00:00:02  * ircretaryquit (Remote host closed the connection)
00:00:10  * kellabytequit (Read error: Operation timed out)
00:00:12  * ircretaryjoined
00:02:58  * kellabytejoined
00:04:23  * julianduquequit (Ping timeout: 246 seconds)
00:08:39  * EhevuTovjoined
00:20:33  * kazuponjoined
00:20:39  * mikolalysenkojoined
00:21:54  * mralephjoined
00:22:32  * mraleph1quit (Read error: Connection reset by peer)
00:24:53  * kazuponquit (Ping timeout: 240 seconds)
00:27:09  * mikolalysenkoquit (Ping timeout: 252 seconds)
00:38:19  * timoxleyquit (Read error: Connection reset by peer)
00:41:48  <MI6>joyent/node: Timothy J Fontaine v0.10 * 4c38742 : test: fix tls-hello-parser-failure on smartos - http://git.io/RBudRg
00:41:54  <isaacs>tjfontaine: 50%?? really?
00:41:59  <tjfontaine>isaacs: really.
00:42:01  <isaacs>tjfontaine: also ^++
00:42:05  <isaacs>:)
00:42:13  <tjfontaine>isaacs: also happy birthday :)
00:42:19  <isaacs>thanks
00:42:26  <isaacs>dude, this retina display... omg...
00:42:30  <isaacs>text looks soooooo good
00:42:36  <tjfontaine>haha, ya
00:42:41  <tjfontaine>they're pretty snazzy
00:42:43  <isaacs>trying to move everything over frantically so i can never use the old one again
00:44:17  <TooTallNate>isaacs: you got a new MBP?
00:44:33  <TooTallNate>isaacs: and happy birthday dude!
00:45:59  <tjfontaine>isaacs: so, for mmalecki they want to rsync /dist, which I'm happy to provide for them, but I wanted to make sure you didn't have a big change planned for distribution via manta
00:46:09  * st_lukequit (Remote host closed the connection)
00:46:09  <isaacs>nah, that can all be yours :)
00:46:15  <tjfontaine>heh ok
00:46:48  <tjfontaine>I'm just going to drop an rsync daemon smf in there, and then if others come by asking for a way to do it we have a solution for them
00:47:07  <isaacs>kewl
00:50:33  * mikolalysenkojoined
00:51:35  <MI6>nodejs-v0.10: #266 UNSTABLE linux-ia32 (2/593) osx-ia32 (217/593) linux-x64 (1/593) http://jenkins.nodejs.org/job/nodejs-v0.10/266/
00:52:04  <tjfontaine>I'm going to have to figure out why that osx ia32 thinks something is holding on to its port
00:56:11  * kellabytequit (Changing host)
00:56:11  * kellabytejoined
00:57:55  <MI6>nodejs-v0.10-windows: #95 UNSTABLE windows-x64 (6/593) windows-ia32 (7/593) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/95/
01:03:23  * kazuponjoined
01:03:29  * loladirojoined
01:08:55  * kazuponquit (Ping timeout: 264 seconds)
01:11:06  <MI6>nodejs-v0.10: #267 UNSTABLE linux-ia32 (25/593) osx-x64 (13/593) smartos-x64 (9/593) osx-ia32 (19/593) linux-x64 (54/593) smartos-ia32 (103/593) http://jenkins.nodejs.org/job/nodejs-v0.10/267/
01:11:57  <tjfontaine>much better, except that full nightly is running and that upsets things as well
01:20:05  * EhevuTovquit (Quit: This computer has gone to sleep)
01:25:18  * TooTallNatequit (Quit: Computer has gone to sleep.)
01:31:14  * Benviequit (Ping timeout: 246 seconds)
01:33:41  * Benviejoined
01:35:50  * brsonquit (Quit: leaving)
01:40:43  * hueniversejoined
01:43:19  * jmar777quit (Remote host closed the connection)
01:43:57  * jmar777joined
01:48:31  * jmar777quit (Ping timeout: 264 seconds)
02:04:46  * julianduquejoined
02:21:41  <kellabyte>does on_body() in http_parser get called multiple times for a request?
02:25:17  * qardjoined
02:28:20  * c4miloquit (Remote host closed the connection)
02:30:00  * julianduquequit (Ping timeout: 246 seconds)
02:31:10  * qardpart
02:32:29  * timoxleyjoined
02:34:28  * timoxleyquit (Read error: Connection reset by peer)
02:48:12  * inolenquit (Ping timeout: 246 seconds)
02:54:13  <trevnorris>tjfontaine: know if we're having a call tomorrow morning?
02:54:53  * defunctzombiechanged nick to defunctzombie_zz
02:54:53  <tjfontaine>when I asked isaacs sunday he said ya
02:55:05  <trevnorris>cool
02:55:14  * julianduquejoined
02:59:01  <trevnorris>tjfontaine: test-debug-break-on-uncaught.js is failing on my box.
02:59:49  <tjfontaine>any output?
02:59:56  <trevnorris>tjfontaine: it's sporadic, but happening. with a "AssertionError: debugger did not pause on exception"
03:00:01  <tjfontaine>ya
03:00:18  * julianduquequit (Ping timeout: 256 seconds)
03:00:24  <tjfontaine>all the debug tests are racey
03:00:47  <trevnorris>hm. just haven't had any debug tests fail on my linux box for a while.
