/2.0.46 (Red Hat) DAV/2 configured -- resuming normal operations
129 2015-02-06 11:26:04 <juliane_> [Sat Jun 24 09:06:22 2006] [notice] Digest: generating secret for digest authentication ...
130 2015-02-06 11:26:06 <juliane_> [Sat Jun 24 09:06:22 2006] [notice] Digest: done
131 2015-02-06 11:26:08 <juliane_> Apache error log (shutdown by TERM signal):
132 2015-02-06 11:26:11 <juliane_> [Thu Jun 22 11:35:48 2006] [notice] caught SIGTERM, shutting down
133 2015-02-06 11:26:13 <juliane_> Apache without resources:
134 2015-02-06 11:26:15 <juliane_> [Tue Mar 08 10:34:21 2005] [error] (11)Resource temporarily unavailable: fork: Unable to fork new process
135 2015-02-06 11:26:17 <juliane_> [Tue Mar 08 10:34:31 2005] [error] (11)Resource temporarily unavailable: fork: Unable to fork new process
136 2015-02-06 11:42:34 <wumpus> wth is that
137 2015-02-06 11:44:42 <waxwing> creatively miscoded botnet? :)
138 2015-02-06 12:38:01 <akstunt600> Is there a build server for windows binaries?
139 2015-02-06 12:38:35 <akstunt600> was about to make a 64 bit of rc4 for windows cuz bored and reasons.
140 2015-02-06 12:56:12 <jonasschnelli> akstunt600, looking for binaries of rc4? https://bitcoin.org/bin/0.10.0/test/
141 2015-02-06 12:57:35 <wumpus> a build server? you can follow the gitian guide (https://github.com/bitcoin/bitcoin/blob/master/doc/gitian-building.md) if you want to build releases
142 2015-02-06 13:01:23 <akstunt600> yeah looks like i will be doing something like this but with mingw and cygwin eww windows pfft
143 2015-02-06 13:03:15 <akstunt600> the windows side of thing is still new to me so this should be intersting
144 2015-02-06 13:04:46 <akstunt600> Im great with windows but havnt daily drove it in years and only made exes/msi a handful of times. Also havnt done having compiliations like this minus some openssl stuff i think
145 2015-02-06 13:09:09 <wumpus> oh, building *on* windows, yes that's pretty much uncharted territory
146 2015-02-06 13:10:51 <jonasschnelli> akstunt600, why not cross compile from linux. Thats works pretty good here. I assume you are faster by installing a linux VM on windows and build there
147 2015-02-06 13:11:43 <akstunt600> iam thanks didnt even thin k about that
148 2015-02-06 13:12:47 <akstunt600> Oh wow now thats not even bad to config either Thanks!!
149 2015-02-06 13:23:42 <jonasschnelli> akstunt600, install VMBox on windows and install a ubuntu. You will be ready to compile bitcoin within 1-2h. https://www.virtualbox.org
150 2015-02-06 13:24:15 <akstunt600> yeah just compiled last rc4 last night on linux
151 2015-02-06 13:24:57 <jonasschnelli> i did the following:
152 2015-02-06 13:25:26 <jonasschnelli> git close https://github.com/bitcoin/bitcoin bitcoin-windows (create another repo clone for windows builds)
153 2015-02-06 13:25:34 <jonasschnelli> then: cd depends; make HOST=i686-w64-mingw32; cd ..; ./configure --prefix=`pwd`/depends/i686-w64-mingw32
154 2015-02-06 13:26:02 <jonasschnelli> work without problems on a debian wheezy and a recent ubuntu version
155 2015-02-06 13:26:10 <jonasschnelli> s/work/worked
156 2015-02-06 13:26:18 <akstunt600> Whoa thats awesome
157 2015-02-06 13:26:26 <akstunt600> i didnt even know that was a thing
158 2015-02-06 13:27:26 <jonasschnelli> compiling *on* windows will turn you crazy and could end up by throwing your computer out of the window.
