00:00:01  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:00:19  <isaacs>tjfontaine: lgtm for master.
00:00:24  <tjfontaine>isaacs: nod
00:00:30  <isaacs>tjfontaine: seems like it works on all the os's we care about.
00:00:33  <isaacs>and it's strictly additive.
00:00:42  <tjfontaine>yup
00:03:59  * bajtosquit (Client Quit)
00:05:01  * bajtosjoined
00:06:18  <MI6>nodejs-master: #201 UNSTABLE smartos-x64 (3/591) osx-ia32 (1/591) http://jenkins.nodejs.org/job/nodejs-master/201/
00:09:14  <bajtos>Now that we can mark tests as skipped, it's time to disable (skip) process_title test on SmartOS. Could you please review the change? https://github.com/joyent/libuv/pull/783/files
00:10:09  <MI6>joyent/node: Ben Kelly master * 8a407f5 : os: Include netmask in os.networkInterfaces() - http://git.io/rx5RAQ
00:11:45  <tjfontaine>bajtos: lgtm, but I of course you probably prefer that from ben or bert :)
00:12:22  <bajtos>tjfontaine: thanks, I am chasing Ben to review it :)
00:12:37  <tjfontaine>bajtos: I mean, it's straight forward enough to me :)
00:18:42  * mikealquit (Quit: Leaving.)
00:20:24  <MI6>nodejs-master: #202 UNSTABLE smartos-ia32 (2/591) smartos-x64 (4/591) linux-ia32 (2/591) http://jenkins.nodejs.org/job/nodejs-master/202/
00:20:57  <MI6>joyent/libuv: Miroslav Bajtoš v0.10 * fe2a315 : test: add error logging to tty unit test (+1 more commits) - http://git.io/nc0Ujw
00:21:44  <bnoordhu1s>ah...
00:22:01  <bnoordhu1s>i'm not sure if it's a good idea to use strerror() on windows
00:22:03  <bnoordhu1s>piscisaureus__: ?
00:22:06  * bnoordhu1schanged nick to bnoordhuis
00:22:56  <bnoordhuis>ah wait, it's a unix-only code path
00:22:58  <bnoordhuis>false alarm
00:23:33  * piscisaureus__quit (Read error: Connection reset by peer)
00:24:09  * piscisaureus_joined
00:31:06  <MI6>libuv-node-integration: #45 FAILURE osx-x64 (1/583) osx-ia32 (2/583) linux-ia32 (1/583) windows-ia32 (13/583) smartos-x64 (2/583) http://jenkins.nodejs.org/job/libuv-node-integration/45/
00:34:11  <MI6>libuv-v0.10: #54 UNSTABLE smartos (7/186) windows (4/187) linux (1/186) http://jenkins.nodejs.org/job/libuv-v0.10/54/
00:34:40  <tjfontaine>that was odd, I'm going to rebuild that
00:36:15  <bnoordhuis>what happened?
00:36:45  <tjfontaine>I think some host madness, I've seen the libuv hrtime test being flakey today
00:38:12  <bnoordhuis>the implementation of generators in v8 seems to be nearing completion
00:38:33  <bnoordhuis>a lot of people will be ecstatic, i don't think
00:38:41  <tjfontaine>heh
00:39:11  <tjfontaine>they're not going to do array comprehensions are they?
00:39:18  <bnoordhuis>hah
00:39:42  <MI6>libuv-v0.10-gyp: #17 UNSTABLE smartos-x64 (3/186) linux-ia32 (3/186) windows-x64 (3/187) smartos-ia32 (4/186) linux-x64 (1/186) windows-ia32 (4/187) http://jenkins.nodejs.org/job/libuv-v0.10-gyp/17/
00:39:53  <tjfontaine>because as discussed before, pain will be visisted upon the first person to try and introduce that to core
00:40:58  <bnoordhuis>V8 version 3.19.0 (candidate) <- i wager 3.19 is around the corner
00:41:11  <tjfontaine>ya trevnorris was mentioning that
00:41:20  <tjfontaine>llvm branched 3.3 today as well
00:41:23  <tjfontaine>tis the season
00:41:30  <bnoordhuis>haha, quite so
00:41:50  <tjfontaine>http://jenkins.nodejs.org//job/libuv-v0.10-gyp/lastCompletedBuild/DESTCPU=ia32,label=linux//tapTestReport/test.tap-11/ how does that even ...
00:45:19  <bnoordhuis>tjfontaine: oh, that's probably a timing issue
00:45:25  <tjfontaine>right
00:45:59  <tjfontaine>almost all those new failures are timing based, which makes me chalk it all up to virtualization
00:46:45  <tjfontaine>osx was green, but that's real hardware on a october-2012 macmini
00:47:31  <bnoordhuis>yeah. it would be nice to get rid of timing sensitive tests
00:47:44  <bnoordhuis>but that one test tests uv_cond_timedwait(), so...
00:55:25  * qardquit (Quit: Leaving.)
01:04:25  * inolenquit (Quit: Leaving.)
01:04:29  <MI6>nodejs-master-windows: #8 UNSTABLE windows-x64 (11/591) windows-ia32 (8/591) http://jenkins.nodejs.org/job/nodejs-master-windows/8/
01:12:42  * defunctzombiechanged nick to defunctzombie_zz
01:14:09  * dapquit (Quit: Leaving.)
01:16:57  <MI6>joyent/node: Robert Kowalski v0.10 * dc92ff8 : doc: document stream.Writable 'error' event - http://git.io/mYqrHw
01:23:41  * qardjoined
01:27:31  * st_lukequit (Remote host closed the connection)
01:31:27  * bnoordhuisquit (Ping timeout: 268 seconds)
01:38:38  * abraxasjoined
01:40:36  * TooTallNatequit (Quit: Computer has gone to sleep.)
01:41:28  * amartensquit (Quit: Leaving.)