03:02:07  * defunctzombie_zzchanged nick to defunctzombie
03:04:36  * inolenjoined
03:04:53  <trevnorris>tjfontaine: do you know if you the fs.write/writeSync arguments are optional? they're listed as required in the docs, but the source allows some to be optional.
03:05:34  * brsonjoined
03:05:37  <tjfontaine> 3 not ok - test-debugger-repl-restart.js
03:05:37  <tjfontaine> 3 not ok - test-http-end-throw-socket-handling.js
03:05:37  <tjfontaine> 4 not ok - test-net-settimeout.js
03:05:37  <tjfontaine> 10 not ok - test-http-client-timeout-event.js
03:05:37  <tjfontaine> 12 not ok - test-http-client-timeout-agent.js
03:05:40  <tjfontaine> 38 not ok - test-debugger-client.js
03:05:51  * timoxleyjoined
03:05:55  <tjfontaine>some of the common failing tests from june
03:06:07  <tjfontaine>trevnorris: lemme look into that
03:07:01  <trevnorris>tjfontaine: nice list. awesome system you've setup. makes testing a lot easier.
03:07:40  <tjfontaine>mjob create -m 'egrep "^(not )?ok" | sed -E "s#ok [0-9]+#ok#" | tr -d "\r"' -r 'sort | uniq -c' < linux-tap-files
03:07:44  <tjfontaine>manta is fun as hell
03:08:01  <trevnorris>heh, cool.
03:08:23  * st_lukejoined
03:09:32  <tjfontaine>trevnorris: the documentation on this doesn't seem very explicit, following our other patterns I would have assumed from the docs then they were likely optional
03:09:40  * julianduquejoined
03:09:45  <trevnorris>yeah. figured so.
03:09:52  <trevnorris>cool.
03:10:13  * trevnorrisback to improving fs writes
03:10:28  <trevnorris>this has been deceptively difficult.
03:11:01  <tjfontaine>ya that path is twisted :)
03:20:47  <trevnorris>oy, since position can be null and not just a number it adds some strange complexity.
03:26:53  * defunctzombiechanged nick to defunctzombie_zz
03:31:10  * mikolalysenkoquit (Ping timeout: 276 seconds)
03:33:16  <MI6>nodejs-v0.10: #268 UNSTABLE smartos-x64 (1/593) linux-x64 (1/593) http://jenkins.nodejs.org/job/nodejs-v0.10/268/
03:35:22  <trevnorris>whoa. only 2 failing. awesome.
03:35:59  * brsonquit (Ping timeout: 240 seconds)
03:37:46  <tjfontaine>ya and they're bouncy ones
03:37:59  <tjfontaine>did you see how much faster the test suit runs?
03:38:02  <tjfontaine>*suite
03:42:16  * dapquit (Quit: Leaving.)
03:43:40  * julianduquequit (Ping timeout: 256 seconds)
03:53:23  * kazuponjoined
03:55:40  * Benviequit (Remote host closed the connection)
03:55:57  * Benviejoined
04:03:37  * mikolalysenkojoined
04:09:15  * TooTallNatejoined
04:09:34  * TooTallNatequit (Client Quit)
04:09:54  <trevnorris>tjfontaine: yeah I did.
04:20:48  * amartensjoined
04:27:50  * defunctzombie_zzchanged nick to defunctzombie
04:37:57  * defunctzombiechanged nick to defunctzombie_zz
04:45:47  <Luqman>hi, in include/uv-private/uv-win.h line 237 on line, looking through git used to be void* srwlock_; since as explained in the comment above it not all toolchains have the type but it seems to have been reverted
04:45:51  <Luqman>was that intentional?
04:47:20  * c4milojoined
04:49:48  <tjfontaine>what toolchain are you using?
04:53:32  <tjfontaine>I'm assuming mingw
04:56:44  * c4miloquit (Remote host closed the connection)
04:59:03  * bradleymeckjoined
05:01:07  * kazuponquit (Remote host closed the connection)
05:03:32  * amartensquit (Quit: Leaving.)
05:07:11  * mikealquit (Quit: Leaving.)
05:11:23  * amartensjoined
05:17:18  * bradleymeckquit (Quit: bradleymeck)
05:22:07  * st_lukequit (Remote host closed the connection)
05:26:07  * mikealjoined
05:36:06  * kazuponjoined
05:45:54  <Luqman>tjfontaine: yep, mingw64
05:46:09  <Luqman>though it seems like -D_WIN32_WINNT=0x0600 was enough to fix that
05:46:39  <tjfontaine>Luqman: I would comment on https://github.com/joyent/libuv/issues/263
05:48:27  <Luqman>tjfontaine: i don't think i have anything to add, since the issue wasn't with libuv's build system itself but in rust
05:49:17  <tjfontaine>ok
05:49:35  <tjfontaine>I thought that sounded like we had handled it in our buildsystem at one point
05:50:43  <Luqman>yep, it was because uv.h is #include'd in rust's runtime. Though it's a bit odd that mingw doesn't complain but mingw64 does
05:51:58  <Luqman>tjfontaine: thanks though!
05:52:42  * Benviequit (Remote host closed the connection)
05:53:03  * Benviejoined
05:53:09  <tjfontaine>mk
06:00:34  * Benviequit (Remote host closed the connection)
06:00:34  * Benviejoined
06:00:48  * julianduquejoined
06:02:03  * defunctzombie_zzchanged nick to defunctzombie
06:02:25  * amartensquit (Quit: Leaving.)