159 2015-02-06 13:27:54 <jonasschnelli> i think here was a similar discussion: http://bitcoinstats.com/irc/bitcoin-dev/logs/2015/01/08#l1420719302
160 2015-02-06 13:28:12 <akstunt600> 
ACTION thinks he can bundles this as exe from ubuntu too

161 2015-02-06 13:28:17 <akstunt600> undle*
162 2015-02-06 13:28:29 <akstunt600> hahahha thanks soo much
163 2015-02-06 13:28:45 <akstunt600> windows... Pfttt why do I even
164 2015-02-06 13:34:28 <jonasschnelli> akstunt600, building with the HOST env var and the --prefix will create a windows exe
165 2015-02-06 13:34:51 <akstunt600> PERFECT!
166 2015-02-06 13:41:54 <akstunt600> Was going to build 64 bit for etra fun. know of anything that will asplode?
167 2015-02-06 13:41:59 <akstunt600> was going to use gcc-mingw-w64-x86-64
168 2015-02-06 13:42:10 <akstunt600> about install my deps on deb
169 2015-02-06 13:50:42 <akstunt600> whoa such confused with minGW why you go x64 w64 mingw32 to i686 w64 mingw32, and i686 w32 mingw64 WUTTTT makes no sense unless it compile in 64 bit, 32 bit apps? and vice versasss. oh this is silly >.< how did i not notice all these different versions before?
170 2015-02-06 13:51:07 <akstunt600> anyway think its all working A++
171 2015-02-06 14:04:33 <akstunt600> Hahaha thats cool, thanks again
172 2015-02-06 14:04:37 <akstunt600> worked great
173 2015-02-06 14:48:10 <akstunt600> jonasschnelli, Hey, well it all look like it work but it broke at ./configure
174 2015-02-06 14:48:26 <akstunt600> i dont see any errors at all looks clean
175 2015-02-06 14:48:48 <akstunt600> there just isnt ./configure in pwd
176 2015-02-06 14:51:40 <akstunt600> ughh wish my scroll back buffer was larger but i dont see anything bomb and its not in depends/x86_x6xxxx
177 2015-02-06 14:54:00 <akstunt600> oh lol
178 2015-02-06 14:54:02 <akstunt600> >.<
179 2015-02-06 15:04:38 <michagogo> akstunt600: ./autogen.sh
180 2015-02-06 15:06:04 <akstunt600> yup yup i got it and then make again
181 2015-02-06 15:25:05 <gdm85> keypool=1000 is still working, right?
182 2015-02-06 15:25:35 <gdm85> which randomness sources does bitcoind use by default? /dev/random, I presume?
183 2015-02-06 15:38:44 <nezZario> gdm85: I think it's an open ssl function
184 2015-02-06 15:39:05 <nezZario> (which probably gets it from /dev/random)
185 2015-02-06 15:39:13 <nezZario> take that with a grain of salt
186 2015-02-06 15:40:08 <gdm85> nezZario: mmh.. ok
187 2015-02-06 16:08:26 <ajweiss> do you guys still need more gitian sigs?
188 2015-02-06 16:37:50 <wumpus> nah, we don't need more, of course more sigs is always better
189 2015-02-06 17:22:10 <dhill> sipa: ?
190 2015-02-06 17:24:39 <sipa> dhill: ?
191 2015-02-06 17:25:10 <dhill> sipa: STRICTENC is strict encoding checks for signatures, hashtype, and pubkeys.  DERSIG is just strict encoding checking for signatures only, correct?
192 2015-02-06 17:25:30 <sipa> indeed
193 2015-02-06 17:25:52 <sipa> though dersig also allows the empty byte vector as invalid signature
194 2015-02-06 17:26:14 <sipa> strictenc is older, never used as a consensus rule, and was originally not softfork-safe
195 2015-02-06 17:28:27 <dhill> ok, thanks
196 2015-02-06 17:29:23 <sipa> dhill: i hope all edge cases are covered in the bip text code and examples
197 2015-02-06 17:29:41 <dhill> well, speaking of that
198 2015-02-06 17:29:52 <dhill> i found IsValidSignatureEncoding only tests about 58%
199 2015-02-06 17:30:13 <dhill> test coverage
200 2015-02-06 17:30:45 <sipa> i got to 100% ...
201 2015-02-06 17:31:00 <sipa> tested with gcov
202 2015-02-06 17:31:15 <sipa> (of that function)
203 2015-02-06 17:31:26 <dhill> maybe the other tests  bring it up.  i only tested with script_valid/script_invalid.json
204 2015-02-06 17:31:57 <sipa> they should cover it; i'll test again
205 2015-02-06 17:32:19 <dhill> so if DERSIG is set, and the signature is 0 length, it should error?