01:47:06  * bajtosquit (Quit: bajtos)
01:49:43  * bnoordhuisjoined
01:57:13  * inolenjoined
02:04:56  * bajtosjoined
02:10:44  <MI6>libuv-node-integration: #46 FAILURE osx-x64 (2/591) osx-ia32 (2/591) linux-ia32 (1/591) windows-ia32 (13/583) smartos-x64 (3/591) http://jenkins.nodejs.org/job/libuv-node-integration/46/
02:10:45  <MI6>nodejs-master-windows: #9 FAILURE http://jenkins.nodejs.org/job/nodejs-master-windows/9/
02:15:15  * c4miloquit (Remote host closed the connection)
02:18:16  * bajtosquit (Quit: bajtos)
02:22:29  * bnoordhuisquit (Ping timeout: 256 seconds)
02:23:57  * piscisaureus_quit (Ping timeout: 276 seconds)
02:24:16  <MI6>nodejs-v0.10: #179 UNSTABLE smartos-ia32 (2/583) osx-x64 (1/583) smartos-x64 (36/583) osx-ia32 (1/583) http://jenkins.nodejs.org/job/nodejs-v0.10/179/
02:26:25  <MI6>nodejs-v0.10-windows: #9 FAILURE windows-ia32 (8/583) windows-x64 (12/583) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/9/
02:27:55  <MI6>libuv-node-integration: #47 FAILURE osx-x64 (1/583) osx-ia32 (3/583) windows-ia32 (13/583) smartos-x64 (1/583) smartos-ia32 (31/583) windows-x64 (12/591) http://jenkins.nodejs.org/job/libuv-node-integration/47/
02:29:26  <tjfontaine>glorious glorious windows
02:32:13  <MI6>joyent/node: Timothy J Fontaine v0.10 * 72c5815 : test: fix pummel/test-net-many-clients.js - http://git.io/eZSeWA
02:32:16  * st_lukejoined
02:36:22  <kellabyte>evening
02:42:26  <MI6>nodejs-v0.10: #180 UNSTABLE smartos-x64 (2/583) http://jenkins.nodejs.org/job/nodejs-v0.10/180/
02:47:43  <MI6>nodejs-v0.10-windows: #10 FAILURE windows-ia32 (8/583) windows-x64 (12/583) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/10/
02:58:01  <tjfontaine>kellabyte: evening
03:04:08  * AvianFluquit (Remote host closed the connection)
03:04:40  <kellabyte>tjfontaine: how are things?
03:05:14  <tjfontaine>not too bad, how are things for you?
03:05:35  <kellabyte>good besides fighting a bout of exhaustion
03:05:46  <tjfontaine>mental or physical? :)
03:06:20  <kellabyte>probably mental, it's been a real struggle lately, crazy tired all day long the last week
03:06:48  <tjfontaine>ah, mental from libuv work? :)
03:07:09  <kellabyte>I dunno, I don't feel like I've been over working myself
03:08:00  <kellabyte>but I have been thinking constantly how to do things so maybe I have no idea lol
03:08:20  <tjfontaine>hehe, you could be not getting good sleep if you're thinking too hard before bed :)
03:08:29  <kellabyte>yeah maybe
03:08:47  <kellabyte>there are some things I don't know how to do and things I know I am doing poorly that are driving me nuts
03:08:55  <kellabyte>and I want to learn how to fix them so maybe thats it
03:09:54  <tjfontaine>well if it's libuv related don't hestitate to ask in here, when folks are around I'm sure they'll do what they can to help
03:10:12  <kellabyte>cool thanks :)
03:10:47  * brsonquit (Ping timeout: 260 seconds)
03:12:34  * TooTallNatejoined
03:12:35  * TooTallNatequit (Client Quit)
03:20:35  * AvianFlujoined
03:24:18  * bajtosjoined
03:24:29  * bajtosquit (Client Quit)
03:26:53  * hueniverse1quit (Read error: Connection reset by peer)
03:49:59  <MI6>nodejs-v0.10-windows: #11 UNSTABLE windows-ia32 (9/583) windows-x64 (8/583) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/11/
03:50:48  * indexzerojoined
03:57:32  * kazuponjoined
04:04:44  * st_lukequit (Remote host closed the connection)
04:17:57  * indexzeroquit (Quit: indexzero)
04:41:41  * abraxasquit (Remote host closed the connection)
04:57:33  * dominictarrjoined
04:57:54  * indexzerojoined
05:10:37  <MI6>nodejs-master-windows: #10 UNSTABLE windows-x64 (7/591) windows-ia32 (9/591) http://jenkins.nodejs.org/job/nodejs-master-windows/10/
05:26:36  * benoitcquit (Excess Flood)
05:31:24  * benoitcjoined
06:09:39  * normanmjoined
06:12:26  * bnoordhuisjoined
06:13:26  <bnoordhuis>jesus h. christ, this hotel wifi is the worst
06:13:48  <bnoordhuis>i don't know how they manage to make such a basic amenity that shitty
06:14:19  * stagasjoined
06:14:59  * bnoordhuisquit (Read error: Connection reset by peer)
06:15:38  * loladiroquit (Quit: loladiro)
06:17:28  * bnoordhuisjoined
06:17:42  <brucem>bnoordhuis: cheaper hotels == better (and usually free) wifi.
06:18:05  <brucem>(at least in the developing world)
06:28:54  * bnoordhuisquit (Ping timeout: 264 seconds)
06:29:14  * kenperkinsquit (Quit: Computer has gone to sleep.)
06:29:42  * bnoordhuisjoined
06:37:17  * rendarjoined
06:39:17  * abraxasjoined
06:42:46  * dominictarrquit (Quit: dominictarr)
06:46:26  * AvianFluquit (Remote host closed the connection)
06:55:43  * bnoordhuisquit (Ping timeout: 260 seconds)
07:07:00  * normanmquit (Quit: Computer has gone to sleep.)