06:06:48  * julianduquequit (Read error: Operation timed out)
06:23:55  * amartensjoined
06:30:17  * wolfeidauquit (Remote host closed the connection)
06:35:33  * Benviequit (Remote host closed the connection)
06:35:52  * Benviejoined
06:38:52  * amartensquit (Quit: Leaving.)
06:47:15  * rendarjoined
06:47:49  * loladiroquit (Quit: loladiro)
06:55:01  * mikolalysenkoquit (Ping timeout: 248 seconds)
06:56:26  * julianduquejoined
06:57:09  * AvianFluquit (Remote host closed the connection)
07:01:36  * julianduquequit (Ping timeout: 246 seconds)
07:02:56  * wolfeidaujoined
07:05:35  * loladirojoined
07:29:27  * defunctzombiechanged nick to defunctzombie_zz
07:33:41  * timoxleyquit (Quit: Computer has gone to sleep.)
07:35:42  <trevnorris>whoot! finally got that sucker working.
07:36:17  * csaohjoined
07:36:36  * timoxleyjoined
07:57:21  * loladiroquit (Quit: loladiro)
08:06:28  * julianduquejoined
08:11:53  * AvianFlujoined
08:17:11  * mralephquit (Ping timeout: 246 seconds)
08:21:18  * dominictarrquit (Quit: dominictarr)
08:24:03  * Benviequit (Remote host closed the connection)
08:24:20  * Benviejoined
08:25:00  * julianduquequit (Ping timeout: 256 seconds)
08:26:41  * mralephjoined
08:29:12  * julianduquejoined
08:30:23  * defunctzombie_zzchanged nick to defunctzombie
08:38:29  * julianduquequit (Ping timeout: 256 seconds)
08:40:01  * defunctzombiechanged nick to defunctzombie_zz
08:47:43  * bnoordhuisjoined
08:51:25  * csaohquit (Remote host closed the connection)
08:54:20  * julianduquejoined
08:56:33  * csaohjoined
09:10:54  * julianduquequit (Ping timeout: 256 seconds)
09:23:43  <MI6>joyent/libuv: Ben Noordhuis master * ddd7e04 : build: switch to autotools - http://git.io/oB67lA
09:24:14  <MI6>libuv-master: #133 FAILURE http://jenkins.nodejs.org/job/libuv-master/133/
09:28:10  <MI6>libuv-master-gyp: #75 UNSTABLE windows-x64 (5/191) smartos-x64 (2/190) smartos-ia32 (2/190) windows-ia32 (3/191) http://jenkins.nodejs.org/job/libuv-master-gyp/75/
09:59:22  * julianduquejoined
10:05:36  * julianduquequit (Ping timeout: 256 seconds)
10:15:16  * wolfeidauquit (Remote host closed the connection)
10:31:50  * udpjoined
10:32:47  * csaohquit (Quit: csaoh)
10:33:53  * julianduquejoined
10:37:11  * kazuponquit (Ping timeout: 246 seconds)
10:38:32  * abraxasquit (Remote host closed the connection)
10:42:17  * udp_joined
10:43:00  * udpquit (Ping timeout: 256 seconds)
10:43:00  * udp_changed nick to udp
10:43:48  <MI6>joyent/libuv: Ben Noordhuis master * d2f67b9 : build: use AM_PROG_AR conditionally - http://git.io/XJzUxg
10:45:53  * julianduquequit (Ping timeout: 240 seconds)
10:46:26  * csaohjoined
10:48:20  <MI6>libuv-master-gyp: #76 UNSTABLE linux-ia32 (1/190) windows-x64 (3/191) smartos-x64 (2/190) smartos-ia32 (2/190) windows-ia32 (3/191) http://jenkins.nodejs.org/job/libuv-master-gyp/76/
10:48:48  * julianduquejoined
10:50:01  <MI6>libuv-master: #134 FAILURE smartos (3/190) http://jenkins.nodejs.org/job/libuv-master/134/
10:56:28  * julianduquequit (Ping timeout: 246 seconds)
11:44:04  <MI6>joyent/libuv: Ben Noordhuis master * d0be852 : test: fix fs_chown when running as root (+1 more commits) - http://git.io/-Wf3_A
11:48:14  <MI6>libuv-master: #135 FAILURE smartos (6/190) http://jenkins.nodejs.org/job/libuv-master/135/
11:50:17  <MI6>libuv-master-gyp: #77 UNSTABLE windows-x64 (3/191) smartos-x64 (2/190) smartos-ia32 (2/190) windows-ia32 (3/191) http://jenkins.nodejs.org/job/libuv-master-gyp/77/
11:51:16  * bnoordhuisquit (Ping timeout: 256 seconds)
11:57:30  * wolfeidaujoined
12:02:43  * wolfeidauquit (Ping timeout: 276 seconds)
12:06:23  * kazuponjoined
12:11:05  * julianduquejoined
12:11:35  * Benviequit (Remote host closed the connection)
12:12:11  * Benviejoined
12:23:17  * julianduquequit (Ping timeout: 256 seconds)
12:33:03  * julianduquejoined
12:33:03  * julianduquequit (Client Quit)
12:33:28  * udpquit (Quit: udp)
12:41:13  * kazuponquit (Remote host closed the connection)
12:47:38  * jmar777joined
12:57:55  * bnoordhuisjoined
13:00:47  * bnoordhuisquit (Read error: Operation timed out)
13:30:55  * hzjoined
13:42:39  * julianduquejoined
13:46:18  * hzquit
13:47:30  * kazuponjoined
13:54:44  * bnoordhuisjoined
13:57:38  * julianduquequit (Ping timeout: 256 seconds)
14:07:51  * kazuponquit (Remote host closed the connection)
14:07:57  * julianduquejoined
14:19:04  * kazuponjoined
14:23:46  * julianduquequit (Ping timeout: 276 seconds)
14:26:08  * udpjoined
14:26:30  * pachetjoined
14:26:30  * pachetquit (Changing host)
14:26:30  * pachetjoined
14:27:52  * mikolalysenkojoined
14:29:40  * kazuponquit (Remote host closed the connection)
14:38:06  <MI6>joyent/libuv: Ben Noordhuis master * 5a3b8f9 : test: fix spawn_setgid_fails and spawn_setuid_fails - http://git.io/zyDe-Q
14:41:47  * wavdedjoined
14:42:17  <MI6>libuv-master: #136 FAILURE smartos (2/190) http://jenkins.nodejs.org/job/libuv-master/136/
14:44:22  <MI6>libuv-master-gyp: #78 UNSTABLE windows-x64 (4/191) smartos-x64 (2/190) smartos-ia32 (4/190) windows-ia32 (3/191) http://jenkins.nodejs.org/job/libuv-master-gyp/78/
14:48:24  * dapjoined
14:49:59  * mikealquit (Quit: Leaving.)
