00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:09  * ircretaryjoined
00:03:09  * groundwaterquit (Quit: groundwater)
00:04:00  * defunctzombie_zzchanged nick to defunctzombie
00:05:33  * dsantiagoquit (Ping timeout: 264 seconds)
00:09:22  * dsantiagojoined
00:13:41  * jmar777joined
00:18:37  * mikealjoined
00:22:31  * mikealquit (Client Quit)
00:28:27  * mikealjoined
00:28:49  * mikealquit (Client Quit)
00:31:23  * wavdedjoined
00:33:44  * julianduquejoined
00:38:27  * defunctzombiechanged nick to defunctzombie_zz
00:40:20  * AvianFlu_joined
00:43:57  * AvianFluquit (Ping timeout: 264 seconds)
01:01:34  * twilightfeelquit (Quit: twilightfeel)
01:02:15  * twilightfeeljoined
01:12:49  * twilightfeelquit (Quit: twilightfeel)
01:13:53  * groundwaterjoined
01:25:13  * st_lukejoined
01:43:55  <isaacs>huh. when stdin is a pipe, just pausing it doesn't unref
01:47:25  <isaacs>that seems like a bug
02:07:06  * defunctzombie_zzchanged nick to defunctzombie
02:13:22  * c4milojoined
02:27:46  * defunctzombiechanged nick to defunctzombie_zz
02:29:48  * c4miloquit (Remote host closed the connection)
02:34:21  * jmar777quit (Remote host closed the connection)
02:34:54  * jmar777joined
02:39:27  * jmar777quit (Ping timeout: 256 seconds)
03:08:40  <MI6>joyent/node: isaacs v0.10 * e48536f : tls: Trivial use_strict fix - http://git.io/uatolQ
03:21:30  <MI6>nodejs-v0.10: #285 UNSTABLE smartos-ia32 (1/594) http://jenkins.nodejs.org/job/nodejs-v0.10/285/
03:22:40  <tjfontaine>so fucking close to a green build
03:27:42  <MI6>nodejs-v0.10-windows: #113 UNSTABLE windows-x64 (8/594) windows-ia32 (8/594) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/113/
03:37:40  * defunctzombie_zzchanged nick to defunctzombie
03:47:45  * defunctzombiechanged nick to defunctzombie_zz
03:54:03  * wavdedquit (Quit: Hasta la pasta)
03:57:55  <kellabyte>I hate linkers
03:58:05  <tjfontaine>amen.
03:58:11  <tjfontaine>especially link.exe
03:59:13  <kellabyte>lol
03:59:37  <kellabyte>./builds/unix/debug/obj.target/lib/libuv/libuv.a(fs-poll.o): could not read symbols: Bad value
04:00:48  <ik>DAYCHANGE!
04:01:14  <kellabyte>lol
04:01:40  <tjfontaine>LOUDBOT: SEARCH UTC
04:01:40  <LOUDBOT>tjfontaine: <YTZ:##church-of-loudbot> SHUTCHER ZOMBIE HOLE. A MOMENT OF SILENCE FOR JACKO BEFORE THE FEASTING BEGINZ PLZ.
04:02:05  <kellabyte>have no idea why the linker is complaining at libuv.a like that
04:02:25  <kellabyte>its erroring on a new shared lib I'm compiling though, which links to libuv
04:07:43  <kellabyte>tjfontaine: can a shared lib link to a static lib? seemed to work fine for VS but not so much GCC maybe
04:15:07  * defunctzombie_zzchanged nick to defunctzombie
04:18:26  * AvianFlu_quit (Remote host closed the connection)
04:23:32  <tjfontaine>kellabyte: yes you should be able to, no particular reason you can't
04:25:23  <kellabyte>tjfontaine: strange, it says "could not read symbols: Bad value", a google says I need to compile libuv with -fPIC
04:26:34  <kellabyte>tjfontaine: I tried -fPIC in my shared lib, but that didn't seem to fix it
04:27:22  <kellabyte>/usr/bin/ld: ./builds/unix/debug/obj.target/lib/libuv/libuv.a(fs-poll.o): relocation R_X86_64_32 against `.text' can not be used when making a shared object; recompile with -fPIC
04:29:30  * c4milojoined
04:30:34  * c4miloquit (Remote host closed the connection)
04:32:59  * defunctzombiechanged nick to defunctzombie_zz
04:36:36  <kellabyte>I wonder if I'm compiling libuv as 32bit and my code as 64? is that what thats saying?
04:48:23  * bradleymeckjoined
05:07:13  * bradleymeckquit (Quit: bradleymeck)
05:20:31  * defunctzombie_zzchanged nick to defunctzombie
05:24:57  * paddybyersjoined
05:28:59  * kenperkinsquit (Quit: Computer has gone to sleep.)
05:30:14  * bajtosjoined
05:48:17  * defunctzombiechanged nick to defunctzombie_zz
06:04:09  * inolenquit (Quit: Leaving.)