07:14:21  * normanmjoined
07:15:40  * normanmquit (Client Quit)
07:19:59  * dominictarrjoined
07:21:10  * qardquit (Quit: Leaving.)
07:27:32  * benoitcquit (Excess Flood)
07:27:35  * normanmjoined
07:30:25  * benoitcjoined
07:39:16  * stolsmajoined
07:43:49  * normanmquit (Quit: Computer has gone to sleep.)
07:54:02  * dominictarrquit (Quit: dominictarr)
08:04:56  * normanmjoined
08:28:45  * indexzeroquit (Quit: indexzero)
08:47:23  * `3rdEdenjoined
08:49:40  * stagasquit (Ping timeout: 256 seconds)
08:51:04  * stagasjoined
08:53:03  * saghulquit (Quit: ["Textual IRC Client: www.textualapp.com"])
09:11:00  * normanmquit (Quit: Computer has gone to sleep.)
09:20:49  * hzjoined
09:42:06  * kazuponquit (Ping timeout: 264 seconds)
09:48:44  * kazuponjoined
10:05:08  * abraxasquit (Remote host closed the connection)
10:07:46  * normanmjoined
10:14:58  * kazuponquit (Remote host closed the connection)
10:18:28  * kazuponjoined
10:19:43  * kazuponquit (Remote host closed the connection)
10:28:04  * timoxleyquit (Quit: Computer has gone to sleep.)
11:10:22  * kazuponjoined
11:38:47  * kazuponquit (Read error: Connection reset by peer)
11:39:19  * kazuponjoined
11:57:12  * kazuponquit (Remote host closed the connection)
12:10:56  * kuebkjoined
12:24:55  * kazuponjoined
12:44:02  * AvianFlujoined
12:54:10  <kuebk>hello
12:54:22  <kuebk>how can i get in touch with bnoordhuis
12:54:24  <kuebk>?
13:00:50  * kevinswiber-awayjoined
13:13:21  * kevinswiber-awayquit (Remote host closed the connection)
13:16:54  * normanmquit (Quit: Computer has gone to sleep.)
13:43:04  * kazuponquit (Remote host closed the connection)
13:51:00  * stolsmaquit (Ping timeout: 256 seconds)
13:59:07  * kevinswiber-awayjoined
13:59:09  * c4milojoined
14:08:56  * kenperkinsjoined
14:19:43  * c4miloquit (Remote host closed the connection)
14:34:17  <tjfontaine>kuebk: have patience, he should be around sometime today
14:49:39  <kuebk>ok
14:50:18  <kuebk>does node not follow symlinked modules in node_module?
14:50:25  <kuebk>node_modules*
14:53:24  * kazuponjoined
14:59:19  * kazuponquit (Ping timeout: 264 seconds)
15:02:37  * c4milojoined
15:06:50  * kuebkpart
15:21:06  * perezdquit (Quit: perezd)
15:26:34  * benoitcquit (Excess Flood)
15:32:57  * benoitcjoined
15:43:56  * normanmjoined
15:48:36  * benoitcquit (Excess Flood)
15:55:58  * benoitcjoined
15:56:41  * dapjoined
15:58:34  * dapquit (Client Quit)
16:00:15  * hueniversejoined
16:02:41  * dapjoined
16:02:41  <isaacs>kellabyte: hey. so, this idea of node.c has been kicking around in my head for several months.
16:02:58  <isaacs>kellabyte: i think it could be a good idea. haywire is an interesting proof of concept, but haywire is a different thing entirely.
16:04:12  <isaacs>kellabyte: the trickiest part is that node's "killer feature" is really three things: 1) a good module system. 2) a powerful module installer/publisher. 3) high-ish level apis for http,tcp,tls,dns,etc.
16:04:38  <isaacs>kellabyte: #3 is pretty easy. #2 is easy if you have #1. #1 probably requires fancy linker changes or some other "impossible" stuff.
16:04:57  <isaacs>kellabyte: so i'll just leave this here and maybe someday you'll decide that it's a worthy challenge ;)
16:05:47  <isaacs>(and no, libuv is not "node.c". the hello world http server should be shockingly few loc.)
16:10:44  * bnoordhuisjoined
16:13:25  * `3rdEdenquit (Remote host closed the connection)
16:22:08  <isaacs>indutny: can you review this? https://github.com/joyent/node/pull/5437
16:26:06  <indutny>hey
16:26:07  <indutny>sure
16:30:13  <bnoordhuis>morning gents
16:30:20  <bnoordhuis>i'm about to become a movie star
16:30:23  <tjfontaine>gmorning bnoordhuis
16:30:47  <tjfontaine>bnoordhuis: action star, or romantic lead
16:30:48  <indutny>isaacs: LGTM
16:31:09  <bnoordhuis>tjfontaine: i think i'm the handsome stranger
16:31:23  <tjfontaine>excellent
16:31:49  <indutny>bring me the head of handsome prince
16:32:19  * kazuponjoined
16:34:04  <tjfontaine>bnoordhuis: kuebk was looking for you earlier
16:34:15  <bnoordhuis>oh dear, what did i do?
16:35:09  <isaacs>indutny: thanks
16:39:09  * loladirojoined
16:42:59  * loladiroquit (Client Quit)
16:43:54  * bnoordhuisquit (Ping timeout: 276 seconds)
16:46:08  * hueniversequit (Quit: Leaving.)
16:46:29  * bnoordhuisjoined
16:48:52  <indutny>isaacs: you're welcome
16:48:57  <indutny>thank you for handling this stuff
16:52:08  <MI6>joyent/node: isaacs master * ec57623 : http: Use writev instead of the hacky hot end (+2 more commits) - http://git.io/InPJwQ
16:52:20  <bnoordhuis>^ "hot end"
16:52:30  <tjfontaine>twss
16:53:52  * amartensjoined
16:56:16  <isaacs>hacky hot end!