14:51:31  * kazuponjoined
14:51:48  * mikealjoined
14:53:53  * mcavagequit (Remote host closed the connection)
15:03:44  * piscisaureus_joined
15:07:10  * stagasjoined
15:09:11  <bnoordhuis>piscisaureus_: sup bertje? how was CA?
15:10:07  * defunctzombie_zzchanged nick to defunctzombie
15:12:32  <mmalecki>half of nodejitsu team is still drunk, I'm going to take an educated guess and say that piscisaureus_ isn't able to respond just yet, bnoordhuis
15:12:53  <piscisaureus_>bnoordhuis: I'm in the sprint planning meeting
15:13:21  <bnoordhuis>oh, is that now? why doesn't calender send me meetings anymore? :-Ss
15:13:27  <bnoordhuis>*reminders
15:13:31  <bnoordhuis>i'll join
15:16:08  <bnoordhuis>upgrading all the things!
15:16:13  <bnoordhuis>provided it compiles
15:18:11  <mmalecki>bnoordhuis: I have this longstanding feeling that I wanted to bother you about something related to uv. do you happen to have an idea what it might be?
15:19:36  <mmalecki>aahhh, https://github.com/joyent/libuv/pull/675
15:19:43  <mmalecki>well, damn, windows :/
15:21:25  <bnoordhuis>mmalecki: btw, uv__new_sys_error is about to disappear
15:21:37  <bnoordhuis>i'm changing all the things to return errors directly
15:21:51  <bnoordhuis>the unix implementation can just return -errno
15:22:04  <mmalecki>oh, any reason behind that?
15:22:32  <bnoordhuis>i want to increase my 'lines touched' score
15:22:56  <mmalecki>I mean, obviously, everyone does
15:25:27  * c4milojoined
15:27:41  <bnoordhuis>mmalecki: one less loop global, easier for bindings author (no uv_err_t), easier to deal with internally, etc.
15:27:44  <bnoordhuis>^ reasons why
15:27:54  <bnoordhuis>it also reduces the line count a little
15:27:57  * dominictarrjoined
15:28:19  <mmalecki>right :-)
15:28:21  <mmalecki>o hai dominictarr
15:29:47  <dominictarr>mmalecki: hey, whats up?
15:30:04  <mmalecki>bnoordhuis: you ever finished that gdb v8 addon?
15:30:20  <mmalecki>bnoordhuis: asking because the book
15:30:26  <mmalecki>*of the book
15:31:21  * timoxleyquit (Quit: Computer has gone to sleep.)
15:34:40  <bnoordhuis>mmalecki: no. gdbjit in v8 kind of works though
15:38:41  <tjfontaine>mmalecki: lemme know if you hit any problems with the rsync
15:39:24  <mmalecki>tjfontaine: I tested it locally, it works :-)
15:39:54  <mmalecki>tjfontaine: so I'm writing this system for quill which'd basically set up a server rsync'ing from nodejs.org
15:40:06  <mmalecki>tjfontaine: can I make this public, or should it be rather private?
15:41:21  <tjfontaine>I don't think it's a problem for it to be public, most people probably don't even know that there is a pure rsync protocol :)
15:41:23  * mikolalysenkoquit (Ping timeout: 246 seconds)
15:42:14  <mmalecki>haha, right
15:43:55  <tjfontaine>it would only really be a problem for bw slamming the first time someone clones
15:44:29  * Benviequit (Remote host closed the connection)
15:44:45  * Benviejoined
15:50:40  <indutny>guys
15:50:45  <indutny>sorry, I'll miss the call
15:51:04  <indutny>going to restaurant with family
15:52:07  * `3rdEdenchanged nick to `3E|DINNER
15:55:40  * qardjoined
15:56:39  <bnoordhuis>indutny: have fun, fedor. anything to report?
15:57:04  * qardpart
15:58:11  <isaacs>tjfontaine: i realized this morning, i could use rsync to make a better version of nave
15:58:47  <tjfontaine>bnoordhuis: aclocal:configure.ac:18: warning: macro `AM_SILENT_RULES' not found in library <-- too old version of the autoconf on the linux slave?
15:58:59  <tjfontaine>isaacs: how does nave work now?