206 2015-02-06 17:32:24 <sipa> no
207 2015-02-06 17:32:34 <dhill> but STRICTENC will
208 2015-02-06 17:32:41 <sipa> the opcode will fail, but the script doesn't necessarily
209 2015-02-06 17:32:47 <dhill> ok
210 2015-02-06 17:33:09 <sipa> you can do checksig not, where an invalid signature is accepted, but an incorrectly encoded one isn't
211 2015-02-06 17:36:50 <dhill> does 0 length signatures pass STRICTENC?
212 2015-02-06 17:36:58 <sipa> i believe so, yes
213 2015-02-06 17:37:16 <sipa> but strictenc isn't consensus critical
214 2015-02-06 17:59:19 <nerdcore> I ran into an issue compiling/linking 0.9.4 test_bitcoin on OpenBSD 5.6. "undefined reference 'main'" linking /usr/lib/crt0.o. I gather this error is generated because test_bitcoin.cpp has no 'main' function, although this is clearly not a problem with other systems / linkers. Any ideas?
215 2015-02-06 18:00:25 <nerdcore> the same system managed to compile bitcoin-cli and bitcoind. Ran bitcoind last night connected to a fully populated node I have on another system and it did this: http://nerdcore.net/mike/debug.log.1423245342
216 2015-02-06 18:01:20 <nerdcore> I'm not even entirely sure what functionality exists within the test_bitcoin binary but I would like to be able to run it now :)
217 2015-02-06 18:02:04 <sipa> it runs the unit tests
218 2015-02-06 18:02:26 <sipa> porting bitcoin core to a new system, without being able to ru nthe unit tests sounds very dangerous
219 2015-02-06 18:02:33 <nerdcore> indeed
220 2015-02-06 18:02:47 <nerdcore> that's why I didn't try to connect it to any other peers :)
221 2015-02-06 18:04:35 <nerdcore> I am curious why this system complains about no 'main 'function while my GNU/Linux system doesn't care... I can provide version info for my build environment if that's helpful
222 2015-02-06 18:05:14 <nerdcore> i've tried using as much stock obsd 5.6 as I can, meaning I have a fairly old gcc etc
223 2015-02-06 18:07:21 <nerdcore> someone suggested I try 0.10rc but I've looked at test_bitcoin.cpp and there's still no 'main' function so I doubt that will solve my problems (also, autogen.sh failed)
224 2015-02-06 18:11:45 <akstunt600> its an openssl issue
225 2015-02-06 18:11:49 <sipa> nerdcore: do you have boost testing installed?
226 2015-02-06 18:12:26 <akstunt600> ooh nm
227 2015-02-06 18:12:33 <akstunt600> wrong window
228 2015-02-06 18:12:50 <nerdcore> sipa: I believe so. I installed boost 1.53.0p5 via a ports package. I think it has everything
229 2015-02-06 18:13:54 <nerdcore> i had never installed/used boost until recently for bitcoind so I'm not too sure how to check. I tried compiling boost from source and it was a disaster so I went with someone's precompiled 1.53.0
230 2015-02-06 18:14:15 <Guest39346> nerdcore: nothing is going to 'fix' it by adding a main function that isn't how the boost test stuff works. But it may work in later versions because of build system changes. (and I know people have been running 0.10rc/git master on openbsd...)
231 2015-02-06 18:16:39 <nerdcore> Guest41310: thx. I will try to get 1.10 up and running
232 2015-02-06 18:16:48 <nerdcore> s/1.10/0.10/ :)
233 2015-02-06 18:21:40 <gmaxwell> (plus even if it doesn't work, we're much more interested in fixing 0.10; we're not likely to do anything about 0.9 now)
234 2015-02-06 18:26:36 <cfields> nerdcore: it'd be much more helpful to get the depends system up and running on bsd
235 2015-02-06 18:27:27 <cfields> (it's our system for self-building the dependencies as-needed)
236 2015-02-06 18:32:24 <nerdcore> cfields: thx for this. There is no mention of the depends/ directory in either INSTALL or doc/build-*
237 2015-02-06 18:33:25 <cfields> nerdcore: well in this case it's not supported. You'd have to help get it working first
238 2015-02-06 18:35:37 <nerdcore> cfields: I just had no idea this depends/ directory was there at all. I'm willing to put some help in,. but I'd looked at INSTALL and then at doc/build-unix.md. I'm stepping out for a bit but I'll read through the READMEs in depends/ and go from there.