16:56:28  <isaacs>sounds dirty.
16:56:33  <indutny>:)
16:56:37  <indutny>that's what she said
16:56:43  <indutny>(sorry)
16:56:50  <bnoordhuis>twtjs
16:56:53  <tjfontaine>:)
17:02:31  <MI6>nodejs-master: #203 UNSTABLE smartos-x64 (4/592) linux-ia32 (1/592) linux-x64 (1/592) http://jenkins.nodejs.org/job/nodejs-master/203/
17:06:56  * brsonjoined
17:07:23  <tjfontaine>we're kinda slipping back into a world where debugger tests are troublesome :/
17:09:24  <isaacs>yeah, seems that way
17:09:26  <isaacs>what changed?
17:09:44  * perezdjoined
17:10:31  <tjfontaine>I'm not sure, we've had a few that seemed to help the breakpoint catching issue, but on smartos64 they fail kinda regularly, and the debugger-client test is starting to fail more again with its econnreset failure
17:10:58  <tjfontaine>the windows failure yesterday were a ton of left over debugger tests hanging around
17:13:24  <isaacs>ircretary: tell piscisaureus How long are our talks supposed to be tonight?
17:13:24  <ircretary>isaacs: I'll be sure to tell piscisaureus
17:19:41  <bnoordhuis>isaacs: bert says 20-25 minutes
17:20:22  <bnoordhuis>isaacs: it's not super strict though, if you have a lot of interesting stuff to tell please do
17:22:25  <bnoordhuis>isaacs: if you want to stick to 15 minutes that's also fine
17:22:33  <bnoordhuis>also beep boop #poopin
17:22:55  <tjfontaine>heh
17:23:05  <bnoordhuis>^ those last four lines were bert as you can no doubt tell by the distinct lack of eloquence
17:24:25  * piscisaureus_joined
17:25:46  * papertigersjoined
17:27:12  * indexzerojoined
17:28:15  <MI6>nodejs-master-windows: #11 UNSTABLE windows-x64 (8/592) windows-ia32 (9/592) http://jenkins.nodejs.org/job/nodejs-master-windows/11/
17:30:17  * TooTallNatejoined
17:31:28  * qardjoined
17:31:53  * indexzeroquit (Client Quit)
17:32:40  * mikealjoined
17:32:56  * mcavagejoined
17:33:21  * c4miloquit (Remote host closed the connection)
17:36:11  * kazuponquit (Remote host closed the connection)
17:37:04  * mjr__joined
17:38:57  <mjr__>Check out this awesome bug: https://gist.github.com/mranney/5549123
17:40:44  <TooTallNate>yikes
17:41:59  * loladirojoined
17:42:41  * indexzerojoined
17:45:13  * c4milojoined
17:47:44  * bnoordhuisquit (Ping timeout: 255 seconds)
17:48:11  <mjr__>Yeah, it's crazy. hrtime works fine though.
17:49:53  * kazuponjoined
17:49:53  <tjfontaine>I don't see any obvious caching going on, I presume external to the node process the system time appears to be advancing?
17:50:55  <mjr__>oh sure
17:51:00  <mjr__>Other node processes on the same machine are fine.
17:54:11  <tjfontaine>mjr__: can you dtrace that process looking for the gettimeofday calls?
17:54:32  <mjr__>oh, probably. Once I google around to find the syntax
17:54:48  <mjr__>Or ask someone else here who already know it.
18:01:13  <mjr__>gettimeofday isn't a sys call. I'm not sure what SmartOS expects to do there
18:04:21  <tjfontaine>asking
18:05:35  <papertigers>mjr__: I can trace it if you want
18:07:26  <papertigers>mjr__: its going to be apart of the pid provider because its a lib func
18:09:20  <tjfontaine>oh with pid I do get it
18:09:21  * inolenquit (Quit: Leaving.)
18:13:02  <papertigers>yeah its getting called a bunch, let me see whats being returned
18:14:19  * bnoordhuisjoined
18:14:20  <tjfontaine>mjr__: ok, so the lib gettimeofday uses hrtime internally, so if there's a discrepency it's either in the v8 layer or something has monkey patched Date.now, but I'm boning further up on my dtrace to get an incantation that will let me verify the results
18:19:59  <isaacs>bnoordhuis: hahah, thanks
18:21:14  <papertigers>tjfontaine: over 10seconds here are the top stacks calling gettimeofday
18:21:15  <papertigers>https://gist.github.com/papertigers/f67c1501bcbfdde947a9
18:22:24  <tjfontaine>papertigers: ya, the Runtime_DateCurrentTime is what we would expect to see
18:22:40  <tjfontaine>I was hoping to use :return and see what the value being returned was
18:23:10  <papertigers>its hard because the struct is passed in on entry and populated. return is just showing us the return value 0 or -1 with the errno
18:23:12  <papertigers>they are returning 0
18:25:13  <tjfontaine>ya, but we need to inspect the timeval struct
18:25:37  <papertigers>tjfontaine: ill grab that right now :)
18:26:19  * kuebkjoined
18:32:14  * defunctzombie_zzchanged nick to defunctzombie
18:34:06  <tjfontaine>this is what I did, but it's wrong, I'm still quite new to the dtrace stuff though https://gist.github.com/tjfontaine/e732c62249b1c376570b
18:35:23  <papertigers>you are on the right track
18:37:48  <papertigers>okay ill gist this
18:37:52  * stolsmajoined
18:37:56  <tjfontaine>I'd appreciate it
18:37:57  * `3rdEdenjoined
18:40:52  <papertigers>tjfontaine: https://gist.github.com/papertigers/f67c1501bcbfdde947a9
18:40:55  <papertigers>the comment
18:41:02  <papertigers>if thats right I'm double checking
18:41:52  <isaacs>piscisaureus_: http://j.mp/2013-road-to-node-10
18:41:58  * kazuponquit (Remote host closed the connection)
18:44:44  <isaacs>piscisaureus_: still editing and probably removing some of the history stuff, because it's already covered elsewhere and i assume most people already know it
18:47:27  <mjr__>papertigers: have you compared those stacks to a process that is getting the dates correctly?