06:24:20  * mikealjoined
06:25:19  * julianduquequit (Ping timeout: 246 seconds)
06:29:02  * hzjoined
06:51:25  * inolenjoined
06:55:23  * mikealquit (Quit: Leaving.)
06:58:12  * defunctzombie_zzchanged nick to defunctzombie
07:02:27  * chrisdickinsonquit (Quit: ZNC - http://znc.sourceforge.net)
07:03:09  * chiltsquit (Ping timeout: 264 seconds)
07:03:45  * chiltsjoined
07:04:26  * chrisdickinsonjoined
07:08:12  * defunctzombiechanged nick to defunctzombie_zz
07:32:53  * hzquit
07:35:21  * rendarjoined
07:37:00  * st_lukequit (Remote host closed the connection)
07:39:03  * csaohjoined
07:55:10  * einarosjoined
07:55:20  * einarosquit (Client Quit)
07:56:35  * einarosjoined
07:56:53  * rendarquit (Ping timeout: 248 seconds)
08:11:06  * rendarjoined
08:28:09  * loladiroquit (Quit: loladiro)
09:00:58  * kazuponjoined
09:02:32  * kazuponquit (Remote host closed the connection)
09:06:16  * mikealjoined
09:06:36  * mikeal1joined
09:06:40  * mikealquit (Read error: Connection reset by peer)
09:10:34  * mraleph1joined
09:10:35  * mralephquit (Read error: Connection reset by peer)
09:11:35  * mikeal1quit (Ping timeout: 264 seconds)
09:37:29  * mikealjoined
09:41:40  * mikealquit (Ping timeout: 246 seconds)
09:45:35  * csaohquit (Ping timeout: 256 seconds)
09:50:08  * csaohjoined
09:56:33  <mmalecki>anyone can think of any reason why binding libssh2 to uv might be a bad idea?
10:04:55  * leonvvjoined
10:08:25  * mikealjoined
10:09:34  * mikealquit (Read error: Connection reset by peer)
10:34:43  * groundwaterquit (Ping timeout: 245 seconds)
10:37:22  * groundwaterjoined
10:39:57  * piscisaureus_joined
10:41:06  * hzjoined
10:44:59  * mikealjoined
10:47:26  * leonvvquit (Remote host closed the connection)
10:49:09  * csaohquit (Quit: csaoh)
10:49:37  * mikealquit (Ping timeout: 256 seconds)
10:50:19  * csaohjoined
11:03:07  * csaohquit (Quit: csaoh)
11:06:16  * csaohjoined
11:19:50  * mikealjoined
11:24:47  * mikealquit (Ping timeout: 264 seconds)
11:29:34  * piscisaureus_quit (Ping timeout: 276 seconds)
11:35:34  * twilightfeeljoined
11:50:23  * mikealjoined
11:55:23  * mikealquit (Ping timeout: 264 seconds)
12:00:08  * piscisaureus_joined
12:06:26  * twilightfeelquit (Quit: twilightfeel)
12:07:29  * twilightfeeljoined
12:21:20  * mikealjoined
12:25:33  * mikealquit (Ping timeout: 245 seconds)
12:33:00  * AvianFlujoined
12:47:45  * dsantiagoquit (Quit: Computer has gone to sleep.)
12:59:19  * jmar777joined
13:20:22  * AvianFluquit (Remote host closed the connection)
13:20:34  * mikealjoined
13:26:10  * AvianFlujoined
13:27:47  * mikealquit (Ping timeout: 264 seconds)
13:31:26  * julianduquejoined
13:47:33  <bajtos>is it an expected behaviour that TLS server does not emit 'newSession' for a TLSv1 client? It still does emit 'resumeSession' though.
13:49:44  <bajtos>I am getting the same behaviour in 0.11.4, 0.11.3, and 0.10.12/13. Adding opts.secureProtocol = 'SSLv3_method' removes the issue.
14:07:23  * twilightfeelquit (Quit: twilightfeel)
14:07:48  * bradleymeckjoined
14:11:16  * c4milojoined
14:15:08  * kenperkinsjoined
14:16:37  * bradleymeckquit (Quit: bradleymeck)
14:24:29  * pachetjoined
14:24:29  * pachetquit (Changing host)
14:24:29  * pachetjoined
14:28:00  * c4miloquit (Remote host closed the connection)
14:30:13  <isaacs>bajtos: that sounds like a bug
14:33:16  <bajtos>isaacs: thanks. I'll write a script to reproduce the problem and submit a GH issue
14:37:57  * mikealjoined
14:38:38  * jmar777quit (Read error: Connection reset by peer)
14:38:53  * jmar777joined
14:40:00  * mikealquit (Client Quit)
14:40:36  * twilightfeeljoined
15:17:15  * mikealjoined
15:19:08  * mikealquit (Client Quit)
15:25:01  * julianduquequit (Ping timeout: 256 seconds)
15:25:28  * c4milojoined
15:35:35  * leonvvjoined
15:38:32  * TooTallNatejoined
15:39:56  * kenperkinsquit (Read error: Connection reset by peer)
15:40:35  * kenperkinsjoined
15:43:38  * pachet_joined
15:43:50  * pachetchanged nick to Guest46410
15:45:31  * pachet_changed nick to pachet
15:45:35  * pachetquit (Changing host)
15:45:35  * pachetjoined
15:45:47  * Guest46410quit (Ping timeout: 264 seconds)
15:56:16  * twilightfeelquit (Quit: twilightfeel)
15:57:09  * loladirojoined
16:07:32  * Damn3dquit (Ping timeout: 250 seconds)
16:13:06  <M28>I embedded lua in my application...