239 2015-02-06 18:35:44 <nerdcore> might be nice to mention it in INSTALL
240 2015-02-06 18:36:17 <cfields> fair enough
241 2015-02-06 18:36:34 <cfields> nerdcore: ping me if/when you're willing to give it a try. You'll need to add a few basic things first.
242 2015-02-06 18:36:55 <nerdcore> cfields: wicked. thanks.
243 2015-02-06 18:42:02 <ajweiss> cfields: what do you think of switching from wget to curl when downloading deps on linux?
244 2015-02-06 18:42:20 <ajweiss> debian wget still complains about bitcoincore.org's certificate
245 2015-02-06 18:42:32 <cfields> curl doesn't?
246 2015-02-06 18:42:36 <ajweiss> nope!
247 2015-02-06 18:42:46 <ajweiss> but it does complain on invalid certificates
248 2015-02-06 18:43:00 <ajweiss> openssl s_client doesn't complain either
249 2015-02-06 18:43:00 <sipa> there's --no-check-ceritifcates
250 2015-02-06 18:43:01 <cfields> because it knows about more certs? or because it doesn't check?
251 2015-02-06 18:43:16 <ajweiss> i threw it at a bad cert and it complained
252 2015-02-06 18:43:21 <ajweiss> all signs point to wget being broken
253 2015-02-06 18:43:30 <cfields> if it's the latter, sounds like a reason _agains_ switching. in that case, i'd prefer what sipa just said.
254 2015-02-06 18:43:32 <ajweiss> moreover, wget on my ubuntu box doesn't complain either
255 2015-02-06 18:43:53 <ajweiss> no, curl complains when there's a bad cert
256 2015-02-06 18:44:29 <ajweiss> i did check that, of course
257 2015-02-06 18:49:34 <cfields> ajweiss: i somewhat prefer wget's failure in this case. curl's silent success is arguably incorrect. though, i suppose there's no reason why we shouldn't allow self-signed certs as long as they're valid
258 2015-02-06 18:49:49 <ajweiss> it's not a silent success
259 2015-02-06 18:49:53 <ajweiss> the certificate is valid
260 2015-02-06 18:50:10 <ajweiss> has a real CA, all that business
261 2015-02-06 18:50:22 <ajweiss> debian wget chokes on it
262 2015-02-06 18:50:22 <cfields> yes, but not signed by a trusted authority (according to the system), no?
263 2015-02-06 18:50:34 <ajweiss> openssl s_client on the same system trusts it
264 2015-02-06 18:52:32 <cfields> ajweiss: then curl is probably using openssl's cert store. if that's the case, switching to curl would be the wrong solution, since it's still a crap-shoot. ultimately if we don't care about CA sigs, we should just disable that check
265 2015-02-06 18:52:58 <ajweiss> wait a minute, i think i screwed something up here
266 2015-02-06 18:53:42 <bliljerk_> https://pscoin.com
267 2015-02-06 18:53:50 <ajweiss> oh right
268 2015-02-06 18:54:07 <ajweiss> openssl s_client works if you explicitly supply /etc/ssl/certs for CApath
269 2015-02-06 18:54:17 <ajweiss> otherwise it fails
270 2015-02-06 18:54:32 <cfields> right
271 2015-02-06 18:55:51 <ajweiss> is there anything being wget'd that's not having it's hash checked?  i don't think so
272 2015-02-06 18:56:35 <cfields> no
273 2015-02-06 18:56:41 <ajweiss> that being the case, https isn't doing anything but waste anyway
274 2015-02-06 18:57:58 <cfields> belt-and-suspenders, as they like to say around here
275 2015-02-06 18:58:29 <ajweiss> fair enough, but --no-check-certificate is a pretty sorry pair of suspenders
276 2015-02-06 18:58:37 <kanzure> uh... https://github.com/richardkiss/pycoin/blob/412efbced6852b586bca8ba2f34f124c974b29c4/pycoin/key/BIP32Node.py#L226
277 2015-02-06 18:59:01 <ajweiss> more like a t-shirt with pictures of suspenders printed on it
278 2015-02-06 18:59:24 <gmaxwell> ajweiss: eventually someone adds something that isn't checked or screws up in some other way. Really it's kinda concerning if https is failing in the "uh do we really want to be fetching from that source at all?" kind of way. :)
279 2015-02-06 19:00:24 <gmaxwell> ajweiss: risk there is compounded by a difficulty of testing the failure conditions.