18:48:15  <tjfontaine>you guys are on 0.8 right? I should build that
18:53:05  * indexzeroquit (Quit: indexzero)
18:53:57  <mjr__>tjfontaine: yeah, we are on 0.8
18:55:10  * loladiroquit (Quit: loladiro)
18:55:17  * bnoordhuisquit (Ping timeout: 252 seconds)
18:59:00  <papertigers>I'm not sure if self->tp is good enough to predicate on
19:00:06  * inolenjoined
19:01:03  * indexzerojoined
19:01:49  <trevnorris>good afternoon
19:02:12  <tjfontaine>papertigers: it's a math problem more or less
19:05:25  <papertigers>tjfontaine: the outputs not looking right
19:13:28  * normanmquit (Quit: ["Textual IRC Client: www.textualapp.com"])
19:23:10  * bnoordhuisjoined
19:23:54  * indexzeroquit (Quit: indexzero)
19:31:10  * bnoordhuisquit (Ping timeout: 272 seconds)
19:33:31  <papertigers>tjfontaine: i have it
19:39:30  * perezdquit (Quit: perezd)
19:43:08  * dapquit (Read error: Connection reset by peer)
19:43:14  * dapjoined
19:43:39  * `3rdEdenquit (Remote host closed the connection)
19:44:24  <papertigers>Here it is: https://gist.github.com/papertigers/f67c1501bcbfdde947a9#comment-828221
19:44:56  <mjr__>So the kernel still knows what time it is, but somehow the time is getting re-used in V8 or node somewhere.
19:52:36  * bnoordhuisjoined
19:56:39  <trevnorris>bnoordhuis: hey. i'm debating on driving or taking caltrain. what are y'alls plans for the day?
19:57:17  * mcavagequit (Remote host closed the connection)
19:57:45  * mcavagejoined
19:59:20  * perezdjoined
19:59:54  <bnoordhuis>trevnorris: i think we move down to the meteor offices around 4 or 5 pm
20:00:03  <bnoordhuis>we're in san mateo now btw
20:00:15  <bnoordhuis>where are you?
20:01:04  <trevnorris>in cupertino
20:01:44  * mcavagequit (Read error: Connection reset by peer)
20:01:44  <bnoordhuis>but you're coming down to sf, right?
20:01:54  * stolsmaquit (Ping timeout: 264 seconds)
20:02:08  <trevnorris>yeah. just depending on the timing I might drive or jump on the train.
20:02:28  <trevnorris>actually, i'll be driving by san mateo on the way up
20:02:51  * mcavagejoined
20:06:48  * abraxasjoined
20:07:01  <bnoordhuis>trevnorris: you're welcome to drop by
20:07:09  <bnoordhuis>it's a bit chaotic though, we're filming
20:07:15  <bnoordhuis>(no joke)
20:07:34  <tjfontaine>papertigers: sorry I was at lunch, my issue was 64bit kernel timeval struct info
20:08:28  <mjr__>tjfontaine: do you understand the V8 macros and such enough to decode this path any further?
20:08:56  <mjr__>We ought to be able to pid provider our way to victory here, if we knew how to read the V8 values.
20:09:11  <mjr__>like, what does this return? v8::internal::Runtime_DateCurrentTime(v8::internal::Arguments, v8::internal::Isolate*)
20:10:04  <trevnorris>bnoordhuis: heh, so you weren't joking about becoming a movie start.
20:10:34  <bnoordhuis>trevnorris: nope. my grand moment in the spotlights is in 20 minutes
20:11:06  <papertigers>tjfontaine: yeah the issue was timeval32 (64bit kernel and 32bit node)
20:11:07  <trevnorris>bnoordhuis: ah, awesome. well good luck. what's the address? I'll swing by on my way up.
20:11:28  <tjfontaine>just as a side note, we know that Date.now isn't monkeypatched right?
20:11:34  * abraxasquit (Ping timeout: 256 seconds)
20:11:36  <bnoordhuis>trevnorris: one sec, let me look it up
20:12:03  <tjfontaine>mjr__: my reading was a Local<Value> but I will look further
20:12:14  * kazuponjoined
20:12:40  <bnoordhuis>trevnorris: 107 S. B Street, San Mateo, CA 94401. we're in suite 220
20:12:51  <trevnorris>bnoordhuis: coolio. thanks.
20:13:22  <mjr__>tjfontaine: I don't know how to know that, but I would certainly be surprised if that were the case.