16:13:10  <M28>that was a huge mistake
16:13:16  * Damn3djoined
16:13:29  <M28>should've gone with v8
16:15:12  <tjfontaine>could have had a v8 http://www.youtube.com/watch?v=ShcbV31tcAk
16:16:43  * defunctzombie_zzchanged nick to defunctzombie
16:24:51  * mikealjoined
16:25:42  * mcavagejoined
16:26:50  * leonvvquit (Remote host closed the connection)
16:31:49  * loladiroquit (Quit: loladiro)
16:39:37  * mikealquit (Quit: Leaving.)
16:40:58  * sblomjoined
16:45:31  * sblomquit (Ping timeout: 246 seconds)
16:47:58  * sblomjoined
16:51:21  <bajtos>piscisaureus_: rember your concern of TLS session store performance on windows? win8 in virtualbox shows +33% performance gain
16:51:42  <piscisaureus_>bajtos: nice - is that with the sessionstore backend?
16:51:51  <piscisaureus_>er- the store-cluster backend?
16:52:18  <bajtos>piscisaureus_: yes. are you in windows now? could you run the benchmark for me on a real HW without VM?
16:53:25  * Benvie_quit
16:55:42  * Benviejoined
17:09:27  * loladirojoined
17:12:41  * c4miloquit (Remote host closed the connection)
17:13:39  * DrPizzajoined
17:14:04  * mikealjoined
17:16:02  * mikeal1joined
17:16:46  * mikealquit (Read error: Connection reset by peer)
17:23:37  <trevnorris>morning
17:23:51  * mikeal1quit (Quit: Leaving.)
17:25:30  * mikealjoined
17:33:27  * mikealquit (Quit: Leaving.)
17:42:03  * mikealjoined
17:42:25  * mikealquit (Client Quit)
17:44:56  * csaohquit (Quit: csaoh)
17:46:34  * mikealjoined
17:49:46  * mikealquit (Client Quit)
17:50:28  * mikealjoined
17:54:15  <trevnorris>TooTallNate: poing
17:54:29  <TooTallNate>trevnorris: sup?
17:54:46  <trevnorris>re the issue about removing the assert
17:54:56  <trevnorris>does it make sense why I'm fine just removing it completely?
17:55:30  * c4milojoined
17:57:44  <TooTallNate>trevnorris: ya that's probably fine
17:58:08  <TooTallNate>removing the assert, wouldn't that allow Buffers pointing to NULL with a .length > 0 though?
17:58:25  <TooTallNate>do we want to guard people against that in core? or do you think addon devs should deal with it?
17:58:33  <trevnorris>yeah. but it's already possible to "char* data = new char(16)" then pass length > 16
17:58:46  <trevnorris>so imo devs already have to deal with it.
17:58:54  <TooTallNate>good point
17:59:03  <TooTallNate>so then +1 on removing it completely
17:59:15  <trevnorris>coolio. i'll take care of that in a bit.
18:00:09  <TooTallNate>im also flattered that that guy is updating ref to the new API for me :p
18:00:22  <TooTallNate>people really <3 generators it seems
18:00:40  <trevnorris>yeah. too bad they're so slow right now. hopefully v8 figures that out.
18:00:46  * sblomquit (Ping timeout: 246 seconds)
18:01:01  <trevnorris>tjfontaine: tell jenkins sorry for me :P
18:01:23  <tjfontaine>what did you do?!
18:01:34  <tjfontaine>oh 3 pushes
18:01:41  <tjfontaine>4
18:01:43  <trevnorris>just have a lot of PR's open that had to rebase after the weekend
18:01:50  <tjfontaine>heh
18:02:49  * sblomjoined
18:02:57  <jmar777>TooTallNate: yup :)
18:04:33  <jmar777>I'm currently arguing for lexically (rather than immediately) scoped yields on es-discuss. would make a lot of scenarios so much easier.