280 2015-02-06 19:00:44 <cfields> gmaxwell: in this case, the unfortunate problem-source is bitcoincore.org
281 2015-02-06 19:01:18 <ajweiss> i don't get why there seems to be two tiers of ca certs installed in debian
282 2015-02-06 19:01:18 <gmaxwell> cfields: whats the actual problem? just doesn't have a signed certificate?
283 2015-02-06 19:01:24 <cfields> so the for-really-real fix to the problem would be to get a more compat cert there. but that's a pain
284 2015-02-06 19:01:33 <gmaxwell> oh there is a cert but its not in debian's ca sore? blah
285 2015-02-06 19:01:40 <kanzure> ca sore :)
286 2015-02-06 19:01:43 <cfields> gmaxwell: it's signed by a CA that older systems don't have in their stores
287 2015-02-06 19:01:44 <cfields> yep
288 2015-02-06 19:01:54 <ajweiss> but the ca cert is on there!
289 2015-02-06 19:01:59 <ajweiss> i didn't install it
290 2015-02-06 19:02:01 <gmaxwell> or nag debian to fix it in security backports on thoses hosts?
291 2015-02-06 19:02:33 <cfields> gmaxwell: according to ajweiss here, it also seems to depend on _which_ store is in use. apparently wget and curl use different stores.
292 2015-02-06 19:02:35 <ajweiss> openssl s_client .... -CApath /etc/ssl/certs works, openssl s_client .... is broken
293 2015-02-06 19:03:33 <cfields> so he's suggesting that we switch to curl. which fixes the immediate problem, but only by chance really.
294 2015-02-06 19:04:11 <gmaxwell> do be fair, if its only old hosts... then its not likely to unfix itself.
295 2015-02-06 19:04:33 <ajweiss> this is a brand new debian-7.7 install
296 2015-02-06 19:04:43 <ajweiss> basically i followed the gitian directions and it broke
297 2015-02-06 19:06:07 <cfields> gmaxwell: sure. but odds are equally good that switching it would unfix some other old host :)
298 2015-02-06 19:07:07 <cfields> ajweiss: i suppose there's little harm in changing and seeing what happens.
299 2015-02-06 19:08:21 <ajweiss> tell ya what, i'm gonna dig around a little more first
300 2015-02-06 19:08:27 <ajweiss> because this is pretty weird
301 2015-02-06 19:08:48 <ajweiss> why are there multiple cert bundles in the first place?
302 2015-02-06 19:09:40 <cfields> libs/apps are free to ship their own if they don't want/like/know what's on the system
303 2015-02-06 19:10:43 <cfields> according to distro policy ofc. I have no idea what those rules are.
304 2015-02-06 19:13:04 <morcos> sipa: i guess there are plenty of things that will now be standard in 0.10 that aren't standard in 0.9, so it shouldn't be a concern that 0-byte sigs are still non-standard in 0.9 (via IsCanonicalSignature)?
305 2015-02-06 19:13:33 <sipa> morcos: yeah, being more strict isn't a problem
306 2015-02-06 19:13:42 <sipa> as long as it's not a consensus rule, of course
307 2015-02-06 19:14:53 <morcos> any thought into whether it'll cause some weird propogation behavior as the network switches over, or we're kind of assuming that previously non-standard txs aren't likely to be very common for a while
308 2015-02-06 19:47:55 <nerdcore> cfields: I've just tried to make in the depends/ directory. ksh (/bin/sh) gives a bunch of "Makefiles: cannot execute - Permission denied"
309 2015-02-06 19:48:09 <cfields> nerdcore: yes, it'll bomb badly without some initial setup
310 2015-02-06 19:48:24 <cfields> nerdcore: head to #bitcoin-build ?