20:16:36  <tjfontaine>ok so first argument to RUNTIME_FUNCTION is the return type, so Runtime_DateCurrentTime is returning MaybeObject*
20:17:31  * kevinswiber-awayquit (Remote host closed the connection)
20:20:49  * `3rdEdenjoined
20:21:23  * kazuponquit (Ping timeout: 252 seconds)
20:21:25  * kevinswiber-awayjoined
20:22:10  * `3rdEdenquit (Remote host closed the connection)
20:23:01  <piscisaureus_>isaacs: sorry, was afk. looking at it
20:23:22  * paddybyersjoined
20:23:28  <piscisaureus_>isaacs: mine will be not extremely technical but be more about the greater context
20:23:44  * papertigersquit (Quit: papertigers)
20:33:05  <tjfontaine>oh papertigers left
20:33:23  <mjr__>he'll be back in a few hours, I believe
20:35:14  <mjr__>tjfontaine: a naive printing of the return value gives you this: https://gist.github.com/papertigers/f67c1501bcbfdde947a9#comment-828246
20:35:30  <mjr__>But I assume that is some V8 pointer to something on their heap or whatever
20:35:38  <tjfontaine>ya
20:35:43  <tjfontaine>ok I'll give you a dtrace incantation for gettimeofday jstack aggregation
20:35:54  <mjr__>I have that already, lemme gist it
20:35:58  <tjfontaine>also just for giggles can we check Date.now.toString()
20:36:38  <mjr__>https://gist.github.com/papertigers/f67c1501bcbfdde947a9#comment-828248
20:37:33  <mjr__>and also https://gist.github.com/papertigers/f67c1501bcbfdde947a9#comment-828249
20:38:11  <tjfontaine>no, actually the function toString()'d, so we can make sure it comes back with function () { [native code] }
20:38:19  <mjr__>ahh
20:38:34  <mjr__>'function now() { [native code] }'
20:38:47  <mjr__>So nobody has swapped it out
20:39:55  <tjfontaine>well not necessarily, var a = function () {}; var b = a.bind(); b.toString(); will return [native code]
20:40:14  <tjfontaine>but at least it's not Date.now = function () {}
20:41:11  <mjr__>Also interesting to note that process.hrtime() works as expected: https://gist.github.com/papertigers/f67c1501bcbfdde947a9#comment-828252
20:42:37  <isaacs>tjfontaine: http://nodejs.org/node.pkg
20:42:42  <isaacs>TooTallNate: ^
20:42:49  <isaacs>can you try that and see if it warns?
20:42:54  <tjfontaine>mjr__: so I presume there's some logging function that's exhibiting this behavior as well as your repl
20:43:37  <TooTallNate>isaacs: https://cloudup.com/ij5xuOvM8oM
20:43:41  <tjfontaine>process_metric seems like a candidate
20:44:50  <isaacs>hmm.
20:44:52  <isaacs>ok
20:45:00  <tjfontaine>do that spctl incantation
20:45:06  <mjr__>tjfontaine: yeah, this process is aggregating timing values, making quantiles, and spitting them out every minute.
20:45:31  <isaacs>TooTallNate: can you do this? spctl -a -vvvv --type install node.pkg --raw
20:46:26  <TooTallNate>isaacs: https://cloudup.com/i2ePUCPJWbZ
20:46:46  <tjfontaine>mjr__: and for a 10 second pass, being executed 667 times seems accurate?
20:47:08  <mjr__>tjfontaine: I guess I'm surprised at how often it runs, but yeah
20:49:07  <mjr__>tjfontaine: it's probably getting called from within the npm "metrics" library, one of the few things we run that we didn't write.
20:49:26  <tjfontaine>ok, I guess I'll look at that
20:51:03  <tjfontaine>hm this doesn't use Date.now, it uses (new Date()).getTime()
20:52:17  <mjr__>The biggest Date.now user in that dtrace output to be the TCP -> HTTP handoff in node itself.
20:53:31  <tjfontaine>right, but we know that gettimeofday is returning the right value, so either v8 or a modified Date is discarding that result
20:54:07  <tjfontaine>but it's unclear to me how that would actually happen, short of the monkey patch mechanism
20:55:05  <mjr__>It's really strange that it doesn't get stuck forever
20:55:12  <mjr__>Every minute or so, it'll advance.
20:55:40  <tjfontaine>is it exactly every min? because that would of course be suspect
20:55:52  <mjr__>checking
20:56:20  * rendarquit
20:56:30  * bnoordhuisquit (Ping timeout: 264 seconds)
20:56:42  <tjfontaine>not that I want to restart the process just yet, but does it happen if you start this process separately?
20:57:23  <mjr__>can't directly start it while one is already running with the same name and port bindings
20:57:32  <tjfontaine>ok
20:59:17  <mjr__>other node processes on the same machine are fine though
21:00:30  <mjr__>looks like it updates the time every 2 minutes and 11 seconds
21:00:31  <mjr__>https://gist.github.com/papertigers/f67c1501bcbfdde947a9#comment-828260
21:01:06  <mjr__>So it is on a very regular interval, but not on an obvious timer boundary
21:01:32  <mjr__>131 seconds?
21:01:56  * Raynoschanged nick to JakeV
21:02:20  * JakeVchanged nick to Raynos
21:03:51  * kevinswiber-awayquit (Remote host closed the connection)
21:06:54  <tjfontaine>nothing about 131 seems magic-number wise
21:12:05  <isaacs>TooTallNate: can you download http://nodejs.org/node.pkg again and try it?
21:12:21  <TooTallNate>yup yup
21:12:51  <TooTallNate>isaacs: damn, no same thing https://cloudup.com/cKe6w9cAo4J
21:12:55  <TooTallNate>i'll try running your command again
21:13:12  <isaacs>ok
21:13:24  <isaacs>fuck apple, srsly.
21:13:24  <isaacs>this is so unnecessarily complicated.
21:13:54  <mjr__>isaacs: you should try publishing an app to the iOS App Store. Totally nuts.
21:13:56  * perezdquit (Quit: perezd)
21:14:17  <isaacs>mjr__: actually, a lot of unnecessary complexity here is in the fact that i'm not using Xcode for everything
21:14:40  <isaacs>i have a feeling that, at least the signing part, would be way easier
21:14:44  <isaacs>of course, then there's the review bs
21:14:52  <tjfontaine>isaacs: you can build, a non-universal, version of node with xcode
21:14:53  <mjr__>I think they just make it a little bit extra hard on purpose to raise the bar past where casual programmers will release random projects that confuse people.
21:15:22  <mjr__>It was only a matter of time before they did this to OSX as well.