18:10:29  * defunctzombiechanged nick to defunctzombie_zz
18:14:06  <MI6>joyent/node: Trevor Norris master * be940b4 : smalloc: remove NULL assert check - http://git.io/Adl07g
18:14:07  <trevnorris>TooTallNate: ^
18:14:14  <TooTallNate>trevnorris: <3
18:16:03  <tjfontaine>blog I'm working on for nodejs.org https://gist.github.com/tjfontaine/00e04f628c0ef55e7cb8
18:17:24  <trevnorris>tjfontaine: i just received an email from travis-ci about node build failing. that anything to do with you?
18:18:29  <tjfontaine>trevnorris: nope, someone else has an old branch or has turned it back on
18:18:37  <trevnorris>ugh. wtf.
18:19:55  <tjfontaine>it shoudl tell you the branch it came from
18:22:18  <trevnorris>tjfontaine: what do you mean branch?
18:22:43  <tjfontaine>the contents of the message from travis should tell you whose repo and branch triggered it
18:23:34  <trevnorris>tjfontaine: says joyent/node
18:23:47  <tjfontaine>gist the email please
18:24:02  <tjfontaine>preferrably in the "show original" format
18:25:46  <trevnorris>tjfontaine: email: https://gist.github.com/trevnorris/6002197
18:25:59  <trevnorris>build: https://travis-ci.org/joyent/node/builds/9106608
18:26:10  <tjfontaine>wtf travis
18:27:55  * pachetquit (Read error: Connection reset by peer)
18:28:05  <tjfontaine>it's not supposed to build if there's no .travis.yaml or whatever, anyway I've disabled it in the service hooks
18:28:12  <trevnorris>thanks
18:28:15  <TooTallNate>tjfontaine: i lol @ the arm-pi build being a good % of downloads
18:28:30  <TooTallNate>well a small %
18:28:37  <tjfontaine>TooTallNate: they show up at all :)
18:28:44  <TooTallNate>ya that
18:28:44  <TooTallNate>haha
18:29:08  <tjfontaine>I'm not done yet, still thinking about other interesting questions, I have one for Homebrew
18:29:19  <tjfontaine>anything you guys want to see included?
18:31:09  <MI6>nodejs-master: #307 UNSTABLE smartos-ia32 (5/612) smartos-x64 (9/612) http://jenkins.nodejs.org/job/nodejs-master/307/
18:35:58  <MI6>nodejs-master-windows: #113 UNSTABLE windows-ia32 (14/612) windows-x64 (17/612) http://jenkins.nodejs.org/job/nodejs-master-windows/113/
18:36:38  * defunctzombie_zzchanged nick to defunctzombie
18:39:47  * bnoordhuisjoined
18:40:39  * bajtosquit (Quit: bajtos)
18:41:49  * sblomquit (Ping timeout: 276 seconds)
18:44:48  * sblomjoined
18:53:09  <trevnorris>well. note to self. don't accidentally run wrk opening 2m connections :P
18:53:44  <TooTallNate>tjfontaine: does Homebrew use a predictable User-Agent?
18:54:38  <tjfontaine>TooTallNate: ya, refresh the gist
18:54:43  <TooTallNate>oh
18:55:27  * werlejoined
18:59:41  * EhevuTovjoined
19:03:06  <trevnorris>this http memory growth issue is really buggin
19:04:40  <tjfontaine>saghul: you around?
19:14:42  <trevnorris>bnoordhuis: ping
19:18:43  <bnoordhuis>trevnorris: pong
19:19:15  <trevnorris>bnoordhuis: so this mem growth thing on http is real. left my machine running for hours and it reached half a gig.
19:19:21  <trevnorris>having set space size to 16 MB
19:19:45  <trevnorris>it's an almost perfectly linear growth. i'm examine it now.
19:20:02  <trevnorris>what I can't determine is the difference between the request of wrk and ab.
19:20:20  <bnoordhuis>trevnorris: i'm still on sick leave, no deep forensics for me tonight
19:20:28  <trevnorris>bnoordhuis: ooh. sorry. :)
19:24:13  <bnoordhuis>trevnorris: np. :) okay, afk again
19:25:28  <isaacs>trevnorris: can you reproduce this on smartos? if so, can you get a core dump? `gcore $pid`
19:25:43  <isaacs>trevnorris: ssh [email protected]
19:26:11  <isaacs>trevnorris: then, we can see what's in the heap with mdb and ::find-js-objects, and track down what's leaking
19:27:29  <mmalecki>bnoordhuis: hey, a little question about your recent change to libuv regarding build systems
19:27:44  <mmalecki>bnoordhuis: I was wondering, how would I include libuv in projects with makefile only?
19:27:54  <mmalecki>bnoordhuis: stuff I came up with was ugly
19:28:39  * bnoordhuisquit (Ping timeout: 260 seconds)
19:30:02  <trevnorris>isaacs: will give it a go.