311 2015-02-06 19:48:33 <nerdcore> k thx
312 2015-02-06 19:59:24 <ajweiss> interesting, it appears the stock gnutls that comes with debian rejects the bitcoincore.org certificate, even when it is given the correct root certificates
313 2015-02-06 19:59:48 <ajweiss> whereas openssl works without issue
314 2015-02-06 20:04:18 <Hasimir> gnutls breaking things ... colour me unsurprised
315 2015-02-06 20:09:03 <nkuttler> ajweiss: wheezy? i don't think it packages new certs
316 2015-02-06 20:09:17 <nkuttler> or if it does it's in some updates repo you might not use
317 2015-02-06 20:15:03 <ajweiss> the certificate is there.
318 2015-02-06 20:15:12 <ajweiss> gnutls doesn't like it
319 2015-02-06 20:15:14 <nkuttler> oh.. ok. that sucks
320 2015-02-06 20:17:24 <ajweiss> ohhhh
321 2015-02-06 20:17:34 <ajweiss> gnutls no support ecdsa/ecdh
322 2015-02-06 20:25:49 <Diablo-D3> Google is giving $6 to Chromecast owners: https://cast.google.com/chromecast/offers/
323 2015-02-06 20:26:50 <theorbtwo> Dammit, I keep pondering getting one of those, but I don't have an actual use for it.
324 2015-02-06 20:28:07 <Diablo-D3> theorbtwo: I have two things plugged into my TV: a chromecast and a bluray player
325 2015-02-06 20:28:23 <Diablo-D3> and if chromecast apps other than netflix could output surround sound, I wouldnt need the bluray player
326 2015-02-06 20:28:26 <Diablo-D3> I'd just rip before watching
327 2015-02-06 20:32:18 <ajweiss> ok, so... debian installs the root certificate just fine, but it also ships with gnutls2.6 which does not have support for elliptic curves.  wget uses gnutls for tls, so it balks on the bitcoincore.org server certificate.  curl uses openssl so there's no problem.
328 2015-02-06 20:39:54 <ajweiss> so can we just use curl now?
329 2015-02-06 20:43:46 <theorbtwo> Diablo-D3: I don't actually have a TV.  I have three monitors on two computers, and a projector that doesn't do HDMI.
330 2015-02-06 20:44:08 <Diablo-D3> I'd toss the projector.
331 2015-02-06 20:44:26 <theorbtwo> Diablo-D3: Er, noted, but it works pretty well, and I'm not made of money.
332 2015-02-06 21:11:25 <akstunt600> Anyone know if blocknotify works in qt wallet?
333 2015-02-06 21:12:15 <sipa> it should
334 2015-02-06 21:12:18 <akstunt600> like displays response in qt? would be cool to see this developed
335 2015-02-06 21:12:33 <sipa> ?
336 2015-02-06 21:12:41 <sipa> blocknotify runs a script whenever a new block is accepted
337 2015-02-06 21:12:47 <sipa> you supply the script
338 2015-02-06 21:12:53 <akstunt600> oh i miss read
339 2015-02-06 21:12:57 <akstunt600> sweet thats perfect
340 2015-02-06 21:14:03 <akstunt600> was just thinking it be cool to have like maybe an admin or advanced wallet coincontrol displayed by default and this built in
341 2015-02-06 21:14:12 <akstunt600> where*
342 2015-02-06 21:14:43 <akstunt600> for GUI people who i am not.
343 2015-02-06 21:14:46 <akstunt600> 
ACTION shrugs

344 2015-02-06 21:34:45 <sipa> dhill: running test_bitcoin, compiled with -O0 --coverage, gives 100% branch coverage in IsValidSignatureEncoding
345 2015-02-06 21:35:07 <dhill> ok
346 2015-02-06 21:56:54 <Luke-Jr> petertodd: there is still no CPFP patchset for 0.10, correct?
347 2015-02-06 21:57:02 <Luke-Jr> err, not CPFP. Replace-by-fee
348 2015-02-06 22:02:31 <phantomcircuit> Luke-Jr, did he ever answer your question about stealth addresses and the contract tool
349 2015-02-06 22:08:05 <ajweiss> actually, i'm full of it.  it has nothing to do with gnutls's lack of ecc crypto, they're all rsa certs
350 2015-02-06 22:08:32 <Luke-Jr> phantomcircuit: not AFAIK
351 2015-02-06 22:08:52 <Luke-Jr> maybe he's out camping again or smth?