15:59:11  <bnoordhuis>tjfontaine: yes
15:59:17  <tjfontaine>bnoordhuis: mk
15:59:26  <bnoordhuis>is it erroring on that or just warning?
15:59:38  <bnoordhuis>warning, complaining
15:59:48  <tjfontaine>configure.ac:18: error: possibly undefined macro: AM_SILENT_RULES
15:59:48  <tjfontaine> If this token and others are legitimate, please use m4_pattern_allow.
15:59:48  <tjfontaine> See the Autoconf documentation.
15:59:59  <isaacs>tjfontaine: it curls
16:00:05  <bnoordhuis>right, easy to fix
16:01:05  <indutny>bnoordhuis: nope
16:01:08  * udpquit (Ping timeout: 252 seconds)
16:01:12  <indutny>was all in rust
16:01:15  <indutny>and voxer stuff
16:01:18  <trevnorris>call today?
16:01:45  <isaacs>ah, right
16:01:47  <isaacs>one second
16:02:30  <MI6>joyent/libuv: Ben Noordhuis master * 3b9cfe2 : build: use AM_SILENT_RULES conditionally - http://git.io/TnnA8Q
16:02:33  <bnoordhuis>tjfontaine: ^
16:02:53  <tjfontaine>bnoordhuis: thanks
16:03:03  <isaacs>oh, we talked about doing a google hangout
16:03:06  <isaacs>i'm going to try that
16:03:20  <piscisaureus_>isaacs: you can add a hangout link to a google calendar invite
16:03:29  <piscisaureus_>there's a checkbox for it
16:03:56  <bnoordhuis>tjfontaine: how old is that automake anyway?
16:04:34  <tjfontaine>bnoordhuis: lemme check
16:06:02  <isaacs>piscisaureus_: ohh, nice
16:06:04  <isaacs>i just invited you all
16:06:06  * TooTallNatejoined
16:06:34  <bnoordhuis>isaacs: what email addy did you use? [email protected] or my @bnoordhuis.nl address?
16:06:56  <bnoordhuis>ah nvm, got the email
16:06:57  <isaacs>Ok, I just added it to the calendar invite
16:06:59  <isaacs>ignore the first email
16:07:02  <MI6>libuv-master: #137 FAILURE smartos (2/190) http://jenkins.nodejs.org/job/libuv-master/137/
16:07:02  <isaacs>follow the second one :)
16:07:38  <isaacs>tjfontaine, trevnorris, piscisaureus_, sblom, TooTallNate, indutny: updated the calendar invite to add a video call. let's try this instead of skype.
16:08:27  <bnoordhuis>isaacs: can you add [email protected]? my @bnoordhuis.nl domain has hangouts disabled
16:08:30  <tjfontaine>which account did you send that to?
16:09:03  <MI6>libuv-master-gyp: #79 UNSTABLE windows-x64 (3/191) smartos-x64 (2/190) smartos-ia32 (2/190) windows-ia32 (3/191) http://jenkins.nodejs.org/job/libuv-master-gyp/79/
16:09:56  <trevnorris>k. just installing plugin
16:24:58  * qardjoined
16:36:23  <tjfontaine>debugger doesn't work in Cygwin (#5779)
16:36:24  <tjfontaine>lul
16:36:50  <bnoordhuis>i was thinking the same thing
16:37:34  <piscisaureus_>bnoordhuis: thinking it, or about it?
16:38:51  <bnoordhuis>piscisaureus_: you're such a rascal
16:38:57  <bnoordhuis>okay, dinner. biab
16:40:07  <piscisaureus_>A member of a criminal gang in Papua New Guinea.
16:40:12  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
16:40:44  <TooTallNate>bert and ben should have like a stand up act where they contradict each other the entire time
16:41:02  <tjfontaine>haha
16:42:46  * mikolalysenkojoined
16:43:24  * bnoordhuisquit (Ping timeout: 252 seconds)
16:46:14  <trevnorris>is it possible to pass a Buffer to fs.write and not have the entire Buffer written to disk?
16:47:39  * TooTallNatequit (Quit: Computer has gone to sleep.)
16:47:54  * bajtosjoined
16:53:15  <isaacs>ircretary: tell piscisaureus_ can you add #manta to slurp's list of rooms?
16:53:15  <ircretary>isaacs: I'll be sure to tell piscisaureus_
17:00:20  * amartensjoined
17:04:44  * TooTallNatejoined
17:05:50  * brsonjoined
17:07:05  <trevnorris>alright. time to start this 12 hr drive. see y'all tomorrow
17:07:12  <tjfontaine>best of luck
17:09:55  <isaacs>trevnorris: drive safe
17:11:51  <MI6>libuv-master: #138 FAILURE windows (4/191) smartos (2/190) http://jenkins.nodejs.org/job/libuv-master/138/
17:18:16  <isaacs>anyone else getting a zillion unused variable _heap warnings on os x?
17:18:29  <isaacs>is this normal? or some xcode setting to be annoying that i can turn off?
17:19:04  <tjfontaine>I am seeing them, and it's probably a relatively new diagnostic I think
17:24:14  * dsantiagoquit (Quit: Leaving...)