19:35:31  * EhevuTov_joined
19:38:09  * EhevuTovquit (Ping timeout: 252 seconds)
19:38:21  * EhevuTov_quit (Client Quit)
19:47:10  * chrisdickinsonchanged nick to erkstyn
19:47:15  * erkstynchanged nick to chrisdickinson
19:48:40  * pachetjoined
19:49:05  <tjfontaine>mmalecki: ugh, we just can't win can we (re: build tools)
19:49:57  <mmalecki>tjfontaine: there's no win in build systems, only tears and blood
19:50:16  <tjfontaine>yes.
19:50:20  <tjfontaine>lots of heartbreak
19:50:38  * julianduquejoined
19:51:14  <trevnorris>isaacs: no mem growth on smartos.
19:51:58  <isaacs>trevnorris: ha!
19:52:01  <isaacs>trevnorris: interesting.
19:52:03  <isaacs>are you sure?
19:52:16  <tjfontaine>what are you using to measure?
19:52:18  <isaacs>trevnorris: how long did you let it run for?
19:52:43  <trevnorris>isaacs: on my machine the memory growth is almost mathematically predictable.
19:52:53  <trevnorris>and that doesn't happen on smartos
19:53:19  <trevnorris>ah, wait.
19:53:30  <trevnorris>it might just be taking a lot longer.
19:53:36  <trevnorris>i forget how fast my machine is
19:54:03  <trevnorris>tjfontaine: printing all of process.memoryUsage()
19:54:11  <tjfontaine>ok
19:54:43  <trevnorris>ok. i'm going to let it run for another 30 mins or so to make sure this is happening.
19:57:17  * Raltquit (Quit: Bye)
19:59:00  * mikealquit (Quit: Leaving.)
20:02:16  * piscisaureus_quit (Ping timeout: 260 seconds)
20:02:18  <trevnorris>isaacs: got it.
20:02:23  * stagasjoined
20:06:07  * indexzerojoined
20:07:22  * piscisaureus_joined
20:17:14  * mcavagequit (Read error: Connection reset by peer)
20:17:37  * mcavagejoined
20:19:50  <trevnorris>isaacs: ping
20:19:56  <trevnorris>tjfontaine: or maybe you can help me
20:20:12  <tjfontaine>sup
20:20:23  * julianduquequit (Ping timeout: 240 seconds)
20:20:23  * hueniversequit (Ping timeout: 240 seconds)
20:20:23  * julianduquejoined
20:20:28  * julianduquequit (Changing host)
20:20:28  * julianduquejoined
20:20:31  <trevnorris>::find-js-leaks isn't working in mdb
20:20:34  * hueniversejoined
20:20:40  <tjfontaine>find-js-leaks isn't a thing, afaik
20:20:43  <trevnorris>yeah
20:20:48  <trevnorris>::findleaks works
20:20:49  <trevnorris>output: 5 oversized leaks, 253952 bytes
20:20:55  <isaacs>pong
20:20:56  <tjfontaine>they're different things
20:20:59  <trevnorris>ah, ok
20:21:10  <trevnorris>isaacs: how am I supposed to use mdb?
20:21:20  <isaacs>mdb the-core-file.cor
20:21:22  <isaacs>*core
20:21:28  <tjfontaine>http://dtrace.org/blogs/bmc/2012/05/05/debugging-node-js-memory-leaks/
20:21:33  <isaacs>trevnorris: then, in there, do ::load v8.so
20:21:38  <trevnorris>tjfontaine: reading that now :)
20:21:46  <isaacs>trevnorris: you might need the full path to v8.so. if that doesn't work, try ::load /home/isaacs/v8.so
20:22:53  <trevnorris>isaacs: ah, cool. working.
20:27:36  <trevnorris>isaacs: after a socket is open, is it supposed to hang around after the connection closes?
20:27:55  * indexzeroquit (Quit: indexzero)
20:31:26  <trevnorris>how the crap long is <value>::findjsobjects -r supposed to take?
20:34:40  * bnoordhuisjoined
20:34:56  <trevnorris>isaacs, tjfontaine: https://gist.github.com/trevnorris/6003201
20:38:57  * bnoordhuisquit (Ping timeout: 246 seconds)
20:40:38  <trevnorris>so that doesn't help me much. why are there so many un-reffered objects?
20:41:04  * jmar777quit (Remote host closed the connection)
20:41:38  * jmar777joined
20:43:24  * AvianFluquit (Read error: Connection reset by peer)
20:44:02  * AvianFlujoined
20:44:35  * indexzerojoined
20:45:38  * indexzeroquit (Client Quit)
20:46:20  * jmar777quit (Ping timeout: 252 seconds)
20:48:46  * defunctzombiechanged nick to defunctzombie_zz
20:55:25  * c4miloquit (Remote host closed the connection)
20:58:51  * leonvvjoined
20:58:52  * mikealjoined
20:59:31  * qardjoined
20:59:32  * qardpart
21:00:19  * mikealquit (Client Quit)
21:02:50  <isaacs>trevnorris: not very long
21:03:25  <trevnorris>isaacs: nope. I set old/new space size to 16, so 98% of the time was spent in gc.