352 2015-02-06 22:09:11 <ajweiss> it looks like the order of the certs in the chain on bitcoincore.org is swapped and gnutls won't tolerate it
353 2015-02-06 22:10:22 <ajweiss> http://0bin.net/paste/aEYntXg2eQYLwyTl#fviflaxmzRmCQ+-IA995MjOgiNFHTVa533R2ewRi5ne
354 2015-02-06 22:21:33 <phantomcircuit> ajweiss, what are you talking about?
355 2015-02-06 22:29:54 <ajweiss> bitcoincore.org
356 2015-02-06 22:31:09 <ajweiss> when you do a make depends it downloads deps, if one of the sites is unreachable it falls back to bitcoincore.org, on linux it uses wget to download https and wget is on top of gnutls2.6 which doesn't tolerate out of order certificate chains, so the build stops on a certificate error
357 2015-02-06 22:48:34 <phantomcircuit> ajweiss, it's in order
358 2015-02-06 22:48:40 <phantomcircuit> maybe it's reversed actually
359 2015-02-06 22:50:00 <ajweiss> i emailed gavin, as he's the technical contact on the domain
360 2015-02-06 22:56:00 <phantomcircuit> ajweiss, the ssl settings for bitcoincore.org are screwed up
361 2015-02-06 22:56:40 <ajweiss> yep.
362 2015-02-06 22:57:09 <cfields> hmm, looking now. the cert order just needs to be switched?
363 2015-02-06 22:57:22 <ajweiss> yes.
364 2015-02-06 22:57:32 <cfields> how did you verify broken/fixed ?
365 2015-02-06 22:57:34 <ajweiss> https://lists.gnu.org/archive/html/gnutls-devel/2012-09/msg00039.html
366 2015-02-06 22:57:47 <phantomcircuit> cfields, there's a few other issues https://www.ssllabs.com/ssltest/analyze.html?d=bitcoincore.org
367 2015-02-06 22:58:57 <ajweiss> if you switch to ecc certs, that will break wget on debian as well
368 2015-02-06 22:59:36 <cfields> phantomcircuit: it says chain order is fine?
369 2015-02-06 23:00:45 <phantomcircuit> cfields, no it's backwards
370 2015-02-06 23:00:51 <ajweiss> openssl s_client -connect bitcoincore.org:443 -CApath /etc/ssl/certs/
371 2015-02-06 23:00:55 <phantomcircuit> #3 was issues by #2
372 2015-02-06 23:01:15 <ajweiss> it's obvious when you look at the chain
373 2015-02-06 23:09:53 <cfields> ajweiss: did that make any difference?
374 2015-02-06 23:23:18 <ajweiss> oddly, no, gnutls still complains
375 2015-02-06 23:23:31 <ajweiss> order is fixed though
376 2015-02-06 23:24:04 <ajweiss> there we go!
377 2015-02-06 23:24:17 <cfields> wget works now?
378 2015-02-06 23:24:20 <ajweiss> yep!
379 2015-02-06 23:24:26 <cfields> hooray!
380 2015-02-06 23:24:31 <cfields> nice detective work :)
381 2015-02-06 23:27:54 <cfields> ajweiss: mind mailing gavin and letting him know we got it taken care of? a double-fix wouldn't be a good thing :)
382 2015-02-06 23:28:47 <gavinandresen> too late
383 2015-02-06 23:29:23 <gavinandresen> cfields: I probably was editing the file at the same time as you.......
384 2015-02-06 23:29:52 <gavinandresen> I’ll stop trying to be helpful now.
385 2015-02-06 23:30:07 <cfields> haha
386 2015-02-06 23:30:41 <cfields> gavinandresen: well i should've just left it alone when he mailed you. but i couldn't resist taking a quick look. sorry bout that
387 2015-02-06 23:31:27 <gavinandresen> no worries
388 2015-02-06 23:34:35 <phantomcircuit> cfields, cd /etc && git init
389 2015-02-06 23:35:36 <cfields> phantomcircuit: i really don't think there's enough complexity on there to warrant that. i'll leave that to gavinandresen though.
390 2015-02-06 23:36:02 <phantomcircuit> i do it basically everywhere just because it's nearly free