17:27:14  <isaacs>k
17:27:18  <isaacs>i'll ignore them, then
17:29:24  * csaohquit (Quit: csaoh)
17:32:12  * dsantiagojoined
17:37:16  <MI6>libuv-master: #139 FAILURE windows (3/191) smartos (2/190) http://jenkins.nodejs.org/job/libuv-master/139/
17:44:34  <MI6>libuv-master: #140 UNSTABLE windows (3/191) smartos (2/190) http://jenkins.nodejs.org/job/libuv-master/140/
17:45:47  * bnoordhuisjoined
17:46:44  <tjfontaine>bnoordhuis: the linux slave is using automake1.9 atm
17:48:30  <bnoordhuis>tjfontaine: ah, that's positively vintage
17:48:34  <tjfontaine>quite
17:48:38  <bnoordhuis>is that an rhel 5 system or?
17:48:58  <tjfontaine>nah, ubuntu12.04 iirc, but oyu know if it wasn't us it'd be someone else :)
17:49:10  <tjfontaine>at least that's what automake in /etc/alternatives points to
17:49:13  <MI6>libuv-master: #141 UNSTABLE osx (1/191) windows (3/191) smartos (2/190) linux (1/190) http://jenkins.nodejs.org/job/libuv-master/141/
17:49:26  <bnoordhuis>ah okay
17:58:23  * wolfeidaujoined
17:59:35  * `3E|DINNERchanged nick to `3rdEden
17:59:38  * wavdedquit (Ping timeout: 246 seconds)
17:59:38  <MI6>libuv-node-integration: #129 FAILURE smartos-ia32 (1/609) smartos-x64 (7/609) http://jenkins.nodejs.org/job/libuv-node-integration/129/
18:00:19  <tjfontaine>this libtool and osx thing is a pain in the ass
18:02:54  * wolfeidauquit (Ping timeout: 256 seconds)
18:10:24  * c4miloquit (Remote host closed the connection)
18:12:11  * wavdedjoined
18:14:42  <MI6>libuv-node-integration: #130 FAILURE smartos-ia32 (1/609) smartos-x64 (7/609) http://jenkins.nodejs.org/job/libuv-node-integration/130/
18:15:15  <bnoordhuis>tjfontaine: libtool?
18:15:23  * bajtosquit (Quit: bajtos)
18:15:35  <bnoordhuis>the autogen script should pick the right libtool on os x
18:15:50  <tjfontaine>it doesn't, libtoolize isn't there in my 10.8
18:16:39  <mmalecki>yeah, it's a pain. you either need to install it from brew or do some serious fucking around
18:17:08  <tjfontaine>I opted for serious fucking around, because I fucking hate, *hate* not using the system as provided
18:17:24  <tjfontaine>brew/ports/pkgsrc etc all will lead to unfun worlds of mixed toolchains
18:17:40  <mmalecki>I traded that for convenience, tbh
18:18:09  <tjfontaine>mmalecki: for a CI though I need to provide generally an environment that matches what we might expect our lusers to be in
18:18:20  <mmalecki>true
18:18:29  <mmalecki>or we could, y'know, not make that harder for them ;)
18:18:55  <tjfontaine>autohell is really only a disaster area on later osx
18:20:29  * Benviequit (Remote host closed the connection)
18:20:57  * Benviejoined
18:27:55  * dominictarrquit (Quit: dominictarr)
18:28:49  <MI6>libuv-review: #23 FAILURE windows-ia32 (3/191) windows-x64 (4/191) smartos-ia32 (2/190) smartos-x64 (2/190) http://jenkins.nodejs.org/job/libuv-review/23/
18:30:37  <bnoordhuis>tjfontaine: we could ship libtool with libuv, it's a single shell script
18:31:55  <bnoordhuis>otoh, any developer on os x probably uses brew
18:32:39  <tjfontaine>doubtful, there are plenty of other purists like me
18:33:08  <tjfontaine>anyway, I have the ci in a world I'm ok with for now
18:43:20  * Benviequit
18:44:27  * Benviejoined
18:47:09  * mikolalysenkoquit (Ping timeout: 252 seconds)
18:47:10  * defunctzombiechanged nick to defunctzombie_zz
18:51:20  * defunctzombie_zzchanged nick to defunctzombie
18:51:55  <MI6>libuv-review: #24 FAILURE windows-ia32 (3/191) windows-x64 (3/191) smartos-ia32 (2/190) smartos-x64 (2/190) http://jenkins.nodejs.org/job/libuv-review/24/
18:52:06  * bnoordhuisquit (Ping timeout: 246 seconds)
18:52:10  * mcavagejoined
18:55:07  * hzjoined
19:04:18  * wavdedquit (Quit: Nighty night)
19:04:57  * EhevuTovjoined
19:07:41  * wavdedjoined
19:07:46  * c4milojoined
19:13:00  * EhevuTovquit (Quit: This computer has gone to sleep)
19:13:30  * EhevuTovjoined
19:13:55  * mikolalysenkojoined
19:14:53  * kazuponquit (Remote host closed the connection)
19:29:15  * wavdedquit (Quit: Nighty night)
19:32:17  * wavdedjoined
19:33:55  * AvianFluquit (Remote host closed the connection)
19:34:15  * dominictarrjoined
19:38:42  <isaacs>tjfontaine: oh, i just realized... i'm going to have to set up my mac os x signing stuff all over again. ToT
19:39:12  <TooTallNate>isaacs: :(
19:39:21  <tjfontaine>isaacs: heh, the good news is you just open up xcode (or reuse that usb key) and export and then on the new one import
19:39:29  <tjfontaine>it's pretty straight forward
19:39:42  <TooTallNate>xcode scares me
19:39:58  <tjfontaine>this part of it is actually relatively smooth
19:58:57  * bnoordhuisjoined
20:00:49  <kellabyte>whats the proper way to handle on_body in http_parser since it gets called multiple times? same approach as the headers? I seem to be getting some funky characters. I've done something wrong!