21:03:41  <trevnorris>was only getting <200 req/sec.
21:03:47  <isaacs>trevnorris: yikes, lol
21:03:53  * paddybyersquit (Ping timeout: 256 seconds)
21:03:58  <isaacs>trevnorris: so, getting any memory over-use is really a problem, then
21:04:10  <isaacs>you were working hard to be spotlessly cleaning up.
21:04:19  <isaacs>like an OCD surgeon before going into the operating room
21:04:26  <trevnorris>heh
21:04:36  <isaacs>way of the FUTURE
21:04:38  <isaacs>anyway
21:04:50  <isaacs>trevnorris: is the script up there? can i play around wth it?
21:05:06  <tjfontaine>can you upload the core somewhere as well? if they appear to not be referred they could be in a closure?
21:05:56  <trevnorris>isaacs: https://github.com/joyent/node/issues/5828#issuecomment-20846439
21:05:59  <tjfontaine>or tell me where the core is on smartos.nodejs.org
21:06:11  <isaacs>also, Object: <unknown> is super annoying
21:06:22  <isaacs>we should put SOMETHING on all th eobjects we creat.
21:06:31  <trevnorris>tjfontaine: it's in /home/trevnorris/gh-5828/core75317
21:06:34  <isaacs>always have a ctor, or an id, or something
21:06:42  <trevnorris>core.75317
21:07:14  <trevnorris>so... yeah. it has no properties, no references.
21:07:38  <trevnorris>isaacs: ah, I bet it's remnant closures from all the open Sockets
21:08:21  <trevnorris>isaacs: but what I don't get is the Socket objects say they aren't referenced anywhere either.
21:11:17  * defunctzombie_zzchanged nick to defunctzombie
21:13:44  <trevnorris>isaacs: oh. it seems there's a strange bug thing.
21:13:57  <trevnorris>isaacs: I ran ::jsfindobjects, then ctrl-c.
21:14:15  <trevnorris>after that I ran it again and it instantly showed a list of objects.
21:14:23  <trevnorris>just did that again, now my objects have different counts.
21:14:27  <trevnorris>:-/
21:14:54  <trevnorris>erm. ::findjsobjects
21:16:57  <saghul>tjfontaine pong :-)
21:17:15  <tjfontaine>saghul: how have you managed to make autotools so unhappy :)
21:17:44  <saghul>tjfontaine using OSX 10.6, resisting the change
21:17:47  <saghul>;-)
21:18:22  <tjfontaine>I have no reason to really think this is it, since it doesn't really do aix work, but can you comment out the pandora check in the configure.ac and retry and see if it solves it?
21:19:10  * paulfryzeljoined
21:19:38  <saghul>it's not related to aix, it's the syntax os that case statement, AFAICT
21:19:44  <saghul>*of
21:20:32  <saghul>I just updated the issue with my findings: https://github.com/joyent/libuv/issues/851
21:20:37  <saghul>tjfontaine ^
21:20:39  <tjfontaine>k
21:22:33  <saghul>seems to be this issue: https://github.com/alanxz/rabbitmq-c/issues/65
21:24:57  <saghul>tjfontaine was shipping a premade configure script considered?
21:25:22  <tjfontaine>saghul: in release tarballs that makes sense, but not something we would do in the git repo
21:26:04  <saghul>tjfontaine I see, damn you OSX!
21:27:27  <tjfontaine>saghul: you might get a different answer from ben, but it's generally bad form to commit generated files to the vcs
21:28:09  <isaacs>tjfontaine: ime, autoconf configure scripts tend to be the exception there.
21:28:20  <saghul>tjfontaine I know, though I've seen many configure configure scripts around
21:28:26  <saghul>what isaacs said :-)
21:28:38  <isaacs>tjfontaine: because they're just so fucking brittle and awful
21:28:47  * werlequit (Quit: Leaving.)
21:29:00  <tjfontaine>it makes sense in release tarballs, it just generates noise in the vcs imo
21:29:20  <tjfontaine>but libuv isn't entirely my responsibility
21:29:52  <isaacs>tjfontaine: the issue is that, if the./configure SHOULD be roughly predictable from the configure.am, which changes only very rarely, committing the configure script significantly reduces the build requirements on users.
21:30:43  <isaacs>just means you have to run `autoconf` or somehting whenever you update that file. i've even seen this as `make configure`, which is a little bit yodawg
21:31:22  <isaacs>not as bad as one project I recall from yahoo, which had `make make` as the first step after checkout (to build the Makefile, obviously)
21:31:23  <saghul>yeah, also since OSX already comes with autotools, homebrew installs it in a different path which you need to include first. I can't just tell people to do that to compile pyuv, and gyp is not really an option since it doesn't work on Python 3 :-S
21:31:50  <tjfontaine>but the configure isn't useful with out the Makefile.in's that's where the noise will be painful
21:32:08  <tjfontaine>and inevitably someone will forget to check things in, and it won't autoreconf
21:32:14  <tjfontaine>and then people get angry
21:32:23  <tjfontaine>there's no winning this
21:32:56  <tjfontaine>saghul: I would hope pyuv is targetting release tarballs now that we generate them?