21:15:24  <tjfontaine>isaacs: ./configure --xcode && open node.xcodeproj
21:16:10  <TooTallNate>isaacs: so, it's different this time at least: https://cloudup.com/cqFhaXlKowD
21:17:34  * kazuponjoined
21:17:38  <tjfontaine>spctl --assess --type install -vvv node.pkg
21:17:38  <tjfontaine>node.pkg: accepted
21:17:39  <tjfontaine>source=Developer ID
21:17:39  <tjfontaine>origin=Developer ID Application: Joyent, Inc (X4ETB2T5LK)
21:18:32  <tjfontaine>this looks good
21:18:35  * perezdjoined
21:18:38  <tjfontaine>10.8.3 here
21:19:22  <isaacs>TooTallNate: your screenshot says that your security settings are overly restrictive
21:19:29  <isaacs>TooTallNate: ie, "only from mac app store"
21:19:36  <TooTallNate>oh shit, is that not the default?
21:19:39  <isaacs>TooTallNate: shoudl be something like "Mac App Store and Trusted Developers"
21:19:40  <isaacs>yeah
21:19:44  <TooTallNate>ok
21:19:44  <TooTallNate>hahah
21:19:47  <TooTallNate>i'm retarded
21:19:48  <TooTallNate>sorry
21:19:50  <isaacs>no worries
21:19:54  <isaacs>so, with that setting, does that works?
21:20:18  <TooTallNate>isaacs: ok, yes, looks good :D
21:20:56  <tjfontaine>mjr__: I'm really at a loss for this, I wish I could figure out a way to reproduce it here
21:21:09  <mjr__>tjfontaine: yeah, me too.
21:21:10  <isaacs>TooTallNate: when you try to install it, then does it work?
21:21:20  <isaacs>TooTallNate: or does it complain about the binaries inside not being signed?
21:21:36  <mjr__>tjfontaine: this actually happens from time to time on different processes, but the side effects of broken time are much harder to spot there.
21:21:51  <TooTallNate>isaacs: you want me to go through the whole installation or is opening the .pkg and seeing the installer good enough?
21:22:02  <mjr__>tjfontaine: because it happened on a metrics server process finally, it messed up a bunch of graphs until we found it.
21:22:05  * papertigersjoined
21:22:06  <isaacs>TooTallNate: if you dont mind, installing it all the way would be awesopme
21:22:10  <tjfontaine>mjr__: but this of course doesn't happen at process start, only after a certain amount of time has passed?
21:22:12  <TooTallNate>sure np
21:22:13  <tjfontaine>mjr__: right
21:22:18  <isaacs>oh, man, i hope this worked.
21:22:19  * kazuponquit (Ping timeout: 260 seconds)
21:22:20  * papertigersquit (Client Quit)
21:22:28  <mjr__>tjfontaine: only after some condition that we don't know what is.
21:22:37  <TooTallNate>done, and i see:
21:22:38  <TooTallNate>☮ ~ (master) ∴ node -v
21:22:38  <TooTallNate>v0.10.6-pre
21:22:40  <TooTallNate>isaacs: ^
21:22:41  <TooTallNate>lgtm
21:23:06  <mjr__>tjfontaine: This process was running for a long time before this happened. Other times we've seen it with processes that have only been running for a couple of days. No obvious pattern to it.
21:24:06  <isaacs>TooTallNate: thanks!!
21:24:09  <isaacs>\o/
21:24:19  <TooTallNate>seet
21:24:21  <TooTallNate>sweet
21:24:23  <TooTallNate>:)
21:24:29  <TooTallNate><3 apple
21:25:07  <tjfontaine>mjr__: can we get a core from the process
21:26:13  <mjr__>sure, there's one already on the machine
21:26:48  <isaacs>mjr__: ok, great.
21:27:04  <isaacs>mjr__: this sounds like it has to be either a bug in V8, or someone's clubbed Date.prototype.getTime()
21:27:16  <isaacs>mjr__: you can test the second pretty easily
21:27:53  * c4miloquit (Remote host closed the connection)
21:28:16  <isaacs>mjr__: add some code in your require.main that does something like this: var orig = Date.prototype.getTime; Object.defineProperty(Date.prototype, 'getTime', { get: function() { return orig }, set: function(n) { throw new Error('SHENANIGANS!') }, configurable: false })
21:28:27  <isaacs>mjr__: if someone IS clubbing it, then that'll throw probably right awa.
21:28:29  <isaacs>*away
21:28:40  <isaacs>and you can test this on a new instance that runs on a different port or something
21:30:26  <mjr__>isaacs: can we do this without restarting the process?
21:30:48  <isaacs>mjr__: i don't see how. if it's already been clubbed, then it's too late.
21:31:13  <mjr__>restart will for sure fix it for some time
21:31:19  <isaacs>yeah, so, that sucks.
21:31:29  <isaacs>but you've established that gettimeofday is returning the correct timeval
21:31:41  <isaacs>and that (new Date).getTime() is calling gettimeofday
21:32:09  <isaacs>so, either it's a fuckup turning that int into a Handle<Number>, and buried in the bowels of V8, or it's a clever module that's clubbing getTime
21:34:38  <TooTallNate>you're not using node-time are you? :p
21:35:23  <isaacs>TooTallNate: i'm going to have one more pkg for you in a sec
21:35:35  <isaacs>TooTallNate: just to make sure it works if i use the developerID to sign all the binaries as well as the pkg itself
21:36:07  <TooTallNate>kk
21:37:53  <trevnorris>isaacs: you going to the meteor office w/ bn?
21:38:37  <TooTallNate>trevnorris: today?
21:38:39  <isaacs>trevnorris: no, i'm planning on just heading to the klout offices
21:38:41  <isaacs>for the meetup
21:38:56  <TooTallNate>ya ya, trevnorris you should go :D
21:39:13  <trevnorris>isaacs: coolio. going to meet up w/ them in san mateo. since I have to drive by on my way anyways.
21:39:14  <isaacs>TooTallNate: ok, go for it
21:39:22  <isaacs>TooTallNate: http://nodejs.org/node.pkg
21:39:37  <isaacs>trevnorris: in san mateo??