20:03:36  * bnoordhuisquit (Ping timeout: 256 seconds)
20:10:53  * wavdedquit (Quit: Nighty night)
20:12:49  * wavdedjoined
20:15:25  * kazuponjoined
20:15:40  * defunctzombiechanged nick to defunctzombie_zz
20:17:55  * defunctzombie_zzchanged nick to defunctzombie
20:18:27  * loladirojoined
20:18:50  <kellabyte>false alarm, null terminator issue :P *noob*
20:20:21  <tjfontaine>pesky \0's getting in the way
20:21:50  <kellabyte>yeah lol
20:22:45  <kellabyte>one thing I'm curious about with http_parser is can I just store the pointer location and length for header values or body and only memcpy the ones used? or will http_parser release some of this memory during parsing the request?
20:23:04  <tjfontaine>"unbounded memory increase" #5780, starts off by saying: "The code caches" <-- well there's your problem.
20:23:18  <kellabyte>trying to figure out if I can essentially not copy strings until the higher level requests them saving a lot of mem copying
20:23:58  <tjfontaine>kellabyte: I haven't really looked into the design of it, but if it's like the rest of our api it's zero copy, meaning you are responsible for allocating and freeing the space
20:24:00  * kazuponquit (Ping timeout: 256 seconds)
20:25:27  <kellabyte>tjfontaine: it seems to work if I don't copy either, and I just store the pointer and length, but I'm not sure if I'm just getting really lucky
20:26:12  <tjfontaine>kellabyte: I can't imagine a world where the parser would free something out from under you
20:26:13  * loladiroquit (Quit: loladiro)
20:26:34  <kellabyte>tjfontaine: but its not going to keep the whole request in memory if its like a multi-megabyte body is it?
20:27:11  <tjfontaine>kellabyte: the "it" that's keeping the memory around is you
20:27:55  <kellabyte>tjfontaine: yeah, thats what I'm saying, I was testing *not* storing the values and only storing the pointers
20:28:03  <kellabyte>tjfontaine: and it seems to work, but I'm not sure if I'm getting lucky
20:29:04  <tjfontaine>kellabyte: are you free'ing the blocks later? if you're explicitly freeing things then you're probably getting lucky, if you're not free'ing things then it's working, unless that's not what you meant to do I guess
20:29:25  <tjfontaine>is there code online for this?
20:30:38  <kellabyte>no but I can gist this part
20:30:44  <tjfontaine>k
20:30:45  <kellabyte>I'm not freeing anything because I'm not allocating anything
20:33:02  <kellabyte>tjfontaine: https://gist.github.com/kellabyte/5912838
20:34:30  <tjfontaine>kellabyte: you are allocating what goes in the _execute side though
20:35:16  * jmar777quit (Remote host closed the connection)
20:35:53  * jmar777joined
20:36:41  <kellabyte>tjfontaine: hmm, its just being fed by uv_read_start() and the callback it calls
20:37:08  <tjfontaine>kellabyte: right, you mean the alloc_cb?
20:37:13  <kellabyte>yeah
20:37:26  <kellabyte>and using the suggested size
20:37:58  <tjfontaine>ya, so you are allocating this memory, libuv shoves some stuff into it, and you then pass it straight to http_parser, so if you're not free'ing it at the end of the road, you're "leaking" it
20:39:26  * defunctzombiechanged nick to defunctzombie_zz
20:39:33  <kellabyte>hmm strange, I don't see a leak, and its requesting 65K, so I would notice
20:39:40  <kellabyte>I've run valgrind and instruments on it
20:40:28  <kellabyte>I'll have to look into that I guess
20:40:40  <tjfontaine>well if you're keeping the buf around in another structure then it's not a full on leak because it assumes you know that you're asking to keep this stuff around
20:40:46  * jmar777quit (Ping timeout: 276 seconds)
20:41:42  <kellabyte>tjfontaine: hmm, ok I'll have to look into it, I'm not holding on to it: https://github.com/kellabyte/Haywire/blob/master/src/haywire/http_server.c#L146
20:42:28  <tjfontaine>kellabyte: https://github.com/kellabyte/Haywire/blob/master/src/haywire/http_server.c#L186
20:42:54  <kellabyte>tjfontaine: ahh, thats the same buffer round tripped?
20:43:37  <kellabyte>ok that makes sense
20:43:57  <kellabyte>tjfontaine: so what happens if something is bigger than 65K?
20:45:24  <tjfontaine>kellabyte: so alloc_cb gets a suggested size (which is always 65k for now) but you can allocate more, or less depending about what you know about the underlying stream, or what resources you have available, from there the actual underlying read will use the size you did allocate so alloc(1MB) read(1MB)
20:46:53  <kellabyte>tjfontaine: does the whole request have to fit into the buffer or if I give it 65K will it just keep asking for more chunks?