21:32:56  <saghul>regenerating the configure could be part of the release script or something, so it would be committed along with the tag or something
21:33:25  <tjfontaine>now that we use autoconf/automake we could change the release script for libuv to just `make dist`
21:33:54  <saghul>tjfontaine pyuv embeds libuv and for now it downloads it using git
21:33:58  <trevnorris>wtf. ::findjsobjects has been running for 30 minutes. what the crap could it be doing?
21:34:22  <saghul>the stable branch uses v0.10 as the base branch and pyuv master targets libuv master
21:35:34  <tjfontaine>trevnorris: is it printing things out? hard to say without seeing it
21:35:55  <trevnorris>tjfontaine: nope. it's just a blank line. like it's thinking really hard.
21:35:58  <tjfontaine>saghul: seems to me like you should be grabbing the release tarballs instead of git for builds
21:36:23  <tjfontaine>trevnorris: odd
21:36:52  <trevnorris>tjfontaine: 90131 trevnorr 1 1 0 39M 22M cpu/11 21:35 3.90% mdb
21:37:13  <saghul>tjfontaine it's really easy for me to build against any commit if I use git, and now I can no longer do it on OSX. My system may not be perfect, but it's not like libuv had releases forever ;-)
21:39:25  * mikealjoined
21:45:33  * julianduquequit (Remote host closed the connection)
21:45:52  * c4milojoined
21:46:50  * julianduquejoined
21:49:56  * EhevuTovjoined
21:57:16  * EhevuTovquit (Ping timeout: 240 seconds)
22:00:01  * pachetquit (Ping timeout: 248 seconds)
22:02:59  * rendarquit
22:04:11  <trevnorris>tjfontaine: ping. it's been running for an hour now at full throttle. think it's like caught in an infinite loop or something?
22:05:25  <tjfontaine>it's possible that has happened, did that happen before?
22:05:33  <trevnorris>tjfontaine: yeah. was just never ending
22:05:41  * paulfryzelquit (Remote host closed the connection)
22:05:43  <tjfontaine>I mean this same core as before?
22:05:46  <trevnorris>tjfontaine, isaacs: after letting it run for almost an hour here are the new results: https://gist.github.com/trevnorris/6003201
22:05:51  <trevnorris>tjfontaine: yeah. same core.
22:06:08  <tjfontaine>I thought you said it wasn't outputting anything
22:06:13  <tjfontaine>oh
22:06:20  <trevnorris>tjfontaine: it doesn't. I ctrl-C it
22:06:21  <tjfontaine>you mean this already been printed
22:06:25  <tjfontaine>ok
22:06:27  <trevnorris>then when I run it again it prints this out
22:07:09  <tjfontaine>ok, so first pass it takes a while, you ctrl-c, then try again and it finally prints things?
22:07:10  * piscisaureus_quit (Ping timeout: 246 seconds)
22:07:17  <trevnorris>yeah
22:07:33  <tjfontaine>ok, this could be it building up its symbol table as it is walking the memory space
22:07:41  <trevnorris>yeah. but it seems to only print what it was able to aquire in the time of the last run.
22:07:50  <tjfontaine>right, caching
22:11:13  * paddybyersjoined
22:11:58  <trevnorris>tjfontaine: you know if tmux or anything like it is running on the smartos server? i want to let the operation run to completion.
22:12:13  * dsantiagojoined
22:12:55  <tjfontaine>screen should be available
22:13:13  <tjfontaine>tmux now as well
22:13:19  <trevnorris>:)
22:13:26  <trevnorris>whoot. thanks.
22:14:32  <tjfontaine>did I ask if this was made with master or v0.10?
22:16:22  <trevnorris>i ran it with v0.10.13
22:16:26  * werlejoined
22:16:26  <tjfontaine>ok
22:16:37  <trevnorris>it's fixed in master, for some reason when ben upgraded to 3.20 it fixed the problem.
22:16:38  <tjfontaine>trevnorris: btw, for more information you can read ::help findjsobjects
22:16:44  <tjfontaine>it talks about the caching
22:16:46  <trevnorris>oh cool. thanks.
22:17:04  <tjfontaine>the core isn't that large though, so it shouldn't have taken an hour in my estimation
22:17:48  <trevnorris>yeah. the website tutorial said it should take a few minutes for a "large" core file.
22:18:19  <tjfontaine>when people here have time to look at it with me we may "debug the debugger"
22:18:29  <tjfontaine>inception time.
22:18:41  <trevnorris>heh
22:20:23  <trevnorris>well, there we go. i'll be interested to see how long that takes this tie.