21:39:39  <mjr__>TooTallNate: not using node-time
21:39:46  <isaacs>trevnorris: i thought they were in SOMA?
21:39:56  <mjr__>isaacs: yes, DTrace says gettimeofday is returning something normal
21:40:06  <isaacs>mjr__: ok, so, there's only two options left.
21:40:27  <TooTallNate>isaacs: i get v0.10.6-pre again
21:40:31  <TooTallNate>so it's good
21:40:32  <isaacs>mjr__: i would strongly encourage you to put the guard in your program.
21:40:36  <isaacs>TooTallNate: W00T!
21:40:40  <isaacs>TooTallNate: thanks again
21:40:44  <isaacs>ok, our pkg's can be non-sucky
21:40:45  <TooTallNate>isaacs: np dude
21:40:52  <TooTallNate>glad i can do something core-related today :p
21:41:04  <trevnorris>isaacs: the address he gave me is 107 S. B Street, San Mateo
21:41:25  <trevnorris>isaacs: and they're heading to the meteor office around 4.
21:41:42  <TooTallNate>san mateo?
21:42:05  <isaacs>trevnorris: whoa
21:42:15  <isaacs>trevnorris: must be near the airport or something?
21:42:22  * kuebzkyjoined
21:42:28  * kuebzkyquit (Client Quit)
21:42:32  <isaacs>trevnorris: i'd recommend parking down there and taking bart
21:43:45  <trevnorris>isaacs: ok. i've only taken caltrain to-date. i'll find a bart stop near by.
21:47:19  <mjr__>tjfontaine: also interesting to note is that this process has been in this state for many days now
21:47:45  <trevnorris>isaacs: oh, even better. there's a caltrain station around the block. good idea. hate parking in sf.
21:51:08  * st_lukejoined
21:51:11  * trevnorrispart ("Leaving")
21:51:16  * trevnorrisjoined
21:52:31  <tjfontaine>mjr__: ok, well my first pass at trying to use findjsobjects resulted in crashing the v8 provider, so I'm trying to get into a better environment to test with it
21:53:38  * trevnorris&
21:53:38  <LOUDBOT>HAVE A BOWL OF MY FORREALLIOS DOG
21:54:42  <isaacs>LOUDBOT: twitlast
21:54:43  <LOUDBOT>http://twitter.com/LOUDBOT/status/332614634869776385 (railbait/#perlcafe)
21:55:33  <isaacs>piscisaureus_: can you review and signoff on https://github.com/joyent/node/pull/5403?
21:55:41  <isaacs>piscisaureus_: or bug ben to?
22:00:24  <isaacs>piscisaureus_: fwiw, it was a green build, so it's strictly feedback on style and API :)
22:02:10  * mikealquit (Quit: Leaving.)
22:03:27  * mjr____joined
22:05:33  * mjr__quit (Ping timeout: 256 seconds)
22:08:45  * c4milojoined
22:08:54  * kuebkquit
22:15:52  <MI6>joyent/node: isaacs v0.10 * 4b69bcf : doc: s/search.npmjs.org/npmjs.org/ - http://git.io/AzF8Bw
22:18:13  * kazuponjoined
22:18:41  * perezdquit (Quit: perezd)
22:18:43  * st_lukequit (Remote host closed the connection)
22:23:04  * kazuponquit (Ping timeout: 256 seconds)
22:28:12  <MI6>nodejs-v0.10: #181 UNSTABLE osx-ia32 (1/583) http://jenkins.nodejs.org/job/nodejs-v0.10/181/
22:28:26  * bnoordhuisjoined
22:35:21  * perezdjoined
22:37:02  * st_lukejoined
22:48:51  <MI6>nodejs-v0.10-windows: #12 UNSTABLE windows-x64 (7/583) windows-ia32 (7/583) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/12/
22:55:50  * dominictarrjoined
22:56:14  <Raynos>there was a sweet node library to find out what's still running in the background that stops my node process from terminating
22:56:55  <Raynos>or maybe a process.binding internal I can peek at
22:57:00  <isaacs>Raynos: process._getActiveHandles
22:57:03  <isaacs>()
22:57:16  <Raynos>yes!
23:00:24  <Raynos>isaacs: what does your intuition say that these are? ( https://gist.github.com/Raynos/eb01a8ff5737dd599afc )
23:01:12  <Raynos>aha! I found it
23:02:01  * mcavagequit (Remote host closed the connection)
23:08:35  * papertigersjoined
23:08:36  * papertigersquit (Client Quit)
23:13:26  <isaacs>Raynos: a net.Socket, and a setTimeout
23:15:48  * c4miloquit (Remote host closed the connection)
23:16:49  * defunctzombiechanged nick to defunctzombie_zz
23:17:00  <Raynos>thanks
23:17:40  * bnoordhuisquit (Ping timeout: 256 seconds)
23:18:01  * perezdquit (Quit: perezd)
23:18:57  * kazuponjoined
23:23:43  * kazuponquit (Ping timeout: 268 seconds)
23:28:08  * st_lukequit (Remote host closed the connection)
23:34:23  <mjr____>isaacs: in case you are wondering, we found the date bug, and it's some low level CPU register bullshit.
23:34:31  <mjr____>Not JS in any way.
23:39:54  * piscisaureus_quit (Ping timeout: 256 seconds)
23:42:49  <isaacs>nice!
23:52:06  * perezdjoined
23:55:09  * defunctzombie_zzchanged nick to defunctzombie
23:56:30  * defunctzombiechanged nick to defunctzombie_zz
23:57:54  <mjr____>isaacs: but it is a truly deep and fascinating mystery
23:59:03  <isaacs>yeah, i'm listening to the discussion here in meatspace
23:59:45  <mjr____>If it wasn't about time for everybody to go home soon, I'd come over there and join in.