20:47:21  <tjfontaine>right, so if there is more to read than what you ask for the callback's just keep happening until you're back in a state where you need to wait
20:47:45  <tjfontaine>or you've called read stop or similar
20:49:23  <kellabyte>ok lemme see if I understood that properly, so if a request is 1MB and the suggested size is 65K, I'll get multiple calls for 65K, and my free(buf) would happen multiple times, so then my zero copy pointers would be gone
20:50:05  <tjfontaine>so, if you allocate 1mb in your alloc_cb it will try and read up to 1mb, the callback lets you know just how much it actually did read (often that will be far less than what you told it to read)
20:50:52  * hzquit
20:51:23  <kellabyte>but your suggesting I make the buffer size the same size as the request, I won't know the size of the request if its bigger than the buffer
20:52:01  <tjfontaine>I'm not suggesting anything, just kinda laying out more about how the system works
20:52:18  <kellabyte>nods
20:52:29  <kellabyte>I'm trying to figure out how to handle a request thats bigger than my buffer
20:52:52  * hzjoined
20:52:56  <kellabyte>it sounds like the way I have it now, only works if the buffer is large enough to fit the whole request, because it'll free the buffer multiple times, losing my pointers
20:56:16  * loladirojoined
20:56:19  <kellabyte>so I'm Doing It Wrong TM
20:56:44  <tjfontaine>right so as you've designed it right now, you'll get what ever http parser cbs you can get from that read cb, and then you free the buffer, so if it doesn't make it all in one read() the pointers will be invalid by the time you try to use them later
20:57:02  <kellabyte>yeha
20:57:04  <kellabyte>err yeah
20:57:37  <tjfontaine>or work out which buffers you need to retain, and hold them until you're ready to close the connection and free them there
20:58:28  <kellabyte>yeah. I suppose it doesn't matter if I retain the data in that buffer or in a memcpy but, if I leave it in the buffer thats one less alloc+copy
20:58:40  <tjfontaine>yup
20:58:46  <tjfontaine>another pattern you should be aware of
20:59:30  <tjfontaine>it's likely that the read() didn't consume all 64k, and nread is less than the length, so you probably want to realloc to the real size, which will prevent you from using up too much extra memory
21:01:38  * julianduquejoined
21:01:39  <kellabyte>ah, thanks for pointing that out. I haven't seen any memory growth much even at 350,000 requests/s but they are really small requests
21:02:05  <kellabyte>probably want to take that into consideration when the requests start attaching a bunch of json or something
21:02:12  <tjfontaine>generally we have lots of memory, and the OS is really good at handing it all out
21:03:07  <kellabyte>ok. thanks a lot for that info! super glad I asked the question :) I'll probably stick to simple right now, but I'm going to write it down cuz I need to tackle that buffer usage
21:03:32  <tjfontaine>nod
21:03:35  <kellabyte>I wonder if I can do something smart where it's likely headers etc fits in 65K, so when I hit the body length
21:03:49  <kellabyte>I could re-allocate at that point if needed maybe
21:05:43  * wolfeidaujoined
21:05:56  <kellabyte>ok lots to think about, thanks so much :)
21:06:09  <tjfontaine>hope I helped and didn't make it worse :P
21:06:33  <kellabyte>lol you made it harder but I learned a lot and its stuff I need to know!
21:06:41  <tjfontaine>hehe
21:14:09  * EhevuTov_joined
21:17:00  * EhevuTovquit (Ping timeout: 246 seconds)
21:20:28  * kazuponjoined
21:23:02  * julianduquequit (Ping timeout: 256 seconds)
21:25:36  * kazuponquit (Ping timeout: 256 seconds)
21:27:52  * rjequit (Ping timeout: 256 seconds)
21:30:08  * rjejoined
21:39:59  * dominictarrquit (Quit: dominictarr)
21:40:25  * defunctzombie_zzchanged nick to defunctzombie
21:46:16  * wavdedquit (Quit: Nighty night)
21:49:47  * rendarquit
21:50:14  * defunctzombiechanged nick to defunctzombie_zz
21:58:38  * EhevuTov_quit (Quit: Leaving)
22:09:24  * mikolalysenkoquit (Ping timeout: 256 seconds)
22:09:36  <tjfontaine>groundwater: btw aszeszo is in #smartos if you wanted to bounce ideas off of him
22:16:20  * dominictarrjoined
22:17:43  * julianduquejoined
22:20:20  <groundwater>tjfontaine: nice, thanks
22:20:44  * timoxleyjoined
22:20:49  <tjfontaine>no problemo
22:21:14  * kazuponjoined
22:21:45  * defunctzombie_zzchanged nick to defunctzombie
22:22:11  <groundwater>tjfontaine: feel free to chime in on the git repo, i think i linked you to it in a tweet
22:22:20  <tjfontaine>groundwater: ya, I'm watching it as well
22:25:56  * kazuponquit (Ping timeout: 256 seconds)
22:27:00  * julianduquequit (Ping timeout: 268 seconds)
22:31:19  * defunctzombiechanged nick to defunctzombie_zz
22:32:55  * brsonquit (Ping timeout: 264 seconds)
22:33:37  * brsonjoined
22:37:56  * c4miloquit (Remote host closed the connection)
22:47:52  * pachetquit (Quit: leaving)
22:51:45  * kazuponjoined
22:56:32  * kazuponquit (Ping timeout: 256 seconds)
23:06:35  * defunctzombie_zzchanged nick to defunctzombie
23:10:51  * kazuponjoined
23:15:13  * defunctzombiechanged nick to defunctzombie_zz
23:16:21  * kazuponquit (Ping timeout: 248 seconds)
23:20:04  * defunctzombie_zzchanged nick to defunctzombie
23:21:24  * stagasquit (Ping timeout: 240 seconds)
23:31:57  * julianduquejoined
23:38:46  * defunctzombiechanged nick to defunctzombie_zz
23:53:30  * julianduquequit (Ping timeout: 256 seconds)
23:57:39  * mikolalysenkojoined