22:20:29  <trevnorris>s/tie/time
22:21:27  <tjfontaine>you should be able to type: !date
22:21:31  <tjfontaine>in your buffer
22:21:43  <tjfontaine>so when it finishes it will also run the date command after
22:23:24  <trevnorris>from the tutorial you posted it says running ::findjsobjects -v will print the number of seconds it took to run the command.
22:24:03  <tjfontaine>probably, but if it happens to die in the mean time you might get lucky :)
22:24:19  <trevnorris>heh, good point. I'll do that :)
22:24:58  <trevnorris>tjfontaine: is there a specific way to setup so I can ssh directly to my account.
22:25:08  <trevnorris>a kitten dies everytime someone ssh's to root :P
22:27:04  <tjfontaine>did you create an account?
22:28:43  <trevnorris>tjfontaine: yeah. and added my pub to .ssh and all that.
22:33:53  <tjfontaine>trevnorris: try now
22:34:40  <trevnorris>tjfontaine: kittens will live because of you!
22:34:44  <trevnorris>what I do wrong?
22:35:41  * AvianFluquit (Remote host closed the connection)
22:36:33  <tjfontaine>trevnorris: there's a "lock" flag that needed to be removed
22:37:12  <trevnorris>ah. that specific to smartos?
22:39:06  <tjfontaine>not particularly, just user management is handled slightly different in a solaris world
22:39:21  <trevnorris>ok cool.
22:39:39  <trevnorris>this node.c api might actually work out. too bad I'm not sure how it'll be implemented :P
22:40:02  <trevnorris>tjfontaine: you know if it's possible to add an a value to an enum at run time?
22:41:40  * rjechanged nick to rje`sick
22:42:51  * paddybyersquit (Ping timeout: 246 seconds)
22:44:06  <mordy__>hrm; more a generic annoying problem; but say i had a cache of idle TCP connections; as i don't want to have these idle connections weigh down the event loop, i'm not watching them until i actually need one
22:44:17  <tjfontaine>trevnorris: huh? no? what do you mean?
22:44:18  <mordy__>with uv, how would i check for 'aliveness' in the connection?
22:44:41  <mordy__>(normally, you try to read/write bytes; errors would be delivered then..)
22:45:01  <trevnorris>tjfontaine: well, i'm trying to figure out a way to do an event system in C. so I have an enum { CLOSE = 0, CONNECTION, DATA, ... } nc_event_action;
22:45:18  <trevnorris>tjfontaine: then I was thinking to do nc_event_emit(CONNECTION, <stuff>);
22:45:38  <trevnorris>but that means no ability to add events afterwards.
22:46:11  <tjfontaine>trevnorris: you either need to use enums (which means predefined) or you need to use strings and then strcmp or hash the values
22:46:51  <trevnorris>tjfontaine: ooh. strcmp sounds like it'd be freakin slow.
22:48:15  * jmar777joined
22:53:04  * leonvvquit (Remote host closed the connection)
22:56:44  * c4miloquit (Remote host closed the connection)
23:00:34  * stagasquit (Read error: Connection reset by peer)
23:02:15  * defunctzombiechanged nick to defunctzombie_zz
23:28:13  * defunctzombie_zzchanged nick to defunctzombie
23:30:20  * defunctzombiechanged nick to defunctzombie_zz
23:37:31  <trevnorris>isaacs: ping
23:40:00  * jonathanpjoined
23:40:53  * jmar777quit (Remote host closed the connection)
23:41:31  * jmar777joined
23:44:15  <trevnorris>tjfontaine: heh, 76 minutes and counting.
23:44:35  <sblom>tjfontaine: I've got jonathanp here to help us fix Windows unit tests. Do you have a good feel for today's hotspots?
23:44:54  <tjfontaine>hm
23:45:16  <tjfontaine>I guess we should focus on master, and what applies to v0.10 can back port
23:45:29  <sblom>I'd say that's right.
23:45:37  * jmar777quit (Ping timeout: 248 seconds)
23:45:50  * sblompulls up current Jenkins status.
23:45:52  <tjfontaine>http://jenkins.nodejs.org/job/nodejs-master-windows/lastCompletedBuild/DESTCPU=x64,label=windows/tapTestReport/
23:45:55  <tjfontaine>is what I would work off of
23:46:26  <tjfontaine>test-debugger-{client, repl} tests can be ignored for now
23:46:34  <tjfontaine>as they generally break everywhere
23:47:08  <sblom>Cool.
23:47:37  * defunctzombie_zzchanged nick to defunctzombie
23:47:45  <tjfontaine>that regress gh-746 and stdout close unref are timing out, they should be achievable
23:48:04  * jmar777joined
23:48:20  <jonathanp>I will focus on those first.
23:48:33  <tjfontaine>cluster-bind-twice should be ignored for now because we have diagnosed that one, and noone has a good answer for what we should do
23:48:35  * jmar777quit (Remote host closed the connection)
23:48:55  <tjfontaine>jonathanp: don't hesitate to ask me if there's something you want to know